l90B5-058'22

5

PPETO' , ^TtTUTE

^sg^g|g||»gaa

MARY GUNN LIBRARY

0000017939

South African National

Biodiversity Institute

Digitized by the Internet Archive

in 2016

https://archive.org/details/bothaliavolume1313unse

BOTH ALIA

Volume 13

Published by the

Botanical Research Institute, Pretoria

Department of Agriculture and Fisheries

Gepubliseer deur die

Navorsingsinstituut vir Plantkunde, Pretoria

Departement van Landbou en Visserye

Edited by — Onder redaksie van

D. J. B. KILLICK

Index prepared by — Indeks voorberei deur

C. F. FOURIE

CONTENTS— INHOUD

Page/Bladsy

No. 1 & 2 1-262

No. 3 & 4 263-592

Date/Datum

18 July/Julie 1980

16 October/Oktober 1981

Page

Bladsy

ARNOLD, T. H. & VORSTER, P. Ecological adaptations and possible convergence in Ficinia arenicola var. erecta

and Mariscus durus 441

Ficinia lucida: An interspecific hybrid between F. cedarbergensis and F. ixioides subsp. glabra 439

Boucher, C. Notes on the use of the term ‘Renosterveld’ 237

Bredenkamp, G. J. & Theron, G. K. A synecological account of the Suikerbosrand Nature Reserve. 13. The

phytosociology of the Ventersdorp Geological System 199

Campbell, B. M. & Moll, E. J. Determination of plot size 575

CODD, L. E. Obituary: Robert Harold Compton (1886-1979) 244

DYER, R. A. Obituary: Adolf Joseph Wilhelm Bayer (1900-1978) 244

A new species of Cyrthanthus from Baviaanskloof , south-western Cape 135

A new species of Huernia from Ovambo 136

The Riocrcuxia flanaganii complex: a reassessment 435

Edwards, D. A note on the extension of the degree reference system for citing biological distribution records to

north of equator and west of Greenwich meridian 574

Ellis, R. P. Leaf anatomy of the South African Danthonieae (Poaceae).

II . Merxmuellera disticha 185

Leaf anatomy of the South African Danthonieae (Poaceae). HI. Merxmuellera stricta 191

Leaf anatomy of the South African Danthonieae (Poaceae). IV. Merxmuellera drakenbergensis and M. stereo-

phylla 487

Leaf anatomy of the South African Danthonieae (Poaceae). V. Merxmuellera macowanii, M. davyi and M.

aureocephala 493

Furness, H. D. & Breen, C. M. The vegetation of seasonally flooded areas of the Pongola River Floodplain 217

Gibbs Russell, G. E. A new combination of Eriochloa 457

Gibbs Russell, G. E. & Robinson, E. R. Phytogeography and speciation in the vegetation of the eastern Cape . . . 467

Glen,H. F. Nomenclature in the genus Mestoklema 455

Guy, P. R. Changes in the herb layer of the riverine woodland in the Singua Wildlife Research Area, Zimbabwe . . 527

Hardy, D. S. & Reid, C. A new variety of Aloe from the Vryheid District 451

Immelman, Kathleen. Notes on South African species of Holothrix 455

KlLLICK, D. J. B. Obituary: John Phillip Harison Acocks (191 1-1979) 239

Kok, P. D. F. Notes on Digitaria in South Africa 457

Liengme, C. A. Plants used by the Tsonga people of Gazankulu 501

Linder, H. P. Taxonomic studies in the Disinae. V. A revision of the genus Monadenia 339

Taxonomic studies in the Disinae. VI. A revision of the genus Herschelia 365

Mac-;ll, R. E. Musci austro-africani II. Bryophyte collections in southern Africa and southern African type

specimens in the National Herbarium, Pretoria 127

Magill, R. E. & VlTT, D. H. The phytogeography and ecology of Macrocoma (Orthotrichaceae, Musci) in

Africa 463

Morris, J. W. Encoding the National Herbarium (PRE) for computerized information retrieval 149

Morris, J. W. & Manders, R. Information available within the PRECIS data bank of the National Herbarium,

Pretoria, with examples of uses to which it may be put 473

Obermeyer, A. A. A new combination in Commelina 437

A new combination in Eriocaulon 450

A new combination in Gerhyllis 136

A new combination in Thuranthos 139

A new species of Crocosmia 450

A new species of Strumaria 435

A new subgenus Rhadamanthopsis and two new species of Rhadamanthus 137

A new subspecies of Gladiolus microcarpus 451

A new subtribe in Liliaceae 137

.

Page

Bladsy

A reappraisal of Urginea altissima 452

The genus Sypharissa (Liliaceae) Ill

The status of Urginea epigea 139

Two new species of Commelina 436

Two new species of Cyanotis 437

Oliver, E. G. H. Some observations on two early Cape florilegia 115

Studies in the Ericoideae. III. The genus Grisebachia 65

Two new species of Ericoideae 446

RETIEF, E. A conspicuous new species of Muraltia 458

A new species of Cyphostemma from the Transvaal 460

A new species of Rhoicissus from Natal 146

REYNEKE, W. F. Three subspecies of Eucomis autumnalis 140

ROSS, J. H. An analysis of the African Acacia species: their distributions, possible origins and relationships 389

A survey of some of the pre-Linnean history of the genus Acacia 95

Rutherford, M. C. Field identification of roots of woody plants of the savanna ecosystem study area, Nylsvley 171

Survival, regeneration and leaf biomass changes in woody plants following spring bums in Burkea africana —

Ochna pulchra Savanna 531

STIRTON, C. H. Natural hybridization in the genus Eriosema (Leguminosae) in South Africa 307

New records of naturalized Rubus in southern Africa 333

Notes on the taxonomy of Rubus in southern Africa 331

Studies in the Leguminosae — Papilionoideae of southern Africa 317

The Eriosema cordatum complex. 13. 33ie Eriosema cordatum and E. nutans groups 281

The genus Dipogon (Leguminosae — Papilionoideae) 327

Taylor, H. C. & Morris, J. W. A brief account of coast vegetation near Port Elizabeth 519

Phytogeography of fynbos 231

Taylor, H. C. & van der Meulen, F. Structural and floristic classifications of Cape Mountain Fynbos on

Rooiberg, southern Cape 557

VAN DER Walt, J. J. A. & Vorster, P. J. Miscellaneous notes on the genus Pelargonium 431

Van Warmelo, K. T. Sexual nuclear division in Neocosmospora 415

Van Wyk, E. A. The identity of Eugenia woodii 142

Verdoorn, I. C. Revision of Hermannia subgenus Hermannia in southern Africa 1

The genus Cola in southern Africa 277

The genus Waltheria in southern Africa 175

Von Broembsen, H. H. A simple method for determining the density of plants in a randomly-dispersed popula-

tion 574

Investigation into the significance of plant dispersion in assessing pasture condition 569

VORSTER, P. On the identity and geographical distribution of Mariscus angularis, M. chersinus, Cyperus bullatus

and C. capensis var. polyanthemus 443

The correct author citation for Cyperus elephantinus 446

The correct author citation for Mariscus dubius 443

The correct author citation for Mariscus macropus 443

The identity and typification of Mariscus dregeanus 444

Weisser, P. J. & PARSONS, R. J. Monitoring Phragmites australis increases from 1937 to 1976 in the Siyai

Lagoon (Natal, South Africa) by means of air photo interpretation 553

Wiens, D. & barlow, B. A. Translocation heterozygosity in southern African species of Viscum 161

Republic of

South Africa

Republiek van

Suid-Afrika

K

}

BOTHALIA

Volume 13, No. 1 & 2

Edited by/Onder redaksie van

D. J. B. Killick

Botanical Research Institute

Navorsingsinstituut vir Plantkunde

Department of Agricultural Technical Services

Departement van Landbou-tegniese Dienste

South Africa/Suid-Afrika

1980

GPS (L)

CONTENTS-INHOUD

Volume 13, No. 1 & 2

Page

B/adsy

1. Revision of Hermannia subgenus Hermannia in southern Africa. I. C. Verdoorn 1

2. Studies in the Ericoideae. III. The genus Grisebachia. E. G. H. Oliver 65

3. A survey of some of the pre-Linnean history of the genus Acacia. J. H. Ross S5

4. The genus Sypharissa. A. Amelia Obermeyer Ill

5. Some observations on two early Cape florilegia. E. G. H. Oliver 115

6. Musci austro-africani II. Bryophyte collections in southern Africa and southern African type speci-

mens in the National Herbarium, Pretoria. R. E. Magill 127

7. Notes on African plants

Amaryllidaceae. R. A. Dyer; A. Amelia Obermeyer 135

Asclepiadaceae. R. A. Dyer 136

Liliaceae. A. Amelia Obermeyer; W. F. Reyneke 137

Myrtaceae. A. E. van Wyk 142

Vitaceae. Elizabeth Retief 146

8. Encoding in the National Herbarium (PRE) for computerized information retrieval. J. W. Morris 149

9. Translocation heterozygosity in southern African species of Viscum. D. Wiens and B. A. Barlow 161

10. Field identification of roots of woody plants of the savanna eco-system study area, Nylsvley. M. C.

Rutherford 171

11. Leaf anatomy of the South African Danthonieae (Poaceae). II. Merxmuel/era disticha. R. P. Ellis 185

12. Leaf anatomy of the South African Danthonieae (Poaceae). III. Merxmuellera stricta. R. P. Ellis 191

13. A synecological account of the Suikerbosrand Nature Reserve. II. The phytosociology of the Ven-

tersdorp Geological System. G. J. Bredenkamp and G. K. Theron 199

14. The vegetation of seasonally flooded areas of the Pongolo River Floodplain. H. D. Furness and

C. M. Breen 217

15. Phytogeography of Capensis. H. C. Taylor 231

16. Miscellaneous ecological notes

Notes on the use of the term “Renosterveld”. C. Boucher 237

Obituaries

J. P. H. Acocks 239

A. W. Bayer 244

R. H. Compton 244

Review of the work of the Botanical Research Institute, 1978/1979 247

Book reviews 259

Bothalia 13,1 & 2: 1-63 (1980)

Revision of Hermannia subgenus Hermannia in southern Africa

I. C. VERDOORN*

ABSTRACT

Observations on the genus as a whole are made and the subgenus Hermannia is defined. Species in southern

Africa which fall in this subgenus are revised in detail. A key is provided to these 93 species and several line

drawings illustrate some of the diagnostic features used in the key.

Ri.SU Mt

REVISION DU HERMANNIA SOUS-GENRE HERMANNIA EN AFRIQUE AUSTRALE

A des observations sur le genre dans son ensemble, cet article joint une definition du sous-genre Hermannia. Les

especes d’Afrique australe qui appartiennent a ce sous-genre sont revisees en detail. II y en a 93, pour lesquelles on

fournit une cle d 'identification et plusieurs dessins au trait qui illustrent certains des caracteres diagnostiques

utilises dans la cle.

INTRODUCTION

This revision is a contribution towards a treatment

of the genus Hermannia for the Flora of Southern

Africa. The format followed is that formerly used for

the F.S.A., which differs in several respects from that

generally employed in Bothalia.

In the course of the work thousands of herbarium

specimens, both in the National Herbarium, Pretoria,

and on loan from overseas herbaria, such as Leningrad,

Vienna, Stockholm and Zurich, were studied.

The genus, like so many others, presents great

difficulties to the taxonomist in his attempt to define

satisfactorily the sub-genera and some of the species.

Intermediates, aberrants and putative hybrids abound.

Before Mahernia L. was put into synonymy under

Hermannia L., Bentham & Hooker in their Genera

Plantarum (1862), separated these genera on the shape

of the filaments. Those of Hermannia were described

as “oblong or dilated at the apex” and of Mahernia as

“dilated in the middle”, the latter termed “cruciform”

by subsequent authors. In 1895 K. Schumann (Natiirl.

PflFam. 3, 6: 80) placed Mahernia L. as section

Mahernia in the genus Hermannia remarking that,

since the main distinguishing feature, namely cruci-

form filaments, did not hold among species subse-

quently put in Mahernia, it could not be upheld as a

genus. Later, in Engl., Monogr. Afr. Pfl. 5, 49 (1900),

K. Schumann raised the section to subgeneric rank as

subgenus Mahernia (L.) K. Schum. This has been

supported in the present study, for not only has

“cruciform” been variously interpreted but it occurs in

groups that are not otherwise related.

Just as it has been found impracticable to separate

these two genera, so the genus Hermannia, in the broad

sense, is not readily grouped into subgenera. Exploring

several features which might be diagnostic, such as

inflated calyces, horned capsules, deeply divided

leaves and different forms of pubescence, it was found

that these characters can only be used to separate

species which occur in a number of small subdivisions.

Forced back to a consideration of the shape of the

filaments, it was found that the species reviewed here

all agreed in having abovate, obtrullate, linear or

narrowly oblong filaments in which the anther

bases overlap the expanded portion of the filaments

(Fig. 1.1-1. 4). On these grounds the genus could

be separated into two subgenera as keyed out

below.

Key to subgenera

Filaments obovate, obtrullate, linear or narrowly oblong, with the anther base overlapping

the expanded portion of the filaments subgenus Hermannia

Filaments abruptly dilated in the upper half or at the apex (more or less cruciform), rarely

linear or narrowly obovate-oblong but then the anther base not overlapping the ex-

panded portion of the filaments (Fig. 1.5-1. 9) subgenus Mahernia

HERMANNIA subgen. HERMANNIA

Hermannia L. subgen. Hermannia

Subgen. Euhermannia Harv. in F.C. 1: 180 (1860);

K. Schum. in Engl., Monogr. Afr. Pfl. 5, 49 (1900).

Subgen. Acicarpus Harv. in l.c. (1860); K. Schum.,

l.c. (1900).

Type species: H. hyssopifolia L.

Herbs, undershrubs or rarely small shrubs, procum-

bent or erect, generally stellate-pubescent often with

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

glandular or simple hairs intermixed. Leaves alternate,

entire, lobed or incised; stipules sometimes foliaceous.

Inflorescence of simple or compound 1 to several-

flowered cymes, axillary or leaf-opposed. Calyx

5-lobed, tube campanulate to globose. Petals 5,

obovate to oblong, often narrowed into a claw at the

base, slightly to very strongly spirally twisted. Stamens

5, filaments expanded, obovate, obtrullate, linear or

narrowly oblong, the expanded portion overlapped by

the anther base, anthers acuminate, ciliate. Ovary

5-loculed, styles 5, cohering. Capsule loculicidally

dehiscent, 5-loculed; locules with rounded apices or

with short to long horn-like appendages.

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Fig. 1. — Stamen types in

Hermannia. 1-4, filaments

as in Hermannia subgen.

Hermannia , all with the

anther base overlapping

the dilated portion: 1,

filament obovate-cuneate,

H.althaeifolia(De Winter

& Verdoorn 9063); 2,

filament more or less

linear-oblong, H. filifolia

( Taylor 6494); 3, filament

narrowly oblong-obovate

to obtrullate, H. angularis

( De Winter & Verdoorn

9097); 4, filament narrow-

ly obovate, dilated at the

apex, H. cristata Compton

30024). 5-9, filaments as

in Hermannia subgen.

Mahernia, variously

shaped but the anther

base not overlapping the

dilated portion: 5, fila-

ment abruptly dilated in

the upper half (cruci-

form), H. aurocoma

( Acocks 12314); 6, fila-

ment abruptly dilated at

the apex (subcruciform),

H. argillicola (D inter

8041); 7, filament broadly

dilated at the apex, H.

grossularifolia ( Marloth

5300); 8, filament more

or less linear, H. oligos-

permum, ( Esterhuysen

15496); 9, filament

obovate-cuneate H. stric-

ta ( Rosch & Le Roux

549).

Fig. 2. — Some diagnostic

characters relating to

capsules and flowers in

the subgen. Hermannia.

1, capsule 5-angled,

fringed on the angles,

H. cristata (/. Thorncroft

3937); 2, capsule with a

pair of horns (developing

with age) at the apex of

each angle, H. spinosa

[Coe t zee & Werger 1739);

3, capsule not fringed on

the angles and not horned

at the apex, H. prisma-

tocarpa ( De Winter <6

Verdoorn 9067) ; 4, flower

with strongly twisted

petals (dolls roses) and

calyx not inflated, H.

prismatocarpa (after

Flower. PI. Afr. t. 1625);

5, calyx inflated, H.

comosa ( De Winter 9107);

6, flowers with petals not

strongly twisted, more or

less' bell-shaped, H.

spinosa ( Schlieben 8841).

I. C. VERDOORN

3

KEY TO THE SPECIES

la Capsules fringed on angles with crowded tentacle-like or filiform processes (Fig. 2.1):

2a Low suffrutex, several stems from a stout rootstock, leaves narrowly elliptic-oblong to lanceolate;

fringe of capsule of fairly stout 7 mm long processes; Eastern Escarpment and Soutpansberg. . . . !

1 . H. cristata

2b Suffrutex up to 2 m tall, leaves large, suborbicular; fringe of capsule of long, filiform processes; S.W.A.

2. H. merxmuelleri

lb Capsules not fringed as above:

3a Stems erect, many from a woody rootstock, short, under 30 cm long, more or less leafless in lower

half with a few large leaves and yellow flowers borne in upper half 3. H. sandersonii

3b Habit not as above:

4a Inflorescence of simple 1 -flowered cymes solitary in axils of leaves on branches of indeterminate

growth, sometimes appearing racemose when upper leaves are much reduced; petals not strongly

contorted, open flowers bell-shaped (Fig. 2.6):

5a Petals shorter than calyx, rarely about as long but then shorter than stamens:

6a Plants with at least outer branches long and trailing; leaves more or less secund:

7a Leaf-margins conspicuously and densely fringed with hoary-pilose hairs 4. H. seineri

7b Leaf-margins not as above :

8a Leaves finely and densely stellate-tomentose at least when young, hairs under 0,5 mm

long 5. H. tomentosa

8b Leaves densely or sparsely coarsely stellate, hairs comparatively long 6. H. eenii

6b Plants bushy, erect or suberect :

9a Branchlets predominantly glandular-pubescent at least in upper half (rarely in H. boragi-

niflora, e.g. Gilliland 746, not conspicuous or absent) :

10a Leaves sparsely to fairly densely stellate-pubescent:

11a Capsule densely stellate-pubescent all over; in addition to glandular hairs, branchlets

mainly appressed stellate-pubescent 7. H. boraginiflora

lib Capsule glandular-pubescent between the sutures, glabrescent:

12a Branchlets with many patent, gland-tipped hairs intermixed with long, pointed,

patent hairs 8. H. glanduligera

12b Branchlets densely pubescent with short (under 2,5 mm), glandular hairs not

obviously gland-tipped 9. H. viscosa

10b Leaves finely stellate-tomentose on both surfaces:

13a Leaves small, mostly under 1,5 cm long; glandular hairs conspicuous on most parts

of plant; S.W.A 10. H. glandulosissima

13b Leaves up to 4,5 cm long; glandular hairs conspicuous only in upper portion of

branchlets; Transvaal 11. H. grisea

9b Branchlets not predominantly glandular-pubescent:

14a Low plant with a few slender branches, calyx-lobes long, subulate; capsule with long,

slender horns 12. H. solaniflora

14b Suffrutices, capsules not or shortly horned:

15a Leaves stellate-tomentose:

16a Leaves very small, mostly under 4 mm long, often recurved, lateral branches sub-

spreading, slender, rigid 13. H. mini mi folia

16b Leaves mostly over 5 mm long and up to 22 mm long; lateral branchlets not as

above:

17a Bush, branched at base, branches long and slender, forming a rounded bush

about 1 m in diameter; leaves drying a light colour; anthers obvious, usually

drying a dark violet; occurs in south-western areas of S.W.A 14. H. engleri

17b Suffrutex, stems erect, branchlets arcuate-ascending or suberect; leaves drying

a darker colour; anthers sometimes dark violet; occurs in the northern areas of

S.W.A. around the Etosha Pan 15. H. guerkeana

15b Leaves stellate-pubescent but not tomentose; found only along east coast of Zululand

and Mozambique 16. H. micropetala

5b Petals as long as or longer than calyx :

18a Gland-tipped hairs obvious, especially on branchlets and petioles:

19a Flowers in long, distinct pseudoracemes; petals usually over 1 ,5 cm long and with a dark

blotch near base showing even on dried specimens; capsule blunt 11. H. amabilis

19b Flowers axillary or in short, not very distinct pseudoracemes; petals usually under 1 cm

long, without a dark spot near base; capsule apiculate or shortly horned at apex of sutures

18. H. modest a

18b Gland-tipped hairs not obvious on branchlets and petioles:

20a Annuals, sometimes biennial or triennial:

21a Weak annual with small flowers usually under 5 mm long; leaves broadly to narrowly

ovate-acuminate 19. H. tigrensis

21b Bi-or triennial; flowers 5 mm or longer; leaves variable 18. H. modesta

20b Suffrutices, perennial, flowers usually over 5 mm long:

22a Leaves very narrow, under 2 mm wide, appearing ericoid, clustered 20. H. linearifolia

22b Leaves not as above:

23a Capsule shortly horned; leaves and branches stellate-tomentose, not laxly stellate-

pubescent :

24a Flowers in pseudoracemes on long, slender branches 21. H. helianthemum

24b Flowers axillary :

25a Calyx 5-ribbed, lobed to less than halfway; pubescence same on calyx and leaves

22. H. damarana

25b Calyx not distinctly 5-ribbed, lobed to beyond middle; pubescence on calyx

longer and coarser than on leaves 23. H. gariepina

4

REVISION OF HERMANNIA SUBGENUS HERMANN1A IN SOUTHERN AFRICA

23b Capsule with long, spreading, pilose horns; leaves and branchlets stellate-pubescent

or stellate-tomentose (Fig. 2.2):

26a Leaves and branchlets stellate-tomentose, at least when young:

27a A small, many-stemmed leafy bush; confined to edge of Namib. . . .24. H. complicata

27b Suffrutex, divaricately branched, branchlets becoming leafless and indurated:

28a Flowers many, secund, in pseudoracemes; western Cape to S.W.A.. .25. H. trifurca

28b Flowers in axils of upper leaves; Kalahari and S.W.A 26. H. affinis

26b Leaves and branchlets stellate-pubescent, not tomentose, or glabrous and minutely

papillose:

29a Virgate, erect suffrutex, leaves and branches appearing glabrous but glutinous . .

27. H. fruticulosa

29b Rounded bush with wiry stems; leaves and branchlets stellate-pubescent, gla-

brescent; peduncles persisting, spreading, indurated, giving spiny appearance

to bush 28. H. spinosa

4b Inflorescence of simple or compound, 1 -several-flowered cymes on same plant, arranged in axils

of upper leaves and at apices of branchlets forming leafy or leafless, racemose or paniculate

cymes, if all cymes 1 -flowered, then in sparse terminal racemose cymes; petals strongly contorted

(“doll’s roses”; Fig. 2.4), open flowers not bell-shaped:

30a Leaves entire or shallowly crenate or toothed on the margins, if somewhat deeply lobed then

not halfway to midrib:

31a Plants with long, decumbent branches, inflorescence and secund branchlets ascending, all

terminating in a leafless inflorescence:

32a Leaves long and narrow 1-3 mm broad in middle, appearing linear but tapering slightly

towards base into an obscure petiole 29. H. linifolia

32b Leaves distinctly petioled, blade oblong to ovate, 5-30 mm broad in middle:

33a Flowers 1 cm long or longer, clustered at ends of branchlets on short pedicels, shorter

than flowers:

34a Pubescence sparse, stellate hairs short and appressed (see also under 30b)

89. H. procumbens subsp. procumbens

34b Pubescence conspicuous, hairs in part long, grey and matted 30. H. decumbens

33b Flowers under 1 cm long, not markedly clustered on short pedicels at ends of branchlets :

35a Branches fairly densely stellate-pubescent, often with long stiff hairs intermingled;

capsule elongating to over 1 cm long (Fig. 2.3) 31. H. prismatocarpa

35b Branches appear glabrous, sometimes sparsely and inconspicuously stellate-pubescent,

hairs few, short or some slightly longer; capsule short, broad, usually under 1 cm

long 32. H. scordifolia

31b Plants not as above but erect suffrutices, sometimes bushy with ascending branches, or low,

woody and sprawling; inflorescences various:

36a Leaves and stipules imbricate, leaves usually longer than internodes, about 12 mm long

or less (see also H. stipulacea under 36b):

37a Calyx and leaves densely and coarsely pubescent with stellate and tufted golden brown

hairs (see also under 36b) 75. H. decipiens

37b Calyx not densely and coarsely pubescent as above:

38a Leaves finely tomentose or tomentellose; calyx lobed to about one third of its length:

39a Leaves distinctly petioled; mature flowers 10 mm long or more; calyx inflated. . . .

33. H. temifolia

39b Leaves sessile or subsessile; open flowers less than 10 mm long; calyx subinflated

34. H. trifoliata

38b Leaves rough with minute fringed scales or hirsute; calyx lobed halfway or beyond:

40a Leaves broadest at apex, rough with minute fringed scales 35. H. eoncinnifolia

40b Leaves narrowly oblong, sparsely hirsute with long, bulbous-based hairs. .36. H. muirii

36b Leaves and stipules not obviously imbricate as above:

41a Leaves generally broadest in lower half or oblong-elliptic to broadly oblong:

42a Pubescence mainly rough, especially on branchlets, mostly of dense or sparse, stiffly

stellate or bulbous-based hairs:

43a Calyx inflated, narrow at mouth, sinuses narrow (Fig. 2.5):

44a Calyx narrowly urceolate, lobes deltoid; leaves coarsely and densely pubescent

with tufted or stellate hairs from a scaly base (see also under 41b). .54. H. salviifolia

44b Calyx globose; leaves mostly basal (also under 42b) 46. H. comosa

43b Calyx not inflated or if subinflated then campanulate to salver-shaped:

45a Leaves mainly basal or in lower half of plant, sometimes some leaves deeply

lobed or pinnatisect; flowers in long racemose or paniculate cymes comprising

the upper half to two-thirds of plant (see also under 30b):

46a Flowers usually small, up to about 6 mm long on dried specimens; leaves

narrowly to broadly oblong or ovate-oblong, coarsely and evenly lobed,

15-20 cm long 91. H. macra

46b Flowers over 6 mm long and up to 10 mm long:

47a Leaves oblong-lanceolate to narrowly lanceolate, dentate to deeply dentate

92. H. juttae

47b Leaves ovate, coarsely crenate to palmatisect 93. H. paucifolia

45b Leaves cauline:

48a Flowers small, up to 6 mm long on dried specimens, numerous and short-

stalked in each inflorescence:

49a Calyx subinflated, about 4 mm long on dried specimens 37. H. floribunda

I. C. VERDOORN

5

49b Calyx not subinflated, under 4 mm long on dried specimens:

50a Leaves distichous, pubescence harsh, mainly of 2 mm long hairs . . 74. H. disticha

50b Leaves not distichous, mainly under 15 mm long; hairs shorter than 2 mm:

51a Leaves white-tomentose on lower surface (see also under 41b). .66. H. alnifolia

51b Leaves not white-tomentose on lower surface, pubescence harsh....

38. H. bryoniifolia

48b Flowers over 6 mm long on dried specimens:

52a Stipules leaf-like (see also under 42b) 44. H. althaeifolia

52b Stipules not leaf-like:

53a Leaves thick-textured, densely pubescent, large, up to 4,5x3 cm; flowers

large, congested in short inflorescences 39. H. cordifolia

53b Leaves and flowers not as above:

54a Pubescence mainly hispid, calyx sparsely so to glabrous; leaves mucro-

nate 40. H. hispidula

54b Pubescence mainly of stellate or grouped hairs, calyx subdensely to

densely stellate-pubescent:

55a Leaves strongly crisped as well as crenate on margin; calyx densely

stellate with hairs rather long (see also under 41b); Western Province

41. H. rugosa

55b Leaves shallowy crenate, not strongly crisped as well, stellate pubes-

cence sometimes dense on calyx but hairs not long; Eastern Province

42. H. althaeoides

42b Pubescence mainly tomentose, especially on branchlets which are tomentose to

smoothly canescent-tomentose or silvery with fringed, pitted scales :

56a Branches silvery with fringed pitted scales (see also under 30b) 88. H. pulverata

56b Branches canescent-tomentose:

57a Flowers mostly under 6 mm long on dried specimens:

58a Leaves tomentose on both surfaces; branchlets always canescent. .43. H. minuti flora

58b Leaves white-tomentose on lower surface only; branchlets not canescent

(see also under 41b) 66. H. alnifolia

57b Flowers mostly over 6 mm long on dried specimens:

59a Stipules leaf-like; calyx hirsute, not narrowed at mouth 44. H. althaeifolia

59b Stipules not large and leaf-like, calyx not hirsute if hairs long then grouped on

stalks :

60a Calyx inflated, narrowed at mouth:

61a Flowers secund; leaves shortly petioled 45. H. johanssenii

61b Flowers not secund; leaves, especially the lower, long-petioled . . .46. H. comosa

60b Calyx not inflated or subinflated:

62a Leaves not normally crisped as well as crenate on margins :

63a Inflorescence of 2-flowered cymes in axils of leaves 47. H. incana

63b Inflorescence of long, terminal, leafless, compound cymes. .42. H. althaeoides

62b Leaves normally crisped as well as crenate on margins :

64a Leaves broadly ovate to sub-orbicular 48. H. vest it a

64b Leaves narrowly ovate to ovate-oblong:

65a Calyx subinflated; branches of inflorescence glabrescent, dark and

slender 49. H. amoena

65b Calyx not subinflated; branches of inflorescence not glabrescent. . . .

50. FT. disermifolia

41b Leaves broadest at apex or generally broadest in upper half but varying in shape from

filiform, narrowly oblong, oblanceolate and broadly oblong-rectangular to subglobose:

66a Calyx inflated, narrowly urceolate to subglobose, narrowed at mouth:

67a Leaves softly tomentose on both surfaces:

68a Tomentum smoothly uniform; plants up to 1 m tall or taller, calyx subglobose

51.//. mucronulata

68b Tomentum intermixed with long, villose hairs; low plant ±30 cm high; calyx

more or less urceolate 52. H. suavis

67b Leaves not, or not normally, softly tomentose on both surfaces:

69a Leaves usually toothed at apex and in upper half, stellate-pubescent with short

hairs, glabrescent; calyx subglobose .53. H. hyssopifolia

69b Leaves entire or occasionally obscurely lobed at truncate apex or in upper third;

stellate pubescence rough, usually of thick and tufted hairs from a glandular or

tubercled base; calyx urceolate (see also under 41a) 54. H. salviifolia

66b Calyx not inflated :

70a Leaves velvety tomentose on both surfaces, entire or very shallowly crenate in

upper half:

71a Stipules not leaf-like:

72a Flowers small, 3-4,5 mm long on dried specimens, crowded in compound

cymes, terminal on lateral branchlets; leaves fairly broad, often suborbicular,

crenate at least at broad apex 55. H. holosericea

72b Flowers normally 5 mm long or longer; leaves mostly narrow, usually entire

and mucronate:

73a Weak twiggy plants; calyx wide, lobed to middle and beyond, scaly, hairs

inconspicuous 56. H. lavandulifolia

73b Virgate shrubs up to 1 m tall or taller; calyx strongly 5-10-ribbed, lobed in

upper half, densely pubescent, hairs conspicuous and foxy-coloured

57. H. odor at a

6

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

71b Stipules leaf-like:

74a Leaves petioled, petiole not completely hidden by sessile stipules; flowers

yellow:

75a Calyx narrowly obconic, strongly 5-10-ribbed; flowers appear narrow-

waisted 58. H. sulcata

75b Calyx not narrowly obconic:

76a Calyx broadly campanulate; flowers in terminal, leafy, paniculate cymes

59. H. velutina

76b Calyx tubular-campanulate, lobed in upper half; cymes 1-2-flowered in

lax, terminal, leafless, racemose inflorescences 60. H. gracilis

74b Leaves subsessile, short petiole hidden by broad, sessile stipules; flowers

usually garnet-red 61. H. diversistipula

70b Leaves not velvety tomentose on both surfaces, pubescence rough, or smooth and

white only on lower surface, if appearing silvery pubescent on both surfaces then

calyx and leaves covered by pitted, fringed scales:

77a Leaves on both surfaces, and calyx covered by pitted, fringed scales giving plants

a silvery grey appearance:

78a Calyx deeply to narrowly campanulate, lobed in upper third; leaves cuneate;

frequent in eastern and western Cape and O.F.S 62. H. cuneifolia

78b Calyx shallowly campanulate to saucer-shaped, lobed almost to middle:

79a Leaves usually deeply divided, if not, then longer than broad, oblong and

only slightly narrowed towards base (see also under 30b) 88. H. pulverata

79b Leaves never deeply lobed, subglobose to obovate-cuneate:

80a Flowers small, up to 5 mm long on dried specimens ; leaves suborbicular,

distinctly crenate 63. H. desertorum

80b Flowers over 5 mm long on dried specimens; leaves obovate-cuneate,

obscurely 3- or more-lobed at broad apex; west coast of Namaqualand

and Namib in S.W.A 64. H. pfeilii

77b Leaves and calyx not silvery grey from the covering of fringed pitted scales :

81a Leaves with a white, stellate or cobwebby, appressed tomentum on lower

surface only :

82a Calyx densely fawn to brown-tomentose with lepidote stellate pubescence,

hairs obvious and straw-coloured 65. H. multiflora

82b Calyx thinly stellate-pubescent, the hairs very short or hispid, not densely

brown-tomentose :

83a Flowers small, under 5 mm long on dried specimens, short-stalked, nume-

rous in short, terminal, compound cymes 66. H. alnifolia

83b Flowers usually over 5 mm long; inflorescence not as above:

84a Flowers in long, terminal, leafless, racemose or paniculate cymes:

85a Plants much branched from base, branches arcuate-ascending

67. H. muricata

85b Tall, virgate shrubs:

86a Inflorescence of lax, paniculate cymes; peduncles and pedicels

well developed 68. H. repetenda

86b Inflorescence of racemose cymes, 1-3-flowered and 1-3 at a node,

peduncle usually aborted 69. H. rigida

84b Flowers in short, few-flowered cymes, terminal on short lateral branchlets

or leaf-opposed along the leafy branches:

87a Leaves petioled; cymes very slender, many, axillary or leaf-opposed

along leafy branches, curved downwards in helicoid fashion

70. H. helicoidea

87b Leaves sessile or subsessile, cymes terminal on short lateral branch-

lets:

88a Leaves fascicled, strongly crisped, mostly narrowly oblong-cuneate

71. H. asp era

88b Leaves single at a node, distichous, broadly obovate-oblong to

ovate-oblong, broadly cuneate or rounded at base ; only sometimes

white-tomentose on lower surface 74. H. disticha

8 1 b Leaves variously pubescent but not with a white, stellate or cobwebby tomen-

tum on lower surface only:

89a Leaves suborbicular or broadly oblong to oblong-obovate, only slightly

longer than broad :

90a Flowers crowded in dense heads, densely hispid from long, tubercle-based

hairs of calyx 72. H. conglomerata

90b Flowers in lax terminal cymes:

91a Small plant with slender intricate branches and small leaves. .73. H. micrantha

91b Shrublet 50-90 cm tall with distichous leaves with an average size of

20x16 mm (leaves rarely white-tomentose dorsally) 74. H. disticha

89b Leaves distinctly longer than broad:

92a Pubescence on calyx dense, golden brown coloured:

93a Bracteoles long, subulate, conspicuous at base of flowers:

94a Leaves small, about 12 mm long, often subimbricate; bracteoles

reaching lower third of calyx 75. H. decipiens

94b Leaves variable, usually over 15 mm long; bracteoles often reaching

to top of calyx 76. H. involucrata

93b Bracteoles not long and subulate:

95a Leaves densely and persistently pubescent on both surfaces . . 77. H. pillansii

95b Leaves laxly to subdensely stellate-pubescent, hairs long from a scaly

base 78. H. stipulacea

I. C. VERDOORN

7

92b Pubescence on calyx not dense and golden brown :

96a Branches glabrous or pubescent to scaly-pubescent, at least on new

growth :

97a Calyx parchment-like, angled at the sinuses, broad-based with broad,

ovate lobes; south western Cape (see also under 96b). . . .85. H. angularis

97b Calyx not as above:

98a Leaves filiform, ericoid, acute, glabrous and clustered; Great and

Little Karoo 79. H. filifolia

98b Leaves not as above, if some appear filiform (conduplicate) then

not completely glabrous or plants restricted to Coastal Fynbos:

99a Branches, leaves and calyx glabrous; Western Cape

80a. H. denudata var. denudata

99b Branches, leaves and calyx not totally glabrous:

100a Leaves mostly acute and dentate in upper half; Transvaal

80b. H. denudata var. erecta

100b Leaves broadest at or near apex which is rounded, mucronate

or lobed :

101a Calyx half or more than half as long as petals, sinuses

V-shaped:

102a Leaves glabrescent at least on upper surface, often flat

and up to 12 mm broad at apex 81. H. flammea

102b Leaves persistently stellate-pubescent on both surfaces,

folded, at most 5 mm broad at apex 82. H. flammula

101b Calyx usually less than half as long as petals, wide and

shallow at maturity, sinuses wide 83. H. joubertiana

96b Branches very rough with prominent scattered tubercle-bases or harsh

stellate pubescence, rarely glabrescent but rudimentary tubercle-bases

can be seen :

103a Inflorescence of long, terminal, leafless, racemose or paniculate

cymes; petals exserted from calyx-tube by about half 84. H. scabra

103b Inflorescence terminal on leafy branches and on very short, leafy

lateral branchlets; petals shortly exserted from conspicuous calices:

104a Calyx glabrous except for small stellate hairs along margins of

lobes 85. H. angularis

104b Calyx rough with tubercle-bases of stellate or grouped hairs. . . .

86. H. rudis

30b Leaves all, or at least some, deeply toothed or lobed to about halfway to midrib or pinnately or

palmately divided:

105a Branchlets or branches of inflorescence canescent with silvery-fringed or stellate scales:

106a Calyx large, inflated, subglobose, up to 10 mm diameter; cauline leaves well developed

87. H. abrotanoides

106b Calyx small, not inflated, about 5 mm long; leaves mainly basal (see also under 30a). . . .

88. H. pulverata

105b Branchlets and branches of inflorescence not canescent, usually laxly and roughly stellate-

pubescent :

107a Stems decumbent with only apical portion ascending; leaves cauline and usually secund,

erect from trailing stem (see also under 30a) 89. H. procumbens subsp. myrrhifolia

107b Stems not decumbent; leaves not as above:

108a Suffrutex, up to 60 cm tall, much branched above, branches slender; leaves basal and

cauline deeply and finely pinnately divided up to 25 mm long; inflorescence of terminal,

leafless, racemose cymes 90. H. confusa

108b Low plants with many stems from a woody base; leaves mainly basal; inflorescence of

long, terminal, leaf-less, paniculate cymes :

109a Flowers small, up to 6 mm long on dried specimens; leaves more or less oblong,

about 1 5-20 mm long, broadly and unevenly lobed, sometimes deeply so but not to

midrib (see also under 30a) 91. H. macro

109b Flowers over 6 mm long:

110a Leaves oblong-lanceolate to narrowly lanceolate, 20-60 mm long, coarsely crenate

or toothed, not palmatisect (see also under 30a) 92. H. juttae

110b Leaves ovate in outline, 15-30 mm long, usually palmatisect with segments irre-

gularly twice-lobed (see also under 30a) 93. H. paucifolia

1. Hermannia cristata H. Bol. in J. Linn. Soc.,

Bot. 25: 156 (1889); Hook. f. in Curtis’s bot. Mag.

t.7173 (1891); K. Schum. in Engl., Monogr. Afr.

Pfl. 5: 53 (1900); R. A. Dyer in Flower. PI. Afr. 30,

1. 1 1 69 (1954). Type: Transkei, Kokstad, Tyson 1689

(BOL, lecto!; Z!; SAM!); Orange Free State,

Cooper 900; Lesotho, Cooper 2001; Natal, Fannin 9.

H. cristata var. geoides Beauv. in Bull. Soc. bot. Geneva

2,3: 133 (1911). Type: Transvaal, foothills of the Drakensberg

near Shilovane, Junod 2856.

Suffrutex with several stems from a stout, woody

rootstock; stems usually less than 30 cm tall, simple

to laxly branched in second year, fairly laxly stellate-

pubescent and with minute glands or minute gland-

tipped hairs intermingled. Stipules linear-filiform or

linear-lanceolate, about 5 mm long and less than, or

up to 1 mm broad, sparsely stellate-pubescent. Leaves

shortly petiolate; blade elliptic, elliptic-oblong, ovate-

elliptic or lanceolate, 2-6,5 cm long, 0, 3-2,2 cm

broad, crenate-dentate, crenations sometimes small and

distant, concolorous, upper surface with impressed

veins, fairly densely to sparsely pubescent with simple

or 2- or more-rayed, bulbous-based hairs, lower sur-

face with prominent veins and sparsely stellate-pubes-

cent with hairs usually shorter than on upper surface;

petiole 1 , 5-6 mm long. Inflorescence axillary and

terminal, flowers usually solitary on long peduncles

up to 5 cm long; pedicels short, about 3 mm long;

bracts linear-filiform, 1-2 on peduncle and 2 at base

of pedicel, about 1-3 mm long. Calyx 8-14 mm long,

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

lobed to midway or just beyond, stellate-pubescent

without and usually intermingled with minute gland-

tipped hairs. Petals 1,1-1, 9 cm long, “crimson”,

“red-orange”, “orange”, “red”, upper third suborbi-

cular, slightly broader than long, narrowing into a

claw with infolded margins. Stamens united at base to

a long stipe, free part of filaments hyaline, narrowly

obovate, broader at apex, about 1 cm long, obscurely

stellate on shoulders; anthers about 1 cm long, ciliate

with basal portion overlapping dilated filaments.

Ovary stipitate, 5-angled, densely stellate-pubescent,

stellate pubescence on keel of angles shortly stalked;

styles 6 mm long, minutely and sparsely hairy with

short patent hairs. Capsule oblong-orbicular in outline,

deeply 5-angled, roughly pubescent with short-rayed,

stellate hairs and fringed on keel of angles with

crowded, tentacle-like processes up to 7 mm long

(developed from stalks of stellate hairs), calyx sub-

persistent at base, ultimately reflexed and withered to

expose a 4 mm long stipe; seeds reniform-orbicular,

plicate. Figs 1.4 & 2 . 1 .

Found in grassveld on rocky slopes, in gorges and

along rivers. Recorded from mountain grassland

following the line of the Drakensberg from Mt Currie

District in the Transkei northwards through Natal

and the eastern Orange Free State, Piet Retief and

Swaziland to the Soutpansberg in northern Transvaal.

Cape. — Mt Currie: near Kokstad, Haggarth s.n.; Tyson 1689

(BOL); Mt Currie, Goossens 354.

O.F.S. — Vrede: 10 km S. of P.O. Vrede, Acocks 21947.

Natal. — Lion’s River: Zwarthof location, Moll 1143.

Newcastle: Normandien, Sim 2884. Nkandhla: Nsuzi, Gerstner

603. Underberg: Springvale farm, Strey 7697. Utrecht: 6 km

S.E. of P.O. Groenvlei. Codd & Dyer 6280.

Swaziland. — Hhohho: Miller’s Falls, Compton 26193;

Pigg’s Peak, Wells 2023.

Transvaal. — Barberton: Saddleback Mtn, Galpin 434.

Belfast: Dullstroom, Noome in TRV 20783. Carolina: Carolina,

Leipoldt in TRV 18627. Ermelo: Athole Pasture Station,

Norval 38. Groblersdal: 1,6 km S. of Nebo, Acocks 20859.

Letaba: east slope Piesangskop, Scheepers 1013. Lydenburg:

12 km N. of Lydenburg, Codd & De Winter 3274. Middelburg:

Tautesberg, Young 4189. Nelspruit: Schagen, Liebenberg 3335.

Pietersburg: Haenertsburg, Potts 4696. Piet Retief: Piet Retief,

Galpin 9633. Pilgrim’s Rest: between Pilgrim’s Rest and Sabie,

Rogers 23216. Soutpansberg: Louis Trichardt, Breyer 24366;

Downs; Junod 4237. Wakkerstroom: North Hill, Galpin 9809.

Until recently this species was unique in Southern

Africa on account of the fringe of long processes on

the angles of the capsule. In 1958 a new species was

found on the Brandberg in South West Africa with

similarly fringed capsule-angles. It was published as

H. merxmuelleri in 1961. The processes on this new

species are longer and more hair-like than those on

H. cristata and in many other respects these two

species differ appreciably. For instance, the leaves of

H. merxmuelleri are suborbicular and the petioles

usually well over 10 mm long, while in H. cristata the

leaves are usually elliptic-oblong and the petioles

I , 5-6 mm long.

From description there are 2 species in America

that resemble ours in having fringed capsules.

2. Hermannia merxmuelleri M. Friedrich in

Mitt. Bot. StSamml., Munch. 4: 167 (1961); M. Fried-

rich et al. in F.S.W.A. 84: 16 (1969). Type: South

West Africa, Brandberg, Zisabschlucht, Merxmiiller

& Giess 1664 (M, holo.; PRE!).

Shrub about 2 m high, branches stellate-tomcntose

with long, patent, minutely gland-tipped hairs inter-

mingled. Stipules subulate, 2-6 mm long, stellate-

pubescent. Leaves petiolate; blade broadly ovate-

orbicular, 1-3,5 cm long, 1,2-3, 5 cm broad, crenate-

dentate except at the subcordate base, apex rounded,

truncate or emarginate, concolourous, densely stellate

on both surfaces with short hairs, veins impressed

above, prominent beneath; petiole 10-30 mm long,

stellate-pubescent, often with patent gland-tipped

hairs intermingled. Inflorescence of 1- to few-flowered

cymes, axillary and terminal; peduncle 5-12 mm long,

pubescent as on petiole. Calyx about 14 mm long,

lobed to just beyond middle, roughly pubescent with

small, stellate hairs, rays mostly short, some longer,

from a minutely bulbous base. Petals glabrous, grey-

green and purplish blue at the apex on inner face,

17-20 mm long, upper third suborbicular, narrowed

into a long claw with infolded margins. Stamens with

hyaline, narrowly obovate-cuneate, apparently

glabrous filaments which are attached to the long

stipe at base, free portion about 6 mm long; anther

8 mm long ciliate, pointed, overlapping filament at

base. Ovary about 5 mm long, densely hairy with

long-stalked, long-rayed, stellate hairs, stalks subulate

or filiform; stipe about 6 mm long; styles 7,5 mm long.

Capsule about 1,5 cm long, deeply 5-angled, crest of

angles furnished with 2 to 3 series of long, filiform

processes (elongated stalks of stellate hairs?), rest of

capsule shortly stellate-pubescent; stipe partly hidden

at base by persistent calyx and reflexed petals.

Found “among rocks on hillsides”. Recorded from

the slopes and gorges of the Brandberg, South West

Africa.

S.W.A. — Omaruru: Brandberg, Zisabschlucht, Merxmiiller &

Giess 1664; White Lady Valley, Macdonald 588 (BM).

To date this remarkable species is known only from

the Brandberg in South West Africa. It is the tallest

known Hermannia in southern Africa, being 2 m tall.

In the fringed angles of the capsule it resembles only

one other species on the subcontinent, namely H.

cristata, in which, however, the processes of the fringe

are slightly stouter. In habit these two species differ

widely, for in H. cristata the stems are under 30 cm

tall and subherbaceous. For further differences see

under H. cristata (p. 7).

3. Hermannia sandersonii Harv. in F.C. 1 : 200

(1860); Wood & Evans, Natal Plants 1: 19, t.20

(1893); K. Schum. in Engl., Monogr. Afr. Pfl. 5: 60

(1900). Type: Natal, “Port Natal”, Sanderson s.n.

(TCD, holo.; PRE, photo.!; K!; S!; PRE!).

Low suffrutex; stems many from a thick, woody

rootstock, simple or laxly branched, more or less

leafless in lower half, subdensely and coarsely hairy

with bulbous-based stellate hairs, rays about 1 mm

long, acute. Stipules ovate, ovate-lanceolate or

linear-lanceolate, about 7 mm long, 1-3,5 mm broad

at base, rather thick-textured (oily?), laxly stellate-

pubescent dorsally. Leaves shortly petiolate; blade

from fairly narrowly to broadly oblong-elliptic, or

some leaves obovate to suborbicular, 2-5 cm long,

0,7-3, 5 cm broad, discolorous, upper surface with

impressed veins, subdensely to laxly stellate-pubes-

cent, hairs few, long, acute and from a bulbous base,

undersurface with prominent veins densely whitish to

yellowish tomentose, tomentum of matted stellate

hairs, margins crenate-dentate; petiole 2-6 mm long,

stellate-pubescent. Inflorescence of a few 1-3-flowered

cymes borne in the upper half; peduncles 2-10 mm

long, shaggily stellate-pubescent; pedicels not readily

distinguishable from peduncles, 3-10 mm long,

pubescence as on peduncle; bracts linear-lanceolate

to subulate, 2-5 mm long. Calyx about 7 mm long,

lobed almost to middle, stellate-pubescent on the

I. C. VERDOORN

9

strong, prominent ribs and margins, lobes deltoid.

Petals yellow, recurved in upper half, about 9 mm

long, about 6,5 mm broad in the broadly rounded

upper third, narrowing to a 2 mm waist and produced

into a claw with infolded margins of which the edges

are membranous and obscurely hairy. Stamens

about 6,5 mm long; filaments joined at base, hyaline,

narrowly obovate, about 3 mm long with 1 or 2

stellate hairs on shoulders; anthers about 4,5 mm

long, overlapping filaments for 1 mm. Ovary 2,5

mm long, 2 mm diam., 5-angled, densely hairy with

sessile and stalked stellate hairs, stipe 5 mm long.

Capsule about 9 mm long, 7 mm diam., 5-lobed and

bluntly 5-umbonate at apex, pubescent with sessile

and stalked hairs, shortly stipitate, with the persistent

calyx like a disc at base. Seeds reniform-orbicular,

black, plicate.

Found on grassy hills and in grassy gorges. Re-

corded from the Camperdown, New Hanover,

Pietermaritzburg and Durban Districts of Natal.

Natal. — Camperdown: Inchanga, Marioth 4090; Medley

Wood 4654; 6536; Umlaas, Maurice Evans 548 A (NH); Shong-

weni Dam, Ross 1292; 413 (NH). Durban: “Port Natal”,

Sanderson 9244; without exact locality, Gerrard 1117 (W).

New Hanover; King’s Hill Halt, Gordon-Gray 6320. Pieter-

maritzburg: in valley 5 km from Table Mtn, Stirton 1024.

This species is readily distinguished by its general

appearance. The many, short, subherbaceous stems

from a woody rootstock are more or less leafless in

the lower half and bear comparatively large dis-

colorous leaves and a few yellow flowers in the upper

half. A collector records that the flowers develop in

advance of the leaves.

H. sandersonii is poorly represented in herbaria

and most of the specimens seen were collected in the

last century.

4. Hermannia seineri Engl, in Bot. Jb. 55: 371

(1919). Syntypes: South West Africa, Otjitjika,

Dinter 2871 (B|; SAM!); Epata, Trotha s.n.; Seiner

284; 390.

Stems procumbent, slender, long, wiry, laxly

branched near base, obviously hoary-pilose intermixed

with appressed stellate pubescence. Stipules narrowly

linear-lanceolate, about 4 mm long, hoary-pilose.

Leaves petiolate; blade narrowly ovate- to obovate

oblong, 2-3,5 cm long, up to 9 mm wide, usually

folded, subrotund at apex, broadly cuneate at base,

upper surface roughly stellate-pubescent, lower sur-

face stellate-pilose with long hoary hairs; margins

obscurely serrate and densely hoary-pilose; petiole

about 1 cm long. Inflorescence of simple 1 -flowered

cymes solitary in axils of leaves; peduncle slender,

straight, 2-2,5 cm long, hoary-pilose; pedicels

cernuous, slender, about 2-3 mm long; bracts about

3 mm long, narrowly linear to subulate. Calyx

hoary-pilose outside and on margins, lobed to beyond

middle, about 7 mm long, lobes 5 mm long. Petals

“orange”, “cream”, about 5 mm long, broadly

oblong to subglobose, narrowed at base into a short

claw, margins inrolled on claw and basal portion of

blade. Stamens with filaments broadly obovate,

narrowing at apex to an acute connective, 2,5 mm

long, sparsely pilose on shou de s; anthers erect,

about 5,5 mm long, basal portion overlapping

expanded filaments, sparsely pilose. Ovary about 2

mm long, sparsely minutely stellate with simple hairs

intermixed; styles about 5 mm long, sparsely pilose.

Capsule 6 mm long, hoary-pilose, 5-umbonate or

shortly horned at apex. Seeds reniform, dark brown,

several in a locule.

Found on level, well-drained sand in grassland and

on loose sand dunes in north-eastern South West

Africa and in Botswana.

S.W.A. Otjitjika: Dinter 2871 (SAM); Omaramba

Omatako, Schoenfelder 5190.

Botswana. — Ngamiland: Pandamatanga road, c. 100 km

S. of Kazungula on road to Nata, Verhagen & Barnard 162.

Before 1977 only two specimens had been seen of

this species, Dinter 2871 (SAM) and Schoenfelder 190

(PRE), both with a hoary pubescence unusual for the

genus. Not only did these specimens answer well to

the description of H. seineri , but Dinter 2871 is a

syntype of the species and Schoenfelder 190 had been

matched with Dinter 2871 in B (now destroyed).

Good material ( Verhagen & Barnard 162) collected in

Botswana during 1977 matched these specimens in

detail and confirmed that H. seineri evidently is a

good, well-defined species.

For H. seineri var. latifolia Engl, see under H.

eenii (p. 10).

5. Hermannia tomentosa ( Turcz .) Schinz ex Engl.

in Bot. Jb. 55: 371 (1919); Burtt Davy, FI. Transv.

1:267 (1926) partly, excluding the synonym H.

johannisburgiana; Wild in F.Z. 1:542 (1960); M.

Friedrich et al. in F. S.W.A. 84:22 (1969), partly.

Type: Transvaal, Magaliesberg, Zeyher 137 (LE!;

PRE!; SAM!; W!; Z!).

Mahernia tomentosa Turcz. in Byull. mosk. Obshch. 31:218

(1858); Harv. in F.C. 1:219 (1860), excluding the description

and the specimen cited (see note). Type: as above.

Hermannia brachypetala Harv. in F.C. 1:202 (1860): K.

Schum. in Bot. Jb. 10; 42 (1889); in Engl., Monogr. Afr. Pfl.

5:86 (1900). Syntypes: Transvaal, Magaliesberg, Burke &

Zeyher s.n. (S; PRE, photo.!); Cape, “Zooloo Country” (sic),

but should be Kuruman area, Miss Owen s.n. (PRE, photo.!).

H. tomentosa var. brevifolia Engl, in Bot. Jb. 55: 371 (1919).

Syntypes: several, including Dinter 277 (Z ! ; SAM!) and Range

1026 (SAM!).

Perennial, branches long, at least the outer decum-

bent and trailing, young branchlets stellate-tomentose,

hairs mostly short, interspersed with longer, tufted

bristles and rarely with short, gland-tipped hairs,

tomentum sometimes disappearing leaving only

tufted bristles. Stipules linear-acuminate to subulate,

2-5 mm long, tomentose. Leaves petiolate; blade ±

oblong to narrowly oblong-elliptic, 8-50 mm long,

5-15 mm (rarely 18 mm) broad, rounded or sub-

truncate at apex, rounded to somewhat cuneate at

base, margin obscurely to distinctly toothed except

in lower third, stellate-tomentose on both surfaces,

hairs very short, tomentum persisting at least on

undersurface; petiole 3-15 mm long. Inflorescence of

simple 1-flowered cymes, solitary in axils of leaves on

branches of indeterminate growth; peduncle straight,

10-15 mm long, tomentose; pedicels cernuous, about

2,5-4 mm long, bracteate at base; bracts 2 or 3,

sometimes 1 of these on peduncle, about 3 mm long,

narrowly lanceolate-acuminate. Calyx lobed almost

to base; tube about 2 mm long, lobes 4,5-6 mm long,

reflexed in fresh flower, tomentose on outside and on

upper portion of lobes within. Petals pink, mauve,

off-white with deep pink base, or “yellow and red”,

shorter than calyx, more or less oblong-globose,

narrowing at base into a claw, margins involute on

claw and sometimes narrowly so on blade also, only

obscurely lobed at base of blade, lobes not thickened

but sometimes darker in colour, with short, appressed

hairs within upper part of claw or base of blade;

blade about 3 mm long, 2 mm broad with the margins

sometimes partly inrolled; claw 1 mm long. Stamens

with filaments broadly obovate, narrowing at apex to

10

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

an acute connective, 2,5 mm long, sparsely hairy on

shoulders; anther erect, about 5,5 mm long, 1 mm

broad near base, narrowing to an acute apex, basal

portion overlapping filament for 2 mm, shortly and

sparsely hairy. Ovary obovate or suborbicular, about

2,5 mm long, 1,75 mm broad near apex, densely

tomentose in upper portion not covered by filaments;

styles very slender, straight, more or less cohering to

each other, up to 5 mm long, sparsely and shortly

hairy in lower half. Capsule shortly and densely

tomentose (off-white tomentum), about 5,5-7 mm

long, 5 mm broad, obovate-oblong, more or less

truncate or crowned with 5 short, rounded processes,

rarely up to 2 mm long. Seeds (on dried specimens)

reniform, brown, corrugated, usually under 2 mm

long, about 3 maturing in a cell.

Found on grassy plains and on slopes of mountains

and hills. Recorded from Transvaal, Orange Free

State, northern Cape and from South West Africa,

especially in southern and central areas; also in

Botswana and Rhodesia.

Cape. — Barkly West: near railway station, Smith 2328.

Gordonia: Gemsbok Park, Mata Mata, Werger 1493; Story

5482. Hay. Wolhaarkop, Esterhuysen 2350. Hopetown: Hope-

town. H. Bolus 1833 (Z). Kimberley: Kimberley, Acocks 43.

Kuruman: Wonderwerk, Esterhuysen 2247. Mafeking: Mafe-

king, Brueckner 518. Prieska: Prieska, Bryant 281. Vryburg:

Palmyra, Rodin 3517.

O.F.S.— Boshof: Boshof, Schweickerdt 1106. Fauresmith:

near Petrusburg, Codd 3401.

Transvaal. — Without locality, Holub s.n. Bloemhof:

Christiana, Burtt Davy in Govt. Herb. 12786. Marico: Zeerust,

Leendertz in TRV 11270. Middelburg: Loskopdam reserve,

Theron 1585. Potchefstroom : Boskop, Louw 736. Pretoria:

Magaliesberg, Zeyher 137; Schlechter 3593; Wonderboom

Reserve, Repton 2887. Rustenburg: Rustenburg, Meeuse 10489.

Schweizer Reneke: Schweizer Reneke, Burtt Davy 12786; 1687.

Soutpansberg: Waterpoort, Moss 4599. Waterberg: Warmbaths,

Collett 515. Naboomspruit: Naboomspruit, Galpin M. 36.

S.W.A. — Gibeon: Codd 5868. Karibib: Karibib, De Winter

2445 A. Okahandja: Okahandja, Dinter 277 (Z). Okavango:

Between Katwitwi & Makambo Camp, De Winter 3862.

Rehoboth: Rehoboth, Basson 6. Warmbad: Ariamsvlei,

Ortendahl 650. Windhoek: Auasberge, Merxmiiller & Giess 786.

This species is characterized by the long trailing

branches with secund leaves and a strong taproot with

long secondary roots spreading horizontally. The

young branches and leaves are stellate-tomentose, the

short fine tomentum usually persisting at least on the

lower surface of the leaves.

Throughout the wide area of distribution, and

especially where it overlaps with that of other species,

occasional specimens are found that appear to be

forms or hybrids of the species, for instance a small-

leafed form occurs fairly frequently. Until a more

in-depth study is possible these are here included under

H. tomentosa. One such specimen is De Winter 3378,

mentioned under H. engleri Schinz in F. S.W.A. 84; 14.

Dinter 5164, also mentioned in F. S.W.A. l.c., ap-

proaches H. minimifolia form or hybrid of that species.

In the past this species has been confused with

H. lancifolia Szyszyl. This is mainly due to Harvey,

who in F.C. 1; 219 cites an unnumbered specimen of

Burke & Zeyher from the Magaliesberg under H.

tomentosa although it is obviously H. lancifolia. It has

been established that H. lancifolia does not occur on

the Magaliesberg and the locality description of the

species must thus be considered to be incorrect.

Burke & Zeyher did collect both these species.

H. tomentosa , which grows on the Magaliesberg, is

represented in their collections by Burke 314 and

Zeyher 137, while H. lancifolia , which they collected

before they reached the mountain, is represented by

Burke 255 and Zeyher 107.

6. Hermannia eenii Bak. f. in J. Bot., Lond.

39; 126 (1901). Type: South West Africa, Damara-

land, Een s.n. (BM, holo.; PRE, photo.!).

H. angolensis K. Schum. in Warb. Kunene-Samb. Exped.

302 (1903); Excell & Mendonca in C.F.A. 1,2:195 (1951);

Wild in F.Z. 1: 543 (1960). Type: Angola, between Kubango

and Kuito, Baum 514 (Z, holo!; PRE, photo!; K; S; W!).

H. longiramosa Engl, in Bot. Jb. 55: 365 (1919). Type: South

West Africa, Grootfontein, Engler 6240.

H. seineri var. latifolia Engl., l.c. 55:371 (1919). Syntypes:

South West Africa, Otjosondjou, Seiner 459; Aris Aukas,

Dinter 725 (SAM!).*

H. brandtii Engl, ex Dinter in Reprium nov. Spec. Regni veg.

18: 25, (1922). Based on Dinter 725 from Otavi,* nom subnud.

H. karakowisensis ined., name on Dinter 7315 (PRE!; Z!).

Perennial, prostrate with long slender runners,

coarsely stellate-pubescent, often intermixed with

bristle-like, -f patent setae from a bulbous base, 1-2

to several in a cluster, especially obvious in young

growth, hairs straw-coloured to bright yellow.

Stipules linear- to narrowly ovate-acuminate, about 3

mm long, sparsely setose, setae up to 1 mm long

from a bulbous base. Leaves petiolate; blade linear-

oblong, lanceolate-elliptic to ovate-elliptic or oblong-

elliptic, varying in size, sometimes on same plant

(leaves on ultimate branchlets much smaller), 1-6,5

cm long, 0,4-2, 5 cm broad, coarsely stellate-pubes-

cent, from densely to sparsely so, more or less rounded

at base, rounded, subtruncate or acute at apex,

obscurely to distinctly crenate at least in upper half;

petiole 2-20 mm long, coarsely, stellate or with tufts

of setae (few hairs in a tuft). Inflorescence of 1 -flowered

cymes, axillary on long branches and crowded at

apices of ultimate branchlets; peduncles slender, up

to 3 cm long; pedicels short, up to 3 mm long; bracts

2-3 at apex of peduncle, linear-lanceolate, 2,5-4 mm

long, 0,5-1 mm broad near base, setose or stellate.

Calyx lobed almost to base, coarsely stellate outside

and on upper portion of lobes within, tube about 2

mm long, lobes about 6 mm long. Petals usually pink,

sometimes white or white with a dark patch near

base, 3-6 mm long, shorter than calyx, oblong or

oblong-obovate, obscurely clawed, about 4 mm long

and 2 mm broad, margins slightly to clearly inrolled

on both blade and claw, shallowly lobed at base of

blade, inner surface of petals with minute appressed

hairs about midway between marginal lobes, apex

rounded to substruncate. Stamens with obovate,

hyaline filaments which are setose on shoulders, 2

mm long, 2 mm broad at apex with a short apicule,

joined to anther near its base; anthers erect, about 6

mm long, basal portion overlapping expanded portion

of filament for 1 mm, thence narrowing to an acute

apex, minutely hairy. Ovary stellate-pubescent with

bristle-like hairs which are sometimes golden-yellow

and often concentrated at apex, minutely papillose

under stellate pubescence; styles very slender, erect,

about 6 mm long, with short, spreading hairs in

lower half. Capsule coarsely stellate-pubescent and

with bristle-like hairs especially at apex, about 5,5

mm long, 5 mm diam., obscurely 5-lobed, lobes with

blunt apices, not produced into horns. Seeds brown,

reniform, corrugated.

Found in sandy parklands, on dunes or in savanna.

Recorded from the northern and eastern regions of

South West Africa and the northern Cape where it

* Dinter 725, collected at Aris Aukas, and the same number

collected at Otavi are represented in SAM. They are obviously

not the same gathering, the former dated 1918-12-29 and the

latter 1908-1 1-29, but they both fall within the range of H. eenii

Bak. f.

I. C. VERDOORN

11

borders on South West Africa and Botswana. Also in

Botswana, Rhodesia, Zambia and Angola.

Cape. — Gordonia: Kalahari Gemsbok National Park, Mata

Mata, Werger 1824.

S.W.A. — Grootfontein: Auros, Dinter 5455; Simkue, Story

6151; Tsumeb, Dinter 7464; Oliewenhof, Merxmiiller & Giess

30141. Okahandja: Hochfeld, Boss in TRV 32832. Okavango:

Makambo Camp, De Winter 3863; Kake Camp, De Winter &

Wiss 4384. Otjiwarongo: N.E. of Otjiwarango, De Winter 2845.

This species is distinguished from H. tomentosa

by the leaves, even the young ones, being coarsely

stellate-pubescent, often intermixed with bristle-like

setae, and not finely stellate-tomentose as in H.

tomentosa. The general colour of the plants is very

variable; some appear grey, others golden yellow. In

fresh specimens the leaves are crystalline. The setae,

which are found in most parts of the plant, and

especially on very young leaves, vary from whitish

to golden yellow.

Forms or hybrids appear to occur in H. eenii, as in

many other species of Hermannia. For example

Hanekom 2230 from Barkly West District appears to

be a hybrid between H. tomentosa and H. eenii.

7. Hermannia boraginiflora Hook., Ic. PI. t. 597

(1843); Harv. in F.C. 1:201 (1860); K. Schum. in

Engl., Monogr. Afr. Pfl. 1 : 267 (1926); Wild in F.Z.

1: 543 (1961). Type: Transvaal, Magaliesberg, Burke

293 (K, holo.; PRE, photo.!; SAM!).

H. tephrocarpa K. Schum. in Notizbl. bot. Gart. Mus. Berl.

2: 304 (1899). Type: Transvaal, Lydenburg, Wilms 68 (K; PRE,

photo. ! ; Z !).

Suffrutex, bushy, much branched from base;

branches and branchlets suberect or ascending,

stellate-pubescent or pubescent with short, patent,

gland-tipped hairs, sometimes with stellate hairs

intermixed, rarely with a few simple patent hairs as

well. Stipules from under 1 mm to 4 mm long, up to

1,5 mm broad, subulate to deltoid-oblong. Leaves

petiolate, blade elliptic, oblong or obovate-oblong,

acute or more often broad at the apex and 3-toothed,

up to 27 x 14 mm, upper leaves sometimes narrowly

elliptic, entire and acute, 6-20 mm long, 3-4 mm

broad, stellate-pubescent on both surfaces, more

densely so on the lower, sometimes some gland-

tipped hairs present and very minute scales; margins

usually crenate-dentate at least in upper portion,

upper leaves often entire; petiole 1,2-10 mm long.

Inflorescence axillary, 1 -flowered; peduncle stellate-

pubescent or with usually gland-tipped, patent hairs,

about 6-22 mm long, 2-3-bracteate at apex, sometimes

with a fourth lower down on peduncle; bracts 1-2 mm

l°ng; pedicels 3-7 mm long, pubescent like the

peduncles. Calyx lobed to beyond the middle, stellate-

pubescent without, often intermingled with gland-

tipped hairs, especially on margins of lobes; tube

about 2 mm long; lobes deltoid-acuminate, about 5,5

mm long. Petals pink or off-white with a pair of

oblong, deep pink patches on lower half of inner

face, oblong-cuneate, about 6 mm long, lobed about

midway, lobes ciliate and some pubescent on inner

face of petal, lower half with inrolled margins.

Stamens with hyaline, obovate-elliptic filaments,

about 4 mm long, ciliate on the shoulders; anthers

acute, about 5 mm long, ciliate, overlapping the

filament for 1,5 mm. Ovary stipitate, stellate-pubes-

cent, stipe 1-2 mm long; styles about 6 mm long.

Capsules densely stellate-pubescent, about 6 mm long,

5 mm diam., truncate at apex with 5 pairs of horns

about 1 , 5-2 mm long) at apex of sutures.

Occurs on rocky hillsides, on mountain slopes, in

red loamy soils and on old lands. Found in ’the

northern Cape, Natal, Swaziland and Transvaal.

Also recorded from Rhodesia, Botswana and Mozam-

bique.

Cape.— Barkly West: Barkly West, Acocks 462. Hay Hay

Griquatown, Drege in SAM 29814. Mafeking: Mafeking

Brueckner 265. Vryburg: Vryburg, Rodin 3496; Codd 1304

52^NH)L' — Ubomlx): coastal be*1 on margins of pans, Tosh

Swaziland. — Lubombo :

29561 (NBG).

Stegi, Blue Jay Ranch, Compton

Transvaal.— Groblersdal: Marble Hall, Codd & Verdoorn

10372. Letaba: Merensky Nature Reserve, Gilliland 784

Lydenburg: West of Blyde River Canyon, Leistner 3264.

Messina: Dongola, Verdoorn 2101. Middelburg: Loskopdam

Reserve, Theron 1800. Petersburg: Pietersburg, Rogers in

TRV 15490. Potgietersrus: Potgietersrus, Galpin 8926. Pretoria:

Magaliesberg, Schlechter 3656. Rustenburg: Rustenburg

Story 6054. Soutpansberg: Soutpan, Obermeyer, Schweickerdt &

Verdoorn, 142. Waterberg: Naboomspruit, Galpin M 35.

Characterized by the roughly stellate-pubescent

calyx and capsule.

In Gilliland 746 from Letaba, gland-tipped hairs

are inconspicuous or absent. This may be an aberrant

form. It approaches the east coast species H. micro-

petala, but its petals are rather long. It falls within the

distribution area of H. glanduligera but has the

stellate pubescence of H. boraginiflora.

8. Hermannia glanduligera K. Schum. in Verh.

bot. Ver. Prov. Brandenb. 30; 232 (1888); in Engl.,

Monogr. Afr. Pfl. 5: 57 (1900); Wild in F.Z. 1: 545

(1961); M. Friedrich et al. in F.S.W.A. 84: 15 (1969).

Type: South West Africa, Amboland, Olukonda,

Schinz s.n. (Z, holo!; PRE, photo.; LE!; BOL!;

GRA!).

H. viscosa sensu Burtt Davy, FI. Transv. 1 : 267 (1926).

Perennial, bushy, much branched from base,

30-60 cm tall; branches and branchlets pubescent

with 1- to several-celled glandular hairs and many

gland-tipped hairs intermixed with some long, patent,

pointed hairs. Stipules 0,75-2 mm long, up to 1 mm

broad. Leaves petiolate; blade on upper branches more

or less narrowly oblong, very variable in size and on

some branches much reduced, ranging from about

6 mm x 1,5 mm to 30x14 mm, rounded or shortly

cuneate at base, usually broadly acuminate towards

apex, margins serrate, coarsely and usually sparsely

stellate-pubescent on both surfaces, with some

glandular hairs intermixed and sometimes gland-

tipped hairs present as well, especially along margins

and with numerous, minute papillae on the surfaces,

petiole 1 ,5-15 mm long, the lower longer. Inflorescence

of 1 -flowered cymes in axils of leaves, solitary but

appearing racemose in specimens where leaves on

ultimate branches are much reduced; peduncle

usually short, 5-14 mm long, occasionally (as in the

type) up to 25 mm long, persistent, 2-3-bracteate at

apex (bracts under 1 mm long), pubescent with

glandular or gland-tipped hairs and sometimes with

long, pointed, patent hairs intermixed; pedicels

2-4 mm long, pubescence like that on peduncles.

Calyx about 7 mm long, lobed to beyond middle,

pubescent on outside mainly with glandular hairs, but

sometimes stellate hairs present and long non-glandu-

lar pointed hairs as well; lobes pubescent inside with

appressed, non-glandular hairs. Petals pale pink to

pale violet, sometimes with a red base, narrowly

oblong-cuneate or suborbicular-cuneate, 2-6 mm

long, 1-2,5 mm broad in upper half, margins inrolled

in lower half only, auricles thickened and infolded

12

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

with a thickened ridge between them, auricles and

ridge densely to laxly setose. Stamens from about

6 to 10 mm long; filaments up to 4 mm long, hyaline,

obovate, setose on shoulders; anthers up to 8 mm

long, erect, pointed, overlapping filaments at base,

sparsely patently pubescent (on a few specimens

anthers aborted, blunt, about 1,5 mm long). Ovary

stipitate at base (stipe up to 1,5 mm long), pubescent

with short glandular hairs and with pointed setae

along the sutures. Capsule up to 7 mm long and about

as wide, glandular-hairy on surface between sutures,

setae along sutures, blunt or with horns eventually

up to 2 mm long; seeds more or less reniform, corru-

gated, granular and with minute papillae.

Found in sandy to semi-sandy bushveld of the

warmer regions, usually in dry riverbeds, along

streams or rivers, sometimes as a weed in lands or

broken ground. Recorded from the drainage basin of

the Okavango and Etosha Pan in northern South

West Africa, through the northern and eastern

Transvaal and Swaziland to Zululand. Also occurs in

Botswana, Zambia and eastwards through Rhodesia

and Malawi to Mozambique.

Natal.— Umfolozi: Umfolozi River, Wager in TRV 22384;

between Black Umfolozi and Nongoma, Acocks 11671.

Swaziland. — Hlatikulu: Usutu Poort, Compton 30099.

Transvaal. — Barberton: Komatipoort, Marloth 5452; Kaap-

muiden, Rogers 25064; Galpin 745; Crocodile River Drift,

Bolus 7685. Letaba: Kruger National Park, Letaba-laagwater-

brug, Van der Schijff 543; 2782; Mamaranda, Breyer in TRV

17607. Lydenburg: Driekop, Meeuse 9297 ; Burgersfort,

Meeuse 9293. Messina: Messina, Pole Evans in Govt. Herb.

2940; 1703. Nelspruit: Research Station, Liebenberg 2725;

Pretoriuskop, Codd & De Winter 4941. Pilgrims Rest: Pilgrims

Rest Ranh & Schlieben 9705. Potgietersrus: Maastroom,

Strey & Schlieben 8661. Sibasa: Punda Milia, Codd 4234.

Soutpansberg: Soutpan, Obermeyer, Schweickerdt & Verdoorn

142; Mara, Schlechter 4616.*

S.W.A. — Okavango: between Shamvura and Kangongo,

De Winter 4210. Ovamboland: Oshigambo, Sylvi Soini s.n.

The isotype gathering represented in the Leningrad

Herbarium is not typical of the species. It must have

been collected in optimum conditions. The leaves are

all well developed, the branchlets straight and the

peduncles much longer than usual.

K. Schumann includes this species under those

with carpels not horned but they are sometimes

shortly horned.

Giess 3178 from Kaokoland may belong here, but

the specimen is incomplete and cannot be identified

with certainty.

The species is characterized by glandular and gland-

tipped hairs intermingled with long, pointed hairs on

the branches, a lax, coarse, stellate pubescence on the

leaves and the capsule with glandular setae only

along the sutures. Noteworthy, too, are the short

horns on the capsule and the thick auricles on the

petals.

9. Hermannia viscosa Hiern, Cat. Afr. PI. Welw.

1:89 (1896); K. Schum. in Engl., Monogr. Afr.

Pfl. 5: 87 (1900), as viscida\ M. Friedrich et al. in

F. S.W.A. 84: 22 (1969). Syntypes: Angola, Welwitsch

4709; 4710 (BM; PRE, photos!); 4704.

Suffrutex, intricately branched; branches predomi-

nantly glandular-pubescent, hairs short, spreading

and only slightly swollen at apex, occasionally with

stellate hairs intermixed. Stipules subulate, glandular-

* Schlechter 4616 (quoted by Engler as a syntype of H.

modesta var. elatior K. Schum. subvar. macropetala Engl.) in

Herb. Z is H. modesta scnsu lat., while the specimens of this

number in SAM and PRE are H. glanduligera.

pubescent, 2-3 mm long. Leaves petiolate; blade 10-25

mm long, oblong-obovate, usually broad at apex,

narrowing slightly towards rounded base, serrate-

dentate except near base, fairly densely pubescent

with stellate and glandular hairs, stellate hairs few-

rayed from a minute central scale, glandular hairs

short, usually slightly longer and more conspicuous

on margins and prominent veins; petiole glandular-

pubescent, 5-20 mm long. Inflorescence of 1 -flowered

cymes, solitary in axils of upper leaves; peduncles

about 10 mm long; pedicels up to 5 mm long; bracts

small, subulate, about 1 ,5 mm long. Calyx 6-10 mm

long, lobed to beyond middle, glandular and stellate-

pubescent without. Petals “scarlet”, or “violet-rose

turning almost purple”, about 6-8 mm long, shorter

or about as long as calyx, more or less oblong,

narrowed in lower third into a claw with infolded

margins apparently glabrous, with 2-coloured ridges

above claw on inner face. Stamens 6-7 mm long,

filaments hyaline, broadly obovate-cuneate, ciliate on

shoulders; anthers longer than filaments and over-

lapping them at base. Ovary densely pubescent, styles

about 3 mm long. Capsule somewhat glabrescent,

about 8 mm long, with 5 pairs of short horns.

Recorded from Kaokoland in South West Africa

and northwards in Angola.

S.W.A. — Kaokoland: west of Otjihu, De Winter & Leistner

5691.

Differs from the related H. glandulosissima princi-

pally in the habit, being a low shrublet with intricate

branches, not erect, with fairly short, slender lateral

branches as in H. glandulosissima. Furthermore, the

leaves of H. viscosa are larger, the petioles longer and

the petals glabrous.

The description is based on the one specimen seen,

De Winter & Leistner 5691. It answers well to the

original description of the species and is cited in

F. S.W.A. under H. viscosa together with Merxmiiller

& Giess 1432 and 1461.

10. Hermannia glandulosissima Engl, in Bot. Jb.

55: 370 (1919); M. Friedrich et al. in F.S.W.A. 84: 15

(1969). Type: South West Africa, Namib, Pforte,

Dinter 182 (not traced).

Suffrutex, erect, over 30 cm tall, much branched,

ultimate branchlets short, slender; branches and

branchlets fairly densely pubescent with short,

patent, glandular hairs, some of them gland-tipped,

very rarely a few long, pointed hairs occur. Stipules

narrowly deltoid to subulate, ± 1 mm long, stellate-

pubescent and usually with one long, pointed hair at

apex. Leaves petiolate; blade narrowly to broadly

oblong-obovate, 2,5-12 mm long, 1-6 mm broad,

toothed in upper portion, densely stellate-pubescent,

hairs short, spreading or suberect (tufted), especially

on veins beneath; petiole 1-4 mm long, pubescent.

Inflorescence axillary, solitary, 1 -flowered; peduncle

5-9 mm long, 2-3-bracteate at apex; bracts about 1 ,5

mm long, more or less subulate, stellate-pubescent

and with a few long, pointed hairs; pedicels 2-4 mm

long, almost straight to cernuous, pubescent like the

peduncle. Calyx about 7,5 mm long, tube about 1,5

mm long, lobes 6 mm long, outside stellate-pubescent

over glandular papillae, inside the lobes are pubescent

with appressed hairs. Petals narrowly to broadly

oblong-spathulate, 3-4,5 mm long, up to 2 mm wide,

margins inrolled on claw dnly, auricles at centre

infoled, fairly thinly fringed with whitish, diaphanous

setae. Stamens about 7 mm long; filament hyaline,

obovate, 2,5 mm long, pubescent on shoulders;

I. C. VERDOORN

13

anthers about 6 mm long, overlapping filaments for

about 1 mm. Capsule 3-5 mm long, shallowly 5-lobed,

mucronate at apex of lobes, glandular-papillate on

surface, stellate or with clustered rather short setae

on sutures; stipitate.

Recorded from Swakopmund District (eastern

border) and Karibib northwards to the borders of

Etosha and the eastern border of Kaokoland.

S.W.A. — Kaokoland: west of Otjitjekua, Giess 3178. Karibib:

“Pforte”, Dinter 181 (SAM), Sphinx, Dinter 8449. Kavango:

Katuitui, Giess 9496. Omaruru: Hunabschlucht, Boss A47.

Swakopmund: Arandis, Boss 2197.

Close to H. glanduligera and H. viscosa but

differs from both mainly in the finer and denser

pubescence on the leaves.

In the original description the locality of the type of

H. glandulosissima, Dinter 182, is given as “Damara-

land, Namib, bei Pforte, Kilom 82”. This means it

cannot be from Pforte in the Outjo district. This is

confirmed by a specimen in SAM, Dinter 181, also

from “Pforte, Kilom 82”, but on this label is the

information that it was collected on the “Reise

Okahandja-Swakopmund”. This places it, presum-

ably, in the district Karibib.

Dinter 8449 was labelled H. rotundifolia Pillans sp.

nov., by N. S. Pillans but this name was never

published.

11. Hermannia grisea Schinz in Vjschr. naturf.

Ges. Zurich 62: 676 (1917); Burtt Davy, FI. Transv.

1:267 (1926). Type: Transvaal, Manaka, Brakrivier,

Schlechter 4631 (Z, holo.!; PRE, photo.!; K, photo.!;

BOL!; PRE!; SAM!; GRA!).

H. sideritifolia Engl, in Bot. Jb. 55: 369 (1919), nom. illegit.

Type: as above.

Suffrutex, about 60 cm tall, branched in lower half,

stems reddish brown but covered with grey-stellate

tomentum; branches virgate, finely grey stellate-

tomentose and leafy in lower half, in upper half

leaves much reduced, and pubescence reduced mainly

to short simple, patent, gland-tipped hairs, some

stellate hairs present. Stipules subulate, under 1 mm

long. Leaves petiolate; blade narrowly oblong-

elliptic, 1,5^4, 5 cm long, 0,2-1 cm broad, densely

and finely stellate-tomentose on both sides, margin

sometimes with short gland-tipped hairs, entire or

dentate except in the lower portion; petiole 1-6 mm

long. Inflorescence axillary, solitary, 1 -flowered, in

axils of upper, much reduced, leaves and so appearing

racemose; peduncles with short, patent, gland-tipped

hairs, 15-21 mm long, minutely bracteate at apex;

pedicels 2-6,5 mm long, pubescence same as on

peduncles, but with stellate hairs as well. Calyx up

to 9 mm long, densely and finely stellate-tomentose,

glandular hairs obscured, lobes up to 6,5 mm long,

inner face lined with appressed hairs. Petals 3-6 mm

long, 2-3 mm broad in upper portion, ± obovate-

spathulate, margin inrolled on claw only, auricles not

thickened but with pointed, diaphanous setae along

the inner edge, inner central face minutely brown-

papillate. Stamens with hyaline, obtrullate (kite-

shaped) filaments, ciliate in upper portion, ± 3 mm

long; anthers about 7 mm long, with the basal portion

overlapping the filaments, sparsely patently pubescent.

Ovary shortly (0,5 mm) stipitate, 2,5 mm long, 1 mm

diam., finely stellate-pubescent and glandular with

tufted setae at the apex; styles ± 5,5 mm long,

sparsely pubescent in basal portion. Capsule 6-8

mm long, about 6 mm diam., densely and finely

stellate-tomentose (glandular hairs or papillae ob-

scured), stipe short, short horns ultimately produced

to about 1,5 mm long; seeds subreniform, corrugated

and minutely granular to tuberculate.

Found on sandy ridges and flats and in brackish

ground. Recorded from Waterberg and Soutpansberg

Districts; also in Botswana.

Transvaal.— Pretoria: Pienaars River, near Makapan’s

Stad, Verdoorn 2510. Soutpansberg: near Salt Pan, Schweickerdt

& Verdoorn 553. Waterberg: Ons Hoop, Codd 8469; Ellisras

Acocks 8812.

Characterized by the finely stellate-tomentose

stems and branches of which the upper portions

become pubescent with short, gland-tipped hairs, the

densely stellate-tomentose leaves which become

reduced to bracts in the upper half giving the appear-

ance of a raceme to the axillary 1 -flowered inflore-

scences and the capsule which is finely stellate-tomen-

tose.

12. Hermannia solaniflora, K. Schum. in Bot. Jb.

10:43 (1888); in Engl., Monogr. Afr. Pfl. 5:58

(1900); M. Friedrich et al. in F. S.W.A. 84: 21 (1969).

Type: South West Africa, Hykamchab, Marloth

1200 (holo., not traced; PRE!).

Low, frutescent plant, 10-40 cm high with a few

slender branches from or from near base; branches

simple, densely to subdensely appressed stellate-

pubescent. Stipules small, up to 1 mm long, subulate

to oblong, acuminate, often obscured by the pubes-

cence. Leaves all, or those in upper portion, often

reducing markedly in size, petiolate; blade elliptic,

narrowing slightly to base and apex or apex truncate

or rounded and crenate, about 8x4 mm, lower leaves

usually broadly elliptic to oblong, crenate in upper

half, often up to 12x9 mm; stellate-pubescent on

both surfaces, densely so beneath; petiole 1-3 mm

long. Inflorescence of simple, 1 -flowered cymes

solitary in axils of leaves and appearing racemose

towards ends of the branches because of reduced

leaves; peduncles slender 10-15 mm long, persistent

stellate-pubescent; pedicels cernuous, 1-3 mm long,

usually more densely stellate-pubescent than pe-

duncles. Calyx lobed to beyond middle, stellate-

pubescent; lobes long, attenuate, 5-8 mm long,

reflexed at one stage in open flower; tube about

2,5 mm long. Petals red, spathulate, about 6 mm

long, 2, 5-3, 5 mm broad in upper half, margins

inrolled on the claw, at one stage in open flower

reflexed from the middle. Stamens with dark purple,

erect anthers about 6-10 mm long, pubescent;

filament obovate, hyaline, no pubescence seen,

about 3 mm long, overlapped by anther base for

about 1 mm. Ovary about 3 mm long, and 2 mm

broad, pubescent with short, stellate or simple

bulbous-based or gland-tipped hairs and at apex

with long, acute hairs; cells 2-horned. Capsule with

short, stellate and glandular hairs, with long, acute

hairs at apex, horns 1-2,5 mm long.

Found in sandy beds of watercourses or ravines

from Swakopmund to Kaokoland in South West

Africa.

S.W.A. — Damaraland: Welwitschia, Galpin & Pearson 7469.

Hereroland: Hereroland, Dinter 1495 (Z); Dinter 16 (Z).

Kaokoland: Anabib, Story 5742. Karibib: Between Okahandja

and Swakopmund, Dinter 180 (SAM). Omaruru: Numas

Ravine, Brandberg, De Winter & Hardy 8219; Giess 3610; 9150.

Outjo: Outjo, Giess, Volk & Bleissner 6279. Swakopmund:

Swakopmund, De Winter & Hardy 8028. Windhoek: between

Windhoek and Walvis Bay, Esdale sub Rogers 15331.

14

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Characterized by short petals which reflex, stellate-

pubescent branches, shortly stellate-glandular and

setose capsules with short horns and stipe, and long

attenuate calyx-lobes.

13. Hermannia minimifolia M. Holzhammer in

Mitt. bot. StSamml. Munch., 1:349 (1953); M.

Friedrich et al. in F.S.W.A. 84: 17 (1969). Type:

South West Africa, Swakopmund, south of Kuiseb,

Strey 2442 (M, holo. ; PRE!).

Suffrutex, erect, up to 60 cm tall, ultimate branch-

lets short, slender, rigid, and divaricate, stellate-

tomentose. Stipules about 1 mm long, stout, tomen-

tose. Leaves not, or very shortly, petiolate; blade

densely stellate-tomentose, on average about 2,5

mm long and 1 mm broad, usually shorter than 5 mm,

broadly elliptic or obovate-cuneate, recurved, apex

acute or truncate and 2-3-lobed, margins entire or

very sparsely toothed in upper half; petiole very

short or absent. Inflorescence axillary, 1 -flowered;

peduncle about 2 mm long, stellate-tomentose;

pedicel about 3 mm long, roughly stellate-tomentose;

bracts caducous. Calyx roughly stellate-tomentose

without, lobed almost to base, tube about 2 mm

long, lobes about 5 mm long, long-acuminate ap-

pressed hairy in upper portion within. Petals “dark

brownish purple”, about 4,5 mm long; limb oblong,

margins somewhat incurved, lobed at the base; claw

with involute margins, ridge on inner face of petal

between lobes minutely setose. Stamens with broadly

obovate, hyaline filaments, hairy on shoulders,

about 2,5 mm long, 2 mm broad; anthers about 6

mm long, ciliate. Capsule about 5 mm long, sub-

globose, roughly stellate-tomentose, not horned;

styles about 6 mm long, minutely hairy. Seeds 1-2

in a cell, reniform, about 2 mm wide.

Recorded from the dunes south of Kuiseb, South

West Africa.

S.W.A. — Swakopmund: south of Kuiseb, Strey 2442.

Characterized by the very small, sessile, or almost

sessile, leaves and the short, rigid, ultimate branches

which become indurated.

Dinter 5164 approaches this species and may fall

within the variation when more material is seen. The

stipules, however, differ in not being stout. Dinter

5164 is mentioned in F.S.W.A. as an atypical repre-

sentative of H. engleri. Giess 13426 also approaches

H. minimifolia having the rather fleshy or stout

stipules but the leaves are longer. Both Dinter 5164

and Giess 13426 differ from H. engleri mainly in the

habit, the latter being a rounded bush, much branched

from the base with long, slender branches.

14. Hermannia engleri Schinz in Vjschr. naturf.

Ges. Zurich 55: 243 (1910), nom. nov. ; M. Friedrich

et al. in F.S.W.A. 84: 14 (1969). Type: South West

Africa, Luderitz, Awichab, Dinter 1085 (Z!; PRE,

photo. !).

H. dinter i Engl, in Bot. Jb. 38:591 (1907), non Schinz. Type: as

above. H. seitziana Engl, in Bot. Jb. 55: 370 (1919). Syntypes:

South West Africa, Sandverhaar, Dinter 1178 (SAM!); Schaefer

285.

Suffrutex, branched at base, forming a rounded

bush 30-90 cm tall and about 1 m in diam.; branch-

lets many, slender, long, ascending, the outer decum-

bent, stellate-tomentose with short and long inter-

mixed hairs. Stipules 1,5-3 mm long, linear or

linear-subulate, tomentose, hairs erect at apex.

Leaves petiolate; blade linear to narrowly oblong,

7-22 mm long, 1,5-7 mm broad, apex broadly

acuminate, rounded or truncate and lobed, base

rounded or broadly cuneate, both surfaces finely

stellate-tomentose with hairs appressed to subspread-

ing (mostly under 0, 5 mm long), veins usually strongly

impressed above and prominent beneath; petiole

1,5-5 mm long, stellate-tomentose with hairs mostly

over 0,5 mm long. Inflorescence consisting of 1-

flowered, axillary cymes, solitary in axils of leaves;

peduncle 2-7 mm long, stellate-tomentose with short

hairs; pedicels 2-4 mm long, stellate-tomentose with

hairs slightly longer than on peduncle; bracetoles 0 or

minute, up to 0,5 mm long. Calyx tube about 1 , 5

mm long, lobes 5-7 mm long, stellate-tomentose

without with hairs both short and long, some being

up to 0,5 mm long, obscurely appressed hairy in

upper part within. Petals mauve, pink, oblong-cuneate,

about 4 mm long, 2 mm broad, claw about 1 , 5 mm

long, margin inrolled on claw and base of blade,

lobed at apex of claw, lobes almost meeting on the

ventral surface which bears a ridge (sometimes

obscure) between the lobes and is glabrous or micro-

scopically hairy on the ridge. Stamens with broadly

obovate, hyaline filaments, about 2 mm long, 2 mm

broad near apex, with hairs on shoulders, apical

connective short; anthers about 5 mm long, over-

lapping filaments at base, ciliate. Ovary densely

stellate-tomentose (with rays sometimes over 0 , 5 mm

long), 4 mm long, 4 mm broad near apex, with 5

blunt horns at apex; ovules probably 6 in a cell.

Capsule densely stellate-pubescent with 5 short, blunt

horns about 1 mm long.

Found in sand in the dry southern areas of South

West Africa.

S.W.A. — Bethanien: Sandverhaar, Dinter 1178 (SAM); 4220;

Pearson 4660 (SAM; BOL). Keetmanshoop: Griindorn,

Dinter 5054. Liideritzbucht : Awichab, Dinter 1085 (Z). Warm-

bad: Swartkop-Noachabeb, Ortendahl 391.

This species is described as forming a low bush up

to 1 m in diameter, a feature which cannot very well

be judged from herbarium specimens. Authentic

specimens in herbaria show many slender branches

from the base and this may help to identify the species.

To date it has been recorded only from the dry

southern areas of South West Africa. Closely related

species are H. tomentosa (see notes under that species)

and H. guerkeana. H. geurkeana is recorded from the

higher rainfall areas further north and does not show

many slender branches from the base.

In the F.S.W.A. two specimens, De Winter 3378

and Dinter 5164, are mentioned under H. engleri as

differing somewhat from the typical. De Winter 3378

could rather be a form of H. tomentosa , while Dinter

5164 is nearest H. minimifolia (see note under that

species).

15. Hermannia guerkeana K. Schum. in Verh.

bot. Ver. Prov. Brandenb. 30: 231 (1888); K. Schum.

in Engl., Monogr. Afr. Pfl. 5: 57 (1900); Wild in F.Z.

1: 542 (1961); M. Friedrich et al. in F.S.W.A. 84: 15

(1969). Type: South West Africa, Omadongo,

Amboland, Schinz 601 (Z!; PRE, photo.!).

H. hereroensis Schinz in Vjschr. naturf. Ges. Zurich 55 : 242

(1910). Type: South West Africa, Waterberg, Dinter 355

(Z! ; PRE, photo.!).

Suffrutex, erect, up to 60 cm tall, with outer

branches sprawling in sand, branches with a fairly

smooth, stellate tomentum, rarely with a few longer

bristle-like hairs intermixed. Stipules linear-subulate,

1-3 mm long, tomentose. Leaves petiolate; blade

linear-elliptic to narrowly oblong or oblong-elliptic,

acute, rounded or truncate at apex, somewhat

I. C. VERDOORN

15

cuneate to rounded at base, finely and thinly stellate-

tomentose on both surfaces, entire or toothed in

upper half, often folded, 10-25 mm long, 3-8 mm

broad; petiole 1,5-4 mm long. Inflorescence of 1-

flowered cymes solitary in the axils of the leaves;

peduncles 9-20 mm long; pedicels 1-2 mm long,

cernuous; bracts 2-5, about 1 mm long, narrowly

deltoid-subulate. Calyx about 7 mm long, divided to

below middle, densely stellate-pubescent, hairs short

from a scaly base. Petals “pink”, “reddish” or

“plum-coloured”, oblong-cuneate, 3 mm long, 1,75

mm broad, margins inrolled in lower, claw-like half

with thickened auricles or horns about midway which

almost meet on inner face of petal, inner face slightly

ridged where blade and claw meet, minutely setose on

the ridge, midrib distinct. Stamens with membranous,

broadly obovate filaments which are setose on

shoulders, about 2 mm long, 1 ,75 mm broad; anthers

about 4-6 mm long, sparsely patently pubescent,

overlapping filaments at base. Ovary densely stellate-

pubescent (hairs short), 5-lobed, lobes rounded at

apex; styles 5, stiff, thread-like, cohering, about 5 mm

long. Capsule densely stellate-tomentose, 3-5 mm

long, 5-umbonate at apex.

Recorded from Kaokoland, Outjo, Otjiwarongo

and Grootfontein in South West Africa, as well as

the northern regions of Botswana. Found on sandy

ground.

S.W.A. — Grootfontein: Karakowisa, Dinter 7301 ; Tsumkwe,

Giess, Watt & Snyman 11033. Kaokoland: Andana Karvapa,

Merxmuller & Giess 1358A. Otjiwarongo: near Otjiwarongo,

Rodin 2719; S. of Okaputa, Tolken & Hardy 957; Waterberg,

Dinter 355 (Z). Outjo: Etosha Pan, Walter 423 (WIND);

Etosha National Park, Le Roux 1109; Namutoni, Le Roux 458.

Botswana. — Maun, Erens 330; Story 4641.

H. guerkeana is near to and may be confused with

H. engleri and H. tomentosa. It occurs in the northern,

slightly moister, regions of South West Africa as

compared with H. engleri which is found in the dry

south, while H. tomentosa occurs in the northern

Cape, the Transvaal, Orange Free State and central

South West Africa.

16. Hermannia micropetala Harv. in F.C. 1: 201

(1860); K. Schum. in Engl., Monogr. Afr. Pfl. 5; 58

(1900); Wild in F.Z. 1:543 (1961). Type: Mozam-

bique, Delagoa Bay, Forbes s.n. (K, holo. ; PRE!).

H. phaulochroa K. Schum. in Notizbl. bot. Gart. Mus. Berl.

2:303 (1899). Syntypes: Mozambique, Schlechter 11576

(PRE!; LE!; Z! ; BOL!; \N\)\ Junod 29.

Suffrutex about 60 cm tall, with slender, terete,

woody stem; branches terete, densely stellate-pube-

scent. Stipules small, about 1-3 mm long, stellate-

pubescent. Leaves petiolate; blade elliptic-oblong to

obovate-oblong, narrowing slightly to base, somewhat

acuminate to apex or broad and rounded or subtrun-

cate at apex, sometimes toothed in upper portion,

stellate-pubescent on both surfaces, sometimes glabre-

scent at least in part on upper surface; petiole 2-6 mm

long, stellate-pubescent. Inflorescence of 1 -flowered

cymes, solitary in axils of leaves; peduncle 7-14 mm

long, usually persistent, stellate-pubescent, minutely

bi-bracteate at apex; pedicels 2-6 mm long, pubescent

with stellate or tufted hairs. Calyx about 6,5 mm

long, lobed to about 1,5 mm from base, densely

stellate-pubescent on outside. Petals pink to reddish,

about 4 mm long, rounded at top and narrowing into

a claw in lower half, blade with margins not inrolled,

claw with margins inrolled and the two lobes at apex

somewhat thickened. Stamens about 7 mm long;

filaments hyaline, obovate, broadly acuminate at apex

into the connective, bearing a few dr obscure hairs

on shoulders; anthers acuminate to an acute apex,

brown-pubescent along margins besides bearing a few’

scattered, patent hairs, overlapping filaments at base!

Ovary very shortly stipitate at the base, stellate-

pubescent, without horns; styles about 9 mm long,

thread-like, cohering in a straight, erect column!

Capsule densely stellate-pubescent with a pair of short,

blunt horns at apices of sutures.

In sandy patches on flats in the Ingwavuma District

of Natal. Also found in the Maputo area and on

Inhaca, Mozambique.

Natal. — Ingwavuma: Ndumu Game Reserve, Ward 1992;

at the pont on Pongola River, Vahrmeijer & Tolken 976-

Ndumu Hill, Pooley 498.

Characterized by the fairly rough stellate pubescence

without obvious gland-tipped hairs, the small petals

with thickened lobes at the apex of the claw, and the

short blunt horns on the capsule which is stellate-

pubescent. It is close to H. boraginiflora from which it

differs mainly in the absence of obvious gland-tipped

hairs.

17. Hermannia amabilis Marloth ex K. Schum. in

Bot. Jb. 10: 42 (1888); K. Schum. in Engl., Monogr.

Afr. Pfl. 5: 54 (1900); M. Friedrich et al. in F. S.W.A.

84: 12 (1969). Type: South West Africa, Hykamkab,

Marloth 1213 (PRE!; BOL!).

H. tenuipes Engl, in Bot. Jb. 55: 363 (1919). Type: South West

Africa, Hoabes, Dinter 1439 (Z!).

Suffrutex, lax to bushy, up to 75 cm tall, leafy in

lower half, in upper half leaves much reduced;

branches sublaxly pubescent with gland-tipped, usually

short hairs. Stipules 1,5-5 mm long, on upper bract-

like leaves down to 1 mm long. Leaves petiolate; blade

narrowly oblong to oblong or ovate-acuminate, 1-6 cm

long, 1,5-10 mm broad or in ovate leaves up to 18 mm

broad near base, upper bract-like leaves 2-10 mm

long, 0,5 mm broad, subulate, stellate-tomentellous

on both surfaces, sometimes intermixed with longer,

gland-tipped hairs especially along margins, rarely

sublaxly stellate-pubescent on both surfaces, petiole

1,5-25 mm long, in upper bract-like leaves about

10 mm long, glandular-pubescent. Inflorescence of

1-flowered, axillary cymes, apparently in long, terminal

racemes owing to the marked reduction of upper

leaves; peduncle very slender, straight, patent to

subpatent, 2-3,5 cm long, glabrous; pedicel cernuous,

sometimes finely pubescent, about 5 mm long. Calyx

about 7,5 mm long, obscurely pubescent without,

lobed to midway or just beyond. Petals longer than

calyx, white to pale pink with a coloured patch

(“pink”, “red” or “mauve”) just below middle,

about 17 mm long, 5 mm board in apical portion and

from there narrowing to base, claw about 4 mm long,

margins inrolled. Filaments obovate, long-cuneate,

mucronate, about 5,5 mm long; anthers about 4,5 mm

long, overlapping filament for almost 1 mm. Ovary

stipitate, oblong-obovate; stipe 1,5 mm; styles fili-

form, about 5-10 mm long, straight. Capsule about

5 mm long and 6 mm diam. near apex, oblong or

narrowing slightly towards the stipitate base, pilose

along sutures, not horned.

Found in gorges and dry riverbeds in South West

Africa.

S.W.A. — Kaokoland: Otjihu, De Winter & Leistner 5680;

Okawerowe, Merxmuller & Giess 1426; Anabib, Story 5755;

Otjinungua, De Winter & Leistner 5772; Sesfontein, De Winter

6 Leistner 5872. Karasburg: Karasburg, Dinter 2802 (SAM).

Omaruru: near Welwitschia, Hardy & De Winter 1452; Brand-

berg, Strey 2650. Outjo: Outjo, Giess 3928. Swakopmund:

Welwitsch, Galpin 7600.

16

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

18. Hermannia modesta ( Ehrenb .) Mast, in F.T.A.

1:232 (1868); K. Schum. in Engl, Monogr. Afr.

Pfl. 5:83 (1900); Burtt Davy, FI. Transv. 1:268

(1926); Wild in F.Z. 1 : 548 (1961); M. Friedrich et al.

in F.S.W.A. 84: 17 (1969). Type: Saudi Arabia, near

Djeddan (or Dschidda) towards Mecca, Ehrenberg s.n.

Tricanthera modesta Ehrenb. in Linnaea 4:402 (1829).

Type: as above.

Hermannia arabica Hochst. & Steud. ex Fisch. in A. B. Mey.

& Ave’-Lall., Ind. Sem. Petrop. 6, Animadv. Bot. 9 (1840);

Hiern, Cat. Afr. PI. Welw. 1:89 (1896). Syntypes: Saudi

Arabia, Dschidda, Ehrenberg s.n.; Schimper 928 (K; PRE,

photo!; LE!).

H. filipes Harv. in F.C. 1:206 (1860). Type: Cape, “Zulu

Country” (sic), in fact Kuruman area, “Miss Owen” s.n. (TCD,

holo.). — var. elatior K. Schum. in Verh. bot. Ver. Prov. Bran-

denb. 30: 235 (1888). Type: South West Africa, Amboland,

Olukonda, Schinz s.n. (not traced but Schinz 585 from Olukonda

is in BOL! & Z!).

H. modesta var. elatior (K. Schum.) K. Schum. in Engl.,

Monogr. Afr. Pfl. 5: 84 (1900). Type: as for H. filipes var.

eliatior. — subvar. virgatissima Engl, in Bot. Jb. 55: 367 (1919).

Syntypes: South West Africa, Damaraland, Tsub, Dinter 2038

(SAM!); Okahandja, Dinter 431 (SAM!; Z!). — subvar. brevi-

cornis Engl., l.c. (1919). Syntypes: Transvaal, Mara, Schlechter

4613 (K; PRE, photo.!; BOL!; Z!); Nazareth, Schlechter 4479

(Z!; BOL!). — subvar. macropetala Engl., l.c. (1919). Syntypes:

South West Africa, Okahandja, Schlechter 4616 (BOL!; Z!*);

Engler 6189 (K; PRE, photo.!); Otavi, Dinter 902 (SAM!). —

subvar. mediipetala Engl., l.c. 368 (1919). Syntypes: South West

Africa, Windhoek, Dinter 1254 (Z!); Okaputa, Engler 6207

(not seen). — var. tsumebensis Engl., l.c. 368 (1919). Type:

South West Africa, Tsumeb, Dinter 3008 (SAM!).

H. kirkii Mast, in F.T.A. 1 : 233 (1868); K. Schum. in Engl.,

Monogr. Afr. Pfl. 5:84 (1900); Wild in F.Z. 1:547 (1961).

Syntypes: Mozambique, Peters s.n.; Lupata, Senna, Kirk s.n.

(K; PRE, photo.!); S. Central Afr., Baines s.n.

H. lugardii N.E. Br. in Kew Bull. 1909: 94 (1909); Burtt Davy,

FI. Transv. 1:268 (1926); Wild in F.Z. 1: 547 (1961), in syn.

Syntypes: Botswana, Kwebe Hills, Lugard 142; Mrs. Lugard

125 (K; PRE, photo.!).

H. atrosanguinea Dinter in Reprium nov. Spec. Regni veg.

18:251 (1922). Syntypes: Waldau, Dinter 551a (SAM!; Z!);

Otjihua, Dinter 551 (SAM!; Z!); Kuibis, Dinter 1182 (SAM!);

Biillsporter Flache, Dinter s.n.

H. holubii Burtt Davy, FI. Transv. 1:42, 268 (1926). Type:

Transvaal, Mara, Schlechter 4613 (K; PRE, photo.! BOL!; Z!)

which is one of the syntypes of subvar. brevicornis).

Suffrutex, appearing annual but often persisting for

a few years, flowering early in first year, stems erect,

branching from near base, branches ascending or

spreading-ascending; stems, branches, petioles and

peduncles usually pubescent with short, patent, gland-

tipped hairs, sometimes intermixed with distinct

stellate hairs or stellate only and occasionally long,

pointed, spreading hairs present as well, rarely some

branchlets glabrescent. Stipules from about 1 mm to

2,5 mm long, linear-acute to subulate, sparsely stellate-

pubescent. Leaves petiolate; blade variable, from

more or less linear, narrowing slightly to base and

apex, to oblong-elliptic, 6-40 mm long, 1-12 mm

broad, usually sparsely stellate-pubescent, glabrescent,

or gland-tipped hairs and bulbous-based hairs present,

margins entire or distinctly toothed, apex acute or

broad and rounded; petiole 1-12 mm long. Inflore-

scence of 1 -flowered cymes, solitary in axils of leaves;

peduncle 10-35 mm long, hardly distinguishable from

the pedicels; pedicels 3,7 mm long, becoming cer-

nuous; bracteoles aborted. Calyx persistent, stellate-

pubescent, becoming translucent with age, sparsely

pubescent with gland-tipped hairs, lobed to about, or

just below midway, about 4,5 mm long. Petals

“blood-red”, “pale red”, “pink”, “deep orange”,

“crimson”, “purple”, mostly appearing deep violet in

* Schlechter 4616 in PRE is H. glanduligera K. Schum., but

in Z it is H. modesta.

pressed specimens, about 5-12 mm long, margin

slightly inrolled on claw only, 4-6 mm broad near

apex, narrowing slightly to the base. Stamens with

obovate, hyaline filaments, pubescence obscure on

shoulders; anthers erect, 4-8 mm long, overlapping the

filament for about 1 mm, acute, sparsely ciliate.

Ovary stipitate (stipe about 1,5 mm long), sparsely

pubescent with gland-tipped hairs and longer hairs

along the sutures and a few stellate hairs inter-mixed

or stellate-pubescent with a few gland-tipped hairs

intermixed, very shortly horned. Capsules 5-12 mm

long, 4-7 mm broad, pubescent with short gland-

tipped hairs and longer acute hairs on the sutures and a

few stellate hairs present or predominantly stellate-

pubescent with a few gland-tipped hairs present; horns

usually under 2 mm long, spreading, stipe about 2 mm

long. Seeds reniform, corrugated and finely granular

with tubercles developing.

Found in dry, warm, sandy regions, mostly in open,

sparse vegetation, often along roads in dry riverbeds

and on old lands. Recorded from Beaufort West in

the Cape northwards through the northern Cape,

parts of the Orange Free State, Natal and Transvaal

and into South West Africa. Also found in Botswana

and northwards to Egypt and Arabia.

Cape. — Beaufort West: Beaufort West, Brueckner 234.

Gordonia: Gordonia, Leistner 1831. Hay: Hay, Acocks 2067;

17675; Upington, Schlieben, 10432. Kenhardt: Kenhardt,

Wasserfall 1138. Kimberley: Kimberley, Leistner 2273. Kuru-

man: Kuruman, Leistner 2214. Prieska: Prieska, Acocks 12641.

Vryburg: Vryburg, Burtt Davy 13675; 13776. Warrenton:

Warrenton, Leistner 1261.

O.F.S. — Bloemfontein: Bloemfontein, Potts 7550. Boshof:

Boshof, Burtt Davy 12893.

Natal. — Lower Umfolozi: Umfolozi Game Reserve, Ross

2026.

Transvaal. — Bloemhof: Bloemhof, Theron s/669. Letaba:

Shilovane, Junod 6220. Lydenburg: Lydenburg, Barnard 488.

Pietersburg: Pietersburg, Schlechter 4479. Pilgrims Rest:

Pilgrims Rest, Acocks 16760. Potgietersrus : Potgietersrus,

Meeuse, 9569. Soutpansberg: Mara, Schlechter 4613; Messina,

Mauve 4415; Salt Pan, Schweickerdt & Verdoorn 486; 652.

S.W.A. — Gibeon: Gibeon, Basson 172; 161; Van Vuuren &

Giess 1073. Gobabis: Gobabis, Wilman 402. Grootfontein:

Grootfontein, MerxmiiUer & Giess 1799; De Winter & Giess

6783; Dinter 5319; Tsumeb, Naegelsbach in TRV 36298;

Dinter 7450. Karibib: Karibib, De Winter & Hardy 7968;

Marloth 1309. Kaokoland: Kaokoland, Story 5867; De Winter

& Leistner 5374; 5166. Keetmanshoop: Keetmanshoop, De

Winter 3253; Dinter 4863. Liideritz: Liideritz, Kinges 2146.

Okahandja: Okahandja, De Winter 2703, Otjiwarongo, Brad-

field 171 ; 419. Okavango: Okavango, De Winter & Marais 4962.

Ovambo: Ovambo, De Winter & Giess 6844. Rehoboth:

Rehoboth, De Winter 3499; Acocks 18159. Swakopmund:

Swakopmund, Giess 3005; Codd 10581. Windhoek: Windhoek,

Dinter 4411; Wilman All ; Merxmiiller 859.

From the long list of synonyms and the several

varieties and subvarieties described, it is obvious that

attempts to circumscribe discrete taxa in this complex

have failed. It is noteworthy, for instance, that the

type of H. holubii Burt Davy, Schlechter 4613, is also a

syntype of H. modesta var. elatior subvar. brevicornis

Engl. ( Schlechter 4613 in BOL differs from a specimen

with the same number in K in having broader leaves).

Much of the variation may be ascribed to hybridi-

zation.

A Venda name “Ishitoni sha” is noted on one of the

specimens.

19. Hermannia tigrensis Hochst. ex A. Rich.,

Tent. FI. Abyss. 1 : 74 t. 17 (1847); K. Schum. in

Engl., Monogr. Afr. Pfl. 5: 85 (1843); Mast, in F.T.A.

1 : 233 (1868); Wild in F.Z. 1 : 546 (1960), as tigreensis.

Type: Ethiopia, dist. Tigre, Schimper 812 (K; PRE,

photo.!; LE!).

I. C. VERDOORN

17

H. mildbraedii Dinter & Engl, in Bot. Jb. 39: 590 (1907);

M. Friedrich et al. in F.S.W.A. 84: 16 (1969). Type: South West

Africa, Okahandja, Dinter s.n. (Z ! ; PRE, photo.!; K (as No.

487).

H. tenella Dinter & Schinz in Vjschr. naturf. Ges. Zurich

55:241 (1910). Type: South West Africa, Waterberg, Dinter

354 (Z! ; PRE, photo.!).

H. pseudo-mildbraedii Dinter & Engl, in Bot. Jb. 55:369

(1919). Syntypes: South West Africa, Okahandja, Dinter 2566

(SAM!); Waldau, Dinter 522 (SAM!).

Annual, very slender, branched from near base;

branches slender, spreading-ascending to arcuate-

ascending, subdensely stellate-pubescent, hairs few

in a group and of different lengths (±0,5 mm long).

Stipules subulate to linear, acuminate, 1,5-2, 5 mm

long, rarely up to 5 mm long (often on same plant).

Leaves petiolate; blade narrowly to broadly ovate-

acuminate, varying much in size on same plant, 10-35

mm long, 3-15 mm broad, rounded, and usually

broadest, at base, broadly acuminate to an acute

apex, obscurely to clearly serrate, sparsely to sub-

densely stellate-pubescent on both surfaces, hairs

few in a group (sometimes reduced to 2) and of

different lengths; petioles 1-4 (-12) mm long. Flowers

axillary, solitary, yellow-orange, dull red, brick-red

or cream with a dark centre, borne on thread-like

peduncles; peduncle about 20 mm long, glabrous or

pubescent in upper portion; pedicel usually short,

from under 1 mm to 2 mm long, rarely up to 5 mm

long, stellate-pubescent; bracts none or very small

and early caducous. Calyx lobed to beyond middle,

translucent, stellate-pubescent without; tube 1-1,5

mm long; lobes deltoid-acuminate, 2-2,5 mm long.

Petals oblong to narrowly oblong, narrowing at base

into a claw, 3,5-5, 5 mm long, 1,5-2 mm broad,

margins not inrolled or obscurely to clearly so on

claw. Filaments obtrullate, about 2 mm long, pube-

scent on shoulders, translucent; anthers erect, about 2

mm long, acute at apex, rounded at base, very

shortly overlapping filament at base (sometimes

coloured). Ovary stipitate, minutely stellate-pube-

scent, about 1 mm long and almost 1 mm broad,

lobes with 2 acute apices; styles very slender, erect,

cohering, apices recurved; stipe about 0,5 mm long,

distinct. Capsule about 5 mm long, 4 mm diam.,

shallowly 5-lobed at the apex, lobes 2-horned; horns

1 ,5-3,5 mm long.

Found in pockets of soil between rocks on koppies

and mountains. Recorded from the northern half of

South West Africa. Also in Angola and northwards

to Ethiopia and Eritrea.

S.W.A. — Kaokoland: near Ohopoho, De Winter & Leistner

5346. Karibib: Ameib, De Winter & Hardy 8086; 8087; Klein

Ameib, Dinter 7076. Okahandja: Okahandja, Dinter 4614;

Waldau, Dinter 522 (SAM). Omaruru: Brandberg, De Winter &

Hardy 8218. Outjo: Paresis Mountain, Barnard 175.

A range of material from South West Africa has

shown that the length of the horns on the capsules

may vary considerably, namely from 1,5-3, 5 mm.

H. mildbraedii, which has been distinguished from

H. tigrensis by its longer horns, is therefore not

upheld. A Rich, in Tent. FI. Abyss uses the spelling

tigreensis, while Hochstetter’s list, Masters in F.T.A.

and Kew Index give tigrensis. F.Z. retains the original

spelling, i.e. tigreensis, as does F.S.W.A. in the note

under H. mildbraedii. It appears that, according to

Article 73 G (b) of the Code, it should be H. tigrensis.

In Z Dinter 457 is indicated as the type of H. mild-

braedii but there is no evidence to support it.

20. Hermannia linearifolia Harv. in F.C. 1 : 205

(1860); K. Schum. in Engl., Monogr. Afr. Pfl. 5; 79

(1900); Burtt Davy, FI. Transv. 1:268 (1926). Syn-

types: Cape, Winterveld, between Nuwejaarsfontein

and Ezelsfontein, Dr'ege s.n. (K; PRE, photo.!; S);

Orange River, Burke & Zeyher s.n. (TCD; PRE,

photo.!); Somerset, Barber s.n. (PRE!).

H. fasciculata Bak. in Refug. Bot. 5, t. 289 (1873) Type'

Cult., originally collected by Cooper in the Cape.

Bushy suffrutex, 15-60 (-100) cm high, usually

heavily browsed; stems minutely and densely grey

scaly, new growth viscid with minute papillae and

occasional stalked glands, otherwise glabrous. Sti-

pules subulate, 0,5—1 mm long, deciduous. Leaves

clustered on short, abbreviated shoots, sessile or

shortly petiolate; blade varying in length in each

cluster, 5 mm to about 12 mm long, very narrowly

obovate, plicate, giving the appearance of being

li tear, microscopically pitted and with minute viscid

papillae, otherwise glabrous; petiole 0 < r up to 3 mm

long. Inflorescence of 1 -flowered cymes in pseudo-

racemes at the ends of the slender, ultimate branch-

lets; peduncles aborted or 1-3 mm long, viscid and

with a few glands; bracteoles minute, about 0,5 mm

long; pedicels usually glabrous, 2-4 mm long. Calyx

about 6 mm long, lobed to middle or just beyond;

glabrous and somewhat viscid with minute papillae,

lobes deltoid-acuminate with the finely and densely

ciliate margins incurved in upper portion. Petals

about 1 cm long, “deep wine-red”, “red”, “brick-

red”, “mauve” or occasionally clear yellow, oblong-

obovate in upper two-thirds, produced into a claw

with incurved margins in lower third, glabrous.

Stamens adhering to stipe at base, free portion about

7 mm long; filaments hyaline, narrowly obovate,

overlapped by anther base about as long as anthers.

Ovary minutely papillate, about 1,5 mm long,

stipitate; stipe about 1,5 mm long; styles 2,5 mm

long. Capsule oblong-globose, about 6 mm long, 5

mm broad, with a 4 mm long stipe at base surrounded

by the persistent calyx-tube, 5 pairs of spreading

horns at apex, horns glabrous, 1-5 mm long.

Found among dolerite rocks, on hill slopes in

calcareous sandy country and near pans. Recorded

from Oudtshoorn and Uitenhage northwards through

Cradock and Middelburg to the south-western Orange

Free State, Christiana in the Transvaal, Kuruman in

the northern Cape and in Botswana.

Cape. — Barkly West: Holpan, Acocks 108; Danielskuil,

Acocks 251. Cradock: near Rayners Koppie, Acocks 11920;

Zebra Nat. Park, Muller 571; Brynard 28; Liebenberg 7238.

Gordonia: Kakaup Pan, Leistner 2047. Herbert: Ramah,

Werger 1300. Kimberley: Merbou Siding, Moran s.n.; Warren-

ton, Adams in PRE 48804; 48806. Kuruman : Kuruman, Marloth

1129. Middelburg: near town, Gill 127; Flanagan 1378; Groot-

fontein Agric. College, Verdoorn 1459; Rosemead Junction, Sim

sub Galpin 5636. Oudtshoorn: near de Rust, Dahlstrand \209.

Richmond: Vlakplaat, H. Bolus 13772. Somerset East: Somerset

East, Barber s.n. Uitenhage: Winterhoeksbergen, Ecklon &

Zeyher 109. Victoria West: Melton Wold Estate, Smith 2438.

O.F.S. — Boshof: Boshof, Leistner & Joynt 2664; Burtt Davy

in Govt. Herb. 10822. Fauresmith: Fauresmith, Pole Evans

1616; Koffiefontein, Codd 3421. Jacobsdal: Honeynestkloof,

Schweickerdt 1154.

Transvaal. — Christiana: Bloemheuwel, Burtt Davy in Govt.

Herb. 11273.

Botswana. — Without exact locality, Holub, s.n. (BOL).

This species is characterized by its clustered, almost

needle-like leaves, and by the more usual stellate-

pubescence being almost entirely replaced by gluti-

nous scales and papillae which glisten in sunlight and

cause the plant to be sticky when crushed. A paste

made of the crushed parts is said to be used to close

wounds and relieve pain and this accounts for the

common name “Pleisterbos”. This species is usually

referred to as “Rooi Pleisterbos”, but since the

flower colour varies in the species and can be pure

18

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

yellow occasionally, it is not very apt. H. linearifolia

belongs to the group in which the petals are not

strongly twisted.

21. Hermannia helianthemum K. Schum. in Bot.

Jb. 10:44 (1889); in Engl., Monogr. Afr. Pfl. 5: 83

(1900); Burtt Davy, FI. Transv. 1:267 (1926), partly

excluding Holub s.n.; M. Friedrich et al. F.S.W.A.

84: 15 (1969). Type: South West Africa, Usakos,

Marloth 1237 (PRE, lecto!; SAM!; BOL!).

Suffrutex 30-60 cm high, virgately branched,

branches slender, woody at base, stellate-tomentose.

Stipules subulate or conduplicate, stellate-tomentose,

1 ,5-5 mm long. Leaves petiolate; blade linear-elliptic

to narrowly ovate-oblong, densely stellate-tomentose

on both surfaces, 16-34 mm long, 4-11 mm broad,

usually broadly acuminate at apex, sometimes rounded

and occasionally, on same plant, dentate at broad

apex, usually rounded at base, margins entire or

obscurely dentate, nerves prominent beneath, 3 basal

nerves of which the two lateral ones (i.e. on each side of

midrib) run only to about halfway up leaf, leaves

reduced to bracts in upper portion of the ultimate

flowering branches; petiole 5-11 mm long. Inflores-

cence of 1 -flowered cymes in the axils of upper, reduced

leaves, appearing racemose, the only distinction

between peduncle and pedicel being that the latter is

more densely stellate-tomentose; peduncle about

5-7 mm long; pedicel 1,5-3, 5 mm long. Calyx about

6 or 7 mm long, lobed to about middle, stellate-

tomentose outside, lobes appressed pubescent within.

Petals “pale” to “deep pink”, about 6-8,5 mm long

and 2 , 5-4 , 5 mm broad, ± oblong-cuneate, the margin

inrolled only along the short, 1,5-2, 5 mm long,

claw which is glabrous. Stamens with obtrullate, very

thin, hyaline filaments, with a few hairs on the shoul-

ders, 2, 5-4, 5 mm long, 1,25-3 mm broad in upper

half; anthers about 5 mm long, ciliate. Ovary densely

stellate-tomentose with longer hairs at the apex, or

subglabrescent in lower portion, about 3,5 mm long,

stipe 1 mm long, with anther bases cohering to form a

ring around it; style filiform. Capsule shortly horned;

Found on sandy flats near Garies and on granite

slopes in the Omaruru, Karibib and Okahandja,

districts of South West Africa, and in Namaqualand,

Cape.

Cape. — Namaqualand: Garies, Esterhuysen 1297.

S.W.A. — Karibib: Anschluss, De Winter & Hardy 7985.

Okahandja: Usakos, Marloth 1237; Okagava, Dinter 6937.

Omaruru: near Uis mine, De Winter 3165. Swakopmund:

Wildreservat, Jensen 202 (WIND).

Characterized by the raceme-like inflorescence, the

stellate tomentum on all parts, short horns to the

capsule, pink, often pale pink, petals drying parch-

ment colour, sometimes with violet patches.

The epithet helianthemum ending in “- um ” is correct,

because it is called after the genus Helianthemum.

In the original description K. Schumann cites only

two specimens, Marloth 1237 and 1238. In Engl.,

Monogr. Afr. Pfl. 583 these two numbers are cited

and indicated as being in the Berlin Herbarium.

Specimens of Marloth 1238 in SAM and BOL are

H. comosa, which could never be confused with

H. helianthemum. It must be assumed that the speci-

men of Marloth 1238 in B, which was destroyed during

the war, represented H. helianthemum.

22. Hermannia damarana Bak.f. in J. Bot., Lond.

39: 127 (1901); M. Friedrich et al. in F.S.W.A. 84: 13

(1969). Type: South West African, Damaraland,

1879, T. G. Een s.n. (BM, holo.; PRE, photo.!).

H. lindequistii Engl, in Bot. Jb. 39: 589 (1907). Type: South

West Africa, between Okahandja and Otzisasu, Dinter s.n.

(not traced).

Suffrutex 30-40 cm tall, stems woody, much

branched, with grey bark; branchlets finely stellate-

tomentose, hairs short, glutinous. Stipules up to 6 mm

long, linear-subulate. Leaves shortly petiolate; blade

narrowly to broadly oblong-obovate, about 3, 5x0, 7

mm to 30x9 mm, finely, densely and shortly stellate-

tomentose (hairs somewhat glutinous), rounded or

acute at apex, narrowing slightly to a rounded or

cuneate base, margin usually entire, midrib and veins

prominent beneath, usually 2 arising from base, one on

each side of the midrib; petiole 1-4 mm long. In-

florescence of 1 -flowered, axillary cymes; peduncle up

to 15 mm long, shortly stellate-tomentose, 2-4-

bracteate at apex; pedicels 2-3 mm long, stellate-

tomentose with slightly longer hairs than on the

peduncle. Calyx about 6 mm long, tube ±3,5 mm,

lobes ±2,5 mm long, shortly stellate-tomentose

without, and lobes with appressed hairs on the inner

face, usually with five prominent nerves running from

apex of lobes to base. Petals orange-red, oblong to

narrowly oblong-spathulate, 7-10 mm long, some

narrowly inrolled only on claw, others in same flower

with somewhat infolded margins all the way, no

auricles at apex of claw, glabrous on inner face.

Stamens with oblanceolate-linear filaments ±4 mm

long, glabrous; anthers ±4 mm long, overlapping

filament for about 1 mm, ciliate. Ovary very shortly

stipitate, ±2,5 mm long, shortly stellate-tomentose,

hairs rather longer at apex; styles about 5 mm long.

Capsule about 6 mm long, 3 mm diam., shortly and

densely stellate-tomentose, eventually shortly horned

at sutures, horns about 1,5 mm long.

Found in limy soils. Recorded from the Omaruru,

Okahandja and Gobabis districts of South West

Africa and eastwards into Botswana.

S.W.A. — Damaraland: Een s.n. (BM). Gobabis: Gobabis,

De Winter 2508; Steinhausen, Dinter 2784 (SAM). Okahandja:

Springbokvlakte, Giess, Volk <& Bleissner 6322a. Okahandja:

Okomita, Dinter 508 (Z). Omaruru: Omaruru, Dinter 7494.

Without precise locality, Bradfield 428.

Botswana. — 60 km N.W. of Serowe, Wild & Drummond

7293.

Characterized by the short calyx lobes, the 5 promi-

nent ribs on the calyx (outside), the fine, short, stellate

tomentum, the three prominent basal nerves on the

undersurface of the leaves and the stellate-tomentose

capsule with very short horns.

23. Hermannia gariepina Eckl. & Zeyh., Enum.

49 (1834); Harv. in F.C. 1 : 202 (1860); K. Schum. in

Engl., Monogr. Afr. Pfl. 5: 80(1900); F.S.W.A. 84: 14

(1969). Type: Cape, Boschmansland, near Orange

River, Ecklon & Zeyher 384 (SAM ! ; S ! ; PRE !).

H. racemosa E. Mey. in Drege, Zwei Pfl. Doc. 192 (1843),

nom. nud. based on Drege 3279 (PRE!; LE!; Z!).

H. exstipulata E. Mey. ex Harv. in F.C. 1: 202 (1860); K.

Schum. in Engl., Monogr. Afr. Pfl. 5: 59 (1900). Type: Cape,

Namaqualand, Orange River Mouth, Drege s.n. (TCD, holo.;

PRE, photo.!; PRE!).

H. truncata Schinz in Vjschr. naturf. Ges. Zurich 55 : 241

(1910). Type: South West Africa, Graspoort, Dinter 1087 (Z!).

H. gariepina Eck. & Zeyh. var. dentata Engl, in Bot. Jb.

55: 364 (1919). Type: Silverfontein, Drege 3279 (PRE!; LE!;

Z!). — var. integrifolia Engl, in Bot. Jb. 55: 364 (1919). Syntypes:

South West Africa, Range 177; 882 (SAM!); Dinter 1124;

1079 (Z!); Engler 6696.

H. nivea Schinz in Vjschr. naturf. Ges. Zurich 70: 218 (1925).

Type: South West Africa, Kleinfonteiner Flache, Dinter 3734

(PRE!; Z!).

Shrublet 22-75 cm tall, much branched, branches

ascending or divaricate, densely and finely stellate-

I. C. VERDOORN

19

tomentose with short, greyish white hairs. Stipules

0,75-2 mm long, thick in texture and densely stellate-

tomentose with short greyish white hairs, sometimes

caducous. Leaves shortly petiolate; blade rather thick

in texture and densely stellate-tomentose with short,

grey hairs, variable in shape and size, elliptic-cuneate

to obovate — or oblong-cuneate, broadest just above

middle or at the apex, about 7 to 20 mm long, margin

entire or crenate at the apex, base cruneate or sub-

rounded, apex broadly acuminate or truncate and

often shallowly 3-lobed; petiole often folded trough-

shaped, 1-5 mm long, rather thick in texture and

densely stellate-tomentose with short hairs. Inflores-

cence of 1 -flowered cymes, solitary in axils of leaves,

upper subtending leaves sometimes much reduced in

size giving the appearance of racemes; peduncles

2.5- 6 mm long, densely stellate-tomentose with short

hairs; pedicels 3-5 mm long, distinguishable from the

peduncle by the coarser stellate pubescence, hairs

over 0,5 mm long; bracts minute or wanting. Calyx

lobed almost to base, stellate-pubescent with hairs

rather long (0,5-1 mm), tube 1,5-3 mm long, lobes

4.5- 7 mm long. Petals pink, “purplish” or “violet”,

6.5- 9 mm long, 2,5-6 mm broad; blade narrowly

oblong to suborbicular, narrowing slightly into a

claw in lower third, margins sometimes minutely

ciliate, glabrous within, claw with slightly inrolled

margins. Stamens with hyaline, obovate filaments

which either narrow gradually or abruptly to apex and

base and are glabrous or setose on shoulders; anthers

3.5- 5 mm long, ciliate with sparse, short hairs, over-

lapping filament for 1,5 mm. Ovary about 2,5 mm

long, 2 mm diam. with short papillose hairs on sur-

faces and setose or stellate along sutures and at apex;

stipe 1,5 mm long; styles 5 mm long, slender, erect,

exserted from the closed mature flower for about

2 mm. Capsule about 5 mm long, usually hidden in the

persistent calyx and petals, obscurely 5-angled, surface

papillose, sutures setose and stellate with a blunt lobe

or short horn at apex, up to 2 mm long; stipe 1,5 mm

long. Seeds brown, reniform, corrugated (and some-

times scaly?).

Found in sandy semi-desert areas. Recorded from

the Kenhardt District and northern Namaqualand in

the Cape northwards to Warmbad, Liideritz and

Kaokoland in South West Africa.

Cape. — Kenhardt: Kakamas Veld Reserve, Acocks 16360;

near Pofadder Esterhuysen 1250. Namaqualand: Springbok,

Dumoulin 4; Silverfontein, Drege 3279; Schakalswater, M.

Schlechter 3; Orange River Mouth, Drege s.n.

S.W.A. — Kaokoland: Cape Frio, Giess 9079 ; south of

Mowebucht, Merxmiiller & Giess 30662. Liideritz: Haalenberg,

De Winter & Giess 6161; Klinghardtgebirge, Dinter 3948; Aus,

Dinter 6039. Warmbad: Vioolsdrift, Merxmiiller & Giess 3671.

Characterized by the dense, fine, whitish, stellate

pubescence of the branches, leaves and peduncles

contrasting with the coarse stellate pubescence of the

pedicels and calyx. The length of the hairs on the

calyx varies considerably being long and shaggy on

some specimens and not so long on others. The leaf

shape and margin varies from elliptic-cuneate with

entire margins to oblong-cuneate with a truncate

and 3-lobed apex. Two forms of petals and filaments

are found. The petals have either a narrow, oblong

blade or a semi-orbicular blade with or without a

ciliate margin, while the filaments which are broadest

near the apex, either narrow gradually to base and

apex or narrow abruptly and so may be taken to be

somewhat cruciform.

Drege 3279, the type of var. dentata, is also the

type of the nomen nudum H. racemosa E. Mey. in

D-ege, Zwei Pfl. Doc. 192 (1843). Specimens of M.

Schlechter s.n. from Schakalwater in Z and LE bear

the name H. maximiliana Schltr. n. sp. In Z and PRE

there are specimens of Dinter 6039 named H. kurtiana

Schinz. This name was never published. The specimen

is H. gariepina. Specimens of Schlechter 11374 in LE

and Z bear the name H. conaclada K. Schum. which

was evidently also not published. It too, is H. garie-

pina. The common name “Koerkassie” is recorded by

Dumoulin.

24. Hermannia complicata Engl, in Bot. Jb.

39: 588 (1907). Type: South West Africa, Jakalswater,

Dinter 1460 (Z, holo.!; PRE, photo.!).

Suffrutescent, leafy, dwarf bush, many-stemmed,

about 20 cm tall, much branched from base, the

whole plant grey stellate-tomentose, part of branch-

lets obscurely lineate. Stipules 1-2,5 mm long,

rather thick and persistent, stellate-tomentose and

with a few long hairs at apex. Leaves petiolate; blade

narrowly to broadly oblong-cuneate, sometimes

broadest at apex, from about 5x2 mm to 20x10

mm, stellate-tomentose on both surfaces; coarsely

crenate in upper half or at apex only, cuneate portion

entire; petiole 2-10 mm long, grey stellate-tomentose.

Inflorescence of 1 -flowered, axillary cymes, usually

for whole length of the branchlet; peduncle 2-5 mm

long, persistent but not indurated and spine-like;

pedicels about 3 mm long, curled, slightly thicker and

with denser tomentum than on peduncle. Calyx

lobed to just beyond middle, stellate-pubescent to

stellate-tomentose outside, lobes sparsely appressed

pubescent within, tube 2-3 mm long, lobes 2, 5-3, 5

mm long. Petals “pink”, about 8 mm long, narrowly

oblong, somewhat cuneate at base, margin narrowly

inrolled in cuneate basal part, thin-textured, not

thickened in centre, glabrous in specimens dissected.

Stamens united at base around stipe; filaments

narrowly obovate, hyaline, pubescent on shoulders,

about 3 mm long; anthers 4 mm long, pointed,

ciliate. Ovary with a 1,5 mm long stipe, subglobose,

about 1,5 mm long, with 10 erect setose horns,

papillose and stellate on surface, setose along sutures;

ovules about 14. Capsule about 3 mm long, 3 mm

broad, stipitate, papillose and stellate-pubescent on

surface, setose along sutures, with 10 long, spreading

to recurved setose or pilose horns 4-10 mm long;

Found on the edge of the Namib in semi-desert

conditions. Recorded from Liideritz District north-

wards to Welwitschia in South West Africa.

S.W.A. — Liideritz: Between Helmeringhausen and Aus,

Kinges 2223. Omaruru: Welwitschia, Boss 2183; Galpin &

Pearson 7575; Brandberg, Giess 1957. Swakopmund: Usakos,

Dinter 6847; between Okahandja and Swakopmund, Dinter 160

(SAM); between Walvisbaai and Swakopmund, Strey 2504.

Characteristically a small, rounded, entirely grey

stellate-tomentose bush about 20 cm tall, branched

from the base with slender branches. The leaves are

coarsely crenate in the upper half and the ten horns

on the capsule are long, pilose and spreading to

recurved.

In F.S.W.A. this species is put in synomy under H.

ajflnis, a species which it resembles closely and which

differs mainly in being a low bushy plant branched

from the base, with branchlets which become less

readily indurated and spine-like.

A specimen in BOL, Miss W. Tucker s.n., from the

sand dunes south of Doornpoort, may be H. com-

plicata, but the material is not sufficient for identifi-

cation.

20

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

25. Hermannia trifurca L., Amoen. Acad. 6, 90

(1760); Sp. PI. ed. 2: 942 (1763), as trufurcata; Syst.

Veg. 610 (1784); Cav., Diss. t. 178, fig. 2 (1788), as

trufurcata-, Jacq., Schoenbr., t. 125 (1797); DC.,

Prodr. 1: 495 (1824), as trifurcata\ Thunb., FI. Cap.

ed. Schult. 503 (1823); Harv. in F.C. 1:205 (1860),

as trifurcata. Type: “Cap. b. Spei”, Linn. Herb. Cat.

No. 854 (LINN, holo.; PRE, photo.!).

H. bicornis Eckl. & Zeyh., Enum. 49 (1834). Type: Cape,

“Piketberg”, Ecklon & Zeyher Enum. No. 385 S; PRE, photo.!

PRE!). H. hilaris (Eckl. & Zeyh.) Hochr. in Annu. Conserv.

Jard. bot. Geneve 11 & 12:2 (1907).

Mahernia incana Eckl. & Zeyh., Enum. 49 (1834). Type: Cape,

between Saldanha Bay and Bergrivier, Ecklon & Zeyher Enum.

No. 388 (PRE!).

M. hilaris Eckl. & Zeyh. l.c. 49 (1834). Type: Cape, “Brack-

fontein”, Clanwilliam, Eckl. & Zeyh. Enum. No. 389 (S; PRE,

photo.!; TCD; PRE!).

Suffrutex, erect, woody, much branched, ultimate

branchlets becoming indurated and spiky; branchlets

scaly to minutely stellate-scaly, ridged in part.

Stipules subulate to linear acute, 1,5-2, 5 mm long.

Leaves sessile or shortly petiolate, clustered on

abbreviated shoots or single on the ultimate elon-

gated shoots; blade linear to linear-cuneate, varying

in size in each cluster, 6-25 mm long, 1,5-5 mm

broad, rounded, acute or truncate at apex and often

3-toothed, densely to sparsely pubescent with, usually,

minute stellate hairs, rays appressed and short,

rarely long, margins entire, rarely toothed below

tridentate apex. Inflorescence of 1 -flowered cymes in

the axils of upper leaves or arranged in fairly long

terminal (pseudo-) racemes, flowers cernuous and

secund; peduncle usually aborted or very short but

occasionally up to 5 mm long, bi-bracteate at apex,

stellate-pubescent; bracts subulate to linear, 1-3 mm

long; pedicels usually 3-5 mm long when peduncle

is not obvious, otherwise ranging from 1,5-4 mm,

stellate-pubescent. Calyx 6-8 mm long, lobed to just

beyond middle, sublaxly stellate-pubescent, usually

denser in lower portion, the hairs short or long,

lobes densely ciliate with tufted hairs. Petals “mauve”,

“purple”, “pink to wine-red, ageing blue”, 7-11 mm

long, suborbicular to broadly oblong-elliptic in

upper half, narrowed below into a claw with infolded

margins, glabrous. Stamens 7-8 mm long; filaments

hyaline, oblong-obovate to narrowly oblong-cuneate,

apparently glabrous, adhering at base to stipe;

anthers ciliate, 4-5 mm long with the base overlapping

the dilated filament. Ovary stipitate, 5-lobed, scaly

and papillose, papillae longer on keels of lobes,

about 1 ,5 mm long; stipe about 1 ,5 mm long; styles

6-7 mm long, eventually shortly exserted. Capsule

5-lobed, about 7 mm long, 6 mm broad, stipitate at

base, apex with 5 pairs of spreading horns, keel of

lobes and horns villous; stipe about 2 mm long;

horns at first short, eventually up to 5 mm long.

Found from Worcester in the western Cape north-

wards through Namaqualand to the Klinghardt

Mountains in South West Africa.

Cape. — Calvinia: S.W. of Soetwater, farm Lokenburg,

De Winter & Verdoorn 9012. Ceres: Ceres, Pienaar s.n. Clan-

william: Clanwilliam, Schlechter 4397; 5056; Galpin 10731;

Brakfontein, Ecklon & Zeyher Enum. No. 389; Citrusdal,

Liebenberg 4336. Malmesbury: “Saldanhabay”, Drege s.n. (S);

Ecklon & Zeyher Enum. No. 388; Darling, Hutchinson 230.

Namaqualand: without precise loc., H. Bolus in Herb. Norm.

441; Springbok, Van der Schijff 8127: 16 km S.E. of Arris,

Marloth 12409; Steinkopf, Marloth 12224. Piketberg: Piketberg,

Ecklon & Zeyher Enum. No. 385; near Bergrivier station,

Boucher 82; between Veldrif and St. Helena Bay, Marsh 1276;

Pickeniers Kloof, Leipoldt 19862. Vanrhynsdorp: Klaver,

Andreae 490; Vredendal, Thompson 1010; Hall 3809. Worcester:

Worcester, Van Breda & Joubert 1811; “Karee Bergen”,

Schlechter 8283; Hex River Valley, Tyson 688.

S.W. A. — Liideritz: Klinghardtsberge, Spitzkuppe Sud,

Wat mo ugh 881.

This woody much branched shrub is characterized

by the linear leaves which are usually 3-toothed at

the apex, although sometimes very shallowly so, and

the drooping, bell-shaped flowers secund towards

the apices of the branchlets. The petals are reminiscent

of flies’ wings and fold together to give a truncate

appearance to the flower. Often the filiform styles

are shortly exserted. The short, broad, 5-lobed

capsules, partly concealed by the persistent perianth,

are made conspicuous by the pairs of pilose horns

spreading horizontally from the apices of the lobes.

A common Hottentot name for this species is

“Koerhassie”.

With regard to the epithet “ trifurca ”, not only did

Linnaeus use it in the original description but, accord-

ing to Latin scholars, it is more correct than “trifur-

cata” (private letter to N. S. Pillans from Capt. Salter).

26. Hermannia affinis K. Schum. in Verh. Bot.

Ver. Prov. Brandenb. 31:180 (1890); in Engl.,

Monogr. Afr. Pfl. 5:81 (1900); M. Friedrich et al. in

F.S.W.A. 84: 11 (1969), for the greater part. Type:

South West Africa, Liideritz, “between Angra

Pequena and Aus”, Schinz 1113 (Z!; PRE, photo.!).

H. windhukiana Engl, in Bot. Jb. 39: 588 (1907). Type:

South West Africa, Windhoek, Dinter 851 (Z; PRE, photo.!).

H. spinulosa Engl, in Bot. Jb. 55: 365 (1919). Type: South

West Africa, Inachab, Dinter 1077 (Z; PRE, photo.!).

H. squarrosa Dinter ex Range in Reprium nov. Spec. Regni

veg. 36: 262 (1934), nom. nud. based on Dinter 6040 (Z!; PRE,

photo.!; PRE!; SAM!).

Suffrutex, with woody stems and branches, branch-

lets varying in colour from pale yellow, metallic,

cinnamon-brown to plum-coloured, thick and squar-

rose to slender and virgate but all with ultimate

branchlets becoming bare, indurated and spine-like

(the persistent peduncles usually not indurating as in

H. spinosa), at first minutely glandular hairy, covered

by dense stellate pubescence which, at least on some

twigs, is cinnamon-coloured and appears powdery,

rarely with some gland-tipped hairs as well. Stipules

I , 5-3 mm long, persistent, tomentose and with a few

long hairs at apex. Leaves petiolate; blade usually

oblong-cuneate, sometimes oblong-elliptic, coarsely

but shallowly lobed in upper half, some leaves entire,

often broadest at apex, 8-30 mm long, 4-15 mm broad

at centre or near apex, densely stellate-tomentose,

stellate hairs from a scaly base; petiole 3-6 mm long,

tomentose. Inflorescence of 1-flowered, axillary cymes;

peduncle 3-4 mm long, pedicel 1,5-4, 5 mm long;

bracts at apex of peduncle small, 0,5 mm long.

Calyx 7-10 mm long, lobed to beyond middle, tube

3 mm long, lobes 7 mm long, stellate-tomentose

without, sparsely pubescent within, ciliate. Petals in

shades of red and purple, about 9 mm long, 2-3 mm

broad, oblong-cuneate, narrowly inrolled along

margin in lower half. Stamens with hyaline, obovate

filaments, sparsely hairy on shoulders, anthers

ciliate, overlapping filaments at base. Capsule 1 mm

long, glandular and sometimes also stellate-pubescent

on surface, setose along sutures; horns spreading to

recurved, pilose, about 8 mm long.

Found on rocky hills and slopes, on quartzite

outcrops, sandy shale cliff's, or in gravelly soils.

Recorded from the northern Cape (Griqualand

West) from Prieska and Kenhardt northwards into

Botswana and South West Africa.

Cape. — Barkly West: Danielskuil, Acocks 257. Hay: near

Griquatown, Leistner 891; Lovedale, Acocks 17674. Herbert:

I. C. VERDOORN

21

Campbell, Leistner 919. Kenhardt: between Pofadder and

Aggeneys, White 10741. Kimberley: Witdam, Acocks H. 849.

Kuruman: Olifantshoek, Hardy & Bayliss 1249; Langberg,

Sitwell 40. Prieska: Buis Vlei, Acocks & Hafstrom 946.

S.W.A. — Bethanien: Inachab, Dinter 1077 (Z). Gibeon:

north of Mariental, Merxmiiller & Giess 764. Keetmanshoop:

Aroab, De Winter 3360. Luderitz: Aus, Dinter s.n. (Z); Dinter

6040; west of Aus, Giess & Van Vuuren 746. Rehoboth: south

of Rehoboth, Hall 991. Warmbad: Klein Karas, Dinter 4856

(Z); Great Karas, Pearson 7908. Windhoek: Auas Mts, Strey

2568; 2433; west of Windhoek, De Winter 2609.

Characterized by the usually stellate-tomentose

leaves, the indurating branchlets, and the long,

pilose horns on the capsules. H. affinis, as defined

here, varies considerably, but it can, nevertheless, be

recognised as an entity and distinguished from

related species. The two species with which it has

been very generally confused are H. spinosa and H.

helianthemum. From H. spinosa it can be distinguished,

among other details, by the branchlets being stellate-

tomentose and becoming indurated or spine-like,

whereas in H. spinosa the branchlets are sparsely

stellate-pubescent or glabrous and it is usually the

persistent, enlarged peduncles which become spike-

like. H. helianthemum differs considerably from H.

affinis, for it has long, slender branchlets which do

not become indurated but terminate in raceme-like

inflorescences.

In F. S.W.A. H. complicata Engl, is placed in

synonomy under H. affinis, a species to which it is

certainly very close, and from which it differs mainly

in habit and habitat, being a low, bushy plant confined

to the edge of the Namib.

The following insufficiently known species may also

be synonymous with H. affinis : H. arenicola Engl.;

H. longicornuta Engl, and the unpublished H. schu-

mannii Schinz based on Dinter 4856 (Z!; PRE,

photo.!).

27. Hermannia fruticulosa K. Schum. in Verh.

bot. Ver. Prov. Brandenb. 30: 233 (1888); in Engl.,

Monogr. Afr. Pfl. 5: 81 (1900): M. Friedrich et al. in

F.S.W.A. 84: 14 (1969). Syntypes: South West

Africa, Bethanien, Schenck, 361 (PRE!; Z!); Guld-

brandtsthal, Schinz s.n. (Z !) ; Schakatsfluss, Schinz

602 (K; PRE, photo.!; Z!; BOL!).

Suffrutex, virgate, 40-60 cm tall, stems with grey or

brown bark, branches becoming woody and leafless,

often glutinous, with numerous minute papillae, other-

wise glabrous, ultimate flowering branchlets short and

congested with leaves and flowers, minutely papillose.

Stipules terete, blunt, usually just under 1 mm long,

papillose. Leaves shortly petiolate; blade oblong-

cuneate, 5-10 mm long, 3-8 mm broad, broadest at

or near apex, dentate in upper part, teeth usually

thickened with glands, upper and lower surface

papillate and sparsely stellate-pubescent (hairs ap-

pressed), nerves prominent beneath; petiole 1-3 mm

long. Inflorescence of 1 -flowered, solitary cymes,

usually in axils of leaves on short congested shoots;

peduncle usually very short, about 1,5 mm long

(rarely on same plant one up to 2,5 mm long),

glabrous; pedicels up to 5 mm long, papillose and

sparsely stellate-pubescent, thickening towards apex;

bracts minute, ±0,5 mm. Calyx usually lobed to

beyond the middle, 4-6 mm long, obscurely papillose

with a few appressed stellate hairs without, ciliate on

margin and pubescent on inner face. Petals broadly

oblong-cuneate, about 8 mm long, 4 mm broad

near apex, claw with narrowly inrolled margin, not

thickened, appearing glabrous. Stamens with obtrul-

late, hyaline filaments (appearing glabrous), about 3

mm long; anthers 3, ciliate. Ovary densely hispid,

stipitate (stipe ±1,25 mm), about 1,5 mm long;

styles about 4 mm long. Capsule from about 4-5

mm long, with 5 laterally compressed carpels, 2,5

mm broad, papillate on surfaces, sometimes stellate-

pubescent too, setose along sutures, with 10 diverging,

pilose horns which are up to 8 mm long.

Broken veld, on shaly, quartz or granite hills and

plateaux or frequently in water-courses. Recorded

from the Gordonia District in the Cape, near the

South West Africa border and in the Keetmanshoop,

Gibeon and Bethanien Districts of South West

Africa.

Cape. — Gordonia: Rietfontein, Werdermann & Oberdieck

2253.

S.W.A. — Bethanien: Aus, Walter 2616 (WIND); Helmering-

hausen, Dinter 8028; Bersaba-Bethanien, Schenk 361; Das-

siesfontein-lnachab, Pearson 7905. Gibeon: Mariental,

Basson 151. Keetmanshoop: Keetmanshoop, Acocks 15619A;

Hoachabeb, Pearson 7905 (BOL).

Distinguishing features are: abbreviated, slender,

congested shoots bearing small, persistent bracts in

lower half, and leaves and flowers above; long,

indurated branchlets, minutely papillose (not stellate-

pubescent) becoming glutinose; peduncle short,

persistent; pedicels longer and stellate-pubescent;

petals longer than calyx, margins inrolled on claw,

not thickened, glabrous on inner face; capsule with

five laterally compressed carpels free to the axis;

horns 10, long, pilose, spreading and recurved.

28. Hermannia spinosa E. Mey. ex Harv. in F.C.

1:205 (1860); K. Schum. in Engl., Monogr. Afr

Pfl. 5: 80 (1900), but with authors as “(Burch.)

E. Mey.”; M. Friedrich et al. in F.S.W.A. 84: 21

(1969). Syntypes: Cape, Gamke River, Burke &

Zeyher s.n.; Nieuweveld, Drege s.n. (K; PRE!).

Mahernia spinosa Burch, ex DC., Prodr. 1:497 (1824).

Type: Cape, Dwaal River, Burchell s.n. (K, holo.; PRE,

photo. !).

Hermannia asbestina Schltr. in J. Bot., Lond. 36: 373 (1895).

Type: Asbestos Mtn, Marloth 2057 (PRE!).

H. aspericaulis Dinter & Engl, in Bot. Jb. 55: 366 (1919).

Type: South West Africa, Quartel, Rehoboth, Dinter 2163

(SAM).

Suffrutex, 20-40 cm tall, usually a rounded bush,

with wiry stems and branches, young branchlets

sparsely stellate-pubescent with short hairs, glabres-

cent, usually metallic coloured and flexuose, armed

with persistent indurated spine-like peduncles. Stipules

1-2,5 mm long, rather thick and sometimes hooked,

stellate-pubescent and with gland-tipped hairs, usually

a few pointed hairs at apex, persistent. Leaves petio-

late; blade oblong-cuneate to narrowly so, coarsely

toothed in upper half or at broad apex only, or elliptic

and entire (both shapes often on the same plant),

4-20 mm long, 2,5-7 mm broad, minutely papillate

and stellate-pubescent on both surfaces, usually

sparsely so, hairs short from a central scale, radiating

and appressed to surface of the leaf; petiole 2-5 mm

long, sparsely stellate-pubescent and with a few

additional glandular hairs or papillae. Inflorescence

of 1 -flowered, axillary cymes; peduncles at first short,

about 2-4 mm long, persistent, glabrescent, becoming

indurated and elongating to about 8 mm, rarely longer,

more or less patently spreading, appearing spiny but

apex blunt; pedicles about 2-5 mm long, nodding,

stellate-pubescent, deciduous. Calyx from about 5 mm

to 7 mm long, lobed to just beyond middle, minutely

glandular and laxly stellate-pubescent outside, lobes

sparsely appressed hairy on inner face. Petals “wine-

red”, “purple-mauve”, “rose-red”, “rose-pink”,

22

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

“brick-red”, oblong-cuneate, 7-10 mm long, margin

inrolled on claw only, apparently glabrous on inner

face (or reddish papillae in a patch). Stamens with

narrowly obovate, hyaline filaments, joined at base,

free portion 2-3,5 mm long, pubescent on shoulders

which are only slightly wider than the overlapping

anther bases; anthers 4-5 mm long, ciliate. Ovary

stipitate, 2, 5-3, 5 mm long, crowned by 10 erect, 1,5

mm long horns, glandular but with setae along sutures

and on horns; styles filiform, erect, cohering. Capsule

up to 1 cm long, papillose between sutures, stellate

with long hairs or setose on sutures; horns ultimately

divergent and recurved, up to 8 mm long, pilose; seeds

about 7 developed in each cell. Figs 2.2 & 2.6.

Found in karroid veld and on stony koppies in the

western Cape from Prince Albert northwards to

Gordonia and from Warmbad northwards to Reho-

both in South West Africa.

Cape. — Beaufort West: Wagonwheel Motel, Mauve 4458;

east of Beaufort West, Adamson D. 166; S.E. of Fraserburg,

Hutchinson 967. Calvinia: Loeriesfontein, Lock 3966.

Carnarvon: Acocks 1744. Ceres: Gemsbokfontein, Hanekom

2165. Gordonia: N. of Upington, Pole Evans 2137; S.E. of

Augrabies, Leistner & Joynt 2854. Hay: Asbestos Mountains,

Marloth 2057. Hopetown: Strydenburg, Acocks 1700. Kenhardt:

Kenhardt, Leistner 2331; N.E. of Pofadder, Van der Sehijff

5710. Prince Albert: Prince Albert, H. Bolus 11449; Marloth

4412; Gamka River, Zeyher 124 (S). Victoria West: Melton

Wold Estate, Smith 2439. Williston: Williston, Comins 684.

O.F.S. — Cult. Fauresmith, originally from Britstown,

Verdoorn 1110; Henrici 2666.

S.W.A. — Bethanien: Helmeringhausen, Giess 8805. Gibeon:

Gamkanas, Dinter 1991 (SAM). Keetmanshoop: W. of Aroab,

De Winter 3387; E. of Koes, Leistner 1798. Maltahohe:

Nassab, Giess, Volk & Bleissner 5194. Rehoboth: Quartel.

Dinter 2163 (SAM). Warmbad: Udabis, Giess, Volk & Bleissner

708a; Ariamsvlei, Schweickerdt 12581.

Distinguished by its habit: a rounded bush habit

with many wiry branchlets and straight, persistent,

spine-like peduncles.

When Harvey described H. spinosa, he stated that

he was not sure that M. spinosa Burch, was the same

species, therefore his name is not a new combination

but a new name.

29. Hermannia linifolia Burm.f., Prodr. FI. Cap.

18 (1768); L., Mant. Alt. 256 (1771); DC., Prodr.

1 : 495 (1824), as H. linifolia L. Type: Cape of Good

Hope, specimen in herb. Burm. f. (G, holo.; PRE,

photo.! Neg. 5185).

Mahernia scoparia Eckl. & Zeyh., Enum. 51, No. 404. (1834).

Type: Cape, “Riedvalley”, Ecklon & Zeyher Enum. No. 404

(K; PRE, photo.!; TCD; PRE; BOL!) — var. glabra Eckl. &

Zeyh. I.c. Type: Cape, Zwartland near Piketberg, Ecklon &

Zeyher s.n. (K; PRE, photo.!; TCD;).

Hermannia scoparia (Eckl. & Zeyh.) Harv., F.C. 1 : 194 (1860);

Adamson in Adamson & Salter, FI. Cap. Penins. 585 (1950).

Suffrutex, stems several from the taproot, slender,

decumbent, with ascending, secund branches, stem

and branches terminating in an inflorescence, bright

green, striate in dried specimens, glabrous or very

sparsely stellate-hairy with rays few, long and pointed;

internodes 8-40 mm long, sometimes longer. Stipules

leaf-like, about half as long as leaves, 4-12 mm long,

linear, usually broad at base, often sparsely ciliate

like the leaves, with distinct, pointed hairs. Leaves

subsessile, appearing linear but, tapering into a petiole-

like base, often acicular, 13-30 mm long, apex acute or

broad and lobed, margins often sparsely ciliate with

long, pointed, bulbous-based hairs. Inflorescence of

1 -flowered cymes arranged in terminal racemose

cymes. Calyx about 6 mm long, lobed to about mid-

way, sparsely stellate-hairy especially on margins and

veins, rays long and pointed, few or solitary from a

bulbous base, glabrescent; lobes deltoid with acute to

acuminate apex. Petals yellow or “orange-yellow”,

glabrous, about 9 mm long, upper third or half oblong-

orbicular, narrowed below into a claw with infolded

margins. Stamens about 6 mm long; filaments hyaline,

oblong, somewhat cuneate, with hairs on shoulders,

united around short stipe; anthers about 3,5 mm long,

overlapping the filment for 1 mm, ciliate, cells acute.

Ovary 2,5 mm long, 5-lobed, narrowing towards base,

stellate-hairy in upper half, especially along sutures;

stipe 1 mm long; styles 5 mm long, cohering, hairy in

basal third. Capsule more or less globose, bluntly

5-umbonate at apex and with a 1 , 5 mm long stipe.

Recorded from the Peninsula and northwards from

the Malmesbury and Piketberg Districts in the Cape.

Found in coastal Fynbos on sandy flats and dunes.

Cape. — Bellville: Tigerberg, Pillans%6\9 (BOL). Malmesbury:

near Saldanha Bay, Drege s.n.; Zwartland, Ecklon & Zeyher

Enum. No. 404b. Peninsula: Peninsula, Pappe s.n. in Herb.

Austr. Afr. 14406 (SAM); Rapenberg, Pillans 3950; Rietvlei

near Milnerton, Adamson 2657 (BOL); “Rietvlei”, Ecklon &

Zeyher s.n.; “Riedvalley”, Ecklon & Zeyher Enum. No. 404.

Piketberg: west of Aurora, Acocks 19806.

When investigating H. linifolia in the sense of

Harvey in Flora Capensis (1 : 195, 1860), it was found

that the description and specimens cited did not

agree with the type of the species but rather with a

variety of H. fiilifolia. Burman’s type of the species

agreed instead with that of H. scoparia (Eckl. & Zeyh.)

Harv. based on Mahernia scoparia Eckl. & Zeyh.

Further study of the species shows that M. scoparia

var. glabra Eckl. & Zeyher is merely a glabrescent form

of the species and is hardly worthy of any taxonomic

rank.

H. linifolia Burm. f. is characterized by the very

sparse, few-rayed, stellate hairs on the branches and

the long-pointed hairs on the margins of the leaves and

the calyx, which arise singly or in tufts from a bulbous

base. It is a bright green, low, slender-stemmed,

procumbent plant with more or less linear leaves and

leaf-like stipules which are almost as long as the leaves.

30. Hermannia decumbens Willd. ex Spreng.,

Neue Entdeck. 1:299 (1880); Link, Enum. 2:181

(1822); Reichb. Ic. Descr. PI. Cult., t. 52, fig. 1 (1822);

DC., Prodr., 1:494 (1824); Eckl. & Zeyh., Enum.

41: No. 328 (1834); Harv. in F.C. 1:185 (1860):

Adamson in Adamson & Salter, FI. Cape Penins.

583 (1950). Type: Cult. Hort. Berlin, Willdenow

12327 (B, holo., PRE, photo.!; C;).

H. pratensis Eckl. & Zeyh., Enum. 41, No. 327 (1834). Type:

Cape, Caledon, near mouth of “Klynrivier”, Ecklon tfc Zeyher

Enum. No. 327 (S; PRE, photo.!; SAM!; PRE!).

H. disermifolia sensu Eckl. & Zeyh., Enum. 41, No. 325

(1834), based on specimen from between “Breederivier and

Duivenshoeksrivier” (S; PRE, photo.!), non Jacq.

H. collina Eckl. & Zeyh., Enum. 41, No. 326. Type: Cape,

near “Langehoogde and Pot and Klynrivier”, Ecklon & Zeyher

Enum. No. 326 (S; PRE, photo.!).

H. argyrata Presl, Bot. Bemerk. 20 (1844). Type: Cape,

between Cape Agulhas and Potberg, Drege 7303 (K; PRE,

photo.!; PRE!; W!).

H. decumbens var. hispida Harv. in F.C. 1:185 (1860).

Type as for H. decumbens Willd. — var. argyrata (Presl) Harv.,

I.c. Type not indicated. — var. collina (Eckl. & Zeyh.) Harv.

I.c. (1860). Type not indicated.

Suffrutex, decumbent; branches up to about 30

cm long, trailing, with inflorescence ascending, with

small, scattered, stellate scales, hairs in parts long

and matted, sometimes scales stalked. Stipules

narrowly lanceolate-elliptic to broadly ovate, nar-

rowed or subcordate at base, acuminate and acute to

subacute at apex, stellate-pubescent with short or

I. C. VERDOORN

23

long, matted, often grey hairs, glabrescent at least

on surfaces, several-veined from base. Leaves petiolate,

variable, blade from narrowly lanceolate-elliptic to

ovate-oblong or very broadly ovate-oblong, 9-50

mm long, 5-25 mm broad, evenly or unevenly crenate

on margin, sometimes entire, stellate-pubescent with

short, or long and matted hairs, often glabrescent at

least on surfaces; petiole 4-15 mm long, stellate-

pubescent with hairs short or long and matted,

sometimes glabrescent in parts. Inflorescence terminal

in pseudocymose racemes or panicles on ultimate

branchlets, flowers clustered in a few, short cymes at

intervals along the ascending branchlets, especially

towards the ends; bracts variable, mostly stipule-like,

from narrowly oblong-elliptic to broadly ovate with

subcordate base, up to 9 x 6 mm long, sometimes the

broad bracts 1-2-lobed at apex, usually glabrescent on

surfaces; bracteoles usually 3, smaller than bracts,

linear to lanceolate-elliptic, often about 3,5x1 mm,

rarely up to 10 x 3 mm and the linear ones 8x1 mm;

peduncles usually short, about 5 mm long, rarely

longer; pedicels 1,5-5 mm long. Calyx campanulate,

subinflated, 6-8 mm long, lobed to almost halfway,

the stellate or grouped hairs usually long and matted.

Petals strongly twisted, yellow to orange, oblong-

orbicular in upper half, abruptly narrowed into

a claw with broad infolded margins, stellate-

pubescent on edges of claw and on inrolled margins.

Stamens about 6-8 mm long; filaments hyaline,

obovate-oblong, minutely stellate-pubescent on shoul-

ders. Ovary about 3,5 mm long, 5-angled or winged,

stellate-pubescent especially on the angles; stipe 0,5

mm long. Capsule enclosed in persistent calyx and

petals, 5-angled, stellate-pubescent.

Found principally in Coastal Fynbos, in dune

bush, on undulating hills and in river plains, in sand

or on shaly ground. Recorded from the Cape, from

the Peninsula and northwards to Mamre Road and

from the Caledon District eastwards along the coast

to Knysna, with a few records in valleys farther

inland.

Cape. — Bredasdorp: Between Cape Agulhas and Potberg,

Drege 7303; near Wydgelee, between Malgas and Bredasdorp,

Ellis & Schlieben 12398. Caledon: near the mouth of the Klein

River, Ecklon & Zeyher Enum. No. 327; 9 km W. of Rietpoel,

Acocks 22579. George: Groene Valley, Burchell 3679 (LE).

Knysna: Lake Pleasant Hotel, Acocks 21527. Malmesbury:

Mamre Road, Compton 18147 (NBG; BOL). Montagu: Kloof

at Montagu Baths, Page 27. Mossel Bay: Mossel Bay, Marloth

7540; 17 km N. of Mossel Bay, Acocks 15392. Peninsula: Cape

Flats, H. Bolus 2999 (BOL); Salter 7469 (BOL). Riversdale:

The Fisheries, Acocks 21344; 14 km S. of Albertinia, De Winter

& Verdoorn 9099; Gouritz River, Schlechter 4395.

This species is characterized by the consistently

decumbent habit, the practically leafless, ascending

inflorescence with clustered flowers, and the hairs on

the calyx and parts of the inflorescence which are

longer and more obvious than on most other parts

of the plant. It is nearest to H. althaeifolia, having the

same semi-inflated calyx and large stipules, but it can

be distinguished by the more or less leafless inflores-

cence with long, spreading, matted pubescence on the

calyx and the usually glabrescent stipules. The areas

of distribution are adjacent but do not overlap. H.

althaeifolia is recorded from the Peninsula and

northwards to beyond Calvinia and eastwards to

Uniondale and northern Mossel Bay; it is generally

found just to the north of the Coastal Fynbos in

which H. decumbens occurs.

Ecklon & Zeyher 324 (SI; SAM!), the type of H.

diversifolia Eckl. & Zeyh. (see note under H. althaei-

folia) seems to be a form or hybrid of this species.

It is cited in F.C. as a synonym of H. althaeifolia.

31. Hermannia prismatocarpa E. Mey. ex Harv. in

F.C. 1:186 (1860); E. Mey. in Drege, Zwei Pfl.

Doc. 103 & 192, nomen (1843); Adamson & Salter,

FI. Cape Penins. 584 (1950); Verdoorn in Flower. PI.

Afr. 41, t. 1628 (1971). Type: Cape, Riebeckskasteel,

Drege s.n. (S, lecto., photo.!; TCD; LE. !).

H. hirsuta sensu Eckl. & Zeyh., Enum. 43, No. 339 (1834)

(in S!) non Schrad.

H. rugosa Adamson in J1 S. Afr. Bot. 10: 123 (1944), partly,

as to Bolus 12619 (BOL!; PRE!); Pillans 4761 (BOL!; PRE')’

Gillett 4162 (BOL!; PRE!).

Suffrutex, branching from base, branches long,

decumbent, terminating in leafless, ascending in-

florescence, up to 60 cm long, laxly to densely pubes-

cent with minute fringed scales or papillae, inter-

spersed with long, bulbous-based hairs, rarely gland-

tipped hairs present; lateral branchlets secund, sub-

erect and terminating in an inflorescence. Stipules

narrowly to broadly ovate, acute to abruptly acumi-

nate, base usually broad, oblique, rounded or cordate,

semi-amplexicaul and sometimes decurrent, 3-10 mm

long and about as broad at base, finely stellate-

pubescent above and below, sometimes also hirsute

with long, pointed hairs. Leaves petiolate; blade more

or less oblong to ovate-oblong, usually broadest below

the middle, slightly cuneate at base, 10-70 mm long,

7-34 mm broad, upper surface finely pubescent with

stellate hairs and fringed scales, glabrescent, lower

surface thinly and finely white-tomentose (sometimes

not obviously white) between nerves, and stellate-

pubescent on raised nerves ; crenate-dentate and plicate

at least when young; petiole 5-20 mm long, densely

pubescent either with fringed scales or tufted, pointed

hairs, or both. Inflorescence of erect, leafless, racemose

or paniculate cymes which terminate main branch and

branchlets; peduncles and pedicels pubescent as on

branchlets; bracts more or less like the stipules, when

in groups of 3 the central one elliptic, in upper parts

of inflorescence united and lobed, often 3-lobed.

Calyx shallowly campanulate, finely stellate-pubes-

cent to tomentose without, about 4,5 mm long,

5-lobed almost to middle, sinuses wide, lobes acute.

Petals twisted, “rich yellow”, about 9 mm long, more

or less oblong, narrowed at middle and eared, margins

half infolded in lower and narrowing to a claw, glabrous

or minutely hairy on margin and inner face. Stamens

with hyaline, obovate-oblong filaments, glabrous or

with minute hairs on shoulders; anthers ciliate, over-

lapping filaments at base. Ovary 5-angled and almost

flat between the angles, stellate-tomentose with rather

short hairs. Capsule about 10-16 mm long, with calyx

and stamens persisting at base, sharply 5-angled,

pubescent with short hairs in tubercle-based tufts,

especially on keels of carpels, umbos or horns up to

2 mm long. Figs 2.3 & 2.4.

On hills in the Cape Peninsula and from Tulbagh

northwards through Malmesbury to Clanwilliam.

Cape. — Bellville: near Philadelphia, Mauve & I. Oliver 126.

Clanwilliam: near Knechtsvlakte, Leipoldt 3671. Malmesbury:

Porterville, Schlechter 4501 ; hills near Saldanha Bay, Hutchinson

276; Kalabaskraal, Hutchinson 195; Moorresberg (sic.),

Bolus 9950. Peninsula: 50 km north of Cape Town, Gillett 4162;

Tigerberg, Pillans 4761. Piketberg: Brittania Bay, De Winter &

Verdoorn 9071; Schlieben 12433. Tulbagh: Saron, Schlechter

4863; Wellington, Haf strom & Acocks 937; Thompson 16.

Characterized by the decumbent habit and the

terminal, ascending, leafless, paniculate cymes, the

pubescence of minute fringed scales mixed with long,

pointed hairs in tubercle-based tufts, the glabrous

petals and the long, sharply 5-lobed capsules which are

pubescent with short hairs in tubercle-based tufts.

Putative hybrids have been observed between this

species and H. multiflora and H. alnifolia.

24

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

What appears to be a form or hybrid of this species

was collected south of Vanrhynsdorp near the bridge

over Doring River on the way to Clanwilliam ( De

Winter & Verdoorn 9034 & 9035). The plants were

somewhat smaller in all parts and the branches,

although spreading, were not trailing on the ground.

In all other respects they agreed with H. prismatocarpa.

This needs further investigation.

32. Hermannia scordifolia Jacq., Hort. Schoenbr.

1:64, t. 120 (1797); Reichb., Ic. Descr. PI. Cult.,

t. 58, fig. 1 (1923); DC., Prodr. 1 : 494 (1824); Harv. in

FC. 1:187 (1860), partly, excl. syn. Iconotype:

Jacq. Hort. Schoenbr. 1, t. 120 (no specimen of actual

plant figured can be traced.

H. scordifolia Jacq. var. integriuscula Harv. l.c. (1860).

Syntypes: Cape, Brandenburg, Zeyher 117 (PRE!; Z!; LE!; W);

Cape, Ebenezer, Olifants River, Drege 7289 (7298 sphalm. in

F.C.) (W!; S!).

Suffrutex, stems many, somewhat flexible, decum-

bent, up to 1,5 m long, bright green to glaucous,

striate, appearing smooth but often pubescent with

stellate or tufted hairs, the hairs short or long, few in a

tuft and bulbous-based, branched especially in upper

parts; branchlets subherbaceous, secund, ascending,

sometimes sparsely stellate-pubescent, glabrescent or

with minute gland-tipped hairs, towards the apex.

Stipules linear to filiform from a broad base, small,

about 4 mm long, early caducous. Leaves sparse,

petiolate; blade discolorous, more or less oblong, from

narrowly to broadly so, sometimes ovate-oblong and

rarely slightly obovate-oblong, 12-50 mm long, 3-28

mm broad, upper surface subdensely to sparsely and

minutely stellate-pubescent, glabrescent, nerves im-

pressed, lower surface whitish stellate-tomentose,

sometimes thinly so, nerves prominent, margins

shallowly to deeply crenate, sometimes lobate-crenate;

petiole 3-20 mm long, stellate-pubescent, densely so in

contrast to the subglabrous branches. Inflorescence

ascending, terminal on branches and branchlets,

forming lax paniculate or racemous cymes; peduncles

10-60 mm long; bracts united, about 4 mm long,

bracteoles filiform or narrowly lanceolate, 1 ,5-5 mm

long, all caducous; pedicels 3-6 mm long, glabrous,

stellate-pubescent or with some tufted hairs, sometimes

gland-tipped hairs are present. Calyx 6-8 mm long,

lobed to about midway, finely stellate-pubescent, the

rays long or short and with black, gland-tipped hairs

intermixed, lobes triangular, sinuses fairly wide.

Petals yellow, fading orange, 8-10 mm long, upper

portion suborbicular, narrowed just above middle into

a claw with inrolled margins which are obscurely

ciliate. Stamens about 7 mm long; filaments hyaline,

oblong-cuneate, 4,5 mm long; anthers ciliate, 3,5 mm

long, overlapping filaments and base. Ovary 2,5 mm

long, stellate-pubescent, especially densely at apex

where hairs are longer; stipe 1 mm long; styles 4,5

mm long, minutely and sparsely hairy in lower portion.

Capsule subglobose, exserted from persistent calyx,

5-lobed, shortly 5-umboned, finely stellate-pubescent,

more coarsly so on keel of lobes, up to 7 mm long.

Found on coastal dunes, in sandy soil in open veld

and on stony outcrops. Recorded from the Cape, from

the Malmesbury District through Clanwilliam and

Vanrhynsdorp to Calvinia.

Cape. — Calvinia: Calvinia, Henrici 3306; Clanwilliam:

Clanwilliam, Leipoldt 390; 24 km W. of Clanwilliam, De

Winter & Verdoorn 9045; Lambert’s Bay, Van Breda 330;

Langevalei, Zeyher 117. Malmesbury: Brittania Bay, Taylor

5188; 3 km N.E. of Vredendal, Hall 3686; 1 ,5 km N. of Vreden-

dal, Thompson 1007. Vanrhynsdorp: 17 km S. of Vanrhynsdorp,

De Winter & Verdoorn 9031; Sandkraal, Acocks 14801.

H. scordifolia is characterized by the long, smooth-

looking, bright green, decumbent branches with termi-

nal, ascending, leafless inflorescences and lateral

branches which are ascending and also terminate in an

inflorescence. The leaves, on fairly long petioles, vary

considerably in shape but are all somewhat dis-

colorous.

H.fistulosa Eckl. & Zeyh., based on Ecklon & Zeyher

Enum. No. 330 (S; PRE, photo!), appears to be a

luxuriant form of H. scordifolia and is not here

upheld as a separate species. Bachmann 2121 (Z!)

seems to be intermediate between Ecklon & Zeyher

Enum. No. 330 and H. scordifolia.

33. Hermannia ternifolia Presl ex Harv. in F.C.

1:197 (I860); H. ternifolia Presl, Bot. Bemerk. 22

(1844), nomen. Type: Cape, between “Groenkloof

and Saldana Bay”, Drege s.n. (S!; PRE, photo.!).

Suffrutex, erect to sprawling, branches slender with

small but prominent leaf-bases, densely covered with

fringed scales, in parts finely tomentose as well.

Stipules leafy, narrowly oblong-cuneate, oblanceolate

or spathulate, from about 7-12 mm long, 1-2,5 mm

broad near apex, on both sides with stellate scales

forming a tomentum. Leaves shortly petiolate; blade

obovate, tapering at base into short petiole, subtrun-

cate or rounded and crenate at apex, from about 7 to

14 mm long and 6-10 mm broad near the apex,

stellate-tomentose as on stipules; petiole 1-7 mm long.

Inflorescence of a few, usually 1 -flowered cymes,

terminal and in axils of upper leaves, “orange-yellow”,

“red to orange” or “marmalade-coloured” ; peduncles

short and stout or up to 5 mm long. Calyx about 7 mm

long, inflated, oblong to subglobose, somewhat 5-

angled, squamulose with minute fringed scales, lobed

to about one-third of its length; lobes broadly ovate,

mucronulate; sinuses narrow. Petals about 13 mm

long, strongly convolute, apical portion spreading to

recurved; limb oblong-obovate, about 6 mm long,

narrowing abruptly into a claw about 7 mm long,

scaly-stellate on sides above and with infolded margins

for most of length below. Stamens with oblong hyaline

filaments about 5 mm long; anthers 3 mm long,

minutely ciliate. Ovary with a very short, 0 , 5 mm long)

stipe, 3,5 mm long, somewhat 5-angled, tomentose

with short stellate hairs except at apex where hairs are

slightly longer; styles cohering, 5 mm long.

Found in coastal scrub, in sand or on limestone, up

to a few miles inland. Recorded from Saldanha Bay

and the Peninsula eastwards along the coast to

Bredasdorp in the Cape Province.

Cape. — Bredasdorp: Bredasdorp, Taylor 4037; Strand

Kloof, Compton 14752; Buffelsjacht, Van Breda 845. Caledon:

Hawston, Marloth 9194; Rooi Els, Leighton 1456, 2106;

Gansbaai, Gillett 4268. Malmesbury: between Groenekloof

and Saldanha Bay, Drege s.n. (PRE, photo.!). Peninsula:

Pringle Bay, Boucher 618; Simonstown Nature Reserve,

Taylor 5953; Swartklip, Leighton 1770.

Characterized by the petioled leaves, the inflated

calyx with broadly rounded, shortly acute lobes

which are often incurved, and the pubescence of

minute fringed scales.

Adamson in Adamson & Salter FI. Cape Penins.,

p. 586, mentions that this species cannot be separated

from H. trifoliata L. With the adequate material now

available, these species can be readily distinguished.

34. Hermannia trifoliata L., Sp. PI. 674 (1753);

Mant. Alt. 431 (1771); Cav., Diss. t. 182, fig. 1 (1788);

DC., Prodr. 1:494 (1824); Harv. in F.C. 1: 198

(1860); Adamson in Adamson & Salter, FI. Cape

I. C. VERDOORN

25

Penins. 586 (1950). Type: Cape, Cult. Herb. Hort.

Cliff. (BM, holo.; PRE, photo.!).

H. imbricata Eckl. & Zeyh., Enum. No. 381 (1834). Type:

Cape, Paardekop near Plettenberg Bay, Ecklon & Zeyher

Enum. No. 381 (S ! ; PRE, photo.!; TCD; PRE!).

Suffrutex, usually under 30 cm tall, often sprawling,

stems and branches stellate-tomentose, ridged or

quadrate, leafy in upper portion right to apex, stipules

and leaves imbricate. Stipules oblong, shortly cuneate,

stellate-tomentose with fringed crater-like scales,

often 2-veined, 5-8 mm long, 2-3 mm broad in upper

half. Leaves thickly tomentose with fringed crater-like

scales, sessile or subsessile, obovate-cuneate, slightly

longer than the subtending stipules, 7-10 mm long,

4-5 mm broad at apex, usually folded, apex broadly

rounded to truncate, sometimes crenate. Inflorescence

of cymes in terminal heads, usually nodding and

surrounded by stipules and leaves; pedicels 0-2 mm

long, with reddish brown fringed scales. Calyx about

7 mm long, lobed in upper third, becoming somewhat

inflated, covered with small crater-like fringed scales,

hairs very short, lobes erect or inflexed, rather broad,

mucronate, sinuses fairly narrow. Petals “red”,

“orange” or “claret-red”, twisted, with upper portion

of the blade spreading to reflexed, about 10 mm long,

suborbicular in upper third, narrowed in middle

forming a long waist, lower third clawed with infolded

margins, ciliate in middle. Stamens about 4 , 5 mm long,

united at base, filaments very thinly hyaline, oblong,

slightly narrowing towards base, nerves brown;

anthers usually less than 2 mm long, ciliate, over-

lapping filaments at base. Ovary pubescent, 5-lobed,

lobes rounded at apex; styles cohering, stigmas

terminal. Capsule enclosed in the somewhat enlarged

calyx, about 5 mm long, 5-angled and 5-umbonate.

Found in Fynbos and Dune Bush and sometimes on

limestone ridges or chalk hills near the sea. Recorded

along the coast from Bredasdorp to Riversdale.

Cape. — Bredasdorp: Cape “Between Aghullas & Potberg”;

Drege s.n.*; Sandhoogte, Smith 2997 ; Bredasdorp, Sidey 1807,

Acocks 15459; Bredasdorp Poort, Compton 14727; Riversdale:

Vermaaklikheid, Acocks 22530; Stille Bay, Muir 1969; Milk-

woodfontein, Galpin 3782; Albertinia, Muir 1777.

This species is characterized by the imbricating

leaves and stipules which are concentrated in the upper

portion of the branches and by the terminal flower

heads which are usually nodding and surrounded at

the base by the upper leaves and stipules. The calyx

becomes rather membranous and subinflated with age.

Cavanille’s figure of H. trifoliata is on t. 182, fig. 1,

not fig. 2, as cited in Flora Capensis and in De

Candolle’s Prodromus.

A specimen of Ecklon & Zeyher Enum. No. 381, the

type number of H. imbricata in PRE has long, exposed

pedicels whereas on photos of the same number in S

and TCD the pedicels are hidden by the leaves.

35. Hermannia concinnifolia Verdoorn in Flower.

PI. Afr. 43, t. 1691 (1974). Type: Cape, Bredasdorp

district, De Mond Forestry Reserve, Taylor 8248

(PRE, holo.!).

Suffrutex, 15-90 cm tall, branching from base;

branches ascending, leaves and stipules usually neatly

arranged in ranks (imbricate), pubescence not con-

spicuous, consisting of minute fringed scales or stellate

hairs, grouped hairs from a tubercled base also

* In Zwei pflanzengeographische Documente, p. 122, this

Drege specimen is numbered 7275, but a specimen in W bearing

this number and no name is H. concinnifolia. Gatherings in LE

and PRE with the same number are H. trifoliata.

sometimes present. Stipules green, subcoriaceous,

7-10 mm long, about 3 mm broad, more or less oblong

or oblong-elliptic, sometimes slightly oblique and

slightly narrowed towards base, mucronate or bi-

mucronate. Leaves subsessile, lettuce-green, sub-

coriaceous, broadest at apex narrowing into a petiole-

like base, plicate, apex rounded and emarginate with a

recurved mucro or more or less truncate with 2 or more

lobes, rough with minute, fringed scales and often

minute, stellate hairs on margin, margin sometimes

with a narrow red-brown rim. Inflorescence of 1-2-

flowered cymes at apex of leafy branchlets and in

axils of upper leaves; peduncle suberect, 5-12 mm

long, with stipule-like bracts at base; pedicels cer-

nuous, 2-3 mm long, with 3 narrow, 4-5 mm long

bracteoles at base. Calyx often suffused or mottled

with red, campanulate, rough with minute stellate

scales and on margins with a few short hairs from a

tubercled base, 6-9 mm long, somewhat 5-10-angled,

5-lobed; lobes about 4 mm long, broadly oblong,

shortly acuminate and mucronate, sinuses narrow.

Petals lemon-yellow, strongly twisted, shortly exserted

from calyx, about 10 mm long, blade broadly oblong,

contracted below middle into a claw with infolded

margins and minute stellate hairs on edges of claw.

Stamens about 5 mm long; filaments hyaline, oblong;

anthers 2,5 mm long, ciliate. Ovary 3-4 mm long,

2, 5-3, 5 mm broad, 5-lobed, stellate-tomentose;

stipe 0,5 mm long; styles 4 mm long. Capsule en-

closed by persistent calyx and petals, about 6 mm

long, 5 mm diam., 5-angled, thinly stellate-tomentose

between angles, more coarsely so on angles.

Found in Coastal Fynbos on limestone formations.

Recorded from the Bredasdorp, Swellendam and

Riversdale Districts in the Cape Province.

Cape. — Bredasdorp: near Wydgelegen, Acocks 23171;

Schlieben & Eltis 12399; De Hoop Provincial Farm, Lewis 6031

(NBG); Windhoek Plato, C. V. van der Merwe 1006; near

Arniston, Marsh 920; De Mond Forestry Reserve, Taylor 8248

(type). Riversdale: S. of Vermaaklikheid Post Office, Acocks

24124; Hanglip, Muir 1961. Swellendam: Cape Infanta,

Walgate 879.

This recently described species is characterized by

the leafy branches with regularly arranged, rather stiff

leaves. The stipules are leaf-like and usually more than

half as long as the leaves. They can be distinguished

from the leaves by their shape which is more or less

oblong and not clearly narrowed towards the base.

The leaves and stipules, which appear to be glabrous,

are minutely stellate-pubescent.

H. concinnifolia has, to date, been found only in

Coastal Fynbos on limestone formations in the

Bredasdorp district and eastwards to Riversdale.

It can be distinguished from the nearly related species

H. rudis and H. flammea principally by the more

densely leafy branches with regularly arranged leaves.

Compared with H. rudis it is also less roughly pube-

scent and from H. flammea it can be distinguished by

the flowers being more compactly disposed, not in

elongated racemose cymes. In the case of H. flammea

the petals are usually dark red, at least on the outside,

and on herbarium specimens they dry darker than the

calyx. This is not usually so in H. concinnifolia which

has been described as having yellow petals. Colour,

however, often varies in most species of Hermannia

and cannot be relied on as a diagnostic feature.

36. Hermannia muirii Pillans in Verdoorn in

Bothalia 10: 571 (1972). Type: Cape, Riversdale

District, Droogeveldvlakte, Muir 1882 (BOL, holo.!;

PRE!).

26

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Shrublet, low, branched from base; branches

subsparsely hispid with long, fairly soft hairs, 2 or

more grouped on a bulbous base, old wood glabres-

cent but rough with minute hair-bases. Stipules

foliaceous, more than half the length of subtending

leaf, oblong, narrowly oblong or ovate-lanceolate,

usually broad-based, 5-10 mm long, hispid especially

on margins. Leaves sessile to subsessile, neatly im-

bricately arranged, more or less oblong, usually

narrowing slightly towards base, 6-10 mm long,

entire or occasionally few-lobed at apex, hispid as on

stipules. Inflorescence of 1-2-flowered cymes in axils

of crowded upper leaves, appearing to grow in

clusters at apices of branches and branchlets; pe-

duncles short, about 1-2 mm long; bracteoles 3,

linear, about 4 mm long; pedicels about 4 mm long,

hispid as in other parts. Calyx about 6 mm long,

lobed to just below middle, laxly pubescent with

long hairs grouped on a bulbous base, tube cyathi-

form, lobes deltoid-acute. Petals white becoming pale

pink, 7-8 mm long, upper portion oblong-obovate,

narrowing just below middle into a claw with infolded

margins, glabrous. Stamens about 4 mm long; fila-

ments oblong-cuneate, hyaline; anthers 2,5 mm long,

overlapping hyaline filaments for 0,75 mm. Ovary

stellate-tomentose, about 4 mm long, 5-lobed, lobes

rounded at apex; stipe 1 mm long; styles 4 mm long.

Capsule more or less enveloped by the calyx and

petals, about 4 mm long, subrotund, 5-lobed, stellate-

pubescent with short hairs between sutures and

longer hairs on sutures.

Found in Coastal Fynbos in sandy valleys between

limestone ridges south of Albertinia; locally common.

Recorded only from the Riversdale and Uniondale

Districts.

Cape. — Riversdale: Droogeveldvlakte, Muir 1882; Ystervark-

fontein, Bayliss 3774; Buffelsfonteinlaagte, Acocks 22890;

about 15 km S. of Albertinia, Acocks 24170.

Characterized by the wiry, dark brown branches,

the light green, neatly arranged, suberect, usually

entire and hispid leaves and stipules, the stipules

which are leaf-like and almost as long as the leaves,

the white flowers which turn pale pink; the calyx

divided to or beyond the middle, drying with brownish

borders and having a cyathiform tube.

37. Hermannia floribunda Harv. in F.C. 1:201

(1860); Burtt Davy, FI. Transv. 1:267 (1926); Wild

in FZ. 1:542 (1960). Type: Transvaal, Vaal River,

Jan. 1842, Burke & Zeyher s.n. (K; PRE!).

H. melissifolia Engl, in Bot. Jb. 55: 357. Type: Lesotho,

Leribe, Dieterlen 555 (B, holo.f; PRE!; Z !).

Suffrutex, 60-120 cm tall, branched at base,

branches long, erect or sprawling, subdensely stellate-

pubescent, hairs fairly long from a minute bulbous

base, often shortly stalked, older glabrescent parts

rough with minute persistent stalks. Stipules from

ovate to ovate-acuminate, ovate-lanceolate or subu-

late, 3-7 mm long, 1-4,5 mm broad near base, soon

withering and falling on old wood, stellate-pubescent.

Leaves petiolate; blade ovate, ovate-cordate or

ovate-oblong, occasionally broader than long, about

2-6 cm long, 1-7 cm wide near base, usually 5-nerved

from base, nerves prominent beneath, stellate-

pubescent on both surfaces, hairs fine and appressed,

from a central scale, some shortly stalked and cadu-

cous, lower surface slightly lighter than upper,

margins unevenly crenate; petiole 1-3 cm long,

stellate-pubescent. Inflorescence of 1- to several-,

mostly 3-flowered, cymes simple or crowded in axils of

leaves and at apices of main branches and short

lateral branchlets; flowers many, small, under 6 mm

long on dried specimens; peduncles about 1-3 mm

long, stellate-pubescent; pedicels 2-5 mm long;

bracts more or less subulate or lanceolate, 1-3 mm

long. Calyx broadly campanulate, becoming thin and

subinflated, 5,5 mm long, stellate-pubescent with

hairs fine and some spreading, lobed to about one

third from top, lobes mucronate. Petals about 6,5

mm long, narrowly oblong-obovate, with infolded

lobes just below middle and narrow infolded margins

on claw. Stamens about 5 mm long, with narrow,

more or less linear, hyaline filaments overlapped by

anther bases. Ovary about 2,5 mm long, densely

stellate-tomentose, 5-lobed; stipe under 1 mm long;

styles about 3 mm long. Capsule about 6 mm long

and broad, exserted from persistent, subinflated

calyx, 5-lobed, lobes acute to mucronate, punctate

and stellate-pubescent, hairs from a microscopic

scale or base, becoming short and wearing off with

age.

Found on stony hill slopes, in shade or open grass-

veld, also on rocky ledges. Recorded from Leribe in

Lesotho, the Potchefstroom District in the Transvaal

and northwards to the Soutpansberg District. Also

occurs in Botswana and Rhodesia.

Transvaal. — Heidelberg: Suikerbosrand, Bredenkamp 442.

Johannesburg: Ormonde, Gerstner 6533. Krugersdorp: Wit-

poortjie, Murray 664. Letaba: Sarahsdrift, Scheepers 943.

Lydenburg: Magnet Heights, Barnard & Mogg 858; Erasmus

Pass, Codd 10499. Pietersburg: Dyer 3164; Blouberg, Strey &

Schlieben 8614. Potchefstroom: Vaal River, Burke & Zeyher

s.n.; Rhenosterfontein, Cohen 301. Potgietersrus: Pyramid

Estate, Galpin 8925. Pretoria: Leeuwhoek, Strey & Leistner

8274; Crocodile and Magalies Rivers, Burtt Davy 198. Sout-

pansberg: Louis Trichardt, Rogers 21149 (Z); Gerstner 5785;

Tshakoma, Obermeyer 991. Zeerust: Linokana, Holub s.n. (Z).

Lesotho. — Leribe, Dieterlen 555.

H. floribunda is characterized by the long branches

which arise near the base of the plant; these may be

erect or sprawling and are stellate-pubescent not

canescent-tomentose ; they bear short, floriferous

lateral branchlets. The specific name is apt, because

in the axils of the leaves and at the apices of both the

main and lateral branches, several cymes with flowers

of different ages are clustered, with several grey,

subglobose buds in each group.

From the description of H. floribunda Harv. given

by K. Schumann in Engl., Monogr. Afr. Pfl. 5:56

(1900), it is obvious that the specimens he had before

him did not belong to this species. In F.S.W.A. 84: 17,

Schumann’s concept is put into synonymy under H.

minutiflora Engl, and this is in all probability correct

although the two specimens cited have not been seen.

H. minutiflora differs from H. floribunda in being a

much smaller plant with smaller leaves and flowers

and with canescent-tomentose branches.

38. Hermannia bryoniifolia Burch., Trav. 2: 258

(1824), as bryoniaefolia; DC., Prodr. 1: 494 (1824), as

bryonifolia; Harv. (under doubtful and little known

species) in F.C. 1:207 (1860); K. Schum. in Engl.,

Monogr. Afr. Pfl. 5: 56 (1900); Burtt Davy, FI.

Transv. 1:267 (1926), partly, excl. loc. Transvaal.

Type: Cape, Postmasburg, Blinkklip, Burchell 2141

(K, holo., PRE, photo.!; PRE!; LE!).

H. rehmannii Szyszyl., Polypet. Thalam. Rehm. 139 (1887).

Syntypes: Cape, Roggeveld, Rietpoort, Rehmann 3249 (Z !) ;

Orange Free State, Bloemfontein, Rehmann 3797 (Z!); Riet-

fontein, Rehmann 3715 (Z !).

Suffrutex, bushily branched, leafy parts ^uniformly

covered with rather coarse stellate hairs from a scaly

base, on lower parts from a tubercled base, rarely

interspersed with obscure, gland-tipped hairs, rays

I. C. VERDOORN

27

many but short, mostly under 0,5 mm long. Stipules

variable, even on same specimen, from narrowly

lanceolate to broadly ovate and then sometimes

subcordate and semi-amplexiaul at base, sometimes

also 3-lobed, from about 2 mm to 5 mm long and 1-4

mm broad. Leaves ovate to ovate-oblong, varying in

size (even on same branchlet), from 5x3 mm near

base to 30x20 mm, distinctly stellate-pubescent on

both surfaces, unevenly crenate-dentate on margin

(twice crenate or shallowly lobed and crenate);

petiole 3-15 mm long, densely and coarsely stellate-

pubescent. Inflorescence axillary of 1-3-flowered

usually 2-flowered) cymes on simple, dichotomous,

ultimate branchlets, with most of leaves much

reduced and so appearing to be racemose or paniculate

cymes; peduncles varying greatly in length, 2,5-8 mm

long, coarsely stellate-pubescent; pedicels similar

but much shorter, 1-4 mm long; bracts linear-

lanceolate, 2-4 at base of pedicels, stellate-pubescent.

Calyx usually thin in texture, coarsely stellate-

pubescent, 5-lobed to slightly above or slightly below

middle. Petals yellow, more or less oblong with

margins infolded in lower half, apparently glabrous

but with a few minute hairs on infolded margins,

narrowed at base into a short claw about 1 mm long.

Stamens about 4 mm long, with hyaline oblong-

obovate filaments, joined at base and with a few

short hairs on shoulders which are overlapped by

anther bases; anthers pointed, ciliate. Ovary ±2 mm

diam., stellate-pubescent, shallowly 5-lobed, ^sessile;

styles 2,5 mm long, sparsely and shortly hairy.

Capsule about as long or slightly longer than calyx,

i.e. 3-4,5 mm, coarsely stellate-pubescent, shallowly

5-lobed, lobes obscurely keeled. Seeds reniform with

folds or wrinkles and minutely granular.

In loamy ground on hills, among rocks, on rocky

ledges in crannies. Recorded from Griqualand West,

the Orange Free State, and across the Vaal River in

the Potchefstroom District.

Cape. — Colesberg: Colesberg, Marloth 49036. Herbert: near

Campbell, Leistner 905; Kimberley: Wimbledon, Badenhorst 53;

Hay: Asbestos Mts, Marloth 2026. Barkly West: Windsorton,

Hafstrom 918 (S), Hafstrom & Acocks H. 918; Postmasburg:

Beeshoek, Leistner & Joynt 2707 ; Blinkklip, Burchett 2141.

Gordonia: Upington, between Olifantshoek and Upington,

Schweickerdt 2524. ?Roggeveld, Rietpoort, Rehmann 3249 (Z).

O.F.S. — Bloemfontein: Rehmann 3797 (Z); Naval Hill,

Moss 4609 (Z). Boshof: Boshof, Brueckner 895. Fauresmith:

Bergplaats, Smith 4378. ?Rietfontein, Rehmann 3715 (Z).

Transvaal. — Potchefstroom: Dassiesrand, Van Wyk 423.

K. Schumann mistakenly cited the locality for

Rehmann 3249 as being “Transvaal, Roggeveld,

zwischen Porter and Trigardsfontein”. It should be

“Roggeveld Rietpoort”, and as the collection has a

relatively low number, it must have come from the

north-western Cape where the other representatives

of this species are found. Acocks writes in connection

with the Rehmann locality “Rietpoort, Roggeveld”:

“There is a Rietpoort a few miles east of Lokenburg,

just to the south of Augustfontein, which could be

regarded as being in the Roggeveld”. Lokenburg is

situated about 40 km south of Nieuwoudtville in the

Cape.

H. bryoniifolia is characterized by the ovate-oblong

leaves which are lobulate-crenate and fairly coarsely

stellate-pubescent. The flowers and capsules are small.

39. Hermannia cordifolia Harv. in F.C. 1: 193

(1860). Type: Cape, Piketberg, “Bonzagen”, Zeyher

111 (TCD, holo. ; PRE, photo.!; S!; K!; SAM!).

Suffrutex, robust, erect, up to 1 m high; branches

ascending, stout, dark brown, densely stellate-pubes-

cent, the hairs of different lengths, sometimes with

gland-tipped hairs intermingled. Stipules ovate, acumi-

nate, often oblique, cordate to semi-cordate and

decurrent at base, about 12 mm long, 7 mm broad

near base. Leaves petiolate; blade large, ovate to

ovate-oblong, 2-5 cm long, 1 ,5-3 cm broad, rounded

to subcordate at base, crenate-undulate on margins,

upper surface hirsute with hairs from a scaly base but

lying in one direction, lower surface grey-tomentose

and stellate-pubescent, nerves prominent beneath;

petiole short, 3-12 mm long, densely stellate-pubes-

cent, the rays of different lengths. Inflorescence of

compound, axillary cymes; ultimate cymes 2-flowered,

congested; peduncles 10-30 mm long; pedicels up to

7 mm long; bracts 25 mm long, 7 mm broad, brac-

teoles like the stipules but smaller and some subulate.

Calyx 7, 5-9, 5 mm long, lobed in upper third,

densely stellate without, tube at first suburceolate

but wide at mouth when mature with capsule pro-

truding, lobes deltoid, erect. Petals 9-10 mm long,

upper half oblong-cuneate, narrowing gradually into

a claw with inrolled margins. Stamens about 6 mm

long; filaments hyaline, obovate; anthers overlapping

filaments for 1 mm. Ovary not seen. Capsule 8 or 9

mm long, densely stellate-pubescent, shallowly lobed,

lobes rounded at apex.

To date the species is known only from the original

gathering, Zeyher 111, and the more recent Pillans

7992. Both were collected in the Piketberg District.

Cape. — Piketberg: “Bonzagen” on the Piketberg, Zeyher 111 ;

Kapiteins Kloof Mountain, Pillans 7992.

The description was drawn up from the few pressed

specimens available {Zeyher 111 in S and SAM and

Pillans 7992 in BOL). H. cordifolia is nearest H.

rugosa, but differs in being a more robust, erect plant

with larger leaves, calyces and capsules, and in having

a clustered inflorescence and a calyx which is subur-

ceolate to campanulate, not somewhat salver-shaped

as in H. rugosa.

40. Hermannia hispidula Reichb., Ic. Descr. PI.

Cult., t. 69 (1823); Walp., Repert. 1:346 (1842);

Harv. in F.C. 1 : 207 under “little known species”

(1860). Iconotype: Cape, cult. Dresden, t. 69 in

Reichb. l.c. (typotype not traced).

Suffrutex with several ascending branches from a

thick tap-root, branches hispid with long, pointed

hairs, one to several from a prominent base, sometimes

interspersed with short gland-tipped hairs. Stipules

2-6 mm long, oblique, broad-based and decurrent,

abruptly narrowed into an acumen up to 4 mm long,

basal portion 1-3 mm broad, sparsely hispid with

bulbous-based or stellate hairs. Leaves petiolate;

blade narrowly to broadly ovate or suborbicular,

mucronate, base broadly cuneate, dentate, upper

surface and margin sparsely hispid with bulbous-

based hairs, lower surface finely and densely stellate

between the nerves, more coarsely so on nerves;

petiole 2-8 mm long, hispid. Inflorescence terminal on

branchlets, a lax racemosecyme, hispid as on branch-

lets and sometimes with gland-tipped hairs inter-

spersed, bracts 3-5, like stipules but those at base of

pedicels much reduced; peduncles 8-30 mm long;

pedicels 2-6 mm long. Calyx about 5 mm long, more

or less cup-shaped, 5 lobed almost to middle, sinuses

wide, very sparsely hispid with one or more hairs from

a bulbous base. Petals yellow, about 8 mm long,

upper portion oblong-orbicular, narrowed just below

middle into a claw with broadly infolded margins.

Stamens about 5 mm long with oblong-cuneate haline

filaments; stamens acute, ciliate, about 2,5 mm long.

28

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Ovary about 2,5 mm long, stellate-hairy, hairs long

at apex, 5-umbonate, 5-angled; stipe 0,75 mm;

styles 3,5 mm. Capsule about 6 mm long, glabrescent

or very sparsely hispid with grouped hairs, 5-angled,

5-umbonate, calyx and stamens persistent and some-

times petals too.

Found on mountain slopes, at the foot of sandstone

ridges in humus-rich pockets of sandy soil. Recorded

from the Piketberg District in the western Cape.

Cape. — Piketberg: Piketberg, Maguire 1182 (NBG); Piket-

berg mountain, Compton 22114 (NBG); near Piketberg, De

Winter & Verdoorn 9059; Bolus 7538 (BOL); near Goed-

verwacht, Bolus 8416 (BOL).

There can be little doubt that this species is repre-

sented by Reichenbach’s figure and his notes confirm

this. Distinguishing features are the acute, dentate

leaves, green and roughly pubescent on both sides and

the obvious hispid hairs on all parts of the plant. The

oblique broadbased stipule is a feature shared with

several other species such as H. prismatocarpa and

H. repetenda.

41. Hermannia rugosa Adamson in J1 S. Afr. Bot.

10: 123 (1944); Verdoorn in duuuuw 11: 519 (1975);

Adamson in Adamson & Salter, FI. Cape Penins. 584

(1950), partly. Type: Cape, Newlands, Salter 8772

(BOL, lower specimen on sheet, lecto. !).

Shrublet, rigid, spreading-erect, raderal, 30-60 cm

tall; branches rough with tubercle- D^.sed hairs.

Stipules narrowly to broadly ovate, acuminate to

abruptly acuminate, rounded or cordate to semi-

cordate and decurrent at base. Leaves petiolate; blade

varying in shape and size, narrowly to broadly ovate

to obovate, rounded or broadly cuneate to cuneate at

base, crenate and crisped on margins, minutely densely

stellate on both surfaces, sometimes whitish tomentose

on lower surface, nerves prominent beneath; petiole

about 3-7 mm long. Inflorescence of racemose cymes,

cymes 1 or more at a node, 1-3-flowered, the terminal

with flowers appearing clustered, branches of inflores-

cence stellate-pubescent with hairs short from a scaly

base or longer from a tubercled base, gland-tipped

hairs sometimes present; bracts like stipules; brac-

teoles deltoid to subulate. Calyx somewnat salver-

shaped, 5,5 mm long, 5-angled in upper half, lobed to

almost midway, sinuses wide, densely stellate-pubes-

cent, the hairs long from a small reddish brown,

tubercular base mixed with short hairs from a centr, 1

scale. Petals “golden yellow”, “orange-red” or

“yellow and red”, 9-12 mm long, twisted, oblong to

suborbicular in upper half narrowing to a claw with

infolded margins. Stamens about 6 mm long with

hyaline, obovate filaments. Ovary about 2,5 mm long,

5-lobed; stipe 0,75 mm long; styles about 5 mm long.

Capsule enclosed by persistent perianth, densely

stellate-pubescent.

Only a few records of this species have been seen.

They come from the Peninsula and from the Paarlberg

and vicinity.

Cape. — Bell ville : Tygerberg Reserve, Loubser 3367. Paarl:

Between Paarlberg and Paardeberg, Drege (S; SAM); Paarl

Mountain, Kruger M 22; Thompson 1756. Peninsula: Newlands

Estate, Salter 8772, partly (BOL); Salter 8669; 8780 (BOL);

8900 (NBG); S.W. of Durbanville, Esterhuysen 17570 (BOL).

The species is characterized by the leaves being

crisped as well as crenate on the margin, the calyx

being somewhat salver-shaped and with a thick

tomentum as in H. multiflora but with longer, softer

hairs intermixed.

Compton 20752 (NBG) from Bellville is nearest

H. rugosa.

There is evidence that H. rugosa may hybridize

with H. multiflora. The following paratypes are

excluded: Adamson 2511, which is probably a hybrid,

Pillans 4761 (cited as 4701), Gillett 4162 and H. Bolus

12619, all three of which are H. prismatocarpa E.

Mey. ex Harv. ; Gillett 3731 which is probably H.

repetenda Verdoorn, and H. Bolus 9948, a putative

hybrid nearest H. scabra Cav.

42. Hermannia althaeoides Link, Enum. Hort.

Berol. 2: 179 (1822); Eckl. & Zeyh., Enum. 42, No.

334 (1834). Type: Cape, cult. Europe (B|, holo.);

Uitenhage, Coega River near “Zwartkopshoogte”,

Ecklon & Zeyher Enum. No. 334 (S, neo!; PRE,

photo.!; TCD, photo.!; PRE!; LE!; NBG!).

H. nemorosa Eckl. & Zeyh., Enum. 42, 335 (1834). Type:

Cape, Uitenhage, between the Sunday’s and Bushman’s Rivers,

Ecklon & Zeyher Enum. No. 335 (K, PRE, photo.!;

NBG!).

H. plicata sensu Eckl. & Zeyh., Enum. No. 341 (S; PRE,

photo.!; NBG!) (1834), non Ait.;

H. mollis sensu Eckl. & Zeyh., Enum. No. 336 (K; PRE,

photo.!; PRE!; Z!) (1834), non. Willd.;

H. discolor Otto & Dietr., Gart. 8: 314 (1840). Syntypes: cult.

Europe (not traced); Uitenhage, “Zwartkopsrivier”, Ecklon &

Zeyher Enum. No. 337 (S; PRE, photo.!; PRE! NBG!) named

“var. discolor ” in S and listed as No. 327 (sphalm.) by Harv. in

F.C. 1 : 186 (1860) under H. candicans var. discolor Harv.

H. candicans sensu Harv., l.c. (1860), pro parte majore, non.

Ait.

Suffrutex, erect or suberect and widely branching

at base; branches ascending, up to 60 cm tall; pubes-

cence on branchlets, inflorescence and leaves varying

from densely and smoothly to roughly and thinly

tomentose, often with numerous minute gland-tipped

hairs interspersed, or glabrescent with scattered

rough stellate or tufted hairs. Stipules small (not

leaf-like), subulate to deltoid-acuminate, 3-8 mm long,

base 1-2,5 mm broad, rounded but not cordate,

stellate-pubescent to stellate-tomentose. Leaves pe-

tiolate; blade variable in size and pubescence, ovate-

oblong or broadly, rarely narrowly, oblong, occasion-

ally suborbicular, 10-50 mm long, 6-30 mm broad,

base broadly cuneate to rounded or somewhat

cordate, apex rounded or rarely broadly acuminate

and acute, margins faintly to distinctly crenate to

base or almost so; pubescence varying from finely to

softly stellate-tomentose on both surfaces to dis-

colourous with upper surface roughly stellate-

pubescent and lower surface softly whitish tomentose.

Inflorescence of 1 or several, 1- to several-flowered

cymes in axils of upper leaves and often forming

terminal, leafless, paniculate cymes; branches of

inflorescence from densely tomentose to glabrescent

with scattered stellate or tufted hairs, but often (on

inland specimens) with numerous, minute, gland-

tipped hairs interspersed. Calyx campanulate, about

3-7 mm long, texture thin, from thinly to densely

stellate-pubescent or tomentose outside, with a few

hairs on inside near margin, lobed in upper third or

almost to half way, lobes deltoid, sinuses wide.

Petals yellow turning red at maturity, not long per-

sistent, 7-11 mm long, upper third suborbicular, 5-6

mm diam., “waist” 1,5-2 mm broad, lower portion

with infolded margins, glabrous or obscurely ciliate

and pubescent within, not tomentose along edges of

claw. Stamens about 5 mm long, with broad, hyaline,

obovate filaments ciliate on shoulder; anthers over-

lapping the filaments at base, ciliate. Ovary 4 mm

long, 5-angled, stellate-pubescent, 5-umbonate at

apex, shortly stipitate at base, styles adhering.

Capsule usually exposed, up to 8 mm long, about as

long as petals.

I. C. VERDOORN

29

Found on rocky slopes, in hillside scrub, grassy

valleys, sandy soil along rivers, coastal thornveld and

dune veld. Recorded only from the Cape from the

coastal area of Humansdorp eastwards to Bathurst

and northwards through Somerset East to Middelburg

and then south eastwards through King William’s

Town to Komga.

Cape. — Albany : Bushman’s River, Story 2349 ; Grahamstown,

Galpin 144; Coombes Valley, Bayliss 4352; Botha’s Hill, Dyer

1451. Alexandria: near Nanaga, Acocks 12800; Quagga Flats,

Burtt Davy 11930; Bushman’s River Mouth, Lanham 132;

Galpin 10673. Bathurst: Bathurst, Sidey 3139; Hutchinson 1557;

Port Alfred, Galpin 2951; Schlechter 2686; Kowie, Britten

680; 672; 2591. Fort Beaufort: Koonap Heights, Britten 2031.

Humansdorp: Kabeljouw, Phillips 3325; Klipdrift, Thode

A2464. Keiskammahoek : Keiskammahoek, Wells 3103;

Acocks 0123. King William's Town: King William’s Town,

Tyson 850. Komga: Komga, Flanagan 487. Middelburg:

Gordonsville, Acocks 17999. Port Elizabeth: Port Elizabeth,

Sidey 3047. Queenstown: Junction Farm, Galpin 8255. Somerset

East: Klein Bruintjieshoogte, Drege 7297d. Uitenhage: Uiten-

hage, Thode A630; near Zwartkops River, Ecklon & Zeyher

Enum. No. 336; near Uitenhage, Ecklon & Zeyher 1982; 1983

(W); Marais 146; Schlechter 2507; Addo, Ecklon & Zeyher

Enum. 334; 337.

The pubescence varies considerably in this species,

from densely and smoothly to roughly and thinly

tomentose. Ecklon & Zeyher Enum. No. 334 in S was

chosen as neotype because it has the ferrugineous

pubescence mentioned in the original description.

This variable species is distinguished from H.

incana mainly by the inflorescence which ends in a

leafless, racemose or paniculate cyme, by smaller

flowers with petals that fall, leaving the maturing

capsule partly exposed, and by the tomentum not

being as uniform as that of H. incana. Moreover, H.

althaeoides is not as tall or robust a shrublet as H.

incana.

Remarking on his four varieties of H. candicans

Ait., which mainly constitute the species here de-

scribed, Harvey writes in F.C. 1: 186 “Very variable

in size and pubescence, but tolerably constant in

other characters. I have no hesitation in uniting

under one head, the various specimens of Ecklon &

Zeyher above quoted.” On examination of represen-

tatives of Ecklon and Zeyher specimens listed the

same conclusion has been reached with the exception

of Ecklon & Zeyher Enum. No. 333 which is H.

incana, the earlier name for H. candicans.

The specimens of Schlechter 6103 in Z and BOL,

the type number of H. prismatocarpoides Engl., fall

within the variation of H. althaeoides Link, and that

species is therefore probably also a synonym of our

species. Note that the Botha’s Hill, at which Schlechter

6103 was collected, is the one near Grahamstown,

not that near Durban.

43. Hermannia minutiflora Engl, in Bot. Jb.

55; 361 (1919); M. Friedrich et al. in F.S.W.A.

84: 17 (1969). Type: Cape, Namaqualand, 1’us

Schlechter 11426 (PRE, lecto.l; W!; LE!; MO!;

BOL!; Z!).

H. floribunda sensu Schum. in Engl., Monogr. Afr. Pfl.

5: 56 (1900), non Harv.

Suffrutex, erect, bushy, 20-60 cm tall, much

branched, in all parts canescent with two layers of

stellate hairs, rays many from a central scale, spreading

or in upper layer tufted on scales which often become

detached, very minute black-headed glandular hairs

often scattered in the tomentum, branches and

branchlets somewhat zigzag. Stipules varying in

size, 1,5-4 mm long, 0,5-1, 5 mm broad. Leaves

petiolate; blade suborbicular or broadly ovate-

oblong, 4-19 mm long, 4-20 mm broad, truncate or

broadly cuneate at base, sometimes some of upper

leaves cuneate in lower third, obscurely 3-5-lobate,

margin crenate, plicate or corrugated at least towards

margin; petiole 2-10 mm long. Inflorescence of 1 or

more 1-3-flowered cymes in axils of upper leaves and

terminal on ultimate branchlets often forming leafy,

cymose racemes, flowers on herbarium specimens

under 5 mm long; bracts and bracteoles about 1,5

mm long, subulate, rather thick. Calyx broadly

campanulate, on herbarium specimens up to 2,5 mm

long, lobed almost to the middle. Petals yellow or

orange, about 5 mm long, oblong-obovate in upper

half, narrowed into a waist and produced below into

a portion with infolded margins narrowing into a

claw below, stellate-pubescent dorsally about middle.

Stamens 3,5 mm long, with broadly obovate, hyaline

filaments about 2 mm long, 2,5 mm broad in upper

portion, with minute hairs on shoulders; anthers

almost 2 mm long, ciliate, overlapping filaments at

base. Ovary stellate-tomentose, shallowly 5-lobed;

stipe 0,5 mm long; styles 2,5 mm long. Capsule

enveloped by persistent calyx and corolla, 5-angled,

at first densely stellate-pubescent then glabrescent

and membranous.

Found on rocky hills, on boulders in valleys, in loose

sandy soil, among grass and on sandstone rocks.

Recorded from South West Africa from Omaruru in

the north, southwards to the border and beyond into

Namaqualand, to Kenhardt, Gordonia and Prieska

Districts in the Cape with one record from west of

LuckhofT in the Orange Free State.

Cape. — Gordonia: north of Kakamas, Wasserfalt 1155.

Kenhardt: 6,2 km N.E. of Pofadder, Hutchinson 940; near

Kakamas, Van der Schijff 8067. Namaqualand: I’us, Schlechter

1 1426. Prieska: no exact locality, Bryant J. 108.

S.W.A. — Bethanien: Helmeringshausen, Merxmidler & Giess

2826. Keetmanshoop: 48 km S. of Narubis, De Winter 3283.

Liideritz: Numeis, Dinter 8077. Maltahohe: Duwisib, Volk

12631 ; Omaruru: Erongo, Boss A65; Brandberg, Numas Valley,

Wiss 1441. Rehoboth: Bullspoort, Strey 2183. Swakopmund:

Remhoogte, Strey 2616. Warmbad: Sperlingsputz, Giess, Volk &

Bleissner 6975; Warmbad, Verdoorn & Dyer 1778. Windhoek:

Auasberge, Dinter 3485.

This species is close to H. vestita, but differs princi-

pally in the smaller flowers and the short hairs of the

tomentum. The main areas of distribution are dis-

tinct. H. minutiflora is recorded from the central

mountain plateau of South West Africa and south-

wards to the Arid Karoo-like southern districts along

the Orange River valley, and across the border into

the north-western Cape where the distribution over-

laps in places that of H. vestita. See under that species

for further notes.

44. Hermannia althaeifolia L., Sp. PI. 673 (1753);

Curtis’s bot. Mag. t. 307 (1795); Jacq., Hort. Schoenbr.

t. 213 (1797); Eckl. & Zeyh., Enum. 41, No. 322

(1834); Harv. in F.C. 1: 184 (1860), excl. syn. H.

diversifolia (in S); Verdoorn in Flower. PI. Afr. 41,

t. 1603 (1970). Type: Cape, Cult. Hort. Cliff. (BM,

holo.; PRE, photo.!).

H. hirsuta Mill., Gard. Diet. ed. 8, No. 8 (1786). Type: Cape,

Miller 8 (BM; PRE, photo.!).

H. plicata Ait., Hort. Kew ed. 1,2: 411 (1789). Type: Cape,

cult. England, seed from Masson (BM, holo.; PRE, photo.!).

Subherbaceous annuals or short-lived perennials

becoming woody at base; stems branching from base

with many lateral, ascending branchlets; branchlets

sometimes densely stellate but pubescence throughout

is of stellate hairs intermixed to a greater or lesser

degree with short, patent gland-tipped hairs and long,

patent, pointed hairs. Stipules leafy, ovate-lanceolate,

the upper larger, about 1,1 cm long (sometimes 2 cm

30

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

long), 6-8 mm broad above base, gradually acuminate

to a subacute apex, usually 3-nerved or faintly 5-

nerved, entire or rarely unequally lobate, pubescent

as on leaves and bracts. Leaves petiolate; blade varying

in size, large ones more or less 6 cm long, 4 cm broad,

ovate, ovate-oblong or ovate-lanceolate, shallowly

cordate, rounded or broadly cuneate at base; lateral

veins about 3, suberect, margins sinuate-crenate,

upper surface rather laxly stellate-pubescent, more

densely so on lower surface, sometimes with gland-

tipped hairs or long, pointed hairs intermixed; petioles

of varying lengths, the longer about 3 cm long, some-

times slightly swollen and kinked near the apex.

Inflorescence of usually 2-flowered cymes, 3 usually

developing successively in axils of leaves and crowded

at apices of branches and branchlets to form leafy

pseudopanicles; peduncles about 25 mm long, pedicels

5 mm long; bracts 3-5 at a node, sessile, narrowly

ovate to linear, varying in size, up to 5 mm long and

1-2 mm broad near base, pubescent as on other parts.

Calyx subinflated, campanulate, about 8 mm long,

lobed to above the middle, obviously pubescent with

three types of hairs, stellate, gland-tipped and long,

patent, acute hairs. Petals cadmium-yellow (and do

not turn red with maturity), twisted, about 9 mm long,

upper third an oblong blade narrowing abruptly into

a claw with broadly infolded margins, stellate-pubes-

cent along sides of claw (not on margins). Stamens

with hyaline filaments about 5 mm long and 1 ,5 mm

broad near apex, narrowly obovate, ciliate on shoul-

ders; anthers 3,5 mm long, acute, ciliate. Ovary very

shortly (0,5 mm) stipitate, about 3 mm long, densely

pubescent with stellate and long, pointed hairs,

shortly 5-lobed at apex; styles cohering, acute, about

5 mm long. Capsule enclosed in persistent calyx and

petals, 6 mm long, pubescence persistent; seeds several.

Fig. 1.1.

Found in Fynbos and Renosterveld in the south-

western Cape, on mountain slopes, hills, in waste

lands and sandy watercourses or in gravelly soil at

roadsides. Recorded from the Peninsula eastwards to

northern Mossel Bay and Uniondale, and northwards

to Calvinia.

Cape. — Calvinia: Hantam Mts, Marloth 10448. Ceres: Cold

Bokkeveld, Marloth 10722. Clanwilliam: between Witte Els and

Lamberts Hoek, Pillans 9075. Ladismith: Sevenweeks Poort,

Bayliss 2437. Malmesbury: Swartberg, Acocks 20657. Montagu:

on road to Barrydale, Ellis & Schlieben 12368. Mossel Bay:

Cloete's Pass, Acocks 14633. Peninsula: Peninsula, Prior 5340;

Camps Bay, Letty 247; Table Mtn, Ecklon in Herb. U.I. 392.

Piketberg: Piketberg, De Winter & Verdoorn 9063. Swellendam:

Barrydale, Galpin 3783. Uniondale: near Joubertina, Marsh

1398. Worcester: Worcester, Marloth 10782.

This marsh-mallow leaved Hermannia is charac-

terized by its many ascending, subherbaceous branch-

lets, the leafy stipules, the fineness of the conspicuous

pubescence and the broadly campanulate, subinflated

calyx. It is related to H. disermifolia and H. amoena

of Namaqualand and where their distribution areas

overlaps it may hybridize with these species. The

distinguishing feature is the appressed pubescence of

H. disermifolia and H. amoena which covers the

branchlets, as opposed to the soft spreading hairs

found on H. althaeifolia. Several putative hybrids

have been noted among specimens under the name

H. althaeifolia.

H. aurea Jacq., Hort. Schoenbr. t. 214 (1797) may

be a synonym but this cannot be judged from the

painting. It seems to have less of the long, patent,

acute hairs and may be a hybrid. A specimen in W,

named H. aurea , does not match Jacquin’s figure and

is H. althaeifolia L. H. diversifolia Eckl. & Zeyh.,

Enum. No. 324 included by Harvey and I.K. as a

synonym of H. althaeifolia is rather H. decumbens

(see specimen in S).

A specimen of Ecklon & Zeyher Enum. No. 323

named “ H . aurea Jacq.” in SAM is H. althaeifolia L.

45. Hermannia johanssenii N.E.Br. in Hooker’s

Icon. PI. 8, t. 2709 (1905). Type: Cape, Calvinia,

“Brand Vley”, Johanssen s.n. (K, holo. ; PRE, photo. ! ;

SAM!).

Suffrutex, low, usually under 30 cm tall, when pro-

tected somewhat taller, sometimes, probably through

heavy browsing, more or less prostrate with rather

thick woody branches; young branches leafy to apex,

completely stellate-tomentose hairs short and fine

from a minute central scale, minute glandular hairs

sometimes present, some stellate hairs very shortly

stalked and readily falling, older branches thinly

pubescent and dark coloured in contrast. Stipules

narrowly lanceolate to oblong-lanceolate, 2-8 mm

long, 1,5-4 mm broad near base, upper usually longer

than those near base of plant, minutely stellate-

pubescent. Leaves petiolate; blade suborbicular to

broadly oblong or ovate-oblong, broadly cuneate or

more rarely cuneate at base, shallowly and unevenly

crenate on margins, 7-19 mm long, 5-11 mm broad,

subdensely stellate-pubescent to tomentose with

short fine hairs from a minute central scale, minute

gland-tipped hairs sometimes present; petiole 2-7 mm

long, sometimes basal leaves with petioles up to 19 mm

long. Inflorescence of short, 1-2-, rarely 3-flowered

cymes in axils of leaves which usually continue to

apices of branches forming long, leafy, terminal,

racemose cymes with secund flowers; peduncles

0-7 mm long, stellate-pubescent; bracteoles 1-2 mm

long; pedicels 1,5-4 mm long, stellate-pubescent,

hairs slightly longer than on branches. Calyx ovoid-

globose to globose, narrowed at mouth, inflated,

pale, about 7 mm long, shortly toothed, teeth deltoid,

up to 2 mm long, sub densely stellate-pubescent with

hairs short and fine from a minute central scale,

some stellate hairs shortly stalked, rays relatively long,

readily falling. Petals twisted, yellow or “golden

yellow”, only shortly exserted from inflated calyx,

about 9 mm long, more or less narrowly oblong in

upper third, narrowing slightly into a waist and pro-

duced below into a claw with infolded margins,

minutely stellate dorsally. Stamens about 5 mm long

with narrow hyaline filaments; anthers overlapping

filaments at base, ciliate. Ovary depressed globose,

5-lobed, about 4 mm long, 5 mm diameter, densely

stellate-pubescent; stipe up to 1 mm long; styles 4 mm

long. Capsule enclosed in persistent calyx.

Found in Western Mountain Karoo, in sandy flats,

in clay, shaly ground, on dolerite ridges “among

vygies”. Recorded from Laingsburg District through

Sutherland to Calvinia District.

Cape. — Calvinia: Akkerendam, Leistner 859; Kree Kom,

Acocks 18559; Laingsburg: Klein Roggeveld, Marloth 10396.

Sutherland: Junction of Quaggafontein and Visrivier roads,

N.W. of Sutherland, Thompson 1800; Waterkloof, Marloth

9675.

Characterized by the floriferous, tomentose branches

which bear secund flowers regularly arranged in the

axils of leaves, in the upper half to two-thirds of

branchlets. Closely related to H. comosa, from which it

differs principally in the leaves, inflorescence and

distribution. H. johanssenii has smaller leaves of more

or less uniform size and arranged at regular, rather

short intervals, reaching to the tops of the branches

and consequently the axillary flowers, which are more

or less of the same age and size, form a uniform

I. C. VERDOORN

31

inflorescence. The distribution area of H. comosa lies

farther to the north-east, in the False Upper Karoo

from Richmond and Colesberg Districts northwards

to South West Africa.

H. leucophylla Presl is a synonym of H. comosa,

not of this species, as some have thought.

46. Hermannia comosa Burch, ex DC., Prodr.

1 : 493 (1824); Harv. in F.C. 1 : 184 (1860); K. Schum.

in Engl., Monogr. Afr. Pfl. 5: 60 (1900); M. Friedrich

et al. in F.S.W.A. 84: 12 (1969); Verdoorn in Flower.

PI. Afr. 41: t. 1605 (1970). Type: Cape, Asbestos

Mountains, Burchell 1683 (K, holo.; PRE!).

H. leucophylla Presl, Bot. Bemerk. 20 (1844). Type: Cape,

Winterveld between Nieuwjaarsveld & Ezelsfontein, Drege s.n.

(named H. plicata) (TCD ; PRE, photo. ! ; PRE! ; LE!).

H. dinteri Schinz in Bull. Herb. Boissier 6: 748 (1898). Type:

South West Africa, between Inachab and Liideritzbucht,

Dinter 6 (Z!).

H. comosa var. crenata K. Schum. in Engl. Monogr. Afr.

Pfl. 5:6 (1900). Type: Cape, between Griquatown and Witte

Waters, Burchell 1989 (B, holo.f). — var. minor K. Schum. in

Engl. Monogr. Afr. Pfl. 5: 60 (1900). Type: Cape, Kimberley,

Bolus in Herb. norm. 324 (B, holo.f).

H. glabripetala Engl, in Bot. Jb. 55: 355 (1919). Syntypes:

South West Africa, Aus, Steingrover 35 (Z !) ; 36 (Z !) ; Dinter

1061 (SAM!); Range 958 (SAM!); Pearson 5272 (SAM!) &

8222 (SAM!).

H. ausana Dinter ex Range in Reprium nov. Spec. Regni veg.

36: 260 (1934), nom. nud. based on Dinter 6058 (Z!).

Subherbaceous perennials, one to several stems

arising from tap-root; stems branched at base;

branches suberect to spreading-ascending, laxly sub-

branched, pubescent to tomentose with fine short

appressed stellate hairs often intermixed with stalked,

tufted hairs. Stipules narrowly deltoid, linear-lance-

olate, lanceolate or ovate-lanceolate, 3-6 mm long,

1-2,5 mm broad near the base, dorsally stellate-

pubescent. Leaves petiolate; blade oblong to oblong-

obovate, rarely oblong-elliptic, broadly cuneate to

rounded at base, 1,5-5 cm long, 0,6-3, 5 cm broad

near base, pubescent with fine stellate hairs on both

surfaces but more often densely so to tomentose

underneath, sometimes partly glabrescent above,

margins coarsely crenate to lobate-crenate or rarely

entire; petiole densely stellate-pubescent to stellate-

tomentose, often with tufted hairs intermixed.

Inflorescence of usually 2-flowered cymes, 2-3 deve-

loping at the apices of branches and in axils of upper

leaves; peduncles 5-20 mm long; pedicels 2-10 mm

long; bracts subulate to linear-lanceolate, usually 3 at

a node, sometimes 2 joined together, 2-5 mm long.

Calyx inflated, white or ivory-coloured and some

flowers suffused with shades of green, yellow or rose,

about 9 mm long, depressed globose, 5-lobed almost

to middle, with lobes incurved, pubescent on outside

with short stellate hairs intermixed with longer,

stalked, tufted hairs (the stalks sometimes up to 1,5

mm long). Petals twisted lemon-yellow, about 10 mm

long, upper third oblong-orbicular, 2,5-4 mm broad

narrowing into a claw 2 mm broad, lower two-thirds

with infolded margins, sparsely to conspicuously

dorsally pubescent with stellate and tufted hairs,

especially along sides, glabrescent. Stamens 5 with

hyaline, obovate-oblong filaments, 4 mm long, joined

at base and sometimes adhering along margins to

almost middle, pubescent on shoulders; anthers acuF,

about 4 mm long, ciliate. Ovary about 4 mm long,

shallowly 5-lobed, stellate-tomentose, hairs fine and

appressed or some, especially at apex, slightly longer

and erect; styles cohering, about 6 mm long. Capsule

enclosed in persistent calyx and petals, deeply 5-angled,

angles produced for 2 mm beyond the style bases,

rounded; seeds reniform, corrugated. Fig. 2.5.

Recorded from the north-western Cape as far south

as Richmond and Colesberg and northwards through

the Orange Free State, the south-western Transvaal

and Griqualand West to South West Africa. Found in

deep sand or gravelly soil on flats or in stony ground

at the foot of hills.

Cape.— Barkly West: Barkly West, Leistner 613; Acocks 237.

Colesberg: Colesberg, Bayliss 3882. Gordonia: Gordonia,

Leistner 2029. Hay: Hay, Acocks 1763. Herbert: Herbert,

Wilman s.n. Hopetown: Hopetown, Smith 2819. Kimberley:

Kimberley, Schlieben 8711; Acocks 16; Marloth 713. Postmas-

burg: Postmasburg, Leistner & Joynt 2783. Prieska: Prieska,

Bryant 112. Richmond: Richmond, J. Wood in TRV 19104.

O.F.S. — Bloemfontein: Bloemfontein, Hanekom 817; Burn

Davy 11770. Boshof : Boshof, Burn Davy in Govt. Herb. 10153;

10346. Edenburg: Edenburg, C. A. Smith 5182. Fauresmith:

Fauresmith, Pole Evans 1822; C. A. Smith 5277. Glen: Glen,

Mostert 528; Hey ink 22. Petrusburg: Petrusburg, Henrici 4299.

Transvaal. — Bloemhof: Bloemhof, Theron 5457; Schweizer-

Reneke: Schweizer-Reneke, Burn Davy in Govt. Herb. 11262.

S.W.A.— Aus: Aus, Marloth 5036; Dinter 6058 (Z); 6041;

6054; 3592. Luderitz: Ltideritz, Van Vuuren 920; Rehoboth:

Rehoboth, Tolken & Hardy 704; Dinter 7977. Warmbad:

Warmbad, Strahl sub Dinter 5030. Windhoek: Windhoek,

Merxmiiller & Giess 803.

Burchell distributed specimens of his 1683 under the

name H. comosa and De Candolle based the species on

that one specimen.

H. comosa is a variable species and several putative

hybrids have been noted. The feature which probably

suggested the specific epithet, namely the stalked

tufts of hairs which are commonly found on the calyx,

is not completely diagnostic. These comas fall readily

and appear to be absent on some plants. There is also

considerable variation in the density of the fine,

stellate pubescence that covers all parts of the plant

and often gives it a silvery grey sheen.

H. cinerascens Engl, in Bot. Jb. 55: 355 (1919) is

most probably a synonym, and it is so treated by

M. Friedrich et al. in F.S.W.A. The type gathering,

Range 286, seen in BOL and SAM, could conceivably

fall within the range of variation of H. comosa.

In Z, Schenk 321, which agrees with H. comosa, is

noted as the type of H. hottentotta Schinz but this

name was evidently never published.

47. Hermannia incana Cav., Diss. 6: 328, No. 473

(1788); Reichb., Ic. Descr. PI. Cult. t. 53 (1822);

Eckl. & Zeyh., Enum. 42, No. 333 (1834). Type:

Cape, cult. D. Cels, Paris (MA, holo.; PRE, photo.!).

H. candicans Ait., Hort. Kew 2:412 (1789); Jacq., Hort.

Schoenbr. t. 117 (1797); Reichb., 1c. Descr. PI. Cult. t. 55

(1822); DC., Prodr. 1 : 493 (1824); Harv. in F.C. 1 : 186 (1860),

pro parte minore. Type: Cape, cult. Hort. Kew (BM, holo.;

PRE, photo. !). — var. incana Harv., l.c., pro parte as to Ecklon &

Zeyher Enum. No. 333 (S; PRE, photo.!; SAM!).

H. praemorsa Wendl., Bot. Beobacht. 51 (1798). Type: not

traced but Wendland & Schrader 1259 in W! may be authentic;

also specimens so named in Herb. Schrader (LE!).

H. mollis Willd., Enum. Hort. Berol. 692 (1809); Reichb., lc.

Descr. PI. Cult. t. 54 (1822); DC., Prodr. 1 : 494 (1824). Type:

Cape, cult. Europe, Willdenow 12308 (B, holo.; PRE, photo.!).

Suffrutex, up to 2 m tall, usually erect; branchlets

whitish grey tomentose with densely fringed scales

with hairs flattened to give a smooth appearance,

often minute gland-tipped hairs scattered in the tomen-

tum. Stipules subulate to narrowly linear or long

acuminate from a 1,5 mm broad base, up to 7 mm

long, 1 to 1,5 mm broad at base, tomentose. Leaves

petiolate; blade ovate-oblong to broadly or narrowly

oblong, broadly cuneate to cuneate at base, 7-35 mm

32

REVISION OF HERMANN/A SUBGENUS HERMANNIA IN SOUTHERN AFRICA

long, 5-20 mm broad, softly stellate-tomentose on

both surfaces, at least when young, faintly to fairly

distinctly undulate-crenate except at the broadly

cuneate base, apex rounded or sometimes shortly

acuminate and acute; petiole 3-15 mm long, softly

tomentose. Inflorescence of 1 -several-flowered, but

mostly 2-flowered, cymes, one or more developing in

axils of upper leaves of ultimate branchlets, forming

leafy, paniculate cymes; peduncles and pedicels

tomentose but not quite as smoothly as on branchlets,

not of equal lengths but both varying, peduncles

from 2,5-20 mm long and pedicels 2-15 mm long;

bracts usually 3 at apex of peduncles, subulate, about

3 mm long, caducous. Calyx stellate-tomentose,

subinflated, campanulate, 6-9 mm long, lobed in

upper half, lobes about 3 mm long, deltoid, sinuses

wide. Petals yellow, persistent 7-10 mm long, broadly

rounded or truncate at apex, narrowing slightly at

waist, claw broad with infolded margins and densely

tomentose along sides. Stamens with hyaline, obovate

filaments, minutely and sparsely pubescent on

margins; anthers ciliate. Ovary about 8 mm long,

5-angled, stellate-pubescent, 5-umbonate at apex and

shortly stipitate at base; styles adhering, at length

shortly exserted. Capsule hidden by long-persistent

petals, 5-angled, densely stellate-pubescent, 5-umbo-

nate at apex.

Found on top of koppies, on slopes of hills and

mountains in stony karroid scrub and valley bushveld.

Recorded from the Peninsula through Worcester and

Ceres eastwards to George.

Cape. — George: Heimans River, Oliver 3579. Ladismith:

Ladismith, Bayliss BS/1092; Amalienstein, Acocks 15304.

Laingsburg: Whitehill, Compton 14875. Boschluiskloof Pan,

Leistner 245. Mossel Bay: east of Gouritz, Burchell 6418 (LE).

Oudtshoorn: E. of De Rust, Dahlstrand 1457. Peninsula:

Tygerberg, Marloth 4478. Prince Albert: Zwartberg Mts, Bolus

in Herb. Austro. Afr. 11720. Riversdale: Gouritz River Bridge,

Story 2881 ; Acocks 21629; Middledrift, Muir 1 383. Swellendam :

east of Kochmanskloof, Ecklon & Zeyher Enum. No. 333 (S).

Worcester: Veld Reserve, Van Breda 10; Olivier 128.

Schlechter 8516, named H. disermifolia, is H. incana

but Vogelfontein in the Clanwilliam District is outside

the known distribution area. Vogelfontein, Beaufort

West is more likely.

Characterized by the smooth tomentum on the

branches, the usually 2-flowered cymes in the axils of

the upper leaves, leaves which are broadly cuneate at

the base, usually broadest just below the middle and

softly tomentose on both surfaces, and the petals

which are broad and persistent, covering the capsule.

Harvey (1860), when describing H. candicans Ait.,

with four varieties, cites a number of Ecklon and

Zeyher specimens listed in their “Enumeratio” under

seven different names. Of these only the specimen

Enum. No. 333, named H. incana , agrees with Aiton’s

H. candicans and is therefore H. incana Cav. The rest

form a variable species here treated as H. althaeoides

Link. See notes under that species.

Judging from Presl’s description of H. dregeana in

Bot. Bemerk. 20 (1844) and from a specimen, Drege

7301, so named in S, H. dregeana Presl may well be

yet another synonym of H. incana Cav.

48. Hermannia vestita Thunb., Diss. Herm. 12

(1794); E. Mey. in Drege, Zwei Pfl. Doc. 60, 62, 64

(1843). Type: Cape, Thunberg s.n. Herb. No. 15499

(UPS, holo. ; PRE, photo.!).

H. candidissima Spreng. f., Syst. Veg. 18 (1828); Eckl. &

Zeyh., Enum. 42, No. 331 (1834); Harv. in F.C. 1: 201 (1860).

Type: “C.B.S., Zeyher (n. 149)” (LE!).

Suffrutex, erect, bushy, up to about 60 cm tall

but often smaller, densely tomentose on most parts

with short-rayed stellate hairs radiating from a central

scale, interspersed with usually longer rayed, grouped

hairs on a scale which becomes detached and often

with very minute, scattered, black-headed, gland-

tipped, hairs; branches or branchlets often zig-zag.

Stipules in upper parts of plant lanceolate to ovate-

lanceolate, 6-10 mm long, 2-4 mm broad just above

base, lower stipules smaller, 3-4 mm long, about 1 mm

broad above base. Leaves petiolate; blade usually

broadly ovate to suborbicular with a truncate or

broadly cuneate base, varying in size even on same

plant, from about 8 mm long and 7 mm broad just

above base, to about 35 mm long and 30 mm broad

(lower-most up to 45 mm long), some leaves in upper

parts of plant narrowly ovate-lanceolate and cuneate

in lower half, stellate-tomentose on both surfaces,

strongly corrugated at least at margins, nerves promi-

nent beneath obscurely 3-lobed as well as crenate at

margin; petiole tomentose, 4-26 mm long. Inflores-

cence of 1 or more, 1-3-flowered cymes in axils of

upper leaves and crowded at apices of ultimate branch-

lets or sometimes forming long, leafy, racemose cymes;

mature flowers on herbarium specimens 6-10 mm

long; bracts and bracteoles subulate to narrowly

lanceolate, 2-10 mm long, up to 3 mm broad near

base; peduncles and pedicels short, mostly under

5 mm long. Calyx not or subinflated with age, urceo-

late to campanulate, narrowed at mouth when young,

lobed almost to middle, stellate-tomentose with short

hairs, often mixed with long hairs, in mature flowers

on herbarium specimens measuring from 4,5 mm to

9 mm long. Petals orange to lemon-coloured, about

6 to 9 mm long, more or less oblong in upper third,

then narowed into a waist with infolded margins

and produced below into a claw, stellate-pubescent in

parts especially in waist. Stamens about 4 mm long,

with broadly obovate to suborbicular hyaline filaments

minutely pubescent on shoulders, anthers about 2,5

mm long, overlapping filament at base, ciliate. Ovary

5-lobed, about 2,5 mm long, stellate-tomentose,

hairs long at apex; stipe about 1 mm long; styles

cohering, about 5 mm long, minutely capitate at apex,

exserted from flowers. Capsule enveloped by long-

persistent calyx and corolla, about 5 mm long,

stellate-tomentose with hairs short and long, 5-

umbonate at the apex. Seeds reniform, corrugated,

1,5x1 mm, minutely papillose.

Found on mountain slopes, hillsides, river banks

and open veld. Recorded from the Karoo, the Cape

midlands, the southern Kalahari and southern Orange

Free State.

Cape. — Barkly West: near Boetsap, Acocks & Hafstrom

H 1387. Beaufort West: Farm Doornboomsfontein, Van Breda

535; Sunnyside, Esterhuysen 2705. Carnavon: Carnavon,

Erik Wall s.n. (S). Colesberg: Colesberg, Brink 247. Cradock:

near Dwingfontein, Story 1318. De Aar: E. of village, Acocks

624. Gordonia: between Olifantshoek and Upington, Schweic-

kerdt 2525. Hay: Excelsior, Acocks 488; east of Postmasburg,

Acocks 471; Dunmurray, Pole Evans 45. Laingsburg: Matjies-

fontein, Cannon 132. Middelburg: Middelburg Theron 18; Con-

way Farm, Gilfillan sub Galpin 5508. Prieska: between Carnavon

and Vosberg, Hafstrom & Acocks 944. Prince Albert: Drooge

Kloof, Marloth 1 1291 ; S. of Zeekoegat, Acocks 22697. Somerset

East : near Annsville, Long 746. Williston : below Jan Swartsberg,

A-ocks 18916. Willowmore: Grootrivier, Trompeterspoort,

Beer Valley, Ecklon & Zeyher s.n.

O.F.S. — Bethulie: Nature Reserve, Roberts 5515. Fauresmith:

Veld Reserve, Pole Evans 1820; Kies 235.

Characterized by the dense tomentum covering

most parts of the plant; the broadly ovate leaves,

which are only slightly longer than they are broad in

I. C. VERDOORN

33

the lower half with the base truncate or broadly

cuneate and the margins crenate and usually wavy, the

subinflated and somewhat urceolate calyx, and the

rays of the grouped hairs, which are longer in parts

than on the rest of the plant. In these features

H. vestita resembles H. minutiflora. As the name im-

plies, the flowers are consistently small, under 5 mm

long on herbarium specimens, as against 5,5-10 mm

in H. vestita. The calyx does not appear to be subin-

flated in H. minutiflora, also the majority of the leaves

is more suborbicular or oblate than in our species.

For a comparison of the areas of distribution see

notes under H. minutiflora.

The type of H. candidissima Spreng. f. is cited as

“Uitenhage, C.B.S. Zeyher (n. 149)”. The Uitenhage

District as delimited at the time when the plant was

collected, extended further north into the distribution

area of the species. The plant is not known, however,

from the Uitenhage District as defined today.

Specimens of Drege 7298 in MO and LE are

obviously H. vestita Thunb. Harvey (1860), however,

cites this number as H. scordifolia Jacq. var. integrius-

cula Harv. This is probably a printing error for

Drege 7289 which is H. scordifolia. Common name

recorded as “Katte Kruie”.

49. Hermannia amoena Dinter ex M. Holzhammer-

Friedrich in Mitt. Bot. StSamml., Munch. 1 : 345

(1953); M. Friedrich et al. in F.S.W.A. 84: 13 (1969)

in syn. Type: South West Africa, Luderitz, Numeis

(not “Uhabis”, see photo of Dinter 8074 at K), Dinter

8074 (K; PRE, photo.!; PRE!).

H. ernesti-ruschii Dinter ex M. Holzhammer-Friedrich, l.c. 346

(1953). Syntypes: South West Africa, Luderitz dist., Wittputs,

Dinter 8089; Wittpiitts — Sendlingsdrift, Wettstein 314 (K; PRE,

photo.!; PRE!).

Suffrutex, about 50-60 cm tall, erect, much branched

in upper half, branchlets densely stellate-tomentose

with interspersed minute gland-tipped hairs, tomentum

smooth, canescent with stellate hairs of varying

lengths, mostly short, radiating from or grouped on a

central scale (some scales readily becoming detached).

Stipules 5-10 mm long, 1-5 mm broad at base, large

ones found towards apices of branchlets and there

ovate-acuminate from a broad clasping base or semi-

ovate and decurrent on branchlets, stellate-tomentose

on both sides, hairs mostly short but sometimes long

on margins, sometimes upper stipules glabrescent on

inner face, in lower parts of plant stipules smaller,

narrowly ovate-acuminate to subulate, about 5 mm

long. Leaves petioled; blade oblong to ovate-oblong,

I, 5-4 cm long, 0,5-2 cm broad, sometimes larger,

stellate-tomentose on both surfaces, rounded at base

and apex, crenate on margins, young leaves corru-

gated, nerves impressed above, prominent beneath;

petiole stellate-tomentose, up to about 2 cm long.

Inflorescence of 2- to 3-flowered cymes arranged in

lax, racemose or paniculate cymes at apices of

branches and branchlets and occasionally in axils of

upper leaves, branches of panicle glabrescent with

sparse, stellate hairs and scattered gland-tipped hairs,

shiny brown in contrast to the canescent branchlets

which they terminate; peduncles about 15 mm long,

glabrescent; pedicels up to 5 mm long, remaining

tomentose; bracts 5-10 mm long, 1-5 mm broad at

base, ovate-acuminate to narrowly lanceolate; brac-

teoles smaller, narrowly lanceolate, linear or subulate.

Calyx about 8 mm long, subinflated, not narrowed at

mouth, subglobose or campanulate, sometimes suf-

fused with pink in parts, lobed to about middle,

stellate-pubescent without, hairs longer and softer

at base of calyx and on margins of lobes. Petals

“golden yellow” or “orange-yellow”, turning orange-

red at maturity, about 11 mm long, glabrous, oblong

to oblong-globose in the upper half, narrowed into a

fairly short and wide waist (about 2 mm wide) and

produced below into a rather wide claw with infolded

margins. Stamens about 7 mm long; filaments hyaline,

obovate, microscopically and sparsely stellate on the

shoulders; anthers overlapping filaments at base,

ciliate. Ovary about 5 mm long, 5-lobed, densely

stellate-pubescent, the hairs longer at apex; styles

about 5 mm long; stipe up to 1 mm long. Capsule

about 10 cm long, hidden by persistent petals and

calyx, stellate-pubescent, 5-lobed, lobes rounded at

apex.

Found on granite hills or in valleys, in sand or

“yellow clay”, on or west of the escarpment in

Namaqualand and South West Africa. Recorded

fequently at Steinkopf and surroundings, northwards

into the Warmbad and Luderitz Districts of South

West Africa and southwards through Spektakelberg,

that is west of Springbok, to the western parts of

Vanrhynsdorp.

Cape. — Namaqualand: Steinkopf, Theron 1303; Marloth

6781; Annenous Pass, Hardy 2646; Steyers Kraal near Kook-

fontein, H. Bolus 6683; Oranjemund, Lavranos 11034; Spek-

takelberg, Van der Merwe 191. Vanrhynsdorp: E. of Vredendal,

Hall 3645.

S.W.A. — Luderitz: 10 km S. of Wittputs, Dinter 8089;

Farm Wittputs, Merxmiiller & Giess 3193; Numeis, Dinter 8074.

H. amoena is characterized by the glabrescent,

slender, dark brown axis of the mature inflorescence,

which contrasts sharply with the canescent tomentum

of the rest of the plant, especially on the branchlets

terminating in the inflorescences. Besides this diag-

nostic feature H. amoena differs from the closely

related H. disermifolia in that the calyx of the latter is

not at all inflated, the upper stipules not so large, and

the hairs of the pubescence in H. amoena are longer

and softer on some parts of the plant, usually at the

base of the calices, giving a villose appearance to these

parts. The main areas of distribution of the two species

differ in that H. amoena usually occurs on or west of

the escarpment, while H. disermifolia is found on the

plateau east of the escarpment. The areas of distri-

bution may overlap in places. Hybrids with H. althaei-

folia are suspected.

Drege 3262 (Z), named H. erosa E. Mey. (a species

evidently never published) appears to be H. amoena.

According to “Zwei Documente” it was collected on

Kamiesberg foothills, Kasparskraal, Ellenbogfontein

and Geelbekskraal.”

Holzhammer-Friedrich, when describing this spe-

cies, cited the type specimen as coming from “Uhabis”.

A photograph of a specimen of Dinter 8074 in Kew

shows a line drawn through “Uhabis” and “Nume-is”

written above it, in what appears to be Dinter's

handwriting. The specimen could have been collected

at either locality.

50. Hermannia disermifolia Jacq., Hort. Schoenbr.

1, t. 121 (1797); Reichb., Ic. Descr. Cult. t. 68, fig. 2

(1822); DC. Prodr., 1 : 494(1824); Harv. in F.C. 1 : 188

(1860). M. Friedrich et al. in F.S.W.A. 84: 13. Icono-

type: Hort. Schoenbr. 1, t. 121 typotype not traced).

H. bryoniifolia sensu Eckl. & Zeyh., (as bryonifolia) Enum.

42, No. 332 (1834), non Burch.

Suffrutex, about 40-60 cm tall, sometimes taller,

erect, branched above; branchlets persistently and

smoothly canescent, stellate hairs appressed, with

short rays radiating from a central scale (some scales

34

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

become detached); tomentum interspersed with

minute, gland-tipped hairs. Stipules subulate to

narrowly ovate-acuminate, upper sometimes up to

7 mm long and 2,5 mm broad near base, lower

smaller, tomentose on both sides, stellate hairs short.

Leaves petiolate; blade oblong to ovate-oblong, about

10 to 20 cm long, densely stellate-tomentose on both

sides, becoming subdensely so in older leaves, surface

corrugated, margin crenate and crisped, sometimes

shallowly lobate as well near base, nerves deeply

impressed above, prominent beneath; petiole 3-9 mm

long, stellate-tomentose as on blade. Inflorescence

of 1-3-flowered cymes terminal on the branchlets and

3 or more developing in the axils of upper leaves

forming a pseudoracemose cyme, branches of in-

florescence stellate-tomentose and hairs short;

peduncles 4-15 mm long; pedicels 2-5 mm long;

bracts linear-subulate, 4-7 mm long, sometimes

narrowly ovate or oblong-acuminate, 5-7 mm long,

2 mm broad near base, occasionally 3 together and

united at base. Calyx not obviously inflated, 5-7 mm

long, lobed almost to middle, densely stellate-tomen-

tose, hairs short, fringing a central scale. Petals about

8 mm long, upper third oblong, narrowed slightly

into a waist and then produced into a claw with

infolded margins, glabrous. Stamens about 5 mm long;

filaments hyaline, obovate; anthers overlapping fila-

ments at base, ciliate. Ovary densely stellate-pubescent,

about 2,5 mm long, shallowly 5-lobed, lobes rounded

at apex; stipe short, up to 1 mm long. Capsule 6 mm

long, densely stellate-pubescent, calyx persistent at

base.

Found on rocky hills, stony slopes, in dry riverbeds

among rocks, and in sandy loam on sandstone hillocks.

Recorded from the north-eastern area of Vanrhyns-

dorp District northwards into Namaqualand as far as

Springbok and a few miles farther north.

Cape. — Namaqualand: without exact locality, Krapohl sub

Marloth 11162. Springbok: Springbok, Van der Schijff 8132;

S.W. of Springbok, Thompson 1070; Reynolds 5446; E.N.E. of

Springbok, Leistner 2538; Garies, Schlechter 11099; Kamies-

bergen, Ecklon & Zeyher Enum. 332 (S). Vanrhynsdorp:

“Hofkraal at Hartebeestrivier”, Zeyher (or Drege) 110 (PRE;

W).

H. disermifolia is characterized by the densely

stellate-tomentose calyx which is not inflated, the very

much corrugated leaves, and the branchlets and axis of

the inflorescence being dr smoothly stellate-tomentose

with the rays of the stellate hairs short. M. Friedrich

et al. in F.S.W.A. (1969) put H. amoena into syno-

nomy under H. disermifolia, but more material of the

former having come to light, it seems reasonable to

keep them apart. The areas of distribution are adjacent

and may overlap in part. H. disermifolia has been

found in the northern portion of Vanrhynsdorp

District and around Garies and Springbok in Nama-

qualand, whereas H. amoena comes principally from

farther north, around Steinkopf and northwards into

South West Africa in the Warmbad and Liideritz

Districts with a few southerly records on or near

Spektakelberg, that is west of the escarpment. For

further distinguishing features see the notes under

H . amoena.

A specimen of Ecklon & Zeyher Enum. No. 332 in

S named H. bryoniifolia Burch, is rather poor, but

does appear to be H. disermifolia, as given in the

Index Kewensis. It is certainly not H bryoniifolia

Burch.

51 . Hermannia mucronulata Turcz. in Byull. mosk.

Obshch. 31 : 217 (1858); Harv. in F.C. 1 : 199 (1860).

Type: Cape, Uitenhage, “Van Stadensriviersberge”,

Ecklon & Zeyher Enum. No. 353 (TCD; PRE,

photo.!; S; SAM!; LEI; W!).

H. salviifolia sensu Eckl. & Zeyh., Enum. 45, No. 353 (1834),

as salvi fjtia, non L.f.

Suffrutex, 30 cm-1,25 m high, branchlets shortly

tomentose with fringed scales interspersed with groups

of short hairs from a scaly base. Stipules leaf-like,

subsessile or petiole up to 1 mm long, about 15x4 mm,

velvety tomentose on both surfaces, mucronate,

somewhat cuneate at base. Leaves petioled; blade

obovate-oblong, 15-30 mm long, 6-15 mm broad in

upper half, entire, velvety pubescent on both surfaces,

mucronate, cuneate at base; petiole 2-4 mm long.

Inflorescence of few-flowered cymes in the axils of

upper leaves; peduncle about 6 mm long; bracts

linear to subulate, 3-10 mm long; pedicels 3-7 mm

long, shortly tomentose. Calyx inflated to subinflated,

subglobose, about 9-10 mm long, lobed in upper third,

subsparsely to densely stellate-pubescent, hairs short,

many in a cluster on a scaly base, lobes 2-3 mm long,

more or less deltoid with narrow sinuses. Petals yellow,

shortly exserted from calyx, about 14 mm long with

the apical portion recurved, narrowed in lower third to

a short waist and produced into a claw with infolded

margins, shortly hairy in waist. Stamens about 8 mm

long with broadly obovate hyaline filaments; anthers

about 4 mm long, overlapping filaments at base.

Ovary about 3 mm long, densely and shortly tomen-

tose; stipe 1 mm long; stigmas 6 mm long. Capsule

enclosed in persistent perianth.

Found on mountains, along slopes and in gorges.

Recorded from Humansdorp, Steytlerville, Uitenhage,

Port Elizabeth and Albany Districts.

Cape. — Albany: Hounslow, Galpin 145. Humansdorp:

Otteriord area, Oliver 4465. Port Elizabeth: “Van Stadens-

riviersberge’’, Ecklon & Zeyher Enum. No. 353: Steytlerville:

Slagboom, Britten 5813. Uitenhage: 25 km up Elands Riv.

road, Acocks 21263; Winterhoek Mountains, Fries, Norlindh &

Weimarck 1055.

Characterized by the leaf-like stipules, the stipules

and leaves which are velvety tomentose on both

surfaces, and by the subglobose, inflated calyx.

When Turczaninov described this species, based on

Ecklon & Zeyher’s concept of H. salviifolia, Enum.

No. 353, he wrote that it differed from H. salviifolia in

the inflated calyx and that H. salviifolia could not be

distinguished from H. involucrata. This suggests that

he had Thunberg’s sense of H. salviifolia in mind which

l have found is H. involucrata. H. salviifolia L. f. has

an inflated calyx but H. mucronulata differs from it

principally in the velvety pubescent, entire leaves.

H. mucronulata is nearest H. ve/utina in leaf charac-

ter but differs in the larger flowers and the calyx

which is lobed in the upper third and not to the middle

or beyond.

Putative hybrids between our species and H. hyssopi-

folia have been observed.

Much confusion exists in connection with speci-

mens of Drege 7267 as seen in various herbaria.

In LE, S and W it is definitely H. mucronulata', in

MO this number is mounted on a sheet with two

distinct elements and two collectors’ labels. The

specimen in the lower right-hand corner (nearest the

Drege label) is H. salviifolia var. grandistipula. The

other label, Krauss 1588 collected on the Winterhoek

Mountains, belongs with the specimens in the upper

half which are H. mucronulata. (See also H. bracteosa

under Species Requiring Further Attention (p. 61). The

type specimen of H. bracteosa Presl, in PR has the

same two elements on it but only one label, Drege 7267 .

I. C. VERDOORN

35

Presl described the right-hand specimen as H. brac-

teosa and the left-hand one as H. velutina. The former

has not been matched and appears to be a hybrid

between H. velutina and H. salviifolia var. grandistipula

while the latter is H. mucronulata.

52. Hermannia suavis Presl ex Harv. in F.C.

1: 198 (1860). Syntypes: Cape, “Zuurebergen”,

northern slopes, Drege 7268 (K; PRE, photo.!; W!;

Z!); between Coega and Sunday rivers, Ecklon &

Zeyher Enum. No. 354 (TCD; PRE, photo.!; S!;

LE!; Z!).

H. involucrata sensu Eckl. & Zeyh., Enum. No. 354 (1834),

non Cav.

Suffrutex, low, up to about 30 cm tall, much

branched, lateral branches spreading, long, appres-

sedly scaly and villose with loose, tufted, spreading

hairs. Stipules broadly oblong to oblong and oblong

elliptic, velvety appressed-pubescent on both faces

with tufted, villose hairs intermingled on lower

surface, usually 3-nerved from base on lower surface.

Leaves crowded towards apices of branches, petioled;

blade obovate to oblong-cuneate or suborbicular-

cuneate, rounded or subtruncate at apex, often folded,

6-12 mm long, 4-7 mm broad (near apex), velvety

tomentose on both surfaces interspersed with grouped

villous hairs from a scaly or tubercled base; petiole

2-6 mm long. Inflorescences of crowded, short cymes

at apices of branchlets and in axils of upper leaves;

pedicels short, up to 3 mm long, villose; bracts (not

conspicuous), caducous, subulate, about 10x3 mm,

dorsally villose, Calyx villose, inflated; tube urceolate

to broadly oblong, 6 mm long, narrowed at the mouth,

5-lobed in upper quarter; lobes 2 mm long, sinuses

narrowly deltoid. Petals yellow, apparently glabrous,

more or less narrowly oblong, about 10 mm long,

apex reflexed, narrowed below middle in a claw with

broad, infolded margins. Stamens about 5 mm long,

filaments oblong, hyaline; anthers about 2 mm long,

overlapping filaments at base, ciliate. Ovary shallowly

5-lobed, about 2 mm long, stellate-pubescent with

pale hairs, hairs longer at apex; stipe 1 mm long;

styles about 3 mm long.

Found on hills and grassy plains from the dune hills

at Cape Recife northwards to the Zuurberg. Recorded

from the Port Elizabeth, Uitenhage and Somerset

East Districts of the Cape.

Cape. — Somerset East: northern slopes of the Zuurberg,

Drege 7268. Uitenhage: Long 1305; between Coega & Sunday

rivers, Ecklon & Zeyher Enum. No. 354. Port Elizabeth: Aloes,

I. L. Drege in Govt. Herbarium 3029; hill and dunes Cape

Recife, Ecklon & Zeyher 1992a; Glendinning Vale Reservoir,

Olivier 461.

This species is close to H. salviifolia var. grandisti-

pula, but differs mainly in its pubescence which is pale

and villose and its leaves, which are velvety tomentose

on both surfaces with scattered villose hairs, whereas

the leaves in H. salviifolia var. grandistipula are densely

but coarsely stellate hairy on both surfaces with some

of the long hairs or the stellate hairs golden yellow.

H. suavis is characterized by leaf-like stipules and an

inflated calyx, besides the pale, villose pubescence

which gives the plant a soft look.

53. Hermannia hyssopifolia L., Sp. PI. 674 (1753);

Cav., Diss. 6, t. 181, fig. 3 (1788); Reichb., Ic. Descr.

PI. Cult. t. 49, fig. 3 (1822); DC., Prodr. 1 : 494 (1824);

Eckl. & Zeyh., Enum. 45, No. 355 (1834); Harv. in

F.C. 1: 197 (1860); Adamson in Adamson & Salter,

FI. Cape Penins. 586 (1950). Type: Cape, Cult., in

Herb. Hort. Cliffortianus (BM, holo. ; PRE, photo.!).

H. angularis sensu Eckl. & Zeyh., Enum. 356 (1834); sensu

Reichb., Ic. Descr. PI. Cult. t. 68, fig. 1 (1822), non Jacq.

Suffrutex, virgate with rigid branches, 36 cm-2 m

high, sometimes over 2 m tall near river, new growth

roughly and densely stellate-pubescent intermingled

with tufted hairs on raised bases. Stipules varying in

size from 2 mm long and subulate near base of plant

to 12 mm long and 3 mm broad in upper portions.

Leaves petiolate; blade broadest at or near apex and

long-cuneate to base or oblong-elliptic, cuneate in

lower half only, 9-30 mm long, 3-15 mm broad,

toothed at apex only or in upper half, finely stellate-

pubescent on both surfaces, sometimes densely so to

velvety tomentose, glabrescent; petiole 5-8 mm long.

Inflorescence of several bracteate cymes usually

crowded at apices of branches and branchlets; bracts

large, 3- or more-nerved from base, about 9x4 mm;

bracteoles smaller, about 3-7 mm long, 1,5-3 mm

broad; peduncles 4-9 mm long; pedicels 2-5 mm

long. Calyx inflated, with 5 obtuse angles, subglobose

to oblong-globose, narrowed at mouth, about 9 mm

long, 6 mm diam., lobed almost to halfway but lobes

broad and conniving, stellate-pubescent outside.

Petals “lemon to yellow”, “citron-yellow”, “pale

greenish” or “cream with red rim”, narrowly oblong,

about 10 mm long, slightly narrowed about midway

into a claw with inrolled margins, glabrous. Stamens

with broad, hyaline filaments about 6 mm long;

anthers 2 mm long, overlapping filaments at base,

minutely ciliate. Ovary about 2,5 mm long, densely

stellate-hairy; stipe 1,75 mm long; styles 3 mm long.

Capsule enclosed in persistent calyx, about 6 mm long,

5-angled; stipe 1,5 mm long.

Widely distributed in Fynbos, Coastal Renosterveld

or Karroid Broken Veld. Found on mountain slopes

or sandy plains. Recorded from the Cape Peninsula

northwards to Malmesbury and eastwards to Port

Elizabeth.

Cape. — Albany: Grahamstown, Britten s.n.; MacOwan 323

(Z). Bredasdorp: Bredasdorp, Smith 3040. Caledon: Caledon,

Schlechter 5412 (3 sheets); Dahlstrand 1175. Ceres: Mitchell’s

Pass, Schlechter 8946. Humansdorp: Humansdorp, Story 2846;

Acocks 15415. Knysna: Knysna, Tyson 993. Malmesbury:

Malmesbury, Marsh 1298. Mossel Bay: Cloete's Pass, Acocks

14637. Paarl: Paarl, Kruger s.n.; French Hoek Pass, Marsh 655.

Peninsula: Doornhoogte, Zeyher 1995; Signal Hill, Mar loth 157;

5543, 5301; Kirstenbosch, Esterhuysen 291. Port Elizabeth:

near Witteklip, Rodin 1040; Enon, Thode A2610. Riversdale:

near Albertinia, Muir 1770. Robertson: east of McGregor,

Marsh 886. Stellenbosch: Stellenbosch, Strey 667 ; Jonkershoek,

Boucher 514; Kruger 140. Swellendam: Appelskraal, Zeyher

1998 c£ 1999; Story 2887. Wellington: Wellington, Doidge s.n.

Worcester: Fairy Glen, Marloth 7546.

Characterized by the inflated calyx which is usually

pale greenish to lemon-yellow, narrowed at the mouth

with usually citron-yellow petals shortly exserted and

reflexed, by cuneate leaves toothed at the apex or in the

upper half, by a fine, close stellate pubescence often

with tufted hairs interspersed, and especially by the

bracteate inflorescences. These bracts are like large

stipules and surround the flowers. The stipules in the

lower part of the plant are the smallest and increase in

size upwards.

Specimens have been observed that appear to be

hybrids between this species and H. mucronulata.

54. Hermannia salviifolia L.f, Suppl. 302 (1781),

as H. salviifolia. Type: Cape, Thunberg s.n., Linn.

Herb. Cat. No. 854.3 (LINN, holo.; PRE, photo.!).

Suffrutex, straggling or erect, 30 cm to 1 ,22 m high,

branchlets and most vegetative parts roughly and

densely to subdensely pubescent with stellate or

36

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

tufted hairs from a scaly, glandular or raised tuber-

cular base, rarely some softer, whitish hairs mingled

with coarse, straw-to golden-coloured hairs. Stipules

oblong or narrowly oblong-acute to narrowly lanceo-

late or subulate, 3,5-10 mm long, 1,5-4 mm broad.

Leaves shortly petiolate; blade obovate-cuneate to

oblong, narrowly oblong or oblong-elliptic, apex

rounded and mucronate or subtruncate and lobed,

base rounded, roughly pubescent as described above;

7-35 mm long and from almost as broad to about four

times as long as broad. Inflorescence of 1-2-flowered

cymes, densely to subdensely clustered at apices of

brachlets; bracts linear to subulate. Calyx inflated,

urceolate, laxly or subdensely stellate-hairy. Petals

somewhat longer than calyx with apical portion

reflexed. Capsule as long as or shorter than calyx,

stellate-pubescent to glabrescent.

The aggregate species is characterized by the in-

flated, urceolate calyx and the pubescence which is

mainly of rather stiff, straw- or golden-coloured hairs

from a scaly or a glandular or raised tubercular base.

With regard to the spelling of the specific name

Recommendation 73G(a) of the Code is followed.

Key to varieties

Leaves from almost as broad as long to about twice as long

as broad, oblong or obovate, entire, or lobed at the

broadly rounded or subtruncate apex, densely stellate-

pubescent:

Stipules oblong, lanceolate or subulate, small to large

but less than half the size of the subtended leaf

(a) var. salviifolia

Stipules leaf-like and half or more than half the size of

the leaf, stellate pubescence dense and fairly soft in

parts (b) var. grandistipula

Leaves narrowly oblong, up to about 4 times as long as

broad, sublaxly stellate-pubescent, hairs short on a

markedly dark, glandular base (c) var. oblonga

(a) var. salviifolia

H. salviifolia L. f., Suppl. 302 (1781); Harv. in F.C. 1: 193

(1860), partly; — var. ovalis Harv., l.c. as to Zeyher 1993 (S; PRE,

photo. ! ; PRE !) & Drege 7280 (K, PRE, photo. ! ; PRE ! ; S ; PRE,

photo. !).

H. micans Schrad. in Schrad. & Wendl., Sert. Hannov. 1 1, t. 5

(1975); Reichb., Ic. Descr. PI. Cult. t. 49, fig. 1 (1822); Eckl. &

Zeyh., Enum. 44, No. 347 (1834). Type: Cult. Hort. Hannov.,

bottom right hand specimen, with the name H. micans written

in Schrader’s handwriting (LE, holol; PRE, photo.!).

H. latifolia Jacq., Hort. Schoenbr. t. 119 (1797). Iconotype:

Figure in Hort. Schoenbr. t. 119.

El. chrysophylla Eckl. & Zeyh., Enum. 44, No. 346 (1834), pro

parte as to some specimens with inflated calices, e.g. Ecklon &

Zeyher Enum. No. 346 (TCD; PRE, photo.!).

Suffrutex up to 1 ,22 m tall; branchlets rough with

tubercle-based, tufted hairs or stellate or fringed

scales, often all three intermingled. Stipules variable,

oblong, narrowly oblong or subulate, 3,5-10 mm long,

1 ,5-4 mm broad, usually much less than half the size

of subtended leaf. Leaves dense towards apices of

upper branchlets, shortly petiolate; blade oblong-

elliptic or obovate-oblong, entire or toothed in upper

portion especially at rounded or truncate apices,

10-30 cm long, 5-15 mm broad, never more than

about twice as long as broad, usually roughly and

coarsely stellate on lower surface, finer and more

densely so on upper surface with minute, black glands

intermingled, hairs straw-coloured or golden yellow,

sometimes exuding a glistening substance which is

clear or golden yellow; petiole up to 5 mm long,

rarely 7 mm. Inflorescence of 1-2-flowered cymes in

axils of crowded upper leaves of ultimate branchlets

forming dense clusters; peduncles and pedicels very

short; bracts linear to subulate usually obvious around

flowers. Calyx urceolate, inflated, about 7 mm long,

laxly to subdensely pubescent with usually long hairs

from a scaly base, lobes connivent about 2,5 mm

long, sinuses narrow. Petals yellow, orange, “marma-

lade-coloured”, orange becoming red with age, about

7 mm long, shortly exserted from calyx with exserted

portion recurved, narrowly oblong in upper half,

narrowing into a claw with narrowly infolded margins,

apparently glabrous. Stamens about 5 mm long;

filaments hyaline, oblong to oblong-cuneate; anthers

overlapping filaments at base. Ovary 5-lobed with a

short, 1 mm long, stipe. Capsule shorter than calyx,

about 4 mm long, 4 mm in diam., enveloped by sub-

persistent perianth, eventually exposed and glabres-

cent.

Found in coastal bush and on slopes, also in passes

north of the coast. Recorded from the Peninsula to

Port Elizabeth and more inland from Montagu to

Willowmore with one record from Graaff-Reinet.

Cape. — Caledon: Kogel Bay, Boucher 498; Groot Hangklip,

Boucher 593. George: Dysselsdorp, Acocks 20590. Graaff-

Reinet: “Cave”, Bolus 490. Humansdorp: Combrink, Acocks

13712; Kabeljaauw, Burtt Davy 12048. Montagu: Cogmans

Kloof, Michell 50. Mossel Bay: Cooper Siding, Acocks 24088;

Gouritz River Mouth, Drege 7280; west of Mossel Bay, Acocks

15402. Peninsula: Elsje's Peak, Pillans 4007; near Llandudnow,

Marloth 12745. Port Elizabeth: Parsons Vlei, Long 871. Rivers-

dale: Muiskraal, Galpin 3784; The Fisheries, Acocks 21343.

Swellendam: Stormvallei, Zeyher 1994; Hesquaspoort, Acocks

22389. Uitenhage: Winterhoekbergen, Zeyher 1993; Drege

9229a. Uniondale: Longkloof, Fourcade 1643; Mannetjiesberg,

Taylor 1463. Willowmore: Aasvogelberg, Andreae 948;

Baviaanskloof, Theron 1806.

In the typical variety the leaves vary considerably in

shape and are not always entire.

(b) var. grandistipula Harv. in F.C. 1: 193 (1860).

Type: not designated.

Differs from the typical variety principally in the

leaf-like stipules which are usually half or more than

half the size of the subtended leaf. It also differs in

the hirsute pubescence which is intermingled with the

dense, harsh pubescence of the typical variety.

Found in a fairly restricted area in the eastern Cape.

Recorded from the Albany, Uitenhage, Somerset

East and Port Elizabeth Districts.

Cape. — Albany: near Grahamstown, MacOwan 360 (GRA;

TCD); Green Hills, Britten 5388; Coombes Valley, Bayliss 4309.

Port Elizabeth: Port Elizabeth, Fries, Norlindh & Weimarck

1067 (partly). Somerset East: Kommadagga, Burchell 3328

(LE). Uitenhage: Uitenhage Zeyher s.n. in herb. Sieber FI.

Cap. No. 64 (W).

Harvey did not designate a type specimen for this

variety. The specimen in TCD of MacOwan 360

(TCD, PRE, photo.!) from near Grahamstown has

the name “7/. salviifolia var. grandistipula ” written on

the label in Harvey’s handwriting and can therefore

be considered authentic, but not a type, for it was

collected after 1860, the date of publication of the

variety. The same collection, MacOwan 360, in GRA

is also this species, but specimens of MacOwan 360 in

K and Z do not have the large stipules and in all

respects appear to be H. salviifolia var. salviifolia.

(c) var. oblonga Harv. in F.C. 1:193 (1860).

Type; not designated.

H. hyssopifolia var. integerrima Schinz in Mitt. bot. Mus.

Univ. Zurich 66: 231 (1921). Type: Cape, George, Victoria Bay,

Schlechter 2397 (Z, holo. ! ; PRE ! ; S ! ; W !).

Differs from the typical variety in the leaves being

consistently narrowly oblong, more than twice as long

as broad and not so crowded. The stellate pubescence

is not nearly so dense and the hairs on the upper sur-

face of the leaves are shorter and arise from a conspi-

cuous, dark, glandular base. The flowers are usually

fewer at the apices of the branchlets.

I. C. VERDOORN

37

Restricted to the coast in the George and Knysna

Districts and in gorges along the northern borders of

these districts.

Cape. — George: near George, Schlechter 5860; Victoria Bay,

Schlechter 2397 . Knysna: Wittedrift, Plettenberg Bay, Pappe s.n.

(TCD; PRE, photo.!); Hontini Pass, Galpin 3785; Buffalo Bay,

Keet 869; Paarde Kraal, Burchell 5168 (LE); between Knysna

and Avontuur, Fries , Norlindh & Weimarck 1637; Goukama,

Heinecken 273. Uniondale: Prince Alfred’s Pass, Acocks

21112; Sapree River, Van Breda 1173.

Harvey did not designate a type but a specimen in

TCD, Pappe s.n., was named by Harvey “var. oblonga ”

and can therefore be considered as authentic.

55. Hermannia holosericea Jacq, Hort. Schoenbr .

1, t. 292 (1797); Reichb., Ic. Descr. PI. Cult., t. 69

(1822); DC., Prodr. 1: 495 (1824); Eckl. & Zeyh.,

Enum. 45, No. 362 (1834); Harv. in F.C. 1:190

(1860). Type: Cape, cult. Hort. Schoenbr., top right-

hand specimen annotated by Jacquin (W, lecto. !;

PRE, photo. !).

H. incana sensu Thunb., FI. Cap. ed. Schult., 505 (1823),

based on Thunberg s.n. Herb. No. 15483 (UPS; PRE, photo.!).

Suffrutex 30-120 cm tall; stem erect, laxly branched

from base, lower branches long and spreading,

ascending; branches slender, at least new growth

velvety grey-tomentose; branchlets numerous, short

and slender. Stipules subulate to narrowly lanceolate,

2-8 mm long, velvety tomentose, early caducous.

Leaves petiolate; Blade from narrowly oblong-

cuneate to oblong-cuneate or suborbicular, 5-25 mm

long, 2-10 mm broad, densely and finely tomentose

on both surfaces, apex truncate or rounded, margins

crenate in upper half or at apex only; petiole 1-10

mm long. Inflorescence of compound cymes, terminal

on upper, ultimate branchlets; flowers sound and

often crowded; peduncles and pedicels velvety

tomentose; bracts linear-subulate to narrowly lan-

ceolate, 1-3 mm long, velvety tomentose on both

surfaces. Calyx about 3,5 mm long, stellate-tomentose

without, tube turbinate, 5-10-ribbed, teeth up to 1,5

mm long, sinuses wide. Petals yellow, about 4,5 mm

long, blade oblong, narrowed into a waist about

midway where it is ciliate, lower third a claw with

infolded margins. Stamens with oblong-elliptic hyaline

filaments, 1 ,75 mm long, united at base and surround-

ing stipe of the ovary; anthers very shortly over-

lapping the filaments, ciliate. Ovary 1,5 mm long,

stellate-tomentose; stipe 1,75 mm long, styles about

2,5 mm long, minutely pubescent. Capsule subglobose,

stellate-tomentose, about 2 mm diam., surrounded

by persistent calyx; stipe short, 0,5 mm long.

Found in Fynbos and Coastal Renosterveld,

grassland, marginal forests, valley scrub, on lime

flats, on hills and mountain slopes. Recorded from

Worcester and Bredasdorp north-eastwards to Graaff-

Reinet and eastwards to Albany.

Cape. — Albany: near Grahamstown, Schlechter 6063 (Z);

Ecklon & Zeyher Enum. 362; Hounslow, Britten 5680; Galpin

143; Alicedale, Rogers 37773 (Z). Bredasdorp: west of Wel-

gelegen, Taylor 3788. George: Wilderness, Levyns 725 (BOL).

Graaff-Reinet: Graaff-Reinet, H. Bolus 488 (Z). Ladismith:

Huis Rivier Pass, Van Niekerk 535 (BOL). Mossel Bay: Gouritz

River, Burchell 6421 (LE, W); De Winter & Verdoorn 9102;

Little Brak River, Sidey 1736; near Herbertsdale, Acocks 16061.

Oudtshoorn: near De Rust, Acocks 18281. Riversdale: Rivers-

dale, Acocks 23997; Swartklip, Muir 1574. Robertson: De

Hoop, Britten 657. Swellendam: Swellendam, Mauve 4695.

Uitenhage: between Coega and Sunday Rivers, Zeyher 1984;

Addo, Ecklon & Zeyher 362. Willowmore: Kouga Mtns

Oliver 463 1 .

Characterized by the velvety tomentose branchlets,

leaves, stipules and branches and bracts of the

inflorescence; the cymose panicles terminal on ulti-

mate branchlets with the flowers secund, small and

numerous. If protected, this species grows into a

large, leafy bush over 1 m high and 1 m in diameter.

H. Andreae sub Marloth 946 from Blydeberg near

Willowmore probably belongs to this species, but

the hairs are longer than usual and some stipules

larger.

56. Hermannia lavandulifolia L., Sp. PI. 674

(1753); Curtis’s bot. Mag. t. 304 (1795); Cav., Diss.

t. 180, figs 1 & lx (1788); Jacq., Hort. Schoenbr. t.

215 (1797); Reichb., Ic. Descr. PI. Cult. t. 56, fig. 1

(1822); Eckl. &. Zeyh., Enum. 45, No. 360 (1834);

Harv. in F.C. 1: 199 (1860), excl. syn. H. odorata.

Type: Cape, specimen in Herb. Hortus Cliff. (BM,

holo.; PRE, photo.!).

H. cavanillesiana Eckl. & Zeyh., Enum. No. 361 (1834),

partly as to reference Cav. Diss., t. 180, fig. 1 (1788).

Suffrutex, weak, twiggy, 30-60 cm tall; branchlets

tomentose with fringed scales and fine stellate pubes-

cence. Stipules linear-subulate, early deciduous, 3-5

mm long, velvety tomentose on both surfaces (as on

leaves). Leaves petiolate; blade oblong-oblanceolate

to oblong-obovate, 6-23 mm long, 2-6 mm broad

near apex, the majority entire, velvety tomentose on

both surfaces, mucronate, rarely some leaves sub-

truncate or with one or more lobes at or near apex,

broadly cuneate at base; petiole 2-3 mm long (ex-

tension of thick midrib obvious below). Inflorescence

of 1-3-flowered cymes, leaf-opposed; peduncle scaly-

pubescent, slender, long, 4-15 mm long, usually

curved downwards; pedicels short, 1-2,5 mm long;

bracts like leaves and stipules but smaller, 0,5-2, 5

mm long. Calyx about 5 mm long, turbinate, lobed to

just beyond middle, scaly-pubescent without, tube

5-angled, lobes with apices often infolded, Petals

yellow, about 6,5 mm long, upper third suborbicular,

narrowed at waist, claw with broad, membranous,

incurved lobes on margin. Stamens about 4,5 mm

long; filaments hyaline, apparently glabrous, oblong-

obovate; anthers ciliate, overlapping filaments at

base. Ovary finely stellate-tomentose, hairs longer at

apex, 5-lobed, 1,5 mm long with a short stipe 0,5

mm long; styles cohering, capitate at the apex,

sparsely pubescent, often exserted. Capsule more or

less hidden by subpersistent perianth, about 4,5

mm long, 5-lobed, stellate-tomentose; styles long

persistent.

Occurs on the south Cape coast in Renosterveld, on

chalky hills, in grassy country and Valley Bush.

Recorded from Bredasdorp and Caledon eastwards

to Mossel Bay or just beyond.

Cape. — Bredasdorp: Between Napier and Oudekraal, Acocks

15494; Fort Beaufort, Marsh 816. Caledon: on slopes of Klein-

berg, 0,3 km north-west of Napier, Thompson 3204. Mossel Bay:

Ruytersbosch, Van Niekerk 41 ; Morau in TRV 19039. Rivers-

dale: Nature Reserve, Taylor 7737; Plattkloof, Muir 471; near

Wyders River, Muir 886. Swellendam: 27 km from Middel-

burg. Story 2888; near Heidelberg, Acocks 21098.

Characterized by entire, velvety tomentose leaves

(tomentum made up of fringed scales and fine stellate

pubescence), distinct but short petioles, small, subulate

stipules, long, slender, leaf-opposed, spreading to

decurved peduncles, and scaly calyces lobed to

beyond the middle.

This species was confused by Ecklon & Zeyher

with H. diversistipula var. graciliflora, which they

referred to as H. cavanillesiana.

Drege 7283 in MO is for the greater part H. lavan-

dulifolia and for the lesser part H. diversistipula and

H. odorata.

38

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

57. Hermannia odorata Ait., Hort. Kew. ed. 1,

2:412 (1789); Reichb., Ic. Descr. PI. Cult., t. 56

(1822); Willd., Sp. PI. 3:597 (1800); Pers., Syn. PI.

2:216 (1807). Type: Cape, cult. Kew., sine coll.,

sine no. (BM, holo.; PRE, photo.!).

Suffrutex up to 150 cm tall with suberect branchlets,

new growth densely clothed with fringed scales,

ribbed in parts. Stipules subulate, 3-8 mm long, up

to 1 mm broad at base, usually less, velvety tomentose

on both surfaces. Leaves cuneate into a very short

petiole; blade oblanceolate to narrowly oblong-

cuneate, 10-26 mm long, 2,5-7 mm broad near apex,

velvety tomentose on both surfaces, usually folded

along midrib which is thick and prominent beneath,

acute and slightly recurved at apex, margin entire,

occasionally with an odd tooth near apex; petiole up

to 2 mm long. Inflorescence terminal on ultimate

branchlets or in axils of, or opposite, upper leaves,

in few-flowered pseudoracemose cymes, flowers

more or less secund; peduncle short and stiff; bracts

subulate, 1-2,5 mm long, usually occurring in threes

(the reduced leaf and 2 stipules), reduced leaves only

twice as long as stipules and stipule-like). Calyx

about 7 mm long, lobed to above middle (sinuses

V-shaped), densely yellow-tomentose without, hairs

fringing pit-like scales. Petals twisted, yellow, about 9

mm long, upper portion oblong, about 5 mm long,

abruptly narrowed into a short, densely ciliate waist

(stellate-pubescent dorsally) and then into a claw

with infolded margins. Stamens with oblong, hyaline

filaments about 3 mm long, anthers pointed, ciliate,

about 2,5 mm long, overlapping filaments at base.

Ovary about 2,5 mm long, 2 mm broad, tomentose,

hairs longer at apex, obscurely 5-lobed and 5-umbo-

nate at apex; styles adhering, with a few short hairs

in lower half, about 3,5 mm long; stipe under 1 mm

long. Capsule about 5 mm long, cupped at base by

persistent calyx, stellate-tomentose, 5-angled, 5-

umbonate.

Found in mountainous Renosterveld, on hills and

rocky slopes. Recorded from the Clanwilliam, Ceres

and Montagu Districts and eastwards through Rivers-

dale, Ladismith and Prince Albert to Uniondale.

Cape. — Ceres: Karoopoort, Marloth 9032; Hafstrom &

Acocks 939; N. of Pienaarspoort, Acocks 23687. Clanwilliam:

“Kerskopp”, Schlechter 8798; Kromme River, Leighton

21572; Esterhuysen 20524. Ladismith: Roodeberg Pass, Lewis

5041. Montagu: North Slope. Dahlgren & Peterson 4 (GB);

near Baths, Michell 144. Prince Albert: Swartberg Pass, Acocks

15526; Zand River Mountains, Muir 4334. Riversdale: near

Garcias Pass, Galpin 3786. Uniondale: Joubertina, Horn SKF

2444; Acocks 20011; Mannetjiesberg, Taylor 1474.

In F.C. 1: 199 H. odorata Ait. is treated as a

synonym of H. lavandulifolia , but Aiton’s type in

BM was found to match the specimens cited here. It

also matches Reichenbach’s figure very well, cited

above, named H. odorata. H. odorata differs from H.

lavandulifolia in the plants being taller, more robust,

the peduncles suberect instead of slender and cer-

nuous, the calyx yellow-tomentose without and not so

deeply lobed. This species is quite distinct from

Mahernia odorata Andr., Bot. Rep. 7: 85 (1800).

M. odorata sensu Ecklon & Zeyher, Enum. No. 405,

is H. scabra Cav. (=//. presliana Turcz.).

Specimens labelled H. fourcadii Pillans ms. belong

to this species.

58. Hermannia sulcata Harv. in F.C. 1 : 200(1860).

Type: Cape, Sondag River near Port Elizabeth,

Ecklon & Zeyher Enum. No. 359 (S, lecto. ; PRE,

photo.!).

H. odorata sensu Eckl. & Zeyh., Enum. 359, non Ait.

Suffrutex up to 60 cm tall, sprawling, much

branched, branches slender, new growth tomentose

with fringed scales. Stipules linear-oblanceolate,

sessile, acute, velvety tomentose on both sides,

5-10 mm long, about 1 mm broad. Leaves shortly

petiolate; blade velvety tomentose on both surfaces,

narrowly obovate, 12-17 mm long, 5-8 mm broad

near apex, often folded, margins usually entire,

occasionally some shallowly crenate in upper portion;

petiole about 2,5 mm long. Inflorescence of leafy

racemose cymes terminal on branchlets and in axils of

upper leaves; peduncle slender, suberect, bracteoles

linear-subulate, in threes (reduced leaf and stipules),

about 2 mm long; pedicels of ultimate pair of flowers

more or less the same length. Calyx narrowly obconic,

narrowed at the mouth, 5-lobed in the upper half,

often with 5 thick, prominent ribs alternating with

5 less prominent ribs, fairly densely pubescent with

small fringed scales and stellate hairs, hairs long on

margins of lobes; lobes about 2 mm long, acute,

mucronate. Petals stongly twisted, yellow, sometimes

partly suffused with dark red, about 8 mm long,

oblong-orbicular and emarginate in upper quarter,

narrowing slightly into a long waist and in lower

quarter into a spur with infolded margins, minutely

stellate-pubescent just behind the margin of narrowed

central portion. Stamens about 4 mm long with

oblong-cuneate hyaline filaments; anthers ciliate,

overlapping filaments at base. Ovary about 1 mm

long, shallowly 5-lobed and very shortly 5-umbonate

at the apex, stellate-pubescent, especially along the

lobes, hairs longer at the apex; stipe under 0,5 mm

long; styles adhering, rather thick; stigma terminal,

capitate.

Found on slopes, in river valleys, in clayey ground.

Recorded from around Port Elizabeth, from near

Grahamstown and in the Steytlerville District.

Cape. — Port Elizabeth: Sundays River near Port Elizabeth,

Ecklon & Zeyher Enum. No. 359 (S); Algoa Bay, Forbes s.n.

(K); Baakens River, Long 476; Walmer, Cruden 316. Albany:

north of Grahamstown, Robbertse 2. Steytlerville: Baviaans-

kloof Mts; Kleinrivier area, N.E. of Smitskraal, Oliver 4560.

The specimen of Ecklon & Zeyher Enum. No.

359 in S has the name H. sulcata written on the label in

Harvey’s handwriting. This specimen has therefore

been selected as lectotype.

It is evidently a species with a restricted distribution.

The specimens cited aoove are the only ones seen.

Apparent hybrids between this species and H.

velutina have been observed in the Steytlerville district

(< Oliver 4517 and 4516). Their calyx is too long for

H. velutina and the infusion of red in the centre of the

petals is present as in H. sulcata. The leaves are more

like those of H. velutina.

59. Hermannia velutina DC., Prodr. 1 : 495

(1824); Eckl. & Zeyh., Enum. 45, No. 357 (1834);

Harv. in F.C. 1: 199 (1860); Sim, For. FI. Cape Col.

146, t. 15 (1907). Type: Cape, Somerset East, “Zwart-

water Poort”, Burchell 3393 (K; PRE, photo.!; PRE!;

LE!).

Frutex, slender, erect, 70 cm to 3 m high, branches

rather slender, new growth pubescent to velvety

tomentose with minute fringed scales and stellate

hairs, glabrescent. Stipules leaf-like, elliptic-cuneate to

oblanceolate-elliptic, 3-20 mm long, 1-6 mm broad,

sessile or shortly petioled, velvety tomentose on both

surfaces, mucronate. Leaves petiolate; blade broadly to

narrowly oblong-elliptic, or oblanceolate-elliptic, 5-50

mm long, 2,5-20 mm broad, velvety tomentose on

I. C. VERDOORN

39

both surfaces, mucronate, often folded along midrib,

midrib impressed above, prominent beneath; petiole

2-10 mm long, Inflorescence of short, paniculate

cymes terminal on branchlets and in axils of upper

leaves, 2,5-6 cm long; bracts early caducous, terete,

up to 9 mm long, tomentose; pedicels short, 2-6 mm

long. Calyx campanulate, 5-7 mm long, lobed to

middle or beyond, pubescent with scale- or tubercle-

based hairs, drying thin. Petals usually only slightly

longer than calyx, from about 6 to 8 mm long, oblong

to suborbicular in upper third, narrowed to a ciliate

waist and produced in lower third into a claw with

broadly infolded margins. Stamens about 5 mm long,

filaments hyaline, obovate, obscurely pubescent on

shoulder, anthers about 3 mm long, ciliate, overlapping

filaments at base. Ovary finely pubescent with longer

hairs at the apex, 5-angled and 5-umbonate at apex;

stipe small, up to 0,75 mm long; styles cohering,

about 2,5 mm long. Capsule enclosed in persistent

perianth, about 5 mm long, finely stellate-tomentose

but hairs longer at apex of capsule.

Occurs on rocky slopes at forest margins, along

rivers, in open grassveld, on sand dunes and in

Coastal Fynbos. Recorded from Humansdorp east-

wards through the Transkei to Natal and inland as far

as Somerset East.

Cape. — Albany: Coombes Valley, Bayliss 4310; Grahams-

town, Rogers 28683; Stone’s Hill Range, Galpin 160. Alexandria:

Nanaga, Acocks 12797; Story 1309. Bathurst: Three Sisters,

Britten 702; Buffalo River, Galpin 1871. Humansdorp: Humans-

dorp, Rogers 2828; Wells 2948; Klipdrift, Thode A2463.

Kentani: Kentani, Pegler 453; Nomaheya Forest, Strey 6712.

Komga: Komga, Flanagan 510. Port Elizabeth: Vaal Vlei

Estate, Mogg 5379; Baakens River, Long 478. Somerset East:

“Zwartwater Poort”, Burchell 3393. Uitenhage: between

Coega and Sundays Rivers, Zeyher 1989; Groendal, Long 1160.

Victoria East: Pepperskop, Alice, Acocks 9761.

Natal. — Durban: Malvern, Medley Wood 11846 (NH).

Isipingo: Isipingo, Ward 893; 5889; between “Omcomas and

Omlas”, Drege s.n. (S).

Characterized by the velvety leaves which are entire

and petioled, the leafy stipules, campanulate calyx and

leafy panicles.

This species usually occurs as a robust plant, but

sometimes it is twiggy with small leaves and very small

flowers. Specimens representing this small form are

Zeyher 1989 in PRE, a PRE specimen originally from

Herb. Sonder, labelled with the Enum. label No. 359

H. odorata, but named H. sulcata by Tych. Norlindh in

1939 and probably also the lower specimen on No.

311334 in W (labelled H. odorata) (the upper is

H. diversistipula).

On many herbarium specimens the author is given

as Burchell, but DC. actually published the species

with Burchell’s specimen as type.

In leaf character, H. velutina approaches H. mucro-

nata Turcz., but differs in the inflorescence and the

smaller flowers with the calyx lobed to the middle.

Specimens which appear to be hybrids between

H. velutina and H. sulcata have been observed.

60. Hermannia gracilis Eckl. & Zeyh., Enum. 45,

No. 358 (1834); Harv. in F.C. 1 : 200 (1860). Type:

Cape, Graaff-Reinet, Ecklon & Zeyher Enum.

No. 358 (SI; PRE, photo.!).

Suffrutex with slender branchlets, 90-120 cm tall,

new growth thinly tomentose with fringed scales,

glabrescent. Stipules leaf-like, narrowly elliptic to

oblanceolate-elliptic, 5-20 mm long, 1-4 mm broad,

acute at apex, cuneate at base, sometimes cuneate into

a petiole up to 3 mm long, velvety tomentose on both

surfaces, sometimes fairly thinly so. Leaves usually

shortly petiolate; blade narrowly elliptic-cuneate to

oblanceolate-elliptic or oblong-cuneate, 10-23 mm

long, 4-8 mm broad, often acute at apex, cuneate at

base, margins entire or rarely with a few teeth in upper

portion, velvety tomentose on both surfaces, some-

times fairly thinly so; petiole 2-5,5 mm long. Inflores-

cence a few-flowered racemose cyme terminal on the

lateral branchlets. Calyx tubular, campanulate, about

7 mm long, tomentose with short stellate hairs, minute

gland-tipped hairs and fringed scales, lobed to above

middle, subglabrescent, lobes acute to subulate from

a deltoid base, about 2 mm long, sinuses wide, u-

shaped, margins pubescent. Petals “yellow and red”,

“dull yellow”, “orange”, “rather pale yellow” or

“brownish yellow”, contorted, about 10 mm long,

exceeding calyx, upper half broadly oblong, emargi-

nate, narrowing into a long claw with infolded

margins, pubescent along sides in centre. Stamens

about 5 mm long, with hyaline, narrowly oblong

filaments united for less than 1 mm at base around

short stipe; anthers acute, ciliate. Ovary about 4 mm

diam., subglobose, obscurely 5-lobed and 5-umbonate

at apex, stellate-tomentose hairs short but longer

towards apex of ovary; stipe almost 1 mm long;

styles adhering; stigma small. Capsule exserted from

calyx remains, about 7-8 mm long, pubescent with

stellate hairs, hairs short except along sutures and at

apex.

Found in karroid bushveld, on rocky slopes, in

river valleys and karroid scrub. Recorded from Steyt-

lerville and Graaff-Reinet Districts eastwards to the

Fish and Kat River valleys.

Cape. — Albany: near Grahamstown, Dyer 519; 1377;

Brak-Kloof, Acocks 12044; Pluto’s Vale, Story 2278; Botha’s

Hill, Schlechter 6086. Bedford: N. of Camerons Glen, Acocks

12802; 12803; Koonop Heights, Britten 2047; on banks of

Baviaans River, Theron 1051. Graaff-Reinet: Wimbledon,

Henrici 4933. Jansenville: N.E. of Jansenville, Comins 818.

Port Elizabeth: Coega, Bayliss BS 2271 (MO). Steytlerville:

N.W. of Patentie, Story 2441. Uitenhage: Kirkwood area,

Hardy 3815; Winterhoek, Krauss 1590 (MO).

61. Hermannia diversistipula Presl ex Harv. in

F.C. 1:198 (1860); Presl, Bot. Bemerk. 21 (1844),

nom. nud. Type: Cape, Swellendam, Breede River,

Drege 121 A (K, holo. ; PRE, photo.!; LE!).

Suffrutex, branches erect to spreading, ascending

or long and distichously arranged, clothed with

fringed scales. Stipules leaf-like, sessile, linear-acute

to elliptic, oblong-obovate or oblong-elliptic, 3-10

mm long, 0,75-3,5 mm broad, scaly-tomentose on

both surfaces. Leaves subsessile or narrowed into a

short petiole; blade 5-20 mm long, 2-10 mm broad at

apex, cuneate at base, apex rounded or truncate,

sometimes lobed, mucronate, usually conduplicate

with small mucro slightly recurved, scaly-tomentose

on both surfaces; petiole up to 3 or rarely 5 mm long.

Inflorescence of short, usually 2-flowered, axillary

cymes, flowers nodding and usually secund. Calyx

lobed to about middle or almost to middle, scaly-

tomentose outside, lobes deltoid or oblong, mucro-

nate, apices often incurved. Petals twisted in upper

portion, garnet-red or chrome-yellow, about 10 mm

long, upper third broadly oblong, narrowing into a

central waist, produced into a claw with infolded

margins in lower third, minutely stellate-pubescent

on margins in centre and on edges of claw. Stamens

with hyaline oblong-elliptic filaments, about 3 mm

long; anthers 1,5-2, 5 mm long, cells acute, minutely

ciliate, overlapping the filaments at base. Ovary 2-3

mm long, obscurely 5-angled and 5-umbonate at apex,

scaly-pubescent and stellate-pubescent at apex; stipe

40

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

almost 1 mm long; styles cohering, minutely pubes-

cent. Capsule 5-7 mm long, 5-umbonate at apex,

finely stellate-pubescent, more coarsely so at apex.

Seeds reniform, 1 ,5 mm long.

Found in stony soil on hill tops with karroid shrubs,

also in river flats, valley bush, Coastal Renosterveld

and Fynbos. Recorded from Tulbagh District and

south-eastwards to George.

The aggregate species is characterized by the

rather small, velvety tomentose, mostly entire leaves,

the leaf-like stipules and the nodding, often secund

flowers which are axillary and occur singly or in pairs

on fairly short peduncles. The tomentum is made up

of fringed scales intermingled with fine stellate hairs.

Key to varieties

Calyx over 5 mm wide at the mouth, with broad lobes; branches

several from a thickened woody rootstock; stipules almost

as long as the leaves (a) var. diversistipula

Calyx tube narrow, under 5 mm wide at the mouth; suffrutex

up to 90 cm tall with slender, distichous branches; stipules

about half as long as the leaves (b) var. gracilifiora

(a) var. diversistipula

Yerdoorn in Flower. PI. Afr. 41, t. 1620 (1970).

H. diversistipula var. nana Harv. in F.C. 1: 198 (1860). Type:

Cape, Storm Valley, Breede River Poort, Zeyher s.n.

Suffruticose perennial with several spreading,

ascending to erect branches from a thickened woody

rootstock; branchlets clothed with fairly coarse,

fringed scales. Stipules oblong-elliptic to oblong-

obovate, 2-8 mm long, 1-3,5 mm broad in upper

half, velvety tomentose on both surfaces with fringed

scales. Leaves sessile to subsessile; blade cuneate

from a rounded or obtuse apex, conduplicate, 3,5

to 12 cm long, up to 6 mm broad at apex, velvety

tomentose on both surfaces with fringed scales, apex

rounded or broadly obtuse and lobed. Inflorescence

axillary in upper leaves (usually only the 3 upper

leaves); peduncles 2-5 mm long; pedicels 2 mm long,

bracts linear, about 2 mm long. Calyx scaly-tomentose,

about 5 mm to 7 mm long, lobed to about middle,

lobes more or less oblong, up to 3 mm broad, apical

portion incurved. Petals red or yellow, about 10 mm

long. Capsule subglobose, shortly exserted from calyx,

5-umbonate at apex.

Found on level stony soil on hilltops with karroid

shrubs, on shale hillocks in valley bush and on river

flats. Recorded from the Malmesbury District east-

ward to Swellendam District.

Cape. — Bredasdorp: between Elandsdrift & Wiesdrift,

Da/dgren & Peterson 440a; (G.B.); Klipdale, Smith 3200;

Kykhoedie, Acocks 22724. Caledon: west of Rietpoel, Taylor

3756; 23 km from Caledon on way to Swellendam, Story 3070.

Malmesbury; Lucasfontein near Oupos Hotel, Acocks 24302.

Swellendam: Breede River, Drege 7274; Bontebok Park,

Acocks 22567. Tulbagh: Ceres Road (Wolseley), Schlechter

8988; Saron, Schlechter 7871 (LE; BOL).

Although the areas of distribution of the varieties

overlap to some extent and are approximate, the

typical variety seems to favour higher regions in

karroid shrub, whereas var. gracilifiora is found at

lower levels in Coastal Renosterveld or Fynbos.

(b) var. gracilifiora Verdoorn in Flower. PI. Afr.

41, t. 1620 (1970). Type: Cape, between Heidelberg

and Witsand, De Winter & Verdoorn 9095, cult.

PRE 20920 (PRE, holol; K!).

H. cavanillesiana Eckl. & Zeyh., Enum. 45, No. 361 (1834)

partly, as specimen Ecklon & Zeyher from between Hassaquas-

kloof and Breede River (TCD; PRE, photo!; PRE!; LE!);

sensu Harv. in F.C. I: 199 (1860) as to Ecklon & Zeyher

specimen cited.

Suffrutex up to 90 cm high with spreading branches

which are distichous and densely to sparsely pubes-

cent with minute fringed scales. Stipules linear-acute

to narrowly elliptic, 3-10 mm long, 0,75-3,5 mm

broad, tomentose on both surfaces with minute

fringed scales and stellate hairs, sometimes with

minute glands intermixed. Leaves subsessile or

shortly petiolate; blade tomentose on both surfaces,

tomentum as on stipules, oblanceolate to obovate,

cuneate at base, usually conduplicate, 4,5-20 mm

long, 2-10 mm broad near apex, entire, apex broadly

acuminate with a small recurved mucro; petiol

up to 3 mm long rarely up to 5 mm. Inflorescence

of usually 2-flowered cymes, axillary, flowers

secund; peduncle 2-5 mm long; pedicels 1-3 mm

long, bracts up to 5 at base of pedicels variable and

linear-lanceolate to ovate-lanceolate, simple or 3-

lobed, lobes up to 3 mm long, acute. Calyx about 5

mm long, lobed almost to middle, obscurely 5-angled,

tomentose without, lobes deltoid, acute or acuminate,

apices sometimes strongly in-curved. Petals red or

yellow, about 9 mm long. Capsule about 5 mm long,

eventually half exserted from persistent calyx, 5-

angled, 5-umbonate at apex.

Found in Fynbos, Coastal Renosterveld and valley

bush along the southern Cape coast. Recorded from

the Caledon District eastwards to George and

Uniondale.

Cape. — Bredasdorp: 33 km E. of Bredasdorp, Marsh

941; Kathoek, Acocks 22597. Caledon: Zwartberg in the

vicinity of the Baths, Zeyher 1988. George: near Camfer

Station, Acocks 22343. Mossel Bay: Mossel Bay, Sidey 1741;

Acocks 15398: near Cooper Siding, Acocks 24094. Oudtshoorn:

De Rust, Acocks 20458. Riversdale: Gouritz River, Drege 7272;

Marsh 576; Tygersfontein, Galpin 3790. Swellendam: between

“Hassaquaskloof and Brede River”, Ecklon & Zeyher s.n.;

between Heidelberg and Witsand, De Winter & Verdoorn

9095; Marsh 935; Mauve 4786; Bonnievale, Marloth 11834.

Uniondale: 1 km N. of Uniondale, Ellis 16396.

Besides the diagnostic characters mentioned in

the key, that is the difference in habit, calyx and

stipules, there are other features that may assist in

distinguishing the two varieties. In var. diversistipula

the fringed scales are usually larger, the leaves, which

hardly exceed the stipules, are often lobed at the

apex and the flowers are usually found in the axils

of only the upper three to five leaves. In var. gracili-

fiora the fringed scales are minute, the leaves which

overtop the stipules by about half are entire at apex

and flowers usually arise in the axils of leaves along

the upper half or two-thirds of the branchlets.

62. Hermannia cuneifolia Jacq., Hort. Schoenbr.

1, t. 124 (1797). Type: Cape., cult. Hort. Schoenbr.,

two flowering twigs on left hand of sheet, with the

name, in Jacquin’s handwriting, on the label (W,

lecto. !; PRE, photo.!).

Suffrutex, usually heavily browsed, varying in

height from 15-90 cm tall, rigid, erect or spreading,

much branched, branchlets very sparsely to very

densely lepidote-stellate and then silvery grey in

appearance. Stipules from small and subulate from

a broad base to broadly ovate in upper parts of plant,

base broad and sometimes subcordate, apex acute or

acuminate, up to about 3,5 mm long and 3,5 mm

broad at base, pubescence as on leaves but with a few

long hairs at apex or also on margin near apex.

Leaves appearing fascicled when crowded on abbre-

viated shoots, shortly petiolate; blade dull grey-green

to silvery grey, concolorous, cuneate, from about

6-18 mm long, 3-10 mm broad at apex, truncate,

bi-lobed or broadly rounded and coarsely crenately

I. C. VERDOORN

41

lobed at apex, sometimes conduplicate, upper and

lower surfaces from sparsely to densely tomentose

with fringed pitted scales (lepidote-stellate), lateral

nerves usually obscure; petiole 1-9 mm long. Inflores-

cence a raceme of secund, 1-2-flowered cymes terminal

on ultimate branchlets, or cymes 3-5 congested at

apices of short, lateral branchlets; peduncles very

short (1 mm long) and spreading or about 4 mm long

and then strictly erect; bracts 2-3 or more at a node

and sometimes with an entire, elliptic bract among

them, broadly ovate, usually about 3-4 mm long,

2-4, 5 mm broad at base, semi-amplexicaul, some-

times united, cucullate, sometimes with bristle-like

hairs on margins and apex; pedicels 1-5 mm long,

cernuous. Calyx campanulate, about 6 mm long,

lobed to about one third the length, laxly to very

densely lepidote-stellate, the scales and hairs straw-

coloured or some light reddish brown. Petals bright

yellow to dark orange-red, up to about 5-10 mm long,

narrowed at base into a claw, claw glabrous to

densely pubescent along margins, lower portion

with inrolled margins. Stamens with hyaline, oblong-

obovate filaments, a few hairs on shoulders; anthers

3 mm long, pointed, ciliate. Ovary densely lepidote-

stellate (hairs short), shallowly 5-lobbed, shortly

stipitate; stipe 1 mm long; styles about 5 mm long.

Capsule lepidote-stellate, about 7 mm long.

Found along the eastern margins of the winter-

rainfall area and from there eastwards to the Great

Fish River and northwards to Namaqualand, the

southern Orange Free State and Lesotho.

Key to varieties

Flowers secund, about 9 mm long; calyx and leaves, at least on

lower surfaces, densely lepidote-stellate . . (a) var. cuneifolia

Flowers 3 to 5 in a cyme, congested at apices of numerous short,

lateral branchlets about 5,5 mm long, leaves fairly sparsely

to sparsely lepidote-stellate on both surfaces

(b) var. glabrescens

(a) var. cuneifolia. Verdoorn in Bothalia 10,

1:79(1969).

H. cuneifolia Jacq., Hort. Schoenbr. 1, t. 124 (1797); Reichb.,

Ic. Deser. PI. Cults, t. 60 (1822); DC., Prodr. 1:495 (1824);

Harv. in F.C. 1: 189 (1860), partly, excluding syn. H. multiflora

and specimen cited. Type: Cape, cult. Hort. Schoenbr., two

flowering twigs on left hand of sheet with the name, in Jacquin's

handwriting, on the label. (W, lecto.; PRE, photo.!).

H. pollens Eckl. & Zeyh., Enum. 48, No. 378 (1835). Harv.

in F.C. 1: 190 (1860). Type: Cape, Uitenhage between Coega

and Sundays Rivers, Ecklon & Zeyher Enum. No. 378 (K;

PRE, photo.! PRE!; S!; NBG!).

H. secundiflora Eckl. & Zeyh., Enum. No. 380 (1835).

Type: Cape, Caledon, Potrivier, Langehoogde et Bontjes Kraal,

Ecklon & Zeyher Enum. No. 380 (PRE!; S!).

H. lepidota Buch. ex Krauss, Flora 27:294 (1844), nom.

nud. based on Krauss s.n. from Winterhoek (W!; PRE, photo.!

LE).

H. membraniflora Schltr. in Bot. Jb. 55: 359 (1919). Type:

Cape, Vanrhynsdorp: Pappelfontein, Onderbokkeveld, Schlech-

ter 10911 (K; PRE, photo.!; PRE! BOL!; S!; LE!).

The typical variety is a more robust plant and is

reported to be more palatable than var. glabrescens ;

flowers are larger and secund on lateral branchlets

and petals usually densely pubescent on margins of

claw. The dense lepidote-stellate pubescence gives

branches and leaves a silvery grey appearance espe-

cially in eastern and northern regions. Under the

microscope this pubsecence is seen to be composed

of pitted scales fringed with hairs.

Found in marginal Fynbos from Bellville in the

south-western Cape northwards to Namaqualand

and eastwards through the Broken Veld and Arid

Karoo to the Alexandria and Middelburg Districts.

Cape.— Alexandria: Addo National Park, Archibald 3751

Beaufort West: Oukloof Pass, Acocks 14332. Bellville: near

Durbanville, Taylor 4963. Bredasdorp: Napier, Compton 8989

(NBG); Klipdale Station, Smith 3196. Caledon: Potrivier,

Langehoogde et Bontjeskraal, Ecklon & Zeyher Enum. No!

380 (S). Calvinia: Calvinia, Acocks 14442; Schmidt 49: 164,

312. Carnarvon: Carnarvon, Leistner 482. Ceres: Ezelsfontein’,

Esterhuysen 20361. Clanwilliam: Clanwilliam, Acocks 19667;

Olifants River near Rondegat, Schlechter 5045. Cradock:

Rayner’s Koppie, Acocks 11922; Dwingfontein, Story 1319;

Baroda, Dyer 1037. De Aar: De Kalk, Henrici 4522. Fraserburg:

Haf strom & Acocks 942. Graaff-Reinet: Klipfontein, Codd

3530. Ladismith: Acocks 14608. Laingsburg: Whitehill, Compton

11202 (NBG). Middelburg: Conway Farm, Gilfillan sub Galpin

5504; 5505. Mossel Bay: Sidey 1742; Acocks 15396. Nama-

qualand: West of Kamieskroon, Acocks 16454. Paarl: Paarl-

berg. Bos 368 A. Richmond: Elandskuilen, Acocks 8732.

Riversdale: east of town, Acocks 21377; Onverwacht, Muir

1608. Robertson: Robertson, Van Breda & Joubert 1921.

Somerset East: Aansville, Long 747; Sheldon, Mrs Hutton

12399; Kommadagga, Bayliss 1619. Stellenbosch: Klapmuts,

Esterhuysen s.n.; Taylor 5034. Steytlerville : Barroe, Story

2307. Swellendam: Swellendam, Martin 249 (NBG); near

Heidelberg, Barker 5567 (NBG; BOL). Uitenhage: Winterhoek,

Burchel! s.n. (LE; W): between Coego & Sundays River,

Ecklon & Zeyher Enum. No. 378 (type gathering of H. pollens).

Vanrynsdorp: Heerelogement, Zeyher s.n. (S); Pappelfontein,

Schlechter 1091 1 ; north of town, Acocks 14744. Victoria West:

Rehm s.n. Willowmore: Theron 1369. Worcester: Mowers

Siding, Acocks 14541.

In the past this species has gone under the name

H. pallens Eckl. & Zeyh., mainly owing to Harvey’s

(1860) misidentification of H. cuneifolia Jacq. See

notes under H. multiflora for particulars.

Although H. cuneifolia var. cuneifolia is widely

spread Irom west to east, and consequently varies in

some respects, it is clearly definable as a unit. It is

evidently very palatable to browsing animals, for most

specimens show signs of having been eaten down

continuously. The flowers are said to be sweetly

scented. Acocks reports that this variety has advanced

and still is advancing eastwards and northwards into

the ecological vacuum caused by the withdrawal of

the grassveld.

This variety is close to H. pfeilii K. Schum., and

H. Bolus t>o81 and Acocks 14994 from Namaqualand

appear to be intermediates, with the pubescence on

the leaves like that of H. pfeilii but the calyx more

like that of H. cuneifolia var. cuneifolia.

H. triphylla sensu Eckl. & Zeyh. Enum. No. 379

(as represented in SAM) is H. cuneifolia.

(b) var. glabrescens {Harv.) Verdoorn in Bothalia

10, 1: 79 (1969). Type: Without locality, Drege s.n.

(K, lecto.!; PRE, photo.!; W!).

H. pallens Eckl. & Zeyh. var. glabrescens Harv. in F.C.

1: 190 (1860), partly as to Drege specimen cited, excluding

Mrs Barber, Albany.

This variety differs from the typical in that the

shrublets are generally 30-60 cm tall, more repeatedly

branched, the branchlets shorter, rigid and early

glabrescent; the leaves are more sparsely lepidote-

stellate, the inflorescence is congested at the apices of

numerous, short, lateral twigs, usually with only 3 to 5

flowers in each cyme; the flowers are smaller, just over

5 mm long (instead of 8-10 mm long), the calyx more

narrowly campanulate and slightly narrowed at the

throat, and the petals glabrous or nearly so and

cuneate into the claw instead of distinctly ciliate to

densely pubescent along the margins and abruptly

narrowed into the claw.

Found in stony ground, mainly in the transitional

zone between Karoo and grassland, which stretches

roughly from the vicinity of Beaufort West north-

eastwards in a widening band through Middelburg to

42

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Aliwal North and through Herbert District into the

Orange Free State, reaching Lesotho in the east.

Cape. — Aliwal North: Aliwal North, Thode A 1840. Beaufort

West: Nieuwveld Mts, Esterhuysen 2748; Marloth 2138;

Sunnyside, Esterhuysen 5056. Colesberg: Colesberg, Botha in

BLFU 7547. Cradock: Cradock. Brynard 43; Modderfontein,

Acocks 12811. Herbert: Thornhill, Leistner 1422. Middelburg:

Middelburg, Horn s.n.; Conway Farm, Gilfillan sub Galpin

2955; 5507. Murraysburg: Murraysburg, Van Heerden 1;

Tyson 271 (SAM). Richmond: Leopard's Vlei, Bolus 13852.

O.F.S. — Bloemfontein: Bloemfontein, Gemmell in BLFU

6440; Thode A521; De Winter 8969; Grants Hill, Wasserfall

219 (BOL). Fauresmith: Fauresmith, Smith 413; 4541; Henrici

1815; 1864; Verdoorn 1140. Philippolis: Philippolis, Smith

4485; 4497. Rouxville: Ecklon & Zeyher loc. 1 14 in Linnaea 19.

Thaba Nchu: Thaba Nchu, Roberts 2666. Zastron: Zastron,

Maree 1.

Lesotho. — Leribe: Dieterlen 755.

According to “Zwei Documente” the Drege speci-

men of H. multiflora , on which Harvey based his

H. pallens var. glabrescens, was collected in the

Hexrivierbergen, Bokkeveld, that is Ceres District,

slightly outside the known distribution but it could

possibly occur there. The specimens agree with those

collected further north.

In Flora Capensis 1: 190 Harvey suggested that

var. glabrescens was possibly H. multiflora Jacq. This

has proved to be incorrect (see notes under H. multi-

flora Jacq.).

Some specimens from outside the known distribu-

tion area appear to be intermediates or hybrids

between these varieties.

According to reports, the leaves when squashed

form a sticky mass which is considered to have a

healing effect when applied as a plaster to sores. This

accounts for the common name “Geel pleisterbossie”,

generally used in the Orange Free State. The prefix

“geel” (meaning yellow) is misleading because the

flowers may be bright yellow on some plants or

orange-red on others.

Burke & Zeyher 122 in S is typical H. cuneifolia,

but it could hardly have been collected on the Orange

River.

63. Hermannia desertorum Eckl. & Zeyh., Enum.

48 (1834); Harv. in F.C. 1: 189 (1860). Type: Cape,

near Graaff-Reinet, Ecklon & Zeyher Enum. No. 382

(K; PRE, photo.!; S!; SAM!).

Suffrutex, usually heavily browsed, if protected,

about 60 cm tall with slender, divaricate branchlets

becoming indurated, branches chestnut-brown, drying

dark brown or with slight metallic sheen, young

growth with a fine, greyish, sericeous tomentum

made up of microscopic fringed scales and stellate

hairs. Stipules subulate from a broad base or ovate-

cuspidate, 2-3 mm long, covered with a thin grey

tomentum. Leaves shortly petiolate; limb suborbi-

cular, broadly cuneate into the petiole, from 5x5

mm to 8 x 7 mm, plicate at first and covered by a thin

greyish rather sericeous tomentum, becoming glabres-

cent on upper surface, older leaves sometimes stellate-

pubescent on both sides, broadly rounded at apex,

deeply and broadly crenate except at base; petiole

2-5 mm long, pubescence as on leaf undersurface.

Inflorescence terminal, in few-flowered, racemose

cymes, (sometimes 1 -flowered and secund-hybrids?);

peduncles and pedicels varying from very short,

about 2 mm long, to 5 mm long, pubescence as on

young branchlets; bracts 1,5-2, 5 mm long, rather

thick, broadly ovate to subulate from a broad base.

Calyx thin-textured in dried specimens, subinflated,

5-gibbose, about 5 mm long, wide at mouth, lobed to

near middle, finely pubescent with minute fringed

scales. Petals yellow, minutely papillose within, upper

half more or less oblong-orbicular, abruptly narrowed

into a waist and produced below in a claw with

margins inrolled, claw sometimes strongly arcuate,

appears glabrous, but a few microscopic hairs present.

Stamens 5mm long, with obovate-oblong, hyaline

filaments which appear to be glabrous but have

microscopic hairs on shoulders, united at base;

anthers ciliate, overlapping filaments at base. Ovary

broad, 5-lobed, very shortly stipitate, finely stellate-

tomentose. Capsule about 4x4,5 mm, stellate-tomen-

tose, shallowly 5-lobed.

Found in karroid broken veld, on the jasper hills

and in pan-like depressions with compact calcareous

sand. Recorded from Laingsberg and Prince Albert

in the south and northwards through Vanrhynsdorp

and Prieska to Witpiitz in South West Africa.

Cape. — Beaufort West: N.W. of Bulwater, Acocks 14140.

Calvinia: N. of Calvinia, Municipal Reserve, De Winter &

Verdoorn 9009. Carnarvon: W. of town, De Winter & Verdoorn

9005. Clanwilliam: Clanwilliam, Boucher 3106. Graaff-Reinet:

near town, Ecklon & Zeyher Enum. No. 382 (SAM). Hay:

Floradale, Esterhuysen 2321; Lanyon Vale, Acocks 1962.

Laingsburg: Zoutkloof, Compton 6434 (NBG). Postmasburg:

near Olifantshoek, Leistner & Joynt 2781. Prieska: Prieska,

Bryant 3383. Prince Albert: Zwartberg Pass area, Stokoe 9036.

Vanrhynsdorp: Komkas, Acocks 23301.

S.W.A. — Liideritz South: Witpiit, Merxmuller & Giess

28770.

Characterized by the fine, pale, stellate pubescence,

the semi-inflated, shallow calyx, 5-gibbose at the base,

the indurating twigs and suborbicular leaves, which

are cuneate into a petiole and crenate-dentate.

Seed of De Winter & Verdoorn 9009 germinated in

Pretoria. One plant survived and grew to 60 cm tall,

flowered profusely during June and July 1970 and was

visited by swarms of bees.

Specimens which appear to be a form of this

species have larger, glabrescent leaves and they occur

west of the escarpment but not in regions with an

annual precipitation below 100 mm and thus not near

the coast. The following appear to belong to this

form or are possibly hybrids nearest H. desertorum :

S.W.A. — Liideritz South, Farm Spitskop, Merxmuller &

Giess 28714; Zebrafontein, Merxmuller & Giess 28787.

Cape. — Namaqualand: between Steinkopf and Port Nolloth,

Rosch & Le Roux 1177; S.W. of Kuboos, Lavranos 11030.

Vanrhynsdorp: S. of town, Thompson 2107. Garies: Garies,

Esterhuysen 5309 (BOL).

64. Hermannia pfeilii K. Schum. in Notizbl. bot.

Gart. Mus. Berl. 2: 304 (1899). Type: Cape, between

Port Nolloth and Oakup, Pfeil 34 (Z!).

H. patellicalyx Engl, in Bot. Jb.55: 358 (1919); M. Friedrich

et al. in F.S.W.A. 84: 19 (1969). Type: South West Africa;

Tafelberg near Buntfeldschuh, Schafer 501 sub Marloth 5244

(PRE!).

Suffrutex, much branched, 20-80 cm tall, leaf-scars

prominent; branchlets thinly grey-tomentose (tomen-

tum of fringed, minute scales and stellate hairs),

glabrescent, indurating and spiky but usually not

sharp-pointed. Stipules from semi-obovate to deltoid

or subulate, falling by a clean cut at base or, if

persisting, becoming woody and brittle, 1-3 mm long,

finely stellate-pubescent. Leaves one or more at a

node, very shortly petiolate; blade obovate to obo-

vate-cuneate, often broader at apex than long, 4-15

mm long, 5-12 mm broad at apex, entire or obtusely

3- or more-lobed at apex, concolorous, sometimes

conduplicate, grey-tomentose on both surfaces, tomen-

tum of fringed, minute, pitted scales; petiole 0,5-5

mm long. Inflorescence of few-flowered, racemose

I. C. VERDOORN

43

cymes terminal on ultimate branchlets, flowers

secund; peduncle usually indistinguishable or up to

3 mm long; bracts and bracteoles from ovate to

deltoid or subulate, 1-3 mm long, rather thick;

pedicels 2-6 mm long, thinly tomentose as on branch-

lets. Calyx patelliform, subinflated, 5-angled, from

densely to sparsely finely stellate hairy without,

lobed to almost middle, lobes erect, broadly triangular,

sinuses wide. Petals bright yellow, about 10 mm long,

blade more or less ovate, narrowed abruptly into a

section with broad, incurved margins and below that

into a spur which is usually curved. Stamens with

broad, hyaline, obovate filaments about 3,5 mm long,

apparently glabrous on the shoulders; anthers

ciliate, about 3,5 mm long, overlapping the filaments

for 1 mm. Ovary about 2,5 mm long, obscurely

5-lobed, finely stellate-tomentose, hairs slightly longer

at the apex; stipe 0,5 mm long; styles cohering,

about 4,5 mm long, shortly exserted, shortly and

patently pubescent; stigmas capitate. Capsule about

5 mm long, stellate-pubescent, obtusely 5-lobed,

lobes rounded at apex.

Found along the west coast of Namaqualand

from Kleinzee, just north of Buffels River, northwards

to the Namib as far as the Liideritz area.

Cape. — Namaqualand: near Kleinzee, Verdoom 1871; Port

Nolloth, Galpin & Pearson 7573; 7518; near Alexander Bay,

Werger 507; E. of Oppenheimer Bridge, Orange River Mouth,

Leistner 3437; Grootderm, Lavranos 11013; Anenous Pass,

Rosch 2167.

S.W.A. — Liideritz: Bogenfels, Schafer sub Marloth 5244;

Klinghardsgebirge, Dinter 3882; 4057; Pomona, Dinter 6348;

Haalenberg, Dinter 3804.

This species approaches H. cuneifolia in several

respects, especially in the cuneate, concolorous leaves

with more or less the same pubescence. Microscopical-

ly the tomentum on both sides is made up of fringed,

pitted scales. In H. cuneifolia the scales are larger

than in our species. The most striking difference is

in the calyx; in H. cuneifolia it is campanulate, not at

all inflated, and shortly lobed, while in H. pfeilii it is,

as the synonym H. patellicalyx implies, shaped like

a little dish and subinflated with lobes almost as long

as the disc. H. pfeilii is also close to H. desertorum but

differs mainly in the larger flowers and the leaves

which are more constantly obovate-cuneate, rather

thick in texture and very shallowly, if at all, crenate.

The areas of distribution are adjacent but distinct.

H. pfeilii is restricted to the coastal area.

The scrap of the type gathering of H. pfeilii in Z, a

twig with a small leaf and a loose, normal-sized leaf,

together with the description, leaves one in no doubt

that H. pfeilii is synonymous with H. patellicalyx.

65. Hermannia multiflora Jacq., Hort. Schoenbr.

I, 1. 128 (1797); Reichb., Ic. Descr. PI. Cult, t.69

(1823); DC., Prodr. 1 ; 495 (1824). Type: Cape, Cult.

Hort. Schoenbr., right hand specimen annotated by

Jacquin (W, lecto!; PRE, photo.!).

H. cuneifolia sensu Harv. in F.C. 1 : 189, partly.

Suffrutex, about 30-75 cm tall, bushy, much

branched, outer branches spreading-ascending;

branchlets lepidote-stellate, the new growth densely so.

Stipules with lower lanceolate-subulate, about 2 mm

long, the upper ovate-acuminate, about 3 mm long,

1 mm broad at base. Leaves petiolate, sometimes

appearing fascicled when crowded on abbreviated

lateral shoots; blade discolorous, usually oblong-

cuneate, sometimes suborbicular-cuneate, variable in

size, from about 3 mm long and 1 ,5 mm broad near

apex, to 20 mm long and 1 1 mm broad, rounded at

apex, crenate-dentate in upper portion, dark and finely

lepidote-stellate on upper surface, pale lepidote-stellate-

tomentose beneath, midrib and subflabellate nerves

impressed above and prominent beneath; petiole 1,5-5

mm long. Inflorescence of 1-2-flowered cymes arranged

in fairly lax, terminal, racemose cymes, one to two at

each node of raceme; peduncles straight, erect or

suberect, varying in length, lower short, the terminal

long, from about 5-20 mm long; pedicels usually

shorter and slightly cernuous; bracts and bracteoles

variable, ovate to lanceolate-acuminate, up to 5 at a

node, often with an elliptic, entire, petioled bract up

to 9 mm long and 1 mm broad among them. Flowers

yellow to orange-yellow, sweetly scented, about 8 mm

long (rarely 5 mm long). Calyx broadly campanulate

to subpatelliform, about 5 mm long, lobed to about

midway, fawn to brown-tomentose, the lepidote-

stellate pubescence dense, the scales reddish brown and

hairs straw-coloured resulting in a mousey colour.

Petals 5-10 mm long, suborbicular to oblong-orbi-

cular in upper half, narrowing into a short waist and

produced below into a claw with inrolled margins,

waist and claw densely to sparsely but obviously

stellate-pubescent on margins. Stamens with filaments

cohering at base of stipe; filaments hyaline, obovate-

cuneate, pubescent on shoulders; anthers ciliate,

slightly shorter than filaments which they shortly

overlap at base. Ovary densely lepidote-stellate,

the hairs long and erect at apex, shallowly 5-lobed,

shortly stipitate; styles cohering, erect, about 2,5 mm

long. Capsule densely lepidote-stellate, about 4 mm

long, shallowly 5-lobed, withering calyx and stamens

persisting at base.

Found on rocky slopes, in mountain Fynbos, on

level sandy plateaux and on hills above the Atlantic

coast. Recorded from the Peninsula northwards to

Vanrhynsdorp and Calvinia.

Cape. — Bellville: Vissers Hok, Leighton 1798. Caledon:

Dwarsberg, Stokoe in SAM 61160. Calvinia: Lokenburg,

Leistner 338; Acocks 17052. Clanwilliam: Olifants River

Bergen, Schlechter 5080; Blaauw Berg, Schlechter 8438. Malmes-

bury: Mamre, Van Niekerk 655; Hopefield, Compton 15967

(NBG). Paarl: Klapmuts, Erik Wall s.n. (S); near Wellington,

Grant 2232. Peninsula: Cape Town, Prior s.n.; Marloth 1558;

Hutchinson 167; Claremont, Schlechter 1512; Cape Flats,

Rehmann 2192 (Z); Witsands, Esterhuysen 12960; Simonstown,

Taylor 3708; Muizenberg, Bolus 308 (SAM, BOL). Piketberg:

Elands Kloof, Hafstrom & Acocks 933. Stellenbosch: Stellen-

bosch, Bos 304. Tulbagh: Saron Flats, Wiese 16 (NBG). Van

Rhynsdorp: Giftberg, Leipoldt 3986; N. of Graaf water, Taylor

3941. Worcester: Orchard, Hex River Valley, Esterhuysen

10299.

This species is characterized by the dense, lepidote-

stellate tomentum on the calyx. It was described and

figured by Jacquin in Hortus Schoenbrunnensis.

The excellent painting makes it possible to match

specimens with it and there are numerous represen-

tatives in many herbaria. Unfortunately, through a

misidentification by Harvey, they are named Herman-

nia cuneifolia. In the Flora Capensis, Harvey writes

that he cannot distinguish H. multiflora from H.

cuneifolia by any tangible characters. H. cuneifolia

is figured on 1. 1 24 of Hortus Schoenbrunnensis.

With the adequate material that is available today

these species can be readily distinguished. The

principle diagnostic characters are that in H. multi-

flora the leaves are discolorous with rounded apices

and prominent, subflabellate nerves beneath and the

flowers are disposed in cymose panicles with acuminate

bracts whereas the leaves in H. cuneifolia are the

same colour on both surfaces, with a silvery hue,

truncate or emarginate at the apex and the nerves

seldom prominent. The flowers are secundly arranged

on short, cernuous pedicels and the bracts are broadly

ovate and abruptly acute but not acuminate.

44

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

The areas of distribution of the two species meet,

but do not overlap.

Among the Hermannia specimens in W, a specimen

was found with the name Hermannia multiflora

written on the label in Jacquin’s hand. This has all

the diagnostic features of the figured plant and is thus

selected as the type specimen.

H. multiflora is very variable in habit and crosses

readily, but the dense, mousey or brownish lepidote-

stellate tomentum on the calyx appears to be a

dominant feature, as are the cuneate leaves, which are

whitish tomentose on the lower surface. One of the

three bracteoles, the central one, sometimes elongates

and is narrowly elliptic or lanceolate and acuminate.

A form with small flowers and slender twigs, known

only from Pakhuis Pass, is represented in some

herbaria (e.g. Compton 9620; Esterhuysen 3367 and

Acocks 15036). It has the diagnostic features of

H. multiflora and is here treated as a form or hybrid

of that species. While superficially resembling

H. helicoidea it differs in the pubescence on the calyx,

the type of inflorescence and the stipules.

In some herbaria specimens of H. multiflora

have been named H. triphylla Cav., that is “sensu

Cav.”, as the type of H. triphylla L. is a species of

Lotononis. With regard to H. triphylla sensu Cav.,

Diss. 333, 1. 1 78, fig. 3, De Winter observed that

H. Bolus 308 compares well with Thunberg’s specimen

in UPS named H. triphylla by Cavanilles. H. Bolus

308 in PRE, Z, SAM & BOL is H. multiflora Jacq.

66. Hermannia alnifolia L., Sp. PI. ed. 1 : 674

(1753); Cav., Diss. 1. 1 79 fig. 1 (1788). Curtis’s bot,

Mag. 9: t. 299 (1795); Jacq., Hort. Schoenbr. 3:291

(1798); DC., Prodr. 1:495 (1824); Eckl. & Zeyh.,

Enum. 46:364 (1834); Harv. in F.C. 1: 189 (1860);

Adamson in Adamson & Salter, FI. Cap. Penins.

585 (1950). Type: Cape, ex Hort. Cliff., Linn. Herb.

Cat. No. 854.5 (BM, lecto.; PRE, photo.!).

Suffrutex, varying from prostrate to semi-erect,

occasionally scrambling, usually about 40 cm tall but

varying from 30 cm to 2 m, branchlets stellate-

pubescent, hairs up to 0,5 mm long, stellate hairs

often raised on tubercles (occasional specimens with

long hairs may be hybrids). Stipules narrowly to

broadly ovate, rounded or deeply cordate at base,

sometimes auricled, lobes (auricles) decurrent on

branchlets, usually abruptly narrowed into a cuspidate

or caudate apex, finely stellate-tomentose. Leaves

subrotund or broadly oblong, cuneate at base, rarely

ovate, 3-25 mm long, 2,5-18 mm broad, finely

stellate-tomentose on both surfaces, lower white-

tomentose, nerves on lower surface subplicate,

prominent, margins crenate, sometimes with a bristle

±5 mm long on each rounded lobe; petiole 1-5 mm

long, stellate-pubescent. Inflorescence of racemose or

sometimes paniculate cymes, ultimate cymes 2-3-

flowered; branches stellate-tomentose; pedicels usually

of unequal length, 1-4,5 mm long; bracts on main

branchlets of inflorescence mostly ovate-acuminate,

sometimes with caudate apex, cordate at base, lobes

sometimes decurrent, about 3 mm long (rarely a long,

oblanceolate bract up to 4,5 mm), bracts at base of

cymes smaller and usually deeply 3-lobed, about 1 ,5

mm long. Calyx campanulate, usually with wide

sinuses between subulate teeth, 3-4 mm long, teeth

about 1,5 mm long, texture rather thin, sparsely to

fairly densely stellate-pubescent without, hairs remain-

ing especially along the 5 main veins. Petals yellow to

orange, about 3-6 mm long, more or less oblong,

broadly rounded at apex narrowing towards the

centre, lower third with inrolled margins, abruptly

narrowed into a short 0,5 mm long claw, inrolled

margins obscurely ciliate or with sparse stellate hairs

near margin. Stamens with hyaline filaments, obovate-

oblong, about 2,5 mm long, with a few hairs on

shoulders; anthers about 2 mm long, cells acute at

apex, ciliate in lower half. Ovary 5-lobed, densely

stellate-pubescent, becoming glabrous in part, lobes

rounded at apex; style about 5 mm long, arising from

centre of lobes, minutely and sparsely hairy in part.

Capsule usually enclosed by the calyx and corolla

remains, about 3 , 5 mm long and 4 mm broad, fairly

shallowly 5-lobed apices of lobes rounded, not

much produced above level of style-base. Seeds

reniform or variously compressed, minutely papillose

(scaly?) becoming very obscure, ±1 mm diam.

Occurs on stony hills, low lying loamy places and

Coastal Renosterveld. Recorded from George and

Oudtshoorn in the east westwards to the Peninsula

and then northwards to Clanwilliam, and rare in

Namaqualand (see Salter 1511 in MO.).

Cape. — Bredasdorp: Bredasdorp, Esterhuysen 3073; Kathoek,

Taylor 3783. Caledon: Caledon Story 3071. Ceres: Gydo Pass,

Hafstrom & Acocks 936; near White Bridge, Van Breda 779.

Clanwilliam: Pakhuis, Esterhuysen 3367; near Clanwilliam,

Sclilechter 5067. George: east of Mossel Bay, Acocks 15385.

Heidelberg: Heidelberg, Acocks 21392. Malmesbury: Darling,

Esterhuysen 12974. Montagu: Montagu Baths, Page 132;

Bushmans Vlei, Pica Survey 576. Mossel Bay: between George

and Mossel Bay, Sidey 1728. Paarl: Paarl, Hutchinson 456;

between Paarlsberg and Paardeberg, Drege (S). Peninsula:

Sea Point, Smith 2934; near Cape Town, Marloth 1749; Vissers

Hok, Esterhuysen 9984. Piketberg: Porterville, Wilman 720;

Pikenierskloof, Penther 2089 (S). Riversdale: Soetmelk River,

Acocks 21374; Gouritz Riv., Schlechter 5712. Robertson:

Wansbek, Van Breda & Joubert 1986. Stellenbosch: near Berg

River, Zwartland, Zeyher 1985 (3 specimens); Welgevallen,

Taylor 5020. Swellendam: Swellendam, Acocks 14591, Bontebok

Park, Liebenberg 6449; 6426; Stormsvalley, Drege 1987 (S).

Tulbagh: hills, Marloth 7128. Wellington: Wellington, Thomson

15. Worcester: Bainskloof, Leighton 1987; Stettyn, Van Rens-

burg 180; Stettynsberg, Esterhuysen 15594; Veld Reserve,

Van Breda 24.

Three not clearly defined groups can be recognized,

(1) a fairly robust, suberect to erect form with normal

internodes, (2) a form with more or less the same

habit but with short internodes, and (3) a prostrate

form with flowers rather smaller and petals only

slightly longer than the calyx. All three these forms

are represented in LINN: the typical form by Cat.

No. 854.5, the form with short internodes by 854.6

and the prostrate form with short petals by 854.7.

All forms are characterized by numerous small

flowers, suborbicular cuneate leaves which are

stellate-tomentose on both surfaces, the lower usually

whitish tomentose, and the rather thin calyx which is

usually laxly stellate-pubescent, with the stellate

hairs occurring mainly on the nerves.

Ecklon & Zeyher Enum. No. 363, named H. multi-

flora in S and LE, is H. alnifolia L.

67. Hermannia muricata Eckl. & Zeyh., Enum.

43, No. 343 (1834). Type: Cape, “Zwellendam”,

Ecklon & Zeyher Enum. No. 343 (S!; PRE, photo.!;

SAM!).

Mahernia dryadiphylla Eckl. & Zeyh., Enum. 51, No. 400

(1834). Type Cape, Clanwilliam, Brakfontein, Ecklon & Zeyher

Enum. No. 400 (TCD; PRE, photo.!; S!; PRE!; W!; LE!;

SAM!).

Hermannia dryadiphylla (Eckl. & Zeyh.) Harv. in F.C. 1 : 191

(1860) as "’dryadifolia" .

H. dryadiphylla (Eckl. & Zeyh.) Druce in Rep. bot Soc.

Exch. Cl. Br. Isl., 627 (1917), comb, superfl.

Sufifrutex, much branched from base; branches

rather slender, outer arcuate-ascending usually with

I. C. VERDOORN

45

long (over 0,5 mm), tubercle-based bristles, one or

more from each tubercle, occasionally short, gland-

tipped hairs present as well, sometimes (in more

northerly and south-eastern plants) the tubercle-based

bristles are displaced by a stellate-pubescence and

minute fringed or stellate scales. Stipules narrowly to

broadly ovate, rounded or cordate at base, often

broad, oblique, amplexicaul or decurrent at base,

acuminate to abruptly acuminate at apex. Leaves

petiolate; blade narrowly oblong, 8-36 mm long,

2-10 mm broad, rounded at base, sometimes slightly

narrowing towards rounded base and broadest near

apex, margins sinuate-crenate to sinuate-dentate,

slightly recurved, upper surface sparsely hairy with

long tubercle-based hairs or sometimes subdensely

stellate-pubescent with hairs short, nerves impressed,

lower surface whitish tomentose, but thinly so on the

prominent nerves; petiole 2-10 mm long, rarely up

to 15 mm long. Inflorescence of long, interrupted,

paniculate cymes, terminal on branchlets; peduncles

and pedicels slender, 10-30 mm long, pubescent as on

branchlets; bracts opposite, broadly ovate, cordate,

abruptly acuminate, 3-5 mm long, 1 , 5-3 mm broad

at base; bracteoles up to 3 at base of pedicels, narrowly

ovate, lanceolate with acuminate apex or subulate,

1,5-3, 5 mm long, 0,75-1,5 mm broad. Calyx

flask-shaped, narrow at base, pubescence as on

inflorescence and branchlets, glabrescent, about 5

mm long, lobed to above middle, lobes sparsely

ciliate or densely and softly so in the eastern form,

sinuses fairly wide. Petals “bright orange-yellow or

orange-red”, about 7 mm long, suborbicular in

upper half, narrowed below into a claw with infolded

margins, glabrous in specimens dissected. Stamens

about 6 mm long, cohering at base around stipe;

filaments hyaline, oblong-cuneate, minutely and

sparsely pubescent on shoulders; anthers ciliate,

overlapping the filaments at base. Ovary shallowly

5-lobed, stellate-pubescent on the sinuses, about 2

mm long, stipe 1 mm long; styles cohering, 4 mm long.

Capsule shallowly 5-lobed, stellate-pubescent along

sutures, minutely so between sutures, glabrescent,

faded calyx persisting at base and obscuring stipe.

Found on hills and mountain slopes in river valleys.

Recorded from Namaqualand southwards to the

Clanwilliam and Worcester Districts (especially in

the Hex, Krom and Olifants river valleys) then

eastward to the George, Uniondale and Steytlerville

Districts.

Cape. — Clanwilliam: Brakfontein, Ecklon & Zeyher Enum.

No. 400; Olifants River, L. Bolus 20328 (BOL); S. of Citrusdal,

De Winter & Verdoorn 9056; Schlieben 12440. George: W.N.W.

of Camfer Station, Esterhuysen 16790 (BOL); Zebra, Compton

24420. Laingsburg: Fissantekraal Valley, Compton 21118

(NBG). Namaqualand: Scully 198 (BOL); Skilpad, Barker

8621 (NBG); Kamieskroon, Esterhuysen 23602. Piketberg:

Porterville, Edwards 188 (Z) and in BOL 28060). Steytlerville:

W. of Wilgehof, Oliver 4572. Swellendam: Swellendam, Ecklon

& Zeyher Enum. No. 343 (S). Uniondale: Kromme River

Heights, Fourcade 2698 (BOL); Mannetjiesberg, Oliver 3600.

Worcester: Hex River near De Dooms, H. Bolus 8007; Tweed-

side, Marloth 10802 (BOL); Mauve & /. Oliver 211.

This species is characterized by the narrowly

oblong leaf-blade sometimes narrowing slightly

towards the base, but with the base rounded, not

cuneate into the petiole, the white-tomentose under-

surface and the sinuate-crenate, slightly recurved

margins which form a narrow dark rim around the

undersurface, by the long, lax, terminal paniculate

cymes with long slender peduncles and pedicels, and

the usually broad-based stipules and bracts. A fourth

characteristic, found on the types and other specimens

from the western Cape, namely the tubercle-based,

long, bristle-like hairs borne laxly on most parts of

the plant, seems to be replaced in specimens from

northerly (Namaqualand) and more easterly regions

by a denser stellate pubescence of short hairs (charac-

ter displacement?).

H. muricata is close to the not very well known

species H. repetenda, which is a more erect plant with

usually broader, obovate or broadly oblong-ovate

leaves cuneate in the lower third, not narrowly oblong

and narrowed towards the rounded base as in H.

muricata.

As in other species, a small-flowered twiggy form

of H. muricata occurs occasionally. Examples are:

Khamieskroon, Esterhuysen 23602 (BOL); Pakhuis

Pass, Esterhuysen 21935 (BOL) and Khamiesberg,

Pearson 6159 (SAM).

68. Hermannia repetenda Verdoorn in Bothalia

10: 77 (1969). Type: Cape, Cult. Hort. Hannov.,

Schrader s.n., (GOET., lecto. !; PRE, photo.!).

H. hirsuta Schrad. in Schrad. & Wendl., Sert. Hannov. 10,

t. 4 (1795); Willd., Sp. PI. 3: 594 (1800); Reichb., Ic. Descr. PI.

Cult. t. 59, fig. 2 (1822); DC., Prodr. 1: 495 (1824); F.C. 1 : 190

(1860), excl. syn., non Mill. (1768). Type: as above.

Suffrutex, erect, virgate, laxly branched above,

ultimate branchlets ending in an inflorescence which

often turns slightly aside with a younger branchlet

developing below and overtopping it, sparsely

pubescent with minute stellate hairs, sometimes with

short, gland-tipped hairs intermingled, often also

subsparsely hirsute with long, pointed, patent hairs

about 1-1,5 mm long arising singly or 2 or more

from a bulbous base, apparently caducous or readily

broken off. Stipules acuminate to abruptly acuminate,

±5-12 mm long, base broad to subcordate sometimes

very oblique and subamplexicaul, pubescence as on

branchlets but sometimes denser. Leaves petiolate;

blade fairly narrowly to broadly oblong, oblong-

obovate or oblong-ovate, cuneate in lower third,

15-40 mm long, 7-20 mm broad, crenate except at

cuneate base, 3-nerved at base, nerves impressed

above, prominent beneath, subflabellately branched,

upper surface finely stellate-pubescent and minutely

papillose and usually sparsely to densely pilose with

long, pointed hairs, undersurface usually, at least

when young, whitish tomentose with fine stellate

pubescence; petiole 3-15 mm long, pubescent.

Inflorescence terminal, of lax, paniculate cymes;

bracts 2- or 3-nate, occasionally the third a long,

subulate bract, otherwise like stipules varying from

±5 mm long and acuminate from a fairly broad base,

to 14 mm long and abruptly acuminate from a broad,

oblique subcordate, decurrent base; peduncles and

pedicels with pubescence as on branchlets. Calyx lobed

to above middle, about 6 mm long, sinuses broad,

texture thin, minutely stellate-pubescent, densely so

at base (on cultivated plants?), sometimes long,

bulbous-based hairs and short, gland-tipped hairs

present. Petals about 7 mm long, narrowed about

midway into a claw, blade suborbicular, claw with

inrolled margins obscurely ciliate. Stamens about

4 mm long, with filaments cohering at base, hyaline,

oblong-obovate, 2,5 mm long, about 1 mm broad

near apex, sparsely stellate on shoulders, anthers 2 mm

long, ciliate, overlapping filament at base. Ovary

5-lobed, stellate-pubescentes pecially along thes utures,

rays short, but with longer bristle-like hairs at apex;

stipe 0,5 mm. Capsule about 6 mm long, minutely and

finely stellate-pubescent between sutures, more

coarsely so on sutures.

Apparently extensively cultivated in the gardens of

Europe in the late 18th and early 19th centuries.

46

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Very rarely found in the wild today. Recorded from

the Piketberg, Clanwilliam and Graaff-Reinet Dis-

tricts.

Cape. — Clanwilliam: Wupperthal, Mauve & I. Oliver 93;

N.E. of Pakhuis, Acocks 15050. Graaff-Reinet: Ouberg Pass,

Oliver 5197. Piketberg: Piketberg, Hafstrom & Acoeks 929;

Rest Mtn, Gillet 3731 (BOL); Elandsberg, Pillans 7952 (BOL).

In the past this species was confused with H. aspera

Wendl., a very distinct species differing principally

in the inflorescence which in H. aspera is a fairly

short, paniculate cyme, terminal on the short, lateral

branchlets. In addition, H. aspera has crowded,

sessile leaves strongly crisped on the margins.

In some respects H. repetenda resembles H. multi-

flora but differs in the fine stellate pubescence or

hispid hairs on the calyx in contrast to the dense,

brownish, stellate tomentum on the calyces of

H. multiflorum.

Another related species is H. muricata (see under

that species for distinguishing features).

Although H. repetenda is well represented in

European herbaria, the specimens having been made

from plants cultivated in gardens, especially in the

Herrenhausen garden, Hannover, but very few,

widely distributed present day records exist. This is

apparently one of the many Hermannia species which

is heavily cropped by browsing animals.

H. repetenda is characterized by the inflorescence

which is a long, lax, paniculate cyme, terminal on the

ultimate branchlets but usually turning aside while a

lower branch develops and overtops it.

When the new name for this species was published

in Bothalia 10: 77 the locality “Van Ryns Pass” was

given, based on De Winter & Verdoorn 9018 collected

there. This specimen has since been found to be

H. rigida and, to date, H. repetenda has not been

recorded from that locality.

69. Hermannia rigida Harv. in F.C. 1 : 188 (1860).

Type: Namaqualand, Zeyher 112b (1126 sphalm.)

(S, holo.!; PRE, photo.!).

H. oligantha Salter in J1 S. Afr. Bot. 12: 99 (1946). Type:

Cape, Clanwilliam District, 8 km east of Graafwater, Salter

2750 (BOL, holo.!).

Shrublet, rigid, fairly laxly branched, virgate, up to

about 1 m tall; branchlets pubescent with fringed

scales, hairs short, intermixed with fairly long hairs

grouped on tubercled bases, minute gland-tipped

hairs usually present as well, apical portion of branch-

lets persisting as dry slender twigs. Stipules narrowly

to broadly ovate-acuminate, often broad-based and

abruptly long-acuminate above, tomentose at least

dorsally and occasionally long hairs present as well,

3-8 mm long, 1-3 mm broad near base, on old wood

only prominent scars left. Leaves petiolate, at first

clustered on undeveloped shoots; blade obovate,

oblong-cuneate, ovate-cuneate or suborbicular in

young leaves, 5-30 mm long, 5-12 mm broad, crenu-

late except on cuneate base, somewhat crisped or

plaited, at first tomentose on both surfaces with

fairly long appressed hairs, undersurface whitish,

older leaves subglabrescent, nerves impressed above,

prominent beneath, somewhat flabellate; petiale

2-10 mm long. Inflorescence of 1- to 3-flowered cymes,

more usually 1 -flowered, arranged in lax, terminal

racemes; peduncle aborted or abbreviated; pedicels

fairly stout, 5-9 mm long; bracts ovate-acuminate

(like the stipules), about 5 mm long; bracteoles

subulate to lanceolate-acuminate, 3-7 mm long.

Calyx finely stellate-pubescent especially at base,

sometimes long hairs present as well, 5-7 mm long,

lobed to above middle, sinuses wide. Petals bright

yellow fading reddish, about 10 mm long, oblong

above, narrowed into a claw with infolded margins,

glabrous except for minute stellate pubescence on

margins of infolded lower portion. Stamens with

oblong-obovate, hyaline filaments about 4,5 mm long,

minutely and sparsely stellate on shoulders; anthers

about 2,5 mm long, ciliate, overlapping filaments

for about 1 mm. Ovary about 2,5 mm long, 5-lobed,

densely stellate-pubescent on angles and with long

erect hairs at apices; stipe 0,75 mm long; styles about

4 mm long. Capsule 5-8 mm long with persistent

calyx at base, 5-angled, 5-umbonate, finely stellate-

pubescent to glabrescent between sutures, more

coarsely so on angles and at apex.

Found on rocky ridges, at the base of sandstone

boulders, in red sandy soil, or in shale. Recorded from

the Clanwilliam and Calvinia Districts and from

Namaqualand.

Cape. — Calvinia: upper part of Van Rhyn’s Pass, De Winter

& Verdoorn 9018; Ellis & Schlieben 12464. Clanwilliam:

Nardouw road, Compton 22822 (NBG); Pakhuis Pass, De

Winter & Verdoorn 9044; Oliver in STE 33288. Namaqualand:

Namaqualand, Zeyher 112b (in F.C. 1126 sphalm.); W. slopes

of Kamiesberg, Acocks 19517 (leaves mostly young and sub-

orbicular); S.E. of Kamieskroon, Compton 11106 (NBG); N.

of Kamieskroon, Anegas, Pearson 6275 (SAM).

Characterized by the terminal pseudo-racemes of

1-3-flowered (mostly 1 -flowered) cymes in which the

penduncles are usually aborted or abbreviated and

the pedicels stout and rigid.

Before seeing the type specimen kindly sent on

loan from Herb. S, it was thought that De Winter

& Verdoorn 9018 and 9044 could fall in the range of

H. repetenda, but more material of both these species

has shown them to be distinct. Among other details,

the lax racemose cymes of H. rigida with aborted

peduncles clearly differ from the paniculate cymes of

H. repetenda with long, slender peduncles and pedicels.

Having studied the type specimen of H. rigida, it

was found too that H. oligantha is that species. Of the

five cited specimens the type, Salter 2750 (BOL) is not

very typical being more twiggy with all leaves young

and suborbicular but it falls within the range of

variation of the species and includes the diagnostic

inflorescence. The same applies to Pillans 7093 (BOL).

The other three are typical H. rigida : Schlechter 8384

(fragment in BOL). Pearson 6275 (SAM) and Compton

1106 (NBG).

70. Hermannia helicoidea Verdoorn in Bothalia

11:288 (1974). Type: Cape, Clanwilliam, Hex River

valley, Pillans 9063 (PRE, holo.!; BOL!).

Suffrutex, much branched; branches slender, young

branchlets pubescent with minute, fringed or stellate

scales and occasionally with a few tufted hairs on a

tubercled base, older glabrescent. Stipules broad,

amplexicaul and decurrent at the base, rounded at

apex with a mucro or abruptly attenuate with a

cuspidate or caudate apex, sometimes shallowly

3-lobed, margins obscurely crenate or dentate,

finely and densely stellate on upper surface, hairs

short, whitish stellate-tomentose below, glabrescent,

I, 5-4 mm long, 2-6 mm broad at base. Leaves

petiolate; blade narrowly to broadly oblong-cuneate,

3-10 mm long, rarely longer, 2-6,5 mm broad in

upper half, finely and densely stellate on upper

surface, whitish stellate-tomentose below, margins

somewhat irregularly crenate as well as undulate;

I. C. VERDOORN

47

petiole 1-3 mm long. Inflorescence axillary or terminal

on lateral branchlets, cymes scorpioid, usually curved

in a helicoid fashion, with very slender, sparsely and

minutely stellate-pubescent branches; peduncles up

to 25 mm long; pedicels 1,5 mm long; bracts and

bracteoles like the stipules but smaller, about 2 mm

long, some appearing subulate from a broad base.

Calyx campanulate, finely stellate-pubescent (hairs

very short), about 4 mm long, 5-lobed to above

middle, sinuses wide. Petals about 6,5 mm long, more

or less orbicular in upper half, narrowed below into a

claw with infolded margins. Stamens about 4 mm long

with obovate-oblong, hyaline filaments which are

overlapped by anther bases. Capsule about 4 mm long,

stellate-pubescent, somewhat 5-lobed, calyx persistent

at base.

Found on hills or mountain slopes in river valleys.

Recorded from the Clanwilliam District in the Hex

and Olifants river valleys.

Cape. — Clanwilliam: without precise locality, Stokoe s.n.;

hills in the Hex River valley, Pillans 9063 ; Olifants River valley,

Schlechter 5092; north of Citrusdal, L. E. Taylor 1025 (MO).

Characterized by the monochasial, helicoid develop-

ment of the slender, lateral cymes. In herbaria these

specimens were mostly named H. rigida, which is a

species of taller, more rigid and robust shrublets

with inflorescences of short, 1- to 3-flowered cymes,

more usually 1 -flowered, arranged in lax terminal

racemes, the peduncles usually absorbed and the

pedicels fairly stout and rigid.

H. helicoidea resembles H. muricata superficially,

but differs from it in leaf-shape and pubescence and

in the inflorescence which is both terminal and axillary,

and usually distinctly helicoid.

In H. muricata the leaves are narrowly oblong,

rounded at the base and the margin sinuate-crenate to

sinuate-dentate with the upper surface usually

sparsely hairy with long tubercle-based hairs, rarely

stellate-pubescent.

71. Hermannia aspera Wendl., Bot. Beobacht. 52

(1798). Type: Cult. Hort. Herrenh., type specimen

not traced (see notes).

H. scabra sensu Jacq., Hort. Schoenbr. t. 127 (1797), as to

figure, and excluding syn. H. hirsuta Schrad. (sphalm. Cav.);

Willd., Sp. PI. 3: 594 (1800), partly as to citation Jacq. t. 127

and “(aspera) Wendl. Obs. 52”; sensu DC., Prodr. 1:495

(1824) as to syn. and citation Jacq. t. 127; sensu Reichb.,

Ic. Descr. PI. 1, t. 59, fig. 1 (1822); sensu Eckl. & Zeyh., Enum.

44, No. 350 (1834), non Cav.

Shrub, stem single, erect, 40 cm-2 m tall, more or

less rigid, ultimate branchlets short, fairly stout, very

rough with tubercle-based, tufted bristles (long or

short, 2 to many). Stipules about 1,7-7 mm long,

rounded to broadly cordate at usually oblique base,

decurrent on one side, abruptly narrowed into a

linear-acute to subulate apical portion. Leaves sessile

or subsessile, crowded, 1 or more at a node, different

sizes at a node, size of leaf varies considerably on

different plants too; blade oblong-cuneate, 5-30 mm

long, 1—20 mm broad, rounded, subtruncate, lobed or

emarginate at apex, rarely apiculate, broadly cuneate

at base, incurved and crenate and crisped on margins,

upper surface minutely papillose and scaly stellate-

pubescent or with tubercle-based hairs, lower surface

finely white stellate-tomentose between the veins and

sparsely stellate on veins; midrib and about 3 suberect

lateral veins prominent beneath; petiole 0-4 mm long.

Inflorescence terminal on ultimate branchlets of

simple or paniculate cymes, short, up to 30 mm long,

3 to 15-flowered; bracts like stipules but smaller,

2, 3 or 4 at a node; pedicels 4-8 mm long. Calyx

submembranous with strong nerves, 5-6 mm long,

lobed almost to middle, glabrescent except along

nerves which are sparsely pubescent with bulbous-

based hairs, lobes sometimes rather densely ciliate

with soft hairs and hairy on inner face. Petals yellow

turning red, 6-9 mm long, 4 mm broad in upper half,

narrowed just below middle into a claw, margins of

claw fairly broadly infolded. Stamens cohering or

joined at base, free portion about 5 mm long; filaments

hyaline, oblong-cuneate, obscurely hairy on shoulders;

anthers pointed, ciliate. Ovary stellate-pubescent,

shortly stipitate; styles about 3,5 mm long. Capsule

small, 4-5 mm long, subglobose, densely stellate-

pubescent (hairs longer towards apex), 5-umbonate.

Found in arid Fynbos, Karroid scrub and Renoster-

veld in the south-western Cape. Recorded from

Worcester northwards to Namaqualand (Kamiesberg)

and through Swellendam and Montagu to Ladismith

in the east.

Cape. — Calvinia: Lokenberg, Acocks 17025; Botterkloof

Pass, Middlemost 1609 (NBG). Ceres: On road between Koue

Bokkeveld and Cedarberg, N. of farm Excelsior, Taylor 5900;

N. of Pienaarskloof, Acocks 23689; Clanwilliam: Pakhuis Pass,

De Winter & Verdoorn 9041; Pillans 7089 (BOL). Ladismith:

Anysberg, Esterhuysen 17075; 25975 (BOL). Montagu: Kiesies-

berg, Lewis in SAM 59090 (SAM); Baden, Pillans 7924. Nama-

qualand: Kamiesberg, Coppejans 1108 (STE). Giftberg, Phillips

7381 (SAM). Swellendam: Anysberg, Stokoe 8222 (NBG &

BOL). Vanrhynsdorp: Bokkeveld Mts, Marloth 7580; Giftberg

Pass, Acocks 14878. Worcester: Hex River, H. Bolus 11864.

This species is plentiful over the area of its distri-

bution and is well represented in most herbaria, due,

probably to its roughness and woodiness which renders

it unpalatable to browsing animals. It is distinguished

from others with somewhat similar leaves (having the

lower surface finely white stellate-tomentose) in that

the leaves are sessile or subsessile and crowded.

Judging from specimens in the Leningrad and Vienna

herbaria this species flourished for several years, at

the turn of the 18th Century, in the botanic gardens

of Europe. To date it has been wrongly identified in

all herbaria as either H. scabra or H. hirsuta. Jacquin

in Vienna published an excellent painting of this

characteristic plant, in his Hortus Schoenbrunnensis,

under the name H. scabra Cav., a wrong identification,

for H. scabra Cav. is a distinct species (see note under

that species). This mistake was subsequently recog-

nized by several botanists, who then referred to it by

the homonym H. scabra Jacq. A specimen of our

species in the Leningrad Herbarium, of which there is

a photograph in Pretoria (PRE Neg. No. 4576), is

labelled Hermannia scabra Jacq. and in the same

handwriting H. aspera Wendl. is given as a synonym.

This led to the investigation of Wendland’s name

which was found to be the correct one for this species

with sessile leaves. Unfortunately no type specimen has

been traced to date, but the Leningrad specimen

mentioned above was collected and labelled in the

gardens of Gottingen at a time when Wendland was

associated with that University and so may be looked

upon as an authentic representative of his species.

The use of the name H. hirsuta stems from the

mistaken conclusion that Wendland’s species,

H. aspera, was the same as that described by Schrader

(1795) as H. hirsuta, now named H. repetenda (see

notes under H. repetenda). This species differs from

H. aspera, among other things, in the longer and

laxer inflorescence and the petioled leaves. It is not

so densely leafy nor as rigidly woody and roughly

pubescent and therefore more palatable. This may

48

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

account for its scarcity in comparison with H. aspera,

in the veld as well as in herbaria.

Certain specimens collected in the Montagu

District, where H. disticha occurs, appear to be a

form or hybrid of H. aspera which is also recorded in

parts of the district. The specimens are De Winter

& Verdoorn 9088, Compton 18359 (NBG) and 18416

(NBG), Lewis in SAM 59089 (SAM), Walgate in BOL

23457 (BOL) and Walgate in BOL 23458 (BOL). All

but the last do not have persistent white tomentum on

the lower surface and so resemble H. disticha , and in

all specimens the margin of the leaf is not so markedly

recurved as in more typical H. aspera. All these are

robust luxuriant plants and this suggests that they

might be hybrids between H. aspera and H. disticha

but nearest to H. aspera.

72. Hermannia conglomerata Eckl. & Zeyh.,

Enum. 44, No. 352 (1834); Harv. in F.C. 1:194

(1860); Saund. Rep. Bot. 4: t. 217 (1871). Type: Cape,

Uitenhage, “Sondags et Zwartkops Rivier”, Ecklon

& Zeyher, Enum. No. 352 (K; PRE, photo.!; PRE!;

SAM!; LE!).

H. glomerata E. Mey. in Drege, Zwei Pfl. Doc. 191 (1843),

nomen.

Suffrutex, suberect, with ascending branches,

10-60 cm tall, branchlets pubescent with long tubercle-

based, few-rayed stellate hairs, with age hairs fall

off and branchlets are rough with scattered tubercles.

Stipules narrowly to broadly ovate-acuminate, cordate

at base, about 8 mm long, with long, few-rayed

stellate hairs. Leaves petiolate; blade subrotund, 1-2

cm diam., pubescent with long, few-rayed, stellate

hairs, sparse and 1- to 2-rayed on upper surface,

several-rayed and more copious on lower surface,

margin crenate; petiole about 5 mm long. Inflorescence

of sessile cymes crowded at ends of branches and

branchlets, interspersed with leaves and bracts;

bracts ovate-, or broadly lanceolate-acuminate, about

9 mm long and 3,5 mm broad near base; bracteoles

linear-lanceolate, about 7 mm long and 1 mm broad

near base. Calyx 6 mm long, 5-lobed to about midway,

densely hispid, at least when young, with few-rayed,

long, tubercle-based hairs. Petals about as long as

calyx, strongly contorted and shortly exserted from

the glabrous, more or less oblong calyx tube, obovate,

rounded at apex, narrowing slightly at centre, lower

half with margins narrowly inrolled and narrowing at

base to a short claw. Stamens about 4 mm long;

filaments hyaline, ± linear, narrowing towards base,

glabrous in the specimens dissected, overlapped by

anther bases; anthers shortly ciliate. Ovary about 2

mm long, 5-lobed, hirsute with long erect hairs.

Among karroid scrub on hillsides in dry river

valleys. Recorded from Alexandria, Albany, Uiten-

hage and Riversdale (Fish River, Sundays River and

Wyders River).

Cape. — Albany: N.W. of Grahamstown, Fish River valley,

Acocks 18397; Schlechter 6113. Alexandria: Addo Bush near

Kenkelbosch, Story 2729. Port Elizabeth: Cape Recife, Zeyher

1990. Riversdale: Wyders River, Muir 978. Uitenhage: Sundays

River, “Sondags and Zwartkoprivier”, Ecklon & Zeyher

Enum. 352.

A sheet with five specimens of this species in the

Bentham Herbarium was photographed by De

Winter. The specimens are labelled as follows: (1 & 2)

Drege 1838, no locality, (3) Schlechter 6113, Fish

River (4) Cooper 1995, western districts and (5)

Cooper 2012, Natal. There is probably some mistake

about the last mentioned locality. No Natal record

exists. Cooper 2012 with the same label is in Z.

The species is characterized by the long spreading

hairs, especially conspicuous on the young flowers

and the glomerate inflorescence.

Drege 7268 and 7292, cited in Zwei Documente

from Zuurberg as A. glomerata , are this species.

73. Hermannia micrantha Adamson in J 1 S. Afr.

Bot. 10: 123 (1944); FI. Cap. Penins. 585 (1950).

Type: Cape, Little Lion’s Head, Salter sub Adamson

1901 (BOL!).

H. intricata Adamson in J1 S. Afr. Bot. 7: 198 (1941), non

Engl. (1919). Type: as above.

Suffrutex small, usually less than 30 cm tall, much

branched; branches slender, divaricate, roughly

stellate-pubescent with rays often over 0,5 mm long,

and with short, gland-tipped hairs intermingled.

Stipules usually in pairs, ovate, acute, cordate or

rounded at base, oblique, one side deeply cordate,

2-3 mm long, 1-1,5 mm broad, coarsely stellate-

pubescent, scaly and occasionally with gland-tipped

hairs intermingled. Leaves petiolate; blade light green

on both surfaces, not whitish beneath, obovate-

oblong, narrowing slightly at base, about 6-10 mm

long, 3-8 mm broad, rarely larger, stellate-pubescent

and scaly on both surfaces, and sometimes with

minute gland-tipped hairs intermingled, margins

crenate; petiole 2-5 mm long. Inflorescence of 1- to

2-flowered cymes, terminal or laxly racemose on

slender ultimate twigs; peduncles vary in length,

often 1 cm long; pedicels 1-2,5 mm long, stellate-

pubescent and with minute gland-tipped hairs;

bracts like stipules ovate-cordate, those at base of

ultimate cyme usually smaller and 3-lobed. Calyx

about 3 mm long, 5-lobed to the centre or slightly

beyond, stellate-pubescent without, and with minute

gland-tipped hairs intermixed, teeth about 1,5 mm

long, acute. Petals lemon-yellow, about 5 mm long,

more or less oblong, rounded at apex, narrowing

slightly to middle with lower portion with inrolled

margins, hardly clawed at base. Stamens 3 , 5 mm long,

with broad, obovate-oblong filaments with a few

hairs on shoulders; anthers 1,5 mm long, cells acute

at apex, ciliate on margins. Ovary and style about 3

mm long; ovary stellate-pubescent with long bristles

at apex; styles with a few short, sparse hairs. Capsule

about 4 mm long and 4 mm broad, deeply 5-lobed,

lobes keeled and apices projecting above by about

one third, stellate-pubescent, especially on keels,

with some rays over 0,5 mm long.

Found on dry southern and western slopes of

Table Mountain, on or in the vicinity of Little Lion’s

Head.

Cape. — Peninsula: Little Lion’s Head, Salter sub Adamson

1901 (BOL); Adamson 2618; Table Mt., foot of cliffs at west

end, Esterhuysen 19251 ; lower west slopes, Pillans 4264 (BOL).

This species is close to H. alnifolia but is dis-

tinguished by the leaves being green on both sides,

obovate and broadly cuneate at the base and not

suborbicular, the stellate pubescence coarser and

interspersed with minute gland-tipped hairs. The

capsule is more deeply 5-lobed with the apices longer

(only different in degree).

In the Bolus Herbarium, the type specimen has

the number 7417 added in pencil after Salter’s name.

74. Hermannia disticha Schrad. in Schrad. &

Wendl., Sert. Hannov. 26, t. 16 (1795-97); Reichb.

Ic. Descr. PI. Cult., t. 60 (1882); Harv. in F.C.

1: 188 (1860). Type: In herb Vahl, missit Schrader

(C, lecto. ; PRE, photo.!).

I. C. VERDOORN

49

H. rotundifolia Jacq., Hort. Schoenbr. 63, t. 118 (1797).

Shrublet, erect, branched, 50-90 cm high; branch-

lets with long, tubercle-based, spreading hairs.

Stipules deltoid to broadly ovate, acuminate, usually

with a broad, cordate base, which is somewhat

decurrent on the branch or stem, with single or few-

rayed bulbous-based hairs, especially on dentate

margin. Leaves distichously arranged, shortly petio-

late; blade broadly ovate-oblong to obovate-oblong,

10x8-20x16 mm or sometimes longer, broadly

cuneate or rounded at base, upper surface with long,

bulbous-based hairs, which are single or few-rayed,

lower surface with similar hairs but many-rayed or

stellate from a bulbous base, at first densely so giving

a whitish appearance to undersurface; margin

distinctly crenate-dentate; petiole very short or 2-5

mm long. Inflorescence a few-flowered, fairly short,

racemose or paniculate cyme terminal on ultimate

branchlets; pedicels short or up to 5 mm long,

pubescence as on branchlets; bracts lanceolate-

acuminate to ovate-acuminate up to 6 mm long, the

upper small and lobed, sparsely pubescent with long,

bulbous-based hairs. Calyx 5-lobed to about middle,

about 5 mm long, sparsely pubescent with long

stellate hairs from a tubercled or scale-like base.

Petals more or less oblong, about 7 mm long, abruptly

narrowed about midway, upper half suborbicular,

lower half with broadly infolded margins, obscurely

ciliate and pubescent, narrowing at base into a claw

about 1 , 5 mm long. Stamens with hyaline filaments

which are oblong, slightly narrowing towards base,

about 3 mm long, with a few hairs on shoulders;

anthers about 3 mm long, cells pointed, ciliate along

sutures. Ovary somewhat 5-angled, pubescent (hairs

erect and long at apex), very shortly stipitate, stipe

about 0,5 mm long; styles minutely pubescent,

about 3,5 mm long. Capsule enclosed in the ±

persistent calyx and corolla, about 2,5x2, 5 mm,

5-lobed, 5-umbonate at apex, stellate-pubescent,

hairs long in upper half and at apex. Seed reniform,

suborbicular, tubercled, ±1 mm diam.

Recorded to date only from the rocky kloofs in the

mountains of Montagu.

Cape. — Montagu: Kloof near Montagu Baths, Page 25;

De Winter & Verdoorn 9083; Kogmanskloof, Acocks 20347;

Kleurkloof, Compton 5749 (NBG).

Characterized by the pubescence of long, patent,

bulbous-based hairs, the broad, distichous leaves

which are not persistently whitish pubescent on the

lower surface and the short, ±4 cm long inflores-

cences.

On the labels in several herbaria the authors are

given as “Schrad. & Wendl.”, but in the original

description and in the Index Kewensis Wendland’s

name does not appear.

75. Hermannia decipiens E. Mey. ex Harv. in

F.C. 1:192 (1860); Phillips in Ann. S. Afr. Mus.

9: 337 (1917); E. Mey. in Drege, Zwei Pfl. Doc. 191

(1843), nomen. Syntypes: Cape, Zwarteberge, Klaar-

stroom Drege s.n. (S!; W!; Z! ; LE !) ; Vrolyk, Drege

1211 (LE!).

Suffrutex up to 1 , 30 mm tall, branches pubescent

with tufted or stellate hairs from a scaly base or on

small raised tubercles, hairs long or sometimes short

and dense. Stipules deciduous, lanceolate-acuminate

to subulate, 4-6 mm long, up to 2 mm broad at base,

stellate-pubescent, hairs long or short. Leaves sessile

or subsessile, sometimes erect, imbricating, oblong,

oblong-elliptic or oblong cuneate, 7-18 mm long, 4—7

mm broad, apex acute, rounded and mucronate or

truncate and lobed, densely but roughly stellate on

both surfaces, sometimes thickly tomentose, hairs

long or short, mostly yellowish; petiole 0-1 mm long.

Inflorescence a short, few-flowered, simple or branched

cyme, terminal or in axils of upper leaves of branch-

lets; pedicels 2-5 mm long; bracts subulate, 2-7 mm

long, seldom reaching more than basal third of calyx.

Calyx not inflated, campanulate, densely and coarsely

pubescent with stellate and tufted hairs, hairs golden

yellow, tube 4-6 mm long, lobes about 2 mm long,

deltoid, acute. Petals 7-8 mm long, oblong in upper

half, narrowed abruptly into a claw with inrolled

margins, dorsally conspicuously stellate from lower

portion of upper half almost to base of claw. Stamens

about 4 mm long with broad hyaline filaments about

3 mm long; anthers ciliate, about 2 mm long. Ovary

2 mm long, 1,5 mm broad, 5-lobed, stellate-pube-

scent; styles 4 mm long, hairy at base; stigmasca pitate.

Capsule enclosed by perianth, shortly exserted from

calyx, about 5 mm long, subglobose, 5-lobed, stellate-

pubescent; stipe short, up to 1 mm long.

Found in mountainous country from the Ladismith

area along the Swartberg eastwards to Willowmore.

Cape. — Ladismith: Seven Weeks Poort, Marloth 2978; Phillips

1400. Ladismith-Laingsburg: “Zwartberge, Klaarstroom, Drege

s.n. Vrolyk, Drege 7277 (LE) Vleiland, Acocks 20497. Willow-

more: Swanepoelspoortberg, Marloth 4127.

This species is closely related to H. involucrata

having the same type of pubescence and leaf-shape

and the same non-inflated calyx which is densely

stellate-pubescent with yellow hairs. It differs from

that species in the smaller leaves and flowers and the

leaves being more erect and fairly closely apressed

to the stem.

76. Hermannia involucrata Cav., Diss. 328, t.

177, fig. 1. (1788). Type: Cape, Thunberg s.n. in

Herb. Cav., (MA, holo.; PRE, photo.!). (The speci-

men so named in the Thunberg herbarium (UPS,

herb. No. 15484), does not agree, having an inflated

calyx, but the one named “//. salvifolia ” ( Thunberg

s.n. in UPS, herb no. 15492) appears to be this species.

H. chrysophylla Eckl. & Zeyh., Enum. 44, No. 346 (1834),

pro parte, as to some specimens with non-inflated calices but

labelled with Enum. No. 346 (PRE!; LE!; W!).

H. salviifolia sensu Cav., Diss. 328, t. 180, fig. 2, non Linn. f.

Suffrutex 60 cm to 1,3 m high, erect or straggling,

branchlets with shiny, tufted hairs from a sessile,

scaly base or a slightly raised tubercle, hairs long or

short and give the new growth, leaves and calyx a

golden hue. Stipules narrowly lanceolate, acuminate,

subulate or linear-acute, 2,5-10 mm long, base 0,5-

1,5 mm broad, roughly and sublaxly to densely

stellate, incurved in upper portion. Leaves subsessile

to shortly petiolate; blade strikingly variable in shape

from broadly elliptic to elliptic or oblong-obovate,

10-37 mm long, 6-14 mm broad, acute and mucronate

to broadly rounded and mucronate or broadest and

toothed at subtruncate apex, roughly and subdensely

or densely stellate-pubescent on both surfaces, hairs

stiff and golden yellow, scattered among groups of

hairs are minute, dark pustules; petiole 1-5 mm long.

Inflorescence of short, few-flowered, branched cymes

in axils of upper leaves, up to 2,5 cm long; peduncles

short, up to 5 mm long, roughly stellate; bracts at

base of pedicels subulate, up to 10 mm long, often

reaching to top of calyx, roughly stellate, arcuate in

upper portion; pedicels 1-2 mm long, roughly stellate.

Calyx subcoriaceous with strong ribs, densely pube-

scent with stellate or tufted hairs from a scaly base,

hairs shiny and giving calyx a golden hue, about 8

mm long, lobed in upper third or almost to middle,

50

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

tube cup-shaped, lobes cuspidate from a deltoid base.

Petals yellow, about 10 mm long, oblong in upper

half, narrowed into a claw with inrolled margins,

apex rounded to subtruncate, recurved, dorsally

finely stellate-pubescent in lower two-thirds. Stamens

about 4-6 mm long with hyaline, obovate-oblong

filaments; anthers about 2,5 mm long, ciliate. Ovary

5-angled, stellate at least on sutures; stipe 0,5-1 mm

long; styles about 4 mm long. Capsule about 7 mm

long, half exserted from persistent calyx, finely

stellate-pubescent between sutures which are coarsely

stellate-pubescent.

Found in gorges, kloofs and on mountain slopes in

false Fynbos. Recorded from Ladismith, Uniondale

and Willowmore eastwards to Uitenhage.

Cape. — Humansdorp: Combrink, Acocks 13698. Steytler-

ville: Waaipoort, Acocks 16014. Uitenhage: Kantiens River,

Oliver 1360; Melkhoutboom, Long 1387; “Langekloof &

Winterhoek” Ecklon & Zeyher Enum. No. 346 (PRE; SAM;

LE; W). Uniondale: Prince Alfred’s Pass, Acocks 21116.

Willowmore: Baviaanskloof, Oliver 4563.

Differs from the closely related and variable species

H. salviifolia in that the calyx is not diaphonous nor

urceolate. Characterized by the tufted, shiny hairs

which give the vegetative parts a golden yellow hue,

the subcoriaceous, strongly ribbed, thickly stellate-

pubescent calyx and the petals which are dorsally,

finely stellate-pubescent. Among the tufted hairs on

the leaves and calyx are scattered minute pustules or

glands that give the appearance of minute holes. This

feature is shown very clearly as black dots on the plate

in Cav., Diss. The leaves vary considerably in shape,

occurring in at least two distinct shapes. The bracts are

usually half as long to longer than the calyx.

In some specimens, e.g. Dahlgren & Peterson 26 and

Acocks 16014, the pubescence is denser and shorter

but still typical of the majority of specimens. This may

be due to some foreign influence, but the main diag-

nostic characters are present.

77. Hermannia pillansii Compton in Trans. R.

Soc. S. Afr. 19:300 (1931). Type: Cape, Witteberg,

Laingsburg Dist., Compton 2797 (BOL, lecto. !; K;

PRE, photo.!).

Suffrutex, low, branching from base, young

branches densely stellate-hirsute. Stipules narrowly to

broadly ovate, acuminate, broad ones abruptly

acuminate at apex and cordate at base, 4-6 mm long,

1-4 mm broad at base, densely stellate-pubescent.

Leaves sessile, obovate-cuneate, up to 10 mm long,

5 mm broad near apex, crenate except on cuneate

base, densely stellate-pubescent to tomentose on both

surfaces with hairs both short and long. Inflorescence

of 1- to 2-flowered cymes clustered at apices of ulti-

mate branchlets; peduncles 4-8 mm long; pedicels

short, 2-4 mm long; bracteoles narrowly ovate-

acuminate, up to 6 mm long. Calyx 5 mm long, lobed

to almost midway, densely stellate-hispid, hairs

rather long, light yellow. Petals about 7 mm long,

more or less oblong in upper third, narrowed into a

densely ciliate waist and produced below into a claw

with infolded margins. Stamens 4,5 mm long; fila-

ments narrowly obovate, not or only slightly broader

than anthers, hyaline, overlapped by anther bases,

anthers 2 mm long, ciliate. Ovary subglobose, shal-

lowly 5-lobed, 3 mm long whitish stellate-tomentose;

stipe 0,5 mm long; styles 4 mm long. Capsule about

4 mm long, enclosed in the calyx, thinly whitish

tomentose.

Found on mountains at 1 300 m or higher. Recorded

from the Witteberg in the Laingsburg District.

Cape. — Laingsburg: Wittebergen, Compton 2797 (BOL; K);

3159 (BOL); 12215 (NBG); foothills, Marloth 11424.

This species closely resembles H. stipulacea, but can

be distinguished mainly by the leaves and stipules

being densely and persistently pubescent on both

surfaces instead of the very characteristic lax and

distinct pubescence on those of H. stipulacea. In

addition, the shape of the leaf is obovate-cuneate in

H. pillansii and the flowers clustered in terminal

cymes with rather short peduncles and pedicels.

Of the four syntypes, Compton 2797, 3159, 3299 and

3559, Compton 2797 is here selected as type because

there is a duplicate of this specimen in K.

78. Hermannia stipulacea Lehm. ex Eckl. &

Zeyh., Enum. 44, No. 349 (1834); Harv. in F.C. 1 : 192

(I860). Type: Cape, near mouth of the “Zwartkops”

River, Ecklon & Zeyher Enum. No. 349 (K; PRE,

photo.!; PRE!; W!; SAM!).

Suffrutex, erect or sprawling with ascending bran-

ches, branches laxly, rarely subdensely, stellate-pube-

scent, hairs few and long, rarely short from a scaly

base. Stipules leaf-like, about half the size of leaves

or over half, ovate-acuminate to ovate-lanceolate,

cordate at base, from 3,5-11 mm long, 1-7 mm broad

near base, laxly and distinctly stellate-pubescent.

Leaves sessile or shortly petiolate, sometimes erect

and somewhat imbricate; blade oblong, oblong-

cuneate, sometimes narrowly so, 4-16 mm long, 2-6

mm broad, apex acute, rounded to broadly rounded

or subtruncate, usually broadest near the apex which

if often lobed, base rounded, laxly but distinctly and

strongly stellate on both surfaces, hairs long; petiole

up to 2 mm long. Inflorescence of 1- to 2-flowered

cymes in axils of upper leaves and at the apices of the

branchlets; peduncle from 3-20 mm long; pedicels

1-7 mm long; bracts usually 3, 1 long and 2 shorter

(the much reduced leaf and 2 stipules), ovate, lanceo-

late or subulate, 1,5-6 mm long, 0,5-3 mm broad,

stellate-pubescent as on leaves, at least along margins.

Calyx about 6,5 mm long, lobed almost to middle,

densely yellow scaly-tomentose and stellate-hispid,

hairs long, lobes about 3 mm long, deltoid-acute.

Petals about 9,5 mm long, subglobose in upper third,

narrowed into a densely ciliate waist and produced

into a claw with infolded margins. Stamens about 6

mm long, filaments hyaline, oblong-cuneate, ciliate on

shoulders, overlapped by anther bases ; anthers 3 mm

long, acute, ciliate. Ovary oblong, about 2 mm long,

shallowly 5-lobed, subglabrous or shortly stellate

below, densely stellate at the apex of each locule,

hairs long; stipe 0,5 mm. Capsule shortly exsertsd

from persistent calyx, about 6 mm long, stellate-

hairy.

Found in Coastal Renosterveld, on stony mountain

slopes, in Fynbos and burnt patches. Recorded from

Riversdale and Ladismith and eastwards to Port

Elizabeth.

Cape. — Humansdorp: Onzer, Drege 7276 (W); Marsh 1351.

Ladismith: Buffels Kloof, Esterhuysen 13998. Montagu:

between Scheepersrust and Barrydale, Esterhuysen 24603

(BOL). Mossel Bay: Cloete’s Pass, Acocks 14636. Oudtshoorn:

Meiring’s Poort, Esterhuysen 24868 (BOL). Port Elizabeth:

near “Zwartkops River” mouth, Ecklon & Zeyher Enum.

No. 349. Riversdale: Soetmelks River, Acocks 22359; Onver-

wacht, Muir 1607. Uniondale: Noll’s Halt, Acocks 16035;

near Joubertina, Thompson 952; Die Hoek, Esterhuysen 10582.

Characterized by the leaf-like stipules and the

sublax, stellate pubescence with stiff hairs. The densely

lepidote-stellate calyx resembles the calyx of H.

involucrata Cav. which has the same golden yellow

hue.

I. C. VERDOORN

51

Harvey cites a Drege specimen from the Orange

River. This must be a mistake, for the Orange River

is far outside the known area of distribution.

H. pillansii is closely allied to this species, but

differs in the pubescence on the leaf (see under that

species). Also, the hairs on the calyx are a lighter

yellow, not so golden, and the stipules more acutely

acuminate.

H. latifolia sensu Eckl. & Zeyh., Enum. 348 (non

Jacq.) in SAM is H. stipulacea Lehm. ex Eckl. & Zeyh.

79. Hermannia filifolia L.f., Suppl. 302 (1781).

Type: Cape, Thunberg s.n.. Herb. No. 15473 (UPS,

holo. ; PRE, photo. !).

Suffrutex, straggling, with slender branchlets or

erect and robust, often heavily browsed. Stipules

leaf-like, somewhat shorter or longer and usually

flatter than leaves. Leaves sessile, fascicled, at first

crowded on abbreviated shoots, glabrescent, filiform

to subulate, sulcate (ericoid), sometimes appearing

narrowly linear or linear-spathulate on pressed

specimens, firmly fleshy or succulent and glaucous,

4- 20 mm long, apex acute or mucronate, not lobed,

fascicles distant or crowded. Inflorescence of 1- or

more-flowered cymes in axils of upper leaves and at

apices of branchlets in racemose cymes. Calyx 4-6

(-10) mm, finely stellate-pubescent or rough with

minute scales or tubercles. Petals strongly twisted,

usually dark red, blade oblong, abruptly narrowed

just below middle into a claw which is pubescent on

edges. Stamens with more or less linear-oblong,

hyaline filaments. Capsule from 6-12 mm long,

5- umbonate at apex, finely stellate-pubescent between

sutures, more coarsely so on sutures and at apex.

Fig. 1.2.

Key to varieties

Leaves short, 4-15 mm long, or up to 20 mm long in var.

grandicalyx but then much reduced towards the

apices of flowering branchlets and there shorter than

internodes:

Calyx 4-6 mm long, leaves firm, internodes rough with

minute scales or tubercles var. filifolia

Calyx about 10 mm long, markedly paler than petals,

leaves sometimes succulent and glaucous, long

internodes rarely minutely scaly, usually glabre-

scent and shiny var. grandicalyx

Leaves long, up to 20 mm long and not much reduced

towards apices of flowering branches, usually longer

than internodes var. robusta

(a) var. filifolia

Verdoorn in Bothalia 10: 572 (1972).

Hermannia filifolia L. f., Suppl. 302 (1781); Cav., Diss. t. 180,

fig. 3 (1788): Jacq., Hort. Schoenbr. t. 123 (1797): DC., Prodr.

1:495 (1824); Eckl. & Zeyh., Enum. 47, No. 372 (1824),

partly; Harv. in F.C. 1: 195 (1860). Type: as above.

H. passerinaeformis Eckl. & Zeyh., Enum. 47, No. 373 (1824).

Type: Cape, Swellendam, Ecklon & Zeyher Enum. No. 373

(K; PRE, photo!, SAM!; W!; S!.).

H. gilfillanii N.E. Br. in Kew Bull. 1906: 100 (1906). Type:

Cape, Middelburg, Conway Farm, Gilfillan sub Galpin 5506

K, holo.; PRE, photo.! PRE!).

H. filifolia var. passerinoides Harv. (sic) in F.C. 1:195

(1860). Syntypes: Cape, Swellendam, Ecklon & Zeyher Enum.

No. 373; between Ecksteen and Stormvalley, Zeyher 2008 (S !).

Straggling plant with thin branchlets and small

flowers 4-6 mm long borne in axils of upper leaf-

clusters and in racemose cymes at apices of branchlets.

Farther east in the Great Karroo bushes are a degree

more robust and somewhat larger flowers are usually

clustered only at apices of branchlets. The branchlets

are rough with minute scales or tubercles. The

specific epithet “filifolia” refers to the most diagnostic

feature of species, the more or less subulate or

ericoid leaves, which however sometimes appear

narrowly linear to spathulate on pressed specimens.

Leaves clustered, sessile, about 4-15 mm long^

glabrous or minutely scaly pubescent, acute at apex’

sometimes ending in a single hair or subobtuse and

mucronate. Stipules almost indistinguishable from

leaves but a close examination shows stipules to be

narrowly linear rather than subulate and usually

with a broader base. Calyx usually about 5 or 6 mm

long, finely stellate-pubescent or in some areas

rough with minute scales or tubercles as in the

related H. flammea. Capsule about 6 mm long.

Found on mountain slopes, in passes, on rocky

hills, and sandstone ridges. Recorded from Worcester,

Ceres and Swellendam in the Little Karoo and in

the Great Karoo from Aberdeen, Graaff-Reinet and

Cradock northwards to the Orange River, and just

beyond to Bethulie in the Orange Free State.

Cape.— Albert: Byrnavon, 50 km W. of Aliwal North,

Werger 1111. Ceres: N. of Pienaarskloof, Acocks 23112.

Cradock: Jakkalsfontein, Acocks 17505. Graaff-Reinet: Farm

Rietvlei, Galpin 995 ; Farm Klipfontein, Codd 3529. Middelburg:

Conway Farm, Gilfillan sub Galpin 5506; N. of Grootfontein,

Van Zinderen Bakker 1100; N. of Middelburg, Sidey 476!

Swellendam: near Barrydale, Acocks 20353; National Bontebok

Park, Barnard 649. Venterstad: Oviston Reserve, Van Schoor

ORFS-27; Fourie ORFS-11. Worcester: near De Dooms,

Bolus 1 3077 ; Pienaarskloof, Taylor 6494.

O.F.S. — Bethulie: Tussen Riviere Game Reserve, Roberts

5511.

Occasional specimens have been found in some of

the localities recorded for this variety which appear

to be intermediates or hybrids.

(b) var. grandicalyx Verdoorn in Bothalia 10: 572

(1972). Type: Cape, Prince Albert, N. of Prince

Albert Road, Acocks 17098 (PRE, holo.!).

H. linifolia sensu Eckl. & Zeyh., Enum. 371 (1834); sensu

Harv. in F.C. 1: 195 (1860), non Burm. f.

The diagnostic feature of this variety is the large

calyx, about as long as, and much paler than the

usually dark red petals, being pale pink or more

often cream-coloured, finely stellate-pubescent and

lobed to just beyond the middle, lobes long-acuminate

to apex. The fascicled leaves appear linear to linear-

spathulate and plicate on pressed specimens but are

described as subterete, sometimes firm in texture but

often succulent and glaucous. Shoots from heavily

browsed plants are characterized by having long,

usually glabrous and shiny internodes and the flowers

in fairly lax, terminal, racemose cymes. The capsule is

larger than in the typical variety, up to 12 mm long.

Found along foothills, stony ridges and on steep

shaly slopes or rocky outcrops. Recorded from the

Little and Great Karoo, from Montagu eastwards

through the northern areas of George to Steytlerville

and northwards to Sutherland and Beaufort West.

Cape. — Beaufort West: Vindragersfontein, Acocks 14328.

George: Heimans River, Olivier 3575. Ladismith: about 10 km

E. by S. of Ladismith, Acocks 14611; Laingsburg: Whitehill,

Compton 11212 (NBG); about 18 km W. of Laingsburg,

Acocks 24351. Montagu: Jakkalsfontein, Van Breda 1238.

Prince Albert: about 5 km N. of Prince Albert Road Station,

Acocks 17098. Riversdale: N. of Waterval (near Ladismith

border), Muir 3754. Steytlerville: Mt. Pellier, Oliver 4578.

Klein Roggeveld, Mar loth 9584. Uniondale: between Hotspring

and Toorwater, Oliver 3648.

H. linifolia sensu Harv. in F.C. fits this variety and

the two specimens cited, Drege 7285 (LE; W.), and

Ecklon & Zeyher Enum. No. 371 (L; S; W), belong

here. H. linifolia Burm. f. (1768) is synonymous with

Hermannia scoparia (Eckl. & Zeyh.) Harv. (1860).

Harvey recognized that his concept of H. linifolia

was nearly allied to H. filifolia and the many specimens

52

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

examined since have shown that it is merely a variety

of that species.

(c) var. robusta Verdoorn in Bothalia 10: 572

(1972). Type: Cape, Port Elizabeth, Glendinning

Vale, 2,4 km N.W. of Donkin Memorial, Olivier

458 (PRE, holo.!).

H. filifolia sensu Eckl. & Zeyh., Enum. 47, No. 372 (1824),

partly, as in BOL, non L.f.

This variety differs from the typical, principally in

the longer, more crowded leaves, the usually larger

flowers and the more robust plant. The plant can be

up to 1 m tall and the branchlets are rough with

minute scales or the tubercled bases of fallen hairs.

The fascicled, ericoid leaves (appearing narrowly

linear on pressed specimens) together with their

leaf-like stipules are 10-20 mm long and are crowded

on branches, usually longer than internodes and not

much reduced towards the apices of the flowering

branches. The flowers are somewhat larger than those

of the typical variety but not quite as long as the

majority of those of var. grandicalyx. From last-

mentioned it can be distinguished mainly by calyx

which is not quite as markedly paler than the petals

and is rough with minute scales or tubercles, not

finely stellate-pubescent.

To date found in disturbed areas and burnt patches

in the Port Elizabeth and Uitenhage Districts.

Cape. — Port Elizabeth: Glendinning Vale, 2,4 km N.W. of

the Donkin Memorial, Olivier 458; 458a; Victoria Park, Long

792; Baakens River Valley, Galpin 9930; Patons Farm, Long 475;

Earn Cliff, Galpin 6368. Uitenhage: Winterhoek Mountains,

Fries, Nordlindh & Weimarck 1085; Groendal Wilderness area,

Scharf 1018.

Ecklon & Zeyher Enum. No. 372 is obviously a

mixed gathering, because of the two distinct localities

mentioned and this is borne out by specimens in

certain herbaria. In LE and BOL one of the specimens

on the sheet with this label is var. robusta and must

therefore be the Port Elizabeth gathering. On other

sheets the specimens with this label are a form of var.

filifolia. In S one of a mixture of specimens is var.

robusta collected by Ecklon near the Lady Donkin

Memorial. Specimen Zeyher 2004 in PRE is H.

filifolia var. robusta, while in S it is H. flammea.

80. Hermannia denudata L.f, Suppl. 301 (1781).

Type: Cape, Sparrman in Linn. Herb. Cat. No.

854.10 (LINN, holo.; PRE, photo.!).

Shrub virgate, 60 cm-2 m tall, vegetative parts

totally glabrous or (var. erecta) at first rough with

minute, fringed scales, occasionally intermixed with

longer, fascicled hairs, glabrescent in parts. Stipules

narrowly lanceolate-acuminate to ovate-acute, broad-

based, sometimes oblique at base, from 3,5 to 17 mm

long, 0,5 to 6 mm broad near base, several-nerved,

glabrous or pubescent with fascicled hairs, sometimes

glabrescent except on margins. Leaves shortly

petiolate, suberect; blade of mature leaves (often

many small leaves present on young shoots) lanceolate

to more or less oblong, cuneate at base, coarsely

dentate in upper half, from about 13 to 65 mm long,

5 to 14 mm broad, glabrous or with stellate or grouped

hairs from a glandular base, glabrescent in part, apex

acute, midrib prominent beneath; petiole 2-6 mm

long. Inflorescence of 1- to 3-flowered cymes terminal

and in axils of upper reduced leaves of ultimate

branches; bracts like much reduced leaves and sti-

pules, linear to subulate. Calyx glabrous or rough

with minute, fringed scales, campanulate, about 4-6

mm long, 5-lobed to midway or almost midway.

Petals yellow to red and yellow, about 8 mm long,

ovate-oblong in upper third, narrowed into a waist

and produced below into a claw with inrolled margins.

Stamens 4,5 to 5,5 mm long; filaments hyaline,

obovate-cuneate, overlapped by anther bases. Ovary

about 2,5 mm long, 5-angled, glabrous or stellate-

tomentose. Capsule exserted from persistent calyx,

up to 7 mm long, 5-angled with rounded umbos.

A distinct species, characterized by the virgate

habit with more or less straight branches and sub-

erect leaves which are distinctly but fairly laxly

dentate in the upper half, the teeth more or less acute.

The two varieties have restricted distributions, most

surprisingly widely separated from each other.

Key to varieties

Vegetative parts entirely glabrous. Western Cape and

Namaqualand (a) var. denudata

Vegetative parts rough with minute, fringed scales and

tufted hairs from a glandular base, glabrescent in

parts. Eastern Transvaal and Swaziland. . . .(b) var. erecta

(a) var. denudata

H. denudata L.f., Suppl. 301 (1781); Cav., Diss. 6: 329,

t. 181, fig. 1 (1788); Ait., Hort. Kew. ed. 1,2:415 (1789);

Jacq., Hort. Schoenbr. 65, & 122 (1797); Reichb., Ic. Descr.

PI. Cult. t. 59 (1822); DC., Prodr. 1 : 495 (1824); Harv. in F.C.

1 : 194 (1860). Type: Cape, Sparrman in Linnean Herb. Cat. No.

854.10 (LINN, holo. ; PRE, photo. !).

The typical variety is distinguished mainly in that

the vegetative parts are completely glabrous, but there

are a few differences of degree which as a rule cannot

be regarded as specific. The stipules of var. denudata

may reach a size of 17x6 mm although they are

often smaller. The largest stipule seen on var. erecta

was 9x3 mm, but the majority measured only 8x1,5

mm. The leaves on the typical variety are usually

slightly longer in proportion to width, for example a

leaf 40 mm long may be 8 to 10 mm broad in this

variety while in var. erecta a leaf of 30 mm long is

usually 9 or 10 mm broad. The flowers in var. denu-

data are recorded as yellow whereas in var. erecta

they are often red or partly yellow and partly red.

Harvey l.c. cites H. quercifolia Eckl. & Zeyh.,

Enum. No. 383 as a synonym in part of H. denudata.

It has not been possible to verify this.

Found in arid Fynbos in the Western Cape and

Namaqualand. Recorded from Clanwilliam,

Vanrhynsdorp, Calvinia and Namaqualand.

Cape. — Calvinia: Boklanskloof, Lokenburg N., Acocks

19438; Ekerdam, Taylor 2747 (BOL). Clanwilliam: Cedarberg,

Sanddrif, Taylor 7533; Nardouw, Stokoe 8224. Vanrhynsdorp:

“Aties” (not located), Pearson in Percy Sladen Mem. Exped.

5387 (BOL); “Addies”, Zeyher 115.

(b) var. erecta ( N.E.Br .) B. Davy & Greenw. in

Burtt Davy, FI. Transv. 1:41 & 267 (1926). Type:

Transvaal, Barberton, Galpin 1346 (K, holo.; PRE,

photo.!; PRE!).

H. erecta N.E. Br. in Kew Bull. 1897: 245 (1897).

Vegetative parts, including calyx, usually rough

with minute, fringed scales plus grouped hairs from a

glandular base, glabrescent in parts. Stipules 3,5 x

0 , 5 mm to 9 x 3 mm. The flowers are described as red,

yellow or partly yellow and partly red.

Found on mountain tops, slopes, granite hillsides,

in deep soil and semi-shade. Recorded from Lyden-

burg, Nelspruit, Barberton and across the border in

Swaziland.

Transvaal. — Barberton: Barberton, Bayliss BS/1535; Rogers

23867; Galpin 1346. Lydenburg: Erasmus Pass, Strey 3806.

Nelspruit: E. of Nelspruit, Marais 275. Lowveld Botanic

Garden, Buitendag 829; Nyamezane Bantu Trust, Nel 5;

suburbs of Nelspruit, De Winter 9403.

I. C. VERDOORN

53

Swaziland. — Mbabane: near Komati Bridge, Compton

28834.

81. Hermannia flammea Jacq., Hort. Schoenbr.

1: 69, 1. 129 (1797); Sims in Curtis’s bot. Mag. 39:

1. 1 349 (1811); DC., Prodr. 1: 495 (1824); Eckl. &

Zeyh., Enum. 46, No. 367 (1834); Harv. in F.C. 1:

196 (1860). Type: Cape, Cult. Hort. Schoenbr. (W,

lecto.!; PRE, photo.!).

H.falcata Eckl. & Zeyh., Enum. 46, No. 365 (1834). Type:

Cape, Uitenhage, “Winterhoeks et Van Stadensrivierberge”,

Ecklon & Zeyher, Enum. No. 365 (PRE!; W!).

H. polymorpha Eckl. & Zeyh., Enum. 46, No. 366 (1834).

Type: Eastern Cape, “Adow”, “Zuurberge”, “Zwartkopsrivier”,

“Krakakamma” and Grahamstown, Ecklon & Zeyher Enum.

No. 366 (PRE!; SAM!; W!).

H. orophila Eckl. & Zeyh., Enum. 47, No. 369 (1834); Harv.

in F.C. 1: 191 (1860). Type: Ecklon & Zeyher No. 369 (PRE,

lecto. ! ; SAM !), not Ecklon & Zeyher Enum. No. 369 (S), see

note.

SufFrutex up to about 65 cm tall, sometimes taller,

erect or diffuse, with slender, usually suberect branch-

lets, branchlets sparsely to densely pubescent with

minute, fringed scales or coarse, stellate hairs, pubes-

cence usually densest on pedicels. Stipules usually

linear-oblong to oblong, mucronate; sometimes

obovate, usually erect but sometimes falcate, 4-11

mm long, 1-2 mm broad, stellate, glabrescent. Leaves

usually fascicled, subsessile or very shortly petiolate;

blade broadest at or near apex, cuneate, from 5-25

mm long, 3-10 mm broad, lowermost sometimes

larger, apex rounded or subtruncate with a small,

slightly recurved mucro and usually 2- to 6-toothed,

stellate-hairy becoming glabrous, sometimes stellate

hairs persistent but then leaves broad, sometimes

persistent on margins only, leaves often rough with

scaly bases of the stellate hairs; petiole up to 2 mm

long. Inflorescence of 1- to 2-flowered cymes arranged

in fairly lax, terminal, racemose cymes or clustered at

apex of branchlets; peduncles 2-4 mm long; bracts

subulate, linear or obovate, about 3 mm long;

pedicels 2-3 mm long. Calyx lobed to about middle;

lobes oblong-deltoid, base broad, very shortly and

broadly acuminate to apex which is often incurved at

first and then at least some spreading, rough with

fairly laxly stellate hairs or with minute scales fringed

with short hairs. Petals red or yellow, “yellow red-

brown”, “orange-red”, “yellow inside, maroon out-

side”, strongly twisted in upper half, 8-9 mm long,

about 5 mm broad at broadly rounded apex, narrow-

ing about midway into a rather broad claw with

infolded margins, densely stellate-pubescent along

the edges of lower half. Filaments hyaline, linear-

oblong, somewhat cuneate, about 3 mm long;

anthers about 2 mm long, ciliate, shortly overlapping

filaments at base. Ovary about 6 mm long, stellate-

pubescent. Capsule long exserted from persistent

calyx, about 8 mm long, 5 mm broad, densely stellate-

pubescent, 5-umbonate at the apex, umbos stellate.

Found on mountain slopes, stony hills and in

valley bushveld. Recorded from western coastal

Renosterbosveld, in the Somerset West, Stellenbosch

and Paarl Districts, and, more commonly, in the

southern coastal belt from Swellendam eastwards

through Port Elizabeth and Albany to Kentani.

Cape. — Albany: near Grahamstown, Schlechter 2637;

Aloes, Drege 3005. Alexandria: Galpin 10758; near Nananga

Store, Story 1310. Bathurst: Hopewell, Acocks 11073; Port

Alfred, Bayliss 2640. East London: Nahoon River, Galpin 3286;

Smith 3762. Humansdorp: Thode A71 1 ; The Glen, Burtt Davy

12084. Kentani: coast, Pegler 37; Kei Mouth, Flanagan 198.

King Williams Town: Tyson 843. Knysna: Knysna, Keet 672.

Mosselbay: Rogers 22791. Paarl: Paardekop, Drege 7305. Port

Elizabeth: slopes “Winterhoeks & Van Stadensrivierberge"

Ecklon & Zeyher Enum. No. 365; Zuurberge near “Zwartkops-

rivier & Krakakamma”, Ecklon & Zeyher Enum. No 366-

Zuurberg Sanatorium, Long 731; towards Wittekop, Rodin

1034. Riversdale: H. Bolus 11219. Somerset East: Zuurberg

Hotel, Story 2383; near Annsville, Story 177. Stellenbosch-

near Hottentotsholland, Ecklon & Zeyher Enum. No. 399;

Stellenbosch Mtn, Bos 133. Stutterheim; Commonage. Acocks

9536; Rogers 12715.

Characterized by the petals usually appearing

darker than the calyx on dried specimens, and by the

leaves being broadest at or near the apex with the

majority toothed at the apex, usually long cuneate

and sublaxly pubescent with scales fringed with

short hairs or coarse stellate hairs, usually glabrescent

and rough with the bases of the hair-groups, or the

stellate pubescence persisting on some broad-leafed

forms or persisting on the margins only; calyx rough

from the tubercled bases of the hairs, lobed to about

midway, the lobes oblong-deltoid or deltoid from a

broad base, often with the apical portion infolded at

first and eventually at least some spreading. This

feature gave rise to the description of the species H.

falcata, now a synonym. The flowers are usually dark

red, but a few yellow forms are recorded. Some

specimens give the impression of being of hybrid

origin. The leaves are usually in fascicles in the wild.

The specimen figured in Hort. Schoenbr. and the

pressed specimen in the Vienna herbarium show the

leaves as not fascicled. This may be due to the plant

being young. Other plants collected later in the same

garden have fascicled leaves.

A specimen from the “Hortus Schoenbrunnensis”

preserved in the National His ory Museum, Vienna,

bearing the name H. flammea, agrees with Jacquin’s

description and plate and is therefore here selected as

lectotype.

H. orophila Eckl. & Zeyh. is based on their No.

369, collected at “Hottentottsholland” (Stellenbosch)

and “Zwarteberg et Klynriviersberge” (Caledon).

Judging by three specimens seen with this number,

none of which has a specific locality, there are two

distinct entities involved. Two specimens, one in

PRE and the other in SAM, are obviously conspecific

with H. flammea, having the dark petals, the charac-

teristic pubescence on the calyx and the fringed scales

on the branchlets. The third sheet, in S, is a smaller

plant with petals obviously yellow and the branchlets

very rough with tubercle-based hairs, matching H.

rudis. In the description of H. orophila, the only item

which points to one or other of the two entities is

the phrase “Flores purpureo-violacei,” which applies

to the PRE and SAM specimens and not to the S

specimen. As the PRE specimen is in better condition

than the one in SAM, the former is selected as the

lectotype, and H. orophila is included in synonymy

under H. flammea.

82. Hermannia flammula Harv. in F.C. 1 : 196

(1860). Type: Cape, Caledon, Zeyher 2009, (K,

lecto; PRE, photo.!; Z! ; PRE!).

H. trifurca sensu Eckl. & Zeyh., Enum. 370, partly excl.

Jacq. t. 125 (see Linnaea 19: 612 and fide Harvey, l.c. 196).

SufFrutex, erect, virgate, with slender branchlets

leafy towards their apices, 25-60 cm tall, rarely 1 mm

tall, branchlets pubescent with tubercle-based stellate

hairs or fringed scales, intermixed with single, thin-

walled, gland-tipped hairs. Stipules linear-oblong or

very narrowly ovate-oblong, 2-8 mm long, 0,75-2

mm broad, stellate-pubescent with a few gland-tipped

hairs intermixed. Leaves subsessile, petiole rarely up

to 2,5 mm long; blade narrowly oblong-cuneate,

usually folded, 7-15 mm long, rarely up to 22 mm

long, 1,5-4 mm broad near apex, apex rounded,

54

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

mucronate or subtruncate and lobed, persistently

stellate-pubescent on both sides, either coarsely so

with hairs short or long, or densely so with long,

thin-walled hairs (tomentose?), simple gland-tipped

hairs usually intermixed. Inflorescence of 1- to 2-

floweied cymes, one or two terminal on branchlets

and appearing axillary when ultimate branchlets are

very short; peduncles 1-4 cm long; pedicels unequal

in a cyme, 2-8 mm long, densely and softly pubescent

with gland-tipped hairs interspersed; bracts usually 3,

linear to subulate, about 3,5 mm long and up to 0,5

mm broad, acute with a few erect hairs at apex.

Calyx about 7 mm long, lobed to just beyond the

middle, stellate-pubescent, densely and with more

thin-walled hairs at the base, some hairs gland-tipped.

Petals twisted, usually dark red, “claret red”, “deep

maroon”, “deep port wine red”, “creamy yellow in

lower half”, about 8 mm long, about 3,5 mm broad

near the somewhat oblique apex, more or less oblong

in upper third, then narrowed into a waist and

produced into a claw with infolded margins and with

stellate tomentum along sides. Stamens with hyaline,

more or less narrowly oblong filaments which are

somewhat longer than anthers and were glabrous in

the specimens dissected, hooked at the point of

attachment; anthers ciliate. Ovary densely stellate-

pubescent, about 2,5 mm long; styles cohering

about 2,75 mm long; stipe 0,75 mm long. Capsule

5-8 mm long, 5-umbonate at apex, stellate-pubescent

especially along thickened sutures where stellate

hairs are on tubercled bases.

Found on hill sides and mountain slopes to Coastal

Renosterveld or Ngongoniveld. Recorded from

Caledon District eastwards to George and, further

inland, from Ladismith to Uniondale.

Cape.— Bredasdorp: Klipdale, Smith 2584; 3198. Caledon:

“Zwartberg”, Zeyher 2009a; Caledon, Compton 14708. Ladi-

smith: Roodeberg, Esterhuysen 17190. Mossel Bay: W. of

Mossel Bay, Acocks 15397. Riversdale: E. of Heidelberg,

Acocks 21101; Milkwoodfontein, Galpin 3789. Swellendam:

Voormansbosch, Zeyher 2009b; Zuurbraak, Acocks 14081.

Uniondale: Joubertina, Van Breda 1189; Buffelsrivier below

Kammanasieberg, Oliver 3623.

Characterized by the persistently stellate-pubescent,

small leaves, the mixed pubescence, some stellate

hairs with thick walls (hard looking) but mostly thin-

walled (soft looking) and single hairs with gland-tips

intermingled, the flowers usually dark red (with pale

calyx), cymes 1-, 2- or 3-flowered, terminal on the

branchlets (some flowering branchlets very short).

83. Hermannia joubertiana Harv. in F.C. 1: 196

(1860). Type: Cape, Bredasdorp, Soetendalsvallei,

Joubert s.n. (S-Herb. Sond., holo; PRE, photo.!).

Suffrutex, 30-90 cm tall, spreading or bushy,

branchlets pubescent with fringed scales or stellate-

pubescent, stellate hairs mostly tubercle-based, gla-

brescent but small tubercles persistent. Stipules

narrowly oblong to ovate-oblong, 2-5 mm long,

1 ,5-2 mm broad at base. Leaves sessile, up to about

12 mm long, narrowly cuneate or spathulate, entire

except sometimes at apex, broadest near apex, about

2,5 mm broad, usually plicate, apex obtuse, mucro-

nate, rarely lobed or toothed, abaxial surface usually

pustulate and with minute stellate hairs usually

crowning the pustules, glabrous on inner face, midrib

broad (translucent?). Inflorescence of 2- to 3-flowered

cymes, cymes clustered at apices of branchlets and

sometimes in axils of upper leaves; peduncles short,

up to about 2 mm long, bracts linear, 2 mm long;

pedicels 2-5 mm long, erect and cernuous only at

apex, stellate-pubescent, sometimes densely so and

with gland-tipped hairs interspersed. Calyx about 4

mm long, lobed to or almost to middle, tube at first

narrowed at mouth but eventually wide and shallow,

lobes acute from a deltoid base, sinuses wide, coarsely

stellate-pubescent outside, sometimes gland-tipped

hairs present as well, rarely glabrescent. Petals red,

partly yellow within, “dark red”, “bright red”, or

“dark purple”, twisted, about 8 mm long, blade

oblong, abruptly narrowed below middle into a claw

with infolded margins, edges of claw thinly tomen-

tose. Stamens about 4 mm long, anthers somewhat

shorter than hyaline, more or less oblong filaments;

filaments shortly overlapped by anther bases. Ovary

about 2 mm long, shallowly 5-lobed and 5-umbonate

at apex, stellate-hairy; stipe under 1 mm long; styles

cohering, about 3 mm long; stigma capitate, usually

conspicuous. Capsule about 4,5 mm long, stellate-

hairy, bluntly 5-angled, 5-umbonate at apex; stipe 1

mm long, umbos 1 mm long.

Found on sandy flats and stabilized dunes in Coastal

Fynbos. Recorded from Caledon through Bredasdorp

to Mossel Bay.

Cape. — Bredasdorp: S.W. of Bredasdorp, Acocks 15458;

Caledon: Hermanus, I. B. Walters 282 (NBG). George: George,

Thorne in SAM 51673 (SAM). Mossel Bay: N.E. of Mossel

Bay, Acocks 15387. Riversdale: S. of Riversdale, Acocks 24097.

Swellendam: Kleinfontein, between Brede and Duiwelshoek

Riviere, Zeyher s.n. (SAM).

Characterized by the cuneate leaves usually folded

and pustulate dorsally, sparsely stellate to glabrescent,

hairs minute, apex mucronate, the average not

toothed or lobed at the apex. The calyx is small and

shallow with acute lobes and wide sinuses. Flowers

usually red, small, petals about twice as long as the

calyx.

84. Hermannia scabra Cav., Diss. 2, t. 182, fig. 2.

(1788); DC., Prodr. 1:495 (1824), partly excl. syn.;

Harv. in F.C. 1: 191 (1860). Type: Cape, specimen in

Herb. Lamarck s.n. (P, holo.; PRE, photo.!).

H. scordifolia sensu Eckl. & Zeyh., Enum. 43, No. 344 (1834).

H. denudata sensu Eckl. & Zeyh., Enum. 43, No. 345 l.c.

H. presliana Turcz. in Byull. mosk. Obshch. 32, 1 : 259

(1859). Type: Cape: “Paarlberg”, Zeyher (sic) 7294 (for Drege

7294) (K; PRE, photo.!; PRE!; S! LE!, W!).

H. patula Harv. in F.C. 1:187 (1860). Syntypes: Cape,

mountain side near Tulbach, Ecklon & Zeyher Enum. No.

344 (S !) ; “Paarlberg”, Drege 7294.

H. pedunculata Phill. in Ann. S. Afr. Mus. 9, 3: 114 (1913),

non K. Schum. Syntypes: Cape, Vanrhynsdorp, Giftberg,

Phillips 7377-7379 (SAM!).

Suffrutex, low spreading bush to stiffly erect

shrublet up to 60 cm tall; branches fairly finely

stellate-pubescent with intermingled gland-tipped

hairs to sparsely and harshly stellate-pubescent, the

hairs appressed or spreading, 1 to few from a basal

tubercle. Stipules ovate to oblong-ovate from a broad

base, sometimes obliquely cordate at base, 2-6 mm

long, 1 , 5-3 mm broad at base, sparsely stellate-

pubescent. Leaves shortly petiolate; blade cuneate

from near apex or from midway, 2,5-40 mm long,

varying in width from 2-15 mm broad at or near

apex, lobed in upper half or at apex only, sparsely

and harshly stellate-pubescent, hairs often 1 from the

tubercled base, especially on the margins, glabrescent

especially on upper surface which is finely dotted

with minute tubercles or glands; petiole 1,5-5 mm

long. Inflorescence of long, leafless, racemose or

paniculate cymes terminal on branchlets, cymes

1- or 2-flowered, internodes long, 10-44 mm long;

peduncles 2-20 mm, stellate-pubescent; pedicels 2-10

mm long; bracts like stipules but reducing upwards,

I. C. VERDOORN

55

lower ovate-cordate or oblique at base, about 8 mm

long, reducing to very narrowly lanceolate bracts

about 2 mm long, usually 3 at apex of peduncle,

central the largest 4 , 5 x 3 , 5 mm, lateral 2 , 5 x 1 , 5 mm.

Calyx 6 to 7 mm long, lobed to about midway, at

first stellate-pubescent especially on nerves, usually

coarsely so with hairs from a raised base, sometimes

more finely stellate with intermingled gland-tipped

hairs, with age becoming glabrescent and scarious;

lobes deltoid often ciliate with long hairs; sinuses

wide, at least in mature flowers. Petals bright yellow

turning reddish brown with age, about 10 mm long,

orbicular to oblong in upper third, narrowing into

a waist which is pubescent, especially on margins

which are usually densely fringed, produced below

into a claw with infolded margins. Stamens about 7

mm long; filaments hyaline, obovate, pubescent on

shoulders, overlapped at apex by anther bases;

anthers ciliate. Ovary 5-lobed, 5-umbonate at apex,

about 3 mm long, stellate especially on sutures;

styles about 4 mm long. Capsule exserted from

persistent calyx, about 6 mm long, 5-lobed and 5-

umbonate at apex, stellate-pubescent, the hairs

longer on the sutures.

Found in Fynbos and Renosterveld, on mountain

slopes, in rocky crevices and stony plains. Recorded

from Bellville and Paarl Districts northwards to

Clanwilliam District.

Cape. — Bellville: Visschers Hoek, Compton 13431 (NBG).

Ceres: Ezelsfontein, Esterhuysen 20343. Clanwilliam: Blaauw-

berg, Schlechter 8466; near Olifants River, Ecklon & Zeyher

Enum. 345 (S). Malmesbury: near Malmesbury, Schlechter

1633; near Hopefield, Mario th 8224. Paarl: “Paarlberg”,

Drege 7294; Wemmershoek Valley, Esterhuysen 17663. Piket-

berg: slopes adjoining the town, De Winter & Verdoorn 9062;

Piekeniers Kloof, Schlechter 4341. Tulbagh: Tulbagh Road,

Rogers 17055; 3,2 km N. of Hermon, Marsh 724. Vanrhyns-

dorp: Giftberg, Phillips 7377; 7379.

Characterized by the harsh, but fairly lax, stellate

pubescence on the strongly cuneate leaves which

vary considerably in size especially in width, and the

lobing which is usually in the upper half but sometimes

at the apex only. The calyx is sparsely stellate-pubes-

cent at first, becoming glabrous and scarious as it

matures. This may account for Turczaninow’s

description of his H. presliana as “calycibus inflatis”,

for on the specimen on which his description is

based the calyx had reached the scarious stage. The

lobes are deltoid, sometimes ciliate with long hairs

and the sinuses wide.

In spite of the variation in pubescence and especially

in the width of the leaves, the species is readily

recognizable. This was confirmed when the specimens

cited by Harvey under H. patula were found to include

the wide range of leaf width reflected in the description

above.

H. myrioclada Diels in Bot. Jb. 55: 359 (1919) may

belong in H. scabra (see also Species Requiring Further

Attention, p. 61).

The specimen figured by Jacq. in Hort. Schoenbr.

t. 127 as H. scabra Cav. is H. aspera Wendl. (see note

under that species).

Specimens of Ecklon & Zeyher Enum. No. 350

named “H. scabra Jacq. non Cav.” are H. aspera,

and the following are H. scabra: No. 344 (named H.

scordifolia ) and No. 345 (named H. denudata). A

specimen of Enum. 351 (in SAM), labelled H. biflora

Eckl. & Zeyh. is nearest H. scabra, but possesses

characters reminiscent of other species, such as a

densely pubescent calyx somewhat resembling that of

H. multiflora.

Story 2990, which Pillans identified as “ H . citrus-

dalensis Pillans sp. nov.”, a name never published,

falls within the range of variation of H. scabra Cav.

85. Hermannia angularis Jacq., Hort. Schoenbr.

1: t. 126 (1797); DC. Prodr. 1:495 (1824); Harv.

in F.C. 1: 197 (1860). Verdoorn in Flower. PI. Afr.

41 t. 1604 (1970). Iconotype: Cape, cult., Hort.

Schoenbr. 1 : t. 126.

H. trifurca sensu Eckl. & Zeyh., Enum. 47, No. 370, partly;

non L.

Suffrutex, 30-60 cm high, branching at base,

branches spreading-ascending, reddish brown in

parts, sub-branched, subdensely harshly stellate-

pubescent, hairs short, 0,5-1 mm long, spreading or

erect and then based on small tubercles, rarely

glabrescent. Stipules narrowly ovate-lanceolate to

linear-acute, with midrib usually obvious, 5-9 mm

long, 1,5-3 mm broad near base. Leaves sessile to

shortly petiolate, fascicled; blade obovate-cuneate,

narrowly oblanceolate or linear-spathulate, 7-30 mm

long, 3-12 mm broad, broadest at or near apex

which is often truncate or rounded at 3- or more-

toothed, sometimes entire and acute, upper and lower

surfaces laxly stellate-pubescent, glabrescent, margins

sometimes ciliate with stellate or tufted hairs. In-

florescence terminal on ultimate branchlets, 1- or

more 2-flowered cymes arising from a fascicle of

leaves, stipules and bracts; peduncles 4-10 mm long,

stellate-pubescent; pedicels 2-6 mm long, stellate-

pubescent; bracts linear-acute, about 5 mm long, 1

mm broad, ciliate. Calyx parchment-like, 8 mm long,

lobed to about midway, net-veined, glabrous except

for stellate pubescence on margins of lobes, 5-angeld,

angles extending from sinuses to base of calyx, the

portion between being concave. Petals about 9-10

mm long, twisted, the shortly exserted portion orange-

red shading to lemon-yellow; blade broadly oblong,

6-6,5 mm long, 5-6 mm broad, abruptly narrowed

into a usually short claw, claw about 3 mm long with

broadly infolded margins, minutely papillose within.

Stamens about 5 mm long; filaments hyaline, oblong-

obovate to obtrullate, apparently glabrous, over-

lapped by filament bases; anthers 2 mm long, ciliate.

Ovary 2 mm long, stellate-pubescent, shortly stipitate

stipe 0,5 mm long; styles cohering in a column, 3,5

mm long. Capsule subglobose, about 6 mm diam.,

usually enclosed by persistent perianth, stellate-

pubescent, hairs short from a scaly base except on

sutures where they are slightly longer. Fig. 1.3.

Found in the south-western Cape, in mountain

passes and at the coast from Somerset West eastwards

to the George District.

Cape. — Caledon : Houw Hoek Pass, Marloth 4806 ; Schlechter

7781. George: Herold’s Bay, Acocks 21232. Knysna: Keurbooms

River, Hutchinson 1373. Mossel Bay: Cloete’s Pass, Acocks

14635. Paarl: French Hoek Pass, Esterhuysen 18886. Riversdale:

Garcias Pass, Acocks 15433; Galpin 3787. Somerset West:

Gordon’s Bay, H. Bolus 8076. Stellenbosch: Sir Lowry's Pass,

H. Bolus 9925. Swellendam: Appelkraal, Zeyher 2006; Hemel

en Aarde, Zeyher 2007b. Worcester: Steynsberg, Esterhuysen

15597.

The distinguishing characters are found mainly in

the flowers. The parchment-like calyx is distinctly

5-angled, the angles extending from the sinuses to the

base of the calyx with the portion between being

concave. The calyx is broad at the base and the lobes

do not spread nor reflex at any time. The s hortly

exserted petals are strongly twisted and lemon -yellow

grading into orange-red. Characteristic too is the

blade of the petal which is broadly oblong and several

times longer than the basal claw.

56

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

The great variation in leaf-shape and pubescence

suggests that this species has crossed with related

species in the same area, the characteristic calyx being

a dominant feature. Muir 1456 (PRE), named “ H .

glabricau/is Pillans n. sp. (ined.)” could, for example,

be of hybrid stock, with H. filifolia var. grandicalyx

as one parent.

H. angularis sensu Eckl. & Zeyh., Enum. 356 and

sensu Reichb., Ic. Descr. PI. Cult. t. 68 fig. 1, is H.

hyssopi folia L.

None of the Ecklon & Zeyher specimens named H.

trifurca L. belongs to the Linnaean species. Some are

H. flammula (e.g. Zeyher 2009 in PRE! and Z!) and

others H. angularis (e.g. Zeyher 2006 and 2007 in

PRE! and Z!).

86. Hermannia rudis N.E.Br. in J. Bot., Lond.

39: 398 (1901); Adamson in Adamson & Salter, FI.

Cape Penins. 585 (1950). Type: Cape Peninsula, near

Simonstown, H. Bolus 4950 (K, lecto.; PRE!; BOL!).

H. rudis var. exserta N.E. Br. l.c. (1901). Type: Cape,

Miller’s Point, S. of Simonstown, Wolley Dod 2997 (K, holo. ;

PRE, photo.!).

H. orophila sensu Eckl. & Zeyh., partly, as to Ecklon &

Zeyher Enum. 369 (S ! ; PRE, photo.!), non Eckl. & Zeyh.

(see note under H. flammea).

Low suffrutex, branching at base, branches spread-

ing, ascending, sub-branched, upper ultimate branches

short, rough, fairly sparsely stellate-pubescent or

hispid with a few stiff hairs from a raised tubercle.

Stipules lanceolate to oblanceolate, narrowly ovate-

oblong or ovate, acute, 3-7 mm long, 0,5-3, 5 mm

broad. Leaves sessile or with a petiole up to 7 mm

long; blade obovate-cuneate, 10-25 mm long 6-15

mm broad, broadly rounded to subtruncate and

coarsely toothed at apex, roughly stellate-pubescent

on both surfaces, glabrescent, often hispid on margins

with stiff hairs from a tubercle. Inflorescence of 1-

to 2-flowered cymes in the axils of upper and sub-

terminal leaves of branches and the many, short,

lateral branchlets; peduncles and pedicels very short,

rarely up to 3 mm long. Calyx often red (not visible

in dried specimens), hispid, especially on veins, with

tubercle-based hairs, 7-8 mm long, lobed to midway,

lobes oblong and abruptly narrowed into an apicule

which is incurved (calyx lobes look truncate). Petals

twisted, “pink”, “pale yellow”, “yellow-orange”,

7-8,5 mm long, scarcely or shortly exserted from

calyx; limb oblong narrowing about midway and

then produced below into a fairly broad claw with

infolded margins, pubescent at middle on rim with

tubercle-based hairs. Stamens with hyaline filaments,

oblong, slightly narrowed to base, about 3,3 mm

long; anthers 2 mm long. Capsule bluntly pentagonal,

lobes rounded at apex, hispid with tufted hairs on

angles and apex, finely stellate-pubescent or glabres-

cent between angles.

Found in Coastal Fynbos or Renosterbos, in dry

sandy soil in rocky areas and on slopes, usually

facing the sea. Recorded from a restricted area in the

southern part of the Cape Peninsula and around

False Bay with outliers farther east at Potberg,

Swellendam District and near Caledon.

Cape. — Caledon: Plat-se-Berg, Esterhuysen 12961; Kogel-

berg, PHtans 9817; Boucher 500; 1391; near Rooi Els, Taylor

4903. Peninsula: Cape Town flower show, Marloth 8814; Fish

Hoek, Adamson 3028; Simonstown, Bolus 4950. Swellendam:

Potberg, Thompson 1129 (STE).

This species is characterized by the rather coriaceous

calyx, which is often red and has the tips of the lobes

infolded giving them a truncate appearance. The

pubescence is harsh, mostly of tufts of stiff hairs on a

raised tubercle. The flowers, which are very short

stalked, grow in the axils of the upper and sub-

terminal leaves and so appear to be in heads at the

ends of the branchlets. The degree to which the

corolla is exserted from the calyx varies and is not

combined with any other difference. It may even be

due to the age of the flower for, while the petals are

still tightly rolled, they seem to be farther exserted.

The variety exserta is therefore not upheld here.

87. Hermannia abrotanoides Schrad., Hort. Goet-

ting. 17 t. 11 (1809); Harv. in F.C. 1 : 204 (I860); M.

Friedrich et al. in F.S.W.A. 84: 11 (1969). Iconotype:

Schrad., Hort. Goetting. 17 t. 11 (1809). (Typotype not

traced).

H. bipinnata Burch., Trav. 1:310 (1822). Type: Cape,

Prieska, Moddergat, Burchell 1627 (K, holo.; PRE photo ’•

PRE!; LE!).

H. multifida DC., Prodr. 1 : (1824), nom. superfl.; Harv. in

F.C. 1 : 204 (1860). Type: as for H. bipinnata Burch.

H. halicacaba DC., Prodr. 1:493 (1824); Harv. in F.C.

1:204 (1860). Type: Cape, Prieska, Asbestos Hills, Burchell

2020 (K, holo.; PRE, photo.!; PRE!; LE!).

Subherebaceous perennial, a low, bushy, leafy

plant with several to many slender stems from a

woody base; stems and branches densely to sublaxly

canescent with silvery strigose and appressedly

stellate hairs, hairs fine and many from a central

scale, minute glandular hairs often present. Stipules

subulate and some upper ones linear-lanceolate,

3-10 mm long, sublaxly stellate-pubescent. Leaves

basal and cauline, petiolate; blade 10-40 mm long,

pinnately to palmately divided to midrib or almost so,

pinnae divided again with ultimate lobes obtuse and

decurrent, densely to sublaxly appressedly stellate-

pubescent; petiole 5-40 cm long. Inflorescence of

1- to 2-flowered cymes, terminal on branchlets and in

axils of upper leaves, forming lax, leafy, racemose

cymes; flowers few and large; peduncles 5-20 mm

long; pedicels 3-7 mm long; bracteoles subulate to

narrowly deltoid, 1,5-5 mm long. Calyx inflated,

subglobose, densely stellate-pubescent with appressed

stellate hairs some of which are very shortly stalked

and readily caducous, leaving glabrescent patches,

green to papery, sometimes suffused with pink or

purple, about 10 mm long when flattened, 5-lobed to

less than halfway; lobes acute. Petals “golden yellow”,

“orange-red”, yellow suffused in part with purplish

pink, twisted, usually only shortly exserted from

inflated calyx, about 1 1 mm long, subglobose in

upper half, narrowed below into a rather broad claw

which is incurved at base and has broad, infolded

margins, margins minutely ciliate. Stamens about 8

mm long; filaments hyaline, more or less obovate,

up to 4 mm broad at the shoulders. Ovary more or

less globose, densely tomentose with short, light

brown, stellate hairs; stipe under 1 mm long; styles

6 mm long. Capsule short, 5 mm long, more or less

globose, overtopped by the persistent perianth.

Found on stony hills, in sand pockets between

quartzite rock, on the edge of lime plateaux, among

asbestos hills. Recorded from the Orange River basin

in the Prieska and Kenhardt Districts northwards

through Hay and Gordonia to South West Africa as

far north as the Windhoek and Gobabis Districts.

Also in Botswana.

Cape. — Gordonia: Upington, Van der Schijff 8060. Hay:

E. of Koegas, Codd 1244; Wolhaarkop, Esterhuysen 2438.

Kenhardt: W. of Kenhardt, Schlieben 8833; S.E. of Pofac'der,

Leistner 2421. Prieska: near Asbestos Hills, Marloth 2022;

Burchell 2020; Moddergat, Burchell 1627.

I. C. VERDOORN

57

S.W.A. — Aroab: N. of Aroab, Acocks 18096. Gobabis:

Witvlei, Mason & Boshoff 2513. Keetmanshoop: Klein Karas-

berg, Dinter 4857 (Z); Galpin 14172. Rehoboth: Farm Bergland,

Merxmiiller 835; between Gollschau and Windhoek, Tdlken &

Hardy 706. Windhoek: Finkenstein, Seydel 1748; Auasberge,

Dinter 3498.

Resembles H. comosa in some respects but is

readily distinguished by the consistently deeply

divided leaves. H. abrotanoides is characterized by the

silvery stellate pubescence, together with the inflated

calyx which is appressedly stellate-pubescent, the

hairs fine and many from a silvery central scale,

some very shortly stalked as well and these are

readily caducous. H. pulverata has the same silvery

stellate pubescence and resembles H. abrotanoides in

the divided leaves but it can be distinguished mainly

by the calyx which is not inflated.

The type figure is of a plant originally from the

Cape, grown in the Herrenhaus en Garden. No

specimen of the plant figured has been traced. There

is a specimen in S which can be looked upon as

authentic. It came from Lehmann’s herbarium

labelled “Missit Wendland. Cult, horto Herren-

husiano”. Unfortunately the specific name is given as

“ abrotanifolia ” instead of “ abrotanoides" , evidently by

mistake.

88. Hermannia pulverata Andr., Bot. Rep. 3:

t. 161 (1801); DC., Prodr. 1:496 (1824) (sphalm

“pulverulenta”); Eckl. & Zeyh., Enum. 47 No. 375

(1834); Harv. in F.C. 1: 203 (1860). Iconotype: Cape,

cult. Hammersmith, t. 161 in Andr. Bot. Rep. 3

(typotype not traced).

H. argentea Sm. in Rees, Cycl. 17 (1819); DC., Prodr. 1 : 496

(1824); Eckl. & Zeyh., Enum. 47, (No. 376 (1834); Harv. in

F.C. 1: 203. Type: Cape, cult. Leyden, Smith s.n. (BM, holo.;

PRE, photo. !).

H. bolusii Szyszyl., Polypet. Thalam. Rehm. 19 (1887). Type:

Orange Free State, Kanonfontein, Rehmann 3544 (Z, holo.!;

PRE, photo.!; S!).

H. cana K. Schum. in Bot. Jb. 10:42 (1888). Type: Cape,

between Kuruman and Boetsap, Marloth 947 (PRE!; SAM!).

Subherbaceous perennial, about 30 cm tall or, if

protected, taller, stems 1 to several from a woody

base (or many after being heavily grazed), erect or

spreading-ascending, sparsely branched above, silvery

canescent with strigose and appressed stellate pubes-

cence, the stallate hairs from a central, silvery scale,

often minute, gland-tipped hairs present. Stipules

subfleshy, up to 10 mm long, ovate to ovate-oblong,

broadly to narrowly so, often cordate at the base and

oblique, silvery appressed stellate-pubescent. Leaves

basal and cauline, petiolate; blade variable in shape

and degree of lobing, broadly to narrowly ovate-

oblong in outline, from coarsely lobed to pinnatipar-

tite, bipinnatipartite or palmatipartite, silvery stellate

pubescent to thinly so, lobes obtuse and decurrent;

petiole 2-12 mm long. Inflorescence of 1- to 2-flowered

cymes, terminal and in axils of upper leaves, forming

racemose cymes; peduncles suberect, 5-30 cm long,

rarely longer, appressed stellate-pubescent and often

with scattered, minute, gland-tipped hairs; pedicels

short, 1-6 mm long, nodding; bracts like stipules but

usually smaller; bracteoles usually 3, 1-2 mm long.

Calyx up to 5 mm long, campanulate, thin-textured,

silvery stellate without, and usually with scattered,

minute, reddish, gland-tipped hairs, lobed to almost

halfway, sinuses wide. Petals usually under 10 mm

long, “khaki yellow”, “indigo to cream”, “dirty yellow

turning golden brown”, tightly curled (flowers

appear half closed and truncate on dried specimens),

orbicular to oblong-orbicular, narrowing in lower

half to a claw with infolded margins, apparently

glabrous. Stamens about 5 mm long with hyaline

oblong-obovate filaments glabrous in specimens

dissected; anthers about 2-5 mm long, slightly

shorter than filaments, ciliate. Ovary about 2,5 mm

big, stellate-pubescent, 5-lobed; stipe short, up to

0,75 mm long; styles 4,5 mm long. Capsule up to 8

m n long, appressed stellate-pubescent especially

along sutures, remains of perianth persisting at base

only.

Found in karroid scrub, karroid broken veld and

false upper karoo, in vleis, shaly river valleys and

rocky kloofs. Recorded from Worcester eastwards to

Alexandria and northwards through Calvinia and

Griqualand West through the Cape midlands to

Fauresmith and Rouxville in the southern Orange

Free State.

Cape. — Alexandria: Addo Elephant Park, Barnard 558.

Barkly West: Holpan, Acocks 109. Calvinia: Calvinia, Schmidt

165. GraafF-Reinet : Graaff-Reinet, Thode A549. Hanover:

Hanover, Sim sub Galpin 5968. Hay: between Campbell and

Griquatown, Acocks & Haf strom H 1043. Herbert: Roben-

fontein, Werger 1430. Kimberley: Blaauwbosdrift, Acocks 1 146.

Laingsburg: S.S.E. of Laingsburg, Acocks 20503. Mossel Bay:

Karoo, “Gauritzrivier”, Ecklon & Zeyher Enum. No. 375 (S).

Middelburg: near Schoombie, Schweickerdt 1270; Grootfontein,

Theron 1182. Oudtshoorn: Oudtshoorn, Rogers 4651. Pearston:

Wildebeest kuil, Hobson 60. Port Elizabeth: “Coegariver”,

Ecklon & Zeyher, Enum. No. 376 (S). Prieska: Vogelstruisbult,

Bryant J. 294. Prince Albert: Prince Albert Road, Bayliss 2443.

Richmond: Vlakplaats, H. Bolus 13773 (Z). Robertson: Robert-

son, Van Breda & Joubert 1923. Steytlerville: De Weg, Oliver

4579. Uniondale: Toorwater, Oliver 3657. Willowmore: S.W.

of Willowmore, Bayliss 4924. Worcester: Hex River Valley,

Marloth 6192.

O.F.S. — Fauresmith: Koksfontein, Henrici 2814. Tromps-

burg: Riet Poort Noord, Verdoorn 2198. Rouxville: Nieuwe-

jaarspruit, Ecklon & Zeyher s.n.

Characterized by the silvery, scaly pubescence

which forms a complete, smooth covering on the

stems in the lower parts and consists of strigose and

stellate hairs, the stellate hairs radiating from a

central silvery scale. In most upper parts the silvery

stellate hairs and scales are less dense and often

scattered, and minute reddish gland-tipped hairs are

present as well. The leaves, like those of several

other species, especially among species that were

formerly in the genus Mahernia, vary considerably in

the degree of lobing. In some specimens the leaves

are merely coarsely lobed while in others, or even on

the same plant, they can be deeply lobed. In this

species they vary from subentire to coarsely lobed to

pinnatipartite, bipinnatipartite or palmipartite.

Andrews, when describing this species, referred to

it as the “powdered Hermannia”.

89. Hermannia procumbens Cav., Diss. 329, t.

177, fig. 2 (1788). Type: Cape, “Groenekloof infra

Leeuwestaart”, Thunberg s.n. (MA, holo.; PRE,

photo.!; UPS, Herb. No. 15490, PRE, photo.!).

Suffrutex, stems one to many from a woody base,

decumbent, slender, sparingly branched at base and

sometimes with a few, short, ascending branchlets

produced in axils of leaves along the trailing stem,

new growth with appressed stellate pubescence,

glabrescent, upper internodes long. Stipules ovate-

oblong, ovate-lanceolate, semi-orbicular, 2-45 mm

long, 1,5-2, 5 mm broad, broadly acuminate to

abruptly narrowed into an acumen, occasionally

rounded at apex. Leaves usually solitary at a node,

upper distant, petiolate; blade 10-35 mm long,

varying in shape and degree of lobing, from oblong-

cuneate and coarsely lobed to bi- or tri-pinnatisect

and then fragile, fairly sparsely stellate-pubescent,

glabrescent; petiole 3-15 mm long. Inflorescence of

58

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

few-flowered, leafless, racemose cymes, terminal on

ascending branches and branchlets; flowers secund

and often crowded at apex; pedicels short, up to 5

mm long. Calyx 5-7 mm long, from broadly campa-

nulate (sub-inflated) and about half as long as petals

to narrowly campanulate (not at all inflated) and

more than half as long as petals, lobed to above the

middle, lobes either broad-based and rounded with an

acute apex, or deltoid to narrowly deltoid from a

narrow base, glabrescent, sometimes with minute

hairs persisting on margins and apex of lobes. Petals

strongly twisted, 8-1 1 mm long, 3-5 mm broad in

upper half, evenly narrowing into claw or broadest

on one side and sometimes auricled on one side, claw

about as long as upper portion, apparently glabrous

or rarely a few hairs on inrolled margins. Stamens

about 6,5 mm long, with narrow to fairly broad,

obovate-cuneate, hyaline, glabrous filaments about

4 mm long; anthers 3 mm long, ciliate. Ovary about

2,5 mm long, densely stellate-pubescent, 5-lobed;

stipe 4-5 mm long. Capsule up to 8 X 7 mm, 5-lobed,

stellate-pubescent, subglabrescent, with the persistent

calyx surrounding base.

Found on the Cape Peninsula and up the coast

from Malmesbury District to near Lamberts Bay in

the Clanwilliam District.

For key to subspecies see key to species (pp. 3-7).

(a) subsp. procumbens

De Winter in Bothalia 11,3: 264 (1974).

H. procumbens Cav., Diss. 6: 329, t. 177, fig. 2 (1788);

Thunb., Diss. Herm. 15 (1794); FI. Cap. 507 (1823); Eckl. &

Zeyh., Enum. 41, No. 329 (1834); Harv. in F.C. 1: 185 (1860).

Adamson in Adamson & Salter, FI. Cap. Penins. 584 (1950).

Type: Cape, “Groenekloof infra Leeuwestaart”, Thunberg

s.n. (MA, holo.; PRE, photo.!; UPS, Herb. No. 15490; PRE,

photo. !).

H. leucanthemoides Presl. Bot. Bemerk. 22 (1845). Syntypes:

Ecklon in Herb. U.I. 395 (PRE!; S ! ; W !); Drege pi. cap. b. spei.

H. zeyheriana Presl, Bot. Bemerk. 22 (1845); Walp., Ann.

1: 108, as zeyheri (1848-49). Type: Ecklon & Zeyher Enum.

No. 329 (S!; LE!; MO!; BOL!).

Characterized by the decumbent habit with one to

several slender stems from a woody base, the lower

leaves oblong-cuneate and fairly coarsely lobed, the

upper often narrowly oblong and lobed down to

half the breadth of the leaf or in specimens from

northern areas even deeper, but not to the midrib

(pinnatifid), the subsparse, appressed, stellate pubes-

cence of the young growth, the inflorescence of few-

flowered, leafless, racemose cymes terminal on the

ascending branches and branchlets, flowers secund

and often crowded at the apex, the broadly cam-

panulate (subinflated?), glabrescent calyx usually

about half as long as the petals and lobed to above

the middle and the lobes broad-based.

Recorded from the Cape Peninsula, around Table

Bay and one record from near Bok Point in the

Malmesbury District.

Cape. — Bellville: Melkbosstrand, Dahlstrand 1044; about

45 km N. of Cape Town on Mamre road, Lewis in SAM 65979.

Malmesbury: near Bok Point, Compton 9410 (NBG). Penin-

sula: Milnerton, Adamson 2605 (BOL); near Greenpoint,

Ecklon in Herb U.I. 395.

Presl, when describing H. leucanthemoides and H.

zeyheriana , stated that Drege 2316 (W!; LE!) and

7310 (W!; LE!) were the true H. procumbens Cav. He

then described Ecklon & Zeyher Enum. No. 329 as H.

zeyheriana Presl and Ecklon in Herb. U.I. 395 as H.

leucanthemoidea Presl. Several sheets in different

herbaria of Ecklon & Zeyher Enum. No. 329 (MO!;

BOL!; S!; LE!) and Ecklon in Herb. U.I. 395 (S ! ;

W!) have been seen and they are true H. procumbens

Cav. The Drege specimens seen by Presl are correctly

cited by Harv. in F.C. 1, No. 2316 (W; LE) under

Mahernia pulchella (L.f.) Cav. and No. 7310 (W; LE)

under Mahernia linearis Harv. Sheets of these Drege

numbers have been seen in W and LE and they are

not H. procumbens Cav.

Previously subspecies procumbens had only been

recorded from around Table Bav which is some

distance from the southernmost record of subsp.

myrrhifolia. But in 1940 a specimen of typical H.

procumbens, Compton 9410, was collected at the 7th

Gate to Bok Point, which lies in the heart of the

distribution area of subsp. myrrhifolia. This supports

the decision to regard these elements only as sub-

species. It may be due to cultivation or chance that

more specimens of both subspecies have not been

recorded from the intervening gap. In habit and

inflorescence the two subspecies are the same and

both show a tendency to turning blackish in parts on

pressed specimens. For the features in which they

differ see subsp. myrrhifolia.

(b) subsp. myrrhifolia {Thunb.) De Winter in

Bothalia 11:264 (1974). Type: Cape, Swartland,

Thunberg Cat. No. 15487 (UPS, holo.; PRE, photo.!).

H. myrrhifolia Thunb., Diss. Herm. 16 (1794); FI. Cap 508

(1825). Type: as above.

H. pinnatisecta Salter in JI S. Afr. Bot. 12: 102 (1946).

Type: Cape, Malmesbury, Mamre Hill, Compton 14928 (NBG,

holo.!); — var. auriculata Salter l.c. 103 (1946). Type: Cape,

Malmesbury, Ysterfontein, Compton 17374 (NBG, holo.!).

This subspecies differs from the typical subspecies

mainly in the leaves and calyx. The leaves are pinna-

tisect with pinnae pinnately lobed, some lobes

decurrent (they are never oblong and coarsely lobed

with upper narrow and pinnatifid as in subsp. pro-

cumbens). The calyx is usually narrowly campanulate

in subsp. myrrhifolia and slightly more than half the

length of petals whereas in subsp. procumbens it is

broadly campanulate and about half as long as petals.

This subspecies is on the whole more fragile than the

typical, especially the leaves; the inter nodes appear

longer and the tendency to turn blackish on pressing

is more apparent.

Found along the Cape west coast from Malmesbury

District to Clanwilliam. Recorded from the Swartland,

especially near Saldhana Bay and northwards to

Lambert’s Bay.

Cape. — Bellville: Monte Vista, Esterhuysen 32491a. Clan-

william: Van Putten’s Vlei, Gillett 4058; Zuurfontein, Schlechter

8549. Malmesbury: Swartland, Thunberg s.n. (S); Mamre Hill,

Compton 14928 (NBG); Ysterfontein, Compton 17374 (NBG);

near Darling, Schlechter 5339 (SAM).

Records of this subspecies are found in very few

herbaria. This is probably owing to the fragile nature

of the plant and to the extensive cultivation in the

Swartland, Malmesbury district, which is evidently

its main distribution area. It would seem that Thun-

berg’s species was lost sight of by most reviewers,

possibly because the name was misapplied by Ecklon

& Zeyher. Their No. 407, mentioned in the Enume-

ratio under the name Mahernia myrrhifolia (Thunb.)

Spreng., is not this species but H. diffusa {=H.

pilosula Harv.).

So it happened that Salter came to describe the

species H. pinnatisecta. His specimens exactly match

the type of H. myrrhifolia Thunb.

90. Hermannia confusa Salter in Jl S. Afr. Bot.

12:99 (1946); De Winter in Bothalia 11,3:263

(1974); Verdoorn in Flower. PI. Afr. 43, t. 1718

(1976). Type: Cape, between Caledon and Babylon’s

I. C. VERDOORN

59

Tower, Ecklon & Zeyher Enum. No. 374 (BOL,

holo. !; S!; LE!; SAM!).

H. tenuifolia sensu Eckl. & Zeyh., Enum. 47, No. 374 (1834);

sensu Harv. in F.C. 1 : 203 (1860), non Sims.

H. coronopifolia sensu Eckl. & Zeyh., Enum. 47 No. 377

(1834), non Link. (S ! ; SAM!).

Mahernia pinnata sensu Eckl. & Zeyh., Enum. 52: 408 (1834),

non L. (S !, poor specimen).

Suffrutex, erect, becoming much branched and

sprawling with age, up to 60 cm tall, branched from

base; branches and branchlets rather slender with

fairly sparse, appressed, stellate hairs and long,

pointed hairs spreading from a minute, bulbous base,

minute, gland-tipped hairs also sometimes present,

the long hairs caducous. Stipules 2-6 mm long, ovate

to narrowly ovate, often broad and oblique at base,

the auricles somewhat thickened. Leaves 10-25 mm

long, one or more at a node, pinnately lobed to

bi-pinnately lobed above, narrowing into a petiole-

like base up to 1 cm long, sparsely stellate-hairy.

Inflorescence of 1- to 2-flowered cymes arranged in

slender, leafless, lax, racemose cymes terminal on

branchlets, cymes 1 to 2 at a node; peduncles 1-4,5

cm long, sparsely stellate-pubescent, glabrous or

with minute, gland-tipped hairs, usually straight and

suberect, usually shorter than internodes; pedicels

short, 3-4 mm long, cernuous; bracts like stipules,

narrowly ovate to ovate with a broad attachment,

shallowly cordate or often oblique with one side

deeply cordate, auricles somewhat thickened, acute

or acuminate, glabrous or with a few long hairs which

fall readily; bracteoles small and often united at base,

1,5-2, 5 mm long. Calyx campanulate, lobed to

almost middle, about 6 mm long, glabrous; lobes

acute. Petals bright yellow turning red-orange with

age, strongly twisted, about 10 mm long, the sub-

orbicular upper half oblique at base where it narrows

into a long claw with infolded margins, minutely

pubescent in central area on inner face. Stamens about

6 mm long with obovate hyaline filaments, pubescent

on shoulders which are overlapped by anther bases.

Ovary about 3 mm long, 5-angled, stellate-hairy;

stipe up to 1 mm long; styles about 8 mm long.

Capsule 6-7 mm long, 5-lobed, shortly exserted from

the persistent calyx, stellate-pubescent at least on

sutures and at apex; styles fairly persistent.

Found on river flats and mountain slopes. Recorded

quite frequently from Caledon eastwards to Swellen-

dam and northwards through Robertson, Worcester

and Stellenbosch to Piketberg and in eastern Clan-

william.

Cape. — Caledon: Zwarteberg, Schlechter 5557; Hartebees-

rivier, Zeyher 2001; Ecklon & Zeyher Enum. 374 (LE; W).

Ceres: Ezelsfontein, Esterhuysen 20359. Clanwilliam: Berg-

valley, Oliver 3893. Piketberg: Piqueniers Kloof, Schlechter

4940. Robertson: near Warsbek, Van Breda & Joubert 2006.

Stellenbosch: Stellenbosch, Ecklon & Zeyher Enum. 408 (S).

Swellendam: Hesquaspoort, Acocks 22391. Tulbagh: Tulbagh

Road, Schlechter 8995. Worcester: Hex River, De Dooms,

H. Bolus 13076.

In the past this species has very generally been

named H. tenuifolia Sims. In 1946 Captain Salter

pointed out that the specimens cited by Harvey in

the Flora Capensis, together with a number of more

recently collected examples, differ from Sims’ species.

Sims states in the text of the Botanical Magazine t.

1348 (1811) that the plate had been prepared some

years before from a plant believed to have been in the

possession of the late Mr Curtis. No specimen was

preserved and, since it has not been exactly matched

with any known South African species, H. tenuifolia

Sims is now considered an insufficiently known

species.

H. confusa differs from its nearest relative, H.

procumbens subsp. myrrhifolia, in its erect habit and

its inflorescence of lax racemose cymes with long,

suberect peduncles which are usually somewhat

shorter than the internodes. It is also characterized by

the fairly shallow, glabrous calyx with the petals

exserted from the tube by more than half, often

showing the claw with its stellate hairs.

In his original description of H. confusa Salter did

not clearly indicate Ecklon & Zeyher Enum. No. 374

as the holotype. He did, however, mention this speci-

men first and added BOL in brackets after it. In the

rest of the work Salter always indicated in brackets

behind the first cited specimen that it is the type and

also the herbarium in which it is housed. It is thus

assumed that the omission of the word type is merely

an error.

91. Hermannia macra Schltr. in Bot. Jb. 55: 362

(1919); M. Friedrich et al. in F.S.W.A. 84: 16 (1969).

Syntypes: Namaqualand, I’us, Schlechter 11403

(PRE!; Z!; W!; S!; LE!); Keuzabies, 18/6/1898,

M. Schlechter s.n. or sometimes given as No. 106

(K!; PRE!; Z!; LE!; S!).

H. rhopalostylys K. Schum. & Schltr. ined., name on some

herbarium specimens of Schlechter 11403 and M. Schlechter 106.

H. paniculata E. Mey., nomen in Drege, Zwei Pfl. Doc. 191

(1843) (W!).

Suffrutex, stems several to many from a woody

base, sparingly branched near base; branches sparsely

to very sparsely and minutely appressed stellate-

pubescent, the hairs radiating from central scales or

glands, scattered minute papillae sometimes present,

glabrescent, sometimes central scales or glands

persistent. Stipules small, narrowly ovate to ovate-

deltoid, broad at base, 1,5-2 mm long, stellate-

pubescent and sparsely ciliate with long, pointed hairs,

glabrescent. Leaves mostly basal, sparse and reduced

above, petiolate; blade more or less oblong from

narrowly to broadly so, or ovate-oblong, coarsely

and unevenly lobed, sometimes deeply so but never

to midrib, lobes usually broad and obtuse, not or

occasionally lobed again, 3-nerved at base, nerves

suberect, appressed stellate-pubescent on both sur-

faces, glabrescent; petiole 5-25 mm long. Inflorescence

of 1- to 2-flowered cymes arranged in erect, terminal,

leafless, racemose or paniculate cymes; peduncles

20-30 mm long, glabrescent; bracteoles like small

stipules, usually 3; pedicles 2,5-6 mm long, stellate-

pubescent with minute, scattered glands or papillae.

Calyx campanulate, about 4,5 mm long, lobed to

almost middle, teeth deltoid and sinuses wide, stellate-

pubescent and with minute, scattered scales or

papillae, glabrescent in upper half only, lobes ciliate

with slightly longer stellate hairs. Petals yellow to

orange-red, about 5,5 mm long, suborbicular in

upper half, usually broader than long, narrowed into

a fairly broad claw with inrolled margins, appears

glabrous but sometimes with minute hairs. Stamens

about 5 mm long, with obovate, hyaline filaments;

anthers almost as long as filaments, overlapping

them at base, acute, ciliate. Ovary about 2,5 mm long,

stellate-pubescent on sutures, papillate between;

stipe about 0,5 mm long; styles 3 mm long, persistent,

often appearing subclavate at apex. Capsule about 6

mm long and almost as broad, shallowly 5-lobed,

stellate pubescence sometimes persisting on sutures,

papillose between hairs, persistent styles still united

and appear sucblavate at apex (hence the abandoned

manuscript name rhopalostylys).

60

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

Found on sandy flats and in dry river beds in the

drainage basin of the lower Orange River. Recorded

from Namaqualand and in South West Africa in the

Warmbad and Liideritz Districts.

Cape. — Namaqualand: I’us, Schlechter 11403; Keuzabies,

M. Schlechter 106; S. of Goodhouse, Schlieben 9100.

S.W.A. — Liideritz: Sendelingsdrift, Merxmiiller & Giess 3261.

Warmbad: Sperlingspiitz, Giess, Volk & Bleissner 7012;

Witpiits, Merxmiiller & Giess 3635.

Closely related to H. paucifolia and differs mainly

in the shape and lobing of the leaf and size of the

flowers. In H. macra the flowers are small, the majority

up to 6 mm long on dried specimens, whereas in H.

paucifolia they are up to 10 mm long. The leaves in

H. macra are more or less oblong, from narrowly to

broadly so, and coarsely and unevenly lobed, rarely

deeply lobed but not to the midrib, whereas in H.

paucifolia the leaves are ovate and the lobing is usually

to the midrib so that most of the lower leaves are

palmatisect. Both these species are characterized by

being fairly small, 30-50 cm tall, with several stems

from a woody base and with leaves mostly basal and

an inflorescence of practically leafless, glabrous,

racemose or paniculate cymes. The styles which

cohere and persist, appear subclavate and this

probably accounts for the unpublished specific name

H. rhopalostylys. The name “ macra ”, the “Meagre

Hermannia” was chosen instead when the species was

published in 1919.

92. Hermannia juttae Dinter & Engl, in Bot. Jb.

55: 356 (1919); M. Friedrich et al. in F. S.W.A. 84: 16

(1969). Type: South West Africa, Bullspoort, Reho-

both, Dinter 2098 (SAM !).

H. rehobothensis M. H-Friedrich in Mitt. bot. StSamml.,

Miinch. 1 : 346 (1953); M. H-Friedrich et al. in F. S.W.A. 84: 16

(1969), in syn. Type: South West Africa, Rehoboth, Volk 2520.

Subherbaceous, stems several to many from a

woody base, erect, sparingly branched just above

base, densely stellate-pubescent at base with long

white hairs, branches sparsely and inconspicuously

stellate-pubescent, often with scattered, very minute

glands, especially in upper portion. Stipules ovate-

lanceolate to linear-lanceolate, somewhat acinaciform,

5-20 mm long, stellate-pubescent. Leaves petiolate,

mostly basal with 1 or 2 distant leaves on lower

half of branches; blade oblong-lanceolate to narrowly

lanceolate, coarsely crenate or toothed, 20-60 mm

long, 8-15 mm broad, occasionally luxuriant plants

with larger leaves ( Giess 13555), stellate-pubescent,

at first densely so with long matted white hairs, later

sparsely so with shorter, appressed hairs, nerves

prominent below, usually 3 from base; petiole 7-27

mm long. Inflorescence of 1- to 2-flowered cymes,

arranged in erect, leafless, lax, terminal, racemose

cymes; cymes 1 to 2 at a node; peduncles suberect,

15-35 mm long, appearing glabrous but usually with

scattered, minute, subsessile glands; bracts like

stipules, though somewhat smaller, bracteoles often

3, 1,5-5 mm long; pedicles slender, 5-15 mm long.

Calyx 8-9 mm long, campunulate, wide at mouth, thin

becoming papery, appears glabrous but has minute

scattered glands and occasionally a few stellate hairs,

toothed in upper half, sinuses wide. Petals golden

yellow becoming orange-red with age, 8-9 mm long,

oblong-oibicular in upper half, narrowing into a

fairly broad claw with narrowly infolded margins,

glabrous in specimens dissected. Stamens 8 mm long

with obovate-oblong, hyaline filaments; anthers

acute, ciliate, slightly longer than filaments. Capsule

about 8 mm long, enclosed in faded petals and calyx,

5-lobed, stellate-pubescent; style persistent.

Found “on the flats”, in deep sandy soil, “among

Acacias”, “among Mesembryanthemum”. Recorded

from the Rehoboth and Maltahohe Districts in South

West Africa.

S.W.A. — Maltahohe: form Schwarzkuppe, Giess, Volk &

Bleissner 5215. Rehoboth: Bullspoort, Dinter 2098 (SAM);

Strey 2085; Liebenberg 5106; Dinter 8321 (S; PRE); Jorro

Flats, Strey 2339; Naukluft, Hardy 1960.

A low, subherbaceous plant with many short,

erect, thin, pubescent stems from a woody base,

sparsely branched just above the base. Characterized

by the basal leaves with only one or two above the

base in the lower half of the plant, and the leafless

upper half which bears the comparatively large flowers

in lax racemose cymes, the calyx, which is campanu-

late, with a wide mouth and appears glabrous although

minute glands and occasional stellate hairs may be

present.

93. Hermannia paucifolia Turcz. in Byull. mosk.

Obshch. 31: 218 (1858); Harv. in F.C. 1 : 203 (1860);

M. Friedrich et al in F. S.W.A. 84: 19 (1969). Type:

Cape, Bitterfontein, Zeyher 118 (K; PRE, photo.!;

PRE; Z!; LE ! ; and in W! as “ Mahernia bipinnata

L.”; in S! as “ H . dissecta Harv.”).

H. chrysanthemifolia E. Mey. ex Harv. in F.C. 1: 204 (1860).

Type: Namaqualand, Kaus Mts, Drege s.n. (W!; LE!).

H. paucifolia var. intermedia Kuntze ex K. Schum. in Verh.

bot. Ver. Prov. Brandenb. 30:233 (1888); in Engl., Monogr.

Afr. Pfl. 5:55 (1900). Type: no type designated but several

specimens cited in 1900, among them Schenck 342 (Z!) and 124

(Z ! ; PRE!). — var. chrysanthemifolia (E. Mey. ex Harv.) Kuntze

ex K. Schum., l.c. (1888); in Engl. l.c. (1900).

H. dissecta Harv. ms. in syn. in F.C. 1:204 (1860). Type:

as for H. paucifolia Turcz.

Suffrutex, stems several to many, bushy, from a

woody base, branched at base and sparingly above,

branches glabrous or with a few hairs or minute

glandular hairs, rarely finely stellate-pubescent. Sti-

pules small, broadly ovate to ovate-lanceolate, 1-3,5

mm long, glabrous or sublaxly pubescent, some hairs

on margins long. Leaves mostly basal with a few

distant and somewhat reduced leaves above, petio-

late; blade 15-30 mm long, usually palmatisect with

segments pinnately and unevenly once or twice lobed,

lobes mostly rounded at apex, leaves rarely ovate and

palmately nerved with margins shallowly to deeply

lobed, upper and lower surface glabrous or sparsely

to densely stellate-pubescent with minute stellate

hairs, upper surface sometimes glabrescent; petiole

10-35 mm long. Inflorescence of 1- to 2-flowered

racemose or paniculate cymes, terminal on main

branches and on a few lateral branches which arise

from distant upper, much reduced leaves; peduncles

suberect, up to 16 mm long; bracts short and broad,

often united at base, about 1 mm long; pedicels 2-3

mm long, often minutely glandular-pubescent. Calyx

campanulate, wide at the mouth, about 6,5 mm long,

lobed almost to middle, minutely stellate-pubescent,

glabrescent with only a few minute glandular scales

at base. Petals yellow to orange, at some stages red,

about 10 mm long, the upper third oblong-orbicular,

narrowing into a long claw with inrolled margins,

rounded or slightly lobed where blade narrows into

claw, appears glabrous but sometimes minute stellate

pubescence and papillae obvious on inner face of

claw. Stamens united at base around stipe, about 7

mm long with oblong hyaline filaments narrowing

slightly to base; anthers ciliate 4 mm long, about as

I. C. VERDOORN

61

long as filaments and shortly overlapping them at

base. Ovary about 4 mm long, entirely stellate-pubes-

cent or only on sutures, papillate between, sutures

narrowing slightly to base; stipe 1 mm long; styles

about 5 mm long, exserted and persistent, appearing

slightly clavate at apex. Capsule 7-10 mm long,

minutely papillose between sutures, sometimes stellate

on sutures and at apex, persistent calyx and stamens

at base obscuring stipe, persistent styles at apex.

Found in the western mountainous karoo, the arid

upper karoo, Namaqualand and south-western South

West Africa. Recorded from Victoria West westwards

to Calvinia and northwards to Namaqualand and

the Ltideritz District in South West Africa.

Cape. — Calvinia: Hantam, Marloth 5634; on road to Wil-

liston, Story 4270; Brandvlei, Comins 679. Namaqualand:

Oograbies Poort, H. Bolus in Herb. Nor. Austr.-Afric. 440;

Kaus Mts, Drege s.n. (LE; W). Vanrhynsdorp: Bitterfontein,

Zeyher 118. Victoria West: Hutchinson, Acocks 9631. Williston:

Matjiesfontein, Foley 160; Victoria West station, Smith 2452.

S.W.A. — Liideritz: Aus, Dinter 6072; Marloth 5079; between

Tsirub and Grasspforte, Schenck 124 (PRE; Z); Gobaxab,

between Aus and the Orange River, Schenck 342 (Z); Kling-

hardt Mts, Dinter 3884.

Characterized by the mainly basal leaves, ovate in

outline and often deeply palmatisect, and the almost

leafless, terminal, paniculate cymes. Closely related

to H. macra. For distinguishing features see under

that species. The distribution of these two species

overlap and some specimens appear to be intermediate.

For the present the treatment in F. S.W.A. is followed

here and the two species are kept separate. Dinter

3884, which in F.S.W.A. is doubtfully included

in this species because of the shallowly lobed, densely

stellate leaves, matches a specimen from Grootderm

collected by Pillans (No. 5324), with the note “luxu-

riant growth result of growing in deep sand in a

watercourse”. Among the specimens examined these

two gatherings are linked to the palmatisect leaves by

Schenck 342 (Z) which has intermediate leaves, some

shallowly, and others on the plant fairly deeply

lobed. Dinter 3884 and PHlans 5324 are therefore

here included in H. paucifolia.

The type gathering, Zeyher 118, is represented in

several herbaria. All those seen are poor specimens.

It is not known which, if any, of these specimens was

seen by Turczaninow.

Drege specimens in LE and W which are this

species have, evidently in error, the specific name

written as “ chrysanthemoides" on the label.

According to Veterinary Services, Onderstepoort’

these plants cause excessive purging in animals-

Common names are Skitterybossie, Purgurbossie,

Scholtzbossie and Governmentsbossie. In the Agri-

cultural Journal (1893), it is claimed that it is used to

cure Bushman of a craving for drink.

SPECIES REQUIRING FURTHER ATTENTION

1. Hermannia bracteosa Presl, Bot. Bemerk. 21 (1844).

The sheet of Drege 7267 in PR comprises two

elements. The specimen on the left is H. mucronulata

Turcz., and from the description in Bot. Bemerk. it

is clear that this is the specimen that Presl referred to

as H. velutina Eckl. & Zeyh. The specimen on the

right which is considered to be the holotype of H.

bracteosa Presl, does not agree with any species

known to me. The leaves are velvety as in H. velutina ,

but the inflorescences matches H. salviifolia var.

grandistipula Harv., which has coarsely stellate

leaves. This seems to point to the holotype being of

hybrid origin.

2. Hermannia incisa Willd., Sp. PI. 3: 599 (1800).

The type specimen of this species is in the Botanical

Museum, Berlin-Dahlem. A photograph of the type

specimen was sent to this Institute and later a flower

and upper leaf was sent on loan. It appeared that in

all probability the species was conspecific with H.

procumbens Cav. and matched fairly well Lewis in

SAM 65979 under H. procumbens subsp. procumbens.

However, it seems wise to investigate further, especial-

ly since the procumbent habit is not mentioned and

cannot be judged from the specimen. The sign on the

specimen indicated that it was woody but not neces-

sarily procumbent.

3. Hermannia myrioclada Diels in Bot. Jb. 55: 359

(1919).

Type: Cape, Cedarberg, Diels 862 (B, holo.f).

From the description this species may be synonymous

with H. scabra.

4. Hermannia polymorpha Eckl. & Zeyh.

The specimen figured in Refug. Bot. 3: 195 together

with a specimen in LE labelled “Ex horto bot.

Petropolitana 66.5” and named H. polymorpha

Eckl. & Zeyh. were most puzzling. In 1975 a specimen

from Kew was brought to my notice. It matched the

above and was said to have appeared spontaneously

in the Royal Botanic Gardens, Melbourne, Australia.

That means no original locality was known for any

of these and since they did not match any of the

thousands of southern African species examined, it

seemed probable that South Africa was not the

country of origin. In 1976 a specimen from the Bolus

Herbarium, that had been overlooked when the

collection from that herbarium was sent on loan to

this Institute, was received here. It matches the

above specimens and is labelled Gillett 1452 from

“Avontuur, edge of Plateau, above Bidou River,

Knysna”. It may be a good species near H. angularis

or a sport or form of that species.

5. Hermannia urceolata Pillans MS.

Pillans evidently originally intended naming this

species H. hollandii , but the name is scored out in his

manuscript and “urceolata” written in its place. The

specimens cited are as follows: “Port Elizabeth

Drive, Cape Road, St. Alban’s Farm, Holland 4062

(type in Bolus Herb.); 15th Mile Cape Road, Long

5056; without precise locality, Cruden 465; Bolus

3055; Baakens River Valley, Kensit 2”. These sheets

have been seen in the Bolus Herbarium. The number

of Long’s specimen is mistakenly printed as “5056”,

which is the genus number. The correct number is

Long 1302. Cruden 465 and Kensit 2 have been found

to be Hermannia suavis. Although Pillans’s MS

species resembles both H. suavis and H. salviifolia var.

grandistipula in the urceolate calyx with rather long,

white, stellate hairs and in the general leaf shape

and size of stipules, it differs from both in the distinct

stellate pubescence on the lower surface, with the

hairs long, whitish and multicellular, giving the leaves

a fringed look around the margin when viewed from

above. To date, field observations have not been

carried out in order to establish whether it is a

distinct species.

62

REVISION OF HERMANNIA SUBGENUS HERMANNIA IN SOUTHERN AFRICA

ACKNOWLEDGEMENTS

I am indebted to Dr B. de Winter, Director,

Botanical Research Institute, for providing me with

working facilities at the Institute and for his kindly

interest in my work, and to Dr O. A. Leistner for

criticizing the text. For the unfailing kindness and help

by other members of the staff, I am truly grateful.

UITTREKSEL

Waarnemings van die genus as geheel word gedoen

en die subgenus Hermannia word omskrywe. Spesies in

Suidelike Afrika wat onder hierdie subgenus sorteer

word in besonderhede hersien. 'n Sleutel tot hierdie 93

spesies word verskaf en 'n aantal pentekeninge illustreer

sommige van die diagnostiese kenmerke wat in die

sleutel gebruik word.

INDEX

Page

Hermannia 1

subgen. Acicarpus Harv 1

subgen. Euhermannia Harv 1

subgen. Hermannia 1

abrotanoides Schrad. 56

affinis K. Schum 20

alnifolia L 44

althaeifolia L 29

althaeoides Link 28

amabilis Marloth ex K. Schum 15

amoena Dinter ex M. Holzhammer-Friedrich 33

angolensis K. Schum 10

angularis Jacq 35

angularis sensu Eckl. & Zeyh 35

arabica Hochst. & Steud. ex Fisch 16

argentea Sm 57

argyrata Presl 22

asbestina Schltr 21

aspera Wendt 47

aspericaulis Dinter & Engl 21

atrosanguinea Dinter 16

ausana Dinter ex Range 31

bicornis Eckl. & Zeyh 20

bipinnata Burch 56

bolusii Szyszyl 57

boranginiflora Hook 11

brachypetala Harv 9

bracteosa Presl 61

brandtii Engl, ex Dinter 10

bryonifolia sensu Eckl. & Zeyh 33

bryoniifolia Burch 26

cana K. Schum 57

candicans Ait 28

var. discolor Harv 31

var. incana Harv 31

candicans sensu Harv 28

candissima Spreng. f. 32

cavanillesiana Eckl. & Zeyh 37

chrysanthemifolia E. Mey. ex Harv 60

chrysophylla Eckl. & Zeyh 36, 49

collina Eckl. & Zeyh 22

comosa Burch, ex DC 31

var. crenata K. Schum 31

var. minor K. Schum 31

complicata Engl 19

concinnifolia Verdoorn 25

confusa Salter 58

conglomerata Eckl. & Zeyh 48

cordifolia Harv 27

coronopifolia sensu Eckl. & Zeyh 59

cristata H. Bo! 7

var. geoides Beauv 7

cuneifolia Jacq 40

var. cuneifolia 41

var. glabrescens (Harv.) Verdoorn 41

cuneifolia sensu Harv 43

damarana Bak. f. 18

decipiens E. Mey. ex Harv 49

decumbens Willd. ex Spreng 22

var. argyrata (Presl) Harv 22

var. collina (Eckl. & Zeyh.) Harv 22

var. hispida Harv 22

denudata L. f. 52

var. denudata 52

denudata sensu Eckl. & Zeyh 54

desertorum Eckl. & Zeyh 42

dinteri Engl 14

dinteri Schinz 31

discolor Otto & Dietr 28

disermifolia Jacq 33

disermifolia sensu Eckl. & Zeyh 22

dissect a Harv 60

disticha Schrad. 48

Page

diversistipula Presl ex Harv 39

var. diversistipula 40

var. gracilifolia Verdoorn 40

dryadiphylla (Eckl. & Zeyh.) Druce 44

dryadiphylla (Eckl. & Zeyh.) Harv 44

eenii Bak. f 10

engleri Schinz 14

ernesti-ruschii Dinter ex M. Holzhammer-Friedrich 33

exstipulata E. Mey 18

falcata Eckl. & Zeyh 53

fasciculata Bak 17

filifolia L.f. 51

var. filifolia 51

var. grandicalyx Verdoorn 51

var passerinoides Harv 51

var. robusta Verdoorn 52

filipes Harv 16

var. eliator K. Schum 16

flammea Jacq 53

flammula Harv 53

floribunda Harv 26

floribunda sensu K. Schum 29

fruticulosa K. Schum 21

gariepina Eckl. & Zeyh 18

var. dentata Engl 18

var. integrifolia Engl 18

gilfillaniil N.E. Br 51

glabripetaa Engl 31

glanduligera K. Schum 11

glandulosissima Engl 12

glomerata E. Mey 48

gracilis Eckl. & Zeyh 39

grisea Schinz 13

guerkeana K. Schum 14

halicacaba DC 56

helianthemum K. Schum 18

helicoidea Verdoorn 46

hereroensis Schinz 14

hilaris (Eckl. & Zeyh.) Hochr 20

hirsuta Mill 29

hirsuta Schrad., non Cav 47

hirsuta Schrad. & Wendl 45

hirsuta sensu Eckl. & Zeyh 23

hispidula Reichb 27

holosericea Jacq 37

holubii Burtt Davy 16

hyssopifolia L 35

var. integerrima Schinz 36

imbricata Eckl. & Zeyh 25

incana Cav 31

incana sensu Thunb 37

incisa Willd 61

intricata Adamson 48

involucrata Cav 49

involucrata sensu Eckl. & Zeyh 35

johannisburgiana Engl 9

johanssenii N.E.Br 30

joubertiana Harv 54

juttae Dinter & Engl 60

karakowisensis ined 10

kirkii Mast 16

lancifolia Szyszyl 10

latifolia Jacq 36

lavaidulifolia L 37

lepidota Buch. ex Krauss 41

leucanthermoides Presl 58

leucophylla Presl 37

lindequistii Engl 18

linearifolia Harv 17

linifolia Burm. f. 22

linifolia sensu Eckl. & Zeyh 51

longiramosa Engl 10

lugardii N.E. Br 16

I. C. VERDOORN

63

Page

macra Schltr / (

melissifolia Engl 26

membraniflora Schltr 41

merxmuelleri M. Friedrich 8

micans Schrad 36

micrantha Adamson 48

micropetala Harv 15

mildbraedii Dinter & Engl 17

minimifolia M. Holzhammer 14

minutiflora Engl 29

modesta ( Ehrenb .) Mast 16

var. elatior (K. Schum.) K. Schum 16

subvar. brevicornis Engl 16

subvar. macropetala Engl 16

subvar. mediipetala Engl 16

subvar. virgatissima Engl 16

var. tsumebensis Engl 16

mollis Willd 31

mollis sensu Eckl. & Zeyh 28

mucronulata Turcz 34

multifida DC 56

multiflora Jacq 43

muricata Eckl. & Zeyh 44

muirii Pillans 25

myrioclada Diels 61

myrrhifolia Thunb 58

nemorosa Eckl. & Zeyh 28

nivea Schinz 18

odorata Ait 38

oligantha Salter 46

orophila Eckl. & Zeyh 53

orophila sensu Eckl. & Zeyh 56

paniculata E. Mey 59

pollens Eckl. & Zeyh. var. glabrescens Harv 41

passerinaeformis Eckl. & Zeyh 51

pateillicalyx Engl 42

patula Harv 54

paucifolia Turcz 60

var. chrysanthemifolia (E. Mey. ex Harv.) Kuntzf

ex K. Schum 60

var. intermedia Kuntze ex K. Schum 60

pedunculata Phill 54

pfeilii K. Schum 42

phaulochroa K. Schum 15

pillansii Compton 50

pinnatisecta Salter 58

var. auriculata Salter 58

plicata Ait 29

plicata sensu Eckl. & Zeyh 28

polymorpha Eckl. & Zeyh 53, 61

praemorsa Wendl 31

pratensis Eckl. & Zeyh 22

pres liana Turcz 54

prismatocarpa E. Mey. ex Harv 23

procumbens Cav 57

subsp. procumbens 58

subsp. myrrhifolia {Thunb.) De IVint 58

pseudo-mildbraedii Dinter & Engl 17

pulverata Andr 57

racemosa E. Mey 18

rehmannii Szyszyl 26

rehobothensis M. Holzhammer-Friedrich 60

repetenda Verdoorn 45

rhopalostylys K. Schum. & Schltr 59

Page

rigida Harv 45

rotundifolia Jacq 49

rudis N.E.Br 55

rudis var. exserta N.E.Br 56

rugosa Adamson 28

salvifolia sensu Eckl. & Zeyh 34

salviifolia L.f. 35

var. salviifolia 36

var. ovalis Harv 36

var. grandistipula Harv 36

var. oblonga Harv 35

salviifolia sensu Cav 49

secundiflora Eckl. & Zeyh 41

sandersonii Harv 8

scabra Cav 54

scabra sensu Jacq 47

scoparia (Eckl. & Zeyh.) Harv 22

scordifolia Jacq 24

var. integriuscula Harv 24

scordiolia sensu Eckl. & Zeyh 54

seineri Engl 9

var. latifolia Engl ]o

seitziana Engl 14

sideritifolia Engl 13

solaniflora K. Schum 13

spinosa E. Mey. ex Harv 21

spinulosa Engl 20

squarrosa Dinter ex Range 20

stipulacea Lehm. ex Eckl. & Zeyh 50

suavis Pres l ex Harv 35

sulcata Harv 38

tenella Dinter & Schinz 17

tenuifolia sensu Eckl. & Zeyh 59

tenuipes Engl 15

tephrocarpa K. Schum 11

ternifolia Presl 24

ternifolia Presl ex Harv 24

tigrensis Hochst. ex A. Rich. (“ tigreensis ”) 16

tomentosa (Turcz.) Schinz ex Engl 9

var. brevifolia Engl 9

trifoliata L 24

trifurca L 20

trifurca sensu Eckl. & Zeyh 53, 55

truncata Schinz 18

urceolata Pillans MS 61

velutina DC 38

vestita Thunb 32

viscosa Hiern 12

viscosa sensu Burtt Davy 11

windhukiana Engl 20

zeyheriana Presl 58

Mahernia L 1

subgen. Mahernia (L.) K. Schum 1

dryadiphylla Eckl. & Zeyh 44

hilaris Eckl. & Zeyh 20

incana Eckl. & Zeyh 20

odorata Andr 38

odorata sensu Eckl. & Zeyh 38

pinnata sensu Eckl. & Zeyh 59

scoparia Eckl. & Zeyh 22

var. glabra Eckl. & Zeyh 22

spinosa Burch, ex DC 21

tomentosa Turcz 9

Tricanthera modesta Ehrenb 16

Bothalia 13,1 & 2: 65-93 (1980)

Studies in the Ericoideae. III. The genus Grisebachia

E. G. H. OLIVER*

ABSTRACT

A revision of the genus Grisebachia Klotzsch in which eight species are recognized is presented. The genus

belongs to the Ericaceae-Ericoideae and is endemic in the south-western part of the Cape Province. The work

revealed a high degree of variability among the species, necessitating the reduction of seven species to infraspe-

cific rank, seven species to synonomy and the rejection of one species as imperfectly known. One new species G.

secundiflora E. G. H. Oliver is described.

rEsumE

Etudes sur les ericoideae. iii. le genre grisebachia

On presente une revision du genre Grisebachia Klotzsch dans lequel huit especes sont reconnues. Ce genre

appartient aux Ericaceae-Ericoideae et est endemique dans le sud-ouest de la province du Cap. Le travail a mis en

evidence un degre eleve de variabilite dans les especes, ce qui a entraine la reduction de sept especes au rang

infraspecifique; sept autres ont ete mises en synonymie et une a du etre rejet ee parce qu 'imparfaitement connue.

On decrit une nouvelle espece, G. secundiflora E. G. H. Oliver.

HISTORICAL OUTLINE

When Klotzsch reclassified the subfamily Ericoi-

deae in 1838, he described the genus Grisebachia

consisting of eight species: G. ciliaris sensu Klotzsch,

G. hispida Klotzsch, G. involuta Klotzsch, G. zeyhe-

riana Klotzsch, G. incana (Bartl.) Klotzsch, G. hirta

Klotzsch and G. plumosa Klotzsch. Of these, only

G. incana and G. plumosa are retained in the present

revision. Later in the same work he described Eremia

parviflora, which is now recognized as a species of

Grisebachia.

The following year Bentham (1839), when revising

the whole family Ericaceae, retained Grisebachia but

in an enlaiged form. He upheld Klotzsch’s eight

species and added G. dregeana and G. serrulata, both

of which have now been reduced to synonomy. He

also included Finkea Klotzsch with two species, as a

section. N. E. Brown (1906) correctly placed the

latter in the genus Acrostemon of Klotzsch.

In 1876 Bentham again revised the family and

included a further two of Klotzsch’s genera, Acros-

temon and Comocephalus, under Grisebachia. The

genus was then divided up into three sections based on

the shape of the corolla, the hairiness of the filaments

and the ovary complement.

In Die NatUrlichen Pflanzenfamilien Drude (1897)

took a very conservative view of the Ericoideae and

placed Grisebachia as defined by Bentham as one of

four sections in the genus Eremia D.Don.

N. E. Brown (1906) in Flora Capensis changed the

system of the earlier workers and adopted Grisebachia

as originally construed by Klotzsch. He retained

Acrostemon as a distinct genus and reduced Como-

cephalus and Finkea to synonomy under it. He

retained all eight of Klotzsch’s species and placed

Eremia parviflora correctly in Grisebachia, but as G.

eremioides which had been named by MacOwan in

1890. He also added ten of his own species of which

only three are upheld as distinct species in the present

revision, namely G. rigida, G. nivenii and G. minuti-

flora.

Phillips (1926) accepted Brown’s work in its

entirety in the first edition of his Genera. In 1944 he

put forward his proposals for a reclassification of the

family in South Africa for the second edition of his

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

Genera (1951). He presumably based his ideas on the

very conservative views of Drude. He recircumscribed

all of the genera and placed Grisebachia under

Eremia together with seven other genera, some of

which are quite unrelated.

The genus has been retained in Dyer’s Genera

(Oliver 1975) in the same form as adopted by N. E.

Brown in Flora Capensis.

When the present revision was undertaken the

genus Grisebachia consisted of 21 species. As a result

of finding numerous variations and overlapping of

characters the number of species has been reduced to

seven. One new species, G. secundiflora, has been

added.

MORPHOLOGY

In habit most of the species of Grisebachia are

erect, often forming compact shrublets. G. parviflora

is usually sparse and spreading among rocks and

vegetation and G. secundiflora is compact but rather

sprawling.

The branches of all species are never entirely

glabrous. Most have pubescent to pilose or tomentose

branches when young, sometimes with simple to

plumose stout hairs intermingled. These may be

gland- tipped.

The leaves are all typically ericoid with no open-

backed forms and are mostly 3-nate. In G. plumosa

subsp. hispida they are always 4-nate, while in subsp.

pentheri they can be occasionally 4-nate inbetween

3-nate. Most leaves are adpressed with one exception,

G. ciliaris subsp. multiglandulosa, where they are

recurved spreading. The indumentum of the leaves is

very variable and disjunctions have been used for

taxonomic division. Stout simple or plumose hairs

may or may not be present on the leaves and may be

confined to the margins or occur on the adaxial

surface as well. In many cases these stout hairs may

fall off and remain only as short stubs, which can

easily be overlooked.

The flowers of all species are terminal either at the

ends of the main branches or more often at the ends

of short lateral branchlets. In G. parviflora these

short branchlets may be aggregated together to form

a loose pseudospike and in G. secundiflora the pseudo-

spike is compacted and secund. The 3-bracteolate

flowers are usually 4—12 in a head or as much as oo

in G. minutiflora.

66

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Fig. 1. — Variation in the size of

the median bracteole from

outer to inner flowers in a

single inflorescence of

Grisebachia plumosa subsp.

plumosa. Drawn xl6 from

Thompson 791 (STE).

The bracteoles vary considerably in shape and size,

particularly in G. plumosa and G. ciliaris where they

are of taxonomic importance. The bracteoles may be

equal to very unequal with the median bracteole

being considerably enlarged with an expanded base.

The variation also occurs within a single inflorescence

where the bracteoles may be very unequal in the outer

flowers to equal in the inner flowers (Fig. 1).

The calyx in all species is 4-lobed or 4-partite. In

G. ciliaris, G. incana, G. rigida and G. nivenii the

sepals are free or very slightly joined at the base,

whereas in G. plumosa, G. parviflora, G. minutiflora

and G. secundiflora they are joined for quarter to

three quarters of their length. In all species the sepals

are more or less equal. Sometimes the lateral pair

may be slightly narrower than the ab- and adaxial

pair. The indumentum of the calyx is very variable

and is used as a taxonomic character. In most species

there are some stout simple to plumose hairs on the

margins of the sepals and sometimes also on the

abaxial surface. These may be gland-tipped in dif-

ferent taxa or in the same taxon in the young stages.

The corolla is of two basic shapes in the genus.

Five of the species, G. ciliaris, G. plumosa, G. rigida,

G. incana and G. nivenii, have a corolla similar to

that in the section Cyatholoma of the genus Erica.

The corolla tube is more or less distinctly con-

stricted in the middle with an ovoid or obovoid base

and cyathiform upper portion including the large

lobes. This is evident in the fresh state and in most of

the species in the pressed material. But care must be

taken with material of G. ciliaris subsp. ciliaris in

which the small flowers easily loose this shape when

pressed. Occasionally, when the constriction is not

very marked, a campanulate shape occurs.

In the three remaining species, G. parviflora, G.

minutiflora and G. secundiflora, the corolla has no

distinct constriction. In the first two species it is

usually funnel-shaped or obconic and in the third it

is tubular or tubular with an inflated middle poition.

In the majority of species especially those with the

constricted corolla the corolla is pubescent to pilose

in the middle region outside and also inside around

the point of constriction.

The number of stamens in all specimens examined

was constantly 4 and is important in the generic

classification. The stamens are free and have pilose

filaments.

The anthers are mostly manifest being arranged

just above the constriction in the corolla. In G.

secundiflora they are manifest to included. The

majority of anthers are characteristically bipartite.

In G. secundiflora t.iere is a tendency foi them to be

bilobed. Awns are present in several species, but may

be absent in anthers of the same flower. This charac-

ter, used in the past for specific recognition, is of no

use taxonomically. The anthers are all dorsally

attached with an expanded apex to the filament.

The pollen in all the material examined occurs as

single grains, which are found in several of the other

minor genera of the Ericoideae. The grains are

tricolporate, the furrows being deeply chanelled and

almost as long as the cell. In shape the grains are

mostly ellipsoid with flattened apices. In a few cases

they are oblate as in G. nivenii and some forms of G.

plumosa and G. parviflora. The sculpturing of the

surface in the first six species is either scabrate or

microscabrate with a tendency for the element rods

to become fused at their distal ends to form tecta. In

the last two species the fusion is complete giving an

almost smooth appearance to the pollen surface

(Fig. 2).

The ovary is mostly 2-celled with a single pendulous

subapical ovule in each cell. Very occasionally 3-celled

ovaries occur, notably in G. parviflora subsp. pubes-

cens. In G. secundiflora the ovary is constantly slightly

obliquely 1 -celled.

Mature fruits that were found in a few species

were hard-walled nuts with the walls often verrucose.

They are apparently indehiscent and contain one or

two very soft juicy seeds. Some fruits on Levyns 1367

{G. ciliaris subsp. multiglandulosa) collected in 1925

still contained soft juicy seeds.

As in most genera of the Ericoideae, the ovary in

Grisebachia is seated on a nectariferous disc which,

in some cases, is very conspicuous. This suggests

that all the species are insect pollinated. On a few

occasions it was noted in the field that plants were

visited by bees.

The stigma varies from simple to capitellate, which

is in accordance with the insect pollination.

DELIMITATION OF THE GENUS

As defined in the present revision the genus Grise-

bachia is characterized by having 3 bracteoles, 4

sepals, which are free or partly fused, a 4-lobed

corolla, 4 free stamens with bipartite anthers and an

ovary with 2, rarely 3 or 1, cells and a single ovule in

each cell. The important characters arfe the stamen

number and bipartite anthers.

The uniformity of the genus has until now been

recorded as very constant. It has been relatively

easy to assign material to the genus when identifying

Ericaceae. The 2-celled ovary, 4 stamens and bipartite

anthers served to be a distinctive combination of

characters.

N. E. Brown recorded occasional 3-celled ovaries

in G. solivaga and G. nodiflora both of which are

E. G. H. OLIVER

67

7

Fig. 2.— Pollen in the genus Grisebachia: scanning electron micrographs of pollen of various species showing the range

of shapes and surface sculpturing. 1, G. minutiflora , xl200; 2, G. incana , xl800; 3 G plumose, subsp. plumose,

x 1800; 4-8, variations in surface sculpturing, all X6000; 4, microscabrate in G. plumosa subsp. irrorata, 5

scabrate in G. rigida\ 6, microscabrate with tendency to form tecta in G. plumosa subsp. plumosa ; 7, scabrate wi

areas large of tecta in G. parviflora subsp. parviflora\ 8, smooth with complete coverage by tecta in G . secimdiflora

(cf. also surface in No. 1 above).

68

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

only known from the type collections. An examination

of a few flowers did not confirm this and without

investigating numerous flowers it was decided to

accept Brown’s observations. A number of flowers

with 3-celled ovaries was, however, found in G.

parviflora subsp. pubescens.

There are several genera in the Ericoideae which

have 2-celled ovaries and 4 stamens, i.e. Simocheilus,

Acrostemon, Thoracosperma, Sympieza, Aniserica and

Coilostigma, but none of them has the distinctive

bipartite anthers found in Grisebachia. They all

have a different appearance from the rather uniform

Grisebachia.

With the discovery of G. secundiflora, this uniformity

was slightly changed. This species was difficult to

place satisfactorily in any of the genera due to its

1 -celled ovary. The species was clearly allied to some

species of Eremia and Grisebachia, but could not be

included in the former on the grounds of having

only 4 stamens and in the latter for having a 1-celled

ovary. As the genus Eremia has recently been emended

(Oliver, 1976) to include ovary variations from

4-celled to 1-celled but with constantly 8 stamens,

it was decided to place this new species under Grise-

bachia next to G. parviflora and emend that genus to

include the 1-celled ovary rather than alter Eremia

even further to include 4 stamens. The 1-celled ovary

brought the species cl >se to Anomalanthus and

Syndesmanhus, neither of which it resembles.

The close similarity between Grisebachia and

Eremia has been mentioned under that genus (Oliver,

1976). G. parviflora is superficially similar to Eremia

curvistyla in flower form and habit. G. secundiflora

looks very much like Eremia totta but, in both cases,

the 4 stamens serve to distinguish them as species

belonging to the two separate genera. The anthers of

all species of Eremia except E. curvistyla are only

bilobed and not distinctly bipartite as occurs in

Grisebachia. In Eremiella outeniquae (Oliver, 1976)

the anthers are also bipartite, but the rest of the

floral characters are very different from Grisebachia.

Grisebachia and Eremia are sympatric to a great

extent in the area from the Cedarberg to the Cold

Bokkeveld. Undoubtedly they are very closely allied

and have possibly envolved from some ancestral

stock, which in turn arose from the genus Erica by

reduction.

PHYTOGEOGRAPHY

The genus Grisebachia is endemic in the south-

western and western parts (Fig. 3) of the Cape Province

corresponding to the South-Western and North-

Western Phytogeographical Centres proposed by

Weimarck (1941). Most of the species occur on and

west of the mountain complex of the Cedarberg,

Cold Bokkeveld and Hex River and are mostly

exclusive to their areas.

The commonest and most widespread species is G.

parviflora, which occurs only on mountains from the

northern Cedarberg southwards to the Villiersdorp

area and eastwards to near Swellendam. It shares the

central plateau of the Cold Bokkeveld with the

geographically restricted species, G. minutiflora and

G. secundiflora. These closely related taxa are ecolo-

gically vicarious. G. minutiflora is confined to sandy

open flat areas.

G. ciliaris and its subspecies occur from the Nieu-

voudtville plateau southwards along the east side of

the Olifants River to the Porterville mountains and

overlap slightly with G. parviflora in the northern

Cedarberg. G. ciliaris ssp. ciliciiflora extends west-

ward into the area of G. plumosa in Boekenberg and

Lambertshoekberg. G. plumosa occupies the west

side of the Olifants River from Clanwilliam south-

wards along the sandy coastal flats to the Mamre

area.

Four closely related taxa are geographically

vicarious. G. ciliaris is a northern species. G. incana

occurs just south of the distribution range of G.

plumosa from near Mamre to Sir Lowry’s Pass. G.

rigida is confined to the valley between Villiersdorp

and Worcester and G. nivenii occurs isolated on the

flats just south-east of Swellendam. In the same

areas as the last two species there are records of

G. parviflora, but these are from rocky higher altitude

slopes.

With the exception of G. parviflora, which appears

to be confined to rocky mountain slopes, all the

species of Grisebachia are confined to sandy flats

either on the coastal lowlands or inland on plateaux

in the mountains. In many cases these sandy areas

may be small and isolated and are usually alluvial

being derived from rocks of the Table Mountain

Series. This is very evident in the localities of G.

rigida, G. nivenii, G. minutiflora, G. secundiflora and

G. ciliaris subsp. ciliaris, subsp. bolusii, subsp.

ciliciiflora and subsp. multiglandulosa. The habitat of

G. ciliaris subsp. involuta is not known, as details are

not given on the three collections and I have not

collected the subspecies, but it would probably fit

the requirements.

On the lowlands of the west coast the sands are

recent and alternate with the heavier clay soils of the

Malmesbury beds. Here the species G. plumosa and

E. G. H. OLIVER

69

G. incana occur. The former is particularly widespread

from Mamre to the Paleisheuwel area. In the north

it is confined to sandy pockets in the mountains. The

isolated localities of G. incana in the area from Mamre

to Sir Lowry’s Pass are ascribed to the occurrence

of suitable sandy restionaceous sites. No species have

been recorded from the extensive sands of the Cape

Flats proper. This is probably due to the sand over-

lying extensive calcrete deposits.

The occurrence of G. ciliaris subsp. ciliaris on the

escarpment at Niewoudtville in the extreme northern-

most extension of more or less continuous fynbos is

interesting. This feature is shared with only one

other species in the minor genera, Simocheilus

klotzschianus Benth. and a few species of Erica.

GRISEBACHIA

Grisebachia Klotzsch in Linnaea 12:225 (1838);

Benth. in DC., Prodr. 7: 700 (1839), pro parte; et in

Benth. & Hook, f., Gen. PI. 2: 592 (1876), pro parte;

Drude in Pflanzenfam. 4,1:63 (1897), pro parte;

N.E.Br. in FI. Cap. 4,1: 337 (1906); Phill., Gen. ed.

1, 461 (1926); E. G. H. Oliver in Dyer, Gen. 1: 433

(1975). Type species: G. plumosa Klotzsch.

Blaeria L.f. Suppl. 122 (1781), pro parte et auct.

mult.

Erica Thunb., Prodr. 73 (1794), pro parte; Salisb.

in Trans. Linn Soc. 6: 339 (1802), pro parte; Thunb.,

FI. Cap. 364 (1823), pro parte.

Eremia Klotzsch in Linnaea 12:498 (1838), pro

parte; Phill. in J1 S. Afr. Bot. 10: 70 (1944), pro

parte; et Gen. ed. 2, 560 (1951), pro parte.

Perennial woody shrublets, erect up to 50 cm,

rarely 1 m, or compact and spreading to prostrate

and spreading. Leaves 3-nate, rarely 4-nate, erect

imbricate to spreading and recurved, pubescent

often with stout subplumose to plumose hairs on the

margins and abaxial surface. Flowers in terminal

heads usually on short lateral branchlets sometimes

forming congested pseudo-spikes. Bracteoles 3, mostly

approximate, small and inconspicuous, the median

sometimes large, glabrous or pubescent often with

stout subplumose to plumose hairs on the margins

and abaxial surface. Calyx 4-partite or lobed, small

to enlarged and conspicuous; segments equal some-

times the laterals slightly narrower, glabrous to

pubescent to crisped pilose with few to numerous

conspicuous stout simple to plumose hairs on the

margins and abaxial surface, the hairs sometimes

gland-tipped. Corolla 4-lobed mostly distinctly con-

stricted about two-thirds of the way up with an

ovoid to obovoid base and cyathiform top or some-

times obconical to tubular with no constriction,

mostly pubescent to pilose in the middle region and

on the inside around the constriction, mostly pink

otherwise white. Stamens 4 free, mostly manifest,

rarely included. Anthers mostly bipartite, rarely

bilobed, muticous, occasionally aristate, scabrous,

with large pores. Pollen grains single, tricolporate,

mostly ellipsoid with flattened apices rarely oblate,

surface scabrate to microscabrate with element

rods free to fused. Ovary 2-celled with a single

subapical pendulous ovule in each cell, rarely 3-celled,

in one species obliquely 1-celled, with a distinct

nectariferous disc. Style exserted. Stigma simple to

capitellate. Fruit a hard apparently indehiscent nut.

1 . G. plumosa

Key to the species

Corolla with a distinct constriction in the middle:

Sepals joined for quarter to three-quarters their length

Sepals free or slightly joined at the base:

Cilia on the calyx as long as or mostly shorter than the width of the sepal:

Sepals less than 2,0x0,65 mm

Sepals more than 2,0x0,65 mm:

Leaves, when young, ciliate with short gland-tipped hairs

Leaves, when young, ciliate with short plumose tufts

Cilia on the calyx longer than the width of the sepal

Corolla obconical or tubular without a distinct constriction in the middle:

Inflorescence globose, 6-36-flowered, not arranged in a pseudospike

Inflorescence 1-4-flowered on short lateral branchlets arranged in a pseudospike:

Leaves not ciliate

Leaves distinctly ciliate with stout plumose hairs

1. Grisebachia plumosa Klotzsch in Linnaea

12: 226 (1838). Syntypes: Doornhoogde on the Cape

Flats, Ecklon & Zevher s.n. (B|); distributed as

268 (G!; MO!; S!; W!); Doornhoogde, Dr'ege s.n.

(B|). Lectotype: Ecklon & Zeyher 268 (MO).

Compact erect shrubs up to 0,5 m, rarely 1 m, high.

Branches numerous erect pubescent to tomentose

with longer stout hairs inbetween, sometimes gland-

tipped, sometimes becoming glabrous and grey.

Leaves 3- or 4-nate inbricate and adpressed to spread-

ing recurved, up to 4 mm long, linear-oblong to

ovate, acute to obtuse, glabrous or puberulous to

tomentose and canopubescent when young and with

short to long stout plumose or simple eglandular or

glandular hairs on the margins only or also on the

abaxial surface becoming glabrous on the abaxial

surface and scabrid with the stout hairs falling off

leaving short truncate stubs, rarely only crisped

....3. G. incana

4. G. rigida

5. G. nivenii

2. G. ciliaris

.7. G. minutiflora

.6. G. parviflora

8. G. secundiflora

pubescent without any stout hairs. Flowers (1) 6-12

(16)-nate in terminal erect or nodding heads; pedicels

short, 0,5-1, 5 mm long, puberulous sometimes with

longer stouter hairs inbetween, sometimes gland-

tipped; bracteoles median to adpressed, markedly

unequal to subequal, the median from lanceolate to

broadly ovate from an expanded base, 1 ,5-3, 5x2, 7

mm, the laterals usually smaller and narrower,

mostly oblong-elliptic, all ciliate with short to long

stout plumose to simple hairs which may be gland-

tipped, with or without an even to sparse covering of

similar shorter or equally sized hairs on the abaxial

surface, sometimes just crisped pubescent. Calyx

joined for one third to two thirds of its length,

campanulate sometimes 4-angled at the base, pink,

glabrous to pubescent with the lobes ciliate with

short to long simple to markedly plumose stout to

soft hairs which may be gland-tipped, occasionally

with an even to sparse covering of similar shorter or

70

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

equally sized hairs on the abaxial surface; lobes

narrowly to broadly deltoid, slightly sulcate at the

apex. Corolla 4-lobed up to 4 mm long, very con-

stricted half to two thirds the way up above an ovoid

to obovoid base cyathiform above, often 4-angled

the angles alternating with the calyx segments,

pubescent to sparsely so in the middle region and

pilose on the inside at the mouth or constriction;

lobes broadly ovate to deltoid, obtuse erect to slightly

spreading, glabrous or slightly pubescent down the

centre outside. Stamens 4, free; filaments narrowly

linear broadened at the point of attachment, pilose,

white; anthers exserted or manifest, 0,5-1 ,3 mm long

obovate, dorsally attached, scabrous, muticous or

rarely minutely awned the awns occurring only in a

few flowers or anthers; pore about half the length of

the cell; pollen grains single. Ovary 2-celled, com-

pressed, ovoid to oblate, obtuse to subacute smooth

to verrucose rarely pilose at the apex otherwise

glabrous, seated on a disc ; style exserted ; stigma small,

subcapitate; fruit hard verrucose. Figs 5-8.

A species forming erect shrublets up to 0,5 m

occurring on sandy coastal flats in the western Cape

Province from Cape Town to Graafwater and on

mountain slopes in the Clanwilliam area, flowering

from June to September.

G. plumosa is characterized by the calyx being

joined for one quarter to three-quarters of its length,

the corolla-tube being distinctly constricted in the

middle and the habit being erect.

In his treatment of the genus Brown recognized

five species in the group with joined calyces, basing

the separation on the form of the stout hairs on the

calyx. The five species were G. plumosa, G. hirta, G.

pilifolia, G. pentheri and G. solivaga. On the small

amount of material available to him this classification

was feasible. But since Flora Capensis numerous

collections of these species have been made. An

examination of all this material showed a degree of

variation in the diagnostic characters sufficient to

warrant the five species being regarded as one single

complex of taxa with discontinuities occurring only

in one character and with partial separation in other

characters between the constituent taxa. The oldest

name applicable to this complex is G. plumosa

Klotzsch.

It was also found that variation in the number of

leaves per whorl in some specimens of G. pentheri

overlapped with the number in the very similar G.

hispicla which had been separated off from the rest

of the species in the genus on this character. G.

hispida therefore had to be included in the G. plumosa

Complex.

The above six taxa were then examined as one

complex group. It was found that the group could be

divided into two form series on the position of the

stout hairs on the leaves, one with the hairs only on

the margins the other with the hairs also scattered

on the abaxial surface. The first series contained G.

plumosa and G. solivaga, the second contained G.

hispida, G. pentheri, G. pilifolia and G. hirta.

In the first form series the variation in calyx hairs

from the very plumose material of G. plumosa in

the Mamre area to the almost simple-haired specimens

in the Aurora area showed an overlap with the type

and only collection of G. solivaga from just west of

Clanwilliam. Material which had been named as G.

hirta was found to constitute a distinct new taxon

more closely allied to G. plumosa on the leaf character.

Furthermore, a collection made in the Gifberg

{Oliver 4951) was found to be nearest to G. plumosa

and, although somewhat anomalous, was referred to

this series which then consisted of G. plumosa (in-

cluding G. solivaga) and the two new taxa.

In the second form series G. pilifolia showed

considerable variation with an overlap in the dis-

tinguishing characters with G. pentheri thus necessitat-

ing its reduction to synonomy. A close relationship

with G. hispida was shown to exist with only a partial

separation on the number of leaves per whorl and a

distinct separation in the plumosity of the hairs.

G. pentheri and G. hirta appeared to be very similar

with only one character showing any disjunction, i.e.

the position of the stout hairs on the abaxial surface

of the calyx. This series therefore consisted of G.

hispida, G. pentheri (including G. pilifolia) and G.

hirta.

Within these two form series recognizable on the

single character difference of leaf hairs, several more

or less distinct taxa could be distinguished again on

various single character differences. The complex

occurs in two main distribution centres, the Mamre

area in the south and the mountains west of the

Olifants River in the north. Regional separation of

the two series in the complex is only partial. The

“plumosa” series is concentrated in the south with

some outliers in the far north and the “hispida”

series in the north with outliers in the south.

Fig. 4. — Distribution of Grisebachia plumosa : © subsp.

plumosa ; ® subsp. hirta', O subsp. irrorata', % subsp.

eciliata; Q subsp. hispida', © subsp. pentheri.

E. G. H. OLIVER

71

Fig. 5. — Grisebachia plumosa subsp. plumosa. 1, flower, x8; 2, corolla, x8; both from Barker 10388 (NBG); 3, young

leaf; 4, old leaf; 5, lateral bracteole; 6, median bracteole of an outer flower; 7, median bracteole of an inner flower;

8, anther, side, front and back views; 9, ovary; all drawn x 16 from the lectotype, Ecklon & Zeyher 268 (MO); 10,

sepal, x 16, drawn from Schlechter 8480 (PRE).

It was decided for reasons of expediency to regard

the complex as one species with six subspecies based

on a single character disjunction occurring with

some degree of regional separation over a wide part

of the distribution range of the species (Fig. 4).

This classification, though not final, attempts to

show the type of variation that occurs in this complex.

Only a thorough biosystematic study of the

populations will solve the problems and should

either confirm or correct the above classification

I have given.

In floral and foliage characters this is a very

variable taxon in which six subspecies are recognized.

Key to the subspecies

Leaves with simple to plumose stout hairs confined to the

margins or without cilia, sometimes in older leaves

remaining only as short truncate setae:

Cilia simple or plumose or setae present on the leaves:

Calyx hairs eglandular (a) subsp. plumosa

Calyx hairs glandular (b) subsp. irrorata

Cilia absent on the leaves, sometimes small compound

tufts present, otherwise shortly lanate. .(c) subsp. eciliata

Leaves with simple to plumose stout hairs on the margins and

abaxial surface at least when young, sometimes in older

leaves remaining only as short truncate setae:

Hairs on the leaves and calyx eglandular:

Leaves 4-nate; calyx hairs densely white plumose

(d) subsp. hispida

Leaves 3-nate rarely also 4-nate on the same branch;

calyx more sparsely hairy, hairs less plumose

(e) subsp. pentheri

Hairs on the leaves and calyx glandular:

Calyx hairs mostly confined to the edges of the sepals

and distinctly larger than the few on the abaxial

surface; bracteoles and sepals usually glabrous

(e) subsp. pentheri

Calyx hairs mostly of equal size and evenly distributed

over the abaxial surface; bracteoles and sepals usually

puberulous (f) subsp. hirta

(a) subsp. plumosa

G. plumosa Klotzsch in Linnaea 12: 226 (1838); Benth. in

DC., Prodr. 7:701 (1839); N.E. Br. in FI. Cap. 4,1:345

(1906). Syntypes: Doornhoogde on the Cape Flats, Ecklon &

Zeyher s.n. (Bj); distributed as 268 (G!; MO!; S ! ; W !) ;

Doornhoogde, Drege s.n. (Bf). Lectotype: Ecklon & Zeyher

268 (MO).

G. serrulata Benth. in DC., Prodr. 7:701 (1839). Type:

Giftberg, Drege 7802 (BM!; K ! ; MO!; P ! ; W!). G. plumosa

var. serrulata (Benth.) N.E. Br. in FI. Cap. 4,1: 346 (1906).

G. solivaga N.E. Br. in FI. Cap. 4,1:347 (1906). Type:

Zeekoe Vlei, Clanwilliam, Schlechter 8480 (BM!; BOL!;

E!; G!; K!; MO!; P!; PRE!; W!).

Branches pubescent to tomentose with short

stout plumose hairs inbetween, rarely just tomentose.

Leaves 3-nate mostly erect and adpressed, pubescent

to canopubescent and ciliate with short, occasionally

long, stout plumose hairs, becoming glabrous and

often scabrid edged with the cilia falling off leaving

short truncate setae. Bracteoles pubescent, rarely

glabrous, ciliate with short stout plumose hairs,

rarely also with similar hairs on the abaxial surface.

Sepals pubescent, rarely glabrous, ciliate with stout

plumose, rarely subplumose hairs, with a few shorter

ones scattered over the abaxial surface. Fig. 5.

Cape. — 3218 (Clanwilliam): Zeekoe Vlei, 122 m (-BA/BB),

Schlechter 8480 (BM; BOL; E; G; K; MO; P; PRE; W).

S.W. of Aurora (-CB), Barker 9712 (NBG; STE); Aurora to

Velddrif, 61 m (-CB), Williams 502 (BM: K); Goergap- Aurora

Road (-DA), Thompson 791 (MO; PRE; STE); Sauer area

(-DC), Barker 5783 (BOL; NBG; STE); Barker 8087 (NBG;

STE); Compton 15488 (NBG); Leighton 167 (BOL; PRE);

Lewis 3991 (SAM; STE); Wilman 865 (NBG; PRE). 3318 Cape

Town: Hopefield (-AB), Bachman s.n. sub Guthrie 2883 (BOL;

K); Darling 244 m (-AD), Bayliss 1663 (B; PRE); Darling

(-AD), Grant 4604 (BM); Between Malmesbury and Darling

(-AD/BC), Hutchinson 205 (BM; BOL; K; PRE); Salter 4410

(K; PRE); Mamre Road (-BC), Barker 5376 (NBG); Compton

19863 (NBG; STE); Hillslopes west of Abbotsdale (-BC)

Oliver s.n. (STE); Burgherspost, Dassenberg, 152 m (-BC),

Rourke 500 (NBG; STE); Mamre area, flats and hills (-AD/BC/

CB/DA), Barker 5375 (BOL; NBG); 10388 (NBG); Compton

6782 (NBG); 7797 (NBG); 9530 (NBG); 13883 (NBG); Garside

4772 (K); Levy ns 3196 (CT); Ry croft 1806 (NBG; STE); Sidey

2289 (MO; S); Stokoe in SAM 62573 (SAM); Between Melk-

72

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

bosch and Mamre (-CB/DA), Salter 1234 (BM; K; MO); 10

miles south of Mamre (-DA), Davis s.n. (SAM; STE); Doorn-

hoogde on Cape Flats (-DC) Ecklon & Zeyher 268 (G; MO; S;

W).

Subsp. plumosa is characterized by having mostly

plumose, occasionally simple, cilia on the leaf margins

and no glands on the calyx. The cilia are present at

least in the young stages as they often fall off leaving

minute truncate setae which can easily be overlooked.

This latter character was used by Bentham in creating

his G. serrulata.

This subspecies is very variable in the degree of

subdivision of the stout hairs on the calyx and

bracteoles. The specimens of Bachmann s.n. from

Hopefield are very plumose. The average degree of

plumosity occurs in the Mamre collections. The

collections from near Aurora have glabrous calyces

with simple stout cilia. These latter collections are

very similar to the type and only collection of

G. solivaga N.E. Br. from just west of Clanwilliam.

(b) subsp. irrorata E. G. H. Oliver, subsp. nov.,

a subspecie typica ciliis glandulis calyce et subspecie

hirta similissime cilis foliorum margine solum

dignoscenda.

Type. — Bokbaai, Taylor 5011 (STE, holo.; PRE).

Branches pubescent becoming glabrous with short

stout plumose hairs inbetween when older. Leaves

3-nate, adpressed but sometimes slightly recurved,

pubescent becoming glabrous on the abaxial surface,

ciliate with short stout plumose hairs, occasionally

gland-tipped when young, sometimes falling off

when older and remaining as short truncate setae.

Bracteoles and sepals pubescent mainly at the apex,

ciliate with simple to sparsely plumose gland-tipped

hairs on the margins and slightly shorter ones on

the abaxial surface. Fig. 6.1.

Fig. 6. — Grisebachia plumosa: subsp. irrorata. 1, leaf, Xl6,

drawn from the holotype, Taylor 5011 (STE): subsp.

eciliata. 2, flower, x8; 3, leaf, xl6; both drawn from

the holotype, Oliver 4951 (STE).

Cape. — 3318 (Cape Town); Bokbaai, 50 m (-CB), Taylor

5011 (PRE; STE); near Driefontein on roadside between

Philadelphia and Mamre (-DA), Pillans 6784 (BOL); Mamre

hills (-AD/BC/CB/DA), Compton 9530 (NBG). Between

Mosselbanks and Berg Rivers (-DA/DB), Burchell 981 (K;

PRE). Without precise locality: Malmesbury, from flower

sellers, Lamb 4097 (SAM). Locality unreliable: Swartberg at

Caledon, Lamb s.n. (BOL).

Subsp. irrorata is recognizable by the gland-tipped

hairs on the calyx and lack of stout hairs on the

abaxial surface of the leaves.

Material of this taxon had been, until now, placed

under G. hirta Klotzsch due to the key character of

a glandular calyx without taking into account the

significant differences in the leaf indumentum. In

subsp. irrorata the leaves are typical of the “plumosa”

kind in which the stout hairs are confined to the

margins of the leaves and do not occur on the abaxial

surface as they occur in subsp. hirta (G. hirta

Klotzsch).

The collection, Compton 9530, is intermediate

between subsp. irrorata and subsp. plumosa in having

glands only on the abaxial surface of the calyx.

There appears to be no intermediate between this

taxon and subsp. hirta. However, after a biosystematic

study this taxon may be shown to be a hybrid between

subsp. plumosa and subsp. hirta.

The three taxa, subsp. plumosa, subsp. irrorata and

subsp. hirta are sympatric on the sandy flats of the

Mamre area. These flats have unfortunately been

decimated by alien vegetation and human activity.

The pressure on the area from industrial and urban

development is now very great.

A thorough biosystematic study of the populations

from this area was not possible and probably never

will be possible for a more objective assessment of

their relationships.

(c) subsp. eciliata E. G. H. Oliver, subsp. nov.,

in specie foliis sine ciliis pagine abaxiali vel marginibus

distinguitur.

Type. — Ordinansiekop, Gif berg, Oliver 4951 (STE,

holo.; BOL; K; NBG; MO; PRE; S).

Branches pubescent with simple hairs only.

Leaves 3-nate adpressed, lanate with simple hairs,

occasionally with a few tufts of stouter hairs on the

margins. Bracteoles pubescent and lanate at the

apex. Sepals very sparsely pilose, ciliate with

irregularly and sparsely plumose hairs, rarely with

a few similar hairs on the abaxial surface.

Fig. 6.2, 6.3.

Cape. — 3118 (Vanrhynsdorp), Ordinansiekop on the Gif berg,

426 m (-DD), Oliver 4951 (BOL; K; NBG; MO; PRE; S; STE).

Subsp. eciliata is distinct in the species for having

leaves which do not possess stout hairs on either the

margins or abaxial surface. The pubescence is very

short, lanate and crisped. There is, however, an

occasional tuft of hairs on the margins of the leaves

but not similar to those found in the rest of the

species.

This taxon is somewhat anomalous in that it has

a similarity to G. plumosa complex in which the

closest affinity is with the material formerly known

as G. solivaga N.E. Br. now forming part of G.

plumosa subsp. plumosa. The crisped pubescence,

lack of distinct stout hairs and the tufts on the

leaves are similar to the condition found in G. ciliaris

subsp. ciliaris which occurs in the same area. The

broad calyx lobes are similar to the broad sepals in

the material formerly known as G. dregeana Benth.

The sepals of the latter are, however, free or very

slightly joined at the base.

It was decided to place this taxon under G. plumosa

on the basis of the fused calyx segments and to

leave G. ciliaris to be characterized by its free sepals.

The taxon is, however, a close link between these

two species and points to the need for a thorough

biosystematic study of all the Olifants River taxa

to understand their relationships.

(d) subsp. hispida ( Klotzsch ) E. G. H. Oliver,

comb, et stat. nov.

G. hispida Klotzsch in Linnaea 12: 226 (1838); Benth. in DC.,

Prodr. 7: 701 (1839); Rach in Linnaea 26: 790 (1855); N.E. Br.

Fig. 7. — Grisebachia plumosa: subsp. hispida. 1, sepal; 2, leaf; both drawn xl6, from the lectotype, Ecklon & Zeyher

267 (G): subsp. pentheri. 3, flower x 10; 4, sepal, x 16; 5, leaf, x 16; all drawn from the isotype, Penther 2925 (STE);

6, sepal, xl6; 7, leaf, xl6; both drawn from Oliver 3869 (STE); 8, cilia variation: a, subsp. pentheri b, subsp.

hispida.

in FI. Cap. 4,1: 339 (1906). Syntypes: near Olifantsrivier and

farm Brakfontein, July, Ecklon & Zeyher s.n. (Bf); distributed

as 267 (G!; S!; W!); mountains at Winterhoek, Witsenberg

and near Vogelvlei; Ecklon & Zeyher s.n. (Bf; LD!; P!).

Lectotype : Ecklon & Zeyher 267 (G).

Blaeria ptilota E. Mey. ex Benth. In DC., Prodr. 7:701

(1839). Type: Drege s.n. (Bf; isos.)

Branches pilose with numerous long stout plumose

hairs inbetween. Leaves 4-nate, imbricate but slightly

spreading, pilose, long ciliate on the edges with

plumose hairs and clothed with similar hairs on the

abaxial surface, ofteh breaking off and remaining

only as short truncate setae. Bracteoles and sepals

pubescent, densely ciliate and clothed on the abaxial

surface with long stout very plumose hairs like those

on the leaves, eglandular. Fig. 7.1, 7.2.

Cape. — 3218 (Clanwilliam): Sandy uplands north of Paleis-

heuwel between Berg Valley and Langevallei (-BC), Acocks

2994 (S); Paleisheuwel (-BC), Barker 6719 (BOL; NBG);

Boekenberg near Paleisheuwel, 549 m (-BC), Compton 4935

(BOL); Zwartbaskraal east of Boekenberg 300 m, (-BC),

Oliver 3890 (B; E; K; NBG; MO; PRE; S; STE); Boekenberg

457 m (-BC), Williams 817 (NBG); near Zwartbaskraal, 300 m,

(-BC/BD), Drege s.n. (BOL; K; PRE; S); between Bergvallei

& Zwartbaskraal 244 m (-BC/BD), Drege s.n. (G-DC);

Bergvallei at Zwartbaskraal 244 m (-BC/BD), Drege 1180 (P);

Olifants River at Brakfontein, (-BD/DB), Ecklon & Zeyher 267

(G; S; W). Without locality: Drege s.n. (BM; G; P; S). Doubt-

ful locality: 3319 (Worcester), Foot of Winterhoeksberg at

Vogelvallei (-AC), Ecklon & Zeyher s.n. (LD; P).

Subsp. hispida may be distinguished by its 4-nate

leaves and calyx thickly covered with densely plumose

hairs. In all the material examined the leaves were

4-nate. In subsp. pentheri some specimens have been

found to possess 4-nate leaves on branches with

mostly 3-nate leaves. The subspecies has the largest

leaves, flower-heads and flowers in the species.

This subspecies is very similar to some forms of

subsp. pentheri , some of which used to constitute

part of what was formerly G. pilifolia N.E. Br.

In these forms the calyx and leaves are eglandular

and the hairs very plumose.

Subsp. hispida appears to be very restricted in its

distribution, occurring only on the sandy hills and

flats near Paleisheuwel associated with dry fynbos

scrub which, according to Acocks’s map of Veld

Types (1953), is classified as True Fynbos and not

Coastal Macchia.

(e) subsp. pentheri ( Zahlbr .) E. G. H. Oliver ,

comb, et stat nov.

G. pentheri Zahlbr. in Ann. Naturh. Mus. Wien. 20:42

(1905); N.E. Br. in FI. Cap. 4,1: 1128 (1909). Type: Elands-

fontein, Clanwilliam, Aug. 1894, Penther 2925 (BM!; BOL!;

STE!; W! holo).

G. dregeana Benth. var. vestita Zahlbr. in Ann. naturh. Mus.

Wien 20: 43 (1905). Type: Olifants River valley. Penther 2917

(W, holo!).

G. pilifolia N.E. Br. in FI. Cap. 4,1 : 346 (1906). Type: near

Clanwilliam, Leipoldt 46 (BM!; BOL!; K! holo; PRE!; SAM!).

G. plumosa var. scabra N.E. Br. in FI. Cap. 4,1; 346 (1906).

Type: Thom s.n. (BOL!)

Branches pubescent, rarely glabrous, with long

stiff plumose eglandular or gland-tipped hairs

inbetween. Leaves 3-nate, rarely also 4-nate, pubescent

when young, ciliate with stout hairs and clothed

with a few similar hairs on the abaxial surface, the

hairs being simple and gland-tipped to plumose and

eglandular, often breaking off and remaining as

short truncate setae. Bracteoles and sepals mostly

glabrous, rarely sparsely pubescent, ciliate with

simple to plumose cilia which may be gland-tipped,

occasionally clothed with a few similar shorter hairs

on the abaxial surface down the middle. Fig. 7.3-

7.7.

Cape.— 3218 (Clanwilliam): Uitkomst, Graafwater, 427 m,

(-BA), Compton 4945 (BOL; NBG); 4949 (BOL); Compton

6789 (NBG; STE); Compton 24218 (NBG; STE); Kanovlei,

east of Graafwater, 396 m (-BA), Oliver 3869 (STE); Die Berg

74

STUDIES IN THE ERICOIDEAE. III. THE GENUS GR1SEBA CHI A

road (-BA/BB), Pamphlett 103 (NBG; STE); Zeekoevlei

(-BA/BB), Pillans 7074 (BOL; K; NBG; PRE); 122 m, Schlech-

ter8480( BM; BOL; E; G; K; MO; P; PRE; STE; W); Krans-

vleiberg (-BB), Compton 20031 (NBG); Summit of ridge south-

west of Kransvleiberg, 640 m (-BB), Oliver 3873 (STE); Top of

Kransvlei Pass, (-BB/BD), Lewis 2724 (SAM); Lambertshoek-

berg, 910 m (-BD), Compton 5490 & 1 (BOL; NBG); Maguire

419 (BOL; NBG); Olyvenboschkraal (-BD), Leipoldt 745

(SAM); Crossroads north of Elandsfontein, 640 m (-BD),

Oliver 3879 (BM; MO; STE); Elandsfontein (-BD), Penther

2925 (BM; BOL; STE; W); Olifants River mountains 610 m,

(-BD/DB), Schlechter 5099 (BM; BOL; K; NBG; NH; PRE).

Without precise locality: Clanwilliam, Leipoldt 46 (BM; BOL;

K; PRE; SAM).

Subsp. pentheri is characterized by its 3-nate

leaves, which are very rarely 4-n.ate on the same

branch, its calyx which is more sparsely hairy and

less plumose than in subsp. hispida and in the

glandular form by having most of the hairs on the

margins of the sepals.

This subspecies is very variable in the form of

the hairs on the leaves and calyx. In Flora Capensis

Brown recognized two separate species based on

these hairs, his own G. pilifolia with its simple to

plumose eglandular hairs and later in the addenda

Zahlbruckner’s G. pentheri with its gland-tipped

simple to subplumose hairs. Since Flora Capensis,

more material of this group has been collected and

has exhibited a complete range between the two

extremes thus necessitating a reduction of G. pilifolia

to synonomy under G. pentheri which itself had to

be reduced to subspecific rank in the G. plumosa

complex.

The relationship between the glandular forms of

subsp. pentheri and subsp. hirta is very close and it

is only with some careful examination that they

can be distinguished. The only character which shows

any discontinuity is the distribution of the hairs on

the calyx. In subsp. pentheri the stout hairs are mostly

confined to the margins of the calyx lobes with the

hairs on the abaxial surface being few and shorter.

In subsp. hirta the hairs are more or less evenly

distributed over the calyx and are of the same length.

This relationship is interesting because the two

taxa are widely separated.

Subsp. pentheri occurs frequently in scattered

populations in sandy areas on the mountains on the

west side of the Olifants River near Clanwilliam.

Unfortunately much of the habitat of this taxon has

been lost to farming practices and all that remains

is in the rocky unusable areas. The fynbos in which

the plants grow may without human and animal

intervention become quite tall and erect plants of

this taxon have been seen up to 1 m high.

(f) subsp. hirta ( Klotzsch ) E. G. H. Oliver, comb

et stat. nov.

G. hirta Klotzsch in Linnaea 12: 226 (1838); Benth. in DC.,

Prodr. 7:701 (1839); N.E. Br. in FI. Cap. 4,1:345 (1906).

Type: saniflats near Groenekloof, Drege s.n. (Bf), distributed

as 7795 (BM!; BOL!; E!; G!; G-DC!; HAM!; K!; MO!;

P!; PRE!; W!). Lectotype: Drege 7795 (P).

Branches pubescent to tomentose with long stiff

plumose gland-tipped hairs inbetween. Leaves 3-nate,

pubescent when young becoming glabrous, ciliate

with long stout simple to plumose gland-tipped hairs

and clothed with similar hairs on the abaxial surface,

erect to spreading-recurved. Bracteoles and sepals

puberulous sometimes sparsely so, ciliate and evenly

clothed on the abaxial surface with numerous short

stout simple to sparsely plumose gland-tipped hairs.

Fig. 8.

Cape. — 3318 (Cape Town): 2 km north-east of Mamre (-AD),

Boucher s.n. (PRE; STE); Mamre (-CB), Baker 2445 (BM;

Fig. 8. — Grisebachia plumosa subsp. hirta. 1, flower, x8; 2,

sepal, xl6; both drawn from Oliver 3763 (STE); 3, leaf,

x 16, drawn from the lectotype, Drege 7795 (P).

K); south of Mamre (-CB), Barker 10388 (NBG); Silverstroom-

strand Road just south of Mamre, 300 m (-CB), Boucher s.n.

(STE); south of Mamre, (-CB) Rycroft 2081 (NBG); 1 km

north of Pella, 300 m, (-DA), Boucher s.n. (STE); Sandflats

near Groenkloof, 60 m (-DA), Drege 7795 (BM; BOL; E; G;

G-DC; HAM; K; MO; P; PRE; W); flats just west of Pella,

228 m (-DA), Oliver 3763 (STE); sandy north-east base of south

end of Dassenberg, (-DA), Pillans 6881 (BOL); Kalbaskraal

(-DA), Van der Merwe 14 (PRE; STE). Without locality:

Thom s.n. (BOL); Drege s.n. (S).

Subsp. hirta is characterized by its glandular stout

mostly simple hairs evenly distributed on the calyx.

It differs only in this respect from subsp. pentheri

and from subsp. irrorata in having gland-tipped

hairs on the margins and abaxial surfaces of the

leaves.

The relationship between subsp. hirta and subsp

irrorata is very close with the flowers being almost

identical. The difference in the leaves, however, is

distinct. In subsp. irrorata the leaves are typical of

subsp. plumosa with short stout plumose cilia. As all

three taxa are sympatric on the sandy flats near

Mamre there is a possibility that hybridisation and

introgression may occur. A thorough biosystematic

study of populations from the area will have to be

carried out to ascertain the relationships of these

taxa.

2. Grisebachia ciliaris ( L.f .) Klotzsch in Linnaea

12:255 (1838) quoad nom., exl. descr. et spec, in

herb. Willd. Type: Herb. Linn. (LINN).

Blaeria ciliaris L.f., Suppl. 122 (1782).

Shrublets mostly low growing and compact or

erect up to 75 cm high. Branches subglabrous to

pubescent, occasionally arachnoid, the hairs thick

and matted, erect or retrorse or very sparse and

erect, sometimes with stouter longer hairs inbetween

which are either plumose or simple and gland-tipped.

Leaves 4-nate, erect and adpressed, sometimes imbri-

cate to spreading recurved, 1-4,5 mm long with the

petiole very short to 0,5 mm long, from linear to

ovate to obovate, very variable in the indumentum,

pubescent with dense crisped retrorse hairs or puberu-

lous with erect hairs, eciliate or rarely with small

compound tufts on the margins or sometimes with

stout plumose cilia with few to many spreading

plume branches, the cilia often falling off and remain-

ing as short setae, all becoming more or less glabrous

with age, sometimes subglabrous to glabrous and

shiny on the abaxial surface and ciliate and clothed

on the abaxial surface with stout simpb gland-tipped

hairs. Flowers 3-12 in capitate, sometimes nodding,

heads at the ends of lateral branchlets; bracteoles 3

subequal to markedly unequal in the outer flowers

of the inflorescences to equal in the inner flowers,

mostly median to remote, adpressed or recurved, the

E. G. H. OLIVER

75

median 1,3-5x0,45-2,3 mm, small and oblong to

narrowly ovate with a relatively large keel-tip and no

markedly expanded base to broadly elliptic or ovate

with a broad flat base and relatively small but distinct

keel-tip, from almost glabrous to puberulous all over,

sometimes with a distinct apical tuft of lanate hairs,

sometimes ciliate with short to long stout simple to

plumose eglandular or gland-tipped hairs, rarely

with a few similar hairs on the abaxial surface at the

keel-tip; the pedicel 1 ,0-2,5 mm long, puberulous to

sparsely glandular pilose. Calyx 4-partite sometimes

slightly joined at the base, 1 , 5-4,3 x 0,4-2, 3 mm,

very variable in size and indumentum, small narrowly

oblong to oblong-ovate to broadly elliptic or large

oblong-elliptic to broadly elliptic and ovate, slightly

keel-tipped occasionally with a knoblike apex,

glabrous to pubescent sometimes with a distinct

apical tuft of lanate or straight hairs, ciliate with

long stout simple to plumose crooked or straight

hairs which are eglandular of gland-tipped, plume

branches long and spreading or short and erect,

sometimes clothed with similar hairs on the abaxial

surface, the apex devoid of cilia or ciliate, the cilia

as long as, mostly longer than, the width of the

sepals. Corolla 4-lobed, 2,5-7 mm long, constricted

in the middle to two-thirds of the way up, sometimes

inconspicuously so, inflated below in the lower part

and often 4-angled, the angles alternating with the

sepals, cyathiform above the constriction, pubescent

outside mainly in the middle region, pubescent to

pilose, rarely subglabrous inside around the constric-

tion; lobes erect to slightly spreading, broadly to

narrowly deltoid, smooth to slightly crenulate, obtuse,

occasionally emarginate. Stamens 4, free; filaments

mostly linear, expanded at the apex at the point of

attachment to the anther, sparsely pilose to villous;

anthers manifest, bipartite, 0,8-1 ,5 mm long, mostly

oblong, scabrid to long scabrid, muticous or aristate;

awns up to half the length of the cell; pore up to

half the length of the cell; pollen grains single.

Ovary 2-celled with a single ovule in each cell, mostly

compressed, ovoid to oblate, obtuse, glabrous to pilose

at the apex; style filiform, glabrous, far exserted;

stigma subsimple to capitellate. Figs 9-16.

A species forming low compact semispreading to

erect shrublets up to 0,5 m occuring in sandy areas

in mountains between Porterville and Niewoudtville

in the western Cape, flowering from August to

November.

G. ciliaris is characterized by having the calyx

segments free or only very slightly joined at the base,

the corolla-tube more or less distinctly constricted

in the middle and the cilia on the calyx longer than

the width of the sepals.

G. ciliaris is one of the oldest described species

among the minor genera of the Ericoideae. Strangely

the species is very isolated and far-removed from

Cape Town where other species more accessible

existed, but were overlooked for so long. Despite

its long standing, the species has been very much

confused until now. Linnaeus, the younger, stated in

the protologue that the species had 3-nate leaves

based undoubtedly on a Thunberg specimen. Thun-

berg himself later published a fuller description from

his own specimen stating that the leaves were 4-nate.

This error was subsequently repeated by numerous

authors until Rach (1853) corrected this.

A similar situation exists with G. ciliaris as occurs

in G. plumosa. In Flora Capensis Brown recognized

six species which he grouped on the character of a

Fig. 9. — Distribution of Grisebachia ciliaris : A subsp. ciliaris ;

O subsp. bolusii ; © subsp. involuta; # subsp. ciliciiflora',

calyx divided to the base and muticous anthers. He

then separated the species on the nature of the

indumentum on the leaves, the sepal size and the

length and form of the sepal hairs. These species

were G. bolusii, G. apiculata, G. involuta, G. vel-

leriflora, G. dregeana and G. zeyheriana. A seventh

species, G. thunbergii (G. ciliaris), he characterized

incorrectly by placing it with those species not having

a distinctly constricted corolla. On the small amount

of material available to Brown the recognition of

these taxa as distinct species was feasible but numerous

subsequent collections have provided a considerable

degree of variation which broke down many of the

existing discontinuities in the median bracteoles

(Figs 10 & 11), sepal hairs (Fig. 12) and leaf glands.

It would appear that we are dealing with an aggre-

gate species of spatially separated noninterbreeding

populations which are in the first stages of evolving

into a number of distinct entities which may eventually

become sufficiently distinct to be regarded as separate

species. At present, similarities are too close to

justify this latter classification.

An important feature and character of use in

delimiting the subspecies is the nature of the cilia

on the calyx, something which is easily observable

and yet somewhat difficult to define (Fig. 11) parti-

cularly in regard to the plume sidebranches.

Bracteole length

76

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Fig. 10. — Grisebachia ciliaris, variation in the median bracteole (above) and sepals (below) of the outer flowers. 1,

subsp. ciliaris', 2, subsp. bolusii', 3, subsp. ciliciiflora', 4, subsp. multiglandulosa', 5, subsp. involuta. All x6.

5

4-

3-

▲

A

S

E

2-

x A

Ox A A

o o * *

o o

x OtD •

• ® O

J3 •

• •

• x

■ ssp involuta

A ssp bolusii

• ssp ciliaris

O ssp ciliciiflora

X ssp multiglandulosa

1 2

bio. 11.— Grisebachia ciliaris, scatter diagram showing the

variation in the median bracteole of the outer 3 flowers in

a single inflorescence. Each dot represents the mean

measurement of each inflorescence.

Fig. 12. — Grisebachia ciliaris, types of calyx cilia, all drawn

Xl6. 1, subsp. ciliaris', 2, subsp. bolusii', 3, sibsp. cilicii-

flora', 4, subsp. multiglandulosa', 5, subsp. involuta.

G. ciliaris is confined mostly to the eastern side

of the Olifants River occuring as far north as the

Bokkeveld Plateau at Niewoudtville and the Gifberg

(subsp. ciliaris). In the northern Cedarberg subsp.

bolusii is confined to the Pakhuis Pass and area around

Heuningvlei and Boontjieskloof with subsp. involuta

recorded only from the Krakadouw area. Two

records of subsp. ciliciiflora occur near Wupperthal

far removed from the rest of the populations which

occur in two centres, one in the Elandskloof mountains

and Piekenierskloof, the other in the west on the

Boekenberg and Lambertshoekberg. Subsp. multi-

glandulosa is confined to the mountains at the southern

end of the main Olifants River Valley. The species

seems to be specific to Grisebachia-type habitats

namely small level open sandy alluvia mostly occur-

ring in the mountains in isolated pockets.

G. ciliaris has several relationships with other

species due to its considerable variability. The

relationships lie with G. plumosa subsp. pentheri and

subsp. eciliata and with G. incana and are discussed

under the relevant subspecies in G. ciliaris.

E. G. H. OLIVER

77

This is a variable taxon in which five subspecies

are recognized.

Key to subspecies

Bracteoles refer to the median bracteoles

of the outer 3 flowers in any inflorescence

Bracteoles small up to 1, 8x1,0 mm mostly 1,2x0, 5 mm,

mostly with a narrow base and relatively large keel-tip:

Leaves canopubescent with crisped hairs at least when young,

not distinctly glandular:

Bracteoles with crisped pubescence, very rarely also with

short stout hairs; calyx hairs short, simple and

straight to crooked and irregularly plumose with

spreading branches (a) subsp. ciliaris

Bracteoles pubescent and with distinct long plumose cilia;

calyx hairs long slender, more or less straight, sparsely

but evenly plumose with short forward-pointing

branches (d) subsp. ciliciiflora

Leaves not canopubescent, gland-apiculate or glandular

hairy;

Glands on short hairs or sessile, apical and/or marginal

only (d) subsp. ciliciiflora

Glands on long hairs, apical, marginal and on abaxial

surface (e) subsp. multiglandulosa

Bracteoles large, more than 2x1 mm with a broad base and

relatively small keel-tip:

Bracteoles less than 3,2 mm long; sepals less than 3 mm long

(b) subsp. bolusii

Bracteoles more than 4,2 mm long; sepals more than 4 mm

long (a) subsp. involuta

(a) subsp. ciliaris

Blaeria ciliaris L.f., Suppl. 122 (1782); Willd., Sp. PI. 1: 631

(1798); Wendl., Collect. 49 (1805); Ait., Hort. Kew ed. 2,

1:249 (1810); Roem. & Schult., Syst. Veg. 3:170 (1818);

Klotzsch in Linnaea 8: 658 (1833); G. Don, Gen. Syst. 3: 805

(1834). Type: Herb. Linn. (LINN, holo.; K, fragm.!). Grise-

bachia ciliaris (L.f.) Klotzsch in Linnaea 12: 255 (1838), quoad

nom. excl. spec, in herb. Willd. no. 2890.

Erica plumosa Thunb., Prodr. 73 (1794); Salisb. in Trans.

Linn. Soc. 6: 339 (1802); Schultes in Thunb. FI. Cap. 364

(1823). Type: Thunberg no. 9393 (UPS). Blaeria plumosa

(Thunb.) Thunb., Diss. Blaeria 9 (1802). Grisebachia thunbergii

Rach in Linnaea 26: 789 (1853); N.E. Br. in FI. Cap. 4, 1 : 347

(1906). Type as for Erica plumosa.

Grisebachia dregeana Benth. in DC., Prodr. 7:701 (1 839) •

N.E.Br. in FI. Cap. 4,1: 342 (1906). Type: Cape Colony, Drege

7803 (K, holo.!; isos.!). *

Branches pubescent with retrorse simple hairs,

sometimes arachnoid. Leaves adpressed, 1-2,5 mm

long, narrowly ovate to oblong to obovate, hairy

with dense crisped retrorse hairs, often becoming

glabrous on the abaxial surface. Bracteoles equal to

slightly unequal, median to remote, often recurved,

the median 1,3-1,8x0,45-0,8 mm, narrowly ovate

to elliptic to oblong with a relatively large keel-tip

and no markedly expanded base, mostly pilose with

crisped hairs, very rarely ciliate with a few short

stout hairs; pedicel up to 1,5 mm long. Sepals

1,5-2,2x0,4-11 mm, narrowly oblong to oblong to

oblong-ovate, often with a swollen knob-like apex,

very slightly keel-tipped, pilose at the base and

sometimes in the upper half, the apex clothed with

a tuft of lanate hairs, ciliate with simple to plumose

stout hairs with irregular short and long spreading

plume branches, eglandular or gland-tipped, cilia

often crooked, as long as or longer than the width

of the sepal, occasionally with some similar hairs

on the abaxial surface, the apex usually devoid of

cilia, rarely with cilia. Corolla up to 3,5 mm long,

distinctly or indistinctly constricted in the middle,

the constriction sometimes not visible in dried

material, pubescent outside in the middle region,

pilose to almost glabrous inside around the constric-

tion. Anthers 0, 8-0,9 mm long, scabrous, aristate

rarely muticous; awns up to half the length of the

cell. Ovary glabrous to pilose at the apex. Fig. 13.

Cape. — 3119 (Calvinia): Van Rhyns Pass (-AC), Barker 9427

(NBG; STE); Compton 20884 (BOL; NBG; STE); Hutchinson

763 (BM; BOL; K; PRE); 790 m, Oliver s.n. (STE); 762 m,

9

Fig. 13. — Grisebachia ciliaris subsp. ciliaris. 1, flower, x8; 2, corolla, x8; 3, bracteoles, xl6. a, laterals, b, median,

c, median inside view; all drawn from the fragment of the holotype (K); 4, flower, x8; 5, corolla, two views, x8,

6, sepal, X 16; 7, anther, side, front and back views, x 16; 8, ovary, X 16; 9, leaf, x 16; all drawn from Oliver 3860

(STE); 10, sepal, x 16; 11, ovary, x 16; both drawn from Marloth 7646 (STE); 12, median bracteoles, x 16, a from

Marloth 7646 (STE) and b from Middlemost 1594 (NBG); 13, anther, back, front and side views, x 16, drawn from

Drege 7803 (MO).

78

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Oliver 3860 (STE; PRE); Salter 1633 (BM; K; MO); L. E.

Taylor 2840 (NBG); Oorlogskloof, 580 m (-AC), Lavis in BOL

19811 (BOL); Bokkeveld mountains behind farm Tabaktuin,

910 m (-AC), Leipoldt 740 (BOL; K); Niewoudtville near the

homestead, 910 m (-AC), Leipoldt 741 (BOL; PRE), Glenridge,

just west of Niewoudtville (-AC), Lewis 5723 (NBG); 700 m,

Oliver 3864 (G; PRE; S; STE); west of Niewoudtville, 750 m

(-AC), Marloth 7646 (PRE; STE); Middlemost 1594 (BOL;

NBG); Willems River, 700 m (-AC), Schlechter 10974 (BM;

BOL; K; MO; STE); Lokenburg, 670 m (-CA), Acocks 17287

(PRE); Ekerdam, Calvinia (?), L. E. Taylor 2970 (NBG). 3118

(Vanrhynsdorp): Kobe Pass, 910 m (-DB), Hall 4514 (NBG;

STE); Gif berg near top of pass, (-DD), Boucher s.n. (STE);

610 m, Oliver s.n. (STE); Gifberg, 610 m (-DC/DD), Drege

7803 (BM; BOL; E; G; K; MO; P; W); Drege s.n. (G-DC);

910 m. Compton 20885 (NBG); Phillips 7585 (BOL; K; SAM).

Without locality: Thunberg s.n. (UPS: S); Thunbergl (LINN;

K; LU). Doubtful localities: Clanwilliam, Bayliss s.n. (NBG);

Cedarberg Middelberg, Leipoldt 741 (BM; BOL; K; SAM);

Nortier Reserve, Clanwilliam, Leipoldt 4365 (BOL).

This subspecies is distinguished by the canopu-

bescent leaves, the hairs being crisped and not

glandular, the crisped pubescence on the bracteoles,

with rarely short stout hairs, and the relatively short

calyx hairs, the hairs being simple and straight to

crooked and irregularly plumose with spreading

branches (Fig. 12).

Variation within the subspecies occurs in the

width of the sepals where the broad form ( Lavis

19811) merges into what was G. dregeana recorded

as a single collection from the Gifberg. Similarly

variation in the pubescence on the ovary apex and

the presence or absence of anther awns provided a

gradation with G. dregeana.

There are three groupings of populations within

the subspecies. The northern group occurs on the

Niewoudtville Plateau and has no plumes on the

leaves nor cilia on the calyx apices. The two southern

populations on the Gifberg and at Lockenberg are

intermediates between subsp. ciliaris and subsp.

bolusii in occasionally having plumes on the cilia

and a few cilia on the sepal apices. Due to the large

spatial separation, a hybrid origin is ruled out.

Another line of relationship occurs with subsp.

ciliciiflora of the Citrusdal area. From this latter it

differs in the type of plumose sepal cilia and in

having short sepal cilia. Sometimes the leaves of

these two subspecies are remarkably similar in having

a crisped retrorse indementum and no cilia.

G. ciliaris ( G . thunbergii Rach) was incorrectly

assessed by Brown, who judged the corollas to be

without any distinct constriction in the middle. This

condition is apparent in the dried material which,

when thoroughly boiled, sometimes shows a slight

constriction. However, all fresh material examined

possessed distinctly constricted corollas.

This subspecies is most closely related to G. incana

from the flats near Cape Town — an unusual distri-

butional relationship. It is distinguished by the

calyx hairs being as long as or longer than the width

of the sepals and, if equal, with a distinct tuft of

apical crisped hairs.

(b) subsp. bolusii ( N.E . Br.) E. G. H. Oliver ,

comb, et stat. nov.

G. bolusii N.E. Br. in FI. Cap. 4,1: 340 (1906). Type: Moun-

tains near Pakhuis Pass, Bolus 8681 (BOL, holo ! ; K!; NH!;

PRE!; STE!; Z!).

Branches puberulous to pilose with simple hairs,

sometimes with short stout plumose hairs admixed.

Leaves extremely variable, 1,2-3, 5 mm, linear-

lanceolate to ovate or obovate, pubescent, sometimes

with dense crisped hairs or almost lanate becoming

glabrous, ciliate with stout plumose hairs or jus,

compound tufts, rarely only pubescent, cilia variablet

mostly straight with few to many long spreading

plume branches, eglandular, often falling off and

remaining as stubs. Bracteoles sub-equal to very

Fig. 14. — Grisebachia ciliaris: subsp. bolusii. 1, flower, x8; 2, lateral bracteole, Xl6; 3, median bracteole, X 16; 4,

leaf, x 16; all drawn from the holotype, Bolus 8681 (BOL); 5, leaf, x 16, drawn from Compton 9634 (STE); 6, leaf,

x 16, drawn from Lewis 2725 (STE): subsp. involuta. 7, median bracteole, x 8; 8, sepal, x 8; 9, leaf, x 16; all drawn

from Bodkin sub Bolus 8680 (BOL).

E. G. H. OLIVER

79

unequal, adpressed to the calyx, the median elliptic to

broadly elliptic or ovate, 2,0-3,2x 1 ,0-1 ,7 mm,

with broad flat base and relatively small but distinct

keel-tip, the laterals oblong-elliptic to obovate

sometimes oblique, glabrous to puberulous, ciliate

in the upper half with stout plumose hairs. Sepals

2, 1-3,0x0,9-20 mm, oblong-elliptic to very broadly

elliptic, pubescent rarely subglabrous, ciliate with

short to long stout plumose hairs, rarely subplumose,

eglandular, plume branches spreading, relatively

long, rarely short and erect, with similar hairs on

the abaxial surface, ciliate at the apex and with an

apical tuft of straight not lanate hairs. Corolla 3-4

mm long, pubescent to pilose inside and outside in

the middle region. Anthers muticous rarely minutely

awned. Ovary glabrous. Fig. 14.1-14.6.

Cape. — 3218 (Clanwilliam) (-BB) & 3219 (Wuppertal)

(-AA): Pakhuis Pass, Barker 5622 (BOL; NBG); Barker 9426

(NBG); Barker 9426 (NBG); Bolus 8681 (BOL; K; NH; PRE;

STE; Z); 610 m, Compton 4319 (BOL; NBG); 914 m, Compton

6864 (NBG); Compton 6927 (NBG); 610 m, Compton 7738

(NBG; STE); 760 m, Compton 9584 (PRE; STE); Esterhuysen

3225 (BOL; K; PRE; SAM); 760 m, Esterhuysen 23753 (BOL;

STE); 610 m, Levyns 3937 (CT); 760 m, Levyns 5052 (CT;

PRE); Middlemost 1731 (NBG; STE); Salter 2765 (BM; BOL;

K); 3219 (Wuppertal): Heuningvlei (-AA), Andrag 130 (STE);

Pakhuis Pass, east slopes (-AA), Lewis 2725 (SAM; STE);

760 m (-AA), Marsh 351 (K; PRE; STE); 914 m (-AA),

Oliver 3002 (PRE; STE); Boontjiesvlei (-AA), Stokoe in SAM

55195 (NBG; PRE; SAM); Pakhuis Pass to Heuningvlei, 96o

m (-AA), Taylor 8537 (STE); 990 m, Williams 899 (NBG);

Bothasberg, 600-900 m (-AA), Thorne in SAM 52644 (SAM).

This subspecies is distinguished by having a large

median bracteole more than 2x1 mm but less than

3,2 mm long with a broad base and relatively small

keel-tip and with marginal cilia, leaves mostly ciliate

with stout plumose hairs at least when young, sepals

less than 3 mm long with very plumose stout hairs

with spreading branches.

This is a very variable taxon particularly as to the

leaves, some being close to subsp. ciliaris with the

short crisped indumentum and only a few tufted

cilia. The majority of specimens has awnless anthers

whereas those of subsp. ciliaris are mostly awned

except in the southern populations. The bracteoles

and leaves with distinct cilia serve to distinguish

subsp. bolusii. The larger linear-leaved form tends

towards subsp. involuta. A superficial similarity

exists between subsp. bolusii and G. plumosa subsp.

pentheri which occurs on the west side of the Olifants

River valley. The free, as opposed to joined, sepals

with numerous or few abaxial hairs serve to distinguish

the two taxa.

Subsp. bolusii is very restricted in its distribution

occurring only in the area around the summit of the

Pakhuis Pass where, as can be expected, it has been

frequently collected.

(c) subsp. involuta ( Klotzsch ) E. G. H. Oliver,

comb, et stat. nov.

G. involuta Klotzsch in Linnaea 12: 227 (1838); Benth in

DC., Prodr. 7: 701 (1839); Drege, Zwei Pfl. Docurn. 72 (1843);

N.E. Br. in FI. Cap. 4, 1 : 340 (1906). Type: Boschkloof, Drege

s.n. (B, holo.f; possible isos. G!; G-DC!) distributed as 7801

(BOL!; K!; P!; S!;). Lectotype: Drege 7801 (K).

Branches pubescent with simple hairs. Leaves

adpressed, up to 4,5 mm long, mostly lanceolate,

straight, sparsely puberulous becoming glabrous,

ciliate with short stout plumose hairs which become

setae. Bracteoles unequal, median, adpressed to the

calyx, the median 4-5x1, 7-2, 3 mm long elliptic

to ovate with an expanded flat base and distinct

keel-tip, the laterals about 3 mm long oblong with

a slight keel-tip, all glabrous except for a few hairs

on the keel-tip, ciliate with long slightly plumose

straight hairs, the plume branches very small and

pointing towards the apex of the cilium. Sepals

4, 3x1, 9-2, 3 mm broadly elliptic, slightly keel-

tipped, glabrous, with apical tuft of short straight

hairs, ciliate with long straight slightly plumose

hairs, plume branches very small and pointing

towards the apex of the cilium. Corolla 6-7 mm long,

constricted two-thirds of the way up, pubescent in

the middle region outside, villous inside at the

constriction. Anthers about 1,5 mm long, muticous,

scabrous. Ovary glabrous. Fig. 14.7-14.9.

Cape. — 3219 (Wuppertal): Krakadouw (-AA), Bodkin sub

Bolus 8680 ( BM; BOL; K; PRE; STE); Boschkloof at the foot

of Blaauwberg, 300-600 m (-AA), Drege 7801 (BOL; K; P;

S); Drege s.n. (G-DC); Pakhuis mountains (-AA), MacOwan

3268 (SAM).

This subspecies is characterized by its overall

larger flowers and inflorescence, the bracteoles being

longer than 4,2 mm and the sepals longer than 4 mm,

both with a broad base and relatively small keel-tip.

The larger size is the only differentiating character

between this subspecies and some forms of subsp.

bolusii. It also has close similarities with some forms

of subsp. ciliciiflora but, again, the size difference

is pronounced and the bracteole shape slightly

different. The sepal cilia (Fig. 12) are more closely

related to those of subsp. ciliciiflora than those of

subsp. bolusii.

Subsp. involuta is very restricted in its distribution

possibly occurring in only one or two populations

on the western side of the Krakadouw range. No

recent collections have been made. These populations

are allopatric to those of subsp. bolusii and subsp.

ciliciiflora making interchange of genetic material

highly improbable.

(d) subsp. ciliciiflora ( Salisb .) E. G. H. Oliver,

comb. et. stat. nov.

Erica ciliciiflora Salisb. in Trans. Linn. Soc. 6: 339 (1802).

Type: Hottentots-Holland, Masson s.n. (BM, holo!). Blaeria

ciliciiflora (Salisb.) G.Don, Syst. Veg. 3: 805 (1834). G. cilicii-

flora (Salisb.) Druce in Rep. Bot. Soc. Exch. Club Brit. Isl.

1916: 625 (1917).

G. velleriflora Klotzsch in Linnaea 12: 227 (1838); Benth. in

DC., Prodr. 7:701 (1839); N.E. Br. in FI. Cap. 4,1:341

(1906). Type: between Twenty-four Rivers and Olifants River,

Drege s.n. (B, holo.f; P!).

G. zeyheriana Klotzsch in Linnaea 12:227 (1838); N.E. Br.

in FI. Cap. 4, 1 : 342 (1906). Type; near Olifantsrivier and Farm

Brakfontein, Ecklon & Zeyher s.n. (B, holo.f) distributed as

269 (G ! ; K!; LU!; MEL!; MO!;P!;S!; SAM!; W!).

G. apiculata N.E. Br. in FI. Cap. 4,1:341 (1906). Type:

mountains near Piekenier’s Kloof, Schlechter 4969 (K, holo!;

BM!; BOL!; G ! ; P!; PRE!; S!; STE!; W!; Z!).

Branches pubescent with erect to retrorse short

simple hairs. Leaves adpressed, 1, 5-3,0 mm long,

linear to ovate, mostly pubescent with adpressed

crisped hairs becoming somewhat glabrous on the

abaxial surface, rarely glabrous when young, occasion-

ally ciliate with short stout gland-tipped hairs or

just apiculate. Bracteoles remote or median, unequal

to subequal, slightly recurved, 1 ,4-2,3 x 0,5-0, 8

mm, oblong to elliptic, the laterals linear to linear-

elliptic, with a distinct keel-tip and flat base, pubescent

to subglabrous but with an apical tuft of crisped

hairs, ciliate with long stout subplumose hairs;

pedicel long pilose. Sepals 1 ,6-2, 5x0, 6-1 ,3 mm,

oblong to broadly elliptic with a slight keel-tip

pubescent to glabrous with or without an apical tuft

of straight hairs, ciliate with long straight plumose

hairs, plume branches small forward pointing,

rarely subplumose, rarely gland-tipped. Corolla

2, 5-3,0 mm long, distinctly constricted and 4-angled

Fig. 15. — Grisebachia ciliaris subsp. ciliciiflora. 1, corolla, x8; 2, leaf, Xl6; both drawn from Oliver 4017 (STE); 3,

lateral bracteole, x8; 4, median bracteole, x8; 5, sepal, x8; 6, anther, front, side and back views, x 16; 7, ovary,

X 16; all drawn from the holotype, Masson s.n. (BM); 8, leaf, X 16, drawn from Stokoe in SAM 54847 (STE); 9,

leaf, X 16, drawn from Schlechter 4969 (STE).

at the base, pubescent to pilose outside and inside

in the middle region. Anthers about 1 mm long,

muticous long scabrous; pore half the length of the

cell. Ovary glabrous. Fig. 15.

Cape. — 3218 (Clanwilliam): Between Berg and Lang Valley

(-BC), Acocks 2995 (S); Boekenberg, 54?T m (-BC), Compton

4947 (BOL); Alexandershoek, 90-120 m (-BC/BD), Schlechter

5127 (BM; BOL; K; P; PRE; SAM; STE; W; UPS; Z);

Lambertsboekberg, 910 m (-BD), Compton 5492 (BOL; NBG);

Between Witelskloof & Lambertshoekberg, (-BD), Pillans 9057

(BOL; K; PRE); Piekenierskloof, 365 m, (-DB), Schlechter

4969 (BM; BOL; G; K; P; PRE; S; STE; W; Z). Without

precise locality: Clanwilliam, Leipoldt 213 (BOL; K; PRE;

SAM); Clanwilliam, Mader 181 (BOL; K). 3219 (Wuppertal):

Citadelkop near Wuppertal (-AA), Compton 24264 (NBG;

STE); Brakfontein, 300 m (-AC) Adamson sub Levyns 1320

(CT); by the Olifants River and near Brakfontein, (-AC),

Ecklon &Zeyher 269 (G; K; LD; MEL; MO; P; S; SAM; W);

Elandskloof, 610 m (-CA), Compton 5325 (BOL; NBG);

1370 m, Compton 16128 (NBG; STE); Compton 20965 (BOL;

NBG; PRE; STE); Leipoldt in BOL 21655 (BOL); Stokoe in

SAM 54847 (SAM; STE); Elandskloof Pass, (-CA), Hdfstrom

& Acocks 1043 (PRE; S; STE); Waterfall between Citrusdal &

Elandskloof, (-CA), Stokoe 7712 (BOL; NBG; NH; PRE);

Williams sub Baker 1821 (BM); Kleinfontein east of Citrusdal

762 m, (-CA), Oliver 4019 (E; K; MO; P; PRE; STE); Allan-

dale, south-east of Citrusdal, 548 m (-CA), Oliver 5007 (BM;

G; S; STE); near Citrusdal (-CA), Rust s.n. (STE); Warmbaths

to Modderfontein, (-CA), Stephens 7042 (BM ; BOL; K; SAM);

between Twenty-four Rivers and Olifants River, 300 m (-CC/

3319 AA), Drege 1179 (P). Without locality: Drege s.n. (BM;

BOL; G; K; MO; W); Forsyth s.n. (BOL; K;); Masson s.n.

(BM; G-DC).

This subspecies is characterized by the small

usually subequal narrow-based bracteoles with con-

spicuous long plumose cilia and by the leaves, when

glandular, with the glands apical or marginal only

and usually short-stalked to sessile and by the sepals

with long plumose hairs having very small erect

branches.

Relationships with the other subspecies are in

three directions and are somewhat difficult to explain

in the case of subsp. ciliaris due to the geographical

isolation of the latter. The relationships with subsp.

involuta and subsp. multiglanclulosa are understandable

due to the reasonably close proximity of the popula-

tions.

There is considerable variation within this sub-

species which necessitated the inclusion of G. apiculata

and G. zeyheriana in the synonomy. The typical

form of subsp. ciliciiflora possesses leaves with a

crisped, sometimes retrorse, indumentum and calyx

with numerous long sparsely plumose hairs with

plume branches forward pointing. In some forms the

leaves possess sessile or subsessile marginal glands

and a large apical gland.

Subsp. ciliciiflora occurs in the Citrusdal area

mostly at lower altitudes on sandy open patches

and on the mountains north-west of the town. The

locality near Wuppertal, Compton 24264, is unusual

and inexplicable.

(e) subsp. multiglandulosa E. G. H. Oliver,

subsp. nov., similis subspecie ciliciiflorae, sed dis-

tinguitur piliis longis glandulis marginibus et paginis

abaxialibus foliorum bracteolarumque.

Type. — Cape, Olifants River Valley above Toorgat

on the farm Grootfontein, Oliver 3972 (STE, holo.l;

K! ; MO!; NBG!; PRE!).

Branches puberulous to subglabrous with stouter

gland-tipped hairs admixed. Leaves recurved-spreading

sometimes straight and adpressed, 1,5-4, 5 mm long

with petiole 0,5 mm long, lanceolate to ovate mostly

glabrous and shiny on the abaxial surface and sparsely

puberulous on the adaxial surface, rarely entirely

glabrous, occasionally puberulous all over when

young, ciliate and clothed on the abaxial surface

with short to long stout simple gland-tipped hairs,

sometimes those on abaxial surface falling off in

erect leaves. Bracteoles median to remote, subequal to

unequal, 0,8-2, 5 mm long the laterals mostly 1,0

mm long, linear to oblong-ovate to oblong-elliptic

with an enlarged keel-tip, glabrous to pilose in the

lower half, sometimes crisped at the apex, ciliate with

stout gland-tipped simple hairs with a few on the

keel-tip; pedicel up to 2,5 mm long, sparsely puberu-

lous with simple and gland-tipped hairs. Sepals

1 ,9-2, 5x0, 5-0, 8 mm oblong to oblong-ovate with

a slight keel-tip, glabrous, rarely with a few scattered

hairs, ciliate with long straight subplumose to simple

hairs with similar hairs on the abaxial surface,

hairs often gland-tipped. Corolla about 4,5 mm

long, distinctly 4-angled, pubescent below the con-

striction sometimes sparsely so and confined to the

angles, villous inside. Anthers about 0,8 mm long,

scabrous, muticous. Ovary glabrous. Fig. 16.

E. G. H. OLIVER

81

Fig. 16. — Grisebachia ciliaris

subsp. multiglandulosa. 1,

flower, x8; 2, lateral

bracteole, xl6; 3, median

bracteole, x 16; 4, leaf, x 16;

all drawn from the holotype,

Oliver 3792 (STE).

Cape. — 3218 (Clanwilliam): Grey’s Pass, 457 m (-DB),

Levyns 1367 (CT; SAM). 3219 (Wuppertal): Olifants River

Valley, Keerom (-CC), Esterhuysen 17889 (BOL); east slope of

Grassruggens, 640 m (-CC), Oliver 3992 (STE); Pillans 8785

(BOL; PRE; STE); Olifants River Valley above Toorgat on

Grootfontein, 548 m (-CC), Oliver 3972 (K; MO; NBG; PRE;

STE); Grootfontein, Ratel River, 426 m (-CC), Oliver 3987

(E; PRE; STE); Porterville mountains, Berghof, 790 m (-CC),

Oliver 3934 (BM; G; NBG; PRE; STE; Z); 3941 (K; STE; W).

This subspecies may easily be recognized by its

leaves which are erect to recurved-spreading mostly

glabrous but with distinct long gland-tipped hairs

on the margins and abaxial surface, by its small

bracteoles up to 1, 8x1,0 mm which have stout

subplumose to simple hairs on the margins and

abaxial surface of the keel-tip and by the simple to

subplumose eglandular or gland-tipped hairs on the

sepals.

The material available varies somewhat in floral

and foliage characters. The leaves are always gland-

ciliate with long stout hairs on the margins and

abaxial surface and are mostly distinctly recurved-

spreading. But a few specimens have erect adpressed

leaves like those of subsp. ciliciiflora. The calyx

cilia may be simple or occasionally plumose with

plume branches like those in subsp. ciliciiflora. It

was found that the only distinguishing character is

the presence of abaxial hairs on the leaves and

bracteoles in subsp. multiglandulosa.

Subsp. multiglandulosa is confined to the mountains

at the southern end of the main Olifants River

valley where it occurs mainly on sandy open flat

areas. The majority of populations is allopatric to

those of subsp. ciliciiflora, occurring at high altitude

only. In the region of Piekenierskloof there is,

however, an overlap. The specimen, Levyns 1367,

is an intermediate very similar to the type and only

collection of G. apiculata (subsp. ciliciiflora). Un-

fortunately all the populations appear to have been

removed in this area by agriculture thus making a

study of the populations impossible.

3. Grisebachia incana ( Bartl .) Klotzsch in Linnaea

12:225 (1838); Benth in DC., Prodr., 7:701 (1839);

N.E. Br. in FI. Cap. 4,1:344 (1906).

Blaeria incana Bartl. in Linnaea 7: 650 (1832). Type: On flats

below Tigerberg at Rietvallei, Ecklon s.n. (B, holo.t; P!).

Lectotype: Ecklon s.n. (P).

G. alba N.E. Br. in FI. Cap. 4, 1 : 344 (1906). Type: without

locality, Grey s.n. (K!).

Small compact shrublets to 30 cm high. Branches

pubescent to tomentose with reflexed hairs, occa-

sionally with stout plumose to gland-tipped sub-

plumose hairs in between. Leaves 3-nate adpressed,

1, 5-2,0 mm long, elliptic to oblong-obovate, pube-

scent becoming glabrous on the abaxial surface,

ciliate with a few very short stout plumose hairs or

with short stout gland-tipped hairs; petiole very

short, pubescent, sometimes with gland-tipped hairs.

Flowers in small terminal heads of 3-6 (9) on the ends

of lateral branchlets, pink, occasionally white; pedicel

up to 1,0 mm long pubescent; bracteoles subequal

to unequal, median but adpressed, 1,0-1, 4 mm

long, narrowly oblong to elliptic oblong often with

an enlarged keel-tip, acute or obtuse, the laterals

linear, pubescent, ciliate with short stout plumose

hairs or subplumose gland-tipped hairs. Calyx

4-partite; lobes 1,2-1,8x0,3-0,65 mm, mostly

narrowly oblong, occasionally elliptic-oblong or

linear, acute, pubescent with longer straight hairs

at the apex, ciliate with stout plumose eglandular

hairs to ciliate with stout subsimple gland-tipped

hairs, all shorter than the width of the lobe, some-

times with similar but shorter hairs on the abaxial

surface. Corolla 2, 4-2, 7 mm long, constricted at the

middle, ellipsoid below, urceolate above, pilose to

villous in the middle region and slightly up the

back of the lobes, pilose inside around the constric-

tion; lobes very broad, obtuse, erect-spreading.

Stamens 8, free; filaments linear, much dilated at

the point of attachment, sparsely pilose; anthers

manifest, about 0,7 mm long with oblong parallel

to spreading cells, scabrid edged, aristate; awns

small to obsolete, arising from the apex of the fila-

ments, scabrid; pore relatively small about one quarter

the length of the cell. Ovary 2-celled with a single

pendulous ovule in each cell, compressed, broadly

ovoid, pubescent on top and seated on a distinct

nectariferous disc; style filiform, glabrous, far

exserted; stigma simple to capitellate. Fruit a hard

verrucose nut. Figs 17 & 18.

A species forming small compact shrublets occurring

in sandy places on the flats between Sir Lowry’s Pass,

Kraaifontein and Mamre, flowering early from

April to July.

Cape. — 3318 (Cape Town), Flats east of Melkbosch (-CB).

Pillans 6671 (BOL; K); Between Melkbosch and Mamre (-CB/

DA), Salter 792 (BM; K); Groenekloof (-DA), Ecklon &

Zeyher 266 (G; MEL; MO; W); Mamre road near Melkbosch-

strand road (-DA), Levyns 9402 (CT); Mamre turnoff south-

west of Olifantskop, 76 m (-DA), Oliver 3753 (STE); Koeberg,

Baasariesfontein, 168 m, (-DA), Oliver 3756 (STE); Melkbosch,

7 km from the sea, (-DA), Wasserfall 167 (K; NBG; PRE);

Rietvallei (-CD/DC), Ecklon s.n. (P); Kraaifontein, 15-30 m,

(-DA), Diimmer 1553 (E); Gravel pits near Kraaifontein,

(-DA), Esterhuysen 18638 (BOL; NBG; PRE); Brackenfell

(-DA), Galpin 12679 (K; PRE; W); Scotsville, Kraaifontein

82

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

leaf, x 16, drawn from Oliver 3750 (STE); 8, median bracteole, x 16; 9, sepal, x 16; both drawn from Oliver 3746

(STE); 10, sepal, X 16, drawn from Stokoe 6062 (BOL); 11, leaf, x 16, drawn from Oliver 3746 (STE); 12, variation

in the hair types on the sepals, x 16.

Fig. 18. — Distribution of Grisebachia incana (3 are doubtful

localities).

(-DA), Markotter s.n. (STE); Koopmanskloof between

Bottelary and Kraaifontein (-DA), Oliver 3746 (STE); 3747

(STE); Base of Tygerberg (-DA), Pilians in BOL 17726 (BOL);

Brackenfell (-DA) Sailer 4395 (BM; BOL; K); Kraaifontein

(-DA), Strey 479 (PRE); Old Tygerberg (-DA), Zeyher s.n.

(SAM; PRE); Penhill Estate, Eersterivier, (-DC), Rain s.n.

(STE); Between the Cape Flats and Stellenbosch, (-DC/DD),

Bnrcliell 8344 (BOL; K; M; P; PRE; W). 3418 (Simonstown),

Sir Lowry's Pass, 305 m, (-BB), Lamb 4091 (SAM). Without

locality: Cape Town Flower Show, Stokoe 6062 (BOL; K;

NBG; PRE); Admiral Grey s.n. (K); Lehmann s.n. (K); Mund &

Maire s.n. (E; G; K; P; W). Doubtful localities: Vogelvlei,

Eck/on <5 Zeyher s.n. (HAM; LU; P; UPS; Z); Dutoitskloof,

Erege 6869 (BM; E; K; W); Hills behind Simonstown, Lamb

3223 (BOL; SAM); Kleinmond, Leipoidt s.n. (BOL; NBG;

P); mountains near Swellendam, Mund s.n. (BOL; K).

G. incana may be distinguished from related

taxa by its small flowers, small narrow sepals less

than 2,0x0,65 mm which have slender straight

cilia as long as but mostly shorter than the width of

the sepal and by the straight hairs forming the

apical tuft on the bracteoles and sepals.

The species affords a good example of geographical

vicarism with its closely related species which are

well separated spatially, e.g. G. ciliaris subsp. ciliaris,

G. rigida and G. nivenii.

Difficulty was experienced in distinguishing G.

incana from G. ciliaris subsp. ciliaris. The former

occurs only on the sandy coastal flats adjacent to

the Cape Peninsula, whereas the latter is confined

to the summits of mountains around Vanrhynsdorp

and Niewoudtville.

N. E. Brown unfortunately misinterpreted the

corolla shape in G. ciliaris subsp. ciliaris and so

isolated it from G. incana in his revision. Several

characters were examined in detail and found to

have a certain degree of disjunction and, when used

in combination, served to distinguish the two taxa.

In G. incana the leaves possess hairs which are mostly

erect as opposed to the crisped retrorse hairs in

G. ciliaris subsp. ciliaris. The leaves are usually

edged with short stout plumose hairs or gland-tipped

hairs, whereas in G. ciliaris subsp. ciliaris this rarely

occurs. The calyx in G. incana is mostly pilose with a

tuft of longer straight hairs at the apex and with cilia

shorter than the width of the sepal. In G. ciliaris

subsp. ciliaris the calyx is very sparsely puberulous

with a distinct apical tuft of long interwoven crisped

hairs and with cilia longer than the width of the sepal.

To the east there occur two closely related species,

G. rigida and G. nivenii, both in restricted separate

areas. G. incana differs from both these species in

the size of the sepals, which are less than 2,0x0,65

mm, and in the texture and indumentum of the

leaves. In the glandular form of G. incana the leaves

are very similar to those in G. rigida but are not so

inflated, are more pubescent and have the glands

confined to the margins.

The sepals in this species are the smallest and

narrowest in the genus; in one specimen being only

E. G. H. OLIVER

83

0,3 mm wide. This feature makes the corollas more

easily visible than in other species. The anthers are

unique in the genus in having the smallest pores

relative to the size of the cell.

Two fairly distinct forms occur in the material

so far collected. The specimens from the north

around Mamre have bracteoles and sepals with

more plumose cilia which are eglandular. Those

from the south in the Kraaifontein area have sub-

plumose gland-tipped hairs on the sepals, bracteoles

and on the leaves. This variation is, however, clinal

with no distinct disjunction between the two extremes.

N. E. Brown described G. alba from a single

collection made by Admiral Grey and based it on

the single character of white flowers including the

anthers. White-flowered forms of G. incana have

been collected, but these have had pale brown anthers.

In all other characters, G. alba is identical to the

glandular form of G. incana and is presumed to be

only an aberrant albino of this species.

G. incana is fairly restricted in its distribution

occurring only on the recent sand deposits on the

coastal flats at Sir Lowry’s Pass, Eerste River, Kraai-

fontein, below Tygerberg and near Melkboschstrand.

The records from Simonstown, Kleinmond, du

Toit’s Kloof and Vogelvlei are very doubtful.

4. Grisebachia rigida N.E. Br. in FI. Cap. 4, 1 :343

(1906). Syntypes: near Brand Vley, Schlechter 9926

(BM!; BOL!; Z!); mountains between French Hoek

and Villiersdorp, Bolus 5193 (BOL!; K!; PRE!; Z!).

Lectotype: Bolus 5193 (K).

Shrublets compact and low to erect up to 50 cm

high. Branches pubescent with simple recurved

hairs, very occasionally gland-tipped, rarely with

stout plumose hairs admixed. Leaves 3-nate up to

3, Ox 1,0 mm, mostly elliptic to narrowly elliptic,

occasionally ovate or oblong-obovate, thick and

fattened, pubescent or minutely scabrous on the

abaxial surface, rarely glabrous and shiny, pubescent

on adaxial surface, ciliate with 7-9 short stout

gland-tipped hairs and with some scattered over the

abaxial surface, petiole very short, shortly glandular

pubescent. Flowers 1-8-nate on the ends of lateral

branchlets, pink, rarely white; pedicels about 1 mm

long, pubescent with some plumose hairs at the apex;

bracteoles equal to slightly unequal with the median

slightly broader, adpressed to the calyx, ovate to

oblong-ovate to elliptic to narrow oblong-elliptic,

obtuse, glabrous or sparsely pubescent, ciliate with

stout simple to very slightly plumose hairs which are

mostly gland-tipped, rarely sparsely pilose inside,

keel-tipped. Calyx 4-lobed, slightly joined at the

base; lobes ovate-elliptic to oblong-elliptic to broadly

elliptic up to 2, 1-2,8 x0, 9-2,0 mm, often with

incurved margins, subacute, keel-tipped, glabrous

or sparsely pubescent mostly in the lower half,

ciliate with broadly based stout hairs almost fimbriate

in places and with similar hairs up the centre of the

abaxial surface, hairs mostly simple or very slightly

plumose, rarely gland-tipped, often crooked. Corolla

up to 4,4 mm long often oblique, distinctly con-

stricted in the middle; tube up to 3 mm long globose

ellipsoid, spreading above the constriction, 4-angled,

puberulous outside with glabrous patches opposite

the sepals, pilose on the inside mainly at the point of

constriction; lobes erect-spreading slightly crenulate

and emarginate about 1 mm long, very broadly

obtuse, pubescent at the base in the middle. Stamens

4; filaments linear, sparsely to densely pilose, up to

2,5 mm long; anthers manifest, attached dorsally

one third the way up, variable in size, 0, 7-1,1 mm

long with oblong to obovate parallel or spreading

cells, almost glabrous to scabrous, occasionally with

some long transparent hairs on the edges, muticous

Fig. 19. — Grisebachia rigida. 1, flower, x8; 2 ovary, x 16; both ^ idTedian^je:

variation drawn x 16: a, Oliver 3297 (STE); b, Salter 4784 (BOL); c, Van Breda 628 (PRE).

84

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

or aristate; awns up to 0,4 mm long or one third

the length of the cell, arising from the filament

apex, spreading laterally to descending, minutely

scabrous; pollen grains single. Ovary 2-celled with a

single pendulous ovule in each cell, 0,6x0,8-0,9x0,9

mm, broadly ovoid to ellipsoid, compressed, obtuse,

variously pilose at the apex, very unevenly wrinkled;

style up to 4 mm long, glabrous, exserted; stigma

slightly capitellate. Fruit a hard verrucose nut.

Fig. 19.

Cape. — 3319 (Worcester), south-west end of Brandvlei,

213 m, (-CB), Oliver 5038 (STE); near Brandvlei, 304 m,

(-CB), Sehlechter 9926 (BM; BOL; Z); Brandvlei Kelders,

(-CB), van Breda 628 (B; PRE); between Worcester and

Villiersdorp (-CB/CD), Compton 22918 (NBG); flats north-east

of Wabooms Farms, 200 m (-CD), Oliver 3299 (STE) ; Kwaggas-

kloof near Moordkuil, 200 m (-CD), Oliver 3297 (MO; PRE;

STE); 13 miles north of Villiersdorp (-CD), Salter 4784 (BM;

BOL; K; SAM); Doornrivier, 304 m (-CD), Walters 1012

(K; PRE; STE); mountain slopes between Villiersdorp and

French Hoek (-CC/CD/ 3419 AA/AB), Bolus 5193 (BOL; K;

PRE; Z). Flowers from August to October.

G. rigida is characterized by its free or slightly

joined sepals which are more than 2,0x0,65 mm

with very broad flattened subplumose or simple

cilia and by its fattened leaves which are ciliate

with short gland-tipped hairs and with a few similar

hairs on the abaxial surface.

The species differs from G. nivenii in the leaf

cilia, less plumose calyx and glabrous inner surface

of the sepals. It is closely related to G. incana from

which it is easily distinguished by its broader sepals

with their broad cilia and by the leaves.

G. rigida varies in the size of the sepals where in

the type, Bolus 5 1 93, they may be as much as 2 , 7 x 2 , 0

mm. The anthers also vary in size, shape and in the

occurrence of awns. A few specimens have anthers

with long colourless hairs, a feature very rarely

seen in the Ericoideae.

The species occurs on the recent sandy alluvial

flats at the eastern base of the Stettyns Mountains

between Worcester and Villiersdorp where isolated

pockets of fynbos grow (Fig. 20). The surrounding

area possesses mountain renosterveld on the shales

and Wittenberg quartzites. In this area the species

is very susceptible to extinction due to encroaching

agriculture and burning.

Bolus’s record between French Hoek and Villiers-

dorp has not been reconfirmed. Although somewhat

removed from the main populations and coming

from a completely different valley system, this record

could be correct due to the numerous sandy alluvial

patches in the area.

5. Grisebachia nivenii N.E. Br. in FI. Cap.

4,1 :343 (1906). Syntypes: Hottentots-Holland, INiven

128 (BOL!; K!); near Zondereinde River, Gill s.n.

(K!); near Swellendam, Mund 3 (K!; PRE!). Lecto-

type : "l Niven 128 (K).

G. ciliaris Benth. in DC. Prodr., 7: 701 (1839), non Klotzsch.

Type: in prov. Swellendam occidentali, collector not cited.

Shrublets compact, erect, up to 50 cm high. Branches

minutely pubescent with retrorse hairs. Leaves

3-nate, adpressed, up to 1,7x1, 2 mm, broadly

ovate to elliptic, very rounded and thick, obtuse or

acute, glabrous on the abaxial surface, pubescent

on the adaxial surface, ciliate with 5-6 short stout

plumose cilia; petiole very short and broad, ciliate.

Flowers in terminal globose heads of 3-8 on the ends

of lateral branchlets, pink, rarely white; pedice’s

very short, sparsely pubescent, sometimes with stout

plumose hairs at the apex; bracteoles adpressed,

Fig. 20. — Distribution of % Grisebachia rigida and

■ Grisebachia nivenii.

about 1,3 mm long, subequal, rarely markedly

unequal, mostly oblong-elliptic sometimes the laterals

obliquely so, the median rarely angular-ovate, all

keel-tipped, obtuse or acute, puberulous outside and

inside, ciliate with stout plumose hairs. Calyx 4-lobed,

sometimes slightly joined at the base; lobes broadly

elliptic to narrowly oblong-elliptic, 2,1-2,9x0,65-1,3

mm, acute, keel-tipped, puberulous on the abaxial

surface mainly towards the base, sparsely puberulous

on the inside, ciliate with stout broad plumose hairs

with similar hairs over the adaxial surface towards

the centre, rarely gland-tipped. Corolla up to 4,0

mm long, sometimes oblique, distinctly constricted

in the middle; tube up to 3 mm long, globose ovoid

to ellipsoid, urceolate above the constriction, pube-

scent to pilose in the middle and lower part and up

the back of the lobes and inside the tube around

the constriction; lobes broadly obtuse, erect or

slightly spreading. Stamens 4; filaments linear,

sparsely to densely pilose; anthers manifest, about

0,8 mm long, oblong, dorsifixed one third of the

way up, papillate, aristate ; pore about a third of the

length of the cell ; awns small, spreading to deflexed,

arising from the filament at the point of attachment

to the anther; pollen grains single. Ovary 2-celled

with a single pendulous ovule in each cell, 0,7-0, 6

mm, ovoid compressed, obtuse sparsely pilose at the

apex; style exserted, up to 4 mm long; stigma capitel-

late or subsimple. Fruit verrucose, hard. Fig. 21.

A species forming compact erect shrublets up to

50 cm high occurring in a very restricted area of

sandy flats south-east of Swellendam, flowering from

July to September.

Cape. — 3420 (Bredasdorp), Bontebok National Park, 90 m

(-AB), Acocks 22371 (PRE; STE); 100 m, Barnard 583, (STE);

581 (PRE), 100 m, Liebenberg 6464 (B; PRE; STE); 90 m,

Oliver 1519 (NBG ; STE) ; Oliver 4305 (STE) ; 4306 (NBG ; PRE ;

STE); sandy places near Swellendam, 180-240 m (-AB), Mund

3 (K; PRE); Buffeljagsrivier (-BA), Zeyher 3330 (SAM).

Without precise locality: Zondereinde River, Gill s.n. (K).

Doubtful locality: Hottentots-Holland, alpine rock places

Niven or Masson 128 (K). (cf. Fig. 20).

G. nivenii may be recognized by its very fattened

shiny adpressed leaves which are glabrous on the

E. G. H. OLIVER

85

Fig. 21. — Grisebachia nivenii. 1, flower, x8; 2, corolla varia-

tion, x8; all drawn from Oliver 4305 (STE); 3, median

bracteole, xl6; 4, sepal Xl6; 5, anther, front and side

views, x 16; all drawn from the lectotype, Niven 128 (K);

6, ovary, x 16; 7, leaf, x 16; both drawn from Oliver 4305

(STE).

abaxial surface and have 5-6 very short stout plumose

cilia and by its more or less free sepals which have

short stout flattened plumose cilia.

The species is closely allied to G. rigida and G.

incana. It differs from G. rigida in having glabrous

shiny leaves, even when young, with plumose cilia

and no gland-tipped cilia. The calyx cilia are more plu-

mose than in G. rigida and there is a sparse pubescence

on the inside of the sepals. From G. incana it differs

in leaf details, the latter having pubescent nonfattened

leaves. The sepals in G. nivenii are broader, more

than 2,0x0,65 mm.

G. nivenii is geographically isolated as it occurs

only on the sandy flats in the hilly country south-

east of Swellendam where it is far removed from

its closest allies, G. rigida and G. incana. Like nearly all

the other species in the genus it is confined to a few

small patches of alluvial sand. The surviving popula-

tions in the area now lie within the boundaries of

the Bontebok National Park.

There is some confusion about the collector of the

lectotype of the species. It was labelled as “C.B.S.

Masson” but N. E. Brown changed this to Niven.

The handwriting is definitely not Niven’s, but matches

that on the type of Eremia brevifolia Benth. which

Brown cited as collected by Masson. Neither of these

labels exactly matches the handwriting of Masson

in the Kew Archives. Brown labelled this sheet as

the type.

Variation within the species is very slight. The

sepals in Zeyher 3330 are somewhat narrower than

in the other specimens and have some gland-tipped

hairs.

The G. parviflora/G. minutiflora group

Along the main chain of mountains and high

level plateaux running in a north-south direction

from the Cedarberg through the Cold Bokkeveld to

the Worcester District, there is a series of vicarious

taxa. Superficially they are very similar in their

low compact to spreading habit, small white sub-

calycine to calycine flowers, manifest anthers and

ciliate calyx lobes.

One very variable taxon (A) is widespread from

the Cedarberg to the mountains south of Worcester

and eastwards to near Swellendam, occurring on dry

stony slopes with short dry restiad/ericoid vegetation.

It usually forms a low compact to sprawling shrublet

with branches spreading amongst the restiads. It

consists of three distinct allopatric vicariads and has

had the names G. parviflora (Klotzsch) Druce ( G .

ermioides MacOwan) and G. similis N.E. Br. applied

to it.

The second taxon (B) is much more restricted,

occurring in the central and southern Cold Bokkeveld

on dry open sandy flats and forms a low compact

shrublet sometimes slightly sprawling and rooting

at the nodes. This has had the names G. minuti-

flora N.E. Br. and G. nodiflora N.E. Br. applied to it.

As with the taxon A, there are two distinct allopatric

vicariads in this taxon.

To my knowledge the two taxa only grow in

reasonably close proximity in the Hartebeeskloof

and Winkelhaak areas where I have observed them.

In the former locality the plants of taxon A ( G .

parviflora) were few and were outliers of the larger

populations higher up the rocky slopes. The plants of

taxon B (G. minutiflora) were locally common on open

sandy patches. In the latter locality taxon B was

flowering three months later than the early flowering

vicariad of taxon A.

The main morphological difference between these

two taxa lies in the form of the inflorescence. In A

the flowers are generally 1-4-nate at the ends of short

axillary branchlets which are often clustered together

in a pseudospike along the main and lateral branches.

In B the flowers are terminal on the ends of short

branchlets with up to 36 flowers forming a cluster.

These clusters usually hang downwards making the

plants less conspicuous than those of taxon A.

Except for one vicariad, plants of taxon B are glan-

dular particularly on the margins of the calyx. The

glands terminate plumose cilia. In taxon A the few

glands are confined to the short simple cilia on the

calyx.

Taking into account the habitat and morphological

differences, I have decided to regard the two taxa

A & B as closely related species referred to G. parvi-

flora and G. minutiflora respectively and to recognize

several subspecific taxa.

6. Grisebachia parviflora ( Kotzsch ) Druce in

Rep. Bot. Soc. Exch. Club Brit. Isl. 1916: 625 (1917).

Type : Hills between Puspas Valley and Kogmanskloof

mountains, Ecklon & Zeyher s.n. (Bf, holo.; isos.?).

Neotype: flats between Witsenberg and Skurfdeberg,

Zeyher 1117 (K!, neo. ; BOL! ; G! ; SAM ! ; STE! ; W!).

Low compact to spreading shrublets up to 20 cm

high, rarely up to 50 cm. Branches erect or spreading

often entwining among the surrounding vegetation,

with numerous short branchlets, pubescent sometimes

with glands admixed. Leaves 3-nate up to 3 mm long

with the petiole, erect to spreading, straight to

markedly recurved, linear to lanceolate rarely ovate,

acute, flat to trigonous, glabrous or at first puberulous,

ciliate sometimes with gland cilia or sessile glands,

sometimes gland-apiculate : petiole adpressed ciliate.

86

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Flowers l^f-nate at the ends of extremely short

branchlets arranged in a spike-like manner along

the branches; pedicels very short, less than 0,5 mm

long; bracteoles 3, equal to subequal, adpressed to

the calyx or slightly spreading, very variable in

size, up to 1,5 mm long, in shape from linear to

elliptic-oblong to ovate, acute to obtuse, glabrous or

pubescent ciliate with or without sessile or stalked

glands white. Calyx 4-lobed for its length,

campanulate 0,6- 1,8 mm long and 0,3-0, 8 mm

wide, glabrous or pubescent, white; lobes erect

variable in shape from elliptic-oblong to ovate to

subquadrate, the apex acute cuspidate or subtruncate

with an apiculus, with or without a distinct keel-tip

and with or without a distinct median ridge, ciliate,

with or without sessile or stalked glands. Corolla

4-lobed, up to 3x1,7 mm long, mostly 2x1,2 mm,

funnel-shaped sometimes broadening more above

the middle, nowhere constricted, occasionally tubular-

ellipsoid; tube glabrous inside sometimes glabrous

outside or puberulous in lower half to fully pubescent;

lobes erect to incurved, rarely slightly spreading,

obtuse, broader than long, glabrous entire. Stamens

4 included or manifest; filaments filiform glabrous;

anthers up to 1,1 mm long with oblong parallel

separate cells, basal, minutely scabrous, aristate;

awns the length of the cell to rudimentary, ciliate;

pore the length of the cell, pollen grains single.

Ovary 2-celled with a single ovule per cell in one

subspecies, rarely 2 ovules per cell and sometimes

3 cells, ovoid to ellipsoid, up to 1x0,7 mm, obtuse

glabrous to puberulous on the top, sometimes

thickly pubescent; style far exserted, straight or

curved, glabrous up to 4 mm long; stigma minutely

capitate. Figs 22 & 23.

A species generally low, sparse and sprawling in

habit, occasionally compact, occurring frequently on

dry stony slopes and flats on mountains from the

Cedarberg to Du Toit’s Kloof eastwards to near

Swellendam, flowering from as early as July to as late

as January.

A very variable taxon in which three subspecies are

recognized.

Key to the subspecies

Leaves erect to spreading, straight or slightly curved, calyx

lobes acute to obtuse:

Corolla tube glabrous or puberulous in the lower $, calyx

usually glabrous (a) subsp. parviflora

Corolla tube pubescent all over, calyx pubescent

(c) subsp. pubescens

Leaves markedly recurved, calyx lobes subquadrate, subtrun-

cate (b) subsp. eglandula

(a) subsp. parviflora

Eremia parviflora Klotzsch in Linnaea 12:488 (1938); N.E.

Br. in FI. Cap. 4,1: 334 (1905). Grisebachia parviflora (Klotzsch)

Druce in Rep. Bot. Soc. Exch. Club Brit. Isl. 1916: 625 (1917).

Erica shalliana Hort. Berol. ex Klotzsch in Linnaea 12: 498

(1838), nom. nudum.

Grisebachia eremioides MacOwan in J.Linn. Soc. 25:392

(1890); N.E. Br. in FI. Cap. 4,1:349 (1906); Compton in

J1 S. Afr. Bot. 1: 151 (1935). Syntypes: Witsenberg and Houw

Hoek Zeyher 1117, near Tulbagh Waterfall, MacOwan 2685

(SAM!); MacOwan sub Herb. Norm. 564 (BM!; BOL!; G1-

K!; P!; PRE!; SAM!; UPS!; W!).

G. eremioides var. pubicalyx N.E. Br. in FI. Cap. 4,1: 349

(1906). Type: Mountains of Tulbagh Kloof, Bolus 5304 (BOL!;

K!; PRE!; SAM!).

G. similis N. E. Br. in FI. Cap. 4, 1 : 350 (1906). Type: Cold

Bokkeveld, Schlechter 8896 (BOL!; K!; numerous isos!).

G. similis var. grata N.E. Br. in FI. Cap. 4,1: 350 (1906).

Type: Cedarberg, near Sneeukop and Wuppertal, Bodkin sub

Bolus 8628 (BOL!; PRE!).

Low erect to spreading shrublet. Leaves erect to

spreading, straight or curved. Calyx glabrous to

sparsely puberulous; lobes ovate, elliptic-oblong to

broadly oblong, apex acute to cuspidate with a

distinct keeltip and median ridge. Corolla up to

2x1,2 mm; tube glabrous to puberulous in the

lower half rarely up to f of the length. Ovary glabrous

to puberulous on the top. Fig. 22.1-22.6.

Cape. — 3219 (Wuppertal): Middelberg Hut (-AC), Barnes

in Bol 19485 (BOL); Sneeukop 1070 m (-AC), Bodkin sub

Bolus 8628 (BOL; PRE); Middelberg Plateau (-AC), Bond 1359

(NBG); Juriesberg (-AC), Compton 6269 (NBG); 7023 (NBG:

STE); Sneeukop 1680 m (-AC), Compton 6270 & 6271 (NBG);

Cedarberg Tafelberg (-AC), Esterhuysen 8158 (BOL); Langberg

1520-1820 m (-AC), Esterhuysen 7315 (BOL; K; PRE);

Hoogvertoon near Sneeuberg hut, 1280 m (-AC), Haynes 815

(STE); Middelberg (-AC), Levyns 2906 (CT); Lewis in BOL

22217 (BOL); Sneeuberg (-AC/CA), Pocock 393 (BOL; STE);

Fig. 22. — Grisebachia parviflora : subsp. parviflora. 1, flower, x 16; 2 corolla, X 16; 3, bracteoles, x 16; 4, sepal, x 16;

5, anther, side, back and front views, X 16; all drawn from an isolectotype, Zeyher 1117 (STE): subsp. eglandula.

7, flower, x 16; 8, sepal x 16; both drawn from Esterhuysen 5924 (PRE): subsp. pubescens. 9, flower, x8; 10, sepal,

x 16; 11, anther, x 1 6 ; all drawn from the holotype, Oliver 4310 (STE); 12, sepal, xl6, drawn from Oliver 4312

(STE).

E. G. H. OLIVER

87

Fig. 23. — Distribution of Grise-

bachia parviflora. % subsp.

parviflora-, © subsp. eglan-

dula\ (J subsp. pubescens\

Variation in floral and sepal

characters for a selection of

specimens is shown. 1,

Stokoe 6567 ; 2, Zeyher 1117;

3, Lewis sub BOL 22009; 4,

Compton 6625; 5, Haynes

815; 6, Esterhuysen 5924; 7,

Schlechter 8818; 8, Oliver

4310; 9, Middlemost 2246;

10, Oliver 311. All X 10.

Koupoort near Krakadouw (-AC), Pocock 546 (STE); Tafel-

berg 1830 m (-AC), Stokoe in SAM 55134 (NBG; PRE);

Sneeukop (-AC), Stokoe in SAM 55154 (SAM); Elandskloof

(-CA), Esterhuysen 3227 (BOL; K; NBG; PRE); Donkershoek

Kop, 1520 m (-CA), Stokoe 9246 (BOL); Duiwelskloof (-CA),

Stokoe in SAM 65522 (SAM; STE); Middeltuin, Cold Bokke-

veld (-CC), Hanekom 792 (K; PRE); Bokkeveld Tafelberg,

1890 m (-CD), Schlechter 10091 (BM; BOL; E; G; K; MO; P;

PRE; S; STE; W; Z); 1370 m (-CD), Bond 702 (NBG); De

Keur (-CD), Compton 6625 (NBG); Compton 16126 (NBG;

STE). Without precise locality: Cedarberg, Primos sub Marloth

11678 (PRE); Primos sub Marloth 11729 (PRE); Koude Bokke-

veld, 1520 m, Schlechter 8896 (BM; BOL; E; G; K; MO; NH;

P; PRE; S; STE: W). 3319 (Worcester): Winterhoek Ridge

Peak, 1650 m (-AA), Andreae 1113 (PRE; STE); 7 miles West

of Gydo Pass (-AA), Hutchinson 1030 (BOL; K); Peak below

Great Winterhoek 1830 m (-AA), Phillips 1817 (BOL; SAM);

Great Winterhoek, 1830 m (-AA), Phillips s.n. (BOL; K);

Visgat (-AA), Pillans 9684 (BOL; PRE); Visgat (-AA), Stokoe

in SAM 63952 (BM; SAM; STE); Sneeugat (-AA), Stokoe

7266 (BOL); Little Winterhoek 2018 m (-AA), Stokoe 6065

(K; NBG); Hartebeest Kloof, Bokberg 1188 m (-AB), Kirsten

442 (STE); Op-die-Berg Village, 980 m (-AB), Oliver 4032

(B; E; MO; PRE; STE; S); Hartebeest Kloof, Valboskloofberg,

1140 m (-AB), Oliver 5147 (STE); top of Gydo Pass (-AB),

Lewis in BOL 22009 (BOL; P); top of Gydo (-AB), Maguire

1778 (NBG; PRE; STE); Inkruip, Witsenberg (-AC), Ester-

huysen 23450 (BOL, K, PRE, STE); Tulbagh Waterfall, 420 m,

(-AC), Bolus 5304 (BOL; K; PRE; SAM); MacOwan 2685

(SAM); 270 m (-AC), MacOwan in Herb. Norm 564 (BM;

BOL; G; K; P; PRE; SAM; UPS; W); Witsenberg near

Steendahl, 400 m (-AC), Bolus 5376 (BOL): Roodesandberg

(-AC), Compton 6624 (NBG); Witsenberg near Tulbagh (-AC),

Leighton 1333, (BOL; PRE); Flats between Witsenberg and

Schurfteberg (-AC/AD), Zeyher 1117 (BOL; G; K; SAM;

STE; W); Slab Peak, 1220 m (-AD), Compton 11960 (NBG);

Slab Peak, Michells Pass (-AD), Esterhuysen 6182 (BOL);

Stokoe in SAM 54996 (SAM); Mostertshoek 1220-1520 m

(-AD), Esterhuysen 9882 (BOL); 300 m (-AD), Guthrie 2414

(BOL), Gorge west of Ceres (-AD), Hutchinson 610 (BM ;

BTL; K; PRE); Ceres, lower slopes, 610 m (-AD), Levyns

4702 (CT); Matroosberg (-BC), A. Bolus s.n. (BOL); Lamb 610

(BOL); 1520 m, Marloth 2215 (BOL; PRE); Marloth 2351b

(PRE); Matroosberg near Lakenvlei 1220 m (BC), Phillips

1983, (SAM); 1980 m (-BC), Phillips 2134 (SAM); Prospect

Peak, Hex River Valley, 760 m (-BC), Esterhuysen 15908

(BOL; PRE); Bokkerivier (-BD), Middlemost 2246 (NBG);

Middlemost 2278 (NBG); Du Toit’s Peak 1220 m (-CA/CC),

Esterhuysen 23775 (BOL); Molenaarsberg 914 m (-CA),

Esterhuysen 30354 (BOL; S; STE); Seven Sisters Wellington

(-CA), Stokoe 6067, (K; NBG; PRE; STE); Audensberg,

1670 m (-CB), Compton 9761 (NBG; STE), Brandwacht

Peak 1828 m (-CB), Esterhuysen 11004 (BOL; PRE) Stettyns-

berg, 1670 m (-CD), Esterhuysen 11057 (BOL); Keeromsberg,

1270 m (-DA), Esterhuysen 9293 (BOL); Wild (Wilge) River

(-DA), Niven s.n. (BM): Saw Edge Peak 3000 (DA), Oliver

3793 (K; MO; PRE; STE); Koo Mountain above Berger's

Pass, (-DB), Oliver 311 (STE); Wildepaardeberg (-DC),

Stokoe 2527 (BOL; K). 3320 (Montagu): Leeurivierberg (-CD),

Esterhuysen 27874 (BOL; PRE; STE); Without locality:

Wordsworth 14034 (K); Niven s.n. (G-DC).

G. parviflora is characterized by the flowers

being 1-4-nate in small heads clustered along the

branches in a congested spike-like manner, the corolla

tube not contracted in the middle and the simple

cilia on the calyx lobes. The species is closely allied

to G. minutiflora and has a remarkable superficial

resemblance to Eremia curvistyla (N.E. Br.) E. G. H.

Oliver differing basically in the number of stamens.

All three species are sympatric.

The true identity of this species has been over-

looked until now as it has nearly always been referred

to the later G. eremioides MacOwan. Klotzsch in

describing it placed the species in Don’s genus

Eremia using as his type an Ecklon & Zeyher collection

from “Hills between Puspas Valley and Kogmans-

kloof Mountains”. Without seeing the type,

N. E. Brown followed Klotzsch in keeping the

species in Eremia. But later in his work on Grise-

bachia he stated that he had seen the type and found

that the species was conspecific with G. eremioides

which he proceeded to retain under the Kew Rule.

This was picked up by Druce in his search through

Flora Capensis for new combinations but never

applied by subsequent botanists.

88

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Unfortunately the holotype in Berlin is no longer

extant and all the Ecklon & Zeyher material

distributed as Eremia parviflora Klotzsch turns out

to be Anomalanthus scoparius Klotzsch. N. E. Brown

on examining the type sent to him on loan stated

“The description of Klotzsch is very erroneous, as

the calyx is not subequal to the corolla, but

considerably shorter than it, and the stamens are 4,

not 8 as Klotzsch states. It is identical with Zeyher

1117, except that the leaves are straighter, like those

of Schlechter 10091”. This statement clears up the

discrepancies in the type description and also paves

the way for the typification of the species. Without

any authentic duplicate material available, I consider

that it is justifiable to rely on N. E. Brown’s com-

parison and therefore select Zeyher 1117 in the

herbarium at Kew as the neotype. The above published

note by N. E. Brown also appears on the sheet

at Kew.

Five duplicates of Zeyher 1117 are located in various

herbaria and all come from “Flats between the

Witsenberg and Skurfdeberg”. The locality of

Houw Hoek on one of the sheets in SAM as cited

by MacOwan in his protologue is erroneous.

To date no additional material of G. parviflora has

been collected in the same area as that visited by

Ecklon & Zeyher. But in the variation pattern,

I have found in the calyx that the broadly

elliptic-acuminate lobes of Zeyher 1117 could well

have occurred in the holotype. The leaf difference

noted by N. E. Brown is not vitally significant as

variation in this character has been noted even on

the same plant.

The basic variation occurring in G. parviflora is

found in the shape of the calyx lobes, the arrange-

ment of the leaves and the indumentum of the flowers,

the widest range being in the shape of the calyx

lobes. In subsp. parviflora three groups occur:

(a) large lobes with a broad elliptic base and

acuminate apex, sparsely gland-ciliate or ciliate,

common, central to southern in distribution

represented by the neotype, Zeyher 1117:

(b) large lobes, narrow elliptic-oblong with an

acute apex, distinctly gland-ciliate, less common

and confined to the central region of the

distributional range (Winterhoek to central

Cold Bokkeveld).

(c) small lobes mostly elliptic-oblong with an

acute apex, northern in distribution centred on

the southern and central Cedarberg.

N. E. Brown described (b) and (c) with their

narrow elliptic-oblong acute lobes in combination

with the possession of erect straight leaves as a

separate variety, var. grata. I have found that neither

the calyx lobe shape nor the leaf arrangement show

any significant differences. There is a definite inter-

grading between the broad based apiculate and the

narrow elliptic-oblong acute sepals as seen in Compton

6624 from Roodesandberg, Stokoe 6067 from the

Wellington mountains and Esterhuysen 11057 from

Stettynsberg. In leaf arrangement variation from erect

straight leaves through to curved slightly spreading

leaves can occur on the same plant. There were

therefore no grounds for keeping the two species

separate.

N. E. Brown also described G. similis var. publicalyx.

The collection Maguire 1778 from Gydo, presumably

from one population, possesses glabrous calyces.

This also occurs in the Compton and Esterhuysen

collections from Slab Peak. The indumentum of both

the calyx and corolla occurs randomly with most

southern collections having a glabrous calyx but

the lack of hairs and pubescence of the collections

from the northern Cedarberg and eastern Bokkeveld

are significant in the delimitation of subsp. eglandula

and subsp. pubescens.

(b) subsp. eglandula (N.E. Br.) E. G. H. Oliver,

stat. nov.

Grisebachia eremioides var. eglandula N.E. Br. in FI. Cap.

4,1:349 (1906). Syntypes: Cedarberg Range at Ezels Kop,

Schlechter 8818 (BM!; BOL!; G!; K!; P!; PRE!; STE!; Z!);

near Clanwilliam, Leipoldt 135 (BOL!). Lectotype: Leipoldt

135 (BOL).

Small shrublet, compact to spreading. Leaves

stiffly trigonous, markedly spreading-recurved when

mature. Calyx lobes subquadrate, subtruncate with

a thickened apiculus, rarely very broadly angular-

ovate, closely ciliate with short simple hairs

occasionally with a few subsessile glands, otherwise

glabrous, rarely sparsely puberulous. Corolla glabrous,

rarely sparsely puberulous. Ovary thickly pubescent

on the upper half. Fig. 22.7, 22.8.

Cape. — 3219 (Wuppertal); Pakhuis (-AA), Barker 4505

(NBG); Esterhuysen 5924 (BOL; PRE); Esterhuysen 21764

(BOL); Krakadouw, 910 m (-AA), Bodkin s.n. (BOL); Stokoe

in SAM 55129 (NBG; PRE; SAM); Stokoe in SAM 56776

(SAM); Rocklands, 790 m (-AA), Kruger 1031 (STE); Eselbank,

1220 m (-AC), Schlechter 8818 (BM; BOL; G; K; P; PRE;

STE; Z); Crevasse Peak, 1220 m (-AC), Taylor 7459 (PRE;

STE). Without precise locality: near Clanwilliam, Leipoldt

135 (BOL); Bokkeveld, 1580 m, Schlechter 8919 (7) (BM;

BOL; E; G; K; MO; P; PRE; STE; UPS; W; Z).

In the Cedarberg, the northern part of the

distribution range of this species, two reasonably

distinct groupings, A & B, of specimens can be made

on the shape of the calyx lobes, the pubescence, the

arrangement of the leaves and the distribution. Group

A has small narrow acute puberulous calyx lobes

with no marked apiculus and erect straight leaves and

is ascribed to subsp. parviflora. Group B has generally

small flattened quadrate to subquadrate glabrous

calyx lobes with a distinct apiculus and markedly

recurved leaves. The pubescence on the ovary is

much longer and denser.

There is very little overlap in these characters

between the two groups in the Cedarberg and both

seem to be fairly distinct. The affinities of group B

appear to lie with the collections much further south.

The leaf arrangements in A and B are very distinct

in the Cedarberg but not between B and some

random southern collections of subsp. parviflora in

which the leaves can be spreading and curved eg.

in Bolus 5403 from Tulbagh. The calyx shapes of B,

although distinct in the Cedarberg, have similarities

in some southern collections.

The distributions of the two groups A & B are

relatively easily separable with A occurring west

and south of the Krakadow-Welbedacht mountain

range and B north and east of the range. A detailed

investigation of the range is necessary to ascertain

whether this separation is in fact true or just due

to lack of records.

There is, then, in the Cedarberg a reasonable

discontinuity in several characters coupled with a

spatial separation. This should warrant recognition at

specific level, but as there are definite similarities with

certain elements to the south, I feel that recognition

is only justifiable at subspecific level.

On two collections N. E. Brown described this

vicariad as var. eglandula. An examination of all the

collections showed that small subsessile glands are

present on the margins of the calyx.

E. G. H. OLIVER

89

N. E. Brown annotated the collection Leipoldt 135

in BOL as the type. It is unfortunate that it is

unlocalized.

The collection Schlechter 8919 given as just Bokke-

veld is undoubtedly this subspecies and I regard the

locality as an error.

(c) subsp. pubescens E. G. H. Oliver, subsp.

nov., a subspecie typica et subspecie eglandula

floribus majoribus, tubo corollae omnino pubescenti,

calyce omnino pubescenti, distributione et flore-

scentia dignoscenda.

Type. — Cape, Ceres Dist: Katbakkies in the

Swartruggens (-DC) Oliver 4310 (STE, holo.; BM;

BOL; E; G; K; MO; NBG; PRE; S).

An erect to spreading shrub up to 50 cm high.

Calyx pubescent; lobes oblong-triangular to broadly

so, ciliate with fine simple hairs and stouter gland-

tipped hairs admixed. Corolla 2-3 mm long and

1,2-1, 7 mm broad; tube pubescent over the whole

length, tubular-ellipsoid. Fig. 22.9-22.12.

Cape. — 3219 (Wuppertal): Schurweberg, east of Bokkeveld

Tafelberg, 1060 m (-CD), Esterhuysen 20651 (BOL; K; NBG;

MO; PRE; S; STE); Zuurvlakte north of Rietvlei in the

Swartruggens, 1060 m (-CD), Oliver 6114 (PRE; STE); Kat-

bakkies in the Swartruggens, 1220 m (-DC), Levyns I860

(CT; SAM); 1188 m, Oliver 4310 (BM; BOL; E; G; K; MO;

NBG; PRE; S; STE); Oliver 4312 (B; C; G; GRA; HAM;

MEL; P; STE; W; Z); 1066 m, Taylor 5890 (PRE; STE). 3319

(Worcester): Winkelhaak, 960 m (-AB), Oliver 4318 (BOL; E;

K; MO; NBG; PRE; STE); Onverwacht in S. Swartruggens,

1220 m (-BA) Acocks 23660 (PRE; STE); Baviaansberg, 1060-

1220 m (-BA), Esterhuysen 29847 (BOL).

There are several collections from the eastern

part of the Cold Bokkeveld which have a different

appearance and earlier flowering time from the rest

of the collections of G. parviflora. The flowers are

generally much larger with the corolla tube completely

pubescent. The Levyns collection from Katbakkies

has the corolla tube pubescent for three-quarters of

its length. In nearly all the collections of G. parviflora

the corollas are glabrous to puberulous and then

usually below the level of the sepals. The collections

of Middlemost from Bokkerivier to the south are

intermediate tending towards the pubescence of the

Levyns collection.

The flowering time of the eastern Bokkeveld

collections is significant being from June to

September for material in full flower. The collections

of Levyns 1860 and Esterhuysen 20651 & 29847

and Oliver 6114 though recorded for Sept., Oct. and

Nov. are of fruiting material. The flowering time

for collections of G. parviflora elsewhere and

particularly in the adjoining areas of the Bokkeveld

are Sept.-Dee. for material in full flower. This

means that cross-pollination, even if the populations

were sympatric, could not take place. The Middlemost

collections from Bokkerivier were collected in

November in full flower.

Despite there being little morphological disjunction

between the group of eastern Bokkeveld collections

and the rest of G. parviflora, I consider that the

geographical and reproductive isolation warrants

recognition of this group as a distinct subspecies of

parviflora.

7. Grisebachia minutiflora N.E. Br. in FI. Cap.

4,1:348 (1906). Type: Cape, near Klein Vlei in

Cold Bokkeveld, Schlechter 10064 (BM!; BOL!; G!;

K, holo!; MO!; P!; PRE!; S!; STE!; W!).

Low compact to semi-spreading shrublet up to

20 cm high. Branches long when spreading and

sometimes rooting at the nodes, pubescent, with

glandular hairs intermingled when young, 3-angled

when young. Leaves 3-nate, up to 3,5 mm long, the

petiole 0,5 mm long, ovate to narrowly oblong,

straight erect or slightly spreading, imbricate or shorter

than the internodes, subobtuse, thick, pubescent

or glabrous when young with a few to numerous

sessile glands on the margins, becoming glabrous

except on the adaxial surface, sometimes terminating

in a sessile gland. Flowers in terminal globose heads

of up to 36 flowers on short branchlets, not forming

congested spikes, white; pedicels almost none up to

0,8 mm long; bracteoles 3, adpressed or recurved-

spreading, equal or very unequal in some outer

flowers in inflorescences, mostly 0,5-1, 7 mm long,

if equal then linear acute or subacute and minutely

keel-tipped, if unequal then the median one large

and leaflike and well keeled, glabrous or pubescent,

ciliate towards the base with simple hairs and a

mixture of simple and larger plumose and gland-

tipped hairs towards the apex. Calyx 4-lobed from

j—f its length, up to 1,9 mm long, pubescent, the

pubescence short over the whole surface or in zones

with sometimes longish hairs; tube obconic or

tubular with spreading lobes; lobes ovate-oblong

to very broadly ovate up to 0,9 mm long and 1 mm

broad, erect or spreading, sometimes with pubescence

on the inside at the top, ciliate with short or long

cilia which are simple or variously plumose from

base to apex, mostly gland-tipped, keel-tipped and

thickened, acute to subobtuse. Corolla 4-lobed,

obconic to tubular with spreading upper half, not

constricted, glabrous inside and outside, up to 2,4 mm

long; lobes short broad obtuse slightly crenulate,

erect or slightly spreading. Stamens 4, manifest or

slightly exserted; filaments filiform, glabrous, sigmoid

at the apex; anthers up to 0,7 mm long with almost

parallel sides, oblong, minutely scabrous, awned

about ^ the way up the back of the cell; awns up

to i the length of the cell sometimes spreading;

pore about ^ the length of the cell. Ovary 2-cel led

with a single pendulous ovule per cell, ellipsoid,

puberulous at the apex; style filiform, glabrous, far

exserted up to 2,2 mm long; stigma simple or

slightly swollen. Fig. 24.

A species forming compact erect to semispreading

low shrublets, occurring in sandy places in the Cold

Bokkeveld north of Ceres, flowering from October to

January.

A variable taxon in which two subspecies are

recognized.

Key to the subspecies

Cilia on the calyx gland-tipped (a) subsp. minutiflora

Cilia on the calyx not gland-tipped (b) subsp. nodiflora

(a) subsp. minutiflora

G. minutiflora N.E. Br. in FI. Cap. 4,1: 348 (1906). Type:

Schlechter 10064 (K, holo!; numerous isos!).

Leaves pubescent when young with several sessile

glands on the margins, becoming glabrous exceps

cn the adaxial surface, or glabrous with numerout

sessile glands over the surface but pubescent on the

adaxial surface. Pedicels almost absent or up to

0,4 mm long; bracteoles usually adpressed, mostly

up to 1,1 mm long, occasionally up to 1,3 mm.

Calyx pubescent in zones, glabrous on inner surface,

cilia plumose or simple, gland-tipped. Figs 24.1-

24.8 & 25.

90

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBA CHI A

Fig. 24. — Grisebachia minutiflora : subsp. minutiflora. 1, flower; 2, corolla; 3, median bracteole; 4, sepal; 5, anther, side,

front and back views; 6, ovary; all drawn from an isotype, Schlechter 10064 (STE); 7, variation in median bracteole,

drawn from Oliver 4105 (STE); 8, sepal, drawn from Oliver 6106 (STE): subsp. nodiflora, form A. 9, flower; 10,

corolla ; 1 1 , median bracteole ; 1 2, sepal ; 1 3, anther, side, front and back views ; 14, ovary ; all drawn from an isolecto-

type, Schlechter 10188 (STE); form B. 15, median bracteole; 16, sepal; both drawn from Schlechter 10188 (STE).

All drawings x 16.

Fig. 25. — Distribution of Grisebachia minutiflora. 9 subsp.

minutiflora', Q subsp. nodiflora.

Cape. — 3219 (Wuppertal), between de Keur and Kromfon-

tein, 914 m (-CD), Oliver 4105 (BM; C; E; G; K; MO; MEL;

P; PRE; S; STE; W; Z); Rietvlei Dam in Swartruggens, 970 m

(-CD), Oliver 6106 (PRE; STE); Stompiesfontein in the Swart-

ruggens (-DC), Esterhuysen 29304 (BOL; STE). 3319 (Wor-

cester), near Sandberg (-AB), Bond 659 (NBG); Esterhuysen

3457 (BOL; PRE); Gydo, 1370 m (-AB), Compton 18722

(NBG); Gydoberg (-AB), Leighton 2254 (BOL; PRE); between

Loch Lynne and Winkelhaak, 960 m (-AB), Oliver 5117 (B;

BOL; E; G; K; MO; PRE; STE); S.W. of Winkelhaak,

960 m (-AB), Oliver 5124 (C; GRA; K; P; PRE; STE); Tuins-

kloof, south of Houdenbek, 940 m (-AB), Oliver 5130 (BOL;

MEL; NBG; PRE; STE); near Klein Vlei; 1220 m (-AB),

Schlechter 10064 (BM; BOL; G; K; MO; P; PRE; S; STE; W);

Baviaansberg, 1676 m (-BA), Bond 1436 (NBG); 1463 m,

Stokoe 4543 (BOL); Stokoe 4550 (BOL; K).

(b) subsp. nodiflora ( N.E . Br.) E. G. H. Oliver,

stat. nov.

G. nodiflora N.E. Br. in FI. Cap. 4,1:438 (1906). Type:

Cape, Schoongezicht in Cold Bokkeveld, Schlechter 10188

(BM!; BOL, lecto ! ; E! ; G!; K!; MO!; P! ; PRE!; S! ; STE!;

W!; Z!).

Leaves pubescent when young, occasionally with a

few sessile glands on the margins. Pedicels up to

0,8 mm long; bracteoles usually adpressed and

recurved towards the apex, mostly up to 1,7 mm

long. Calyx pubescent in zones or evenly with

pubescence also on the inner surface in the upper

quarter, cilia plumose to the apex and not gland-

tipped. Fig. 24.9-24.16.

Cape. — 3219 (Wuppertal), Schoongezicht, 1370 m (-CC),

Schlechter 10188 (BOL & numerous duplicates).

G. minutiflora is characterized by the globose

inflorescences of 6-36 flowers scattered along the

branches, the corolla-tube not contracted in the

middle, the plumose cilia on the calyx lobes and the

general glandular condition of the flowering branches.

It is closely allied to G. parviflora.

When Brown revised the genus he had two

collections of Schlechter to examine and justifiably

described them as two new species, G. minutiflora

and G. nodiflora, basing them on the prescence or

absence of gland-tipped plumose or simple cilia on

the calyx lobes.

Since then no further material referrable to G.

nodiflora has been collected, but there have been

12 collections of G. minutiflora. The type of G.

nodiflora possesses no glands on the calyx lobes,

whereas in the collections of G. minutiflora there are

always some glands p esent on the cilia, but these

may be extremely small in some flowers. The material

E. G. H. OLIVER

91

was then examined for other characters. Brown used

the differences in the size of the inflorescence heads,

but this is invalid as the heads in Oliver 4105 are

much larger than in Schlechter 10188. He also used

the degree of feathering on the cilia, but this is very

variable in G. minutiflora from simple to almost

fully plumose.

Slight differences were found in the length of the

pedicel which is absent to 0,4 mm long in G. minuti-

flora and up to 0,8 mm long in G. nodiflora. The

bracteoles of the former are usually adpressed whereas

in the latter they are approximate but curved-

spreading. I decided to reduce G. nodflora to sub-

specific level under G. minutiflora, because of the

difference in the glandular state of the cilia and the

slight discontinuities in the bracteoles and pedicel

characters coupled with the allopatric distribution.

At first examination I found that there was a

distinct difference in the type of pubescence on the

calyces with G. minutiflora having zones that were

pubescent and glabrous and with G. nodflora having

an evenly pubescent calyx. Close examination of

Schlechter 10188 duplicates showed there to be two

distinct forms. In one the pubescence is zoned as in

G. minutiflora, in the other it is evenly distributed

over the calyx, a condition not found in G. minuti-

flora. This variability suggests a closer relationship

between the two taxa than was previously accorded

to them.

The material on sheets of Schlechter 10188 can

easily be separated into two forms, A & B, on the

type of pubescence and the shape of the calyx lobes.

In form A the pubescence is zoned and the calyx

lobes are oblong to elliptic-oblong acute. In form B

the pubescence is denser and evenly distributed over

the calyx the lobes of which are transversely broadly

elliptic obtuse. These variations can only be recognised

as forms at present as they presumably came from

one population. This has not been rediscovered and

until such time as it is, no further status can be

given to this variation.

The specimens examined in detail have been

assigned to the two forms as follows:

Herbarium Form A Form B

BOL 1 twig —

BOL 2 twigs —

E — 4 twigs

MO 1 twig 2 twigs

PRE 2 twigs 1 twig

STE 1 twig 1 twig

K 1 twig 1 twig

N. E. Brown did not designate a holotype but

labelled one sheet in BOL (form A only) and one in

K (form A & B) as types. From the protologue

two characters can be pinpointed to determine which

form he used to describe his species, i.e. “ca yx

lobes oblong, acute”. These undoubtedly refer to

form A. I have therefore chosen the Schlechter sheet

labelled as the type in BOL as the lectotype.

8. Grisebachia secundiflora E. G. H. Oliver sp.

nov., in genere singulari ovario unicellulari et

inflorescentiis secundis densis sed affinis G. parvi-

florae (Klotzsch) Druce.

Frutex humilis compactus vel eflfusus ad 0,8 m

altus. Rami multi ascendentis foliis tantum ad

extrema, pubescentes demum canopubescentes et

lamelliformes. Folia 3-nata imbricata demum patentia

3-4 mm longa petiolo 1 mm, anguste ovata ad

elliptica, pubescentia glabrescentia, ciliata ciliis

brevibus crassis plumosis et apiculata demum

serrulata, interdum ciliis glandulis parum glauca ad

olivacea; petiolo quam lamina pubescentiora. Flores

3- nati in remulis axillaribus brevissimis, aggregate in

pseudospicam secundam dense farotam ad 3,5 cm

longam versus extrema ramorum; pedicellis brevis-

simis 0,5 mm longis; bracteolis 3 adpressis, mediana

ad 1,0x0, 6 mm saepe minore quam lateralibus

ovatotriangulare, lateralibus ad 1 ,3x0,6 mm anguste

ovatis, interdum obliquis, omnibus glabris ad

pubescentibus, ciliatis ciliis simplicibus vel plumosis

interdum glandulis, apiculatis cilio longo plumoso,

albidis ad chlorinis. Calyx 4-lobatus campanulatus

ad 2, 1 mm longus junctus quadrante ad tertia parte,

ad dimidio tubi corollae glaber ad puberulus albidus

ad chlorinus; lobis in latitudine inaequalis, ab- et

adaxialibus latioribus quam lateralibus, late ovatis ad

obovatis parum cucullatis ciliatis ciliis plumosis

crassis interdum glandulis cum ciliis simplicibus

immixtis, apiculatis, indistincte sulcatis apice. Corolla

4- lobata 3, 5-4, 5 mm longa et 1,3-1, 6 mm lata,

angusto tubulosa et parum inflata in parte tertio e

basi, ad breviore multo inflata in dimidio interiore,

pubescens in parte medio et intra glabra, albidus;

lobis 0,6x0, 6-1 ,0x0,9 mm erectis ad parum

patentibus, obtusis, glabris. Stamina 4 libera; fila-

mentis anguste linearibus glabris albidis; antheris ad

1,0x0, 5 mm, inclusis ad manifestis, subbasaliter

dorsifixis, laevigatis ad minute scabridis, aristatis;

aristis dorsalibus, deorsum currentibus, cellulis plus

minusve dimidio brevioris; pollinis singularibus.

Ovarium 1-cellulare ovulo uno pendulo apicale,

rarissime 2-ceIlulare, parum obliquum, 0,7-0, 6 mm

glabrum viride; disco distincto rubro; stylo filiformi

4,0-4, 5 mm longo, exserto glabro; stigmate

capitellato.

Type. — Cape, Ceres District, Swartruggens in the

Cold Bokkeveld, Oliver 6105 (STE, holo.; BM ;

BOL; E; G; K; MO; NBG; P; PRE: S; W).

Low compact but sprawling shrublets up to 0,5 mm

high, much branched. Branches numerous ascending

with leaves only towards the ends, finely pubescent

becoming grey pubescent and flaky with age. Leaves

3-nate, closely imbricate, spreading when older

mostly 3-4 mm long with the petiole 1 mm long,

narrowly ovate to elliptic, pubescent becoming

glabrous, ciliate with short stout plumose hairs and

apiculate with a long stout plumose hair, becoming

serrulate, occasionally with gland-tipped cilia when

young, slightly glaucous to olive-green; the petiole

more pubescent than the lamina. Flowers 3-nate on

very short axillary branchlets crowded into a densely

packed secund pseudo-spike up to 3,5 cm long

towards the ends of the main branches; pedicel very

short about 0,5 mm long; bracteoles 3 adpressed

to the calyx, the median up to 1,0x0, 6 mm often

smaller than the laterals, ovate-triangular, the

laterals up to 1 ,3 x 0,6 mm narrowly ovate sometimes

oblique, all glabrous to pubescent, ciliate with simple

and plumose cilia sometimes gland-tipped, apiculate

with a long plumose hair, white to pale yellow-green.

Calyx 4-lobed campanulate up to 2, 1 mm long joined

from quarter to one third of its length, reaching

halfway up the corolla-tube, glabrous to puberulous

white to pale yellow-green; lobes unequal in width,

1-1,7 mm, the ab- and adaxial being broader than

the laterals, broadly ovate to obovate slightly

cucullate, ciliate with stout plumose cilia which may

be gland-tipped with simple cilia inbetween, apiculate,

indistinctly sulcate at the apex. Corolla 4-lobed

3, 5-4, 5 mm long and 1,3-1, 6 mm wide, narrow

tubular and slightly inflated one third of the way up

92

STUDIES IN THE ERICOIDEAE. III. THE GENUS GRISEBACHIA

Fig. 26. — Grisebachia secundiflora. 1, flower, x 8; 2, corolla, x 8; 3, three bracteoles, x 16; 4, sepals, lateral and abaxial,

xl6; 5, anther, back, front and side views, xl6; 6, gynoecium, xl6; 7, leaf, x8; all drawn from the holotype,

Oliver 6105 (STE); 8, flower, x 8, all drawn from Oliver 5044 (STE).

Fig. 27. — Distribution of Grisebachia secundiflora.

to short tubular and much inflated in the lower half,

finely pubescent in the middle region and glabrous

inside, white: lobes varying from 0,6x0, 6-1, Ox

0,9 mm, erect to slightly spreading, obtuse, glabrous.

Stamens 4, free; filaments narrow linear, glabrous

white; anthers up to 1,0x0, 5 mm, included to

manifest, subbasally attached on the dorsal surface

smooth to minutely scabrid, aristate; awns dorsal

pointing downwards, about half the length of the

cell, minutely ciliate, white; pore about half the

length of the cell; pollen grains single. Ovary 1 -celled

with a single pendulous apical ovule, very rarely

2-celled, slightly oblique, 0,7x0, 6 mm, glabrous,

greenish, seated on a distinct dark red nectariferous

disc; style filiform, 4,0-4, 5 mm long, exserted,

glabrous; stigma capitellate. Figs 26 & 27.

Cape. — 3219 (Wuppertal); Rietvlei (-CD), MacGregor in BOL

31286 (BOL); Swartruggens, Rosendal area, north-east of the

farmstead, 960 m (-CD), Oliver 6105 (BM; BOL; E; G; K;

MO; NBG; P; PRE; S; STE; W); 6107 (B; C; PRE; STE);

1066 m, Oliver 6115 (NBG; PRE; STE). Locality uncertain;

Ceres Wildflower Show, Oliver 5044 (PRE; STE). Flowering in

October.

G. secundiflora is very distinct in the genus with

its erect secund pseudospicate inflorescences, its

considerably bare branches and 1 -celled ovary.

The species was first seen by me at the Ceres

Wildflower Show in October 1974 when I was doing

the naming of specimens. The collector and locality

unfortunately could not be traced. A few weeks

later the material collected the previous year in the

Swartruggens by Dr J. MacGregor was sent to me

for identification.

An examination of the above collections showed

that they constituted a new and very distinct species

which I was not able to place satisfactorily in any

known genus. Following Dr MacGregor’s directions,

I visited the Swartruggens and located the species

in three disjunct sparse populations. A range of

material was collected and examined for variations

in the critical character, the 1 -celled uniovulate

ovary. An examination of numerous flowers produced

only a few with unequally 2-celled ovaries.

The species is remarkably similar to Eremia totta

(Thunb.) G. Don in the outward appearance of the

flowers and leaves but cannot be placed near that

species which has 8 stamens and a 4-celled ovary.

Eremia has been considerably amended to include

the 1-celled Eremia curvistyla (N.E. Br.) E. G. H.

Oliver but still retains the constant character of

eight stamens (Oliver, 1976).

To a lesser extent the species is similar in outward

appearances to Grisebachia parviflora (Fig. 22) and

G. minutiflora (Fig. 24) both of which it grew with

in the Swartruggens. They all possess four stamens.

Until this revision, all species of Grisebachia possessed

E. G. H. OLIVER

93

2- celled ovaries with very few exceptions having

3- celled ovaries.

There were thus four ways of dealing with the new

species: (1) placing it under Eremia and having to

amend the generic circumscription even further to

include the 4-stamened condition, thus causing a

breakdown in the distinction between Eremia and

Grisebachia', (2) placing it under Grisebachia and

amending the generic circumscription to include this

1-celled species; (3) placing it in either Anoma-

lanthus or Syndesmanthus, genera with 4 stamens

and a 1-celled ovary, but which have no resemblance

to it; (4) describing the species as a separate mono-

typic genus.

Taking into account the implications of the above,

it was decided to broaden the circumscription of

Grisebachia to include the 3-celled variations parti-

cularly in G. parviflora subsp. pubescens and the

1-celled, rarely 2-celled, condition occurring in the

new species.

In the Swartruggens G. secundiflora was found

in three separate populations consisting of only a

few scattered plants each. At the lower altitude the

plants were growing on sandy flats with a population

of G. minutiflora nearby. Higher up the mountain

they were growing on sandy, rocky slopes together

with some plants of G. parviflora subsp. pubescens.

In all cases G. secundiflora formed decumbent yet

compact shrubs up to 6,5 m high and up to 1 m

across with numerous ascending to decumbent

branches. The branches were strikingly bare and

devoid of leaves except towards their ends. The white

conspicuous secund inflorescences were subterminal.

This contrasted strongly with the very compact low

shrublet of G. minutiflora or the sparse spreading

procumbent to semi-erect plants of G. parviflora ssp.

pubescens which were all in fruit.

Some variation in floral characters occurs. The

pubescence on the calyx may be present or absent

on different twigs in the collections Oliver 6105,

6107 and 6115 and is present in MacGregor s.n.

It is absent in Oliver 5044 from the flower show.

The corolla tube in the collections of Oliver and

MacGregor from the Swartruggens are short, tubular

and inflated in the lower half whereas in the material

from the flower show it is distinctly longer and

narrower. As this latter material is unlocalized no

further comments on its status can be given.

SPECIES NON SATIS COGNITAE

1. Erica nodiflora Salisb. in Trans. Linn. Soc.

6: 340 (1802). Type: Hibbert s.n. (?).

Blaeria nodiflora G. Don., Syst. Veg. 3: 805 (1834);

Klotzsch in Linnaea 12: 246 (1838) sub. sp. non

sat. cog.

Bentham places this species under G. plumosa

Klotzsch and N. E. Brown accepts this. No type

specimen has been seen and the description could

fit most species of Grisebachia.

2. Erica capitata Salisb., Prodr. 293 (1796), non

L. Type: ?. Bentham and N. E. Brown place this

species under G. plumosa, presumably because in the

Kew copy of the Prodromus Salisbury has written

in pencil ‘Nodiflora MS’. Salisbury gave no description

quoting only Linnaeus incorrectly.

3. Blaeria ciliaris sensu Klotzsch in Linnaea 8: 658

(1933), non G. ciliaris (L.f.) Klotszch in Linnaea

12:225 (1838); N.E. Br. in FI. Cap. 4,1: (1906).

Type: Willdenow Herb. no. 2890 (B!).

Klotzsch wrongly ascribed the Willdenow specimen

to the species which had, up until then, been cited

as Blaeria ciliaris L.f. In his description he stated

that the leaves were 4-nate, possibly repeating the

slip made by Thunberg and copied by most sub-

sequent authors. The Willdenow specimen has in

fact 3-nate leaves.

Rach (1855) in examining Klotzsch’s specimen

and that of Thunberg stated that they were not of

the same species. Brown noted that Thunberg’s

specimen was identical to the type in the Linnaean

Herbarium.

I have been able to examine the Willdenow

specimen in the Berlin Herbarium. The material

certainly does not belong to G. ciliaris (L.f.) Klotzsch

subsp. ciliaris and is only in young bud stage from

which it is not possible to identify it with any certainty.

SPECIES EXCLUDED

Grisebachia eriocephala (Klotzsch) Benth. in DC.,

Prodr. 7:702 (1839) based on Finckea eriocephala

Klotzsch in Linnaea 12:238 (1838) = Acrostemon

eriocephalus ( Klotzsch ) N.E. Br. in FI. Cap. 4,1: 335

(1906).

Grisebachia bruniades (Klotzsch) Benth. in D.C.,

Prodr. 7: 702 (1839) based on Finckea bruniades

Klotzsch in Linnaea 12: 238 (1838) = Acrostemon

eriocephalus (Klotzsch) N.E. Br.

UITTREKSEL

Hierdie is 'n hersiening van die genus Grisebachia

Klotzsch waarin 8 spesies erkenning geniet. Die genus

behoort tot die Ericaceae-Ericoideae en is endemies

in die suidwestelike deel van die Kaapprovinsie. Die

ondersoek het getoon dat 'n groot mate van variasie

onder die spesies voorkom wat dit noodsaak om 7

spesies tot infraspesifieke rang en 1 spesies tot sino-

nieme te verlaag, terwyl 1 spesie as 'n takson met 'n

onduidelike identiteit beskou word. Die nuwe spesies,

G. secundiflora E. G. H. Oliver, is beskryf.

REFERENCES

Bentham, G., 1839. Ericaceae. In DC., Prodr. 7: 580-733.

Paris.

Bentham, G., 1876. Ericaceae. In Benth. & Hook, f., Gen. PI.

2: 577-604. London: Reeve.

Brown, N. E., 1906. Grisebachia. In W.T. Thistleton-Dyer,

FI. Cap. 4,1: 337-350. Ashford: Reeve.

Brown, N. E., 1909. Addenda and Corrigenda, Ericaceae. In

W.T. Thistleton-Dyer, FI. Cap. 4,1: 1123-1129. Ashford:

Reeve.

Drude, O., 1897. Ericaceae-Ericoideae. In Engl. & Prantl,

Pflanzenfam. 4,1: 57-65. Leipzig: Engelmann.

Klotzsch, J. F., 1838. Ericearum, genera et species. Linnaea

12: 211-247.

Oliver, E. G. H., 1975. Ericaceae. In Dyer, Gen. 1: 429-439.

Pretoria: Botanical Research Institute.

Oliver, E. G. H., 1976. Studies in the Ericoideae. I. The genera

Eremia and Eremiella. Bothalia 12: 29-48.

Phillips, E. P., 1926. Ericaceae. In Gen. ed. 1, 457^465. Pre-

toria: Division of Botany and Plant Pathology.

Phillips, E. P., 1944. Notes on the minor genera of the

Ericaceae. Jl S. Afr. Bot. 10: 69-73.

Phillips, E. P., 1951. Ericaceae. In Gen. ed. 2, Pretoria:

Division of Botany.

Weimarck, H., 1941. Phytogeographical groups, centres and

intervals within the Cape flora. Lunds Univ. Arsskr. 37:

1-143.

Bothalia 13, 1 & 2: 95-110 (1980)

A survey of some of the pre-Linnean history of the genus Acacia

I. H. ROSS*

ABSTRACT

The pre-Linnean history of the plants referred to the genus Acacia to some extent reflects the development

of botanical description, classification and illustration. Attention is drawn to some of the earliest references to

plants known to belong to the genus Acacia and to references in selected herbals and publications up until

Philip Miller’s description of the genus in the fourth abridged edition of his Gardeners Dictionary in 1754.

r£sum£

REVUE DE QUELQUES £l£MENTS DE L ’ HISTOIRE PR£-LINN£ENNE DU GENRE ACACIA

L’histoire pre-linneenne des plantes rattachees an genre Acacia reflete jusqu’a im certain point le developpe-

ment de la description, de la classification et de l’ illustration en botanique. On attire 1’ at tent ion sur certaines des

references les plus anciennes a des plantes connues comme appartenant au genre Acacia ainsi qu 'a des references

puisees dans un choix d’herbiers et de publications anterieures a la description faite de ce genre par Philip Miller

dans la quatrieme edition abregee de son “Gardeners Dictionary” en 1754.

INTRODUCTION

Although of no standing in present-day nomen-

clature, it is nevertheless of considerable interest to

trace the pre-Linnean history of the plants now

referred to the genus Acacia as to some extent it

mirrors the development of botanical description,

classification and illustration.

From the beginning, plants, particularly those of

utilitarian value, attracted the attention of man and

the use of plants for medicinal purposes long pre-

ceeded any description of the plants themselves. Since

very early times a variety of herbs was used as healing

agents and it had become necessary to study them in

detail in order to be able to differentiate the kinds

employed for different purposes. In the words of

Stearn (1958), “Botany as a science was fashioned

out of herb-lore at Athens when Theophrastus (370-

285 B.C.) applied to the vegetable kingdom the

principles of classification based on logic associated

with his teachers Aristotle and Plato.”

Attempts were made to classify plants in the earliest

works on natural history. Theophrastus in his Enquiry

into Plants considered the principles of classification

suggesting that the vegetable kingdom by classed into

trees, shrubs, under-shrubs and herbs and that minor

divisions should be based on differences such as those

between flowering and flowerless and deciduous and

evergreen plants. In addition, he hinted at an ecological

classification.

A number of manuscript herbals was written in

western Europe during the centuries that elapsed

between the end of the classical period and the end

of the fifteenth century. Theophrastus, Dioscorides

and Pliny either gave no descriptions to the names

of the plants or they described them so inadequately

that it was probably difficult even then, as it still is

now, to identify many of the plants referred to in

their works. The writers of the early herbals sought

to recognize in the plants of their own country those

of classical antiquity named by Theophrastus, Dio-

scorides and Pliny as it was at first assumed that

the plants described by the Greek physicians grew

wild throughout Europe. As a consequence, each

author identified a different native plant with one

* National Herbarium of Victoria, Birdwood Avenue, South

Yarra, 3141, Victoria, Australia.

mentioned by Theophrastus or Dioscorides or others

thereby creating much confusion so that the reader

of one work can in many instances never be sure

whether the plant referred to by a certain name is

the same as a plant with the same name in the work

of another author. A description of a plant during

the early sixteenth century is therefore usually

accompained by a critical enquiry as to whether the

usage of the name agrees with the use to which it

was put by other authors. Many of the early works

showed little originality being copies of copies of

yet earlier copies. During this copying process errors

were introduced and descriptions of quite common

plants were borrowed from earlier works and em-

bellished with superstitions so that many departures

were made from the original texts.

As many of the herbalists were medical men prob-

ably one of the objects which early herbalists had

in mind when writing their books was to enable the

reader to identify the herbs used in medicine. How-

ever, until the sixteenth century was well advanced

the illustrations generally provided in herbals were

often so stylized and the descriptions so inadequate

that it must have been extremely difficult to identify

many of the plants solely by reference to these works.

Arber (1938) suggested that the knowledge of plants

was transmitted by word of mouth and that the

herbals were only used as reference works in which to

seek information about plants whose identity was

already known to the reader.

A significant advance occurred when the authors of

herbals and other works based their descriptions on

the actual plants that they had before them instead

of copying earlier descriptions. The descriptions were

not very methodical initially but they slowly became

more systematic. The herbals of the late sixteenth

century mostly contain descriptions of plants known

to the author from the immediate environment of

his native land. Later authors endeavoured to present

a more comprehensive account in each herbal by

recording all plants noted by predecessors whether

or not they had seen them and adding the previously

unknown plants that they had seen themselves. In

constrast with previous centuries, the merit of each

new herbal came to depend upon the number of

plants added from the authors’ own observations and

not on what the author had copied from predecessors.

As each author wished to include in his work as many

96

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

new plants as possible, the number of plants described

grew fairly rapidly. Fuchs (1542) described about five

hundred species but by 1623 the number of species

enumerated by Caspar Bauhin in his Pinax theatri

botanici had risen to six thousand.

As a product of the process of compiling descrip-

tions the similarities and differences between plants

became more apparent to authors along with the

realization that some of the affinities had little to do

with the medicinal properties or agricultural im-

portance of the plants. A significant advance occurred

when information relating to medical superstition

was omitted from the descriptions and the perception

of natural affinities among plants awakened a desire

to distinguish more precisely whatever was different

and to bring together more carefully whatever was

similar (Sachs, 1890). This perception of resemblances

and differences of form developed and led in turn to

the idea of natural relationships and systems of

classification. The recognition of natural groups is

found in the later herbals from the late sixteenth

century onwards, and the series of works published

between 1530 and 1623, from Brunfels to Caspar

Bauhin, reflects how the perception of a grouping of

affinity grew more and more distinct.

Caspar Bauhin (1623) considered the arrangement

of plants in his Pinax theatri botanici to be of the

greatest importance and his system was far ahead of

those of his predecessors. He employed the system

which de l’Obel had used in 1576 in his Plantarum seu

stripium historia but carried it out more thoroughly.

Caspar Bauhin consistently used the binary system of

nomenclature, which Linnaeus is often thought to

have founded, each plant bearing a generic and a

specific name, although sometimes a third or even a

fourth descriptive word was added. However, these

additional words are apparently only auxiliary and

not essential. In his Pinax, Caspar Bauhin also sought

to put an end to the nomenclatural confusion which

had arisen by listing for each species known to him

all of the names that had been applied previously

by earlier writers.

The art of botanical illustration and the develop-

ment of plant descriptions proceeded at different

rates and to an extent independently of one another.

The first millenium of the history of plant illustration

shows no steady advance from primitive work to

naturalistic, but rather a gradual decline which was

not fully arrested until the early sixteenth century

with the appearance of Weiditz’s illustrations in

Brunfels’s Herbarium vivae eicones (Blunt, 1955).

Remarkable examples of some very early large-scale

brush drawings are found in the Codex Aniciae

Julianae of Dioscorides’s work. This work was made

at Constantinople about the year 512 A.D. but it

appears that some of the illustrations were derived

from those made by Crateuas who was personal

physician to King Mithridates (120-63 B.C.) (Arber,

1938; Stearn, 1954; Blunt, 1955).

As species of Acacia occur in the Nile Valley in

Egypt and in the middle east it is not surprising that

references to plants now known to belong to this

genus can be traced back almost to the earliest

recorded history. Reference to the genus is found in

texts of the ancient Egyptians, in the Bible, and in the

writings of classical antiquity. Of course, as is to be

expected, many of the plants referred to by the names

Acanthus, Acanthos, Akakia, Acatia and Acacia were

unrelated and many are excluded from the present

generic concept of Acacia.

Attention is drawn in this paper to some of the

earliest references to plants known to belong to the

genus Acacia and to references in selected herbals and

publications leading up to Philip Miller’s description

of the genus in 1754.

THE ANCIENT EGYPTIANS

Species of Acacia have flourished on the banks of

the river Nile in Egypt for thousands of years and,

according to Rochebrune (1899), the ancient Egyp-

tians were familiar with A. nilotica (L.) Willd. ex Del.

and A. seyal Del. and very probably A. tortilis

(Forssk.) Hayne and perhaps even A. Senegal (L.)

Willd. The name of A. nilotica is found in contemporary

texts of the pyramids and, of all of the Acacia species,

its name occurs most commonly in inscriptions in

religious and historical texts and in literary and

medical papyri. The most frequently used symbol to

depict the species is a pod which represented a figure

in hieroglyphics. Rochebrune noted that A. nilotica

is represented on the tomb of Menephtha of the

eighteenth dynasty at Beni-Hassan.

Fliickiger & Hanbury (1874) record that the Egyp-

tian fleets brought gum arabic from Arabia as early

as the seventeenth century B.C. and that there were

representations of the trees, together with heaps of

gum, in the treasury of King Rhampsinit (Ramses

III) at Medinet Abu. The symbol used to signify gum

arabic, which was largely used in painting, is fre-

quently encountered in Egyptian inscriptions.

The ancient Egyptians used the flowers of Acacia

for crowns and garlands, some of which adorned the

mummies of certain kings. A. nilotica was sometimes

placed among the offerings on the altars of the Gods

but there is no evidence of its having been sacred,

while Acacia wood is reputed to have been used to

clamp shut mummy-coffins made of sycamore.

THE BIBLE

There is almost universal agreement that the plant

referred to in the Bible by the Hebrew word “shittah”

(singular) or “shittim” (plural) is a species of the

genus Acacia, three or four of which occur in biblical

lands (Moldenke & Moldenke, 1952). The Bible

(The Authorized Version of King James) contains

numerous references to shittim-wood particularly in

connection with the ark of the Tabernacle which was

ordered to be made of this wood. For example, in

Exodus 25: 5, 10, 13, 23 and 28 —

“And rams’ skins dyed red, and badgers’ skins, and

shittim wood. . . And they shall make an ark of shittim

wood: two cubits and a half shall be the length thereof,

and a cubit and a half the breadth thereof, and a cubit

and a half the height thereof. . . And thou shalt

make staves of shittim wood, and overlay them with

gold . . . Thou shalt also make a table of shittim

wood . . . And thou shalt make the staves of shittim

wood."

Smith & Fuller (1893) record that the predominant

use of the plural form of the word in the Scriptures,

that is “shittim” rather than “shittah”, is probably

because the trees are usually gregarious and seldom

occur singly. In the Revised Version of the Bible the

terms “acacia tree” and “acacia wood” are used.

According to Moldenke & Moldenke (1952), most

authorities are of the opinion that A. seyal or A. tortilis

are the most likely species involved in these references.

Both species are seemingly able to flourish in dry

areas and A. tortilis is the largest and commonest

tree on the deserts of Arabia where the Israelites

wandered for forty years and is especially conspicuous

J. H. ROSS

97

on Mount Sinai. Although usually shrubby or twisted

and gnarled in desert areas, in favourable localities

A. tortilis may attain a height of 15 metres. Its wood

is very hard, close-grained and durable and thus

admirably suited for use in the construction of the

ark of the Tabernacle.

Other authors feel that A. seyal is more probably

the species referred to while A. nilotica may also be

involved. The almost complete absence of references

to the “shittah” tree in the later books of the Bible

suggests that the tree was not a native of northern

Palestine.

It has also been suggested (Feliks, 1971) that A.

albida Del. may be the species in question as it grows

in the Jordan Valley near the mouth of the river

Yarmuk. A. albida is an erect tree with hard light

wood which would have provided timber of suitable

lengths for the construction of the ark.

In the books of Numbers, Joel, Joshua and Micah

the word “shittim” is used as a place name probably,

according to the Authorized Version, because of the

abundance of acacias at those places at that time.

The “Abel-shittim” of Numbers 33:49 literally means

“the meadow (or moist place) of the acacias.”

The acacia of the Bible is not Robinia pseudo-acacia

L., the common black locust of eastern North

America. This species was confined to North America

in biblical times and was only introduced into

Palestine at the end of the seventeenth or the beginning

of the eighteenth century.

While there is little doubt that the word “shittim”

refers to a species of Acacia, there is much controversy

about the following verses of Exodus 3: 2-4:

“And the angel of the Lord appeared unto him in a

flame of fire out of the midst of a bush: and he looked,

and, behold, the bush burned with fire, and the bush

was not consumed. And Moses said, I will now turn

aside, and see this great sight, why the bush is not

burnt. And . . . God called unto him out of the midst

of the bush."

One of the possible explanations discussed by

Moldenke & Moldenke l.c. is that the “flame of fire”

may have been the brilliant crimson-flowered mist-

letoe, Loranthus acaciae Zucc., which grows in pro-

fusion on Acacia species in Sinai and in biblical

lands. The crimson flowers of the mistletoe stand out

conspicuously against the green foliage and yellow

inflorescences of the host plant and some authorities

are of the opinion that the story of Moses and the

“burning bush” may be an allegory referring to the

flame-like appearance of the mistletoe among the

branches of an Acacia.

THE WRITINGS OF CLASSICAL ANTIQUITY

Scientific botany owes its origins to curiosity about

the medicinal properties of plants. Theophrastus

(370-285 B.C.), the distinguished Greek philosopher,

who was first a disciple of Plato and afterwards the

favourite pupil of Aristotle, applied to the vegetable

kingdom the principles of classification based on logic

associated with his teachers (Stearn, 1958). This is

revealed in his work which has come down to us

entitled The Enquiry into Plants. The Enquiry into

Plants is chiefly concerned with the plants of the

Mediterranean region around Greece, but it also

contains some of the observations made during

Alexander the Great’s military expedition into Asia

in the years 331-323 B.C. It is not known from what

source Theophrastus first became acquainted with

species of Acacia but the following mention is made

in the Enquiry of acacias which he would have had

an opporuinity of seeing in Egypt during his visit

to the country at the invitation of Ptolemy:

“Thus in Egypt there are a number of trees which

are peculiar to that country, the sycamore the tree

called persea the balanos the acacia and some others.”

(Theophrastus, Enquiry IV, ii, 1 ; transl, Hort 1 : 291

1916).

Theophrastus continued (IV, ii, 8; transl. Hort

1: 299, 1916):

“The akantha (acacia) is so called because the whole

tree is spinous ( akanthodes ) except the stem; for it has

spines on the branched shoots and leaves. It is of

large stature, since lengths of timber for roofing of

twelve cubits are cut from it. There are two kinds,

the white and the black; the white is weak and easily

decays, the black is stronger and less liable to decay:

wherefore they use it in shipbuilding for the ribs.

The tree is not very erect in growth. The fruit is in a

pod, like that of leguminous plants, and the natives

use it for tanning hides instead of gall. The flower is

very beautiful in appearance, so that they make

garlands of it, and it has medicinal properties, where-

fore physicians gather it. Gum is also produced from

it, which flows both when the tree is wounded and

also of its own accord without any incision being

made. When the tree is cut down, after the third year

it immediately shoots up again; it is a common tree,

and there is a great wood of it in the Thebaid . . .”

The plant referred to in the latter part of the quota-

tion is apparently A. nilotica.

More than two centuries elapsed after the death

of Theophrastus before a reference is again found to

a plant that is alleged to be an Acacia. This reference

is in Georgies, the work of Virgil (70-19 B.C.), the

celebrated Roman poet. Elfriede Abbe in The Plants

of Virgil’s Georgies 129 (1965) translated Georgies

2: 118-119, namely, “quid tibi odorato referam sud-

antia ligno balsamaque et baccas semper frondentis

acanthi”, as follows:

“Why should I tell you of the balsam that sweats

from the fragrant wood and the berries of the ever-

leafy Acacia?”

Elfriede Abbe was of the opinion that the Acacia

referred to in these lines is A. nilotica, and maintained

that by “baccas” Virgil meant either the round heads

of flowers or the moniliform pods which resemble

a string of beads.

However, the above interpretation is at variance

with some earlier opinions. Parkinson 1549 (1640)

wrote:

“Some have thought that the Acanthus baccifera of

Virgill, mentioned in the second of his Georgickes,

in these words Quid tibi odorato . . ., should be this

tree (A. nilotica), as Servius Grammaticus, and Christo-

ferus Landus both of them commenters upon Virgill

say; but without true judgement as Guilandinus noteth

it, who would referre it to the Acanthus Aegyptia of

Athanaeus; ...”

According to Wood in Rees 1 (1802), Virgil had

two different plants under this name (Acanthus).

Wood continued:

“The acanthus with which he adorns the handles of

Alcimedon’s cups, in the 3d Eclogue, and places

in the Corycian’s garden, in the 4th Georgic, and

the Egyptian acanthus of Theophrastus, are two very

different plants. Virgil mentions another acanthus as

being an ever-green plant, and producing berries, or

a small round fruit; baccas semper frondentis acanthi,

are his words; and Theophrastus tells us, that his

Egyptian acanthus is a prickly tree, and bears pods

like those of beans ... It is plain, that the acanthus of

Theophrastus is the acacia, a tree, from some species

of which we have the gum arabic now in use: and the

acanthus of Virgil, mentioned in the places above

cited, is a garden herb, . . . The other acanthus

mentioned by Virgil in the fourth Eclogue, and

second Georgic, is the acanthus of Theophrastus.”

98

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

Two important botanical works appeared during the

1st century A.D., namely, the Natural History of Pliny

the Elder, and De Materia Medica of Dioscorides.

The Natural History or Historia Naturalis of Pliny

the Elder, Caius Plinius Secundus (A.D. 23-77), the

distinguished Roman writer, is regarded as one of the

most valuable relics of classical antiquity. The work

is in effect a vast encyclopaedic compilation in Latin

from the writings of Greek authors of the knowledge

of his time and contains a large section devoted to

plants.

Pliny made the following reference to Acacia:

“In the same countrey there groweth a thornie plant,

which the inhabitants make great account of: and

especially that which is in colour blacke, because

it will abide the water, and never rot nor putrifie in

it: and therefore excellent good for the ribs and sides

of ships. As for the white thorn of this kind, it will

soone corrupt and be rotten. But both the one and

other, is full of prickles even to the very leaves. The

seed lieth in certain cods or huskes, wherewith cur-

riers use to dresse their leather instead of gals. The

flower that this thorne beareth, is beautifull, whereof

folke make faire guirlands and chaplets ; profitable also

besides and good for many medicines. Out of the

barke of this tree there commeth a gum likewise. But

the cheefeth commoditie and profite that it yeeldeth is

this, cut it down when you please, it will be a big

tree againe within three yeares. It groweth plentifully

about Thebes in Aegypt, among Okes, Olives and

Peach-trees, for the space of three hundred stadia from

Nilus: where the whole tract is all woods and forrests,

and nathelesse well watered with fountaines and

springs among.” (Pliny the Elder, Natural History 13,

9; transl. Philemon Holland 1:390, 1601).

Rackham 4: 137 (1945) was of the opinion that the

black-thorn is A. nilotica and suggested that the white-

thorn may be A. albida.

Philemon Holland 1: 391 continued with Chapter

1 1 of the 1 3th book of Pliny as follows :

“The best gum in all mens judgement is that which

commeth of the Aegyptian thorne Acacia, having

veines within of checkerworke, or trailed like wormes,

of colour greenish, and cleare withall: without any

peeces of barke intermingled among, and sticking

to the teeth as a man cheweth it. A pound thereof is

commonly sold at Rome for three deniers.”

Chapter 12 of the 24th book of Pliny contains a

lengthy discourse on gums and their varied uses.

De Materia Medica of Pedanios Dioscorides (1st

century A.D.), the celebrated Greek physician and

botanist from Anazarba in Asia Minor, is an encyclo-

paedic herbal in which are described the plants then

reputed to have healing properties. It provided a

valuable record of Greek herb-lore being based on

his own observations and experience and on the

writings of others including Crateuas, personal

physician to King Mithridates (120-63 B.C.) (Stearn,

1954). As Pliny noted, Crateuas not only wrote about

herbs; he also painted them in colour.

No contemporary version of the manuscript sur-

vived but the work has descended to us by the copying

of copies of yet earlier copies. Consequently there

exist manuscript versions of varying ages, com-

pleteness, accuracy and authenticity. Of these illu-

strated manuscripts of Dioscorides’s work, the most

important is the Codex Aniciae Julianae (also known

as the Codex Vindobonensis and Codex Constantino-

politanus). This work was made at Constantinople

about the year 512 A.D. as a gift for the lady Anicia

Juliana, the daughter of Flavius Anicius Olybrius,

Emperor of the West in the year 472. A number of the

illustrations in the Codex Aniciae Julianae appear to

be derived from those made by Crateuas about two

thousand years ago.

In the facsimile edition of Dioscorides’s Codex

Aniciae Julianae published in Leiden in 1906 Book 1,

Chapter 133 is translated as follows:

“Akakia. Acacia growes in Egypt. It is a Thorne,

growing well neere to the bignesse of a tree, the fruit

of it lying in cods as that of the Lupin.”

After the work of Dioscorides there is little botanical

history for about 1500 years. During this long period

Dioscorides’s herbal was venerated and uncritically

accepted as the infallible authority. Then, in the

sixteenth and seventeenth centuries, the correction and

extension of Dioscorides’s work became one of the

main preoccupations of the many herbalists as a

result of which numerous illustrated herbals were

published.

THE PERIOD 1500-1754 A.D.

Among the first herbals to appear in which reference

is made to Acacia or to plants referred to by this name

was that of Otto Brunfels. The first edition of Brun-

fels’s Novi herbarii tomus II was published by

Johannes Schott in Strassbourg in 1531 and the

following reference to Acacia and discussion on gum

arabic appears on p. 9 :

“ACACIA succus spinae crescentis in Aegypto.

Resudat ex eo gummi quod Officinae gummi

Arabicum, Celsus sine epitheto Gummi appellat.

Vulgus Medicorum hodie ignorat quid sit Acacia, &

pruna ilia sylvestria quae in spinis proveniunt, pro

vera Acacia interpretantur, gravi errore: cum ijs

prorsus Dioscor, descriptio non respondeat. Sed de

his alias.”

According to Riddle, in Dictionary of Scientific

Biography 4: 121 (1971), by 1544 approximately 35

editions of Dioscorides’s translations and com-

mentaries had been produced. The most illustrious

edition was Pierandrea Mattioli’s which was first

published in Italian in Venice in 1544 under the title

Di Pedacio Dioscoride Anazarbeo libri cinque della

historia & materia medicinale. Reference to A.

nilotica is found on p. 84, and the gradual improve-

ment of this work occupied much of the remainder of

Mattioli’s life. It was translated into many languages

and appeared in a long series of editions.

A Latin version entitled Commentarii in sex libros

Pedacii Dioscoridis was published in Venice in 1554.

Reference to Acacia appears on p. 113 and on p. 114

there is an illustration of “Acacia altera”. However,

the plant figured is not an Acacia but a member of

Papilionaceae (tribe Genisteae).

Reference is made to Acacia on p. 129 of the edition

published in 1560 and on p. 51 of the 1562 edition of

the work. Once again, the plant figured in these

editions is not an Acacia but a member of Papilion-

aceae. The illustration from the latter edition is

reproduced here as Fig. 1.

The same papilionate was illustrated under Acacia

on p. 64 of Mattioli’s Kreiiterbuch published in 1563,

but there is some discussion devoted to gum arabic.

In yet another edition entitled I Discorsi nelli sei

libri di Pedacio Dioscoride Anazarbo della materia

medicinale published in 1581 after Mattioli’s death,

reference to Acacia and illustrations of “Acacia

prima” and “Acacia seconda” appear on p. 162.

Neither species illustrated is an Acacia: both are

papilionates. The figure of “A. seconda” is the same

as that which appeared on p. 1 14 of the 1554 edition

under the name “Acacia altera”. Two papilionates

are illustrated under the name Acacia on pages

210-211 of the 1585 edition.

J. H. ROSS

99

CrnFo *t>y Strom prc/polmj

jr,

Acacia.

LT

.f rembu ScfeU^aibom.

fc'fr pvi m.i'/muo j (c b«bw it a mgftc prareefyo tyctam Bjqwuis', H-tylcpfJy mijfg Im6>w«

g-jt rwr)-flami a reomw,

<J PKwjenq a tTIocy.

C pArojcm'i g<ff (kibmebo a liirfx^o /dyu d trpEe a ;mmaa% tVToc gcgi) gc(i itctco jbuflaoaa,

til a xwffaymxa. iDavca |c bo Ictarftretf Ctcrcj fe 6 ©cun (?rogiwagi/pro j/tfiroog Cot»:2llc mu.-

fv gq prwr bob’-' • n> JUioioree tvbbi rcypran. TOjittaui gc(I mK y proti rojgibancym Hcffrewiy

(cm/obytobar :/ CyiiSu ru ©fu/<j prou o(uti) au(t/ ITfclycrty »?|ijtmgc/pugm y pnHabamjm/

3 i q tnatfti

Fig. 1. — A species of papilionaceae illustrated under the name

“Acatia” in Mattioli, Commentarii in sex libros Pedacii

Dioscoridis 51 (1562).

Sanitaria, €ap,j£tu toctnflf

«L Irlatilfte.

©tebfobreeg enbe wncfilrg cap &iufe(i<xm) fi'jrj conf enbe bioogfjeoag natocreg

(ot ig beg twtebei) graeDi/cnbe wot fubrtj! pag fubffaritieg.

« £tacht enoe ibctchingje.

3Die groyne Mabereg wg ^>aufef>oorq Dienrg onj &aafcg boer af fe mokeg f of beg **

fpbfetj jjkcltick bat ■tmrckde/ende Die faofe Dir baet of gbrmaecki moibt/eo peccoelen*

be enbe maecht appeti it/enSe eo feer goet beg gbeneg Die oerbtf enbe coKfachftch fug.

3Die brfiekme vat) -SoBfeboon) (topper) beg loop 9e« biiptp/enbe fletpi Me o uerotoe ©

j)gbe cfoet oag beg ownsoeg/enbe a!le onnatnerliidte Koetganck,

JOie tMifefeg nag £>aufekoorg ig loogjje ghetpeptlit mokeg Oaf (apj geel/olfm? bat (G

jopi 0 aet mebe Ditkwife wofebt,

Tlcacia. £ap ♦

Ac«», CCfattoert.

ItfTfldaee eegffekenDeenbeboomoc^ti'cfi gbeitwe/maernfef eetftf opvaf

fenbe/biogbenoe oeel tachegbiemet feerpe boienegbrfet fbg.&ilg Mocm

kena fng iCTt.lTfoet ea fcaeef gbefijck erg ffupine/enbe trafi ig iggfie fan

tprg/enbe baec Bt petffrneg bat fap/baf ftt'tct ea enbe oock Ttcaoa gbe*

heeteg rcoibt. C (©jaetfe.

31cacia irxifttg Higppteg ata 3Dio(co«bea fctgff .

C/ftacm.

©ifffekenSegbctraetboibfghehretegig <2>riecp enbe fg Saftjg Aceoe/rnbeait*

Dew gheeneg naeig ea one behenf.Jg Die Jtpotrheg ec/i oock onbekent / hoe reel nock*

•ana bat bie nang baet ghebleueg ea/ bie ig bie Tlpofehe g(egben?ip<»bt beg fope tag

ttilbe pjupnikena oat on eetbtelgtk Ae«« ghebtefeg twibt,

6k6 ft jBomete.

Fig. 2. — A species of Acacia (possibly A. nilotica) illustrated in

Dodoens, Cruijdeboeck 741 (1554).

A reference to Acacia (reproduced here as Fig. 2)

appears on p. 741 of Rembert Dodoens’s Cruijdeboeck

published by Van der Loe in Antwerp in 1554.

Dodoens’s real name was Rembert van Joenckema and,

according to Florkin in Dictionary of Scientific

Biography 4: 138 (1971), he changed it to Dodoens

(son of Dodo), Dodo being a form of the first name

of his father. The name was latinized into Dodonaeus,

from which the French further transformed it into

Dodonee. The plant figured appears to be a species of

Acacia and circumstantial evidence would suggest

A. nilotica but it is not possible to identify it with any

certainty.

Three years after the publication of this first Flemish

edition, a French edition with numerous additions

appeared under the title of Histoire des plantes, the

translation being carried out by Charles de Escluse

(Carolus Clusius). Dodoens supervised the production

of this book and in view of the numerous corrections

it is in reality a second edition of the Cruijdeboeck.

The French edition of 1557 was itself translated into

English by H. Lyte in 1578 and appeared under the

title “Dodoens, A Nievve herball, or historic of

plantes”. The reference to Acacia in this work appears

on p. 685 where two species are illustrated. The illu-

stration of “Acatia Aegyptica” is the same one as

that which appeared in the first edition of the Cruijde-

boeck, while the other species figured under the name

“Acatia altera” is a papilionate.

In 1576 Mathias de 1’Obel (De Lobel or Lobelius)

published his Plantarum seu stirpium historia in

Antwerp which was in effect an enlarged version of

his Stirpium adversaria nova published in 1570-1.

De l’Obel devised a system of classification in which

the different groups were distinguished by the charac-

ters of their leaves which was a significant advance

on previous efforts. He thus distinguished roughly

between the classes now known as monocotyledons

and dicotyledons. Reference to Acacia is found on

p. 536 where two species are illustrated (reproduced

here as Fig. 3). The figure on the left, “Spina Acatiae

. . .”, is A. nilotica, the figure being similar to that

which appeared on p. 741 of the first edition of

Dodoens’s Cruijdeboeck in 1554 except that a frag-

ment of a pod and a seed have been added. The figure

on the right referred to as “Acatia altera” is a papi-

lionate.

De l’Obel’s work was translated into Flemish and

published in 1581 in Antwerp under the title of

Kruydtboeck where the same two species were

illustrated on p. 110 under Acacia.

100

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

1 1 1 a 51 y «

Sfbu Ac at! ^ Diofc. Ainaf.pq. 409. A c a t I a ‘her* DitJioriJii.

AtontlJHiTbeopb. An^uil. Atluerf. pag. 410.

Dioscor, Aut'uin Acgjpto nafotur jpittj in far arhorityfrutuaft, r.cnfein reaum a!toSer:s,fciem haht

C4ndidltiHi& [men lupno jbntic infollitulu ex quo ftccsu expreffus pa jr.tr tit vmbra, tiiger ex nut/trofennn-, fuL -

rufiu cx yirtdi. V.tgj'ur viodttc rufti< & odoratus , vt ttt l}JC ariose effe potell. Ahqui exprtmunt ex feiiu & [tnane

[ticiun: , Mjit.it c? guwm ex c.x fpnu. vis et ad Jj'iJfwduni » rcfitgcr*iidt*wq. effeax . Succus octtlorum v edicjvieiuv

vttUs.v.tlet adtgvcru ficrum, vkera qua ferpunt . pet utonc s, pterygia, & erts vU era. proadtntesoutlos reprint, tren-

fiuni abuud.tuii.t;r fiflit.procidu.tw vuluain cobibet. at ant iluum fttppumtt ,aut pouts aut falter tnjitus. captllcsdem-

gr.u . i trttt: ttr ocular urn medic ament is ,mto in aqua.mox effufo quod comteuaat%donee pur a extet aqiu . <r ua t cgittir

in p.tftid'js. vritur etiam in (Tud.t fidelu , donee fiflilefornactbm percoqujtur-. torreturq, m carbonum balitu per fata.

Spina decocl urn folutos art us feta comnnttu. Prafertur eitu JpinxgumwitquodM vem'uuUrum Jpniew lontraiiatur,

& vwi mode perluceatyltgni expers. proxtmum eft candidum. for Mum feti ac refmofum, inutile. Cut it meatus ob-

firtitt. & medic ameutcrum quilus admtfcetur, aaimonim bebetat. ambu His excuotiluum , pusiulas nonpatirur

er umpire.

Gal. Acatu; Sc plan uipla.iccrba eft, & fru<fhis&' fucctis.qui lerus quidem &imbccillior A: minus

mordax redditur, vrpotc acrimoniam quondam in lauationedcponcns. Porro (i parti alicui fanx illina-

tur , proiinus«Mmi& ficetorem &contra&.im cfticict .nullum quidem caloris fenfum imichens • fed

ncc frigons admodum valencem. vndc cooftat mcdicamentum id efle fhgidum Sc terrofum, iirimfta

qu.id.im etiam cOcmiaaqnca. Et fan£ conic&uracft, non eflefimilarc, verifin quafdatn in fefedilper-

las habere panes tenues & calidas , quae in ip&ablutiotie legregcntur. 1 ftoqueA' hoctcriij otdmis

* ii!otu veto. «iccantium,& f-cundi refrigeranuym, vbi quidem elotum tucrir, *abfquc eo verb pi imi.

A c a t i a altera Diofc.

Succuseitts adftringir, viribus inferior* & ocuiatibm medicamemis inyrilis.

Fit AX INULA

Fig. 3. — Illustrations of “Spina Acatiae . (A. nilotica) and

of “Acatia altera . . . (a species of Papilionaceae) in de

l’Obel, Plantarum seu stirpium historia 536 (1576).

In the first edition of Dodoens’s Stirpium historiae

pemptades sex sive libri xxx published by Christophe

Plantin in Antwerp in 1583 reference to Acacia

appears on pages 739-741. The first species is A.

nilotica, the figure (reproduced here in Fig. 4) being

identical to the figure which appeared on p. 536 of

De l’Obel’s Plantarum seu stirpium historia (see

Fig. 3), and the second species figured under the name

“Acacia altera” is a papilionate. The latter figure is

similar to the figure which appeared in De l’Obel’s

work and was possibly based on it but is not identical.

Dodoens, De l’Obel and Clusius permitted the use of

their wood-blocks in each others work which explains

the occurrence of the identical figure in the worxs of

de l’Obel and Dodoens.

In the second edition of Dodoens’s Stirpium historiae

pemptades sex sive libri xxx published by C. Plantin

in Antwerp in 1616 A. nilotica is discussed and illu-

strated on p. 752.

In 1587 the Historia generalis plantarum, sometimes

referred to as Historia plantarum lugdunensis, a book

which formed a compendium of much of the botany

of the late sixteenth century, was published in Lyon.

Although no author’s name appears on the title page,

the work is attributed to Jacques Dalechamps (Quinby

HISTORIC

vnam fiue colhm krrentibus compofita,

vclun funt Lends: florcsaibidulobi liuc liliqu^

plana: lata: quidem inftar Lupmi qua

parte (emen continent : quod modo vna.quan-

doque duabuslobi particulis anguftiorc colio

coh^rombuscomprchenditurdcmcnglabrum

Srrelplendens.ExpruTiitur autem cx Ins fuccus,

qui in vnibra ixccatus: nigerex maruriselbusnj-

r-opfos&uc lubruffus ex lmmatuns : alsqui verb

cx loins Sc fru&u luccum colligunc.

Gumrai verb Sc c)f hac ftirpe profluir.

Na{cjtur,«iic Diofoondes, Acacia uiAEgypto.

Inter fruticcs autcni ac arborcs pcrpcruovi-

rentes Acaaam ctiam numer.it Petrus Bclio-

ni us Singu 1. lib. i .cap. x l i i iu

Hatic ipm.im Grxci ai«£x/<weciam noftra x-

mcvocant: Latinis ftrmlicer Acacia eft: Dici-

rur Sc /Egvptia fpina.

Succus Acacia: quoqucnoincn habet. Offi-

cinas Gcrmani? huiusloco luccum cxlylucftri-

bus Prumsexpreflum lubftituunt : quem &: A-

caciam appellant.

GumnninOificims Arabicu gumi dieftur.

Matthioluspro Acacia Arborem Iuda: rcce-

nonbus didam, piclura cxhibet.cui praarer ve-

titarem fpinas addiidit,vr Acaciam menttrecur.

fedtamen nccfic Diokor;dis ddcriptioni re-

Ipondcntcm reddidit.

Acacia’ aueem (uccus.vtGalcntis aic.noneft

partium fimilanum, led cum tubitantialrigsda

Sc rcrrca.cui & quadam aqueaconmnfta.quaC.

dam ettam in (ele difperfas partes habet tenues

ficcaltdas. Ellicaq;tcrtij ordinis exiccantium,

refngerantium veronon lotus prinn : kcundi

vero vbi lotus fueric. Lotionc emmacnmonia

ac partes ealidiorcs depomc.

Coducit autem, inquit Diofeondcs.adocu-

lorum mcdicamcta,&: ad eryhpclata, herpetas,

chimctla.pterygUjOris vlcera: proptolcsoculo-

lorum repnmit : in ailium abund anti am iiftit:

procidencem vtcruin connahit: alui fluorem

fuppritmtjtum potus , turn per clyftcrcm mdi-

tus; capillosdcmgr.it. Ad ocuiorum autem mc-

dicamcntalauatur.

GunHad(lt;ngens&:modicc rcfngcrans eft:

cmplafticaip vero vna lacultace habet.quaacri-

moriiam medicanictorum, quibus admiicctur,

rctundit. Ambuftis cum ouo ilhtuin.vcfieas ex-

atari prohibet: Diolcoridcs.

I)e A Cite i a altera. Cap. x i i i i.

Ltr r a Acacia fpina* /Egyptueafli milis,

vc Diolcondcs ait , led longe minor tene-

norque.ftirps bumilissaculcatislpmisvonuarj.i-

r.i : loiia habens Ruta : lemen n\ ftliquistrium

aut quaruor capacibus Lcntc minus. Etpro hac

quidem, exhibita ohm a Match, olo fust, ihrps

lit'nofiS&aculcanstamis.folns terms eoliarcn-

tibuSjRuta: graueolcntis non valdc ddrinuli-

bus;

Fig. 4. — Illustrations of Acacia (A. nilotica) and “Acacia altera”

(a species of Papilionaceae) in Dodoens, Stirpium historiae

pemptades sex sive libri xxx 740 (1583).

1: 165, 1958; Stafleu & Cowan 1: 591, 1976). A

detailed discussion and a summary of the know-

ledge of plants referred to Acacia up until this time

appear on pages 160-163. The three species illustrated,

namely, Acacia Aegyptiaca, Acacia Matthioli and

Acacia Altera Matthioli, are reproduced here as

Figs 5 & 6. Of the three, only Acacia Aegyptiaca is

an Acacia, probably A. nilotica. The figure of this

species is curious in that some of the blank spaces

have been decorated with insects and falling leaves.

The figure of Acacia Altera Matthioli is likewise

decorated with insects.

In 1592 Prospero Alpini published a treatise in

Venice entitled De plantis Aegypti which was a pioneer

study of the Egyptian flora. Alpini, a doctor, accom-

panied the Venetian Consul, Giorgio Emo, to Egypt

where he took advantage of the opportunity to study

the local flora (Arber, 1938). Alpini’s medical training

led him to approach the new flora in the traditional

manner of attempting to correlate the plants he

encountered with the names and descriptions found

in classical sources. However, when this was im-

possible, he described the plant under its local name

and based the description upon specimens that he

p:rsonal'y examined (Stannard in Dictionary of

Scientific Biography 1: 124-125, 1970).

J. H. ROSS

101

QV IN DVMETIS R.EPER. LII

« Acacia

ACypuaca.

<A C A C I a

i. ^MtU&noU.

pcco ,cx quofcminc in horco Patauino , aliifque quamplurimis Vcnctorum

locis fato, orta eft ifta Acacia. Accepit & Morganus plantam nuperex niaris

Occidui mfulis Perunianis, ne quisputec peculurcm ZEgypto aut Arabix. In

neruulis pra:longisfc>l»ola func Scorpioidislcguminisfimilis, aucherba: Sferra

cauallo nominal*, in coco virgulto , gracili , aliquot rigidis Spirals in furculis

cxcrtis. Scminis tota ftiiqua non maior vno alceroue Lupino ftmul mudis,

fed fuss alueolis ib ft inch. Qiurc Matchiolus non Acaciam, fed ft Spinas

drmas,arbon Inda: limilcm pictamdcdiftc vulctuGcuiusfiliqua^fiue lobi non

func Lupim fimiles,lcd Genift*,duplo laciores, vcScnar coniprdfxia Pena hie

deferipea verior. Altcrum Acacix genus, y£gyptix Spina: eftlimilc, longcmi- Fimr

nils, tenerius, hum ile, aculcorum vallo munitum : Folia habet Rune : iemen

Lencicula*,mtnus,Autiimno,in loculis conuexis,tcrnum,quarcrmimuc capaci-

bus. Acaciam altera genuina cllc hicappidacenfent Periri Hcrbarii, loliis Ru-

ta: am Cytifi,tcrnis, ftiiqua Gcniftella*, aut vulgaris Glycyrrhiz?,nouaailx ehi-

gie, dorlo obtuftore, Scalrcro latere, quaft acieacutiorc, truaut quaruorctiam

Gcniftella: feminadura continence, qua* nondummatura flauenr,pofteanign-

canc.HacTyrrhenum & Ligufticum.arque Medircrrancuni liras fcarec, &: plc-

raque alia Italia- loca. Dc Acacia, eiulque genenbus hare prodidic Theophraft. i.ih.<

Spina ex co nomcn acccpir,qubd tora arbor aculeis horreat, cxccpco caudicc: ‘-TF

nam&fnper gcrmina,foliaqucaculeos habet. Altitudinc proccraeft)vt qua ad

duodecim cubicos pcrucniac.Matencs cx ea ca-ditur tedis idonea.Eiusduoge-

ncra:qua.'dam Candida, quxdam nigra.Candidaimbecillis,&: facile purrefeens,

nigra robuftior, fie incorrupra.Quarc in nauibus fabricandis ad carum coftas

ilia veuntur. Redanon valdc afliirgit, Frudus in ftliqtu modo ieguminum,

quo incolxcoriapcrhciunc Gall* vicc.Flosvfqucadcoaipedu pulcber, vt ex

co coronas facianr,eft Sc mcdicamcntis vtilis : quamobrem a mcdicis colligi-

tur.Manat&Gummicx ip£i,tumvulncrata , rum fpontc line vlla plaga. Cum c„,n,

Fig. 5. — Illustrations of “Acacia Aegyptiaca” (probably

A. nilotica) and “Acacia Matthioli” (a species of Papilio-

naceae) in Dalechamps, Historia generalis plantarum 161

(1587).

Alpini discussed A. nilotica on pp. 4-6 and his t. 4

is reproduced here as Fig. 7. Inflorescences and fruits

of A. nilotica were depicted for the first time in this

figure so it represents a significant advancement on

previous attempts to illustrate the species.

Some of Alpini’s original descriptions were included

in the writings of Linnaeus who regarded Alpini with

sufficient esteem to name the genus Alpinia (Zingi-

beraceae) in his honour.

The year 1597 saw the publication by John Norton

in London of the first edition of John Gerard’s The

Herball or Generali Historie of Plantes. It appears

(Arber 129, 1938; Quinby 1: 188, 1958) that Norton

had commissioned a Dr Robert Priest to translate

Dodoens’s Stirpium historiae pemptades sex, which

was first published in 1583, but Dr Priest died before

the work was published. His manuscript came into

the possession of Gerard who altered Dodoens’s

arrangement to that of De l’Obel, added some of his

own comments, and published the work as his own.

De l’Obel was requested by the printer to correct

Gerard’s more obvious errors while the book was in

the press, but Gerard’s impatience with the correc-

tions prompted him to stop De l’Obel and insist on

immediate publication.

In his discussion of the “Aegyptian Thorne” on

p. 1149 Gerard stated: “Dioscorides hath made

mention of two sorts of Acacia, this whose figure we

have set downe is the right Acacia”. However, the

plant described and illustrated as “Acacia Dioscoridis,

The Aegyptian Thorne” is in fact a papilionate and

the same plant which was figured under the name

“Acacia altera” in Dodoens’s Pemptades (1583).

The 1633 edition of Gerard’s Herball was enlarged

and amended by Thomas Johnson who succeeded in

correcting many of the errors in the 1597 publication.

Two species are illustrated on p. 1330 of the 1633

edition. The illustration on the left “Acacia Dioscori-

dis. The Aegyptian Thorne” is of A. nilotica, the

figure of the species being similar to that which

appeared in Plantin’s edition of Dodoens’s Stirpium

historiae pemptades in 1583. By changing the species

illustrated under this name, Johnson succeeded in

correcting the error made by Gerard. The illustration

on the right “Acacia alteratrifolia Thorny Trefoile”

is the same species of Papilionaceae as that figured by

Gerard in 1597 but a different illustration was used.

Aldinus, Exactissima descriptio rariorum plantarum

Romae in Horto Farnesiano 2-7 (1625), provided,

under the name Acacia Indica Farnesiana, a very

detailed description and two illustrations of a plant in

cultivation in the garden of Cardinal Farnese in

Rome. The illustration on p. 2 (reproduced here as

Fig. 8) shows the habit of the plant and the illustration

:S1 Hli i ORIji OMN1VM PLAN 1 ARV M,

adafuerit,tcrtioanno flatim renafcicur.Eius

copu & fylua mgens circa agru Thebamim

eft., vbi & Qucrcus , &. Perlea Olea quoque

eodem Joco gignitur, non aquis flumj riguo

(dill at entm plufquam trccenus ftadns) fed

manantibus fpontc aquis, multi enim fontes

in eo func tradu. Hate Theophralli , Plinius

eo quern fupraindicauiraus loco trosfcripftt,

atquevertit.Eadem dc re alio loco idem ftc

fcripfit,quo vulgati Codices cx vetuftifsimis

fic lunc emendandi. Eft & Spina? fuccus Aca-

cia.Ftt in yfgypto aiba,nigraquc arboreritem

c icmine vindi,aut maturo, fed longc melior

c priori. Fit & in Galatia, ceneriore fpinofaq;

arbore. Semen hums lencicula: fimilerminorc

eft tantum dc grano, & folliculo. Colligitur

Aurumno , ante colledum mmio validius.

Spiflatur fuccus ex foihcuhs aqua cadefti per-

fufis : mox in pila tufts cxprimitur orgams:

denfatufquc foie in morcariis cogicurin pa-

ftillos.Fic dc ex foliis minus effteax. Ad coria

perhcicndafeminc pro Galla vtuntur. Folio-

rum fuccus & Galacicx Acacia? nigerrimus .

improbatur:icem qui valdc rufus. Hie Plmius alteram Acacia? fpcciem , quam

in Cappadocia Sc Ponto naici ait Diofcoridcs,Galatiani appeliat, &» cum prio-

re,yEgyptia,fcilicct,confundit,ft femen omnium, & non Galatia? tantum Len-

ticufte ftmilefacit. Dioicorid enim prions femen Lupino comparat, altenus

Cummi minus Lente eftc ait.Gummi quod cx AcaciaSptna ftillat,id pradcrcnr,qubd in

vcrmiculorum fpeciem concrahicur,& vim modo pcrlucecjigm expers-.proxi-

mum eft candidum-.fordidum verb ac refmofum inutile.Id ctiam opnmumef-

fe trad it Plinms,vcamcu!atum , colore glauco , purum, finccqruce,dentibus

Lffi.9 H'ft- ajjlirrens.jlle0phraftus lacrymofum humorem banc Spinam Ferre tradition

t<P 1 c corncc/cd in folliculo.Serapio Arabicum nuncupar,quod cx Arabia regione

Lib.i.Diof. ^gypto hnitima (uo tempore impoftarccur,inquit MatthioL Scire autem

cap■,l' oporter Gummi Arabicum Ofikinis didum,multum a Spina* zEgy pcia* Gum-

fi*“xoAcft nil difetcmcqucenim concradis vcrmiculisfimilc eft, led grumis verftcolo-

Gunu Ava- nbus conftat. Aduchinon ita pridcm cccpit ad nos ex atgypto v era Acacia,

blct!lTU m paranda Thenace, aliifquc componendts mcdicamcntis valde expetita,

qua nos ad hoc vfque cempus caruimus. Id Gummi Galenus aliquando

Lib. 7. me. Su fM Jv&ukov vocauic , ob id forrafl'e quod Theophrailo audorc in agro

Thcpano magna fit fpinofx huius arbons copia , & ingeos fyiua. Spina?

4^ ‘ yEgypcice fuccus , inquic Diofcorides, oculorum medicamenns vtilis : valet

vucs tu>;« xgnem facrum , vlcera qua ierpunt, pcrnioncs , pterygia , oris vlccra:

proctdcnces oculos reprimit : mcnftum abundanuam lillic , prociduam

vu’uam cohibct: citam aluum ftftic > aut potus,aut fubter inditus:capiilos

demgrat. In qua RuelUj iranslarione id mcrico reprehendendum videtur,

qaoo»jp%4 potw yuuawTov , mcnftum abundantiam liftir, interprciecur,cum

ftuxum muliebrcm dicer c dcbuifler.Dtffertenim fluxus mulicbns a menlibus,

Lib^.ca.<?j. feumciiftruapurgationcimmodica, vt pcrfpicuc docet Paulus.Scd Plinium,

Ruellius imitatus eft, quo loco cadcm qua: Dioicor.habet.Acacia purpurea, aut

lom, zAC AC 1 A ALTERA

Fkj. 6. — Illustration of “Acacia Altera Matthioli” (a species of

Papilionaceae) in Dalechamps, Historia generalis plantarum

162 (1587).

102

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

DE PLANTIS AEGYPTI

ACATIA, SANT, ET KAKIA.

Fig. 7. — Illustration of “Acatia . . . .” (A. nilotica) in Alpini,

De plantis Aegypti t. 4 (1592).

on p. 4 (reproduced here as Fig. 9) is of a twig bearing

flowers and fruits. Aldinus recorded that seeds of the

plant were received from the island of St Domingo

and were germinated in the year 1611. This appears to

be the first direct reference to a species of Acacia

indigenous to the western hemisphere or from an

area other than the mediterranean or middle east.

The quality of the figure reproduced here as Fig. 9

greatly exceeded that of any previous illustration of

an Acacia species, and of many which appeared in

subsequent works. It is a faithful representation of the

species now known as Acacia farnesiana and is of

historical importance as reference to Aldinus’s work

is made by Linnaeus in the protologue of Mimosa

farnesiana L. in his Species Plantarum ed. 1 : 521

(1753), the basionym of Acacia farnesiana (L.)

Willd., Sp. PI. 4: 1083 (1806). Analysis of the proto-

logue of M. farnesiana indicates that Linnaeus

relied to some considerable extent on Aldinus’s

description and illustration of Acacia Indica Farne-

siana for his concept of M. farnesiana, and that the

epithet “farnesiana” was taken from Aldinus. In the

absence of any specimen on which Linnaeus could

have based his phrase-name of M. farnesiana, the

Aldinus p ate reproduced here as Fig. 9 was selected as

the lectotype of A. farnesiana (Ross, 1975b).

Aldinus discussed A. aegyptiae (A. nilotica ) in

detail on p. 7 of his work together with the characters

that enabled the species to be distinguished from

Acacia Indica Farnesiana.

It is perhaps as well to mention that there is some

controversy over the authorship of the work here

attributed to Aldinus. Pritzel, Thesaurus Lit. Botanicae

ed. 2: 58 (1871), attributes the work to Castellus and

notes “Operis “Exactissima descriptio” autor est

Petrus Castellus, atque falso sibi vindicavit Aldinus;

typographus enim hisce etiam verbis: “In gratiam

Tobiae Aldini scripsi cuncta” profitetur, Aldinum

auctorem non esse. Seguier”. Aldinus was Cardinal

Farnese’s physician and so the work may well have

been dedicated to him. Saccardo, La botanica in

Italia: 12 (1895), credits Aldinus with the work. In

the Catalogue of the Library of the British Museum

(Natural History) 1 : 26 (1903) the work is attributed to

Aldinus but there is a note reading “By some this has

been considered to be really the work of P. Castelli.”

The first edition of Caspar Bauhin’s Pinax theatri

botanici was published in 1623 in Basle. In this work,

which included all of the plants known to western

botanists up until this time, he listed for each species

known to him all of the names (i.e. synonyms) that

J. H. ROSS

103

had been applied previously by earlier authors.

Bauhin arranged the plants according to their natural

affinities as he saw them and not merely in alpha-

betical order. A description of the genus Acacia

appears on p. 391 and reference is made on p. 392 to

two species, namely, “1. Acacia foliis scorpioidis

leguminosae . . . .” (i.e. A. nilotica), and “2. Acacia

trifolia . . . .” (i.e. a papilionate). Linnaeus drew

heavily on Bauhin’s Pinax and made constant reference

to it in his Species Plantarum.

Volume 1 of Jean Bauhin’s major botanical work,

Historia plantarum universalis, was published in

1650 after his death under the co-authorship of J. H.

Cherler and D. Chabrey (Stafleu & Cowan, 1976).

A generic description of Acacia and a long dissertation

appear on p. 426, and on p. 429 there is a description

and illustration of Acacia vera (reproduced here as

Fig. 10). The upper figure was clearly based on the

illustration (t. 4) in Alpini’s De plantis Aegypti

(1592) which is reproduced here as Fig. 7. This

upper figure featured by Bauhin was in turn repro-

duced in Duplain’s Historie des plantes de l’Europe 2:

715 (1737) under the name Acacia Aegyptica.

The year 1635 saw the publication in Paris of

J.-P. Cornut’s Canadensium plantarum, aliarumque

nondum editarum historia which contains an early

record of about 30 north-east American plants. The

4 Rariores plants

ACACIAE INDICAE FOLIA, FLORES, ET SILIQ.VAE.

Fig. 9. — Illustration of “Acacia Indica Farnesiana” (A. farne-

siana) in Aldinus, Exactissima descriptio rariorum planta-

rum Romae in Horto Farnesiano 4 (1625).

LIBER XII

gc f.Ol D;oCoi,J,,c,Wn [ Herb, median,-

c !w . .

joif (? . Non 'gnoI« Amatui Onbaf. dc Aetmm Acx-

tuuiiu «•- Ci^aibukulamium Arabic* fiueAegyptia fptria,

idem faccre ,quos Ruclltut fccutu* in , fed non

line errore. Reprchendit Agricola Dmfcoodcm,

Stt**- qu°dm Acgypto lolttm nafci roemorat, quum

tamcn Si alibi prouenut : Nara & Synam fpmam

diciconfbrc: Errare etiam infupcrcos dicu qui

Z»cr*»U Acaaaro fieri puceni cx fpma ArabicaiGammi A

A. ra^:cam > quod ii qnibuldam Babylonicam dica-

,**«»*>. ®“»S«rapioR«n cx arbore Acacia promanarr fen.

bcre,qualitfpina Acgyptia cuius mcrameru Pli-

ruur Jib. ij-cap. n. Launeigitur vocan hoegutru

mi , Acaas gurami , yel gummi Acgyptur Spi-

r£,4*ir Ataciam modb fiuticem(eorn*

UiifUM4. bnf if) medo arborem Aegyptucam fpinolkmap-

peilac rmnih attend fuccu <? porais cspnmi fbi-

fcu. Acacia arbor non fert poma ,ftdliliquas <?

qtuhm fuccus cxpnmi poteft.

Fnfchlini errorem fcqumir Hclfricus , qui Sc

jpfc nonnoutc Arabicum gummi ellc cx cadcm

arbore.

Aquiuoca Ad Acaai iqumocitioncs & fabftuura ordi-

nedcucmarrtus. /k,«irfliqiudcmt/c74, ab A cm*

TAmAr,n~ r«^«»Tamarindos Cordodufta, non parum c-

Ut uarnt , dc hac iain dnftimi, Acaaam Matthioli

lu- mbilaJiudcIfcquiiriiiu/ait^j^^rnw, cut appi-

fa fpinc obfcntauim*,priufquam qnilquara id

(cnpiis mandartede qua poftcatuo c.ipite.Mu!ci

tamcn credidcrunt clic Acaaam Prumuu fy'ue-

Deem { 0d>lcJ>£ H born Acaaam Gcmtanicnn vo-

tat , Dodon. ] fed hi maximein errore vetfantur.

Smamrtu Et hmcmagno errore perhiafom eft Euft*dii,3c

”»'» b‘H- Euchar. RccC, Hai^h. Guidoni.PLc.-ario&nnn-

nullisAcacum die kiccum Pruneftoruru (yluc

ftnnm. Nobilcum fcniit Cordu-. in Du.'cor

Trag. I.omc Btunlcll. lo.Agric. Hcrnv-I. A

4(*,ia A- plurcs alii. Eft & Acacia altera Vtteubus vt Di.T

nrM. condi, quatmukts dicunr ftuccx trdolius , qui

abis ecu Sc nobis exiftnnatur Aff*l*rbiiifemnda d.

AffAu- qUa pjj(^ n> |,jJ_ + Conliderandum an Ar-

W.” ^ Pu,llaix AcmaAtcrj find* Si^rpio Theopbrafti,

j.rjMhtt lit Alpalathus z.Mon!pelicnfium <lc qua lbft n.

i '.Attr/f. Iibeodcm. Ac. fia( mut.ua liter* c. in I. ) eft

Canter, in horto Cjniulunlui caralcus.

Sublb tun n>aior pars M d comm Sc Officina

rumAcicij ddkienl z nm.nmf/htjtui

bmeut [ quern nonnulit Acaaam Ofticinarum vocant,

v.HMvum nial^ ) quorum fentcntix neutiquam fubfen

O1" bendum, beet Macthiol. in Hi It Germ. Acacia

itribat.Quodmiruracthmn ptioribus in Diolc,

comm. Vcterum potulsopiniorcm Jtqtutur, vt

Diofcondis.quicius vice turn folia Rhois , turn

H) fttifa. foliorum lu c cum , rum Hypoolbden

nicdicametttisadmifcct ,quibus vtendum potids

focrit CumDiofcoridc Acaciar !<<co fupponunt

Hypociftm Auic. Sylu. fuchfin aniibal. Col-

lcg. Florer.t. Piacot. Bnrgur. Marthiolun epift. 1.

Lacuna mDtolcor. Monacbi in Mel. Fcrncl

M.M. Fauard. in Gal.de fimpl. Camer. in Hift.

Matthtol. Germ. Croncnburg. lbft. Lugd. Sc

.*»nu< Anbd.Romanum: Succum Lemilci.Auic Mat-

thiol. Lacuna. Monachi in Mcf. Apollo rad. i.

GaLdclubft. Cairu-r. apud.Matth. Germ. Cto-

nenburg.& Hilt. Lugdcncnfis : Rhois folia , ( aitodore effe nonmgrato Cayn quidem ffauj

xim-ft/n. M-utitiol. in Diofcor. Placoc. Lacuna , Braftau. vclpallido: In montams verb Arabic.albo : Ar-’J^-

kktuf»!f* Bacchand. lo Agnc Croncnburg. A: Hi ft. Lug- ' borem magninidneMon,ramilque circlacftscx- cXl v^.

duneniis : Puipara Rhois, Collegium Florcn- 1 pan.ii»,caudice Prnni.trunco^quah.

t,nura & Borgarucns : Succum femims aut fobo- 1 Dc Acacia loquitur Rauv volft. ct)m ait : Grca

rnro Monachi in Mcf. 3c Apollo. Non placet ran- \ HalepumconfpiararSpmolus ftetes Acaae.qui ictu'

Torn. L Nn j ^

Fig. 10.— Illustration of Acacia vera (A. nilotica) in J. Bauhin,

Historia plantarum universalis 1 : 429 (1650).

Acacia Americana Robini described on p. 171 and

illustrated on p. 172 is Robinia pseudo-acacia.

John Parkinson’s Theatrum Botanicum was

published in London in 1640. Parkinson, who was

honoured with the title of Herbalist to Charles I,

took advantage of Bauhin’s Pinax which enabled him

to give the detailed nomenclature of each plant, and

in this respect his work shows improvements on

Gerard’s Herbal. However, the rudimentary system of

classification adopted by Parkinson was inferior to

the system used by De l’Obel or Bauhin and serves to

illustrate the lack of progress made in classification by

the herbalists.

Parkinson’s discussion of Acacia commenced on

page 1 547 as follows :

“ Dioscorides hath made mention of two sorts of

Acacia, the one of Egipt, and the other of Cappadocia,

and Pont us: Theophrastus also speaketh of two sorts,

blacke and white: that of Egipt is reasonable well

knowne, but of that sort of Pontus, there is some

controversie among Writers, some taking one bush

to be it, and others denying it to be it, the differences

of Theophrastus sorts are onely expressed in the

wood, . .

Parkinson recognized and provided descriptions of

three species. The first species enumerated was

dem lotibemun Afiadxftibftixuere Gummi Aim-

bicurn : quod ( licitex cadcm planu ,cx qua A-

cacia , prouer.iat ) turn f apore , turn /inbtu mui-

tiim diffeiac. Scd fajtc bretmer <tn no title fuf£>

ACACIA VERA.

Cacic pla- Ddcriptio

IcfcciHiorctn

cx femme na-

cam Pauun vi-

di, (ptrui hor-

rentem, firruli.

bui feri fpinu

CrcMm , fiuc

j B.rb'.m valgb

^ di^lx codcin-^

tut h*bc-

:ont-g-t.Lu- Sdij.

ii Sd»quis L"?’

. quodammodo

Lk "'s'111-

ituc furcc,

".“'P >

piunbus quan

abpbus loleic

ifthmis di ft i li-

no habcoquod

dicam ) c6quc fragili , vhicuiquc loculo ad latera -

apcrtili fcmen inct: vmeum ,mir.ujquitu Carta-

but fiuc S;bqus,fpadiccum, drculo in vngucm

du^rijvcluu in Tamarindij notacum ,t>blongu,

uu contcmpcnci duricic mco medulla. C?-

, Sdiquam intrinlecus lucangit pellicula

ruffeftens inter quara 3c cxiimum corcccmcon-

hquor apparct .qual.sfer^ in Cap®/ , fed

valdc piucus. Tlorcs pokherrimi funt Iphenca ru™

rocunditate rglobuli magnitndanu maioris Pifi,

quiconlbnt velutilanuginc molli, cxfcpuluc-

rem remitten tej, colons funt flaui, pcdiculi fingu-

iorum ebbuloram, tenues, unaaies • Almnn«

104

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

“ Acacia sive Spina Aegyptia vera. The true Acacia,

that is Egiptian thorne or binding Beane tree”

(i.e. A. nilotica), and the illustration which ac-

companied his description is reproduced here as

Fig. 1 1 . His description reads as follows :

“The Egiptian Thorne groweth in some places to be a

great tree, and rather crooked then straight or rising

high, covered with a blackish barke, spreading abroad

great armes and branches, full of sharp thornes, with

many winged leaves set on both sides of them, that is,

with foure wings of leaves on a side, made of sundry

small ones, set opposite on a middle rib, without any

odde one at the end, although it be so expressed,

Bellonius saith that he counted 350 of those small

leaves, that were upon the whole branch, and yet all

of them might but cover his thumbe: the flowers grow

among the branches, like flockes of wooll, of a whitish

yellow colour, where after come somewhat large and

thicke huskes, like unto the Lupine or flat beane cods,

blacke when they are ripe, and bunched forth against

the places where the seedes lye, in some three or foure,

and in some more, each as bigge as a small wild Beane,

round, and of a grayish or ash-colour, almost shining:

the tree abideth alwayes with greene leaves thereon,

and yeeldeth of it owne accord a white gumme in

small curled peeces like great wormes, and greater

round peeces if it be wounded.”

The second species described was “ Acacia Americana

Farnescena. The West Indian Acacia or binding Beane

tree” (i.e. A. farnesiana), and the illustration of it i ;

reproduced here in Fig. 11. It is at once apparent that

this figure of A. farnesiana was based on the figure

first published by Aldinus (1625). The third species,

namely, “ Acacia secunda sive altera Dioscoridis, The

true second Acacia of Dioscorides ” is a papilionate.

Herbaria Nuovo by Castore Durante was first

published in Rome in 1585. The work was translated

and reprinted for many years and the reference to

Acacia on p. 3 of the edition published in Venice in

1667 is reproduced here as Fig. 12. It will be noticed

that the description of Acacia del Matthioli (which is a

papilionate and not an Acacia) is written in the form

of a poem. Acacia D’Egitto is probably A. nilotica.

A good illustration of A. farnesiana, similar to the

one originally published by Aldinus (1625), appeared

under the name “Acacia Americana” in Plate 35 of

Abraham Munting’s Waare Oeffening der Planten

(1672) and the plants referred to the genus Acacia

were discussed on p. 32.

A discussion of Acacia vera (i.e. A. nilotica ) appeared

on p 398 of F. Hoffmann’s Clavis Pharmaceutica

Schroederiana (1681), while reference was made on p.

399 to Acacia Germanica (i.e. Prunus sylvestris).

And now, for the first time, our attention turns to

the southern hemisphere. During his stay at the Cape

of Good Hope, the High Commissioner Hendrik

Adriaan van Reede tot Drakenstein, Lord of Myd-

recht, authorized the Commander, Simon van der

Stel, to explore the Copper Mountains of Namaqua-

land (Reynolds, 1950). Van der Stel’s expedition left

on 25th August 1685 and reached the Copper

Mountains some 300 miles to the north on 21st

October. The expedition returned to the Cape on

26th January 1686 after exploring part of the coast.

It may be assumed that the artist Hendrik Claudius

was a member of Van der Stefs party and to him are

attributed the seventy-one pages of coloured drawings

including those of plants encountered during the

expedition.

The official record of Van der Stefs expedition to

Namaqualand was removed from the Dutch East

India Company’s Archives in 1691 or 1692, and all

trace of the manuscript (‘Dag Register’) was lost

until 1922 when it was identified by Professor G.

Waterhouse in the Catalogue of the Fagel Collection

acquired by Trinity College, Dublin in 1802. The

first part of the manuscript consists of the Journal of

Fig. 1 1 . — Illustrations of “Acacia

vera sive spina Aegyptiaca”

(A. nilotica) and “Acacia

Americana Farnesiana” (A.

farnesiana) in Parkinson,

Theatrum Botanicum 1548

(1640).

J. H. ROSS

105

P E L D V R A N T E.

AB VTILLO. A

Frjngit ABVTILLVM return, trabil itquehpiUos,

ynnamque achmuUct reaumque dolores.

NOMI Lat .AhniHlum, ahbfi alter j.Ial.M.i!uau'[co

b if hr do. :

FORMA. Ha le fogliedi zucca.ma minon.lilce , &

ci’viu for^jlilfima lanugine ricouertejfa il tufto alco vn

' eombico , 6l mezo , & quachc voita piu. Ifioriaurei

pertuctoil fulto .daquali nafee i! feme nero aencro a

certigufei come capidi papaueri. Halaradicclonga

con molce radiccttc attornq.

j LOCO.Nafce ne i campi,& fponcaneamente nc gh

[ QVALrrA', 8c VIRTV'.Dx dentro.ll femebctiuto al c

pclodvnadrainmae meza con vino , caccia fuori le

pietre,8c le rendle.Pcouoca 1‘vriaa » 8c mitiga la dim-

culta,8c i dolori chc per cio fopragiongono.

l.’ACQVAcheda cutcala piantaii dclhla ugii

effcrri lnederimi.

acacia

del Matthioli .

A mbuflis prodefl,ocultfque ACACIA, ficroque

lgni,tm fiflu meHfes.iluumque fluentem,

Serpenteu rnorbos fvutytwgnque capillum ,

Hfc eidemj'iriagit //>$**> vtthumque repomr,

Tu(iulaque blticciuim^lengn, pernio cedunt.

MbicM . . , |

SPFTIE.E 1’ acacia di due fpccie,cioc nuggiorc>&:

minoro.

• FORMA. Nafee a guifadifpinafruticofa non s’ in ’

alzando:Ha il fior bianco 8c il feme co.n j ii lapino ne‘

i baccclMal quale fi Iprcrn: il fucco, at ieccali a Pom-

bra, 8c chiamali col medefimo no;ne della pianca. j

LOCO. Nafee l'Acaciaprincipalmcnte in Hgitto

QVALirA'.Il fucco della maggiorccondcniato el

medtcamenco frigido nel fecondo gradoil lauato;A il

no lauaco c frigido del prim o 8c fecco ncl terzo grade. 1

VlRTV' 0/ .iMC^.Beuuto il fucco, 3c mdfone i chci-

fticri fcriiaa i fltiffi delle Donne , rimette la madricc

dislocaca , 8c riibgna i dtilli del corpo : A 1 flnllt del

ventre (i di con acqua rofa p oluerizata I’Acacia, l’Hi-

poquiltidc,8c pietra ematitc ana fcropo.vno. Ji/aor’. j

K llfuccoe vtileai medicamcnti deghocchi , iqitali j

ridncc.fc efcono del luogo loro, vale al loco facr o,all’

vlccrc ferpiginofe alii ptcrigijddiedica , Aairvlcerc j

della bocca: fa neri i capelli : fomentandofi con la dc- |

cottionditutta lapianrale gionturc finollefi ridil- ;

conoal luogo loro : Lafuagommanoneligotnma •

arabicache quclla^noncakroche vn mifeug iogi piu

gominc d' albcri ; che per non pocerfi por tar 1 Acacia

e da penfare,chc non li ci porti ancora la fua go.nma, '

laqualcha virtu di riempire 8c riferrare iporrt ticila

came, 8c impiaftrata con otiation lafeia ,iar . e vciichc

allc cotturc del fuoco.Sc faiu lc fpcronaglie. Al vomi- ■

to colcrico fi prende acacia,goma arabica, 8c dra^anti

con chiarad’ ouo.li fa nellapaddlafrittata,che n ma-

gia, 8c s’ applica alio ftomacho,Al Hullo d j i metlrui,o

jaague di nafo fi fa foppofta con acacia,8c fucco di po-

ligono,c qualche voita bifogna aggiongcrui del gclfo.

Falli dcll'acacia anche impiallro peril voinito,&. (icr

il flulfo del ventre co olio rofato,chiara d'ouo, acacia,

nullicc, 8c sague di drago.il fumacco e miglior fuccc-

dancodeH acacia,che no e il fucco de pr uni laUiatichi.

ACACIA D’EGITTO.

L'ACAClAcarbore.che nafee inluoghi loncani

dal mare ; ve nefonogranquantitadi quelli arbo-

ri nclli monri del Sinai , pofti quafi viemo al mar

rodo. Crdcono quafi della grandezzad' vn moro

ramofi , allargandofi al di fopra: ilchc c' auucrci-

fcc Diofcoride, chc non fileuanoinaltodritti. Il

A 2 tronco

807. Acacia Giraffaf. Willd. (?).

Fig. 12. — Illustrations of “Acacia del Matthioli” (a species of

Papilionaceae) and “Acacia D’Egitto” (possibly A.

nilotica) in Durante, Herbaria Nuovo 3 (1667).

Fig. 13. — Illustration of an “Acacia” of unknown identity

attributed to Claudius (T.C.D. No. 807). (Reproduced

from Waterhouse, Simon van der Stel’s Journal of his

Expedition to Namaqualand 1685-6).

the expedition, and the second part of the coloured

drawings. In 1932 Waterhouse published a book on

Van der Stel’s expedition to Namaqualand which

included half-tone reproductions of the Claudius

drawings.

One of the plants illustrated on Van der Stel’s

expedition (TCD No. 807) is reproduced here as

Fig. 13.

A translation of the notes accompanying the

drawing TCD No. 807 (Waterhouse, 1932) reads as

follows:

“This tree grows in such abundance in Namaqualand

that almost all the forests are composed of it. On

account of its multitude of hurtful thorns we call it

Thorn Tree, whereas the natives call it Choe. It is

moderately tall and large but crooked, and it has

good, hard, useful wood. It is found only along

rivers and brooks. Its flowers have a remarkably

pleasant smell and they are followed by a pod

containing a few flat seeds, the effects of which are so

far unknown.”

Along the route followed by the Van der Stel

expedition Claudius would certainly have encountered

the plant that is now known as Acacia karroo Hayne.

The only other Acacia species armed with paired

stipular spines and with flowers in round heads that

he may possibly have encountered was A. erioloba E.

Mey. However, the illustration attributed to Claudius

(Fig. 13) bears little actual resemblance to A. karroo,

to A. erioloba, or to any other South African Acacia

species. The leaves are shown to be consistently

imparipinnately compound whereas in all of the

indigenous South African Acacia species the leaves

are always paripinnately compound, and the pods

illustrated are at variance with those of A. karroo and

of A. erioloba. Father Tachard, who visited the Cape

in 1685, is quoted by Karsten 89 (1951), as having

said of Claudius that “He draws and paints animals

and plants to perfection.” As Claudius was an artist

of such high repute it seems odd that his illustration is

inaccurate in several obvious and significant respects

and bears so little actual resemblance to any of the

Acacia species. That is, of course, if the painting was

executed by Claudius and at present there is no reason

to doubt that it was not.

The figure published by Plukenet in his Phyto-

graphia t. 123, fig. 2 (1692), and reproduced here as

Fig. 14, is almost identical to the illustration executed

by Claudius on the Namaqualand expedition.

Plukenet’s illustration differs chiefly in that it has

been reversed from left to right, i.e. the leaves,

inflorescences and pods are depicted facing in the

other direction. In addition, Plukenet has added a

loose inflorescence, a loose pod and two more loose

106

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

seeds. The Claudius drawings are known to have been

copied and the copies copied and a set of drawings

was presented to Henry Compton, the Right Reverend

the Bishop of London from 1675-1713, while his

lordship was attending a Congress in Amsterdam in

1691. Both Petiver and Plukenet had access to the

drawings in Bishop Compton’s possession. The close

similarity between the Claudius and Plukenet illustra-

tions suggests that Plukenet copied Claudius’s

drawing: no specimen on which Plukenet could

have based the illustration has been located in the

Sloane Herbarium in the British Museum (Natural

History), although this does not, of course, provide

proof that Plukenet copied the Claudius illustration.

It does, however, strengthen the argument that

Plukenet copied an illustration and not an actual

specimen. Aloe and Gladiolus paintings prepared by

Claudius are known to have been copied by Petiver

and by Plukenet (Reynolds, 1950; Lewis et al., 1972)

and it is therefore a reasonable assumption that the

Plukenet figure reproduced here as Fig. 14 was also

copied.

The identity of the plant depicted by Claudius

(Fig. 13) and subsequently copied by Plukenet remains

uncertain which is unfortunate because Mimosa

capensis Burm.f., Prodr. FI. Cap. 31 (sphalm. 27)

(1768) was based on Plukenet t. 123, fig. 2. This

inability to positively identify the plant depicted has

led to the name Mimosa capensis being rejected as a

name of uncertain application (Verdoorn, 1954;

Ross, 1971, 1975a).

It is interesting and perhaps significant that the

plant depicted by Plukenet in his Phytographia t.

123, fig. 1 (see Fig. 14) is A. karroo. The figure was

based on a sterile twig of A. karroo. Herb. Sloane Vol.

99, fol. 3 in the British Museum (Natural History),

and is a good representation of it.

Plukenet’s Phytographia t. 123, fig. 1, was cited by

Plukenet in his Almagestum botanicum 3 (1696) under

Fig. 14. — Illustrations of “Acacia Africana, spinis candicantibus

horrida ” as t. 123 fig. 1 (A. karroo) and “Acacia

Africana, Abruae foliis, aculeata, . . . .” as t. 123 fig. 2

(identity unknown) in Plukenet, Phytographia (1692).

Acacia vera, by Linnaeus, Species Plantarum 1 :

521 (1753), in synonymy under Mimosa scorpioides,

and by Burm.f., Prodr. FI. Cap. 31 (1768), under the

name Mimosa nilotica, but in each case it was an

incorrect identification.

The first volume of John Ray’s Historia Plantarum,

which contains descriptions of all plants then known,

was published in London in 1686. The natural

system employed by Ray depended in part on the

differences on the formation of the embryo, that is,

plants were divided roughly into monocotyledons and

dicotyledons. “Acacia vera J.B.’’ is discussed in some

detail on p. 976 and there is reference to gum arabic,

and “Acacia Indica Farnesiana Aldini” is discussed

on the following page.

Horti academici lugduno-batavi catalogus by

Paul Hermann, the director of the Leiden Botanic

Garden from 1679-1695, was published in Leiden in

1687. Hermann considered two of the species referred

to Acacia by previous authors, namely Acacia

trifolia and Acacia Germanica vulgo, to be sufficiently

distinctive to exclude them from his concept of the

genus, although he retained Acacia Americana

(i.e. Robinia pseudo-acacia) in the genus Acacia.

Further reference to Acacia appears on page 36 of

Jacob Breyne’s Prodromus fasciculi rariorum planta-

rum secundus, which is in effect a catalogue of

plants observed by the author in gardens in Holland,

published in Danzig in 1689.

Leonard Plukenet’s Phytographia was published in

three parts in London in 1691-2 and species referred

to Acacia are illustrated in plates 121-123. Plukenet’s

works are of importance for purposes of typification

because Linnaeus frequently cited them in his Species

Plantarum, making reference to Plukenet’s illustra-

tions which were usually based on specimens in the

latter’s herbarium (now part of the Sloane Herbarium

in the British Museum). Many of the species described

by Linnaeus were known to him only by Plukenet’s

figure and brief descriptive note.

The plants illustrated in figures 3-6 of Plate 121 of

Plukenet’s Phytographia were referred to Acacia.

Linnaeus, Sp. PI. 1: 521 (1753), based his Mimosa

horrida, the basionym of Acacia horrida (L.) Willd.,

Sp. PI. 4: 1082 (1806), on Plukenet’s Phytographia t.

121, fig. 4 (1692). The descriptive phrase quoted by

Linnaeus “Acacia maderaspatana, foliis parvis,

aculeis e regione binis praegrandibus horrida, cortice

cinereo” appeared at the foot of Plukenet’s plate, and

was repeated, without additional information, in

Plukenet’s Almagestum botanicum 3 (1696). Although

Linnaeus never saw the actual specimen drawn,

there is nothing in his diagnostic phrase that could

not have been obtained from a study of Plukenet’s

figure. The specimen on which Plukenet’s t. 121, fig.

4 was based, is preserved in the Sloane Herbarium

Vol. 95, fol. 3 in the British Museum (Natural History).

Plukenet’s t. 121, fig. 5, was cited by Linnaeus, Sp.

PI. 1: 520 (1753), under Mimosa cinerea [i.e. Dichro-

stachys cinerea (L.) Wight & Arn.], but the identity of

the other two plants referred to Acacia in t. 121 is

uncertain.

Four species referred to Acacia were illustrated in t.

122 one of which, namely fig. 1 , was cited by Linnaeus,

Sp. PI. 1: 520 (1753), under Mimosa cornigera [i.e.

Acacia cornigera (L.) Willd., Sp. PI. 4: 1080, 1806],

and a further three in t. 123.

Reference has already been made to t. 123 fig. 1,

the caption of which is as follows: “Acacia Africana,

Fig. 16.— Illustration of “Acacia similis, spinis corniformibus

mexiocana” ( A . cornigera) in J. Commelin, Horti medici

amstelodamensis 1 1. 107 (1697).

Fig. 15. — Illustration of “Acacia altera vera” (A. nilotica ) in

PJnVpnft Almagestum botanicum t. 251 fig. 1 (1694).

The year 1689 saw the publication of Jan Comme-

lin’s Catalogus plantarum horti medici amstelodamen-

sis in Amsterdam. In this catalogue of the Amsterdam

Physic Garden, of which Jan Commelin was Director,

the plants are arranged alphabetically and on pages

3-4 four species of Acacia were listed. Of these, only

the first two are referable to Acacia , namely, “Acacia

vera J. Bauh ” (i.e. A. nilotica) and “Acacia

indica Farnesiana . . . .” (i.e. A. farnesiana).

In 1697 volume 1 of Jan Commelin’s Horti medici

amstelodamensis was published posthumously by his

nephew Caspar Commelin in Amsterdam. In this

fine work figures 105-107 were devoted to Acacia, but

of the three figures only the latter is readily identified

as a species of Acacia. Figure 107, which is reproduced

here as Fig. 16, depicts under the name “Acacia

similis, spinis corniformibus mexiocana” one of the

swollen -thorn Acacia species grown from seed

collected in Cuba. As indicated by Blunt (137, 1955),

the illustration of this Acacia, like many of the plants

figured, rises stiffly from the soil and masquerades

as a little tree. The figure in question was cited by

Linnaeus under Mimosa cornigera in his Species

Plantarum 1: 520 (1753). Commelin compared the

plant with Acacia Americana Aldini (i.e. A. farnesiana)

and discussed the reports of the small black ants

associated with the swollen spines.

H. ROSS

spinis candicantibus horrida, subrotundis foliis,

odoratissima.” The sterile twig depicted is referable

to the species now known as Acacia karroo.

This figure was later cited by Plukenet in his

Almagestum botanicum 3 (1696) under Acacia vera as

follows:

“Acacia vera, s. Spina Aegyptiaca, subrotundis foliis,

flore luteo, siliqua brevi pauciorib. isthmis glabris,

& cortice nigricantibus, donata . . . hujus Icon

exhibetur in Phytogr. nostr. Tab. 123 fig. 1 sub titulo

Acaciae Africanae, spinis candicantibus horridae,

subrotundis foliis odoratissimae”.

Reference has also been made to t. 123, fig. 2,

which was apparently copied from a drawing prepared

by the artist Claudius who accompanied Van der Stel

on his expedition to Namaqualand in 1685 and

represents a plant of unknown identity. Plukenet’s

caption to this figure “Acacia Africana, Abruae

foliis, aculeata, spinis longissimis horrida. ...” was

repeated in his reference to the species in Almagestum

botanicum 3.

Plukenet made reference in his Almagestum

botanicum 3 (1696) under the name Acacia altera

vera (i.e. A. nilotica) to the illustration of the species

which appeared in his Almagestum botanicum as t.

251, fig. 1, in 1694 (reproduced here as Fig. 15).

This figure was cited by Linnaeus, in Hortus Cliffortia-

nus 209 (1738), under the name “Mimosa spinis

geminatis, foliis duplicato-pinnatis” and later in the

synonymy of Mimosa Senegal, in Sp. PI. 1: 521

(1753), which illustrates that to his earlier concept of a

species armed with paired spines in Hortus Cliffortia-

nus Linnaeus subsequently added in the Species

Plantarum the diagnostic phrase name of a species

armed with spines (actually prickles) in threes.

108

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

The three volumes of J.P. Tournefort’s Institutiones

rei herbariae, editio altera with fine copper engravings

by Claude Aubriet, one of the classics of systematic

botany, were published in 1700 in Paris, a work

described as a second edition of his Elemens de

botanique published in 1694. Tournefort adopted the

concept of genera and species formulated by Caspar

Bauhin but, unlike Bauhin, he placed the main

emphasis on the genus. Whereas Bauhin gave only the

name of the genus and supplied the species with

descriptions, Tournefort consistently provided the

genera with names and descriptions and added the

species without providing special descriptions. Tourne-

fort distinguished genera primarily on the characters

of the corolla and fruit but he also accepted genera

differing from allied genera by vegetative characters

which he termed ‘genera of second rank.’ A generic

description of Acacia , and of Mimosa, is given on p.

605 of Volume 1 together with an enumeration of

species, and both genera are illustrated in t. 375 of

Volume 3 of the work.

The two volumes comprising Hermann Boerhaave’s

Index alter plantarum quae in horto academico

Lugdano-Batavo were published in Leiden in 1720.

On p. 56 of the second volume the genus Acacia is

attributed to Tournefort and twelve species are

enumerated under the genus, some of which belong to

other genera.

An alphabetical list of the plants in the Botanical

Garden at Pisa is contained in Catalogus plantarum

horti Pisani by Michele Angelo Tilli published in

Florence in 1723. Several species of Acacia are

discussed on p. 2 of the catalogue and illustrations

appear in Plate 1.

The year 1737 saw the publication in Regensburg of

the first volume of Johann Weinmann’s Phytanthoza

Iconographia. A generic description of Acacia appears

on p. 8 and several species attributed to Acacia are

enumerated, many of which belong to other genera.

A hand-coloured illustration of A. nilotica, referred

to as “Acacia Aegyptiaca Vera” is figured in t. 10.

Another work which appeared in 1737, in addition

to the first edition of Linnaeus’s Genera Plantarum,

was Thesaurus Zeylanicus by Johannes Burman

published in Amsterdam. Plants are arranged in

alphabetical order in this work and the enumeration

of species referred to Acacia appears on pp. 2-6.

Burman used polynomials and the first species

described on p. 2 under the name “Acacia aculeata,

multiflora, foliis pennas avium referentibus” and

illustrated in Table 1 (reproduced here as Fig. 17) is

Acacia pennata (L.) Willd. Linnaeus cited this plate

under Mimosa pennata in his Sp. PI. 1: 522 (1753).

The other Acacia illustrated in Table 2 under the

name “Acacia spinosa ex alis spicata, foliis pennas

avium referentibus” is Dichrostachys cinerea (L.)

Wight & Arn. and the plate was cited by Linnaeus

l.c. : 520 under Mimosa cinerea. Many of the other

species referred to Acacia belong to other genera.

In an appendix to the Thesaurus Zeylanicus (1737),

Catalogi duo plantarum africanorum, Burman listed

a number of plants collected at the Cape by Hermann

among which were three Acacia species. The second of

these referred to as “Acacia Africana, angustifolia,

spinis majoribus, fiore odoratissimo. Acacia Africana,

spinis candicantibus horrida Plukn." is apparently

A. karroo, but the identity of the other two is not

clear from the descriptions.

In Adrian van Rooyen’s Florae Lugdensis Pro-

dromus published in Leiden in 1740, which comprises

Fig. 17. — Illustration of “Acacia aculeata, multiflora, foliis

pennas avium referentibus” (A. pennata) in J. Burman,

Thesaurus Zeylanicus t. 1 (1737).

a list of the plants in the Leiden Botanical Garden, the

genus Acacia, along with the genus Inga, was treated

as a synonym of Mimosa in keeping with the generic

concept adopted by Linnaeus.

Linnaeus did not employ the name Acacia in a

generic sense in the first edition of his Genera Plan-

tarum published in 1737 (or in subsequent editions in

1742, 1743, 1752, 1754 or 1764), in Hortus Cliffor-

tianus (1738), Hortus Upsaliensis (1748), in the

first edition of Species Plantarum in 1753 or in the

second edition of 1763. In these publications the

genus was relegated to synonymy under Mimosa, and

in the synonymy of Mimosa in Genera Plantarum

Linnaeus attributed the genus Acacia to Tournefort.

Linnaeus had a much broader generic concept than

TournefoFt and some of his successors being influenced

primarily by characters of the androecium and

gynoecium.

Linnaeus did use Acacia in a generic sense in his

Flora Zeylanica 217 (1747) and his name has been

associated with the genus from this publication.

However, as this was prior to 1753, the starting point

of modern botanical nomenclature, the genus Acacia

L. has no standing in present nomenclature.

Philip Miller (Gard. Diet, abridg. ed. 4, 1754) was

the first author to employ the name Acacia in a

generic sense subsequent to 1753 and is, therefore,

regarded as the author of Acacia. Philip Miller used

the name Acacia in a generic sense from the first

edition of his Gardeners Dictionary in 1731 to the

seventh edition in 1759, and in the first to the fifth

abridged editions published between 1735 and 1763.

Miller’s taxonomic knowledge was considerable and

J. H. ROSS

109

he was slow to accept Linnaeus’s views on onmen-

clature and classification. Although the first edition of

Linnaeus’s Genera Plantarum was published in 1737

and the first edition of Species Plantarum in 1753,

Miller did not accept all of Linnaeus’s generic and

specific concepts uncritically, but retained as distinct

many genera defined by Tournefort and suppressed by

Linnaeus. It is largely on account of his departures

from Linnaeus’s concepts that Miller’s works pub-

lished subsequent to 1753 derive their nomenclatural

importance (Stearn, 1969).

In the seventh edition of The Gardeners Dictionary

(1759) Philip Miller adopted the phrase-names from

Linnaeus’s Species Plantarum wherever applicable

and provided new ones where required for species

not known to Linnaeus. Miller wrote under his

treatment of the genus Acacia in the seventh edition:

“Dr. Linnaeus has joined the plants of this genus, and

also the Inga of Plumier, to the Mimosa , or sensitive

plant, whereby he has multiplied the number of the

species greatly, and occasioned some confusion. I shall

choose, therefore, to refer them to their former genera

again; for as all sorts of Mimosa have articulated pods,

and their leaves move on being touched, so the

Acacias, which have neither of these properties,

may very reasonably be made a distinct genus, and

hereby the ancient officinal name will be preserved.”

He then proceeded to enumerate the characters of

the genus Acacia.

It was not until the eighth edition of The Gardeners

Dictionary in 1768 that Miller finally accepted

Linnaeus’s binomial nomenclature for species. In his

preface of this edition he stated:

“In the last edition of this work, the author adopted

in a great measure the system of Linnaeus, which was

the prevailing method of ranging plants then in use

among botanists; but as many of the plants which

were treated in the Gardeners Dictionary, were not to

be found in any of Linnaeus’s works then published,

Tourneforts system was also applied to take in such

as were not fully known to Dr Linnaeus; but since

that time the learned professor having made great

additions to his works, and those additions being

generally consulted for the names of plants, the author

has now applied Linnaeus’s method entirely, except

in such particulars, where the Doctor not having had

an opportunity of seeing the plants growing, they

are ranged by him in wrong classes, . . .”

Thus Miller belatedly converted to Linnaeus’s

system in the eighth edition of his Dictionary and

relegated Acacia to synonymy under Munosa where

he noted:

“The Acacias are so nearly allied to the Mimosas

in their characters, that Linnaeus has joined them

in the same genus; and as his system is now generally

followed, so in compliance with that I have done

the same.”

Subsequent authors did not follow Linnaeus’s

broad generic concept and treated Acacia and Mimosa

as distinct genera, although the limits of the genera

remained ill-defined for a long time. The generic

limits of Acacia were finally clarified by Bentham

(1842) and have not been seriously in doubt since.

ACKNOWLEDGEMENTS

I am most grateful to Mr R. Zabeau, Royal Botanic

Gardens, Kew, England and Mr P. Fay, Dept, of

Crown Lands and Survey, Victoria, Australia, for

photographing the illustrations, and to the Directors

of the Royal Botanic Gardens, Kew and of the Royal

Botanic Gardens and National Herbarium of Victoria

for permission to reproduce the illustrations.

UITTREKSEL

Die voor-Linnaeiese geskiedenis van die p/ante

toegewys aan die genus Acacia weerspieel in 'n sekere

mate die ontwikkeling van botaniese beskrywing,

klassifikasie en Ulustrasie. Aandag word gevestig op

sommige van die vroegste verwysings na plante bekend

as behorende tot die genus Acacia en op verwysings in

uitgekose kruieboeke en publikasies tot met Philip

Miller se beskrywing van die genus in die vierde

verkorte uitgawe van sy Gardeners Dictionary in 1754.

REFERENCES

Arber, A., 1938. Herbals, their origin and evolution’, a chapter in

the history of botany, 1470-1670. 2nd ed. Cambridge:

Cambridge University Press.

Bentham, G., 1842. Notes on Mimoseae, with a synopsis of

species. Hook., Lond. J. Bot. 1 : 318-392.

Blunt, W., 1955. The art of botanical illustration. 3rd ed.

London: Collins.

Blunt, W., 1958. The illustration of early botanical works.

In J. Quinby, Catalogue of botanical books in the collection

of Rachel McMasters Miller Hunt 1: xli-xlvii. Pittsburgh:

Hunt Botanical Library.

Feliks, J., 1971. Acacia. In Encyclopaedia Judaica 2: 198-199.

Jerusalem : Keter Publising House.

Fluckiger, F. A. & Hanbury, D., 1874. Pharmacographia — a

history of the principal drugs of vegetable origin pp. 206-216.

London: Macmillan & Co.

Fuchs, L., 1542. De historia stirpium commentarii insignes.

Basel.

Fulton, J. F., 1958. Medical aspects of early botanical works.

In J. Quinby, Catalogue of botanical books in the collection

of Rachel McMasters Miller Hunt 1 : xxxii-xxxv. Pittsburgh :

Hunt Botanical Library.

Karsten, M. C., 1951. The Old Company's Garden at the Cape

and its superintendents Cape Town: Maskew Miller.

Lewis, G. J., Obermeyer, A. A. & Barnard, T. T., 1972. A

revision of South African species of Gladiolus. Jl S. Afr.

Bot., Suppl. Vol. 10. Cape Town: Purnell.

Moldenke, H. N. & Moldenke, A. L., 1952. Plants of the

Bible. Waltham: Chronica Botanica Company.

Parkinson, J., 1640. Theatrum Botanicum, London.

Quinby, J., 1958. Catalogue of botanical books in the collection

of Rachel McMasters Miller Hunt 1. Pittsburgh: Hunt

Botanical Library.

Rackham, H., 1945. English translation of Pliny's Natural

History, Vol. 4.

Rees. A., 1802. The cyclopaedia; or universal dictionary of arts,

sciences and literature 1. London

Reynolds, G. W., 1950. The aloes of South Africa. Johannes-

burg: the Aloes of South Africa Book Fund.

Rickett, H. W., 1958. Botany from 840 to 1 700 A.D. In J.

Quinby, Catalogue of botanical books in the collection of

Rachel McMasters Miller Hunt 1: xxiii-xxxi. Pittsburgh:

Hunt Botanical Library.

Rochebrune, A. T. de, 1899. Toxicologie Africaine 2. Paris.

Ross, J. H., 1971. Acacia karroo in' Southern Africa. Bothalia

10: 385^01.

Ross, J. H., 1975a. Notes on African Acacia species. Bothalia

1 1 : 443^147.

Ross, J. H., 1975b. The typification of Mimosa farnesiana L.

Bothalia 11 : 477 -472

Sachs, J. von, 1890. History of botany, 1530-1860. Translated

from the German by H. E. F. Garnsey and revised by I. B.

Balfour. Oxford: Clarendon Press.

Smith, W. & Fuller, J. M. (eds), 1893. A dictionary of the

Bible comprising its antiquities, biography, geography and

natural history . 2nd. ed. London: Murray.

Stafleu, F. A., 1967. Taxonomic literature. Regnum Vegetabile.

Vol. 52. Utrecht: I APT.

Stafleu, F. A., 1969. Miller’s 1754 Gardeners Dictionary.

Taxon 18: 713-715.

110

A SURVEY OF SOME OF THE PRE-LINNEAN HISTORY OF THE GENUS ACACIA

Stafleu, F. A. & Cowan, R. S., 1976. Taxonomic literature.

1: A-G. Regnum Vegetable Vol. 94. Utrecht: Bohn,

Scheltema & Holkema.

Stearn, W. T., 1954. Codex Aniciae Julianae: the earliest

illustrated herbal. Graphis (Zurich) 10: 322-329.

Stearn, W. T., 1957. An introduction to the Species Plantarum

and cognate botanical works of Carl Linnaeus. Prefix to the

Ray Society facsimile of the first edition of Linnaeus’s

Species Plantarum 1753. London: Ray Society.

Stearn, W. T., 1958. Botanical exploration to the time of

Linnaeus. Proc. Linn. Soc. Lond. 169: 173-196.

Stearn, W. T., 1961. Botanical gardens and botanical literature

in the eighteenth century. In A. Stevenson, Catalogue of

botanical books in the collection of Rachel McMasters

Miller Hunt 2, 1 : xlii-cxl. Pittsburgh : Hunt Botanical

Library.

Stearn, W. T., 1969. Introduction to Philip Miller, The Gar-

deners Dictionary ("Abridged Ed. 1754). Lehre, Germany:

Verlag von J. Cramer.

Stevenson, A., 1961. Catalogue of botanical books in the

collection of Rachel McMasters Miller Hunt 2. Pittsburgh :

Hunt Botanical Library.

Verdoorn, I. C., 1954. The nomenclature of the Cape Acacia.

Bothalia 6 : 409^11 3.

Waterhouse, G., 1932. Simon van der Steps journal of his

expedition to Namaqualand, 1685-6. London.

Bothalia 13,1 & 2: 111-114 (1980)

The genus Sypharissa (Liliaceae)

A. AMELIA OBERMEYER*

ABSTRACT

The genus Sypharissa Salisb. (Liliaceae) is resuscitated and four species belonging to this genus are

validly published.

rFsumF

LE GENRE SYPHARISSA ( LILIACEAE )

On ressuscite le genre Sypharissa Salisb. ( Liliaceae ), auquel cette publication rattache validement quatre

especes.

A group of species in the Liliaceae, all bearing

caudate floral bracts, has been variously placed under

the genera Drimia, Urginea, Urgineopsis, Thuranthos,

Rhadamanthus, Rhodocodon and Litanthus. In 1977

Jessop sank Urginea, Urgineopsis and Thuranthos

under Drimia. However, the species now all placed

under this genus form a most heterogeneous

assemblage and further study may assist in arriving

at a more satisfactory arrangement. It was observed

that a group of related species placed in a separate

genus by Salisbury (1866), differed sufficiently from

Drimia sensu lato and Urginea to be recognized as a

genus on its own. The genus Sypharissa of Salisbury

(1866) was not taken up by other systematists.

Steinheil (1834) had established the genus Urginea,

which consisted of two groups of species, one from

the Mediterranean Region with four species and the

other from the Cape Winter-rainfall Region with

three species. The generic name Urginea was derived

from Ben Urgin, the name of an Arab tribe. Steinheil

illustrated the common species from Algiers, U.fugax

(Moris) Steinheil, which has been selected as the

lectotype.

The three Cape species placed by Steinheil under

Urginea were illustrated and named by Jacquin in

1794-1795 as Anthericum filifolium, A. exuviation and

A. fragrans. Their flowers resemble those of Antheri-

cum, such as A. bipeduncu/atum Jacq. [now Chloro-

phytum triflorum (Ait.) Kunth], which led Jacquin

to place them in this genus. Subsequent systematists,

however, aware of the fact that the underground

parts were bulbous, not rhizomatous, removed them

from Anthericum to Albuca (Ker-Gawler, 1805)

to Urginea (Steinheil, 1834) and to Ornithogalum

(Kunth, 1843). Eventually Salisbury (1866) placed

them in a new genus, Sypharissa, the name being a

reference to the striate, membranous, sheathing

cataphylls, which reminded him of the slough of a

snake. Baker in 1873 returned them to Urginea as

the tribe Sypharissa.

A close examination of the three species illustrated

by Jacquin, and one other recently described, all from

the Winter-rainfall Region, revealed a number of differ-

ences with Urginea Steinheil s.s. and Drimia Jacq. The

bulb is of the Tulipa- type, whereas that of Drimia

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

t Three species of Ornithogalum, namely O. anguinum

Leight. ex Oberm., O. monophyllum Bak. (section Urophyllon)

and O. zebrinum Bak. ex Oberm. (section Osmyne) appear to

possess the Tulipa- type of bulb as well. They, too, form trans-

versely striate cataphylls enveloping leaf and peduncle. The

genus Gethyllis in the Amaryllidaceae also possesses the

Tulipa-lype of bulb, with well-developed cataphylls.

and Urginea is of the Narcissus- type. The shoot,

consisting of synanthous deciduous leaves and a

raceme, is surrounded by long, sheathing cataphylls

which form very characteristic membranous apices

that are markedly striate with raised dark transverse

ridges. The leaves and racemes are deciduous while

the bases of the cataphylls become swollen to form

the bulb-scales. (In the Narcissus-type the leaf-bases

swell up to form the bulb-scales. )f Other characters

shared by these three species are the stout, erect

stamens with basifixed anthers surrounding the

ovary and the declinate, exserted style.

SYPHARISSA

Sypharissa Salisb., Gen. PL: 37 (1866). Type

species: S. exuviata (Jacq.) Salisb. ex Oberm. (lecto-

type).

Anthericum sensu Jacq. partly as to S. exuviata,

S. filifolia and S. fragrans.

Urginea Steinheil in Annls Sci. nat., ser. 2,2: 322

(1834), partly as to South African species.

Urginea, tribe Sypharrissa Bak. in J. Linn. Soc.,

Bot. 13: 216 (1873).

Drimia sensu Jessop in J1 S. Afr. Bot. 43:269

(1977).

Bulb of the Tulipa-type. The young cataphylls

prolonged above into cylindric membranous sheaths

enveloping the new shoot, their length depending

on depth of bulb; apical exposed part with raised,

transverse, purple or brown ribs, becoming worn

with age, lower part persisting, forming the swollen

white bulb scales; caudex often woody in old bulbs

and bearing long, stout roots in some species or

many thinner ones. Leaves synanthous, deciduous,

l-oo, long, terete. Raceme simple, few to many-

flowered; peduncle smooth, terete; lower bracts

boat-shaped, from a wide base, acuminate, with a

long, acute, basal spur, upper reduced, their spurs

rudimentary. Flowers diurnal, open for part of the

day, flowering for several days, sweetly scented.

Perianth stellate, the 6 tepals spreading, usually

white with a dark, reddish brown or green midrib.

Stamens 6, erect, surrounding ovary, filaments firm,

anthers basifixed, opening by longitudinal slits.

Ovary 3-locular, oblong-globose with many biseriate

axile ovules; style terete, exserted (at least when

receptive) declinate; stigma globose, papillate.

Capsule 3-locular, the valves with thickened margins;

seeds oval, flat, with a loose dark shiny testa.

A genus of 4 species found in the Cape Winter-

rainfall Region, east as far as Port Elizabeth and

west as far as southern South West Africa. Flowering

September-December.

112

THE GENUS SYPHARISSA (LILIACEAE)

Key to species

Leaves c. 4, coriaceous, linear, semi-circular or grooved, 3-4 mm in diam., glaucous-green; seed oval, c.

8 mm long; coarse plants with thick, long roots; raceme about as long as leaves or shorter. ...1.5. exuviata

Leaves many (1-few in juvenile plants), soft, 0,5-2 mm in diam., green; seeds 3-6 mm long; plants more

slender; roots various; raceme as long as leaves or exserted:

Leaves c. 1-1,5 mm in diam.:

Raceme many-flowered, elongated, exserted above the bunch of leaves; roots thick, long, woody;

seeds c. 6 mm long 2. S. fragrans

Raceme about as long as leaves; flowers usually close together, shortly racemose to subcorymbose

(in juvenile plants); seeds c. 4 mm long 3. 5. filifolia

Leaves c. 0,5 mm in diam., filiform, very many 4. 5. multifolia

Lig. 1. — SEM photographs of testa of Sypharissa seeds. 1, 5. exuviata: a, x700; b, X2000.

2, 5. fragans: a, X700; b, x2000 3, 5. filifolia : a, X700; b, X2 000.

A. AMELIA OBERMEYER

113

1. Sypharissa exuviata ( Jacq .) Salisb. ex Obenn.

comb, nov.* Iconotype: Cape, Jacq., Icon. PI. Rar.

2 (13), t. 415 (1794).

Anthericum exuviatum Jacq., Coll. Suppl. 89, t. 14, fig. 2

(1797), Icon. PI. Rar. l.c.: Willd., Sp. PI. 2: 136 (1799). Phalan-

gium exuviatum (Jacq.) Poir., Encyc. 5:243 (1804). Albuca

exuviata (Jacq.) Ker-Gawl. in Curtis’s bot. Mag. 22: t. 871

(1805). Urginea exuviata (Jacq.) Steinh. in Annls. Sci. nat., ser.

2,2: 330 (1834); Bak. in FI. Cap. 6: 466 (1897); Duthie in Ann.

Univ. Stell. 4, A, 2: 6 (1928); Adamson in Adamson & S liter,

FI. Cape Penins. 193 (1950). Ornithogalum exuviatum (J icq.)

Kunth, Enum. PI. 4: 369 (1843). Drimia exuviata (Jacq.)

Jessop in J1 S. Afr. Bot. 43: 276 (1977).

Coarse plants up to 1 m tall. Bulb ovoid with loose,

broad, fleshy scales, the inner extended above into

cross-barred, purplish, membranous sheaths up to

150 mm long; caudex elongated in older bulbs, hard,

bearing many thick, long roots. Leaves 1-5, suberect

or sprawling, semicircular becoming grooved when

dry, up to 1 m long, 3 mm broad, leathery, glaucous.

Raceme about as long as leaves or shorter, many-

flowered; peduncle erect, stout in fruit; bracts long-

caudate in lowest flowers, upper reduced; pedicels

c. 10 mm long, up to 15 mm in fruit. Flowers typical;

tepals up to 15 mm long and 5 mm broad. Capsule

oblong-ovoid, 15-25 mm long; seeds typical, c.

10 mm in diam. Fig. 1 .1.

Recorded from the Cape Winter-rainfall Region,

from Namaqualand to Port Elizabeth and Grahams-

town in sandy, peaty habitats or in humus-rich

rock crevices, often locally frequent; flowering in

September-October, but according to Adamson, only

after fires.

Cape. — 2917 (Springbok): Steinkopf (-BD), Meyer sub

Marloth 13320 (PRE). 3119 (Calvinia): Klipkoppie, Nieuwoudt-

ville (-AC), Barker 9763 (NBG). 3317 (Saldanha): Kreefte

baai (-BB), Rourke 587. 3318 (Cape Town): Malmesbury

Commonage (-BC), Lewis 3621: Stellenbosch Flats (-DD),

Duthie 1750', Camps Bay (-CD), Zeyher 4655 (S\M). 3325

(Port Elizabeth): Zwartkops River, Zeyher 105b (SAM).

3326 (Grahamstown): Bushman’s River Bridge (-CB), Archi-

bald 5314 (PRE). 3419 (Caledon): Rivierzondereinde (-AB),

Zeyher 4248 (SAM). 3421 (Riversdale): 16 km S. of Albertinia

(-BA), Acock s 22883.

2. Sypharissa fragrans {Jacq.) Salisb. ex Oberm.

comb. nov. Type: Jacq., Hort. Schoenbr. 1, t. 86.

Anthericum fragrans Jacq., Hort. Schoenbr. 1, t. 86 (1797).

Willd. Sp. PI. 2: 135. Phalangium fragrans (Jacq.) Poir., Encyc!.

5: 247 (1804). Albuca fugax Ker-Gawl. in Bot. Register 4:

t.3 1 1 (1818), nom. nov. for Anthericum fragrans Jacq. (not to be

confused with Urginea fugax Steinh.). Urginea fragrans (Jacq.)

Steinh. in Annls Sci. nat. ser. 2,2: 328 (1834); Bak. in FI. Cap. 6:

465 (1897). Ornithogalum fragrans (Jacq.) Kunth, Enum. PI. 4:

366 (1843).

Drimia exuviata sensu Jessop in J1 S. Afr. Bot. 43: 276 (1977).

Plants usually found in groups. Bulb forming a

long, woody caudex bearing long, thick, hard roots,

the outer scales remaining fairly loose and soft;

young inner tunics forming a short to long, white,

membranous sheath (its length depending on depth

of bulb) the upper part exserted, with an acute purple

fluted apex. Leaves c. 20, terete, erect or somewhat

flexuose, c. 300 mm long, soft. Raceme exserted

above bundle of leaves, c. 40-flowered, elongating

with age; peduncle firm, becoming woody in fruit.

Flowers typical; tepals c. 10 mm long. Capsule

narrowly oblong, 16 mm long, valves fairly thin;

seed oblong, 8 mm, typical. Figs 1.2 & 2.

Recorded from Malmesbury to Calvinia, in sandy

areas, locally abundant, in colonies; flowering

September to November.

* Since Salisbury did not validly publish the combination,

this is done here.

Fig. 2 — Sypharissa fragrans. Raceme, x2, Hanekom 2086.

Cape. — 3118 (Vanrhynsdorp): summit of Gifberg (-DA),

Hall 3906 ; Klawer (-DC), Godfrey VH-1254. 3119 (Calvinia):

Lokenburg (-CA), Acocks 18572. 3218 (Clanwilliam): Het

Kruis (-DA), Barker 2597. 3219 (Wuppertal): Citrusdal,

Thee Rivier (-CC). Hanekom 2086. 3318 (Cape Town): near

Hopefield (-AB), Letty 38. 3319 (Worcester): Saron (-AA),

Herre in STE 26768.

This is the tallest species in the genus; it is many-

flowered, attractive and strongly scented.

3. Sypharissa filifolia {Jacq.) Salisb. ex Oberm.

comb. nov. Iconotype: Cape, Jacq., Icon. PL Rar.

2(15): t. 414 (1794).

Anthericum filifolium Jacq., Coll. Suppl. 93 (1797), Icon.

PI. Rar. 2 (15): 18, t. 414 (1794): Willd., Sp. PI. 2: 135 (1799).

Phalangium filifolium (Jacq.) Poir., Encycl. 5:242 (1804).

Albuca filifolia (Jacq.) Ker-Gawl. in Bot. Register 7: 557 (1821).

Urginea filifolia (Jacq.) Steinh. in Annls Sci. nat. ser. 2,2: 329

(1834); Bak. in FI. Cap. 6: 466 (1897); Duthie in Ann. Univ.

Stell. 6A (2): 6 (1928); Adamson in Adamson & Salter, FI.

Cape Penins. 193 (1950). Ornithogalum filifolium (Jacq.) Kunth,

Enum. PI. 4: 369 (1843).

Anthericum spiratum Thunb., Prodr. 62 (1794). Syntypes:

“Cap. b. spei”, Thunberg s.n. (UPS — 8413, 8414, micro-

fiche!).

Urginea unifolia Duthie in Ann. Univ. Stell. 6A (2): 8-9

(1928); Adamson in Adamson & Salter, FI. Cape Penins.

193 (1950). Type: Cape, Stellenbosch Flats, Duthie in STE

1891 (STE, holo. !).

Urginea duthieae Adamson in J1 S. Afr. Bot. 8: 239 (1942)

sine descr. latine and in FI. Cape Penins. 194 (1950). Type:

Cape, Stellenbosch Flats, Duthie in STE 1790 (STE, holo.!; K),

nom. nov. for Urginea ecklonii sensu Duthie in Ann. Univ.

Stell. 6A (2): 6 ( 1 °>28) non Bak. Drimia duthieae (Adamson)

Jessop in J1 S. Afr. Bot. 43: 278 (1977).

Urginea flexuosa Adamson in J1 S. Afr. Bot. 8: 240-241

(1942), FI. Cape Penins. 193 (1950). Type: Cape Peninsula,

Smitswinkel Bay, Adamson 3099 (BOL, holo.!).

Drimia exuviata sensu Jessop in J1 S. Afr. Bot. 43 : 276

(1977), non Jacq.

Plants 100-300 mm tall. Bulb ovoid, c. 30-40 mm

in diam. compact, often enclosing 2-3 daughter

bulbs, or outer scales disintegrating and scales then

114

THE GENUS SYPHARISSA (LILIACEAE)

spreading (in juvenile plants the small bulbs consist

of c. 4-6 rounded succulent opposing scales); cata-

phylls membranous, soon disintegrating; roots many,

swollen during active growth. Leaves numerous,

slender, filiform, glabrous, straight or laxly flexuose,

up to 280 mm long and c. 1 mm in diam. (in juvenile

plants with 1-3 leaves). Inflorescence racemose, about

as long as leaves, 6-30-flowered (in juvenile plants

the few flowers often congested to subcorymbose);

peduncle terete; bracts of lowest flowers bearing a

long sharp spur which is reduced in upper. Perianth

with tepals 8-12 mm long, white or flushed with

purple and with a dark midrib. Stamens typical.

Ovary green, style white or purplish, declinate.

Capsule oblong-globose, coriaceous, c. 10 mm long,

brown; seeds numerous, ovate, 2-3 mm wide with a

loose wide membranous wing, shiny black. Fig. 1 .3.

Frequent on the Cape Peninsula to Vanrhynsdorp

and east to Bredasdorp, in sandy or gravelly soil,

locally common as scattered plants, flowering

September to December. Adamson records it as

flowering frequently after fires.

Cape. — 31 18, (Vanrhynsdorp): S. of Vredendal (-DA),

Acocks 19713, 3119 (Calvinia): Lokenburg (-CA), Acocks

17264. 3219 (Wuppertal): Citrusdal (-CA), Barker 7395.

3318 (Cape Town): Camps Bay (-CD), Zeyher 134; Stellen-

bosch Flats (-DD), Duthie in STE 652 ; Taylor 5147. 3319

(Worcester): Visgat-Agterwitzenberg (-AA), Emdon 254;

Tulbagh (-AA), Barker 9230. 3320 (Montagu): Poort N. of

Pienaarskloof (-AA), Acocks 23709. 3418 (Simonstown):

Bergvliet Farm (-AB), Purcell s.n. 3420 (Bredasdorp): 16 km

SE of Buffelsjagtsrivier (-BA), Acocks 24268.

The SEM photos of the testa of the seeds of

S. filifolia and Urginea unifolium Duthie revealed

that they belong to one species. This was confirmed

by a transverse section of the leaf. Duthie (1926)

in her informative study of the Urginea species of

the Stellenbosch Flats on Plate 4, figs 3, 5, 7 and 9,

drew sections of the leaves of U. exuviatum and

U. filifolium. It was seen that the transverse section of

the leaf of U. unifolium was similar to that of S.

filifolia. The testa of the seed of U. duthieae (SEM,

x 700, x 1 000) matched that of Sypharissa filifolia.

There is a problem about the dates of Anthericum

filifolium Jacq. and A. filifolium Thunb., Prodr. 62,

two different species which were both published in

1794. However, Roemer & Schultes (1829), and

later Kunth (1843), accepted that Jacquin’s name

was older and Thunberg’s plant was renamed

Anthericum nematodes Roem. & Schlt. (= Urginea

nematodes Bak.) and Ornithogalum thunbergii Kunth.

4. Sypharissa multifolia (Lewis) Oberm. comb. nov.

Type: W. Cape, 42 km S. of Springbok. Lewis 2302

(SAM, holo.!).

Urginea multifolia Lewis in Ann. S. Afr. Mus. 40: 9 (1952).

Drimia multifolia (Lewis) Jessop in J1 S. Afr. Bot 43 • 278

(1977).

Bulb globose, c. 450 mm in diam. with pallid

scales, often dividing and forming small clumps;

roots thick; shoot surrounded by 1 -several elongated,

membranous, fluted, sheathing scales. Leaves very

numerous (30-50), filiform, thin, up to 100 mm long,

0,5 mm broad, forming loose spirals. Raceme up

to 250 mm long, exserted, firm about 30-flowered;

lower bracts spurred; pedicels patent-erect, up to

8 mm long, thin. Flowers typical, with sweet scent

(resembling that of Lippia citriodora fide Lewis).

Capsule and seeds unknown.

Apparently rare in S.W. Cape and Namaqualand;

on dolerite ridges.

Cape. — 2917 (Springbok): 42 km S. of Springbok (-DD),

Barker 6310. 3017 (Hondeklip Bay): Kamieskroon (l’Aus),

Schlechter 11222. 3119 (Calvinia): Kareekom, 40 km N.W.

of Calvinia (-AB), Leistner 477 .

UNCERTAIN SPECIES

A collection from South West Africa, Giess 12857

from farm Kubub, LU 15, sandy flats (2616 CB),

is too incomplete for positive identification.

UITTREKSEL

Die genus Sypharissa Salisb. ( Liliaceae ) is weer

erken en die vier soorte wat daaronder val is wettig

gepubliseer.

REFERENCES

Adamson, R. S. & Salter, T. M., 1950. Flora of the Cape

Peninsula. Cape Town: Juta.

Baker, J. G., 1873. Revision of the genera and species of

Scilleae and Chlorogalae. J. Linn. Soc., Bot. 13: 209-292.

Jacquin, N. J., 1786-97. Collecteana. Vienna: Kraussiana.

Jessop, J. P., 1977. Studies in the Bulbous Liliaceae in South

Africa: 7. J1 S. Afr. Bot. 43:265-319.

Ker-Gawler, J. B., 1805. Albuca exuviata. Curtis's bot.

Mag. 22: t. 871.

Ker-Gawler, J. B., 1821 , Albuca filifolia. Bot. Register 7: t. 557.

Kunth, C. S., 1843. Enum. PI. 4. Stuttgart & Tubingen:

Cottae.

Poiret, J. L. M., 1804. Phalangium. In Lamarck & Poiret,

Encyclopedic methodique botanique 5. Paris.

Salisbury, R. A., 1866. The genera of plants. London: Van

Voorst.

Stearn, W. T., 1978. Mediterranean and Indian species of

Drimia (Liliaceae): a nomenclatural survey with special

reference to the medicinal squill, D. maritima (syn. Urginea

maritima). Ann. Musei Goulandris 4: 199-210.

Steinheil, A., 1834. Urginea. Annls Sci. nat. Bot. Ser. 2,

321-330.

Thunberg, C. P., 1794-1800. Prodromus plantarum Capensium.

Uppsala: J. F. Edman.

Bothalia 13,1 & 2: 115-125 (1980)

Some observations on two early Cape florilegia

£. G. H. OLIVER*

ABSTRACT

A number of early Cape florilegia and codices exist in libraries in Europe and South Africa. Four of

these florilegia are closely related and are housed in the Brenthurst Library, Johannesburg, the Botanical

Research Institute, Pretoria, the Bodleian Library, Oxford and the Rijksherbarium, Leiden. The first two

are compared and discussed in detail in this article. Arising from this comparison, a new interpretation of

the interrelationship and origins of the four florilegia is proposed. The key volume is the florilegium in the

Botanical Research Institute, Pretoria.

RtSUMl

CERTAINES OBSERVATIONS SUR DEUX ANCIENS FLORILEGIA DU CAP

Un certain nombre d'anciens florilegia et codices du Cap se trouvent dans les bibliotheques d' Europe et d'Afrique

du Sud. Quatre de ces florilegia sont etroitement apparentes et se trouvent a la bibliotheque Brenthurst de Johannes-

burg, a rinstitut de Recherche Botanique de Pretoria, a la Bodleian Library d'Oxford et au Rijksherbarium de

Leiden. Les deux premiers sont compares et discutes en details dans cet article. Emergeant de cette comparaison,

une nouvelle interpretation de l' inter-relation et des origines des quatre florilegia est proposee. Le volume clef

est le florilegium de Vlnstitut de Recherche Botanique de Pretoria.

INTRODUCTION

The Cape flora has for three centuries excited the

interest of botanists. With the remarkable increase

in the exploration of the world from the fifteenth

century, new areas with fascinating plants and animals

became known to the educated and knowledge-

hungry world of Europe. The Cape of Good Hope

became a vital stopping-over point for the many

ships going to the East in search of spices and riches.

It was thus inevitable that plants growing at the

Cape found their way back to Europe. Soon the

demand for these unusual plants, both to grow and

to possess as dried specimens or paintings, increased

considerably.

During the governorship of Simon van der Stel

from 1679 to 1699, there began a period of

considerable exploration and expansion. This was

accompanied by a significant increase in the scientific

knowledge of the indigenous flora and fauna with

Van der Stel himself as the worthy patron. It is

known that he commissioned the artist Claudius to

record by means of sketches the natural history of

the area and also the gardeners, Oldenland and later

Hartog, to build up “one of the most beautiful and

curious gardens I have ever seen” (Tachard, 1686).

Illustrated books on the Cape flora were not

available at the time and so collections of paintings,

florilegia or codices of animals were also included,

were produced for influential patrons of natural

history. Among these florilegia or codices were the

Codex Witsenii, Codex Bentingiana, Codex Comp-

toniana, Dolneus’s Florilegium, Van der Stel’s own

Collection and the Codex accompanying the official

report of his Expedition to Namaqualand in 1685/6.

Gunn & Codd (1980) give a fine overview of early

Cape botanical history in which they discuss these

works and their significance.

Examples of early Cape florilegia and codices exist

today in various institutes and libraries in Europe

and South Africa, namely:

Botanical Research Insti- Brenthurst Library, Johan-

tute, Pretoria (BRI) nesburg (BFC)

Rijksherbarium, Leiden Bodleian Library, Oxford

(LD) (OXF)

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001

South African Museum,

Cape Town (SAM)

Africana Museum, Johan-

nesburg (IPA)

British Museum (Nat. Hist.),

London (BM)

South African Public Lib-

rary, Cape Town (SAPL)

Trinity College, Dublin

(TCD)

The works mentioned in this paper will be referred

to by the abbreviations given above for the institutes

and libraries where they are housed. Much has been

published about these works by Waterhouse (1932

Barnard (1947), Smith (1952), Jessop (1965), Edwards

(1965), Kennedy (1967), Macnae & Davidson

(1969), Gunn & Du Plessis (1978), Waterhouse (1979)

and Gunn & Codd (1980).

The florilegia in the first four institutes mentioned

above are of particular interest to me because

of their close relationship, one of them being in

the library of the Botanical Research Institute,

Pretoria. Jessop (1965) published a detailed account

of this florilegium, but unfortunately at the time

did not know of the existence of the other three

florilegia. Gunn & Du Plessis (1978) edited and wrote

the introduction to the so-called Flora Capensis of

Jakob and Johann Philipp Breyne, housed in the

Brenthurst Library and which was reproduced in toto

by the Brenthurst Press, Johannesburg. They were

unable to give a detailed comparison of the four

florilegia. Resulting from my close examination of

BFC and BRI while writing a review of the

BFC reproduction for Bothalia (Oliver, 1980),

I have been able to extend the published notes and

observations on the two florilegia in South Africa.

I have not had the opportunity to examine the

florilegia in Oxford and Leiden and accept the

statements made by Gunn & Du Plessis (1978),

who have consulted them. In the discussions which

follow, reference is made to the numbering of the

plates (as arranged by Gunn & Du Plessis) in the

Brenthurst reproduction of the Breynes’s Flora

Capensis

BINDING

The most noticeable difference between BFC and

BRI is in the size of the volumes, where the page

size in BFC is 310x195 mm and in BRI 393 x250 mm.

This difference in size is significant as will be seen

when the origins of the two florilegia are discussed

later on. The Brenthurst reproduction of BFC has

116

SOME OBSERVATIONS ON TWO EARLY CAPE FLORILEGIA

been reduced to an even smaller size, the title page

being 230 mm long instead of 270 mm (Gunn &

Du Plessis, 1978). Reference to the list of plates at

the beginning of BFC shows that 47 of the

102 paintings had to be reduced.

The binding of BFC is full red morocco leather

with gold tooling including the Breynes’s coat-of-arms.

The pages have been gilded. On the title page it is

stated that the volume was bound by the younger

Breyne in 1724. BRI on the other hand is plainly

bound in vellum, much like the codices SAM and

SAPL; the tooling is blind and the edges of the pages

have been spattered with red and blue ink. On the

front cover there is a distinct erect capital P done in

slightly faded black ink. Barnard (1947) mentions

a capital Q on the front cover of SAM. I examined

SAM and find that the Q slants obliquely to the

right, is smaller and is placed at the top of the

front cover.

Gunn & Du Plessis mention the binding of the

pages of BFC, but give no details about the

gatherings. It was not possible for me to ascertain

the exact binding sequence as the original book had

recently been restored. They do, however, mention

that forty-nine leaves have no fold-marks and forty-

nine have fold-marks which indicate that these

folded leaves were painted on at the Cape, folded and

then despatched to Europe. They also mention the

extra pieces of paper which had been glued into

these folded leaves, when binding took place, in

order to bring all the pages up to the same size.

Some of the paintings on the larger sheets had in

turn been cut during binding. This is most noticeable

on the coloured frontispiece and plates BFC 30

and 42.

BRI is bound into gatherings of 5 or 6 sheets with

one gathering of 7 sheets. These sheets are folded

giving gatherings of 10, 11, 12 or 13 pages with three

instances of single pages having been tipped in

during binding. Jessop (1965) points out that the

same cutting of paintings occurred in BRI and feels

that the paintings must have been executed before

the book was bound. The cut paintings occur only

on BRI 8, 68, 86 and 109 to any extent, but far less

than in BFC. One’s interpretation of the term

“binding” is, to my mind, important. I feel that the

book was made up with a softish cover, as is found in

SAPL, to form a working volume into which the

paintings could be executed. At a later stage the

volume was slightly trimmed when the soft cover

was replaced by a proper vellum binding for placing

in some person’s library. The 68 blank pages at the

end of BRI give a very strong indication that the

water-colours were painted into the book. Another

point which supports this view is the statement by

Tachard (1686) that Claudius “had already completed

two thick volumes of divers plants, painted from

nature”.

The other florilegia in South Africa are rather

similar to BRI in appearance. IPA is most like BRI

being similar in size and vellum binding. The paper

is of the same thick quality and the paintings are

done on the recto of each page with binding in

gatherings. In SAM and SAPL the volumes are much

smaller, like BFC. In these two florilegia the binding

of the sheets is interesting. The leaves are folded

double and bound singly giving a lot of 2-paged

gatherings. The paintings occur on the recto of the

first page with the descriptions on the verso in SAM

and the recto of the second page in SAPL.

Both Gunn & Du Plessis and Jessop give details

of the watermarks found in the respective florilegia

and discuss the possible origin and date of manu-

facture of the paper. In BFC there is a number of

different types of paper. Gunn & Du Plessis did not,

however, mention the connection between water-

marks and countermarks presumably due to the

difficulty of unravelling the binding sequence of so

many varied sheets. They do mention the similarity

of the watermarks in BFC, SAM and TCD. BRI,

on the other hand, has very uniform paper of only

two types. Jessop gives an illustration of the main

watermark, the Strassburg lily with coronet, mantling

and 4WR and the main countermark, IHS/DYSVLI

(cf. left-hand pair, Fig. 1 in the present article). He

noted that a few pages possessed a different counter-

mark, IHS/RM. This is, in fact, the countermark

for a slightly different watermark which is more

crudely produced and of smaller size (cf. right-hand

pair, Fig. 1). The countermark itself is very much

cruder, as well, and this suggests either an

inexperienced papermaker or paper of earlier origin

before the refinement of the watermarking technique.

This cruder watermark occurs in paper that takes up

two bound gatherings of 22 pages between paintings

111 and 132.

The main watermark occurs a number of times in

IPA, but with different countermarks. It also occurs

in the Claudius animal paintings in the Africana

Museum in Johannesburg, as figured by Smith (1952).

It can also be seen in the binding paper of Commelin’s

Horti Medici Amstelodamensis (1697-1701 ) and in the

body of the book where different mantling and the

countermark of Pieter van der Ley occur. Both the

main watermark and countermark of BRI occur

once in BFC.

Churchill (1935) states that the Strassburg Lily

with 4WR was first used in 1636 and is of German

origin. Jaffe (1930) and Heawood (1950) say it was

made at the papermill of one Wendelin Richel

(Riehel), which began production in 1583 without

countermarks. This watermark became associated

with quality paper and was much copied in Europe.

Jaffe also states that the countermark, IHS sur-

mounted by a cross, “/« Hoc Signo", was first used in

the Lombardy area in Italy from 1481-1580 before

being taken up in the Lothringen area of Germany.

The version with DYSVLI was one of the many

countermarks used by J. Villedary (Vildary) whose

mills produced paper over a period of 150 years

from 1658-1812. The countermark with RM is

unknown, but noted by Heawood (1950).

It is not possible, using currently available

references, to put any date to the manufacture of

the paper in BRI. Jessop (1965) feels that the paper

was made in about 1700. Gunn & Du Plessis (1978),

however, were able to use the BFC watermarks and

countermarks more usefully and they dated most of

the BFC paper to the last half of the 1600’s. This led

them to the assumption that BFC was the oldest

volume in the series of four similar florilegia. A more

exact date can be deduced by reference to certain

statements made by J. P. Breyne and Seyler in the

Prodromi of 1738. Jakob Breyne was said to have

left his collection of paintings, later bound into the

volume BFC, to his son on his death in 1697.

Therefore, if no additions were made by the son,

one can deduce that the paintings must have been

executed before 1697. Jakob Breyne had in 1678

produced his Centuria containing among the 100

plants, illustrations of 48 Cape species, including

E. G. H. OLIVER

117

Fig. 1. — Watermarks and countermarks in the paper in the BRI forilegium, x0,5.

Erica cerinthoides, but did not use any BFC paintings

for these plates. Thus it can be assumed that he did

not possess any BFC paintings at the time. This

means that he must have received the paintings

from the Cape between 1678 and his death early

in 1697.

THE PAINTINGS

The BFC volume contains water-colour paintings

of 102 species of which 66 occur in BRI and 36 are

exclusive to BFC. Included in the last set are two

paintings which are known to have occurred in

BRI, but are now missing (see under Numbering).

Gunn & Du Plessis (1978) recognized two different

qualities of art-work and grouped the paintings into

Group A of superior quality and Group B of poorer

quality. They ascribe 37 paintings to Group A and

65 to Group B. I would, however, have placed the

two non Cape paintings, BFC 60 and 69, in Group A.

I would also regard BFC 18, Spiloxene alba, as being

a Group A painting and BFC 42, Chasmanthe

aethiopica, as Group B. This leaves all the 36 Cape

species exclusive to BFC belonging to the Group B

painting quality. I have followed the same arrange-

ment of groups for convenience in this comparison.

The BRI volume contains water-colour paintings

of 148 species of which there are the 66 shared with

BFC and 82 exclusive to BRI. In my opinion the

148 paintings can be divided into 115 being Group A

and 27 being Group B with 6 paintings being

impossible to place with certainty. All of the 82

paintings exclusive to BRI are of Group A quality.

The Group A paintings, however, can be subdivided

into a possible two or even three different styles.

The majority matches the Group A style paintings in

BFC. At the beginning of BRI there is a series of

paintings, namely 3, 5, 6, 7, 8, 9, 10, 17, 18, 19 and

20, which are in a completely different style, rather

vague and lacking precise details and yet not crude

like some of the Group B paintings. These paintings

are undoubtedly the work of a different artist and

strike me as being originals. Their quality and colours

remind me of the work of the “second artist” in IPA.

Most of the Group A paintings are accurate and

beautifully executed. Towards the end of the series

of paintings the quality of the paint seems to change,

the greens being much deeper in colour with a bluer

tinge and the paint is much thicker. These paintings

could also be the work of another artist. One painting,

BRI 135, stands out as completely different from the

rest of the numbered plates. It is of Crassula coccinea

and is classified as a Group B painting, but gives the

strong impression of having been executed in the

Claudius style like the paintings occurring in SAM

and SAPL. The paint is of quite a different texture

and the painting lacks any perspective. The extra

painting of Erica cerinthoides, an unnumbered one

occurring with BRI 132, is mentioned by Jessop as

also unlike any others in style and quality of paint.

This paint is very reminiscent of the powdery type

used in the Group A paintings in BFC as it is

smudging slightly and has left an imprint on the

recto of the preceding page. This feature also occurs

on BRI 99 in the browns of the rather large tuber

of the Bulbine tuberosa.

The differences between the Group A and Group B

paintings in BRI are very marked, particularly where

the two styles occur on the same page, e.g. BRI 55,

Lobelia pinifolia (A) and 56, Drosera cistiflora (B);

BRI 71, Pelargonium longifolium (A) and 72 Moraea

tricuspidata (B); BRI 75, Mesembryanthemaceae (A)

and 76, Galaxia ovata (B) and 77, Dorotheanthus

bellidiformis (B). In these examples it is always the

Group A painting which is on the left-hand side of

the page and numbered first. This would indicate that

the small Group A painting was done first leaving

space on the page for additional species to be

illustrated at a later stage which, in the above cases,

turned out to be rather crude Group B paintings.

This feature of the two distinct qualities of paintings

on the same page is important in respect to the

problem of the origin of these florilegia.

Gunn & Du Plessis (1978) are of the opinion that

the paintings of BFC are the originals in the set of

four very similar florilegia. They based their opinion

on the dating of the paper — “The paper used in

the Breynes’s ‘Flora Capensis’ is the earliest and of a

period consistent with the possibility that it is the

original set”. This might be true of the paper, but

not of the paintings, as the following points obtained

from a detailed comparison of BFC and BRI will

show. Gunn & Du Plessis (1978) state that the

volumes in Oxford and Leiden are obviously inferior

copies, which opinion I must accept not having had

the opportunity to examine them myself.

The most noticeable difference between the BFC

and BRI paintings in the Group A series is the

shortened dimensions of many of the paintings in

BFC. The floral parts of the paintings are painted

the same size as those in BRI, but the vegetative parts,

particularly the stems, have been reduced to be

118

SOME OBSERVATIONS ON TWO EARLY CAPE FLORILEGLA

able to fit onto the smaller size of the BFC paper, in

some cases so drastically as to make the painting

disproportionate. Good examples of this feature

may be seen when comparisons are made of BFC 19

and BR1 125, Spiloxene capensis; BFC 20 and BRI

126, Moraea aristata\ BFC 40 and BRI 109, Gladiolus

maculatus and BFC 38 and BRI 68, Gladiolus carneus.

The last example is illustrated in Fig. 2 where a

glance will show the more natural proportions of the

BRI painting. This feature would indicate that the

BRI paintings could not have been copies from

BFC and that the reverse is the case.

Fig. 2 also illustrates another important feature,

namely the lack in BFC of certain details present in

BRI paintings. In the BRI paintings of Gladiolus

carneus there are an additional flower bud, terminal

bracts and a leaf. There is also another very important

and significant feature about this BFC painting,

and that is the lack of paint in the region of the

ovary of the basal flower, where the artist forgot to

fill in the colour. This also occurs in BFC 39, BRI

122, Gladiolus hyalinus. Further examples of parts

of the plants being left out in the BFC paintings can

be found in paintings which lack roots, hairs, flowers,

Fig. 2. — Gladiolus carneus Delaroche. Left, painting number 68 from the BRI florilegium, x0,5; right, painting number 38 from the

BFC florilegium (Brenthurst reproduction), x0,75.

E. G. H. OLIVER

119

corms or branches. An additional very marked

example of this feature is shown in Fig. 3, Polygala

bracteolata.

There are, however, a few examples where this

loss of details in BFC is reversed and one finds BFC

p/' tings with more parts. An example is shown in

.g. 3, Empodium plicatum, where the lateral flower

has many more tepals than the species should have,

but this is more the result of an inaccurate copier.

In some of the species of Moraea in Group B, the

BFC paintings have the old leaf bases included.

In BFC 70, BRI 89, Leonotis leonurus, additional

flowers and another inflorescence are depicted. All of

these last examples are paintings of Group B and

these I regard as copies in both florilegia, that is,

copies from another set of paintings which may or

may not have been the originals.

The act of copying is always fraught with the

possibility of slips, misinterpretation or plain re-

interpretation by the copier. This is evident in the

fig. 3 .-Polygate bracteoh.a L. Left, painting number 63 from the BRI itorijwum; centre JjW

florilegium (Brenthurst reproduction), both x0,5. Empodium plicatum (L. .) /R' ’tPrenro^uct;on) \ 0 70

florilegium, x0,5; centre bottom, painting number 62 from the BFC florilegium (Brenthurst reproduction), XU, /u.

120

SOME OBSERVATIONS ON TWO EARLY CAPE FLORILEGIA

differences that occur between some of the BRI and

BFC paintings. In copying the artist has produced

extra curliness or waviness into the organs that he

was copying. This is clearly seen in the leaves of

Empodium plicatum shown in Fig. 3 and in the very

stylized leaves of Polygala bracteolata in the same

figure. Further examples of this type can be found

in BFC.

As I am interested in ericas, I was particularly

struck by the differences between the two similar

renderings of Erica cerinthoides. In BRI 43 the leaves

are shown in distinct whorls and are themselves

depicted trigonous whereas in BFC 90 the leaves are

randomly scattered and executed rather poorly and

haphazardly by the single stroke of a brush. One

can also see that the copier in BFC could not easily

interpret the BRI flowers which must have been

somewhat passe when painted.

Another interesting and important feature, which

was also noted by Jessop (1965), is that there are

several paintings in BRI which have faint pencil

outlines still remaining on the pages. These are of

additional parts of the plants which the artist must

have decided not to use. Two clear examples are

BRI 12, Adenandra villosa, and BRI 109, Gladiolus

maculatus. This feature gives a strong indication that

the paintings are originals and not copies.

The strangest anomaly found in the comparison

of the paintings occurs in the paintings of Sparaxis

bulbifera, BRI 102, BFC 32, both Group B paintings.

The paintings have their lower halves reversed.

Reversing of paintings is of course commonly

encountered in engravings used for printing plates.

From the above comparison of the paintings it is

my opinion that the Group A paintings in BRI are

the originals and that the BFC paintings were all

copied from BRI. An examination of the paints used

lends additional weight to this view. In BRI the

water-colours are of very good quality with most of

the colours still unchanged. In BFC the paintings

were executed with a very powdery paint which, as

Gunn & Du Plessis (1978) noted, has changed

colour in a number of cases, but due to its powdery

nature, it is being smudged and rubbed off with

time. As a result the coarseness of the paper is

accentuated and clearly seen in the Brenthurst

reproduction particularly on BFC 19, Spiloxene

capensis (BRI 125).

The quality of the BRI Group A paintings, most

of which are petaloid monocotyledons, is outstanding.

The perspective in the flowers is extremely good and

is far superior to any that I have seen in the other

early florilegia. Some of the bulbs and corms have

been painted in very fine detail, for example, BRI

125, Spiloxene capensis ; BRI 132, Homoglossum

watsonium and the leaf-base in BRI 49, Urginea

duthieae. But many bulbs, corms and bases of plants

have been done in much less detail. This seems, in

my opinion, to indicate that the leaves and flowers

were painted in the field with only pencil sketches of

the vegetative parts followed by a completion of the

colour work at a later stage, perhaps at camp in the

evening or even at home.

Several persons such as Petiver, Witsen, Burman

and the Breynes stated that their paintings were

executed from live plants at the Cape. The demand

for paintings in the late 1600’s and early 1700’s

must have been due to the lack of colour

reproductions in books and it is certain that copies of

originals were made to satisfy this demand. Here the

enigma of the Codex Witsenii and Claudius paintings

is the prime example. Copying, whether at the Cape

or back in Europe, was common. A glance through

IPA or the reproduction of the plates by Kennedy

(1967) will show that the painting of a legume occurs

twice with only a few pages separating them. The

folded paper of the Group B paintings in BFC points

to the copying having been done at the Cape.

It was suggested by Gunn & Du Plessis (1978) and

by Jessop (1965) that some of the paintings could

have been executed from plants cultivated in Europe.

As many of the plants in BFC and BRI are geophytes

they could easily have been grown in Europe

at the time. However, one or two features point

to a wild origin for the plants, certainly of the

Group A species. These include the damage to the

leaves caused by insects and other animals, and so

accurately portrayed by the artist. The best example

of this feature is BRI 42, BFC 35 of Babiana tubiflora

which has its leaves almost completely chewed off by

some grazing animal. As Gunn & Du Plessis (1978)

point out “This is a very clear indication that the

painting was made from a plant which grew wild at

the Cape and not from a cultivated plant”. Further

examples of this type may be found in BRI 45,

Wachendorfia paniculata; BRI 52, Ornithogalum

thyrsoides and BRI 92, Ixia paniculata. Also, the

natural dimensions of the vegetative parts of the

plants suggest that the subject was a wild plant rather

than one grown under glasshouse conditions in

Europe.

In this discussion of the paintings in BFC and BRI

it is worth mentioning that in BRI there are various

pieces of plant debris lodged in between the pages.

In the fold with BRI 131, on which the unnumbered

painting of Erica cerinthoides occurs, there are three

leaves of Erica cerinthoides. These leaves could

either have become lodged in the fold when the

species was being painted or at a later stage when the

owner of the volume was perhaps comparing a

specimen with the painting.

NUMBERING

The paintings in all of the four similar florilegia

are numbered, according to Gunn & Du Plessis

(1978). These numbers provide an important feature

for comparison. The important volume is BRI.

In this volume nearly all the paintings are numbered

near the base of each plant in a consecutive sequence

up to 142. Strangely the first painting is not actually

numbered. Number 142 is followed by 10 un-

numbered paintings and then the last two in the set

numbered as 143 and 155. The sequence of sheets in

the binding gatherings has not been interrupted.

Four pages have, however, been cut out of BRI

leaving gaps in the numbers, viz. BRI 13, 14 and 15;

27; 50; 90 and 91. As Gunn & Du Plessis (1978)

state, comparison with BFC gives the identity of

two of these missing paintings, because copies of

BRI 27 and 90 occur in BFC. An additional three

pages have been cut out of BRI near the end of the

paintings, but, as no numbers are missing in the

sequence, one may assume that these pages were

removed before the numbering was done as stated

by Jessop (1965). These pages may or may not have

had paintings. Despite the anomalies in the numbering

towards the end of the paintings htere should have been

a total of 155 paintings, but with the loss of seven

there now remains the total of 148

E. G. H. OLIVER

121

The first set of numbers up to 98 is written in a

very neat small writing with black ink while the

numbers 99-143 are written in a different slightly

larger writing and paler ink. Number 155 is written

in another handwriting and number 57 in yet another.

In several cases already mentioned, two completely

different styles of painting occur on the same page

and also on the same sheet of paper but separated due

to the binding. These are all numbered in the sequence.

Why there are two different main handwritings in the

numbering and who did them cannot be answered.

The numbers might have been done by the artists.

Why there is the batch of 10 unnumbered paintings

near the end is also a mystery. The fact that the

paintings were numbered consecutively regardless

of the painting styles shows that the numbering must

have been done directly into the volume and must

therefore be a series exclusive to BRI.

BFC has 85 pages numbered consecutively from

2-86. These numbers have been written all in the

same handwriting in the top right-hand corner of the

recto of each page. These numbers refer to the pages

and not the paintings as on 14 pages there are two,

three or four paintings. Gunn & Du Plessis (1978)

give references to these ‘Folio’ numbers. In BFC

there are 35 paintings bearing numbers near the base

of the plant depicted and these numbers are the same

as those on the BRI paintings of the same species.

All the paintings bearing these numbers in BFC

belong to the superior Group A, while all the Group B

paintings in BFC are unnumbered (cf. Fig. 2).

Gunn & Du Plessis (1978) miss the real significance

of this very important point which gives additional

proof that the BFC Group A paintings must have been

copied from BRI. None of the 31 paintings in BFC

belonging to Group B and shared with BRI is

numbered. This suggests that the BFC Group B

paintings were copied from some originals in another

volume without any numbering and that the Group A

paintings were later copied from BRI together with

the noting of the BRI numbers.

Two paintings with numbers stand out as unusual.

BRI 63, Polygala bracteolata, a Group A painting,

occurs in BFC as a Group B painting (cf. Fig. 3).

If the copying of the Group A paintings had been so

good why had this species been copied so poorly.

The other unusual painting is BFC 13, Hessea

cinnamomea. Group A. In BFC it bears the number

distinctly written as 105, but BRI 105 is of a Group

B leguminous species. H. cinnamomea in BRI is

108, but the 8 is rather indistinctly written and at

a glance could well be mistaken for a 5 which is

most probably what the copier did.

Gunn & Du Plessis (1978) state that the two

volumes in this series of four similar florilegia, OXF

and LD, both contain paintings they regard as

inferiorly executed copies. They state that “The

numbers at the base of all the paintings . . . are in

the first instance related to this volume ‘BFC’ where

the numbers appear chronologically and related to

page numbers”. The OXF volume is vellum-bound

and has the paintings pasted onto the page and

signed “A.B. del”. They mention that one of the

younger Breyne’s daughters had signed her own

paintings “A.B.”. However, Edwards (1964) states

that this volume forms SHERARD MS. 188 and is a

collection of paintings no doubt drawn by Anthonie

van Breda, of plants growing in gardens in Holland.

Boerhaave bought the volume for Sherard from

Levinius Vincent who through marriage acquired van

Breda’s famous museum. This relationship of the

volume with van Breda and its very strong similarity

with BRI and BFC is very confusing and needs

further investigation.

The LD “volume” is in fact a set of loose paintings

of inferior quality which were acquired by D. van

Rooyen at two auctions in 1778 and 1779. Gunn &

Du Plessis (1978) suggest a possible link between

some of these paintings and the collections of Seba.

There occurs in the BRI volume with the numbering

on only 17 paintings in the first 96, a plus sign.

In some cases this plus sign looks as though it was

done by the numberer, in others by Burman (cf.

Fig. 4). All of the 17 paintings belong to Group A.

The significance of this sign is as yet not understood.

COMMON NAMES

BFC and BRI have a number of paintings accom-

panied by a common name in Dutch (cf. Fig. 3)

or occasionally a Latin polynomial. They are all

written in the same, but very different, handwriting

in each volume. These common names were given

to nearly all of the Group B paintings but a few

were given in BRI to Group A paintings by the

same person (cf. Fig. 4, left). In BRI, 27 paintings

have common names, 23 Group B, 4 Group A with

the remainder of Group A without a common name.

BFC has the same 27 paintings with common names

except that BFC 8, Ornithogalum thrysoides, is a

quite different painting of Group B as opposed to

BRI 52 which is one of the 4 Group A paintings.

BFC has 28 extra paintings of Group B with common

names. One anomally is the occurrence of 7 shared

paintings of Group B with common names. In

BRI 43, Erica cerinthoides, there is an additional set

of words written in the same handwriting “ Erica

coris folio hispido cerinthoides africana Breynia",

which is the Latin polynomial given to the species

by Jakob Breyne in his Centuria prima of 1678.

The handwriting in BRI is very distinctive and

could well be the same as that used to write the notes

accompaying the animal paintings in the Codex

Witsenii, SAM; cf. FOL. 160 reproduced by Barnard

(1947). This handwriting is not the same as that

occurring on the botanical paintings which, according

to Smith (1952) quoting a former Chief Archivist,

Graham Botha, is identical with that in IPA. The

handwriting of the notes in SAPL is also of a similar

style. This could suggest that the artist wrote the

names in BRI and the notes in the SAM animals,

which are regarded as having been copied at the

Cape for Nicolaas Witsen some time before 1692

(Barnard 1947).

A number of the common names differs slightly

between BFC and BRI, most being differences in

spelling which could be attributed to the home

language of the copier, e.g. middags/middaghs;

bloem/blom; Ringel/Rengel; — aanse/ — aense; sterre/

starre. Some peculiar differences are noted here:

BFC 90: Peloaan Bloem

BFC 78: Heutel

BFC 17: Hyacins tuberosa

peruanus

BFC 70: Piramus infralia

BRI 43: Pelicaen blom

BRI 62: Huetel

BRI 85: Hyasinta tuber-

osus peruanus

BRI 89: Piramus in jtalia

These anomalies would indicate copying from the

same original, but in BFC by a rather poorer copier

of words.

Jessop (1965) pointed out that the common names

in BRI must have been added to the bound volume,

as many of them have been blotted on the verso of the

preceding page.

122

SOME OBSERVATIONS ON TWO EARLY CAPE FLORILEGIA

FIG. A.-Wachendorfia paniculate Burm. Top left, painting number 45 from the BRI Aorilegium xO 36; top right, TaN IX from

Breynes’s Prodromi (1739), x0,5. Monsonia speciosa L. Bottom left, Tab. XXI from Breynes s Prodrom. (1739), XU,5, Dottom

right, painting number 22 from the BRI florilegium, x0,36.

E. G. H. OLIVER

123

ANNOTATIONS

The BRI volume was at one time in the possession

of Johannes Burman as it was inscribed by him on

3 August 1755 [cf. fig. 1 of Jessop (1965)]. Burman

also annotated every painting in BRI. Most of his

annotations were probably done about the same time

as the same dark ink and style of writing was used

by him. This was done in the volume as is evidenced

by the blotting of the ink on the verso of the

preceding pages. He quotes many times Breynes’s

Prodromi of 1739 and his own Rariorum Africanarum

Plantarum of 1738/39 using Latin polynomials.

However, on BRI 112, Antholyza ringens , he cites

Linnaeus’s Species Plantarum of 1753 and quotes his

description. Binomials were added by Burman using

a finer pen. Additional annotations were also added

in a larger clumsier handwriting using paler ink.

PUBLISHED WORKS

Paintings from both BFC and BRI are known

to have been used in published works, namely,

Breynes’s Prodromi (1739) and Burman’s Rariorum

Africanarum Plantarum (1738/39). Mention has

already been made of these by Jessop (1965) and by

Gunn & Du Plessis (1978). However some extra

observations not noted by them throw a different

light on the relationships of the florilegia and the

above publications.

The younger Breyne undoubtedly used the BFC

paintings as the originals from which 15 of the

engravings in the Prodromi of 1739 were copied.

These paintings occur in both BFC and BRI but a

very careful comparison of the water-colours and

the engravings shows that the engravings were made

from BFC. All of these originals fall into the Group

A paintings. Seven engravings can be linked to

water-colours which occur only in BFC and these are

all Group B paintings.

Of significance are three engravings which can be

linked to water-colours of Group A found only in

BRI. Tab. VII fig. 2 of Gladiolus carneus is taken

from BRI 47. Gunn & Du Plessis (1978) compared

this engraving with another quite different water-

colour of the same species which occurs in both

florilegia and is illustrated in the present article in

Fig. 1. Tab. IX, fig. 1 in Breynes’s Prodromi is of

Wachendorfia paniculata which is taken from BRI 45

and is illustrated in the present article in Fig. 4.

Of particular interest is the third example, Monsonia

speciosa on Tab. XXI, fig. 2, which is taken from

BRI 22 (cf. Fig. 4 in the present article). On the same

Tab., fig. 1 is of Senecio cymbalarifolius which is

represented in both florilegia, Group A paintings,

and in the text is cited as being “ ex Flora nostra

Capensis". But in the text for Monsonia the

Breynes state “in Flora nostra Capensi ”. This water-

colour does not exist in BFC. From their statement

it would appear that the Breynes had had access to

a copy of this species, which had not been bound

into BFC, or even access to BRI. If the latter were

the case why then had the Breynes not reproduced

more of the superb water-colours.

J. Breyne had collected together the paintings for

BFC before his death in 1697. His son, J. P. Breyne,

when mentioning the paintings in the Prodromi did

not give exact details of their origin other than that

they came from the Cape, e.g. Tab. XII, fig. 1 —

“ Huius iconem accepit Parens ex Capite bonae spei

vivis coloribus pictam ”.

Flo. 5. -Pelargonium cucullatum (L.) L’Herit. Centre, Tab. XXXV, Fig. 3, from Burman’s Rariores Africanarum Plantarum (1738),

x0,5; left, painting number 103 from the BRI florilegium, x0,25; right, painting number 54 from the BFC flonlegium

(Brenthurst reproduction), x0,30.

124

SOME OBSERVATIONS ON TWO EARLY CAPE FLORILEGIA

Burman’s Rariorum Africanarum Plantarum con-

tains descriptions and engravings of some Cape plants.

Burman attributes 92 of them to the Codex Witsenii,

34 to Herbarium Witsenianum and 33 to the Collection

or Codex Simon van der Stel. Of all these figured

plants six can be identified as being almost identical

to water-colours in BRI. In the text accompanying

the plates he attributes five of the six species to the

Collect, van der Stel. e.g. Tab. XII, fig. 2, “ atque

haec in Collect, van der Stel eleganter depicta mihi

obvenit, unde hanc exhibemus" . All of these species

depicted are Group B paintings. The remaining

28 species attributed to the Collect, van der Stel do not

occur in BRI. Therefore it must be assumed that

Burman had all these plants reproduced as engravings

from a volume which he knew was the Collect, van

der Stel and that the five Group B copied water-

colours in BRI came from that source. As stated by

Jessop (1965) and by Gunn & Codd (1980) this

volume is not traceable.

The remaining BRI water-colour depicted in

Burman’s work is of Oxalis purpurea on Tab. XXVII,

fig. 3 and is attributed to the Codex Witsenii. This is

the figure that Jessop was concerned about because

of the hairiness and stamens which Burman had

added. Reference to the text shows that Burman

referred to other works in which the species was

mentioned, namely those of Commelin who described

his species as glabrous, of Breyne as hirsute (in fact

only the calyx) and of Herman also as hairy and had

thus adapted his figure accordingly. There is, of

course, the possibility that the copier for BRI had

merely omitted the hairs and stamens, which were

present in the original.

In Fig. 5 of the present article showing Pelargonium

cucullatum, the centre illustration is taken from

Burman’s work, Tab. XXXV. He cites in the text

“& ex Collect. D. van der Stel hanc publici juris

facimus". On the left in Fig. 5 there is the same

species as depicted in BRI 103 and on the right is

BFC 54. The similarity between the BRI water-colour

and the Burman engraving is obvious. The BFC

water-colour is rather far removed, but can be seen

to bear some resemblance. Both water-colours belong

to the Group B copied paintings. It has been shown

earlier in this article that the BRI copies are probably

truer copies of the originals than the BFC copies.

One would then assume that the BRI and BFC

paintings were copied from the same original which

was in the Collect, van der Stel. The other two species

of Pelargonium illustrated in this plate were copied

from the Codex Witsenii and are almost identical

to the water-colours attributed to Claudius in SAM,

IPA and TCD.

One plate, Tab. LXXV, in Burman’s work is of

special interest. In it are depicted Crassula capensis

(fig. 4) and an unidentifiable composite (fig. 5).

The former is present in both BFC and BRI, but the

latter is only in BFC. In the text Burman states for

the Crassula “ Fructus nec semina adpicta sunt in

Collect. D. van die Stel, unde hanc cum subsequente

producimus", the subsequent figure being the com-

posite. This would indicate that the Breynes had had

their painting copies from the Collect, van der Stel.

Barnard (1947) gives a detailed description of

SAM which he says is part of the Codex Witsenii

particularly as it is autographed by Witsen. The

Codex contains 12 water-colours which Burman

figured and attributed to the Codex Witsenii. These

engravings were taken from SAM and not from the

almost identical paintings in IPA, SAPL or TCD,

as a detailed comparison of the paintings and

engravings shows a closer match with SAM. Many

of the plates which Burman attributes to the Codex

Witsenii can be traced to that very fine Codex, IPA,

in Johannesburg. This has been discussed by Macnae

& Davidson (1969).

CONCLUSIONS

The four early florilegia housed in Libraries in

Pretoria, Johannesburg, Oxford and Leiden consist

of the same basic set of water-colour paintings of

Cape plants. These paintings can be divided into two

distinct groups, A and B, on the style of painting

and the quality of detail. The key volume is the

florilegium, BRI, housed in Pretoria.

In BRI the quality of the Group A paintings is

outstanding, as the paintings have more natural

proportions, in some cases contain more and better

details and were executed with good quality paint.

They must be regarded as originals. The Group B

paintings are all reasonable copies taken from another

or, perhaps, several sources. The paper on which

the paintings were executed is all of similar make and

quality and is unfolded. All the paintings were

executed at the Cape directly into the volume, as is

indicated by the occurrence of Groups A and B

randomly distributed through the volume, some on

the same sheet and others even the same page.

The volume was later properly bound in vellum.

The paintings were nearly all numbered consecutively

giving a series of numbers relevant only to BRI.

In BFC the Group A paintings are of poorer quality

with sometimes fewer details depicted, they have

altered unnatural proportions and were painted with

a poorer quality water-colour paint. They all bear a

number which corresponds to that in BRI, but are

randomly arranged at the beginning of BFC. These

paintings are undoubtedly copies of some of the

Group A paintings in BRI, and are done on unfolded

paper. The Group B paintings of which there are

many more than in BRI, are likewise copies taken

from a similar source as BRI. These are all done on

folded paper which indicates that the copying was

done at the Cape. The Group A and B paintings were

copied separately on several different types of paper

and then bound into a volume as late as 1724, with

Group A paintings first and Group B’s second.

At this stage the pages or folios, not the paintings,

must have been numbered consecutively.

The volumes, OXF and LD, are both sets of inferior

copies with their numbering related to the folio

numbers of BFC. This indicates that they must have

been copied from BFC after 1724. The copies in

LD were acquired at different times, the Group A

paintings in 1779 and the Group B paintings in 1778,

possibly from the estate of Seba according to

Gunn & Du Plessis (1978).

Burman used some Group B paintings from which

engravings were made and published in his Rariorum

Africanarum Plantarum of 1738/39. He stated that

they came from the collection of Simon van der

Stel. This could indicate that all the Group B

paintings were copies for BRI and BFC from one

of the volumes of paintings owned by Van der Stel

and now untraceable.

My conclusions from a comparison of BRI and

BFC are that the BRI volume was painted first with

Group A paintings painted from live plants and the

Group B paintings copied at the Cape from Van

der Stel’s collection of paintings. They were then

E. G. H. OLIVER

125

number consecutively. The same artist, or perhaps

another one, must have copied some of the BRI

Group A paintings for Jakob Breyne, and yet another

artist copied the Group B paintings for Breyne

possibly from the same collection of Van der Stel.

These paintings were eventually bound in 1724 and

the pages numbered. The other two florilegia, OXF

and LD must then have been copied from BFC.

The main questions that remain are — who were the

artists and when was the first florilegium painted?

The first question will probably remain as a point

for speculation and remain unanswered for ever, as

none of the artists active at the Cape in its early

days ever signed a copy of his work. When the BRI

florilegium was executed can be roughly deduced from

certain facts. Jakob Breyne must have acquired his

collection of paintings before he died in 1697. Thus

for BFC Group A paintings to have been copied

from BRI, BRI must have been in existence before

1697. As Breyne did not use any of the BFC

paintings to illustrate the Cape plants depicted in

his Centuria of 1678, it can be assumed that he

acquired BFC after 1678. Gunn & Du Plessis (1978)

mention several artists who were active at the Cape

from the mid- 1680’s to the mid-1690’s. It is reasonable

to assume that BRI was painted during that period.

As Gunn & Codd (1980) state, botanists in South

Africa must be grateful to the Brenthurst Press for

publishing the complete BFC florilegium in colour.

This statement I certainly endorse. Already available

as reproductions are the superb facsimile edition in

colour of TCD (Waterhouse, 1979), the sepia reproduc-

tion of IPA (Kennedy, 1967) and the rather poor black

and white reproduction of the SAM paintings (Bar-

nard, 1947), all of which give botanists an idea of the

paintings in those volumes and something with which

to make comparisons. As yet BRI, SAPL, OXF and

LD have not been reproduced in any form to make

them generally available to researchers. Gunn & Codd

(1980) also point to the possibility of the existence of

“undiscovered” manuscripts and volumes in libraries

and archives in Europe, particularly in the rich collec-

tions at Leiden. It is hoped that this article will add

to the increasing literature on early Cape florilegia and

that at some time in the future new information will

come to light that will solve some of the unanswered

problems.

ACKNOWLEDGMENTS

Sincere thanks are due to the Brenthurst Library

in Johannesburg for allowing examination of the

original copy of the Breynes’s Flora Capensis.

This also applies to the Africana Museum in

Johannesburg and to the South African Museum

and South African Public Library in Cape Town for

permitting examination of the original codices in

their collections.

Useful discussions were held with Dr L. E. Codd

and with Miss M. D. Gunn. Mrs A. Romanowski is

thanked for her reproduction of the water-colour

paintings.

UITTREKSEL

'n Aantal vroee Kaapse florilegiums en kodekse

bestaan in biblioteke in Europa en Suid-Afrika. Vier

van hierdie florilegiums is naverwant en word in die

Brenthurst-biblioteek, Johannesburg, die Navorsings-

instituut vir Plantkunde, Pretoria, die Bodley-biblioteek,

Oxford en die Rijksherbarium, Leiden, gevind. Die

florilegiums van die eerste twee word met mekaar

vergelyk en in detail in hierdie artikel bespreek.

Die artikel het ontstaan na aanleiding van 'n resensie

van die Brenthurst-weergawe van Jakob en Johan

Philipp Breyne se Flora Capensis wat vir hierdie

uitgawe van Bothalia opgestel is. Voortspruitende uit

die vergelyk ing word 'n nuwe verklaring vir die onder-

linge verwantskappe en die oorsprong van die vier

florilegiums voorgestel. Die sleutelvolume is die

florilegium in die Navorsingsinstituut vir Plantkunde,

Pretoria.

REFERENCES

Barnard, K. H., 1947. A description of the Codex Witsenii

in the South African Museum. Jl S. Afr. Bot. 13: 1-52.

Breyne, J., 1678. Exoticarum Aliarumque Minus Cognitarum

Plantarum Centuria Prima, etc. Danzig: Rhetius.

Breyne, J. & Breyne, J. P., 1739. Prodomi Fasciculi Rariorum

Plantarum Primus et Secundus, etc. Danzig.

Burman, J., 1738-1739. Rariorum Africanarum Plantarum, etc.

Amsterdam: H. Boussiere.

Churceiill, W. A., 1935. Watermarks in paper in Holland,

England, France etc. in the XVII & XV III centuries and

their interconnections. Amsterdam: Menno Hertzberger.

Edwards, P., 1964. Some manuscripts relating to South

African botany in the William Sherard Collection in the

Bodleian Library at Oxford. Jl S. Afr. Bot. 30: 103-105.

Gunn, M. D. & Du Plessis, E., 1978. Introduction and text in

The Flora Capensis of Jakob and Johan Philipp Breyne.

Johannesburg: Ad. Donker for the Brenthurst Press.

Gunn, M. D. & Codd, L. E., 1980. Botanical collectors in

Southern Africa. Cape Town: Balkema. In press.

Heawood, E., 1924. The use of watermarks in datin old maps

and documents. Geogl. J. 63: 391-410.

Heawood; E., 1950. Watermarks mainly of the 1 1th and 18 th

centuries. Hilversum: The Paper Publications Society

(Monumenta Chartae Papyraceae, No. 1).

Jaffe, A., 1930. Zur Geschichte des Papieres und Seiner

Wasserzeichen. Pfalzische Heimatkunde (26 Jahrgang),

1930, Heft 3/4: 3-29.

Jessop, J. P., 1965. A volume of early water-colours in the

Library of the Botanical Research Institute, Pretoria.

Jl S. Afr. biol. Soc. 6: 38-52.

Kennedy, R. F., 1967. Catalogue of pictures in the Africana

Museum, Vol. 2, C307-738. Johannesburg.

Macnae, M. M. & Davidson, L. E., 1969. The volume “ leones

Plantarum et Animalium" in the Africana Museum,

Johannesburg; and its relationship to the Codex Witsenii

quoted by Jan Burman in his “ Decades Rariorum

Africanarum Plantarum" . Jl S. Afr. Bot. 35: 65-81.

Oliver, E. G. H., 1980. Book review: The Flora Capensis of

Jakob and Johan Philipp Breyne, ed. M. Gunn & E. du

Plessis. Bothalia 1 3 : 000.

Smith, A. H., 1952. Notes on the Claudius water-colours in the

Africana Museum. Johannesburg: Africana Museum,

Frank Connock Publication No. 1.

Tachard, G., 1686. Voyage de Siam, etc. Paris: Arnould

Seneuze & Daniel Horthemels.

Tachard, G., 1689. Second Voyage du Pere Tachard, etc.

Paris: Daniel Horthemels.

Waterhouse, G., 1932. Simon van der Stel's Journal of his

Expedition in Namaqualand, 1685-86. Dublin: Longmans

Green.

Waterhouse, G., 1979. Simon van der Stel's Journey to Nama-

qualand in 1685. Cape Town & Pretoria: Human & Rous-

seau.

Bothalia 13,1 & 2: 127-133 (1980)

Musci austro-africani II. Bryophyte collections in southern Africa and

southern African type specimens in the National Herbarium, Pretoria

R. E. MAGILL*

ABSTRACT

A brief review of bryological collections and collectors in southern Africa introduces a catalogue of southern

African type specimens housed in the National Herbarium, Pretoria. The type catalogue, arranged alphabetically

by basionym, includes correct names, type status and label data.

RiSUMt

MUSCI AUSTRO-AFRICANI II. COLLECTIONS DE BRYOPHYTES EN AFRIQUE AUSTRALE ET

SP£CIMENS-TYPES DE L’AFRIQUE AUSTRALE DANS L’HERBIER NATIONAL A PRETORIA

Apres une introduction consacree a une breve revue des collections bryologiques et des recolteurs en Afrique

australe, on presente un catalogue des specimens-types de l' Afrique australe qui sont conserves dans I’Herbier

National a Pretoria. Ce catalogue est dispose en ordre alphabetique par basionyme; il inclut les noms corrects, le

statut du type et les donnees des etiquettes.

It has become obvious, through examination of

several recent revisions dealing in part with southern

African taxa, that the distribution and the content

of bryophyte collections housed in South African

herbaria are not fully understood. Although most

southern African herbaria have some bryophyte

material, most of the specimens housed in the Flora

of Southern Africa area, are at the five larger herbaria

(BOL, GRA, NBG, NH, PRE). In connection with

the current research on southern African mosses,

the curators of these herbaria have made their

collections available to the Botanical Research

Institute (PRE) on temporary loan.

Each of these herbaria has a unique collection

with only a limited amount of overlap. For reasons

discussed below, the National Herbarium (PRE) has

the greatest overlap with each of the other herbaria,

as well as its own distinctive collection (Table 1).

A survey of each of these collections substantiates

Sim’s (1926) supposition that specimens gathered

by the early collectors are not represented in South

Africa. Collectors such as W. J. Burchell, J. Breutel,

J. F. Drege, W. H. Harvey, C. W. L. Pappe, C.

Thunberg and F. Wilms made small but very

important collections of bryophytes, that were

apparently sent to Europe for investigation and

distribution. Dr A. Rehmann is the only early collector

with his specimens represented in substantial numbers

in southern African herbaria. After each of his

collecting trips to southern Africa, Rehmann dis-

tributed his mosses in an exsiccata entitled “Musci

Austro-africani” (cf. Dixon & Gepp, 1923).

Rehmann’s contact with P. MacOwan, H. Bolus

and J. Medley Wood and the contributions made by

these botanists to the exsiccata, accounts for the

presence of sets at BOL, GRA and NH. Each of

these sets, although containing numerous identical

specimens, contains several collections not found

in the other two sets. A pre-distribution set of the

exsiccata was apparently sent to C. Muller, who

in 1899, described a large number of the specimens

as new species. It is indeed unfortunate that inconsist-

ency exists between the citation of specimen numbers

by Muller and Rehmann, since the specimens of the

exsiccata probably represent the only extant type

material.

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

TABLE 1. — Distribution of major or historic collections of

South African bryophytes

Key: 1 — First set or major collection.

2 — Major duplicate set.

3 — Duplicates.

4 — Exsiccata.

5 — Representative specimens.

H. A. Wager (1917) and T. R. Sim (1915) began

the first collections of bryophytes to be retained in

southern Africa. Nevertheless, specimens of both

collectors are also well-represented in the British

Museum, because of their correspondence with

H. N. Dixon.

Wager made a random collection of southern

Africa mosses. His specimens are rarely numbered

and location data are generally restricted to the listing

of a nearby town. On the other hand, Sim made

extensive collections throughout southern Africa

and Rhodesia. His specimens are mostly numbered

and frequently have labels with data in addition to

locality, i.e. habitat, association. Most of Sim’s

collections were identified or verified from duplicates

sent to Dixon. Sim’s enthusiasm as a bryologist was

apparently infectious as he was able to encourage a

128

MUSCI AUSTRO-AFRICANI II. BRYOPHYTE COLLECTIONS IN

AFRICAN TYPE SPECIMENS IN THE NATIONAL

SOUTHERN AFRICA AND SOUTHERN

HERBARIUM, PRETORIA

large number of amateur and professional botanists

to collect bryophytes for him. The collections of

both Wager and Sim were bequeathed to the National

Herbarium upon the deaths of these first South

African bryologists.

Bolus Herbarium

The Bolus Herbarium has maintained curatorial

and research activities on the bryophyte collection

since Sim (1926) published his Bryophytes of Southern

Africa and, before the recent renovation of the

collection at PRE, was the only South African

herbarium to incorporate nomenclatural and taxono-

mic changes. The bryophyte collection at BOL

contains c. 7 000 specimens including a set of the

Rehmann exsiccata ( pro parte), an important collec-

tion of Cape Peninsula bryophytes by N. S. Pillans,

east African collections by R. A. Diimmer, F. Eyles

and E. A. Schelpe and a large extra-African reference

collection. A substantial backlog of approximately

1 500 unidentified South African collections by E.

Esterhuysen, S. Garside and E. A. Schelpe is currently

being identified as part of the author’s research and

the first set of these specimens will be housed at BOL.

The bryophyte collections at BOL are in packets

mounted on standard herbarium sheets, with one

to several packets per sheet. The sheets occasionally

have notes and illustrations attached. The Rehmann

specimens are kept separate in a vertical file packet

system.

Albany Museum

The bryophyte collection at the Albany Museum

(GRA) began with the collections of P. MacOwan

and his exchange with several overseas bryologists.

Although most of his own collections were sent out

on exchange, a few of his specimens are present in

the collection. The herbarium contains c. 1 300

specimens, including a set of the Rehmann exsiccata

( pro parte) and a small extra-African reference

collection. The specimens are in packets mounted

individually on standard herbarium sheets and the

collection has been annotated to conform with Sim’s

(1926) Bryophytes of Southern Africa. The Rehmann

exsiccata are maintained separately in individual

unmounted packets and the MacOwan specimens are

mounted, exposed or in packets, in light-weight

paper folders, approximately the size of herbarium

sheets. Most of the modern collection is made up of

duplicates from Wager and Sim, however GRA has

recently received specimens collected by A. Jacot

Guillarmod from the eastern Cape.

Compton Herbarium

The Compton Herbarium (NBG) obtained its

bryophyte collection through acquisition of the

South African Museum Herbarium. This collection,

considered the oldest in southern Africa and contain-

ing numerous phanerogam collections of the early

collectors cited above, holds only a few bryophyte

specimens of Ecklon, Pappe and Zeyher. As curator

of the collection, MacOwan was responsible for the

build-up of the very large extra-African reference

collection, but he left less than a dozen of his own

specimens at SAM, when he moved to the Albany

Museum. For the most part, the collection is made

up of specimens gathered during the 1920-1930’s

by K. H. Barnard, C. Thorne and W. Tyson, primarily

in the western Cape and South West Africa. The

c. 2 500 specimens, including a set of the east African

collections of R. A. Diimmer, are mounted in packets,

one to several per standard herbarium sheet. Most

of the South African collections were identified or

annotated by Sim and therefore follow his nomen-

clature and taxonomy.

Natal Herbarium

The South African specimens at the Natal Herba-

rium consist primarily of the Rehmann exsiccata,

however these are supplemented by specimens collected

in Natal by J. Medley Wood, H. Bryhn and L. Title-

stad. A large extra-African collection was obtained

through an exchange with N. Bryhn and includes a

set of his Canary Island specimens. The specimens are

mounted, exposed or in packets, on half herbarium

sheets, one specimen per sheet. The southern African

specimens have been annotated to conform nomen-

claturally and taxonomically with Sim (1926).

National Herbarium, Pretoria

The bryophyte collection at the National Herbarium

(PRE) is the largest in southern Africa, both in

respect to African and extra-African holdings.

The majority of the collection is made up of the

private collections of Sim and Wager and the holdings

of the Transvaal Museum acquired in 1953. The

collection has been increased over the years by the

work of several botanists, e.g. S. Arnell, M. Bosman,

A. M. Bottomley, E. M. Doidge, R. E. Magill and

P. J. Vorster and through an intermittent cryptogamic

exchange programme.

Wager deposited duplicates of his collections at

PRE, GRA and the Transvaal Museum. These

specimens correspond in number and nomenclature

to his checklist of South African mosses (1917).

Unfortunately, these specimens have no collection

numbers and only brief location data. His personal

herbarium was left to PRE and incorporated in 1951.

These specimens generally have collection numbers

corresponding to those used in several publications

by Dixon (1920, 1929). Wager’s practice of incorporat-

ing any subsequent collection in a single packet for

each species is regrettable, as this has resulted in a

few type specimens being mixed with other collections.

The Sim specimens represent the largest and most

extensive collection of southern African bryophytes.

In preparation for his text, “The Bryophyta of South

Africa”, Sim collected throughout southern Africa

and Rhodesia and persuaded many collectors to

send him specimens. During his research, Sim had

access to each of the above collections and frequently

kept duplicates in exchange for identifications. One

important note in this context was Sim’s annotation

and up-dating of the Rehmann exsiccata for BOL,

GRA and NH. Sim kept a scrap, whenever possible,

and this has resulted in an almost complete set of

the exsiccata at PRE.

The bryophyte collection at PRE had been mounted

in various sized packets with one to 10 packets per

herbarium sheet (Fig. 1). In an effort to make the

collection more accessible, the entire collection has

been repacketed and shifted to a vertical file packet

system (Figs 2 & 3). The specimens are now filed by

family in phylogenetic order (cf. Crosby & Magill,

1977 and Magill & Schelpe, 1979).

The species are filed alphabetically under the genus

and each genus, when possible, has been divided into

southern African, other African and extra-African

groups. In addition, the southern African species

are colour-coded by country and/or province. The

mosses have been annotated to follow the nomencla-

ture of Index Muscorum or recent revisions and the

types are filed under annotated names.

R. E. MAGILL

129

Fi'G. 1. — Three herbarium sheets illustrating various packet sizes and mounting procedure prior to renovation

of PRE bryophyte collection.

130 MUSCI AUSTRO-AFRICANI II. BRYOPHYTE COLLECTIONS IN SOUTHERN AFRICA AND SOUTHERN

AFRICAN TYPE SPECIMENS IN THE NATIONAL HERBARIUM, PRETORIA

Fig. 3. — Close-up of specimen box, illustrating filing system

and new packet.

In connection with the current research on the moss

flora, a card index of all southern African names

has been compiled and the South African collections

searched for type material. A total of 206 southern

African types (284 specimens) has been identified at

PRE and are listed below as an aid to researchers

interested in taxa present in southern Africa. The

citation is alphabetical by basionym; when necessary,

this is followed by the correct name. The status of the

PRE specimen(s) and label data, collector and

numbers are also listed under each name. No attempt

to choose lectotypes has been undertaken here:

citation of a single syntype indicates only current

holdings at PRE.

CATALOGUE OF SOUTHERN AFRICAN TYPE SPECI-

MENS IN THE NATIONAL HERBARIUM, PRETORIA

Acanthocladiella transvaaliense Ther. & Dix.

= Heterophyllum transvaaliense (Ther. & Dix.) Ther. & P.

Varde

Isotype: Transvaal, Woodbush, Wager s.n.

Amblystegium filiforme Wag. & Wright

=Pseudoleskea leskeoides (Par.) C. Mull.

Isotype: Natal, Van Reenen, Wager s.n.

Anoectangium assimilis Broth. & Wag.

=A. wilmsianum (C. Mull.) Par.

Isotype: Natal, Wager s.n.

Aongslroemia abruptifolia C. Mull.

=Dicranella subsubulata (C. Mull.) Jaeg.

Isotype: Cape, Esternek above Knysna, Rehmann 25.

A. gymnomitrioides Dix.

=Cladophascum gymnomitrioides (Dix.) Dix.

Iso-syntypes: Rhodesia, Zimbabwe, Sim 8747; Matopos,

Sim 8772, 8850; Khami, Sim 8838.

Aptychus sphaeropyxis C. Mull.

=Sematophyllum sphaeropyxis (C. Mull.) Broth.

Isotype: Natal, Inezanga, Rehmann 372.

Archidium acanthophyllum Snider

Paratype: Natal, Wager s.n., July 1908.

A. julicaule C. Mull.

Iso-syntype: Cape, Cape Town, Rehmann 426.

Barbula afroruralis C. Mull.

=Tortula ruralis (Hedw.) Gaertn., Meyer & Scherb.

Isotype: Cape, near Stinkwater, Rehmann 114.

B. brachyaichme C. Mull.

=Tortula hildebrandtii (C. Mull.) Broth.

Isotype: Cape, Cape Town, Rehmann 107.

B. deserta C. Mull.

=Desmatodon convolutus ( Brid .) Grout

Isotype: Cape, Cape Town, Rehmann 96.

B. eutrichostomum C. Mull.

=Tortella humilis (Hedw.) Jenn.

Isotype: Cape, near Blanco, Rehmann 91.

B. natalensis C. Mull.

=B. indica (Hook.) Spreng.

Isotype: Natal, Port Durban, Rehmann 104.

B. oranica C. Mull.

=Tortula hildebrandtii (C. Miill.) Broth.

Isotype: Orange Free State, Bethlehem, Rehmann 126.

B. reticularia C. Miill.

=Tortula papillosa Wils. in Spruce

Isotype: Cape, Cape Town, Rehmann 106.

B. salisburiensis Dix.

Isotype: Rhodesia, Salisbury, Eyles 596.

B. trichostomacea C. Miill.

=Trichostomopsis australasiae (Hook. & Grev.) Robins.

Isotype: Cape, Rondebosch, Rehmann 97.

B. trivialis C. Miill.

=Trichostomopsis trivialis (C. Mull.) Robins.

Isotype: Orange Free State, Kadziberg, Rehmann 99.

Bartramia africana C. Miill.

=Philonotis africana (C. Miill.) Rehm. ex Par.

Isotype: Natal, Inanda, Rehmann 93.

B. afrofontana C. Mull.

=Philonotis afrofontana (C. Miill.) Par.

Iso-syntype: Orange Free State, Kadziberg, Rehmann 192.

B. afro strict a C. Miill.

=B. substricta Schimp. in C. Miill.

Iso-syntypes: Cape, Cape Town, Rehmann 203, 204.

B. pernana C. Miill.

=Philonotis androgyna (Hampe) Jaeg.

Isotype: Cape, near Belvedere, Rehmann 191.

B. squarrifolia Sim

Syntypes: Transvaal, Witpoortje, Moss s.n.; Cape, Disa

Gorge, Sim 9166; Paarl Mountain, Sim 9639; Schoone-

kloof, Pil/ans 4089; Table Mountain, Mitchell s.n.;

Bews s.n.

B. subasperrima C. Miill.

=B. compacta Hornsch.

Isotype: Cape, Cape Town, Rehmann 213.

Brachythecium afroalbicans Dix.

=Juratzkaea leptura ( Tayl .) Weber ex Buck

Isotype: Cape, Blinkwater Ravine, Table Mountain, Bews

s.n. (Sim 8634).

B. afrosalebrosum C. Miill.

=B. implicatum (Hornsch.) Jaeg.

Iso-syntype: Orange Free State, Kadziberg, Rehmann 383b.

B. afrovelutinum C. Mull.

=B. implicatum (Hornsch.) Jaeg.

Isotype: Cape, mountains near Rondebosch, Rehmann 379.

B. erythropyxis C. Miill.

=B. implicatum (Hornsch) Jaeg.

Isotype: Natal, Inanda, Rehmann 382.

B. knysnae C. Miill.

=B. implicatum (Hornsch.) Jaeg.

Iso-syntype : Cape, between Knysna and Belvedere, Rehmann

387.

B. pinnatum Dix.

Isotype: Cape, Knysna, Wager 520.

Braunia maritima C. Miill.

=Hedwigidium integrifolium (P. Beaav.) Dix. in C. Jen.

Iso-syntype: Cape, Table Mountain, Rehmann 306.

B. peristomata Dix.

=Leucodon maritimus (Hook.) Wijk & Marg.

Iso-syntypes: Rhodesia, Zimbabwe Ruins, Sim 8750, 8778,

8793, 8809; Fort Victoria, Sim 8843.

Breutelia angustifolia Rehm. ex Sim

Holotype: Transvaal, Mac-Mac, McLea s.n. (Rehmann 538).

B. tabularis Dix. in Sim

Iso-syntypes: Cape, Platteklip Ravine, Table Mountain,

Sim 9277; Pillans 3335, 4899.

Bryum acuminatum Sim, horn, illeg.

=B. simii Schelpe

Holotype: Natal, Cathkin, Owen 15.

B. aterrimum C. Miill. ex Sim

Holotype: Cape, between Knysna and Belvedere, Rehmann

235.

B. liliputanum C. Miill.

=Brachymenium dicranoides (Hornsch.) Jaeg.

Isotype: Cape, Cape Town, Rehmann 241.

B. oranicum C. Miill.

=B. argenteum var. australe Dix.

Isotype: Orange Free State, Kadziberg, Rehmann 260.

B. rigidicuspis Dix.

=B. bicolor Dicks.

Syntypes: Natal, Van Reenen’s Pass, Wager 74; Rhodesia,

Zimbabwe, Sim 8790; Khami Ruins, Sim 8839.

Calymperes victoriae Dix.

Isotype: Rhodesia, Victoria Falls, Sim 8879.

Calyptothecium africanum Broth.

=C. hoehnelii (C. Miill.) Argent

Iso-syntype: Cape, Knysna, Rehmann 332.

R. E. MAGILL

131

Campylopus angustinervis Dix.

=Microcampylopus perpusillus (Mitt.) Broth.

Iso-syntypes : Transvaal, Belfast, Wager 884; Rhodesia,

Matopos, Sim 8862; Zimbabwe, Sim 8806.

C. bewsii Sim.

Syntypes: Natal, Knoll, Hilton Road, Sim 9838; Mount

aux Sources, Sim 9891.

C. echinatus Sim

=C. introflexus ( Hedw .) Brid.

Syntypes: Cape, near Cape Town, Rehmann 67; Camps Bay,

Rehmann 69; Knysna, Rehmann 72; Table Mountain, Bews

s.n. (Sim 8604); Pirie Forest, Sim 8581B; Natal, Murchison

Flats, Sim 9837.

C. edwardsii Sim

=C. olivaceonigricans (C. Mull.) Par.

Holotype: Transvaal, Johannesburg, Edwards s.n. (Sim 9836)

C. pseudojulaceus Sim, horn, illeg.

=C. simii Schelpe.

Holotype: Orange Free State, Kadziberg, Rehmann 58.

C. symonsii Sim

Syntypes: Natal, Giant’s Castle, Symons s.n. (Sim 9843,

9835); Mount aux Sources, Gibb s.n.

Chamaebryum pottioides Ther. & Dix.

Syntypes: Cape, Cape Town, Wager 633 cfr. 654, 653.

Cupressina anotis C. Mull.

=Ectropothecium regulare (Brid.) Jaeg.

Isotype: Cape, mountains near Estemek, Rehmann 413.

Dicranella rigida Dix. in Sim

Holotype: Cape, Paarl, Sim 9633.

D. symonsii Dix.

=Anisothecium symonsii (Dix.) Broth.

Isotype: Natal, Giant’s Castle, Symons s.n. (Sim 8665).

Dicranoloma entabeniense Magill

Holotype: Transvaal, Soutpansberg, Entabeni Forest,

Bottomley PRE-CH3381.

Dicranum atroluteum C. Mull.

=Campylopus atroluteus (C. Mull.) Par.

Iso-syntype: Cape, Cape Town, Rehmann 63.

D. bartramiaceum C. Miill.

=Campylopus atroluteus (C. Miill.) Par.

Isotype: Cape, Cape Town, Rehmann 37.

D. catarractilis C. Mull.

=Campylopus catarractilis (C. Miill.) Par.

Isotype: Cape, Devil’s Peak, Rehmann 64.

D. chlorotrichum C. Miill.

=Campylopus chlorotrichus (C. Miill.) Par.

Iso-syntypes: Cape, Montagu Pass, Rehmann 53; Knysna,

Rehmann 53b.

D. inandae C. Mull.

=Camplyopus purpurascens Lor.

Isotype: Natal, Inanda, Rehmann 43.

D. leucobasis C. Miill.

=Campylopus introflexus (Hedw.) Brid.

Isotype: Cape, Montagu Pass, Rehmann 71.

D. longescens C. Miill.

=Campylopus clavatus (R. Br.) Wils.

Isotype: Cape, near Estemek, Rehmann 41.

D. purpureoaureum C. Miill.

=Campylopus purpurascens Lor.

Iso-syntype: Orange Free State, Liebenbergsvley, Rehmann

59.

D. serridorsum C. Miill.

=Campylopus catarractilis (C. Miill.) Par.

Isotype: Cape, Table Mountain, Cape Town, Rehmann 45.

D. stenopelma C. Miill.

=Campylopus stenopelma (C. Miill.) Rehm. ex Par.

Iso-syntype : Cape, near Estemek, Rehmann 52.

D. tenax C. Miill.

=Campylopus stenopelma (C. Miill.) Rehm. ex Par.

Isotype: Cape, near Blanco, Rehmann 54.

Didymodon afrorubellus Broth. & Wag. ex Dix.

Isotype: Natal, Van Reenen’s Pass, Wager 79.

D. knysnae Rehm. ex Sim

=D. xanthocarpus (C. Miill.) Magill

Holotype: Cape, Port Elizabeth, Rehmann 83.

D. pottsii Dix.

=D. dimorphus (C. Miill.) Broth.

Isotype: Orange Free State, Bloemfontein, Eagles Nest,

Potts s.n. (Sim 8663).

D. subfontanus Dix. in Sim

Isotype: Transvaal, Witpoortje, Moss 10322.

Dimerodontium africanum C. Miill.

Iso-syntype: Cape, mountains near Rondebosch, Rehmann

354.

D. carnifolium C. Miill.

=Dimerodontium africanum C. Miill.

Isotype: Cape, mountains near Rondebosch, Rehmann 358.

Diplostichum africana C. Miill.

=Eustichia africana (C. Miill.) Par.

Isotype: Orange Free State, above Kadziberg, Rehmann 279.

Distichophyllum taylorii Sim

Holotype: Cape, near Wilderness, Taylor s.n. (Sim 10281).

Ditrichum spirale Dix.

Isotype: Cape, Gaika’s Kop, D., B. & M. Henderson 232.

Drepanocladus hallii Broth. & Dix.

Isotype: Cape of Good Hope, Hall 7.

D. sparsus C. Miill.

Isotype: Orange Free State, Kadziberg, Rehmann 398.

Drepanophyllaria caudicaulis C. Miill.

=Hygroamblystegium caudicaule (C. Miill.) Broth.

Isotype: Cape, Devils Peak, Rehmann 404.

Ectropothecium brevisetum Dix., horn, illeg.

=E. brachycarpum (Dix.) Magill

Isotype: Mozambique, Shirindjen, Junod 331.

intodon cymbifolius Wag. & Dix.

Isotype: Transvaal, Moorddrift, Waterberg, Wager 408.

E. natalensis Rehm. ex C. Miill.

Isotype: Natal, Van Reenen’s Pass, Rehmann 331.

Ephemerella nervosa Dix.

=Ephemerum nervosum (Dix.) Schelpe

Iso-syntype: Cape, King William’s Town, Wager 1082b.

Erpodium distichum Wag. & Dix.

Iso-syntype: Transvaal, Barberton, Wager 279.

E. transvaaliense Broth. & Wag. in Dix.

Isotype: Transvaal, Wolhuter’s Kop, Wager 189.

Fabronia waged Dix.

Isotype: Cape, Cape Town, Wager 5.

F. victoriae Dix.

Isotype: Rhodesia, Victoria Falls, Sim 8943.

Fissidens aciphyllus Dix.

Isotype: Natal, Port St Johns, Wager 927.

F. amblyophyllus C. Miill.

Iso-syntype: Natal, Inanda, Rehmann 285.

F. aristatus Sim

=F. simii Schelpe

Syntypes: Natal, Maritzburg, Sim 9903, 9907, 9909.

F. borgenii Hampe var. obtusifolius Dix.

Isotype: Natal, Van Reenen, Wager 166.

F. brevisetus Sim

=F. scleromitrius (Besch.) Broth.

Holotype: Natal, New Hanover, Sim 9906.

F. calochlorus Dix.

Iso-syntypes: Rhodesia, Victoria Falls, Sim 8891, 8882.

F. corrugatulus Dix.

Iso-syntypes: Rhodesia, Victoria Falls, Sim 8885, 8904.

F. dubiosus Dix.

Isotype: Rhodesia, Victoria Falls, Palm Grove, Sim 8819.

F. enervis Sim.

Syntypes: Natal, Maritzburg, Sim 9899, 9900; Nottingham

Road, v. d. Bijl s.n. (Sim 8648).

F. hoeegii P. Varde.

Iso-syntype: Natal, Umgeni near Pietermaritzburg, Hoeg

423, 430, 432.

F. hyalobasis Dix. in Sim

Isotype: Transvaal, Moorddrift, Wager 406.

F. ischyrobryoides C. Miill.

=F. marginatus Schimp. in C. Miill.

Isotype: Cape, Devils Peak, Rehmann 290.

F. latifolius Dix.

Iso-syntypes: Rhodesia, Zimbabwe, Sim 8766, 8807, 8753,

8761, 8768; Khami Ruins, Sim 8841; Matopos, Sim 8856.

F. microandrogynus Dix.

Isotype: Rhodesia, Bulawayo, Wager 895.

F. papillifolius Dix.

=F. urceolatus Wag. & Dix. in Sim

Isotype: Natal, Albert Falls, Umgeni Nook, Sim 8709.

F. parvilimbatus Sim

Syntypes: Natal, Greenkoppies, J. Sim s.n. (Sim 9915); New

Hanover, Sim 9905; Albert Falls, Sim 9912.

F. pectinidens Dix.

Isotype: Natal, Port St Johns, Wager 936.

F. perpaucifolius Dix. & Sim

Isotype: Cape, Stellenbosch, Wager 647.

F. pycnophyllus C. Miill.

Isotype: Cape, Cape Town, Rehmann 293.

F. rehmannii C. Miill.

=F. glaucescens Hornsch.

Isotype: Natal, Inanda, Rehmann 282d.

F. stellenboschianus Dix. in Sim

Isotype: Cape, Stellenbosch, Wager 612.

F. urceolatus Wag. & Dix.

Isotype: Transvaal, Pretoria, Wager 264.

F. wageri Dix. in Wag.

Isotype: Natal, Umkomaas, Wager s.n.

Fontinalis duthieae Sim

=Wardia hygrometrica Harv. in Hook.

Holotype: Cape, Cape Town, Platteklip, Sim 9389.

Funaria longicollis Dix.

Iso-syntypes: Rhodesia, Zimbabwe, Sim 8735, 8796, 8/y/,

Khami Ruins, Sim 8842.

132 MUSCI AUSTRO-AFRICANI II. BRYOPHYTE COLLECTIONS IN SOUTHERN AFRICA AND SOUTHERN

AFRICAN TYPE SPECIMENS IN THE NATIONAL HERBARIUM, PRETORIA

Glyphomitrium marginatum Wag. & Dix.

=Ptychomitrium marginatum (Wag. & Dix.) Dix.

Iso-syntype: Transvaal, Kaapsche Hoop, Wager 298.

Grimmia austropatens C. Miill.

=Grimmia apocarpa Hedw.

Isotype: Cape, Table Mountain, Rehmann 137.

G. caffra C. Miill.

=G. apocarpa Hedw.

Isotype: Orange Free State, Witteberge above Kadziberg,

Rehmann 130.

G. drakenhergensis Sim

=G. pulvinata (Hedw.) Sm.

Syntypes: Natal, Giant’s Castle, Symons s.n. (Sim 9962);

Mount aux Sources, Sim 9963; Edwards s.n.

G. leptotricha C. Miill.

=G. pulvinata (Hedw.) Sm.

Isotype: Cape, Somerset East, Mt Boschberg, MacOwan s.n.

G. senilis Sim

=G. laevigata (Brid.) Brid.

Syntypes: Cape, Cookhouse, Sim 9956; Orange Free State,

Springfontein, Sim s.n.

Gymnostomum bewsii Sim in Dix.

=Anoectangium wilmsianum (C. Miill.) Par.

Syntypes: Natal, Tugela Gorge, Bews s.n. (Sim 8660);

Giant’s Castle, Symons s.n. (Sim 8661).

G. gracile Dix., hom illeg.

=G. wageri Schelpe

Isotype: Transvaal, Pretoria, Wager 91.

G. lingulatum Sim

=Didymodon lingulatum (Sim) Magill

Holotype: Transvaal, Lechlaba, Houtbosch, Rehmann 437.

Gyroweisia amplexicaulis Sim

=Husnotiella latifolia (Dix.) Zander & Magill

Holotype: Natal, Van Reenen, Wager 4\4.

G. latifolia Dix.

=Husnotiella latifolia (Dix.) Zander & Magill

Isotype: Rhodesia, Victoria Falls, Sim 8931.

Harrisonia eckloniana C. Mull.

=Rhacocarpus purpurascens (Brid.) Par.

Iso-syntype: Cape, Table Mountain, Rehmann 314.

H. rehmanniana C. Miill.

=Rhacocarpus rehmannianus (C. Miill.) Wij'k & Marg.

Iso-syntype: Cape, Table Mountain, Rehmann 313.

Hookeria tristis C. Miill.

=CallicosteIla tristis (C. Miill.) Broth.

Isotype: Natal, Inanda, Rehmann s.n.

Hymenostomum eurybasis Dix.

Iso-syntypes: Rhodesia, Matopos, Eyles 940, 941; Hellet’s

Concession, Magude, Sim 8989.

H. opacum Wager

=Hyophila zeyheri (Hampe) Jaeg.

Holotype: Natal, Van Reenen, Wager s.n.

Hyophila basutensis Sim

Holotypes: Cape, Rhenosterberg, McLea s.n. (Rehmann 458).

H. erosa Sim

=Oreoweisia erosa (C. Miill.) Kindb.

Holotype: Natal, Muller’s Farm, Drakensberg, McLea s.n.

(Rehmann 412).

Isopterygium brachycarpum Dix.

=Ectropothecium brachycarpum (Dix.) Magill

Iso-syntypes: Cape, Knysna, Wager 512; Transvaal,

Rietfontein, Wager 234.

I. punctulatum Broth. & Wag.

Isotype: Transvaal, near Rustenburg, Wager s.n.

I. taylorii Sim

Holotype: Cape, Wilderness, Taylor s.n. (Sim 10287).

Leptotrichum brachypodum C. Miill.

=Ditrichum brachypodum (C. Miill.) Broth.

Isotype: Orange Free State, Kadziberg, Rehmann 86.

Leucobryum gueinzii C. Miill.

Iso-syntype : Cape, Montagu Pass, Rehmann 74.

L. perfalcatum Sim, nom. illeg.

=L. rehmannii C. Miill.

Holotype: Cape, mountains near Esternek, Knysna, Rehmann

75.

L. rehmannii C. Miill.

Isotype: Cape, mountains near Esternek, Knysna, Rehmann

75.

Lindbergia haplocladioides Dix.

Isotype: Natal, Rydal Mount, Wager 29.

L. viridis Dix.

Isotype: Transvaal, Kaapsche Hoop, Wager 310.

Macromitrium confusum Mitt.

=Macrocoma lycopodioides (Schwaegr.) Vitt

Iso-syntypes: Cape, Table Mountain, Rehmann 165b, 166;

Claremont, Rehmann 162; Camps Bay, Rehmann 159.

M. schlotheimiaeforme Par.

Isotype: Cape, Devils Peak, Rehmann 151.

Meteorium rehmannii C. Miill.

=Squamidium biforme (Hampe) Broth.

Iso-syntype: Cape, Montagu Pass, Rehmann 323.

Microthamnium cavifolium Dix.

= Mittenothamnium cavifolium (Dix.) Wijk & Marg.

Syntypes: Natal, Inanda, Rehmann 368; Inanda, Wood s.n.

(Rehmann 654); Pietermaritzburg, Wager s.n.

Microthamnium ctenidioides Dix.

=Mittenothamnium ctenidioides (Dix.) Schelpe

Isotype: Cape, near Hogsback, Henderson & Henderson 220.

Mielichhoferia subnuda Sim

Syntypes: Natal, Mooi River, Sim 10217, 10197; Goodoo,

Wager 788; Bergville, Wager 715; Ladysmith, Wager 686.

Nanobryum drummondii Dix.

=N. gladiolus (Mitt.) Bizot

Iso-syntype: Cape, Port St Johns, Wager 955.

Neckera pseudocrispa C. Miill.

=CaIyptothecium acutifolium (Brid.) Broth.

Isotype: Natal, Van Reenen’s Pass, Rehmann 328.

Oedipodium australe Wag. & Dix.

=Oedipodiella australis (Wag. & Dix.) Dix.

Iso-syntypes: Natal, near sea, Wager 3; Cape, Pirie Forest,

King William’s Town, Wager 823.

Orthotrichum macleai Sim

Holotype: Cape, Graaff-Reinet, McLea s.n. (Rehmann 514).

O. mirum Lewinsky

Holotype: Natal, Scheepers’s Nek, Sim 10104.

O. piliferum Sim.

=0. glaucum Spreng.

Lectotype: Cape, Uitenhage, Sim 9001; Isotypes: Eastern

Cape, McLea s.n. (Rehmann 516); King William’s Town,

Sim 7053; Jager’s Drift, Sim 10099.

O. transvaaliense Rehm. ex Sim

Holotype: Transvaal, Houtbosch, Rehmann 517.

Oxyrrhynchium confervoideum Sim

Holotype: Natal, Maritzburg, Sim 10278B.

O. subasperum Sim

Syntype: Zululand, Mtunzini, Edwards s.n.

Philonotis afrocapillaris Dix. ex Sim

Isotype: Cape, Wilderness, Taylor s.n. (Sim 10153).

P. laeviuscula Dix.

Isotype: Rhodesia, Umtali, Teague 253.

Physcomitrium succuleatum Wag. & Wright

Isotype: South Africa, Wager s.n.

Plagiothecium hendersonii Dix.

=Entodon dregeanus (Hornsch.) C. Miill.

Iso-syntypes: Cape, Hogsback, D. Henderson 366, 365.

P. membranosulum C. Miill.

Isotype: Cape, Table Mountain, Rehmann 389.

P. rhynchostegioides C. Miill.

Iso-syntype: Cape, Hex River Mountains, Rehmann 396.

P. selaginelloides C. Miill.

=P. membranosulum C. Miill.

Iso-syntype: Cape, mountains near Rondebosch, Rehmann

390.

Platygyrium afrum C. Miill.

=Hypnum cupressiforme var. capense Pleisch.

Isotype: Cape, prope Rondebosch, Rehmann s.n.

Polytrichum flexicaule C. Miill.

=P. commune Hedw.

Iso-syntype: Cape, near Esternek, Rehmann 275.

P. natalense Sim.

Syntypes: Natal, Upper Bushman’s River, Sim 8671; Mount

aux Sources, Edwards s.n. (Sim 9696).

P. trichodes C. Miill.

=P. commune Hedw.

Isotype: Natal, Inezanga, Rehmann 211.

Porothamnium woodii Sim

=Stereophyllum woodii (Sim) Magill

Isotype: Natal, Wood 285.

Porotrichum natalense C. Miill.

=Porothamnium natalense (C. Miill.) Fleisch.

Isotype: Natal, Inanda, Rehmann 334.

P. pennaeforme C. Miill.

=Porothamnium pennaefrondeum (C. Miill.) Card.

Iso-syntype: Cape, mountains near Blanco, Rehmann s.n.

Pottia afrophaea C. Miill.

=Hyophila afrophaea (C. Miill.) Warnst.

Isotype: Orange Free State, Bethlehem, Rehmann s.n.

P. subplanomarginata Dix.

Isotype: Cape, Stellenbosch, Wager 671.

Pseudoleskea macowaniana C. Miill.

=P. leskeoides (Par.) C. Miill.

Isotype: Cape, Somerset East, Mt Boschberg, MacOwan 1877.

Psilopilum afrolaevigatum Dix.

Isotype: Natal, Rosetta, Sim 8068.

P. wageri Broth, in Dix.

Isotype: Natal, Rydal Mount, Wager 45.

R. E. MAGILL

133

Pterygoneurum macleanum Warnst.

Isotype: Cape, Graaff Reinet, McLea s.n. ( Rehmann 461).

Ptychomitriopsis africana Dix.

Isotype: Transvaal, Soutpansberg, Lake Funduzi, Wager

1122.

Ptychomitrium eurybasis Dix.

Iso-syntypes: Rhodesia, Matopos, Sim 8851; Zimbabwe,

Sim 8808; Macheke, Eyles 1994.

Racomitrium drakensbergense Sim

=R. crispulum (Hook. f. & Wils.) Dix.

Holotype: Natal, Giant’s Castle, Symons s.n. (Sim 8697).

Rhapidorrhynchium zuluense Sim

=Sematophyllum zuluense (Sim) Magill

Holotype: Zululand, Ngoye Forest, Sim 10285.

RJiynchostegium afrostrigosum C. Miill.

=Rhynchostegiella zeyheri (C. Miill.) Broth.

Isotype: Cape, Cape Town, Rehmann 364.

Schwetschkea rehmannii C. Mull.

=Helicodontium lanceolatum (Hampe & C. Miill.) Jaeg.

Iso-syntype: Cape, Blanco, Rehmann s.n., 1875.

Sciaromium capense Mitt, ex Dix.

Isotype: Cape, near Hogsback, D., B. & M. Henderson 213.

Sematophyllum wageri Wright ex Wag.

Isotype: Natal, Wager s.n.

Semibarbula elongata Hilp.

=Barbula zambesii Magill

Iso-syntypes: Rhodesia, Victoria Falls, Sim 8895, 8897, 8898.

Stereophyllum natalense Sim

Holotype: Zululand, Ngoye Forest, Sim 10275.

Syrrhopodon uncinifolius C. Mull.

Isotype: Cape, Montagu Pass, Rehmann 129.

Thamnium afrum C. Miill.

=Porothamnium hildebrandtii (C. Miill.) Fleisch.

Iso-syntype: Natal, Van Reenen's Pass, Rehmann 329.

Thuidium amplexicaule C. Mull.

=Haplocladium angustifolium (Hampe & C. Miill.) Broth.

Isotype: Natal, Van Reenen’s Pass, Rehmann 392.

T. sublaevipes Dix.

=T. ramusculosum (Mitt.) Jaeg.

Iso-syntypes: Cape, Tjumie, D. Henderson 358; Transvaal,

Kaapsche Hoop, Wager 295.

T. thamniopsis Sim

=Rigodium kilimandscharicum (Broth.) Par.

Holotype: Transvaal, Mac-Mac, McLea s.n. (Rehmann 647).

Timmiella pelindaba Magill

Holotype: Transvaal, Pelindaba, Bosman 1607.

Tortella obtusifolia Dix.

=Barbula umtaliensis Magill

Isotype: Rhodesia, Umtali, Eyles 1741.

Tortella opaca Dix.

=Weissia opaca (Dix.) Magill.

Iso-syntypes: Rhodesia, Victoria Falls, Sim 8884, 8890.

T. petrieana Sim

=T. humilis (Hedw.) Jenn.

Holotype: Natal, Glynn Falls, Sim 10063.

Tortula irregularis Sim

=T. hildebrandtii (C. Miill.) Broth.

Holotype: Natal, Edendale Falls, Sim 10064.

Trachyphyllum maximum Dix.

Isotype: Rhodesia, Makoni, Nobbs 1317b.

Trematodon africanus Wag. & Dix.

Syntypes: Natal, Wager s.n.; Transvaal, Tzaneen, Wager s.n.

T. ligulatus Rehm. ex Roth.

=T. mayottensis Besch.

Isotype: Natal, Oakford, Rehmann 22.

T. pillansii Dix. in Sim

Iso-syntypes: Cape, Platteklip Ravine, Sim 9282; Table

Mountain, Sim 9297 ; Miller’s Point, Pillans 4058, 4060.

Trichostomum afrofontanum C. Mull.

=Barbula afrofontana (C. Miill.) Broth.

Isotype: Natal, Van Reenen’s Pass, Rehmann 82.

T. atrovirens C. Miill.

=Hyophila atrovirens (C. Miill.) Broth.

Isotype: Natal, Van Reenen’s Pass, Rehmann 119.

T. cyathiforme Dix.

=Hypnodontium dregei (Hornsch.) C. Miill.

Isotype: Rhodesia, Victoria Falls, Sim 8934.

T. rehmannii Sim

=T. tortuloides Sull. & Lesq.

Holotype: Transvaal, Pilgrim’s Rest, McLea s.n. (Rehmann

471).

Triquetrella strictissima C. Mull.

=T. tristicha (C. Miill.) C. Miill.

Isotype: Cape, Wellington, Rehmann 144.

Webera depauperata Sim

=Pohlia depauperata (Sim) Schelpe

Holotype: Natal, Giant’s Castle, Symons s.n. (Sim 10208).

W. macleai Sim, nom. illeg.

=Pohlia macleai (Sim) Schelpe

Holotype: Cape, Rhenosterberg, McLea s.n. (Rehmann 548).

W. revoluta Sim, horn, illeg.

=Pohlia simii Schelpe

Holotype: Transvaal, Houtbosch, Rehmann 566.

Weissia dieterleniae Ther.

Isotype: Lesotho, Leribe, Dieterlen s.n.

Weisiopsis pulchriretis Dix.

Isotype: Natal, Royal Natal National Park, Wager 739.

Zygodon africanus Sim

Holotype: Transvaal, Mac- Mac, McLea s.n. (Rehmann 497).

Z. dixonii Sim

Holotype: Natal, Cathkin Peak, Sim 10004.

Z. leptobolax C. Miill.

Isotype: Cape, near Rondebosch, Rehmann 379 (499, fide

Dixon & Gepp. 1923).

Z. runcinatus C. Miill.

Iso-syntype: Cape, Table Mountain, Rehmann 150b.

Z. strictissima (Rehm.) C. Miill.

=Triquetrella tristicha (C. Miill.) C. Miill.

Isotype: Cape, Wellington, Rehmann 144.

Z. transvaaliensis Sim

=Z. intermedius B.S.G.

Isotype: Transvaal, Houtbosch, Rehmann 500.

UITTREKSEL

'n Kort oorsig van briologiese versamelings en

versamelaars in suidelike Afrika lei 'n katalogus in

van tipe eksemplare van suidelike Afrika wat in die

Nasionale Herbarium te Pretoria gehuisves is. Die

tipe-katalogus, alfabeties gerangskik volgens basioniem,

sluit korrekte name, tipe status en etiket data in.

REFERENCES

Crosby, M. R. & Magill, R. E., 1977. A dictionary of mosses.

St Louis: Missouri Botanical Garden.

Dixon, H. N., 1920. New and interesting South African mosses.

Trans. R. Soc. S. Afr. 8: 179-224.

Dixon, H. N. & Gepp, A., 1923. Rehmann’s South African

mosses. Kew Bull. 6: 193-238.

Dixon, H. N. & Wager, H. A., 1929. New and noteworthy

mosses from South Africa. Trans. R. Soc. S. Afr. 18:

247-261.

Magill, R. E. & Schelpe, E. A., 1979. The bryophytes of

southern Africa: An annotated checklist. Mem. bot.

Surv. S. Afr. 43: 1-39.

Muller, C., 1899. Contributiones ad bryologiam austro-afram.

Hedwigia 38: 52-155.

Sim, T. R., 1915. Check list of the bryophytes of South Africa.

Pietermaritzburg: Royal Society of South Africa.

Sim, T. R., 1926. The bryophyta of South Africa. Trans. R.

Soc. S. Afr. 15: 1-475.

Wager, H. A., 1917. A check list of the mosses of South Africa.

Pretoria.

Bothalia 13,1 & 2: 135-147 (1980)

Notes on African plants

VARIOUS AUTHORS

AMARYLLIDACEAE

A NEW SPECIES OF CYRTANTHUS FROM

Cyrtanthus labiatus R. A. Dyer sp. nov., C.

inaequali O’Brien affinis, foliis latioribus, pedicellis

longioribus, perianthii tubo angustiore, lobis minoribus

differt.

Bulbus subglobosus plusminusve 35 mm diam.

Folia 3-4, synantha, usque 300 mm longa et 20 mm

lata basin et apicem versus gradatim contracta,

glabra. Pedunculus ±300 mm longus, cylindricus,

±10 mm diam. fistulosus, ±8-florus, floribus paten-

tibus; spathae 2, ±40 mm longae, ±5 mm latae;

pedicelli, plusminusve erecti, usque 22 mm longi.

Perianthium zygomorphum labiatum, corallinum,

55-60 mm longum; tubus ±35-40 mm longus, e basi

2,5 mm diam. sensim 6 mm diam. ampliatus; lobi

4 supra, 2 infra, ±20 mm longi, exteriores 6 mm

interiores 7 mm lati. Filamenta biseriata, ±15 mm

longa. Stylus cum staminibus 4 arcuatus, ad apicem

perianthii fere attingens.

Type. — Cape, 3324 (Steytlerville), 19 km from

Patensie on Cambria Rd (-DA), in Baviaanskloof, fl.

24/1/1979 in cult. Roy Bayliss 5660 (PRE, holo.).

Bulb subglobose, about 35 mm diam., contracted

into a short neck, covered with dry membranous

brown scales. Leaves 3-4, contemporary with the

inflorescence, up to about 300 mm long and 20 mm

broad, gradually contracted to apex and base, slightly

concave on upper surface, glabrous, slightly glaucous.

Peduncle about 300 mm long, cylindric, ±10 mm

diam., hollow, about 8-flowered, with the flowers

more or less spreading (in cult.); spathes 2, ±40 mm

long, ±5 mm broad; pedicels more or less erect, of

different lengths, up to about 22 mm long. Ovary

cylindric 6-7 mm long. Perianth zygomorphic, labiate,

55-60 mm long, light coral-red; tube about 35 mm

long, 2 , 5 mm at base and evenly expanded to mouth

6 mm diam., very slightly fluted from mouth; lobes

4±2, four upper ones (2 outer, 2 inner) overlapping

and forming a hood, 20 mm long, outer 6 mm broad,

inner 7 mm broad; two lower ones (1 outer 1 inner),

spreading decurved; filaments (4 ±2) attached more

or less in perianth throat, biseriate, somewhat shorter

than the perianth; anthers ±3 mm long when

dehiscing. Style becoming slightly longer than the

stamens under the hood and decurved from slightly

below stigma; stigma subcapitate very slightly

3-grooved (Figs 1 & 2).

As regards the distribution, only the type colony

is recorded. The collector, Roy Bayliss, states that

the plants grow in complete shade on the vertical

face of a road cutting. The face of the rocks of

Table Mountain Sandstone conglomerate were

covered with Coccinia quinqueloba and the Cyrtanthus

leaves were growing through this. There were many

offsets in the clumps, but no flowers at the time of

collection. He goes on to say that permanent shade

is given by Ficus capensis and Celt is africana.

The type specimen flowered in a pot in Barbara

Jeppe’s greenhouse in Johannesburg in January

1979. These conditions would be far different from

BAVIAANSKLOOF, SOUTH-EASTERN CAPE

Fig. 1. — Cyrtanthus labiatus, depicting the holotype from the

Baviaanskloof in the south-eastern Cape, cultivated by

B. Jeppe in Johannesburg.

Fig. 2 .—Cyrtanthus labiatus showing the 8-flowered umbel of

the holotype; the perianth-lobes have an unusual labiate

spread, 4 lobes comprising the upper hood-like portion

and 2 lobes comprising the lower lip.

those of plants flowering in the wild and this one

must bear in mind when interpreting and comparing

C. labiatus with related species. The epithet labiatus

seemed an obvious choice because of the 4 and 2

spread of the perianth lobes simulating the 2-lobed

corolla of the Lamiaceae (Labiatae).

R. A. Dyer

136

NOTES ON AFRICAN PLANTS

A NEW COMBINATION IN GETHYLLIS

Gethyllis namaquensis ( Schonl .) Oberm., comb.

nov.

Klingia namaquensis Schonl. in Rec. Albany Mus. 3: 178

(1919). Type: Cape, Namaqualand, without precise locality,

Kling s.n. (GRA, holo.).

This elusive species was first collected by the

Reverend H. Kling of Steinkopf in the Richtersveld,

northern Cape, who sent two bulbs to Schonland at

the Albany Museum in Grahamstown in 1918.

Schonland placed the species in a new monotypic genus

Klingia, which he named in honour of the collector.

He separated the genus from Gethyllis on the grounds

of the six filaments which are fused at the base and

expanded into a wide cup or corona.

The species was again collected in October 1978

near Rosh Pinah in South West Africa/Namibia by

Mr D. S. Hardy (No. 4924). It flowered at the Botani-

cal Research Institute in November 1979 and afforded

the author the first opportunity to study live

flowers.

The androecium of Gethyllis is extremely variable,

the stamens doubling or trebling in some species.

Wilsenach, in Plant Life 21 : 82-88 (1965), investigated

the idiograms of Klingia and Gethyllis and found them

to be identical, but he did not sink Klingia under

Gethyllis. However, it is clear from the evidence that

the two genera are congeneric. Klingia namaquensis

has, in fact, for some years been known as Gethyllis

namaquensis, but the combination has not been

published. This omission is now rectified.

A. A. Obermeyer

ASCLEPIADACEAE

A NEW SPECIES OF HUERNIA FROM OWAMBO

Huernia owamboensis R. A. Dyer, sp. nov.,

caulibus 40-50 mm altis, 8-15 mm diam., 4-5-

angulatis, corolla ±10 mm longa 20 mm expansa,

rosacea interne papillosa et minute rubro-punctata,

tubo campanulato ±5 mm longo ±10 mm diam.,

corona exteriore aeque 10-dentata, distincta.

Planta succulenta basi ramosa; caules erecti,

40-50 mm alti, 8-15 mm diam., glabri, virides, 4-5-

angulati, dentibus deltoideis apiculatis. Pedicelli

±12 mm longi. Sepala lineari-lanceolata, ±2,5 mm

longa. Corolla ±10 mm longa, 20 mm diam., extus

glabra, intus basin versus plana, supra papillosa,

incarnata, minute rubro-punctata; tubus campanu-

latus, ±5 mm, longus, 10 mm diam.; lobi patentee,

±triangulares, 5 mm longi, acuti. Coronae lobi

exteriores aeque 2-dentati (5 x 2), lobi interiores

breviter incumbentes, obtusi. Pollinia oblongo-

elliptica, 0,25 mm longa.

Type. — S.W. Africa, Owamboland, sandveld,

March 1974 (Flowered at PRE, April-May 1979)

J. Vahrmeijer PRE 57730 (holo.).

Stems succulent, densely branched from the base

and extending into clumps, 40-50 mm high, 8-15 mm

diam., 4-5-angled, green; angles tuberculate with

tubercles deltoid, apiculate, ±2 mm prominent.

Flowers produced from near base of young branches

(apparently opening in succession). Pedicels ±12 mm

long, spreading. Sepals linear-lanceolate, ±2,5 mm

long. Corolla ±10 mm long, ±20 mm diam. from

tip to tip of expanded lobes, closely ribbed on outer

surface, pale pink within, with minute maroon spots

except on lower half of tube; tube narrowly cam-

panulate with nearly parallel sides, ±5 mm deep,

±10 mm diam., smooth within basal portion of tube,

minutely papillate above and on lobes; lobes

spreading, with minute teeth at the sinuses, ±trian-

gular, ±5 mm long, acute. Corona united at base

spreading on corolla into 10 (5x2) equal teeth,

maroon, inner lobes adpressed on back of anthers

somewhat swollen, obtuse, not extending over top of

staminal column, amber-coloured. Pollinia 0,25 mm

long, elliptic-oblong, with thickened amber margin

2/? of its length, with short basal connectives to small

winged carrier (Fig. 3).

Fig. 3. — Huernia owamboensis. 1, plant, natural size; 2, flower,

x3 ( Vahrmeijer in PRE 57730, holotype). Photo: A

Romanowski.

VARIOUS AUTHORS

137

Only one plant was found and collected at the

time of discovery. It made slow progress in its new

environment at BR1 under the care of Mr D. Hardy.

When it eventually flowered in April-May 1979, it

proved to be undescribed. The small size of the stems

and flowers, their delicate pink colour and the

evenly 10-toothed outer corona are distinctive

features.

R. A. Dyer

LILIACEAE

A NEW SUBTRIBE IN LILIACEAE

Subtribus Caudibracteateae Oberm., subtribus

nova, Scilleae affinis, sed bracteis caudatis differt.

Type genus: Drimia Jacq.

In the J1 S. Afr. Bot. 43:265-319 (1977) Jessop

revised a group of closely related genera of the tribe

Scilleae, all bearing spurred bracts and including

some other features as well. These genera, namely

Drimia , Urginea, Sypharissa , Thuranthos, Rhada-

manthus, Litanthus, Schizobasis, Bowiea (and Rhodo-

codon from Madagascar) form a natural subtribe

and comprise the new subtribe Caudibracteateae.

It appears to be an ancient group in which the species

have become stable and in this it differs from the

rest of the Scilleae tribe, for instance the genera

Ornithogalum and Ledebouria, where the majority

of the species exhibit variability. In the subtribe

Caudibracteateae some floral parts, especially the

androecium, have become more complex in certain

species that otherwise are close to species with the

normal type of stamen. In Thuranthos macranthum

C. H. Wright in Curtis’s bot. Mag. 142: t.8680 (1916)

for instance, the stamens have evolved into a cage-like

body, while in its other characters the species resembles

T. basuticum (Phill.) Oberm. (p. 139). In Drimia

hyacinthoides Bak (1874) the lobes of the perianth

have remained short and erect, while in other respects

it closely resembles other members of Drimia with

long and recurved lobes.

A. Amelia Obermeyer

A NEW SUBGENUS RHADAMANTHOPSIS AND TWO NEW SPECIES OF RHADAMANTHUS

When plants of the two species described below

came into flower in the nursery of the Botanical

Research Institute, Pretoria, in 1976-77, it was at

first uncertain to which genus they belonged.

The two new species differ from typical Rhada-

manthus only by the locules of the anther, which open

loculicidally instead of by apical pores. When

Nordenstam revised the genus [Bot. Notiser 123:

155-182 (1970)], he pointed out that in some species,

e.g. R. fasciatus B. Nord. and R. albiflorus B. Nord.,

the dehiscence is by “apical slits, which eventually

proceed down to below the middle of the thecae”.

It was decided to place the two, possibly more

primitive, new species, in a new subgenus Rhada-

manthopsis.

Subgen. Rhadamanthopsis Oberm., subgenus

novum, Rhadamantho subgenui typico affine, sed

thecis antherae longitudinum dehiscentibus differt.

Type species: Rhadamanthus namibensis Oberm.

Rhadamanthus namibensis Oberm., sp nov., a

R. karooico Oberm. affinis, sed foliis erectis longiori-

bus et glaucescentibus differt.

Bulbus ovoideus compactus c. 50 mm in diam.

tunicis laevis lucidis succulentis; radices crassae.

Folia 2-4, hysterantha erecta linearia c. 140-240 mm

longa at 10-20 mm lata glauca firma. Racemus

simplex ad 0,5 m altus multiflorus; pedunculus

erectus teres; bracteae caudatae minorae caducae;

pedicelli sub anthesi decurvati c. 8 mm longi postea

erecta 12 mm longi. Perianthium campanulatum

c. 9 mm longum pallide lilacinum nitidum; lobi

c. 5 mm longi marginibus transparentibus; basi

interiori atrovinosi. Stamina 6 brevia inclusa conni-

ventia ad basin perianthii affixia; antherae introrsae

thecis loculis longitudinaliter dehiscentibus. Ovarium

oblongo-globcsun; ovulae c. 8 prope basin affixae;

stylus teres; stigma obtusum. Capsula oblongac. 8 mm

longa; semen oblongum, 5 mm longum nigrum

nitidum.

Type. — S.W. Africa, 2716 (Witputz): Witpiitz-Suid

1 km S.E. of Police Station (-DA), Giess 13781 (PRE,

holo. ; WIND).

Bulb c. 50 mm in diam. with wide white to pale

mauve, fleshy scales topped by the persistent, trans-

versely striped leaf-bases, which become white.

13781. Plat e by R. Weber.

138

NOTES ON AFRICAN PLANTS

Fig. 5. — Rhadamanthus namibensis. 1, 3 perianth lobes and

3 stamens, x3; 2, gynoecium and 2 s'amens, x3; 3, part

of raceme with one capsule, sterile flowei s and 2 seeds.

Fig. 6. — Rhadamanthus karooicus. Holotype in leaf. Leg. J. van

Zanten sub PRE 45560.

crinkled and dry with age. Roots many, swollen.

Leaves hysteranthous, 2-4 per shoot, erect, linear,

c. 1 50-240 mm long and 1 5—25 mm broad, canaliculate,

firm, glaucous, margin smooth, dying back from

the tip. Raceme simple, up to 0,70 m tall, many-

flowered; peduncle firm, terete, c. 0,4 m long,

rhachis about equally long; flowers laxly spaced on

rhachis; bracts small, varied, tailed, the lower part

pale lilac, somewhat fleshy; pedicels c. 8 mm at

at anthesis, pendulous in bud, erecto-patent in flower

and in fruit, lengthening to 12 mm. Perianth cam-

panulate, c. 9 mm long, pale mauve, shiny, with a

dark base, the lobes c. 5 mm long with transparent

margins. Stamens short, erect, connivent around

ovary, attached to base of perianth; filaments

1,5 mm long; anthers versatile, dorsifixed near base,

introrse, apiculate, the locules opening by longi-

tudinal slits. Ovary oblong-globose, 3-locular,

placentae basal; ovules c. 8, axile; style terete,

shallowly 3-grooved, slightly longer than ovary,

gradually thickening towards the obtuse apical

stigma, shorter than stamens. Capsule narrowly

ellipsoid, 5-7 mm long, apiculate, loculicidal; the

seed semi-discoid to narrowly obovoid, 4-5 mm long,

black, shiny. Figs 4 & 5.

At present the plant has only been recorded from

the Witpiitz District in southern South West Africa/

Namibia. It was previously collected by De Winter

6304 in the same locality and flowered in the nursery

in November 1960. Later two leaves were produced

measuring up to 240 mm long and 25 mm wide.

The type flowered in October 1977; capsules with

ripe seeds were produced in November; the new

leaves emerged in January and reached maturity in

March 1978^

Rhadamanthus karooicus Oberm., spec, nov., a

R. namihensi affinis, sed minoribus ; foliis patentibus

mmoribus et bulbo ad apicem applanato differt.

Bulbus oblongus applanatus, c. 30 mm latus tunicae

viridae ad apicem asperatae et siccae; radices

tenuiores. Folia 4-6 hysteranthia patentes c. 6,

tenuiter oblonga c. 40 mm longa et 7 mm brevi

atroviridia. Racemus simplex c. 0,2 m altus; flores

c. 30 cernuis. Perianthium campanulatum inflatum

c. 6 mm longum pallide lilacinum nitidum. Stamina

6 brevia inclusa conniventes ad basin perianthii

affixia; antherae introrsae thecis longitudinaliter

dehiscentibus Ovarium oblongo-globosum; ovulae

c 8 prope basin affixiae; stylus teres stigma minora.

Capsula ignota. Figs 6 & 7.

Fig. 7. — Rlndamanthus karooicus. Holotyi e flowering. Leg.

J. van Zanten sub PRE 45560.

VARIOUS AUTHORS

139

Type. — Cape, 3320 (Montagu): Laingsburg, Farm

Keurfontein (-BB), J. van Zanten sub PRE 45560

(holo.)

Bulbs forming colonies, oblong, broad and flat

above, up to c. 30 mm broad; tunics rough and dry

at the apex, smooth and green below; roots fairly

thin. Leaves spreading, up to 6, arranged in 2-3

opposing pairs, spreading, narrowly oblong, up to

40 mm long, 7 mm broad, glabrous, shiny, dark

green. Raceme simple, c. 0,2 m tall, erect, bearing

c. 30 pendulous flowers; bracts small, caudate,

membranous, vinaceous, early caducous; pedicels

patent, filiform c. 3-8 mm long. Perianth cam-

panulate, c. 6 mm long, swollen, pale lilac, shiny

with a faint green keel. Stamens 6, included, conni-

vent, attached to base of perianth; anthers c. 1 mm

long, orange, dehiscing introrsely by longitudinal

slits. Ovary ovoid, pale green, style cylindrical, stigma

apical, exserted above stamens but included. Capsule

unknown.

So far this is the only collection that could be

studied both in flower and in leaf. It did not set seed.

A specimen collected by Bayliss 6327 at Kaboega

in the Somerset East District on 1974-01-06 may

belong here, but the flowers were described as pale

yellow with an indigo streak. Leaves unknown. Noted

to be fairly common.

A. Amelia Obermeyer

A NEW COMBINATION IN THURANTHOS

Thuranthos basuticum ( Phill .) Oberm., comb.

nov.

Urginea basutica Phill. in Ann. S. Afr. Mus. 16: 306 (1917).

Type: Lesotho, Dieterlen 854a.

Drimia angustifolia Bak. FI. Cap. 6:489 (1897): Jessop in

J1 S. Afr. Bot. 43: 273 (1977), non Drimia angustifolia Kunth

Enum. 4: 340 (1843); the latter appears to be a species of

Ledebouria.

This species (Fig. 8) closely resembles the other

two species T. macranthum (Bak.) C. H. Wright

(1916) and T. nocturnale R. A. Dyer (1964), but the

filaments of the stamens are less specialized. They

also curve inwards around the style forming a cage-

like body, but the upper spreading part remains

short.

A. Amelia Obermeyer

Fig. 8. — Thuranthos basuticum. 1, stem and young leaves,

x0,6; 2, flowering stem, x0,06; 3, filaments and anthers,

xl,2; 4, anther, x3; 5, ovary and style, xl,2; 6, fruit,

x0,6; 7, longitudinal section, x0,6; 8, transverse section,

x0,6; 9, seed, x0,2. Strey 3935. Note: the flower was

figured during the morning when the periant-lobes had

started to close; at night they are fully recurved.

THE STATUS OF URGINEA EPIGEA

In his revision of the bulbous Liliaceae of South

Africa Jessop (1977) sank Urginea under Drimia.

While I agree that some species are better placed

with the Drimia species of Jacquin, there is a number

of others that I prefer to retain in Urginea, namely

those with small, white, stellate flowers and spreading

stamens which Jacquin considered to belong to

Anthericum. In his revision, Jessop (1977) erred in

placing Urginea epigea R. A. Dyer as a synonym of

Drimia altissima (L.f.) Ker Gawl. from the Cape.

While the latter is best placed in Drimia, Dyer’s

species is an Urginea, closely related to U. micrantha

(A. Rich.) Solms-Laub. (1867) from tropical east

Africa, first described as Scilla micrantha A. Rich, in

Tent. FI. Abyss.: 2:328 (1843).

There is also the taxonomic value of the nature

of the bulbs to be considered. In the gregarious

species Urginea epigea, the bulbs consist of loose

scales and grow above ground, whereas apparently

the majority of plants identified as U. micrantha,

have compact hypogeal bulbs and are usually solitary.

They are widely distributed in southern to tropical

Africa, whereas U. epigea has so far only been

recorded from the Transvaal and South West Africa.

From herbarium sheets with incomplete material

and meagre notes one cannot distinguish between

these two taxa. This problem requires further study

in the field.

A. Amelia Obermeyer

140

NOTES ON AFRICAN PLANTS

THREE SUBSPECIES OF EUCOMIS AUTUMN ALIS*

Within the genus Eucomis L’Herit. there is a group

of plants of problematic circumscription. These plants

vary in height from 200-500 mm and are a homo-

geneous green in colour. As a result, the different

taxa of this group have never been clearly defined

and various names such as E. autumnalis (Mill.)

Chitt., E. albomarginata Barnes, E. amaryllidifolia

Baker, E. clavata Baker, E. macrophylla hort., E. regia

sensu Ait., E. robusta Baker and E. undulata Ait.

have been assigned to them.

In an unpublished monograph of the genus

(Reyneke, 1972), it was concluded that the group

consisted of only three taxa, all varieties of

E. autumnalis, namely var. autumnalis, var. amarylli-

difolia and var. clavata. In this note the three taxa

are published as subpecies, since they occupy fairly

distinct geographical areas.

Eucomis autumnalis {Mill.) Chitt. in Royal Hort.

Soc. Diet. Gard. 2: 787 (1951). Type: apparently

none preserved; neotype: eastern Cape, Flanagan

2590 (BOL; NBG, PRE, isoneo.).

Bulb pyriform with dark brown or red-brown

membranous tunics. Leaves plain green, without a

purple colour, usually linear, sometimes lanceolate

or ovate; with a distinct midrib; margin undulate;

cuticle on margin dentate or crenate. Inflorescence a

homogeneous green raceme with 50-125 flowers.

Peduncle cylindrical or clavate. Scape 60-200 mm

long. Bracts linear, shorter than the flowers except

for those situated at the apex which are longer.

Coma with 10-45 sterile bracts, usually pendulous

over the inflorescence; bracts oblong, linear or ovate,

margin undulate, cuticle on margin undulate and/or

dentate. Pedicel erecto-patent, 3-9 mm long. Flowers

50-1 25 ; white, yellow-green or light green immediately

after anthesis, older flowers green. Perianth with 6

segments lanceolate to ovate, persistent. Stamens 6,

arising from the base of perianth, the filaments

triangular-acuminate, their broad bases fused below

to form a cup 3-4 mm high. Gynoecium white or

yellow-green in young stage, green when older;

style cylindrical; stigma simple; ovary with three

lateral indentations. Capsule trilocular; pericarp

membraneous, lies iy or hard, opaque with 2 distinct

layers. Flowering period December to February.

Eucomis autumnalis is usually found in open

grassland, but also in sheltered places e.g. in rock

crevices, under trees and shrubs and even in swamps.

The morphology of the peduncle, whether clavate

or cylindrical, is a distinctive character which does

not change in different habitats and is used to

distinguish between the three subspecies. A cylindrical

scape is characteristic of the subsp. autumnalis, while

the subspp. amaryllidifolia and clavata both have

clavate scapes (Fig. 9).

Key to the subspecies

Leaves linear 15-40 mm broad and 130-300 mm long,

margin undulate; peduncle clavate; perianth segments

6-10 mm long; capsule papery . .(b) subsp. amaryllidifolia

Leaves ovate to lanceolate, seldom linear, 40-130 mm

broad, 150-550 mm long, margin undulate; peduncle

cylindrical or clavate; peranth segments 10-16 mm

long:

Peduncle clavate; capsule with a hard double-layered

pericarp (c) subsp. clavata

Peduncle cylindrical; capsule with a thin, sometimes

inflated, pericarp (a) subsp. autumnalis

* Part of an M.Sc. thesis submitted to the University of

Pretoria.

(a) subsp. autumnalis

Fritillaria autumnalis Mill., Gard. Diet. ed. 8 (1768); Eucomis

autumnalis (Mill.) Chitt. in Royal Hort. Soc. Diet. Gard.

2:787 (1951).

F. longifolia Hill in Hill's Hort. Kew. ed. 2, 354, 1. 1 5 (1769);

Vitman, Sum. PI. 2:299 (1789-1792).

Eucomis undulata Ait., Hort. Kew. ed. 1, 1:433 (1789);

Don, Hort. Cant. ed. 2, 40 (1796); Willd., Sp. PI. 2: 93 (1799);

Gawler in Curtis's bot. Mag. 27: 1. 1 083 (1808); Willd., Enum.

364 (1809); Ait., Hort. Kew. ed. 2, 146 (1811); Don, Hort.

Cant. ed. 6, 86 (1811); Trattinick, Thes. Bot. t.53 (1819);

Thunb., FI. Cap. ed. 2, 317 (1823); Hook., Cat. PI. R. Bot.

Gard. 1:26 (1825); Spreng., L. Syst. Veg. 2:76 (1825);

Roem. & Schult., Syst. Veg. 7: 622-623 (1829); Kunth, Enum.

PI. 4: 302 (1843); Baker in J. Linn. Soc. (Bot.) 13: 225 (1873);

Nicholson, Ulus. Diet. Gard. 1: 538 (1885); Engler & Prantl,

Pflanzenfam. 2: 67, fig. 47 (1888); Bak. in FI. Cap. 6: 476-477

(1897); Engler & Drude, Veg. der Erde, 2: 301, fig. 203 (1908);

Pole Evans in Flower. PI. S. Afr. 6: t.220 (1926); Grey, Hardy

Bulbs 3:236 (1938); Batten & Bokelmann, Wild Flow. E.

Cape Prov. 12 (1966); Gledhill, East Cape Wild FI. 72 (1969).

Ornithogalum undulatum (Ait.) Thunb., Prodr. 1, 62 (1794).

Basilaea undulata (Ait.) Mirb., Hist. Nat. PI. 8: 339 (1802-

1806). Type: specimen in BM.

E. regia sensu auct. sequent.: Ait., Hort. Kew. 1: 433 (1789);

Soland. in Salisb. Prodr. 218 (1796), as Eucomea regia; Don,

Hort. Cant, ed.2, 40 (1796); Willd., Sp. PI. 2: 93 (1799); Red.,

Liliac. 4: 1. 1 75 (1807); Ait., Hort. Kew. ed.2, 2:245 (1811);

Roem. & Schult., Syst. Veg. 7: 623 (1829); Kunth, Eium. PI.

4: 302 (1843); non L’Heritier. Basilaea regia sensu Mirb., Hist.

Nat. PI. 8: 339 (1802-1806).

Basilaea coronata Lam., Encycl. 1 : 382 (1789); Lam., Encycl.

1:590 (1810); Poiret, Encycl. suppl. 1, 590 (1810-1811); Lam.,

Tabl. Encycl. t.239, fig. 2 (1823).

Eucomis clavata sensu Van der Spuy, Die Groot Veld-

blommeboek 175 (1971), non Baker.

Bulb pyriform. Leaves linear to ovate, 60-130 mm

broad and 150-550 mm long, margin undulate.

Inflorescence 40-55 mm in diameter with 70-110

compactly arranged flowers. Peduncle cylindrical,

sometimes slightly clavate, 60-200 (300) mm long.

Scape 60-200 mm long. Coma 70-150 mm in diameter,

10-45 bracts; bracts oblong, 30-70 mm long and

20-40 mm broad and 4-5 mm in diameter. Perianth

segments 10-13 mm long. Capsule with membranous

pericarp. Figs 9 & 10.

Subsp. autumnalis usually grows as single plants

in open grassland, on mountain slopes and forest

margins of Malawi, Rhodesia, northern Transvaal

and the eastern Cape.

Malawi. — 1234 (Kota Kota): Kota Kota (-CC), Benson

89 (PRE).

Rhodesia. — 1832 (Umtali): Inyanga (-BA), Juliesdale,

Rutherford-Smith 505 (PRE), World’s View (-BA), Plowes

2680 (PRE). 1932 (Melsetter): Umtali, Vumba Mountain

(-BA), Obermeyer 2138 (PRE), Castle Beacon (-BA), Farrar

4041 (PRE), Plowes 2240 (PRE).

Transvaal. — 2330 (Tzaneen): Duiwelskloof (-CA), West-

falia Estate, Scheepers 297 (PRE), 2427 (Thabazimbi): Krans-

berg (-BC), farm Groothoek, Codd 6314 (PRE). 2529

(Witbank): Loskop Dam (-AD), Theron 1769 (PRE & PRU).

Cape. — 3024 (De Aar): Hondeblafriver (-BC), “Horses

Grave”, Burchell 2701 (K). 3029 (Kokstad): Cedarville (-AC),

Mconyani, Bandert 1 1 1 (GRA). 3224(GraafF-Reinet): Sneeuberg

(-AA), Burke, s.n. (K). 3225 (Somerset East): Cradock Berg-

kwagga Park (-AB), Muller 660 (PRE). 3227 (Stutterheim):

Moordenaarskop (-BD) near Komga, Flanagan 2590 (BOL,

NBG & PRE). 3325 (Port Elizabeth): Uitenhage (-CD),

Ecklon & Zeyher 102 (K).

No specimen was apparently preserved at the

time of the original description of this plant, which

was grown in England by Miller from seeds which

were sent to him from the Cape of Good Hope

therefore, in the absence of a figure, a neotype,

Flanagan 2590, is selected.

Within subsp. autumnalis, originally described

from the eastern Cape, two forms may be distin-

guished, namely plants with a southern distribution

VARIOUS AUTHORS

141

Fig. 9.— Drawings of 1, Eucomis

autumnal is subsp. autumnalis;

2, subsp. amaryllidifolia', 3,

subsp. clavata.

Fig. 10. — The known distribution of Eucomis autumnalis

subsp. autumnalis.

Fig. 11. — The known distribution of Eucomis autumnalis subsp.

amaryllidifolia.

from the eastern Cape showing affinity with subsp.

clavata in growth and leaf forms, and those with a

northern distribution from Malawi, Rhodesia and

the northern Transvaal showing affinity with E.

pallidiflora Bak. except in stature. These two forms

could possible be regarded as distinct subspecies, but

there are so many intermediates that such a distinction

would be tenuous.

(b) subsp. amaryllidifolia {Bak.) Reyneke, stat.

nov.

E. amaryllidifolia Bak. in Gard. Chron. 2, 10: 492 (1878);

Nicholson, Illus. Diet. Gard. 1:537 (1885); Bak. in FI. Cap.

6:477 (1897); Grey, Hardy Bulbs 3:232 (1938); Chitt. in

Royal Hort. Soc. Diet. Gard. 2: 787 (1951). Type: Eastern

Cape Bosberg, MacOwan 1907 (K, holo.!; GRA, iso.!).

Small plants not more than 350 mm high. Bulb

pyriform. Leaves linear, 130-300 (500) mm long,

1 5^40 mm broad, margin undulate. Inflorescence

25-40 mm diam. with 30-50 compactly arranged

flowers. Peduncle clavate, 60-130 (230) mm long.

Scape 30-70 mm long. Coma with 13-20 bracts;

bracts lanceolate, 24-80 mm long, 20-35 mm diam.

Pedicel 2-5 mm long. Perianth segments 6-8 mm long,

4 mm diam. Gynoecium with yellow ovary and white

style just after anthesis. Capsule with transparent

membranous pericarp which is often slightly inflated.

Figs 9 & 1 1.

Subsp. amaryllidifolia usually grows in groups

between rocks on mountain slopes of the western

Orange Free State and eastern Cape Province.

O.F.S. — 2727 (Kroonstad): Valsrivier (-AC), Pont 415

(PRE). 2827 (Senekal): “Bell's Pass” (-AC) Reyneke 318

(PRU). 2925 (Jagersfontein): Fauresmith, koppie in southern

part of Botanical Reserve (-CB), Verdoorn s.n., 2209 (PRE),

Reyneke 107 (PRU). 2927 (Maseru): Kommissiepoort (-AD),

Tylden s.n. (NBG).

Cape. — 2823 (Griekwastad): Griekwastad (-CC), Karree-

fontein, Wilman 2147 (GRA). 3026 (Aliwal North): Aliwal

North (-DA), Elandshoek, Bolus 270 (BOL). 3125 (Steyns-

burg): Grootfontein College of Agriculture (-AC), Theron 1057

(PRE), Hofmeyer (-DB), False Karoo, Acocks 16336 (PRE).

3323 (Willowmore): Uniondale Poort (-CA), Acocks 20325

(PRE). 3326 (Grahamstown): Grahamstown (-BC), Cherry

941 (GRA), Dyer 420 (GRA & PRE).

(c) subsp. clavata {Bak.) Reyneke, stat. nov.

E. clavata Bak. in Saunders's Ref. Bot. 4: t.238 (1871);

Lucas & Pike, Wild Flow. Witwatersrand (1971). Type: Orange

Free State, Cooper 1196 (K, holo.!; NU, iso.!).

E. regia sensu auct. sequent: Bak. in J. Linn. Soc. (Bot.)

13:225 (1873); Bak. in FI. Cap. 6:477^178 (1897); Turrill

in Gard. Chron. 8:75 (1921); Grey, Hardy Bulbs 3:236

(1938); Chitt. in Royal Hort. Soc. Diet. Gard. 2: 787 (1951);

Pam in J. Royal Hort. Soc. 69: 164 (1951); non L'Heritier.

E. macrophylla hort., Bak. in J. Linn. Soc. (Bot.) 13:225

(1873).

E. robusta Bak. in FI. Cap. 6: 477 (1897); Grey, Hardy

Bulbs 3:236 (1938); Chitt. in Royal Hort. Soc. Diet. Gard.

2: 787 (1951). Type: Natal near Koenigsberg, cultivated

specimen introduced by Dammann & Co. of Naples, 1894

(BOL, holo.!).

E. albomarginata Barnes in S.A. Gdng Country Life 20: 115

(1930); Grey, Hardy Bulbs 3:231 (1938). Type: Orange Free

State, Clarens, De Leeuw s.n. sub NBG 1876/26 (BOL; holo.!;

NBG, iso.!).

E. undulata sensu Letty, Wild Flow. Transv. 28 (1962);

Trauseld, Wild Flow. Natal Drakensberg 25 (1969).

142

NOTES ON AFRICAN PLANTS

Fig. 12. — The known distribution of Eucomis autumnalis

subsp. clavata.

Bulb dome-shaped. Leaves lanceolate to ovate,

sometimes linear, 150-450 (600) mm long, 60-130 mm

broad, margin undulate; cuticle on leaf-margin

dentate. Inflorescence 50-70 mm diam., 50-125

flowers which are exceptionally compactly arranged.

Peduncle mostly clavate, 70-120 (230) mm long.

Scape 70-130 mm long. Coma 130-180 mm diam.,

pendulous over inflorescence, 15-30 sterile bracts;

coma bracts ovate, 50-80 mm long, 30-50 mm diam.

Pedicel 3-8 (10) mm long. Perianth segments 12-17

mm long, 5-6 mm diam. Capsule with hard opaque

pericarp in which two layers can be distinguished.

Figs 9 & 12.

Subsp. clavata usually grows in groups in open

grassland and marshes. Recorded from Botswana,

Transvaal, Orange Free State, Swaziland and Natal.

Botswana. — 2525 (Mafeking): Lobatsi (-BA), Watt &

Breyer-Brandwijk 1886 (PRE).

Transvaal. — 2429 (Zebediela): Zebediela (-AD), summit of

mountain, Reyneke 101 (PRU). 2430 (Pilgrim’s Rest): Marieps-

kop (-BD), Reyneke 100 (PRU), Bedford foot-path. Van der

Schijff 4689, 6254 (PRE & PRU). 2526 (Zeerust): Zeerust

(-CA), Jenkins 11691 (PRE). 2528 (Pretoria): Pretoria (-CA),

Pretoria University Experimental farm, Codd 3692 (PRE);

14 miles south east of Pretoria (-CD), Codd 6214a (PRE).

2530 (Lydenburg): Belfast Forest Reserve (-CA), Reyneke 92

(PRU). 2531 (Komatipoort): Louw’s Creek (-CC), Brayshaw

170 (NU). 2626 (Klerksdorp): Lichtenburg (-AA), Hesse 1599

(PRE). 2628 (Johannesburg): Johannesburg (-AA), Holden

28053 (PRE). 2630 (Carolina): The Gem (-BC), Walker 1212

(PRE). 2730 (Vryheid): near Piet Retief (-BB), Compton

22335 (NBG).

O.F.S. — 2827 (Senekal): Rosendal (-BD), Rhebokkop,

Goossens 1883 (PRE). 2828 (Bethlehem): 6 miles south east of

Bethlehem (-AB), Scheepers 1883 (PRE).

Swaziland. — 2631 (Mbabane): Utukula rocks (-AC),

Compton 24782, 24865 (NBG). 2632 (Bela Vista): Lebombo

mountains (-CC), Strey 4661 (NH & PRE).

Natal. — 2729 (Volksrust): Charlestown (-BD), Smith 5755

(PRE). 2730 (Vryheid): Mooihoek (-AC), Devenish 403

(PRE); near Grootspruit (-BC), Strey 8058 (NH & PRE).

2828 (Bethlehem): Royal Natal National Park, Mont-aux-

Sources (-DB), Pardoe s.n. (PRE). 2829 (Harrismith): Van

Reenen (-AD), Bews 616 (NU), Van Reenen’s Pass,

Schweickerdt 952 (PRE); Cathedral Peak Forest Station (-CC),

Killick 1250, 1264, 1294 (PRE), Reyneke 50 (PRU). 2930

(Pietermaritzburg): Rietvlei (-AB), Greenwich Farm, Fry

5730 (PRE). 3030 (Port Shepstone): between Ixopo and

Umzinto (-AB), Werdermann & Oberdieck 1209 (PRE).

Lesotho. — 2828 (Bethlehem): Butha Buthe (-CC), Jacot

Gaillarmod 2140 (RUH), Leribe (-CC), Phillips 1227, 549

(NBG). 2927 (Maseru): Mafeteng (-CC), Watt & Breyer-

Brandwijk 2411 (PRE).

The author wishes to express his sincere gratitude

to Mrs A. A. Mauve of the National Herbarium,

Pretoria for generous help and advice. This work was

supported in part by the Research and Publications

Committee of the University of Pretoria as well as

the C.S.I.R.

REFERENCE

Reyneke, W. F., 1972. 'n Monografiese studie van die genus

Eucomis L' Herit. in Suid-Afrika. Unpublished M.Sc.

thesis, University of Pretoria.

W. F. Reyneke*

* Department of Botany, University of Pretoria, 0002

MYRTACEAE

THE IDENTITY OF EUGENIA WOODII

Ever since Eugenia woodii Diimmer (1912) was

described, some doubt has existed as to whether it is a

distinct species. In his original diagnosis Diimmer

distinguished E. woodii from E. zuluensis Diimmer by

its broadly elliptic or obovate leaves with the midrib

impressed above (prominently raised in E. zuluensis )

and the hypanthium covered with white appressed

hairs. He also noted that the ultimate branchlets

were slightly sericeous towards their apices, while

the axillary and apical buds were appressedly

pubescent.

Engler and Von Brehmer (1917) included E.

woodii in their enumeration of African Myrtaceae,

but did not mention some of the outstanding features

of this species. In fact, they described the branches

and the hypanthium as being glabrous, a character

not found on any of three specimens of the syntype.

Wood 1 32, investigated by the present author (Fig. 13).

Examination of the material of Eugenia in the

National Herbarium, Pretoria and the Natal

Herbarium, Durban, revealed that almost all the

specimens of E. woodii were misidentified. Most

material was placed under E. natalitia Sond., but

some specimens from the Transvaal were not

identified to species. Three specimens in the National

Herbarium, however, were correctly identified as

E. woodii by Dr G. J. H. Amshoff of Wageningen in

1960 or 1961 and have apparently escaped the

attention of subsequent workers. Dr Amshoff was

obviously aware of the correct identity of E. woodii,

as she had previously referred to the characteristically

2-3-flowered “pedicels” of this species in one of

her papers (Amshoff, 1958).

The confusion of E. woodii with E. natalitia

probably arose because of the superficial resemblance

in leaf shape. Because of the undue emphasis placed

upon leaf shape, the taxonomic significance of the

pubescence and other characters of E. woodii was

apparently either ignored or treated as trivial in

nature. This probably led to the inability to

distinguish between the two species, and the sub-

sequent incorporation of E. woodii as a synonym

under E. natalitia by Palmer & Pitman (1973).

This concept was taken one step further by White

(1977), who also added E. zuluensis as a synonym

and at the same time reduced E. natalitia to the

VARIOUS AUTHORS

143

Fig. 13. — Eugenia woodii. Part of the syntype. Wood 132 (BM),

showing the densely pubescent hypanthia (A) and charac-

teristically 3-flowered cymules (B).

status of subspecies, viz. E. capensis (Eckl. & Zeyh.)

Sond. subsp. natalitia (Sond.) F. White.

The present study has shown that E. woodii is a

very distinct species, which is not closely related to

E. natalitia. The comparative anatomy (Van Wyk,

1978) and external morphology revealed additional

distinguishing characters which had hitherto been

overlooked. Some of these are now included in

Table 1, which will enable one to distinguish most

specimens of E. natalitia and E. woodii.

It must be emphasized that morphological

characters in the genus Eugenia are extremely variable.

Taxonomic conclusions must therefore be based on

as many different characters as possible. The most

reliable characters that can be used to distinguish

between the two species are the nature of the disc

in bisexual flowers, the degree of pubescence of the

hypanthium, the nature of the seeds and the position

of the first-formed periderm in the stem. By using

the phloroglucinol/'hydrochloric acid test on freehand

sections of fresh or rehydrated material, the nature

of the periderm can be ascertained quickly and

easily.

Although E. woodii shows some similarity to E.

zuluensis Diimmer, E. albaner.sis Sond., E. zeyheri

Flarv. and E. pusilla N.E. Br., it seems to be more

closely related to the recently described E. erythro-

phylla Strey from southern Natal and Transkei, with

which it is sometimes confused. However, E. erythro-

phylla differs from E. woodii in its more coriaceous

and often larger leaves and the much larger flowers,

which are usually shorter pedicellate and not

aggregated in 3-flowered cymules. The young leaves

of E. erythrophy/la are often densely whitish pilose

on both sides with the upper surface of the lamina

usually becoming glabrous before the lower; the

reverse is found in E. woodii. The pubescence on the

hypanthium of E. erythrophy/la also tends to be

denser and more spreading than that of E. woodii.

The following amplified description of E. woodii is

based on the more abundant material now at hand.

Eugenia woodii Diimmer in Gdnrs’ Chron. ser.3,

52: 192 (1912); Engl. & Von Brehmer in Bot. Jb.

54:333 (1917); Von Breitenbach in Indig. Trees S.

Afr. 4: 845 (1965). Syntypes: Natal, without precise

locality, Gerrard 1643 (K!); between bushes near

Durban, Wood 132 (K!; BM!; PRE!).

Eugenia natalitia sensu Palmer & Pitman in Trees S. Afr.

3: 1669 (1973), pro parte quoad E. woodii ; sensu Compton

in FI. Swaziland: 396, pro parte quoad Compton 25175.

Eugenia capensis (Eckl. & Zeyh.) Sond. subsp. natalitia

(Sond.) F. White in Kirkia 10: 402 (1977), pro parte quoad

E. woodii; sensu Coates Palgrave in Trees S. Afr.: 689, pro

parte.

Tree up to 20 m high. Bark dark brown to grey

or whitish, corky and rough, typically flaking off in

irregular pieces. Branchlets reddish-brown to brown

becoming grey when mature, flattened, sparingly to

densely covered with appressed hairs, and glabrascent

later; buds densely pubescent; internodes (10-)

20-40 (-55) mm long. Leaves decussate, rarely in

threes, petiolate, lamina conspicuously bronze or

pinkish when young, becoming dark green and

shiny above, pale whitish green and dull below,

initially densely whitish pilose above, usually

sparingly pilose to glabrous beneath, soon becoming

glabrous with age, usually elliptic to broadly elliptic

or obovate to broadly obovate, 35-90 mm long,

20-60 mm wide, with apex bluntly or obtusely

cuspidate, tapering from about the middle into the

petiole, coriaceous, with revolute margin in dried

and fresh leaves; venation pinnately net veined,

midrib in dried leaves deeply concave above, strongly

elevated below, concave above and prominently

elevated below in fresh leaves; primary lateral veins

alternate or opposite, (6-) 8-12 (-14) pairs, spreading,

raised on both sides in dried leaves, slightly raised or

flat on both sides in fresh leaves, fused into a longi-

tudinal lobed marginal vein about 1-4 mm from the

margin of the lamina; tertiary veins slightly raised

on both sides in dried leaves, obscure in fresh ones;

petiole (4-) 5-8 (-10) mm long, ventrally canaliculate

and sometimes sparingly covered with appressed

hairs. Inflorescences rarely short 2-4-flowered racemes

mainly on the older wood, flowers usually solitary

of in 3-flowered cymules in the axils of bracts or

leaves on the first few nodes of the new seasons

growth. Staminate flowers usually with pedicels

(3-) 8-15 (-20) mm long; bracteoles 2, attached at

the base of the hypanthium, often absent in the

lateral flowers of a 3-flowered cymule, about as long

as the hypanthium, c. 0,4-0, 8 mm wide, lanceolate,

acute, usually densely appressed pubescent, eglandular

or with 1-4 glands. Sepals 4, subrotund with the

apices tending to be acute, 2 large, c. 1 ,5-2 mm long,

c. 2 mm wide, 2 small, c. 1-1 ,5 mm long, c. 1 ,5 mm

wide, outer surface sparingly to densely pubescent

and sparingly gland-dotted, margins usually ciliolate.

Petals 4, very rarely 5, white to greenish-white or

pinkish, usually elliptic, sometimes oblong or more

or less oval, c. 4-6 mm long, c. 3-4 mm wide, margins

usually ciliolate, eglandular or with a few obscure

glands. Disc with a central depression, surface even,

fleshy and usually densely pubescent. Stamens usually

20-30, arising from the disc; filaments of various

lengths, c. 3-6 mm long; anthers 2-thecous,

1x0,75 mm, all fertile. Hypanthium more or less

144

NOTES ON AFRICAN PLANTS

TABLE 1 . — Organographic and anatomical differences between Eugenia woodii and E. natalitia. Characters regarded as most significant

are marked with an asterisk

obconical, c. 1-2 mm long, densely appressed

pubescent. Ovary aborted; style rudimentary,

sometimes split into two, c. 0,5-1 mm long or absent;

stigma absent. Bisexual flowers with the pedicels,

bracteoles, sepals and petals as in staminate flowers.

Disc convex with an even surface, fleshy, usually

densely pubescent. Stamens usually 10-20, resembling

those of the staminate flowers. Hypanthium obconical,

c. 2 mm long, covered with whitish appressed hairs.

Ovary fused to the lower part of the hypanthium,

2- locular; ovules usually 2 per locule, 1 or 2

developing; style filiform, terete, glabrous or with a

few scattered hairs, c. 4-6 mm long; stigma small,

somewhat capitate, covered with small papillae.

Fruit a fleshy berry, at first yellow, becoming bright

red when ripe, obovoid to subglobose, c 15-25 mm

diam., glabrescent with persistent calyx lobes at the

apex; flesh of pericarp reported to be cream-coloured.

Seed globose with a smooth surface; testa woody

and tough with a fibrous texture, c. 1 mm thick,

brownish; embryo with cotyledons partly fused,

apparently eglandular but sometimes with a few

obscure glands mainly associated with the radicular

protuberance. Fig. 14.

E. woodii occurs as a tree in forest and associated

woodland in Natal, Transvaal and Swaziland probably

extending into Mozambique. It is locally common in

some localities, especially the forests of the north and

north-eastern Transvaal where it often occurs in

association with E. natalitia Sond. Flowering takes

place mainly from September to November.

For the present I am referring to E. woodii all the

specimens (mainly from PRE and PRU) cited below.

Unfortunately most collections are without bisexual

flowers or fruits, therefore these identifications must

be considered as tentative. There are indications

that a thorough study of the fresh fruits of E. woodii

may eventually prove that the species can be separated

into several infraspecific taxa.

Transvaal. — 2229 (Waterpoort): Wylliespoort (-DD), Van

Wyk 903 ; 904 ; 905 (PRU). 2230 (Messina): Entabeni (-CC),

Poynton s.n. sub PRE 50706 (PRE); Tate Vondo Forest Reserve

(-CD), Hemm 22 (PRE). 2329 (Petersburg): Lejuma near

Louis Trichardt (-AB), De Winter 6003 A; Hanglip Forest

Reserve (-BB), Poynton s.n. sub PRE 50630 (PRE), Van Wyk

907 (PRU). 2430 (Pilgrim’s Rest): 10 km from the Ofcolaco-

Trichardtsdal junction on the road to the Downs (-AA), Van

Wyk 2168 (PRU); Cyprus Farm (-AB), Renny 182; 226 ; 245

(PRE); Van Wyk 2165 (PRU); Welgevonden Forest Reserve

(-DB), Loock s.n. sub PRE 57403 (PRE); Blydepoort Nature

Reserve (-DB), Botha 1972 (PRU; PUC); Van Wyk 825

PRU); Mariepskop, near Reitz’s grave (-DB), Van der Schijff

6013 (PRE; PRU); Mariepskop, Blyde River picnic spot

(-DB), Van der Schijff 6091, 6394A (PRE; PRU); Van Wyk

2146; 2150 (PRU); Lothian Forest near Bushbuck Ridge

(-DD), Forest Officer 35 (PRE). Grid. ref. unknown: Sout-

pansberg Mountains, Poynton s.n. sub PRE 50653 (PRE).

VARIOUS AUTHORS

145

Fig. 14. — Eugenia woodii com-

pared with E. natalitia. E.

woodii : 1, leafy twig with

male flowers, x 0,5; 2, longi-

tudinal section of male

flower, x 6 ; 3, bisexual flower

with front petals and stamens

removed, x6; 4, disc with

calyx (other floral parts

removed), x 6. E. natalitia : 5,

bisexual flower with front

petals and stamens removed,

x6; 6, disc with calyx (other

floral parts removed), x6.

(1 from Renny 245; 2 from

Van Wyk 2146; 3 and 4 from

Van Wyk 2416; 5 and 6 from

Van Wyk 1119).

Swaziland. — 2631 (Mbabane): Mbuluzi Falls (-AC),

Compton 25175 (PRE); 4 km NE of Mbabane (-AC), Kemp

1048 (PRE); c. 1,5 km NW of Mbabane (-AC), Miller SI 138

(PRE); Sibanyone Hill (-CA), Miller S/264 (PRE).

Natal. — 2632 (Bela Vista): Amanzimnyana, 10 km E of

Maputa (-DD), De Winter & Vahrmeijer 8605 (PRE), near

Kosi Bay Nature Reserve (-DD), Edwards 2553 (PRE), Kosi

River (-DD), Moll & Strey 3833 (PRE); N bank of Nswa-

manzi River, near Mhlange Lake (-DD), Tinley 328 (PRE).

2731 (Louwsburg): Sokosoko Forest (-DC), Gerstner 4909

(PRE). 2732 (Ubombo): Gwalaweni Forest (-AA), Botha &

Van Wyk 949 ; 1122 (PRU); Vahrmeijer & Hardy 1672 (PRE);

Sibayi Dune Forest (-BC) Sibayi Project 327 (PRE)' Venter

5812 (PRU); Ngoboseleni Lake (-DA), Ross & Moll 5074

(NH: PRE). 2831 (Nkandla): emGangado (-BB), Gerstner

5031 (PRE); Eshowe (-CD), Thode A1237 (NH; PRE). 2832

(Mtubatuba): Mapelan Forest (-AD), Venter 5573 (PRU);

Banghazi Lake (-BA), Venter 5700 (PRU); Enseleni Nature

Reserve near Richards Bay (-CC), Venter 5913 ; 5914; 6099

(PRU). 2930 (Pietermaritzburg): near Durban (-DD?), Wood

132 (K; BM; PRE); Westville, Palmiet Nature Reserve (-DD),

Ward 8207 (PRE). 2931 (Stanger): King Hamlyn’s Farm,

Darnall (-AD), Moll 3611; 5503 (NH; PRE). 3030 (Port

Shepstone): Isipingo Beach (-BB), Ward 1000 (PRE).

Locality unknown. — Cultivated plants on the campus of

the University of Pretoria, Van Wyk 2416; 2419 (PRU).

Over the whole of its distribution range E. woodii

shows considerable variation in leaf shape and leaves

from more northerly plants are often larger than those

from the south. Specimens from open woodland

also possess smaller and more coriaceous leaves and

tend to be more shrubby than their forest counterparts.

The fruits of E. woodii are edible and reporetd to

have a pleasant taste. They are preferred to those of

E. natalitia , which are less fleshy and have a mealy

after-taste.

Common names: iJobe (Sw); umBomvane (Z);

’stawatawane (V). However, some of these names are

also recorded for E. zuluensis and E. natalitia.

REFERENCES

Amshoff, G. J. H., 1958. Notes on Myrtaceae. VII Myrtaceae

of French Equatorial Africa. Acta bot. neerl. 7: 53-58.

Dummer, R. D., 1912. Eugenias of South Africa. Gdnrs' Chron.

Ser. 3, 52.

Engler, A. & Von Brehmer, W., 1917. Myrtaceae africanae.

Bot. Jb. 54: 229-341.

Palmer, E. & Pitman, N., 1973. Trees of Southern Africa

III. Cape Town: A. A. Balkema.

Van Wyk, A. E., 1978. ’« Taksonomies-anatomiese ondersoek

van verteenwoordigers van die genus Eugenia L. ( Myr-

taceae) in Suid-Afrika. Unpublished M.Sc. thesis, Pot-

chefstroom University for C.H.E.

White, F., 1977. Some new taxa in African Myrtaceae. Kirkia

10,2: 401-404.

A. E. van Wyk*

* H. G. W. J. Schweickerdt Herbarium, Department of

Botany, University of Pretoria, Pretoria, 0002.

146

NOTES ON AFRICAN PLANTS

VITACEAE

A NEW SPECIES OF RHOICISSUS FROM NATAL

Rhoicissus sessilifolia Retief, sp. nov., R. rhom-

boidea (E. Mey. ex Harv.) Planch, affinis, sed foliis

sessilibus et petiolulo glabro differt.

Planta scandens caulo verrucato; cirrhis glabris.

Folia spiraliter disposita, sessilia, glabra, 3-foliolata;

foliolis terminalis petiolulatisque distinctis petiolulis

10-20 (27) mm longis, lamina obovata ad elliptica,

(35-) 43-70 (-80) mm longa, 19-32 (-40) mm late,

apice acuta vel obtusa, base cuneata, margine integro

vel 2-3 (aliquando 1 ) dentato in latero uno vel utrinque ;

foliolis lateralibus petiolulis brevioribus, petiolulis

(1,2-) 2-3,5 (6) mm longis, lamina ovata, 20-35

(-47) mm longa, 12,5-35 mm late, acuminata, base

asymmetrica, margine integro vel 2-3 (aliquando 1)

dentato in latere uno vel utrinque; pagina inferiore

foliolorum omnium venis majoribus prominentibus

formantibus venatiorem conspicue reticulatum;

acarodomatiis interdum in axillis venarum inferarum

lateralium. Inflorescentia cyma, folia opposita vel in

cirrhis, partibus juvenibus pubescentibus. Calyx

cupulatus, Petala 6, glabra, eburnea. Stamina 6,

supra ovarium flexa, antheris basifixis, oblongis,

0,5 mm longis. Ovarium in disco immersum; stylo

lineare; stigmate simplice, obtuso. Fructus bacca

ovoidea, 6-9 mm longus, 8-10 mm latus, ruber

juventute, atropurpurascens maturitate. Semina

globosa c. 5-7 mm longa, 5-6 mm lata, rugosa, sulco

longitudinali, latere uno complanato ubi semina

binatim.

Type. — Natal, 2732 (Ubombo): eastern shores of

Lake Sibayi (-BC), Moll & Nel 5619 (PRE, holo.)

Climber with warty stems; tendrils glabrous.

Leaves spirally arranged, sessile, glabrous, 3-foliolate ;

terminal leaflets distinctly petioled, petiolules 10-20

(27) mm long, lamina obovate to elliptic, (35-)

43-70 (-80) mm long, 19-32 (-40) mm broad; lateral

leaflets with shorter petiolules, petiolules (1,2) 2-3,

5 (6) mm long, lamina ovate 20-35 (-47) mm long,

12,5-35 mm broad, acuminate, base asymmetrical,

margin entire or with 2-3 (sometimrs 1) dentations

on both sides, under surface of all lei fl :ts with major

veins raised to form a conspict ou ;ly reticulate

venation; acarodomatia sometimes present in axils

of lower lateral veins. Inflorescence a cyme, leaf-

opposed or on tendrils, young parts hairy. Calyx

cupshaped. Petals 6, glabrous, creamy white. Stamens

6, bending over ovary, anthers basifixed, oblong,

0,5 mm long. Ovary immersed in a disc; style linear;

stigma simple, obtuse. Fruit an ovoid berry, 6-9 mm

long, 8-10 mm broad, red when young, purplish

black when ripe, Seeds globose, c. 5-7 mm long,

5-6 mm broad, rugose, with a longitudinal furrow,

one side flattened when seeds in pairs. Fig. 15.

Natal. — 2632 (Bella Vista): coastal dune forest north of

Kosi Bay (-DD), Stephen & Van Graan 1295. 2732 (Ubombo):

Mandozi Forest, Lake Sibayi (-BC), Vahrmeijer & Hardy 1606 ;

eastern shores of Lake Sibayi (-BC), Moll & Nell 5619 ; Sor-

dwana Bay (-DA), Tinley451. 2831 (Nkandla): Umlalazi Nature

Reserve (-DD), Moll 4979; Guy & Ward 62. 2382 (Mtubatuba)

along Charters Creek, Lake St Lucia (-AB), De Winter 8742;

Richards Bay (-CA), Strey 9910; mesic forests near quarry

(village area), Richards Bay (-CC), Venter 5549.

R. sessilifolia is endemic to Natal, occurring only

Fig. 15.— Rhoicissus sessilifolia ( Moll & Nel 5619, PRE, holo- in a restricted area (Fig. 16) of coastal dune forest,

type). ’ The first record of the species in the National

VARIOUS AUTHORS

147

Fig. 17. — Rhoicissus sessilifolia. 1, leaf, x 1 ; 2, part of a branch with an inflorescence on a tendril, x0,5.

Herbarium was from the Umlalazi Nature Reserve,

collected during January 1959 ( Guy & Ward 62).

For nearly twenty years this remarkable species of

Rhoicissus has been known, but never correctly

assessed until now. All the material cited above has

either been identified as R. rhomboidea or R. revoilii

or placed amongst the unnamed specimens. Although

R. rhomboidea occurs in the distribution area of

R. sessilifolia , it can be distinguished from the latter

species by the distinctive petioles (sometimes up to

35 mm) and long reddish brown hairs on the stems,

petioles and peduncles. R. revoilii is, in addition a

woodland species, often recorded from rocky

hillsides.

R. sessilifolia differs from all the other known

species of the genus in having sessile compound

leaves (Fig. 17). The tendrils of R. sessilifolia usually

consist of two internodes with an inflorescence often

developing at the node. This is also the case in several

other species of the genus Rhoicissus. According to

various authors [see Wild & Drumm. in FI. Zamb.

2:493 (1963), Dyer, Gen. 1:350 (1975)], the genus

Ampelocissus Planch, differs from Rhoicissus in

having tendrils which bear inflorescences. The

presence or absence of inflorescences on tendrils

cannot therefore be used to distinguish between

Ampelocissus and Rhoicissus.

Elizabeth Retief

Bothalia 13,1 & 2: 149-160 (1980)

Encoding the National Herbarium (PRE) for computerised

information retrieval

J. W. MORRIS*

ABSTRACT

This paper describes the operation of encoding information from 470 000 extant herbarium labels for

accession to a computer data bank with a view to aiding curators and systems analysts who may wish to apply

similar procedures in their herbaria. The preparation of the herbarium, encoding procedures used, verification

of encoded information, costs and remuneration of encoders are described. A complete list of items encoded

and their code equivalents is given. Some of the difficulties faced during the operation are discussed and the

success of the project to date is evaluated briefly. Its success is attributed to a bonus incentive scheme and

availability of a comprehensive instruction manual.

RlSUMt

ENCODA GE DE L 'HERBIER NA TIONAL (PRE) PO UR EXTRA CTION DE L 'INFORM A TION PAR

ORDINATEUR

On decrit ici Vencodage de V information contenue dans 470 000 etiquettes d’herbier pour leur incorporation

a une banque de donnees d'ordinateur; cette publication vise a aider les conservateurs et analystes qui souhaiteraient

appliquer des procedures similaires a leurs herbiers. La description comprend: la preparation de I’herbier, les

procedes d’encodage utilises, la verification de I’information codee, le cout de I’operation et la remuneration des

encodeurs. On donne une liste complete des elements encodes et de leurs equivalents dans le code. Certaines des

difficultes rencontrees en cours d'execution sont discutees et le succes du projet a ce jour est evalue brievement.

Ce succes est attribue a un plan d’ encouragement financier et a la disponibilite d’un manuel d’ instructions complet.

INTRODUCTION

One of the most expensive and logistically-com-

plicated parts of the operation to set up a computerised

information storage and retrieval system for the

National Herbarium was the encoding of the extant

470 000 specimen sheets. While it was acknowledged

that without inclusion of these specimens, known

below as the backlog, a data bank of new accessions

would have limited value, the complete encoding

of a herbarium of this size had not previously been

attempted (Morris, 1974). In this paper, the pro-

cedures developed for the encoding of backlog

material are described in detail, as it is considered

that these will be of use to all curators of collections

and systems analysts contemplating such an operation

in future. The benefits being accrued from com-

puterisation are such that it is hoped that other

herbaria will follow this lead (see Brenan et al., 1975).

Preparation of the herbarium, training of staff for

encoding and the development of a remuneration

system for the part-time bonus incentive encoding

scheme were some of the essential aspects which

required attention and are reported here. Another

reason for recording these procedures, and the

explanations accompanying items and their codes in

particular, is that users of the data bank will then

know the conventions adopted during coding.

Without this knowledge, they may have difficulty

formulating their queries for optimum results.

Backlog encoding was one aspect of the develop-

ment of an integral computer-based system for the

retrieval of information from specimen labels,

taxonomic data and literature references. The whole

system, as it exists at present, is described by Morris

& Glen (1978). The part described here deals only

with that information recorded by collectors on their

labels together with the state of the specimen and

its scientific name, including only specimens collected

prior to the completion of the project reported here.

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag XI 01, Pretoria, 0001.

THE HERBARIUM AND ITS PREPARATION FOR

ENCODING

The National Herbarium is located in four wings

of the building housing the headquarters of the

Botanical Research Institute in Pretoria, South

Africa. Specimens are housed in some 500 cabinets,

each with 24 to 36 shelves, and are arranged in the

taxonomic sequence of Dyer (1975, 1976). Specimens

are filed in blue, heavy paper species covers which

are themselves contained in light card, brown genus

covers. Type specimens are generally located at the

start of a species and are filed in easily-identifiable

special folders with red edges. Within a species,

specimens are arranged by province and country

with the oldest specimen from a province or country

at the top and the newest accession at the bottom.

Only one province or country is contained in a blue

cover and the province or country is written on the

outside of the cover. It was decided that all material of

indigenous taxa from the Flora of Southern Africa

Area (Ross et al, 1977) as well as all African type

material and photographs of African type material,

when the actual specimen was housed elsewhere,

would be encoded. Specimens having labels with

fewer than two items of information were not encoded.

A twelve-digit code number, consisting of seven

digits for the genus and five digits for the species

was developed (Morris & Leistner, 1975; Morris &

Glen, 1978). A computer card was punched for each

valid combination and four lists of codes, one for

each wing of the herbarium, were printed by computer.

It was found that there were more than 17 000

Southern African species names included in these

master lists. Once these lists had been checked and

corrected they were used to number species and

genera folders. All name changes accepted by the

herbarium from that time were then entered by hand

on these lists.

Before encoding commenced in a wing, curatorial

assistants paged through the entire collection, checking

that all the herbarium sheets within a species cover

150 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

were of the same taxon and that the specimens were

correctly identified to genus, species and sub-specific

epithet (where appropriate). Where a name could

not be assigned, assistants ensured that the unidentified

material within a folder was all of the same taxon.

The species code number was written on the outside

of the species covers and the genus code number was

written on the outside of the genus covers.

All herbarium cabinets were numbered by wing

and cabinet to facilitate replacement of specimens

after encoding.

ENCODING PROCEDURES, VERIFICATION AND

REMUNERATION OF ENCODERS

Introduction

Once it had been decided that the backlog housed

at PRE would be encoded, the Workstudy Section

of the Department of Agricultural Technical Services

together with officers from the Botanical Research

Institute, investigated ways in which the task could

be efficiently, rapidly and cheaply completed (Pieters,

1974). Requirements of the operation were that it

was to be carried out within a relatively short period,

say two to three years, and that minimal disturbance

be caused to routine herbarium curatorial activities.

The procedure outlined below was considered

superior to the alternative of appointing a smaller

number of encoders to work during office hours

without a bonus incentive. Direct encoding from

specimens onto computer punch cards by means of

IBM 029 punches and onto magnetic tape by means

of Olivetti key-punch stations were also tried but

found to be inefficient, relatively expensive and

encoded information was difficult to verify in com-

parison with encoding onto data sheets and sub-

mitting batches of data sheets for punching as a

separate task.

Procedures, verification and remuneration

Twenty-eight people were employed to carry out

the encoding operation for up to four hours each

evening, Mondays to Thursdays. Of these, five

teams, each consisting of four encoders and a team

checker, completed and checked data sheets,

respectively. Two bonus controllers were responsible

for vetting a 10 per cent sample of the completed

data sheets and a supervisor was responsible for

training new encoders, interpreting the encoding

instructions written by Morris & Du Toit (1976),

when necessary, deciphering unreadable handwriting

on labels, translating foreign-language labels and

resolving differences of opinion between encoding

teams and bonus controllers. Staff consisted of

volunteers drawn from the Public Service and no

botanical background was expected, except in the

case of the supervisor who was expected to have

had experience with the project and to have taxonomic

knowledge.

Although most of the encoders had no prior

botanical background, it was found that, in general,

they could be trained within 12 hours (three evenings).

Clear, written instructions were a prerequisite for

both training and the bonus incentive and penalty

scheme and as a result a handbook, which went

through a number of editions (Morris & Du Toit,

1976), was compiled. Such a manual also ensured that

encoding standards and conventions were main-

tained throughout the period. Labels in Latin and

German were particular problems as they were next

most common after English and Afrikaans and most

encoders were not familiar with these languages.

Lists of words commonly occurring on labels with

their meanings and code equivalents (see below)

were drawn up for use by encoders (Tables 1 & 2).

These ad hoc lists should not be considered as definitive

translations (the ablative case of nouns, for example,

is often given as the most common case occurring

on labels was usually listed) but indicate the use

made of such aids by encoders.

Another problem faced by encoders was that of

illegible handwriting and obvious copy typing errors.

Where encoders could not decipher writing, the

supervisor was asked to assist. In some cases, words

could not be deciphered and were not encoded.

TABLE 1.— Alphabetical list of Latin terms and their English

equivalents and numeric codes. Nouns are given in the

case (often ablative) in which they are most often encoun-

tered on labels

ad ripam— on river bank (moisture regime 07)

ad rivulum= on stream bank (moisture regime 07)

ad viam= at road (biotic effect 06)

albilalbus= white (colour)

anno= year (date of collection)

apertis= open (not coded except apertis «/vae=vegetation 5)

arenosus= sandy (soil type 02, substrate 1)

argillaris=white clay (soil type 04, soil colour A 1)

aridulus=slightly dry soil (substrate 1)

aridus= dry soil (substrate 1)

campester= flat area (marsh) (moisture regime 05)

capensis planitie==Qape Flats (locality)

CBS=C.B.Esp. (=erance)=Cape of Good Hope (locality)

c/>ca=about (not coded)

c//vwj=slope (not coded)

clivus supra= upper slopes (not coded)

collibuslcollinus= hill (not coded)

convalle=n\eT valley (not coded)

corotla= part of flower (flower colour)

corona= part of a flower (flower colour)

rfec/;w'.v=downward slope (not coded)

ericifruteceta=fynbos (vegetation 8, Veld type 69)

ericetis= fynbos (vegetation 8, Veld type 69)

/?av/=yellow (colour)

7?ore=flower (possibly flower colour)

fiumen=nvtv (moisture regime 08)

fructus/fructificatio= fruit (possibly fruit colour)

fruticis=scrub (vegetation 6)

frutex= shrub (life form 02)

fusci rubi= dark red (colour)

graminosis= grass (vegetation 3)

humidus= damp ground (marsh) (not coded)

lactei= white (colour)

lapidos(is)— rocky (substrate 2)

/e^.=collector (collector)

litor{al)ibus= sea shore (moisture regime 12)

/«re/=yellow (colour)

luteus= clay (soil type 04, substrate 1)

mar is— sea (moisture regime 12)

maritimus—%ea shore (moisture regime 12)

montanus= mountain (not coded)

wo«n7w«=mountain (not coded unless part of locality)

wo«f/s=mountain (not coded)

occidentalis=western (possibly locality)

orientalis=e astern (possibly locality)

paludibus=( in) marsh (moisture regime 05)

paludosus=marshy (moisture regime 05)

Planitie=Rats (possibly locality)

planitie=f\at (slope 1)

pratis= meadow (vegetation 3)

purpurei=purp\e (red) (colour)

ripis=f\\ev bank (moisture regime 07)

r/v«/«i=stream (moisture regime 08)

rubi= red (colour)

sabulosis=marsh (moisture regime 05)

ia/to5/j:=bushy rocky mountain (vegetation 4, substrate 2)

jaxoi/i=rocky (substrate 2)

j//vae=forest (vegetation 7)

y//v;.s=forest (vegetation 7)

5.«.=no collector’s number — assign a PRE number

suffrutex=sbv\ib (life form 02)

Tabularis=Tab\e (Mountain) (locality)

umbrosis= shady area (not coded)

virides— green (colour)

J. W. MORRIS

151

TABLE 2.— Alphabetical list of German terms and their

English equivalents and numeric codes

ausdauernd=ptTtrm\dL\ (annual etc. 4)

Felsen= rocks (substrate 3)

<7erc>7/=gravel (substrate 1, soil type 1)

Glimmerschiefer= mica slate (rocky) (substrate 2)

haeufig= abundant (abundance 7)

Halbstrauch= halfshrub (life form 03)

Kies= gravel (substrate 1, soil type 1)

/ac/z.y=salmon pink (colour)

£>«e//e=fountain (moisture regime 07)

7?/««,ra/=watercourse (moisture regime 10)

rosa= pink (colour)

S'e«fce=depression (moisture regime 03)

5/aWe=shrub (life form 02)

Sumpf= marsh (moisture regime 05)

7b«=clay (substrate 1, soil type 04)

Tor=gate/poort (not coded unless part of locality)

B//we=grassland (vegetation 3)

zart= soft (not coded)

zwischen= between (not coded unless part of locality)

Encoders were further instructed to correct obvious

typographical errors.

Encoders collected specimens from the herbarium

cabinets, a shelf of specimens at a time. For each

shelf a Daily Production Sheet (Fig. 1) was completed

which accompanied the specimens, held in a parcel by

means of a rubber band, through the process until

the specimens were returned to the cabinet after

encoding and checking. As some shelves required

more than one evening to encode, provision was made

on the Sheet for encoding to cover as many as four

evenings. Most items required to be entered on the

Sheet do not require explanation. New encoders

were allowed 12 hours in which to work at their

own speed without being financially penalised for

errors and speed. During this period they were

considered as ‘learners’ (see Fig. 1). Where more

than one specimen was mounted on a sheet, a coding

form for each specimen was completed. Information

from all the labels attached to a specimen were used

and in cases of contradiction, information from the

apparently-oldest label was favoured.

After being encoded, the specimens, together

with their encoding forms, were passed to the team

checker. The checker was responsible for checking

all the encoded information and correcting any

discovered errors. No penalty was incurred for errors

discovered and corrected by the team checker. The

checker also rubber-stamped each specimen with a

small ‘encoded’ stamp.

From the team checker, specimens were submitted

to the two bonus controllers. They took a sample

selected at random of 10 per cent of the encoding

forms and made a careful comparison of the speci-

mens with the completed data sheets. Errors dis-

covered were of two degrees of severity: full errors

(being such things as date of collection omitted or

province being incorrectly encoded) and minor errors,

counting as one fifth of a full error. A list of full

errors was drawn up and it was understood that all

other errors counted for one fifth of an error. The

number of specimens checked (10 and 3 in Fig. 1)

as well as the number of errors found (} and f) were

recorded on the Daily Production Sheets. These errors

were corrected by the bonus controllers. Specimens

were returned to the cabinets by the bonus control-

lers. Daily Production Sheets were collected until the

end of the encoding period, which was usually three

or four, four-evening weeks.

At the end of each encoding period the production

of each encoder and of each team was calculated and a

Monthly Team Summary was drawn up (Fig. 2) for

each team. For each encoder, the number of specimens

encoded and the number of errors found by the bonus

controllers was recorded. The norm was the number

of specimens expected from the encoder, based on

the number of hours worked and the hourly norm.

At the termination of encoding the norm was 16

specimens per hour. Initially it was 10 and it was

raised gradually by the supervisor in consultation with

the bonus controllers so that the average production

per cent of all teams remained in the range from 120

to 140. Production and error per cents were calculated

according to the formulae given in Fig. 2. Gross

bonus per cent was read from a degressive bonus

graph (Fig. 3), using the team’s total production

per cent. The penalty per cent (using Table 3) was

subtracted from the gross bonus per cent to give the

net bonus per cent for the team. Bonus controllers

made use of a table based on the degressive bonus

graph instead of the graph itself for greater speed

and accuracy. Although production and error per

cents were calculated individually, the team average

net bonus per cent was used to calculate payment

DAILY PRODUCTION SHEET

NAME

E King-

TEAM

3

027

o, £»

If Learner, mark

1 3

Fig. 1. — Example (slightly re-

duced) of completed Daily

Production Sheet.

Initials of

Coder

Initials of

Checker

152 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

MONTHLY TEAM SUMMARY

TEAM NUMBER

PERIOD

Ol )<? -/?7fcoa./3

PRODUCTION % = (SPECIMENS ENCODED/NORM) x 100

ERROR % = (ERRORS/NUMBER INSPECTED) X 100

Fig. 2. — Example of completed

Monthly Team Summary.

Fig. 3. — Degressive bonus graph for estimation of gross

bonus per cent from production per cent: (a) Linear

relationship, (b) degressive relationship used in this

project.

for all members of a team. Thus a fast, accurate

encoder raised the bonus per cent for the whole team

while a slow, inaccurate encoder had the reverse

effect. Individual statistics were calculated and

tabulated (Fig. 2) so that the supervisor could in-

vestigate slow or inaccurate encoding at monthly

intervals. The degressive bonus graph (Fig. 3) was

proposed by the Workstudy Section as a necessary

part of such a bonus incentive scheme (Pieters, 1974).

Beyond a production of 120 per cent, the team benefits

the employer to an ever increasing extent with this

relationship. The area between lines a and b (Fig. 3)

TABLE 3. — Penalty corresponding to error per cent. See text

for explanation

represent the profit of the employer at the expense of

the employee. Such a graph is used principally to

protect the health of the encoders (Pieters, pers comm.

1978) and to maintain the gross bonus at 20-30 per

cent. Bonus calculation per team and not individual is

another prerequisite of such a scheme. The penalties

corresponding to error per cent (Table 3) were devised

so that a small error had at most a very limited effect

on the net bonus while at higher error per cent levels

the net bonus was drastically reduced and finally no

bonus was paid for an error per cent greater than 10.

The scale was intended to encourage accurate encoding

above mere speed.

The number of hours worked by each encoder and

each team checker was entered daily on an Attendance

Register and Pay Sheet (Fig. 4). At the end of the

encoding period the hours worked were totalled and

payment calculated according to the formula given

in Fig. 4. Net bonus per cent was obtained from calcu-

lations described above. Basic hourly rates of payment

at the termination of encoding were: encoder: R2,50;

team checker: R3,20; bonus controller and super-

visor: R3,75. Only encoders and team checkers

participated in the bonus incentive scheme.

DATA CODING FORM AND KEY-PUNCHING OF DATA

An example of a completed encoding form is given

in Fig. 5. The form was specifically designed to expe-

dite the encoding operation and be suitable for key-

punching on 80-column cards. The first line of the

form was completed for the first sheet in each species

cover only and this information was duplicated to

following forms by computer until replaced by the

next occurrence of a first line. Information common to

all the specimens in a species folder, including identi-

J. W. MORRIS

153

Fig. 4. — Example (slightly re-

duced) of completed Attend-

ance Register and Pay Sheet.

ATTENDANCE REGISTER AND PAY SffPPrr

month |Trp«ur*>/ -February t<plo | team | 3

PERSONNEL NUMBER | £) 1007^X4-3/ j

SURNAME AND INITIALS £ _ K I W 6r

D => A

(A

HOURLY RATE

NO. OF HOURS

NET BONUS %

TOTAL PAY

| Rx,so »|

1 48 b|

I l+l, (a c[

| Rib9n D|

CERTIFIED CORRECT : BONUS CONTROLLER

Fig. 5. — Example (reduced from

A4 size) of completed Back-

log Coding Form.

I'.alflil hi.Vj',1!'

7 ,5ife.g i oo.&o,o,3P\o, s,»,

za

S(«ui ii?i

►- gJ j

(A.I.B^'c. 1.0-4) LXT 4 d°’m nInt0"^"00 A,'"ud* = < f, ff L> r«ord hj-J

[~~r.1q3|/,7lc>fel(,9|5,s| .3i9.sJ/ll3.Zi/.3l . .oloLz

’ 13. / i2J i/. / 1

I J! ilijfi

if Ji

iMt!

i Irin?

-W-

1091

m

Common ntm, Ltn|uaf«

fication of encoder, genus and species code numbers

and cabinet and shelf numbers, was included in this

line. Subsequent lines contained information unique

to the specimen and had to be completed for each

specimen.

Shortly after encoding commenced, a 240-character

per record key-to-tape data entry system replaced the

ibm 029 card punches. If it had been known that

such a system was due, a form with longer records and

no ‘auto dup’ fields would have been designed. It is

fortunate that it was not necessary to punch computer

cards throughout the project as more than 2 300 boxes

of cards would have been required and the task of

processing and storing them would have been much

more difficult than that of a few hundred magnetic

tapes.

Encoding forms were numbered consecutively by

the encoding supervisor by means of a numbering

machine (Fig. 5) and then stapled into bundles of

about 50. Each bundle was given a header form and

consecutive bundle number. Batches of 20 bundles

were submitted for key-punching and computer

processing.

Specimens were accessioned piecemeal to computer

disk packs in three groups of 150 000 and the remain-

der (about 60 000). The accessioning programme

edited each specimen and listed errors. Errors were

corrected by means of the same encoding form pre-

ceded by a different header form. Another pro-

gramme finally converted correct and corrected speci-

mens to the precis data base.

154 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

ITEMS AND THEIR CODE EQUIVALENTS

It was decided that as far as possible, items would

be given a numeric code (see Morris, 1974 for dis-

cussion). Exceptions were collectors’ names, common

names and locality names. Some descriptive text

would be entered in full but would not be searchable.

A list of items captured during the backlog encoding

is given in Table 4 and an example of a complete

TABLE 4. — Items encoded during backlog encoding

(U=uncoded, C=coded, T=text)

5. Quarter-degree grid

The latitude and longitude part of the code was

entered as it appeared on the specimen. The letters,

A, B, C and D, were converted to 1, 2, 3 and 4,

respectively, before encoding. Where half- and quarter-

degree codes were missing, a code of ‘0’ was used.

6. Latitude and Longitude

Seconds, if given, were rounded to minutes before

encoding. Where a range was given, the midpoint

was coded.

7. Region

Black States within South Africa were coded as

major or minor locations and for pragmatic reasons

were given the region code of the province of which

they were previously part. The codes used were:

1 Encoding date 26

2 Encoder 27

3 Taxon number 28

4 Cabinet and shelf number 29

5 Quarter-degree grid (C) 30

6 Latitude and longitude (U) 31

7 Region (C) 32

8 Date collected

9 Altitude and units (C) 33

10 Abundance (C) 34

1 1 State of the specimen (C) 35

12 Special record (C)

13 Type status (C) 36

14 Label language (C) 37

15 Substrate (C) 38

16 Moisture regime (C) 39

17 Slope (C) 40

18 Aspect (C) 41

19 Soil colour (A and B 42

horizons) (C) 43

20 Presence of grey mottles (C) 44

21 Soil type (C)

22 Biotic effects (C) 45

23 Vegetation type (C)

24 Veld type (Acocks, 1953) 46

25 Main flower colour (C)

Main fruit colour (C)

Life form (C)

Annual/perennial (C)

Woody/herbaceous (C)

Evergreen/deciduous (C)

Weed/encroacher (C)

Actually or potentially

used (C)

Collector’s number (U)

Height and units (C)

Diameter at breast height

and units (C)

Utilized by (C)

Economic property (C)

Collector’s name (U)

Major locality name (U)

Minor locality name (U)

Precise location (T)

Miscellaneous notes (T)

Economic narrative (T)

Common name (U) and

language (C)

Number of duplicate

sheets

Specimen or photograph

of specimen (C)

01 Angola

02 Botswana / Bechuana-

land/Ngamiland

03 Cape Province / Kaap-

provinsie

04 Lesotho/Basutoland/Ba-

soetoeland

05 Mozambique / Mosam-

biek

British Bechuanaland (usual-

ly Mafeking and surround-

ings)=03

Boesmanland (without men-

tion of Province or coun-

try )=03

Great Namaqualand=09

Klein Namaqualand=03

Maputoland=06

8. Date collected

07 Orange Free State /

Oranje- Vrystaat

08 Rhodesia/ Rhodesie

09 South West Africa/

Caprivi/Namibia

10 Swaziland

11 Transvaal

12 Not these

00 Unknown

Namaland=09

Namaqualand=03

n’Gamiland=02

Northern Ngamiland=09

Sekukuniland=l 1

Tembuland=03

Tongaland=06

Zululand=06

backlog coding form is given in Fig. 5. Codes used

under each item are listed below, together with notes

on their use and conventions used by the encoders,

where appropriate. The sequence is that of the en-

coding form.

06 Natal

Code 12 was used for other countries, in which case

the name of the country was given as a major or

minor location. The following region code conven-

tions were adopted :

1. Encoding date

Two-digit codes for the month and day of encoding

were recorded. The year of encoding was obtained

from the batch header form.

2. Encoder code

Each encoder was assigned a unique, consecutive,

three-digit code and each of the five encoding teams

had a one-digit code. Thus 0394 represents encoder

39 from team four.

The date on which the plant was collected was

encoded in full. Missing dates, or parts of dates, were

coded as ‘00’. In the case of a range being given, the

earlier date was coded. The date on which the plant

had flowered, according to the label, was used if the

specimen had flowers and no other date was given.

If the specimen was not flowering and only flowering

date was given, date collected was entered as

‘00000000’ (i.e. unknown) and flowering date was

given as a miscellaneous note.

3. Taxon number

The seven-digit genus code and four-digit species

code (see Morris & Glen, 1978) were taken from the

outsides of the folders containing the specimens.

An extra trailing zero was added to the species

number by computer on conversion to allow for the

future subdivision of an extant species into 100 new

specific and sub-specific categories instead of only 10.

The following special species codes were used:

9999 Tropical African type specimens and southern African

material identified to genus only

9998 Hybrids (some hybrids also coded ‘9999’)

9997 Cultivated, exotic species

Although a few were encoded, it was decided that in

general exotic species would be omitted.

4. Cabinet number

Cabinets in each of the wings were given consecu-

tive three-digit numbers, starting with ‘001’ in each

wing. Within each cabinet, shelves were numbered

from top to bottom and left to right with two-digit

numbers.

9. Altitude and units

Altitude was entered, right-justified. The mid-

point of an altitude range was coded. The codes T’ =

feet and ‘3’=metres were used for the units code.

10. Abundance

Four basic codes were used, each of which could

be prepositioned by ‘locally’. ‘Locally’, ‘scattered’ and

‘local’ without accompanying indication of abundance

were, however, not coded. The codes were:

1 rare/very occasionallscarce/uncommon/seldsaam/skaars

2 locally rar e/plaaslik seldsaam

3 occasional /infrequent/ toevallig/ matig/ nie volop me

4 locally occasional/very localized /plaaslik toevallig

5 common/fairly common/co-dominant/frequent/u/^emeen/

wydversprei/ volop

6 locally common/localized /plaaslik algemeen

7 abundant/frequently abundant/very frequent/6a/c volop

8 locally abundant /plaaslik baie volop

Additional codes for very rare and locally very rare

have been added subsequently and distribution (i.e.

local or widespread) has been separated from abund-

ance. (In all cases where alterations have been made

J. W. MORRIS

155

to code lists the necessary alterations have been made

to the data bank and the ‘new’ codes will be used

exclusively in future.)

1 1 . State of specimen

The presence and maturity of flowers, fruits, roots

and leaves were coded after inspection of the speci-

men. The code for ‘present’ was not used if the encoder

could determine whether the organ was mature or

immature. For specimens of grasses and other families

with reduced or small flowers, flowers and fruits were

coded ‘present’ if either was visible. State of specimen

of bryophytes was always coded ‘0000’. Side roots

and root hairs had to be present before roots were

coded ‘present’, ‘mature’ or ‘immature’. Photographs

of types were always coded ‘0000’. The following

codes were used for the state of flowers, fruits, roots

and leaves:

1 absent/ afwesig

2 immatur e/jonk/onvolwasse

3 mature./ volledig ontwikkel / volgroei

4 present/ teenwoordig

12. Special record

A code was allocated for specimens collected for a

special purpose. A rubber stamp or special label on

the herbarium sheet usually indicated a special record.

Codes allocated were :

000 not special record jnie spesiale versameling nie

001 Eland food study / Elandkosstudie

002 SKF alkaloid study / SKF alkaloiedstudie

003 Lamziekte survey /Lamziekte opname

004 Stijfsiekte survey /Stijfsiekte opname

005 Dunsiekte survey /Dunsiekte opname

006 Bovine staggers survey / Stootsiekte opname

007 Pollen studies (UOFS )/Stuifmeel studies ( UOVS )

008 Anatomy study /Anatomiese studie

009 CSIR alkaloid survey /WNNR alkaloied opname

010 Bushmen food study / Boesmankosstudie

01 1 Cancer research/ Kankernavorsing

012 Economic plants of the Kung Bushman /Ekonomiese

plante van die Kung-Boesman

013 Study of Tswana names and uses/ Studie van Tswana

name en gebruikte

13. Type status

All specimens found in type covers as well as

specimens with a note of type status in ordinary

folders were allocated a type status code. Where more

than one type status was given for a specimen, the

lowest code number was assigned, except for ‘isotype’

which was coded in preference to ‘neotype’, ‘lecto-

type’ and ‘type’. The codes were:

0 no status/ geen status nie 4 neotyp e/neo-tipe

1 holotyp e/holo-tipe 5 lectotyp e/lekto-tipe

2 syntype/cotyp e/syn-tipe/ko- 6 type/ tipe

tipe 7 isotyp e/iso-tipe

3 paratyp e/para-tipe

14. Label language

The language in which the majority of the origint 1

label was written was coded as follows:

1 Afrikaans

2 English /Engels

3 Latin /Latyn

4 Dutch/JVeder/ands

5 French/Frans

6 German/ Duits

7 Italian /Italiaans

8 Portuguese/ Portugees

9 Spanish /Spoons

0 other/unknown /ander/on-

bekend

15. Substrate

This item was coded as follows (present code

numbers in parenthesis):

1 soil/mud/sandy flats /grond/modder/sandvlaktes (01)

2 stony soil/between rocks /klipperige grond (02)

3 bare rock I blootgestelde rots (03)

4 talus/scree (04)

5 cliff face/rock crevices/kranswand (05)

6 termite mound/termitaria/ termiethoop (06)

7 beach dunes/dune forest /kusduine/duinwoud (07)

8 desert dunes /woestynduine (08)

9 other/plant growing on another plant/ander/plant groei op

ander plant (10)

An additional code, ‘in water’ (09) has been added

subsequently. In the case of coding ‘other’, details

were entered under miscellaneous notes. Plants noted

to be growing in or on granite, dolomite, sandstone

or another rock type were coded ‘2’ and if the colour

of the rock was given, it was coded as the colour of

the B-horizon soil (see below). Sandy flats were coded

T’ for substrate, ‘02’ for soil type and T’ for slope.

Dune forest was also coded as vegetation ‘7’. Code T

(soil) was only used if one of the following words,

or their Afrikaans equivalents, occurred on the label:

clay, gravel, sand, sandy, soil, turf.

16. Moisture regime

This item was coded in conjunction with substrate,

using the following codes:

01 poorly drained so\\/ swak gedreineerde grond

02 well-drained soil /goed gedreineerde grond

03 pan/depression/edge of pan /holte

04 seepage area /syfergebied

05 marsb/swamplbog/vlei/moeras

06 floodplain /vloedvlakte

07 river/streambank/near river /rivier/spruitoewer

08 r'wer jstream/bum/ rivierj stroom

09 river or stream bed /rivier- of spruitbedding/ omurumba/

oshona

10 ditch/donga/furrow/water course/ sloot/spoelsloot

1 1 Xake/dam/webr/meer/dam/stuwal

12 estuary/sea/lagoon/river mouth/littoral/mangroves/s/raw^-

meer/ri viermond/ see

13 in water

14 otber/ander

Code ‘07’ was also used for ‘next to river’, ‘near

creek’, ‘riverine’ (except ‘riverine forest’ and ‘riverine

bush’, which were coded ‘7’ under vegetation only,

‘waterfall’ and ‘stream bank’. ‘Above river’ and place

names (e.g. Tugela Mouth) were not coded here. All

specimens coded ‘13’ have been given substrate code

‘09’ and code ‘13’ has been removed from moisture

regime.

17. Slope

The codes for slope were:

1 plain/flat/sandy flats /vlakte/gelykte

2 gentle/ effens

3 moderate//na%e

4 steep/s/e//

‘Plateau’ and ‘slope’ without qualifiers were not coded.

18. Aspect

Aspect was coded on the following eight-point

scale, provided it was clear that the collector was

referring to aspect and not location (i.e. south of . . .):

1 north/ noord 5 south /suid

2 north east /noordoos 6 south west /suidwes

3 east /oos 7 west /wes

4 south east /suidoos 8 north west/ noordwes

19. Soil colour

On very few occasions, soil colour was mentioned

on labels. Where the soil horizon was not named,

the colour was coded as that of the A-horizon. Rock

colour was coded as that of the B-horizon. The codes

used were:

1 white/ wit

2 light grey/% grys/vaal

3 greylgrey-brown/beige/iawn/grys/grysbruin

4 y e\\ow-br own/ geelbruin

5 red /rooi

6 black /swart

156 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

20. Grey mottles

A code for the presence of grey mottles within the

soil profile was included on the advice of soil scientists

but as this information was virtually never given by

collectors, the item was not recorded after about

50 000 specimens had been encoded. It is now included

in the data bank as code ‘11’ of soil type.

21. Soil type

The soil type was described by up to four of the

following codes:

01 gravel/shale/gritty/^raw’w sandjgruisgrond

02 sand/sandy/jaw/

03 loam /teem

04 clay /turf/ klei/ turf

05 humus-rich/peaty /humus-vrugbaar

06 salty /brak

07 on calcrete/limestone/calcareous soil /kalkklip

08 on laterite/ferricrete/oiA//>

09 disturbed soil /versteurde grond

10 eroded /geero deer

11 other lander (12)

‘Grey mottles’ (11) were subsequently added and

code ‘11’ was changed to ‘12’. ‘Sandveld’, ‘sand

flats’ and ‘sand forest’ were coded ‘02’ ; ‘kalkveld’ was

coded ‘07’ under this item and vegetation type was

coded ‘4’. ‘Sand on calcrete’ was coded ‘02’ and ‘07’.

‘Dolerite soil’ and ‘alluvial soil’ were coded as sub-

strate T’ only. ‘Disturbed soil’, included as code ‘09’

of soil type was not coded again under biotic effects.

22. Biotic effects

Biotic effects noted on labels were coded by means

of up to three of the following codes :

01 cultivated land/ploughed! landerye

02 abandoned land/fallow/cw/a«</

03 planted pastur e/aangeplante weiding

04 plantation/plantation margin /p/antasie

05 garden/lawn/rwzw

06 roadsidc/railwayside/7a«^.s’ pad fangs treinspoor

07 heavily grazed / 1 ra rn p led /.y waar bewei/oorbeweiluitgetrap

08 recently burnt /onlangs gebrand

09 disturbed-other/ versteurd-ander

10 no effect seen/undisturbed/rested/^een invloed waar-

geneem nie

The following conventions were adopted in the

coding of biotic effects. A firebreak was not considered

as ‘08’ and disturbed soil was not ‘09’. In a National

Botanic Garden, or similar institution, code ‘05’ was

only used if it was stated that the specimen was culti-

vated. If the locality was a domestic garden, or the

locality could be assumed to be other than a National

Botanic Garden or similar institution then code ‘05’

was assigned even though the specimen was not

specifically said to be cultivated. Code ‘08’ was used

for burns of up to one year old, if the age of the burn

was given.

23. Vegetation

The following codes were used for describing the

vegetation in which the plant was collected:

1 desert/semi-desert/ woestyn

2 karoo /karooveld

3 grassland/sand veld Igrasveldl veld

4 savanna/bushveld/thornveld/tree veld/open veld/parkland/

grasveld met borne /kalkveld

5 woodlandlbushlboomveldlbos

6 scrub/thicket/among shrubs /digte bos/kreupelhout/ruigte

7 forest/sand forest/dune forest/riverine forest/forest margin/

woud

8 fynbos/heath/macchia/sclerophyll scrub Ifynbos

9 other lander

The term ‘bush’ was coded ‘5’ unless associated with a

measurement of height, in which case the term was

taken as a life form. If vegetation had been recently

burnt, the vegetation type was assumed to be grass-

land unless another vegetation type was given.

24. Veld type

The two-digit veld type number assigned by Acocks

(1953) was coded when given on the label.

25. Flower colour

The dominant flower colour, as noted on the label,

and not as observed on the dried specimen, was coded

as follows:

01 white/cream/ w/r/roow

02 grey/silver/^rys

03 yellow/saffron/^ee/

04 pink/rose/cerise/pz'ew/fc

05 orange/amber/oran/e

06 red/magenta/scarlet/maroon/rooi/persrooilskarlaken/wyn-

kleurlkarmosyn

07 mauve/purple/violet/heliotrope/lilac/vermilion/perc

08 blu e/blou

09 green /groen

10 brown/buff/Ara'ra

11 black/.swwr

26. Fruit colour

The flower colours given above were also used for

fruit colour. The codes were in each case greater by

20. Thus, white fruit was ‘21’ and black ‘31’.

27. Life form

The following codes were used :

01 tree/ boom

02 shrub/undershrub/suffrutex/stm'A:

03 dwarf shrub/semi-shrub/half shrub/dwergstruik/halfstruik

04 herb/forb/hemicryptophyte/&raYf

05 geophyte/bulb/bulbous plant /geofiet/bo/plant

06 epiphyte/ epifiet

07 climber / creeper / scandent / vine / twiner / liane / klim-

plant/ranker

08 parasit e/parasiet

09 succulent/semi-succulent/vetp/awf

10 aquatic/hydrophyte/ waterplant

Plants with succulent or fleshy leaves were not coded

‘09’ unless life form was also specified.

28. Annual etc.

The life cycle of the specimen was coded as follows:

1 annual! eenjarige

2 ephemeral lefemeer

3 biennial / 1 weejarige

4 perennial/perennial rootstock/ meerjarige

When ‘herbaceous’ was coupled with ‘woody root-

stock’ the specimen was not coded as ‘4’.

29. Woody etc.

This item was coded as follows:

1 woody /houtagtig

2 herbaceous! kruidagtig

‘Woody rootstock’ was not coded under this item. An

additional code, half-woody, has been added subse-

quently.

30. Evergreen etc.

This item was coded as follows:

1 evergreen/ immer groen

2 semi-deciduous/Aa// bladwisselend

3 deciduous/ bladwisselend

3 1 . Weed etc.

This item was coded as follows:

1 weed / ruderallonkruidjpuinhoopplant

2 encroacher /indringer

3 weed & encroacher /onkruid en indringer

J. W. MORRIS

157

An additional code for exotic plants has been added

subsequently.

32. Actual I potential use

Actual uses as noted on labels (e.g., used as a

poison) were separated from potential uses (e.g.,

could be a food plant) by means of this code. Use of

this item meant that utilized by and economic property

had to be entered as well. ‘Poison’ without further

elaboration was taken as an actual use. Codes were:

1 actually used/ werklik gebruik

2 potential us e/potensieel bruikbaar

33. Collector’s number

The collector’s number, exactly as given on the

label, was entered, left-justified. If no collector’s

number was given on the label, a PRE accession

number, prefixed by ‘PRE’ was used for this item. If

the specimen had a printed label of the Transvaal

Museum (TM), Herbarium of the University of the

Witwatersrand (MOSS), South African Museum

(SAM), National Botanic Gardens (NBG), Herba-

rium of the University of the Orange Free State

(BLF), Stellenbosch Unit of the National Herbarium

(STE), Durban Unit of the National Herbarium

(NH), Herbarium of the Forestry Department (FD)

or Bolus Herbarium (BOL) and the number was that

of the herbarium and not the collector, the ‘collector’s

number’ was prefixed with the letters given above in

parentheses. Where specimens had both a personal

collector’s number as well as a herbarium accession

number, the former was always used. If the initials

of the collector were given as part of the number,

they were omitted.

34. Height and units

The height of the plant, rounded to three digits if

necessary was coded. The mid-point of a range was

entered. When two or more distinct heights were

given, the greater height was used. Units, for both

height and diameter at breast height, were coded as

follows:

1 feet /voet 3 metres/ meter

2 inches /duim 4 centimetres /sentimeter

35. Diameter at breast height and units

The conventions noted under height also applied to

the coding of this item.

36. Utilized by

Up to five codes were used to describe the uses of

the plant. If more from one code was used then only

one economic property was allowed, and vice versa.

The following codes were applicable:

81 man /mens

82 stock/vce

83 cattl e/beeste

84 sheep/skape

85 goats /bokke

86 horses/ donkeys/ perde/donkies

87 other mamma\s/ game /ander soogdiere

88 birds /voels

89 fishes/reptiles/amphibians /visse/reptiele/paddas

89 fishes /reptiles /amphibians/ visse/reptiele/amfibiee

90 invertebrates (excluding bees)/ongewerweldes ( uitsluitende

bye)

91 honey bees /heuningbye

Codes ‘90’ and ‘91’ were subsequently reversed so

that ‘bees’ became ‘90’ and ‘other invertebrates’ ‘91’.

Unless otherwise specified, ‘eaten’ and ‘not eaten’ on

labels were assumed to apply to stock (code ‘82’)

and ‘poison’ was not coded under the utilized by

item unless qualified.

37. Economic property

Up to five of the following codes were used to

describe economic properties of specimens (present

code numbers in parentheses):

01 poison/^//(01)

02 poison : arrow /pylgif (02)

03 poison: criminal/#//: kriminele doeleindes (03)

04 medicinal /medisyne (06)

05 drug/ verdo wingsmiddel (07)

06 irritant/aWergy /prikkelmiddel/allergie (08)

07 tainting e.g. milk /bysmake aan melk ens. (09)

08 magic/ritual/roor kims/ritueel gebruik (10)

09 eaten/ palatable/^eeer (11)

10 not eaten /nie geeet nie (12)

11 beverage/ drank (13)

12 cordage/ tone /vlegwerk (14)

13 paper /papier (15)

14 cloth'mg/ klere (19)

15 structural/baskets/mats/broorns/^eirw/A: vir konstruksie/

mandjies/ matjies/ besems (20 and 21)

16 aromatic/snuff /aromaties/snuif (22)

17 cosmetic / grimering (23)

18 beads/ krale (24)

19 soap/ seep (26)

20 oil /olie (27)

21 gum/resm/ gom/harpuis (28)

22 dy e/kleurstof (29)

23 fuel /brandstof (30)

24 sand binder /sandbinder (31)

25 ground cover/lawn grass/ grondbedekking (32)

26 hedgel'M'mdbreak/heining/windskerm (33)

27 shade/ sk ad uwee (34)

28 garden ornamental /sierplant (35)

30 other — see miscellaneous notes /ander — sien algemene notas

30 other — see miscellaneous notes /ander — raadpleeg algemene

notas

Where more than one distinct economic note was given

for a specimen, the first was coded and the subsequent

notes entered in the economic property narrative

field only. The following codes have been added

subsequently:

medicine — internal (4) other building (18)

medicine — external (5) other decorative (25)

thatching (16) other horticultural (36)

timber (17) crop (37)

38. Collector(s )

In the case of one-person collectors, the surname,

followed by initials without full-stops were entered.

Where more than one collector’s name was given on

the label, surnames only were coded, each separated

by two blank spaces. Where no collector’s name was

given, the collector was coded as ‘PRE’. When a

specimen was collected by one person (usually a lay-

man) for another (usually a well-known botanist) and

both names appeared on the label, the name and

number of the botanist were encoded and the name of

the actual collector was given in a miscellaneous note.

39. Major location name

Difficulty was experienced by encoders with the

choice of major and minor location names. Pages of

examples were drawn up to illustrate some of the

many possibilities.

A major location name was only given if a minor

location name was also coded; otherwise the major

location name was entered in the minor location name

field. It was intended that a large geographical unit,

but not a region (see above) be entered. Examples are

mountain ranges and magisterial districts.

40. Minor location name

It was intended that a small geographical unit be

coded as the minor location name. Examples would

be towns and cities. In general, a minor location is

158 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

TABLE 5. — Alphabetical list of abbreviations used in major, minor and precise locality fields

situated within a major location. The collecting site

had to be within the arbitrary limit of 16 km (10 miles)

from the minor location or ‘near’ the minor location,

if an exact distance was not given. Farm names were

not used for major and minor locations except in

South West Africa and Botswana, or where no other

locality name was given. A list of abbreviations was

prepared (Table 5) and these were used throughout

for the spelling of location names. Points of the com-

pass in names (e.g. Pretoria North) were not abbre-

viated. Specimens collected between two towns or on

the boundary of two magisterial districts were not

given major or minor location names through the

use of these town or district names.

41. Precise location

The precise location, as given on the label, was

recorded on two lines of the coding form. This entry

was made as complete as possible without unneces-

sarily repeating the major and minor locations. The

standard list of abbreviations was used in this field

as well.

42. Miscellaneous notes

The most important information not coded any-

where on the data form was entered in this field.

43. Economic narrative

The utilized by and economic property items were

elaborated upon in this field.

44. Common name and language

The common name was entered in this field as it

appeared on the label. The language of the entered

common name was coded from the following list

(present code numbers in parentheses):

01 Afrikaans (01)

02 YingWshl Engels (02)

03 German/ Duits (03)

04 Dutch /Nedertands (04)

14 Basotho /Basoetoe (05)

05 Bushman /Boesman (06)

06 Damara (07)

07 Herero (08)

08 Himba (09)

09 Kavango (10)

10 Mafwe (11)

1 1 North Sotho/Sipedi/AoorJ-

Sotho (12)

12 Ovambo (13)

18 Shangaan (19)

13 South Ndebde/ Suid-Nde-

bele (14)

14 South Sotho/ Suid-Sotho

(15)

15 Subia (16)

16 Swazi (17)

17 Tjimba (18)

18 Tsonga (19)

19 Tswana (20)

20 Venda (21)

21 Xhosa (22)

22 Zulu /Zoeloe (23)

00 other/unknown /ander/on-

bekend

The label language codes were kept separate from

the common name language codes.

45. Number of duplicate sheets

If there was more than one complete specimen of a

collector’s number within a species folder, the number

of duplicates was entered in column 43-44 of the last

line of the form. Where parts of a specimen were

mounted on more than one sheet, these were not

considered as duplicates and all such sheets were taken

into account in the determination of the state of the

specimen (see above).

46. Photograph code

Specimens encoded from photographs mounted on

herbarium sheets and filed in species covers were

identified by means of a code ‘PH’ on the data form

in columns 45—46 of the last line of the form.

COSTS AND DURATION OF OPERATION

It was anticipated that encoding of all the estimated

500 0 )0 specimens housed in the National Herbarium

would take the encoding team two to three years.

Coding was begun during June 1974 and was effective-

ly completed by December 1976. As a result of con-

tinuing specimen collection (some 20 000 new acces-

sions per annum) encoding will continue until all

collectors are using the new collector’s form (see Mor-

ris & Glen, 1978: Fig. 3). A total of over 470 000

specimens were encoded (Table 6) and 38 500 man-

hours were worked. The average cost per specimen

was R0,258 and the average cost per man-hour was

R3,14.

TABLE 6.— Statistics of the backlog encoding project

Costs and statistics for each of the encoding

periods for which Monthly Team Summaries were

calculated are summarised in Table 7. Initially, costs

per specimen were high as a result of intensive training

of encoders without botanical background and the

use of inefficient procedures. Costs dropped rapidly

and stabilised at R0,23 to R0,26 per specimen while

the norm was gradually increased. Towards the end

of the period, costs rose slightly as a result of basic

pay rate increases being awarded to encoders.

J. W. MORRIS

159

TABLE 7. Herbarium backlog encoding monthly statistics

Report

number

Cost per

specimen

1

2

3

4

5,

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23..

24.

25.

26..

27.

28.

1347.5

1804.0

2106.0

1486.0

1605.0

2070.0

2064.0

1560.0

1995.0

1013.0

1669.0

1613.0

1144.5

1520.5

1580.0

1800.5

1219.0

1609.0

1643.0

1132,0*

1062,5*

767,0*

1165,5*

956,0*

625,0*

715,5*

864,0*

420,5*

R

2926.50

5604.05

6919.96

4811.30

4836.50

6379.30

6503.06

4914,75

6560.96

3557.06

5136,24

5119,18

3584,12

4763,84

5137.60

6198.43

3591,89

4699.44

4913.60

3453,48

3346,38

2507,17

3633.97

3024.77

2163.98

2635,29

2776,53

1515.77

0,79

0,39

0,27

0,28

0,24

0,22

0,29

0,26

0,25

0,22

0,25

0,23

0,25

0,24

0,27

0,23

0,25

0,27

0,24

0,25

0,27

0,26

0,27

Encoding for 3 hours per evening instead of 4 as previously.

DISCUSSION

The success of the project is attributed primarily to

the bonus incentive scheme and the availability of a

comprehensive encoder instruction manual. The bonus

scheme and the part-time staff of 28 ensured that the

task was completed within a relatively short period.

For about the same cost, the operation could have

been carried out in about 10 years by a smaller, full-

time staff. Under those circumstances, it is likely that

interest in the project would have waned and also

that encoding standards and conventions would

have altered more than under the circumstances

actually imposed. Availability of an instruction

manual facilitated the training of staff, enabled

encoding standards to be maintained and allowed

the use of an enforceable penalty scheme.

The codes, including synonyms and Afrikaans

equivalents are given in full in this paper together with

the coding instructions given to encoders. In this

way, users of the data bank are informed about the

conventions adopted and generalisations made. Any

future operation of this kind should find this a useful

foundation on which to build. The tables of Latin and

German equivalents are included for the same reasons.

It is not intended that where more than one word is

given for a code they should be considered as syno-

nyms in the strict sense but they indicate the range of

variation within the concept included in the code.

Translation lists from foreign languages should be

used to obtain an understanding of the philosophy

behind the use of ite n codes; my lists are not diction-

aries. Use was made of inferences and assumptions

in order to define codes clearly and to spell out encoder

instructions. In a few cases, the codes indicated will be

incorrect or misleading to a trained botanist in the

context of all the information contained on a label.

It was, however, impossible to define all possible

situations and a few such errors will obviously have

escaped detection.

Most problems were experienced with the computer

processing of the data and the physical handling of

large volumes of data sheets, printouts and punch

cards. Although computer processing falls outside

the scope of this contribution, these problems are

mentioned here. At initiation, this was the largest

system on the Burroughs B5700 computer and al-

though the computer was large enough, the operations

room was not capable of handling the volume effi-

ciently and keeping concise records of processed

data tapes and system backup tapes. Although the

problems diminished as experience was gained, a great

deal of time was lost as a result. The need for efficient

computer room control procedures with a project

such as this cannot be over-emphasized. The volume

of paper output to be scanned for errors was also

underestimated and as a result handling of large

stacks of paper was difficult. The single printout of

every specimen which was made for manual checking

purposes used over 100 boxes of computer paper.

Storage of paper and punched data sheets in an

accessible manner was also difficult to plan. Solutions

would have been the use of computer output micro-

film and a re-designed output format requiring fewer

lines per specimen so that more specimens could

have been listed on each page. It is fortunate that

after about 200 boxes of computer cards had been

punched a switch was made to key-punching onto

magnetic tape. As mentioned earlier, handling of

magnetic tapes is far easier than boxes of cards. It

was found that computer processing could not keep

pace with encoding and, similarly, error correction

could not keep pace with computer processing.

Processing and data correction therefore lagged

further and further behind as the project continued.

160 ENCODING THE NATIONAL HERBARIUM (PRE) FOR COMPUTERISED INFORMATION RETRIEVAL

It was hoped that manual vetting and correction of

computer printouts could be made part of the bonus

incentive scheme so that it could be completed in a

short period as a distinct operation but this was not

possible and vetting was carried out by the small

team of permanent encoders who have taken over

from the backlog encoding team.

The data bank is open-ended in that new accessions

(about 20 000 a year) are being added continually.

A special field data sheet has been designed for use by

collectors who regularly submit specimens to the

National Herbarium to obviate the need for encoding.

A small team of permanent encoders has been respons-

ible for encoding donations and exchanges not

accompanied by the new field data sheet since the

conclusion of the project reported here. This team is

also responsible for the continuing correction of

errors in the precis data bank.

The error detection systems included in the bonus

incentive scheme were successful in reducing the

errors to an acceptable level. With the large volume

involved, however, a certain number of errors were

bound to slip through. A continuing programme of

error correction will be maintained for a few years

while the data bank is used and errors are discovered.

The global editing of localities and collectors’ names

was planned from the outset and is proceeding satis-

factorily. From alphabetical lists of all collectors and

localities, mis-spellings are being removed by a

qualified botanist. The addition of quarter-degree

grids by computer from the list produced by Leist-

ner & Morris (1976) at a later stage is being considered.

The costs given in this paper do not include those of

key-punching, computer program writing and com-

puter processing time. Also excluded are the costs of

the manual checking of computer printouts and

completion of error correction data forms. The costs

of the systems analysis and time spent by the writer

on the planning of the project are also omitted. These

costs are not available but it is estimated that, if

included, they would double or treble the expenses of

encoding reported here. The benefits which the

botanical community will gain from the project are,

however, such that the expense will be amply justified.

ACKNOWLEDGEMENTS

It is most unlikely that this project would have been

initiated without the enthusiastic support of Dr B. de

Winter, Director of the Botanical Research Institute.

The Workstudy Section of the Department of Agri-

cultural Technical Services, and Messrs G. P. Kleyn-

hans, F. J. S. van Biljon and J. L. Pieters in particular,

assisted by streamlining procedures and proposing the

bonus incentive scheme. Mr G. J. Smit and the staff

of the Data Processing Section of the Department

contributed to the success of the project by making

computer-related facilities available when required.

The Algol computer program for accessioning data

was written by Mrs E. de Bruyn. The staff of the

National Herbarium, and in particular the then Cura-

tor, Dr O. A. Leistner, and Mrs E. van Hoepen, were

of assistance in the preparation of the herbarium for

encoding. I thank Drs J. H. Ross and A. V. Hall for

encouragement and advice during the planning stages

of this project and Drs O. A. Leistner and H. F. Glen

for commenting on drafts of this manuscript.

UITTREKSEL

Die enkodeering van inligting vanaf 470 000 bestaan-

de herbariumeksemplare vir gebruik in ’n rekenaar-

databank word beskryf as hulpmiddel vir kuratore en

stelselontleerders wat gelyksoortige prosedure in hul

eie herbariums wil toepas. Die voorbereiding van die

herbariums, die enkoderingsprosedure , die nagaan

van geenkodeerde eksemplare, koste en vergoeding van

enkodeerders word beskryf. ’n Volledige lys van geenko-

deerde items en hul kodes word gegee. Van die probleme

wat aangetref is, word bespreek en die sukses van die

projek tot op datum word kortliks geevalueer. Sukses

word toegeskryf aan die bonus-aansporingsstelsel en die

beskikbaarheid van 'n volledige instruksieboek.

REFERENCES

Acocks, J. P. H., 1953. Veld types of South Africa. Mem. bot.

Surv. S. Afr. 28: 1-192.

Brenan, J. P. M., Ross, R. & Williams, T., 1975. Computers

in botanical collections. London: Plenum Press.

Dyer, R. A., 1975. The genera of Southern African flowering

plants. 1. Dicotyledons. Pretoria: Botanical Research

Institute.

Dyer, R. A., 1976. The genera of Southern African flowering

plants. 2. Gymnosperms and Monocotyledons. Pretoria:

Botanical Research Institute.

Leistner, O. A. & Morris, J. W., 1976. Southern African

place names. Ann. Cape Prov. Mus. 12: 1-565.

Morris, J. W., 1974. Progress in the computerization of

herbarium procedures. Bothalia 11: 349-353.

Morris, J. W. & Du Toit, P. C. V., 1976. Handbook for the

encoding of herbarium specimens. Pretoria: Botanical

Research Institute (unpublished report).

Morris, J. W. & Glen, H. F., 1978. precis the National

Herbarium of South Africa (PRE) computerized informa-

tion system. Taxon 27 : 449-462.

Morris, J. W. & Leistner, O. A., 1975. Progress with com-

puterization of the National Herbarium, Pretoria. Bois-

siera 24: 411-413.

Pieters, J. L., 1974. Verslag oor die ondersoek na die kodering

van plantmonsters in die Nasionale Herbarium van die

Navorsingsinstituut vir Plantkunde met die oog op reke-

narisering gekoppel aan ’« direkte finansiele aansporing-

stelsel. Pretoria: Department of Agricultural Technical

Services (unpublished report).

Ross, J. H., Leistner, O. A. & De Winter, B., 1977. A guide

to contributors to the Flora of Southern Africa (F.S.A.).

Pretoria: Department of Agricultural Technical Services.

Bothalia 13,1 & 2: 161-169 (1980)

Translocation heterozygosity in southern African species of Viscum

D. WIENS* and B. A. BARLOWf

ABSTRACT

Sex-associated and floating translocation complexes are characteristic of dioecious species of Viscum ,

but are virtually absent in monoecious species. The majority of dioecious species has fixed sex-associated

translocation complexes with the male being the heterozygous sex. The sex-associated multivalent is usually

04 (ring-of-four) or 06, rarely 08. Dioecious species without sex-associated translocations are much less

common. Most of the dioecious species are also polymorphic for floating translocations, producing one or

more additional multivalents ranging from 04 to 012. Floating translocations may be more frequent in

species that do not have sex-associated translocations. Supernumerary chromosomes are also present in several

species. Sex ratios are at unity in most dioecious species, but female-biased ratios may occur in some species.

The high correlation between dioecy and translocation heterozygosity suggests that translocations are primarily

associated with the origin and establishment of dioecy. Any role in the maintenance of biased sex ratios through

meiotic drive is probably secondary. Sex-associated translocations may serve to stabilize dioecy by bringing the

sex factors into close linkage. Subsequent structural rearrangements within a sex-associated translocation

complex may bring the sex factors together in one chromosome pair, releasing floating translocations. The

high frequencies of floating translocation heterozygosity in some species indicate that such heterozygosity

also has adaptive value.

RESUME

TRANSLOCATION HEtEROZYGOTE CHEZ DES ESPECES SUD-AFRICAINES DE VISCUM

Des complexes de translocation lies an sexe e t flat I ants sont caracteristiques des especes dioiques chez Viscum, mais

n ’existent pratiquement pas dans les especes monoiques. La plupart des especes dioiques ont des complexes de trans-

location fixes et lies au sexe, avec I’heterozygotie du cote male. Le multivalent associe au sexe est habituellement

04 ( anneau de quatre) ou 06, rarement 08. Les especes dioiques sans translocations liees au sexe sont beaucoup

moins repandues. La plupart des especes dioiques sont egalement polymorphiques pour des translocations flottantes,

produisant un ou plusieurs multivalents additionnels qui s’etendent de 04 a 012. Des translocations flottantes

peuvent se rencontrer plus frequemment chez des especes qui n'ont pas de translocation liee au sexe. Des chromo-

somes surnumeraires se rencontrent egalement chez plusieurs especes. Dans la plupart des especes dioiques la sex-

ratio est de 1, mais dans certaines on peut trouver une deviation du cote femelle. Le degre eleve de correlation

entre la dioecie et I’heterozygotie de la translocation suggere que les translocations sont associees primairement

a Vorigine et a l’ etablissement de la dioecie. Tout role que pourrait jouer la poussee meiotique dans le maintien de

sex-ratios faussees est probablemei t szcondaire. Des translocations liees au sexe peuvent servir a stabiliser la

dioecie en about issant a une association etroite entre les facteurs sexuels. Des rearrangements struct urels ulterieurs

dans I’enceinte d’un complexe de translocation lie au sexe peuvent rassembler les facteurs sexuels dans une seule

paire chromosomique en relachant des translocations flottantes. La frequence elevee des translocations heterozygotes

flottantes chez certaines especes indique que pareille heterozygotie possede egalement une valeur adaptative.

Viscum is a genus of mistletoes comprising about

100 species with an extensive range in the Old World.

Its species parasitize a wide range of dicotyledonous

trees and shrubs in a variety of forest and woodland

habitats. Major centres of species diversity are

Africa and Madagascar, but there is also a significant

development of the genus in tropical and subtropical

Asia. The extremes of its geographical range are

reached in Europe, southern Africa, temperate Asia,

and Australia. In southern Africa the genus is

represented by at least 17 species, of which 9 are

dioecious, 7 monoecious and one ( V . capense) that

has both monoecious and dioecious subspecies.

The flowers in Viscum are small and consistently

unisexual. The basic inflorescence unit is a cymule,

usually consisting of three minute flowers (1-3 mm

wide), but in some species the flowers are solitary or

in larger clusters. The monoecious species typically

have the central flower of the cymule of one sex

and the two lateral flowers of the other.

Dioecious species are common in Africa and

Madagascar, and a few others are widely distributed

in Europe and temperate Asia, but monoecious

species are the rule in tropical Asia. In the dioecious

species the cymules are still typically three-flowered,

so that dioecy presumably involves a possibly simple

conversion of all flowers to the same sex for all

* Department of Biology, University of Utah, Salt Lake

City, Utah 84112, USA.

t School of Biological Sciences, Flinders University of

South Australia, Bedford Park, South Australia 5042.

cymules of a plant. Dioecy appears to be highly

stable in its expression in Viscum, and mixed cymules

are unknown in any of the dioecious entities we

have studied.

The most common chromosome number among

the African species of Viscum is x—\4, and this is

apparently the basic number for the genus (Wiens,

1975). The Madagascan species so far examined

mostly have x=13, but more extensive data are

needed before firm conclusions can be drawn. Some

of the Madagascan species are more closely related

to the species of Viscum of the Asian region than

to those of Africa. The southern African species are

especially interesting because of the high incidence

of derived aneuploid chromosome numbers of x=15,

12, 11 and 10 occurring in this region.

Translocation heterozygosity was first reported

for Viscum in the dioecious species, V. fischeri Engl,

from eastern Africa (Wiens & Barlow, 1973; 1975).

This species has a unique system of translocation

heterozygosity in which male plants have a

chromosome number of 2n=22>, and consistently

produce seven bivalents and an open multivalent

chain of nine chromosomes at meiosis. Female

plants have a chromosome number of 2n=22, form

1 1 bivalents at meiosis, and are thus chromosomally

homozygous for the 11-chromosome genome.

Translocation heterozygosity is therefore presumed

to be sex-associated, with the male being the hetero-

zygous sex and the 11- and 12-chromosome genomes

female-determining and male-determining, respec-

tively (Barlow & Wiens, 1976). Additional studies

162

TRANSLOCATION HETEROZYGOSITY IN SOUTHERN AFRICAN SPECIES OF VISCUM

also revealed the presence of a strongly female biased

sex ratio of approximately 1 : 2.

Further studies have indicated that sex-associated

translocations are widespread in dioecious species of

Viscum, but typically absent in monoecious species.

A more extensive cytogenetic survey of the southern

African species of Viscum is presented in this paper.

The occurrence and distribution of translocaton

heterozygosity and supernumerary chromosomes are

described and considered in relation to the origin

and establishment of dioecy and to variations in

the sex ratio.

METHODS

Flower buds were fixed in ethyl alcohol/acetic

acid (3:1) and stored in absolute alcohol at —15 °C

whenever possible. Chromosomal constitution was

determined from aceto carmine squash preparations

of pollen mother cell, or embryo sac mother cell

meiosis in some female plants. The latter involved

dissection of a block of tissue containing the sporo-

genous cells from the base of the flower bud (there

are no discrete ovules in Viscum), maceration in a

solution of 1% HC1 in 45% acetic acid on a slide

for 2 min at 60° C, and light squashing.

The sex ratio for most species was determined

directly from established populations in the field.

Plants were mostly removed from the host for

counting so that difficulties caused by clumped

distributions and autoparasitism could be minimized.

Voucher specimens for each chromosome deter-

mination are deposited at the State Herbarium of

South Australia (AD), Botanical Research Institute,

Pretoria (PRE), and the Garret Herbarium, University

of Utah (UT).

RESULTS

The chromosomal constitutions of 359 individuals

were determined for 16 of the 17 species of Viscum

currently recognized for southern Africa (Wiens &

Tolken, 1979). Details of these determinations are

given in the Appendix and selected characteristic

configurations are illustrated in Fig. 1.

Monoecious species

All of the monoecious species were studied

karyologically with the exception of V. schaeferi.

In all but one of these the floral distribution is

essentially constant and generally typical of

monoecious species, with each cymule bearing both

male and female flowers, with a central flower of one

sex and lateral flowers of the other. In V. capense

subsp. hoolei, however, the expression of monoecy

is altered. Some plants produce mostly female flowers

which possess basically female morphology, while

other plants produce mostly flowers of typical male

morphology and mature few fruits. The vegetative

characters of V. capense subsp. hoolei agree generally

with the dioecious V. capense subsp. capense ; thus

in subsp. hoolei monoecy probably represents a

derived state resulting from modification of sex

determination in a dioecious species. In other

monoecious species of Viscum, e.g. V. obovatum,

male and female flowers are produced together only

at certain times, while at other times only female

flowers occur in the bracteal cups (Wiens & Tolken,

1979). This is apparently a relatively common

occurrence in monoecious species in other parts of

the world (Danser, 1941).

The chromosomal data for the monoecious entities

are summarized in Table 1. There is a general absence

n

3 4

Fig. 1. — Pairing configurations of metaphase I chromosomes in African species of Viscum. 1, V. subserratum 911 04 (W5305h).

2, V. subserratum 811 06 (W5305f). 3, V. continuum 1011 04 (W5377b). 4, V. obscurum 1011 04 06 (W5372b).

D. WIENS AND B. A. BARLOW

163

of translocation heterozygosity in these species, its

only occurrence being recorded in a single individual

of V. capense subsp. hoolei which has chromosome

associations of eight bivalents and a ring of four

chromosomes (811 + 04). This is in striking contrast

to the situation in the dioecious species where trans-

location heterozygosity is probably universal, at

least in some male plants (see below). As previously

mentioned, in V. capense subsp. hoolei monoecy

may be derived from dioecy, so that the translocation

heterozygosity present may be the same as that

found in the dioecious subspecies.

TABLE 1. — Summary of chromosome constitutions in southern

African monoecious species of Viscum

* atypical monoecy : see text.

Chromosome numbers among the monoecious

species are n= 10, 12, 14 and 23, but /?= 28 occurs

in some monecious species in eastern Africa (Wiens,

1975). Viscum spragueanum is probably consistently

polyploid, but one tetraploid individual was

discovered in V. capense subsp. hoolei, which is

otherwise diploid. This situation contrasts sharply

with that in the dioecious species, where only a single

triploid plant was discovered in V. verrucosum.

Although alternative explanations are possible, the

rarity of polyploidy in the dioecious species could be

related to a mechanism where sex is determined by

a balance between strongly female- and male-

determining X and Y factors, as originally proposed

by Muller (1925).

Dioecious species

The chromosome constitutions of all of the

dioecious southern African species of Viscum were

analysed (Table 2). Translocation heterozygosity

occurs in all 10 of the dioecious species, and these

fall into two groups. The larger group (with 6 species)

exhibits patterns in which the male plants are all

heterozygous for at least one translocation complex.

Only a few female plants were analysed for pairing

relationships in the southern African species, but

these were consistent with the assumption that some

translocation rings are sex associated, occurring in

male plants only, and that female plants are

predominantly chromosomally homozygous and pro-

duce mostly bivalent associations. Thus where male

plants occur with more than one translocation ring,

the additional rings presumably also occur in female

plants. The sample sizes in the southern African

species are not large enough to predict the

frequencies with which these floating translocations

occur in female plants, but there is no reason to

assume they are not identical to those in the male

plants, as previously demonstrated in an east African

species, V. hildebrandtii (Wiens & Barlow, 1979).

In that species the population from Riandu, Kenya,

has male plants which all possess 06 (the sex-

associated complex), but some male plants have an

additional 04 and/or 06. In the female plants these

floating 04 and 06 were recorded as well as a

homozygote with 14 II. This strongly supports our

earlier suggestion (Barlow & Wiens, 1975) that in

this group of dioecious Viscum there is one trans-

location ring which is sex-associated and fixed in

males as the heterozygous sex, while the other rings

are floating and not sex-associated.

In some respects the species in this group are

chromosomally relatively uniform. With the exception

of V. subserratum («=11), all species in the group

have «=14 and a generally symmetrical karyotype.

The most common sex-associated rings are 04 and

06, and in three species both rings are present.

A floating 04 is present in at least five of the six

species.

TABLE 2. — Summary of choromosome constitutions in dioecious species of Viscum in southern Africa

164

TRANSLOCATION HETEROZYGOSITY IN SOUTHERN AFRICAN SPECIES OF VISCUM

A few differences are superimposed on this common

pattern. A sex-associated 08 occurs in V. com-

breticola, and in two additional species in eastern

Africa (Wiens & Barlow, 1979). Its occurrence may

be linked with the presence of a sex-associated

06, which appears to be common in these three

species. In V. combreticola there is also evidence for

two independent floating 04, and a floating 06

occurs in V. menyharthii.

In this species group, then, the sex-associated

complex consistently appears in males as 04 and

06, and in addition, most species have a floating

04. This regular pattern suggests that the sex-

associated translocations (and possibly also the

floating ones) may be the same in most or all of the

species in the group. If so, the translocation system

may be relatively old and possibly established prior

to the differentiation of the extant species. The

accumulation of linked complexes through the

translocations may have promoted rapid differen-

tiation of biotypes, and the uniformity of chromosome

number among these species is consistent with this

possibility.

The second group includes the remaining four

species which all have different chromosome numbers

(« = 10, 12, 14, 15). In these species some of the male

plants produced only bivalents, indicating that

fixed sex-associated complexes are not present. There is

evidence, however, of extensive floating translocation

heterozygosity, presumably in female plants as well

as males, but additonal data on this point are needed.

In fact, floating translocations are apparently

more frequent in this species group than in the first.

In V. oreophilum two floating Q4s occur, but V.

obscurum contains the most extreme complexity,

where associations up to 012 exist, and individuals

with two rings are relatively common. The minimum

number of floating translocations needed to produce

these associations in V. obscurum is five, assuming

that all configurations observed represent recom-

binations of the same translocations.

Perhaps the species in this group represent a

transitional stage in the evolution of the sex-associated

translocation system, with the floating translocations

acquired first and later linked with the sex-deter-

mining chromosomes by subsequent translocations.

However, this group contains both highly specialized

and diverse species as indicated by their morphology

and range of chromosome numbers. They may,

therefore, also represent a derived state in which

sex-associated multivalents are released from linkage

with the sex-determining chromosomes by sub-

sequent translocations. These would then become

floating multivalents. This hypothesis is consistent

with the observation that the species in this group

have a greater diversity of floating multivalents than

the species in the preceding group. In other words,

sex-associated multivalents were possibly acquired

early in the evolution of the genetic system, and

later converted to floating associations by subsequent

exchanges. This appears to us as the best explanation

and is discussed further.

A third class of translocation heterozygosity occurs

in African Visca, but it is limited to the single species,

V. fischeri, which is geographically restricted to

eastern Africa. This translocation system was

described earlier in this paper and extensive discussion

of this translocation system is unnecessary and it is

mentioned here only for conceptual completeness.

This species is unique in having different chromosome

numbers in males and females and also unusual in

its uniformity, having the same sex-associated multi-

valent in all male plants examined, and no floating

multivalents in either sex.

Frequencies of floating translocations

The data indicate that the frequencies of floating

translocations vary between populations in some

species. Estimates of the frequencies of floating 04

for a few species in which some analysis is possible

are given in Table 3.

In V. verrucosum the frequency of the heterozygote

with the floating 04 may vary significantly between

populations, from zero to the theoretical maximum

of 0,5, with corresponding variations in gametic

frequencies of the translocation complexes. Similar

variations appear to occur in some of the other

species listed. However, because of the small sample

sizes it is not possible to determine the frequencies

TABLE 3. — Frequencies of floating 04 in southern African species of Viscum

* theoretical maximum.

** significant at p<0.025.

D. WIENS AND B. A. BARLOW

165

TABLE 4. — Sex ratios in some dioecious southern African species of Viscum

* p<0.05.

with high levels to confidence, and the calculated

frequencies in Table 4 can only be regarded as a

general indication of the scale of variation.

If the high frequencies of heterozygosity recorded

for some populations are real, then positive selection

favouring the heterozygotes is indicated. If the

translocation complexes maintain adaptive gene

combinations, there must be interaction between

suites of genes in the complementary chromosome

complexes, thus generating heterozygote advantage.

Bloom (1977) has shown in Clarkia that translocation

heterozygosity is maintained because of the levels of

inbreeding which might be imposed on natural

populations. This argument could apply to Viscum,

which is a generally host-specific parasite, often

occurring in small, relatively isolated local populations

probably derived from a few founder individuals.

Supernumerary chromosomes

A single supernumerary (B-) chromosome was

observed in some individuals in a few of the dioecious

species. In V. verrucosum, the B-chromosome was

recorded from several populations throughout the

range of the species in South Africa, and occurred in

approximately half the plants sampled from each

population in which it was found. The B-chromosome

in this species was visible at meiotic metaphase as a

small rounded unit usually just off the equatorial

plate. Earlier in prophase it showed a constriction

near one end, indicating it was acrocentric. The B-

chromosome usually moved undivided to one pole

at first anaphase and divided at second anaphase,

and was thus transmitted to two of the four pollen

grains. Occasionally it lagged at first anaphase, and

the elimination of the B-chromosome may be

correlated with the occurrence of a small microcyte

present in less than 10 per cent of tetrads. In some

cases the B-chromosome in V. verrucosum appeared

to be associated with aberrations in the A-chromo-

somes. Sometimes one bivalent did not orient at

first metaphase, behaving as if it were monocentric.

In other cases chromosome fragmentation appeared

to occur in cells carrying the B-chromosome.

In V. capense from the Springbok area the B-

chromosome was of similar shape and behaviour to

that in V. verrucosum, except it was larger. In V.

obscurum a B-chromosome was observed in only one

population, but it was much larger than those observed

in any other species. Thus different species appear to

have distinctive B-chromosomes. In addition, at

least three different B-chromosomes are widespread

over the range of V. album in Europe and Asia (Barlow

et al., 1978; Barlow & Wiens, unpublished data),

suggesting they have a regular behaviour which

ensures their persistence.

Although B-chromosomes in Viscum may be

confined to the dioecious species, it is unlikely that

they have a direct association with the dioecious

state. Since the dioecious species are also characterized

by high frequencies of translocations, it seems likely

that the B-chromosomes have originated as centric

fragments resulting from chromosome breakage,

reunion and segregation. Such an origin has been

suggested for other groups of plants (John, 1976;

Darlington, 1974; Jackson, 1960) and particularly for

Clarkia, where B-chromosomes may arise de novo

rather frequently in populations with high levels of

translocation heterozygosity (Wedberg et al., 1968).

166

TRANSLOCATION HETEROZYGOSITY IN SOUTHERN AFRICAN SPECIES OF VISCUM

Sex ratio

The data on sex ratio are summarized for several

species in Table 4. Sample sizes are occasionally

relatively small, due to the difficulty of collecting some

species. A few species of Viscum have deviations

from a sex ratio of unity. An excess of female plants

was previously recorded for V. fischeri in east Africa

(sex ratio 0.52; Barlow & Wiens, 1976). In V.

obscurum the total for all samples deviates significantly

from an expectation of unity, and the individual

samples are homogeneous and consistent with a

female predominant sex ratio of about 0.75. A

similar excess of female plants may occur in V.

subserratum, although the populations sampled gave

results just within the level of significance. The only

case of a possible excess of male plants is in the

east African species, V. hildebrandtii where the sex

ratio may be about 1.4 (Barlow & Wiens, 1975).

In most cases, however, the data agree with a sex

ratio of 1.0, and where more than one population

of a species was sampled, the data appear to be

homogeneous. The biased sex ratio first observed

in V. fischeri therefore does not appear to be a general

feature of the African species. No other species

examined from Africa shows the distortion to the

same degree, either in favour of male or female

plants.

In V. fischeri we originally suggested that the

biased sex ratio might be due to the sex-associated

translocation heterozygosity, with the two trans-

location complexes having different transmission

rates through the pollen. We believed this might

be the principal role of the translocation hetero-

zygosity in this species. The patterns obtained in

southern Africa now indicate that this explanation is

unlikely, since there is no regular relationship between

sex ratio and translocation heterozygosity. If the

translocations have a role in the maintenance of

biased sex ratios, it is probably a secondary function

of the translocation system and is established only

in those species in which distortion of the sex ratio

may be of significance for other reasons. Situations

in which biased sex ratios might be favoured in plants

are described by Bawa & Opler (1977), Mulcahy

(1967) and Kaplan (1972). In Viscum an excess of

females might achieve maximum seed set in

populations which are space-limited, efficiently

pollinated, and perhaps subject to heavy seed

predation. An excess of male plants may be favoured

in open situations where pollination efficiency might

be limited by the distance between male and female

plants.

DISCUSSION

This survey shows that translocation hetero-

zygosity occurs in all the dioecious species of Viscum

in southern Africa, but is apparently rare among

the monoecious species. Since the dioecious species

V. album L. and V. cruciatum Boiss. in Europe and

Asia, and V. alniformosanae Hayata in Taiwan show

patterns of translocation heterozygosity similar to the

African species (Barlow et ah, 1978; unpublished

data), this relationship probably extends to all

dioecious members of the genus. The only remaining

area rich in dioecious species of Viscum, which is not

well surveyed, is Madagascar. The Madagascan

species, as opposed to the African ones, are apparently

based predominantly on x=13 instead of x=14

(Wiens, 1975), but there is little reason to suspect

that their genetic systems are different.

The translocation heterozygosity in Viscum thus

emerges as a phenomenon which is characteristic of

the dioecious state. As suggested above, some of

the translocations may be common to several species,

and the translocation systems were possibly

established prior to, or during the differentiation of

related species. The translocations, in fact, may have

played a fundamental role in the establishment of

dioecy in the ger.us.

There is little doubt that dioecy in Viscum is

derived from monoecy. Unisexual flowers are

characteristic of all Viscaceae, and monoecy is fixed

or predominant in all genera. In most other respects,

however, Viscum conforms with the observatio s or

predictions of dioecy in tropical forests made by

Bawa & Opler (1975). They concluded that dioecy

evolved in response to selection for the enforcement

of outcrossing, and suggested a number of factors

which may have favoured dioecy as the outcrossing

mechanism. These include (1) the likely genetic

simplicity of dioecy compared with self-incom-

patability (Baker, 1967), (2) high levels of reproductive

failure in self-incompatible hermaphrodite or

monoecious species because of the small foraging

ranges of the pollen vectors, and (3) escape from seed

predation caused by altered size and distribution of

seed set. Possibly all of these factors have contributed

to the evolution of dioecy in Viscum.

If dioecy is established by mutations affecting

different hormone systems which favour maleness

and femaleness (Audus, 1972; Bose & Nitsch, 1970),

then such genes would likely be nonallelic. This

possibility was argued on a priori grounds for angio-

sperms generally by Ross & Weir (1976) and

Charlesworth & Charlesworth (1979). These authors

also point out that mutations for dioecy are not

likely to accumulate simultaneously, and that in

many cases gynodioecy, through male sterility, is an

intermediate step. If the genes for full dioecy were

not linked, recombination could produce males,

females, hermaphrodites and neuters, and Charles-

worth & Charlesworth (1979) suggest there might

often be a “linkage constraint ’, such that unlinked

genes for dioecy may be selected against. Thus genes

for dioecy should occur on the same chromosome,

with no crossing-over between loci, so that particular

chromosomes become identified with sex deter-

mination. If these constraints apply in Viscum, then

translocations should have a role in the evolution of

dioecy, namely in bringing the nonallelic sex factors

into close linkage. Our data suggest that in many

species of Viscum dioecy is stabilized in this way, and

that this system for maintaining dioecy is conserved

during the differentiation of new species. The problem

is discussed more fully elsewhere (Wiens & Barlow,

1979).

Even if this model for the evolution of dioecy in

Viscum is correct, translocations are perhaps of

further adaptive importance in the genus. Floating

translocations are common even in species like

V. hildebrandtii which retain the sex-associated multi-

valent, and they are also retained in species like

V. obscurum which may have lost the sex-associated

multivalent. Their persistence in high frequency

indicates a more basic role in the genetic system of

Viscum, presumably as a means of maintaining high

levels of heterozygosity. But why have translocations

accumulated only in dioecious species and not in

monoecious species? Perhaps the fixation of trans-

location heterozygosity by sex-association allows time

• o

.o *—

•CO

00 CO

r— LO

•c. 5!

} -Q fO

J CM

) lO VO

JO to

VO VO

TD O

CM I

co eo

»— CM

CM CO

CO r—

CM

u CO

TO tO *

co cn <

CM CM <

•O O

m in

Sco

co

in m

r— • — cm •—

to o

o o

r— CO

o o o

•o- ^

in in in

<D CD

co co

01-0 c O’

O0 co co CO

(O X3

CO CO

CM CM

VO lO CO CO

oO oO

3 3

CO CO CO CO CO

CO CO

CO CO CO CO CO

o0o0oQ«0«0

(O -O

CO vO VO

oun m

T3

CM CM CM

CO CO CO

in in in

(O UTJi — CD

*3- *3-

CO 00 CO 00 00

in m in in in

jo -o to to to 0)

r-v. r-^ r- in in in

co co co co co co

in in m in in in

»— VO

i/i o vo m

c cm to in

a> co <r

CM CM cm cm

^ vy

in in in in

co co co co

o o o

CO CO CO

in in in

co co

22

aiTjc io

ssss

CO CO CO CO

o •*-

OJ -O *0 *0 T3

O O CO CO CO

(O CD

co CO

I U C 0*0-0

i in in vo ot to

i m m vo vo vo

) CO CO CM CM CM

> in in in in in

<o o

o a> o.

sg =

8£8

to XT o

O U 1/1

Soo — ~cx>cx>

5^2 2 2 2 2 25co

E E E

r- CM

f- o

CO c

I -C

r-. o

CM ~0

E E

CO r-

r— CM

E E*-

lOINr-

■SS

C

0) o

a_ -*->

0) >»

o. to

to co

o

sn

<v c

ex <o

<3 i.

(O

• • JC

c o

o

— E

I I

OO OO

3UJ

2 2

o OM p p

r— , CO

£ ;

*—

o o

*-• c

ex •*-

<C Q-

O O

ex ^

U- f-

<£ 3

>— cn

CO CD

C3>

►-« <3>

O •-«

lO

CD

*-« CM

CD CD

CM O

r— r— r— CO

J "

5 I

-f->

< E

o *o

*— « x:

cc <o

•a: ex

o s-

*-« <D

CX 4->

< 3

CD C7-

O CX O

M4J •— > CD • S-

ex s- ex o 01

Lx 3 u_ O U-

<C -Q < CX <t

E E EE EE

in r- lO »— CO CM

O 4-

cx <0

lx c*

< C-i

0 S_ O <-> s-

o .. g

1 M « «■<=

££ y ££ „ <u

CX ’TO

Lx 3

O

O S-

*— <D

CX JO

< 5

3 3

vo «—

OS- o S-

O CD O 07

— v c o

r— Lx tO Lx- O

_ CD

^ CM CMOH

r— r— r— r— CD

4-> <D

VO vo

CD CD

CD VO ^3- VO

VO "O' «3- ID ® QD CD O'

o ® esajo-o-o- a>

^ >1 »-i ^ o o o *-•

~ 5 5

CD 00

r- ®

CD CD

CD

in co o •

cj> *3-

1 >— CD

mOH

*3 ^ ^-1 r- r- r— r— r— Ot r— r— Ol

«v3- VO

CD CD CO

^3- «^r cd (

CDCDCD^f—

_ , _ _ , CD CD

^ O — — •

® ® a>

10 co o co ... ^ ^

CD ®

in CO r—

EE E E E E E

r— r— CM VO ^ CM r—

,-vMr-r- CM r- CM

E E E E E E

COr-r-f- CM r— r—

EEEEE EEE EEE

CM r— CM r- *— r— r— CM CO CM »—

o

E EE

r— r— lO •—

r— *— CM

s s

*3* vo

CD CD — I CD ® CD CD CD

cn 1 — !EZ cn CD CD co f — 01 n

- 3

33

CD CD hm m

CD CD CD ^r

CD

CM r— O '

E E H- V

VO CM CM »

je •*-

*0 to

CX 4->

EEE

in *— cm

E E

vo in

EE E E E E E

CM CM CNJr-5fr- w—

s -

EE E

9— r— in

ia <y co *•"

o •*- O

<co «t vn

5 ^

8.S

*tJ O

o »—

•• E

CC *o

E

— < a> *-*

QJ

&s

o o

4->

<: E

o «o

•-* s=.

ex *0

: 0

CM O

< E

O «o

*—i x:

ex *0

lx S-

C CD

CO I —

ex o

Lx

< 3

10 CO

>

S la

03 O O

s- je

- £ S

<cl

0 3 0

»-.£■*->

CX <0 0)

Lx >, S-

in co to co

o •*-

ex ^

<3

O. c

*0 5

o o

O 0(0

•-« xr

ex ex ro

IX lx X

C CD

tO S-

o S-

a>

•• 4-»

< U)

O 0)

*— o

ex s-

Lx O

<31 3

CX r-

<c m

C LU

SJ 6

co *—

CM •—

O *—

-r- 0)

c s- •»-

2 5-2^

ex r- ex

Lx OJ ixl-

< Z «£ O

o "O 0(0

*0

ex c

<£ ex

CX CD

> X= O <D

• o •-* >

: •*- ex t-

. S- lx 3

: t— <0

«o ex

>

v/i -o

c c

«t <0

o x:

*-1 o

ex f-

lx S-

I—

<c en.

O f—

*-< a»

ex a>

< to

z

5

>lx

c •—

«o <0

S- x>

H— C

-8

(O •*-

CX CD

- 3

*«r o tve>3

CD CD . — CD CD <3

E E

CO CM

EEE

*— CO •—

• J

CO

(O

z

5

ex

z

i

•*- o

a! s- *0

000

c o

S. 5

8 -

2 5

z •*-

0 §

^ S

V.

.

'

TRANSLOCATION HETEROZYGOSITY IN SOUTHERN AFRICAN SPECIES OF VISCUM

for the accumulation of adaptive gene combinations

in the translocation complexes, thus giving them

immediate selective value when released as floating

translocations. Such complexes would have more

likelihood of being conserved than raw exchanges

directly exposed to selection, as would be the case

in monoecious species.

ACKNOWLEDGEMENTS

This study was completed while the first author

was a Research Fellow at the Botanical Research

Institute, Department of Agricultural Technical

Services, Pretoria, South Africa. We are especially

indebted to the Institute for financial support, the

use of laboratory and field facilities, and the generous

co-operation of the staff, particularly the Director,

Dr B. de Winter. We gratefully acknowledge the

assistance of many individuals who otherwise con-

tributed to the study, especially Mr P. van Wyk for

field assistance and Ms C. Wiens and W. H. Busby

for indispensable help in both the laboratory and

field. The study also received support from the United

States National Science Foundation and the Flinders

University Research Committee.

UITTREKSEL

Geslagsgekoppelde en drywende translokasiekom-

plekse is kenmerkend van tweehuisige ( diesiese )

Viscum-spesies, maar is feitlik afwesig in die een-

huisige spesies. Die meerderheid tweehuisige spesies

het vasgelegde translokasiekomplekse wat geasso-

sieerd is met geslagsbepaling; met die manlike geslag

die heterogametiese geslag. Die geslagsmultivalente is

gewoonlik 'n ring van vier 04 of 06, maar soms 08.

Tweehuisige spesies sonder geslagsgekoppelde trans-

lokasies kom baie selde voor. Die meeste tweehuisige

spesies is ook polimorfies vir drywende translokasie

deurdat een of meer addisionele multivalente kan

voorkom, wat wissel van 04 tot Q\2.

Drywende translokasies word meer dikwels aangetref

in spesies wat nie geslagsgekoppelde translokasies besit

nie. Bi-chromosome is ook teenwoordig in verskeie

spesies. Die verhouding manlike tot vroulike spesies

is gelyk in die meeste tweehuisige spesies, maar hierdie

verhouding mag verskuif; meer in die guns van die

vroulikes in sommige spesies. Die hoe korrelasie

tussen tweehuisigheid en translokasieheterosigositeit is

'n aanduiding dat translokasies primer geassosieerd

is met die oorsprong en vestiging van tweehuisigheid.

Enige rol in die afwyking van die verwagte verhouding

manlikes tot vroulikes deur “ miotic drive", is heel-

waarskynlik sekonder. Geslagsgekoppelde translokasies

mag tweeslagtigheid stabiliseer, deurdat dit die geslags-

faktore in noue koppeling bring. Daaropvolgende

strukturele herrangskikkings binne geslagsgekoppelde

translokasie chromosome mag die geslagsfaktore

geneties koppel in een chromosoompaar, met die

totstandkoming van drywende translokasies. Die hoe

frekwensie van drywende translokasieheterosigositeit in

sommige spesies dui aan dat sulke heterosigositeit

ook voordelig in die plante is.

REFERENCES

Audus, L. J. ,1972. Plant growth substances. London: Leonard

Hill.

Baker, H. G., 1967. Support for Baker’s Law — as a rule.

Evolution 21 : 853-856.

Barlow, B. A. & Wiens, D., 1975. Permanent translocation

heterozygosity in Viscum hildebrandtii Engl, and V. engleri

Tiegh. (Viscaceae). Chromosoma (Berl.). 53:265-272.

Barlow, B. C. & Wiens, D., 1976. Translocation heterozygosity

and sex ratio in Viscum fischeri. Heredity 37: 27—40.

Barlow, B. A., Wiens. D., Wiens, C., Busby, W. H. &

Brighton, C. 1978. Permanent translocation heterozy-

gosity in Viscum album and V. cruciatum : sex association,

balanced lethals, sex ratios. Heredity 40: 33-38.

Bawa, K. S. & Opler, P. A., 1975. Dioecism in tropical forest

trees. Evolution 29: 167-179.

Bawa, K. S. & Opler, P. A., 1977. Spatial relationships

between staminate and pistillate plants of dioecious tropical

forest trees. Evolution 31:64-68.

Bloom, W. L., 1977. Translocation heterozygosity, genetic

heterozygosity, and inbreeding in Clarkia speciosa.

Evolution 31:256-264.

Bose, T. K. & Nitsch, J. P., 1970. Chemical alteration of sex

expression in Luffa acutangula. Physiologia PI. 23: 1 206—

1211.

Charlesworth, B. & Charlesworth, D., 1979. A model for

the evolution of dioecy and gynodioecy. Amer. Nat. (in

press).

Danser, B., 1941. The British-Indian species of Viscum revised

and compared with those of South-Eastern Asia, Malaysia

and Australia. Blumea 4: 261-319.

Darlington, D. D., 1974. Chromosome botany. London:

Allen and Unwin.

Jackson, R. C., 1960. Supernumerary chromosomes in Haplo-

pappus gracilis. Evolution 14: 135.

John, B., 1976. Population cytogenetics. London: Edward

Arnold.

Kaplan, S. M., 1972. Seed production and sex ratio in anemo-

philous plants. Heredity 28: 281-285.

Mulcahy, D. L., 1967. Optimal sex ratio in Silene alba.

Heredity 22 : 41 1—423.

Muller, J., 1925. Why polyploidy is rarer in animals than

plants. Amer. Nat. 59: 346-353.

Ross, M. D. & Weir, B. S., 1976. Maintenance of males and

females in hermaphrodite populations and the evolution

of dioecy. Evolution 30: 425-441.

Wedberg, H. L., Lewis, H. & Venkatesh, C. S., 1968. Trans-

location heterozygotes and supernumerary chromosomes

in wild populations of Clarkia williamsonii. Evolution

22: 93-107.

Wiens, D., 1975. Chromosome numbers in African and Mada-

gascan Loranthaceae and Viscaceae. Bot. J. Linn. Soc.

71: 295-310.

Wiens, D. & Barlow, B. A., 1973. Unusual translocation

heterozygosity in an East African mistletoe ( Viscum

fischeri). Nature New Biol. 243 : 93-94.

Wiens, D. & Barlow, B. A., 1975. Permanent translocation

heterozygosity and sex determination in East African

mistletoes. Science 187: 1208-1209.

Wiens, D. & Barlow, B. A., 1979. Translocation hetero-

zygosity and the origin of dioecy in Viscum. Heredity 42

(in press).

Wiens, D. &Tolken, H. R., 1979. Viscaceae. In O. A. Leistner,

Flora of Southern Africa. 10,1: 1-59.

Bothalia 13,1 & 2: 171-184 (1980)

Field identification of roots of woody plants of the savanna ecosystem

study area, Nylsvley

M. C. RUTHERFORD*

ABSTRACT

A key for the field identification of fresh root material of 21 woody plant species on the savanna ecosystem

study area, Nylsvley, South Africa, is given. Descriptions of macroscopic features of roots as well as photo-

graphic descriptions of roots and root systems are provided.

rEsum £

IDENTIFICATION SUR LE TERRAIN DE RACINES DE PLANTES LIGNEUSES DANS LA REGION

DE L'ECOSYSTEME DE SAVANE £TUDl£ A NYLSVLEY

On dome une cle pour I’indentification sur le terrain de materiel radiculaire frais concernant 21 especes de

plant es ligneuses rencontrees dans la zone d' etude d'un ecosysteme de savane a Nylsvley en Afrique du Sud. Les

caracteres macroscopiques des racines sont decrits et Ton y joint des photographies des racines et des syst ernes

radiculaires.

INTRODUCTION

During the course of root investigations in the

Burkea africana- dominated savanna of the South

African savanna ecosystem project study area on the

Nylsvley Nature Reserve in the northern Transvaal,

the need for species identification of roots has

frequently arisen. In the nutrient-poor, often

deep (> 2 m) sandy soil of this area, it has been

found that roots of woody plants often extend well

beyond the tree canopy’s ground projection area. In

some species the lateral roots commonly extend

linearly up to seven times the extent of the canopy,

that is, an area about 50 times the area covered by

the canopy. The roots of some species do not often

radiate symmetrically from the stem position, but may

all lie to one side of this position, for example, some-

times in Ochna pulchra. The result is often a relatively

high degree of different species root interpenetration

(interdigitation) that is usually extremely difficult

to predict from only a study of the distribution

of the aboveground parts of the species. Species

identification of roots is needed to increase the quality

of root data derived from techniques including soil

core and block or monolith sampling. Identification

is also important when exposing entire root systems

for determination of root distribution and biomass

distribution of a limited set of species. Here positive

identification is needed to ‘weed out’ roots of other

species that occur in between the roots of the species

under study as field work proceeds. Because many tons

of soil would often need to be carefully removed to

trace a particular root to its parent plant stem for

identification, and more importantly since it has been

found that it is usually impracticable to trace roots of

several species simultaneously without destroying

some of the roots that are still required, root identi-

fication based on easily observable characteristics in

relatively small root samples is required.

The drawing up of keys to identify roots to the

species based on macroscopic features in the field

appears to have been seldom attempted. One work

that has been found provides a key for the identifica-

tion of roots of 1 1 conifer species in certain forest

areas of north-western America (Gilbertson et al.,

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

1961). These workers report not finding any reference

to such identification prior to their work. One of the

great problems, also found by the above quoted

writers, is that diagnostic characteristics for species

identification of roots tend to disappear on small-

diameter roots. Another factor that has undoubtedly

discouraged devising field root identification keys is

that below ground plant organs are usually far less

clearly differentiated compared to the high degree of

above ground plant organ differentiation that pro-

vides many possible diagnostic features for clear

identification. Most taxonomic works on woody plant

species seldom contain descriptions of roots. Although

microscopic features undoubtedly provide much more

information for root identification, such features are

usually impracticable for use in rapid field identifica-

tion.

In the present study, description of species roots is

limited to roots usually greater than 5 mm in diameter.

The species that were included were selected on the

basis of a plant ecological survey carried out by

Coetzee et al. (1976) on the site. All woody species

that occurred with a frequency of > 20% in their

defined “ Eragrostis pallens — Burkea Tree Savanna”

were included. Their sample plots were so large (each

about 0,8 ha) that some of the species with a fre-

quency of > 20% were not common. The 19 species

included were therefore not limited to the most com-

mon species. Since two species ( Lannea edulis and

Fadogia monticola) with a frequency of just under

20% but probably in smaller plots of 0,02 ha, are

known despite their non-woody above-ground habit

to have large and extensive woody below-ground

organs, these two species were also included resulting

in a total species number of 21. These species also

included those 1 1 species which together constituted

more than 98% of the total above-ground woody

plant biomass of the site (Rutherford, 1979). Since not

all woody species are included, however, it is possible

that in some situations, roots of other species may be

present. Therefore inspection for above-ground

evidence of rarer excluded species should be done and

the roots of such species first examined to prevent

misidentification using the key. This principle will

obviously also apply when using the key in other

similar Burkea africana communities. In the event of

a particular difficulty arising when using the key, it

should be borne in mind that often several species

1 72 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA

NYLSVLEY

may be eliminated on their absence (viewed on above-

ground evidence) within a radius of about 50 m from

the sampling point.

Not all underground organs are roots, for example,

the underground stems of Fadogia monticola and

some other “rhizomatous” geoxylic suffrutices (White,

1976). To avoid the cumbersome repetition of “under-

ground organ”, the term “root” is used instead to

indicate that part of the plant below the soil surface.

After much secondary thickening has taken place, the

difference between root and underground shoot

becomes much reduced, so that particularly from a

functional viewpoint, these two organ types may be

regarded as much alike. Pith refers to the apparent

original centre of growth in the root. Outer bark

refers to the outer, dry and usually very thin, layer

around the root. Inner bark refers to the rest of

the bark and is distinguished by being moister and

usually much thicker than the outer bark. In the key

and descriptions, bark thickness (sum of two bark

thicknesses over total root diameter) given is for roots

further away from the stem position and not neces-

sarily for the very thick roots near the plant base.

Often in fresh root cross sections, pores are not visible

or scarcely visible but, upon drying out (for example,

in some of the photographs given), become clearly

visible sometimes together with rays and rings. All

cross section views (cut with pruning shears) and

surface views in the photographs are of roots between

1-3 cm diameter. The root system views are scaled

with a pole marked off in 10 cm divisions. The degree

of tap root development indicated in some of the

photographs depends upon the age of the individual.

In smaller younger trees, tap roots are usually pro-

minent, but in larger, older trees lateral and heart

roots in most species develop far more than does the

tap root. In the selection of root systems for photo-

graphing, juvenile individuals were avoided. Each

root system was exposed by first removing all soil

and roots vertically up to a point 30 cm in front of

the stem(s) position, then using water under pressure

to remove all soil up to a point immediately below

the stem(s) and half a metre on each side of the (main)

stem and down to about one metre depth. All roots

of other species were pruned from this region. The

photographs are only intended to give on a standard

comparative basis, the type of root system in the im-

mediate area of th; stem base and in most cases only

represents a very small proportion of the whole root

system.

The roots of certain species are very variable in

appearance and are sometimes keyed out more on

negative characters than on positive characters.

One character that may be particularly variable is the

intensity (and pattern) of the colour of the outer bark,

which often depends on the moisture content of the

soil at the time of sampling. In a few species the inner

bark or wood of the roots changes colour markedly

after a relatively short period of exposure. Such

characters have been employed in the key, although

this inevitably implies that only totally fresh root

material can be identified and that there are built-in

time delays while using the key and inspecting for

time-dependent colour changes. The key and descrip-

tions do not apply to dead roots.

The following species are included and are listed

alphabetically:

1. Burkea africana Hook. (Caesalpinioideae)

2. Combretum molle R. Br. ex G. Don (Combretaceae)

3. Combretum zeyheri Sond. (Combretaceae)

4. Dichapetalum cymosum (Hook.) Engl. (Dichapetalaceae)

5. Dichrostachys cinerea (L.) Wight & Arn. (Mimosoideae)

6. Dombeya rotundifolia (Hochst.) Planch. (Sterculiaceae)

7. Euclea natalensis A. DC. (Ebenaceae)

8. Fadogia monticola Robyns (Rubiaceae)

9. Grewia flavescens Juss. (Tiliaceae)

10. Lannea discolor (Sond.) Engl. (Anacardiaceae)

11. Lannea edulis (Sond.) Engl. (Anacardiaceae)

12. Ochna pulchra Hook. (Ochnaceae)

13. Ozoroa paniculosa (Sond.) R. & A. Fernandes (Anacar-

diaceae)

14. Parinari capensis subsp. capensis Harv. (Chrysobalan-

aceae)

15. Pygmaeothamnus zeyheri (Sond.) Robyns (Rubiaceae)

16. Securidaca longipedunculata Fresen. (Polygalaceae)

17. Strychnos cocculoides Bak. (Loganiaceae)

18. Strychnos pungens Soler. (Loganiaceae)

19. Terminalia sericea Burch, ex DC. (Combretaceae)

20. Vitex rehmannii Guerke (Verbenaceae)

21. Ximenia caffra Sond. (Olacaceae)

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

Field key to roots of woody species

Bark emits readily recognizable odour of menthyl salicylate (oil of wintergreen)

16. Securidaca longipedunculata

Bark has no discernible odour of menthyl salicylate 2

Outer bark consists of very corky ridges 17. Strychnos cocculoides

Outer bark not markedly corky 3

Bark exudes copious amounts of white latex when cut 13. Ozoroa paniculosa

Bark not exuding white latex when cut 4

Bark papery in very loose multi-layered shedding, pale cream (not brown) layers. . 5. Dichrostachys cinerea

Bark not papery in very loose multi-layered shedding pale cream layers 5

Inner bark a bright orange colour (not reddish) especially when outer bark lightly scratched away with

thumbnail 3. Combretum zeyheri

Inner bark not a bright orange colour 6

Inner bark with uniformly smooth texture in cross section appears yellowish green colour, often becom-

ing a deeper green colour after about one minute exposure. Thumbnail scratch colour not pink to red

20. Vitex rehmannii

Green not the predominant colour of the inner bark 7

Inner layer or all layers of inner bark deep yellow with the inner bark easily stripped in long extremely

tough fibrous lengths (virtually unbreakable by hand tension) 19. Terminalia sericea

Inner bark not deep yellow and not extremely tough 8

Wood pithy and loose grained and often presenting a speckled appearance in cross section

18. Strychnos pungens

Wood not pithy and not speckled in cross section 9

Light yellow colour of inner bark and wood in cross section changes to deep intense yellow very rapidly

within half to one minute after exposure. This deep yellow colour is then in characteristically strong

contrast to the very dark outer bark 7. Euclea natalensis

Inner bark and wood do not clearly change to a deep yellow colour within half to one minute after

exposure 10

Wood in cross section has very characteristic alternation of light yellow and dark yellow wedges

4. Dichapetalum cymosum

Wood in cross section does not have characteristic alternation of light and dark coloured wedges 11

M. C. RUTHERFORD

173

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Markedly eccentric radial growth often present (test at several spots on the root sample) with the pith

often on the extreme edge of the wood. Wood cross section with alternating clear lighter and darker

yellow bands (not thin rings), outerbark very dark brown often with reticulate striations, or grooved

9. Grewia flavescens

Radial growth usually not extremely eccentric and no alternating bands in wood cross section and

outerbark not very dark brown 12

Non-continuous orange strands set amongst yellowish (not reddish) tissue visible when deep scratches

are made through the outer bark well into the inner bark 15. Pygmaeothamnus zeyheri

Orange strands set amongst yellow tissue not visible in deep scratches into the inner bark 13

Outer and inner bark together extremely thin 14

Outer and inner bark together not extremely thin 15

Protrusions on outer bark in whorls of three and all lying in the same direction longitudinally

8. Fadogia monticola

If outer bark protrusions present, not in whorls of three 2. Combretum molle

Roots usually distinctly contorted, outer bark light brown to reddish brown and easily visible fine alter-

nating radial lines of lighter and darker yellow in cross section of the very close grained wood ....

12. Ochna pulchra

Roots seldom contorted and outer bark not light brown to reddish brown and fine alternating radial

lines of lighter and dark yellow in cross section of wood not easily visible 16

Inner surface of inner bark (at cambium layer) seen when bark partially stripped away from wood clearly

changes from whitish to a deep brownish yellow within three minutes of exposure

6. Dombeya rotundifolia

Inner surface of inner bark not clearly changing from whitish to deep brownish yellow within three

minutes of exposure 17

Outer bark overall colour whitish grey (not cream) with brown blotches, roots may be contorted. Woody

cambium layer often markedly striated, almost furrowed. Appearance of pores in wood cross

section are often characteristic 14. Paranari capensis

Outer bark not lighter than a creamish colour; if whitish usually without very obvious regular pattern of

lenticellular structures; wood at cambium layer usually not markedly striated so as to almost form

furrows 18

Bark thickness > 25 % of root diameter; inner bark fleshy with relatively high water content 19

Bark thickness < 25 % of root diameter; inner bark not particularly fleshy 1 . Burkea africana

Inner bark thick and rubbery and bark will not strip in long lengths but breaks off leaving clear yellow

or orange fibres extending from inner half of inner bark 21. Ximenia caffra

Inner bark not rubbery and bark will peel in long lengths 20

Bark often peels closely away from wood; outer bark often cracking in large block shapes; Pale tan

coloured lenticellular structures often present on outer bark; pith often softer than remainder of

wood 11. Lannea edulis

Bark often starts to peel cleanly away from the wood but then leaves behind some innerlayered bark

attached to the wood; outer bark seldom cracking in block shapes; lenticellular structures rare or

absent; pith usually not softer than rest of wood 10. Lannea discolor

1 . Burkea africana Hook. (Figs 1 & 2)

Outer bark: Dark brown to cream; lenticellular

structures often present and then scattered fairly

evenly over bark and are a darker colour than mo t

of the background. When bark is dark brown it often

has a reticulate pattern of lighter and darker parts.

When root is cream, bark is usually smoother tut

with many large lenticellular structures almost always

present. There is a full gradation between these two

root appearances. Inner bark: White on thin roots

(<1 cm diameter) through light salmon pink to purple

pink but a very pale colour on the innermost parts.

Fig. 1. — Root system of a Burkea africana individual 5,6 m

high with a stem diameter of 13,3 cm at 20 cm above

ground.

Fig. 2. — Exterior view and cross-section of (a, b) the darker

striated type and (c, d) the light coloured lenticellular

type of Burkea africana root.

Thumbnail scratch colour has the same range. Bark

not hard or particularly soft, not fleshy, thickness

commonly between 10 and 25% of root diameter.

Samples from 1 5 trees gave a mean bark thickness as

a percentage of root diameter as 15,1 with a standard

deviation of 2,3. Bark peels relatively easily but strips

in long lengths usually only on thin roots, bark of

medium-tensile strength. Wood: Creamy white to

yellow. Pith usually coloured differently to rest of

174 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA,

NYLSVLEY

wood, pith colour varies from very light salmon pink

through to purple pink which may spread over much

of the wood section. Wood hard. Rings usually

evident, rays not readily visible. Wood at cambium

layer generally smooth or with small striations.

General: Root system with lateral branching at base;

roots usually fairly straight.

Because the appearance of Burkea africana roots

may vary considerably, they are sometimes difficult

to identify. Inner bark and wood colouration is usually

distinctive, but may be confused with that of Ximenia

caffra, aid the two Lannea species, but it has a

significantly (P =0,001) lower relative bark thickness

than that of these three species and it has a usually

drier inn^r bark than that of these species.

2. Combretum mode R. Br. ex G. Don (Figs 3 & 4)

Outer bark: Variable blotches of brown and dark

blackish brown, golden brown and cream; irregular

protrusions sometimes present; golden brown single

layer papery flakes sometimes present. Inner bark:

No inner bark apparent; thumbnail scratch colour

shows the yellow wood ; bark peels very difficultly and

does not strip in long lengths, bark thickness com-

Fig. 3. — Root system of a two-stemmed individual of Com-

bretum molle 5,7 m high with stem diameters of 12,2 cm

and 1 2 , 5 cm at 20 cm above ground.

Fig. 4. — Exterior view and cross-section of a Combretum

molle root.

monly not more than 5% of root diameter. Wood:

Yellow pith colour often slightly darker than that of

wood; hard; rings with pores sometimes clear, rays

not clearly visible; wood at cambium layer generally

smooth. General: Root system extensively and

laterally branched at base; roots usually fairly

straight.

Extremely thin bark with a lack of protrusions

arranged in whorls characterizes this species.

3. Combretum zeyheri Sond. (Figs 5 & 6)

Outer bark: Varying proportions of cream and

brown blotches; usually smooth with few or no pro-

trusions; usually has powdery fine scales that may be

rubbed off slightly. Inner bark: Orange; thumbnail

scratch colour very bright orange ; not hard, medium

fleshy; bark thickness commonly between 20 and

35% of root diameter; bark removable but not strip-

ping in long lengths; low tensile strength. Wood: Yel-

lowish to orangish; pith colour sometimes reddish;

wood fairly close-grained; pores usually visible; ring-

structures usually evident, rays not; wood at cambium

layer with imperfectly parallel striations. General:

Root system with branching near base; roots usually

fairly straight.

Fig. 5. — Root system of a Combretum zeyheri individual 6,3 m

high with a stem diameter of 15,8 cm at 20 cm above

ground.

Fig 6. — Exterior view and cross-section of a Combretum

zeyheri root

M. C. RUTHERFORD

175

The bright orange colour that appears with the

thumbnail scratch mark is unmistakable and clearly

distinguishes this from the other species.

4. Dichapetalum cymosum {Hook.) Engl. (Figs 7

& 8)

Outer bark: Creamy yellow, protrusions sometimes

present; fairly smooth with some evenly coloured

reticulation pattern sometimes present.

Inner bark: White to light yellow; thumbnail scratch

colour usually white; bark not hard nor particularly

soft; not fleshy; thickness commonly between 10 and

20% of root diameter; bark peels relatively easily, but

usually only strips in short lengths; bark of low

tensile strength. Wood: Light and dark yellow; wood

not hard; cross-section shows regular alternation of

dark yellow and light yellow wedges; wood at cam-

bium layer has striations that change in width.

General: Root system with shallow lateral connections

between individuals; roots usually less than 1 cm in

diameter and fairly straight.

Fig. 8. — Exterior view and cross-section of a Dichapetalum

cymosum root.

Wood cross-section with its very characteristic

alternation of light yellow and dark yellow wedges

and the non-corky outer bark characterize/), cymosum.

5. Dichrostachys cinerea (L.) Wight & Am.

(Figs 9 & 10)

Outer bark: Cream; no protrusions; very loose

flakes layered on top of one another forming irregular

stacks; outer bark usually as thick as inner bark.

Inner bark : Greenish white; thumbnail scratch colour

whitish, sometimes greenish; bark soft, not fleshy;

total bark thickness in the order of ‘.0/o of the root

diameter (flaking makes measurement approximate);

bark peels easily, stripping in long lengths; at least

medium tensile strength. Wood: Orangy yellow; wood

close-grained ; rings sometimes evident, rays not ; wood

at cambium layer has orange and yellow striations.

General: Root system extensively branched at base;

roots usually straight.

Fig. 9.— Root system of a two-stemmed individual of Dichro-

stachys cinerea 2,8 m high, with stem diameters of 5,7 and

4 , 7 cm at 20 cm above ground.

Fig. 10. — Exterior view and cross-section of a Dichrostachys

cinerea root.

This species is clearly distinguished from Combretum

molle, which may also have papery bark, by its

cream bark colour, thicker bark and multi-layered

papery bark layers.

176 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA,

NYLSVLEY

6. Dombeya rotundifolia ( Hochst .) Planch. (Figs

11 & 12)

Outer bark: Creamy with various usually brownish

blotches; lenticellular protrusions often in short rows

at right angles to the longitudinal axis of the root.

Inner bark: Light pink (sometimes very light pink

with orange spots under protrusions) in the outer area

to clear, reddish or whitish in the inner area; the pink

in the outer areas is often concentrated in longitudinal

‘canals’; thumbnail scratch colour pale pink to very

bright red; bark not hard, fairly fleshy; thickness

commonly between 30 and 50% of the root diameter;

bark peels relatively easily and strips in long lengths;

medium tensile strength, but breakable by hand.

Wood: Pale yellow; pith colour sometimes pink,

particularly in those roots with deeper coloured bark;

wood close-grained; very fine pores sometimes visible;

rings sometimes clear, rays less evident; wood at

cambium layer generally smooth or finely striated.

General: Root system extensively branched at base;

roots usually straight.

This species may be confused with Ximenia caffra,

Lannea edulis and Lannea discolor and some forms of

Burkea africana roots. However, the inner surface of

the inner bark seen when the bark is partially stripped

Fig. II. — Root system of a Dombeya rotundifolia individual

6,2 m high with a stem diameter of 25,6 cm at 20 cm

above ground.

Fig. 12. — Exterior views and cross-section of a Dombeya

rotundifolia root. A root with wood freshly exposed and

exposed for 3 minutes indicates the differential colouration

away from the wood changes from whitish to a deep

brownish yellow within three minutes of exposure.

The yellow colour of the wood also deepens with

exposure. This character distinguishes Dombeya

rotundifolia from the associated species.

7. Euclea natalensis A. DC. (Figs 13 & 14)

Outer bark: Blackish, dark brown to very dark

brown; usually no protrusions; smooth but with

slight reticulate longitudinal fissures. Inner bark:

Light yellow, becoming deep yellow within half-a-

minute to one minute of exposure; thumbnail scratch

colour light yellow, deeper scratch light yellow be-

coming dark yellow half-a-minute to one minute of

exposure; orange strands visible in deeply scratched

bark but within half-a-minute of exposure change

together with the rest of the bark to a deep yellow

colour; bark not particularly soft or hard; fairly

fleshy; thickness commonly between 30 and 40% of

the root diameter; bark peels relatively easily, but

does not strip in long lengths; low tensile strength.

Wood: Light yellow becoming very deep yet bright

yellow within half-a-minute to one minute of ex-

posure; pith sometimes slightly darker than rest of

wood; wood close-grained; rings may be visible, but

rays not obvious; wood at cambium layer has stria-

tions more clearly seen within the first half minute of

Fig. 13. — Root system of a multi-stemmed individual of

Euclea natalensis 1,5m high.

Fig. 14. — Exterior view and cross-section of a Euclea natalensis

root.

M. C. RUTHERFORD

177

exposure. General: Root system extensively branched

at base; roots usually fairly straight.

Inner bark and wood colour changes form light

yellow to deep yellow on exposure at a rate so rapid

that one can see the colour changing; this character

distinguishes E. natalensis from the other savanna

species.

8. Fadogia monticola Robyns (Figs 15 & 16)

Outer bark : Light brown to brown; many regular

protrusions all facing in one direction longitudinally

and positioned in whorls of three; longitudinal fissures

usually present. Inner bark: Too thin to see colour;

thumbnail scratch colour shows whitish, light yellow

colour of wood; bark not fleshy; thickness commonly

less than 5% of root diameter; bark removes dif-

ficultly and does not strip in long lengths. Wood:

Yellow; close-grained; rings and rays occasionally

visible; wood at cambium layer usually smooth.

General: Root system with shallow lateral connec-

tions between individuals; roots fairly straight to

contorted.

Fig. 15. — Root systems of interconnected individuals of

Fadogia monticola 0,5 m high.

Fig. 16. — Exterior view and cross-section of a Fadogia monticola

root.

Protrusions in whorls of three characterize this

species.

9. Grewia flavescens Juss. (Figs 17 & 18)

Outer bark: Dark brown to blackish brown; pro-

trusions usually not obvious; reticulate pattern often

present, sometimes grooved. Inner bark: Pinkish;

thumbnail scratch colour pale pink to white; bark not

particularly soft or hard; not fleshy; very fibrous;

thickness very variable and depends on position of

eccentricity of radical growth; bark peels easily and

strips in long lengths; high tensile strength. Wood:

Deep yellow; wood hard and close-grained; radial

growth in most cases markedly eccentric, with the

pith often at the edge of the wood; the eccentric

growth is also reflected in the bark; pores sometimes

visible; wood cross-section with alternating lighter

and darker eccentric yellow bands; usually no rays

visible; wood at cambium layer fairly smooth or with

light striations. General: Root system with much

branching at base; roots usually straight.

Fig. 17. — Root system of a multi-stemmed individual of

Grewia flavescens 1,6m high.

Fig. 18. — Exterior view and cross-section of a Grewia flavescens

root.

Usually markedly eccentric radial growth together

with lighter and darker alternating yellow bands in

wood cross-section together with the usually dark

brown outer appearance of the roots distinguishes

this species from Terminalia sericea.

10. Lannea discolor ( Sond .) Engl. (Figs 19 & 20)

Outer bark: Creamy blotches, though sometimes

fairly evenly light brown to cream; usually at least

some irregularly distributed rounded protrusions (cor-

responding to protrusions of the wood at the cambium

layer); usually very few such protrusions on thinnest

178 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA

NYLSVLEY

Fig. 21. — Root systems of interconnected individuals of

Lannea edulis.

Fig. 22. — Exterior view and cross-section of a Lannea edulis

root.

rotundifolia; its inner bark is thicker than that of

Burkea africana. The bark of this species usually peels

in longer lengths than that of Ximenia cajfra. Some-

times, however, it is difficult to distinguish this

species from Lannea edulis, although Lannea discolor

roots typically do not have the outer bark cracking

in large regular block shapes, the bark often does

not peel more cleanly away from the wood, pale

tan coloured lenticellular structures are seldom pre-

sent, the wood at the cambium layer has slightly less

pronounced pale thread-like striations and the pith

is usually not softer than the rest of the wood.

1 1. Lannea edulis ( Sond .) Engl. (Figs 21 & 22)

Outer bark: Creamy to dark brown; often have

large rounded protrusions and pale light brown

lenticels with no corresponding marks on the wood

surface below the bark; bark often cracking in large

square patches on thicker roots. Inner bark: Pale to

dark pink or purple and red on the outside to white

on the inside; red and white longitudinal canals in

the outer layers together with large orange specks;

thumbnail scratch colour pale pink through to bright

red; bark soft, very fleshy, thickness commonly in

the order of 25 to 40% of root diameter; bark peels

relatively easily and strips in long lengths without

radial taper; medium tensile strength but breakable

Fig. 20. — Exterior view and cross-section of a Lannea discolor

root.

(less than 5 mm diameter) roots; bark often flaking

slightly. Inner bark: Reddish purple in the outer

layers to pinkish, yellow, light yellow to yellowish

clear in the inner layers; in very thin roots no red

colour in inner bark only a uniform white colour;

alternation of reddish and white longitudinal strands

(occasionally with irregular large orange specks set

in between) in outer layers; thumbnail scratch colour

bright reddish purple, sometimes pink; bark soft;

fleshy with relatively high water content; thickness

commonly between 30 and 50% of root diameter;

bark peels easily, although often tearing into a taper-

ing (in thickness) strip; bark stripping in long lengths

on thin roots; fairly high tensile strength. Wood: Yel-

lowish white; on thicker roots purplish pink pith;

close-grained; very fine pores sometimes visible in

cross-section; rings sometimes visible; rays not

obvious; wood at cambium layer generally smooth,

but with very fine parallel striations. General: Root

system with extensive lateral branching; roots usually

fairly straight.

This species may be confused with Dombeya

rotundifolia, Ximenia cajfra and Lannea edulis as well

as some forms of Burkea africana roots. However,

it does not have the colour reaction of Dombeya

Fig. 19. — Root system of a Lannea discolor individual 6,7 m

high with a stem diameter of 28,7 cm at 20 cm above

ground.

M. C. RUTHERFORD

179

by hand. Wood: Pale yellow to yellow; pith usually

light pink (pith colour sometimes spreading through

most of the wood) and appears to be softer or more

loosely packed than the rest of the wood; ring

structures and rays not readily visible; pores some-

times clear in cross-section; wood at cambium layer

has fine raised whitish thread-like striations. General:

Root system with lateral connections between indivi-

duals; roots usually fairly straight.

This species may be confused with Dombeya

rotundifolia, Ximenia caffra and Lannea discolor

as well as some forms of Burkea africana roots.

However, it does not have the colour reaction of

Dombeya rotundifolia; its bark is thicker than that of

Burkea africana. The bark of this species usually peels

in longer lengths than that of Ximenia caffra. Some-

times, however, it is difficult to distinguish this species

from Lannea discolor, although Lannea edulis roots

typically have the outer bark cracking in large

rectangular block shapes, the bark often peels more

cleanly away from the wood, pale tan coloured

lenticellular structures are often present and the wood

of the cambium layer has slightly more pronounced

striations than that of Lannea discolor.

12. Ochna pulchra Hook. (Figs 23 & 24)

Outer bark: Light brown to reddish brown; no

regular protrusions; sometimes large flakes present

especially on thicker roots. Inner bark: Purple to

pink; thumbnail scratch colour reddish purple; bark

Fig. 23. — Root systems of individuals of Ochna pulchra with

mean height of 2,2 m and mean diameter at 20 cm above

ground of 6,4 cm.

Fig. 24. — Exterior view and cross-section of an Ochna pulchra

root.

hard; not fleshy; thickness commonly in the order

of 25% of root diameter; bark removal possible but

not readily and does not strip in long lengths; bark

of low tensile strength. Wood: Light yellow; pith

often pinkish or purplish; wood close-grained; rings

usually not obvious but distinctly visible fine lines of

alternating lighter and darker yellow rays; wood at

cambium layer generally smooth with very fine stria-

tions. General: Root system very variable but usually

with strong lateral development at about 30 cm depth

with connections between individuals sometimes pre-

sent; roots usually contorted.

Sometimes may be confused with roots of Burkea

africana, mainly based on similar colouration, how-

ever, Burkea africana roots are very seldom contorted

and do not have easily discernible wood rays.

13. Ozoroa paniculosa ( Sond .) R. & A. Fernandes

(Figs 25 & 26)

Outer bark: Creamy light brown; protrusions usu-

ally present in irregular pattern; smooth sometimes

with a trace of reticulated fissures. Inner bark:

Reddish; thumbnail scratch colour is pinkish orange

or reddish with latex if scratch not very shallow;

copious exudation of white latex from cut surfaces;

bark soft, fleshy and thickness commonly in the order

Fig. 25. — Root system of a multi-stemmed individual of

Ozoroa paniculosa 1,3m high.

Fig. 26.— Exterior view and cross-section of an Ozoroa pani-

culosa root.

1 80 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA,

NYLSVLEY

of 40% of root diameter; bark peels relatively easily

but does not strip in long lengths; bark of low tensile

strength. Wood: Light yellow; pith often pinkish;

wood hard; rings and rays not always readily visible;

wood at cambium layer with fine striations. General:

Root system with lateral branching at base and roots

usually straight.

The copious exudation of white latex from severed

bark characterizes this species.

14. Parinari capensis Harv. (Figs 27 & 28)

Outer bark: Whitish grey with brown blotches;

usually no protrusions; smooth with small wrinkles

in places. Inner bark: Reddish purple; thumbnail

scratch colour purple but changing to light salmon

pink on thinner roots; bark neither hard nor soft and

not fleshy; thickness commonly in the order of 20 to

25% of the root diameter; bark comes away relatively

easily but does not strip in long lengths; bark of low

tensile strength. Wood: Reddish near the centre to

whitish on the outside; pith often reddish; wood

close-grained; many evenly spread pores clearly

visible in cross-section; few broad bands of rings

sometimes visible; rays not readily visible; wood at

cambium layer strongly striated with very clear and

parallel striations (almost furrowed) and sometimes

Fig. 27. — Root systems of interconnected individuals Parinari

capensis.

with reddish irregular patches on the wood parti-

cularly on thicker roots. General: Root system with

lateral connections between individuals; roots some-

times contorted.

Possibly may be confused with the whitish form of

Burkea africana roots, hoewever, Burkea does not

have the same striations. Freshly excavated roots of

Parinari capensis may then be separated ' from all

other species on the basis of the overall colour of the

outer bark being more white than cream and on

having almost furrowed striations on the wood

surface.

15. Pygmaeothamnus zeyheri ( Sond .) Robyns

(Figs 29 & 30)

Outer bark: Brown to dark brown; irregular

rounded protrusions usually present. Inner bark: Yel-

low with orange specks; thumbnail scratch colour

yellow, deeper scratches expose longitudinal non-

continuous orange strands set in yellow (pale orange

in thinner roots); bark neither hard nor soft and not

fleshy; thickness commonly in the order of 20-25%

of the root diameter; bark removable but does not

strip in long lengths; low tensile strength. Wood:

Yellow; pith sometimes purplish red and sometimes

only a dark spot; wood close-grained; rings not

readily visible but rays sometimes visible; wood at

cambium layer has clear but often fine striations.

Fig. 29. — Root systems of interconnected individuals of

Pygmaeothamnus zeyheri.

Fig. 28. — Exterior view and cross-section of a Parinari capensis

root.

Fig. 30. — Exterior view and cross-section of a Pygmaeothamnus

zeyheri root.

M. C. RUTHERFORD

181

General: Root system with shallow lateral connections

between individuals; roots sometimes contorted.

May be confused with Lannea discolor, Lannea

edulis, Ximenia caffra or Dombeya rotundifolia. How-

ever, all these species have orange specks set in or

positioned near reddish-pinkish tissue in the bark and

the shape of the orange specks are often more round

than those in Pygmaeothamnus zeyheri where they

form longer strands set in yellowish (not reddish)

tissue.

16. Securidaca longipedunculata Fresen. (Figs 31

& 32)

Outer bark: Uniform light to deep yellow colour;

protrusions very seldom present; smooth but some-

times with circumferential wrinkles on the thicker

roots. Inner bark: Yellow, thumbnail scratch colour

very light yellow; bark soft and fairly fleshy; thickness

commonly between 50 and 70% of the root diameter;

bark with very distinctive smell of menthyl salicylate

(oil of wintergreen) ; bark peels easily and sometimes

strips in long lengths; low to medium tensile strength.

Wood: Yellow but darker than the bark; not par-

ticularly close-grained; pores visible in cross-section;

rings and rays not obvious; wood at cambium layer

smooth or sometimes very slightly striated. General:

Root system mainly a tap root but with several

lateral roots. Roots normally straight.

Fig. 31. — Root system of a Securidaca longipedunculata indi-

vidual 6,1 m high with a stem diameter of 19,5 cm at

20 cm above ground.

Fig. 32. — Exterior view and cross-section of a Securidaca

longipedunculata root.

Cannot be confused with any other species, because

of the distinctive smell of menthyl salicylate in the

bark.

17. Strychnos cocculoides Bak. (Figs 33 & 34)

Outer bark: Yellowish beige; deeply furrowed

corky bark. Inner bark: Creamy yellow; inner bark

only ^ to \ as thick as the outer bark; bark very

soft, not fleshy; bark thickness commonly in the order

of 35% of root diameter; bark does not remove

easily and does not strip in long lengths; very low

tensile strength. Wood: Creamy yellow; pith often a

darker yellow than the wood; wood close-grained;

ring structures clear and light yellow radial lines

alternate regularly with darker yellow wedges; wood

at cambium layer appears smooth. General: Root

system with tap root and relatively little lateral

branching at base; roots usually straight.

Fig. 33. — Root system of a Strychnos cocculoides individual

7,2 m high with a stem diameter of 30,0 cm at 20 cm

above ground.

Fig. 34. — Exterior view and cross-section of a Strychnos

cocculoides root.

May always be distinguished from other species by

the extremely thick and corky outer bark.

1 82 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREA

NYLSVLEY

18. Strychnos pungens Soler. (Figs 35 & 36)

Outer bark: Light brown; protrusions sometimes

present; fairly smooth but sometimes wrinkled on

thicker roots with a slightly reticulated furrow pattern

on thinner roots. Inner bark: Yellow; bark thickness

commonly in the order of 20 to 40% of root diameter;

bark removable and does not strip in long lengths;

low tensile strength. Wood: Creamy yellow to yellow:

wood not close-grained and very pithy, sometimes

monocotyledonous appearance in cross-section; no

rings or rays readily visible; wood at cambium layer

striated. General: Root system a tap root with

relatively little lateral branching; roots straight to

curved.

Fig. 35. — Root system of a Strychnos pungens individual 6,2 m

high with a stem diameter of 28,5 cm at 20 cm above

ground.

Fig. 36. — Exterior view and cross-section of Strychnos pungens

root.

May always be distinguished from the other species

by its particularly pithy wood.

19. Terminalia sericea Burch, ex DC. (Figs 37 & 38)

Outer bark: Cream to brown; sometimes with

numerous pale lenticellular type structures; some-

times with longitudinal fissures. Inner bark: Yellowish

but on thicker roots purple in the outer layers and

yellow in the inner layers; thumbnail scratch slight

pinkish to dark purplish; bark is fairly soft, not

fleshy; thickness commonly between 30 and 50% of

the root diameter; bark peels relatively easily and

strips in long lengths; very high tensile strength.

Wood: Yellow; pith sometimes darker shade than

wood; wood hard; ring structures usually clear but

rays not readily visible; wood at cambium layer

smooth on thin roots but with definite striations on

thicker roots. General: Root system with lateral

branching usually also with very shallow lateral roots

[also found by Cole & Brown (1976) in Botswana];

roots generally straight.

Fig. 37. — Root system of a Terminalia sericea individual

5,7 m high with a stem diameter of 12,9 cm at 20 cm

above ground.

Fig. 38. — Exterior view and cross-section of a Terminalia

sericea root.

Thinner roots may be sometimes confused with

those of Grewia flavescens, however, the wood of

Terminalia sericea, seen in cross-section, does not

have lighter and darker yellow bands.

M. C. RUTHERFORD

183

20. Vitex rehmannii Guerke (Figs 39 & 40)

Outer bark: Yellowy cream; no obvious protru-

sions; smooth with some small very soft flakes.

Inner bark: Uniformly smooth when cut in cross-

section; almost a translucent yellowish green, the

green sometimes becoming more intense with ex-

posure; thumbnail scratch colour is white; also often

becoming greenish with exposure ; soft and very fleshy ;

thickness commonly between 30 and 40% of the root

diameter; bark separates easily but does not strip

in long lengths; low tensile strength. Wood: Yellow;

wood hard; rings and rays sometimes obvious; wood

at cambium layer has fine striations. General: Root

system extensively branching at base; roots usually

fairly straight.

Fig. 39. — Root system of a Vitex rehmannii individual 5,1 m

high with a stem diameter of 8 , 2 cm at 20 cm above ground.

Fig. 40. — Exterior view and cross-section of a Vitex rehmannii

root.

May always be distinguished from other species by

the definite green colour contained in the fleshy inner

bark. The bark is also non-flaky and has no discern-

ible smell of menthyl salicylate.

21. Ximenia caffra Sond. (Figs 41 & 42)

Outer bark: Blotchy creamy to light brown; pro-

trusions sometimes present; fairly smooth; some-

times with faint longitudinal wavy lines. Inner bark:

Reddish purple on thicker roots to white on thinner

roots with the deeper colour outside and paler towards

the centre; thumbnail scratch colour white (on small

roots) through pale pink to pinkish purple (usually

mottled); deep scratch mark shows fine orange

strands or yellow fibres set in red or pink tissue; bark

soft to very soft, rubbery and fleshy; thickness com-

monly between 25 and 45% of the root diameter;

peels relatively easily but does not strip in long

lengths (breaks off in short chunks leaving yellowish

orange or yellow strands emerging from those layers

of the inner bark closer to the wood); low tensile

stength. Wood: Deep yellow to yellowish brown;

wood hard; rings and rays seldom obvious but rings

visible on thicker roots; very many fine pores usually

visible in cross-section; wood at cambium layer has

clear fine parallel striations; pith sometimes darker

on thicker roots. General: Root system with branching

at base; roots usually fairly straight; fleshy root galls

sometimes on roots.

Fig. 41. — Root system of a multi-stemmed individual of

Ximenia caffra 1,2 m high.

Fig. 42.— Exterior view and cross-section of a Ximenia caffra

root.

1 84 FIELD IDENTIFICATION OF ROOTS OF WOODY PLANTS OF THE SAVANNA ECOSYSTEM STUDY AREAS,

NYLSVLEY

The roots of this species may be confused with

those of Dombeya rotundifolia, Lannea discolor, Lan-

nea edulis and some forms of Burkea africana. How-

ever, it does not have the colour reaction of Dombeya

rotundifolia, its bark is thicker than that of Burkea

africana and it can be distinguished from both species

of Lannea by having a more rubbery inner bark,

that will not strip in long lengths but often breaks off

in short chunks with characteristic yellow-orange or

yellow strands or fibres emerging from the inner bark.

ACKNOWLEDGEMENTS

I thank Messrs P. S. Carr and M. D. Panagos for

assistance in the fieldwork and Mr R. P. Ellis for sug-

gestions on the root descriptions.

UITTREKSEL

’n Sleutel vir die uitkenning in die veld van vars

wort el materiaal van 21 houtagtige plantsoorte op die

savanne ekosisteemprojek studie area, Nylsvley, Suid-

Afrika, word gegee. Beskrywings van makroskopiese

eienskappe van wortels asook fotografiese beskrywings

van wortels en wortelstelsels word aangetoon.

REFERENCES

Coetzee, B. J., Van der Meulen, F., Zwanziger, S., Gon-

salves, P. & Weisser, P. J., 1976. A phytosociological

classification of the Nylsvley Nature Reserve. Bothalia 12:

137-160.

Cole, M. M. & Brown, R. C., 1976. The vegetation of the

Ghanzi area of western Botswana. J. Biogeog. 3 : 169-196.

Gilbertson, R. L., Leaphart, C. D. & Johnson, F. D., 1961.

Field identification of roots of conifers in the Inland

Empire. For. Sci. 1 : 352-356.

Rutherford, M. C., 1979. Aboveground biomass subdivisions in

woody species of the savanna ecosystem project study

area, Nylsvley. South African National Scientific Pro-

grammes Report No. 36, pp 33.

White, F., 1976. The underground forests of Africa: a pre-

liminary review. Gardens' Bulletin 29: 57-71.

Bothalia 13,1 & 2: 185-189 (1980)

Leaf anatomy of the South African Danthonieae (Poaceae).

II. Merxmuellera disticha

R. P. ELLIS*

ABSTRACT

The anatomical structure, of the leaf blade as seen in transverse section, and of the abaxial epidermis, of

Merxmuellera disticha (Nees) Conert is described and illustrated. Three distinct anatomical “forms” are recog-

nized viz. typical M. disticha, the Drakensberg form and the alpine bog form. These three anatomical groups

also appear to have differing environmental requirements and probably warrant taxonomic status.

RESUME

ANATOMIE FOLIAIRE DES DANTHONIEAE ( POACEAE ) D’AFRIQUE DU SUD. II.

MERXMUELLERA DISTICHA

La structure anatomique du limbe foliaire en coupe transversale et celle de I'epiderme abaxial de Merxmuellera

disticha (Nees) Conert sont decrites et illustrees. On reconnait trois "formes” anatomiques distinctes, soit la

forme M. disticha typique, la forme du Drakensberg et la forme du marecage alpin. Ces trois groupes anatomiques

semblent egalement avoir des exigences de milieu dijferentes et il est probable qu’elles meritent un statut taxono-

mique.

INTRODUCTION

Merxmuellera disticha (Nees) Conert (1970) ( =Dan -

thonia disticha Nees) is probably the best known and

most distinctive southern African representative of

this genus. It is a wiry, tussock grass and is economi-

cally relatively important (Acocks, 1971) as it may

become dominant, and completely usurp the position

of better grazing grasses in certain areas. It occurs over

extensive areas along the south coast and eastern

mountain ranges and is an important constituent of

the following veld types: Themeda — Festuca Alpine

Veld, Stormberg Plateau Sweetveld and Karroid

Merxmuellera Mountain Veld (Acocks, 1975).

This species is easily recognized by the inflorescence,

which is an oblong, uninterrupted, distichous spike.

It is the only species of this genus which can have

2-flowered spikelets and both the upper and lower

glumes 3-nerved (Chippindal, 1955). M. disticha is

thus distinct morphologically and only the plants with

2-flowered spikelets may be confused with Pentaschis-

tis basutorum Stapf (Chippindal, 1955).

It was, therefore, most unexpected to discover that

three distinct anatomical “forms” are present in this

species. In addition, from the sample examined in this

study, it appears that each of these three “forms” has

different habitat requirements. For convenience,

M. disticha sens. lat. has been sub-divided into three

“forms” in the following descriptions and discussions:

typical M. disticha, the Drakensberg form and the

alpine bog form. Each of these forms exhibits charc-

teristic leaf anatomy and epidermal structure.

In the anatomical descriptions which follow, the

following abbreviations will be used:

vb/s — vascular bundle/s

l’vb/s — first order vascular bundle/s

2’vb/s — second order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — outer bundle sheath; parenchyma sheath.

For definitions of terminology used see Ellis

(1976, 1979).

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

ANATOMICAL DESCRIPTION OF MERXMUELLERA

DISTICHA SENS. LAT.

Leaf in transverse section

Leaf outline: infolded with an elliptical or U-shaped

outline. Symmetry about the median bundle not

perfect with the l’vbs of opposite halves of the lamina

alternating. This tendency particularly marked in the

Drakensberg form (Fig. 4). Leaves narrow (< 1 , 1 mm

wide) when folded. Adaxial channel normally a deep,

narrow cleft. Ribs and furrows: adaxial furrows

narrow, cleft-like and located between all vbs except

the penultimate 3’vb which lies at the base of a furrow.

Adaxial ribs flat-topped and angular; l’vb ribs larger

than those overlying 3’vbs. Abaxial surface without

ribs or furrows. Median vascular bundle: present;

indistinguishable structurally from other l’vbs. Vascu-

lar bundle arrangement: no 2’vbs present; one 3’vb

between consecutive l’vbs; l’vbs decrease in size

towards margin. Vascular bundle structure: vbs

circular or elliptical in shape; phloem adjoins ibs;

lysigenous cavity and protoxylem vessels present;

metaxylem vessels circular, extremely narrow with

thickening of the walls. Vascular bundle sheaths: obs of

3’vbs circular or with slight abaxial interruptions in the

Drakensberg form (Fig. 5); no sheath extensions.

Cells rounded or elliptical, inconspicuous and much

smaller than the mesophyll cells; thin walled; chloro-

plasts few or absent. Ibs indistinct, complete and with

uniformly thickened walls. Obs of l’vbs elliptical or

horse-shoe shaped with wide adaxial and abaxial

interruptions; no extensions. Sheath cells incon-

spicuous; smaller than both the mesophyll and the

ibs cells; rounded or, more often, elliptical; thin-

walled or may resemble ibs cells with u-shaped wall

thickenings in typical M. disticha only (Fig. 1). Few or

no chloroplasts present. Ibs complete; inner radial

and tangential walls thickened; often adaxial ibs

cells larger than lateral cells. Sclerenchyma: small,

shallow adaxial strands associated with 3’vbs.

Inversely anchor- or T-shaped adaxial girders associated

with l’vbs; width and length of girder stem variable

(Table 1); girder interrupts obs. Abaxial girders not

associated with 3’vbs except in Drakensberg form

(Figs 4 & 5). l’vb girders usually trapezoidal narrow-

ing toward the bundle; fibres interrupt the obs.

Continuous abaxial hypodermal sclerenchyma well

186 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). II. MERXMUELLERA DISTICH A

Figs 1-9. — Leaf blade outline of Merxmuellera disticha sens lat. in transverse section. 1-3, typical M. disticha form.

All X 160. (1, Ellis 2572; 2, Ellis 2564; 3, Ellis 669.) 4-6, Drakensberg form. All x 160. (4, Du Toit 675; 5, Ellis

1404; 6, Ellis 3152.) 7-9, Alpine bog form. All x 250. (7, Ellis 3315; 8, Ellis 3306; 9, Ellis 3316.)

developed in typical M. stricta (Fig. 1) but only a

shallow interrupted hypodermal layer present in the

alpine bog form (Figs 7-9) (Table 1). Fibres with both

lignified and cellulose cell walls present in individual

sections. Margin: relatively small, pointed, scleren-

chyme cap developed. Mesophyll: not radiate; cells

isodiametric, regular and tightly packed. Mesophyll

tissue not continuous between all vbs; U-shaped

chlorenchyma groups occupy sides and bases of

furrows between consecutive l’vbs. No colourless

cells. Adaxial epidermis: bulliform cells poorly

developed; basal cells of furrows may be slightly

enlarged to form small, fan-shaped bulliform cell

groups. Epidermal cells inflated with the outer wall

slightly thickened. No macro-hairs, hooks or prickles

developed. Cells variously papillate; one papillus per

cell. Abaxial epidermis: no bulliform cells present.

Cuticle and epidermal cell thickening differs in the

three forms. Hooks, prickles and papillae absent;

macro-hair bases present in alpine bog form (Figs

7 & 8).

Abaxial epidermis

Intercostal zones: not differentiated. Stomata:

absent throughout abaxial epidermis. Papillae: absent

except in some specimens of the Drakensberg form

where long cells are inflated and tend towards obliqi e

papillae (Fig. 15). Prickles: absent. Hooks: none

observed. Micro-hairs: absent except in the alpine

bog form where bicellar hairs with short basal, and

elongated, tapering distal, cells occur (Fig. 16).

Macro-hairs: absent except on all specimens of alpine

bog form (Figs 16 & 17). Uni-cellular; 2-3 specialized

epidermal cells associated with hair base; base swollen

in relation to hair thickness; short >1,25 mm long.

Silica bodies: variable (Table 1); rounded in typical

M. disticha (Figs 10-12) to tall and narrow in the

Drakensberg form (Figs 13-15). Present throughout

abaxial epidermis; granules present; width approxi-

mately the same as that of adjacent costal long cells

especially in typical and Drakensberg forms. Costal

<cglls: silico-suberose couples alternate with single

costal long cell throughout abaxial epidermis; cork

cells crescentic and enfolding the silica body except for

tall and narrow cork cells associated with tall and

narrow silica bodies in the Drakensberg form; costal

long cells rectangular with straight to slightly inflated

walls; slightly undulated.

Specimens examined:

Typical M. disticha

O.F.S. — 2828 (Bethlehem): Golden Gate National Park,

Brand wag Peak (-DA), Ellis 2391.

Cape. — 3124 (Hanover): Lootsberg (-DC), Theron 486.

3126 (Queenstown): Jamestown (-BB), Ellis 2603. 3225

(Somerset East): Groot Riet Vlei (-AD), Acocks 11961.

3226 (Fort Beaufort): 40 km from Tarkastad on Adelaide road

(-AD), Ellis 2572. 3325 (Port Elizabeth): Addo Elephant

National Park (-BC), Liebenberg 6656, 7713; King Neptune

Beach (-DC), Ellis 2564. 3420 (Bredasdorp): between Swellen-

dam and Riviersonderend (-AA), Ellis 1263', Potberg (-AD),

Ellis 669 ; Cape Agulhas (-CC), Loxton 246.

R. P. ELLIS

187

■

Figs 10-18. — Abaxial epidermis of Merxmuellera disticha sens. lat. All x 400. 10-12, M. disticha form. (10 , Liebenberg

7713; 11, Ellis 1263; 12, Ellis 669.) 13-15, Drakensberg form. (13, Ellis 3157; 14, Ellis 3152; 15, Ellis 1404.) 16-18,

Alpine bog form. (16, Ellis 3183; 17, Ellis 3192; 18, Ellis 3313.)

TABLE 1 . — The anatomical differences between the different forms of Merxmuellera disticha as seen in transverse sections of the leaf

blade and on epidermal preparations in surface view

1 88 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). II. MER XM UELLERA DISTICHA

Drakensberg form

O.F.S. — 2828 (Bethlehem): Golden Gate National Park,

Brandwag Peak (-DA), Du Toit 675. Wodehouse Peak, Ellis

2382\ Witsieshoek — Mont-aux-Sources area (-DB), Ellis 3133,

3152, 3157.

Natal. — 2829 (Harrismith): Cathedral Peak, Organ Pipes

Pass (-CC), Ellis 1404, 3185, 3303, 3305. 2929 (Underberg):

Giants Castle Game Reserve (-AD), McAllister 112', Banner-

mans Pass, Ellis 3312.

Lesotho. — 2929 (Underberg): Mokhotlong (-AC), Coetzee

834.

Alpine bog form

Natal. — 2829 (Harrismith): Cathedral Peak, Organ Pipes

Pass summit (-CC), Ellis 3183, 3184 3192, 3306, 3309. 2929

(Underberg): Giants Castle, summit of Bannermans Pass

(-AD), Ellis 3313, 3315, 3316', top of Sani Pass (-CB),

Du Toit 699.

Lesotho. — 2929 (Underberg): above Sani Pass (-CA),

Du Toit 2207.

Fig. 19. — Distribution of Merxmuellera disticha in South Africa.

□ — typical M. disticha form; 0 — Drakensberg form;

O — Alpine bog form. Shaded symbols represent localities

of specimens studied anatomically. Compiled from speci-

mens at the National Herbarium, Pretoria (PRE).

DISCUSSION AND CONCLUSIONS

From the foregoing anatomical description of

M. disticha sens, lat., and from the accompanying

photomicrographs (Figs 1-18), it is evident that

numerous anatomical differences exist between the

three different forms of M. disticha recognized in this

study (Table 1). These differences are of considerable

magnitude and are much more obvious than are the

anatomical differences among some of the other

closely related Merxmuellera species e.g. M. macowanii

(Stapf) Conert, M. davyi (C. E. Hubb.) Conert and

M. aureocephala (J. G. Anders.) Conert or M. drakens-

bergensis (Schweick.) Conert and M. stereophylla

(J. G. Anders.) Conert. In addition, the anatomical

differences between M. drakensbergensis and M.

stereophylla, for example, are merely a matter of

degree and there is a tendency for the characters to

grade from one species into the other. The differences

among the three M. disticha forms, on the other hand,

are distinctly disjunct with the characters being

structurally different e.g. adaxial ribs, sclerenchyma

girders, silica bodies etc. Furthermore, a number of

correlated characters, from both the leaf blade in

section and the epidermis, characterize each of the

three forms. These diagnostic characters constantly

occur in combination and with the detection of any

single diagnostic character the remainder can safely

be inferred. This evidence, based solely on leaf

anatomical criteria suggests, therefore, that each of the

forms of M. disticha, recognized in this study, warrants

taxonomic status, possibly sub-specific rank.

In certain spikelet characters, differences are also

exhibited among these three forms of M. disticha.

From the sample examined, it appears that, in both the

Drakensberg and alpine bog forms, only 2-flowered

spikelets are found. In typical M. disticha, all speci-

mens had three or more florets. In typical M. disticha

and in the Drakensberg form the length of the upper

and lower glumes is greater than 12 mm whereas, in

the alpine bog form they are 1 1 mm or less in length.

In both typical M. disticha and the Drakensberg form

the lower glume was always distinctly 3-nerved,

whereas in the alpine bog sample only a single

prominent nerve was present with two poorly de-

veloped lateral nerves sometimes being evident.

From the small sample examined, it would appear,

that a detailed study of the spikelet morphology of

M. disticha sens. lat. should confirm the anatomical

groupings and assist in reaching a taxonomic decision

The three forms of M. disticha show distinct

vegetative differences as well, and can be readily

recognised in the field. The alpine bog form has very

narrow, short, setaceous leaves between 100-200 mm

long and forms fine, delicate, compact but low

tussocks. The younger green leaves are normally

more or less straight and erect and the older, dry

leaves curl, thus forming an irregularly matted

“cushion” out of which the needle-like green leaves

project. Typical M. disticha plants have a similar

structure, except on a much larger scale with the

leaves being from 300-500 mm long. The setaceous

leaves are thicker and much more rigid and fibrous,

with a very high tensile strength. The old, dry leaves

of the tussock form a dense curly mass. The Drakens-

berg form, on the other hand, has leaves which are

often not setaceous, but are open and up to 3,5 mm

wide. This is especially conspicuous under conditions

of reduced radiation such as on misty, cloudy days.

In this state, the Drakensberg form is unmistakeable

and it is regrettable that in the preparation of her-

barium vouchers the leaves become infolded and

setaceous and the herbarium specimens resemble

typical M. disticha. However, the green leaves are

softer, more flexible and more easily torn and the

old leaf blades are exceptionally curly.

Differences in habitat requirements between the

three forms became evident while collecting material

in the field for the anatomical investigation. Thus,

the alpine bog and Drakensberg forms occur in the

Drakensberg mountains but only in basaltic soils

above the cave sandstone layers. On the summit of

this escarpment they often occur in close proximity to

one another. The alpine bog form is restricted to

saturated, humic soils in shallow bogs or seepage

areas and is often found in water about 100 mm deep.

The Drakensberg form may be found on raised

mounds in the selfsame seepage areas but obviously

requires better drainage conditions. It is also more

widespread being found down to altitudes of about

2 000 m, whereas the bog form is restricted to the

summit at over 3 500 m.

Typical M. disticha has a wide distribution in the

Cape Province (Fig. 19) being found at low altitudes

along the southern coast and then throughout the

eastern and north-eastern Cape mountains. Inexplic-

ably it appears to be absent from the Transkei and

Natal, but is found in Lesotho and the sandstone

mountains of the eastern Orange Free State. Thus at

Golden Gate typical M. disticha occurs on the slopes

of Brandwag Peak (Ellis 2931), but higher up the

same mountain, on the basalt cap known as Wode-

R. P. ELLIS

189

house Peak, the Drakensberg form is found in black,

peaty soils (Ellis 2382; Du Toil 675).

The type of intraspecific variation described here,

has also been observed in M. stricta (Schrad.) Conert

in similar habitats in the same mountains (Ellis, in

prep.). Possibly the other closely related Merxmuellera

species of the summer rainfall areas also represent

the outcome of similar diversification and speciation

e.g. M. drakensbergensis and M . stereophylla or

M. macowanii and M. aureocephala. Within this

genus, therefore, there appears to have been consider-

able adaptive radiation associated with altering

environmental conditions. This applies particularly

to altitude effects along this mountain range. These

factors must be borne in mind when final taxonomic

decisions are taken.

ACKNOWLEDGEMENTS

The author is grateful to the following organizations

for permission to collect material on their property:

the Department of Forestry, the National Parks

Board and the Natal Parks, Game & Fish Preservation

Board. The capable technical assistance of Miss R.

Manders is gratefully acknowledged.

UITTREKSEL

Die anatomiese struktuur van die blaar in dwarsnee

en die abaksiale epidermis van Merxmuellera disticha

word beskryf en geillustreer. Drie afsonderlike ana-

tomiese “ vorme ” word erken: tipiese M. disticha, die

Drakensberg vorm en die alpienevlei vorm. Dit blyk ook

dat hierdie drie anatomiese groepe ook verskillende

omgewingsbenodigdhede het en waarskynlik takso-

nomiese status verdien.

REFERENCES

Acocks, J. P. H., 1971. The distribution of certain ecologically

important grasses in South Africa. Mitt. bot. StSamml.,

Munch. 10: 149-160.

Acocks, J. P. H., 1975. Veld types of South Africa, 2nd ed.

Mem. bot. Surv. S. Afr. 40: 1—27.

Chippindall, L. K. A., 1955. In D. A. Meredith, The grasses

and pastures of South Africa. Johannesburg: CNA.

Conert, H. J., 1970. Merxmuellera, eine neue Gattung der

Gramineen. Senckenberg Biol. 51 : 129-133.

Ellis, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

Bothalia 13,1 & 2: 191-198 (1980)

Leaf anatomy of the South African Danthonieae (Poaceae).

III. Merxmuellera stricta

R. P. ELLIS*

ABSTRACT

The anatomical structure, of the leaf blade as seen in transverse section, and of the abaxial epidermis, of

Merxmuellera stricta (Schrad.) Conert is described and illustrated. In this variable species four distinct ana-

tomical “forms” are recognized viz. the typical M. stricta form, the Cathedral Peak form, the Drakensberg

form and the alpine form. The alpine and Cathedral Peak forms have recently been described as M. guitlar-

modiae Conert (1975). The degree of anatomical differentiation of these “forms” resembles the situation

described in M. disticha (Nees) Conert (Ellis, 1980). Populations of both M. stricta and M. disticha from the

Drakensberg mountains display extensive anatomical diversification which appears to be correlated with

environmental factors. In addition, morphological differences are exhibited as well and the anatomical “forms”

of M. stricta probably warrant taxonomic recognition.

RLSUML

ANATOMIE FO LI AIRE DES DANTHONIEAE ( POACEAE ) D'AFRIQUE DU SUD. III.

MERXMUELLERA STRICTA

La structure anatomique du limbe foliaire en coupe transversale et celle de I’epiderme abaxial de Merxmuellera

stricta (Schrad.) Conert sont decrites et illustrees. Dans cette espece variable on reconnait quatre "formes”

anatomiques distinctes, soit la forme M. stricta typique, la forme Cathedral Peak, la forme du Drakensberg et la

forme alpine. Les formes alpine et Cathedral Peak ont recemment ete decrites sous le nom de M. guillarmodiae

Conert (1975). Le degre de differenciation anatomique de ces " formes ” ressemble a la situation decrite chez

M. disticha (Nees) Conert (Ellis, 1980). Dans les montagnes du Drakensberg des populations tant de M. stricta

que de M. disticha montrent une large diversification anatomique qui est apparemment en correlation avec des

facteurs de milieu. En outre, des differences morphologiques sont egalement manifestes et il est probable que les

“formes” anatomiques de M. stricta meritent d’etre reconnues taxonomiquement.

INTRODUCTION

Merxmuellera stricta (Schrad.) Conert (1970)

[=Danthonia stricta (Nees) Schrad.] is widely distri-

buted in the southern mountainous areas of South

Africa. It occurs from Namaqualand in the north-

west southwards to the south-western Cape, then

eastwards to the north-eastern Cape mountains from

where the distribution continues in a northerly direc-

tion along the Drakensberg mountains to Lesotho

and the eastern Orange Free State. Together with

M. disticha (Nees) Conert, M. stricta is an important

constituent of the Karroid Merxmuellera Mountain

Veld along all the higher mountains of the False

Karoo and the Central Upper Karoo (Acocks, 1975).

To the west of Beaufort West, M. stricta replaces M.

disticha as the dominant grass of the Mountain

Renosterbosveld (Acocks, 1975). M. stricta is also

common in the fynbos communities of the south-

western Cape.

M. stricta is a variable perennial, forming coarse,

wiry tufts. Chippindall (1955) states that “There is

considerable variation in the plants referred to D.

stricta, and it is possible that they comprise more

than one variety”. In the north-west M. stricta may be

confused with M. dura (Stapf) Conert, but M. stricta

can be recognized by the glabrous condition of the

lemma at the point of insertion of the central awn

and is distinct anatomically (Ellis, in prep.). In the

north-east, in the Drakensberg mountains, a situation

exists, similar to that observed in M. disticha (Ellis,

1980), with three additional anatomical “forms” being

present.

These anatomical “forms” appear to be correlated

with morphological characters and habitat differences.

Certain of the spikelet differences and other morpho-

logical characters must be of considerable magnitude

* Botanical Research Institute, Department of Agricultural

Technical Services, Private Bag X101, Pretoria, 0001.

as Conert (1975) has independently described a new

species, M. guillarmodiae Conert, from M. stricta

collections from the alpine region of the Drakensberg.

Unfortunately, the specimens cited by Conert (1975)

as belonging to M. guillarmodiae fall into two of the

anatomical categories recognized in the present study,

while the remaining two “forms” fall in M. stricta as

currently constituted. If the precedent created by the

description of M. guillarmodiae is to be followed, it

implies that a further two species require description.

Similarly, by the same token, two new species, present-

ly referred to M. disticha, also warrant description

(Ellis, 1980).

In the present context M. stricta is viewed in its

widest sense, and for convenience M. stricta sens. lat.

has merely been sub-divided into four “forms” for des-

criptive purposes: the typical form (M. stricta), the

Cathedral Peak form ( M . guillarmodiae), the Drakens-

berg form (M. stricta) and the alpine form (M.

guillarmodiae). Each of these “forms” exhibits charac-

teristic leaf anatomy and epidermal structure which

will be described and discussed according to the

terminology of Ellis (1976; 1979).

In the anatomical descriptions which follow, the

following abbreviations will be used:

vb/s — vascular bundle/s

Fvb/s — first order vascular bundle/s

2’vb/s — second order vascular bundle/s

3’vb/s — third order vascular bundle/s

ibs — inner bundle sheath; mestome sheath

obs — other bundle sheath; parenchyma sheath.

ANATOMICAL DESCRIPTION OF MERXMUELLERA

STRICTA SENS. LAT.

Leaf in transverse section

Leaf outline: infolded with an elliptical or U-shaped

outline. Permanently infolded in typical and Cathedral

Peak forms (Figs 1-6) but regular opening to 180

192 L EAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). III. MERXMUELLERA STRICT A

occurs in Drakensberg form (Fig. 9) and to 45° in the

alpine form (Fig. 12). Lamina symmetrical about the

median vb. 9-13 vbs present in leaf section (Table 1)

with the alpine form always with 9 vbs (Figs 10-12)

and typical M. stricta with 11 or 13 (Figs 1-3).

Adaxial channel an extremely narrow and deep cleft

in the typical and Cathedral Peak forms and is

narrower than the lamina thickness in the Drakens-

berg and alpine forms when infolded. Leaves narrow

(<1,1 mm wide) when folded. Ribs and furrows:

adaxial furrows of variable depth but constant for

each of the “forms” (Table 1); narrow, cleft-like.

Similar ribs over all vbs when present; rounded or

flat-topped with one vb per rib. Abaxial surface

smooth except in typical M. stricta where grooves are

present on either side of the median vb (Figs 1-3).

Median vascular bundle: present; indistinguishable

structurally from lateral l’vbs. Vascular bundle arrange-

ment: no 2’vbs; 3’vbs absent between lateral 1’vbs.

3 or 4 l’vbs in each half of the lamina in typical form

but decreasing to two in alpine form. All bundles

located in centre of blade. Vascular bundle structure:

vbs circular or elliptical in shape; xylem and phloem

distinguishable in all vbs; phloem adjoins ms; often

divided by intrusion of fibres in typical M. stricta

(Figs 2 & 3) and the Cathedral Peak form (Fig. 5).

Lysigenous cavities present. Metaxylem vessels

thickened, circular and very narrow. Vascular bundle

sheaths: obs circular or elliptical but normally horse-

shoe shaped due to wide abaxial interruptions. Adaxial

extensions present but cell composition differs in

each of the “forms”. Obs cell shape variable from

round to elliptical but all cells in a given leaf similar in

shape. Obs cells conspicuous but not larger than the

Figs 1-12. — Leaf blade outline of Merxmuellera stricta sens. lat. as viewed in transverse section. 1-3, typical M. stricta

form, all X 160. Note abaxial grooves on either side of median bundle only. (1, Ellis 2476; 2, Ellis 2445; 3, Ellis 2441.)

4-6, Cathedral Peak form, all X 160. Adaxial groove, but no furrows present. (4, Ellis 3295; 5, Ellis 2372; 6, Trauseld

833.) 7-9, Drakensberg form. Very deep adaxial furrow on either side of median vascular bundle. Leaf not permanently

infolded. (7, Ellis 3318, x250; 8, Ellis 3322, Xl60; 9, Ellis 3321, x250.) 10-12, alpine form, all X250. Furrows

developed between all vascular bundles. (10, £//»- 3181 ; 1 1, Ellis 3317; 12, Ellis 3308.)

R. P. ELLIS

193

;

_r~Q4 ~

' % ji f #

Figs 13-24. — Abaxial epidermis of Merxmuellera stricta sens. lat. as seen in surface view. 13-15, typical M. stricta form.

All with round silica bodies. (13, Roberts 2034, x250, note absence of intercostal zones; 14, Ellis 1156, X250, note

stomatal files and intercostal zones; 15, Ellis 2317, x640.) 16-18, Cathedral Peak form. Silica bodies dumb-bell

shaped. (16, Ellis 2372, x250; 17, Galpin 10357, x250; 18, Ellis 3289, <400.) 19-21, Drakensberg form. Dumb-bell

shaped silica bodies. (19, Acocks 22069, x 160; 20, Ellis 1428, x400; 21, Ellis 3290, x400.) 22-24, Alpine bog form.

Silica bodies crescent-shaped, round or irregularly dumb-bell shaped. Macro-hairs, micro-hairs and prickles common.

All X 160. (22, Du Toil 2206; 23, Ellis 3181 ; 24, Ellis 1393.)

mesophyll cells; without chloroplasts (Fig. 26). Ibs

complete; small cells with U-shaped thickenings;

adaxial cells larger than lateral cells (Fig. 26). Scleren-

chyma: Adaxial sclerenchyma variable from minute

strands in the Cathedral Peak form (Figs 4-6) to well-

developed, inversely anchor-shaped girders in typical

M. stricta (Figs 1-3) or the alpine form (Figs 10-12)

(Table 1). Abaxial girders well developed; either

trapezoidal (Fig. 1 & 7) or narrower than the vb

(Figs 2 & 5). In certain specimens of typical M.

stricta individual girders fuse forming a continuous,

abaxial, hypodermal band (Fig. 1). Margin: small,

triangular, sclerenchyma caps developed. Mesophyll:

not radiate; composed of regular, small, isodiametric,

tightly packed cells (Fig. 26). Tall, narrow groups of

chlorenchyma tissue between consecutive vbs sepa-

rated by sclerenchyma girders. No colourless cells.

Adaxial epidermis: bulliform cells poorly developed;

basal cells of furrows may be slightly enlarged to

form small, fan-shaped groups (Table 1). Epidermal

cells inflated with outer wall slightly thickened. Macro-

hairs and prickles absent; hooks occur irregularly.

Cells variously papillate; one papillus per cell. Adaxial

papillae best developed in typical M. stricta (Fig. 1).

Abaxial epidermis: no bulliform cells developed.

Hooks, prickles, macro-hairs and papillae lacking.

Outer tangential wall flattened with smooth, con-

tinuous cuticle.

Abaxial epidermis in surface view

Intercostal zones: differentiated except in typical M.

stricta specimens with continuous hypodermal scle-

renchyma layer (Figs 1 & 13). Long cells medium to

elongated with parallel, slight to moderately undu-

lating, side walls. Silico-suberose couples between

successive long cells. No bulliform cells. Stomata:

absent in alpine form (Figs 22-24) and in those typical

M. stricta specimens with hypodermal sclerenchyma

layer (Fig. 13). Low dome-shaped, 1-2 files of stomata

in centre of each intercostal zone; files adjacent to

one another. One interstomatal cell between successive

stomata. Intercostal short cells: silico-suberose

couples; cork cell tall and narrow to crescentic;

occur between most intercostal long cells. Papillae:

194 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). III. MERXMUELLERA STRICTA

TABLE 1. — The differences in leaf anatomy and spikelet morphology between the different forms of Merxmuellera stricta

absent. Prickle-hairs : only present on alpine form

(Fig. 22) (Table 1). Barbs short. Micro-hairs : present

but very rare in typical M. stricta and then only in

specimens with intercostal zones; common in other

three “forms”. Bicellular, elongated (especially in

alpine form); basal cell slighly longer than distal cell;

tapering distal cell thin-walled (Fig. 18). Macro-hairs :

absent except on all specimens of alpine form; occur

in groove on either side of median vb in one specimen

of Drakensberg form (Fig. 19). Unicellular, inflexible;

2-3 specialized epidermal cells associated with base of

hair; base somewhat swollen in relation to hair

thickness. Silica bodies: differ in different “forms”

(Table 1). Costal bodies rounded or elliptical in

typical M. stricta (Figs 14 & 15) and in certain speci-

mens of Drakensberg form; dumb-bell shaped in

Drakensberg and Cathedral Peak form (Figs 18, 20,

21); small rounded to elliptical and associated with

crescentic cork cell in alpine form (Figs 22 & 24).

Granules present. Width same as adjacent costal long

cells except in alpine form. Costal cells: silica cells

alternate with costal short cells except in alpine form

where silico-suberose couples alternate with costal

short cells; files with silica cells alternate with files of

costal long cells in all forms except alpine form.

Specimens examined.

Typical M. stricta form

Cape. — 3119 (Calvinia): Van Rhyns Pass (-AC), Ellis 1139,

1/40', Kobe Mts (-CA), Ellis 2445, 2447*. 3126 (Queenstown):

Bushmanhoek Pass (-AD), Ellis 2577* ; Hangklip Mt (-DD),

Roberts 2034*. 3218 (Clanwilliam): Pakhuis Pass (-BB),

Ellis 1705*. 3219 (Wuppertal): Pakhuis Pass (-AA), Ellis 1151,

1156', Buffelberg Pass (-CA), Ellis 1189. 3225 (Somerset East):

Grootfontein valley (-CB), Van der Walt 184*. 3318 (Cape

Town): Jonkershoek, Stellenbosch (-DD), Ellis 2241, 2242,

2258. 3319 (Worcester): Gydoberg (-AD), Ellis 2476*; Karoo-

poort, Ceres (-BC), Hafstrom & Acocks 100*. Franschoek Pass

(-CD), Ellis 687, 688. 3320 (Montagu): 23 km from Barrydale

on road to Montagu (-DC), Ellis 643. 3322 (Oudtshoorn):

Swartberg (-AC), Ellis 2579*; Robinsons Pass (-CC), Ellis

2583*. 3323 (Willowmore): Uniondale (-CA), Ellis 1641*.

3325 (Port Elizabeth): Gamtoos Valley (-CC), Acocks 16112.

3326 (Grahamstown): Peddie (-BC), Fairall 240. 3418 (Simons-

town): Red Hill (-AB), Ellis 2317, 2318. 3419 (Caledon):

Kleinmond (-AC), Ellis 2514, Van Heerden 75. 3420 (Bredas-

dorp): Bredasdorp (-AA), Ellis 1265*.

Cathedral Peak form (M. guillarmodiae)

O.F.S. — 2828 (Bethlehem): Golden Gate National Park,

Brandwag Peak (-DA), Ellis 2371, 2372.

Natal. — 2829 (Harrismith): Royal Natal National Park

(-CB), Galpin 10351; Cathedral Peak Forest Reserve (-CC),

Killick lllOt 1576t Ellis 3289, 3295, 3297. 2929 (Underberg):

Giants Castle Game Reserve (-AB), Trauseld 833; Nyiginye

(-BA), Du Toit 2516; Highmoor Forest Reserve (-BC), Du

Toil 2500.

Drakensberg form ( M . stricta)

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(-CC), Ellis 1428, 3290, 3299. 2929 (Underberg): Giants

Castle Game Reserve, Bannermans Pass (-AD), Ellis 3318,

3321 ; Loteni Game Reserve (-DA), Ellis 3322.

•Specimens with continuous abaxial hypodermal scleren-

chyma layer and without stomata.

t Specimens cited by Conert (1975) as being M. guillarmodiae.

R. P. ELLIS

195

Lesotho. — 2929 (Underberg): Sehlabathebe National Park

(-CC), Du Toil 2631, 2642.

Cape. — 3028 (Matatiele): Quachas Nek (-BA), Ellis 243*;

Ramatselisonek (-BB), Acocks 22069. 3226 (Fort Beaufort):

Great Winterberg (-AD), Story 4522*

Alpine form (M. guillarmodiae)

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve,

summit above Windy Gap (-CC), Ellis 1393, 3181, 3182 , 3308.

2929 (Underberg): Giants Castle Game Reserve, summit of

Bannermans Pass (-AD), Ellis 3317; top of Sani Pass (-CA),

Du Toil 2206, 2242, 2286.

Lesotho. — 2828 (Bethlehem): Butha Buthe (-CC), Roberts

5844, Lubke 305; Tsehlanyane valley (-CD), Jacot-Guillarmod

3127\, 3734f.

Cape. — 3027 (Lady Grey): Doodmans Kranz, Barkly East

(-DC), Galpin 6906 f. 3028 (Matatiele): Quachas Nek, Temrock

Peak (-BA), Liebenberg 5729.

Fig. 33. — Distribution of Merxmuellera stricta in South Africa.

□ — typical M. stricta; A — Cathedral Peak form;

V — Drakensberg form; O — Alpine form. Shaded symbols

represent localities of specimens studied anatomically.

Compiled from specimens at the National Herbarium,

Pretoria (PRE).

DISCUSSION AND CONCLUSIONS

All the specimens examined in this study have the

lateral first order vascular bundles located adjacent to

one another and lack the single third order bundle

interspaced between them which is characteristic of

M. disticha (Ellis, 1980). M. dura, considered to be

closely allied to M. stricta (Chippindall, 1955; De

Wet, 1960), does not share this characteristic and has,

therefore, been excluded from the present considera-

tions.

Anatomically the leaf structure of the typical M.

stricta form is remarkably constant throughout its

wide distributional area (Fig. 33). The only anatomi-

cal difference worthy of mention is the development of

a continuous abaxial hypodermal sclerenchyma layer

(Fig. 1) with the resultant exclusion of the intercostal

zones and stomata (Fig. 13). This structure was

present in 40% of the specimens examined in this

study, and occurs in plants from widely scattered

localities, throughout the range of this form, and does

not appear to be correlated with any obvious environ-

mental factor. Morphologically there is a definite

tendency for these specimens to have shorter glumes

[(11-) 12,1 mm (-15)] than the remainder of the

typical M. stricta specimens [(15-) 17,5 mm (-22)].

These two characters, therefore, appear to be con-

* Specimens with round or elliptical silica bodies,

t Specimens cited by Conert (1975) as being M. guillarmodiae.

stantly linked and, on the basis of glume length, these

shorter spikelets tend to resemble the Cathedral

Peak form of M. guillarmodiae (Table 1). However,

the M. stricta form has distinctly longer awns —

averaging 11,1 mm as against 9,8 mm for the Cathedral

Peak form. This form, however, does not have a

continuous hypodermal sclerenchyma layer (Fig. 4)

and has numerous stomatal files (Figs 16-18).

The anatomy of the Cathedral Peak form is un-

mistakeable, due to the shape of the adaxial groove

which lacks ribs and furrows, (Figs 4-6) and due to

the occurrence of dumb-bell shaped silica bodies

(Figs 16-18). This type of anatomical structure

appears to be strongly correlated with M. guillar-

modiae spikelet characters, such as the presence of

hairs on the back of the lemma up to the point of

insertion of the central awn (Conert, 1975). However,

as Conert (1975) points out the hair arrangement of

the cited Cathedral Peak form specimens ( Killick

1100, 1576) resembles that of M. guillarmodiae, but

the hairs themselves, are longer. The typical M.

guillarmodiae (alpine form) specimens examined have

hairs up to 2,0 mm in length with the lemmas being

sparsely hairy but the Cathedral Peak form specimens

are densely hairy on the back of the lemmas with the

hairs being from 2, 0-3, 2 mm long.

All morphological indications are, therefore, that

this Cathedral Peak form is closely linked to the

typical M. guillarmodiae or alpine form. However,

one specimen, Trauseld 833, does not exhibit any of

these M. guillarmodiae spikelet characteristics.

Although the adaxial groove and silica bodies con-

form, it differs anatomically from the remainder of

the Cathedral Peak form sample in that it lacks

adaxial bundle sheath extensions and thus the meso-

phyll distribution is continuous adaxially and not

restricted to separate tall and narrow groups (Fig. 6).

This specimen was collected at Giants Castle and,

although this area was revisited, no plants with

Cathedral Peak form anatomy or morphology were

located. The Drakensberg form, which Trauseld 833

resembles morphologically, is common in this area,

but all specimens collected had the distinctive Drakens-

berg form anatomy. Specimens, conforming anatomi-

cally and morphologically with the Cathedral Peak

form, do occur at Highmoor to the south of Giants

Castle, e.g. Du Toil 2500.

Further collections from the southern Drakensberg

are required to help clarify the relationships of this

Cathedral Peak form. It has been grouped under M.

guillarmodiae (Conert, 1975) together with the alpine

form, which it closely resembles morphologically.

Anatomically, however, it has most in common with

the Drakensberg form. Both these latter “forms”

have similar sclerenchyma girders and bundle sheath

extensions but, most significantly, both have dumb-

bell shaped silica bodies.

The Drakensberg form, on the other hand, although

resembling the Cathedral Peak form anatomically,

differs significantly morphologically. The back of the

lemmas are glabrous along the centre, as in typical

M. stricta, and the lower glumes are much longer

(Table 1). Anatomically the most important difference

is the two deep adaxial furrows on either side of the

median vascular bundle (Fig. 7) and the laminae are

capable of opening to 180° (Fig. 9) especially during

conditions of low irradiance. This ability is undoubted-

ly functionally significant and was possibly partly

responsible for the relative success of this Drakens-

berg form in the north-eastern mountains.

196 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). III. MERXMUELLERA STRICT A

Figs 25-32. — A comparison between the leaf anatomy of the setaceous-leaved, summer rainfall Pentaschistis species

with the anatomy of the leaves of the Merxmuellera stricta group. 25, Pentaschistis basutorum, Ellis 2368, x 160.

Note arrangement of large and small vascular bundles. 26, Merxmuellera guillarmodiae, Ellis 3289, X400. Detail of

mesophyll and bundle sheath structure typical of the M. stricta group. 27-28, Pentaschistis fibrosa. (21 , Du Toil 646,

x 160, note vascular bundle arrangement; 28, Du Toit 2301, X400, outer bundle sheath cells thin-walled and in-

conspicuous.) 29-30, Pentaschistis tysonii. (29, Ellis 3302, x250, third order bundles absent; 30, Ellis 3319, X400,

bundle sheath cells distinct.) 31-32, Pentaschistis sp. (31, Ellis 1409, X100, no third order bundles between lateral

first order bundles; 32, Ellis 3292, x400.)

R. P. ELLIS

197

This anatomical difference, although distinct and

consistent, may, however, not be as significant as it

initially appears. It can be simply explained by

opposite developmental trends in respect of only a

single character— a single adaxial furrow on either

side of the median vascular bundle. The epidermis

also closely resembles that of the Cathedral Peak

form as both have dumb-bell shaped silica bodies

(Figs 16-21).

The anatomical indications are, therefore, that the

Drakensberg form resembles the Cathedral Peak

form more closely than either of the other two forms.

These similarities are not corroborated by spikelet

structure, however, and in this respect the Drakens-

berg form displays typical M. stricta structure.

Some Drakensberg form specimens from the eastern

Cape ( Ellis 243; Story 4522), however, anatomically

demonstrate this relationship with typical M. stricta.

These specimens have round silica bodies, with

characteristic M. stricta type epidermal structure but

in section the anatomy is typical of the Drakensberg

form. This observation, therefore, provides an import-

ant clue as to the affinities of this form and provides

an anatomical link to support the morphological

evidence. A degree of morphological and anatomical

gradation appears to occur between these two forms

in the areas of sympatry in the north-eastern Cape

and once again further collecting is required to help

confirm affinities between the Drakensberg, Cathedral

Peak and typical M. stricta forms.

The anatomy of the alpine form specimens compares

very favourably with that of the type specimen of

M. guillarmodiae ( Jacot-Guillarmod 3734). This form

appears to be distinct, both anatomically and morpho-

logically, from M. stricta (Conert, 1975) although the

Cathedral Peak form is intermediate in most spikelet

characters (Table 1). In addition it has specialized

habitat requirements, being restricted to the higher

alpine zone of the Drakensberg, often associated with

boggy conditions. It is, therefore, not spatially

associated with any of the other forms.

This alpine form of M. guillarmodiae exhibits

striking resemblances, in habitat preferences, growth

form and anatomy, with the alpine bog form described

in M. disticha (Ellis, 1980). Anatomical similarities

are rib and furrow distribution and form, mesophyll

configuration and epidermal structure. The only

difference is in the pattern of arrangement of the

various orders of vascular bundle along the width of

the lamina. This arrangement differs in the two

“forms” but corresponds with the patterns found

in either typical M. stricta or typical M. disticha. This

difference is correlated with differences in inflore-

scence characters — a contracted panicle in M. stricta

and a distichous spike in M. disticha — and appears to

indicate the relationships of the alpine “forms”. This

seems to be an excellent example of convergent

evolution in response to similar environmental

conditions.

The most obvious morphological differences

between the specimens of the four “forms” of M.

stricta, recognized and examined in this study, are

briefly summarized in Table 1. Typical M. stricta

and the Drakensberg form have significantly longer

lower glumes than do the alpine and Cathedral Peak

forms. The longer glumes are associated with longer

awns on the lower lemma except in the Cathedral

Peak form which is intermediate between the alpine

form and the other two. The Cathedral Peak and

typical M. stricta forms are glabrous on the back of

the lower lemma along the central vein up to the

base of the central awn, although, occasionally, a

few scattered hairs may be present in M. stricta type

specimens. In both the alpine and Cathedral Peak

forms the back of the lemma is hairy-sparsely hairy

with hairs up to 2 mm long in typical M. guillar-

modiae (the alpine form) but densely hairy with longer

hairs (2, 0-3, 2 mm long) in the Cathedral Peak form.

Morphological characters, therefore, indicate two

groups in M. stricta sens. lat. — the typical M. stricta

and Drakensberg forms sharing certain characters

whereas, the same characters differ considerably in

the alpine and Cathedral Peak forms. This grouping is

not confirmed by anatomical evidence which indicates

close relationships between the Cathedral Peak and

Drakensberg forms with typical M. stricta and the

alpine forms being distinct. However, a few specimens

with anomalous anatomy tend to break down the

rigid anatomical divisions and, thereby, add substance

to the morphological grouping.

The acceptance of the morphological groupings as

reflecting affinities, implies that silica body structure

is of no significance in this instance. This would

normally be considered unlikely as silica bodies are

usually of considerable value taxonomically (Metcalfe,

1960). Typical M. stricta specimens have classic

rounded or elliptical festucoid-type silica bodies

whereas, both the Drakensberg and Cathedral Peak

forms have panicoid-type, dumb-bell shaped bodies

(Clifford & Watson, 1977). All other indications are

that typical M. stricta and the Drakensberg form are

closely related except for silica body shape. In fact,

these two “forms” of the same species possess silica

bodies supposedly characteristic of different tribal

groupings (Prat, 1932; 1936). De Wet (1954, 1956,

I960) notes the mixed character of the epidermis in

M. stricta but actually refers to the association of

micro-hairs (a panicoid character) with the festucoid

elliptical silica bodies. Two panicoid epidermal

characters may, therefore, occur in M. stricta sens,

lat. but this is not supported by other anatomical

evidence.

The recent description of M. guillarmodiae (Conert,

1975) as a separate species initially appears justified

on the anatomical evidence presented in this paper.

However, certain M. guillarmodiae specimens,

included here in the Cathedral Peak form, differ

dramatically from the type of M. guillarmodiae (the

alpine form) and, applying the same criteria appear to

merit specific status in their own right. The Drakens-

berg form would then also warrant specific status.

This implies that a further two species require descrip-

tion, as is the case in M. disticha where an almost

identical anatomical situation exists (Ellis, 1980).

However, especially in M. stricta, these anatomical

differences do not correlate very well with morpho-

logical differences.

This anatomical diversification within Merxmuel-

lera populations in this restricted area of high alti-

tudes and of climatic extremes also appears to be

reflected by the description of M. stereophylla.

(J. G. Anders.) Conert (Anderson, 1962) from M.

drakensbergensis (Schweick.) Conert. The recognition

of M. stereophylla was based on morphological and

ecological grounds but, in this case, the anatomical

differences are minimal in comparison with the

differences between the anatomical “forms” of M.

stricta and M. disticha.

In addition, two of the four setaceous-leaved

Pentaschistis species from this same area, display

remarkable anatomical similarities with M. stricta

198 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). III. MERXMUELLERA STRICT A

sens. lat. (Figs 25-32). P. basutorum Stapf (Fig. 25)

and P. fibrosa Stapf (Figs 27 & 28) have typical

Pentaschistis anatomy with thin-walled outer bundle

sheath cells and somewhat diffuse mesophyll. The

bundle sheath extensions and sclerenchyma associated

with the third order bundles differ considerably from

M. stricta sens. lat. (Fig. 26). P. tysonii Stapf (Figs

29 & 30) and Pentaschistis sp (Figs 31 & 32), on the

other hand, display typical Merxmuellera type

anatomy, and, on anatomical grounds only, appear to

have more in common with Merxmuellera and espe-

cially the M. stricta group than they do with Pentas-

chistis.

The following Pentaschistis specimens were

examined:

P. basutorum

O.F.S. — 2828 (Bethlehem): Golden Gate National Park

(-DA), Ellis 2367, 2368, 2369, 2370, 2373 ; Liebenberg 7454.

P. fibrosa

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(-CC), Ellis 3300', Sentinel (-DB), Du Toit 646. 2929 (Under-

berg): Highmoor Forest Reserve (-BC), Ellis 3169\ Sani Pass

(-CB), Du Toit 2301.

P. tysonii

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(-CC), Ellis 3296, 3302 ; Killick 2280. 2929 (Underberg):

Giants Castle Game Reserve, Bannermans Pass (-AD), Ellis

3314, 3319.

Pentaschistis sp.

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(-CC), Ellis 1409, 3291, 3292. 2929 (Underberg): Giants Castle

Game Reserve, Bannermans Pass (-AD), Ellis 3320.

For the above reasons the description of numerous

new Merxmuellera species from this summer rainfall

area is to be cautioned against for the time being.

The indications are that this temperate region of high

altitudes has only relatively recently been colonized

by these typically winter rainfall grasses from the

south. Adaptive radiation appears to be actively

continuing and the taxonomic picture is not at all

clear. Further studies, especially those of a bio-

systematic and autecological nature, are needed,

within Merxmuellera and closely related genera in the

Danthonieae, before reliable taxonomic conclusions

can be reached. At present, the assigning of specific

rank to any of these anatomical “forms” cannot be

fully justified. However, for practical purposes, each

of these anatomical “forms”, described in both M.

stricta and M. disticha (Ellis, 1980), deserve taxonomic

recognition but infraspecific groupings are recom-

mended until the status of this genus in this area is

better understood.

In arriving at a final conclusion, it must be remem-

bered, that, in both M. stricta and M. disticha, the

anatomical differences between the “forms” are of

considerable magnitude, disjunct and are correlated

with other anatomical, morphological and ecological

characteristics. In many instances these differences

are, therefore, greater than are those between other

Merxmuellera species and even between some of the

genera of the Danthonieae.

ACKNOWLEDGEMENTS

The author is grateful to the following organizations

for permission to collect material on their property:

the Department of Forestry, the National Parks

Board and the Natal Parks, Game and Fish Preser-

vation Board. The capable technical assistance of

Miss R. Manders and Miss L. Breytenbach is grate-

fully acknowledged as is the advice and assistance

given by Miss L. Smook on the identification and

morphology of the specimens studied.

UITTREKSEL

Die anatomiese struktuur, van die blaar in dwarsnee

en die abaksiale epidermis, van Merxmuellera stricta

(Sc hr ad.) Conert word beskryf en geillustreer. Vier

afsonderlike anatomiese “vorme” word erken: die

tipiese M. stricta vorm, die Cathedral Peak vorm, die

Drakensberg vorm en die alpiene vorm. Die alpiene en

Cathedral Peak vorme is onlangs beskryf as M. guillar-

modiae Conert (1975). Die mate van anatomiese

dijferensiasie van al hierdie “vorme” toon ’n ooreen-

koms met die toestand wat in M. disticha beskryf is

(Ellis, 1980). Populasies van, beide M. stricta en M.

disticha, van die Drakensberge, vertoon anatomiese

diver sifikasie op groot skaal wat ook gekorreleer is

met omgewingsfaktore. Boonop is morfologiese verskille

ook sigbaar en hierdie anatomiese “vorme” van

M. stricta verdien vermoedelik om taksonomiese

erkenning te verkry.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa, 2nd ed.

Mem. bot. Surv. S. Afr. 40: 1-127.

Anderson, J. G., 1962. Notes on African plants: Gramineae.

Bothalia 7 : 419^420.

Chippindall, L. K. A., 1955. In D. Meredith, The grasses and

pastures of South Africa. Johannesburg: CNA.

Clifford, H. T. & Watson, L., 1977. Identifying grasses.

St. Lucia: Univ. of Queensland Press.

Conert, H. J., 1970. Merxmuellera, eine neue Gattung der

Gramineen. Senckenberg. Biol. 51 : 129-133.

Conert, H. J., 1975. Merxmuellera guillarmodiae Conert n.sp.

Senckenberg Biol. 56: 145-152.

De Wet. J. M. J., 1954. The genus Danthonia in grass phylo-

geny. Am. J. Bot. 41 : 204-21 1 .

De Wet, J. M. J., 1956. Leaf anatomy and phylogeny in the

tribe Danthonieae. Am.J. Bot. 43: 175-182.

De Wet, J. M. J., 1960. Leaf anatomy and morphology in the

South African species of Danthonia. Bothalia 7 : 303-310.

Ellis, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-672.

Ellis, R. P., 1980. Leaf anatomy of the South African Dantho-

nieae (Poaceae). II. Merxmuellera disticha. Bothalia 13:

185-189.

Metcalfe, C. R., 1960. Anatomy of the monocotyledons. I.

Gramineae. Oxford: Clarendon Press.

Prat, H., 1932. L’fipiderme des Graminees. fitude anatomique

et systematique: Annls Sci. nat. (Bot.) ser. 10 14: 117-324.

Prat, H., 1936. La systematique des Graminees. Annls Sci. nat.

(Bot.) ser. 10 18: 165-258.

Bothalia 13,1 & 2: 199-216 (1980)

A synecological account of the Suikerbosrand Nature Reserve.

II. The phytosociology of the Yentersdorp Geological System

G. J. BREDENKAMPf and G. K. THERONJ

ABSTRACT

The vegetation of the Ventersdorp Geological System of the Suikerbosrand Nature Reserve is analysed and

classified according to the Braun-Blanquet method. Descriptions of the plant communities include description

of habitat features, the identification of differentiating species groups as well as the listing of prominent and less

conspicuous species for the tree, shrub and herbaceous layers. The habitat features that are associated with

differences in vegetation include altitude, aspect, slope, rockiness of soil surface, soil depth and soil texture.

rEsumE

COMPTE-RENDU SYNECOLOGIQUE DE LA RESER VE NA TURELLE DU SUIKERBOSRAND. II.

PHYTOSOCIOLOGIE DU SYSTEME GEOLOGIQUE DE VENTERSDORP

La vegetation du systeme geologique de Ventersdorp dans la Reserve Naturelle du Suikerbosrand est analysee

et classee selon la methode de Braun-Blanquet. Les descriptions des communautes vegetales incluent la description

des caracteristiques d’habitat, l' identification de groupes d'especes qui se differencient ainsi que le catalogue des

especes en vue et des especes moins apparentes dans les strates arborees, arbustives et herbacees. Les caracte-

ristiques d’habitat associees a des differences dans la vegetation comprennent l’ altitude, T aspect, la pente, la

nature rocheuse de la surface du sol, la profondeur et la texture du sol.

INTRODUCTION

The first part of the account of the synecology of the

Suikerbosrand Nature Reserve dealt with the phyto-

sociology of the Witwatersrand Geological System

(Bredenkamp & Theron, 1978), while this, the second

and final part, deals with the phytosociology of the

Ventersdorp Geological System.

THE STUDY AREA

The Ventersdorp System occupies the western part

of the Reserve and covers approximately 8 000 ha

(60%) of the Reserve (Fig. 1). This system is volcanic

in origin (Du Toit, 1954) and includes a chain of

mountains with vast, undulating grassland plateaux,

1 800 m to 1 900 m above sea level and flat grassland

plains at the foot of the mountains, 1 500 m to 1 650 m

above sea level. Rocky outcrops are found scattered

on the plateaux. The north-, northwest- and northeast-

facing slopes of the mountains are mostly gentle and

covered with grassland. The south-facing slopes, on

the contrary, are steep, especially where cliff's of

andesitic lava are exposed. These steep, usually rocky

slopes are mostly covered with dense bush (Fig. 2).

The plateaux are drained by a number of kloofs,

some of which cut deeply into the mountains resulting

in steep slopes, characteristically covered with woody

vegetation. The lower valleys in the bigger kloofs are

sheltered and contain deep, fertile clayey soils of

alluvial and colluvial origin. The soils of the study area

however, are, mostly fairly shallow with a sandy clay

or sandy clay loam texture.

The climate of the area was described in Part I of

this account (Bredenkamp & Theron, 1978).

Fig. 1. — A simplified geologi-

cal map of the Suikerbos-

rand Nature Reserve.

* Partly based on an M.Sc. thesis, University of Pretoria.

t Department of Botany, University of the North, Private Bag X5090, Pietersburg, 0700.

% Department of Botany, University of Pretoria, Hillcrest, Pretoria, 0002.

200 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE II THE PHYTOSOCIOT OGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM '

Fig. 2. — Dense bush on south-

facing slopes of the Ven-

tersdorp System moun-

tains.

METHODS

The methods applied are those used by Bredenkamp

& Theron (1978). Thirteen different physiographic

and physiognomic units represented in the study area

were delineated on aerial photographs. One hundred

and ninety three sample plots were divided pro rata

on an area basis among the thirteen units (Table 1),

and were placed at random within these units.

THE PLANT COMMUNITIES

The floristic composition of plant communities is

represented in phytosociological Tables 2 & 4 and

the prominent species in each community are indicated

in Tables 3 & 5. The symbol “P” indicates species

covering at least 5% of the area represented in the

releves in at least 25% of the releves representing a

community, while “/?” indicates species covering at

least 1 % of the area represented in the releves, in at

least 50% of the releves representing a community.

The symbol “t” in Tables 2 & 4 equals the symbol “r”

in normal Braun-Blanquet tables.

The vegetation may be divided into two major

communities:

The Euclea crispa — Rhoicissus tridentata Bush and

Savanna Communities (Tables 2 & 3) and

The Trachypogon spicatus — Themeda triandra Grass-

land Communities (Tables 4 & 5).

1 . Euclea crispa — Rhoicissus tridentata Bush and

Savanna Communities

These communities are found in the kloofs and

valleys and on the foothills and slopes of the mountains

within the study area and their floristic composition is

represented in Table 2. This vegetation is characterized

by the large Rhoicissus tridentata species group

(Table 2) with Euclea crispa as the most prominent

species (Table 3) and is divided into five principal

communities described under 1.1, 1.2, 1.3, 1.4 and 1.5.

1 . 1 Rhus pyroides — Leucosidea sericea Bush and

Savanna Communities

These communities include the dense forests in the

kloofs, as well as the isolated patches of forest and

savanna on the steep south-facing slopes of the

mountains. All these communities are characterized

by the Rhus pyroides species group (Table 2), with

Euclea crispa and Rhus pyroides very prominent.

Two communities each with two variations are

distinguished:

1.1.1 Rhus pyroides — Rhamnus prinoides Forest

(Fig. 3).

This dense forest vegetation is characterized by the

Rhamnus prinoides species group (Table 2) and is

mainly found in the sheltered kloofs within the

TABLE 1. — The distribution of the 193 sample plots in the 13 physiographic-physiognomic units

Physiographic-physiognomic units

* N=north-facing, etc.

TABLE 2. — A phytosociological table of the Euclea crispa — Rhoicissus tridentata Bush and Savanna Communities

o

'

■

G. J. BREDENKAMP AND G. K. THERON

TABLE 3. — Prominent species in the Euclea crispa—Rhoicissus tridentata Bush and Savanna Communities

203

Explanation of “P” and “p” in text.

Fig. 3. — Rhus pyroides —

Rhamnus prinoides Forest

in a kloof.

204 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. II. THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

study area, but isolated patches are found on steep

south-facing slopes. According to the Fosberg (1967)

classification, this community is an evergreen broad

sclerophyll forest (1A1/6).

Two variants are distinguished :

(a) Rhus pyroides — Rhamnus prinoides — Cassinop-

sis ilicifolia Variant

This variation is characterized by the presence of the

Dolichos falciformis species group and is negatively

associated with the Acacia karroo species group

(Table 2). It occurs in sheltered moist kloofs on sandy

loam soils at an altitude of 1 740-1 770 m.

The total canopy cover of the vegetation is 90-95 %

and an average of 28 species was recorded per releve.

The tree stratum is 8-10 m tall and has a canopy

cover of 70-90%. Euclea crispa (100%)*, Rhus

pyroides (100%), Diospyros lycioides (100%) Leuco-

sidea sericea (80%), Rhamnus prinoides (80%),

Heteromorpha arborescens (80 %), and Celtis africana

(80%) are prominent trees (Table 3). Cassinopsis

ilicifolia (80%), Buddleja salviifolia (60%), Olea

africana (60%), Cussonia paniculata (40%) and

Maytenus heterophylla (40 %) are also often present.

A definite shrub stratum, 2 m tall and with a canope

cover of 30-50 % is dominated by Diospyros lycioides

(100%) and Myrsine africana (Table 3), both forming

very dense thickets locally. Other shrubs includy

Diospyros whyteana (80 %), Psoralea polysticta (80 %),

Grewia occidentals (60%) and Calpurnia intrusa

(40%). Artemisia afra (60%) and Felicia filifolia

(40 %) are conspicuous in less dense patches.

Lianes are well represented in the tree and shrub

strata, with Rhoicissus tridentata (100%), Dolichos

falciformis (100%), Clematis brachial a (80%), Rubia

petiolaris (40 %) and Asparagus africanus (40 %) often

present.

The herbaceous layer is fairly open, with 20-50%

canopy cover, and is mostly less than 0,2 m tall,

but prominent individual shrublets of Isoglossa grantii

(40%) and Clutia hirsuta (40%) may be up to 1 m tall.

Galopina circaeoides (100%), Ehrharta erecta (80%),

Senecio isatideus (80%), Achyranthes aspera (80%),

Argyrolobium rupestre (60%), Carex spicato-paniculata

(40 %), Bromus leptoclados (40 %) and other species are

present.

(b) Rhus pyroides — Rhamnus prinoides — Acacia

karroo Variant

This variation occurs in drier kloofs, on clay loam

soils at an altitude of less than 1 740 m. It is cha-

racterized by a negative association with the Dolichos

falciformis species group, but the presence of some

species of the Acacia karroo species group (Table 2).

An average of 27 species was recorded per releve, and

the total canopy cover for this community is 80-95 %.

The tree stratum is often 10 m tall, with a canopy

cover of 50-90%. Euclea crispa (100%), Rhus pyroides

(89%), Celtis africana (78%) and Rhamnus prinoides

(78%) are prominent trees (Table 3). Other trees

include:

Heteromorpha arbo-

rescens 67 %

Acacia karroo 67%

Maytenus heterophylla 67%

Olea africana 56%

Buddleja salviifolia. . . 56%

Leucosidea sericea. ... 44 %

* In the following treatment where the symbol % follows the

name of a species, the figures indicate the constancy of that

species in the community.

A well-defined shrub stratum, 2 m tall and with a

canopy cover of 40-60% occurs in this community.

Prominent shrubs include Rhus pyroides (89 %),

Myrsine africana (78%), Acacia karroo (67%) and

Buddleja salviifolia (56 %) (Table 3). Other conspicuous

shrubs include Asparagus suaveloens (56 %), Diospyros

whyteana (44%), Grewia occidentals (44%), Lantana

rugosa (33 %), Calpurnia intrusa (33 %), Rubus rigidus

(22%), Lippia javanica (22%) and Diospyros austro-

africana (22 %).

Lianes often present in the tree and shrub strata

include Rhoicissus tridentata (89 %), Clematis brachiata

(78%), Rubia petiolaris (44%), Dolichos falciformis

(22%), Asparagus aethiopicus (22%), A. africanus

(22 %) and A. asparagoides (22 %).

The herbaceous layer varies considerably. It is

mostly up to 0,5 m tall, and the canopy cover varies

from 5%, where the upper strata are dense, to 70%,

where the upper strata are more open. Prominent

herbs include Ehrharta erecta (100%), Galopina

circaeoides (78 %) and Isoglossa grantii (78 %) (Table 3)

but Achyranthes aspera (78%) and Teucrium capense

(67 %) are locally conspicuous. Other herbs which may

be present include :

1.1.2 Rhus pyroides — Protea caffra Savanna

This community is characterized by the Berkheya

setifera species group (Table 2), and can further be

distinguished from the Rhus pyroides — Rhamnus

prinoides Forest by the presence of the Brachiaria

serrata species group. It occurs on south-facing

slopes, mostly at fairly high altitudes of more than

1 700 m above sea level. The vegetation is an ever-

green broad sclerophyll shrub savanna (1K1/2) of

Fosberg (1967).

Two variants are distinguished:

(a) Rhus pyroides — Protea caffra — Chrysanthemoides

monilifera Variant

This variation occurs on very steep (often more than

20°) rocky south to southeast facing slopes, and is

distinguished from the Rhus pyroides — Protea caffra —

Harpochloa falx Variant by the absence of the

Harpochloa falx species group (Table 2).

An average of 34 species was recorded per releve

and the total canopy cover of the vegetation varies

between 60 % and 90 %.

The trees are up to 6 m tall, and grow singly or in

groups to form scattered bush thickets. Although

Rhus pyroides (80%), Euclea crispa (80%) and

Maytenus heterophylla (80 %) are the most prominent

trees (Table 3), Protea caffra (40%) is conspicuous in

this community. Other trees which are often present in

this community include Leucosidea sericea (60%),

Buddleja salviifolia (60%) and Celtis africana (40%).

Although many shrubs occur in this community,

their total canopy cover never exceeds 30 %. Diospyros

lycioides (100%) and Myrsine africana (60%) are the

most prominent shrubs. Other shrubs include Cassine

aethiopica (60%), Rhus eckloniana (40%), Canthium

gilfillanii (40%), Diospyros whyteana (40%), Felicia

Fagara capensis 33%

Cussonia paniculata 33%

Pittosporum viridi-

florum 22%

Rhus leptodictya.. . . 22%

Ziziphus mucronata 22%

G. J. BREDENKAMP AND G. K. THERON

205

filifolia (40%), Psoralea polysticta (40%) and Celtis

africana (40%).

The lianes Rhoicissus tridentata (100 %) and Clematis

brachiata (100%) are constantly found in the tree and

shrub strata.

The herbaceous layer is often up to 1 m tall, and has

a canopy cover of up to 80%. The tall Cymbopogon

marginatus (80%) is generally conspicuous, but

Eragrostis curvula (40%) and especially Hyparrhenia

hirta (40%) are locally prominent (Table 3). Other

herbaceous plants which occur in at least two of the

five releves representing this community include:

(b) Rhus pyroides — Protea caffra — Harpochloa falx

Variant

This variation occurs in scattered patches on

southwest-facing slopes that are not as steep as those of

the previous community.

The Harpochloa falx species group (Table 2) is

characteristic of this community. An average of 44

species was recorded per releve and the total canopy

cover of the vegetation is more than 90 %.

The canopy of the tree stratum is often less than

10%, and Protea caffra (100%), Rhus pyroides (75%)

and Euclea crispa (75%) are the most prominent trees

(Table 3). Individuals of Rhus leptodictya (50%),

Celtis africana (50 %) and Cussonia paniculata (50 %)

are found scattered in this community.

Shrubs are poorly represented but Rubus rigidus

(100%) is constantly present and forms impenetrable

thickets locally. Rhus eckloniana (100%), Diospyros

lycioides (100%), Myrsine africana (75%) and Lippia

javanica (50%) are usually small inconspicuous

shrublets.

The herbaceous layer is well defined, up to 0,7 m

tall, and is usually very dense with a canopy cover of

70-90%. Trachypogon spicatus (100%) dominates the

herbaceous layer, while Harpochloa falx (100%),

Brachiaria serrata (100%), Senecio isatideus (75%),

Tristachya hispida (75%), Hyparrhenia hirta (75%)

and Themeda triandra (50%) are locally very con-

spicuous (Table 3). In spite of the relatively low

canopy cover of Berkheya setifera (100%), Rhus

discolor (100%) and Aster peglerae (100%) they are

constantly present and very conspicuous. The

following other herbs were present in at least two of

the releves representing the community:

1.2 Euclea crispa— Maytenus heterophylla— Setaria

nigrirostris Savanna

This community is situated on north-facing slopes

of kloofs at an altitude of 1 740-1 800 m. Scattered

trees and shrubs occur, but the herbaceous layer is

dense and well defined. The vegetation is an ever-

green broad sclerophyll shrub savanna (IK 1/2) of

Fosberg (1967). The community is characterized by

the Vernonia oligocephala species group (Table 2).

The simultaneous presence of the Phyllanthus parvulus

and Brachiaria serrata species groups, as well as

Aloe marlothii (Table 2) is another characteristic

feature of this community. The total canopy cover of

the vegetation is 60-90% and an average of 33

species was recorded per releve.

The tree stratum is sparse, with a canopy cover of less

than 20%. The trees are 3-4 m tall. Aloe marlothii

(66%), Euclea crispa (66%) and Maytenus hetero-

phylla (66%) are the most prominent trees (Table 3),

while Ziziphus mucronata (66%) and Cussonia

paniculata (66 %) are often present.

The shrub stratum is also sparse, but its canopy

cover is locally as high as 25 %. As in the case of the

tree stratum, the most prominent species are Aloe

marlothii (100%), Euclea crispa (100%) and Maytenus

heterophylla (100%), but Rhus eckloniana (100%),

Diospyros lycioides (66 %), Lippia javanica (66 %) and

Lantana rugosa (66 %) are also common.

The herbaceous layer is well defined, up to 0,5 m

tall and has a canopy cover of 60-80%. Hyparrhenia

hirta (100%) and Themeda triandra (66%) are the

most prominent species, but Trachypogon spicatus

(66%) and Cymbopogon marginatus (66%) are also

very conspicuous (Table 3). Other herbs present in

at least two of the three releves representing this

community include the following:

Setaria nigrirostris.. . . 100%

Vernonia oligocephala 100%

Aloe davyana 100%

Pellaea calomelanos. . 100%

Ledebouria marginata 100%

Athrixia elata 66%

Brachiaria serrata. ... 66%

Chaetacanthus bur-

chellii 66%

Phyllanthus parvulus 66%

Becium obovatum. . . 66 %

Diheteropogon am-

plectens 66 %

Pentanisia angusti-

folia 66%

Hypoxis rigidula. . . . 66%

Anthospermum ri-

gidum 66%

Elephantorrhiza ele-

phantina 66%

1.3 Euclea crispa — Maytenus polyacantha —

Canthium gilfillanii Bush and Savanna Communities

These communities include the dense bush and open

savannas which mostly occur on north-, west- and

east facing slopes of the mountains in the western

part of the study area. The Maytenus polyacantha

species group (Table 2) is characteristic of these

communities, and a prominent feature of it is that

Maytenus polyacantha and Canthium gilfillanii locally

form impenetrable thickets.

Three communities are distinguished and described

under 1.3.1, 1.3.2 and 1.3.3.

1.3.1 Maytenus polyacantha — Heteropogon contortus

Savanna

This community (Fig. 4) is mostly found on rocky

slopes of less than 22° between 1 600 and 1 700 m

above sea level. Rocks cover 21-60% of the soil

surface. The vegetation of releves 36, 40, 73 and 75.

situated on somewhat steeper slopes, included species

from the Buddleja saligna and Aloe marlothii species

groups (Table 2). An affinity with the Acacia caffra —

206 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. II. THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

Pig. 4. — Maytenus polyacan-

tha— Heteropogon confor-

ms Savanna.

Aloe marlothii Bush, normally found on the steeper

slopes, is thus indicated.

The community is characterized by the Heteropogon

contortus species group (Table 2). The trees and

shrubs often occur in scattered bush clumps. The

vegetation is an evergreen broad sclerophyll scrub

(1B1/4) of Fosberg (1967). The total canopy cover

varies from 60-95 % and an average of 35 species was

recorded per releve.

The tree stratum is up to 5 m tall with a canopy

cover of 5-40 %. This stratum is often dominated by

Canthium gilfillanii (94%) and Euclea crispa (89%)

(Table 3). These two trees often form groups of dense

thickets. Other trees present in this community include

Fagara capensis (78%), Cussonia paniculata (61%),

Celtis africana (44%), Dombeya rotundifolia (44%),

Maytenus heterophylla (39%), Ehretia rigida (27%),

Rhus leptodictya (22 %) and Ziziphus mucronata (22 %).

The shrub stratum has 10-40 % canopy cover, with

Maytenus polyacantha (78 %) the most prominent

shrub (Table 3). Canthium gilfillanii (94%) and

Euclea crispa (89%) are also often very prominent

(Table 3), in which case it becomes difficult to differen-

tiate between the tree and shrub strata. Other shrubs

include:

Fagara capensis 72 %

Diospyros lycioides. . 72%

Cassine aethiopica.. . . 67%

Rhus eckloniana 56 %

Isoglossa grantii 44%

Myrsine africana 44%

Maytenus hetero-

phylla 39%

Grewia occidentals 22%

Pavetta assimilis. ... 22%

Tarchonanthus cam-

phoratus 22%

Carissa bispinosa ... 1 7 %

The liane Rhoicissus tridenta (100%) is constantly

prominent in the tree and shrub strata, while Rubia

petiolaris (50%) and Clematis brachiata (33%) are

locally present.

The herbaceous layer is well developed, 0,5 m tall,

with canopy cover often up to 80%. Heteropogon

contortus (89%) and Themeda triandra (89%) are the

most prominent species (Table 3), but Elionurus

muticus (94%) is also very conspicuous. Commelina

africana (72%) and Ruellia cordata (61%) are well

represented in the shade of the dense bush clumps.

Other herbs present in this community include:

1.3.2 Euclea crispa — Maytenus polyacantha — Sco-

lopia zeyheri Bush

This community occurs in dense, isolated patches on

steep northeast- and northwest-facing slopes at an

altitude of 1 645-1 690 m. These dense bush clumps

are examples of evergreen broad sclerophyll scrub

(IB 1/4) of Fosberg (1967) and are characterized by the

Canthium mundianum species group (Table 2), and the

prominence of Euclea crispa (100%) (Table 2). An

average of 22 species was recorded per releve and the

total canopy cover is 70-90 %.

The tree stratum is 3-4 m tall and is very dense, with

a canopy cover of 60-90%. Together with the very

prominent Euclea crispa (100%), Canthium mundianum

(75%), C. gilfillanii (75%) and Rhus leptodictya

(75%) are also prominent trees (Table 3). Other

trees include Scolopia zeyheri (100%) and Dombeya

rotundifolia (100 %).

The shrub stratum covers 20-40 % and is dominated

by Euclea crispa (100%), Canthium mundianum

(75%), C. gilfillanii (75%), Felicia filifolia (100%) and

Maytenus polyacantha (75%). Other shrubs include

Scolopia zeyheri (100%), Lippia javanica (75%) and

Lantana rugosa (50 %).

Rhoicissus tridentata (100%) is a prominent liane

in the tree and shrub strata.

Owing to the dense tree and shrub strata, the

herbaceous layer is poorly developed, often not

taller than 0,2 m and with a canopy cover of only

G. J. BREDENKAMP AND G. K. THERON

207

5-10%. Conspicuous species present in this layer

include:

Commelina africana. . 100%

Aloe davyana 100%

Eustachys mutica . . . . 75%

Elionurus muticus. ... 75 %

Pavonia burchellii. ... 50%

Hyparrhenia hirta.. . . 50%

Trachypogon spica-

tus 50%

Themeda triandra. . . 50%

Eragrostis curvula. . . 50 %

Kalanchoe panicula-

lata 50%

1.3.3 Euclea crispa — Buddleja saligna — Acacia caffra

Bush

These communities are mostly found on steep,

rocky slopes (more than 22°), at an altitude of 1 660-

1 770 m. The dense vegetation ranges from Fosberg’s

(1967) deciduous thorn savanna (112/3) to micro-

phyllous deciduous thorn scrub savanna (1K2/5),

and is characterized by the Buddleja saligna species

group (Table 2).

Two variations are recognized :

(a) Acacia caffra — Aloe marlothii Variant (Figs 5 & 6)

This variation is situated on steep, rocky north-,

northwest- and northeast-facing slopes, where more

than 60 % of the soil surface is covered by large rocks.

The Aloe marlothii species group (Table 2) is charac-

teristic of this variant. The total canopy cover is

60-95 % and an average of 25 species was recorded per

releve.

The trees of this variation are up to 10 m tall and

the canopy cover of the tree stratum is 30-80%.

The most conspicuous trees are Aloe marlothii

(87%), Acacia caffra (80%), Rhus leptodictya (93%),

Tarchonanthus camphoratus (53%) and Buddleja

saligna (53%) (Table 3). Other trees include Dombeya

rotundifolia (73%), Euclea crispa (73%), Fagara

capensis (67%), Canthium gilfillanii (53%), Acacia

karroo (47%), Ziziphus mucronata (47%) and Celt is

africana (47 %).

The canopy cover of the shrub stratum is 10-50%.

Maytenus polyacantha (87%), Aloe marlothii (87%)

and Isoglossa grantii (80%) are the most prominent

shrubs in this community (Table 3). The large patches

of Maytenus polyacantha forming impenetrable bush

Fig. 5. — dense Acacia caffra —

Aloe marlothii Bush on

steep north-facing slopes

(a); Acacia karroo — Teu-

crium capense Savanna in

the valley (b); and Berk-

heya setifera — Koeleria

cristata — Pentanisia pru-

nelloides Grassland on the

south-facing slopes (c).

Fig. 6. — Acacia caffra — Aloe

marlothii Bush.

208 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. II. THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

thickets are a conspicuous feature of this vegetation.

Other shrubs include the following:

The lianes Rhoicissus tridentata (93 %) and Dioscorea

sylvatica (47 %) are often found in the tree and shrub

strata.

The herbaceous layer varies considerably as a

result of the local variation in the canopy cover of the

tree and shrub strata. The canopy cover of the

herbaceous layer is as low as 5% under the denser

woody layer, but may be up to 50% where the upper

layers are more open. Herbaceous plants often

present in this community include:

Eragrostiscurvula. . . . 80% Pellaea calomelanos 47%

Eustachys mutica 73% Ledebouria sp 40%

Aristida junciformis. . 67% Sidadregei 33%

Commelina africana. . 67% Ruellia cordata 27%

Themeda triandra. . . . 60% Aloedavyana 27%

Kalanchoe paniculata 53 %

(b) Euclea crispa — Buddleja saligna Variant

This variation is situated on rocky south- and east-

facing slopes with large rocks covering 21-60% of

the soil surface. The variation is distinguished from the

Acacia caffra — Aloe marlothii Bush by the absence of the

Aloe marlothii species group as well as the absence of

Acacia karroo (Table 2). An average of 27 species per

releve was recorded and the total canopy cover varies

from 70 to 80%.

The tree stratum is up to 8 m tall and has a canopy

cover of 50-70%. The most prominent species include

Buddleja saligna (86%), Euclea crispa (86%), Rhus

leptodictya (86%), Canthium gilfillanii (71%), Acacia

caffra (57 %) and Heteromorpha arborescens (43 %)

(Table 3). Other trees include Fagara capensis (71 %),

Ziziphus mucronata (57%), Celtis africana (57%),

Olea africana (43%), Cussonia paniculata (43%) and

Rhus pyroides (29 %).

The shrub stratum has a 20-50 % canopy cover and

is mainly composed of the following species:

Lianes found in the woody strata include Rhiocissus

tridentata (100%), Asparagus aethiopicus (57%) and

Dioscorea sylvatica (29 %).

The herbaceous layer is normally poorly repre-

sented, but Themeda triandra (71 %), Eustachys mutica

(57 %) and Eragrostis curvula (43 %) are fairly abundant

locally (Table 3). Other herbs include Commelina

africana (71%), Ledebouria sp. (71%), Kalanchoe

paniculata (57%), Aloe davyana (57%), Sida dregei

(57%) and Pellaea calomelanos (43%).

1 . 4 Acacia caffra — Teucrium capense Savanna

This community occurs on rocky south- and west-

facing slopes of the Holhoek Kloof. The vegetation

varies from the microphyllous deciduous thorn

shrub savanna (1K2/5) to the evergreen broad

sclerophyll forest (1A1/6) of Fosberg (1967). The

combination of Acacia caffra, Rhus pyroides, Teucrium

capense, Cynodon dactylon, Ehrharta erecta and

Tarchonanthus camphoratus is characteristic of this

community (Table 2). The total canopy cover of the

vegetation is 60-90% and an average of 28 species

was recorded per releve.

In the open areas the trees are up to 6 m tall, with a

canopy cover of 50-60%, but in denser areas the

canopy cover may be as high as 80% and the trees are

up to 12 m tall. The most prominent trees include

Acacia caffra (100%), Rhus pyroides (75%), Euclea

crispa (75%), Celtis africana (75%), Ziziphus mucro-

nata (75%), Tarchonanthus camphoratus (75%) and

Maytenus heterophylla (75 %) (Table 3).

The shrub stratum covers 10-30% and the most

prominent shrubs include Maytenus heterophylla

(75%) and Felicia filifolia (75%) (Table 3).

Other shrubs which may be present are Diospyros

lycioides (75%), Asparagus suaveolens (75%), Lippia

javanica (75%), Isoglossa grantii (50%) and Ehretia

rigida (50%).

The canopy cover of the herbaceous layer is only

15-40%, but the layer may be up to 1 m tall owing to

tall Cymbopogon validus (75%) and C. marginatus

(50%). Eragrostis curvula (25%) is locally prominent

(Table 3). Other herbs often found in this community

include Achyranthes aspera (100%), Ledebouria sp.

(100%), Ehrharta erecta (75%), Aristida congesta

(75%), Solanum rigiscens (75%), Pellaea calomelanos

(50 %) and Pavonia burchellii (50 %).

1.5 Acacia karroo — Teucrium capense Savanna

Communities (Fig. 5)

These open Savanna Communities occur at rela-

tively low altitudes, on clay loam soils, in sheltered

valleys and often along dry stream banks. These

communities are characterized by the Acacia karroo

species group (Table 2). Acacia karroo is by far the

most prominent species in this community, which

corresponds to the deciduous microphyll thorn

savanna of Fosberg (1967). Three variations are

distinguished and described under 1.5.1, 1.5.2 and

1.5.3.

1.5.1 Acacia karroo — Teucrium capense — Conyza

podocephala Savanna

This variation is situated on fairly rocky northeast-,

south- and southwest-facing slopes of less than 9°,

usually at the foot of adjacent rocky hills. The

Conyza podocephala species group (Table 2) is

characteristic of this variation. An average of 30

species was recorded per releve and the total canopy

cover of the vegetation is 60-80%.

The tree stratum is up to 6 m tall with 20-40%

canopy cover. Acacia karroo (100%) is the most

prominent tree (Table 3). Rhus pyroides (50%),

Euclea crispa (50%), Celtis africana (50%) and

Ziziphus mucronata (50 %) are locally present (Table 2).

The shrub stratum has a 5-20% canopy cover and

is composed of Acacia karroo (100%), Diospyros

lycioides (100%), Asparagus suaveolens (100%),

Lippia javanica (75%), Euclea crispa (50%), Felicia

filifolia (50 %) and Lantana rugosa (50 %).

The herbaceous layer is well developed with 50-70%

canopy cover. Eragrostis curvula (100%) is often the

most conspicuous species in the herbaceous layer

(Table 3), but Themeda triandra (75%), Hyparrhenia

TABLE 4. A phytosociological table of the Trachypogon spicatus—Themeda triandra Grassland Communities 209—21 0

Xi

.

G. J. BREDENKAMP AND G. K. THERON

211

hirta (75%) and Teucrium capense (100%) are locally

abundant. Other species include:

Conyza podocephala 100%

Hermannia depressa. 100%

Helichrysum rugulo-

sum 75%

Crabbea acaulis 75%

Solanum incanum.. . . 75%

Heteropogon contor-

tus 50%

Elionurus muticus*. . . 50%

Pollichia campestris 50%

Osteospermum mu-

ricatum 50%

Helichrysum nudifo-

lium 50%

Eragrostis chlorome-

las 50%

E. racemosa 50%

Setaria sp 50%

1.5.2 Acacia karroo — Teucrium capense — Felicia

muricata Savanna

This variation is usually situated on flat, rockless

areas on black clay soils, and is characterized by the

Albuca setosa species group (Table 2). The area was

previously overgrazed and trampled. Surface erosion

occurs on large areas while more severe donga

erosion is found locally. An average of 23 species was

recorded per releve and the vegetation has a total

canopy cover of 60-80 %.

The tree stratum is up to 8 m tall with a canopy

cover of 20-60%. This stratum is mostly fairly open,

but denser thickets appear locally. Acacia karroo

(100%) is the most prominent species (Table 3) but

Ziziphus mucronata (57 %), Euclea crispa (28 %) and

Rhus pyroides (28 %) are also present in this variation.

The open shrub stratum has a canopy cover of

10-20% and consists mostly of younger individuals of

Acacia karroo ( 100 %).

Local bush encroachment of Acacia karroo may

become a serious problem as a result. Other shrubs

present in this variation include Asparagus suaveolens

(100%), Pavonia burchellii (100%), Isoglossa grant ii

(86%), Lantana rugosa ( 71 %), Felicia filifolia (57%),

Maytenus polyacantha (43%), Lippia javanica (43%)

and Grewia Occident alis (29 %).

The herbaceous layer is often up to 0,8 m tall and is

often well developed with a canopy cover of up to 70 %.

Eragrostis curvula (100%) and Themeda triandra

(71%) are conspicuous (Table 3). However, many

bare patches are present in this community where the

herbaceous layer has a canopy cover of only 20%.

The pioneer species Felicia muricata (86%), Aristida

congesta (86%), Cynodon dactylon (71%), Aristida

junciformis (43 %) and Conyza podocephala (29 %)

are conspicuous here. Other herbs include (Table 2):

Teucrium capense. ... 100%

Albuca setosa 100%

Delosperma mahonii 71%

Bulbine narcissifolia. . 57%

Talinum caffrum 57%

Setaria flabellata 57%

Sida dregei 57%

Achyranthes aspera 43 %

Commelina africana 43 %

Kalanchoe rotundi-

folia 29%

K. paniculata 29%

Hypoxis rooperi .... 29 %

Antizoma sp 29%

1.5.3 Acacia karroo — Acacia caffra — Teucrium ca-

pense Savanna

This variation is found on rocky slopes of 8-17°

on the hills adjacent to the Acacia karroo — Teucrium

capense — Felicia muricata Savanna. This community

is differentiated from the other variations of the

Acacia karroo — Teucrium capense Savanna by the

absence of the Conyza podocephala and Albuca setosa

species groups (Table 2). An average of 22 species vas

recorded per releve and the total canopy cover of the

vegetation is 60-70%.

The tree stratum is up to 6 m tall and has a canopy

cover of 20-60%. Acacia karroo (80%) is the most

prominent species in this fairly open savanna, but

* Elionurus muticus ( =E. argenteus)

local bush encroachment of A. karroo results in dense

bush thickets. A. caffra (40%) is locally also prominent

(Table 3) and other trees present include Ziziphus

mucronata (60%), Celtis africana (60%) and Euclea

crispa (40%).

The shrub stratum is usually fairly open, with 10-

20% canopy cover, and is dominated by Acacia karroo

(80 %). Asparagus suaveolens (80 %) is also conspicuous

in this stratum (Table 3). Other shrubs include

Fippia javanica (80%), Diospyros lycioides (60%),

Felicia filifolia (60%), Isoglossa grantii (40%),

Maytenus polycant ha (40 %) and Ruellia cordata (40 %).

Rhoicissus tridentata (60%) is often found

scrambling in the tree and shrub stn ta.

The herbaceous layer has a canopy cover of 40-

60%, and is often up to 1 m tall. Eustachys mutica

(40%) and Cynodon dactylon (80%) are often promi-

nent (Table 3). The high canopy cover of the hardy

pioneer species Aristida congesta (100%), Cynodon

dactylon (80%), Sida dregei (60%), Eragrostis pseu-

dosclerantha (60%) and Aristida junciformis (40%)

is an indication of previous mismanagement of this

vegetation. Other herbs include the following:

Eragrostis curvula. . . . 100% Pavonia burchellii. . . 60%

Themeda triandra. .. . 60% Delosperma mahonii 40%

Elionurus muticus. .. . 60% Aloedavyana 40%

Teucrium capense. .. . 60% Eragrostis chlorome-

Achyranthes aspera . . 60% las 40%

2. Trachypogon spicatus — Themeda triandra Grassland

Communities

The communities of this grassland are summarized

in Table 4. These communities cover the high altitude

plateaux and low altitude flats, but also occur on

gentle rockless slopes. This grassveld is characterized

by the Trachypogon spicatus species group (Table 4)

often with Trachypogon spicatus, Themeda triandra

and Heteropogon contortus as the most prominent

species. Three major communities are distinguished

and are described under 2 . 1 , 2 . 2 and 2 . 3.

2.1 Rhynchelytrum setifolium — Monocymbium cere-

siiforme Grassland

This community occurs on very shallow soils on

the rocky outcrops sporadically distributed at altitudes

of 1 830-1 870 m, on the mountain plateaux. Large

rocks cover more than 80% of the soil surface and the

soil and vegetation are restricted to small patches

between the rocks. The canopy cover of the vegetation

is often less than 50% and the vegetation corresponds

to Fosberg’s (1967) seasonal orthophyll short grass

(1M2/1). An average of 23 species was recorded per

releve.

This community is characterized by the Mono-

cymbium ceresiiforme species group (Table 4), while

the presence of the Rhynchelytrum setifolium species

group, which is also present in the Berkheya setifera —

Rhynchelytrum setifolium Grassland, is also an

important feature of this community. Another

important characteristic of this community is its

negative association with the Acalypha punctata

species group, which occurs in all the other Trachypo-

gon spicatus — Themeda triandra Grassland communi-

ties represented in the study area.

A scanty shrub stratum is represented by a few

dwarfed shrubs. The only woody species which is

fairly constantly present is Rhus eckloniana (86%).

Small individuals of Euclea crispa, Maytenus hetero-

phylla, Myrsine africana, Diospyros lycioides and

Rhoicissus tridentata are locally present.

212 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. II. THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

TABLE 5. — Prominent species in the Trachypogon spicatus — Themeda triandra Grassland Communities

Explanation of “P” and “/>” in text.

The herbaceous layer is dominated by Rhynchely-

trum setifolium (100%) and Cymbopogon marginatus

(100%), but Themeda triandra (86%) is often locally

abundant (Table 5). Other species include the follow-

ing:

Monocymbium cere-

siiforme 100%

Pel laea calomelanos . . 1 00 %

Crassula setulosa 86 %

Sutera caerulea 86 %

Streptocarpus vande-

leurii 86%

Ursinea nana subsp.

leptophylla 86%

Senecio oxyriifolia.. . . 86%

Leonotis microphyl-

lum 71%

Selaginella dregei 71 %

Diheteropogon am-

plectens 57%

Cheila nthes hirta. ... 57 %

Oldenlandia herba-

cea 43%

Pelargonium luridum 43%

Trachypogon spica-

tus 43%

Eragrostis racemosa <3%

Ipomoea ommaneyi. 4 %

Microchic a caffra. . . 43%

Raphionacme hirsu-

ta 43%

2.2 Themeda triandra — Berkheya setifera — Rhus dis-

color Grassland Communities

These grassland communities are situated on the

plateaux (Fig. 7) and adjacent slopes, and are charac-

terized by the Berkheya setifera species group. Other

important features of these communities are the

presence of the Vernonia natalensis species group and

the high canopy cover and constancy of many species

of the family Asteraceae (Table 4). The vegetation

corresponds to a seasonal orthophyll tall grass

(1 L2/1) of Fosberg (1967). Four different communities

are distinguished and described under 2.2.1, 2.2.2,

2.2.3 and 2.2.4.

2.2.1 Berkheya setifera — Rhynchelytrum setifolium

Grassland

This community is usually found on the fairly

rocky patches, scattered on the plateaux. Rocks

cover 20-60% of the soil surface and the soils are

seldom deeper than 300 mm. The vegetation is up to

1 m tall, and usually very dense; canopy cover is

often as high as 95%. An average of 32 species was

recorded per releve.

The Vernonia galpinii species group (Table 4)

characterizes this community, and the Rhynchelytrum

setifolium species group, also characteristic of the

Rhynchelytrum setifolium — Monocymbium ceresiiforme

G. J. BREDENKAMP AND G. K. THERON

213

Grassveld on the rocky outcrops, is likewise an

important feature of this community.

Low-growing shrubby plants such as Rhus ecklo-

niana (53%), Athrixia elata (47%), Chrysanthemoides

monilifera (47%) and Elephantorrhiza elephanlina

(20%) are found scattered in this community, but

they are inconspicuous, because they are dwarfed by

the dense, tall-growing herbaceous layer.

Themeda triandra (80%) is often the most prominent

species. Rhynchelytrum setifolium (93 %), Berkheya

setifera (80%) and Heteropogon contortus (47%) are

also conspicuous (Table 5), while Eragrostis racemosa

(80%), Tristachya hispida (60%) and Rhus discolor

(53%) are locally abundant. Other species include the

following:

Brachiaria serrata. ... 100%

Acalypha punctata ... 73 %

Cymbopogon margi-

natus 73%

Vernonia galpinii 60%

V. natalensis 60%

Pentanisia angustifolia 60%

Senecio coronatus. ... 60%

Kohautia amatymbica 60%

Anthospermum rigi-

dum 60%

Trachypogon spicatus 60%

Lotononis calycina. . . 53%

Monsoniaattenuata. . 47%

Bewsia biflora 47%

Tephrosia capensis . . . 47%

Eragrostis capensis. . . 47 %

Rhynchosia totta 47 %

Becium obovatum. ... 40 %

Hypoxis rigidula 40 %

Crabbea acaulis 40 %

Aj uga ophrydis 40 %

Chaetacanthus bur-

chellii 40%

Helichrysum adscen-

dens 33%

2.2.2 Berkheya setifera — Koeleria cristata— Pen-

tanisia prunelloides Grassland { Fig. 5)

This very dense grassland occurs on south-facing

slopes that are mostly steeper than 10° and usually

occur at altitudes of more than 1 700 m. Large rocks

are locally present, but never prominent. The com-

munity is characterized by the Koeleria cristata

species group (Table 4). The vegetation is up to 1 m

tall with 60-95% canopy cover. A few scattered

individual shrublets of Rhus pyroides, Artemisia

afra, Cussonia paniculata, Athrixia elata and Psoralea

polysticta are dwarfed by the very dense herbaceous

layer.

The grasses Themeda triandra (87%) and Trachypo-

gon spicatus (87%) are very prominent (Table 5) and

cover large areas. Berkheya setifera (73%), Vernonia

natalensis (73%) and Rhus discolor (93%) are locally

conspicuous. Harpochloa falx (53%), Aster peglerae

(33%), Digitaria diagonalis (47%), Setaria nigrirostris

(33 %) and Helictotrichon turgidulum (33 %) are less

constantly present, but are locally abundant. Other

species present include:

2.2.3 Berkheya setifera — Hyparrhenia hirta Grass-

land

This community occurs at altitudes of 1 740-1 870m

on moderate to steep (3°-16°) north-, east- and west-

facing slopes (Fig. 8). The vegetation is up to 1,5 m

tall and has a canopy cover of 90-95%. An average of

26 species was recorded per releve.

Individual shrubs of Rhus eckloniana, Athrixia

elata, Lippia javanica, Elephantorrhiza elephantina

and Ziziphus zeyherana are found scattered in the

dense herbaceous layer.

The constant high canopy cover of Hyparrhenia

hirta (100%) (Table 4) is characteristic for this

community. Other prominent species are Trachypogon

spicatus (100%) and Themeda triandra (58%).

Ipomoea crassipes. . . 33 %

Bulbostylis burchel-

lii 33%

Lotononis foliosa. . . 33 %

Diheteropogon am-

plectens 33 %

Aloe davyana 33 %

Pelargonium luridum 27%

Helichrysum nudifo-

lium 27%

Ledebouria margina-

ta 27%

Cyanotis speciosa. . . 27 %

Eragrostis curvula. . . 20 %

Setaria nigrirostris. . 20%

Silene burchellii 20%

Microchloa caffra. . . 20%

Andropogon schiren-

sis 20%

Pleiospora cajanifo-

lia 20%

Cussonia paniculata 20%

Aster harveyanus.. . . 20%

Turbina oblongata. . 20%

Gnidia capitata 20%

Fig. 8. — Berkeya setifera —

Hyparrhenia hirta Grass-

land on moderate north-

facing slopes.

214 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. IE THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

Other species present include the following:

Grassland

This community is restricted to the less rocky

areas of the plateaux, at altitudes of 1 780-1 870 m.

Large single rocks occur sparsely in this community.

The presence of the Berkheya setifera species group

combined with the absence of the Rhynchelytrum

setifolium, Vernonia galpinii, Koeleria cristata and

Hyparrhenia hirta species groups (Table 4) is characte-

ristic of this community. The vegetation is 1 m tall

and has a canopy cover of 80-95 %. An average of 26

species was recorded per releve.

Themeda triandra (100%) and Trachypogon spicatus

(75%) are the most prominent species, but Berkheya

setifera (100%), Rhus discolor (88%), Digitaria

diagonalis (63%) and Heteropogon contortus (75%)

are locally conspicuous (Table 5).

Other species include the following:

2.3 Elionurus muticus — Themeda triandra Grassland

Communities

These grassland communities occur on the relatively

low altitude flats and on the undulating hills at the

foot of the mountains, and also in the lower parts of

the mountain slopes. They are characterized by the

Elionurus muticus species group (Table 4). Most parts

correspond to seasonal orthophyll tall grass (1L2/1),

but several patches appear to be seasonal orthophyll

short grass (1M2/1) of Fosberg (1967). Three com-

munities are distinguished and described under 2.3.1,

2. 3. 2 and 2.3.3.

2.3.1 Elionurus muticus — Themeda triandra —

Trachypogon spicatus Grassland

This community occurs at altitudes of I 600-

1 750 m, on the lower parts of the mountain slopes,

and clearly represents a transition between the

Themeda triandra — Berkheya setifera — Rhus discolor

Grassland Communities of the higher altitudes and

the remainder of the Elionurus muticus — Themeda

triandra Grassland Communities of the relatively low

altitudes. This is illustrated by the combination of the

Elionurus muticus species group typical of the low

altitude communities, with the Vernonia natalensis

species group, typical of the high altitude communities

(see Table 4). The vegetation is up to 0,5 m tall with

60-95 % canopy cover. An average of 32 species was

recorded per releve.

The shrubs Diospyros lycioides, Cussonia paniculata,

Maytenus heterophylla, Leucosidea sericea, Cassine

aethiopica, Euclea crispa and Rhoicissus tridentata

occur on rocky outcrops scattered throughout the

community.

The most prominent grasses in this community are

Trachypogon spicatus (94%), Elionurus muticus (94%),

Heteropogon contortus (82%), Eragrostis racemosa

(88%) and Eragrostis chloromelas (41 %) (Table 5).

Other species present include:

2.3.2 Elionurus muticus — Themeda triandra —

Teucrium capense Grassland

This grassland occurs on the relatively low altitude

(1 580-1 660 m) flats at the foot of the mountains,

and is usually situated adjacent to the Acacia karroo—

Teucrium capense Savanna Communities (1.5 above).

It probably represents a transition between these

Savanna Communities and the Elionurus muticus —

Themeda triandra — Eragrostis curvula Grassland

(2.3.3 below).

The Elionurus muticus — Themeda triandra — Teu-

crium capense Grassland is characterized by the

Senecio asperulus species group (Table 4). The total

canopy cover of the vegetation is 60-95%. An

average of 21 species was recorded per releve.

A conspicuous feature of this community is the

absence or low constancy of Trachypogon spicatus,

Acalypha punctata, Brachiaria serrata, Diheteropogon

amplectens, Aloe davyana, Ajuga ophrydis and Era-

grostis capensis, which are all constantly present in the

other Trachypogon spicatus — Themeda triandra Grass-

land Communities. The presence of pioneer species

such as Helichrysum rugulosum (88%), Senecio

asperulus (75%), Hermannia depressa (75%), Antho-

spermum rigidum (75%), Felicia muricata (50%) and

Conyza podocephala (38 %), as well as the local surface

erosion present in this community, suggest that this

vegetation has resulted from previous mismanagement.

G. J. BREDENKAMP AND G. K. THERON

215

Fig. 9. — Donga erosion in

Elionurus muticus — The-

me da triandra — Eragros-

tis curvula Grassland.

Shrubs that occur scattered in this community

include the following: Acacia karroo, Diospyros

lycioides, Ziziphus mucronata, Ehretia rigida, Canthium

gilfillanii, Euclea crispa, Asparagus suaveolens, Grewia

occidentalis, Rhus leptodictya and Rhoicissus triden-

tata.

The herbaceous layer is up to 1 m tall with Themeda

triandra (100%) the most conspicuous species.

Eragrostis chloromelas (63 %) and Elionurus muticus

(100 %) are locally very abundant (Table 5).

Other species present include the following:

Eragrostis curvula Grassland

This grassland occurs on the undulating hills and

flats at the foot of the mountains, at altitudes of

1 640-1 800 m. Surface erosion and more severe

donga erosion occur locally in this community

(Fig. 9). The presence of the Elionurus muticus species

group together with the absence of the Senecio

asperulus and Vernonia natalensis species groups is

characteristic for this community (Table 4). The

most prominent species include Themeda triandra

(93 %), Eragrostis curvula (86 %), Elionurus muticus

(86%), Eragrostis racemosa (79%), Heteropogon

contortus (79%) and Trachypogon spicatus (64%).

Other species present in this community include the

following:

Hermannia depressa 93 %

Anthospermum rigi-

dum 79%

Tephrosia capensis ... 79 %

Chaetacanthus bur-

chellii 29%

Andropogon appen-

diculatus 29%

DISCUSSION

As was pointed out by Coetzee (1974) and Breden-

kamp & Theron (1976; 1978), the polythetic nature of

the Braun-Blanquet method provides a very natural

classification of vegetation, where plant communities

are closely related to a specific set of environmental

conditions. This is again emphasized in the present

study.

Habitat conditions strongly associated with

differences in vegetation include altitude, aspect,

slope, rockiness of the soil surface, soil depth and

soil texture.

Closely related communities are grouped into larger

units, which may be of great practical value in a

management programme.

Although the communities are arranged into a

hierarchical system, the data obtained in this study

are not sufficient to determine character species or to

ensure correct ranking of the syntaxa represented in

the Bankenveld (Acocks, 1975).

ACKNOWLEDGEMENTS

The Nature Conservation Division of the Transvaal

Provincial Administration is thanked for the oppor-

tunity to work in the Reserve. The comments of

Mr B. J. Coetzee and Dr M. J. A. Werger, both

previously of the Botanical Research Institute,

Pretoria, are highly appreciated. The first author is

indebted to Prof. D. R. J. van Vuuren, Department

of Botany, University of the North, for his interest in

this project.

216 A SYNECOLOGICAL ACCOUNT OF THE SUIKERBOSRAND NATURE RESERVE. II. THE PHYTOSOCIOLOGY

OF THE VENTERSDORP GEOLOGICAL SYSTEM

UITTREKSEL

Die plantegroei van die Sis teem Ventersdorp van die

Suikerbosrandnatuurreservaat is volgens die Braun-

Blanquet-metode bestudeer en geklassifiseer. Beskry-

wings van plantgemeenskappe shut beskrywings van

habitateienskappe, totale floristiese samesteUing, lyste

van differensiele spesiegroepe asook prominente spesies

in die boom-, struik- en kruidstratums in. Habitat-

eienskappe wat met verskille in die plantegroei geasso-

sieerd is, is hoofsaaklik hoogte bo seespieel, aspek,

helling, klipperigheid van die grondoppervlak, grond-

diepte en grondtekstuur.

Die produk van die kiassifikasie is natuurlike plant-

gemeenskappe. Die range van sintaksa is nie bepaal nie,

omdat kennis van die plantegroei van die Bankenveld

{Acocks, 1975) onvoldoende is.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed.

Mem. bot. Si/rv. S. Afr. No. 40.

Bredenkamp, G. J. & Theron, G. K., 1976. Vegetation units for

management of the grasslands of the Suikerbosrand

Nature Reserve. S. Afr.J. wildl. Res. 6,2: 113-122.

Bredenkamp, G. J. & Theron, G. K., 1978. A synecological

acount of the Suikerbosrand Nature Reserve. I. The

phytosociology of the Witwatersrand geological system.

Bothalia 12: 513-529.

Coetzee, B. J., 1974. A phytosociological classification of the

Jack Scott Nature Reserve. Bothalia 11 : 329-347.

Du Tojt, A. C., 1954. The geology of South Africa. 3rd ed.

Edinburgh: Oliver & Boyd.

Fosberg, F. R., 1967. A classification of vegetation for general

purposes. In C. F. Peterken, IBP Handbook 4, Guide to

the checksheet for IBP areas pp. 73-120. Oxford: Blackwell

Scientific Publications.

Bothalia 13,1 & 2: 217-231 (1980)

The vegetation of seasonally flooded areas of the Pongolo

River Floodplain

H. D. FURNESS* and C. M. BREEN*

ABSTRACT

The Braun-Blanquet method of phytosociological analysis has been used to identify the plant communities

of the seasonally flooded areas of the Pongolo River Floodplain. Six communities and two sub-communities,

whose distribution is closely related to relative periods of exposure and inundation, were recognized. The struc-

ture and interrelationships of the communities are considered and the role of human activities in determining

their extent is commented upon. Comparison is made with similar situations elsewhere in Africa.

RtSUMt

VtGlTATION DES ZONES PERIODIQUEMENT INONDTES DANS LE LIT MAJEUR DE LA

RIVIERE PONGOLO

On a utilise la methode d’ analyse phytosociologique de Braun-Blanquet pour identifier les communautes

vegetales des zones periodiquement inondees dans le lit majeur de la riviere Pongolo. On y a reconnu six com-

munautes, dont la distribution est en correlation etroite avec les periodes relatives d' exposition et d'inondation. On

considere la structure et les interrelations des communautes et Ton commente le role joue par les activites humaines

dans la determination de leur amplitude. La situation est comparee a d’autres du me me genre qui se presentent

ailleurs en Afrique.

INTRODUCTION

The impoundment of rivers inevitably results in

changed conditions downstream and where this

involves a floodplain, the effects are likely to be of

considerable importance (Attwell, 1970; Begg, 1973;

Phelines, Coke & Nichol, 1973; Townsend, 1975).

This may be particularly pertinent in the African

context since some of the most important fisheries

are located on floodplains (Welcomme, 1974).

The object of this study was to define and to deter-

mine the factors influencing the distribution of the

plant communities of the seasonally flooded areas of

the Pongolo River Floodplain. It forms part of a

multi-disciplinary research programme whose objec-

tive is an assessment of the long-term effects of

impounding the waters of the Pongolo River on the

floodplain ecosystem.

This paper presents the results of a Braun-Blanquet

analysis carried out between June and September 1974.

THE STUDY AREA

The Pongolo River Floodplain (27°S, 37°E) is

situated on the low-lying coastal plain (75 m above

mean sea level) in the north-eastern corner of Natal,

South Africa (Fig. 1). It is approximately 60 km long

and 0,8 km wide. Although the floodplain covers

10 416 ha (Welcomme, 1974), this study only con-

sidered 8 800 ha, because the floodplain in the vicinity

of Nhlanjane Pan, which had not been flooded since

construction of the Pongolopoort Dam (1972) (Fig. 1),

was excluded.

Flooding is predominantly a summer phenomenon,

and prior to construction of the Dam, the major

floods were of relatively short duration: 28 cumec

flows lasted for periods of up to 100 days and caused

minimal flooding; 56 cumec flows for up to 60 days;

and flows in excess of 85 cumecs for only between

two and ten days (Table 1). The major floods, arising

from either the very high flows or lower flows for

longer periods, annually recharge a number of shallow

* Pongolo River Research Group, Departments of Botany and

Zoology, University of Natal, P. O. Box 375, Pietermaritzburg,

3z00.

218

THE VEGETATION OF SEASONALLY FLOODED AREAS OF THE PONGOLO RIVER FLOODPLAIN

TABLE 1. — The duration of different flow rates on the Pongolo

River Floodplain. High flows have the shortest duration.

Data compiled from information provided by the Depart-

ment of Water Affairs, Pretoria

waterbodies, locally referred to as pans (Hutchinson,

Pickford & Schuurman, 1932). These have a total area

of c. 2 600 ha (Musil, Grunow & Bornman, 1973),

but this is reduced considerably by evaporation to

less than 1 000 ha during the dry season.

Fluctuating water levels have been shown to

influence both the species composition and the distri-

bution of individual species in wetland areas (Howard-

Williams, 1972; Howard-Williams & Walker, 1974;

Howard-Williams, 1975). On the Pongolo River

Floodplain some 8 000 ha lies between high flood level

(HFL) and maximum retention level (MRL) of the

pans, i.e. the level at which pan and river lose contact

(Breen et al., 1978), and is therefore normally only

flooded for relatively short periods. A further approxi-

mately 1 000 ha becomes exposed gradually during

the dry season as levels drop below MRL, and is

therefore inundated for longer periods than that

between HFL and MRL. The natural flooding

regime has been altered by the impounding of the

river at Jozini (Fig. 1), and the effects of these changed

conditions on the floodplain vegetation will be dis-

cussed in a subsequent paper.

Inhabitants of the higher area immediately around

the floodplain make extensive use of the alluvial soils

for subsistence agriculture and most of the floodplain

vegetation is disturbed (Fig. 2). However, since a

portion of the floodplain was incorporated in the

Ndumu Game Reserve (Fig. 1) when it was established

in 1924, a relatively undisturbed area was available

for comparison.

METHODS

The Braun-Blanquet method of sampling and

synthesis as described by Werger (1974) was adopted.

Using aerial photographs, physiognomically distinct

vegetation areas were demarcated and field checks

were carried out to ensure that they were correctly

assigned. Optimal plot size was determined by increas-

ing the area until the number of species recorded

remained more or less constant. Plot size was deter-

mined as 100 m2 (5 m x 20 m) for the forest community

and, although it may be adapted to vegetation type

(Werger, 1974), all communities were sampled by

100 m2 plots. A total of 106 plots was investigated

during August and September 1974. All species

present were listed and, since most annuals only

appear after the floods have receded (usually in April),

it is probable that most were still recognizable at the

time of the survey.

After communities had been checked and demar-

cated on the aerial photographs (1 :25 000), they were

subjectively transcribed onto a map of the same scale.

The following site factors, soil pH in 1 N KC1

(Jackson, 1958), and proportions of clay, silt and

sand (Black, 1965), were determined. With the aid of

1 : 6 000 contour maps the height of each plot above

MRL and below HFL was estimated.

Plant nomenclature follows Ross (1972).

THE PLANT COMMUNITIES

Six communities have been recognized and they

may be grouped according to their relative periods of

exposure and inundation: (1) the communities (two)

of high-lying areas, which are only inundated for

short periods; (2) those (three) of low-lying gently

sloping areas, which only become exposed as water

approaches and drops below MRL, i.e. those having

the longest submergence and shortest period of

exposure; and (3) the community which occupies the

intermediate areas, where the slope is slightly steeper

Fig. 2. — An aerial view of

part of the Pongolo River

Floodplain showing the

extensive cultivation

along the levees and

riverbank. Only small

pockets of the Ficus

sycomorus — Rauvolfia

caffra Community remain

outside the Ndumu Game

Reserve.

219—220

Flo. 3._Map showing the distribution of the plan, communities of the seasonal., flooded areas of the Pongo.o River Floodplain between Mzinyeni Pan and the confluence of the Pongolo and Usutu Rivers.

21*

T A

wa

Pi.

of

bu

les

inf

bu

W)

He

Fit

(H

pai

(Bi

flo

ma

the

bei

re£

riv

coi

cu:

the

for

Fig. 2. — An aerial view of

part of the Pongolo River

Floodplain showing the

extensive cultivation

along the levees and

riverbank. Only small

pockets of the Ficus

sycomorus — Rauvolfia

caffra Community remain

outside the Ndumu Game

Reserve.

H. D. FURNESS AND C. M. BREEN

221

so that water drains off fairly rapidly thereby pre-

venting the development of communities typical of

the low-lying areas.

1 . Communities of the high-lying areas (Fig. 2)

Two communities may be recognized: the Ficus

sycomorus — Rauvolfia caffra Community with two

subcommunities, occupying 406 ha, which is

restricted to levees on either side of the main river

channel, and the Acacia xanthophloea — Dyschoriste

depressa Community, 128 ha in extent, which is

confined to narrow bands along the margin of the

floodplain (Figs 3 & 4).

1.1 The Ficus sycomorus — Rauvolfia caffra Com-

munity

This community forms a distinct vegetational unit.

Since most of the species present do not occur in any

other community, they may be regarded as character

or differential species (Table 2). They include Ficus

sycomorus, Rauvolfia caffra, Trichilia emetica, Entada

spicata, Syzygium guineense, Adina microcephala,

Allophylus decipiens, Kraussia floribunda and Monan-

thotaxis caffra.

Outside the Ndumu Game Reserve (Table 3:

releves 90, 65, 96, 81, 80, 87, 83 & 104), the community

usually has only two strata: a tall tree stratum

(12-15 m), formed principally by Rauvolfia caffra,

Ficus sycomorus and Trichilia emetica with cover of

60-80%, and a 2-3 m shrub stratum of Allophylus

decipiens, Grewia caffra, Monanthotaxis caffra, Ficus

capreifolia, Syzygium guineense and Adina microcephala

with cover of 10-20%. Within the reserve a further

two strata of shade tolerant species are evident

(releves 73, 72, 68, 69, 70, 71 & 74), the taller (0,75-

1,25 m, cover 15-20%) of Dicliptera heterostegia and

Setaria chevalieri, while in some releves (81 & 87)

the grass Oplismenus hirtellus forms a low stratum

(0,1-0,25 m) with cover not exceeding 10%. The

tendency for these two strata to be developed within

the Reserve reflects the agricultural disturbance on

the levees outside the Reserve.

Two sub-communities are recognizable:

(a) The Ficus sycomorus— Eriochloa meyeriana Sub-

community

This develops where the impact of agriculture is

severe. Because of clearing, the upper stratum (8-10 m,

cover 40 %) is only occasionally present (Table 3:

releves 93, 91, 49 & 48), and the middle shrub and

lower herb strata are poorly developed. The second

stratum of Dicliptera heterostegia, Monanthotaxis

caffra, Kraussia floribunda and Setaria chevalieri is

Fig. 4. — Diagrammatic representation of a cross-section of the Pongolo River Floodplain showing the relative

positions of the different plant communities.

TABLE 2. — Character* and differential! species of the Floodplain communities

222

THE VEGETATION OF SEASONALLY FLOODED AREAS OF THE PONGOLO RIVER FLOODPLAIN

0,75-1,25 m tall with slightly higher cover (15%)

compared with the Ficus sycomorus — Rauvolfia caffra

Community. The lower stratum (0,1-0,25 m, cover

15-60%) contains a number of species which is found

in other communities e.g. Sida alba, Eriochloa

meyeriana and Commelina africana. In moister areas,

Echinochloa pyramidalis forms quite dense mats

(releves 49 & 48). The presence of opportunist species

such as Sida alba and Commelina africana, and the

extent (2 142 ha) and the distribution of this sub-

community (Fig. 3) suggest that it results from dis-

turbance. It is thus present along the levees ani

between the old and present river course north and

east of Mzinyeni and south of Pongolwani Pans, sites

that are favoured for agriculture and where, under

undisturbed conditions, the Ficus sycomorus — Rauvol-

fia caffra Community would normally be present.

(b) The Ficus sycomorus — Acacia xanthophloea Sub-

community

This sub-community has a very restricted distribu-

tion (10 ha, Fig. 2) and is represented only by two

small stands (Table 3: releves 86, 85 & 58).

Four strata are present, the upper (8-10 m, cover

40-50%) being comprised of Ficus sycomorus, Acacia

xanthophloea and two climbers, Ipomoea digitata and

Jasminum fluminense. The 2-4 m shrub stratum,

dominated by Ficus capreifolia and Grewia caffra ,

but also containing Acacia xanthophloea, has relatively

low cover (10-20%). A third stratum of approximately

0,75 m, comprising mainly Kraussia floribunda,

Mimosa pigra and young Acacia xanthophloea, is

present with cover not exceeding 30%. The lowest

stratum (0,4 m) is characterized by both the greatest

species diversity and, at times, the highest cover.

Dominant species are Dyschoriste depressa, Cissam-

pelos mucronata, Eriochloa parvispiculata, Cardio-

spermum halicacabum, Hemarthria altissima, Erio-

chloa meyeriana and Echinochloa pyramidalis.

The two sites where this sub-community has

developed, to the south-west of Khangazini and west

of Mengu Pans (Fig. 2), are areas where the river

levees and the margin of the floodplain are close

together. Such a situation would facilitate an inter-

grading of the Ficus sycomorus — Rauvolfia caffra and

the Acacia xanthophloea — Dyschoriste depressa Com-

munities. This sub-community is therefore regarded

as transitional.

1.2 The Acacia xanthophloea — Dyschoriste depressa

Community (Fig. 5)

Acacia xanthophloea and Dyschoriste depressa are

confined to this community and the Ficus sycomorus —

Acacia xanthophloea Sub-community and they may,

therefore, be regarded as either character or diffe-

rential species (Table 2: releves 99, 18, 53, 23, 16, 11,

13, 15, 17, 45, 102, 30, 26, 32, 36, 100, 98 & 75). A

striking feature of this community is that it is formed

of two strata only, the tree stratum being composed

entirely of Acacia xanthophloea (8-12 m, cover 20-

50%). The second stratum is made up of low-growing

(0,15-0,45 m) herbs with sparse cover (usually 7-10%).

Dyschoriste depressa is the most prominent, others

being Sida alba, Heliotropium ovalifolium, Helio-

tropium indicum, Cynodon dactylon, Ambrosia arte-

misiifolia and Cardiospermum halicacabum. These

latter species tend to be widely distributed and have

low cover-abundance values (r or +). Many may be

regarded as opportunists, reflecting the unstable

conditions brought about by flooding and grazing.

In some areas outside the releves investigated a few

well-established Ficus sycomorus are present. These

stands are, however, not considered typical. They do,

however, serve to indicate the close relationships

between this community and the Ficus sycomorus —

Rauvolfia caffra Community.

1.3 Interrelationships

Both communities of the high-lying areas develop

on widely varying soils, usually slightly acidic (pH

4-6,6) and with extremely variable clay (9-68%), silt

(3-49%) and sand (16-62%) fractions. This suggests

that some other factor(s) predominate in determining

their distribution. Because of topographical variation

along the floodplain, different pans at MRL have

different elevations with regard to both the river

level and HFL. This has a striking influence on the

vegetation, those communities intolerant of flooding

being closely related to HFL, where those communi-

ties, which are more hygrophilous and more tolerant

of flooding, have a distribution that is closely related

to MRL irrespective of its height relative to HFL.

Fig. 5. — A view of the Acacia

xanthophloea — Dyscho-

riste depressa Community

taken during the extended

summer floods of 1976.

At this time the herb layer

becomes incorporated in-

to the aquatic system.

TABLE 3. — Communities of the seasonally flooded areas of the Pongolo River Floodplain

223—224

Phragmites

australis

Community

RELEVE NUMBER

HEIGHT ABOVE MAXIMUM RETENTION LEVEL (M)

HEIGHT BELOW HIGH FLOOD LEVEL (M)

TOTAL NUMBER OF SPECIES

TOTAL COVER (%)

CLAY ( % )

SAND (%)

SILT ( % )

pH (N KC1)

PHRAGMITES AUSTRALIS (CAV. ) TRIN. EX STEUD.

PHRAGMITES MAURITIANUS KUNTH

ECHINOCHLOA PYRAMI DALIS (LAM.) HITCHC . & CHASE.

POLYGONUM PULCHRUM BLUME

LUDWIGIA STOLONIFERA (GUILL . & PERR.) RAVEN

POLYGONUM SALI Cl FOLIUM WILLD .

POLYGONUM SENE.GALENSE MEISN .

' CYPERUS FASTIGIATUS ROTTB .

ALTERNANTHERA SESSILIS (L.) DC.

ERIOCHLOA MEYERIANA (NEES) PILG.

RANUNCULUS MULTIFIDUS FORSK .

CENTELLA ASIATICA (L.) URBAN

CYNODON DACTYLON (L.) PERS .

GNAPHALIUM PENSYLVANICUM WILLD.

GRANGEA MADERASPATANA (L.) POIR.

GLINUS LOTOIDES L.

SENECIO MADAGASCARIENSIS POIR.

POLYGONUM AVICULARE L.

HEMARTHRIA ALTISSIMA (POIR.) STAPF & C.E. HUBB .

RORIPPA MADAGASCARIENSIS (DC.) HARA

COMMELINA AFRICANA L.

CONYZA BONARIENSIS (L.) CRONQUIST

COTULA AUSTRALIS (SPRENG.) HOOK. f.

CARDIOSPERMUM HAL I C ACABUM L.

AMBROSIA ARTEMISIIFOLIA L.

HELIOTROPIUM INDICUM L.

EUPHORBIA INAEQUILATERA SOND .

COTULA ANTHEMOIDES L.

CAPERONIA STUHLMANNII PAX

CYPERUS SEEDLING

HELIOTROPIUM OVALIFOLIUM FORSK.

LEGUME SEEDLING

PASPALUM COMME RSON 1 1 LAM.

SIDA ALBA L.

ERIOCHLOA PARVISPICULATA C.E. HUBB.

CISSAMPELOS MUCRONATA A. RICH.

ACACIA XANTHOPHLOEA BENTH .

DYSCHORISTE DEPRESSA NEES

FICUS SYCOMORUS L.

RAUVOLFIA CAFFRA SOND.

TRICHILIA EMETICA VAHL

DICLIPTERA HETEROSTEGIA PRESL EX NEES, NON CHEV.

ENTADA SPICATA (E.MEY.) DRUCE

ADINA MICROCEPHALA (DEL) HIERN.

SYZYGIUM GUINEENSE (WILLD.) D.C.

SET ARIA CHEVALIERI STAPF EX STAPF & C.E. HUBB.

IPOMOEA DIGITATA L.

AGERATUM CONYZOIDES L.

ALLOPHYLUS DECIPIENS (SOND.) RADLK .

GREWIA CAFFRA MEISN.

FICUS CAPREIFOLIA DEL.

KRAUSSIA FLORIBUNDA HARV.

OPLISMENUS HIRTELLUS (L.) BEAUV.

MONANTHOTAXIS CAFFRA (SOND,) VERDC .

JASMINUM FLUMINENSE VELL.

OTHER ACCOMPANYING SPECIES

ABUTILON GUINEENSE (SCHUMACH.) BAK . f & EXELL

SESBANIA SESBAN (L.) MERRILL

MIMOSA PIGRA L.

PSILANTHES MAURITIANA (PERS.) DC.

GOMPHRENA CELOSIOIDES MART.

PHYLA NODIFLORA (L.) GREENE

CHENOPODIUM ALBUM L.

XANTHIUM STRUMA RIUM L.

SOLANUM NIGRUM L.

ECLIPTA PROSTRATA (L.) L.

GNAPHALIUM LUTEO- ALBUM L.

INFREQUENT SPECIES IN RELEVfiS OF TABLE 3.

AMBROSIA ARTEMISIIFOLIA L. (95: + )

AZIMA TETRACANTHA LAM. ( 1 3 : + , 15:+, 17:r)

BARLERIA SP . (11: r)

CENTEMA SUBFUSCA (MOQ.) LOPR. (32: 1)

COTULA AUSTRALIS (SPRENG.) HOOK. f. (890:+,

EUCLEA UNDULATA THUNB . (15: r)

GNAPHALIUM SP . (96 :+, 92:+, 94 :+)

GREWIA SUBSPATHULATA N.E. Br. (16 :1, 13 : r)

HYGROPHILA AURICULATA (SCHUMACH.) HEINE (75 :+,

LEPIDIUM SULUENSE MARAIS (15 : +)

LUDWIGIA STOLONIFERA (GUILL. & PERR. ) RAVEN (20

MAYTENUS SENEGALENSIS (LAM.) EXELL (6 :+)

SPOROBOLUS SMUTSII STENT (35 :+)

76 77 92 94 95

0 0 0,3 0,6 0,3

2,7 2,7 2,1 1,8 1,8

2 2 7 8 5

90 90 50 85 85

38 28 16 43 33

48 65 73 27 51

14 19 || 30 |6

6,4 6,3 5,6 5,6 4,9

5 5 2 2 1

2 12 2

74 :+, 96 :+)

96 : 1, 99 :+)

: r, 28 : r)

Phragmites

mauritianus

Community

51

0,6

52 57 101

0 0,6 0

1,2 1,8 0,6 1,8

4 5 6 7

100 100 100 90

71 46 56 39

20 31 22 21

9 23 22 40

88 64

0 0

3,0 3,0

7 3

90 100

69 66

25 18

5,3 4,4 4,1 4,9 3,6 5,0

Cyperus fastigiatus

Echinochloa pyramidalis Community

37

0,6

1.2

7

40

58

30

! 2

4,4

34 3 33 2 14 61 46 54 40 I 38 103 39 66 78 24 79 63

0,6 0,3 0 0,6 0,3 0,3 0,3 0,3 0,3 0,3 0,6 0,3 0,6 0,3 0,3 0,9 1,8 0,3

1,8 1,8 0,9 0,9 0,9 2,4 1,5 2,4 3,3 2,7 1,8 2,1 2,7 2,4 1,2 2,4 2,4 2,7 1,8 1,5 2,7

14

85

54

21

25

10

75

68

21

I I

6

70

72

I 2

35

7

40

64

15

21

13

95

63

18

19

I 3

90

51

30

I 9

12

90

70

16

14

12

80

51

36

I 3

60

66

22

12

7

75

72

4,2 3,8 3,9 4,8 3,4 3,7 4,5 4,0 4,3 5,3 4,0 5,2 4,9 3,4 5,3 3,8 4,4 4,0 3,4 3,9 4,1

6

90

68

16 17

16 24

4,9 4,0 4, I

I I

10

65

17

18

3,9

Nodum of

indeterminate

rank

56 2 1

0,3 0

19 35 20

0 0,9 0

1,5 2,1 2,4 1,2 2,4

17 23 24 20 17

60 55 70 70 75

46 45 69 22 62

33 35 19 71 20

21 20 12 7 18

6,1 4,3 4,8 4,7 4,3

Cynodon dactylon

Community

8 22

9 28

31 29

0,9 1,2 1,5 1,8 0,6 0,9 0,3 0,3

25 6 27

0,3 0,6 1,2

1,8 2,4 1,2 0,3 3,3 2,7 0,9 3,0 1,5 3,3 2,7 1,8

4 7

75 55

29 63

4

80

49

17 14

50 60

13 9 5 9 5 11

55 70 85 65 75 50

56 57 50 58 50 73

55 63

14 13

20 25 38

1 7 26 6

47 2 1

8 2 I

20

30

47 56

42 38

I I 6

4,8 6,8 4,6 4,9 5,7 6,2 6,3 4,8 4,5 5,4 5,5 5,9

Acacia xanthophloea - Dyschoriste

Community

depressa

F.sycomorus

A.

xanthophloea

Sub-

Community

99 43 18 53 23 16

1,5 3,0 1,5 1,2 2,7 2,1

3,3 0 0,3 0,6

12 II 12 6

50 60 70 40

63 28 51 26

30 62 17 45

7 10 32 29

4,1 5,3 4,5 5,1

8 15

45 65

60

20

53

30

17

Ficus sycomorus -

Eriochloa

Meyeriana

Sub - Community

+ i

i + +

3 3 2 2 3

+ 2 +

Ficus sycomorus - Rauvolfia coffra

Community

+ + +

+

+ +

■

.

H. D. FURNESS AND C. M. BREEN

225

Fig. 6a reveals that although the distribution of both

communities is relative to height below FIFL, and

therefore to the period of inundation while the river

is in flood, the Ficus sycomorus—Rauvolfia caffra

Community is the more sensitive to flooding in that it

develops on the highest areas. Clearly, however,

distribution of the community is not only determined

by period of inundation, because it is absent from

areas around the margin of the floodplain, even

where they may have an almost identical flooding

regime to that of the levees. It seems likely that season-

al availability of soil moisture acts in conjunction with

inundation, because proximity to the river would

prevent the development of the very dry conditions

which arise along the edge of the floodplain during the

dry winter months (unpublished data). The seasonally

drier conditions along the margin of the floodplain

eliminate Monanthotaxis caffra and other hygro-

philous woody species, while the presence of occa-

sional Ficus sycomorus suggests that it may be some-

what more tolerant. Alt lough more favourable soil

moisture conditions might be found at lower eleva-

tions along the floodplai 1 margin, development of the

woody component is prevented by the longer period of

inundation. The combination of these factors is

probably responsible for the distinct separation of the

Ficus sycomorus — Rauvolfia caffra and Acacia xan-

thoph'oea — Dyschoriste depressa Communities.

The Acacia xanthophloea — Dyschoriste depressa

Community tends to develop in slightly lower-lying

areas than the Ficus sycomorus- Rauvolfia caffra

Community, but above MRL (Fig. 4). It is, therefore,

inundated only while the river is in flood, and the

period of inundation is greater than that experienced

by the F. sycomorus — R. caffra Community.

2. The communities of low-lying areas

Three communities were recognized (Table 3). The

Phragmites australis and the P. mauritianus Com-

munities together occupy an area of c. 234 ha, most of

which (65%) is in the Ndumu Game Reserve. The

Cyperus fastigiatus — Echinochloa pyramidalis Com-

munity is one of the largest, covering c. 2 471 ha, with

particularly extensive stands occurring west of Tete

and Nsimbi Pans and in the Ndumu Game Reserve

(Figs 3 & 7).

2. 1 The Phragmites australis Community

This community is dominated by Phragmites

australis, which may be regarded as the character/

differential species (Table 2). Under protection in the

Ndumu Game Reserve (Table 3: releves 76 & 77), it

forms dense stands growing to a height of 2, 0-3,0 m

with high cover-abundance values and few other

species (releves 76 & 77). Where its vitality is reduced

by cutting and burning (releves 92, 94 and 95),

invasion by other species occurs, principa'ly Echino-

chloa pyramidalis and Eriochloa meyeriana that form

a stratum between 0,3 and 0,45 m.

2.2 The Phragmites mauritianus Community

Phragmites mauritianus is the dominant and dif-

ferential species (Table 2), growing to a height of

1,5-3, 5 m and forming dense stands with high cover-

abundance values (Table 3). It is usually associated

Sequence of Releves in Phytosociological Table

Fig. 6. — Position of the releves relative to: (a) height below high flood level;

tion level. Releves are ordered in the sequence defined by the phytosociological Table 3. Nodum ot

indeterminate rank.

226

THE VEGETATION OF SEASONALLY FLOODED AREAS OF THE PONGOLO RIVER FLOODPLAIN

with Echinochloa pyramidalis and Alternanthera

sessilis, although other species are also found in this

community (e.g. Eriochloa meyeriana, Sida alba and

Cynodon dactylon), particularly outside the Ndumu

Game Reserve. Only one releve (64) was, however,

examined within the Reserve. Where present, these

species form a low-growing stratum (0,15-0,4 m) with

low cover-abundance values.

2.3 The Cyperus fastigiatus — Echinochloa pyramidalis

Community (Fig. 7)

Cyperus fastigiatus and Echinochloa pyramidalis

are differential species (Table 2), for this community

(Table 3) and, when they have high cover-abundance

values (releves 59 & 105), few other species are present.

Where cover-abundance of the differential species is

lower, opportunist species such as Glinus lotoides,

Heliotropium indicum and Cardiospermum halicacabum

are more common. Three strata may be recognized,

an upper comprising mainly Cyperus fastigiatus (1-2

m; cover 20-80%), intermediate (0, 3-0,4 m; cover

15-80%) dominated by Echinochloa pyramidalis ; and

a layer formed by prostrate species such as Cynodon

dactylon and Polygonum aviculare (0,1-0,15 m;

cover 5-20%).

2.4 Interrelationships

The three communities of the low-lying areas all

develop on acidic soils with a pH of 3, 3-6,4 and of

rather variable texture (Table 3). It seems that the

Phragmites australis Community develops on soils

that are more sandy (27-73%) than those of the

Phragmites mauritianus and Cyperus fastigiatus —

Echinochloa pyramidalis Communities (10-57%).

The development of all three communities is closely

associated with MRL (Fig. 6b). This suggests that it is

not so much inundation as the presence of a more

stable water supply that controls the position of these

communities. The Phragmites Communities tend to

occupy the lower positions, P. australis favouring

swampy areas as opposed to P. mauritianus, which

prefers sites where there is water movement, such as

river banks. The Cyperus fastigiatus — Echinochloa

pyramidalis Community develops only on flat and

gently sloping draining areas adjacent to pans, and in

depressions that remain wet for most of the dry season

(Fig. 4).

The Cyperus fastigiatus — Echinochloa pyramidalis

Community shows marked affinities with the Phrag-

mites Communities through the extension of the distri-

bution of Echinochloa pyramidalis into these semi-

aquatic habitats, and with the Cynodon dactylon and

Acacia xanthophloea — Dyschoriste depressa Communi-

ties because of the presence of Cyperus fastigiatus in

these higher-lying areas. The degree to which these

rather broad distributional patterns reflect the distri-

bution under natural and undisturbed conditions is

debateable, because reduction in the cover of the

Phragmites Communities might favour intrusion of

Echinochloa pyramidalis, whereas grazing of Cyperus

fastigiatus seems to favour an increase in the Cynodon

dactylon component. Where both differential species

are grazed, particularly in the areas west of Sivungu-

vungu and to the south and south-west of Khangazini

Pans, their cover-abundance values are lower and a

wider variety of species is present (Table 3: releve 46,

38 & 39).

3. Communities of intermediate areas

Only one community, the Cynodon dactylon Com-

munity, has been recognized, covering 171 ha of the

floodplain. It is generally found on gently sloping

areas that become exposed gradually as the flood-

waters recede (Table 3, Fig. 6). The soils vary from

acidic (pH 4,5) to almost neutral (pH 6,8), with very

variable texture: clay 29-73%, silt 6-36% and sand

11-63%. Cynodon dactylon is extremely tolerant of

extended periods of dry conditions foi owing exposure,

and of submergence. Not surprisingly, therefore, the

distribution of the C. dactylon Community does not

show marked relatior ships with either height above

MRL or below HFL (Fig. 6a & b). It therefore extends

from below maximum retention level to above high

flood level (Fig. 4). C. dactylon may be found in all

the communities, particularly where disturbance has

occurred.

3.1 The Cynodon dactylon Community (Fig. 8)

Cynodon dactylon, the differential species (Table 2)

for this community, forms dense, almost pure, stands

Fig. 7. — A good example of

the Cyperus fastigiatus —

Echinochloa pyramidalis

Community. This stand

on the north-western edge

of Tete Pan has been

reduced in size and in

height. Most of the area is

now a mixture of Cyperus

fastigiatus, Echinochloa

pyramidalis and Cynodon

dactylon forming a stra-

tum of about 50 cm in

height.

H. D. FURNESS AND C. M. BREEN

227

Fig. 8.— The Cynodon dacty-

lon Community is grazed

extensively during the

late winter period when

pasturage elsewhere is

low.

of considerable extent around some pans (e.g. Nama-

nini-Bumbe 42 ha, east of Mthikeni and Nsimbi

27 ha). It forms a single stratum (0,2 m) with up to

90% cover where conditions are most favourable

(Table 3: releves 5, 8 & 10). Elsewhere, where cover is

reduced (releves 27 & 28), and grazing more intense,

the height may be less (0,05 m) and the weeds ( Am-

brosia artemisiifolia, Conyza bonariensis and Poly-

gonum aviculare ) become more prominent. These

species form an ill-defined stratum of 0,2-0, 3 m in

height.

3.2 Interrelationships

The close relationships of the Cynodon dactylon

Community with other communities on the flood-

plain are emphasized not only by the distribution of

C. dactylon, but also by five releves (56, 21, 19, 35 &

20) that form a nodum of indeterminate rank between

the C. dactylon and Cyperus fastigiatus-Echinochloa

pyramidalis Communities.

It seems probable that this community would

normally develop in areas where the inundation

period was too long for the woody communities and

where it became too dry for communities characteris-

tic of the low-lying areas (Fig. 6). Since it presently

extends into areas that clearly show relics of other

communities, it must be concluded that disturbance

has increased the extent of this community. At the

lower levels it is principally by replacement of Cyperus

fastigiatus, which is adversely affected by grazing and

trampling and, at higher levels, by clearing, which

reduces competition and allows more direct illumina-

tion, both of which favour Cynodon dactylon.

4. Releves of undetermined affinity

Six releves (50, 12, 47, 41, 60 & 62) do not appear to

fit satisfactorily into the communities outlined above.

Although they could have been allocated to specific

communities, thereby improving the information on

total floristics, the degree of affinity did not, in our

opinion, justify this action. They are included in the

table because an advantage of the Braun-Blanquet

method is that new releve data may continuously be

added, thereby facilitating recognition of communities

that may not be presently recognizable (Werger 1973).

Riparian Forest, which included the disturbed and

undisturbed vegetation of the levees and high-lying

margins of the floodplain, and the Aquatic and Margi-

nal Pan Vegetation. This broad classification, which

does not take into account the observed marked

influence of the flooding regime, does not agree well

with the community distinctions drawn up in this

study.

More recently, De Moor et al. (1977) and Pooley

(1978) have reported on the vegetation within the

Ndumu Game Reserve (Fig. 1). Only a portion of this

Reserve includes the floodplain of the Pongolo River.

De Moor et al. (1977) using the system of Fosberg

(1967), recognized six formations (Table 4) of which

five corresponded well with those recognized in this

study. The microphyllous deciduous shrub savanna

is, however, difficult to relate to our communities,

because it includes species such as Sesbania sesban,

Echinochloa pyramidalis and Phragmites australis,

which appear to exhibit markedly different responses to

inundation, and it therefore probably forms a transi-

tion between “forest and grasses” as has been sug-

gested by Pooley (1978). Neither De Moor et al.

(1977) nor Pooley (1978) have, however, attempted to

relate the plant communites to flooding regimes.

TABLE 4. — Comparison of the formations determined by

De Moor et al. (1977) with communities of this study

De Moor et al. (1977) Present study

Formation

1 Dry season deciduous forest

4 Tall evergreen graminoid

marsh

4a Tall evergreen graminoid

marsh

4b Seasonal orthophyll tall grass

4c Microphyllous deciduous

shrub savanna

5 Seasonal orthophyll meadow

Community

Ficus sycomorus — Rauvolfia

caffra

Phragmites mauritianus

Phragmites australis

Cyperus fastigiatus — Echi-

nochloa pyramidalis

Probably transitional be-

tween Cyperus fastigia-

tus— Echinochloa pyra-

midalis and Ficus syco-

morus— Rauvolfia caffra

Cynodon dactylon

DISCUSSION

The studies by Tinley in 1958 (published in 1976)

are the earliest reports on the Pongolo River Flood-

plain vegetation. He recognized two formations, the

Werger (1974a) has suggested that in areas exposed

to extreme conditions, emphasis should be placed on

permanently recognizable species in phytosociological

analysis. This concept was applied to the floodplain

228

THE VEGETATION OF SEASONALLY FLOODED AREAS OF THE PONGOLO RIVER FLOODPLAIN

vegetation because, particularly above MRL where

both inundation and exposure have to be tolerated,

conditions may be considered to be extreme. It

reduces the number of species from 64 to 25, but does

not alter the definitions of the communities (Table 5).

None of the communities recognized on the Pon-

golo River Floodplain is unique. They have been

recorded from widely separated areas on the African

continent, on floodplains and areas of fluctuating

water levels.

Phragmites australis and P. mauritianus have a

world-wide distribution (Clayton, 1967; Fernandes

et al., 1971), although P. mauritianus tends to be more

tropical and is frequently encountered in swampy and

seasonally flooded areas (Gordon-Gray & Ward,

1971; Howard-Williams & Walker, 1974; Rzoska,

1974). The distinction in ecological preference between

P. australis, which favours standing water, and P.

mauritianus which prefers moving water, supports the

observations of Gordon-Gray & Ward (1971).

Howard-Williams & Walker (1974) reported similar

environmental conditions in stands of P. mauritianus

in Lake Chilwa, although it was also present in

alkaline swamps. In other parts of Central Africa,

Vesey-Fitzgerald (1963) recorded Phragmites in

Riverine Grasslands and “lakes” where it formed

dense stands on silt banks, sand bars and in lagoons.

He did not record it from “pans” or “alkaline swamps

and flats”, but this may, in the former instance,

reflect the fact that the pans being investigated were

shallow depressions that normally dried up during the

dry season. They are clearly different from those of

the Pongolo system. Vesey-Fitzgerald did not record

the specific names of Phragmites and from the distri-

bution (Clayton, 1967; Fernandes et al., 1971) it

seems likely that, although both P. australis and P.

mauritianus could have been present, P. mauritianus

would be the more common.

Cyperus fastigiatus is widespread in the province of

Natal in South Africa (Ross, 1972), but has not been

recorded in tropical areas (Vesey-Fitzgerald, 1955;

1963; Dean, 1967; Cook, 1968; Howard-Williams &

Walker, 1974; Imevbore & Bakare, 1974; Rzoska,

1974). Ross (1972) claims that C. fastigiatus is closely

allied to the more tropical C. auricomus Sieber ex

Spreng., which is also placed with C. digitatus Roxb.

subsp. auricomus by Kiikenthal (in Ross, 1972).

Howard-Williams & Walker (1974) reported C.

digitatus from relatively acidic soils (pH below 6,0) in

their neutral to acidic marsh vegetation type. These

conditions are similar to those in which C. fastigiatus

is found on the Pongolo system.

Unlike Cyperus fastigiatus, Eehinochloa pyramidalis

is widespread in wet areas where it is often associated

with a variety of other species (Vesey-Fitzgerald,

1955; 1963; Dean, 1967; Cook, 1968; Imevbore &

Bakare, 1974; Howard-Williams & Walker, 1974).

The description of its growth in floodplain grassland

by Vesey-Fitzgerald (1963) aptly describes its beha-

viour in the Pongolo system: “Growth starts at the

onset of the rains but the stature depends on the

extent of flooding. Under optimum conditions the

previous season’s accumulation of rough rots away

in the water and the new growth is very vigorous”.

Although E. pyramidalis is grazed by hippopotamus

(Scotcher et al., 1978), the numbers of hippopotamus

outside the Ndumu Game Reserve on the Pongolo

Floodplain are small and they do not exert a marked

effect. However, as soon as the floods recede and the

marshy areas become dry enough for cattle, both

E. pyramidalis and C. fastigiatus are grazed. As in

tropical areas, even when drying out and being grazed,

node shoots remain green until quite late in the season.

In areas where grazing is particularly heavy and where

drainage is slightly more rapid following the floods,

E. pyramidalis may form a mosaic with Cynodon

dactylon as has been observed by Vesey-Fitzgerald

(1955) and Dean (1967).

The development of Cynodon dactylon Communities

under conditions of fluctuating water levels is well

documented, both as pure stands (Lea & Van V.

Webb, 1939; Vesey-Fitzgerald, 1955; Dean, 1967;

Greenway & Vesey-Fitzgerald, 1969) and in associa-

tion with other species (Lea & Van V. Webb, 1939;

Burnett, 1951 ; Anderson & Herlocker, 1973; Howard-

■ Williams & Walker, 1974). These short-grass lawns

develop on a wide range of soils from acid to alkaline

in areas where prolonged flooding is not experienced.

Dean (1967) noted that C. dactylon was easily killed

by flooding. On the Pongolo, it regularly tolerates

periods of submergence of up to 150 days, apparently

without much detrimental effect.

During summer the C. dactylon around the pans

may be inundated, thereby becoming an integral

part of the aquatic system. Even if it is not completely

inundated, the substrate is generally too wet for access

by cattle and goats. Thus it is only during the drier

parts of the year that these lawns become accessible

to terrestrially based grazers, for which it provides a

valuable source or pasturage.

The distribution of Acacia xanthophloea along

tropical and sub-tropical river courses and in damp

depressions is well documented (Anderson & Her-

locker, 1973; Greenway & Vesey-Fitzgerald, 1969).

Vesey-Fitzgerald (1974) concluded that the cyclic

and serai status of groves of A. xanthophloea were

substantially influenced by drainage conditions. A

consequence of this is that trees along the lake-shore

may succumb during periods of high lake levels

(Greenway & Vesey-Fitzgerald, 1969), a situation

observed on the Pongolo Floodplain particularly in

the vicinity of Mzinyeni Pan (Figs 1 & 5) during the

unnaturally extended floods caused by discharge of

water from the Pongolopoort Dam after its construc-

tion.

Ficus sycomorus occurs throughout Africa along

river courses, swampy areas and even arid areas where

the water table is high (Palmer & Pitman, 1972). Site

preference seems to be for those areas where drainage

is quite good and yet water is freely available, because

the best stands on both the Pongolo Floodplain (in

the Ndumu Game Reserve) and in Lake Manyara

National Park (Greenway & Vesey-Fitzgerald, 1969)

are along the tops of the river banks and levees.

This study has illustrated the importance of the

flooding regime in the development of the vegetational

communities. As a result, changes in the pattern of

flooding, both of frequency and perhaps more import-

antly of duration of inundation and exposure, can be

expected to exert a profound effect on the communi-

ties, as has been observed elsewhere (Dean, 1967;

Townsend, 1975, Attwell, 1970). With the very sandy

nature of the soils of the floodplain, particularly of

the levees and around the pans, destruction of vegeta-

tion, resulting either directly from changes in the

flooding pattern, or indirectly as a result of allowing

cultivation and grazing in areas that were formerly

too wet, could increase erosion, thereby having a

profound impact on the floodplain system as a whole.

TABLE 5. Communities of the Pongolo River Floodplain delineated on the permanently recognizable species only

H. D. FURNESS AND C. M. BREEN

229

/^.lunonnoo-qns

eireija./Cani *3

- sruoaiooXs *3

i^Timimnoo-qns

eaoxqdoqxuBX *v

-snaooiooyCs *3

A^TunoiraoD

Bssaudap

aqsiaoqos^a

-Baoxqdoqqirex

btobov

Tlqrunraraoo

uox^oep uopoiL/Co

jfirea

a4.Buxau9x.apuT

30 ampoM

/xTunairaoo

snuBixTunBai

saqiaiSBjqd

^iTunairaoo *-*

• CM

STXBjqsnB JJ

saxxaiBBjqd ^

Jasmirouro fluminese Veil

230

THE VEGETATION OF SEASONALLY FLOODED AREAS OF THE PONGOLO RIVER FLOODPLAIN

ACKNOWLEDGEMENTS

This research was funded by a grant from the

Inland Water Ecosystems Section of the National

Programme for Environmental Sciences. Dr J.

Morris of the Botanical Research Institute assisted

with tabulation and Drs D. Edwards and C. Howard-

Williams commented on the manuscript. The Natal

Parks Board provided accommodation in the Ndumu

Game Reserve.

UITTREKSEL

Die Braun-Blanquet-metode van fitososiologiese ana-

lise is gebruik om die plantgemeenskappe van die

seisoen-oorstroomde gebiede van die Pongolorivier-

vloedvlakte te identifiseer. Ses gemeenskappe en twee

subgemeenskappe, waarvan die verspreiding non in

verband staan met die periodes van relatiewe blootstel-

ling en onderwatersetting, is onderskei. Die struktuur

en onderlinge verwantskappe van die gemeenskappe is

in ag geneem en kommentaar is gelewer oor die invloed

van menslike aktiwiteite op die omvang daarvan.

Vergelykings met soortgelyke situasies elders in Afrika

is getref.

REFERENCES

Anderson, G. D. & Herlocker, D. J., 1973. Soil factors

affecting the distribution of the vegetation types and their

utilization by wild animals in Ngorongoro Crater, Tanzania.

J. Ecol. 61 : 627-651.

Attwell, R. I. G., 1970. Some effects of Lake Kariba on the

ecology of a floodplain of the mid-Zambezi valley of

Rhodesia. Biol. Conserv. 2: 189-196.

Begg, G. W., 1973. The biological consequences of discharge

above and below Kariba Dam. Proceedings of the Con-

ference of the International Commission on Large Dams.

Madrid. 1973.

Black, C. A. (editor-in-chief), 1965. Methods of soil analysis.

Madison, Wisconsin: American Society of Agronomy.

Breen, C. M., Furness, H. D., Heeg, J. & Kok, H., 1978.

Bathymetric studies on the Pongolo River Floodplain.

J. Limnol. Soc. sth. Afr. 4 : 95-100.

Burnett, G. F., 1951. Field observations on the behaviour of

the Red Locust, Nomadacris septemfaciata (Serville) in the

solitary phase. Anti Locust Bull. 8 : 1-37.

Clayton, W. D., 1967. Studies in the Gramineae XIV. Kew

Bulletin 21 : 111-118.

Cook, C. D. K., 1968. The vegetation of the Kainji Reservoir

site in northern Nigeria. Vegetatio 15 : 225-243.

Dean, G. J. W., 1967. Grasslands of the Rukwa Valley. J. Appl.

Ecol. 4: 45-57.

De Moor, P. P., Pooley, E., Neville, G. & Barichievy, J.,

1977. The vegetation of Ndumu Game Reserve, Natal:

a quantitative physiognomic survey. Ann. Natal Mas.

23: 239-272.

Fernandes, A., Launert, E. & Wild, H., 1971. Flora Zam-

besiaca 10: 91-94.

Fosberg, F. R., 1967. A classification of vegetation for general

purposes. In G. F. Peterken, Guide to the check sheet for

IBP. areas pp. 73-116. Oxford: Blackwell Scientific

Publications.

Gordon-Gray, K. D. & Ward, C. J., 1971. A contribution to

the knowledge of Phragmites ( Gramineae ) in South Africa,

with particular reference to Natal populations. Jl S. Afr.

Bot. 37: 1-30.

Greenway, P. J. & Vesey-Fitzgerald, D. F., 1969. The

vegetation of Lake Manyara National Park. J. Ecol. 57:

127-149.

Howard-Williams, C., 1972. Limnological studies in an

African swamp: seasonal and spatial changes in the

swamps of Lake Chilwa, Malawi. Arch. Hydrobiol. IQ-

379-391.

Howard-Williams, C., 1975. Seasonal and spatial changes in

the composition of the aquatic and semi-aquatic vegetation

of Lake Chilwa, Malawi. Vegetatio 30: 33-39.

Howard-Williams, C. & Walker, B. H., 1974. The vegetation

of a tropical African lake: classification and ordination of

the vegetation of Lake Chilwa (Malawi). J. Ecol. 62 •

831-854.

Hutchinson, G E., Pickford, G. E. & Schuurman, J. F. M.,

1932. A contribution to the hydrobiology of pans and other

inland waters of South Africa. Arch. Hydrobiol. 24: 1-154.

Imevbore, A. M. A. & Bakare, O., 1974. A pre-impoundment

study of swamps in the Kainji Lake Basin. Afr. J. Trop.

Hydrobiol. and Fish. 3 : 79-93.

Jackson, M. L., 1958. Soil chemical analysis Englewood Cliffs,

N.J.: Prentice-Hall.

Lea, A. & Webb, D. van V., 1939. Field observations on the

Red Locust at Lake Rukwa in 1936 and 1937. Sci. Bull.

Dep. Agric. For. Un. S. Afr. 189.

Musil, C. F., Grunow, J. O. & Bornman, C. H., 1973. Classi-

fication and ordination of aquatic macrophytes in the

Pongolo River pans, Natal. Bothalia 1 1 : 181-190.

Palmer, E. & Pitman, N., 1972. Trees of Southern Africa. Cape

Town : Balkema.

Phelines, R. F., Coke, M. & Nicol, S. M., 1973. Some biolo-

gical consequences of the damming of the Pongolo River.

Proceedings of the Conference of the International Com-

mission of Large Dams. Madrid, 1973.

Pooley, E., 1978. A checklist of the plants of Ndumu Game

Reserve, North-eastern Zululand. Jl S. Afr. Bot. 44:

1-54.

Ross, J. H., 1972. The flora of Natal. Mem. bot. Surv. S. Afr

No. 39.

Rzoska, J., 1974. The Upper Nile swamps, a tropical wetland

study. Freshwat. Biol. 4: 1-30.

Scotcher, J. S. B., Stewart, D. R. M. & Breen, C. M., 1978.

The diet of the hippopotamus in Ndumu Game Reserve,

Natal, as determined by faecal analysis. S'. Afr. J. Wildl.

Res. 8:1-11.

Tinley, K. L., 1976. The ecology of Tongaland. Durban: The

Natal Branch of the Wildlife Society of S. A.

Townsend, G. H., 1975. Impact of the Bennett Dam on the

Peace-Athabasca Delta/. Fish. Res. Bd. Can. 32: 171-176.

Vesey-Fitzgerald, D. F., 1955. The vegetation of the outbreak

areas of the Red Locust ( Nomadacris septemfasciata Serv.)

in Tanganyika and Northern Rhodesia. Anti Locust Bull.

20: 1-31.

Vesey-Fitzgerald, D. F., 1963. Central African grasslands.

J. Ecol. 51 : 243-274.

Vesey-Fitzgerald, D. F., 1974. The changing state of Acacia

xanthophloea groves in Arusha National Park, Tanzania.

Biol. Conserv. 6: 40-47.

Welcomme, R. L., 1974. The fisheries ecology of African flood-

plains. Rome: Food and Agriculture Organisation of the

United Nations.

Werger, M. J. A., 1973. Phytosociology of the Upper Orange

River Valley , South Africa. Ph.D. thesis, University of

Nijmegen, Holland.

Werger, M. J. A., 1974. On concepts and techniques applied in

the Zurich-Montpellier method of vegetation survey.

Bothalia 1 1 : 309-323.

Werger, M. J. A., 1974a. Applicability of Zurich-Montpellier

methods in African tropical and subtropical range lands.

In R. Tuxen, Handbook of Vegetation Science p. 13.

The Hague: W. Junk.

Werger, M. J. A., Kruger, T. J. & Taylor, H. C., 1972. A

phytosociological study of the Cape fynbos and other

vegetation at Jonkershoek, Stellenbosch. Bothalia 10:

599-614.

Bothalia 13,1 & 2: 231-235 (1980)

Phytogeography of fynbos

H. C. TAYLOR*

ABSTRACT

Previous classifications of the vegetation of the Cape Floristic Region, or Capensis, are outlined. The

distinctive features of the Cape flora such as species diversity, endemism and distribution patterns are discussed

in an attempt to elucidate the origins and evolution of the principal vegetation type of the region, known today

as fynbos. Evidence suggests that the present species-rich Cape flora has mainly evolved in and radiated from

the southwestern part of Capensis, an area where a true mediterranean-type climate is found.

RlESUMfE

PHYTOGlOGRAPHIE DU FYNBOS

On rappelle schematiquement des classifications anterieures de la region floristique du Cap, ou Capensis. Les

caracteres distinctifs de la flore du Cap, tels que la diversite des especes, I'endemisme et les modeles de distribution

sont discutes pour tenter d’elucider les origines et revolution du principal type de vegetation de la region, connu

aujourd’hui sous le nom de fynbos. Les indices disponibles suggerent que la flore du Cap actuelle, avec sa richesse

en especes, a principalement evolue dans le sud-ouest de la province du Cap et a rayonne a partir de cette region

qui posse.de un climat de vrai type mediterraneen.

INTRODUCTION

In recent works both Takhtajan (1969) and Good

(1974) allocated the rank of Kingdom to the Cape

flora despite its small area on a world scale, thus

according it equivalent phytogeographical importance

to such vast regions as the Holarctic Kingdom which

encompasses the whole of the temperate and arctic

northern hemisphere.

The Cape Floral Kingdom is concentrated in the

region known today as Capensis, the Cape folded-

mountain belt, that comprises the distinctive temperate

floral area of the southwestern and southern Cape

Province occurring between latitudes 31° and 35°

south and longitudes between 18° and 27° east. The

western part has a distinctly mediterranean-type

climate with dry summers and wet winters but east-

ward the rainfall becomes increasingly non-seasonal.

Over the area as a whole, rainfall varies from extremes

of 300 to 3 000 mm.

Capensis is bounded to the west and south by the

coast and to north and east principally by the Karoo-

Namib Region (Werger, 1978a) together with some

outliers of the Sudano-Zambezian (Werger & Coetzee,

1978) and Afromontane Regions (White, 1978). The

Karoo-Namib flora penetrates into Capensis in the

Little Karoo, an arid area between the coastal and

inland mountains (Werger, 1978a). On mountains in

the Knysna region Capensis intergrades with the Afro-

montane flora, and several Cape species such as

Berzelia intermedia, Diospyros glabra, Leucadendron

eucalyptifolium and Protea cynaroides become domi-

nant as one ascends from foothill forest to Mountain

Fynbos (White, 1978).

The vegetation of Capensis consists principally of

fynbos, a broad category of diverse evergreen

sclerophyllous shrublands comprising Acocks's (1975)

veld types 47 (Coastal Macchia), 69 (Macchia) and

70 (False Macchia), but includes two transitional veld

types, Coastal Renosterbosveld and Strandveld, that

contain a mixture of Cape and other floristic elements.

The broad phytogeographic demarcation of

Capensis began in the last century when botanical

travellers included “the region of the Cape flora” in

their descriptions of vegetation formations and floristic

* Botanical Research Unit, P.O. Box 471, Stellenbosch, 7600.

kingdoms. Among these pioneers were Schouw (1823),

Drege & Meyer (1843), Grisebach (1872), Rehmann

(1880), Engler (1882), Drude (1887, 1890) and

Schimper (1898). Their vegetation descriptions and

maps, based on scant information, were largely

conjectural, and are of mere historical interest today.

From the turn of the century, resident botanists

like Bolus (1886, 1905) and Marloth (1906, 1908)

began to describe and map the vegetation of South

Africa. Fynbos delineation was gradually refined by

Bews (1916), Pole Evans (1936), Adamson (1938a)

and others, until finally Acocks (1953) in his well-

known map that is still in use today, recognized the

three fynbos types mentioned above. In his discussion

of these and later works, Werger (1978b) concludes

that reasonable unity of opinion on the zoogeographi-

cal boundaries in Africa was reached much earlier

than was the case with phytogeographical boundaries.

The latter are, indeed, still being debated (e.g. Axelrod

& Raven, 1978) and a clear picture will only emerge

when the taxonomy and distribution of present

floras, including lower plant groups, are known in

detail.

CHARACTERISTICS AND AFFINITIES

The singular biogeographic features that led

phytogeographers to give the Cape flora such a high

status in their classifications include the great concen-

tration of species, the high degree of endemism, the

characteristic distribution patterns of typical elements

despite a general lack of species dominance, and the

predominance of certain families and genera. The

Cape flora “is noted for its richness in species, both

in small areas and over its whole range” (Taylor,

1978), and has been claimed as being one of the

richest in the world for its size (Oliver, 1977). For

diversity at the community level (alpha diversity),

Taylor (1972) has recorded 121 species of flowering

plants in a single 100 m2 quadrat in a homogeneous

stand of Mountain Fynbos. This figure, although

probably not the final tally, considerably exceeds the

average level of alpha diversity in equivalent shrub-

land communities of mediterranean-type ecosystems

elsewhere (Kruger et ale. paper delivered at the Second

International Symposium on African Wildlife Manage-

ment and Research, Pretoria, July 1977). Plant species

richness at the landscape level, such as a mountain

232

PHYTOGEOGRAPHY OF FYNBOS

range (gamma diversity — Whittaker, 1972), is much

greater than that of southwestern Australia, the

richest floristic zone in that continent (Kruger,

1977) ; while at the level of a floristic kingdom or

province, Goldblatt’s (1978) figure of 8550 species of

vascular plants in the Cape Floristic Region compares

well with the figure of 7 000 given by McLarty (1952)

for the whole of Western Australia, an area about

twenty times that of Capensis.

Two of the three families that are characteristic of

the Cape flora, Restionaceae and Proteaceae, have a

pronounced austral distribution. Both of them occur

on all three southern continents and have very few

representatives north of the equator. The third Cape

family, Ericaceae, has no connections with either of

the other two southern continents. Its subfamily

Ericoideae that occurs in the Cape has a strictly

north-south distribution. All three families have

strong concentrations of taxa in Capensis. Outliers

occur on mountains not only in adjacent dry areas

to the north, especially Namaqualand (Adamson,

1938b), but also intermittently along the eastern

highlands of the continent as far north as central

Africa (Oliver, 1977). Many other typical Cape taxa

have similar distribution patterns. In fact there are

so many in the Drakensberg that both Killick (1963,

1978) and Edwards (1967) have described a type of

“fynbos” for that area, and Boughey (1957), Hedberg

(1965) and Wild (1968) have recognized a “Cape

element” on the mountains of Rhodesia and East

Africa. But the floras of high mountains in southern

and central Africa are incompletely known and inten-

sive plant collecting is needed to provide data for a

full phytogeographic study of these interesting

migration routes.

Families that are richest in genera at the Cape but

widely distributed in other parts of the world include

Fabaceae, Iridaceae, Rosaceae and Thymelaeaceae.

In contrast, many taxa of high rank are endemic to

the Capensis region or nearly so. These include

Bruniaceae, Penaeaceae, Stilbaceae, Grubbiaceae,

Roridulaceae and Geissolomataceae, the tribe

Diosmae of the Rutaceae and large genera such as

Aspalathus, Phylica, Muraltia and Cliffortia (Oliver,

1977; Axelrod & Raven, 1978; Taylor, 1978). Wei-

marck (1941) found that of 282 genera with their

centre of origin in Capensis, 212 were endemic, and

he estimated that more than 3 500 species were

endemic out of the total of 4 200 that he surveyed.

In examining local endemism, Weimarck found that

the two westernmost centres together contained 45 , 5

percent of the total number of endemics represented

within the “Cape proper.”

Taxonomic studies show that both palaeo-endemics,

or relics, and neo-endemics, or recently evolved

species that are still limited to a small area, occur in

the Cape flora. These types have local or disjunct

distributions or both. The Proteaceous species

Orothamnus zeyheri, Mimetes hottentoticus and

Leucadendron argenteum are, by virtue of their

distribution and biology, considered to be palaeo-

endemics. Rourke (1972) reports vicariad groups of

neo-endemic Leucospermum species on coastal low-

lands of the southern Cape and Levyns (1954) shows

that several Muraltia species occupying the same

coastal strip are also youthful endemics. From

phylogenetic evidence, Oliver (1977) considers some

species in the minor genera of the Ericaceae, and even

some of the genera themselves, to be neo-endemic.

Marloth (1929) and Goldblatt (1972), among others,

have attributed this local endemism and disjunction

partly to the diverse topography of the region, in

which the concomitant diversity of soils and local

climates, and the appearance of new land surfaces

along the coast, gave opportunities for recent specia-

tion, and partly to the great age of the flora, members

of which may have been able to survive climatic

changes by retreating into favourable localities.

Weimarck (1941), Levyns (1938, 1952), Dahlgren

(1963), Williams (1972) and others have shown that

typical families and genera of the Cape flora have a

characteristic distribution pattern in which the highest

numbers of species per area are concentrated in the

western part of the region (see Fig. 1). Levyns (1952)

went so far as to state it as a “rule that all genera of

the Cape Flora show a concentration in the south-

Map 3. Concentration of Aspalathus species. The figures express the number of species represented in the area of each square.

The map is based on dot maps of all the species. The distribution in Natal and the Khamiesberg area are excluded (cf. map 1).

It is seen that the squares with the greatest concentration of species cover the mountain regions in the southwest. The number

in the square covering most of the Cape Peninsula is surprisingly high considering the area occupied by the Cape Flats,

which are poorer in species, and the sea. A high number of species is found in all the squares covering the western and

southern mountain ranges. The great contrast between the relative richness in the Witteberg — Zwartberg mountains and the

poverty of species in the Little Karroo deserts has only partly been possible to demonstrate.

Fig. 1. — Map 3 from Dahlgren (1963).

H. C. TAYLOR

233

west. Any apparent constituent, which does not show

this particular pattern of distribution, may be sus-

pected of being an invader.” The Restionaceae,

Proteaceae and most of the Ericaceae show this

pattern, as do the endemic families Bruniaceae and

Penaeaceae and the larger genera Cliffortia,

Aspalathus, Phylica, Muraltia and Leucadendron.

On the other hand, genera like Babiana, Ferraria

and Pteronia that are fairly common in Capensis have

their maximum concentration in Namaqualand.

Among other “apparent constituents” that do not

show the typical southwestern distribution are species

of Aloe, Erepsia, Carpobrotus, Crassula and Zygophyl-

lum. The presence of such species within the Cape

flora and the presence of fynbos outliers on high

mountains in Namaqualand as far north as Springbok

suggest an intermingling of the Cape flora with

those floras abutting it to the north.

In contrast, certain austral forest elements that

also do not have the typical southwestern con-

centration occur within Capensis only as relic patches

of forest vegetation in moist or sheltered habitats

and not as constituents of fynbos. Such genera as

Podocarpus, Cunonia, Platylophus and Curtisia are

examples of these (Levyns, 1962). Though their

presence suggests a previous wider distribution of

forest, there is little evidence that they are becoming

adapted to the typical rigorous habitats occupied by

fynbos.

The Cape flora also contains elements that are

found more commonly in more distant lands. The

grass Hyparrhenia hirta has a wide but disjunct

distribution in the coastal Mediterranean, in east

Africa and in southern Africa, linked only by high

mountain outliers in the Sahara (Quezel, 1978).

Genera such as Anemone, Rubus, Scabiosa, Geranium

and Dianthus have their main centres of concentration

in the northern hemisphere. Aloe, Euphorbia and the

Asclepiadaceae are prominent members of other

African floras. Gladiolus is widespread elsewhere in

Africa and beyond. Rhus and Euclea have many more

species in subtropical forest, scrub and savanna than

in fynbos (Taylor, 1978). Similarities at the generic

level suggest comparatively recent migrations and

intermingling. Similarities at a higher taxonomic level

suggest older affinities. Adamson (1958) considered

that pairs of taxa like Selaginaceae (Cape) and

Globulariaceae (Mediterranean), Dimorphotheca and

Calendula, Lobostemon and Echium, Crassula and

Sedum, and Widdringtonia and Tetraclinis provided

evidence of a once widespread flora that became

fragmented as the climate changed and then evolved

in isolation.

The affinity of the Cape flora with that of south-

western Australia is striking but more remote than its

affinities with northern hemisphere floras. The

Thymelaeaceae, Haemodoraceae and Droseraceae are

common to both continents and the endemic Cape

family Roridulaceae is closely paralleled by the

Australian Biblidaceae. Diosmae (Rutaceae) of the

Cape has its counterpart in the Australian tribe

Boroniae (Bolus & Wolley-Dod, 1904) and the genera

Tetraria and Gahnia, and Phylica and Cryptandra are

closely related (Adamson, 1958).

A picture emerges from this account of a present

flora with high alpha and gamma diversities, a flora

uniquely characterized by three widespread families

and some endemic ones, and by many endemic taxa

of lower rank, some young, some old, some widespread

within the region, some restricted or disjunct in

distribution. The flora has a high concentration of

species in the west, it shows some close taxonomic

affinities with abutting floras and with the central

African mountain flora, obvious but more distant

affinities with the flora of southwestern Australia, and

tenuous relationships with northern hemisphere

floras. In all, the Cape flora is floristically and phyto-

geographically unique. Despite migrations and inter-

minglings, it appears to have been isolated for a long

time and to have suffered vicissitudes that have

encouraged speciation, radiation and hybridization at

a singularly high rate.

ORIGINS AND EVOLUTION OF FYNBOS

Such features suggest a long and varied history of

geology and climate. On these grounds and because

dominance by one or more species is a rare pheno-

menon in mature fynbos, the flora has hitherto been

generally regarded as an ancient one (Marloth, 1915;

Bews, 1925; Weimarck, 1941; Levyns, 1952; Adam-

son, 1958; Dyer, 1966). Yet despite general agreement

on its age, there has been much controversy about the

origin of the Cape flora. One school postulated an

origin in the northern hemisphere, another in the

southern, while yet a third conten ied that it originated

somewhere in central Africa.

Until very recently the third theory has seemed

most plausible, mainly as a result of the perceptive

work of Levyns (1938, 1952, 1958, 1964), summarized

by Van Vuuren (1973). She pointed out that very

many members of the Cape flora, though they are

concentrated in the Capensis region, show clear

traces scattered throughout Africa, mainly on moun-

tains as far north as Ethiopia. Proceeding southwards

these “islands” become more frequent until, south of

the Swartberg, all the scattered mountains of the

Little Karoo have cappings of Cape plants while the

flora of the lowlands is entirely different (Levyns,

1950). Levyns showed, too, that the more primitive

members of Cape plants in many groups are to be

found in mountain outliers within the tropics, whereas

in the southwestern Cape many of the species are

advanced and occupy restricted geographical ranges.

These distribution patterns suggest that a flora of the

Cape type was once widespread in central Africa, and

that this flora retreated southward when the climate

became unfavourable in the north, leaving traces on

the northern mountains and speciating in the favour-

able temperate conditions found in the southwestern

Cape.

This subject has been recently reviewed by Taylor

(1978) who quoted further evidence suggesting that

the presumed central African origin in fact represented

a secondary centre of establishment for a flora that

originated in austral lands. A review by Axelrod and

Raven (1978), which appeared at the same time as

Taylor’s, presents evidence to support this theory.

The evidence strongly suggests that the summer-dry

climate is of recent origin in southern Africa and

probably only appeared at the beginning of the

Pleistocene some 2,5 million years ago. But already

in the early Miocene, rapid speciation probably took

place in South Africa with the broad warping and

uplift of the continent. Further study of fossil floras

lice those of Coetzee (1978a, b) is needed to sub-

stantiate this. At about this time increasing glaciation

in Antarctica brought the cold water of the Benguela

Current to the west coast, accentuating the trend to

increased summer drought on the western land

surfaces. Then, when strengthening high pressure

234

PHYTOGEOGRAPHY OF FYNBOS

systems brought this drier climate to the interior,

sclerophyllous taxa that had lived earlier under

summer and winter rain were adapting to increasingly

dry summers. With dry summers spreading from the

west, the taxa that required summer rain were

gradually restricted eastward. This left the western

environment open for the sclerophylls with tolerance

to withstand summer drought, and many of those

that survived in this new habitat had great oppor-

tunities for evolutionary radiation. This supports

the findings of Levyns and others, mentioned earlier,

that typical members of the Cape flora show a

concentration of taxa and of endemics in the west.

But not all fynbos species originated in their

present area. Axelrod & Raven (1978) postulate that

even during the most recent climatic changes in the

Pleistocene, fynbos in its present area may have been

restricted by the expansion of forest and other

vegetation, and could then have been displaced to

the north, into the regions now occupied by desert

and semi-desert. At the end of the Tertiary when

the area of dry climate expanded, fynbos would

have retreated to its present area and, during this

retreat, interchange between fynbos and isolated

pockets of relic sclerophyll vegetation may have

contributed directly and through hybridization to

the overall diversity of the flora. Thus, the rich flora

of the present Capensis region “may represent but

a remnant of a much richer sclerophyllous flora

that ranged over the present desert and steppe areas

into the Pleistocene” (ibid., 1978).

Though Axelrod and Raven’s hypothesis is attrac-

tive, the diversity and high rate of speciation in

fynbos may not simply be the result of mass plant

migrations following climatic change. As Levyns

(1963) has pointed out, a vegetation category, like

fynbos, is not “a flock of sheep” but an association

of taxa that extend and diminish their ranges, not

collectively but individually, in response to different

factors to which the taxa are variously adapted.

For example, the fact that fynbos is largely restricted

to nutrient-poor soils would preclude its movement

en bloc into the Karoo-Namaqualand area, even

though individual taxa may adapt to the richer

soils there.

The pulse of alternating cooler and warmer, and

at times wetter, climates during the Quaternary

would also have contributed to the high diversity in

the Cape flora. At times of moister climate, some

taxa of the sclerophyllous Cape flora could have

spread widely over the present Karoo region and

speciated there. “As drier climate returned, the

flora shifted coastward into its present area, bringing

new taxa with them and leaving relic stands in

moist situations” (Axelrod & Raven, 1978, p.116).

Thus, even during recent climatic changes in the

Quaternary, the Cape flora may have been far more

widespread (Levyns, 1938 in Axelrod & Raven, 1978),

and has only been restricted to its present environment

for a relatively short time.

An interesting feature in support of this view is

that some Cape plants enter into a period of rapid

vegetative growth towards the end of summer at a time

when water supplies in the southwestern Cape are at

their lowest. This strangely ill-adapted growth rhythm

suggests that the ancestors of these plants “evolved

in some place having a summer rainfall. The same

phenomenon has been recorded for South Australia,

where a similar change in climate is postulated to

account for the same, apparently inexplicable, features

of growth” (Levyns, 1964). This suggests that the

mediterranean climate is not ancient, but is so youthful

that the plants have not yet fully adapted to it.

Further research on South African plants is needed

to clarify this phenomenon.

As stated earlier, a true mediterranean climate is

present only in the western part of Capensis. It is

this western part that was first colonized by primitive

sclerophylls, and in the southwestern corner which

has the highest rainfall and most diverse topography,

speciation has been most active, producing the greatest

concentration of taxa and endemics. This rich south-

western centre can be regarded as the true home of the

Cape flora from whence it has radiated. To the north

its spread is limited beyond Vanrhynsdorp by an

arid climate, but eastward along the well-watered

south coast it extends into the regions of non-seasonal

and summer rain as far as Grahamstown. The eastern

extension, having relatively low diversity and few

endemics, is presumably younger than the western

part. Fynbos would probably only have started

colonizing this eastern area when the coastal temperate

forest was reduced by a drying climate (cf. Acocks,

1975, Maps 1 & 2), but it has expanded its range faster

within historic times owing to the destruction of

forest by man (Von Breitenbach, 1972), and veld

mismanagement is now encouraging its spread

further eastward into mountain grassland (Trollope

& Booysen, 1971).

ACKNOWLEDGEMENTS

This paper was originally written for a Council for

Scientific and Industrial Research publication syn-

thesizing present knowledge on fynbos. I appreciate

the Council’s permission to republish this slightly

expanded version, and I am grateful to the referees

and other colleagues for useful comment.

UITTREKSEL

’n Oorsig word gegee van vorige klassifikasies van

die plantegroei van die Kaapse Floristiese Gebied of

Capensis. Die onderskeidende kenmerke van die Kaapse

flora, soos verskeidenheid van spesies, endemisme en

verspreidingspatrone word bespreek in ’n poging om

lig te werp op die oorsprong en evolusie van die ver-

naamste plantegroei-tipe van hierdie gebied wat vandag

as fynbos bekend staan. Daar bestaan bewyse dat die

huidige ryk Kaapse flora ontwikkel het in die suid-

westelike deel van Capensis, ’n gebied met ’n egte

Mediterreense klimaat.

REFERENCES

Acocks, J. P. H., 1953. Veld types of South Africa. Mem.

bot. Surv. S.Afr. No. 28.

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed.

Mem. bot. Surv. S.Afr. No. 40.

Adamson, R. S., 1938a. The vegetation of South Africa. London:

British Empire Vegetation Committee.

Adamson, R. S., 1938b. Notes on the vegetation of the Kamies-

berg. Mem. bot. Surv. S.Afr. No. 18.

Adamson, R. S., 1958. The Cape as an ancient African flora.

Pres. Add. Sect. K(Botany), Adv. Sci. 15: 118-127.

Axelrod, D. I. & Raven, P. H., 1978. Late Cretaceous and

Tertiary vegetation history of Africa. In M. J. A. Werger,

Biogeography and ecology in Southern Africa 1: 77-130.

The Hague: Junk.

Bews, J. W., 1916. An account of the chief types of vegetation

in South Africa, with notes on the plant succession. J. Ecol.

4: 129-159.

Bews, J. W., 1925. Plant forms and their evolution in South

Africa. London: Longmans.

Bolus, H., 1886. Sketch of the flora of South Africa. In Official

Handbook of the Cape of Good Hope. Cape Town.

Bolus, H., 1905. Sketch of the floral regions of South Africa.

In Science in South Africa pp. 199-240. Cape Town: Maskew

Miller.

H. C. TAYLOR

235

Bolus, H. & Wolley-dod, A. H., 1904. A list of the flowering

plants and ferns of the Cape Peninsula, with notes on some

of the critical species. Trans. S.Afr. Phil. Soc. 14: 207-373.

Boughey, A. S., 1957. The origin of the African flora. Oxford:

University Press.

Coetzee, J. A., 1978a. Late Cainozoic palaeoenvironments of

Southern Africa. In E. M. van Zinderen Bakker, Antarctic

glacial history and world palaeoenvironments pp. 115-127.

Rotterdam: Balkema.

Coetzee, J. A., 1978b. Climatic and biological changes in

southwestern Africa during the Late Cainozoic. Palaeoeco-

logy of Africa 10/11: 13-29.

Dahlgren, R., 1963. Studies on Aspalathus. Phytogeographical

aspects. Bot. Notiser 116: 431-472.

Drege, J. F. & Meyer, E., 1843. Zwei Pflanzengeographische

Documente nebst einer Einleitung von Dr E. Meyer.

Besondere Beigabe zur ‘ Flora ’, 2.

Drude, O., 1887. Atlas der Pflanzenverbreitung. In Berghaus,

Physik. Atlas, Karte 6. Gotha.

Drude, O., 1890. Handbuch der Pflanzengeographie. Stuttgart:

Engelhorn.

Dyer, R. A., 1966. Impressions on the subject of the age and

origin of the Cape Flora. S.Afr. J. Sci. 62: 187-190.

Edwards, D., 1967. A plant ecological survey of the Tugela

River Basin. Mem. bot. Surv. S.Afr. No. 36.

Engler, A., 1882. Versuch einer Entwicklungsgeschichte der

extratropischen Florengebiete der suedlichen Hemisphere.

Leipzig: Engelmann.

Goldblatt, P., 1972. “Iridaceae” in Lectures on the Cape

Flora at the University of Cape Town’s Public Summer

School, Jan-Feb. 1972.

Goldblatt, P., 1978. An analysis of the flora of Southern

Africa: its characteristics, relationships and origins. Ann

Mo. bot. Gdn. 65 : 369-436.

Good, R., 1974. The geography of the flowering plants. 4th

ed. London: Longman.

Grisebach, A., 1872. Die Vegetation der Erde. 2 vols. Leipzig:

Engelmann.

Hedberg, O., 1965. Afroalpine flora elements. Webbia 19:

519-529.

Killick, D. J. B., 1963. An account of the plant ecology of the

Cathedral Peak area of the Natal Drakensberg. Mem. bot.

Surv. S.Afr. No. 34.

Killick, D. J. B., 1978. The Afro-alpine Region. In M. J. A.

Werger, Biogeography and ecology in Southern Africa 1 :

515-560. The Hague: Junk.

Kruger, F. J., 1977. Fynbos communities. In Council for the

Habitat Conf. on Mountain Environments of S.Afr.,

Johannesburg, May 1976.

Levyns, M. R., 1938. Some evidence bearing on the past

history of the Cape Flora. Trans. R. Soc. S.Afr. 26: 401-424.

Levyns, M. R., 1950. The relations of the Cape and Karoo

floras near Ladismith, Cape. Trans. R. Soc. S.Afr. 32:

235-246.

Levyns, M. R., 1952. Clues to the past in the Cape flora of

today. S'. Afr. J. Sci. 49: 155-164.

Levyns, M. R., 1954. The genus Muraltia. Jl S. Afr. Bot.

Suppl. 2.

Levyns, M. R., 1958. The phytogeography of members of

Proteaceae in Africa. Jl S.Afr. Bot. 24: 1-9.

Levyns, M. R., 1962. Possible Antarctic elements in the South

African flora. S.Afr. J. Sci. 58: 237-241.

Levyns, M. R., 1963. The origins of the flora of South Africa.

Lantern 13, 1: 17-21.

Levyns, M. R., 1964. Migrations and origin of the Cape Flora

Trans. R. Soc. S.Afr. 37: 85-107.

Marloth, R., 1906. The phytogeographical subdivisions of

South Africa. Rep. Brit. Assoc. Adv. Sci. 1905: 589-590.

Marloth, R., 1908. Das Kapland. Jena: Fischer.

Marloth, R., 1915. The effects of drought and some other

causes on the distribution of plants in the Cape reeion

S.Afr. J. Sci. 12: 383-390.

Marloth, R., 1929. Remarks on the realm of the Cape flora

S.Afr. J. Sci. 26: 154-159.

Mclarty, R., 1952. Foreword. In C. A. Gardner, Flora of

Western Australia 1. Perth: W. H. Wyatt.

Oliver, E. G. H., 1977. An analysis of the Cape flora. In

Proc. 2nd National Weeds Conf. of S.Afr. pp. 1-18. Cape

Town: Balkema.

Pole-evans, I. B., 1936. A vegetation map of South Africa.

Mem. bot. Surv. S.Afr. No. 15.

Quezel, P., 1978. Analysis of the flora of Mediterranean and

Saharan Africa. Ann. Mo. bot. Gdn. 65: 479-534.

Rehmann, A., 1880. Geobotanische Verhaltnisse von Sud-

Afrika. Bot. Centralbl. 1: 1119-1128.

Rourke, J. P., 1972. Taxonomic studies on Leucospermum

R. Br. Jl S.Afr. Bot. Suppl. 8.

Schimper, A. F. W., 1898. Pflanzen-Geographie auf physio-

logischer Grundlage. Jena : Fischer.

Schouw, J. F., 1823. Grundziige einer allgemeinen Pflanzen-

geographie und pflanzengeographischer Atlas. Berlin:

Reimer.

Takhtajan, A., 1969. Flowering plants : origin and dispersal.

Edinburgh: Oliver & Boyd.

Taylor, H. C., 1972. Fynbos. Veld & Flora 2: 68-75.

Taylor, H. C., 1978. Capensis. In M. J. A. Werger, Biogeo-

graphy and ecology in Southern Africa 1: 171-229. The

Hague: Junk.

Trollope, W. S. W. & Booysen, P. de V., 1971. The eradication

of macchia (fynbos) vegetation on the Amatole Mountains

of the Eastern Cape. Proc. Grassld. Soc. Sth. Afr. 6: 28-38.

Van Vuuren, D. R. J., 1973. Die oorsprong en verwantskappe

van die Suid-Afrikaanse flora. Publ. v.d. Univ. v.d. Noorde,

Reeks C, No. 25.

Von Breitenbach, F., 1972. Indigenous forests of the Southern

Cape. J. bot. Soc. S. Afr. 58: 1 8—47.

Weimarck, H., 1941. Phytogeographical groups, centres, and

intervals within the Cape Flora. Lund. Univ. Arsskrift Avd.

2. Bd. 37 Nr. 5.

Werger, M. J. A., 1978a. The Karoo-Namib Region. In M. J. A.

Werger, Biogeography and ecology of Southern Africa

1: 231-299. The Hague: Junk.

Werger, M. J. A., 1978b. Biogeographical division of Southern

Africa. In M. J. A. Werger, Biogeography and ecology in

Southern Africa 1 : 145-170. The Hague: Junk.

Werger, M. J. A. & Coetzee, B. J., 1978. The Sudano-Zam-

bezian Region. In M. J. A. Werger, Biogeography and

ecology in Southern Africa 1: 301—462. The Hague: Junk.

White, F., 1978. The Afromontane Region. In M. J. A. Werger,

Biogeography and ecology in Southern Africa 1: 463-513.

The Hague: Junk.

Whittaker, R. H., 1972. Evolution and measurement of

species diversity. Taxon 21 : 213-251.

Wild, H., 1968. Phytogeography in South Central Africa.

Kirkia 6: 197-222.

Williams, I. M. J., 1972. A revision of the genus Leucadendron

(Proteaceae). Contr. Bolus Herb. 3: 1—425.

Bothalia 13, 1 & 2: 237 (1980)

Miscellaneous ecological notes

VARIOUS AUTHORS

NOTES ON THE USE OF THE TERM “RENOSTERVELD”

These notes have been compiled to clarify and

standardize botanical reference to the vegetation

referred to as Rhenosters bosch (Van der Stel in

Waterhouse, 1932), Rhenosterveld (Marloth, 1908;

Bews, 1916; Levyns, 1929, 1935 & 1972; Acocks,

1933, 1953 & 1975), Rhenoster-veld (Adamson, 1938),

Renosterbosveld (Jordaan, 1946; Acocks, 1953 &

1975), Rhenosterbosveld (Acocks, 1953 & 1975) and

Renosterveld (Smith, 1966; Taylor, 1978).

In 1685, Simon van der Stel found that the Olifants

River Valley . . is bewassen met Rhenosters bosch

alhier so genaemt om dat de selve daer gemeenlyck in

legeren, . . (Waterhouse, 1932). This passage has

been translated by Waterhouse (1932) as: “The above-

mentioned valley is overgrown with rhinoceros wood,

so-called here because these animals are usually

found in it.” This is therefore the first reference to

the vernacular name for “renosterveld”. The unit

“bosch” having been replaced by “veld” and the

Afrikaans spelling “renoster” for the rhinoceros having

been introduced.

Levyns (1972), in contrast, suggests that the grey,

uneven appearance of a “rhenosterveld” community,

when viewed from a distance, resembles the wrinkled

hide of the rhinoceros.

Although the vernacular name “renosterbos”, for

the plant, Elytropappus rhinocerotis, is an undisputed

fact, Smith (1966) has wrongly referred to the above

passage in Van der Stefs journal as the first use of

this name for the plant, while Van der Stel was, in

fact referring to the vegetation unit. Smith, quoting

Waterhouse (1932), has wrongly spelt “Rhenosters

bosch” as “Rinocerbosch” (sic).

The recent tendency to refer to this vegetation as

“rhenosterbosveld” or “renosterbosveld” might result

from E. rhinocerotis often becoming dominant follow-

ing disturbance of the vegetation. The absence of any

rhinoceros ( Diceros bicornis), through their extinction

in this habitat, may be a contributing factor to

emphasis being placed on the renosterbos plant

instead of on the rhinoceros itself. Acocks (1975),

however, uses the terms “rhenosterveld” and

“rhenosterbosveld” indiscriminately.

The term “renosterbosveld” could be misconstrued

as “renoster-bosveld”. Bushveld or “boschveld”, the

parklands of the northern Transvaal (Carpenter, 1938),

are absent in the Cape Province.

It is proposed here that the original and most

widely used concept be retained using the correct

Afrikaans spelling “renosterveld”.

1 thank Mr M. J. Wells and Dr D. Edwards for

providing the stimulus for this note and Mr H. C.

Taylor and Mrs M. F. Rand for their valued com-

ments.

REFERENCES

Acocks, J. P. H., 1933. Vegetation of a portion of the Cape

Flats and a list of the plants so far found there. Unpubl.

M.Sc. thesis, Univ. Cape Town.

Acocks, J. P. H., 1953. Veld types of South Africa. Mem. hot.

Surv. S. Afr. 28: 1-192.

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. hot.

Surv. S. Afr. 40: 1-128.

Adamson, R. S., 1938. The vegetation of South Africa. London:

British Empire Vegetation Committee.

Bews, J. W., 1916. An account of the chief types of vegetation

in South Africa with notes on plant succession. J. Ecol.

4: 129-159.

Carpenter, J. R., 1938. An ecological glossary. New York:

Hafner.

Jordaan, P. G., 1946. Plantegroei van die distrikte Bredasdorp

en Caledon. Tydskr. Wetensk. en Kuns, nuwe reeks 6:

47-58.

Levyns, M. R., 1929. The problem of the rhenoster bush.

S. Afr. J. Sci. 26: 166-169.

Levyns, M. R., 1935. A revision of Elytropappus Cass. Jl S.

Afr. Bot. 1 : 161-170.

Levyns, M. R., 1972. The rhenosterbush. Veld & Flora 2 (1):

7-9.

Marloth, R., 1908. Das Kapland. Jena: Gustav Fischer

Verlag.

Smith, C. A., 1966. Common names of South African plants.

Mem. bot. Surv. S. Afr. 35: 1-642.

Taylor, H. C., 1978. Capensis. In M. J. A. Werger, The

biogeography and ecology of southern Africa. The Hague:

Junk.

Waterhouse, G., 1932. Simon van der Stel's journal of his

expedition to Namaqualand 1685-6. Dublin: Longmans,

Green.

C. Boucher

\

OBITUARIES

239

JOHN PHILLIP HARISON ACOCKS (1911-1979)

John Acocks (originally Acock) was born in Cape

Town on 7th April 191 1 (Fig. 1). He received most of

his schooling at the South African College School

(SACS) and then entered the University of Cape

Town, where he graduated with B.A. & B.Sc. in 1932,

majoring in Botany, Physics, Chemistry and Latin,

In botany he studied under Professors R. S. Adamson,

R. H. Compton, Dr M. R. B. Levyns and Miss E. L.

Stephens. In 1933 he was awarded the M.Sc. degree

for a thesis entitled “Vegetation of portion of the

Cape Flats”. Subsequently he worked for two years

towards a Ph.D. degree on the effect of burning,

resting, clearing and cultivation of the humid fynbos

of Kirstenbosch and the more arid fynbos of Tierbos,

Hout Bay. This work was undertaken at the instigation

of Dr I. B. Pole Evans, then head of the Division of

Plant Industry, who was looking around for suitable

pasture research officers. Owing to lack of essential

apparatus apparently not supplied in time by the

Department of Agriculture, Acocks was unable to

finish his project. It was about this time that Acocks

(letter to Mr L. S. Richfield of 1976-12-03), while

looking at fynbos on Sir Lowry’s Pass, observed that

there “were a lot of species that one already knew at

sight with certainty, there were a lot more in collectable

condition and the remainder would sooner or later

become collectable. Eventually one would know them

all and would simply list them, collecting specimens

only when in doubt.” Thus was evolved his well-known

listing method of plant survey.

In January 1936 Acocks joined the Division of

Plant Industry in Pretoria and worked under Dr

Fig. 1. — Mr J. P. H. Acocks.

J. W. Rowland. Within a few months he was sent to

Kimberley to make a survey of the Griqualand West

area (Fig. 2) in connection with the vermeerbos

problem and to establish a research station there. He

chose a suitable site for a research station (at Koop-

mansfontein), but was relieved of the task of establish-

ing it. Based at the McGregor Museum in Kimber-

ley, Acocks carried on with his survey of vermeerbos

and took the opportunity of applying his listing

method. He widened his horizons by surveys of

farms in other parts e.g. Carnarvon, Petrusburg and

Middelburg, Cape. He had a break of four months

when he was assigned to the Swedish plant collector,

Adolf Hafstrom; from 17th September-1 5th Novem-

ber 1938 they undertook a plant collecting expedition

“by devious routes from Cape Town to Victoria Falls

and back”, travelling in an enormous black Chrysler

sedan (Fig. 3). The joint collection amounted to 2 340

specimens.

In 1939 Acocks returned to Pretoria, where he

became involved in the botanical analysis of pastures,

at Rietvlei Pasture Research Station just outside

Pretoria and Leeuwkuil Pasture Research Station near

Vereeniging. He applied his estimated frequency list-

ing method and devised two new methods of pasture

analysis, namely the ring and rod method (Fig. 4)

and the basal cover calipers method, all of which were

significant enough to be fully described in the Com-

monwealth Agricultural Bureaux Bulletin No. 42

(1954). He studied the vegetation of the Kakamas

Veld Reserve in the northern Cape and accompanied

Dr Pole Evans and others on several trips to the

Transvaal Bushveld to take photographs illustrating

good and bad farming practices.

During 1941-1942 he was stationed at Towoomba

Pasture Research Station near Warmbaths. There he

made detailed surveys of the pastures on the station

involving four or five successive stages in 16 veld

types and sub-types (see Diagram 1, p. 2 in Veld

Types of South Africa, 1975.). Together with Mr

L. O. F. Irvine O/C of the Station, he travelled

widely in the northern Transvaal examining the effects

of various grazing systems on pastures. At this time

Acocks contributed to Professor J. M. Hector’s

“premature” attempt to compile a detailed vegetation

map of South Africa to replace the rather over-

simplified vegetation map produced by Pole Evans

in 1935. However, as Acocks pointed out (letter to

author, 1969-02-26), “we found we could distinguish

140 vegetation types, but none of us concerned with

it could draw the boundaries except in the few regions

we happened to know well, e.g. parts of the Trans-

vaal in the case of Hector, Irvine and others. Natal

in the case of Pentz, and Griqualand West in the case

of myself. This meant that somebody would have

to undertake a survey of most of the country, and he

(Hector) wanted the Department to second me to

him to do the job.”

In 1942 Acocks was based at Dohne Research

Station near Stutterheim in the eastern Cape, where

he carried out point quadrat analyses of pastures and

collected generally in the region.

1943 saw Acocks at Estcourt Pasture Research

Station in Natal. As on the previous stations, he was

occupied with botanical analyses of pastures. In

addition, he did much collecting and most important

of all, conceived the idea of the veld type. To quote

Acocks (letter to author, l.c.), “The idea of the Veld

240

Fig. 2. — Acocks’s camp under

Acacia erioloba at Witsand,

south-west of Postmasburg

in Griqualand West.

Fig. 3. — The two Swedish collec-

tors, Adolf Hafstrom & Erik

Wahl, who Acocks joined on

an expedition from Cape

Town to the Victoria Falls

in 1938. Photograph taken at

Buisvlei north-west of Pries-

ka.

241

Fig. 4. — The ring and rod method of pasture analysis being

used in the Kakamas Veld Reserve (1939).

Type was developed, while I was stationed at Est-

court, under the influence of Pentz with his idea that

the vegetation, being an indicator of the whole

ecological environment, is also an indicator of cor-

rect land use.” Acocks defined the veld type as “a

unit of vegetation whose range of variation is small

enough to permit the whole of it to have the same

farming potential.” In July 1945, while still at Est-

court, Acocks was transferred to the newly-created

Botanical Survey Section of the Division of Botany

and Plant Pathology. Largely at his own insistence,

he was given the huge task of preparing a compre-

hensive vegetation map of South Africa which,

according to Acocks (letter to Mr Richfield, l.c.),

“was what all the previous work appeared to be

calculated to lead up to.”

Acocks remained at Estcourt and surveyed the

eastern part of the country. In the survey Acocks

employed his relative abundance listing method and

used as his unit of vegetation, the veld type. In 1948

the western part had to be surveyed “in less detail

as the need for the map became more urgent”, and

consequently Acocks was transferred to Middelburg

in the Cape, where he was based at Grootfontein

College of Agriculture. In 1948 and 1949 Acocks

travelled all over the country as a member of the

Desert Invasion Committee.

Acocks’s mapping method proved so successful

that within eight years he had sufficient data to prepare

a vegetation map and accompanying text, which was

published as “Veld types of South Africa, Mem, bot.

Surv. S. Afr. No. 28, 192 pp. (1953). An accomplished

artist, Acocks did all the colour work of the original

map. During his survey Acocks examined 1533 stands

and collected over 25 000 specimens (i.e. numbers).

In “Veld types” Acocks recognized 70 veld types and

75 variations. These were described in varying detail,

the veld types in the eastern half of South Africa (as

explained earlier), receiving greater attention than

those in the western half. Considerable accent was

placed on the spread of Karoo as a result of incorrect

veld management. Included in the memoir was a series

of maps showing the vegetation of South Africa in

AD 1400, 1950, 2050? (showing the possible spread

of karoo eastwards almost as far as Vereeniging and

of desert almost to Bloemfontein) and a map showing

what the vegetation would be like if scientifically

managed. The memoir was updated in 1975, the

second edition containing 104 photographs illustrating

the veld types. “Veld types of South Africa” is a

classic work in the field of South African botany and

the information contained therein has been used by

scientists of many disciplines. The map has stood the

test of time and even after 26 years does not require

radical revision. The map has been described by some

as essentially a land use map and this is one of the

reasons for its great value.

Since 1953 Acocks attempted to equalize his treat-

ments of the veld types. He concentrated chiefly on

the karoo and karroid types and had examined an

additional 2467 stands involving some 177 000 species

records (Figs 5, 6 & 7). The revision of these veld

types will be published in the Institute’s journals.

During the course of his survey, Acocks collected

over 25 000 numbers of beautifully pressed specimens

(with the labels written in his copper-plate hand-

writing), which have enriched herbaria both here and

overseas. At the same time he added much to our

knowledge of the South African flora (Fig. 8): many

new species were discovered and new distribution

records made. Acocks had an exceptionally good

“eye” for plants and was able to recognize plant

species in all their various growth and ecological

forms. He often corrected herbarium identifications

and was frequently consulted by taxonomists with

problems concerning species limits.

Acocks amassed a tremendous amount of data on

plant distribution and frequency and this is all

meticulously indexed on cards. The Department of

Agricultural Technical Services considered his data

so valuable, that all his field note books were micro-

filmed several years ago. The possibility of com-

puterizing his data is being explored.

Apart from botanical work, Acocks played a

significant role in pasture science in South Africa.

Mention has already been made of his contributions

to pasture analysis. As early as 1945 Acocks came to

the conclusion that what the veld needed was non-

selective grazing with short, heavy gazing followed by

long rests. This, he felt, would help to restore the

climax vegetation in many parts of the country. After

the publication of “Veld types”, he was able to devote

more time to the subject and eventually in 1965 he

formulated his principle of non-selective grazing

(NSG) which formed the basis of the Acocks-Howell

method of grazing management. A document setting

out Acocks’s views in detail was sent to all pasture

research workers in South Africa in August 1965 and

comments were invited. Pasture scientists, on the

whole, opposed the method, some suggesting that

there was nothing new in the method while others

doubted that it would work in all types of grassland.

The extra cost of fencing and provision of \yatering

points was also mentioned. However, a considerable

number of farmers, particularly in the Karoo, sup-

ported the method, some even testing it and refining

it and achieving apparently successful results. The

method received considerable publicity and was even

242

243

Fig. 7. — Acocks listing plants in

Coastal Macchia at Danger

Point in 1962.

exported, modified and applied with success in

Zimbabwe. The well-known wagon-wheel method of

grazing management devised by the Zimbabwe

ecologist, Alan Savory, is derived from the Acocks-

Howell method. It is true to say that the Acocks-

Howell method, while to-day not generally accepted in

its original form, led to the acceptance of the feasibility

of multi-camp systems; it led to the reappraisal of

existing systems of grazing management and provided

a stimulus to the development of new systems.

Acocks retired from the Department of Agricultural

Technical Services in April 1976, but was re-employed

to complete his revision of “Veld types of South

Africa”.

Mention must be made of Acocks’s guidance to a

long line of ecologists at the Botanical Research

Institute and elsewhere. He gave freely of his know-

ledge and experience and this has benefited South

African ecology immensely.

Acocks did not receive the recognition he deserved,

partly because he was stationed in Middelburg in the

Cape, far from the main scientific centres and partly

because he was by nature, a quiet, solitary and indivi-

dualistic worker. He repeatedly refused promotion,

because he wished to carry on with research.

Acocks’s publications, though few (list appended)

represent extremely valuable contributions to South

African botany, especially “Veld types of South

Africa”. Clearly reflected in Acocks’s papers were his

chief attributes as a scientist, namely a capacity for

painstaking collection and synthesis of data and a

capacity for original thinking.

Honours came to Acocks in his twilight years. In

1975 he was awarded a gold medal by the Fertilizer

Society of South Africa for his outstanding contri-

butions to agriculture. In 1976 he was awarded three

medals: the medal of the Wildlife Society of Southern

Africa for notable contribtuions to conservation; the

244

South African Medal for Botany awarded by the

South African Association of Botanists and the

Senior Captain Scott Memorial Medal by the South

African Biological Society for outstanding scientific

research.

Acocks died at Middelburg in the Cape on 20th May

1979.

In conclusion, I would like to quote Dr R. A. Dyer,

one of his previous directors, who wrote: “In his

specialized field of plant geography, Acocks is with-

out peer amongst botanists both past and present”.

We salute a brilliant botanist and a good friend.

PUBLICATIONS

Acocks, J. P. H., 1952. The grassy veld types of South Africa.

In Veld Gold, Report of the Southern African Grass

Conference, September 1952. 22-29. Johannesburg:

National Veld Trust.

Acocks, J. P. H., 1953. Veld types of South Africa. Mem bot.

Surv. S. Afr. No. 28, 192 pp.

Acocks, J. P. H., 1966. Agriculture in relation to a changing

vegetation. S. Afr. J. Sci. 52: 101-108.

Acocks, J. P. H., 1964. Karoo vegetation in relation to the

development of deserts. In D. H. S. Davis, Ecological

studies in Southern Africa. 100-112. The Hague: W. Junk.

Acocks, J. P. H., 1966. Non-selective grazing as a means of

veld reclamation. Proc. Grassld Soc. Sth. Afr. 1 : 33-39.

Acocks, J. P. H., 1967. Vegetation of South Africa. Lantern

16: 50-56.

Acocks, J. P. H., 1971. The distribution of certain ecologically

important grasses in South Africa. Mitt. bot. StSamml.,

Munch. 10: 149-160.

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed.

Mem. bot. Surv. S. Afr. No. 40, 128 pp.

Acocks, J. P. H., 1976. Riverine vegetation of the semi-arid

and arid regions of South Africa. Jl S. Afr. biol. Soc. 17:

21-35.

Acocks, J. P. H., 1979. The flora that matched the fauna.

Bothalia 12: 673-709.

D. J. B. Killick

ADOLF JOSEPH WILHELM BAYER (1900-1978)

Born at Hermannsburg, Natal, on the 8th January

1900, Adolf Bayer, except for vacations and a study

tour in America, spent his life in the province of his

birth. His school days were at Durban Boy’s High and

his university career was completed at the then Natal

University College, Pietermaritzburg. He majored in

botany and obtained his D.Sc. degree under the

world renowned plant ecologist, J. W. Bews. He was

appointed lecturer under Bews in 1925 and in 1939

succeeded him in the professorial chair, which he

occupied with distinction until his retirement in 1967.

At this juncture, he was awarded the title of Emeritus

Professor of Botany.

It was not surprising that Adolf Bayer’s scientific

leaning was to plant ecology to which aspect of

botany he specialized as a background for his well-

illustrated university lectures. In addition to his pro-

fessorial duties, he served on the Senate and University

Council and performed the duties of Dean of the

Faculty of Science on several occasions. In 1970 he

was appointed acting Vice Principal of the University

of Natal in Pietermaritzburg and awarded the degree

of Doctor of Science Honoris Causa.

Apart from his strictly university duties, Adolf

Bayer (Fig. 9) took an active part in the affairs of a

number of outside societies, served on numerous com-

missions and many advisory bodies. He was Sectional

President of the South African Association for the

Advancement of Science (S2 A3) in 1942 when he

delivered an address on “Thornveld trees (Natal): a

note on plant adaptation”, and was President of the

Fig. 9. — Professor A. W. Bayer.

Society in 1971, when his address was entitled,

“Aspects of Natal’s botanical history.”

In recognition of his contributions to botanical

science he was elected a Fellow of the Royal Society

of South Africa and Volume 36 (1970) of the Journal

of South African Botany was dedicated to him in

eulogistic terms.

It has been said of Adolf Bayer that in spite of his

personal contributions to science he preferred to be

recognized for the many distinctions gained by his

students. The Botanical Research Instituted has over

the years boasted a strong contingent of staff who

qualified under his tutelage.

As a contemporary student, lifelong colleague in

botanical science and personal friend, I can testify to

his unassuming kindly nature, staunchness, dedication

and absolute integrity. In these times when old age

often presents social problems, one cannot mourn

the peaceful passing of one who has had a long and

fruitful life to the end. In a letter written in the week

of his death he wrote: “Daphne and I have had a

better year healthwise than any since 1973. I have

finished my contribution to the book on Katharine

Saunders’s flower paintings and the printer hopes to

send proofs during the month. Anyway I can relax

awhile. Yours Adolf.”

He died at his home in Kloof on Friday, 8th

December 1978.

R. A. Dyer

ROBERT HAROLD COMPTON (1886-1979)

Professor R. H. Compton, former Director of the

National Botanic Gardens, Kirstenbosch, died in

his 92nd year in Cape Town on 11 July 1979 after a

career devoted almost entirely to studying the South

African flora (Fig. 10).

245

Fig. 10. — Professor R. H. Compton.

He was born at Tewkesbury, Gloucestershire, on

6 August 1886 and was educated at Mill Hill School,

London, entering Gonville and Caius College, Cam-

bridge in 1905. In 1909 he graduated with a double

first class and distinction in botany, later taking the

M.A. degree. From 1911-13 he was Demonstrator

in Botany at Cambridge University and during 1914

participated in a field expedition to New Caledonia,

collecting about 2 500 specimens of the rich flora,

including new genera and a number of new species.

In 1915 he married Kathleen Askin Sealy of Sydney,

Australia, and they had two children, a daughter and

a son.

After some years of war service from 1915-18, he

came to South Africa in March 1919 to become

Director of the National Botanic Gardens at Kirsten-

bosch and Harold Pearson Professor of Botany

at the University of Cape Town, posts which he

held for the next 34 years. While at Cambridge

his main contributions to botanical literature dealt

with anatomy and morphology of Gymnosperms,

Pteridophytes and Angiosperm seedlings but, from

his arrival in South Africa, his interests turned to the

taxonomy of the South African flora. Most of his

papers were published in the Journal of South African

Botany which he initiated in 1935 and edited until his

retirement.

In 1921 he established the first subsidiary garden

of the National Botanic Gardens when he acquired

a site at Whitehill near Matjiesfontein to cater for

the rich succulent flora and, in 1931, he published a

paper on the flora of the Whitehill District in Trans.

Roy. Soc. S. Afr. 19: 269-329 in which two new

genera and many new species were described. In

1945 the Karoo Garden was moved to a more

convenient location near Worcester. He colla-

borated with the artist Elsie Garrett Rice in her

production of Wild Flowers of the Cape of Good

Hope, Cape Town, 1951, and in 1965 published a

history of the first 50 years of the National Botanic

Gardens under the title Kirstenbosch, Garden for a

Nation.

On his retirement in 1953 he settled in Swaziland

and, at the request of the Swaziland Government,

supported by the British Colonial Development and

Welfare Fund, he undertook a botanical survey of

the territory. Preliminary results were published as

“An annotated checklist of the flora of Swaziland”

in J1 S. Afr. Bot. Suppl. 6 (1966). After his return to

Cape Town in 1971, he enlarged this to a full scale

“Flora of Swaziland” which appeared as J1 S. Afr.

Bot. Suppl. 1 1 (1976) on his 90th birthday.

Many honours were conferred upon him. He was

elected F.R.S.S. Afr.; was President of the S. African

Association for the Advancement of Science in 1957

and received their medal and grant; an Honorary

fellow and medalist of the Royal Horticultural Society;

twice President of the South African Museums As-

sociation; President of the Mountain Club of South

Africa for eleven years and was awarded an honorary

D.Sc. by the University of Cape Town in 1968.

He was one of the most prolific collectors in South

Africa, his numbers exceeding 35 000, of which about

8 000 were collected in Swaziland. As may be expected,

these contained many novelties and two genera are

named after him: Comptonella Bak. f. from New

Caledonia and Comptonanthus B. Nord. from South

Africa. In addition, he is commemorated in the names

of about 20 species. When it was decided to remove

the Bolus Herbarium from Kirstenbosch to the

University of Cape Town in 1938, he vigorously set

about building up a new herbarium at Kirstenbosch

which, appropriately, is named the Compton Her-

barium in his honour.

L. E. Codd

247

REVIEW OF THE WORK OF THE BOTANICAL RESEARCH INSTITUTE, 1978/1979

CONTENTS

Introduction 253

Reports of sections 253

Staff list 260

Publications by the staff. 266

INTRODUCTION

Progress with botanical research has been good and

the rate of publication has been maintained at a high

level, a usually reliable indication of the state of

field. The steady increase of the volume of material

submitted for publication in the botanical journals of

the Institute has made it necessary to investigate

alternative methods of publication and funding to

offset the steep rise in costs.

The need for more research in the plant-taxonomic

and ecological fields is becoming increasingly evident

and pressures building up will require a positive

response in addition to the efforts already being made

by the Department. The real awareness of environ-

mental problems and their vital importance to man,

which already exists and is growing, will make a

large input into biological research, especially plants,

unavoidable in the near future. Nevertheless, the

contributions to both the fields mentioned, as is

evident from this report, have been considerable.

The control of weeds particularly from a pasture

point of view will require careful research planning

as well as increased efforts to control, in order to

deal with a potentially dangerous situation. In this

respect, nassella tussock, australian acacias as well

as jointed cactus and other members of this family,

are examples of what can happen if weeds are allowed

to get out of hand. Research on botanical aspects

of weeds have therefore been regarded as of high

priority. Amongst other activities, the compilation

of a national weed list, which is now complete, is

the first step towards monitoring weeds and establish-

ing the status of potentially harmful plants with a

view to control before the exponential stage of

distribution expansion is reached.

The use of electronic devices and computers in

research and information services is a fast expanding

activity. The vast and complicated data bank for

herbarium type information is now, after several

years of activity, nearing the productive stage. It will

provide a strong stimulus for types of research, which,

without this facility, would not have been possible,

such as the large-scale production of distribution

maps and check lists of a range of geographical areas.

The trend towards the integration of botanical

research into environmental planning and develop-

ment of the country has been maintained but, as said

earlier, research will have to be expanded to keep

pace with the demand for information.

REPORTS OF THE SECTIONS

HERBARIUM SERVICES SECTION

The five herbaria of the Institute continued to

identify plants and provide information for a wide

range of people including officers of the Institute,

various State and Provincial Departments, universities

and the public both in South Africa and neighbouring

countries.

National Herbarium, Pretoria ( PRE )

A total of 17 575 specimens was named and 766

visitors dealt with. Accessions to the herbarium

numbered 26 571. During the year 8 300 specimens

were sent out on loan to local and overseas researchers.

A few minor expeditions were undertaken to the

eastern Transvaal. The main expedition was organised

in conjunction with Prof. H.- D. Ihlenfeldt and Dr

H. Hartman of Hamburg. Originally it has been

intended to mount a major expedition to the southern

Namib Desert in the area between Liideritz and the

Orange River; as a result of the poor rains only

Messrs D. S. Hardy and S. Venter went from the

Institute.

The repacketing of the moss collection has been

completed and a start has been made on the liverworts.

When these have been completed, the lichen collection

will be tackled. The move to the new cryptogamic

herbarium in the basement will take place during

September 1979. It is hoped that the Herbarium will

obtain the services of a lichenologist.

A new herbarium procedure was introduced with

the establishment of a service room in the basement.

This has centralized such activities as the receipt and

dispatch of specimens for identification, loans,

exchanges and the preparation of herbarium labels.

On December 6th the Director and Mr E. G. H.

Oliver attended the official opening of the Venda

Herbarium at Tate Vondo in the eastern Soutpansberg

in the Venda Homeland. The opening was performed

by the Venda Minister of Agriculture and Forestry

and attended by the Chief Minister and remainder of

the Cabinet.

Among the numerous visitors who came to consult

the collections and staff were the following: Prof.

H.- D. Ihlenfeldt and Dr H. Hartman (Hamburg),

Dr Juliet Prior (London: Swaziland Archeological

Association), Prof. D. and Dr U. MiiLer-Doblies

(Berlin), Mr R. B. Drummond and Mr L. C. Leach

(Salisbury), Dr O. J. Hansen (UNDP, Botswana)

and Dr H. van Gils (Botswana).

Wing A: Mr P. J. Vorster left the Institute in

February to take up a post in the University of

Stellenbosch. He completed his work on Mariscus

for which he obtained his doctorate. Mrs E. van

Hoepen took over control of this Wing.

Miss Smook continues with work on grasses and

will assist Dr E. G. Gibbs-Russell with the treatment

of grasses for the Flora.

248

Miss C. Reid began in February and is concentrating

on petaloid monocots.

Dr R. E. Magill produced a check list of the mosses

(591 spp.) and liverworts (316 spp.) for southern

Africa and is continuing with revisions for the Flora.

His assistant Mr J. van Rooy is dealing with most

identifications and is working on Bryum.

Wing B: Mr T. H. Arnold left the Herbarium in

November on transfer to the Economic Botany

Section. He completed the curating and updating of

the Cyperaceae resulting from his investigations at

Kew.

Mr G. Germishuizen resumed charge of the Wing

and has now taken up Polygonaceae for his research

project. He has also been working on the botanical

text for a book on medicinal and edible plants.

Mr D. A. Davies joined the staff" in January and has

been doing general naming and curating.

Wing C: Miss E. Retief has begun with a study of

the Campanulaceae of Southern Africa. She delivered

a paper on seed collections at the Annual Congress

of the South African Association of Botanists in

Stellenbosch.

Mr P. P. J. Herman joined the staff in March. He

had already begun doing an M.Sc. degree on the

anatomy of Pavetta.

Mrs M. Crosby moved to this Wing in January to

assist with general identifications. Mrs J. L. M.

Grobler also moved back to this Wing as a result of

Miss M. Evans having left in February. Mr S. Venter

left in February to resume his studies.

Mr C. Hildyard has decided to take up botany as a

career and is studying at the University of Pretoria.

He is employed on a part-time basis.

Wing D: Miss W. G. Welman continues to be

responsible for the Wing. She has recently taken over

as South African extractor for Excerpta Botanica.

Mrs S. Smithies gained her M.Sc. degree with

distinction at the Witwatersrand University with a

thesis entitled ‘Studies in the Middle Ecca (Lower

Permian) Flora from Hammanskraal, Transvaal, with

emphasis on the Glossopterid fructification Ottokaria

Zeiller’.

Mr G. Goosen joined the staff in February and will

be going to university next year to do a B.Sc. degree.

Service Room: Mrs I. Ebersohn began in January

and Mrs G. L. Radmacher in July 1979. The work in

the Service Room is controlled by Mrs E. van Hoepen.

Natal Herbarium, Durban ( NH )

A total of 1 459 specimens was named and 414

visitors dealt with. Accessions to the herbarium

numbered 3 167.

Mr P. C. V. du Toit is continuing with his work on

Pentaschistis (Poaceae). Mrs H. M. A. du Toit left

the staff last year and her place was taken in February

by Mrs B. J. Pienaar who worked for the first six

months in the National Herbarium in Pretoria gaining

experience. Miss A. Wright took over as technical

assistant from Miss A. M. King who left in October.

Prof. K. Schlosser of Kiel, West Germany, again

visited the Herbarium in connection with his work on

trees and associated folklore.

Albany Museum Herbarium, Grahamstown (GRA)

A total of 2 528 specimens was named and 598

visitors dealt with. Accessions to the herbarium

numbered 1 181.

Mrs E. Brink continues as Curator of the Her-

barium. Miss G. V. Britten was transferred to the

half-day post after serving for 57 years on the full-time

staff. Colonel R. A. Bayliss left at the end of Decem-

ber.

Government Herbarium, Stellenbosch (STE)

The number of specimens named totalled 4 853

with 333 visitors requiring information. Accessions

to the herbarium numbered 1 696.

Mrs M. F. Rand (nee Thompson) is continuing with

her work on Hypoxidaceae and completed a revision

of the small genus Pauridia. She is now dealing with

Spiloxene. She delivered a paper on generic differences

in the family at the S.A.A.B. Congress in Stellenbosch.

Miss L. Hugo completed her work on the section

Campylia in Pelargonium and will submit it for an

M.Sc. degree at the University of Stellenbosch at

the end of 1978.

Mrs A. M. Pietersen left the staff at the beginning

of 1979 and was succeeded by Miss G. Garwood.

Several local collecting trips were undertaken, in

particular to the site of the large Theewaterskloof

Dam to record the flora before final inundation.

S.W.A. Herbarium, Windhoek (WIND)

A total of 6 017 specimens was named and 366

visitors dealt with. Accessions of 2 386 specimens for

the year brought the holdings up to 39 167.

Mr M. A. N. Miiller continues to run the herbarium

and work on a revision of Eriocephalus. His duties

include lectures at the Neudamm Agricultural College

and the Tsumis Agricultural College. He is also

responsible for the issue of phytosanitary certificates

and export permits for local products among which

are the dried roots of Harpagophytum procumbens

(Pedaliaceae).

Mr H. J. W. Giess continues his work on the local

flora in conjunction with the Botanische Staats-

summlung, Miinchen. He was recently awarded the

Certificate of Merit by the South African Association

of Botanists for his outstanding and valuable con-

tributions to botany in South West Africa/Namibia.

Mrs J. L. van Aswegen and Mrs H. E. J. Stoffberg

are still on the staff, the latter now in the newly

created full-day post.

FLORA RESEARCH SECTION

The main task of this section is the preparation of

floras of the subcontinent dealing with flowering

plants, and cryptogams as well as fossils.

Flora of Southern Africa

The third introductory volume to this series was

published: a Bibliography to South African Botany

(up to 1951) by Mr A. A. Bullock, formerly of Kew.

This work of 194 pages contains some 10 000

references and consists of two parts, an author index

and an index to plant groups. It comprises mainly

taxonomic literature but deals also with ecology and

other fields of botany.

Two volumes are in preparation for the series on

Cryptogams.

Vol. 1 1 : A check list of the 907 species of mosses

known in the region was published by Dr R. E.

Magill. These species belong to 282 genera and 87

families. First drafts of the treatment for the

Flora were completed for 71 species, and a total

of 86 species was figured in detailed pencil drawings

by Mrs Rita Weber.

249

Vol. 12: In preparation for this volume on ferns a

check list of the 241 known species and varieties

was produced by Prof. E. A. Schelpe of the Bolus

Herbarium and circulated. Most illustrations for

this work have been completed by Mrs R. C.

Holcroft and others and the greatest portion of

the research work has already been done.

In the series on Flowering Plants the following

volumes or parts thereof are at an advanced stage

of preparation:

Vol. 3: A revision of the 65 species of Mariscus

(Cyperaceae) and related genera by Dr P. J. Vorster

was accepted as a D.Sc. thesis. In the complex

genus Ficinia (Cyperaceae) more than 30 of the

estimated 70 species have been researched in depth

by Mr T. H. Arnold. A study of the pattern of

silica deposition in the epidermis of seeds is pro-

viding an important aid in this investigation.

Vol. 4, part 2: Work on Xyridaceae, Eriocaulaceae,

Pontederiaceae and Juncaceae, totalling 42 species,

was completed by Mrs A. A. Mauve. Commeliaceae,

the family still outstanding for this part, is being

dealt with jointly by Prof. J. P. M. Brenan, Dr R.

Faden and Mrs A. A. Mauve.

Vol. 5: In the Liliaceae Mrs A. A. Mauve progressed

well with work on genera related to Urginea and

she completed a publication on the genus Sypharissa.

Vol. 10, part 1 : This part on Loranthaceae and

Viscaceae by Prof. D. Wiens of the University of

Utah and Dr H. R. Tolken, formerly of the BRI,

now at Adelaide, South Australia, has gone to

press and should be available before the end of

1979.

Vol. 1 1 : Revisions of Lampranthus and Gibbaeum

(Mesembryanthemaceae) have been submitted for

publication in the Contributions of the Bolus

Herbarium by Dr H. F. Glen, who is now studying

the sub-tribe Ruschiinae.

Vol. 14: In this volume on Crassulaceae by Dr

H. R. Tolken revisions of the following genera

were completed: Cotyledon, Tylecodon, Kalanchoe

and Adromischus. The revision of the genus

Crassula, which was published in the Contributions

of the Bolus Herbarium last year, is being converted

to Flora format and the entire volume should be

ready for the printers early in 1980.

Vol. 21 : A treatment of the family Tiliaceae written

many years ago by Prof. H. Wild is being prepared

for publication. The 93 species of subgenus Her-

mannia of Hermannia written up by Dr I. C.

Verdoorn were submitted for publication in

Bothalia.

Vol. 25: A revision of Griesebachia (Ericaceae) by

Mr E. G. H. Oliver will be going to press soon and

work on Thamnus and Platycalyx of the same

family was completed.

Vol. 27, part 4: This part which is being prepared by

Dr R. A. Dyer deals with a total of 130 species

belonging to Brachystelma, Ceropegia and Rio-

creuxia (all Asclepiadaceae). It has gone to press

and should appear before the end of 1979.

Vol. 28: Work on Lamiaceae, a family comprising

about 250 species, should be completed by Dr L. E.

Codd before the end of 1981.

Palaeoflora of Southern Africa

A revision of Dicroidium, a genus of seed-ferns, is

being prepared for publication in a volume on the

Upper Triassic Molteno Formation by Drs J. M. and

H. M. Anderson. Complete copy for this volume is

being typeset on a compositor in the Institute. Large

composite photographic plates are compiled and will

be reduced to the eventual A4 size. A world-wide

review of genera of megaplants of the Permo-Triassic

is being compiled by the same two authors with the

co-operation of various authorities from different

countries. It is planned to publish this work as an

introductory volume to the series.

Botanical collectors in Southern Africa

For this work, which is being written by Miss M. D.

Gunn and Dr L. E. Codd, the encyclopaedia dealing

with plant collectors in alphabetical sequence was

completed except for a few late entries which must still

be included. An historical introduction was written

to provide a chronological account of early collecting

activities and of literature dealing with southern

African plants, with emphasis on illustrations, up

to the time of Linnaeus.

Register of plant taxonomic projects

A new edition of this register, listing some 350

current projects on African plants, is in the last

stages of preparation and will again be distributed

world-wide in form of microfiche cards.

Southern African Plants

Material for a brochure on 20 of the most important

water weeds in the region was compiled. For each

species two to three colour photos and a distribution

map are provided. The text comprises a short descrip-

tion, a comparison to related species, a note on

distribution and paragraphs on ecology, importance,

derivation of the scientific name and literature.

Pretoria Flora

With the completion of a further 298 line drawings

only 40 species remain to be illustrated. Special

attention was given to the large and difficult families

Poaceae and Fabaceae. The 235 species of grasses

were figured by Mrs Hester Wouda-du Toit while

Mr P. C. V. du Toit wrote the text. Dr H. R. Tolken

compiled the text for Fabaceae while Mrs R. C.

Holcroft did most of the illustrations.

Ceropegia and related genera

A fully illustrated account of the genera Ceropegia,

Brachystelma and Riocreuxia by Dr R. A. Dyer is

in the final stages of preparation. It will provide

information additional to that given in Volume 27,4

of the Flora of Southern Africa and will cater for the

world-wide interest in the group by amateurs.

Liaison Officer, Kew

The Officer, Mr C. H. Stirton, was active along a

broad front, especially in the fields of weed research,

the taxonomy of Fabaceae and pollination biology.

He participated in several symposia and scientific

gatherings in the United Kingdom and on the Con-

tinent. Numerous herbaria were visited in the course

of his research on the weed genera Lantana and

Rubus as well as the genera Eriosema, Psoralea and

related groups. Problems from different fields of

botany were researched for scientists and institutes in

South Africa and abroad and information was

exchanged with botanists around the globe.

Plant anatomy

Mr R. P. Ellis continued his work on the anatomy

of the Kranz syndrome in the southern African

grasses. Very good correlations were found between

the distribution of ‘malate’ and ‘aspartate type C4

250

grasses in South West Africa and the rainfall. Malate

formers were shown to increase with increasing pre-

cipitation whereas aspartate formers decreased and

were most common in the Namib and pre-Namib

areas. This is the first time that an ecological relation-

ship has been demonstrated for these sub-types of

the C4 photosynthetic pathway. This has only been

possible because of the extensive collection of grass

leaf blade anatomical preparations assembled during

the course of this project.

A further two field trips were undertaken to the

Drakensberg mountains in Natal to collect and study

members of the genus Merxmuellera. Anatomical

results clearly demonstrate that, in both M. disticha

and M. stricta, three distinct anatomical “forms”

occur in the Drakensberg mountains in addition to

the typical “form” found further south in the Cape

Province. These “forms” all have characteristic leaf

anatomy, display morphological similarities as well

as having similar habitat requirements, and appear

to warrant taxonomic status.

The most serious infestations of the noxious weed,

Opuntia aurantiaca, occur in the eastern Cape area.

Dr T. B. Vorster and Mr T. H. Arnold visited this

area and determined, by cytogentical studies, that in

South Africa this problem plant originating from

Argentina, is represented by only one biotype. In

Argentina O. aurantiaca forms a complex consisting

of hundreds of biotypes, but Dr Vorster is of the

opinion that the plant which has invaded the eastern

Cape is the vegetative progeny of a single plant, or

a few genetically identical plants.

BOTANICAL SURVEY SECTION

The function of the Botanical Survey Section is to

study the vegetation of South Africa and its ecological

relationships. This work covers three main fields of

activity: (1) The identification, description and

mapping of various vegetation classes; (2) the study

of the ecological relationships between different types

of vegetation — mutually and with the environment —

and also of the various processes and mechanisms that

govern the behaviour of plant communities; (3) the

development of various methods and techniques

required for ecological studies of vegetation. In this

work there is close co-operation with the Data

Processing Section, to which ecosystems research has

been transferred.

Veld types of South Africa

At the time of Mr J. P. H. Acocks’s death in May

1979, the revision of the veld types of the western half

of the country was approaching completion. He had

completed writing up all the 25 Karoo and False

Karoo types together with the west- and south-coastal

types, and part of the Kalahari Thornveld. These

amount to 32 veld types in all, subdivided into 89

variations. The revision excludes the Mountain

Fynbos and the False Fynbos. Part of the revised

treatment of Kalahari Thornveld remains to be done.

Western Transvaal bushveld survey

The final report on this project was presented as

a Ph.D. thesis by Dr F. van der Meulen and published

in the series Dissertationes Botanicae as Band 49

(1979), the title being “Plant sociology of the western

Transvaal Bushveld, South Africa: a syntaxonomic

and synecological study”.

Zululand coastal dune survey

The impact of mineral prospecting on dune vegeta-

tion in Reserve 10 (Zululand) was uneven, the sea-

ward slope of the dune being the most sensitive area.

Regeneration of vegetation on the sampling lines

was good. A preliminary evaluation of conservation

priorities by Dr P. J. Weisser has shown most of the

area to be of secondary character and low in con-

servation priority.

Wilderness Lakes study

The macrophyte encroachment problem of the

Wilderness Lakes in the Cape Midlands was studied

by Dr P. J. Weisser at the request of the Lake Areas

Development Board (George). An increase of macro-

phytes such as Potamogeton pectinatus, Phragmites

australis and Typha latifolia subsp. capensis in some

areas was found and mechanical control suggested.

Vegetation survey of the Cape of Good Hope Nature

Reserve

For the assessment of the rate of infestation of

pest-plants (or plant invaders) during the study period

starting in 1966, Mr H. C. Taylor carried out the re-

enumeration of pest plant density on 29 sample plots

during the year. The geographic position of 10 of

these permanent samples, previously lost, had to be

re-established by compass and distance-measure.

A primary survey of Rooiberg Mountain Catchment

Reserve near Ladismith, Cape

The completion of the quantitative floristic survey

by Mr H. C. Taylor confirmed the expected high

degree of correlation between the physiognomic and

phytosociological classifications reported on last year.

The results of the reconnaissance survey, including

management recommendations, were embodied in a

47-page report, with five appendices, 28 photographs,

two maps and a phytocoenon table submitted to the

Secretary for Forestry, as requested.

A classification based on structural-functional charac-

teristics of the vegetation

In this, the first year of the project, Mr B. M.

Campbell spent a considerable time on methodology.

He completed a literature survey, tested ideas for the

classification in the field, compiled a computer-

package for the manipulation and analysis of struc-

tural-functional data, and wrote a paper on metho-

dology. In the latter part of the year, he began his

sampling programme. A third of the necessary data

has now been collected.

Distribution and autecological aspects of the Hakea

pest plants

Mr S. R. Fugler has drawn and analysed distribu-

tion maps for the three South African pest Hakea spp.

and improvements in the present H. sericea mechanical

control programme have been suggested. Evidence

indicates that all infestations of H. suaveolens and H.

gibbosa can be eradicated in South Africa except the

H. gibbosa infestation in the Kleinriviersberge and

adjoining mountains. A management programme for

these mountains has been proposed. Methods for esti-

mating the density and fruit load of H. sericea for a

given infestation have been designed and the pheno-

logy of the pest Hakea species has been studied.

Fynbos biome project

In October the Task Group searched for and found

a suitable site to replace the proposed west coast

study site which had been burnt.

Mr H. C. Taylor prepared a chapter on “The

Phytogeography of Fynbos” for a volume to be

published by the CS1R synthesizing pesent knowledge

on the Fynbos Biome.

251

Semi-detailed survey of the vegetation of the western

and southern Cape coastal lowlands

Mr C. Boucher’s sampling of the natural vegetation

along transects through the first priority western

coastal foreland progressed as follows: 1. Langebaan

Lagoon to Saron — 100 km long; 150 floristic samples;

preliminary sampling complete. 2. Buck Bay to

Bainskloof — 70 km long; 100 floristic and 100 physi-

ognomic samples; 22 km sampled. 3. Table Bay to

Stellenbosch — 35 km long; 26 samples; 10 km

sampled. 4. Orange River mouth to Table Bay littoral

zone — 82 samples; sampling complete. Existing com-

puter programmes were modified and new pro-

grammes were written to analyse these data. Approxi-

mately half of the data has been coded for analysis.

Littoral vegetation of the southern Cape

To provide quantitative data for a contribution to

a book on dry coastal ecosystems (in Elsevier’s Series

Ecosystems of the World), a phytosociological study

of littoral vegetation of the southern and western

Cape is being conducted. During the year floristic data

from 163 releves were collected along the southern

Cape coast from Cape St Francis to Cape Point.

Provisional indications are that the sand-dune flora

is remarkably uniform but there is considerable

variation in the flora of the rocky coasts.

Orothamnus project

Mr C. Boucher visited the known populations of

the marsh rose ( Orothamnus zeyheri) in the Kogelberg

State Forest during December 1978. Population size

and development of selected individuals are monitored

annually. Preliminary results indicate that it takes at

least eight years for half the individuals in a popula-

tion to reach flowering maturity. A total of 1 180

individuals was recorded in the area.

Aquatic ecology

The Braun-Blanquet classification of the “water-

plant” communities of Natal together with a paper

entitled: “A preliminary classification of the water-

plant communities of Natal ”, intended for publication,

have been completed by Mr C. F. Musil. Water-plant

communities have been set out in a synoptic table and

each described in detail. The communities follow a

broad salinity and pH gradient and have been divided

into 6 major groups. These are (1) marine, (2) estuarine,

(3) brackish water, (4) moderately fresh to slightly

brackish water, (5) fresh-water, and (6) communities

found in fast-flowing fresh-water.

Mr C. F. Musil has completed more than a year’s

field measurements, at fortnightly intervals, of growth

rates and chemical composition of water-hyacinth

plants, together with chemical and physical para-

meters of the water and environment. Present syn-

thesis of data has revealed an apparent linear rela-

tionship between diffuse radiant flux and growth rate

of water-hyacinth plants. This suggests that light is

the primary factor limiting water-hyacinth growth,

where previously the effect of light was considered to

be minimal. Mr Musil has practically completed a

detailed report, critically evaluating the literature,

methods and culture techniques employed, that will

contribute to the “methods” section of the final report.

NAKOR National Conservation Plan

Dr J. C. Scheepers and Miss B. K. Drews continue

to co-ordinate the NAKOR National Conservation

Plan for the Department of Environmental Planning

and Energy, in co-operation with an officer of that

Department. A staff change in the latter Department

has interrupted continuity, but progress continues.

Approval has been obtained and funds earmarked

for the setting up of a data bank for efficient storage,

retrieval and processing of the rapidly accumulating

data, using the computer of the Department of Agri-

cultural Technical Services.

DATA PROCESSING AND ECOSYSTEM STUDIES

SECTION

The mandate of this relatively new section is the

provision of data processing facilities for research

purposes to the rest of the Institute as well as under-

taking plant ecological research at the ecosystem

(function) level. The largest data processing task

under development, the computerization of the

National Herbarium, is almost complete. Our con-

tribution to the Savanna Ecosystem Project at Nyls-

vley, determination of biomass relations and sea-

sonal biomass change in dominant tree and shrub

species has made good progress and a report on

biomass relations has been published. The Savanna

Ecosystem Project data bank is administered by this

Section, under the control of Dr J. W. Morris. He is

also responsible for co-ordination of modelling acti-

vities as well as research in the Decomposer Com-

ponent of the Project.

The National Herbarium Data Bank has changed

to the status of a production system, although the

last few tests of programmes forming part of the

system are still being done. An intensive process of

correcting errors in the data is under way and will be

completed soon by Dr H. F. Glen. The production of

realistic maps showing the distribution of specimens

selected from the data bank is a facility which is in

the process of being added to the system.

In a project being undertaken for the Savanna

Ecosystem Project by Dr. M. C. Rutherford, it has

been found that a considerable proportion of photo

synthate is “lost” to the formation of wood material

in savanna vegetation but that water relations are

important in interpreting radial stem growth measure-

ments. In another project, he has found that in-

creased fire damage to Ochna pulchra canopies results

in a greater proportion of basal coppice material.

Tree height has been found to influence the relative

wood-leaf composition of basal coppice material and

to influence several other plant properties following

upon fire.

ECONOMIC BOTANY SECTION

The work programme of the section, under Mr

M. J. Wells consists of: plant utilization research

(including ethno- and palaeo-ethno botany), weed

research, plant geography, the conservation of plant

resources, and the provision of an information service

spanning these fields. In addition the section hosts

work on the new palaeo-flora of southern Africa.

Until recently the accent of the section’s work was

weed research, but this has diminished as a result of:

the temporary transfer to Kew of the leader of the

weed research team (Mr C. H. Stirton); the conclusion

of Mr Wells’s investigation of nassella tussock — for

which control measures have now been instituted;

the decision of the Department to establish a Weed

Research Centre, which will in future co-ordinate work

(including botanical work) on weeds.

At present the section is completing its work on a

number of weed projects that will be basic to the work

of the new Weed Research Centre (e.g. the provision

of a national weed list), and is preparing to register

252

new projects in the field of plant utilization, and the

conservation of plant resources. These projects will

fall under the newly appointed team leader, Mr

T. H. Arnold.

Plant utilization

Mr Arnold has commenced a literature study on

the utilization of South African plants, as a pre-

liminary to registration of an umbrella project. The

first component of this is a project on barrier plants.

Other projects will include research on the origins

of indigenous crop-plants. Mrs B. van Gass planted

out, and harvested seed of all entries in our Citrullus

(water-melon) collection, under the direction of Dr

B. de Winter. Several new entries, including one from

the Seychelles, were added to the collection which

now includes an interesting range of wild species and

eco-types.

Miss C. A. Liengme returned after completing her

B.Sc. Hons at Natal University, Pietermaritzburg.

She finalized her report on the ethnobotany of the

Tsonga of Gazankulu, and completed a literature

study of the ethnobotany of tribal peoples in South

Africa. This will form the basis for registration of an

umbrella project on ethnobotany, and will provide

the background for more detailed research on Tsonga

ethno-botany, which she is about to begin.

Conservation of plant resources

The first stage of a project to conserve germ plasm

of primitive crop cultivars and of rare and endangered

indigenous spiecies is under way, and a survey of

over 200 literature references has been completed by

the project leader, Mr Arnold. Work on this and on

the related, plant utilization project will be carried

out by two new appointees: Miss K. J. Duggan, from

the University of Natal in Durban; and Miss L.

Henderson, from the University of the Witwatersrand.

Weed research

A first national weed list was drawn up, mainly

from the literature, by Mr G. B. Harding, prior to

his leaving for Natal University (PMB) where he

has been seconded for his honours course. The list

has since been added to by other stalf members,

notably Mr A. A. Balsinhas, and now includes the

names of over 700 plants of which about 30% are

indigenous.

This list is to provide a basis for sorting w ;eds into

catagories, distinguishing weed-corn* le ^s, making

weed status evaluations, and can ultimately provide

the basis for legislation.

Material of all the exotics on the list, and the species

of world-wide distribution, has been sent to Mr C. H.

Stirton at Kew, for checking of identification, so that

the weed status of these species can be evaluated on

the basis of available literature.

Pilot studies have also been completed on two

methods of carrying out weed surveys to provide

qualitative and quantitative information for future

weed status and weed-complex evaluations. In the

first of these we experimented with sampling agro-

nomic weeds. Messrs Harding and Balsinhas and

Mrs van Gass carried out this experiment: collecting

soil samples in cultivated fields in winter, germinating

weeds from the soil samples under nursery conditions,

and then comparing the species found in the field with

those germinated in the nursery. This method needs

considerable refinement, but the results obtained

suggest that it can be used to enable us to carry out

a country-wide survey with the help of local collec-

tors, thus reducing survey costs and the use of expert

manpower.

The second of these projects, carried out by Misses

Henderson and Dugfetlii, was a survey of woody,

exotic invaders in the Pretoria-Witwatersrand-Ver-

eeniging area of the Transvaal highveld. Approxi-

mately 30 woody species were found to be encroaching

in streambank and roadside habitats and in the open

veld. Melia azedarach (syringa), Acacia dealbata

(silver wattle) and Acacia mearnsii (black wattle) are

the commonest woody, roadside invaders, whilst

Salix babylonica (weeping willow) and Populus alba

(white poplar) are the commonest woody, streambank

invaders. Arundo donax (Spanish Reed) is common

in both habitats, and is replacing Phragmites in some

places.

Mrs W. G. Gaum continued her work on a cyto-

genetic evaluation of the Lantana camara complex,

under the diection of Dr T. B. Vorster. Chromosome

counts have been completed for all the variants so far

found in South Africa. The accent is now on the search

for embryo-sac abnormalities. The investigation is

still incomplete, but at this stage there are no grounds

for considering any of the variants to be safe to

cultivate.

A quick survey of local nurseries, carried out by

Mrs D. M. C. Fourie and Mr Balsinhas showed that

most of them are still selling Lantana camara although

it is a declared-noxious weed.

Mr Stirton will complete the bio-taxonomic aspects

of the Lantana study when he returns from Kew in

1980.

A highlight of the year was the appearance of the

book ‘Plant invaders: beautiful but dangerous’ com-

piled and edited by Mr Stirton for the Cape Depart-

ment of Nature and Environmental Conservation. It

enables the layman to identify 26 of the worst plant

invaders of the Cape Province, and provides informa-

tion on how to counter the threat that they pose.

A short study of Prosopis (mesquite) in the northern

Cape, carried out by Mr Harding, showed that, of the

three species recorded as having been introduced to

South Africa, two are invasive, namely P. glandulosa

(honey mesquite) and P. veluti (velvet mesquite). The

third species, P. pubescens (screw bean), does not

appear to be causing trouble.

The long-awaited book on poisonous plants by

Mr J. H. Vahrmeyer is still at the printers, awaiting

publication. The text has been translated from Afri-

kaans into English by Mrs E. Brink, of the research

unit at Grahamstown, and Mr Wells has edited both

versions.

Miss S. M. Bulley, a new staff member from Natal

University in Duruan, has started a two year aut-

ecological study of nassella tussock ( Stipa trichotoma).

She is stationed at Cape Town University, where she

is under the direction of Dr A. V. Hall. The object of

this study is to find ways of modifying and improving

our control strategies for this plant.

Plant geography

Dr J. M. Anderson continued his survey of the

distribution of woody plants in the Transvaal. Two

hundred and eighty field listings of woody plants were

made. Listings have now been made in nearly 1 700

of the ±1 000 1 /16th ° squares that occur in the survey

area. The first objective, a 10% sampling cover, has

been achieved for nearly the whole area, whilst the

253

ultimate objective, a sampling cover of 25% has been

achieved in over half the area. The listings are being

used to plot the distributions of all of the ±900

species of indigenous woody plants that occur in the

Transvaal.

The palaeoflora of southern Africa

It has been decided to describe the fossil plants of

southern Africa in a ‘flora’ as a companion series to

the extant flora series. This work is in the hands of

the husband and wife team Drs J. M. and H. M.

Anderson. Good progress has been made on an

introductory volume and one on the Molteno Forma-

tion. Already some 5 000 fossil plant specimens have

been collected, the photographic plates illustrating

about 1 000 specimens have been prepared, and the

text, tables and maps are nearing completion. This

project is of world-wide interest since, in order to

cover the South African fossil flora adequately, the

whole of Gondwanaland will have to be treated.

The information service

Mrs Fourie, scientific information officer for the

Institute, handled 283 requests for information about

economic plants, and their utilization or control.

These included about 440 identifications. Particular

interest was shown in the newly declared noxious

weeds: Sesbania punicea and Pereskia spp.; and in the

utilization of indigenous Euphorbia spp. and the

Mexican rubber bush (or guayule) Parthenium

argentatum.

Colour-slide collection

Nine hundred and fifty labelled colour-slides of

plants were added to the Institutes’ collection which

is curated by Mr Balsinhas.

GARDEN SECTION

Mr J. Erens continued in charge of the garden

whilst Mr D. S. Hardy was in charge of the nursery.

There were 1 127 accessions to the garden during the

year including Mr Hardy’s collection of 500 succulents

from South West Africa. Mrs B. C. de Wet and Mrs

K. P. Clarke, continued with the mammoth task of

labelling and record keeping.

In the garden the most striking development was

in the savanna biome area, where Mr H. J. de Villiers

and Mr T. A. Ankiewicz landscaped the area adjacent

to the Silverton road, building rock-covered koppies

that will soon be the home of wild figs and other

lowveld species. All the beds in this biome area have

now been demarcated, and most of them have been

sown to Eragrostis curvula, which provides an attrac-

tive, soil-binding cover. The areas between the beds

have been planted with lawn grass, and 120 savanna

trees were planted out. The water reticulation system

of the savanna biome area is now complete, except

for the ‘ Aloe koppie’.

Progress in the coast forest biome area, where

Mr Ankiewicz is in charge, has also been rapid. Most

of the excavation of soil, and landscaping of an area

for swamp forest development has been done. The

reticulation system has been completed and, as in

the savanna biome area, paths to the plantings serve

a dual function by also providing access to the water

pipes. One hundred and fifty trees and shrubs were

added to the coast forest plantings.

The fynbos biome has not fared so well — replant-

ings of Proteaceae having suffered during the hot,

dry summer of 1978/9. The summer display, consisting

of about 25 000 Pelargonium, Gazania, Arctotis,

Dimorphotheca, Diascia, Felicia and mesem plants,

was effective, but also suffered from the drought,

whereas the earlier display of spring annuals was

outstanding.

Floral decorations were provided for about 80

special occasions during the year. We were also glad

to have contributed to the South African gold medal

exhibit at the Chelsea Flower Show.

254

BOTANICAL RESEARCH INSTITUTE

Scientific, Technical and Administrative Staff

(30th June 1979)

Director

B. de Winter, M.Sc., D.Sc. (Taxonomy of

Poaceae, especially Eragrostis, and of

Hermannia; plant geography)

Deputy Director

D. J. B. Killick, M.Sc., Ph.D., F.L.S.

(General taxonomy and mountain

ecology)

Assistant Director

D. Edwards, M.Sc., Ph.D. (Ecological

methodology; aquatic plants, remote

sensing and vegetation structure and

physiognomy)

ADMINISTRATION

Administrative Officer

Senior Administrative As-

sistant (Personnel)

Assistant Accountant

Senior Administrative As-

sistant

Administrative Assistants

Senior Clerical Assistant

Clerical Assistants

Receptionist

Typists

Driver

D. F. M. Venter

Mrs G. E. Hussem, B.A.

A. Smith

J. Conradie

Mrs J. Rautenbach

C. J. Smith (temporary)

Mrs T. Creffield (Regis-

try)

Mrs C. A. Bester

Mrs I. J. Joubert* (Regis-

try)

Mrs M. E. M. Venter

Mrs N. Miller*

Mrs S. M. Thiart*

Mrs J. Gerke*

H. L. Reyneke

HERBARIUM SERVICES SECTION

Officer in Charge E. G. H. Oliver, M.Sc.

National Herbarium, Pretoria (PRE)

Chief Professional Officers

Technical Assistant

E. G. H. Oliver, M.Sc.

(Curator; Ericaceae)

Mrs E. van Hoepen,

M.Sc. (Assistant Cura-

tor; supervision of

identifications and en-

quiries)

Mrs E. D’Alebout

Wing A (Cryptogams — Monocotyledons)

Chief Professional Officer

Senior Professional Officer

Professional Officer

Technician

Technical Assistants

Mrs E. van Hoepen

R. E. Magill, M.S., Ph.D.

(Musci)

Miss C. Reid, B.Sc. Hons

Miss L. Smook, B.Sc.

(Poaceae)

Mrs P. W. van der Helde

Mrs L. R. Filter*

Mrs A. M. Fourie*

J. van Rooy (Musci)

Wing B (Piperaceae — Oxalidaceae)

Senior Professional Officer G. Germishuizen, B.Sc.

Hons (Polygonaceae)

Technician D. A. Davies, B.Sc.

Technical Assistant Mrs I. R. Leistner*

Wing C (Linaceae — Asclepiadaceae)

Senior Professional Officer

Professional Officer

Technician

Technical Assistants

Miss E Retief, M.Sc.

(Campanulaceae)

P. P. J. Herman, B.Sc.

Hons

Mrs M. J. A. W. Crosby,

B.Sc.*

Mrs J. L. M. Grobler*

C. Hildyard, B.Sc. (Elec.

Eng.) (part-time)

Wing D (Convolvulaceae — Asteraceae)

Chief Professional Officer Miss W. G. Welman,

M.Sc.

Professional Officer Mrs S. J. Smithies, M.Sc.

Technical Assistants G. J. Goosen

Mrs K. A. Kleynhans*

Service Room

Technical Assistants Mrs I. Ebersohn

Mrs G. L. Radmacher

Typist Mrs A. M. Verhoef

Photographic Room

Photographer Mrs A. J. Romanowski

Natal Herbarium, Durban (NH)

Senior Professional Officer

Technician

Technical Assistant

P. C. V. du Toit, M.Sc.

(Curator; Poaceae

especially Pentaschistis,

and general identifica-

tions)

Mrs B. J. Pienaar, B.Sc.

Miss A. L. Wright

Albany Museum Herbarium, Grahamstown (GRA)

Senior Professional Officer Mrs E. Brink, B.Sc. (Cu-

rator; general identi-

fications)

Technical Assistants Vacancy

Miss G. V. Britten*

Government Herbarium, Stellenbosch (STE)

Senior Professional Officers Mrs M. F. Rand, M.Sc.

(Curator; Hypoxid-

aceae and general iden-

tifications)

Miss I. Hugo, B.Sc. Hons

( Pelargonium and gene-

ral identifications)

Technical Assistan Miss G. D. Garwood

Clerical Assistant Miss E. N. Pare

Windhoek Herbarium (WIND)

Senior Professional Officer M. A. N. Muller, M.Sc.

(Curator; Eriocephalus

and general identifica-

tions)

Half- day

255

Technician H. J. W. Giess (Poaceae

and trees of South

West Africa/Namibia)

Technical Assistants Mrs H. E. J. Stoffberg

Mrs J. L. van Aswegen

FLORA RESEARCH SECTION

Officer in Charge O. A. Leistner, M.Sc.,

D.Sc., F.L.S.

Flora of Southern Africa Team

Chief Professional Officer O. A. Leistner, M.Sc.,

D.Sc., F.L.S. (Tax-

onomy, especially Mal-

vaceae)

Senior Professional Officers L. E. W. Codd, M.Sc.,

D.Sc. (Taxonomy, es-

pecially Lamiaceae,

and early plant collec-

tors)

R. A. Dyer, M.Sc., D.Sc.,

F.R.S.S.Af., F.A.S.S.

(Taxonomy, especially

Asclepiadaceae)

Mrs G. E. Gibbs Russell,

B.S., Ph.D. (Tax-

onomy, especially aqua-

tic plants and Cyper-

aceae)

Mrs A. A. Obermeyer-

Mauve, M.Sc. (Tax-

onomy, especially Mo-

nocotyledons)

C. H. Stirton, M.Sc.

(Liaison Officer, Kew;

taxonomy, especially

Fabaceae and weeds)

Technicians Mrs R. C. Holcroft*

(artist)

Mrs A. M. Weber (artist)

Technical Assistant Mrs C. F. Fourie*

Anatomy — Cytogenetics Laboratory

Chief Professional Officers R. P. Ellis, M.Sc. (Anat-

omy of southern Afri-

can grasses)

T. B. Vorster, M.Sc.,

D.Sc. (Agric.) (Cyto-

genetics of Eragrostis

curvula Complex)

Technician Miss R. Manders, B.Sc.

Technical Assistant Miss E. van Gylswyk

BOTANICAL SURVEY SECTION

Officer in Charge J. C. Scheepers, M.Sc.,

D.Sc.

Chief Professional Officers J. C. Scheepers, M.Sc.,

D.Sc. (Vegetation ecol-

ogy, especially of

forest / woodland /

savanna / grassland

relationships; conser-

vation and land-use

planning; phytogeo-

graphy)

H. C. Taylor, M.Sc. (O/C

Botanical Research

Unit, Stellenbosch;

mountain fynbos and

forest ecology; Braun-

Blanquet approach and

techniques; conserva-

tion)

C. Boucher, M.Sc. (Low-

land fynbos ecology

and phytosociology;

conservation and land-

use planning; Braun-

Blanquet approach

and techniques)

Senior Professional Officers C. F. Musil, M.Sc.

(Aquatic ecology and

survey of aquatic

plants, especially in

Natal; ecophysiologi-

cal studies on Eich-

hornia crassipes and

Salvinia molest a)

P. J. Weisser, Ph.D.

(Reedswamp ecology;

ecological planning

and environmental im-

pact studies; Zululand

coast dune vegetation;

conservation)

Professional Officers B. M. Campbell, B.Sc.

Hons (Physiognomic-

structural classification

of fynbos; fynbos

phytosociology; quan-

titative methods)

S. R. Fugler, B.Sc. (Eco-

logical survey of in-

vasive aliens in fynbos,

especially Hakea spp.)

R. H. Westfall, B.Sc.

Hons (Ecology and

phytosociology of

Transvaal Bushveld)

Miss B. K. Drews, B.Sc.

Hons (Ecological plan-

ning and environmen-

tal impact studies;

conservation)

P. A. Shepherd, B.Sc.

(Lowland fynbos ecol-

ogy and phytosoci-

ology, threatened spe-

cies)

Mrs M. Engelbrecht,

B.A. Hons (Fine Arts)

(Draughtsmanship and

cartography; artwork,

layout and design)

Miss A. J. Naude, B.Sc.

(Ecological data pro-

cessing and presenta-

tion, ecological litera-

ture, nature conserva-

tion, air-photo inter-

pretation, cartography)

Miss Y. Myburgh

(Draughtsmanship and

cartography; artwork,

layout and design)

Technicians,

Technical Assistant

Half-day

256

DATA PROCESSING AND ECOSYSTEM

STUDIES SECTION

Officer in Charge

Chief Professional Officers

Senior Professional Officer

Technician

Learner Technician.

Technical Assistants,

Administrative Assistant

Clerical Assistants

J. W. Morris, M.Sc.,

Ph.D.

J. W. Morris, M.Sc.,

Ph.D. (Data process-

ing and quantitative

ecology)

M. C. Rutherford,

M.Sc., Ph.D., Dipl.

Datamet. (Biomass

and production studies

in savanna and fynbos)

H. F. Glen, M.Sc., Ph.D.,

F.L.S. (Data process-

ing, numerical tax-

onomy and Mesembry-

anthemaceae)

B. Curran

M. D. Panagos

Vacant

Mrs J. H. Jooste*

Mrs L. E. Oosthuizen*

Mrs N. Nigrini*

Mrs J. Mulvenna

Mrs S. Smit*

Mrs B. Rooke*

Senior Technician

Research Technicians. . . .

Technical Assistant

Acting Curator (Garden)

Acting Curator (Nursery)

First Technicians

Senior Technicians,

Technical Assistants,

Learner Technician.

Farm Foremen

T.N. 4 Supervisor...

T.N. 6

* Half-day

Miss L. Henderson, B.Sc.

Hons (Conservation of

germ plasm and plant

utilization)

Miss S. M. Bulley,

B.Sc. Hons (Weed

research)

Mrs H. M. Anderson,

Ph.D. (Palaeo-flora)

Mrs D. M. C. Fourie,

B.Sc. (Scientific infor-

mation service and

identification of exo-

tics)

Mrs W. G. Gaum, B.Sc.

(Cytogenetics)

Mrs B. van Gass, N.D.

Hort. (Garden utiliza-

tion)

A. A. Balsinhas (Plant

collecting)

J. Erens

D. S. Hardy

D. S. Hardy (Nursery

supervision, succulents

and orchids)

H. J. de Villiers, NTC

III (Hort.), Dipl. Rec.

P.A. (Development

savanna biome)

J. Erens (Garden super-

vision and administra-

tion)

T. A. Ankiewicz, Dip.

For. (Administration,

stores and purchases,

development coastal

forest biome)

Mrs B. C. de Wet*, B.A.

(Garden records)

Mrs K. P. Clarke*

(Garden records)

Miss Y. Mennin

G. J. Stolz

H. N. J. de Beer

L. C. Steenkamp

J. P. Booysens

ECONOMIC BOTANY SECTION

Officer in Charge M. J. Wells, M.Sc.

Chief Professional Officer M. J. Wells, M.Sc.,

(Weed research, pa-

laeo-ethnobotany and

botanical horticulture)

Senior Professional Officers J. M. Anderson, Ph.D.

(Palaeo - ethnobotany,

plant geography and

paleo-flora)

T. H. Arnold, M.Sc.

(Conservation of germ

plasm and plant utili-

zation)

Professional Officers Miss C. A. Liengme,

B.Sc. Hons (Ethnobo-

tany)

G. B. Harding, B.Sc.

(Weed research)

Miss K. J. Duggan, B.Sc.

Hons (Conservation of

germ plasm and plant

utilization)

PRETORIA BOTANICAL GARDEN

PUBLICATIONS BY THE STAFF

Acocks, J. P. H., 1979. The flora that matched the fauna.

Bothalia 12: 673-709.

Anderson, H. M., 1978. Podozamites and associated cones

and scales from the upper Triassic Molteno Formation,

Karoo Basin, South Africa. Palaeont. Afr. 21: 57-77.

Anderson, J. M. & Cruickshank, A. R. T., 1978. The

biostratigraphy of the Permian and Triassic, part 5 — a

review of the classification and distribution of the permo-

triassic tetra-pods. Palaeont. Afr. 21: 1 5—44.

Arnold, T. H., 1979. Ficinia radiata (L.f.) Kunth. Flower.

PI. Afr. 45: t. 1784.

Boucher, C. & Boucher, D. A., 1978. Uitheemse indringer-

plante bedreig die Kaapse Weskus. Elsenburg J. 2 (1): 1-11.

Boucher, C. & Stirton, C. H., 1978. Acacia cyclops. A.Cunn.

ex G.Don. In C. H. Stirton, Plant invaders — beautiful, but

dangerous pp. 40-43. Cape Town: Dept, of Nature &

Environmental Conservation of the Cape Provincial

Administration.

Boucher, C. & Stirton, C. H., 1978. Acacia longifolia (Andr.)

Wild. In C. H. Stirton, Plant invaders — beautiful, but

dangerous pp. 44—47. Cape Town: Dept, of Nature &

Environmental Conservation of the Cape Provincial

Administration.

Boucher, C. & Stirton, C. H., 1978. Acacia saligna (Libill.)

Wendl. In C. H. Stirton, Plant invaders — beautiful, but

dangerous pp. 60-63. Cape Town: Dept, of Nature &

Environmental Conservation of the Cape Provincial

Administration.

Boucher, C. & Stirton, C. H., 1978. Acacia mearnsii de Wild.

In C. H. Stirton, Plant invaders — beautiful, but dangerous

pp. 48-51. Cape Town: Dept, of Nature & Environmental

Conservation of the Cape Provincial Administration.

Boucher, C. & Stirton, C. H., 1978. Acacia pycnantha Benth.

In C. H. Stirton Plant invaders — beautiful, but dangerous

pp. 56-59. Cape Town: Dept, of Nature & Environmental

Conservation of the Cape Provincial Administration.

257

Bullock, A. A., 1978. Bibliography of South African botany

(up to 1951). Flora sth. Afr. Introductory volume, pp. 194.

Campbell, B. M., 1978. Similarity coefficients for classifying

releves Vegetatio 37 101-109.

Codd, L. E., 1979. Plectranthus praetermissus. Flower. PI. Afr.

45: t. 1791.

De Winter, B., 1979. Die bydrae van taksonomie en bio-

sistematiek in die ontwikkeling van Suid-Afrika. Bothalia

12: 742-745.

De Winter, B., Killick, D. J. B. & Edwards, D., 1979.

Terrestrial ecology in South Africa. S. Afr. J. Sci. 75:

59-61.

Downing, B. H., Robinson, E. R., Trollope, W. S. W. &

Morris, J. W., 1978. Influence of macchia eradication

techniques on botanical composition of grasses in the

Dohne Sourveld of the Amatole Mountains. Proc. Grassld.

Soc. Sth. Afr. 13: 111-115.

Dyer, R. A., 1978. The changing scene in Ceropegia and

Brachystelma. Cactus Succ. J., Los Ang. 50. (May-June),

112-114.

Dyer, R. A., 1978. Kalanchoe robusta. Flower. PI. Afr. 45: t.

1783.

Dyer, R. A. 1978, Cyrtanthus odorus. Flower. PI. Afr. 45: t.

1787.

Dyer, R. A., 1978. Pelargonium acetosum. Flower. PI. Afr. 45:

t. 1800.

Dyer, R. A., 1979. Looking back. Cactus Succ. J., Los Ang.

51 (May-June): 116-117.

Dyer, R. A., 1979. A new combination in Cyrtanthus. Bothalia

12: 627.

Dyer, R. A., 1979. New records of Brachystelma. Bothalia 12:

627-629.

Dyer, R. A., 1979. New records of Ceropegia. Bothalia 12: 629.

Dyer, R. A., 1979. The Riocreuxia flanaganii Complex. Bothalia

12: 632.

Dyer, R. A., 1979. A new species of Stapelia. Bothalia 12: 632.

Edwards, D., 1977. Is there a need for strict control of nature

reserves? Koedoe suppl. 95-105.

Edwards, D., 1979. The role of plant ecology in the develop-

ment of South Africa. Bothalia 12: 748-751.

Edwards, D., 1978. The role of plant ecology in the develop-

ment of South Africa. Environment RSA 5 (8): 5-6 & 5

(9): 6-8.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in

surface view. Bothalia 12: 641-671.

Ellis, R. P., Manders, R. & Obermeyer, A. A., 1979, Anatomi-

cal observations on the peduncle of Xyris capensis. Bothalia

12: 637-639.

Furniss, P. R. & Morris, J. W., 1979. Current modelling

activities within Savanna Ecosystem Project. 5. Afr. J.

Sci. 75: 108-109.

Glyphis, J., Moll, E. J. & Campbell, B. M., 1978. Phyto-

sociological studies on Table Mountain, South Africa:

1. The Back Table. Jl S. Afr. Bot. 44: 291-295.

Harding, G., 1978. Prosopis glandulosa Torrey. In C. H.

Stirton, Plant invaders — beautiful , but dangerous pp.

128-131 Cape Town: Dept, of Nature & Environmental

Conservation of the Cape Provincial Administration.

Heeg, J., Breen, C. M., Colvin, P. M., Furness, H. D. &

Musil, C. F., 1978. On the dissolved solids of the Pongolo

flood plain pans. J. Limnol. Soc. Sth. Afr. 4 (1): 59-64.

Huntley, B. J. & Morris, J. W., 1978. Savanna Ecosystem

Project: Phase I summary and Phase II progress. 5. Afr.

Nat. Sci. Prog. Rep. 19: 1-52.

Killick, D. J. B., 1978. More remarks on specific epithets.

Veld & Flora 64: 36.

Killick, D. J. B., 1978. The Mary Gunn Library. Veld &

Flora 64: 125-128.

Killick, D. J. B., 1978. The Mary Gunn Library. The Mafeking

Mail 1 (5): 7,-A.

Killick, D. J. B., 1978. Roadside recognition of trees. AA

outdoor holiday guide of South Africa 6-11.

Killick, D. J. B., 1979. Notes on Flacourtiaceae. Bothalia 12:

635-636 (1979).

Killick, D. J. B., 1979. Omission of Hypericum oligandrum

from Flora of Southern Africa. Bothalia 12: 633 (1979).

Killick, D. J. B., 1979. Review: Mountain splendour by

R. O. Pearse. Veld & Flora 65: 63-64 (1979).

Killick, D. J. B., 1979. The contributions of the Botanical

Research Institute to botany in South Africa. Bothalia 12:

740-742.

Laidler, D., Moll, E. J. & Campbell, B. M., 1978. Phyto-

sociological studies on Table Mountain, South Africa: 2.

The Front Table. Jl S. Afr. Bot. 44: 291-295.

Leistner, O. A., 1979. Southern Africa. In R. A. Perry &

D. W. Goodall, Arid-land ecosystems: structure, functioning

and management 1: 109-143. Cambridge: Cambridge

University Press.

Leistner, O. A., 1979. Review: A field guide to the trees of

of Southern Africa by Eve Palmer. Bothalia 12: 759.

Leistner, O. A., 1979. Review: Trees of Southern Africa by

Keith Coates Palgrave. Bothalia 12: 759.

Leistner, O. A., 1979. Review: Heil-und Giftpflanzen in

Siidwestafrika by Eberhard von Koenen. Bothalia 12:

759-760.

Linder, P. & Campbell, B. M., 1979. Towards a structural-

functional classification of fynbos: a comparison of

methods. Bothalia 12: 723-729.

Magill, R. E., 1979. Amendment to proposal 440: Conserva-

tion of Entodontaceae Kindberg. Taxon 27: 553.

Magill, R. E. & Schelpe, E. A., 1979. The bryophytes of

Southern Africa. An annotated checklist. Mem. bot. Surv.

S. Afr. No. 43, pp. 38.

Moll, E. J., McKenzie, B., McLachlan, O. & Campbell,

B. M., 1978. A mountain in a city — the need to plan the

human usage of the Table Mountain National Monument,

South Africa. Biol. Cons. 13: 117-132.

Morris, J. W., 1978. Botanical research in perspective — -

terrestrial ecology. S. Afr. J. Sci. 74: 236.

Morris, J. W., 1979. Review: Multivariate analysis in vegetation

research. Second enlarged edition by L. Orloci. Bothalia

12: 760.

Morris, J. W. & Glen, H. F„ 1978. PRECIS the National

Herbarium of South Africa (PRE). Computerised Informa-

tion System. Taxon 27 : 449^-62.

Neser, S. & Fugler, S. R., 1978. Silky Hakea. In C. H. Stirton,

Plant Invaders — beautiful but dangerous pp. 76-79. Cape

Town: Dept, of Nature & Environmental Conservation

of the Cape Provincial Administration.

Obermeyer, A. A., 1979. Aloe macleayi. Flower. PI. Afr. 45:

t. 1789.

Obermeyer, A. A., 1979. A new species of Gladiolus from the

Transvaal. Bothalia 12: 636.

Obermeyer, A. A. & Bogner, J., 1979. Gonatopus rhizomatosus.

Flower. PI. Afr. 45: t. 1782.

Oliver, E. G. H., 1978. Review: The South African Acacias

by J. D. Carr. Bothalia 12: 577.

Retief, E. & Finkelstein, N., 1979. The correct name for

Fagara ovatifoliolata. Bothalia 12: 637.

Ross, J. H., 1979. Notes on Prosopis in southern Africa.

Bothalia 12: 635.

Rutherford, M. C., 1979. Aboveground biomass subdivisions

in woody species of the Savanna Ecosystem Project study

area, Nylsvley. S. Afr. Nat. Sci. Prog. Rep. 36: 1-33.

Rutherford, M. C., 1979. Allometric biomass relations applied

to succulent plant forms. Bothalia 12: 737.

Rutherford, M. C. & Kelly, R. D., 1978. Woody plant basal

area and stem increment in Burkea africana — Ochna

pulchra Woodland. S. Afr. J. Sci. 74: 307-308.

Scheepers, J. C., 1979. Review: Biogeography and ecology of

Southern Africa. (Monographiae Biologicae Vol. 31)

edited by M. J. A. Werger. Bothalia 12: 761-762.

Smook, L. & Stirton, C. H., 1979. Naturalized Agrostis in

South Africa: Agrostis avenacea and A. montevidensis.

Bothalia 12: 637.

Stirton, C. H., 1978. Lantana camara L. In C. H. Stirton,

Plant Invaders — beautiful, but dangerous pp. 88-91. Cape

Town: Dept, of Nature & Environmental Conservation of

the Cape Provincial Administration.

Stirton, C. H., 1978. Nerium oleander L. In C. H. Stirton,

Plant invaders — beautiful, but dangerous pp. 100-103. Cape

Town: Dept, of Nature & Environmental Conservation of

the Cape Provincial Administration.

Stirton, C. H., 1978. Nicotiana glauca Grah. In C. H. Stirton,

Plant invaders — beautiful, but dangerous pp. 104-107.

Cape Town: Dept, of Nature & Environmental Conserva-

tion of the Cape Provincial Administration.

Stirton, C. H., 1978. Plant invaders: a threat to the flora of

the Cape Province. In C. H. Stirton, Plant invaders —

beautiful, but dangerous pp. 28-29. Cape Town: Dept, of

Nature & Environmental Conservation of the Cape

Provincial Administration.

Stirton, C. H., 1978. The vegetation of the Cape Province.

In C. H. Stirton, Plant invaders— beautiful, but dangerous

pp. 12-27. Cape Town: Dept, of Nature & Environmental

Conservation of the Cape Provincial Administration.

Stirton, C. H., 1978. Why are plant invaders so successful?

In C. H. Stirton, Plant invaders-— beautiful, but dangerous

pp. 36-37. Cape Town: Dept, of Nature & Environmental

Conservation of the Cape Provincial Administration.

Stirton, C. H., 1978. Why is the flora of the Cape Province

worth saving? In C. H. Stirton, Plant invaders — beautiful,

but dangerous pp. 38-39. Cape Town: Dept, of Nature &

Environmental Conservation of the Cape Provincial

Administration.

258

Stirton, C. H. & Gordon-gray, K. D., 1978. The Eriosema

cordatum complex 1. The E. populifolium group. Bothalia

12: 395-404.

Stirton, C. H., & Neser, S., 1978. Control of plant invaders.

In C. H. Stirton, Plant invaders — beautiful, but dangerous

pp. 144-148. Cape Town: Dept, of Nature & Environ-

mental Conservation of the Cape Provincial Administra-

tion.

Stirton, C. H., & Trainor, H. M., 1978. How you can help

control plant invaders. In C. H. Stirton, Plant invaders —

beautiful, but dangerous pp. 154-156. Cape Town: Dept,

of Nature & Environmental Conservation of the Cape

Provincial Administration.

Taylor, H. C., 1978. Albizia lophantha (Willd.) Benth. Stink-

bean. In C. H. Stirton, Plant invaders — beautiful, but

dangerous pp. 64-67. Cape Town: Dept, of Nature &

Environmental Conservation of the Cape Provincial

Administration.

Thompson, M. F., 1979. Studies in the Hypoxidaceae III.

The genus Pauridia. Bothalia 12: 621-625.

Tolken, H. R., 1979. Review: Flora Zambesiaca. Volume 4.

S.Afr. J. Sci. 75: 142.

Tolken, H. R., 1979. A re-evaluation of the Cotyledon orbiculata

Complex. Bothalia 12: 615-620.

Tolken, H. R., 1979. Three new taxa and a new combination.

Bothalia 12: 633-634.

Tolken, H. R. & Wiens, D., 1979. A new species of Tapinanthus

(Loranthaceae). Jl S. Afr. Bot. 45: 223-225.

Van Der Meulen, F., 1979. Plant sociology of the western

Transvaal Bushveld, South Africa: a syntaxonomic and

synecological study. Dissert. Bot. 49: 1-192. Lehre: J.

Cramer.

Van Der Meulen, F., Morris, J. W. & Westfall, R., 1978.

A computer aid for the preparation of Braun-Blanquet

tables. Vegetatio 38: 129-134.

Van Der Meulen, F. & Westfall, R. H., 1979. A vegetation

map of the western Transvaal Bushveld. Bothalia 12:

731-735.

Verdoorn, Inez C., 1979. Aloe corallina. Flower. PI. Afr. 45:

t. 1788.

Vorster, P., 1979. A note on Mariscus sumatrensis. Bothalia

12: 634.

Vorster, P., 1979. Review: Praktiese plant sistematiek deel I:

Die kriptogame deur A. Eicker, M. I. Claassen, W. F.

Reyneke en N. Grobbelaar. Bothalia 12: 760.

Weisser, P. J. & Marques, F., 1979. Gross vegetation changes

in the dune area, between Richards Bay and the Mfolozi

River, 1937-1974. Bothalia 12: 711-721.

Wells, M. J., 1978. History of control of plant invaders. In

C. H. Stirton, Plant invaders — beautiful but dangerous

pp. 149-153. Cape Town: Dept, of Nature and Environ-

mental Conservation of the Cape Provincial Administra-

tion.

Wells, M. J., 1978. Stipa trichotoma Nees. In C. H. Stirton,

Plant invaders — beautiful but dangerous pp. 140-143. Cape

Town: Dept, of Nature & Environmental Conservation

of the Cape Provincial Administration.

Wells, M. J., 1978. What is a weed? Environment RSA 5

(11): 6-7.

Wells, M. J., 1978. A note on Fagus sylvatica var. purpurea.

Forestry News 4 (78): 6-7.

Wells, M. J., 1979. Review: Pennington’s Butterflies. South

African Weed Science Society, Newsletter 3.

Wells, M. J., 1979. The flora of the Pretoria area. Environment

RSA 6 (4): 3-4.

Wells, M. J., 1979. The role of economic botany in the develop-

ment of South Africa. Bothalia 12: 751-754.

Welman, W. G., 1979. Review: Compositae in Natal by

O. M. Hilliard. Bothalia 12: 760-761.

259

Book Reviews

The Flora Capensis of Jakob and Johan Philipp Breyne

edited by Mary Gunn and Enid du Plessis. Johannesburg:

Ad. Donker for the Brenthurst Press. 1978. Pp. 218, 255 x265

mm, 90 colour plates. Edition limited to 1 000 copies of which

25 were bound in full leather, 125 in half leather at R225 each

and 850 in cloth as the standard edition at R85 each.

The publication of a book on any aspect of the South African

flora is an event to be welcomed by both botanist and plant

enthusiast alike, as such a book, particularly with colour

illustrations, is bound to win an appreciation of our fascinating

and unique flora. The last few decades have seen many such

books appear, which now grace the bookshelves of Africana

collectors, botanists and plant enthusiasts, but none, in my

opinion, can be compared with the one under review.

During the 17th and 18th Centuries artists painted and made

collections of paintings of Cape flowers. Some of the collec-

tions found their way into libraries of eminent and wealthy

people in Europe and aroused keen interest in the flora of the

Cape “that inexhaustible storehouse of botanical riches” (Lod-

diges). This resulted in a number of published works depicting

some of these paintings. Amongst these works was that of

Jakob Breyne of Danzig and his son Johan Philipp. A collection

of colour paintings of Cape flowers was acquired by Jakob

Breyne some time before his death in 1697, and his son had

them bound in book form in 1724 entitling it Flora Capensis.

This book must have passed to many owners and was eventually

acquired by Sir Ernest Oppenheimer in 1959 for his own Brent-

hurst Library. The current owner, Mr Harry Oppenheimer, has

said “I have been concerned to share some of the library’s

unpublished Africana with a wider audience. Hence my decision,

five years ago, to support the creation of the Brenthurst Press.

If its productions make other people as happy as they do me,

then I am not going to grumble about its cost” (S.A. Panorama,

June 1979). The Breynes’ Flora Capensis is the fourth in the

series emanating from the Brenthurst Press and was edited for

them by Miss Mary Gunn, formerly Librarian of the Botanical

Research Institute, and Mrs Enid du Plessis of the C.S.I.R.

The title ‘Flora Capensis’ as given by the Breynes is some-

what of a misnomer as the volume is a florilegium (collection

of paintings of flowers) with no text in the style of a flora. The

editors have therefore given a detailed account of the back-

ground to the work with a most interesting introduction dealing

with florilegia and the early artists and botanists who were or

may have been associated with them. Then there is a most

valuable chapter on the life and works of the Breynes, father

and son, followed by another chapter giving the details about

the Brenthurst florilegium with comparisons of almost identical

florilegia in the libraries of the Botanical Research Institute,

Oxford and Leiden. This text covers 31 pages; included in it

are some reference plates, the title page of the original Flora

Capensis, and a photograph of Jakob Breyne taken from the

Prodromi of 1739 published by Johan Philipp Breyne.

The text is followed by the reproduction of the 85 coloured

folios of the Flora Capensis florilegium rearranged by the

editors. Each plate is accompanied by some brief notes about

the species depicted. Here I would have liked to see much more

useful information such as typification where relevant to the

Breyne’s works. After the paintings there is a section on selected

biographies of persons mentioned in the text, notes on cited

references, and an exhaustive bibliography. One would have

hoped for an index, but reference has to be made to the numer-

ical listing of the plates right at the beginning of the work.

The first thing that strikes one is the quality of production.

The imported paper gives that extra touch of quality. One

glance at the last page of the book where full details of the

paper, printing and binding are given, a rarety in publishing,

shows that this is a book with a difference; but one difference

I found most disconcerting was the format.

The ‘modern’ format seems unsuited to the reproduction of

an old work. The ‘trimmed’ frontispiece and several plates and

the shape of the reproduced title-page look rather out of place.

After a study of the list of plates one finds that just over half

of the plates had to be altered in size to fit the format of repro-

duction. This I find disappointing in a quality reproduction.

It seems to be a general practice that reviewers of books

should show their diligence of examination by quoting printing

errors. This I would not readily have wished to do with this

book, being far more interested in the subject, but realizing

that it is a quality production and an expensive one I was

most surprised to find a number of printing and even manu-

script errors. Spelling mistakes occurred as, for example,

Hessia for Hessea (PI. 13), Penacaceae for Penaeaceae (PI. 53),

pedicles (PI. 17), corms ovid (PI. 39), O. Stopf for O. Stapf

(p. 215) and Plunkenet twice on p. 217. Some bad proof-reading

occurred on PI. 62 where the caption “this is not a Cape plant”

is misplaced and on PI. 78 where the lower two captions are

reversed. Some oddities are “medical” instead of “medicinal”

properties, “lightly packed” instead of “tightly packed” (PI. 55)

and sentences such as “collected by Carl Thunberg at the Cape

in 1772; this plant was also described by Linnaeus” (PI. 37), or

“in the south-western Cape and as far north as Port Elizabeth”

(PI. 51). One botanical statement I disagree with, occurs on

PI. 9 concerning Lachenalia contaminata, a common lowland

species; “The leaves only appear when the flowers are fully

open and sometimes only when the flowers are already going

to seed.” The situation is exactly opposite. The leaves have

mostly died down and even disappeared by the time the plants

have come into flower which is in early summer when the moist

habitat has completely dried out. These are minor matters and

enough has been said.

One could argue about some of the determinations, but then

in the difficult cases of the Group B paintings this may be a

matter of opinion. One real error is in the only non-Cape plant

depicted on PI. 69 which is definitely not Orchis sp. but Ophrys

fuse a.

A number of papers has been published on some of the

early florilegia of Cape flowers notably the Codex Witsenii and

Codex van der Stel and the ‘originals’ from which authors like

the Breynes, Burman, Boerhaave, Petiver, Plukenet, and

Tachard obtained their engravings. The editors of the Breynes’

Flora Capensis have added yet another puzzle, not hitherto

known to recent botanists, to the bibliographical problems.

Frequent mention is made by the editors to the florilegium in

the Botanical Research Institute. This bound volume was most

fortunately acquired by Miss Gunn in 1949 for the Botanical

Research Institute.

This work aroused in me a fascination for these early florilegia

and a desire to examine and compare the originals of as many

as I could. This resulted in some interesting observations, which

have been gathered into a short article published elsewhere in

this journal (pp. 115-125).

The Brenthurst Press must receive a sincere vote of apprecia-

tion for making this work available and the editors must be

thanked for their part in the production of this very fine work.

They have done a great service to botany in South Africa. The

very high price of the book will unfortunately confine it to

Africana collectors and to libraries where botanists will be able

to consult it for references to early botanical history and

taxonomy.

E. G. H. Oliver

Pennington’s butterflies of Southern Africa. Ed. C. G. C.

Dickson with the collaboration of D. M. Kroon. Johannesburg:

Ad. Donker. 1978. Pp. 670, 198 colour plates, 1 map. Price

R51,00.

Not since D. A. Swanepoel’s ‘Butterflies of South Africa’

(1953), which included 474 species, has there been an attempt

to illustrate and treat all the butterflies of South Africa in a

single volume. It is sixteen years since the last volume of the

late Dr George van Son’s uncompleted taxonomic treatise ‘The

Butterflies of Southern Africa’ (1949-1963) appeared, and

there has been a great need for an up to date coverage of the

butterflies of the whole sub-region to provide a platform for

further entomological work, to stimulate inter-disciplinary

research and as a tool for ecosystem studies.

‘Pennington’s Butterflies of Southern Africa’ includes 780

species — all the species recorded from Africa south of an

imaginary line drawn from the Cunene River in the west to the

Zambesi River in the east. Each species is illustrated and has a

short text covering aspects such as: distribution, whether the

species is single, double or continuously brooded, flying times,

foodplants and feeding habits. The result is a fine (if pricey) book

that will be of great value, not only to entomologists, but also

to botanists, pollination biologists, production ecologists, weed

scientists and conservationists.

This is not a taxonomic treatise, it does not include taxonomic

descriptions or drawings of genitalia, but great attention has

been paid to characters that separate closely allied species and

this, together with the scope and excellence of the illustrations

make Pennington’s Butterflies the layman’s ultimate guide to

the identification of the butterflies of the sub-region via wing-

markings.

At best (e.g. no. 29) the plates are superb, but in some (e.g.

no. 126) the blue background is too harsh and competitive—

an aesthetic drawback that does not affect the usefulness of

the plates. I cannot see these plates easily being bettered, and

for this reason alone ‘Pennington’s Butterflies’ will remain a

standard reference work for very many years.

260

My chief criticisms of the text are: the retention of outdated

plant names (rectified in the foodplant list); and the excessive

use of abbreviations for butterfly genera e.g. ‘A’ for Aloeides.

In large genera one may have to turn back 25 pages in order to

find the genus heading (which could have been repeated at the

top of each text column). This criticism extends to the food-

plant list where the combination of genus abbreviations and

lack of page references is annoying, and forces one to match

up genus abbreviations and species names in the general index,

in order to find page references to the relevant text.

The distribution records cited are somewhat scanty and, in

the absence of distribution maps, no clear picture of species

distributions emerges. In the case of rare species needing

protection this is not only understandable but commendable.

In other cases it is trying and I feel that small distribution maps

to give a generalized picture of distribution at a glance would

have been a great asset, although adding considerably to the

bulk and cost of an already substantial volume.

If the species accounts are somewhat spartan, there is com-

pensation to be found in specialist chapters. For those interested

in production studies and in host/predator relationships the

full set of life-history paintings of the Hesperiid (skipper)

butterflies by the late Mr Gowan Clark have been included and

full text references are given to published life-histories of other

South African butterflies. There is a chapter on the intricate

relationships between butterfly larvae, ants and other insects,

and plants (including fungi and lichens) — a necessary reminder

that we have to do with complicated ecosystems, not simple

plant/insect relationships.

The fine mimicry by butterfly larvae and pupae of their host

plants, provides mute evidence of their long history of associa-

tion. Botanists will be interested in the incidence of endemism

and in parallels in distribution between butterflies and vegetation

areas such as the S.W. Cape, the Karoo and the eastern coast-

belt (with tropical affinities).

The list of known foodplants of butterfly larvae, compiled

by Dr D. M. Kroon, is far more comprehensive than anything

previously available. It contains the names of over 400 plants —

the foodplants of well over half the butterfly species of the sub-

region. The list enables one to pick out those, usually the rarer

species that have only a single foodplant, those such as some

of the Papilio spp. that have a number of foodplants (often

restricted to a plant genus or family) and those plants that are

host to a number of butterfly species e.g. there are 9 species

recorded as feeding on Acacia karroo, whilst 34 species (from

9 genera) have been recorded on the grass Ehrharta erecta.

There are some omissions e.g. Cryptocarya latifolia recorded

as a foodplant of Papilio euphranor (in the J. Ent. Soc. S.Afr.

20: 117), and Gymnema sylvestre mentioned on p. 34 of

‘Pennington’s Butterflies’ do not appear in the list. There are

also some errors, such as that resulting from the omission of

the family name Polygonaceae, one line below Polygalaceae on

p. 633; the mis-spelling of spicaeformis, which should read

spiciformis on p. 610; the mis-spelling of ruwenzorensis, which

should read ruwensorensis on p. 613; and there are incon-

sistencies in the endings of specific epithets especially those

taken from the names of persons, but on the whole the list is

conspicuously free of gremlins.

Criticisms aside, I cannot see ‘Pennington’s Butterflies’ being

improved on as a general reference work for many years and

with so many habitats and species being threatened this book

must also (unfortunately) fall into the category of ‘instant

Africana’.

M. J. Wells

Mountain Splendour. Wild Flowers of the Drakensberg

by R. O. Pearse. Cape Town'. Howard Timmins. 1978. Pp. 239

with c. 400 colour photographs. Price R20,00.

For one who, as a young botanist almost straight from uni-

versity was sent in 1950 on transfer to Cathedral Peak Forest

Station in a rather reluctant Willys Americar to carry out a

botanical survey for the Department of Forestry, who found

himself perched alone in a wooden cottage on the Little Berg

without furniture and with only a few tins of canned food,

kilometres from any human habitation and with only the

baboons and antelopes as neighbours, and who initially viewed

the Drakensberg scene with a somewhat jaundiced and lonely

eye, but who in time came to love the mountains and their

flora, returning again year after year, it is a pleasure to review

Mr Pearse’s book, “Mountain Splendour”.

The foreword to the book is fittingly contributed by the late

Professor A. W. Bayer, who also edited the text. It is largely

due to the encouragement and interest of Professor Bayer that

the flora of the Drakensberg is as well known as it is to-day.

In his introduction Mr Pearse outlines the object and scope

of the book. He then deals with the topography and climate,

fires, plant nomenclature, Zulu medicinal usages, African names

and finally the Natal Wild Flower Protection Ordinance.

Preceding the text proper is a double page colour photograph

of a branch of Halleria lucida with red, rain- and light-dappled

flowers. The caption to the photograph is a quotation from

Rupert Brooke’s well-known poem, “The Great Lover”, in

which he describes the things he has loved. The photograph

and caption, to me, set the tone for what is to follow, namely

an intimate look at some of the most common and striking

plants of the ’Berg, photographed with technical skill and

artistry and written in impeccable prose. At the same time the

work, with minor exceptions, is botanically correct. And

through it all, Mr Pearse’s great love of the flowers of the

Drakensberg is clearly evident.

The species are dealt with under family and genus. Mr

Pearse describes all three taxa briefly and informally and gives

the derivation of their names. He records any information of

special interest concerning the plants and mentions their

medicinal uses. Mostly Mr Pearse has used old family names

(sanctioned by long usage), for example Compositae (Aster-

aceae), Labiatae (Lamiaceae) and Leguminosae (Ft baceae),

but surprisingly has used the “new” alternative fam.ly name

Clusiaceae for Guttiferae.

On p. 166 he explains that Plectranthus calycinus has been

transferred to Rabdosia, but states that in this instance it has

been thought better to include it in the genus Plectranthus. In

fact, although he includes it under the heading Plectranthus, he

consistently refers to it by its correct name Rabdosia calycina.

It would have been preferable to include it under a separate

heading Rabdosia.

On p. 140 Mr Pearse writes that the derivation of the name

Hypericum is uncertain. Dr N. K. B. Robson of the British

Museum, world authority on Hypericum and co-author with

the reviewer of the revision of Hypericum in the Flora of

Southern Africa, considers that the name is derived from

the Greek name for St Johns Wort, viz. hypereikon, i.e.

hyperei+kon=above and image. The Greeks used plants of

this genus to decorate religious images in order to ward off evil

spirits, especially around midsummer’s eve (St John’s eve).

On a few occasions Mr Pearse’s names are out of date. For

example, most authorities now regard Vellozia viscosa as a

species of Xerophyta and Vellozia elegans as Talbotia elegans.

Mr Pearse is aware of these changes, but has not accepted them.

Buddleia (p. 154) is now Buddleja following the new Code (1978)

and Buddleja corrugata is now B. loricata.

Much more could have been written about Helichrysum

tenax. This plant grows up to 2 m high (not about 1 m high)

and from a distance bears a striking resemblance to the arbore-

scent Senecios of East African mountains. It usually grows

along streambanks and in fynbos, but when there is man-made

disturbance such as road construction, it is often dominant in

the resultant rock rubble, e.g. Mike’s Pass at Cathedral Peak

Forest Station in the 1950’s. Mr Pearse mentions that the plant

has sticky leaves. This is certainly true and led to speculation

in Natal newspapers many years ago that the plant was in-

sectivorous which, of course, it is not. While on the subject of

Helichrysum, it should be pointed out that apart from a different

flowering time, H. tenuifolium may be distinguished from H.

trilineatum by its greater stature (plants up to 2 m high) and the

usually narrower leaves.

On p. 192 Mr Pearse discusses the derivation of the specific

epithet of Pentanisia prunelloides and correctly states that it

refers to the resemblance of the plant to Prunella of the Lami-

aceae. He then remarks on the coincidence that there is a Latin

word pruna meaning “pertaining to fire” and that the Afrikaans

name of P. prunelloides is Sooibrandbossie and the Zulu name

i-Cimamlilo, “to put out the fire”. However, this is mere

coincidence. The Africans in the ’Berg use the roots to “put out

the fire” in their stomachs, while the Afrikaans name Sooi-

brandbossie meaning “heartburn shrublet” suggests a similar

use. The derivation of Pentanisia, which Mr Pearse is not sure

about, is straightforward: it comes from the Greek words for

five and unequal in reference to the “five very unequal corolla

segments of the flowers” (Wittstein, 1856).

On p. 51 I had to rub my eyes to see whether they were

registering correctly. The excellent photograph of Sandersonia

aurantiaca which many regard as the most beautiful flower in

Natal, bears the caption Littonia modesta and vice versa. As an

editor of botanical journals, only too familiar with printing

gremlins, I sympathize with Mr Pearse. Thirty-five pages later,

I rubbed my eyes again: the “mystery flower”, which Mr

Pearse refers to and which the experts cannot identify with any

certainty, is labelled Watsonia densiflora! Another error is on

p. 69: Vellozia elegans (fig. 3) is really Aristea cognata.

261

At the risk of being labelled a one-track botanist, who has

no time for exotics, I query the inclusion of a weed such as

Phytolacca octandra, which I find ugly, however well photo-

graphed. Bidens formosa, Cosmos, is also a weed, but it is so

attractive a plant of the Drakensberg lowlands in autumn that

its inclusion is understandable. Personally I would have pre-

ferred to see some of the elegant ’Berg grasses included.

I would like to voice a criticism that applies to many authors

(including Mr Pearse) of popular and semi-popular books on

natural history in South Africa, and that is the failure to

acknowledge information obtained from other published works

usually of a scientific nature. Scientists are obliged to acknow-

ledge in the text and in the list of references at the end of their

papers, information obtained elsewhere. It may be argued that

frequent reference citations in the text of popular books would

interrupt the smooth flow of the written word. This may be

true, but it would be quite acceptable and certainly courteous to

mention sources in the Acknowledgements.

In conclusion, “Mountain Splendour” is a well-written and

beautifully illustrated book which is a must for all lovers of the

Drakensberg flora. The book is well-presented, the paper is

of excellent quality and the colour-reproduction is superb.

The pencil sketches by Muriel Zonneveld of individual flowers

and the mountains add to the charm of the book. The slightly

larger than A4 format makes the book more suitable for the

coffee table than the field. I am sure that readers of Bothalia

will enjoy the book as much as I did.

D. J. B. Killick

Plantkunde: Anatomie en Fisiologie by N. Grobbelaar,

P. J. Robbertse, J. V. VAN Greuning & J. H. Visser. Durban:

Butterworth. 1979. Pp. 182, figures numerous. Price R6,95.

This plant anatomy and physiology booklet is one of a series

of five Afrikaans language botany text-books designed speci-

fically for first-year university students, especially those with

virtually no biological background at all. The authors— all

members of staff of the Department of Botany at the Uni-

versity of Pretoria — and the publishers, are to be congratulated

on effectively achieving their stated objective with a product

which compares very favourably with any comparable English

language work.

This book is neatly but not extravagantly produced and,

being soft-covered, should be within the means of the average

student, bearing in mind that a set of five is required to fully

cover the first year botany course. The design is modern and the

layout is clear and uncluttered with the grey chapter title pages

being easy to locate. Each chapter is also preceded by a relevant

and concise summary of the study objectives of the ensuing

chapter. The 173 pages of Afrikaans text are easy to read and

yet comprehensive and even beginners should readily compre-

hend and assimilate the contents and to the English speaker it

will serve as an extremely useful introduction to Afrikaans

terminology and spelling! All figures and photographs are

original and a truly refreshing innovation is the use of examples

drawn from our indigenous South African flora wherever

possible. The printing of the excellent photomicrographs is

adequate and all details referred to are clearly visible. Indeed, a

book of this nature virtually makes a teacher redundant and to

my mind this text-book is ideal for correspondence courses.

This handy series undoubtedly fulfills a very real need in our

particular South African context — to provide a basic botanical

text-book for Afrikaans-speaking students requiring botany as

an ancillary course in their university curriculae. This text-book

is, therefore, more than adequate for a first, and only, course in

plant anatomy and physiology for medical, veterinary and

agricultural students. I am sure that the authors have directed

this book primarily at this category of student which un-

doubtedly comprises the bulk of university first year botany

enrolment and includes many students with little or no botanical

background. However, if the training of professional botanists

in South Africa is to be upgraded, this booklet must be seen

only as a sound foundation for the first course in a three-year

biology degree but needs to be supplemented at least by further

practical courses in techniques and applied aspects. Obviously

the authors have had to contend with shortcomings in our educa-

tional system in this respect, but perhaps it is opportune to

propose here that matriculation biology (as well as physical

science and mathematics) be made a prerequisite for acceptance

to a biological degree at South African universities just as

science and mathematics are minimum requirements for

chemistry and physics majors. To have to bring beginners up

to matric level in the first year leaves an insurmountable teaching

load for the second and third years.

In common with most other botanical text-books, I feel that

the applications of plant anatomical and physiological informa-

tion, in fields other than basic, theoretical botany, are not

sufficiently stressed. To the young, unenlightened student plant

anatomy and physiology must still appear as isolated, uninte-

grated, academic subjects. For students not intending to

major in botany the use of anatomical and physiological

knowledge in solving everyday problems in disciplines other

than botany would be stimulating and of possible benefit in

their future careers The inclusion of D. F. Cutler’s Applied

Plant Anatomy in the lists of suggested further reading at the

end of each chapter in the anatomy section would have over-

come this shortcoming to a certain extent. Cutler’s book,

together with the book under review, complement each another

and together would help stimulate student interest in anatomy

and assist students in relating to the subject.

Very few minor errors were detected although Fig. 5 5 C is

definitely not a transverse section of a Cynodon sp. leaf. This

is a thorough and well-balanced book and appears assured of

the success it definitely deserves.

R. P. Ellis

Plantkunde: Genetika en Kriptogame deur A. Eicker,

M. I. Claassen, W. F. Reyneke en N. Grobbelaar. 1979’

Durban: Butterworth. Bladsye 157. Prys R6,95.

Hierdie boek is een van ’n reeks van vier boeke wat gepubli-

seer is met die doel om eerstejaarplantkundestudente met hulle

studies te help. Die werk word eenvoudig uiteengesit sodat

iemand met geen biologiese agtergrond dit ook kan verstaan.

Die studiedoelstellings aan die begin van elke hoofstuk lei die

eerstejaarstudent sodat hy na elke hoofstuk, deur selftoetsing,

kan bepaal of hy op ’n sinvolle wyse daaraan verantwoording

kan doen.

Genetika

Die eerste hoofstuk is bondig en sistematies uiteengesit,

sodat die leerstof vinnig en maklik bemeester kan word. Elke

meiosestadium se belangrikste kenmerke kon miskien net

puntsgewys aangestip gewees het, wat die instudering daarvan

sou bespoedig. Elke stadium in meiose is goed beskryf, alhoewel

hier en daar belangrike kenmerke bygevoeg kan word. In die

beskrywing van leptoteen kan genoem word dat die kernmem-

braan teenwoordig is en dat die nukleolus baie prominent is

gedurende hierdie stadium.

Sinne soos “diploiede organismes het twee stelle chromosome

in elke sel” (bl. 5), kan miskien tot verwarring lei by die student

wat nog geen onderrig in biologie gehad het nie. Dit sou duide-

liker wees om te se: “Diploiede organismes het twee stelle

chromosome in elke somatiese sel (liggaamsel)”, wat geslag-

selle dus uitsluit.

In Fig. 1.2 sal ’n paar veranderinge dit dalk meer duidelik

maak. Die nukteolus kan so geteken word dat dit verklein in

elk daaropvolgende profase I — stadia — vanaf sigoteen totdat

dit uiteindelik verdwyn wanneer metafase I bereik word. In die

figuur kan die terminalisering van die chiasmata gedurende

diakinese duideliker ge'illustreer word, sodat daar gesien kan

word dat die chiasmata neig om na die eindpunte van die

bivalent te beweeg om aan die bivalent (gedurende diakinese),

sy kenmerkende ronde voorkoms te gee. In dieselfde figuur sou

dit ’n baie helderder beeld aan die eerstejaarstudent gee as die

terminalisasie van die chiasmata gedurende diakinese sodanig

aangedui word, dat die nuwe chromatiedsegmente wat ontstaan

het as gevolg van die breking in die chromatiede en gevolglike

uitruiling van chromatiedsegmente, aangetoon word, ’n Inlas-

skets ter illustrering sou miskien die geskikste wees.

Gedurende metafase I van Fig. 1 . 2 kan die skets die feit dat

die chiasmata van die bivalente op die ekwatoriaalplaat le en

die sentromere georienteer in die rigting van die teenoorge-

stelde pole aan beide kante van die ekwatoriaalplaat, duideliker

aangetoon het. Hierdie sal dan ook aan die eerstejaarstudent

die verskilpunt van metafase I van meiose en metafase van

mitose (waar die sentromere op die ekwatoriaalplaat le),

helderder na vore bring en uitlig. Twee verskillende vorms van

bivalente kon ook aangetoon word. Byvoorbeeld ’n ringvormige

bivalent (by die metasentriese chromosome), waar chiasmata

in beide arms van die chromosome plaasgevind het en ’n staaf-

vormige bivalent, wat kon ontstaan het by die akrosentriese

chromosome waar chiasmata in die langer arm van die bivalent

kon plaasgevind het. Die voorstelling behoort die leerstof vir die

eerstejaarstudent duideliker uit te spel. Verder kan die onder-

skrif van die figuur vir volledigheid lui : “Die verskillende stappe

van meiose by ’n organisme met 2 pare chromosome (2n=4).”

By die beskrywing van anafase I van meoise kan spesifiek

gewys word op die belangrike verskilpunt met mitose. Dit is,

dat gedurende anafase I van meiose geen deling van die sen-

tromere plaasvind soos by anafase van mitose nie, maar slegs

’n skeiding van 'onaf hanklike sentromere.

262

Daar is op ’n eenvoudige en insiggewende wyse verduidelik

watter belangrike bron van variasie die anafaseskeiding van

meiose I onder die nageslag meebring. Die skrywers het ’n

baie goeie beskrywing en verduideliking van Mendel se wet

van segregasie en rekombinasie, asook die wet van onafhanklike

sortering weergee. Koppeling en uitwisseling is baie goed aan

die hand van baie duidelike sketse aangebied en dit geld ook

vir die beskrywing en diagramme oor die tipe kruisings wat

gedoen kan word om bepaalde genotipiese en fenotipiese ver-

houdings in die F2-generasie te verkry.

Hierdie hoofstuk beantwoord op ’n kort, eenvoudige en

interessante wyse aan die voorafopgestelde studiedoelstellings.

Enige eerstejaarstudent behoort, nadat hy hierdie hoofstuk

deurgewerk het, die teks baie goed te begryp en sy studiedoel-

stellings sinvol te kan weergee. Die leerstof word op ’n stimu-

lerende wyse oorgedra en is deurspek met interessanthede wat

die leser tot verdere opleeswerk behoort aan te vuur. Deur

gebruik te maak van die voorgestelde 8 boeke as aanvullende

leesstof, is die inhoud van hierdie hoofstuk baie deeglik gedek.

Ongelukkig is daar ’n drukfout in die teks op bl. 21. Die

letters C en c moet vervang word met G en g.

Kriptogame

Die verskillende terme word kortliks verduidelik en in die

meeste gevalle word dit verder met uitstekende diagrammatiese

voorstellings, deurstraal-aftaselektronmikrograwe of fotomikro-

grawe soos veral in die geval van die alge, toegelig. Ongelukkig

word geen skaal by enige van die elektronmikrograwe of foto-

mikrograwe aangegee nie. Die student het dus geen begrip van

die grootte van die organisme nie. Byskrifte by die mikrograwe

ontbreek, sommige is selfverduidelikend, terwyl by ander, soos

byvoorbeeld in die geval van die Bryophyta dit nie so is nie. In

sommige gevalle van die Bryophyta-mikrograwe word die

leser egter verwys na vorige figure vir die byskrifte. Hierdie

mikrograwe is eintlik ’n herhaling van dieselfde struktuur.

As gevolg van die bondigheid van die inhoud word daar soms

opmerkings gemaak wat die verkeerde inligting weergee, byvoor-

beeld die alge word in ses afdelings verdeel, maar daar is geen

verklaring waarom die Cryptophyta nie ingesluit word nie.

Volgens die inhoud van die boek is algeselle of naak of dit

besit ’n selwand wat in die meeste gevalle uit twee lae bestaan.

Geen melding word gemaak van die ander tipe selomhulsels

soos skubbe of lorikas wat ’n belangrike rol speel in die identifi-

kasie van wiere nie.

Die boek voorsien in ’n behoefte wat daar bestaan het en

sal van groot nut vir enige eerstejaarstudent wees. Aangesien

die strukture en lewenssiklusse van hierdie laer groepe so goed

uiteengesit word, word daar gehoop dat meer studente gein-

spireer sal word om hierdie organismes te bestudeer — ’n saak

wat in die verlede grootliks verwaarloos is.

W. Gaum en R. Glen

Plantkunde: Organografie en Sitologie deur W. F.

Reyneke, L. A. Coetzer en N. Grobbelaar. Durban: Butter-

worth. 1979. 133 bladsye. Prys R6,95.

Hierdie publikasie, wat handel oor die organografie en sito-

logie van die Anthophyta (Blomplante), verteenwoordig een

van ’n reeks van vier teoretiese en een praktiese handleiding wat

deur die personeel van die universiteit van Pretoria, vir die

gebruik van eerstejaarplantkundestudente, geskryf is. ’n Student

het nie matriek onderrig in biologie nodig om hierdie boek te

verstaan nie. Die geskiedkundige sitologiese agtergrond neem

die student kortliks deur ’n historiese toer vanaf die baan-

brekerswerk, gedoen deur vroeere wetenskaplikes, met behulp

van hulle eenvoudige lense, tot die waarneming van die gedetail-

leerde ultrastruktuur van selle met behulp van verbetering in

resolusiekrag, wat verkry is met die elektronmikroskoop.

Die teks is duidelik uiteengesit met goeie taalgebruik en is

maklik leesbaar en nie te tegnies van aard nie. Vetgedrukte

letters word deurgaans gebruik om sleutelwoorde vir die

student te benadruk en aan die begin van elke hoofstuk word

studiedoelstellings ook weergegee wat aan die studente ’n

geleentheid skenk om spoedig sleutelwoorde te noteer en der-

halwe horn in die regte rigting te lei. Aanvullende leesstof word

aan die einde van elke hoofstuk aangedui en bied aan die student

’n wye spektrum van inligtingsbronne wat geraadpleeg kan word

indien verdere inligting verlang word. Die sistematiese en logiese

wyse van aanbieding is stimulerend vir studiedoeleindes.

Die boek bevat goeie illustrasies wat deur middel van foto’s

asook diagrammatiese voorstellings die teks breedvoerig

toelig. Byvoorbeeld, die gepaarde membrane van die golgi-

apparaat en endosplasmiese retikulum is goed waarneembaar

aangebring en die funksie van die ribosome word met behulp

van ’n skets gei'llustreer wat die verduideliking in die teks in ’n

neutedop aanvul. Vir duidelikheid oor die morfologie van ’n

chromosoom in, byvoorbeeld, laat profase, kon ’n skets met

byskrifte egter ingesluit gewees het sodat terme soos chroma-

tied, suster-chromatiede en half-chromatiede gei'llustreer kon

word. Wat die weergawe van foto’s betref, dien daarop gelet

te word dat waar foto’s sekere voorbeelde uitbeeld, elkeen

duidelik is om sodoende die leser ’n goeie idee van die onder-

werp te laat vorm.

Die ordelikheid, voortreflikheid, bondigheid en interessant-

heid van aanbieding is treffend, sodat dit op ’n aangename wyse

deur ’n student geabsorbeer kan word. Om verwarring te voor-

kom kon ’n paar punte egter duideliker gestel word. By

anafase kon die sametrekking van die spoeldrade in die rigting

van die pole as rede aangevoer word waarom die dogter-

chromosome na teenoorgestelde pole beweeg. Om die idee

dat mitose presies in metodies-afgebakende stadia plaasvind

uit te skakel, kan egter vooraf genoem word dat mitose ’n

dinamiese, deurlopende proses is waar een stadium in ’n

ander oorgaan sonder definitiewe af bakeningslyne.

W. Gaum en G. Germishuizen

GUIDE FOR AUTHORS

GENERAL

Bothalia is a medium for the publication of botanical papers

dealing with the flora and vegetation of Southern Africa. Papers

submitted for publication in Bothalia should conform to the

general style and layout of recent issues of the journal (from

Vol. 11 onwards) and may be written in either English or

Afrikaans.

TEXT

Manuscripts should be typed, double-spaced on one side

of uniformly-sized A4 paper having at least a margin of 3 cm

all round. Latin names of plants should be underlined to indicate

italics. All other marking of the copy should be left to the editor.

Metric units are to be used throughout. Manuscripts should be

submitted in duplicate to the Editor, Bothalia, Private Bag X101,

Pretoria.

ABSTRACT

A short abstract of 100-200 words in both English and

Afrikaans should be provided. In the abstract the names of new

species and new combinations should not be underlined.

FIGURES

Black and white drawings, including graphs, should be in

jet-black Indian ink preferably on bristol board or plastic film.

Lines should be bold enough to stand reduction. Indicate the

desired lettering lightly in pencil : the printer will insert the final

lettering. If authors prefer to do their own lettering, then use

some printing device such as stencilling, letraset, etc. It is

recommended that drawings should be twice the size of the

final reduction.

Photographs submitted should be of good quality, glossy,

sharp and of moderate, but not excessive contrast. Photograph

mosaics should be composed by the authors themselves: the

component photographs should be mounted neatly on a white

card base leaving a narrow gap between each print; number the

prints using some printing device.

Figures should be planned to fit, after reduction, into a width

of 8 cm, 11 cm or 17 cm with a maximum vertical length of

24 cm.

The number of each figure and the author’s name should be

written on the back of the figure using a soft pencil.

Captions for figures should be collected together and typed on

a separate page headed Captions for Figures. A copy of each

caption should be attached to the base of each figure.

Authors should indicate in pencil in the text where they would

like their illustrations to appear.

TABLES

Tables should be set out on separate sheets and numbered

in Arabic numerals.

CITATION OF SPECIMENS

In citing specimens the grid reference system should be used

(Technical Note: Gen. 4). Provinces /countries should be cited

in the following order: S.W. Africa, Botswana, Transvaal,

Orange Free State, Swaziland, Natal, Lesotho and the Cape.

Grid references should be cited in numerical sequence. Locality

records for specimens should preferably be given to within a

quarter-degree square. Records from the same one-degree square

are given in alphabetical order i.e. (-AC) precedes (-AD), etc.

Records from the same quarter-degree square are arranged

alphabetically according to the collectors’ names; the quarter

degree references must be repeated for each specimen cited.

The following example will explain the procedure:

Natal. — 2731 (Louwsburg): 16 km E. of Nongoma (-DD),

Pelser 354; near Dwarsrand, Van der Merwe 4789. 2829 (Harri-

smith): near Groothoek (-AB), Smith 234; Koffiefontein (-AB),

Taylor 720; Cathedral Peak Forest Station (-CC), Marriott 74;

Wilgerfontein, Roux 426. Grid ref. unknown: Sterkstroom,

Strydom 12.

Records from outside Southern Africa should be cited from

north to south i.e. preceding those from Southern Africa.

The abbreviation “distr.” should be added to all district names

e.g.:

Kenya. — Nairobi distr.: Nairobi plains beyond race course,

Napier 485.

GIDS VIR SKRYWERS

ALGEMEEN

Bothalia is ’n medium vir die publikasie van plantkundige

artikels wat handel oor die flora van Suidelike Afrika Artikels

wat voorgele word vir publikasie in Bothalia behoort ooreen

te stem met die algemene styl en rangskikking van onlang e

uitgawes van die tydskrif (vanaf Vol. 11). Dit mag in Engels of

in Afrikaans geskryf word.

TEKS

Manuskripte moet getik wees in dubbelspasiering slegs op

een kant van ewegroot A4-papier, met reg rondom ’n rand van

minstens 3 cm breed. Latynse name van plante moet onderstreep

word om aan te dui dat dit kursief gedruk moet word. Alle

ander merke moet aan die redakteur oorgelaat word. Metrieke

eenhede moet deurgaans gebruik word. Manuskripte moet in

tweevoud ingedien word by die Redakteur, Bothalia, Privaatsak

X101, Pretoria.

UITTREKSEL

’n Kort uittreksel van 100-200 woorde moet voorsien word,

beide in Engels en Afrikaans. In die uittreksel moet die name

van nuwe soorte en nuwe kombinasies nie onderstreep word nie.

AFBEELDINGS

Wit en swart tekeninge, insluitende grafieke, moet met

pikswart Indiese ink geteken word, verkieslik op “bristol

board” of plastiekfilm. Lyne moet dik genoeg wees om verklein

te kan word. Dui die verlangde byskrifte liggies in potlood aan:

die drukker sal die uiteindelike byskrifte invoeg. Indien skrywers

verkies om hulle eie byskrifte te maak, gebruik dan een of ander

hulpmiddel soos letraset of ’n sjabloon. Dit is wenslik dat

tekeninge tweemaal so groot as die uiteindelike verkleining sal

wees.

Foto’s wat ingedien word, moet van hoe kwaliteit wees—

glansend, skerp en van matige maar nie oordrewe kontras.

Fotomosaieke moet deur die skrywer self saamgestel word:

die afsonderlike foto’s moet netjies monteer word op ’n stuk

wit karton met ’n smal strokie tussen die foto’s; nommer die

foto’s met behulp van een of ander druk-hulpmiddel.

Afbeeldings moet so beplan word dat hulle na verkleining

sal pas in ’n breedte van 8 cm, 11 cm of 17 cm met ’n maksimum

vertikale lengte van 24 cm.

Die nommer van elke afbeelding sowel as die skrywer se

naam moet op die rugkant van die afbeelding geskryf word met

’n sagte potlood.

Onderskrifte vir afbeeldings moet bymekaar getik word op ’n

afsonderlike bladsy met die opskrif Onderskrifte vir Afbeeldings.

’n Afskrif van elke onderskrif moet aan die onderkant van elke

afbeelding vasgeheg word.

Skrywers moet met potlood in die teks aandui waar hulle

graag hulle afbeeldings wil he.

TABELLE

Tabelle moet op afsonderlike velle papier kom en genommer

word met Arabiese nommers.

SITERING VAN EKSEMPLARE

Wanneer eksemplare siteer word, moet die ruitverwysing

stelsel gebruik word (Tegniese Nota: Gen. 4). Provinsies/

lande moet in die volgende volgorde siteer word: Suidwes-

Afrika, Botswana, Transvaal, Oranje-Vrystaat, Swaziland,

Natal, Lesotho en die Kaapprovinsie. Ruitverwy sings moet

in numeriese volgorde siteer word. Lokaliteitsrekords vir

eksemplare moet verkieslik tot binne kwartgraadvierkante

gegee word. Rekords uit dieselfde eengraadvierkant word in

alfabetiese volgorde aangebied, nl. (-AC) kom voor (-AD)

ens. Rekords uit dieselfde kwartgraadvierkant word alfabeties

gerangskik volgens die versamelaars se name, en die kwart-

graadverwysings moet herhaal word vir elke eksemplaar wat

siteer word. Die volgende voorbeeld sal die metode verduidelik:

Natal.— 2731 (Louwsburg): 16 km O. van Nongoma (-DD),

Pelser 354; naby Dwarsrand, Van der Merwe 4789. 2829

(Harrismith) : naby Groothoek (— AB), Smith 234; Koffiefontein

(-AB), Taylor 720; Cathedral Peak Bosboustasie (-CC),

Marriott 74; Wilgerfontein, Roux 426. Ruitverwysing onbekend:

Sterkstroom, Strydom 12.

Rekords van buite Suidelike Afrika moet siteer word van

noord na. suid, d.w.s. dit gaan die van Suidelike Afrika vooraf.

Die afkorting “distr.” behoort by alle distriksname gevoeg

te word, bv:

Kenya- — Nairobi-distr.: Nairobivlakte anderkant die ren-

baan, Napier 485.

REFERENCES

References in the test should be cited as follows: “Jones

(1955) stated . . ,”or“. . .(Smith, 1956)” when giving a reference

simply as authority for a statement. The list of references at

the end of the article should be arranged alphabetically and

the literature abbreviations used should conform to the World

List of Scientific Periodicals (1965) or the list of Literature

Abbreviations (Technical Note: Tax. 6/3 AN 1) issued by the

Botanical Research Institute, thus:

Brown, N. E., 1909. Asclepiadaceae. In W. T. Thiselton-

Dyer, FI. Cap. 6,2: 518-1036. London: Lovell Reeve.

Hutchinson, J., 1946. A botanist in Southern Africa. London:

Gawthorn.

Kruger, F. J., 1974. The physiography and plant communities

of the Jakkalsrivier Catchment. M.Sc. (Forestry) thesis,

University of Stellenbosch (unpublished).

Morris, J. W., 1969. An ordination of the vegetation of

Ntshongweni, Natal. Bothalia 10: 89-120.

If, as in many taxonomic papers, periodicals or books are

mentioned in the text, usually in the species synopsis, they should

be cited as in the following examples : Gilg & Ben. in Bot. Jb. 53 :

240 (1915) and Burtt Davy, FI. Transv. 1 : 122 (1926).

REPRINTS

Authors receive 75 reprints gratis. If there is more than one

author, this number will have to be shared between or among

them.

VERWYSINGS

Verwysings in die teks moet as volg siteer word: “Jones

(1955) beweer . . .” of “. . . (Smith, 1956)” wanneer ’n verwysing

slegs as outoriteit vir ’n stelling gegee word. Die verwysingslys

aan die einde van die artikel moet alfabeties gerangsicik wees

en die literatuurafkortings wat gebruik word, moet in ooreen-

stemming wees met die World List of Scientific Periodicals

(1965) of die lys van Literatuurafkortings (Tegniese Nota: Tax.

6/3 AN 1) wat uitgegee is deur die Navorsingsinstituut vir

Plantkunde, as volg:

Brown, N. E., 1909. Asclepiadaceae. In W. T. Thiselton-

Dyer, FI. Cap. 6,2: 518-1036. London: Lovell Reeve.

Hutchinson, J., 1946. A botanist in Southern Africa. London:

Gawthorn.

Kruger, F. J., 1974. The physiography and plant communities

of the Jakkalsrivier Catchment. M.Sc. (Bosbou) tesis,

Universiteit van Stellenbosch (ongepubliseerd).

Morris, J. W., 1969 An ordination of the vegetation of

Notshongweni, Natal. Bothalia 10: 89-120.

Wanneer, soos in baie taksonomiese artikels die geval is,

tydskrifte of boeke in die teks genoem word, gewoonlik in die

soortsinopsis, behoort hulle siteer te word soos in die volgende

voorbeelde: Gilg & Ben. in Bot. Jb. 53: 240 (1915) en Burtt

Davy, FI. Transv. 1: 122 (1926).

HERDRUKKE

Skrywers ontvang 75 herdrukke gratis. Wanneer daar meer

as een skrywer is, sal hierdie aantal tussen hulle verdeel moet

word.

Republic of

South Africa

Republiek van

Suid-Afrika

BOTHALIA

Volume 13, No. 3 & 4

Edited by/Onder redaksie van

D. J. B. Killick

ISSN 0006-8241

Botanical Research Institute

Navorsingsinstituut vir Plantkunde

Department of Agriculture and Fisheries

Departement van Landbou en Visserye

South Africa/Suid-Afrika

1981

GPS (L)

CONTENTS— INHOUD

Volume 13, No. 3 & 4

Page

Bladsy

1. Revision of Melhania in southern Africa. I. C. VERDOORN 263

2. The genus Waltheria in southern Africa. I. C. Verdoorn 275

3. The genus Cola in southern Africa. I. C. Verdoorn 277

4. The Eriosema cordatum complex. II. The Eriosema cordatum and E. nutans groups. C. H. Stirton 281

5. Natural hybridization in the genus Eriosema (Leguminosae) in South Africa. C. H. Stirton 307

6. Studies in the Leguminosae — Papilionoideae of southern Africa. C. H. Stirton 317

7. The genus Dipogon. C. H. Stirton 327

8. Notes on the taxonomy of Rubus in southern Africa. C. H. Stirton 331

9. New records of naturalized Rubus in southern Africa. C. H. Stirton 333

10. Taxonomic studies in the Disinae. V. A revision of the genus Monadenia. H. P. Linder 339

1 1. Taxonomic studies in the Disinae. VI. A revision of the genus Herschelia. H. P. Linder 365

12. An analysis of the African Acacia species: their distribution, possible origins and relationships.

J. H. ROSS 389

13. Sexual nuclear division in Neocosmospora. K. T. vanWarmelo 415

14. Miscellaneous notes on the genus Pelargonium. J. J. A. van der Walt and P. J. Vorster 431

15. Notes on African plants

Amaryllidaceae. A. A. Obermeyer 435

Asclepiadaceae. R. A. Dyer 435

Commelinaceae. A. A. Obermeyer 436

Cyperaceae. T. H. Arnold; T. H. Arnold and P. Vorster; P. Vorster 439

Ericaceae. E. G. H. Oliver 446

Ericoaulaceae. A. A. Obermeyer 450

Liliaceae. D. S. Hardy and C. Reid; A. A. Obermeyer 451

Mesembryanthemaceae. H. F. GLEN 454

Orchidaceae. K. L. Immelman 455

Poaceae. G. E. Gibbs Russell; P. D. F. Kok 457

Polygalaceae: J. A. R. Paiva 458

Vitaceae. E. Retief 460

16. The phytogeography and ecology of Macrocoma (Orthotrichaceae, Musci) in Africa. R. E. Magil L

and D. H. Vitt 463

17. Phytogeography and speciation in the vegetation of the eastern Cape. G. E. Gibbs Russell and

E. R. Robinson 467

18. Information available within PRECIS data bank of the National Herbarium, Pretoria, with exam-

ples of uses to which it may be put. J. W. MORRlSand R. Manders 473

19. Leaf anatomy of the South African Danthonieae (Poaceae). IV. Merxmuellera drakensbergensis

and M. stereophylla. R. P. Ellis 487

20. Leaf anatomy of the South African Danthonieae. (Poaceae). V. Merxmuellera macowanii, M. davyi

and M. aureocephala. R.P.Ellis 493

21. Plants used by the Tsonga people of Gazankulu. C. A. Liengme 501

22. A brief account of the coast vegetation near Port Elizabeth. H. C. Taylor and J. W. Morris 519

23. Changes in the herb layer of the riverine woodland in the Sengwe Wildlife Research Area, Zim-

babwe. P. R. Guy -'-7

24. Survival and regeneration and leaf biomass changes in woody plants following spring burns in

Burkea africana — Ochna pulchra Savanna. M. C. Ruteierford -51

25. Monitoring Phragmites australis increases from 1937 to 1976 in the Siyai Lagoon (Natal, South

Africa) by means of air photo interpretation. P. J. WElSSERand R. J. Parsons 553

26. Structural and floristic classifications of Cape Mountain fynbos on Rooiberg, southern Cape.

H. C. Taylor and F. van der Meulen 557

Page

Bladsy

27. Miscellaneous ecological notes

Investigation into the significance of plant dispersion in assessing pasture condition. H. H. von

Broembsen 569

A simple method for determining the density of plants in a randomly-dispersed population.

H. H. von Broembsen 574

A note on the extension of the Degree Reference System for citing biological distribution

records to north of Equator and west of Greenwich meridian. D. Edwards 574

Determination of plot size. B. M. CAMPBELLand E. J. Moll 575

Review of the work of the Botanical Research Institute, 1979/80 577

Book Reviews 589

Bothalia 13, 3 & 4: 263-273 (1981)

Revision of Melhania in southern Africa

I. C. VERDOORN*

ABSTRACT

The 13 species of Melhania in southern Africa are revised. Among the problems resolved is a long-standing one

concerning the identities of M. rehmannii Szyszyl. and M. griquensis H. Bol., the latter in its strict sense being

relegated to synonymy of the former.

RESUME

REVISION DU MELHANIA EN AFRIQUE AUSTRALE

Les 13 especes de Melhania d’Afrique australe son/ revisees. Parmi les problemes resolus il en est an qui fut de

lohgue duree et qui concerne les identites du M. rehmannii Szyszyl. et de M. griquensis H. Bol., ce dernier, dans le

sens strict, etant relegue la synonymie du premier.

MELHANIA

Melhania Forssk., FI. Aegypt.-Arab. 107: 64

(1775); Harv. in F. C. 1: 221 (1860); Arenes in FI. de

Madagascar 131, Sterculiaceae: 160 (1959); Wild in

F. Z., 1, 2: (1961); M. Friedrich et al., F.S.W.A. 84:

23 (1969); R. A. Dyer, Gen. 1: 364 (1975).

Brotera Cav., Icon. 5: 19, t. 433 (1799).

Shrublets or herbaceous plants with a woody root-

stock, the new growth usually pubescent; pubescence

of stellate or tufted unicellular hairs often matted to

form a thin or dense tomentum or in the case of long

tufted hairs forming a hispid or villose pubescence,

sometimes scales and short glandular hairs present as

well. Leaves simple, petioled, stipulate. Inflorescence

axillary, or terminal, racemose or cymose, often

reduced to a solitary flower; flowers bisexual, rarely

polygamous. Epicalyx-bracts 3 persistent, united at

the base, inserted subunilaterally, pubescent on both

surfaces. Calyx 5-lobed, persistent. Petals yellow, 5,

obovate, somewhat unequal-sided, convolute, hypo-

gynous, early deciduous from the base but the faded

flower persisting, like a cap, with convolute apices,

on the capsule. Stamens 5, alternating with 5 ligulate

staminodes, connate at the base forming a short

annulus or collar. Ovary stellate-tomentose, 5-celled,

ovules 1 to many in a cell; style with 5 stigmatic

branches, sometimes abortive in essentially male

flowers. Capsule loculicidally dehiscent. Seeds 3 or

more sided usually with minute raised broken lines or

dots on the surface.

The name Melhania is derived from Mt Melhan, in

Arabia.

KEY TO SPECIES

la Upper surface of leaves fairly sparsely pilose with rather long unicellular subappressed simple (or very rarely

2 to 3 together) hairs lying in the same direction:

2a Leaves shallowly dentate, lateral veins excurrent in the teeth, lamina broad, mostly over 2,5 cm and up to

4 cm broad 1 . M. didytna

2b Leaves with entire margins very rarely a few teeth present on some leaves, upper surface early glabrescent,

lamina mostly under 2,5 cm broad (see also under lb) 2. M. prostrata

lb Upper surface of leaves glabrous (early glabrescent) or distinctly stellate pubescent or densely to sparsely

stellate-tomentose:

3a Leaves with entire margins, lamina mostly under 2,5 cm broad:

4a Leaves velvety-tomentose on the upper surface, style long 6-8 mm long 3. M. integra

4b Leaves early glabrescent on the upper surface, style short 2-3,5 mm long:

5a Hairs when present on the upper surface simple subappressed lying in the same direction, veins

impressed above 2. M. prostrata

5b Hairs when present on the upper surface minutely stellate, net veins on glabrous leaves (Middelburg

form) prominent above 4. M. randii

3b Leaves serrate or crenate, obscurely so in M. polygama, but then lamina broad up to 5,5 cm broad:

6a Epicalyx bracts broadly ovate to ovate-acuminate, rounded or cordate at the base (rarely sublanceolate

in M. polygama, which occurs only in N.E. Natal):

7a Epicalyx as long as or longer than the calyx:

8a Epicalyx broad, shortly and broadly acuminate, grey-tomentose; flowers bisexual; style short up to

3,5 mm long; plant ferrugineous pubescent in part 5. M.forbesii

8b Epicalyx long-acuminate, grey tomentose; flowers polygamous, the male with long styles ± 7 mm long

and abortive style-branches, female with short style, t 3 mm, and branches as long as or longer than

the style; plant sulphur-grey tomentose 6. M. polygama

7b Epicalyx shorter than the calyx: sometimes only slightly shorter in M. transvaalensis:

9a Epicalyx abruptly acuminate; style long, up to 7 mm long; leaves tomentose on both surfaces

7. M. acuminata

aggregate species

10a Leaves usually broadly ovate-oblong, tomentum predominantly grey, interspersed with short

light brown clustered hairs 7a. var- acuminata

10b Leaves usually narrowly ovate-oblong, tomentum grey but copiously interspersed with rather long

dark brown clustered hairs 2b. var. agnosia

9b Epicalyx not or gradually acuminate in the upper half; style short:

1 la Leaves with the upper surface distinctly and coarsely stellate-pubescent throughout, not tomentose,

usually more than twice as long as broad; flowers usually over 1 cm long, style short up to 5 mm

ione 8. M. transvaalensis

'c/o Botanical Research Institute, Department of Agriculture and Fisheries, Private Bag X 1 0 1 , Pretoria, 0001.

264

REVISION OF MELHANIA IN SOUTHERN AFRICA

lib leaves with upper surface stellate-tomentose to glabrescent usually less than twice as long as

broad; flowers under 1 cm long 9. M. rehmannii

6b Epicalyx-bracts linear-subulate to lanceolate; if fairly broad not broadest at the base:

12a Leaves not ovate, not broadest at or near the base; low plants under or up to 35 cm high with several to

many stems arising from the base:

13a Epicalyx-bracts linear-subulate; flower-stalk short, ± 4 mm long, shorter than the petiole and usually

1 - flowered 10. M. virescens

13b Epicalyx-bracts lanceolate, mature flower-stalk as long as or longer than the petiole, 10-40 mm

long, 1-3-flowered 11. M. burchellii

12b Leaves usually ovate to ovate-oblong, variable in shape but usually broadest at or near the base; shrub-

lets usually over 30 cm tall, main stems 1 to a few, woody branched above:

14a Strong perennial, producing shoots from a reduced persistent woody base; leaves varying in size,

2- 7 cm long, 1-4,5 cm broad 12. M. damarana

14b Weak perennial, not producing roots from a woody base; leaves varying in size, but usually under

5 cm long, thin-textured and minutely tomentose on both surfaces, silvery grey on the under

surface 13. M. suluensis

1. Melhania didyma Eckl. & Zeyh., Enum. 52

(1834) Szyszyl. Polypet. Thalam. Rehm. 137 (1887)

excl. var.; Harvey in F. C. 1: 222 (1860); K. Schum.

in Engl., Monogr. Afr. Pfl. 5: 10 (1900) excl. var.

Syntypes: Uitenhage, Winterhoek Mountains, Eck-

lon & Zevher s.n. (SAM!); Zuurbergen, Ecklon &

Zevher s.n.

Vialia macrophylla Vis., Ind. Sem. Patav. in Linnaea 15, Litt.

103 (1840) Described from plant grown from seed, origin un-

known.

Melhania leucantha E. Mey. in Drege, Zwei Pfl. Doc. 201, 133

& 134 (1843) nom. nud.

Shrublet, 20-60 cm tall, sometimes up to 1 m tall,

branched from near the base, erect or, when

browsed, forming a low bushy plant, main stem

woody; branchlets tomentose, the grey tomentum in-

termixed with light brown, clustered, villose hairs.

Stipules subulate, tomentose, up to 14 cm long.

Leaves pubescent above with subappressed, long,

single hairs or very occasionally 2 or 3 hairs from the

same base, greyish-stellate-tomentose beneath inter-

mixed with clustered light brown hairs, especially on

the nerves, ovate-oblong to ovate-lanceolate, up to

10 cm long and 4 cm broad, margins shallowly den-

tate, sometimes obscurely so; petiole 0,7-1, 5 cm

long. Inflorescence axillary, peduncle rigid, up to 5

cm long, usually 2-flowered at the apex, pedicels up

to 1 cm long. Epicalyx-bracts, broadly ovate, acumi-

nate, cuspidate at the apex, cordate at the base,

usually slightly longer than the calyx and petals,

densely tomentose to villose-tomentose on both

sides, up to 2,5 cm long, 1,3 cm broad, usually longer

than the calyx. Calyx lobed almost to the base, lobes

ovate-lanceolate, up to 1,5 cm long, hirsute-hispid

without. Petals about as long as the calyx, broadest

at the apex, pale yellow, fading brownish yellow.

Stamens 5, alternating with 5 linear-ligulate

staminodes which are about twice as long as the

stamens. Ovary subglobose densly villose; style about

3-5 mm long, branches 1,75-2 mm long; ovules 4-5

in a cell. Capsule hispid with tufted hairs. Fig. 1.

Occurs on grassy slopes, thorn veld or margins of

open forests, often near rivers. Recorded from the

eastern Cape, Swaziland, Natal and the eastern

Transvaal.

Transvaal*. — Nelspruit: Kaapsche Hoop, Rogers 20875.

Swaziland. — Mbabane: Hlane Wildlife Sanctuary, Stephen

1451.

Natal. — Alfred: Izingolweni, Strey 6146. Durban: KrausslM.

Eshowe: near Dhlangubo store, Codd & Verdoorn 10175. Est-

court: Jones’s Kloof, West 1548. Hlabisa: Hluhluwe, Wells 2120.

Lower Umfolozi: Umfolozi Game Reserve, Strey 4958. Mapu-

mula: 16 km NE of Thring’s Post, Codd & Verdoorn 10180. Port

Shepstone: Umzimkulu ‘horseshoe’ slopes, Strey 7459. Umzinto:

Imhambanyoni Valley, King 114.

Cape.— Albany: Kudu Nature Reserve, Fort Brown, Gibbs

Russell 3652. Alice: Woburn, Acocks 9823. East London: Nahoon

River Valley, Smith 3753. Fort Beaufort: Fort Fordyce, Story

2114. Kokstad: Umzimvubu Cutting, Strey 10707. Komga:

Flanagan 9. Mqunduli: Pegler 578. Queenstown: Junction Farm,

Galpin 8078. Uitenhage: Groendal, Long 1162.

Fig 1. — Melhania didyma. Differ-

ences between hairs on upper

and lower surfaces of leaves. 1

& 2, simple and tufted hairs

from upper surface; 3 & 4,

stellate hairs with slender

rays, side and dorsal views,

from lower surface. All x 50.

* The specimen citations in this and the following two papers are

arranged alphabetically according to districts.

I. C. VERDOORN

265

According to Harvey in F. C. 1: 222, M. leucantha

E. Mey. nom. nud. is this species. The Drege speci-

mens at Kew which Meyer named M. leucantha have

been examined and they are M. didyma.

M. didyma resembles M. forbesii in some respects

and especially in the size and shape of the epicalyx-

bracts, but it is readily distinguished by the upper

surface of the leaf being pubescent with single subap-

pressed hairs, while in M. forbesii it is tomentose or

with short, grouped or stellate hairs.

The specimens from Botswana listed by O. B. Mil-

ler in J 1 S. Afr. Bot. 18: 57 (1952) as M. didyma are

M. forbesii and most probably Dinter 3036 from

Tsumeb, listed as M. didyma in Feddes Reprium 19:

96 (1924) is, judging by the locality, also this species.

For the relationship between M. didyma and M.

prostrata see the notes under the latter species.

2. Melhania prostrata DC., Prodr. 1: 499 (Jan.

1824); Burch., Trav. 2: 263 (1824); Harv. in F. C. 1:

222 (1860); K. Schum. in Engl., Monogr. Afr. Pfl. 5:

9 (1900); Burtt Davy, FI. Transv. 1: 260 (1926); Wild

in F. Z. 1, 2: 530 (1961). Type: Cape, Kimberley,

Klipfontein, Burchell 2153 (PRE, iso.!).

M. linearifolia Sond. in Linnaea 23: I 8 (1850); Harv. in F. C. 1:

222 (1860). Type: Natal, Durban Gueinzius 532 (S, holo. ! ; PRE

photo.!).

M. didyma var. linearifolia (Sond.) Szyszyl., Polypet. Thalam.

Rehm. 137 (1887); K. Schum. in Engl., Monogr. Afr. Pfl. 5: 10

(1900); Burtt Davy, FI. Transv. 1: 261 (1926).

M. prostrata forma latifolia Bak. f. in J. Bot., Lond. 37: 425

(1899). Type: Zimbabwe, Bulawayo, Rand 24.

Shrublet, a weak to fairly strong perennial, usually

branching near the base; branches upright, spreading

or prostrate, usually branched again 20-60 cm long,

new growth grey-stellate-tomentose, tomentum inter-

spersed in parts with scattered red-brown, bunched

hairs. Stipules subulate 4-14 mm long. Leaves linear-

lanceolate linear-oblong, narrowly oblong or nar-

rowly ovate-oblong, rounded at base and apex, rarely

apex truncate and very shallowly lobed, 2,4-12 cm

long, 0,4-2 cm broad, rarely 2,5 cm broad, dis-

colourous, lower surface grey-tomentose with

golden-brown, lepidote, stellate-pilose scales bearing

short or long hairs upper surface fairly sparsely

pubescent with mostly rather long single, (or rarely

about 3 in a cluster), subappressed hairs, lying t in

the same direction or glabrous; margins entire, very

rarely shallowly lobed in part; lateral veins few, im-

pressed above, prominent beneath, not excurrent;

petiole 0,2-1, 2 cm long, rarely 1,5 cm long, grey

tomentose. Inflorescence axillary 1-2-flowered;

peduncles 3-32 mm long, ultimately longer than the

petioles which range from 2-12 mm, frequently in

the single flowered inflorescence the peduncle and

pedicels are indistinguishable. Epicalyx-bracts from

narrowly to broadly ovate-acuminate, shallowly to

deeply cordate at the base and abruptly narrowed in a

short basal claw, about 8-15 mm long and 2-6,5 mm

broad, thinly to densely stellate-tomentose on both

surfaces. Calyx with lobes ovate-lanceolate, acumi-

nate, shorter or longer than the epicalyx, 8-14 mm

long, 2,5-4 mm broad, dorsally grey-tomentose,

sparsely to very densely interspersed with bunched or

stellate golden-brown, bristle-like hairs. Petals

yellow, broadest at the apex cuneate, about 12 mm

long and 8 mm broad at apex. Stamens about 4 mm

long, alternating with linear-ligulate staminodes

somewhat longer than the stamens. Ovary subglo-

bose densely tomentose and bristly with bunched or

stellate hairs; style 1,5-3,75 mm long with branches

half to three quarters as long; ovules 6-8 in a cell.

Fig. 2.

K '-4 '/■ fta, yc.

Fig 2. — Melhania prostrata. Burchett 2153, isotype in PRE.

Found on rocky slopes, sand flats, grasslands and

bushveld from Griqualand West, northern Cape,

throughout the Transvaal into Botswana, Swaziland

and Zululand. Also recorded from Zimbabwe and

Mozambique.

Botswana. — South East District: Dikomo di Ki, Wild 5 1 79.

Transvaal. — Barberton, Pott 5634. Brits: Jacksonstuin, Van

Vuuren 418. Heidelberg: Suikerbosrand Nature Reserve, Lam-

brechts 183. Johannesburg: Schoongezicht, Mogg 22932. Letaba:

Hans Merensky Nature Reserve, Oates 391. Lydenburg: Kruger’s

Post, Burtt Davy 7303. Marico: 4,8 km S. W. Zeerust, Story 795.

Middelburg: Loskop Dam, Codd & Verdoorn 10366. Nelspruit:

Lowveld Botanic Gardens, Buitendag 775. Pietersburg: Chunies-

poort, Codd & Verdoorn, 10478. Pilgrims Rest: Erasmus Pass,

Killick & Strey 2522. Potchefstroom: Venterskroon, Van Wyk

1428. Potgietersrust: Pyramid Estate, Galpin 8923. Pretoria:

Wonderboom, Thode A 396. Rustenburg: Rustenburg Nature Re-

serve, Jacobsen 1 146. Soutpansberg: 37 km N. of Louis Trichardt,

Magill 9630A. Vereeniging: Suikerbosrand, Bredenkamp 532.

Waterberg: Kransberg, Germishuizen 280.

O.F.S. — Senekal: Doringberg-oos, Mutter 1866.

Swaziland. — Stegi: Blue Jay Ranch, Compton 29746.

Natal. — Hlabisa: Hluhluwe Game Reserve, Ward Mil.

Ingwuvuma: Josini Dam, Strey 5283. Ngotshe: Makatini flats,

Wells 2219. Ubombo: Lebombo Mts, near Ubombo, Acocks

13124.

Cape. — Barkly West: Klipvlei, Esterhuysen 2082. Hay: Berge-

naars Pad, Acocks 2087,. Kuruman: 17,2 km N. W. of Kuruman,

Leistner 596. Philipstown: near Petrusville, Jooste 285.

266

REVISION OF MELHANIA IN SOUTHERN AFRICA

The decision to follow the authors of Flora Zam-

besiaca and put M. linearifolia Sond. into synonymy

under M. prostrata and not as a variety of M. didyma

as Szyszylowicz did, was made after seeing the type

specimen of the former, Gueinzius 532. This speci-

men compares well with the type of M. prostrata

from Griqualand West and the many specimens of

this species found throughout the Transvaal and in

parts of Natal.

M. prostrata is closely related to M. didytna, but

may be distinguished by the much narrower leaves

with entire margins (very rarely a single leaf shallow-

ly lobed in part). The pubescence on the upper sur-

face of the leaves is, when present, like that of M.

didyma of rather long usual single subappressed

hairs, but in M. prostrata the upper surface is early

glabrescent.

Where the distribution of these two species

overlap, intermediates are found. It is not clear

whether these intermediates are of hybrid origin or

whether the two taxa are fairly recent segregates from

a common stock.

The specimen of Schlechter 4322 in the Transvaal

Museum Herbarium, now in the National Herb-

arium, is a mixture of M. prostrata and M. transvaal-

ensis. The latter has the upper surface of the leaves

distinctly stellate throughout.

When Burchell collected the type of this species,

Burchell 2153, he evidently thought it was a Dom-

beya, for in his catalogue he has listed it as Dombeya

prostrata.

3. Melhania integra Verdoorn in Bothalia 8: 177

(1964). Type: Transvaal, Pietersburg, Wolkberg,

Codd & Verdoorn 10407. (PRE, holo.!).

Shrublet, 20-45 cm tall with several slender

woody stems from a woody rootstock, laxly branch-

ed; branches shortly and densely tomentose, some-

times, in new growth reddish-brown lepidote stellate

scales are obvious. Stipules subulate, 6-15 mm long.

Leaves on flowering branches clustered towards the

apex, entire (very rarely lower leaves shallowly tooth-

ed in part), finely and densely velvety tomentose on

the upper surface, greyish tomentose beneath with

reddish, lepidote, stellate scales usually obvious, es-

pecially on the nerves, narrowly lanceolate-elliptic,

mostly 2-6,5 cm long and 0,4-1, 2 cm broad,

sometimes up to 1,8 cm broad, nerves prominent

beneath, obscure above, basal nerves 3, rarely 5, base

subrounded, apex subacute or rounded, mucronate;

petiole 0,4-1 cm long, tomentose with a few to many

reddish-brown, lepidote stellate scales obvious. In-

florescence axillary in the upper leaves, 1-3-flowered;

peduncle 4-20 mm long, pedicels 3-14 mm long,

both tomentose with scattered reddish-brown stellate

scales obvious in parts. Epicalyx-bracts ovate

acuminate, rounded to cordate at the base, about 12

mm long, 6 mm broad near the base, tomentose on

both surfaces with the lepidote scales obvious on the

outside. Calyx-lobes lanceolate, acuminate, about 15

mm long, 3,5 mm broad near the base, tomentose

and lepidote without. Petals canary yellow, turning

brownish at maturity, broadest at the apex, about 16

mm long, 10-14 mm broad at apex. Stamens with

filaments 1 mm long, anthers 3-4 mm long,

staminodes 9 mm long. Ovary subglobose, about 3

mm diam., white tomentose; style about 6-8 mm

long, branches 1,5-3 mm long; ovules 7 in each

locule. Capsule about 10 mm long, 8 mm diam.,

shortly tomentose and scaly, the scales bearing short

hairs. Fig. 3.

Fig. 3. — Melhania integra. Codd <& Verdoorn 10407, isotype in

PRE.

Recorded from the Wolkberg, the northern ex-

tremity of the Transvaal Drakensberg and at the top

of Abel Erasmus Pass, growing on rocky hillsides

among grass in shallow soil over dolomite.

Transvaal.— Letaba: Dalton Farm, above Dublin Mine,

Fourie 601; Pietersburg: Wolkberg, Paardevlei, Codd & Verdoorn

10407; Verdoorn 2474; Pilgrims Rest: Abel Erasmus Pass, Mauve

4790.

This species occurs abundantly in the area which

lies south-east of Boyne in the Pietersburg District

and on the borders of the Letaba District. It grows on

the grassy slopes among dolomite rocks. The entire,

velvety-tomentose leaves, together with the long style

and large petals, characterize the species.

It resembles M. randii Bak. f. from Zimbabwe in

the entire leaves but differs, among other things, in

the pubescence on the upper surface of the leaves, the

larger flowers and the ovate-cordate epicalyx. The

specimen from Abel Erasmus Pass is not quite as

velvety as the Wolkberg specimens.

4. Melhania randii Bak. f. in J. Bot., Lond. 37:

425 (1899); K. Schum. in Engl., Monogr. Afr. Pfl. 5:

6 (1900); Wild in F. Z. 1,2: 530 (1961). Type: Zim-

babwe, Salisbury, Rand 439 (BM, holo.).

Shrublet, 6-60 cm tall with several slender woody

stems from a thick, woody rootstock, branches at

first finely stellate-tomentose with minute reddish

brown scales intermingled. Stipules narrowly

lanceolate-subulate up to 8 mm long. Leaves entire,

rarely some shallowly dentate in part, finely minutely

I. C. VERDOORN

267

stellate-tomentose and with scattered reddish brown

scales (the scales sometimes stalked and all bearing

short hairs) on the under surface, the upper surface

finely stellate-tomentose and in some forms early

glabrescent, narrowly oblong-elliptic or lanceolate-

elliptic to ovate-oblong, 2-9 cm long, 0,2-1, 5 cm

broad, rounded at the base, subacute to rounded at

the apex, mucronate; petiole up to 1 cm long, tomen-

tose and lepidote. Inflorescence axillary, tomentose

and lepidote, 1-3-flowered; peduncle 7 mm long and

indistinguishable from the pedicel or up to 4 cm long

and pedicels up to 1,2 cm long. Epicalyx-bracts nar-

rowly lanceolate, narrowly ovate to ovate-acuminate,

from 6-13 mm long and 1-6 mm broad (large bracts

found mostly on the specimens from the eastern

mountainous country), tomentose on both surfaces

and with minute reddish scales intermixed on upper

surface. Calyx-lobes lanceolate-acuminate, up to 11

mm long, tomentose and with conspicuous reddish

brown scales on outer surface. Petals yellow, ‘bright

golden yellow’ about 9 mm long, 7 mm broad at the

oblique apex. Stamens with filaments 1 mm long, an-

thers 2-3 mm long, staminodes 3-5 mm long. Ovary

stellate tomentose and with straight bristly hairs

especially in upper portion, about 2 mm diam., style

short, at maturity up to 2 mm long, branches when

spreading up to 3 mm long. Capsule up to 1 cm long,

stellate-pubescent and scaly.

Recorded from the Middelburg District, where it

grows in open mixed grassveld in iron-rich soil and a

taller form from the Barberton District on the expos-

ed, wind-swept summit of a mountain near Barber-

ton. Also occurs in Zimbabwe and Mozambique.

Transvaal. — Barberton: Sheba Hills, Russel’s Beacon,

Scheepers 1247; Verdoorn 2479. Middelburg: Roos Senekal, Ver-

doorn 2492; 2501; 2502; 2503; 2504.

M. randii has been recorded from four widely

separated areas of distribution. Probably the long

period of isolation accounts for the plants in any one

of these colonies differing in some respects from

those in another. The specimens seen to date from

around Salisbury, the type locality; are low bushes

with long narrow leaves on which the pubescence per-

sists on the upper surface. In the mountainous

eastern regions of Zimbabwe and the Transvaal the

plants are over 30 cm tall with slender erect stems;

they also have long leaves with persistent pubescence

on the upper surface, and differ only slightly from

each other, for instance the epicalyx bracts on the

Transvaal specimens are usually broader than those

on the Zimbabwean specimens. The form found in

the Middelburg District of the Transvaal differs from

all the others in having shorter and broader leaves,

which are early glabrescent on the upper surface. In

this form the bushes are usually under 30 cm tall.

M. randii resembles M. prostrata and M. integra in

the entire leaves. It can be distinguished from the

former, among other things, by the upper surface of

the leaves being finely stellate-pubescent before

becoming glabrous as opposed to having long simple,

appressed hairs on the upper surface, before becom-

ing glabrous. From M. integra it may be distinguish-

ed by the persistent minutely stellate pubescence of

the upper surface of the leaf as well as the short style

and smaller flowers, the styles being 1,75 mm long as

opposed to 8 mm long in M. integra and the flowers

1,1 cm long as opposed to being 2 cm long.

5. Melhania forbesii Planch, ex Mast, in F.T.A.

1: 231 (1868); K. Schum. in Engl., Monogr. Aft. Pfl.

5: 12 (1900); Burtt Davy, FI. Transv. 1: 261 (1926);

Wild in F.Z. 1, 2: 531 (1961); M. Friedrich et al.,

F.S.W.A. 84: 25 (1969). Syntypes: Mozambique,

Cupanga, Kirk (K); without precise locality, Hutton

(K).

M. serrulata R. E. Fr., Wiss. Ergebn. Schwed. Rhod.-Kongo-

Exped. 1: 157 (1914). Type. Zimbabwe, Victoria Falls, Fries 74

(UPS).

Shrublet about 60 cm tall, sometimes taller, stem

erect, branched; branches suberect the upper portion

usually appearing rust-coloured from the clusters of

rather long ferruginous hairs, which are sometimes

stalked and occur more or less densely intermingled

with the short greyish stellate tomentum. Stipules

subulate 10-20 mm long. Leaves greyish tomentose

with short, fine grouped or stellate hairs especially

dense on the lower surface, and sometimes with

clusters of ferruginous hairs on the nerves beneath,

more or less ovate-oblong, 3-1 1 cm long, 1, 5—6 cm

broad, rounded at the apex, sometimes mucronate,

base rounded to cordate, margin crenate-serrulate;

petiole 1-2,5 cm long, usually densely covered with

clusters of ferruginous hairs. Inflorescence axillary,

peduncles straight, tomentose and with numerous

clusters of ferruginous hairs, suberect, 1-6 cm long,

branched near the apex; raceme 1-4-flowered, usual-

ly 3-flowered; pedicels 2-7 mm long. Epicalyx-bracts

from about 12-18 mm long and 10-14 mm broad,

ovate, sometimes broadly acuminate to an acute apex

(not abruptly so), cordate at the base, longer than the

calyx and petals (in dried specimens obscuring the

calyx), tomentose on both surfaces. Calyx-lobes

greyish villose-tomentose dorsally, glabrous within,

lanceolate about 1 cm long, apex acute. Petals about

as long as the calyx-lobes, obovate, broadest at the

apex (the faded petals twisted at the apex and per-

sisting like a cap on the capsule may appear to be

longer than the calyx but they are not attached at the

base). Stamens about 12 mm long, filaments and an-

thers about 5 mm long. Ovary subglobose about 9

mm diam., densely tomentose; style about 2 mm

long, style-branches about 4 mm long. Seeds up to 6

in a cell.

Found along rivers in alluvial soil or on dolomitic

hillsides, sandy flats and parkland in red soil.

Recorded from Natal, Swaziland, the eastern and

northern Transvaal; and northern South West

Africa. Also occurs in Mozambique, Zambia, Zim-

babwe and Angola.

S.w. A.— Eastern Caprivi: Linyanti area, Killick & Leistner

3169, Grootfontein: foot of Aha Mtns, Story 6376. Okavango:

Nianzana, Dinter 7252. Ovambo: Oshihanga, Loeb 563.

Botswana. — Central District: Orapa, Allen 126. Ngamiland:

Plains near Tsodilo Hills, Banks 59.

Transvaal. — Barberton: Komatipoort, Marloth 5457. Letaba:

Hans Merensky Nature Reserve, Gilliland 782. Nelspruit: E. of

Skukuza, Codd & de Winter 5015. Pietersburg: Chuniespoort,

Codd & Verdoorn 10460. Pilgrims Rest: Dindinnie Farm, Mauve

4793. Soutpansberg: Punda Milia, Schlieben 9334. Waterberg:

Krantzberg area, Germishuizen 267.

Swaziland.— Hlatikulu: near Sipofaneni, Compton 29891.

Tabankula: Barrett 312. Tshanani: Swaziland Irrigation Scheme,

Riches 12.

Natal.— Hlabisa: False Bay Park, Scott-Smith & Ward 70. In-

gwavuma: Kosi Estuary, Vahrmeijer & Tolken 914. Ubombo:

Mkuzi Game Reserve, Ward 4123.

In the shape of the epicalyx M. forbesii closely

resembles M. didyma, but can be readily distinguish-

ed by the upper surface of the leaves being tomentose

and minutely stellate and not pubescent with single,

long, sub-appressed hairs as in M. didyma. The

specimens from Botswana listed by O. B. Miller in J1

S. Afr. Bot. 18: 57 (1952) as M. didyma are M.

forbesii and, most probably, Dinter 3036 also listed

268

REVISION OF MELHANIA IN SOUTHERN AFRICA

as M. didyma from Tsumeb, in Feddes Reprium 19:

96 (1924), is this species. (See also notes under M.

acuminata var. acuminata.)

The native name for M.forbesii is moulhwadambo

meaning ‘setting sun’.

6. Melhania polygama Verdoorn in Bothalia 8:

178 (1964) Type: Natal, Umfolozi Game Reserve,

Strey 4957b (PRE, holo.!).

Shrublet about 35 cm tall, with a woody rootstock.

Stems many, rather robust, in dried specimens 3-5

mm diam., laxly branched, new growth with a sul-

phur-grey stellate tomentum, the hairs of different

lengths and grouped on scales. Stipules subulate,

5-10 mm long. Leaves finely and densely sulphur-

grey stellate-tomentose on both surfaces, upper sur-

face somewhat darker than the lower; lamina broadly

ovate-trullate or ovate-elliptic, 3-9, 5 cm long,

1,5-5, 5 cm broad, broadly cuneate or subrounded at

the base, narrowing slightly to the obtuse or subtrun-

cate apex, shallowly crenate-dentate except at the

base; petiole 1-2 cm long, densely sulphur-grey

tomentose, fairly stout. Inflorescence in axils of up-

per leaves; peduncle up to 6 cm long, stellate-

tomentose, 1-3-flowered at apex; pedicels 0,3- 1,5

cm long. Flowers polygamous. Epicalyx-bracts

ovate, cordate at base, gradually attenuating to a

subulate upper portion and acute apex, longer than

calyx, about 14 mm long, 4-12 mm wide, stellate-

tomentose on both surfaces. Calyx with lobes

deltoid-acuminate, about 11 mm long, outer surface

stellate-tomentose. Petals yellow, broadest at the

apex, in essentially male flowers about 14 mm long,

in female shorter, about 10 mm long. Stamens, in

male flowers, with anthers about 4 mm long, in

female abortive, about 2 mm long; ligulate

staminodes about 6,5 mm long. Ovary densely beset

with long, grouped hairs on scales; style in male

flowers long and slender, about 7 mm long, branches

aborted or short, style in female flowers thick and

short, about 2,5 mm long, branches longer than the

style, about 5 mm long, sometimes decurrent on the

style; ovules 5 in a cell. Capsule broadly oblong,

about 9 mm long, tomentose and with groups of

rather long, sulphur-coloured hairs. Fig. 4.

To date recorded only from the Umfolozi Game

Reserve, Zululand, where it grows on grassy hill-

slopes.

Natal. — Lower Umfolozi: Umfolozi Game Reserve, Ward

3324; Dengeseni Beacon Hill, Strey 4957a; 4957b; 4957c; 4957d.

This is the first known record of a species of

Melhania with polygamous flowers. The essentially

male flowers are more showy with longer petals,

large stamens and an aborted style. The female

flowers have short petals, small stamens and a style

with long, stout branches. The broad ovate-trullate

leaves, that is, leaves, which are usually broadest

about a third of length above the base, are a distin-

guishing feature.

7. Melhania acuminata Mast, in F.T.A. 1: 231

(1868); Wild in F.Z. 1, 2: 532 (1961). Type: Mozam-

bique, Sena, Kirk s.n. (K, holo.).

Shrublet about 65 cm tall; stem erect from a com-

paratively slender tap-root, branched; branches erect

and spreading, new growth greyish stellate-tomen-

tose and sometimes with bunched light to dark brown

hairs intermixed. Stipules subulate, 0,5- 1,7 cm long.

Leaves greyish stellate-tomentose on both surfaces,

sometimes more thinly so on the upper surface, lower

ET3 '•’* ”

Fig 4. — Melhania polygama. Strey 4957, holotype in PRE.

surface sometimes with light to dark brown bunched

hairs, especially on the prominent nerves, broadly to

narrowly ovate-oblong, 3,5-10 cm long, 1,5-6 cm

broad, narrowing towards a broadly rounded apex,

rounded to subcordate at the base, margins crenate-

serrate; petiole 1-2,5 cm long, stellate-tomentose.

Inflorescence axillary, peduncle straight, suberect up

to 5 cm long, 3- or more-flowered near the apex;

pedicels 0,6-1 cm long. Epicalyx-bracts ovate (rarely

narrowly so), shorter than the calyx, rounded at the

base and with a very short broad claw, abruptly

acuminate above into a caudate-like upper half,

tomentose on both surfaces. Calyx with lobes

lanceolate, gradually narrowing to the acute apex,

1-1,5 cm long, stellate-tomentose on the outside on-

ly. Petals usually longer than the epicalyx, obovate,

broadest at apex. 1-1,5 cm long. Stamens with

filaments and anthers about 3,5 mm long;

staminodes 7 mm long. Ovary subglobose about 8

mm diam., densely tomentose; style 6-10 mm long,

branches 1-2 mm long; ovules 6 in a cell.

For key to varieties see key to species.

(a) var. acuminata. Wild in F.Z. 1, 2: 532 (1961).

M. acuminata Mast, in F.T.A. 1: 231 (1868); K. Schum. in

Engl., Monogr. Afr. Pfl. 5: 13 (1900); Burtt Davy, FI. Transv. 1:

261 (1926); M. Friedrich et al., F.S.W.A. 84: 24 (1969); M.

velutina sensu Excell & Mendonfa, C.F.A. 1, 2: 190 (1951), pro

parte.

The typical variety is characterized by the leaves

usually being broadly ovate to ovate-oblong, the

flower large, and the rather thick tomentum being

predominantly grey with fairly obscure, bunched,

I. C. VERDOORN 269

brown hairs intermixed. There is a fair number of

herbarium specimens, which approach var. agnosia

either in the leaves being narrower than usual or the

bunched hairs somewhat more conspicuous. This

supports the decision to reduce M. agnosta K.

Schum. to varietal rank under M. acuminata.

Found in sandy soil, on flats, in bushveld and on

grassy wooded slopes. Recorded from the northern

Cape, the northern and eastern Transvaal, South

West Africa and Botswana. Also occurs in Zambia,

Zimbabwe, Angola and Mozambique.

S.W.A.— Gobabis: Farm Davis, Merxmuller 1206. Grootfon-

tein: Neitsas, Schonfelder 1051. Kavango: E. of Runtu, De Winter

3788. Otjiwarongo: Quickborn, Bradfield 366. Outjo: Chorab,

Dinter 5352. Owambo: S.E. of Oshikango, Rodin 9118. Tsumeb:

Farm Kumhause, Giess 14975.

Botswana.— Central Dist. : Seleka Ranch, Hansen 3070.

Chobe: on track to Sajawa, Smith 2627. Glanzi N.W. of

Malepolole, Story 4973. Kgalagadi: Kang, Mott 291. Kweneng:

between Gaberones and Thamaga, Revneke 329. Ngamiland:

Okavango Delta, Smith 387.

Transvaal. — Groblersdal: Loskop Dam to The Hell, Codd &

Verdoorn 10366b. Letaba: K.N.P. Van der Schijff 2779. Lyden-

burg: N.E. of Penge Mine, Leistner 3465. Nelspruit: Malelane

Camp, Acocks 16726. Pilgrims Rest: Olifants River bank, Pole

Evans H. 17021. Potgietersrus: on Limpopo River bank near

Usutu, Van Graan & Hardy 451. Waterberg: N.E. of Nylstroom,

Codd & Verdoorn 10435. Soutpansberg: E. of Wylies Poort,

Meeuse 9207 .

Cape. — Vryburg: near Malope River, Mostert 1244.

M. acuminata var. acuminata is sometimes con-

fused with M. forbesii, which has more or less the

same habit and partly the same distribution. These

taxa are mainly distinguished by the shape and length

of the epicalyx-bracts around the length of the style.

According to Wild in F.Z., 1, 2: 532 (1961), M.

velutina sensu Exell & Mendonca in Conspectus

Florae Angolensis 1, 2: 190 (1951) is partly M.

acuminata.

A note on Rodin 9291 from Kwanzama, South

West Africa, reads, ‘leaves are stamped and sap is

put on wounds to heal them’.

(b) var. agnosta (K. Schum.) Wild in Bolm Soc.

broteriana ser. 2, 33: 36 (1959). Type: Transvaal,

Makapaansberg, Strydpoort, Rehmann 5490 (K,

iso.!; PRE, photo.!). Fig. 5.

M. agnosta K. Schum. in Engl., Monogr. Afr. Pfl. 5: 1 1 ( 1900);

Burtt Davy, FI. Transv. 1: 261 (1926). Type as above.

M. ferruginea sensu Szyszyl., Polypet. Disc. Rehm. 139 (1887),

pro parte as to Rehmann 5490.

M. obtusaN.E. Br. in Kew Bull. 1906: 99(1906). Type: Zimbab-

we, near Bulawayo, Cecil 94 (K, holo.).

This variety differs from the typical in the leaves

being narrowly oblong or narrowly ovate-oblong and

the tomentum on the new growth being densely inter-

spersed with bunched, dark-brown hairs. The flowers

are smaller on the whole and usually clustered at the

apices of the branchlets. It agrees with the typical in

the characteristic abruptly-acuminate epicalyx-

bracts, the long style, general habit and the long

peduncles bearing 1 to 4 flowers at the apex.

Recorded from Botswana and the Transvaal. Also

occurs in Zambia, Zimbabwe and Mozambique.

Botswana. — Kanye: near Pharing, O. B. Miller B/586.

Transvaal. — Petersburg: Strydpoort, Makapaansberg,

Rehmann 5490.

A specimen collected in 1978 on the Pilanesberg,

Rustenburg District, Peelers, Gericke & Burelli 590

may be this variety although the leaves are somewhat

broader than in the type specimen.

Fig 5. — Melhania acuminata var. agnosta. Rehmann 5490, iso-

type in K.

8. Melhania transvaalensis Szyszyl., Polypet.

Thalam. Rehm. 138 (1887); K. Schum. in Engl.,

Monogr. Afr. Pfl. 5: 11 (1900); Burtt Davy, FI.

Transv. 1: 261 (1926). Type: Transvaal, Elandsriver

drift, Rehmann 4940.

Shrublet 14-60 cm tall, stems many arising an-

nually from the base, laxly branched, new growth

grey or off-w'hite tomentose, tomentum interspersed

with bunched hairs which are sometimes yellowish-

brown. Stipules subulate, tomentose, 4-14 mm long.

Leaves oblong, narrowly oblong or ovate-oblong,

usually more than twice as long as broad, 2-9 cm

long 0,7-3 cm broad, rounded to subtruncate at the

apex, more or less rounded at the base, margins fairly

distantly crenate dentate, upper surface with distinct

coarse stellate pubescence only (not tomentose with

fine bunched or stellate hairs), under surface w hite to

grey stellate-tomentose; petiole stellate-tomentose,

from about 5-20 mm long. Inflorescence axillary in

upper leaves usually 1 -flowered, stalk 1-2,5 cm long,

often with an articulation somewhere about midway

distinguishing the pedicel from the peduncle.

Epicalyx-bracts ovate-lanceolate, long acuminate,

subcordate at the base, 10-12 mm long, about 3,5

mm broad, stellate-tomentose on both surfaces.

Calyx with lobes lanceolate, slightly longer and nar-

rower than the epicalyx-bracts, 12-14 mm long, 3

mm broad, long acuminate, stellate-tomentose, with

clustered yellow-brown hairs interspersed on the

outer surface, subglabrous on inner. Petals yellow,

about 13 mm long. Stamens slightly shorter than the

ligulate 8 mm long staminodes. Ovary more or less

ovoid, stellate-tomentose; style 2, 5-3, 5 mm long,

branches 2,5 mm long. Capsule ovoid, about 12x8

mm, with off-white tomentum mixed with rather

long grouped, yellowish hairs.

Found on stony hills and grassy slopes on dolomite

formation. Recorded from restricted areas in the nor-

thern Cape and the Transvaal.

270

REVISION OF MELHANIA IN SOUTHERN AFRICA

Transvaal. — Petersburg: Chuniespoort, Code! & Verdoorn

10461. Potchefstroom: Oudedorp, Van Wvk 1473. Potgietersrust:

about 3,5 km NE of Potgietersrust, Codd & Verdoorn 10386.

Pretoria: Hennops River, Verdoorn 2468.

Cape. — Griqualand West: Knochbarragh, Bruckner 246.

A study of Rehmann’s route when he collected the

type of this species shows that the locality ‘Elands-

river Drift’ was most probably in the northern part of

the Pretoria District, north-east of Rust de Winter.

The herbarium sheet of Schlechter 4322 from the

Transvaal Museum, now in the National Herbarium,

Pretoria, has mounted on it a mixture of this species

and M. prostrata. Another sheet of the same

number, Schlechter 4322, which has been in the Na-

tional Herbarium for many years, is purely M.

trartsvaalensis, i.e. all the leaves are crenate-dentate

and coarsely stellate on the upper surface (not entire

with simple subappressed, long hairs or glabrous on

the upper surface).

In Engler’s Monographieen Afrikanischer Pflan-

zen, K. Schumann describes M. trartsvaalensis Szys-

zyl. as having the upper surface of the leaves ‘ap-

presse pilosis’ which describes those of M. prostrata.

It is difficult to explain this for he cites only Reh-

mann 4940, the type number. The sheet in the Kew

Herbarium of Rehmann 4940 has the leaves stellate-

pubescent on the upper surface as described by

Szyszylowicz, the author of the species.

9. Melhania rehmannii Szyszyl., Polypet.

Thalm. Rehm. 138 (1887)*, sphalm. rehmanii ; K.

Schum. in Engl. Monogr. Afr. Pfl. 5 — 10 (1900);

Burt Davy, FI. Transv. 1: 46 & 261 (1926): Wild in

F.Z. 1: 53 (1961); M. Friedrich et al., F.S.W.A. 84:

26(1969). Type: Transvaal, Klippan, Rehmann 5220.

M. griquensis H. Bol. in J. Linn. Soc., Bot. 24: 172 (28th Nov.

1887) excl. Orpen sub Bolus 6045 for the greater part. Type: Cape,

Asbestos Mtns in kloof, Burchett 2050 (K, holo.; PRE!).

M. rupestris Schinz in Bull. Herb. Boissierser. 2,2: 1008(1902);

M. Friedrich et al., F.S.W.A. 84: 26 (1969). Type: S.W. Africa.,

Oanop near Rehoboth Fleck 19a (Z).

Shrublet, low, usually under 30 cm tall, many

stems from a woody base, new growth tomentose

with ashy-grey, bunched, stellate and simple rather

long hairs. Stipules subulate, about 7 mm long,

usually turning a dark colour. Leaves ovate-oblong,

oblong or obovat e-oblong, 1-4 cm long, 0,8-2, 3 cm

broad, on both surfaces thickly ashy-grey stellate-

tomentose at first, becoming thinly so, more or less

truncate at the apex, rounded to broadly cuneate at

the base, coarsely crenate-dentate; petiole 0,3- 1,5

cm long. Inflorescence axillary, flowers usually

solitary on slender cernuous pedicels, 5-10 mm long.

Epicalyx-bracts broadly ovate about 5 mm long, odd-

ly shaped, usually with a short basal claw, apex acute

or rounded, ashy tomentose on both surfaces. Calyx

with lobes lanceolate, about 7 mm long and 2,5 mm

broad, dorsally tomentose pubescent, glabrous on in-

ner face except at the tip. Petals orange-yellow,

about 6 mm long, usually shorter than the calyx.

Stamens almost as long as the ligulate, 3 mm long

staminodes. Ovary subglobose, tomentose in upper

half; style about 2-5 mm long, branches about 1,5

mm long; ovules 2-3 in a cell. Capsule densely

tomentose with mixed pubescence.

Found in bushveld, sandy flats or rocky ridges in

dry country. Recorded from the north-western Cape,

Swaziland, the central and northern Transvaal, Bot-

*lt has not been possible to establish the exact date in 1887 when

the description of the species M. rehmannii Szyszyl. appeared, but

it is assumed that it would not have been later than 28th November

when M. griquensis H. Bol. was published.

swana and South West Africa. Also occurs in Zim-

babwe.

S.W. A. — Gobabis: Liebenberg 4627; Keetmanshoop: Brunau,

Farm Kochena, Giess & Mutter 11889.

Botswana. — Central District: N.E. corner of the Makarikari

Pan; O. West 9864. Ganzi District: gate between S.W. Africa &

Botswana, Mamuno, Brown Manuno 15. Ngamiland District:

Mawabu Pan, Buerger 1010. S.E. District: Boteti delta, N.E. of

Mopipi, Standish- White 15.

Transvaal. — Groblersdal: near Marble Hall Codd & Verdoorn

10870. Lydenburg: Sekukuneland, farm Lordskraal, Barnard &

Mogg 911. Pilgrims Rest: K.N.P., Acornhoek, Gertenbach 5413.

Potgietersrus: 21 km from Roedtan on road to Grass Valley, De

Winter 2222. Rustenburg: ± 5 km S.E. of Rooibokkraal, Leistner

3176. Soutpansberg: near Sandrivier, Schlechter 4590; Dongola

Reserve, Verdoorn 2112.

Swaziland. — Big Bend, Baytiss BS 2720.

Cape. — Beaufort West: ± 21 km S. of Layton, Acocks 23541.

Hay: Asbestos Mts, Burchett 2050; Postmasburg, Esterhuysen

881. Herbert: Christian’s Drift, Leistner 1998. Kimberley: Davis’s

Drift, ± 53 km W. of Kimberley, Leistner 1220. Kuruman: be-

tween Khosis & Maremane, Esterhuysen 2396.

As explained by Wild in Flora Zambesiaca, the

confusion between M. rehmannii and M. griquensis

(now pro parte under M. virescens) was initiated by

H. Bolus, the author of M. griquensis who, although

basing his description on the Orpen specimen from

Griquatown (hence griquensis), cited Burchell 2050

from the Asbestos Mountains as the holotype: he saw

this specimen at Kew in 1881 and mistakenly con-

sidered it to be conspecific with the Orpen specimen.

Subsequent taxonomists such as K. Schuman, Wild

and Burtt Davy all agreed that two taxa were invol-

ved, for although closely related, they were readily

distinguished by the epicalyx bracts which were

subulate to linear in the Griquatown specimen and

broadly ovate in Burchell’s specimen from Asbestos

Mountains. Following these taxonomists, it has

become necessary to relegate M. griquensis sensu

stricto to synonomy of M. rehmannii, and M. gri-

quensis (two parte as to the description and Orpen

specimen) to M. virescens (K. Schum.) K. Schum.,

the first published synonym.

For further distinguishing features and notes on

distribution see under M. virescens.

10. Melhania virescens (A-. Schum.) K. Schum.

in Engl., Monogr. Afr. Pfl. 5: 6 (1900): M. Friedrich

et al., in syn. in F.S.W.A. 84: 25 (1969). Type: South

West Africa, Walvis Bay, Otjatambi, Luderitz 117.

M. griquensis H. Bolus in J. Linn. Soc., bot. 24: 172 (28 Nov.

1887) partly as to Orpen sub Bolus 6045 for the greater part; —

var. virescens K. Schum. in Verh. bot. Ver. Prov. Brandenb. 30:

239 (1888). Type as for M. virescens.

M. bolusii Burt Davy. FI. Transv. 1: 46 & 261 (1926).

Small bushy shrublet with a perennial rootstock,

5-30 cm tall, annual growth arising from the base,

young branches grey tomentose. Stipules subulate,

5-10 mm long, usually pubescent with grey crisped

hairs. Leaves with the blade often silvery grey-

stellate-tomentose, the upper surface becoming thin-

ly so or glabrescent, lower surface densely grey

stellate-tomentose and sometimes with scattered

glands as well, oblong-elliptic, 2x1, 2-5 x 2,5 cm,

varying in texture from rather thin to rather thick,

subtruncate or subrounded at the apex, rounded to

somewhat cuneate at the base, margins crenate-

dentate, lateral veins prominent and running into the

lobes; petiole 5-15 mm long. Inflorescence axillary,

usually 1 -flowered, rarely 2-flowered, peduncle short

and straight 2-5 mm long, pedicels, if present, up to

2,5 mm long. Epicalyx-bracts linear-subulate, 5-10

mm long, shorter than the calyx-lobes, grey stellate-

tomentose on both surfaces. Calyx with lobes

lanceolate, acuminate, 8-14 mm long, 3 mm broad,

I. C. VERDOORN

271

grey tomentose dorsally. Petals bright yellow, 6-12

mm long, shorter than the calyx. Stamens slightly

shorter than the ligulate staminodes. Ovary globose,

densely stellate-tomentose; style 2-7 mm long,

sometimes with a twist, branches 0,5-1 mm long.

Capsule densely stellate-tomentose; seeds usually 4 in

a cell.

Found in limy soils in the northern Cape, the

Transvaal, Botswana and South West Africa.

S.W.A. — Gobabis: Witvlei, Basson 102. Grootfontein:

Gautscha’s Pan, Story 6224. Grunau: Gemsbokvlakte, Le Roux

663. Karibib: ± 4,3 km E. of Wilhelmstal, De Winter 2684.

Maltahohe: Buellspoort Mtn, Strev 2334. Okahandja: Bradfield

165. Okavango: Okosongomingo, Holzhammer 206. N.E. of Ot-

jiwarongo, De Winter 2832. Outjo: Outjo Townlands, De Winter

3040. Rehoboth: Kalkrand, Acocks 18162. Windhoek: Farm Gam-

mans, Wanntorp 122.

Botswana. — Kwening District: Takatshanane Pan, Wild 5089.

Nagamiland District.: Malie Rogoni, Cursor l 170.

Transvaal. — Potgietersrus, 12,2 km N. of Grass Valley on

road to Zebediela, De Winter 2237. Lydenburg: Farm Boskloof,

Fourie 5/98.

Cape.— Barkly West: ± 3,2 km W. of Borrelskop edge of Kaap

Plateau, Acocks 219. Griqualand West: Boetsap, Brueckner 286.

Hay: Floradale, Esterhuysen 2304. Kuruman: Cotton End,

Esterhysen 2170. Mafeking: Molopo Native Reserve, Peelers,

Cericke & Burelli 224. Taung: Thoming, Rodin 3423. Vryburg:

Taljaard Nature Reserve, MacDonald 77/22.

The confusion that existed between this species and

M. rehmannii is understandable for, until one has

recognized the features that distinguish them, these

species look very much alike. The specimen on which

Bolus based the greater part of his description of M.

griquensis, i.e. Orpen sub Bolus 6015, is a mixed

gathering. Four of the portions have the linear-

subulate epicalyx bracts as described by Bolus, and

the fifth, on the left hand side of the sheet, the ovate

bracts of M. rehmannii. the distribution of these two

species overlap in the northern Cape and parts of the

Transvaal. A further distinguishing feature is that the

peduncles in M. virescens are short (usually shorter

than the petiole) and suberect and the pedicels, if pre-

sent, are also short and suberect, whole in M.

rehmannii the peduncles are somewhat longer (about

as long as or longer than the petiole) and the pedicels,

if present, longer, thinner and cernuous.

With regard to the distribution, as mentioned

above, these species overlap in the northern Cape and

the Transvaal, but M. virescens extends further

westward in South West Africa. According to collec-

tors’ notes M. virescens is restricted to limy soil,

while M. rehmannii is not so restricted and is much

more widespread.

The specimens added by N. E. Brown to Bolus’s

original description of M. griquensis, namely Bur-

chell 2385, Holub s.n. Rehmann 5220 are all M.

rehmannii.

11. Melhania burchellii DC., Prodr. 1: 499

(1824); Harv. in F.C. 1: 222 (1860); Burtt Davy, FI.

Transv. 1: 260 (1926); Wild in F.Z. 1, 1: 530 (1960);

M. Friedrich et al. in F. S.W.A. 84: 24 (1969). Type:

Cape, Kuruman District, Burchell 247.

M. serrata Schinz in Bull. Herb. Boissier ser. 2, 2: 1007 (1900).

Type: Okahena, Dinter 499.

M. albicans Bak. f. in J. Bot. Lond. 39: 123 (1901); Burtt Davy

FI. Transv. 1: 260 (1926). Type: Transvaal, ‘Pilgrims Rest’ (more

likely near Buttons Kop, Pietersburg District), Greestock s.n. (K.

holo.; PRE, photo.!).

M. dinteri Engl, in Bot. Jb. 55: 350 (1919). Syntypes: S.W.

Africa, several specimens including Dinter 645 & 2705 from

Gobabis.

Plant subherbaceous, from a woody rootstock,

low, up to about 90 cm tall, branches many, erect

ascending new growth densely stellate tomentose

with long and short hairs, hairs sometimes foxy.

Stipules subulate 3-10 mm long, tomentose but soon

drying, curving and deciduous. Leaves elliptic-

oblong (on flowering branches) narrowly oblong to

linear oblong (on shoots), blade 3-10 cm long, 0,4-2

cm broad, shallowly to fairly distinctly toothed in the

upper half stellate-tomentose on both surfaces, with

short and long greyish hairs; petiole 3-17 mm long.

Inflorescens axillary, 2-several-flowered, sometimes

reduced to 1 flower, peduncle 1-6 cm long, pedicels

0,4-1 cm long; in some single-flowered in-

florescences the peduncle and pedicel are in-

distinguishable, in others the articulation is obvious.

Epicalyx-bracts linear-lanceolate to lanceolate, 7-11

mm long, up to 3 mm broad, the central usually the

broadest stellate-tomentose on both surfaces with

long and short hairs. Calyx with lanceolate lobes,

slightly longer than the epicalyx 10-15 mm long,

stellate-tomentose without, glabrous within. Petals

yellow about 5 mm long, ligulate staminodes about 7

mm long, staminal tube about 1,5 mm long. Style

2—4 mm long, branches about as long as the style.

Ovary tomentose with short and long off-white hairs.

Capsule sub-globose to oblong-ovoid, stellate-

tomentose with short and long hairs, up to 12 mm

long.

Found in hot, dry parts of the north-western Cape,

the Transvaal, Orange Free State, South West Africa

and Botswana. Also occurs in Zimbabwe.

S.W.A. — Gobabis: 80 km S.E. of Sandveld, Basson 98. Outjo:

Etosha National Park, Le Roux & Grobler 1238.

Botswana. — Ghanzi: 8 km N. of Kang, Wild 5052. Kgalagadi:

Kalahari Park, E. of Swartpan, Van der Walt 5748. Kweneng:

Khutse Reserve, Liebenberg 8986.

Transvaal. — Pietersburg: Commonage, Acocks & Hafstrdm

962. Potgietersrus: about 16 km NE of Potgietersrus, Codd & Ver-

doorn 10390.

O.F.S. — Hoopstad: Bultfontein, O’Connor 126.

Cape. — Barkly West: between Jacobs Rush & Sydney, Acocks

1556. Gordonia: Gemsbok National Park, Leistner 1007. Herbert:

near Jacobs Rush, Acocks & Hafstrdm H.1016. Kuruman: Cotton

End, Esterhuysen 2237. Vryburg Tosca, Peelers, GerickeA Burelli

236.

In the Transvaal this species is very common be-

tween Potgietersrus and Pietersburg. The form which

occurs here agrees with the description of M. albi-

cans, but no grounds can be found for separating it

from M. burchellii in spite of the disjunct distribu-

tion and the difference in habitat. In the Transvaal it

grows on a heavy gravelly loam soil which is very dif-

ferent from the loose sand in which it is found in the

north-western Cape. It may be noted that the type

locality of M. albicans is given as Pilgrims Rest. The

Rev. William Greenstock, who collected the original

specimen, spent some time with Edward Button at

Eersteling Mine, between Potgietersrus and Pieters-

burg, before proceeding to Pilgrims Rest. The plant

has not been found at Pilgrims Rest again and it is

considered that Greenstock collected this specimen in

the neighbourhood of Eersteling. Several similar

cases are known where species based on Greenstock

specimens have been attributed to Pilgrims Rest, but

subsequent investigation has shown that they were

without doubt collected in the Eersteling area.

In the Flora Capensis under M. burchellii Miss

Owen’s specimen is cited as coming from ‘Zooloo

Country’; this is probably another case of the several

instances where Miss Owen’s specimens, collected

after she had left Zululand (that is, after the Dingaan

massacre), were labelled as from Zululand, whereas

they probably were from the Kuruman or Zeerust

Districts.

272

REVISION OF MELHANIA IN SOUTHERN AFRICA

Regarding another citation in the Flora Capensis,

namely Zeyher from the ‘Aapies River’, it is doubtful

whether this locality is correctly cited, since this

species has not been found near Pretoria.

Although none of the syntypes of M. dinteri has

been seen, the description best fits the specimens in

this taxon.

12. Melhania damarana Harv. in F.C. 2: 590

(1862); M. Friedrich et al., F.S.W.A. 84: 25 (1969).

Type: South West Africa, Damaraland, Elliott s.n.

M. ovata var. oblongata K. Schum. in Engl., Monogr. Pfl. 5: 7

(1900), pro parte as to Marioth 1314.

Shrublet, 30-60 cm tall, with 1 or more erect or

suberect stems from a woody base; branchlets long,

erect or ascending, new growth stellate-tomentose

with minute reddish stellate scales obvious in parts.

Stipules subulate 5-10 mm long. Leaves variable in

texture, shape and size, broadly to narrowly ovate-

oblong or narrowly oblong, 2-7 cm long, 1-4,3 cm

broad, from irregularly and coarsely crenate to finely

and regularly toothed, rounded or obtuse at the base

and sometimes obscurely cordate at the point of at-

tachment, broadly rounded or slightly narrowing to

an obtuse or acute apex, densely to subdensely

stellate-tomentose or finely stellate-pubescent on

both surfaces; petiole 1-2,5 cm long. Inflorescence

axillary, 1-3-flowered; peduncle straight, persistent,

1,5-4, 5 cm long, pedicels 5-1,5 cm long. Epicalyx-

lobes lanceolate to broadly lanceolate-acuminate,

acute, 7-11 mm long, 2,5—3 mm broad, tomentose

with grouped hairs on both surfaces. Calvx-lobes

lanceolate-acuminate, acute or narrowed into a

subulate apical portion, 8-16 mm long, 3-3,5 mm

broad, tomentose on the outer surface. Petals yellow,

sometimes pale yellow, almost as long as the calyx or

slightly longer, 7-20 mm long. Stamens with

filaments about 2 mm long, anthers 5 mm long,

ligulate, staminodes 6-11 mm long. Ovary sub-

globose, stellate-tomentose with short and long,

creamy to pale yellow hairs; style 2-7 mm long,

branches, 2-3 mm long; ovules 6 in a cell. Capsule

globose to oblong-globose, stellate-tomentose, 5-12

mm long.

Occurs throughout the length of western South

West Africa with one record to date from the nor-

thern Cape. Found mostly in dry conditions on the

western side of the plateau at the Namib fringe.

S.W.A. — Kaokoland: W. of Etanga, De Winter & Leistner

5411. Karasburg: Mundis, Auret 5601. Luderitz: farm Weissen-

born, Kinges 2377. Maltahohe: Abendruhe, Oliver, Muller &

Steenkamp 6568. Omaruru: Numas Valley, Wiss 1415. Outjo: Ot-

jiwarongo, Giess 8524. Rehoboth: Buffelspoort, Tolken & Hardy,

668. Tsumeb: 6 km S.W. Tsumeb Giess 8663. Walvis Bay: foot-

hills Erosgebirges, MerxmuUer & Giess 3566.

Cape. — Hay: Rietkloof, Acocks 8520.

This species approaches M. suluensis from eastern

Natal. For main distinguishing features see the key,

p. 264.

M. damarana also resembles forms of M. ovata

auct. found in tropical Africa with the type from

South America, the principal distinguishing feature

being the shape of the epicalyx-bracts, which are

linear-subulate in M. ovata and lanceolate to broadly

lanceolate in M. damarana.

13. Melhania suluensis Gerstner in J1 S. Afr.

Bot. 12: 37 (1946). Type: Natal, Eshowe, near

Dhlangubo store, Gerstner 2888 (BOL, holo.).

Suffrutex, main stem erect with many slender

branches (no new growth from the base), 60-90 cm

tall, new growth off-white tomentulose Stipules

subulate, about 3 mm long, persistent. Leaves com-

paratively thin-textured, broadly ovate to ovate-

oblong, sometimes broadly oblong, 1-4,5 cm long,

0,7-2, 5 cm broad, very shortly stellate-tomentose on

both surfaces, lower surface silvery-grey and

sometimes with reddish brown stellate hairs from

lepidote scales scattered on the raised nerves, upper

surface darker and more uniformly stellate-tomentu-

lose, rounded to shallowly cordate at the base,

broadly rounded or rarely subacute at the apex, mar-

gins shallowly crenate-dentate; petiole 0,5-1, 5 cm

long. Inflorescence axillary 1-2-flowered, peduncles

slender, 0,5-2 cm long, pedicels 0,3-1, 2 cm long, in

single-flowered inflorescence the peduncle and

pedicel are indistinguishable and up to 2 cm long.

Epicalyx-bracts lanceolate-acuminate to ovate-lan-

ceolate (not broadest at the base), 6-9 mm long,

2, 5-3, 5 mm broad, felted on both surfaces with

short off-white stellate tomentum. Calyx with lobes

lanceolate-acuminate, about 9 mm long and 3 mm

broad, dorsally stellate-tomentulose. Petals lemon-

yellow, about 10 mm long, 5-8 mm broad at the

apex. Stamens with filaments about 0,5 mm long,

anthers about 2,5 mm, united portion, 1 mm long,

staminodes ligulate, about 5 mm long. Ovary subglo-

bose, stellate-tomentose; style 3-7 mm long, bran-

ches about 1,5 mm long; ovules about 5 in a cell.

Capsule subglobose, about 6x6 mm, densely and

shortly stellate-tomentose.

Found in alluvial soil or on rocky slopes in dry

bushveld at altitudes of 3-170 m. Recorded from the

coastal area of Natal.

Natal. — Camperdown: Umgeni Valley, Forbes 1259 (NH).

Entonjaneni: Nogeya, Venter 3728. Eshowe: near Dhlangubo

store, Codd & Verdoorn 10174. Hlabisa: HluhluweGame Reserve,

Scott-Smith 71. Ngotshe: Itala Nature Reserve, Brown & Shapiro

445. Ubombo: Mkuze Poort, Ward 4071.

M. suluensis is closely related to M. damarana,

which occurs in South West Africa, west of the

escarpment. Besides this difference in distribution

the main distinguishing features are given in the key

on p. 264.

ACKNOWLEDGEMENTS

When reviewing the genus Melhania in 1962, I was

fortunate in being able to enlist the help of Dr O.A.

Leistner, who was the South African Liaison Officer

at the Royal Botanic Gardens, Kew: he kindly ex-

amined specimens for me at Kew and the British

Museum (Natural History) and solved the problems

submitted to him. I am indeed grateful to Dr Leistner

for his assistance.

I am also indebted to Prof. H. Wild of the Univer-

sity of Zimbabwe and Mr R. B. Drummond of the

Salisbury Herbarium, Zimbabwe, for advice and the

loan of specimens of Melhania randii.

The loan of herbarium material from Rhodes

University and the University of the Orange Free

State was much appreciated.

UITTREKSEL

Die 13 spesies van Melhania in suidelike Afrika is

hersien. Een van die probleme wat opgelos is, is ’n ou

een betreffende die identiteite van M. rehmannii

Szyszyl. en M. griquensis H. Bol. M. griquensis, in

die enge sin, word in die sinonimie van M. rehmannii

geplaas.

I. C. VERDOORN

273

INDEX

Page

Melhania Forssk 263

acuminata Mast 268

var. acuminata 268

var. agnosia (A\ Schum.) Wild 269

albicans Bak. f 271

bolusii Burtt Davy 270

burchellii DC 271

damarana Harv 272

didyma Eckl. & Zevh 264

didyma var. linearifolia (Sond.) Szyszyl 265

dinteri Engl 271

forbesii Planch, ex Mast 267

griquensis H. Bol. s.str 270

griquensis H. Bol. p.p 270

var. virescens K. Schum 270

integra Verdoorn 266

Page

leucantha E. Mey 264

linearifolia Sond 265

obtusa N.E.Br 269

ovata spreng. var. oblongata K. Schum 272

polygama Verdoorn ! 265

proslrata DC 265

randii Bak. f. 266

rehmannii Szyszyl 270

rupestris Schinz 270

serrata Schinz 271

serrulala R. E. Fr 267

suluensis Cerstner 272

transvaalensis Szyszyl 269

virescens (A". Schum.) K. Schum 270

Via Ha macrophylla Vis 264

.

Bolhalia 13, 3 & 4: 275-276 (1981)

The genus Waltheria in southern Africa

I. C. VERDOORN*

ABSTRACT

Waltheria indica L., the only species of Waltheria represented in southern Africa, is revised. This species, which

occurs throughout the tropics and substropics of the world, is found abundantly in the northern Cape, Swaziland,

northern Natal, Transvaal and northwards through South West Africa/Namibia and Botswana. Thoughout its wide

distribution the species is uniform. A scrutiny of herbarium specimens revealed that what appeared as a distinct

species or subspecies was without doubt an abnormality, probably caused by insect injury.

RESUME

LE GENRE WALTHERIA EN AFRIQUE AUSTRALE

Waltheria indica L, la seule espece de Waltheria representee en Afrique australe, est revisee. Cette espece qui se

trouve d’un bout a I’autre des tropiques et sub-tropiques du tnonde, est trouvee en abondance au Cap septentrional,

au Swaziland, au nord du Natal, au Transvaal et vers le nord, a trovers le Sud-Quest africain/ Namibia ainsi qu’au

Botswana. Malgre sa distribution Ires dispersee, I’espeee est uniforme. En scrutant les specimens des herbiers, il se

revele que ce qui apparaissait comme des especes ou sous-especes distinctes est sans aucun doute une anomalie prob-

ablement causee par degats d’insecte.

WALTHERIA

Waltheria L., Sp. PI. 673 (1753); Gen. PI., ed. 5:

304 (1754); Harv. in F.C. 1: 180 (1860); Benth. &

Hook, f., Gen. PI. 1: 224(1862); Wild in F.Z. 1: 536

(1960); M. Friedrich et al. in F.S.W.A. (1969); R. A.

Dyer, Gen. 1: 365 (1975).

Lophanthus Forst., Char. Gen. 27, t. 14 (1776).

Astropus Spreng., Neue Entd. 3: 64 (1822).

Shrublets, subherbaceous at first, pubescent with

stellate, tufted and simple hairs. Leaves simple,

crenate-dentate, petioled, stipulate. Inflorescence

axillary and terminal, flowers in cymes, often conges-

ted, or heads of flowers racemously or paniculately

disposed. Bracts often present. Calyx 5-lobed. Petals

5, marcescent. Stamens 5, united at the base; anther-

cells parallel. Ovary 1-celled; ovules 2 anatropus;

style somewhat excentric, clavate or fimbriate at the

apex. Capsule usually 1-seeded; seed ascending, en-

dospermous; embryo straight.

Found in the tropics and subtropics of both the

east and the west with the greatest concentration of

species in South and Central America. Of the 67

known species only 7 occur in the Old World and one

of these is common to the Old and New World. This

common species, W. indica, is the only species found

in southern Africa.

The generic name was given in honour of Augustin

F. Walther of Leipzig, a contemporary of Linnaeus,

remembered especially for his botanic garden.

Waltheria indica L., Sp. PI. 673 (1753); R. Br.

in Tuckey, Narrat. Esp. river Zaire, App. 5: 484

(1818); Mast, in FI. Brit. India 1: 374(1874); Harv. in

F.C. 1: 180 (1860) Wild in F.Z. 1: 536 (1960);

F.S.W.A. 84: 28 (1969). M. K. Scott in Bothalia 12

452 (1978). Type: India, Linn. Herb. 852.2.

W. americana L., Sp. PI. 673 (1753); Mast, in F.T.A. 1: 235

(1868); K. Schum. in Engl., Monogr. Afr. Pfl. 5: 45 (1900); Burtt

Davy, FI. Transv. 1: 268 (1926). Type: America, Linn. Herb.

852.1. — var. indica (L.) K. Schum., l.c. 547 (1900). Type as for

W. indica L. — var. subspicata K. Schum., l.c. 547 (1900). Types:

several syntypes including Schinz s.n., South West

Africa/Namibia.

* c/o Botanical Research Institute, Department of Agriculture

and Fisheries, Private Bag X101, Pretoria, 0001.

Shrublet, subherbaceous at first, stems slender,

hirsute, with stellate, tufted or simple bulbous based

hairs, erect to spreading and bushy, 30 cm to 1 m tall,

sometimes taller. Stipules linear-acute, 3-10 mm

long. Leaves petioled; blade ovate, ovate-oblong,

oblong to narrowly oblong, 2-10 cm long, 1-4,5 cm

broad, crenate-dentate, rounded, subtruncate or

broadly subacute at the apex, rounded, broadly

cuneate or cordate at the base, nerves prominent

beneath, laxly to densely stellate pubescent or hirsute

with bulbous-based hairs on both surfaces; petiole

hirsute, 0,6-3, 5 cm long. Inflorescence axillary,

cymose, usually tightly congested rarely (abnormal)?

sublax, sometimes heads of flowers appear to be

racemosely or paniculately disposed because the sub-

tending leaves are much reduced or absent. Bracts

usually 3 to a flower, linear acute, 3,5-5 mm long,

differing slightly in width but mostly under 1 mm

wide, dorsally hirsute. Calyx hirsute, 10-nerved; tube

turbinate, about 2,5 mm long; lobes deltoid acute to

deltoid acuminate, about 2 mm long. Petals yellow,

‘orange yellow’, turning reddish brown at maturity,

oblong-cuneate to spathulate, about 4 mm long, 1 ,25

mm broad, dorsally sparsely hairy, adhering at the

base to the stamen base, deciduous from the base but

persisting for some time around the capsule, the

lower portion held within the calyx-tube. Stamen op-

posite the petals, shorter than the full grown petals;

filaments membrane-margined at the base or, more

usually, almost to the apex, more or less united along

the membrane edge, the apical unwinged filiform

portion shortly and sparsely pubescent and at certain

stages contorted; anthers erect or (especially in ab-

normal specimen) horizontal, cells parallel, subacute

at both ends. Ovary 1-celled, obovoid slighly flatten-

ed, broadest at the apex, densely hirsute in upper

portion; style excentric, about 3 mm long, sometimes

rather contorted, fimbriate at the apex; ovules 2,

ascending, anatropus. Capsule thin-walled, about 4

mm long, almost 2 mm broad at the apex, hirsute on

upper portion, 1-seeded or rarely the second ovule

developing as well; seed obovoid; embryo straight.

Found in open grassland, on rocky slopes, along

rivers, in waste places, and rarely in woodland.

Recorded from the northern Cape and Natal north-

wards through the Transvaal, South West Africa/

Namibia and Botswana to tropical Africa. Also oc-

276

THE GENUS WAL THERIA IN SOUTHERN AFRICA

curs elsewhere in the tropics and subtropics of the

Old and New Worlds.

S.W. A. /Namibia. — East Caprivi: Katima Mulilo, Killick cS

Leistner 3058. Gobabis: Farm Amasib, Merxmiiller 1088. Groot-

fontein: Amrib, Schoenfelder S.554. Karibib: Farm Ameib,

Kinges 3306. Kavango: Tsotzana, Giess 9958. Okahandja:

Quickborn, Bradfield 85. Owambo: Oshigambo, Sylvi Soeni s.n.

or near Ruacana, De Winter & Giess 7084. Tsumeb: 50 km N. of

Tsumeb, Goldblatt 1938.

Botswana. — Central distr. : Dikolodi East, Kerfoot 8015.

Chobe: Lower Ngwezumba, Miller 13187. Ganzi: Mamono Ridge,

Mason & Boshoff 292. Kgalegadi: 16 km N. W. of Tsabong,

Leistner 3069. Kweneng: 412 km N. W. of Molepolole, Story 4974.

Ngamiland: Boro Floodplain, Biggs M. 588. Ngwaketse: Plaring,

Miller B. 865. South East: 8 km N. of Gaborone, Cox 5059.

Transvaal. — Barberton: Tonetti, Thorncroft 76. Carolina:

Nelsberg, Taylor 1990. Groblersdal: Marble Hall, Pienaar 519.

Letaba: Westfalia Estate, Scheepers 302. Lydenburg: Farm

Sterkspruit, Galpin 12174. Marico: Lekkerlach, Louw 251.

Nelspruit: near Pretoriuskop, Codd & De Winter 4917.

Pietersburg: Shilovane, Junod 4891. Pilgrims Rest: Sabie River,

Van der Schijff 41. Potgietersrust: Percy Fyfe Nature Reserve,

Huntley 1679. Pretoria: N. of Pretoria at turn off te Rust de

Winter Dam, Tolken 1224 (partly abnormal). Rustenburg: 3,2 km

S. of Rooibokkraal P. O., Leistner 3210. Soutpansberg: Tokwe,

Dreyer in TRV 21616 (partly abnormal); S. of Punda Milia,

Schlechter 9291. Waterberg: Nylstroom, Repton 530. Witbank:

Mapochsdrift, Du Plessis 411.

Swaziland. — Hlatikulu: Big Bend, Compton 30278. Manzini:

Sipofaneni, Compton 29713.

Natal. — Entonjaneni: Melmoth, Gerstner 4306. Hlabisa:

Hluhluwe Game Reserve, Ward 1855. Ingwavuma: Ndumo Game

Reserve, Hancock 7. Lower Tugela: Mandeni, Edwards 1360.

Lower Mfolozi: Mfolozi Game Reserve, Ward 4614. Marico:

Motswedi, Peeters, Gericke & Burelli 478. Ngotshe: I tala Nature

Reserve, Brown & Shapiro 309. Ubombo: Mkuze, Galpin 13706.

Cape. — Hay: Rietkloof, Acocks 8518. Kuruman: Seremoneng,

Acocks 2280. Mafikeng: ‘Ferndale’, Brueckner 451. Postmasburg:

Nchwaneng, Leistner 2205. Vryburg: Vryburg, Mogg 8889.

This species, which is widely spread in all tropical

and subtropical regions of the world, is fairly

uniform in its South African localities. Characterized

by small yellow flowers, with hirsute calyces, con-

gested in axillary glomerules, it is readily recognized.

Sometimes the glomerules or ‘heads’ of flowers ap-

pear to be racemosely arranged or in panicles. This is

because they grow on slender side branches with the

subtending leaves either much reduced or absent.

Among the many specimens examined, there is one

from the northern Transvaal in parts of which, the

major part, the flowers are in sublax cymes up to 6

cm long. This at once gives the specimen a different

aspect. It was found, too, that the parts of the

flowers also differed from the normal. The filaments

were not membrane-margined for most of their

length, but only at the very base if at all, the anthers

were horizontal, not erect, and the ovary somewhat

acuminate to the oblique apex instead of broadest

and more or less truncate at the top. This gave the

impression that a distinct species was involved. But

right at the base of the specimen on the same branch

was a typical, congested cyme and the flowers on it

are typical. What caused the abnormality and why it

should be accompanied by a difference in flower

structure remains a problem, but it may be due to in-

sect injury.

Waltheria indica L. and W. americana L. were

published simultaneously in 1753. Since R. Brown

was apparently the first to unite the two taxa under

the name W. indica, his choice must be followed

(Art. 57 of ICBN, 1978) and this is so done in this

treatment.

ACKNOWLEDGMENTS

1 am indebted to members of the staff of the

Botanical Research Institute for their very helpful

assistance with dissections, drawings and excursions

to examine the species in the field.

UITTREKSEL

Die enigste spesie van Waltheria wat in suidelike

Afrika verteenwoordig is, word hersien. Hierdie

spesie wat dwarsdeur die tropiese en subtropiese dele

van die werelcl voorkom, word dikwels in die

noordelike Kaap, Swaziland, noordelike Natal,

Transvaal en noordwaarts dear Suid-Wes Afrika/

Namibia en Botswana aangetref. Hierdie spesie

varieer min dwarsdeur sy wye verspreidingsgebied.

’n Ondersoek van herbariumeksemplare het ge-

toon dat sekere monsters wat gelyk het na verteen-

woordigers van ’n afsonderlike spesie of subspesie in

werklikheicl sonder twyfel ’n abnormaliteit vertoon

wat dear insekbeskadiging veroorsaak is.

Bothalia 13, 3 & 4: 277-279 ( 1981 )

The genus Cola in southern Africa

I. C. VERDOORN*

ABSTRACT

The two species of Cola that occur in southern Africa are reviewed. C. natalensis Oliv. is endemic, but C. green-

way i Brenan has mainly a tropical east African distribution with the eastern Transvaal and northern Zululand speci-

mens being the most southerly records. A study of the available material from tropical east Africa and southern

Africa supports the view that C. microcarpa Brenan is synonymous with C. greenwayi.

RESUME

LE GENRE COLA EN AFR1QUE AUSTRALE

Les deux especes de Cola que Eon trouve en Afrique auslrale sont revisees. C. natalensis Oliv. esl endemique,

mais C. greenwayi Brenan a surtout une distribution tropicale Est africaine avec les specimens du transvaal oriental

el du Zululand septentrional etant ceux qui ont ete enregistres comme les plus meridionnaux. Une etude du materiel

disponible d’Afrique tropicale de I’Est et de V Afrique australe soutient /’opinion que C. microcarpas Brenan est

synonyme de C. greenwayi.

COLA

Cola Schott & Endl., Melet. Boi. 33 (1832); Benth.

& Hook, f., Gen. PI. 1: 218 (1862); K. Schum. in

Engl., Monogr. Afr. Pfl. 5: 1 10 (1900); Wild in F.Z.

1: 558 (1961); nom. cons. (LC.B.N. 1956); R. A.

Dyer, Gen. 1: 365 (1975). Type species: C. acuminata

(Beauv.) Schott & Endl.

Trees dioecious or monoecious, or occasionally

with bisexual flowers. Leaves alternate, entire or lott-

ed, petiolate; petioles often with a swollen apical por-

tion (pulvinus). Inflorescence cymose or flowers in

the axils of the leaves, or on branchlets between the

leaves, solitary or clustered, sometimes on old wood;

unisexual, occasionally bisexual. Calyx 4— 5(— 6) lob-

ed Petals absent. Male flower: stamens usually 10,

united into a column, bearing sessile anthers in 1 or 2

rings around the apex, vestigial carpels often sunk in

the top of the androphore column. Female flower:

carpels (3)— 4— 5 (-10), cohering at first with a ring of

rudimentary stamens at the base; ovules several to

many per carpel; styles as many as the carpels. Fruit

splitting into 4-5 carpels or sometimes 1 or more

aborting. Seeds exendospermous; cotyledons thick, 2

or more.

Species over 100, all African. Two species occur in

South Africa.

KEY TO SPECIES

Petiole with apical swollen portion elabrescent; fruit obovate, up to about 4 cm long, rind hard and thick

I . C. natalensis

Petiole with apical swollen portion tomentose; fruit subglobose to oblique ellipsoid, 1 x 1,3-1, 8x2 cm, rind at

maturity rather thin and brittle 2. C. greenwayi

1. Cola natalensis Oliv. in Hooker’s Icon. PI.

14: 70, Plate 1390 (1882); K. Schum, in Engl.,

Monogr. Afr. Pfl. 5: 114 (1900); Sim, For. FI. Cape

Col. 144, Plate 17, f.3 (1907). Type: Inanda, Natal,

Medley Wood 32 1 .

Tree 3-15 m high, evergreen, monoecious or

dioecious, ‘bark flaking off in small patches’;

branchlets glabrescent. Stipules early caducous, not

seen except on seedlings, subulate-acuminate, ± 6

mm long. Leaves simple, entire; blade obovate-

elliptic, usually narrowly so, shortly and broadly nar-

rowing to, or acuminate at, the apex, narrowing

from above the middle to the base, 7-20 cm long,

1-6,5 cm broad, glabrous at maturity, young leaves

sparsely stellate with short hairs, midrib prominent,

especially on under surface, with 7-12 or more

distinct lateral nerves; petioles 1-4 cm long, with a

slight thickening (pulvinus), 5-10 mm long, at the

apex; pulvinus glabrescent. Inflorescence 1 -flowered,

axillary, solitary or fascicled, or from below leaves

and on old wood, usually clustered on a much ab-

*c/o Botanical Research Institute, Department of Agriculture

and Fisheries, Private Bag X101, Pretoria, 0001.

breviated shoot, 2 to several in a cluster; bracts at

base, 1,5-2 mm long, 1,5-2 mm broad, early

deciduous; pedicels 3-7 mm long, up to 1 mm diam.,

articulated about midway, and with scars of 1-2

bracts; densely stellate-tomentose at first, hairs

short, less than 0,5 mm long, light or dark brown.

Calyx 5-lobed almost to base, densely brown stellate-

pubescent without; lobes 4-8 mm long 1,5-4, 5 mm

broad, lepidote within and with a few scattered

brown stellate hairs in upper half. Male flower with

staminal-tube glabrous or minutely stellate in basal

portion, 1,5-5 mm long, anthers 8 in 1 row at the

apex, cells parallel, 1 mm long, rudimentary ovary

and 5 styles in anther ring. Female flower with sessile

ovary, 4-5 carpellate, about 2 mm diam., with a ring

of rudimentary stamens round the base; styles 5,

sometimes fewer, with club-shaped recurved,

stigmas; ovules 2 or 3 in each cell. Carpels 4—5, or by

abortion fewer, obovate-globose, about 4 cm long, 3

cm diam., densely microscopically scaly and stellate-

pubescent, hairs very short or worn off, rind hard

(often tuberculate through insect activity), ripe fruit

orange to salmony yellow, glutinous inside; seeds 2,

or sometimes 3, cotyledons thick, creamy white and

THE GENUS COLA IN SOUTHERN AFRICA

Fig 1 • Cola natalensis 1-6. I, flowering twig with male flowers, x 0,8 ( Ward 5623); 2, male flower, x 5 (Ward 5622); 3,

androphore with front stamen removed to expose the rudimentary ovary, x 12 ( Ward 5622); 4, female flower, x 5 (Moll

3211); 5, fruiting twig, x 0,8 ( Ward 2623); 6, longitudinal section of carpel, X 0,8 (Ward 2623). C. greenwayi 7-8.

7, fruiting twig with comparatively smaller carpels, x 0,8; 8, longitudinal section of carpel showing the comparatively

thin rind, x 0,8. (7 & 8, Garland sub PRE 47565.)

I. C. VERDOORN

279

deep pinkish red tinged on inner face and with resi-

nous ducts, in part minutely stellate-pubescent. Fig. 1 .

Found in dense forest along coast in the Transkei

and Natal as far north as Mtunzini.

Transkei —Lusikisiki: Intafufu, Campbell, sub Sim 2627. Port

St Johns: Noxolweni Forest, Mogg 13066; St Johns River, Pole

Evans H. 18047.

Natal. — Durban: Bluff (?), Medley Wood s.n; Inanda: Medley

Wood 1500; Umhlanga, Marais 797; Story 4152. Watmough 443;

Cheadle, Johnson & Wells 686; Moll 1810. Mtunzini: Ngoya

Forest, Edwards 121.

The fruit is said to be inedible. Wood durable.

Related to the tropical species C. clavata, but differs

principally in the large, less numerous flowers. In C.

clavata, the flowers arise from numerous abbreviated

shoots on the old wood as well as in the axils of the

leaves. It differs from C. mossambicensis, the other

species that occurs in Mozambique, mainly in the tex-

ture of the leaves, the subglabrescent fruits and the

glabrescent pulvinus.

2. Cola greenwayi Brenan in Kew Bull. 1956: 147

(1956); Wild in F.Z. 1: 560 (1960); Drummond in

Kirkia 10: 260 (1975); Palgrave, Trees of Sth. Afr.

598 (1977). Type: Tanzania, Mkuzi, Greenway 7891

(K, holo.!).

C. microcarpa Brenan in Kew Bull. 1956: 147 (1956); Palmer &

Pitman, Trees of Sth. Afr. Vol. 2, 1491 (1972). Type: Tanzania,

Turiana, Sensei 1466 (K, holo.!).

Tree 3-20 m high, monoecious or dioecious;

young branchlets russet to light cinnamon brown or

grey tomentose at first, mixed with stellate or bunch-

ed hairs which fall readily. Stipules very soon falling,

subulate-lanceolate to linear, 3-6 mm long, dark to

light-brown or grey tomentose. Leaves simple, entire,

blade elliptic to narrowly obovate-elliptic, narrowing

to base and apex, 4-15 cm long, 1,4-5 cm broad,

glabrous at maturity, midrib slender, prominent on

both surfaces, with usually 7-18 main lateral veins,

distinct to prominent on lower surface; petiole

0,5-5, 5 cm long with a thickened pulvinus at the

apex, tomentose throughout, with some long hairs as

well, becoming glabrous except for the pulvinus

which is persistently tomentose. Inflorescence

1 -flowered, axillary, solitary or fascicled and crowd-

ed on branchlets between the leaves; flowers unisex-

ual, rarely some bisexual, apparently dioecious (may

be monoecious), arising from sessile imbricating

bracts which are 1,5-3, 5 mm long, densely stellate-

tomentose without subpersistent; pedicels about

7-20 mm long, (sometimes obscurely articulate),

about midway, densely stellate-pubescent, hairs dark

to light brown, short or long (as long as the pedicel is

broad). Calyx 4-6-lobed almost to the base at

flowering time, dorsally stellate-pubescent; lobes

from about 5-10 mm long, 2, 5-3, 5 mm broad, inner

face stellate-pubescent at least in upper half lepidote

below. Petals 0. Male flower with staminal tube

about 2, 5-4, 5 mm long, densely pubescent to

subglabrous; anthers up to 8 in one row, sessile in a

ring around the apex of the staminal-tube, rudimen-

tary carpels and styles in the centre of the ring.

Female flower with ovary sessile, densely stellate-

pubescent, about 2 or 3 mm diam., with a ring of

rudimentary stamens around the base, carpels 4-5

(3?); styles arising laterally with clavate recurved,

papillose stigmas. Carpels 4-5 or less by abortion,

orange yellow to deep orange with touch of vermilion

when ripe, obliquely subglobose obliquely ellipsoid

or obovate, 1 x 1,3 cm to 1,8x2 cm rounded on top

with the remains of the style-base lateral and

obscure, or forming a mucro, stellate pubescent in

parts (rubs off easily) with dark to light brown hairs,

long bristles, rind at maturity rather thin and brittle;

seeds 1-2 (coat like that of fruit); cotyledons 2 (one

smaller than the other, reddish pink tinge between

them in fresh fruit); plumule stellate-pubescent.

Found in dense woodland or forests, ranging from

sea level to steep mountain slopes at altitudes of

1 100 m or more. Recorded from northern Natal and

eastern Transvaal. Also occurs in Mozambique and

northwards in tropical east Africa.

Transvaal. — Barberton: Bearded Man Mountain, farm

Duurstede Buitendag 1 1 10; Twello Forest Estate, ± 12kmS.E.of

Barberton on road to Flavelock, Arnold 1621; Pedlar’s Bush, Bui-

tendag & Kruger 0000.

Natal. — Fllabisa: False Bay, Ward 3664, 3676; near Charters

Creek, Rochat 14; Hluhluwe Game Reserve, Ward 2966; Duku-

duku Forest Strey 5590. Ingwavuma: Lake Sibayi, Vahrmeijer

1080; Garland 5801. Ubombo: Lake Sibayi, Vahrmeijer 720.

When describing the two species, C. greenwayi and

C. microcarpa, Brenan (1956) mentioned that the dif-

ferences between them are small but the altitudes at

which they grow differ considerably, C. microcarpa

occuring at about 460 m, whereas C. greenwayi is

found at altitudes between 1 600 and 2 000 m. Drum-

mond in Kirkia (1975) sank C. microcarpa under C.

greenwayi without comment. Drummond is followed

here because, whereas the small differences noted

between these two species are merely of degree, such

as the colouring and length of the pubescence, the

comparable size of the fruit and the mucro being

more obvious on the carpels of one than on the other,

the features considered to be diagnostic are the same.

These features are the general size and shape of the

flowers and the carpels and the brittle rind of the ripe

fruit. According to collectors’ notes, both species are

at times, and in small patches, expecially on the

ripening carpels, suffused with ‘reddish crimson’ or

‘orange vermilion’, but this is not obvious on herb-

arium specimens.

The presence of a species of Cola in the Barberton

District, probably the southernmost limit of the

genus, has only recently (1976) come to light. This is

surprising, since that is an area particularly well col-

lected as shown by the large Thorncroft collection.

The probable explanation is the inaccessible nature

of the forest patches on the steep mountain slopes.

UITTREKSEL

Die twee C ol a-spesies wat in suidelike Afrika voor-

kom word hersien. C. natalensis Oliv. is endemies,

maar C. greenwayi Brenan kom hoofsaaklik in

tropiese Oos- Afrika voor met eksemplare uit Oos-

Transvaal en Noord-Zoeloeland as die mees suidelike

verspreidingsrekords. ’n Studie van die beskikbare

eksemplare uit tropiese Oos-Afrika en suidelike

Afrika staaf die opvatting dat C. microcarpa Brenan

’n sinoniem van C. greenwayi is.

Bothalia 13, 3 & 4: 281-306 (1981)

The Eriosema cordatum complex. II. The Eriosema cordatum and

E . nutans groups

C. H. STIRTON*

ABSTRACT

The Eriosema cordatum E. Mey. complex is segregated into a number of species. E. cordatum E. Mey. is retained

as a polymorphic species and two allied species, E. lucipetum C. H. Stirton and E. zuluense C. H Stirton are

described as new. Four additional species E. buchananii Bak. f., E. nutans Schinz, E. psoraleoides (Lam ) G Don

and E. parviflorum E. Mey. are revised and excluded from the E. cordatum group.

RESUME

LE COMPLEXE ERIOSEMA CORDATUM. // LES CROUPES ERIOSEMA CORDATUM ET E. NUTANS

Le complexe Eriosema cordatum E. Mey. est divise en un certain nombre d’especes. E. cordatum E. Mev est

retenu comme une espece polymorphique et deux especes, allies E. lucipetum C. H. Stirton el E. zuluense C. H. Stir-

ton sont decrites comme nouveUes. Quatre especes additionnelles E. buchananii Bak /., E. nutans Schinz,

E. psoraleoides (Lam.) G. Don et E. parviflorum E. Mey. sont revisees et exclues du groupe E. cordatum.

INTRODUCTION

Verdcourt (1971a) included seven species in the

Eriosema cordatum E. Mey. complex. The identity

of two of these species, E. populifolium Harv. and E.

distinctum N.E.Br., has already been clarified (Stir-

ton, 1978). This paper deals with the remaining

species of the complex: E. cordatum E. Mey., E.

nutans Schinz, E. buchananii Bak. f., E. parviflorum

E. Mey. and E. psoraleoides (Lam.) G. Don.

In 1975 Stirton concluded that E. cordatum could

be segregated into five taxa which, pending further

study, he named A to E. Taxon A is retained as a

polymorphic species. Taxon B and taxon C are

united into a new species E. lucipetum. Taxon D is

described new as E. zuluense. Taxon E has already

been described as E. gunniae C. H. Stirton (Stirton,

1981), so will not be dealt with further. The remain-

ing species are correctly identified as E. buchananii,

E. nutans, E. parviflorum and E. psoraleoides.

GROSS MORPHOLOGY OF THE SPECIES

Vegetative parts

Habit. This complex comprises perennial herbs, suf-

frutices and shrubs. E. psoraleoides is the only true

shrub and often grows up to 1,5 m high. The suf-

frutices are low growing woody plants with annually

regenerated growth points; E. nutans, E. buchananii,

and E. parviflorum. The remaining taxa all die back

completely during the winter months to regenerate

aerial parts each spring. Growth form may be erect,

ascending, decumbent or prostrate.

Rootstocks. All species develop perennial root-

stocks which, being quite variable in form, provide

excellent key characters for fieldwork, especially in

the absence of flowers and fruits. E. cordatum is

readily separated from all other species by its distinc-

tive daucate rootstock (Fig. 1). E. psoraleoides has

the most robust and branched rootstock (Fig. 2).

There is usually a short pseudo-stylopodium at the

apex of the rootstock. Root nodulation occurs in all

the species treated in this paper (Grobbelaar, Van

Beyma & Todd, 1967; Grobbelaar & Clarke, 1972;

Stirton, 1975). Freshly exposed rootstocks, with

roots and nodules still intact, will, if left in water for

*Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

24 hours, produce a pungent mustard-like odour.

The nodules swell considerably.

Stems. The perennial herbs have strict stems that are

either terete or angular. Up to fifteen stems may arise

from the short pseudo-stylopodium at the apex of the

perennial rootstock. Complex branching is common

in the suffrutices and in the shrub E. psoraleoides.

Two species are worth describing in detail.

The main stem in E. parviflorum is very short.

This results from the suppression of the leader con-

comitantly with the production of lateral branches.

The lateral branches grow out perpendicularly to the

main stem and are produced close to the ground.

Their growth begins once the leader has reached ap-

proximately 50 cm in height. The side branches grow

very quickly. If the leader dies back the plant

assumes the shape of a low bowl. The characteristic

low spreading nature of this suffrutex is possibly at-

tributable to the sudden retardation of growth in the

leader stem.

In contrast to that of E. parviflorum the stem of E.

psoraleoides is erect and continues to lengthen once

branching begins. However, if the leader is damaged,

the lateral branches soon compete to lead. This

results in the production of a short rounded shrub in

contrast to the usual tall, erect, somewhat spreading

shrub. Short rounded shrubs are commonly found in

Kwazulu and on closer investigation the leader will be

found to have been damaged by a borer insect. E.

psoraleoides also produces tall, densely branched

shrubs, with a compact appearance. This form

results from the production of numerous watershoots

that arise directly from the pseudo-stylopodium and

this form can be expected to be found in areas subject

to burning.

Vesture. The nature of the hair covering is determin-

ed by the length, direction, form, and quantity of the

hairs considered collectively. As St earn (1973) in-

dicated, these characters should be stated individual-

ly as there are more types of vesture than there are

terms to designate them. It is most unfortunate that

there is no standard hair terminology for, as Roe

(1971) pointed out, this would enhance the tax-

onomic value of hairs. As vesture forms a useful

character in the characterization of taxa in the pre-

sent study, I have decided to follow Grear’s (1970)

treatment of the American species of Eriosema by

adopting Lawrence’s (1960) definitions of hair types.

282

THE ERIOSEMA CORDATUM COMPLEX. IE THE ERIOSEMA CORDATUM AND E. NUTA NS GROUPS

STIGMA

•f T

ih m

hi' $ >\

•il'.l J

}H w

I f.f I vrs?<

CALYX

LOBE

CALYX

TOOTH

CALYX ]

TUBE I

5b >

My; f:

OVARY

PISTIL

HEIGHT

or

CURVATURE

Figs 1-9. — 1, Eriosema cordatum, daucate rootstock, side branches absent. 2, E. psoraleoides, multibranched rootstock.

3, basic types of stipules found in the genus Eriosema: a, side view of free stipules with petiole in plane of paper, and T/S

of stem at point indicated by dotted line showing how stipules flank petiole; b, side view of fused stipules clasping stem

with petiole lying in the plane of the paper, and T/S of stem showing how stipules are fused and lie opposite the petiole. 4,

marginal hair line in stipules of Eriosema: 1 , free stipules; a, hair line extending to base of stipules, 2, fused stipules show-

ing a split down the middle; b, end of hair line; c, uneven edge showing direction of rupture. 5a, diagramatic representa-

tion of an Eriosema calyx showing component parts; b, side view of an Eriosema flower showing the relative position of

the three types of lobes: 1, keel lobe; 2, lateral lobes; 3, vexillar lobes. 6, diagramatic representation of an Eriosema

gynoecium 7, E. cordatum, fruit cut away showing seed attachment to valves. 8, E. cordatum, plants showing a range of

variation in length of inflorescene (5 plants collected from one population at the Scottsville Race Course, Pietermaritz-

burg). 9, known distribution of the E. cordatum complex in South Africa.

In Eriosema all hairs are simple. Both glandular

and non-glandular hairs are represented. There are

two types of glandular hairs: firstly, very short

stalked globose glands (white, yellow, orange or red)

that vary considerably in size among the different

species; secondly, uniseriate hairs with bulbous

bases. The non-glandular hairs are uniseriate and

vary in colour from hyaline to white, grey, tawny or

ferruginous.

Leaves. Trifoliolate leaves are more common than

unifoliolate leaves. Occasional tetrafoliolate leaves

occur in E. psoraleoides. Cotyledons are always

unifoliolate. In all taxa the first two leaves of each

stem are nearly always unifoliolate. Subsequent

leaves may be unifoliolate or trifoliolate. Grear

(1970) reported a similar situation in America. He

mentioned that the ‘first-formed leaves in trifoliolate

species, even in strong shoots of old plants, are

C. H. STIRTON

283

always unifoliolate, even up to a fifth or higher

node’. The South African species commonly referred

to as unifoliolate are now known to include trifolio-

late-leaved plants as well. Since many herbarium col-

lections are made before the fourth internode has

elongated, it is often difficult to decide whether the

first formed three unifoliolate leaves are the basal

leaves of a unifoliolate or a trifoliolate species.

Similarly, it commonly happens that the basal leaves

have abscised by the time the fourth and higher inter-

nodes have elongated. Such plants are often referred

to as trifoliolate-leaved plants. These so-called

trifoliolate-leaved species may include unifoliolate-

leaved plants, but less rarely so than the so-called

unifoliolate-leaved species including trifoliolate-

leaved plants. For these reasons the use of leaf struc-

ture as used by previous authors has been dropped as

a key character for the delimitation of taxa in the E.

cordatum complex.

The size, shape, vesture and texture of the leaves

have been found to be useful definitive characters.

However, as will be indicated later, both size, shape

and vesture exhibit a remarkable variation under dif-

ferent environmental conditions. An appreciation of

this variation alone will do much to help in the under-

standing of the anomalies encountered in the keys of

Harvey (1862) and Burtt Davy (1932), both of whom

relied heavily on the shape, foliolation and vesture of

leaves as diagnostic criteria.

The shapes of leaves seen in this complex have

been classified according to Hickey’s (1973)

‘Classification of dicotyledonous leaves’ (see also

Dilcher , 1 974). As this system is not readily available,

a length-breadth ratio is sometimes provided in

brackets after the description of each shape. This

allows for a rapid conversion of Hickey’s terms to

the equivalents of any other system. All the leaf

measurements given are those of mature terminal

leaflets, or if there are no trifoliolate leaves, then of

the mature unifoliolate leaf.

Petioles. Petioles are either subsessile or elongated.

Most are channelled and ridged on the adaxial sur-

face. Length has been found to be a useful character

in some instances, but in others this parameter shows

marked plasticity.

Stipules. Stipules are a key character in Eriosema.

They may be free and laterally positioned in relation

to the petiole, that is, if the petiole was held tightly

against the stem the stipules would flank it (Fig. 3a);

or they are connate and opposite the petiole, that is,

would not flank the petiole if it was held tightly

against the stem (Fig. 3b).

The use of the stipule as a key character requires

careful observation as in some species the connate

stipules are wont to split to the base, and may

therefore be mistaken for a pair of free stipules. This

disadvantage is readily overcome however by a

careful inspection of the hair line along the margin of

the stipule (Fig. 4). It can be seen that in free stipules

the hair line extends uninterruptedly to the base (Fig.

4.1a) whereas in connate stipules that have split, the

hair lines end abruptly (Fig. 4.2b). The split margin is

also uneven (Fig. 4.2c). Connate stipules usually split

as a result of the expansion of nodal tissue.

The stipules are either tightly clasped to the stem,

patent, or recurved. In some taxa the stipules persist

but senesce rapidly almost before the young leaflets

expand, whereas in other taxa the stipules persist but

remain green until almost the end of flowering. In E.

psoraleoides the stipules are often caducous. These

are valuable field characters.

Reproductive parts

Inflorescences. In all taxa the inflorescence is an ax-

illary raceme bearing ( 1 — )— 50 flowers. Flower

number is too variable to be of much diagnostic sig-

nificance. The length of the peduncle is useful in

diagnosing E. psoraleoides as in this species it is

always less than one third of the length of the

raceme. Flower bracts are either caducous ( E . psora-

leoides) or persistent (all other spp.); shorter (£.

nutans ), or equal to (£. cordatum) the calyx. Bract

vesture may be sericeous, pilose, pubescent or stiff

patently hairy.

Flowers. The component parts of the flower vary

considerably in shape, colour, pubescence, gland en-

crustment and dimensions. Flower colour has proved

to be a useful distinguishing character. Flowers are

consistently either pure yellow, or red and yellow, or

pale pink with purple stripes, or red, yellow and

orange, or orange and yellow. Herbarium collections

of all but the yellow-flowered taxa turn purple on

drying: yellow flowers retain their colouration.

Flowers usually hang downwards before they open.

As they open they tend to rise into the horizontal

plane but return to their former position when they

close.

(a) Calyx. All taxa have a five-lobed campanulate

calyx with distinct triangular or lanceolate teeth (Fig.

5a). Three parts of the calyx are referred to in this

study: calyx lobes, calyx teeth and calyx tube. The

fused portion of the calyx is called the calyx tube

(Fig. 5a) whereas the unfused portions are called the

calyx teeth (Fig. 5b). The fused and unfused portion

of each segment together is called a calyx lobe. The

lobes are measured from the apex of each tooth to

the base of the tube directly beneath each tooth.

Other authors use the term lobe instead of tooth.

Three types of calyx lobes have been distinguished.

1. Keel lobe (Fig. 5a. 1) This lobe lies directly

beneath the keel and is thus abaxial (Fig. 5.1).

2. Lateral lobe (Fig. 5a. 2). There are two lateral

lobes. These lobes are situated one on each side of

the flower (Fig. 5b. 2). They are always equal in

length and upcurved towards the vexillar lobes.

3. Vexillar lobe (Fig. 5a. 3). There are two vexillar

lobes. These lobes are situated on the upper

(adaxial) part of the flower. The vexillar lobes

may be markedly connate or free and this is a

useful character in some taxa.

The calyx teeth may be triangular (£. parviflorum)

or lanceolate (£. cordatum)-, and shorter (E. parvi-

florum), or longer than, or equal to the calyx tube (£.

nutans). The calyx lobes may be shorter than half the

length of the flower (E. psoraleoides, E. parviflorum

and E. cordatum) or equal ( E . nutans) or longer.

Vesture on the calyx ranges from shortly pubescent

(£. parviflorum) to stiff long ferruginous-haired (£.

cordatum).

The length, pubescence and degree of fusion of the

calyx lobes; the shape and length (in relation to calyx

tube) of the calyx teeth; and finally the length of the

calyx lobes (in relation to the length of the flower) are

all useful diagnostic criteria.

(b) Standard. The standard varies in colour, size,

shape and structure. Hickey’s (1973) classification

has been used to determine shapes. Length-breadth

(1/b) ratios are given as was done for leaves. Shape

varies from narrow obovate (2:1) to wide obovate (1 .

2:1 , E. psoraleoides), or oblong (2: 1 , E. buchananii).

The apex may be rounded, emarginate or hooded.

284

THE ERIOSEMA COR DA TUM COMPLEX. II. THE ERIOSEMA CORD ATOM AND E. NUTANS GROUPS

Auricles vary in size, and range from prominent in-

flexed flaps (£■. buchananii) to small slightly inflexed

flaps ( E . cordatum). The appendages of the standard

are important diagnostically. These are situated on

the inner surface just above the auricles ( E . cor-

datum), well above auricles but less than half-way up

the blade ( E . zuluense), or high up and above the

middle of the blade ( E . buchananii). The appendages

may be free (E. buchananii), partly connate at base

( E . parviflorum) or connate (E. nutans). E.

psoraleoides and E. lucipetum (anomalous popula-

tion Stirton 1340) are both characterized by the

absence of any appendages.

The back of the standard is nearly always pubes-

cent and glandular. In E. psoraleoides it may also be

glabrate. Two unusual specimens Strey 5048 and

Vahrmeijer 1121, which have been included in E. cor-

datum also have glabrate or glabrous standards.

The colour of the standard varies but it is constant

within the species under study. It is yellow in E. cor-

datum, E. lucipetum, E. zuluense, E. nutans, E. par-

viflorum and E. psoraleoides. In E. buchananii it is

pale pink lined with purple. E. cordatum has a dull

red outer surface with the inner surface orange or

yellow and distinctly purple-veined.

A characteristic of all taxa is the presence of a

semi-circular patch of intense yellow situated just

above the appendages. This patch is noticeable even

in the yellow flowered taxa where it tends to be a

more intense yellow. It is referred to as a ‘nectar

guide’ in this study.

(c) Wings. At first glance the wings appear to be

uniform. Their infrequent use as a diagnostic

character has probably been due to the difficulty en-

countered in describing them. The wing consists of a

lamina which may be cultrate (E. nutans), cultriform

(E. cordatum), or narrow oblong (E. psoraleoides).

The single auricle may be straight ( E . psoraleoides),

or forward sloping ( E . nutans). The claw is either at-

tenuate (E. nutans) or strongly recurved (E.

zuluense).

The wings may be shorter, equal to or longer than

the keel blades. Colour varies from yellow, yellow

with flushes of pink, to orange.

(d) Keel blades. The keel blades show various degrees

of fusion. The shape is difficult to describe but

ranges from narrow oblong narrowing towards the

claw (3: 1, E. psoraleoides) to variously curved to

almost falcate abaxially. The apex may be obtuse,

falcate or rostrate. The auricle is less defined than in

the wings. It may be recurved, vertical or forward

sloping. The claw is attenuate. The keel blades are

usually wider than the wings. The keel blades may be

equal to (E. psoraleoides), shorter than (most

species) or longer than (E. parviflorum) the wings.

The lamina may be sparsely or densely encrusted with

glands. The base line of the keel blade may be

glabrous or hairy. The keel blade is generally

greenish white in colour.

(e) Androecium. The stamens are diadelphous with

the vexillar stamen free. The vexillar stamen has a

knee-like thickening near the base. The thickness of

this knee is usually proportional to the depth of the

claw of the standard. The fused stamens are collec-

tively referred to as the staminal sheath. The anthers

form two whorls when the sheath is closed as alter-

nate filaments are unequal in length. Anthers are uni-

form, dehiscing longitudinally.

(0 Gynoecium. The length of the gynoecium and the

variation in its components are important characters.

The ovary may be stalked, subsessile or sessile. It is

biovulate (rarely triovulate in E. psoraleoides) with

two campylotropous ovules. The type and degree of

vesture on the ovary wall is a useful character. The

hairs may extend less than half the length of the

valves. The style is glabrous, or minutely hairy and

glandular, or eglandular. It is variously thickened at,

or beyond, the point of curvature. The height of cur-

vature is a useful character and is calculated as the

vertical distance from the base line of the style to the

level of the stigma (Fig. 6). The erect portion of the

style may be straight or incurved. The stigma is capi-

tate, inserted.

(g) Discoid floral nectary. Whilst floral nectaries

have been frequently reported in papilionoid

legumes, particularly in the tribe Phaseoleae (Waddle

& Lersten, 1974), their presence in the genus Erio-

sema is poorly documented. This is probably owing

to their small size and also to the difficulties involved

in the interpretation of their structure. Phillips (1951)

mentioned that the ovary in Eriosema is ‘usually sur-

rounded by a small cupular disc’. The only other

report found was Waddle et al’s (1974) report of the

presence of a discoid floral nectary in Eriosema

rufum Don. There are no reports on the presence of

floral nectaries in the American species (Grear,

1970). Discoid floral nectaries have been found in all

the species studied in the present work.

The ontogeny of discoid floral nectaries remains

obscure. Moore (1936) thought that they resembled a

staminal sheath and were more closely associated

with the staminal whorl than with the gynoecium.

Waddle (1968) suggested alternatively that the disc

nectary was not a reduced whorl but an outgrowth of

the receptacle. The solution to this puzzle of mor-

phological interpretation will involve, as Waddle et

al. (1974) pointed out, a careful ontogenetic study.

The discoid floral nectaries in all the taxa studied

here were securely adnate to the base of the staminal

sheath. In view of the lack of ontogenic evidence the

discoid flora nectary has been considered in the des-

cription of species as in independent structure.

The sizes and shapes of the floral nectaries are

variable. The apical margin may be wavy, erose,

smooth, finely dentate or revolute. The type of

margin appears to be correlated with the quantity of

nectar secreted. Nectaries with irregular margins

were seen, in the field, to produce more nectar at one

time than the smooth margin nectaries. Frey-

Wyssling (1955) has reported that the relative propor-

tions of xylem and phloem in nectaries affects the

type of secretion. Nectaries which are supplied most-

ly by phloem tissue tend to secrete nectar with a high

sugar concentration. It is difficult to interpret the

meaning of the lack of vascular tissue reported in the

nectary of Eriosema rufum Don (Waddle et al.,

1974). A study is needed to correlate types of vascu-

lar tissue with nectar secretion in Eriosema and then

to correlate the nectaries with pollination mechan-

isms. A useful study could be made of the structure,

morphology, stomatal distribution and ontogeny of

nectaries in papilionoid legumes. I am not aware that

such a survey has been made for the family as a

whole.

Fruits. The fruits are compressed 2-valved, 2-

seeded dry dehiscent pods. The valves are obliquely

beaked, oblong (2:1) to wide oblong (1, 5:1), or

trapeziform. The valves are smooth and shiny inside,

and usually sericeous, velutinous, pilose hairy or

shortly pubescent outside. The valves dehisce along

upper and lower sutures until they remain attached

C. H. STIRTON

285

only at the base. Most of the South African species

dehisce with an explosive action. This may be due to

an environmental influence as pods of the same

species have been seen to dehisce very slowly during

cloudy overcast conditions. The separated valves coil

very tightly once the seed has been shed.

Seeds. Mature seeds are smooth, dull or shiny, com-

pressed, transverse or oblique. Shape may be oblong,

reniform or somewhat mango-shaped. Seed colour

varies from black or deep purple to chestnut brown

to khaki, with or without speckles and mottles.

Polymorphic seeds are commonly produced in a

species but although both types are never found on

the same plant they may occur in the same popula-

tion. In E. parviflorum the black seeds are thicker

and narrower than the speckled khaki seeds (see

under Insect infestation).

The hilum of the Eriosema seed is linear and ex-

tends almost the whole length of the seed. The sub-

horizontal funicle is attached at the inner extremity.

The thick bilabiate flange flanking the hilum is refer-

red to as a strophiole (after Grear, 1970). The attach-

ment of the seed of the valve wall can be seen in Fig.

7.

Seeds swell to almost double their size when they

imbibe water. This effect is also seen in the unripe

pod. When the pod is green and fully developed the

two seeds almost fill it, but as the pod blackens and

dries out the seeds shrink to almost half their

previous size.

ECOLOGY

Throughout this study field observations have

been aimed primarily at assessing the range of varia-

tion within the taxa under study. If the field observer

repeatedly meets a particular phenomenon, or range

of variation, in the individuals or populations under

study and this is frequently associated with a par-

ticular set of environmental or local habitat condi-

tions, there is some justification for associating these

where either one, or the other, is encountered in new

field situations. Therefore a field observer experienc-

ed in the study of a particular group of plants, should

be able to draw upon his experience when meeting

new plants of the same genus in new localities. Under

these conditions his intuitive assessment of variation

in relation to general habitat conditions should be

reasonably accurate, despite his inability to relate a

particular feature of variation to a particular en-

vironmental factor.

If, in the consideration of variation within the

species in this study, an environmental factor, or

plexus of factors, are suggested as being related to

some particular morphological form, or feature of it,

this is the outcome of field observation and is regard-

ed as a reference point to be kept under considera-

tion, rather than as a factual statement of cause and

effect.

This section collates various field observations that

have been useful in understanding the species under

study. Selected examples of infraspecific variation

are presented in order to provide as broad a picture

of the species as possible. Fortunately most of the ex-

amples presented are indicative of the types of varia-

tion, and phenomena, to be found in the whole E.

cordatum complex. Where relevant more detailed ac-

counts are given under each species.

Field observations

Plants growing in full sun tended to be shorter and

more compact than plants of the same species grow-

ing under different intensities of shade. Similarly

plants that grew in burnt veld had a more stunted

form than did plants of the same species which grew

in adjacent unburnt veld. These phenomena, which

apply to all species in the E. cordatum complex, were

fully discussed in a previous paper (Stirton, 1977).

E. cordatum was observed to exhibit a markedly

greater variability than any other species in the com-

plex. Plants that grew in the same population showed

a variation in the size, shape and foliolation of the

leaves, degree of compactness of flowers on the in-

florescence, length and thickness of the peduncle,

shape and senescence of the flower bract and finally

the size of the standard. Fig. 8 shows the variation

which can occur in the length of the inflorescence

within the same population.

It was found that plants of E. cordatum that grew

in tall grass had bigger leaves, longer internodes,

shorter inflorescences and smaller standards, than

those plants that grew in short grass in more open

areas. In other populations the peduncle was very

long in tall grass and short in short grass.

Allsopp (1947) wrote that plants of E. cordatum

become suppressed if the veld remains unburned, or

is heavily grazed and trampled. This information ap-

pears to have been based only on the inspection of

field sites.

The combined influence of wind and sun appear to

affect Eriosema plants. In prolonged hot windy

weather these plants tended to die back from the

growing points and to shed their basal leaves. This

observation is important since the first dichotomy of

the available keys differentiates between unifoliolate

and trifoliolate leaves. It is now known that many

species, E. cordatum in particular, produce their first

trifoliate leaf at the fifth or higher node. Such a

plant, if collected before the trifoliolate leaves had

been produced, would be difficult to key out.

Similarly for a trifoliolate-leaved plant that had

already shed its basal leaves. The loss of unifoliolate

leaves by a predominently trifoliolate-leaved plant

may of course be due to natural ageing.

Hot windy days have also been observed to wither

flowers. Two populations of E. cordatum were

observed to have failed to set seed for an entire

season because the flowers withered just as the cor-

ollas emerged from the calyx.

Insect infestation

Insect larvae, particularly Agromyzidae, cause ex-

tensive damage to various plant parts. Eleven dif-

ferent types of larvae have been collected and await

identification and I am indebted to Dr B. R.

Stuckenberg of the Natal Museum for his advice. He

informed me that most species of Agromyzid flies lay

their eggs in particular plants and are host specific to

those plants. The observation that some larvae of

these insects are found only in certain species has

been a useful field guide for the field indentification

of species. These studies when pursued promise to be

most rewarding. The different Agromyzid larvae

form distinctive patterns (mines) beneath the cuticle

or epidermis. Such patterns are often most noticeable

on herbarium specimens.

Further studies are also needed on the insects

which cause extensive damage to seeds. In E. par-

viflorum both laboratory and field observations

revealed that the production of polymorphic seeds

was closely associated with the infestation of insects

in fruit pods. Pods with black seeds were rarely af-

fected, whereas in pods with speckled, greenish

286

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

yellow seeds, the seeds rarely reached maturity

because of heavy insect damage. These seeds should

be analysed comparatively for chemicals that inhibit

predation.

GEOGRAPHICAL DISTRIBUTION

The geographical distribution maps presented are

based mainly on the author’s field collections and

available herbarium material. This commonly used

approach to the compilation of distribution maps has

certain pitfalls that should be recognized. Firstly the

distribution map is not necessarily a reflection of the

taxon’s extant distribution. It stands to reason that

the compilation of a distribution map is an act based

on an a priori hypothesis that the map will provide

the basis for some deductive argument. If this is so,

then a map compiled from herbarium collections

spanning a century or more is indeed a shaky base for

any type of interpretation. This awareness of a space-

time relationship is seldom considered in the com-

pilation and interpretation of distribution maps. The

use of unchecked specimen citations gathered from

the literature hardly needs any comment. The second

pitfall is that a map compiled from all the available

herbarium material is open to severe criticism if, as

White (1971) so testily states, its interpretation con-

cerns unwarranted speculations as to its inclusive

variation. The reason is that the known complexity

of genetic systems, including hybridization, poly-

ploidy and apomixis, is such that any particular

expression of a phenotypic character could be theor-

etically possible over any interval of time and at any

particular point of a taxon’s range. Such expression

may have little or no significance in an understanding

of the taxon’s origin or distribution pattern. The

chances of this situation occurring are greater the

older are the specimens used. The third pitfall con-

cerns the frequent attempts by botanists to discover

variables of ecological significance by a reference to

distribution maps. Although not known as such, this

phytocartogeographic approach has an equivalent in

the zoocart ogeographic approach of zoologists. An

example of the problem and the confusion that may

develop is shown in the critical assessment by

Stuckenberg (1969) and Van Dijk (1971) of

Poynton’s (1964a, 1964b) faunal study of amphibia

in southern Africa. The final pitfall is obvious and

common to most approaches. It concerns the correct

delimitation of taxa. Even one incorrectly named

specimen in a distribution map can lead to incorrect

interpretations.

The known distributional ranges of all the species

studied in this complex is given in Fig. 9. This com-

plex occurs mainly in the Transvaal, Swaziland,

Natal and the eastern Cape Province. Very few col-

lections are known from Lesotho, Botswana and the

Orange Free State. No representatives of the complex

occur in South West Africa/Namibia or the western

and northern Cape Province.

Eriosema cordatum and E. psoraleoides, have the

most widespread distribution of the group in

southern Africa, whereas E. nutans is less widely

dispersed. The only really restricted distributions are

those of E. buchananii, E. zuluense, and E.

lucipetum. Four species of the complex extend

beyond southern Africa. These are E. buchananii, E.

nutans, E. parviflorum and E. psoraleoides.

TAXONOMY

The descriptions of Eriosema species that follow

are based on the author’s examination of herbarium

material and of populations in the field in South

Africa.

A selected citation of representative specimens of

each species recognized in this study is given in the

enumeration of species. This was prepared by the

selection of a single voucher specimen from amongst

all specimens studied to serve as a record of each

quarter degree square through which the taxon was

naturally distributed.

KEY TO SPECIES

Stipules connate:

Calyx teeth triangular, shorter (rarely equal to) than calyx tube; pubescent or tomentose, rarely with few scat-

tered longer hairs; appendage of standard situated well above auricles; flowers yellow 1 . E. zuluense

Calyx teeth lanceolate, always longer than calyx tube, stiffly shaggy or pilose; appendage of standard situated

low down or just above auricles or if higher then extending to auricles; flowers red and yellow . . 2. E. cordatum

Stipules free:

Perennial herbs; calyx teeth lanceolate, longer than calyx tube 3. E. lucipetum

Shrubs or suffrutices; calyx teeth shortly triangular, mostly equal to or less than calyx tube:

Plants erect shrubs; stipules minute < 5 mm long; appendages absent on standard; peduncle of inflorescence

< 'A length of inflorescence 4 .E. psoraleoides

Plants ascending, decumbent or erect suffrutices; stipules > 5 mm long; appendages present on standard;

peduncle of inflorescence > 'A length of inflorescence:

Stipules 10-16 mm long, 2, 5-4,0 mm wide, oblong-lanceolate, base narrower than middle; flowers pale

pink lined with purple; appendages of standard high up above middle; free, shallow ridged,

half moon shaped, vertical to axis of standard 5. E. buchananii

Stipules 5-8 (10) mm long, 1-2 mm wide, linear lanceolate or triangular, base wider than middle;

flowers yellow; appendages of standard situated low down just above the auricles, connate,

upcurled flaps, horizontal to axis of standard:

Stipules mostly erect, clasping stem; petiole 3, 1-7,0 mm long; leaf rachis (7) 8-11 (12) mm long;

flowers 7,5-10,0 mm long; flower bracts lanceolate; calyx (3,0) 3, 5-4, 3 (4,5) mm

long; standard oblong; wings longer than keel blades. Leaves not photonastic.

(Transvaal.) 6. E. nutans

Stipules recurved and spreading when old; petiole 1,0-2, 2 (3,0) mm long; leaf rachis 2-6 mm

long; flowers 5-6 mm long, flower bracts shortly triangular; calyx 2-3 mm long; standard wide

obovate; wings shorter than keel blades. Leaves photonastic. (Natal, one record from N. E.

Transvaal) 7. E. parviflorum

C. H. STIRTON

287

Fig. 10. — Eriosema zuluense. 1, habit; 2, stem with fruits and flowers, x 1,2; 3, flower bract, x 9, 4, flower x 3,8;

5, calyx opened out, x 9; 6a, standard opened out, x 4,6; 6b, standard closed x 4 6; 7, wing, x 4,6, 8 keel, x 4,6, 9,

vexillar stamen, x 6,4; 10, staminal sheath, x 6,4; 1 1 , discoid floral nectary, x 18; 12, gynoecium, x 6,4, 13, stigma,

40.

288

THE ERIOSEMA CORDATUM COMPLEX. IE THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

1. Eriosema zuluense C. H. Stirton, sp. nov., E.

cordato E. Mey. affinis, sed dentibus calycinis

triangularibus, tubo brevioribus (raro eum aequan-

tibus), floribus luteis, appendiculo interno vexillari

manifeste supra auriculos sito distinctam.

Herbae perennes, prostrati vel repentes vel,

decumbentes, 10-12 cm altae. Folia 1-foliolata, raro

infima 3-foliolata, 4, 0-5, 5 cm longa et lata, late

ovata, apice obtusa cum apiculo, basi cordata, utrin-

que praesertim secus venas sparse sericea, parce glan-

dulosa, margine revoluto. Stipu/ae connatae. Racemi

axillares, 20-30-flori, folia excedentes. Flores lutei,

10-12 mm longi, bractis dimidio longitudinis

florum, cito caducis. Calycis lobi aequilongi, den-

tibus triangularibus, tubo brevioribus, molliter

pubescentibus. Vex ilium alis aequilongum, omnibus

carinam valde glandulosam excedentibus. Fructus

12x10 mm, oblongus, prominentes rostratus,

sericeus, tenuiter pubescens.

Type. — Natal, 2831 (Nkandla): Ntambanana

(-DA), Acocks 12980 ( K, holo.!; PRE).

Perennial herb, 10-12 cm tall, flowering in spring.

Stems trailing, prostrate or decumbent, finely pubes-

cent with reflexed yellowish hairs. Leaves 1-foliolate,

rarely the lowest leaves 3-foliolate, 4, 0-5, 5 cm long

and wide, broadly ovate, apex blunt with a point,

base cordate, both sides sparingly sericeous especially

along the veins, sparsely glandular, margin revolute.

Stipules 5-7 (10 mm long, lanceolate, connate but

occasionally splitting when old, sericeous. Petioles

2-4 mm long. Racemes axillary, (10)

20-30-flowered, exceeding the foliage, peduncle 5—6

cm long. Flowers yellow, 10-12 mm long, bracts ±

half the length of the flower, rapidly caducous. Calyx

4 mm long, lobes equal in length; teeth triangular,

shorter than (rarely equalling) the tube, softly pubes-

cent and sparingly glandular. Standard ± 12 mm

long, 5 mm wide, oblong, equal in length to the wing

petals but always exceeding the keel; appendages

slightly fused, free from auricles. Wings 12 mm long,

3 mm wide, oblong, auriculate, pouched, longer than

the keel. Keel 8 mm long, 3 mm wide, strongly glan-

dular, sparingly pubescent along base, pocket pre-

sent. Stamina! sheath 7 mm long, free stamen

geniculate; anthers uniform, alternately long-

basifixed and short-medifixed. Gynoecium 7 mm

long; ovary 4 mm long with 1 mm long gynophore,

densely sericeous, curvature of style 3 mm high;

stigma capitate. Nectary present, high as gynophore;

margin erose. Fruit 12 mm long, 10 mm wide, oblong

with a prominent 3 mm long beak, sericeous and

finely pubescent. Seed not seen. Fig. 10.

Eriosema zuluense is endemic to Zululand and

stretches from Josini Dam and Hluhluwe in the north

to Mhlabatini, Hlabisa, Melmoth, Eshowe and

Ngoye Forest in the south (Fig. 11).

Natal. — 2732 (Ubombo): Hilltops on main road to Josini

(-CA), Pooley 276; 6 km from Ubombo (-CA), Stirton 505;

Ubombo Mountain Summit (-CA), Venter 1768, 4308. 2831

(Nkandhla): Mhlabatini (-AB), Gerstner 4185, 4232; Nkwaleni

(-BA), Lawn 1548; 8 km south of Hlabisa (-BB), Codd 2004 and

Acocks 13067; 25 km south-east of Melmoth (-CB), Codd 1838;

Eshowe (-CD), Lawn 1164; between Nkwaleni and Eshowe

(-DA), Stirton 426; 9 km west of Ntambanana (-DA), Codd

1885; 8 km north-west of Ntambanana (-DA), Acocks 12980;

Ngoye Forest Reserve (-DD), Venter 406. 2832 (Mtubatuba):

Hluhluwe Game Reserve (-AA), Ward 1548. Without precise

locality: Gerstner 4656. This is a mixed sheet which also has a

specimen of E. cordatum.

This species is somewhat variable and additional

collecting is needed. It is best exemplified by Acocks

12980, Codd 1885, Codd 2004, Ward 13067 and

Ward 1548. E. zuluense is characterized by its

Fig. 11. — Known distribution of Eriosema zuluense in South

Africa.

vesture, habit, branching and its triangular calyx

teeth shorter than the calyx tube. Acocks 13067 is

unusual in that the stipules are rapidly caducous and

the leaflets narrower.

The plants grow in dense grassland on rocky

hillsides, particularly dolerite outcrops, between

300-400 m. Flowering occurs between October and

December.

The specific epithet zuluense refers to the region

throughout which they are distributed.

2. Eriosema cordatum E. Mey., Comm. 128

(1836); Meisn. in J. Bot., Lond. 2: 80 (1843); Harv.

in FI. Cap. 2: 259 (1862); Burtt Davy, FI. Transv. 2:

413 (1932); Verde, in Kew Bull. 25: 118 (1971); Ross,

FI. Natal, 208 (1972). Type: Natal, between Gekau

and Basche, in grassland (V, b, 22) Drege s.n. (PRE;

K, iso.).

E. guenzii Sond. in Linnaea 23:34 (1840); Drege, Zwei Pfl. Doc.

145 (1843); A. Dietr., FI. universalis II: 187 (1861), icon. Syntypes:

South Africa, Natal, Port Natal, Guenzius 27, 634. (PRE; K,

isosyn.).

E. cordatum E. Mey. var. guenzii (Sond.) Harv. in FI. Cap. 2:

259 (1862); Burtt Davy, FI. Transv., 2: 413 (1932); Verde, in Kew

Bull. 25: 118 (1971).

Perennial herb 5-60 cm tall with 1-20 stems from

short stylopodium of underground rootstock.

Woody rootstock vertical, daucate, up to 50 cm

deep. Stems prostrate, decumbent to semi-erect;

terete or trigonous; variously ribbed; vesture hispid,

hirsute, velutinous or pilose; hairs red, white or

tawny; if glandular then orange, red or yellow gland-

ed. Leaves unifoliolate, or trifoliolate with basal

leaves unifoliolate; stipules (4,0) 5,0-15 (23) mm

long, 3-8 (12) mm wide; narrow ovate (2:1) to

lanceolate (3:1), often falcate; semi-connate, either

splitting or tightly clasping stem; glabrescent or

hairy. Petiole (2) 4-7 (20) mm long. Leaflets 6-10

(14) cm long, (2,5) 3, 0-5,0 (9,2) cm wide, laterals

smaller; base cordate or obtuse; apex obtuse, round-

ed or acute; terminal leaflet symmetrical, ovate

(1,5:1) to very wide ovate (1:1 or less), elliptic (2:1) to

suborbiculate (1, 2:1), laterals asymmetrical, obli-

que, oblong, (2:1) to narrow oblong (3:1), to ovate

(1,5:1) to narrow ovate (2:1); vesture strigose,

sericeous, hirsute, pilose or velutinous, erect or ap-

pressed. Young leaflets either densely glandular or

eglandular, with red foxy or white hairs. Petiolules

2-4 mm long. Rachis (9) 10-15 (23) mm long, rarely

with two persistent acrorachial stipels. Racemes 3-27

C. H. STIRTON

289

flowered; peduncle 5-12 (21) cm long; rachis 2-5 cm

long; flowers 8-10 (13) mm long, 3—5 mm wide;

reflexed vertical or with apex recurved until touching

calyx; red and yellow, or red, yellow and orange;

bracts lanceolate, persistent or deciduous. Calyx

(3.0) 4, 0-6,0 (7,8) mm long, half length of flower,

teeth equal or unequal, longer than tube; lanceolate;

stiffly shaggy or pilose; glabrous or finely hairy in-

side; glandular; tube (1,5) 2, 0-4, 5 (5,0) mm long;

vexillar lobes (3,8) 4, 0-6,0 (7,8) mm long; lateral

lobes (3,0) 4, 0-5, 5 (6,5) mm' tong; keel lobe (3,8)

4. 0- 6,0 (7,5) mm long often thinner or incrassated.

Standard (8,0) 10,0-11,0 (12,3) mm long; 4, 4-7,0

(9.0) mm wide; oblanceolate (3,1) to wide obovate

(1,2:1), narrowed towards auricles; apex obtuse,

ret use or emarginate; usually densely white pubescent

and glandular on back, but occasionally glabrescent;

claw 2-3 mm long; auricles inflexed, prominent; ap-

pendages bifurcate, between or just above auricles,

either low ridges or large upcurled flaps, (3,2)

3. 8- 4,0 (4,8) mm from base of claw. Wings (7,8)

9.0- 11,5 mm long, 2, 1-3,5 (4,0)mm wide at max-

imum; longer than keel blades; claw 1, 2-2,0 mm

long; auricles 1, 2-2,0 mm high. Keel blades (6,2)

6. 8- 8,0 (8,5) mm long, 3, 0-4, 6 wide at maximum;

yellow, white or orange glanded; claw 1,2-2, 8 mm

long; auricles 1,2-2, 8 mm high. Stamina l sheath

6. 0- 7, 2 mm long, 2, 5-4,0 mm wide at maximum,

vexillar stamen 5, 0-7, 2 mm long. Gynoecium (5,0)

6, 5-7,0 (7,5) mm long, ovary 2, 5-4,0 mm long; with

stalk 0,3-1 ,0 mm long, hairs extending to half length

of pistil; style incrassated at or above point of flex-

ure, curvature 2-3 mm high; stigma level with an-

thers, capitate. Discoid floral nectary smooth slightly

rippled or erose irregular. Fruits (11) 15-16(17) mm

long, 9-11 mm wide, wide oblong (1,5:1), strongly

compressed; beak straight, downward or upward

pointing; with scattered golden brown hairs mostly

patent but appressed on margins and often glabrate

on sides, apex oblique. Seeds 3, 0-5, 2 mm long,

1,7-2, 4 mm wide, chestnut brown or khaki, with or

without speckles, hilum white or brown. (See Fig. 7,

Stirton, 1978.)

Eriosema cordatum is common and widespread in

Swaziland and eastern parts of South Africa. It oc-

curs less commonly in Zimbabwe, Mozambique and

Lesotho.

This difficult species has long been the dumping

ground for many specimens of Eriosema in South

African herbaria. The two factors which have pro-

bably contributed most to this confusion are the oc-

currence of marked phenotypic plasticity within the

taxon and the occurrence of extensive hybridization

(Stirton, 1981) of this species with other species.

E. cordatum remains the most heterogenous group

in the E. cordatum complex and includes within its

range of variation the previous varieties cordatum

and guenzii. This taxon is the most widespread

Eriosema in southern Africa (Fig. 12). Preliminary

Fig 12. — Eriosema cordatum.

Pictorialized distribution map

showing variation in leaflet

shape and size, length of peti-

ole and length of rachis. a,

Galpin 1031; b, Stirton 1468;

c, Stirton 1462; d, Strey 5048;

e, Stirton 1620; f, Baijnath

419; g. Weeks s.n.; h, Huntley

90; i, Schelpe 1755; j, Mogg

15756; k, Scheepers 487.

Where two leaflets are shown,

the first is the terminal leaflet

of a trifoliolate leaf and the

second leaflet is one of the

laterals of the same leaf. If

single leaves are shown, only

unifoliolate leaves occur. The

left hand bar refers to length

of petiole and the right hand

bar refers to length of leaf

rachis. Area 1, area in which

unifoliolate leaved plants are

likely to occur in the Trans-

vaal; 2, area where erect

plants are likely to occur; 3,

area where unifoliolate cor-

date leaved plants grow in

Natal.

290

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

fieldwork has enabled me to outline some of its com-

plex variation. Fig. 12 is a pictorialized distribution

map showing variation in leaf shape and size, relative

lengths of leaf petiole and leaf rachis, stem vesture

and shape of calyx lobes (Fig. 13). The areas 1—3

shown in Figs 1 2 & 1 3 are merely the known ranges of

certain types of variation and do not necessarily in-

clude all the collections found within them.

Plants tended to be decumbent and predominantly

unifoliolate along the coastal areas (3). There was a

general increase in size of the leaves from south to

north. Coastal plants growing south of Port Edward

were both unifoliolate and trifoliolate. In the

Mpangazi area in northern Zululand there were a few

collections which were notable for their larger

flowered and sparsely haired, unifoliolate condition.

This area requires further investigation. Inland from

the Natal coast plants became smaller and

predominantly trifoliolate. More stems were produc-

ed. Area 2 is an approximation of the area in which

erect trifoliolate leaved plants can be expected to oc-

cur [for example, Stirton 1462 (Fig. 14)]. Area 1 is

the range of small semi-erect unifoliolate leaved

plants which differed from the coastal unifoliolate-

leaved plants in their pubescence, leaf shape and

habit. These plants, for example Stirton 1465 & 1468

were restricted to the high altitude areas of the

eastern Transvaal. They appeared to have a close af-

finity to E. gunniaeC. H. Stirton. Although included

here, their identity remains open. They warrant fur-

ther investigation.

Transvaal plants were predominantly decumbent,

trifoliolate, and small-leaved. Leaflet apices were

more acute than in the Natal populations. Lateral

leaflets were narrower, more asymmetrical and more

oblique than lateral leaflets of Natal plants. As ex-

ceptions occurred randomly throughout the area no

safe conclusions could be drawn from this data.

Stem vesture was too variable to be reliable but,

considered as a whole, it was noted that the majority

of plants had either tawny or ferruginous, upward or

downward pointing hairs. This was a feature which,

considered with the presence of a daucate rootstock,

connate stipules (invariably splitting), strongly

reflexed red and yellow flowers, stiff haired calyx

and flower bract, and with the wing greatly exceeding

the rostrate keel, could be used to separate this

species from all other Eriosemas in southern Africa.

The morphological diversity of this species has

been discussed in some detail elsewhere (Stirton,

1981) and in the section on Ecology. In summary it

may be said that no useful purpose will be served at

present if it is divided into a number of sub-groups.

Fig. 13. — Eriosema cordatum.

Pictorialized distribution map

showing variation in shape

and size of calyx, and stem

pubescence. The specimens il-

lustrated and areas demar-

cated are as indicated for Fig.

12.

C. H. STIRTON

291

Fig 14. — Eriosema cordatum. Erect plant growing along roadside

between Mbabane and Malkerns Agricultural Research

Station.

If it is recognized that the species is polymorphic and

if future field studies are aimed at understanding pat-

terns within the variation encountered, much will

have been done towards understanding the complex

as a whole.

Transvaal. — 2329 (Pietersburg): Louis Trichardt (-BB),

Breyer 24203; Pigeon Hole (-CD), McCollum PRE 32658; Wood-

bush (-DB), Jenkins PRE 32656; Haenertsburg (-DD), Srirton

1424. 2330 (Tzaneen): Westfalia Railway Siding (-CA), Slirion

1429; Modjadjes (-CB), Rogers 18019; New Agatha (-CC), Mc-

Collum s.n. 2331 (Phalaborwa): Letaba (-DA), Swiertstra 4384.

2428 (Nylstroom): Palala (-BC), Breyer PRE 18124. 2430

(Pilgrims Rest): The Downs (-AA), Moss & Rogers 252, 412, 850.

2431 (Acornhoek): 9 km from Bushbuckridge to Nelspriut (-CC),

Srirton 1453. 2526 (Zeerust): Roster (-DD), fiwm Davy 7178. 2527

(Rustenburg): Rustenburg Nature Reserve (-CA), Jacobsen 1626;

Bokfontein (-DB), Jenkins 7536; Masgaliesburg (-DC), Story &

Innes 1397; Uitkomst (-DD), Coetzee 466. 2528 (Pretoria):

Groenkloof ( -CA), Barker 10096; Irene ( — CC), Smith 5155. 2529

(Witbank): Loskop Dam (-AD), Theron 2111. 2530 (Lydenburg):

Lydenburg (-AB), Wilms 5856; 35 km from Nelspruit to Sabie

(-BB), Stirton 1478; Lowveld Botanic Garden (-BD), Stirton

1453; Waterval Onder (-CB), Young 26495; 15,8 km from

Nelspruit to Kaapsche Hoop (-DB), Grobbelar 847; Cythna Letty

Nature Reserve (-DD), Muller 2254. 2531 (Komatipoort); White

River (-AC), Gillett 1036; 12,3 km from Nelspruit to Barberton

(-CC), Stirton 1462; Barberton ( -CC) Williams 7645. 2627 (Pot-

chefstroom): Krugersdorp ( -BB), Cowgill s.n. J 33026; Losberg

(-CB), Theron 924; Sterkfontein Caves (-DD), Mogg 35301. 2628

(Johannesburg): Robertsham (-AA), Stirton 1300; Modder East

(-AB), Louw 862; 11,3 km south of Heidelberg (-CB), Acocks

20854. 2630 (Carolina): Redhill near Oshoek (-BB), Stirton 1311;

Mavieriestad (-CA), Pott 5074; Athole Pasture Research Station

(-CB), Norval 68; 29 km from Lothair to Amsterdam (-DA), Stir-

ton 1352; Piet Retief (-DD), Collins 13200. 2729 (Volksrust):

Volksrust (-BD), Mogg 7499. 2730 (Vryheid): Mooihoek (-BA),

Devenish 70; 20 km from Paulpietersburg to Piet Retief (-BB),

Stirton 1314; Wakkerstroom (-AC), Beeton 243.

Swaziland. — 2531 (Komatipoort): Devils Bridge (-DC),

Compton 31722. 2631 (Mbabane): Bremersdorp (-AD), Bolus

11850; Malkerns (-CA), Compton 26164; Kubuta (-CD), Comp-

ton 32420. 2731 (Louwsburg): Goedgegun (-AA), Compton

32409.

Natal. — 1730 (Vryheid): Farms Tweekloof and Altemooi

(-AD), Thode A 185; Paulpietersburg (-BD), Stirton 1324; 18

km from Vryheid to Paulpietersburg (-DB), Stirton 1322; Vryheid

(-DD), Stirton 1332. 2732 (Ubombo): Pongola Poort (-CA), Stir-

ton 489; Mpangazi (-DA), Strey 5048. 2828 (Bethlehem): Mahai

Valley (-DB), Schelpe 1755. 2829 (Harrismith): near Van Reenens

Pass (-AD), Schweikerdt 909; Elandslaagte (-BD), Shirley NU

32891. 2830 (Dundee): Dundee (-AA), Shirley 246; Buyahlanga

Mountain (-AC), Stirton 1362; Tugela Ferry (-CD), Stirton 1330;

12 km from Greytown to Dundee (-DC), Grobbelaar 1630. 2831

(Nkandhla): Eshowe (-CD), Lawn 1213; Ngoye Mountain (-DC),

Stirton 472: Mtunzini (-DD), Stirton 414. 2832 (Mtubatuba):

Hluhluwe Game Reserve (-AA), Hitchins 599; Charters Creek

area (-AB), Ward 2839; Hlabisa (-AC), Harrison 252; Dukuduku

Forest (-AD), Hitchins 68; eastern shores of Lake St Lucia (-CC),

Taylor 97; Enseleni Nature Reserve (-CC), Venter 6501. 2930

(Pietermaritzburg): Estcourt Pasture Research Station (-AB),

Acocks 9868; Noodsburg (— BD), Hilliard 1220; Pietermaritzburg

(— 1 CB), Stirton 1235; Effingham (— CC), Oliver 558; Baynesfield

(— CB), Stirton 702; Inanda (-DB), Wood 272; Chelmsford Park

(-DD), Hilliard 1845. 2931 (Stanger): Thring’s Post (-AA), Moll

2248; Tugela Bridge (-AB), Stirton 1268; Umhlali (-AC), Acocks

10331; Sheffield Beach (-AD), Grobbelaar 1641; Tongaat Beach

(-BA), Stirton 388; Tugela Beach (-BB), Johnson 382. 3030 (Port

Shepstone): near Ixopo (-AA), Hilliard 1771; Nhlavini Store

(-AB), Stirton 1224; Port Shepstone (-AC), Weeks 51; Umzinto

(-BC), Baijnath 419; Greenhart turnoff (-CD), Stirton 1403.

Lesotho.— 2828 (Bethlehem): Leribe (-CC), Dieterlen 221,

6845. 2927 (Maseru): Masoeling (-CB), Jacol Gui liar mod 1580.

Transkei. — 3127 (Lady Frere): near Engcobo (-DB), Flanagan

2819; between Nuamkwe and Engcobo (-DD), Bolus 8894. 3129

(Port St. Johns): Libode (-CA), Schonland 3890; Coffee Bay

(-CC), Tyson 20; Port St Johns (-DA), Baker 14164.

Vernacular names for this species include: Uqontsi

(Bryant, 1909), Zulu; Leshetla — soft bone, Lesapo

le letsehali — the female bone, and Setloli se sehola

— the big jumper (Phillips, 1917; Jacot Guillarmod,

1971), Sesotho. Both Zulus and Sesothos are

reported to use this plant for medicinal purposes.

Bryant (1909) has an interesting chapter on impoten-

cy and barrenness. He wrote: ‘With all primitive

peoples, all that pertains to the sexual functions, in-

volving as it does the propagation of the species and

the piCservation of the tribe, is a matter of para-

mount inportance. Impotency on either side is with

them more than a disgrace, it is a calamity. Should

the male organs fail altogether to produce the

seminal fluid, the roots of the imPindisa ( Rubia cor-

difolia) are boiled and drunk at bedtime, resulting in

an early emission. A hot milk infusion of the roots of

the Uqontsi ( Eriosema cordatum and E. saligrtum)

herb has a similar effect’.

Both Phillips (1917) and Jacot Guillarmod (1971)

reported that the Sesotho mixed the Leshetla with

other plants and used this as a medicine to stimulate

bulls in spring. Phillips (1917) noted further that E.

cordatum E. Mey. is less powerful a stimulant than

E. salignum E. Mey.

Eriosema cordatum is a plant of diverse habitats

and is quick to colonize disturbed habitats; areas

where it readily hybridizes with other species (Stir-

ton, 1981). Flowering occurs throughout the summer

months.

3. Eriosema iueipetum C. H. Stirton, sp. nov.,

E. cordato E. Mey. affinis stipulis liveris et floribus

luteis diversa.

Herbae perennes aut 1-foliolata erectae aut 3-

foliolatae decumbentes 10-30 cm altae, prope basin

ramosae. Foliola 8,5-11,0 cm longa, 4, 5-7,0 cm

lata, lateralia minora, basi cordata, apice acuta, ter-

minalia et 1-foliolata symmetrica, anguste usque late

ovata, lateralia minora, obliqua, strigosa vel sericea,

pagina inferiore opaca vel nitida, glandulis pallide

flavis vel rubris obsita. Stipulae liberae,

amplectentes. Flores lutei, raro rosei et lutei, bracteis

caducis, florum dimidio longitudinis. Calycis lobi ae-

quales, dentibus triangularibus vel lanceolatis tubo

longioribus ferrugineis vel sparse pubescentibus.

Vexillum et alae carina longiores. Fructus 15-17 mm

longus, 10-13 mm latus, oblique oblongus, pilis

patentibus 1,0-2, 5 mm longis vestitus. Semina 6-7

mm longa, 3 mm lata.

Type. — Natal, 2829 (Harrismith): 1 km from Fort

Mistake to Ladysmith (-BB), Stirton 1417 (PRE,

holo.).

292

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

Perennial herb, 10-30 cm tall, flowering in early

summer. Stems erect or decumbent, branching near

the base, tawny or ferruginous. Leaves mostly

1-foliolate on erect plants and 3-foliolate on decum-

bent plants, basal leaves always 1-foliolate. Leaflets

8,5-11,0 cm long, 4, 5-7,0 cm wide, laterals smaller,

base cordate, apex acute, terminal and 1-foliolate

leaves symmetrical, narrowly to broadly ovate,

laterals smaller and oblique, strigose or sericeous,

lower surface dull or shiny, covered with either light

yellow or red glands. Stipules 10-15 mm long, 4-5

mm wide, ovate-lanceolate, oblique, free, clasping

the stem, pubescent. Petiole 5-7 mm long. Racemes

axillary, (10) 20-40- flowered, exceeding leaves,

peduncle 5-12 cm long. Flowers yellow, rarely pink

and yellow, bracts more than half the length of the

flower, caducous. Calyx 6-7 mm long, lobes equal,

teeth triangular or lanceolate, longer than the tube,

C. H. STIRTON

293

ferruginous and sparsely pubescent. Standard (6)

8-12 mm long, 5-8 mm wide, oblong to obovate; ap-

pendages present, situated low down, free or merging

into auricles, rarely absent. Wings 7-10 mm long,

2-3 mm wide, cultrate, auriculate, sculpturing ab-

sent but slightly indented near auricle, longer than

the keel. Keel 6-8 mm long, 2-3 mm wide, densely

glandular, hairy along base, slightly pocketed.

Stamina l sheath 6-7 mm long, free stamen

geniculate; anthers uniform, alternately long basifix-

ed and short medifixed. Gynoecium 6 mm long;

ovary 2-3 mm long, with 0,75 mm long gynophore,

densely sericeous, curvature of style 2-3 mm high;

stigma capitate. Nectary present, high as gynophore,

margin undulating. Fruit 15-17 mm long, 10-13 mm

wide, with a downward pointing 2 mm long beak;

obliquely oblong, covered in 1-25 mm long patent

hairs. Seeds 6-7 mm long, 3 mm wide, polymorphic,

either black or light grey and speckled (only one

population inspected however). Fig. 15.

Eriosema lucipetum is distributed mainly in nor-

thern Natal and northern Kwazulu but also extends

as far north as Wakkerstroom in the Transvaal and

westwards to Paulina and Rensburgskop in the

Orange Free State (Fig. 16).

Transvaal.— 2730 (Vryheid): Wakkerstroom (-AC), Beeion

243.

Natal.— 2729 (Volksrust): 28 km from Newcastle to Volksrust

(-BD), Marsh 69; 10 km from Newcastle to Memel ( -DA), Stirton

1422; Ingogo (-DB), Shirlevs.n. NH 30759; 5 km from Newcastle

to Memel (-DB), Stirton 1420; 15 km from Newcastle to Lady-

smith (-DB), Stirton 1306; Koenigsburg (-DD), Thocle 4416. 2829

(Harrismith): Fort Mistake (-BB), Stirton 1417; Elandslaagte

(-BD), Shirley 303; Mpaleni Kop ( -DA), £9; 3 km from

Colenso to Paulpietersburg ( -DB), Stirton 1412. 2830 (Dundee).

Kelvin Grove (-AA), Medley Wood 5184; 8 km from Dundee to

Washbank (-AA), Stirton 1344; Mpati Mountain (-AA), Shirley

s.n. 2930 (Pietermaritzburg): New Hanover (-BC), Stirton 1309.

O.F.S.— 2828 (Bethlehem): Paulina (-AD), Thode 4422. 2829

(Harrismith): Rensburgskop (-AC), Jacobz 702. Without precise

locality, Thode 3332.

This species is easily separated from E. cordatum

by its free stipules. In areas where the two species

overlap there exist plants which may be intermed-

iates. Such plants, because they have free stipules,

are included here in E. lucipetum. They tend to have

pale pink and yellow flowers (for example Stirton

1416). Further studies may indicate that they are

hybrids.

Eriosema lucipetum is rather variable. Plants from

the westernmost part of the range are characterized

by their decumbent habit, leaves with upper surface

Fig. 16. — Known distribution of Eriosema lucipetum in South

Africa.

strigose and under surface shiny with a dense cover-

ing of small reddish glands, and calyx ferruginous

hispid (Fig. 17). The northern populations usually

comprise erect plants with sericeous leaves, the under

surface dull and variously covered in white or pale

yellow glands; the calyx being white or tawny veluti-

nous. The western plants may bear yellow, pink and

yellow and very occasionally red and yellow flowers

whereas northern and eastern plants are always

yellow-flowered. It is the western plants which

overlap in distribution with E. cordatum.

Stirton 1344 is a problem specimen to place in E.

lucipetum. Its general facies is that of the erect form

of the northern and eastern parts. It differs from

these however in its sharply acute ovate leaves and its

differently shaped standard petal that lacks appen-

dages on the inner face. Among the erect plants of

the eastern populations there is also an increasing

tendency towards the decumbent habit the further

east and south one goes. These areas are poorly col-

lected however and it may be that more than one

species may be involved. Collections are still needed

between the range of E. lucipetum and that of the

group of yellow-flowered specimens included tem-

po; arily under E. zuluense. (See under that species.)

The latter specimens are included under E. zuluense.

owing to their fused stipules and shortly-triangular

calyx teeth.

E. lucipetum is a plant of rocky grassveld, especial-

ly rocky banks and along roadsides, usually between

1 900-2 100 m. Flowering begins in October with a

peak in Novemer-December and then tails off

during March.

The specific epithet lucipetum refers to the seem-

ingly light-seeking habit of the inflorescences as they

emerge from their rocky habitat.

4. Eriosema psoraleoides (Lam.) G. Don, Gen.

Syst. 2: 348 (1832); Bak. f. & Haydon, Leg. Trop.

Afr. 508 (1929); Staner & De Craene in Annls Mus.

r. Coneo beige ser 6, 1: 52 (1934); Hauman in F.C.B.,

6: 206 (1954); Cufod. in Bull. Jard. bot. Etat Brux.,

25 suppl; 328 (1955); Hepper in F.W.T.A. ed 2, 1:

557, fig. 167 (1958); White, For. FI. N. Rhod. 151

(1962); Torre in Conspectus Flor. angol. 3: 337

(1966); Verde, in F.T.E.A. Legum.— Papil.: 772

(1971); Jacques-Felix in Adansonia, ser 2, 11: 157

(1971); Ross, FI. Natal 208 (1972). Type:

Madagascar, Commerson (P, holo.; K, photo.).

Crotalaria psoraleoides Lam., Encycl. 2: 201 (1786), as

‘psoraloides’.

Rhynchosia cajanoides Guill. & Perr. in Guill., Perr. & A.

Rich., FI. Sen.: 215 (1832). Type: Verdcourt cites type as Gambia,

near Gilfre not far from Albreda, Leprieur (P, holo.), but Felix

gives type as Senegal, Perrotet, (P, holo.).

Eriosema polystachyium E. Mey., Comm. 130(1836); Meisn. in

J. Bot., Lond. 2: 80 (1843). Type: South Africa, Caffraria, Drege

s. n. (K, iso.!).

Eriosema cajanoides (Guill. & Perr.) Hook. f. in Hook., Niger

FI.: 314(1849); Baker in F.T.A. 2: 227(1871); Harv. in FI. Cap. 2:

261 (1862); Medley Wood, Handbook FI. Natal 43 (1907); Harms

in Eng, Pflanzenw. Afr. 3, 1: 673 (1915); Bews, FI. Natal

Zululand 113 (1921); F.W.T.A. 1: 403 (1928); Bum Davy, FI.

Trans. 2: 413 (1932).

Eriosema proschii Briq. in Annu. Conserv. Jard. bot. Geneve,

6: 4 (1902); Bak. f. & Haydon, Leg. Trop. Afr. 497 (1929). Type:

Zambia, Barotseland, Sefula, Prosch 33 (E, holo.).

Perennial shrub or subshrub, 0, 6-2,0 m tall. Stems

erect or suberect, single or few arising from woody

rootstock, finely pubescent. Branches angular,

strongly ribbed; softly white pubescent or fulvocan-

escent, tomentose, more so on ridges, giving a cha-

294

THE ERIOSEMA CORDA TUM COMPLEX.

II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

Fig 17. Enosema lucipetum (western population). 1, habit; 2, stem with fruits and flowers, x 0,6; 3, free stipules, x 0 9; 4

under surface view of leaflet with glands densely packed along margin, x 1 ,2; 5, flower bract, x 9; 6, calyx opened out’

x 6,4; 7a, standard openend out, x 4,6; 7b, standard closed, x 4,6; 8, wing, x 4,6; 9, keel, x 4,6; 10, vexillar stamen’

x 6,4; 11, staminal sheath, x 6,4; 12, discoid floral nectary, x 18; 13, gynoecium, x 6,4; 14, stigma’, x 40.

C. H. STIRTON

295

racteristic zebra effect; densely or lightly inter-

mingled with numerous small white to yellow glands,

orange in older material. Leaves trifoliolate. Stipules

free, up to 5 mm long, often minute, patent or sharp-

ly down-flexed, becoming scariose even before sub-

tended leaf expands. Petioles 1-4 (5) mm long. Leaf-

lets (3,0) 4, 5-9,0 (9,5) cm long, 1,5-2, 2 (2,8) cm

wide, laterals smaller; subcoriaceous; narrow obo-

vate (2:1) to narrow oblanceolate (6:1), or elliptic

(2:1) to narrow elliptic (3:1), or oblong (2:1) to nar-

row oblong (3:1), base cuneate; margin almost

revolute; apex rounded or obtuse, less often acute or

emarginate; terminal leaflet symmetrical, laterals

almost symmetrical, base not oblique; finely

sericeous above, characteristically penninerved,

discoloured, velutinous beneath giving either a very

smooth silvery appearance or a buff and grey ap-

pearance; numerous orange glands present, often

obscured by dense indumentum. Young leaflets

hoary. Petiolules 2, 0-3,0 mm long, more densely

hairy than petiole or rachis, slightly swollen often

glandular. Rachis 1-5 mm long. Racemes 10-50-

flowered; subdistichous, usually lax; peduncles

2, 5-4,0 cm long, nearly always less than one third of

raceme, finely pubescent, rachis 6, 0-9, 5 cm long.

Flowers yellow, 11-13 mm long, about 5 mm wide.

Pedicels ± 2 mm, semi-erect 30° to axis. Pedicel

bract 2, 0-5,0 mm long, deciduous at bud stage.

Calyx 4, 0-5, 5 (6,0) mm long, less than half the

length of flower, lobes all equal, teeth length ± equal

to tube, triangular or minaret shaped, with keel tooth

sometimes narrower; puberulous inside, shortly

pubescent or sericeous outside, with dense orange

glands or eglandular. Standard 10,3-12,1 mm long,

(7.7) 8,0-10,3 mm wide; wide obovate (1,2:1); apex

retuse; glabrate or lightly puberulous; claw 1, 9-2,0

(2.8) mm long, width between auricles 3,4-5, 0 mm,

auricles 0,5 mm wide, prominent; appendages ab-

sent. Wings 10,1-11,5 mm long, 2, 1-2,7 (3,0) mm

wide; oblong (2:1) to lorate (6:1), cultrate, narrower

than but same length as keel blades; claw 2, 0-3,0

(3,5) mm long; auricle 1, 7-2,0 (3,0) mm high; heel

present. Keel blades (9,8) 10,0-11,0 (11,4) mm long,

(4.1) 4, 7-5, 8 mm wide; deep pocketed beneath auri-

cle; oblong (2:1); apex subacuate to rounded; claw

2, 1-3,2 mm long; auricle (1,8) 2, 0-3,0 mm high;

covered with numerous small orange glands, often

with scattered hairs along base line. Stamina! sheath

9,0-10,5 mm long, (2,5) 3, 5-4,0 mm wide, oblong

(2.1) ; vexillar stamen (8,8) 9,1-11,2 mm long, knee

up to 2 mm long. Gynoecium 9,5—11,0 mm long;

ovary 3, 5-4,0 mm long with 0,5- 1,0 mm stalk,

densely hairy and glandular but hairs hardly ex-

ceeding ovary; style very long, partly hairy on base

line, incrassated at point of upward flexure; cur-

vature 3, 0-3, 8 (4,0) mm high, stigma very small,

semi-globose, inward angled or erect, inserted.

Discoid floral nectary 0,5-0, 9 mm high, margin

smooth, slightly wavy. Fruit wide ovate (1,2:1) to

very wide oblong (1,2:1), 13-14 mm long, (7) 9-10

mm wide, coriaceous; beak down-turned, up to 1mm

long; densely white or ferruginous villous haired,

hairs erect on margins, mostly appressed on sides.

Seeds khaki or chestnut brown, 5, 4-5, 9 mm long,

3, 4-3, 9 mm wide, 1, 9-2,0 mm thick, rim-aril cream.

Eriosema psoraleoides extends from West Africa

across Central Africa to Sudan, then southwards

from east Africa across parts of Zambia to Angola,

through Zimbabwe, Mozambique, Botswana, Swazi-

land into South Africa (Transvaal, Natal). It also

occurs in Madagascar, the type country (Fig. 18).

Transvaal. — 2229 (Waterpoort): Wylies Poort (-DD), Hut-

chinson 2096. 2230 (Messina): Tate Vondo Forest Reserve (—CD),

Hemm 342; Makondo (-DA), Wespha I 38; Lukandwane store,

north of Sibasa (-DC), Codd 6881. 2329 (Pietersburg): 8 km

above Louis Trichardt (-BB), Rodin 3991. 2330 (Tzaneen):

Albasini Dam (-AA), Stephen 291; Tshakoma (-AB), Obermeyer

1054; Westfalia Estate (-CA), Scheepers 141; Tzaneen (-CC),

Rogers 12588. 2427 (Thabazimbi): Kransberg (-BC), Germis-

huizen 220; Rankins Pass (-DB), Acocks 23587. 2428

(Nylstroom): Near Alma (-AC), Clarke 36; 6 km south-east of

Palala (-BC), Story 1662; Nylstroom (-CBJ, Rogers 75; 12 km

from Nylstroom to Naboomspruit (-DA), Coetzer 16. 2429 (Zebe-

diela): Lunsklip (— AA), Maguire 1362. 2430 (Pilgrims Rest): Shil-

ouvane (-AB), Junod 766. 2431 (Acornhoek): 12,3 km from

Greenvalley to Nelspruit (-CA), Stirton 1450. 2526 (Zeerust);

Groot Marico (— CB), Holland 405/27. 2528 (Pretoria): Wonder-

boompoort (-CA), Leendertz 704; Bronkhorstspruit (-DD),

Grobbelaar 499. 2529 (Witbank): between Grobblersdal and

Marble Hall (-AB), Grobbelaar 32; Loskop Dam (-AD), Mogg

17242. 2530 (Lydenburg): Sabie Valley (-BB), Galpin s.n.;

Wonderkloof Nature Reserve (-BC), Elan Putlick 222; Nelspruit

(-BD), Buitendag 127; Mashonamini (-DA), Gross 87. 2531

(Komatipoort): Pretoriouskop (-AA), Van der Schijff 266; Mlam-

bane Dam (-AB), Brynard & Pienaar 4464; Plaston (-AC),

Graham 27; Mala Mala (-BA), Johnson 461; Malelane (-BC),

Brynard & Pienaar s.n.; Komatipoort (-BD), Rogers 428; Barber-

ton (-CC), Rogers 28568. 2628 (Johannesburg): Modderfontein

(-AA), Haagner s.n.

Swaziland. — 2631 (Mbabane): 15 km west of Mankaiana

(-CA), Compton 28662; Kubuta (-DC), Compton 28128.

Natal. — 2632 (Bela Vista): Ndumu Game Reserve (-CC), Ross

1972; Usutu Flood Plain (-CD), Tinley 532. 2730 (Vryheid):

Pongola (-BC), Marguts 115; Zungwini (-DB), Shirley s.n. 2731

(Louwsburg): I tala Nature Reserve (-AC), Brown & Shapiro 423;

Mbuzana River, 22 km from Magudu (-DA), Ross 1099. 2732

(Ubombo): Pongola river (-AA), Galpin s.n.; Mkuze (-CA),

Galpin 1371; False Bay (-CD), Gerstner 6756; Sodwana Bay

(-DA), Balsinhas 3223. 2830 (Dundee): Ingezeraan (-AB), Wood

15069; Qudeni Road to Tissiman’s farm (-DA), Hilliard 1379.

2831 (Nkandla): Buxedini (-BB), Guy 90; Eshowe(-CD), Gerstner

4157; 18 km from Eshowe to Gingingdlovu (-DC), Stirton 1296;

near Empangeni (-DB), Codd 1882; Mtunzini (-DD), Lawn 427.

2832 (Mtubatuba): Hluhluwe Game Reserve (-AA), Ward 1903;

Charters Creek (-AB), Baker 10010; Dukuduku Forest (-AC),

Strey 5539; Monzi Settlement (-AD), Strey 6567; Richards Bay

(-CC), Venter 5458. 2930 (Pietermaritzburg): Inanda (-DB),

Wood 160; Umlaas (-DA), Wood 5659; Umlazi (-DD), Wood

11147. 2831 (Stanger): 42 km from Stanger to Mtunzini (-AB),

Stirton 408; Chaka’s kraal (-AD), Thode 4415; Amatikulu (-BA),

Mogg 6289.

Eriosema psoraleoides is one of the most common-

ly collected Eriosema species in southern Africa. It is

a very distinctive species and is easily recognized. It

has been confused occasionally with E. salignum E.

Mey., E. burkei Benth. and E. nutans Schinz, but is

readily identified by its leaf shape and pubescence,

minute stipules, absence of appendages on the stand-

ard, and its congested, short peduncled inflores-

cence.

Fig 18.— Known distribution of Eriosema psoraleoides in South

Africa.

296

THE ERIOSEMA CORDATUM COMPLEX. IT THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

F'G l? ' ~-Erlose™a Psora leoides . I, branch with fruits, x 0,6; 2, inflorescence, x 0,6; 3, flower, x 4,6; 4, calyx opened out,

x 3,7, 5a, standard opened out, no appendages, x 4,6; 5b, standard closed, x 4,6; 6, wing, x 4,6; 7, keel x 4 6- 8 vex-

lllar stamen x 4,6; 9, starnmal sheath, x 4,6; 10, discoid nectary gland, x 18; 1 1, gynoecium, x 6,4; 12, stigma, x 40;

13, fruit pod, x 1,8; 14, seed with strophiole showing hilum, x 2,7.

C. H. STIRTON

297

The species is restricted to the lower-lying areas of

Swaziland, the Transvaal and Natal. It is usually

found along roadsides, disturbed crop lands and

forest margins, and along coastal sandflats in coarse

grassland with scattered shrubs. It has also been

recorded in swampy conditions. Field studies have

shown that although variable in different environ-

ments, it was nevertheless consistent in its important

diagnostic characters. Following the road from the

Malkerns Agricultural Research Station in Swaziland

to Mbabane, I noticed a decrease in size, branching,

and number of flowers per inflorescence. A similar

gradient occurs between coastal and inland plants.

Eriosema psoraleoides can produce flowers all year

round but its flowering is heaviest between October

to May with peaks between October to December and

March to May. During the period of the second peak

the Natal populations, as a whole, flower much later

than the Transvaal populations, reaching a peak in

April as compared with March. In the first peak the

Natal populations flower earlier in September as

compared with October - November for the Trans-

vaal.

On the Natal coast plants growing south of Man-

dini have been observed to flower at different times

to those plants growing north of the Tugela River (C.

J. Ward, 1975, pers. comm.).

The Venda common name for this species is muta-

ngasiwa ( Westphal 38, Codd 6881). In Afrikaans it is

referred to as the geelkleurtjie, an allusion, according

to Smith (1966), to their yellow flowers and general

appearance to the European Laburnum.

Both Gerstner (5002) and Moll (5345) recorded

that Tonga and Zulu tribesman cook and eat the ripe

seeds. A medicinal use has been observed by Codd

(6881) who noted that roots were used to cure inter-

nal disorders. These plants apparently are invoked by

some Swazi tribesman during storms as a protector

against lightning. A less obvious usefulness is the

reputed efficacy of their leaves and roots when used

as a remedy against venereal diseases; instances hav-

ing been noted in the Congo, Zambia and Nigeria

(Watt & Breyer-Brandwijk, 1962). The same authors

also report that in West Africa peasants rub leaves on

their dogs in order to control lice.

5. Eriosema buchananii Bak.f. in J. Bot., Lond.

33: 145 (1895); Bak. f. & Haydon, Leg. Trop. Afr.

505 (1929); Hauman in F.C.B. 6: 211 (1954); Verde,

in F.T.E.A. Legum.-Papil.: 776 (1971): Agnew,

Upland Kenya Wild Flowers 288, icon. (1974). Type:

Malawi, Mt Zomba, Buchanan 214 (K, holo.!).

E. richardii Bak. f. & Haydon var. ovatum Staner & De Craene

in Rev. Zool. Bot. afr. 24: 286 (1934). Type: Tanzania, Rungwe

District, Ukinga Mountains, Madehani, Stolz 2312 (K, holo.!).

Perennial herb or subshrub up to approximately

0,6 m, with simple or sparsely branched, erect stems

from a woody rootstock. Stems shallowly grooved

with scattered orange to red spherical glands, densely

covered with spreading ferruginous hairs, longer on

the ridges and shorter, erect to appressed, in the

grooves. Leaves trifoliolate, lower leaves often

unifoliolate. Stipules free, 1,0-1, 6 mm long, 2, 5-4,0

mm wide; lanceolate, base narrower than middle,

acinaciform (scimitar-shaped), often falcate; glab-

rous inside, ferruginous hairy outside. Petiole 2-8

mm long. Leaflets 3, 2-6, 8 cm long, 1,0-3, 5 cm

wide; laterals smaller 1,4-5, 9 cm long, 1,0-2, 5 cm

wide; elliptic (2:1), otherwise ovate (1,5:1) to ovate

lanceolate (3:1 or more) base slightly cordate, round-

ed, or truncate; apex acute, mucronate; margin en-

tire, revolute; veins below raised, chestnut coloured

and thickly appressed pilose in contrast to shorter,

erect pilose areas between veins; veins above thinly

appressed pilose; numerous small orange glands scat-

tered below, occasionally above; terminal leaflet

symmetrical; lateral leaflets asymmetrical, base obli-

que, width ratio 1:2. Young leaves ferruginous pilose

on veins, intervening areas packed with orange

glands and softly pubescent to pilose; petiolules

2. 0- 3,0 mm long, acropetiolar parts thickened,

blackish, hairs denser. Rachis 0,5— 1 ,5 mm long,

canaliculate. Racemes up to 40-flowered; peduncle

5-23 cm long, grooved, glandular, densely covered

with a mixture of long and short ferruginous hairs;

rachis 2-6 cm long; pedicels erect about 60° to axis,

1.0- 1, 5 mm long, recurved at apex, glabrous at base

just above bract abscission scar. Calyx (4,0) 4, 5-5,0

(5,7) cm long; externally spreading pilose and gland

dotted, with shorter hairs near the base, and longer

denser hairs on the lobes and creases; internally fine-

ly hairy on lobes but restricted mainly to veins and

the horn lobes; tube 2, 0-3,0 mm long, being longer

between the horn lobes, 2, 3-3, 3 mm long, circumfer-

ence 5, 0-5, 3 mm, tissue distinctly bulged between

the horn lobes when the calyx is flattened; lobes

triangular unequal, horn lobes partly connate,

4. 2- 5, 2 mm long; lateral lobes 3, 9-5,0 mm long,

keel lobe 4, 0-5, 7 mm, longest; lobes more or less

equalling the tube. Standard very pale pink lined with

dark purple; 7,7-10,7 mm long, 3, 0-4, 6 mm wide;

oblong (2:1), narrowed in the middle, appendages oc-

cur above the middle and are slight ridges often diffi-

cult to see; claw 2, 2-3, 6 mm; auricles prominent, in-

curved, 3, 5-4, 5 mm apart. Wings 7, 3-9,0 mm long,

1, 8-2,0 mm wide, longer than keel, ratio 1:1,2; nar-

rowly oblong (3:1); claw 2, 5-3, 5 mm, auricle

1 .2- 1,9 mm high. Keel blades 6, 1-8,0 mm long,

2. 0- 2, 6 wide at maximum; claw 2, 0-3,0 mm long; a

few hairs along baseline, glandular. Staminal sheath

5. 2- 6, 5 mm long, 1,9-2, 6 mm wide at maximum,

vexillar stamen 4, 5-6, 2 mm long, basal knee 0,5-0, 8

mm long. Gynoecium 5, 0-7, 2 mm long; ovary

2. 0- 3,0 mm long, with stalk 0,3—1 ,0 mm long; hairs

on ovary wall up to 2,5 mm long, distinctive, wavy,

not stiff; style slightly bent and thickened towards

end; curvature 1,1 -2,0 mm high; stigma small

capitate, inserted. Discoid floral nectary 0,3-0, 6 mm

high, margin slightly wavy, often revolute. Fruits and

seeds not seen. Fig. 20.

In South Africa this species is confined to a few

localities in the northern Transvaal (Fig. 21). It ap-

pears to be restricted to the North-eastern Mountain

Sourveld Veldtype. It occurs between 900 and 1 490

m. The Transvaal is the southernmost limit of distri-

bution. It is also recorded from Kenya, Uganda, Tan-

zania, Zambia, Malawi and Zimbabwe.

Transvaal.— 2329 (Pietersburg): Haenertsburg (-DD),

Eastwood 13 & Pott 4635. 2330 (Tzaneen): Duivelskloof (-CA),

Scheepers814; Magoebaskloof (-CA), Papendorf 275; Woodbush

(-CC), Mogg 13992.

Verdcourt (1971) in his treatment of Eriosema for

the Flora of Tropical East Africa reported that he

had seen only one specimen of var. buchananii from

South Africa. This specimen, Scheepers 814, has pro-

ved however to be the most atypical of the South

African specimens. It has larger more distinctly ovate

leaves, longer peduncles as well as a shorter lighter

patent indumentum. This may be due to environmen-

tal influences as similar attributes have also been

noted in field studies in other species in the E. cor-

datum complex. Scheepers 814 groups well with the

298

THE ERIOSEMA CORDATUM COMPLEX. IE THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

Fig 20.— Eriosema buchananii var. buchananii. 1 , branch with flowers, x 0,6; 2, stipule, x 4,6; 3, flower bract, x 9; 4, calyx

opened out, x 6,4; 5a, standard opened out, x 4.6; 5b, standard closed, x 4,6; 6, wing, x 4,6; 7, keel, x 4,6; 8, vexillar

stamen x 4,6; 9, staminal sheath, x 4,6; 10, discoid floral nectary, x 18; 1 1, pistil, x 6,4; 12, stigma, x 40.

C. H. STIRTON

299

following specimens from outside South Africa:

Chaje 682 (Zimbabwe); Corby 1430 (Zimbabwe);

Hilliard & Burtt 4139 (Malawi); Pawek 1564, 8046

(Malawi); Richards 2362/A (Tanzania); Semsei 1652

(Tanzania). These specimens might deserve some in-

fraspecific ranking. Verdcourts inclusion of var.

buchananii in the E. cordatum E. Mey. complex can

be understood from his comments on the Scheepers

814 specimen. He referred to Scheepers 814 as being

a ‘small rounded shrub of upright habit with

numerous ascending to suberect branches from the

base, corolla very pale pink, veined with dark red

purple’. In differentiating between E. cordatum and

E. buchananii in South Africa, he made some rele-

vant remarks, based as it turned out, on a specimen

atypical of its range. He commented: ‘This is most

interesting because the same collector has collected

true E. cordatum at Letaba [Scheepers 487, 3 Oct.

1958 (K; PRE)]; he described the flowers as yellow

veined with dull red, the standard dull red outside.

The two are strikingly similar in general appearance

save that Scheepers 814 has much denser patent in-

dumentum on the stem. The two are, however, equal-

ly strikingly different in their standards; not only is

there the difference in the appendages already

pointed out, but the shape is also quite different in

each. E. buchananii has a distinctly rectangular stan-

dard with the auricles pronounced, above which it is

very slightly narrowed; E. cordatum has a distinctly

obovate standard, gradually narrowed into the claw,

the auricles being less obvious. Moreover the claws of

the other petals are much longer in E. buchananii

than they are in E. cordatum’.

Scheepers (1974 pers. comm.) felt sure his Nos 487

and 814 were different — ‘certainly they are very

distinct in habit’. He commented on E. buchananii’ s

‘more stiffly upright habit’ as opposed to E. cor-

datum’s ‘characteristically laxly sprawling habit’.

These comments hold for all the other specimens

of E. buchananii in South Africa which are clearly

distinct and easily separable from E. cordatum. I

suspect that their relationship is probably not as close

as previously envisaged.

In South Africa E. buchananii has been confused

more often with two other species rather than with E.

cordatum. Here it has been known under the name E.

nutans Schinz [ = E. polystachyum (A. Rich.) Bak.]

to which it has a very strong resemblance, and also to

E. montanum Bak. f. which does not extend as far

south as South Africa. E. buchananii may be readily

separated from the rest of the complex by the

Fig. 21.— Known distribution of Eriosema buchananii in South

Africa.

presence of an appendage situated above or at the

middle of an oblong standard, the large incurved

flattened auricles and the falcate stipules with the

base narrower than the middle. The appendage is

characteristic as it is divided into two shallow,

crescentiform ridges which lie closely one on either

side of the central axis of the standard. E. nutans is

separated by its smaller auricles and its undivided ap-

pendage which occurs just above the auricles merging

into them.

Apart from the comment by Mogg (13992) that his

plant was ‘common in grassland ridges of gneiss’,

there is very little ecological information for the

species in South Africa. Scheepers 814 has the

following note ‘north facing slope, intense sunlight

to misty or windy, periodically moist well drained

and aerated to dry, arid shallow soil tending toward

lateritic type’.

E. buchananii flowers from November to January,

with a peak in December.

There is another variety in east Africa, var sub-

prostratum Verde. This does not occur in South

Africa. Gillett 17669 (K) from Tanzania, which has

been included under var. subprostratum seems to me

to be worthy of some rank. It has smaller flowers,

lacks the typical buchananii stipules and has a dif-

ferent pubescence.

6. Eriosema nutans Schinz in Bull. Herb.

Boissier ser 2, 8: 629 (1908); Bak. f. Leg. Trop. Afr.

1: 505 (1929); Burtt Davy, FI. Transv. 2: 15, 413

(1932); Verde, in Kew Bull. 25: 121 (1971). Type:

South Africa, Transvaal, Shilouvane, Junod 2165 (Z,

holo.!).

E. richardii Bak. f. & Haydon, Leg. Trop. Afr. 505 (1929).

Type: Ethiopia, without locality, Schimper 708 (K, holo.!).

E. richardii Bak. f. & Haydon forma ellipticum Staner & De

Craene in Rev. Zool. Bot. afr. 24: 286 (1934). Type: Kenya Gishu,

Harvey 71 (K; holo.! PRE, iso.!).

E. buchananii Bak. f. var. richardii (Bak. f. & Haydon) Staner

forma ellipticum (Staner & De Craene) Staner in Kew Bull. 278

(1935).

E. polystachyum sensu auett. e.g. Bak. in F.T.A. 2: 225 (1871);

Burtt Davy, FI. Transv. 2: 413 (1932); non (A. Rich.) Bak., nec E.

Mey., non Rhynchosia polystachya A. Rich.

Perennial herb or subshrub 30-90 cm tall from

underground rootstock. Stems erect or decumbent,

branched, grooved, densely or lightly covered with ±

spreading white or ferruginous hairs and shorter cur-

ly hairs on ridges, ± glabrous in troughs. Leaves

trifoliolate; stipules free, 5-9 mm long, 1,0-1, 5 mm

wide, linear lanceolate or triangular, straight or

falcate, closely veined, not recurved, appressed hairy

outside, glabrous inside ± glands. Petioles 3-7 mm

long. Leaflets 4-8, cm long, 1 ,3-2,8 cm wide, laterals

much smaller; narrow ovate (2:1) to lanceolate (3:1),

or elliptic (2:1); apex acute or rounded; base cuneate

or truncate, margin revolute; terminal leaflet sym-

metrical, laterals almost symmetrical, base mostly

oblique; shortly appressed pilose above, appressed

pilose on raised venation beneath but sparsely erect

pilose between veins; both surfaces covered with

either large orange or small yellow glands or both

without. Young leaflets shaggy white or ferruginous,

± densely packed with glands. Petiolules 1-2 (3) mm

long. Rachis (0,7) 0,8-1, 1 (1,2) cm long. Racemes

axillary, 10-43-flowered; peduncle 4,0-12,2 cm

long, densely or sparsely ferruginous or white hairy;

rachis (3,2) 4, 0-7, 2 (11,1) cm long; Flowers yellow

(7,0) 7, 5-9,0 (10,0) mm long, (2,5) 3, 0-3, 4 mm

wide, reflexed; bract 3, 0-5, 8 mm long, lanceolate

slightly boat-shaped, caducous, hairy outside,

300

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

Fig 22. — Eriosema nutans. 1, habit; 2, branch with fruits and flowers, x 0,6; 3, flower bract, x 9; 4, flower, x 4,6; 5, calyx

opened out, x 9; 6a, standard opened out, x 4,6; 6b, standard closed, x 4,6; 7, wing, x 4,6; 8, keel, x 4,6; 9, vexillar

stamen, x 9; 10, staminal sheath, x 9; 11, discoid floral nectary, x 18; 12, gynoecium, x 9; 13, stigma, x 40; 14, fruit

pod, x 2,8.

C. H. STIRTON

301

glabrous inside; calyx (3,0) 3, 5-4, 3 (4,5) mm long,

lobes ± equal, deltoid or triangular, ± yellow or

orange glands, shortly pilose, long hairs if present

mostly at tips or along veins; tube (1,5) 2,0 (2,9) mm

long, longer between horn lobes (1,9) 2, 2-2, 8 (3,6)

mm long; lateral lobes (3,0) 3, 5-4,0 (4,5) mm long;

keel lobe (3,0) 3, 5-4, 2 (4,5) mm long, horn lobes

longest (3,1) 4, 0-4,3 (4,8) mm long, tips straight, oc-

casionally incurved; basal circumference 4, 0-5,0

mm. Standard (6,7) 7, 0-9, 2 (10,0) mm long, (2,5)

2, 8-4,0 (4,5) mm long, mostly oblong (2:1), densely

hairy on back, glands present or absent; claw (1,0)

1, 5-2,0 (2,8) mm long; width between auricles (2,2)

3. 0- 3, 3 (4,0) mm; auricles prominent; appendage

situated low down and extending across auricles,

bifurcate, two upcurled flaps 2, 7-3, 6 (4,0) mm from

base of claw. Wings (4,2) 6, 9-7, 8 (8,8) mm long,

(0,9) 1, 0-1,1 (1,6) mm wide at maximum, about

same length as keel blades, sparsely glandular and

hairy on outside, claw (1,7) 2, 0-2, 5 (3,2) mm long;

auricle 1,0- 1,2 (1,6) mm high. Keel blades (6,0)

7. 0- 8,0 mm long, (0,9) 2, 5-2, 8 (3.3) rnm wide at

maximum, 2-3 times wider than wing, either densely

packed with glands or eglandular, prominently

hairy along base line up curvature to apex, claw

2. 0- 2, 5 (3,0) mm long, auricle 1,2-1, 5 (1,8) mm

high. Staminal sheath (5,5) 6, 0-7,0 (7,1) mm long,

(1.2) 1,9-2, 2 mm wide at maximum, vexillar stamen

(5.2) 6, 0-6, 8 (7,0) mm long. Gynoecium (5,4)

6. 0- 7, 2 (8,5) mm long; ovary (2,2) 2, 8-3,0 mm long

with short haired stalk, 0,2-0, 5 mm long; glandular

with long silky sericeous hairs reflexing at tips and

extending as far as thickening in style; style filiform

but thickened at point of flexure, curvature (1,5)

2.0- 2, 1 (2,2) mm high; stigma minutely capitate, in-

serted. Discoid floral nectary with margin crenulate.

Fruits ovate-oblong, 10-12 mm long, 6-8 mm wide,

beak oblique 0,5-1, 1 mm long, covered with long

soft ferruginous hairs and short hairs, longer hairs

mostly at top and sides, short appressed hairs distinc-

tive along base line as it incurves towards stem. Seeds

3. 0- 3, 5 mm long, 1, 6-2,0 mm wide; black,

strophiole cream or white. Fig. 22.

In South Africa E. nutans is confined to the

Transvaal and Swaziland (Fig. 23). Elsewhere it oc-

curs in Zaire, Sudan, Ethiopia, Eritrea, Kenya,

Uganda, Tanzania, Malawi, Zambia, Mozambique

and Zimbabwe.

Transvaal.— 2329 (Pietersburg): Louis Trichardt (-BB),

Breyer s.n.; Bloodriver (—CD), Hay s.n. 2330 (Tzaneen): Elim

(-AA), Schlechter 4550; Westfalia Estate (-CA), Grobbelaar 426;

Modjadjes Reserve near Duiwelskloof (-CB), Krige 29; New

Fig 23. — Known distribution of Eriosema nutans in South Africa.

Agatha (-CC), McCallum s.n. 2428 (Nylstroom): Tarentaalpas

(-AD), Clarke 360. 2430 (Pilgrims Rest): The Downs (-AA),

Junod 4354, Shilouvane (— AB), Junod 2165, 4896; Pilgrims Rest

(-DD), Rogers 14510. 2528 (Pretoria): Pretoria (-CA), Repton

241. 2530 (Lydenburg): Lydenburg (-AB), Galpin 12194; Mount

Anderson (-BA), Galpin 13 720. 2627 (Potchefstroom):

Krugersdorp (-BB), Jenkins 9228. 2628 (Johannesburg): Melville

Koppies (-AA), McNae 1455; Boksburg (-AB), Murray s.n. 2629

(Bethal): Standerton (-CD), Rogers 4813.

Swaziland. — 2631 (Mbabane): Mliwane Game Reserve (-AC),

Arnold 832.

Burtt Davy (1932) accepted both E. nutans Schinz

and E. polystachyum (A. Rich.) Bak. He remarked

that E. nutans Schinz ‘is a close ally of E.

polystachyum (A. Rich.), Bak. f. ms’. Under E.

polystachyum (A. Rich.), Bak. he stressed its close

affinity to E. richardii Benth. ex Bak. f. Later,

however, he cited Rogers 14510 as E. polystachyum

but labelled it E. richardii. His treatment of these

various taxa can be understood and appreciated after

reading a note on pp 15-16 of his flora. In this note

on ‘Synonyms and References’ Burtt Davy com-

mented on some of the difficulties he encountered

with synonomy. His example fortuitously concerned

Eriosema polystachyum. He said: ‘For the sake of

space, also, only so much synonomy has been given

as seemed requisite to correlate the species with the

“Flora Capensis” and the “Flora of Tropical

Africa”, or with more recent revisions and

monographs. The principle has been adopted that the

same name should not be used for more than one

plant, even though the first name so applied has been

reduced to the rank of a synonym. This rule has

much to commend it in view of the fact that dif-

ferences of opinion will always exist as to the correct-

ness of reducing a name to the status of a synonym,

and the consequent liability that the plant to which it

was first applied will be restored to specific rank. An

example in point is that of the Tropical African

Eriosema polystachyum (A. Rich.) Bak. (1871);

Rhynchosia polystachya A. Rich. (1847); this is

antedated by Eriosema polystachyium E. Mey.

(1835), which is a distinct species, of the Transkei.

Although Meyer’s name has been reduced to a

synonym of E. cajanoides (Guill. & Perr.) Benth.

Rhynchosia cajanoides Guill. & Perr. (1833), it is

quite possible that the Transkeian E. polystachyium

is not conspecific with E. cajanoides, in which case

Meyer’s name would have to be restored and

Richard’s name would have to be changed, and with

consequent confusion. Propably anticipating this,

Bentham proposed in mss, the name Eriosema

richardii, which is adopted here’. [N.B. E.

polystachyium E. Mey. is a synonym of E. psora-

leoides (Lam.) G. Don].

E. nutans has been interpreted differently in dif-

ferent parts of Africa, a not unusual result of the

slow, spasmodic development of African botany.

This name has been used rarely in South Africa. Here

the most popular name has been E. polystachyum

(A. Rich.) Bak. In referring to practice outside South

Africa, Verdcourt (1971) made the following com-

ment: ‘During the past thirty years this (E. nutans

Schinz) has mostly been accepted as a variety of E.

buchananii but I am certain that the two are best kept

distinct, the latter differing in large leaves, much

broader stipules, different position of the standard

appendages and less constancy in flower colour. The

flowers of E. nutans are yellow' but only a few plants

of E. buchananii have been seen with yellow flowers;

they are mostly cream, pink, or purple lined with

purple’.

302

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND £. NUTANS GROUPS

As the presence of E. buchananii has now been

firmly established in the Transvaal, I might add that

Verdcourt’s comments hold for South Africa, except

that here plants of E. buchananii have not been

recorded as yellow-flowered.

Over 50 specimens had been incorrectly named in

the past. This material had been referred to E.

psoraleoides (E. cajanoides), E. parviflorum, E.

zeyheri and E. squarrosum. E. psoraleoides with its

very small stipules, wide obovate standard without

appendages, different keels, wings and pistil is easily

separated from E. nutans which has an oblong stan-

dard with appendages and long lanceolate stipules.

The two species differ also in that in E. nutans the

peduncle is longer than half the length of the in-

florescence. In E. psoraleoides the peduncle is less

than one third the length of the inflorescence. E. par-

viflorum is distinguishable from E. nutans by its wide

obovate (1,2: 1) standard, and its allopatric distribu-

tion (almost restricted to Natal in S. Africa).

The most noticeable variation which occurs in E.

nutans is the presence or absence of glands. Two

types of glands have been found; large and small red

glands and yellow glands. Gland colour may be the

result of ageing or an artefact of the poisoning of

herbarium specimens. Glands are mostly confined to

specimens from the northern Transvaal. Glandular

and non-glandular plants both extend from Tanzania

to the Transvaal. Calyces are quite variable ranging

from sparsely short pilose and short toothed, to long

toothed and densely short pilose. Leaf shape and size

are most variable. Stipules range from short

triangular and slightly falcate, to long linear-

lanceolate and straight. This variation may be worthy

of infra-specific categorization. However, in view of

the numerous difficulties of synonmy which have

resulted from regional demarcations of infra-specific

ranks, I have left this to later workers who may see

the species in a broader African perspective.

E. nutans grows mainly in the bushveld and open

grasslands and has been found on south-facing

hillsides, in disturbed areas such as dipping tanks, in

cultivated fields and near marshy ground. It flowers

from November to May with a definite peak in

February to March.

The Venda name for this plant is reported by the

Reverend Rogers (No. 4813) to be Mutzikedzi.

7. Eriosema parviflorum E. Mey., Comm. 130

(1836); Meisn. in J. Bot., Lond. 2: 80 (1843) Harv. in

FI. Cap. 2: 260 (1862); Bak. in F.T.A. 2: 225 (1871);

Bak. f. in J. Bot., Lond. 33: 142 (1893); Wood &

Evans, Natal Plants 1: 73-74 (1899); Harms in

Warb., Kunene-Samb. Exped.: 265 (1903); Medley

Wood, Handbook FI. Natal 43 (1907); Bews, FI.

Natal Zululand 1 13 (1921); Bak. f. & Haydon in Leg.

Trop. Afr. 504 (1929); Staner & De Craene in Annls

Mus. r. Congo beige ser 6, 1: 60, fig. 9 (1934); et in

Rev. Zool. Bot. afr. 24: 288 (1934); Rossberg in Fed-

des Reprium 33: 166 (1936); Hauman in F.C.B. 6:

211 (1954), pro parte; Hepper in F.W.T.A. ed. 2, 1:

557 (1958), pro parte; Verde, in Kew Bull. 25:

124 - 125 (1971); Verde. in F.T.E.A.

Legum.-Papil.: 778 (1971); Jacques-Felix in Adan-

sonia, ser 2, 11: 169- 170 (1971); Ross, FI. Natal:

208 (1972). Lectotype: South Africa, Natal, between

Umzimkulu River and Umkomaas River, Drege (K ! ).

The synonomy of Eriosema parviflorum is puzz-

ling. There is little agreement among the three latest

studies, and a: the chosen example below indicates, it

is a problem the solution of which rests in a full

African study. The chosen example concerns the

treatment of the name Eriosema parviflorum var.

sarmentosa Staner & De Craene. Keay (1973) assign-

ed this to E. parviflorum subsp. parviflorum,

whereas Verdcourt (1971a) placed it under E. par-

viflorum subsp. podostachyum (Hook, f.) J. K. Mor-

ton. Jacques-Felix (1971), however, excluded it from

E. parviflorum completely and made it a synonym of

E. spicatum Hook. f. Unlike both Keay and Verd-

court, he did not recognize infraspecific categories in

E. parviflorum. This example forms part of a con-

fused pattern. Rather than add to this confusion I

have decided not to recognise any infraspecific

categories in the region under study until the species

has been studied over its entire range.

Perennial spreading suffrutex up to 60 cm high,

arising from underground woody rootstock. Stem

erect, branching near base, apical growth arrested

with development of branches. Branches ascending

or decumbent, up to 1 m long, slightly ridged; thinly

white pubescent or densely shaggy with deflexed fer-

ruginous hairs on older parts but admixed with

smaller white hairs on younger; white or red glan-

dular on younger parts. Leaves rarely unifoliolate (at

base), trifoliolate, exhibiting nastic response.

Stipules free, (4) 5-8 (10) mm long, 1, 0-2,0 mm

wide, spreading and reflexed when old turning red

brown as leaflets expand; lanceolate, linear, or if

tear-shaped then blackish at base. Petiole (1,1)

1,3-2, 2 (3,0) cm long, canaliculate, vesture thickest

on ridges. Leaflets 2, 0-6, 2 cm long, 1,0-3, 2 cm

wide, laterals smaller but 1/w ratio less, elliptic (2:1)

or wide elliptic (1,5:1), occasionally tending to sub-

orbiculate (1,2:1), narrow obovate (2:1); base round-

ed, if tapered then narrow truncate or narrow cor-

date; apex obtuse to rounded, rarely acute; margin

entire; terminal leaflet symmetrical, laterals asym-

metrical, base oblique; finely pubescent on both sur-

faces, with white or orange glands. Young leaflets

densely pubescent, hairs longer appressed on raised

veins, with densely packed red or white glands.

Petiolules 1-2 mm long, swollen, terete, shaggy and

glandular. Rachis 0,2-0, 6 mm long. Racemes 5-54

flowered; peduncles elongate, (5) 7-13 (16) cm long,

shortly densely spicato-racemose at summit, deflexed

white or ferruginous hairy; rachis (1,2) 2-5 (6) cm

long. Flowers yellow, 5-6 mm long, ± 2,5 mm wide;

pedicels 1,0-1, 5 mm long, recurved; bracts 2, 5-4,0

mm long, shortly triangular, caducous. Calyx lobes

(2,0) 2, 5-3,0 mm long, half length of flower; lobes

equal; teeth shorter than or equalling tube, triangular

with keel tooth often pinched into finger-like pro-

cess, pubescent with short appressed hairs, occa-

sionally longer haired especially near extremities,

glandular. Standard (5) 6-7 (8) mm long, 3, 0-4, 2

(4.6) mm wide, wide obovate (1,2:1) occasionally

narrow obovate (2:1), apex ± rounded; 1/b ratio

1,5- 1,8; glabrescent or sparsely pubescent and glan-

dular outside; claw 1, 9-3,1 mm; width between

auricles 1, 0-2,0 mm, inflexed auricles very promi-

nent; appendage situated low down and extending

across auricles, (2,0) 2, 2-2, 4 (2,7) mm from base of

claw. Wings (4,9) 5, 0-6,0 (6,7) mm long; 2, 0-2, 5

mm wide at maximum, shape variable when in posi-

tion, shorter than keel blades, glabrous or pubescent,

claw (1,0) 1, 8-2,0 (2,1) mm long; auricle 1, 0-1,1

(1,3) mm high, ‘step’ present. Keel blades 5, 2-6,0

(6.6) mm long, 2, 0-2, 6 mm wide at maximum, ob-

tuse slightly incurved, densely glandular; claw (1,1)

2,0-2, 1 mm long; auricle 1, 0-1,1 (1,5) mm high.

Stamina! sheath (4,0) 5, 0-5, 5 (5,8) mm long, 2, 0-2, 2

(2.7) mm wide at maximum, oblong (2:1); vexillar

stamen (4,0) 5, 0-5, 3 (6,0) mm long. Gynoecium

C. H. STIRTON

Fig 24 —Eriosema parviflorum. 1, habit, 2, branch with fruits and flowers, x 0,6; 3, leaves epinastic, night position, x 0,6;

4 flower bract x 9; 5, flower, x 4,6; 6, calyx opened out, x 9; 7a, standard opened out, x 4,6; 7b, standard closed, x

4 6- 8 wing x 4 6- 9 keel x 4,6; 10, vexillar stamen, x 6,4; 11, staminal sheath, x 6,4; 12, discoid floral nectary, x

1 8 • ’ 13 gynoecium ’ x’ 6,4; 14, stigma, x 40; 15, fruit, x 2,1; 16a, black seed with strophiole, face view, x 6,4; 16b,

black seed with strophiole, marginal view showing hilum, x 6,4; 17a, khaki, grey speckled seed with strophiole, tace

view, x 6,4; 17b, khaki, grey speckled seed with strophiole, marginal view showing hilum, x 6,4.

304

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORD A TUM AND E. NUTANS GROUPS

5,0-6, 1 mm long; ovary (2,l)-2,6 (3,0) mm long,

stalk 0,2-0, 5 mm long shortly hairy but glabrous

underneath; densely covered with glands and yellow

or white hairs extending almost to point of curvature;

style glabrous, variable in shape, aberrations com-

mon, thickened at curvature, incurved or erect, cur-

vature (1,0) 1 ,4-1,8 (2,1) mm high; stigma small,

capitate, sometimes incurved; inserted. Discoid floral

nectary folded, slightly rippled. Fruits 0,9-1, 2 cm

long, 0,6-0, 7 cm wide; obliquely oblong (2:1) to

wide oblong (1,5:1), beak straight or slightly upturn-

ed, up to 1 mm long; strongly compressed; shaggy

with foxy hairs and fine pubescence, hairs longer and

wispier on sides, glandular. Seeds 3, 5-3, 7 (4,0) mm

long, (2,3) 2, 5-2, 6 mm wide, 1,1-1, 5 (2,1) mm

thick, polymorphic, khaki with purple flecks and

speckles, or chestnut, or black, rim aril white black

seeds longer but thinner and narrower than speckled

seeds. Fig. 24.

Eriosema parviflorum extends eastwards from

west Africa across the Camerouns and Congo to the

Sudan and east Africa, then sourthwards through

Zambia and Mozambique into South Africa (Trans-

vaal and Natal) (Fig. 25). It also occurs in Madagas-

car.

Transvaal. — 2230 (Messina): Sibasa, north of Pepiti (-CD),

Smuts & Gil let t 3287.

Natal. — 2632 (Bela Vista): Kosi Estuary (-DD), Moll & Strey

3685. 2732 (Ubombo): Sordwana Bay (-BC), Stephen, Van Graan

& Schwabe 1095; Lake Sibayi (-DA), Vahrmeijer & Hardy 1625.

2831 (Nkandla): Eshowe (-CD), Lawn 84; Empangeni (-DB),

Venter 1894; Ngoye Forest Reserve (-DC), Huntley 250; Mtunzini

(-DD), Stirton 413. 2832 (Mtubatuba): Palm Ridge Farm (-AC),

Harrison 382; Richard Bay (-CC), Stirton 534. 2930 (Pieter-

maritzburg): Shongweni (-CD), Ross 778; Inanda Mountain

(-DB), Hilliard 2043; Isipingo Flats (-DD), Ward 6807. 2931

(Stanger): Between Blythedale Beach and Tugela Bridge (-AD),

Stirton 396, 795; Berea, Durban (-CC), Medley Wood 1448. 3030

(Port Shepstone): 8 km from Winklespruit along Eston road

(-BB), Stirton 1123; St Michaels-on-Sea (-CD), Nicholson 773;

Hibberdene (-DA), Grobbelaar 67.

Transkei.— 3129 (Port St Johns): Magwa Falls (-BC), Galpin

10985; Port St Johns (-DA), Hilliard 1066, 1111. 3130 (Port

Edward): Mnyameni Mouth (-AA), Acocks 13396.

Eriosema parviflorum E. Mey. is mostly restricted

to the alluvial coastal plain of Natal. It grows in open

moist grassland, coastal forest margins, damp

marshland and bushy sand dunes. It seems to favour

a south-west aspect and an altitude range of 5-650

m.

There is a variation in robustness and leaf size in

this species. Figs 26 & 27 show the ranges possible.

Fig. 26. — Huntley 250, large-leaved form of E. parviflorum col-

lected in the Ngoye Forest Reserve (1963-02-07).

Fig 25. — Known distribution of Eriosema parviflorum in South

Africa.

Fig. 27. — Ross 778, small-leaved form of E. parviflorum collected

at Shongweni (1964-03-08).

C. H. STIRTON

305

The large leaved forms are rare. Stem pubescence,

gland colour, size of stipules and number of flowers

also vary to some extent, but apart from this the

species is fairly constant throughout its range.

In Natal E. parviflorum has a tendency to produce

floral aberrations. I have recorded and preserved an

unusually large number of cases in which up to three

standards were present in the same flowers. The most

common aberration found was a variation in the

shape of the pistil. Two forms occurred, sometimes

in the same plant. The area of thickening may be at

the point of flexure or above it. Beyond the point of

flexure the style may be erect or strongly recurved.

This may constitute an example of cleistogamy.

Another common aberration is the lack of, or partial

fusion of the stamens into a staminal sheath. Keel

blades are often absent. This information is recorded

in case difficulties are encountered in keying out the

species. The wing is an unreliable character.

An unusual feature is the occurrence of ‘sleep

movements’. I have been unable to determine

whether these are photonastic or thermonastic. They

are definitely not thigmonastic. Fig. 24.2 and 24.3

shows the same plant during the day and later during

the night. The leaflets expanded during the day can

be seen to have closed tightly to the petioles at

nightfall.

This species was recorded by Phillips (1917) for

Lesotho. He based this citation on Dieterlen 866. A

study of this specimen has revealed that it has

characters intermediate between those of E. cor-

datum var. cordatum and E. salignum E. Mey. The

possibility of hybridization cannot be ruled out (Stir-

ton, 1981). Plants such as Dieterlen 866 were pro-

bably common since Phillips (l.c.) referred to it by its

Lesotho name, ‘Leshetla la loti’, a name different

from those he recorded for the putative parents. I am

certain that his specimen is not E. parviflorum E.

Mey. and that this species does not occur in Lesotho.

E. parviflorum has no close affinities in South

Africa. It has often been confused with E. nutans

Schinz, but differs from that species in the shape of

the standard and the shape of the seeds.

Plants flower from early September to as late as

April.

ACKNOWLEDGEMENTS

The author wishes to thank the following: the

curators and staff of PRE, NH, GRA, STE, NU,

BOL, NBG, P, BM, K, and G for their help and loan

of much valuable material; the Department of

Agriculture and Fisheries for financial assistance; L.

Cowan, B. Pascoe and P. Waugh for illustrations;

the staff of the Botany Department, University of

Natal (where most of the study was completed)

especially L. Smook, D. Tunnington, M. Brand, J.

Oliver, M. Ram, M. Moodley, T. H. Arnold, S.

Wood and O. M. Hilliard. The South African liaison

botanists J. H. Ross and H. R. Tolken; M. D. Gunn

and R. G. Strey for information on early collectors;

J. C. Scheepers, C. J. Ward, F. B. Wright and N.

Grobbelaar for field observations; the staff of the

Natal Museum for insect identifications; the Natal

Parks Board, Department of Forestry, the Curator

of the Lowveld Botanic Garden, and I. Garland for

accommodation and assistance; and finally to K. D.

Gordon-Gray to whom I express my warmest thanks

for her help, interest, exactitude and the conscien-

tious standard she demanded.

UITTREKSEL

Die Eriosema cordatum E. Mey. kompleks word in

’n aantal spesies gesegregeer. E. cordatum E. Mey.

word as ’n veelvormige spesies behou, en twee ver-

wante spesies, E. lucipetum C. H. Stirton en E.

zuluense C. H. Stirton word as nuut beskryf. Vier

bykomende spesies E. buchananii Bak. /., E. nutans

Schinz, E. psoraleoides (Lam.) G. Don en E. parvi-

florum E. Mey, word hersien en uitgesluit uit die E.

cordatum groep.

REFERENCES

Allsopp, R. J., 1947. Notes on the ecology of associated grassveld

plants of Pietermaritzburg, South Africa. M.Sc. thesis,

University of Natal (unpublished).

Bryant, A. T., 1909. Zulu medicine and medicine men, Ann.

Natal Mus. 2: 1-104.

Burtt Davy, J., 1932. A manual of the flowering plants and ferns

of the Transvaal with Swaziland, South Africa. 2. London:

Longmans Green.

Corby, H. D. L., 1974. Systematic implications of nodulation

among Rhodesian legumes. Kirkia. 9: 301-329.

Dilcher, D. L., 1974. Approaches to the indentification of

Angiosperm leaf remains. Bot. Rev. 40, 1: 1-157.

Frey-Wyssling, A., 1955. The phloem supply to the nectaries.

Acta bot. neerl. 4: 358-369.

Grear, J. W., 1970. A revision of the American species of

Eriosema (Leguminosae-Lotoideae), Mem. N. Y. bot. Gdn

20, 3: 1-98.

Grobbelaar, N., Van Beyma, M. C. & Todd, C. M., 1967. a

qualitative study of the nodulating ability of legume species:

List 1. Pubis Univ. Pretoria. New ser 38: 1-9.

Grobbelaar, N & Clarke, B., 1972. A qualitative study of the

nodulating ability of legume species: List 2. J! S. Afr. Bot. 38,

4: 247.

Harvey, W. H., 1862. Eriosema DC. In W. H. Harvey & O. W.

Sonder, Flora Capensis, 2: 259-262. Dublin: Hodges, Smith.

Hepper. E. N., 1973. Eriosema DC. In J. Hutchinson & M.

Dalziel, Flora of West Tropica! Africa, ed. 2 revised by

R. W. Keay 1 , 2: 555-559 ( 1958). London: Crown Agents for

Overseas Govts. & Admins.

Hickey, L. J., 1973. Classification of the architecture of

dicotyledonous leaves. Am. J. Bot. 60, 1: 17-33.

Jacot Guillarmod, A., 1971. Flora of Lesotho (Basutoland).

Lehre: Cramer. 474 pp.

Jacques-Felix, H., 1971. Observations sur les especes du genre

Eriosema de Republique centrafricaines, du Cameroun et

d’Afrique occidentale. Adansonia ser. 2, 11: 141-199.

Lawrence, G. H. M., 1960. Taxonomy of vascular plants. New

York: Macmillan. 823 pp.

Moore, J. A., 1936. The vascular anatomy of the flower in the

papilionaceous Leguminosae. Am. J. Bot. 23: 349-354.

Phillips, E. P., 1917. A contribution to the flora of the Leribe

Plateau and environs. Ann. S. Afr. Mus. 16: 1-379.

Phillips, E. P., 1951. The genera of South African flowering

plants, ed. 2, Mem. bot. Surv. S. Afr. 25: 923 pp.

Poynton, J. C., 1964a. The amphibia of southern Africa: a faunal

study. Ann. Cape Prov. Mus. 2: 252-272.

Poynton, J. C., 1964b. The biotic division of southern Africa, as

shown by the amphibia. 206-218. In D. Davis, Ecological

studies in southern Africa. The Hague: W. Junk. 415 pp.

Roe, K. E., 1971. Terminology of hairs in the genus Solanum.

Taxon 20: 501-508.

Smith, C. A., 1966. Common names of South African plants.

Mem. bot. Surv. S. Afr. 35. 642 pp.

Staner.P. & DeCraene, A., 1934. Les Eriosema de la Flore Con-

golaise. Annls Mus. r. Congo beige ser. 6, 1,2: 1-90.

Stearn, W. T., 1973. Botanical Latin. Newton Abbot: David &

Charles. 566 pp.

Stirton, C. H., 1975. A contribution to knowledge of the genus

Eriosema (Leguminosae, Lotoideae) in southern Africa (ex-

cluding Mozambique and Rhodesia). M.Sc. thesis, University

of Natal, Pietermaritzburg, 182 pp. (unpublished).

Stirton, C. H., 1977. The identity of Eriosema nanum. Bothalia

12: 199-203.

Stirton, C. H., 1978. The Eriosema cordatum complex. 1. The

Eriosema populifolium group. Bothalia 12: 395-404.

Stirton, C. H., 1981. Studies in the Leguminosae-Papilionoideae

of southern Africa. Bothalia 13: 317-325.

Stirton, C. H. 1981. Hybridization in Eriosema (Papilionoideae)

in southern Africa. Bothalia 13: 307-315.

Stuckenberg, B. R., 1969. Effective temperature as an ecological

factor in southern Africa. Zoologica Afr. 4, 2: 145-197.

Van Dijk, D. E., 1971. The zoocartographic approach to Anuran

ecology. Zoologica Afr. 6, 1: 85-117.

306

THE ERIOSEMA CORDATUM COMPLEX. II. THE ERIOSEMA CORDATUM AND E. NUTANS GROUPS

Verdcourt, B., 1971a. Studies in the Leguminosae-Papilionoideae

for the flora of tropical east Africa, Kew Bull. 25: 65-169.

Verdcourt, B., 1971b. Eriosema DC. In E. Milne- Redhead &

R. M. Polhill In Flora of Tropical East Africa, Leguminosae-

Papilionoideae, 761-805. London: Crown Agents for

Overseas Govt & Admins.

Waddle, R. M., 1968. See Waddle, R. M. & Lersten, N. R. 1974.

Waddle, R. M. & Lersten, N. R., 1974. Morphology of discoid

floral nectaries in Leguminosae, especially tribe Phaseoleae

(Papilionoideae). Phytomorphology 23: 152-161.

Watt, J. M. & Breyer-Brandwijk, M. G., 1962. The medicinal

and poisonous plants of southern and eastern A frica. 2. Edin-

burgh & London: Livingstone. 1457 pp.

White, F., 1971. The taxonomic and ecological basis of

chorology. Mitt. bot. StSamml., Miinch. 10: 91-112.

Borhalia 13, 3 & 4: 307-315 (1981)

Natural hybridization in the genus Eriosema (Leguminosae)

in South Africa

C. H. STIRTON*

ABSTRACT

Both spontaneous and introgressive hybridization occur naturally in Eriosema in South Africa. One case of

hybrid swarming is reported and a catalogue of six hybrids is presented and discussed in detail.

RESUME

HYBRIDATION NATURELLE DANS LE GENRE ERIOSEMA (LEGUMINOSAE) EN AFRIQUE DU SUD

L’hybridation naturelle et introgressive survient naturellement chez /’Eriosema d’Afrique du Sud. Un cas

d’essaimement d’hybride est rapporte dans un catalogue de six hybrides et est presente et discute en detail.

INTRODUCTION

Although hybridization is today widely recognized

as having played a major role in the evolution of the

plant kingdom, there is still disagreement (Heiser,

1973) as to how it should be defined. For example,

Sibley (1957) defined it as interbreeding between

populations in secondary contact, regardless of tax-

onomic rank, whereas Solbrig (1970) differentiated

clearly between the crossing of different taxa, calling

that between species, or taxa of higher rank,

‘hybridization’, and that within species ‘recombina-

tion’. In this paper the term hybridization is used in

Mayr’s (1942) sense of ‘the crossing of individuals

belonging to two unlike natural populations that

have secondarily come into contact’.

Most authors who have written on this subject

recognize two main types of hybridization: spon-

taneous hybridization and introgressive hybridiza-

tion. Naturally occurring interspecific hybrids,

presumably individuals of the first filial generation

(henceforth referred to as the F,), are known in all

major groups of plants and in all well studied floras

(Grant, 1971). According to this author the Fj

generation may be fertile, semi-sterile, highly sterile,

or completely sterile, and in all but the last case it can

produce some later-generation progeny. The partially

fertile F, hybrid may reproduce sexually, be selfing,

or there may be crossing with sister hybrid plants, or

backcrossing to one or both parental species. The

resulting second-generation progeny can then go on

to cross with one another and with the original

plants. This results in a hybrid swarm, an extremely

variable mixture of species, hybrids, backcrosses,

and later-generation recombination types.

In this study the term ‘spontaneous hybridization’

is used in reference to the production of occasional or

sporadic natural hybrids, whereas the term ‘intro-

gressive hybridization’ refers to the repeated back-

crossing of a natural hybrid to plants of one, or both,

parental populations.

Three hybrid situations have been encountered in

the present study. These are:

(a) spontaneous hybridization between two

species, with the hybrid progeny sterile;

(b) introgressive hybridization between two spe-

cies, with the hybrid progeny semi-sterile or

fertile;

(c) hybrid swarming between four species with the

hybrid progeny of variable fertility.

♦Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

Representative examples of spontaneous and intro-

gressive hybridization have been selected from the

hybrid catalogue and are presented in detail. The

only hybrid swarm encountered is briefly referred to,

as its detailed analysis was considered beyond the

scone of this work. The three hybrid stituations listed

above form the framework on which the hybrid cata-

logue has been compiled. It is presented in the last

section.

The structure of populations of plants derived

through hybridization have been evaluated almost

exclusively on the basis of morphological criteria

(Levin, 1967). Their analysis has been possible by the

development of a number of techniques such as

Anderson’s (1949, 1953, 1956, 1957) pictorialized

scatter diagrams and hybrid indices and to a lesser ex-

tent Hatheway’s (1962) weighted hybrid index. How-

ever, as Levin (1967) clearly pointed out, these tech-

niques, although they provide considerable informa-

tion about the gross population structure, the direc-

tion and extent of gene flow, and the correlation and

segregation of species characters, they nevertheless

may fail to provide a basis of ascertaining the specific

nature of each hybrid. As a consequence of domi-

nance, character cohesion, epistasis, complex mode

of inheritance, varying degrees of expressivity and

phenotypic plasticity Levin (1967) considers any

judgements concerning parentage of individual

organisms to be unreliable. Clearly comparative mor-

phology alone is inadequate in portraying the dimen-

sions and significance of hybridization: it can

however, indicate its physical presence and, from a

practical taxonomic viewpoint, can provide an ade-

quate preliminary tool of analysis.

For over 30 years Anderson’s paradigm of charac-

ter coherence, the basis of his concept of introgres-

sive hybridization, has been very useful to taxono-

mists dealing with hybrid populations. There is now,

however, sufficient data to suggest that this para-

digm should be considerably modified (Flake, Rud-

loff & Turner, 1969; Adams & Turner, 1970; Heiser,

1973; Anderson & Harrison, 1979; Grant, 1979;

Wells, 1980). The Andersonian theory has been suc-

cinctly summarized by Grant (1979) as follows:

‘Character coherence is a diagnostic feature of

natural hybrid populations. The character coherence

is due primarily to multifactorial linkage and secon-

darily to pleitropy and environmental selection.

Selection in nature normally favours parental types

and thus works in the same direction as linkage’.

What has troubled Grant (1979) and other authors is

the limiting scope of Anderson’s theory as well as the

misleading interpretations that can arise from its

faulty techniques. The latter has been clearly shown

308

NATURAL HYBRIDIZATION IN THE GENUS ERIOSEMA (LEGUMINOSAE) IN SOUTH AFRICA

by Wells (1980). The biggest stumbling block in

Andersonian theory is the finding (Grant 1979) that

different hybrid populations may have marked dif-

ferences in character coherence. Grant has suggested,

therefore, that character coherence is not a constant,

but rather a variable feature of natural hybrid popu-

lations. Two new techniques are now available to

detect this (Grant 1979; Wells, 1980). The nub of

Grant’s conclusion is that multifactorial linkage and

recombination are opposing tendencies and that in

hybrid populations they work in balance such that

their respective expressions of character coherence

and character recombination will reflect a balance or

compromise between opposite extremes. He suggests

that ‘the balance is affected by various internal and

external factors. The components of the recombina-

tion system of the plant group comprise the internal

factor (see Grant 1975, Chap. 3); the chief external

factor is natural selection. The interplay between

these factors determines the point of equilibrium be-

tween character coherence and character recombina-

tion and so will vary from hybrid population to

hybrid population depending on the controlling fac-

tors involved.

This study has utilized Andersonian techniques as

these were all that were available at the time. Unfor-

tunately the present author has no further opportuni-

ties to continue this study and presents the results in

the hope that someone will reasses the problem in

greater depth using chemical and cytological data

and a revised Andersonian paradigm.

Eriosema is one of the few papilionoid genera in

southern Africa which is known to have undergone

hybridization. The reason for this propensity is un-

known. It is interesting that Rhynchosia, a genus

close to and often confused with Eriosema, has not a

single recorded case of hybridization. Judging from

reported cases hybridization among papilionoid

legumes is an apparently rare phenomenon. This

may, however, be due to a lack of appreciation by

taxonomists of its role in plant variation. The

chemotaxonomic studies of Baptisia by Alston and

Turner and co-workers remain the classic case of

hybridization analysis in the Papilionoideae (Alston,

1959; Alston, Turner, Lesters & Horne, 1962; Alston

& Turner, 1963; Alston & Hempel, 1964; Alston,

1965) and a useful model to follow if suitable

facilities are available.

1. SPONTANEOUS HYBRIDIZATION

While collecting specimens on vacant commonage

at Hayfields, Pietermaritzburg (Fig. 1), I found a

large population of yellow flowered E. salignum E.

Mey., growing below a ridge on a gentle slope. South

of this was an extensive population of prostrate,

bright red, orange and yellow flowered E. cor datum

E. Mey. In the adjacent area between the two species

and also within the E. cordatum population were

about twenty semi-erect, pale pink and yellow-

flowered plants (Fig. 2), most of which bore withered

flowers and fruits. Some of the fruits were fully

formed but the enclosed seeds had shrivelled. The

vesture of hairs of these twenty proved to be inter-

mediate between those borne by plants of E. salig-

num and E. cordatum that grew in the same general

area. I decided to analyse the entire population.

After considering the possible influence of slope,

shade and soil type, I laid out four transects that pro-

vided an adequate sample of the area. I collected all

plants of Eriosema that grew within 1 m of these

lines. At 50 m intervals, 1 then made right-angled

transects along which I also collected all plants within

1 m distance from the transect.

As fruiting and flowering were almost completed

at time of first locating the population, emphasis had

to be placed upon the morphological form of vegeta-

tive characters. A record was made of each plant’s

habit of growth, the degree of senescence of the

stipules and the type and density of stem and leaf ves-

ture. The length and width of the terminal (upper)

leaflet of the mature trifoliolate leaf subtending the

first produced inflorescence, and the length of the

petiole of the same leaf were also noted. The qualita-

tive morphological characters are listed in the left

hand column of Table 1. Each character was divided

into three states. Those states representative of plants

of E. salignum were given the value 0; those represen-

tative of plants of E. cordatum carried the value 2,

while intermediates were represented by value 1. One

hundred and fifty two plants were scored for these

character states. The sum of the index values for all

character states of an individual comprises its total

index value. Theoretically the total index value for

plants of E. salignum was thus 0, that for plants of E.

cordatum was 10, while for hybrid plants scores from

1-9 were possible. Plants with similar total index

C. H. STIRTON

309

Figs 1 — 5. — Spontaneous hybridization. 1, fallow commonage near Scottsville. 2, Stirton 736: plant collected at Hayfields,

Pietermaritzburg (1974-02-21). 3, frequency distribution of total index values of the Eriosema population at Hayfields,

Pietermaritzburg (S, E. salignum; H, ‘hybrid’; C, E. cordatum). 4, pictorialized scatter diagram of population of E.

salignum and E. cordatum, and putative hybrids growing at Hayfields, Pietermaritzburg (sampled 1974-12—01). 5, Stir-

ton 1601: plant collected at Hayfields, Pietermaritzburg (1975-01-22), probably a hybrid between E. cordatum and E.

salignum — 1, stem with flowers, x 0,5; 2, glabrous abaxial surface of lateral leaflet, x 3; 3, hairy adaxial surface of ter-

minal leaflet, x 3; 4 semi-erect hairs of stem and adaxial surfaces of leaflets; 5 free stipules at node 2, x 2,8; 6, fused

stipules at node 3, x 2,8; 7, flower, x 4,3.

310

NATURAL HYBRIDIZATION IN THE GENUS ERIOSEMA (LEGUMINOSAE) IN SOUTH AFRICA

values were grouped into a frequency distribution

(Fig. 3). The 21 intermediate plants all fell into class 5

of the frequency distribution. The 61 plants which

had individual total index values of 0 were considered

to belong to E. salignum, whereas those plants which

had a total index value of 10 were considered to

belong to E. cordatum. Only one group of plants was

found to be intermediate, namely those falling within

class 5.

Quantitative parameters were combined with the

qualitative character states by constructing a pictor-

ialized scatter diagram (Fig. 4). In this diagram the

value of the ratio of leaflet length/leaflet width was

plotted against the length of the petiole.

In Fig. 4 the hybrids (open circles) are seen to

occupy an area of the graph intermediate between the

main area of variation occupied by each putative par-

ent. The hybrid is not entirely discrete spatially as a

number of specimens of both putative parents fall

within its range.

Fig. 4 and Table 1 both indicate the presence of a

group of plants with characters that are not those of

E. salignum nor those of E. cordatum but more or

less intermediate between these.

A resampling of this population on the 22-10-1975

produced a plant (Fig. 5) which, although clearly

within the range of the intermediate plants referred

to above, showed some features that are worth

special mention. The stipules, instead of senescing,

persisted and remained green. They also varied from

free and overlapping at the base (Fig. 5.5) to fused

for two-thirds their length (Fig. 5.6). It would be ad-

vantageous to monitor an area such as the one des-

cribed, for hybrids and hybrid variation throughout

several years to attempt to obtain information relat-

ing to maternal and pollen parents of the hybrids that

grew and flowered at different times and especially to

ascertain whether the nature of the spontaneous

hybrids varied over time.

In the Hayfields area the location of the hybrid

plants in relation to the populations of both putative

parental species suggested that the hybrids arose

from seed shed by E. cordatum. Plants of E. cor-

datum always set abundant seed, whereas those of E.

salignum produced a poor, erratic seed set. This level

of seed set has been found to be consistent for these

taxa throughout their range. Bruchid damage was

also heavier in seeds of E. salignum.

2. INTROGRESSIVE HYBRIDIZATION

Of common occurrence in Eriosema in South

Africa are populations in which two species are read-

ily recognized but in which a range of ‘intermediates’

are also present.

The example discussed in this section was located

near Sobantu village, Pietermaritzburg (Fig. 6). Two

species occurred: E. cordatum and E. preptum C. H.

Stirton.

The habitat consisted of heterogeneous soil condi-

tions that ranged from clay in burnt grassveld to deep

sand on the lip of a bare roadbank. The latter site

formed the limit of distribution. The roadbank was

fully exposed to the sun, whereas the grassland had a

fair number of plants of Acacia that provided shade.

Plants were analysed as in the example decribed

previously. A field assessment of their range of varia-

tion revealed two characters with marked quantita-

tive variation, namely the length of the inflorescence

and the length of the terminal leaflet of a mature

trifoliolate leaf. Measurements were recorded for 136

plants.

The qualitative characters are listed in the left hand

column of Table 2. Each character was divided into 3

or 4 states. The numerical value of the character

states provides an index value. Total index values

were compiled for each plant and then used in the

compilation of a frequency distribution (Fig. 7) by

the method used previously. The scoring of individ-

ual plants was such that a typical plant of E. preptum

would score 0, one of E. cordatum would score 9,

whereas plants with characters of either putative

parent would score from 1 to 8.

Fig. 7 shows that nearly 50 % of the plants scored

fell between the two putative parents. Fig. 8 is typical

of the plants falling in group 4 in Fig. 7. From Fig. 7

it appears that there was backcrossing of the interspe-

cific hybrids with plants of E. preptum.

The quantitative characters were combined with

the various qualitative character states and are pre-

sented in a pictorialized scatter diagram (Fig. 9). It

TABLE 2. — Characters and index values used in scoring the hybrid population at Sobantu Village, Pietermaritzburg

Characters 0 12 3

C. H. STIRTON

311

Total index value

STEM PUBESCENCE

O short white hairs with longer appressed hairs

interspersed

£ short white hairs with spreading yellow hairs

interspersed

• stiff, spreading, ferruginous hairs

GLAND COLOUR ON CALYX

O yellow

©-orange

2. INTROGRESSIVE

©cwhite

HYBRID

10

COR DATUM

PLANT HABIT

O erect

-O semi-erect or ascending

prostrate

UNDERSURFACE OF LEAVES

PREPTUM

O dense white woolly with longer appressed hairs

on the veins. Yellow glands interspersed

y sparsely woolly with longer erect hairs, and

yellow glands interspersed

^ dull with orange glands and scattered erect

hairs interspersed

• shiny with white glands and scattered erect

hairs interspersed

CORDATUM

PREPTUM

Figs 6-10. — Introgressive hybridization. 6, open grassland with scattered Acacia trees near Sobantu Village, Pietermaritz-

burg. 7, frequency distribution of total index values of the Eriosema population near Sobantu Village, Pietermaritzburg

(P, E. preptum; H, ‘intermediate plants’; C, E. cordatum) 8, Stirton 1602: putative hybrid plant between E. cordatum

and E. preptum. 9, pictorialized scatter diagram of hybridizing population of E. preptum and E. cordatum, near Sobantu

Village, Pietermaritzburg (sampled 1974-11-20); the terminal leaflet of the leaf subtending the second produced in-

florescence was measured. 10, representative specimens of the Eriosema population near Sobantu Village, Pietermaritz-

burg: 1, E. preptum; 2, part of hybrid intermediate showing hybrid vigour; 3, E. cordatum.

312

STUDIES IN THE LEGUM1NOSAE — PAPILIONOIDEAE OF SOUTHERN AFRICA

can be seen that plants of E. cordatum mostly had

longer terminal leaflets than did plants of E. prep-

tum. The tendency of the hybrids to produce longer

inflorescences than those of either putative parent is

also apparent. What is not clearly shown is the

hybrid vigour of the putative hybrids, but this may be

seen in Fig 10.

The hybrid plant (half the specimen was photo-

graphed) shown in Figs 8 & 10 is a robust intermedi-

ate, and of all the different ‘introgressants’ is, with

others like it, the least likely to be found in the prox-

imity of either parent.

The hybrids were found scattered throughout the

total population but tended to be aggregated in

micro-localities not particularly favoured by either

parent.

Although the overall population at Sobantu Vil-

lage has been interpreted as exhibiting signs of intro-

gressive hybridization, it is quite likely that this may

be an incorrect explanation as a number of alterna-

tives are theoretically plausible. Firstly, individuals

having characters of two species may represent the

remnants of the ancestral population out of which

the two species differentiated (Dobzhansky, 1941).

This does not appear to be the case here as the intro-

gressants were mostly sterile and were found in the

more recently disturbed areas of the total population.

A second alternative might be that the variation

arose through mutation. Heiser (1973) pointed out

that this is probably not uncommon and quoted

Mayr (1942) as saying that it is known that the basic

potentialities of related species tend to be similar and

so the mutational channels are therefore more or less

prescribed. This, however, also seems unlikely in

view of the numbers of ‘intermediates’ observed.

The third possible explanation is more complex. It

involves population intergradation. There are two

types of intergradation. Primary intergradation

(Mayr, 1942) occurs in populations that are in con-

tinuous contact, whereas secondary intergradation

occurs between populations that have come together

after a previous separation. Heiser (1973) pointed out

that the former involved recombination and the later

hybridization. Secondary intergradation could result

in introgression. But as Mayr (1963) also pointed out,

these two types of intergradation are not easy to dis-

tinguish, and because of this, the effects of primary

intergradation would be difficult to distinguish from

those of introgressive hybridization. Heiser (1973)

quoted Barber & Jackson (1957) as stating that ‘in a

region of great ecological change one can expect

simultaneous clinal variation in the frequencies of

genes at a number of loci. Variability under the con-

trol of selection may reach a peak and decay on both

sides of this peak’. This would lead to the loose

association of variables that Anderson (1953) con-

sidered diagnostic of introgression. This would then

mean that highly variable populations resulting from

primary intergradation would be quite similar to

those resulting from introgression. Both putative

parents are widely distributed in the drier areas of

Pietermaritzburg. For the most part they are allopat-

ric but at least three localities are known where they

overlap. In all the overlapping localities there is

hybridization of the type reported in this section. It

seems that these examples conform to Mayr’s (1963)

statement, that ‘belts of highly variable populations

in a meeting zone between rather uniform popula-

tions are almost invariably zones of secondary in-

tergradation’.

Heiser (1973) mentioned two further possible inter-

pretations of what is known as introgression. The

first involved segmental allopolyploidy where segre-

gates approaching one or both parents may occur.

From the available cytological evidence (Baudet,

1977; Turner & Fearing, 1959), it seems that poly-

ploidy is absent in the subtribe Cajaneae to which

Eriosema belongs. It seems unlikely, therefore, that

the intermediates of the Sobantu population were

segregates produced by a polyploid species.

Heiser’s second interpretation concerned the pro-

duction of introgression-like effects which might

result from inbreeding and selection in autogamous

plants following an initial hybridization. Heiser (l.c.)

stressed that although occasional hybridization bet-

ween autogamous species did allow for the possibility

of introgression, the lack of backcrossing after the

initial hybridization would imply that the variation

observed, while like introgression, was not the out-

come of this phenomenon.

Field and laboratory experiments which involved

bagging and emasculation of flowers revealed that

both putative parents of the Sobantu Village popula-

tion failed to set seed if the inflorescences were

isolated from pollinators.

One real difficulty in interpreting the Sobantu pop-

ulation is that there is, as Heiser (1973) pointed out,

some confusion between the definition of introgres-

sion and of hybrid swarming. Whether the Sobantu

population consisted of hybrid populations out of

genetic contact with the putative parents, or whether

it consisted of hybrid plants repeatedly backcrossing

to the parental forms clearly cannot be solved

without detailed genetic experiments. The unlikely

possibility also exsists that the introgressants are

nothing more than F, hybrids showing a marked

degree of environmental plasticity.

3. HYBRID CATALOGUE

In the hybrid catalogue I have endeavoured to in-

clude all the putative hybrids that have been found in

Eriosema in southern Africa (Fig. 11). This hybrid

list will, I hope, encourage further studies within

Eriosema. I hope it will also indicate to future

workers that hybridization probably contributes

materially to the extensive variation represented

within Eriosema in southern Africa.

Each pair of putative parents has been named and

numbered as representing a case of interspecific

hybridization.

Under the numbered putative parents (for example

H|), each individual hybrid specimen known has been

cited, together with its putative parents where these

have been suspected. Each set of citations if followed

by a general discussion.

HI Eriosema cordatum E. Mey. x E. salignum E.

Mey. (yellow-flowered form)

Lesotho. — 2828 (Bethlehem): Malaoaneng (-CC), Dieterlen

866 (PRE); Leribe (-CC), Dieterlen 104 (PRE).

South Africa. — 2628 (Johannesburg): Germiston (-AA),

Rogers I2l99e (BOL). 2732 (Ubombo): top of Pongola Poort

(-CA), Stirton 506 (PRE), putative parents Stirton 503 & 507

(PRE). 2830 (Dundee): Buyahlanga Mountain, between Wasbank

and Elandslaagte (-AC), Stirton 1364 (PRE), putative parents

Stirton 1362 & /363(PRE). 2831 (Nkandla): Eshowe(-DC), Lawn

2271 (NH); Mtunzini (-DD), Stirton 417, 1271 (PRE), putative

parents Stirton 414 & 419 (PRE). 2930 (Pietermaritzburg): Hilton

Road (-CB), Ford s.n. (NH); Hayfields, Pietermaritzburg (-CB),

C. H. STIRTON

313

Stirton 736 (PRE), putative parents Slirton 737 & 738 (PRE); also

Stirton 1169 (PRE), and Stirton 1368 (PRE) with putative parents

1367 & 1369 (PRE); Voortrekker Road, Clarendon (-CB), Stirton

1240( PRE), putative parents Stirton 1238 & 1239 { PRE); Baynes-

field (-CB), Stirton 704 (PRE), putative parents Stirton 702 & 701

(PRE). 3029 (Kokstad): 2 km from Harding to Umzimkulu (-DB),

Story 646 pro parte (PRE). 3030 (Port Shepstone): Hazlewood,

Umzinto (-BC), Haijnath 418 (NU), putative parent Baijnath 419

(NU).

Dieterlen 866, a hybrid, was the basis for the incor-

rect inclusion of Eriosema parviflorum E. Mey. in

the flora of Lesotho. Dieterlen 104 (PRE) has five

specimens on the sheet but only the robust specimen

is a hybrid. Rogers 121 99e is included with uncer-

tainty.

Putative hybrids between E. cordatum and E.

salignum (yellow-flowered form) have been found

over a wide range yet the plants that are of suspected

hybrid progeny are everywhere remarkably uniform.

No specimens have been found which might be con-

sidered backcrosses between the intermediate hybrid

and either of the two putative parents.

Hybrids, in general, are not common and thus are

likely to be rare in any locality. The Stirton 736

Hayfields population, however, numbered over 20

plants spread over four hectares. Some of these

plants had very large rootstocks and were considered

to be at least five years old.

Fig 11. — Known distribution of hybrid populations ot Eriosema

in southern Africa.

Field studies in the Hayfields population showed

that E. salignum and E. cordatum usually had very

little overlap in the time and duration of anthesis but

on cloudy days both species showed fully reflexed

flowers (i.e. flowers ready for pollination with the

standard folded back) throughout the day. Sporadic

hybrids were found to occur in the locality ot this

population and may well have been the result of such

an environmental influence. It is possible that

isolating mechanisms other than time of flower

reflexion may be in general operation to maintain the

integrity of the two parental species. This needs fur-

ther study.

Field studies during the present work have sug-

gested that plants of E. cordatum and E. salignum

hybridize more frequently than any other known

cases of hybridizing species in the genus in southern

Africa.

E. cordatum E. Mey. x E. salignum E. Mey.

(yellow-flowered form) can be separated from its

putative parents by the characters listed below.

E. salignum

(yellow-flowered form)

Plants erect

Stipules becoming

senescent, before

young leaflets expand

Stipules free

Flowers yellow

Hybrid

Plants semi-erect

or ascending

Stipules senescent

to halfway up stem

Stipules slightly

connate at base, to

— 14 total length

Flowers orange (or

pale pink) and

yellow

E. cordatum

Plants decumbent

Stipules persistent,

green

Stipules connate, at

base to V* total

length

Under surface of

leaflets shiny with

sparsely scattered

erect hairs

Stem pubescence

shaggy ferruginous

or white, patent.

Flowers red and

yellow

Under surface of leaf- Under surface of

lets woolly with longer leaflets dull with

appressed hairs on semi-erect hairs on

veins veins and in inter-

costal areas

Stem pubescence dense. Stem pubescence

appressed, finely dense, shaggy and

hairy appressed.

H2 Eriosema cordatum E. Mey. x E. salignum E.

Mey. (red and yellow flowered form)

South Africa. — 2831 (Nkandla): Ngoye Forest (-DD), Stirton

1278 (PRE), putative parent Stirton 1279 (PRE); Twinstreams

Farm, Mtunzini (-DD), Stirton 1285 (PRE), putative parents Stir-

ton 1283 & 1284 (PRE); University College of Zululand (-DD),

Venter 690 (ZULU), putative parent Venter 691 (ZULU). 2930

(Pietermaritzburg; Kloof (— DD), Galpin s.n. (24-6-1932. BOL),

mixed sheet with E. salignum. 2931 (Stanger): Tugela Beach

(-AB), Johnson 407 (NBG).

The suspected hybrids of putative parents E. cor-

datum and E. salignum (red and yellow-flowered

form) are different from those cited under Case HI.

Fig 12.— Rootsystems of 1, Eriosema salignum; 2, hybrid E.

cordatum x E. salignum; 3, E. cordatum.

314

NATURAL HYBRIDIZATION IN THE GENUS ERIOSEMA (LEGUM1NOSAE) IN SOUTH AFRICA

This evidence suggests that there may be two distinct

taxa presently included in E. salignum and this pro-

bability is still being investigated. Case H2 hybrids

are more erect with very dark green shiny upper leaf

surfaces and have ovate leaflets with the bases cor-

date. The leaflets of Case H2 hybrids are more com-

monly unifoliolate than in the Case HI hybrids.

The hybrid situation on Mr I. Garland’s farm at

Mtunzini was particularly interesting. The hybrids

were numerous and occurred wherever the putative

parents were sympatric. There was a very large

hybrid population at the bottom of the front lawn of

his house. The hybrids were intermediate in nearly all

characters, the most striking being the intermediate

root system (Fig. 12). In this figure, the hybrid plant

lies between the two putative parents: it can be seen

to have features both of the straight daucate

rootstock of E. cordatum and of the constricted

rootstock of E. salignum. (Intermediate rooting

systems are very characteristic of hybrids in the genus

as a whole.) There was a marked hybrid vigour in this

Mtunzini population. The hybrids were tall, suberect

or ascending and stood out against other plants of

the population.

H3 Eriosema cordatum E. Mey. x E. preptum

C. H. Stirton

South Africa. — 2930 (Pietermaritzburg): Hilton College farm-

lands (-CB), Khan s.n. (NU 45857); behind S.P.C.A. kennels,

Scottsville, Pietermaritzburg (-CB), Stirton 713 (PRE, this is a

mixed collection containing both putative parents) and Stirton

1243 (PRE), putative parents Stirton 1241 & 1242 (PRE); near

Sobantu Village ( -CB). Stirton 1411 (PRE), putative parents Stir-

ton 1409 & 1410 { PRE).

The Stirton 1409-1411 population was analysed

under Introgressive Hybridization. In the Sobantu

Village and Scottsville populations there was marked

hybrid vigour and an intermediate type of rooting

system. The main differences between the hybrids

and the putative parents are summarized below. It

must be stressed, however, that in dealing with an in-

trogressed population the choice of characters are to

some extent arbitrary. Overlap must therefore be ex-

pected.

E. cordatum Etybrid

Plants prostrate Plants semi-erect

or ascending,

very robust

Stem vesture consisting Stem vesture con-

of stiff, spreading, sisting of short

ferruginous hairs white hairs with

spreading yellow

hairs interspersed

Under surface of leaf-

lets shiny with white

glands and scattered

erect hairs inter-

spersed

White glands on calyx

Rootstock carrot-like

with occasional right-

angled extensions;

short stylopodium

Under surface of

leaflets either

sparsely woolly

with longer erect

hairs and yellow

glands inter-

spersed or dull

with orange

glands and scat-

tered erect hairs

interspersed

Orange glands

on calyx.

Long central

carrot-like

rootstock often

very thick and

constricted with

or without

numerous small

beaded off-

shoots.

E. preptum

Plants erect

Stem vesture consis-

ting of short white

hairs with longer

appressed white

hairs interspersed

Under surface of

leaflets dense white

woolly with longer

appressed hairs on

the veins. Yellow

glands interspersed

Yellow glands on

calyx.

Rootstock with long

stylopodium: thin

and beaded when

young but tending

towards slightly

wavy or constricted

but carrot-like when

old

Flowers orange with

red veins, or yellow-

orange

Flowers red and yellow Flowers pink,

orange and yellow

H4 Eriosema cordatum E. Mey. x E. kraussianum

Meisn.

South Africa. — 2730 (Vryheid): 18 km from Vryheid to Paul-

pietersburg (-DB), Stirton 1323 (PRE), putative parent Stirton

1324 (PRE), E. kraussianum seen but not collected). 3030 (Port

Shepstone): Turnoff to Greenhart on Port Shepstone Road

(-DD), Stirton 1402, 1405 (PRE), putative parents Stirton 1403,

1404 (PRE).

The Stirton 1402-1405 population shows charac-

teristics associated with introgression. Two clear-cut

intermediates were noted in the field. Stirton 1402

showed hybrid vigour and was morphologically close

to E. cordatum, whereas Stirton 1405 was a stunted,

multistemmed plant approximating E. kraussianum.

The first plants that I found on locating the popula-

tion were ‘introgressants’ that I was unable to relate

to any species. They grew in deep sand along a road.

After studying them for some time two factors drew

my attention, namely, their pinkish flowers, and

their abortive or irregular seed set. the overall facies

of these plants led me to postulate that they were

hybrids and that one of the putative parents was E.

cordatum. After a search down the left-hand side of

the road I found a large colony of E. cordatum grow-

ing on a sandy bank near the end of the road, about

500 m from the first ‘introgressants’. I returned

down the opposite side and discovered a strange col-

lection of very stunted plants that gave an immediate

impression of E. kraussianum. These orange and

yellow-flowered plants were growing on the damp

sandy lip of an irrigation canal. After further sear-

ching I found a small colony of E. kraussianum

within the same general locality. In trying to recon-

struct how the various plants became distributed, I

noticed that the soils of the two hybrid colonies had

been removed from the general vicinity of the paren-

tal populations during road-building operations. The

two hybrid colonies were established in local habitats

that were different in both water availability and

aspect. It was possible that the hybrids had resulted

from seed or roots being transported in the sand. If

putative parents had also been transported, it could

be that these were unable to survive the harsher en-

vironmental conditions of the hybrid localities.

The second example, Stirton 1323, was an isolated

plant growing in a very disturbed intermediate habi-

tat. It had numerous ascending to erect stems and

showed marked hybrid vigour.

H5 Eriosema transvaalense C. El. Stirton x E.

angustifolium Burtt Davy

South Africa.— 2329 (Pietersburg): near the Magoebaskloof

Hotel, (-DB), Stirton 1442a, 1442b, putative parents Stirton 1445

& 1444 (PRE).

The facies of the Stirton 1442-1445 population is

similar in many ways to that described for the Stirton

1409-141 1 population of E. cordatum E. Mey. x E.

preptum C. H. Stirton. In both instances the habitat

had been subjected in the past to burning. The soil

cover and type was also rather variable. Stirton 1442a

(sheet 1) showed the greatest hybrid vigour. This

population may be backcrossing to either parent as

Stirton 1442a (sheet 2) and Stirton 1442b were in-

termediates between Stirton 1442a (sheet 1) and E.

transvaalense (Stirton 1445) and E. angustifolium

( Stirton 1444) respectively. The hybrids were spread

randomly throughout the population and were readi-

ly distinguishable on habit and flower colour. E.

transvaalense has pink and yellow flowers and is- a

prostrate, softly pubescent densely matted plant

whereas E. angustifolium is an erect, stiffly haired

yellow-flowered species. The hybrids tended to have

C. H. ST1RTON

315

pale pink flowers and to vary from prostrate, ascen-

ding to erect in habit.

E. transvaalense and E. angustifolium are sym-

patric species but this small population is the only

one in which I have found hybridization between the

two species.

H6 Hybrid swarm

South Africa.— 2931 (Stanger): 1 km past Gingingdlovu to

Durban (—BA), Stirton 1259, 1260 (PRE), putative parents E.

salignum [Stirton 1257 (PRE)], E. psoralioides [Stirton 1258

(PRE)], E. cordatum [Stirton 1261 (PRE)] and E. preptum [Stir-

ton 1262 (PRE)].

This population was spread over a disturbed, re-

cently-felled Eucalyptus plantation. Stirton 1259 had

yellow flowers and was semi-erect, whereas Stirton

1260 was pale pink-flowered, numerous stemmed

and prostrate. These two collections formed the

limits of the range of variation. There was a wide

range of ‘intermediates’ between the four ‘putative’

parents. I attempted to analyse this population using

the techniques described earlier but could not make

any sense of the results. As there was no time avail-

able to pursue the problem it is reported here in case

difficulties may be encountered in identifying plants

collected in the Gingindlovu area.

ACKNOWLEDGEMENTS

The bulk of this study was undertaken in 1974 as

part of a Masters thesis at the University of Natal,

Pietermaritzburg. 1 am grateful to the former Head

of Department of Botany, Professor C. H. Born-

man, and his staff for facilities and assistance; to the

Secretary of the Department of Agricultural Techni-

cal Services and the Director, Botanical Research In-

stitute for one year’s study leave; to L. Cowan for

Figs 5 & 8 and for assistance in the field, and finally

to Professor K. D. Gordon-Gray for her enthusiastic

supervision and inspiring introduction to biosyste-

matic botany.

UITTREKSEL

Beide spontane en introgressiewe verbastering kom

in die natuur in Eriosema in Suid-Afrika voor. Daar

word verslag gegee van een geval van hibriede-ge-

swerm en ’n lys van ses hibriede word aangebied en

tot in besonderhede bespreek.

REFERENCES

Adams, R. & Turner, B., 1970. Natural populations of Juniperus

ashei Buch. Taxon 19: 728-751.

Alston, R. E., 1959. Application of paper chromatography to

systematics: recombination of parental biochemical com-

ponents in a Baptisia hybrid population. Nature 184: 285-286.

Alston, R. E. Turner, B. L., Lester, R. N. & Horne, D., 1962.

Chromatographic validation of two morphologically similar

hybrids of different origins. Science 137: 1048-1050.

Alston, R. E. & Turner, B. L., 1963. Natural hybridization

among four species of Baptisia (Leguminoseae). Amer. J.

Bm. 50: 159-173.

Alston, R. E. & Hempel, K., 1964. Chemical documentation of in-

terspecific hybridization. J. Hered. 55: 267-269.

Alston, R. E., 1965. Flavonoid chemistry of Baptisia: a current

evaluation of chemical methods in the analysis of interspeci-

fic hybrids. Taxon 14, 8: 268-274.

Anderson, R. C. & Harrison, T., 1979. A limitation of the hybrid

index using Quercus leaf characters. Southwest, nat. 24, 3:

463-473.

Anderson, E., 1949. Introgressive hybridization. London: Wiley.

109 pp.

Anderson, E., 1953. Introgressive hybridization. Biot. Rev. 28:

280-307.

Anderson, E., 1956. Natural history, statistics and applied

mathematics. Amer. J. Bot. 43: 883-889.

Anderson, E., 1957. A semigraphical method for the analysis of

complex problems. Proc. Natn. Acad. Sci. (U.S.A.) 43:

923-927.

Barber, H. & Jackson, W. 1957. Natural hybridization in action

in Eucalyptus. Nature, Lond. 1979: 1267-1269.

Baudet, J. 1977. Recherches sur la classification generique des

papilionaceae — Phaseoleae. Ph. D. thesis. University Pierre

& Marie Curie, Paris 213 pp.

Dobzshansky, T., 1941. Genetics and the origin of species. New

York: Columbia University Press. 446 pp.

Flake, R., Rudloff, E. & Turner, B., 1969. Quantitative study of

clinal variation in Juniperus virginiana using terpenoid

data. Proc. Nat. Acad. Sci ( U.S.A. ). 64: 487-494.

Grant, V., 1971. Plant speciation. New York: Columbia Univer-

sity Press. 435 pp.

Grant, V., 1979. Character coherence in natural hybrid popula-

tions in plants. Bot. Gaz. 140, 4: 443-448.

Grant, V., 1975. Genetics of flowering plants. New York, Colum-

bia University Press.

Hatheway, W. H., 1962. Natural hybridization between Phlox

maculata and Phlox glaberrima and its evolutionary signifi-

cance. Amer. J. Bot. 50: 714-720.

Heiser, C. B., 1973. Introgression re-examined. Bot. Rev. 39, 4:

347-366.

Hilliard. O. & Burtt, B., 1971. Streptocarpus: an African plant

study. Pietermaritzburg: University of Natal Press. 410 pp.

Levin, D. A., 1967. Hybridization between annual species of

phlox: population structure. Amer. J. Bot. 59, 9: 1 122-1 130.

Mayr, E., 1942. Systematics and the origin of species. New York:

Columbia University Press. 334 pp.

Mayr, E. 1963. Animal species and evolution. Cambridge:

Belknapp Press. 797 pp.

Sibley, C. G., 1957. The evolutionary and taxonomic significance

of sexual dimorphism and hybridization in birds. Condor

59: 166-191.

Solbrig, O. T., 1970. Principles and methods of plant

biosystematics. London: Macmillan. 226 pp.

Turner. B. L. & Fearing, O. S., 1959. Chromosome numbers in

the Leguminosae: African Species, including phyletic inter-

pretations. Amer. J. Bot. 46,1: 49-57.

Wells, H., 1980. A distance coefficient as a hybridization index:

An example using Mimulus longiflorus and M. flemingii

(Scrophulariaceae) from Santa Cruz Island, Callifornia.

Taxon 29: 53-65.

Bothalia 13, 3 & 4: 317-325 (1981)

Studies in the Leguminosae— Papilionoideae of southern Africa

C. H. STIRTON*

ABSTRACT

Six African species of Psoralea are transferred to Cullen Medik.: C. biflora (Harv.) C. H. Stirton, C. holubii

(Burtt Davy) C. H. Stirton, C. drupacea (Bunge) C. H. Stirton, C. jaubertiana (Fenzl) C. H. Stirton, C. obtusifolia

(DC.)C. H. Stirton and C. plicala (Del.) C. H. Stirton. Psoralea patersoniae Schonl. based on an introduced garden

plant is placed under synonomy of Cullen corylifolia (L.) Medik. The following new names are published: Lebeckia

waltersii C. H. Stirton of subgenus Plecolobium C. H. Stirton; Bituminaria bituminosa (L.) C. H. Stirton of

subgenus Bituminaria and B. acaulis (Stev.) C. H. Stirton of subgenus Christevenia Barneby ex C. H. Stirton;

Rhynchosia aridaC. H. Stirton; Eriosema gunniaeC. H. Stirton, E. preptumC. H. Stirton and£. transvaalense C.

H. Stirton. Eriosema capitatum E. Mey. is placed in synonomy with Psoralea tomentosa Thunb., but as P. tomen-

tosa Thunb. is a later homonym of P. tomentosa Cav. it should be referred to P. sericea Poir.

RESUME

ETUDES SUR LES LEG UMINOSAE-PA PILIONOIDEAE D'AFRIQUE AUSTRALE

Six especes africaines de Psoralea sont transferees a Cullen Medik.: C. biflora (Harv.) C. H. Stirton, C. holubii

(Burtt Davy) C. H. Stirton, C. drupacea (Bunge) C. H. Stirton, C. jaubertiana (Fenzl) C. H. Stirton, C. obtusifolia

(DC.) C. H. Stirton et C. plicata (Del.) C. H. Stirton. Psoralea patersoniae Schonl. basee sur une plante de jardin

introduite est placee sous la synonymie de Cullen corylifolia (L.) Medik. Les nouveaux noms suivants sont publies:

Lebeckia waltersii C. H. Stirton et du sous-genre Plecolobium C. H. Stirton, Bituminaria bituminosa (L.) C. H.

Stirton du sous-genre Bituminaria et B. acaulis (Stev.) C. H. Stirton du sous-genre Christevenia Barneby ex

C. H. Stirton; Rhynchosia arida C. H. Stirton, Eriosema gunniae C. H. Stirton, E. preplum C. H. Stirton et E.

transvaalense C. H. Stirton. Eriosema capitatum E. Mey. est place en synonymie avec Psoralea tomentosa Thunb.,

mais comme P. tomentosa Thunb est un dernier homonvme de P. tomentosa Cav. il devrait se refere a P. sericea

Poir.

CONTENTS

1. Cullen Medik.

2. Bituminaria Heist, ex Fabricius

3. Lebeckia Thunb.

INTRODUCTION

This paper is the first in a series of notes on the

taxonomy of the Leguminosae — Papilionoideae for

the Flora of Southern Africa series. Included also are

nomenclatural changes that appertain to floras lying

beyond this area.

1. CULLEN Medik.

Recent investigations have shown that the South

African representatives of Psoralea L. sensu Forbes

(1930) should be rearranged into the genera Psoralea

L., Otholobium C. H. Stirton and Cullen Medik.

(Stirton, 1980). Psoralea L. emend. C. H. Stirton is

now restricted to 17 Cape species. The new genus

Otholobium comprises about 28 species widespread

over the southern and eastern parts of southern

Africa, with one species extending as far north as

Kenya. New combinations will be made in Otholo-

bium once the complicated nomenclatural problems

have been resolved. The remaining seven African

species of Psoralea L. sensu Hutch, are here trans-

ferred to Cullen Medik. The Asian and Australasian

material of Cullen, comprising over 40 species, falls

outside the scope of this investigation.

1. Cullen biflora (Harv.) C. H. Stirton, comb.

nov.

Psoralea biflora Harv., FI. Cap. 2: 157 (1862). Type: South

Africa, Burchell 1720 (K, holo.!)

2. Cullen corylifolia (L.) Medik, in Vorles,

Churpf. Phys.-Oek. Ges. 2: 380 (1787).

Psoralea corylifolia L. Sp. PI. 764 (1753). Type: India, Herb.

Linn. 928. 24 (LINN, holo.!).

*Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

4. Rhynchosia Lour.

5. Psoralea L.

6. Eriosema (DC.) G. Don

Psoralea patersoniae Schonl. in Rec. Albany Mus. 3:54 (1914).

Type: South Africa, Redhouse, Paterson 383 (K, holo.!) syn. nov.

3. Cullen drupacea (Bunge) C. H. Stirton,

comb. nov.

Psoralea drupacea Bunge in Arb. Nat. Ver. Riga 221 (1847).

Type: U.S.S.R., between Buchara and Samarkand, Lehmann s.n.

(LE, not seen).

4. Cullen holubii (Burtt Davy) C. H. Stirton,

comb. nov.

Psoralea holubii Burtt Davy, FI. Transv. 2: XXIX (1932). Type:

South Africa, Matebe, Holub s.n. (K, holo.!).

5. Cullen jaubertiana (Fenzl) C. H. Stirton,

comb. nov.

Psoralea jaubertiana FenzTm Flora 26: 392(1843). Type: Syria,

between Aleppo and Orfar, Kotschy (not seen).

6. Cullen obtusifolia (DC.) C. H. Stirton,

comb. nov.

Psoralea obtusifolia DC., Prodr. 2: 221 (1825). Type: South

Africa, Burchett 1214 (P, holo.!; K, iso.!).

7. Cullen plicata (Del.) C. H. Stirton, comb.

nov.

Psoralea plicata Del., FI. Egypt. 252, t. 27, fig. 3 (1812). Type:

between Qournah and Medynetabou, Delile s.n. (MPU, not seen).

Psoralea odorata Blatt. & Halb. in J. Bombay nat. Hist. Soc.

26: 238 (1918). Type India, Jodhpore, Barmer, Blatter 7005 (K,

holo., photo.!).

2. BITUMINARIA Heist, ex Fabricius

Psoralea bituminosa L. and P. acaulis Stev. are

two widely cultivated species of Psoralea L. sensu

lato. The recent decision to confirm the subdivision

318

STUDIES IN THE LEGUMINOSAE — PAPILIONOIDEAE OF SOUTHERN AFRICA

of Psoralea L. into a number of genera (Stirton,

1980) has made it necessary to search for a generic

name to accommodate them, the name Psoralea now

being reserved for some 17 species of plants endemic

to the southern tip of Africa. Most authors have used

Aspalthium Medik (1787) as a segregant generic

name to accommodate P. bituminosa whenever it

was thought to be distinct from Psoralea (e.g. Meikle

1977). According to Dandy (1967), however, the

name Bituminaria Heist, ex Fabricius (1759) is a

synonym of Psoralea sensu lato. It predates

Aspalthium and becomes available. There is however

some controversy about selecting Fabricius generic

names based on Heister’s works. Some authors such

as Holub (1970) and Rauschert (1968) consider

Fabricius’s generic names to be uninomials and in-

validly published. Dandy (1967) and Stafleu &

Cowan (1976) disagree with this considering that the

typographical distinction of the uninomials clearly

distinguishes them as generic names. Fabricius

(Enum, 1759), however, consistently distinguishes

generic names by capitals. After consulting Dr R. K.

Brummitt and Mr R. D. Meikle (Royal Botanic

Gardens, Kew) and in view of the conservation of a

number of Fabricius’s names, I have accepted that

Bituminaria Heist, ex Fabricius, as the protologue

shows, is validly published.

‘BITUMINARIA Heister. Trifolium asphaltites f.

bituminosum Rpp. & Dod. Psoralea Linn. Planta fere

arborescens. Flores spicati pediculo communi longo

sustinentur, 3. foliolis in 3. vel. 4 lacinias dissectis 3.

utpl. flores ex ala sua emittentibus petiolatos. Perian-

thum tubulosum quinquefidum, una lacinia maxima.

Vexillum purpuro violaceum oblongum sursum et ad

latera reflexum in medio faciei internae 2. appendiculis

instructum. Alae breviores albae rectae concavae.

Carina brevior purpurea anterius, postyrius alba ob-

tusa bipetala, reliqua ut in alia trifoliis. Sed fructus sit.

legumen compressum rotundo falcatum setosum ultra

perianthium prominens, et continens semen unicum

magnum figura phaseoli’.

Bituminaria Heist, ex Fabricius comprises two

species and is closely related to or may be congeneric

with Pediomelum Rydb. from North America. Pen-

ding further study which may indicate otherwise, I

recognize two distinct subgenera: Bituminaria and

Christevenia Barneby ex C. H. Stirton.

Bituminaria Heist, ex Fabricius, Enum. 165

(1759); Dandy, Regn. Veget. (1967). Type species:

Bituminaria bituminosa ( L .) C. H. Stirton.

Psoralea L., Sp. PI. 1: 762 (1753), pro parte.

Aspalthium Medik. Vorles. in Churpf. Phys.-Oek. Ges. 2: 380

(1787).

Key to subgenera

Plant caulescent; leaflets entire; peduncles axillary (Medi-

terranean Europe and Macaronesia) . . . .subgen. Bituminaria

Plant acaulescent from superficial caudex; leaflets denti-

culate; penduncles scapiform (north-eastern Turkey

and W. Transcaucasus) subgen. Christevenia

Subgen. Bituminaria

Plant perennial, caulescent. Leaves pinnately

trifoliolate, entire. Flowers shortly spicate on

elongated axillary peduncles; bracts at each node of

the inflorescence united through half their length or

more, those at the lower nodes into a flabellate

3(5)-toothed blade, succeeding ones often narrower

or the uppermost wanting; calyx ebracteolate. Fruit

indehiscent, with a long sword-shaped beak which at

length breaks off, furnished with glabrous spinulose

processes; pericarp adnate to seed.

The single species in this subgenus, B. bituminosa,

is very distinct yet remarkably variable throughout its

range. It has been widely cultivated throughout

Europe and to a lesser extent elsewhere. A survey of

the available material would indicate that it com-

prises a number of distinct taxa, some of which are

quite localized. This subgenus needs to be in-

vestigated anew over its entire range. Particular at-

tention should be paid to floral dissections. Psoralea

morisiana Pignatti & Metlesics belongs here and may

be quite a distinct species. Its status will have to be

determined in context of the overall range of B.

bituminosa (see: Bol. Soc. Sarda Sci. Nat. 15: 53,

1975).

1. Bituminaria bituminosa (L.) C. H. Stirton,

comb. nov. Type: in Siciliae, Italiae, Narbonae col-

libus maritimis.

Psoralea bituminosa L., Sp. PI. 1 : 763 (1753); Boiss., FI. Or., 2:

187 (1872); Stuart Thompson in J. Bot., Lond. 44: 306 (1906);

Post, FI. Palest., ed. 2, 1: 367 (1932); Davis, FI. Turkey, 3: 264

(1970); Zohary, FI. Palest., 2: 50, t. 66 (1972).

Aspalthium frutescens Medik. in Vorles, Churpf. Phys.-Oek.

Ges., 2: 380 (1787); Aspalthium herbaceum Medik., l.c. 2: 381

(1787). Aspalthium bituminosum (L.) Fourr. in Ann. Soc. Linn.

Lyon, ser. 2, 16: 365 (1868); Kuntze in Post & Kuntze, Lex. Gen.

Phan., 48 (1903) as Asphalthium bituminosum; Meikle, FI.

Cyprus, 1: 489 (1977).

Subgen. Christevenia Barneby ex C. H. Stirton,

subgen. nov.

Planta perennis, acaulis. Folia subdigitatim tri-

foliolata, denticulata. Flores capitati, pedunculo

elongato scapiformi; bracteae in quoque nodo inflo-

rescentiae usque ad basin distinctae; calyx basi un-

trinque bracteola lineari instructus. Fructus indehis-

cens, rostro ensiformi processibus mollibus pubes-

centibus armato.

Type species: Bituminaria acaulis (Stev.) C. H.

Stirton.

Plant perennial, acaulescent. Leaves subdigitately

trifoliolate, denticulate. Flowers capitate on elon-

gated scapiform peduncles; bracts at each node of the

inflorescence separate to the base; calyx furnished at

the base, each side with a linear bracteole. Fruit in-

dehiscent, with a sword-shaped beak armed with soft

pubescent processes.

The single species in this subgenus, B. acaulis

(Stev.) C. H. Stirton, occurs in north-eastern Turkey

and the west Transcaucasus.

2. Bituminaria acaulis (Stev.) C. H. Stirton,

comb. nov. Type: in iberia occidentali (W. Georgia).

Psoralea acaulis Stev. ap. Hoffmn. in Comm. Soc. Phys.-mat.

Mosq. 1: 47 (1806); M. B. FI. taur.-Cauc. 2: 206(1808); Ldb., FI.

Ross. I: 563 (1842); Boiss., FI. Or. 2: 187 (1872); Grossg., FI.

Kavk. 2: 291 (1930); Vasil’chenko in FI. U.S.S.R. 1 1: 226 (1945),

Eng. transl. 1971; Davis., FI. Turkey 3: 264 (1970). Aspalthium

acaulis (Stev.) Hutch., Gen. FI. PI. 1: 420 (1964).

3. LEBECKIA Thunb.

For a number of years an undescribed Cape

legume has been circulating under the manuscript

name ‘Waltersia heleniae’. The correct generic place-

ment of this undescribed species has remained unre-

solved ever since it was first discovered and collected

by Dr I. B. Walters. It had been variously referred to

Buchenroedera, Wiborgia and Lebeckia. Apart from

the collections cited and the one in Dr Walters’s per-

C. H. STIRTON

319

sonal herbarium, this species has apparently not been

collected by the early collectors. This seems surpris-

ing considering the plant’s rather distinctive features

and, so, notwithstanding its localized distribution, it

is to be expected that additional collections may still

be found misfiled in undetermined covers among

various genera. I have referred it to Lebeckia not-

withstanding the un-Lebeckia like presence of promi-

nent involucrate stipules. The only other legume in

South Africa which I know to have similar stipules is

Argyrolobium involucratum. The alternative is to

describe a new monotypic genus. Such an approach

would be somewhat premature until Lebeckia and its

segregates have been adequately revised and until

more is known about this species which I am describ-

ing as Lebeckia waltersii. This species is easily sepa-

rated from all other Lebeckia species by its unique

stipules and plicate pods. I am therefore establishing

the subgenus Plecolobium to accommodate it. This

name refers to the concertina-like pods.

Subgen. Plecolobium C. H. Stirton, subgen. nov.

Frutices parvi valde ramosi, stipulis conferrumina-

tis petiolo adnatis ramulos omnino vaginantibus.

Fructus plicati, indehiscentes, modice lignosi.

Type species: Lebeckia waltersii C. H. Stirton.

Small profusely branched shrubs with fused stip-

ules adnate to the petiole and completely sheathing

the branches. Fruits indehiscent, plicate, somewhat

woody.

Lebeckia waltersii C. H. Stirton, sp. nov., af-

finitate incerta.

Frutex erectus ramosisimus 50-60 cm altus. Stipu-

lae conferruminatae petiolo adnatae, ramulos om-

nino vaginantes. Folia trifoliolata, argentea; foliola

4-10 mm longa, usque 2 mm lata, subsessilia, aequa-

lia, anguste obovata, curvata, conduplicata. Inflores-

centia 1-3-flora, axillaris. Flores 10 mm longi,

pedicello 3-5 mm longo. Calycis dentes tubo

breviores, lobis vexillaribus maxime fissis, tenuiter

pubescentes. Vexillum 8,6 mm longum, usque 9,5

mm latum, unguiculatum, dorso sericeum. Petala

alaria carinalibus subaequilonga. Pistillum 10-11-

ovulatum; ovarium sericeum. Stamina monadelpha,

axialiter usque ad basin fissa, antheris dimorphis.

Stigma minutum. Fructus plicatus. Semina renifor-

mia, late brunnea.

Type. — Cape, 3319 (Worcester): Worcester Com-

monage (-CB), Rourke 1484 (K, holo.!).

Erect much-branched shrub 50-70 cm high. Stip-

ules fused and wholly adnate to the petiole and com-

pletely sheathing branches, becoming bifurcate in

leaves that subtend inflorescences, sericeous. Leaves

trifoliolate, very shortly petiolate. Leaflets 4-10 mm

long, 1 ,5—2,0 mm wide, subsessile, equal in size, nar-

rowly obovate, somewhat recurved, base cuneate,

apex obtuse, conduplicate, rarely flattened, serice-

ous. Inflorescences axillary, 1-3-flowered, borne on

short lateral branches. Flowers 10 mm long, bright

yellow (less yellow than in L. cytisoides, however),

each subtended by a very small erect bract, ebracteo-

late; pedicel 3-5 mm long. Calyx 4 mm long;

triangular teeth shorter than the 3 mm tube, vexillar

lobes less connate than lateral and keel lobes; finely

pubescent outside, glabrous inside. Standard 8,5 mm

long, up to 9,5 mm wide, claw 3 mm long, broadly

ovate, auricles and appendages absent, apex emar-

ginate, back silky. Wing petals 10,5 mm long, 4,0

mm wide, claw 3 mm long, cultrate, sparsely pubes-

cent, equal in length to keel; sculpturing upper basal

and left central, finely lamellate-lunate; auriculate.

Keel petals fused, 9,5 mm long, 3,5 mm wide,

somewhat pubescent. Pistil 7 — 8 mm long; ovary 5

mm long, subsessile, flattened and most hairy above

and below, sparsely laterally style erect, glabrous,

height of curvature 2,5 -3,0 mm; stigma minute,

very finely penicillate. Stamens monadelphous,

sheath split adaxially; anthers dimorphic, basifixed

anthers narrowly ovate and 2,5 mm long, dorsifixed

anthers rounded and 1,3 mm long. Fruit 15-20 mm

long, 4-5 mm wide, plicate, indehiscent, somewhat

woody, persisting on the plant for more than a year.

Seeds 3 mm wide and long, reniform, greenish

brown. 2n = 32 (count by Christine Brighton, Jod-

rell Laboratories, Kew). Fig. 1.

Lebeckia waltersii is endemic to the south-western

Cape (Fig. 2) and is found growing in renosterbos-

veld on Table mountain sandstone conglomerate in

association with Pteronia and Elytropappus. Flower-

ing occurs in July.

Nw tf Ci? *’

“•» - 7-i«T5 ****■■ Mir

Fig. 1. — Lebeckia waltersii. Rourke 1484, holotype in K.

Cape. — 3319 (Worcester): Worcester (-CB), Rourke 1484 (K;

NBG; PRE); Esterhuysen 35048 (K; BOL); Walters s.n. (NBG,

Herb. Walters).

Hutchinson 253 (K) collected between Vredenburg

and Hoetjies Bay may be conspecific with Lebeckia

waltersii. It may even represent another species. It

has overall a similar appearance to Rourke 1484 but

differs in its non-conduplicate, wider leaves, more

than 1 -flowered inflorescence, less prominent stipule

320

STUDIES IN THE LEGUMINOSAE — PAPILIONOIDEAE OF SOUTHERN AFRICA

hg. 2. — Known distribution in southern Arrica oi Lebeckia

waltersii ( ■), Rhynchosia arida ( ★ ), Eriosema gunniae ( T),

E. preptum ( • ) and E. transvaalense (O).

scars and disjunct distribution. I have included it ten-

tatively under L. waltersii as it fits no where else. It

had been incorrectly named Wiborgia sericea Thunb.

Hutchinson 253 is without fruits, which are required

for a firm decision. I have seen no further collections

of it.

L. waltersii is an attractive shrub which, with its

mass bloom of small yellow flowers, should delight

gardeners. There is a very real danger however that

this species may become extinct unless the citizens of

Worcester, its locality, take concerted action to en-

sure its protection. It is limited to probably only

some ± 60 individual plants divided into a few col-

onies (Walters, 1980, pers. comm.) Perhaps Worces-

ter could adopt it as their town flower? If other

towns and villages throughout the country similarly

adopted a rare or endangered species specific to their

area perhaps much could be done to ensure the ade-

quate protection and propagation of many species of

plants across the length and breadth of South Africa.

The specific epithet waltersii is given in recognition

of the determined campaign that Dr I. B. Walters of

Worcester has waged in ensuring that L. waltersii was

at least named before it disappeared! It is now up to

the citizens of Worcester.

4. RHYNCHOSIA Lour.

Rhynchosia arida C. H. Stirton, sp. nov., affini-

tate incerta.

Frutex deciduus usque 1 m altus et latus verne

florens; ramuli virgati, puberuli et sparsim glandu-

losi. Folia trifoliolata; foliola ovata vel elliptica,

apice uncinata, basi truncata, 10-20 mm longa, 9-13

mm lata, atrovirentia; lateralia inaequalia, basi obli-

qua, terminali symmetrico breaviora; supra pube-

rula, infra glabrescentia, in utroque superficie glan-

dulosa. Stipulae subulatae, glandulosae. Petioli 9-13

mm longi. Racemi axillares, 2-4-flori, foliis longio-

res; pedunculus 13-20 mm Iongus. Flores lutes-

centes, 14-16 mm longi; bractea 4-5 mm longa,

caduca, leviter navicularis. Calyx 15-16 mm Iongus,

inaequaliter lobatus, sparsim pilosus, glandulis

magnis basi bulbosis conspicue tectus; tubus 5 mm

Iongus; lobus carinalis ceteris longior, lanceolatus,

acuminatus; lobi laterales falcati; lobi cornuti per

dimidium longitudinis conferruminati. Vexillum

unguiculatum, 11-12 mm longum, 10 mm latum,

late ovatum, glabrum et eglandulosum, vix reflexum,

carina brevius sed alis longius, callis leviter evolutis et

bene supra parvas auriculas sitis. Alae glabrae,

10-11 mm longae, carina breviores, marsupio

evoluto, auricula bene evoluta, sculptura secus cris-

tam marsupii leviter evoluta, lamellatae. Carina

glabra, laminis 13 mm longis, usque 7-8 mm latis,

basi curvatis, apice obtusis, marsupioevoluto.

Vagina staminalis 12-13 mm longa, stamine discreto

14 mm longo, antheris uniformibus, filamentorum

breviorum medifixis, longiorum basifixis, dehiscen-

tia longitudinali. Gynoecium 13 mm longum,

ovarium 4 mm longum, cum gyophoro 1 mm longo;

pubescentia brevi, patente; curvatura 5 mm alta;

stigma capitatum, ultra stamina exsertum. Nec-

tarium 0,4-0, 6 mm ahum, margine undulato. Legu-

mina 3-4 cm longa, 1 cm lata, falcata, lignosa,

margine undulato, subtiliter pubescente. Semina ig-

nota.

Type. — Cape, 3118 (Vanrhynsdorp), Mount Mat-

sikamma (-DB), Acocks 15125 (K, holo.; PRE,

iso.!).

Deciduous shrub up to 1 m high and broad,

flowering in spring; branchlets virgate, puberulous

and sparsely glandular. Leaves trifoliolate; leaflets

ovate or elliptic, apex uncinate, base truncate, 10-20

mm long, 9-13 mm wide, dark green; laterals

unequal-sided, oblique at base, smaller than sym-

metrical terminal leaflet; puberulous above, glabres-

cent below, glandular on both surfaces. Stipules

subulate, glandular hairy. Petioles 9-13 mm long.

Racemes axillary, 2-4-flowered, longer than leaves,

peduncle 13-20 mm long. Flowers yellow, 14-16 mm

long; bracts 4-5 mm long, caducous, slightly boat-

shaped. Calyx 15-16 mm long, unequally lobed,

sparsely pilose, conspicuously covered with large

bulbous-based glands; tube 5 mm long; keel lobe

longest, lanceolate, acuminate, laterals falcate, vex-

illar lobes fused for half their length. Standard

unguiculate, 11-12 mm long, 10 mm wide, broadly

ovate, glabrous, eglandular, scarcely reflexed,

shorter than keel but longer than wings, appendages

weakly developed and situated well above small

auricles. Wings 10-11 mm long, glabrous, shorter

than keel, pocket present, auricle well developed,

sculpturing weakly developed along ridge of pocket,

lamellate. Keel blades 13 mm long, 7-8 mm wide at

maximum, glabrous, incurved, apex obtuse, pocket

present. Stamina! sheath 12-13 mm long, free

stamen 14 mm long, anthers uniform, medifixed and

basifixed, dehiscence longitudinal. Gynoecium 13

mm long, ovary 4 mm long with gynophore 1 mm

long; clothed with short patent pubescence; cur-

vature 5 mm high; stigma capitate, exerted beyond

stamens. Nectary 0,4-0, 6 mm high, margin un-

dulate. Fruits 3 - 4 cm long, 1 cm wide, falcate,

woody, upper margin undulate, finely pubescent.

Seed unknown. Fig. 3.

It is now over thirty years since the first and only

specimen of this species was collected in semi-succu-

lent karoo along the lower slopes of Mt Matsikamma

(Fig. 2). R. schlechteri Bak., R. bullata Benth. ex

Harv., R. ferulaefolia Benth. ex Harv., R. pinnata

Harv. and R. viscidu/a Steud. are the only species of

Rhynchosia that are distributed in Mediterranean

areas of the Cape and are therefore outliers in an

essentially subtropical genus. These rare, unrelated

species, including R. arida, are endemic to the Cape.

All have narrow disjunct distributions with few

relatives elsewhere in the genus.

The specific epithet arida, meaning becoming dry,

is in reference to the semi-succulent karoo vegetation

C. H. STIRTON

321

13804

Fig. 3. — Rhynchosia arida. Acocks 15125, holtype in K.

in which the species grows. As far as I know, R. arida

is the only Rhynchosia which is restricted to this

habitat.

5. PSORALEA L.

Eriosema capitatum E. Mey., Comm. 130 (1836),

was treated by Harv., FI. Cap. 2: 262 (1862), as a

species unknown. A close study of the protologue

shows that E. capitatum cannot be an Eriosema.

‘Caule erecto antrorsum appresso pubescentes canes-

centes, folioliiss lanceolato-oblongo utrinque acutius-

culis subtus incano-sericeis, capitulis subglobosis in-

volucratis longe pedunculatis (italics mine)’.

There is no species of Eriosema in South Africa

which has an involucrate subglobose flower-head.

Furthermore, the type locality Ruigtervallei falls out-

side the distribution range of Eriosema. The proto-

logue does, however, indicate features that might

suggest Psoralea L.

A study of the literature indicates that E. capita-

tum should be placed in synonomy under Psoralea

tomentosa Thunb., a distinctive Cape species. A pro-

blem arises, however, in that P. tomentosa Thunb.,

Prodr. 2: 135 (1800), is a later homonym of P.

tomentosa Cav., Icon. 3: 21, t. 240 (1795), a validly

described species from Mexico, now accepted as

Dalea tomentosa (Cav.) Willd. The earliest available

name for P. tomentosa Thunb. is P. sericea Poir.

Psoralea sericea Poir., Diet. 5: 687 (1804); in

DC., Prodr. 2: 218 (1825); Meisn., in J. Bot., Lond.

2: 81 (1843); Drege, in Linnaea 19: 645 (1846); PresI,

Bot. Bemerk. 60 (1844).

Rhynchodium sericeum Presl, Bot. Bemerk 60 (1844).

Psoralea tomentosa Thunb., Prodr. 2: 135 (1800) non Cav.

(1795); in DC'., Prodr. 2: 218(1825); Harv., FI. Cap. 2: 156(1862).

P. pedunculatu Ker-Gawl. in Bot. Register t. 223 (1817) non

Poir. (1816) nec Vail. (1891); Meisn., I.c. 2: 81 (1843).

Eriosema capitatum t. Mey., Comm. 130 (1836) syn. nov.;

Meisn., I.c. (1843); Presl, Bot. Bemerk. 60 (1844); Harv., I.c. 2-

262 (1862).

Rhynchosia cephalotes Steud. Norn. 2, 2:588 (1841), syn. nov.

6. ERIOSEMA G. Don

1. Eriosema gunniae C. H. Stirton, sp. nov., E.

cordato E. Mey. affinis, a qua imprimis ramulis flo-

riferis erectis brevioribus, floribus luteis et bracteis

persistentibus aequantibus differt.

Herba perennis, 10-15 cm aha, verne florens.

Rami erecti, pilis velutinis dense obtecti. Folia pro

maxima parte 1-foliolata; 6-9 cm longa, 2, 5-3,0 cm

lata anguste elliptica vel anguste ovata usque

lanceolata, apice acuta, cuneata, utrinque (sed

praecipue in nervaturis) sericeo-pubescentia, glan-

dulosa, margine aliquantum revoluto. Stipulae

15-20 mm longae, lanceolatae, coalitae, glandulosae

et pilosae. Petioli 10-20 mm longi. Racemi axillares,

1 0— 1 5-flori, foliis aequilongi; pedunculus 6-9 cm

longus. Flores lutescentes, 8-10 mm longi; bracteae

± 8 mm longae, persistentes. Calyx 5-7 mm longus,

lobis aequalibus, pilis stramineis patentibus usque

2,5 mm longis dense obtectus, tubo 2 mm longo; lobi

deltoidei, acuminati, lobus carinalis longior, laterales

falcati, lobi cornuti non coaliti. Vexillum 9-10 mm

longum, 4, 5-5, 5 mm latum, unguiculatum, reflex-

urn, obovatum, extra glandulosum et pilosum, carina

et alis longius; calli bene evoluti, conferruminati

cucullati, ab auriculis prominentibus liberi. Alae

8. 5- 9,0 mm longae, 2, 0-3, 2 mm latae ad maximum,

oblongae, auriculatae, carina longiores. Carina glan-

dulosa et pilosa, laminis 6-7 mm longis, 3 mm latis

ad maximum, marsupio evoluto. Vagina staminalis

5. 5- 6,0 mm longa, stamine discreto 5, 4-6,0 mm

longo, antheris uniformibus. Gynoecium 5 mm

longum, ovarium 2,5 mm longum, cum gynophoro

0,5 mm longo, longe pubescens; curvatura 2,0 mm

alta; stigma capitatum, ultra stamina exsertum. Nec-

tarium evolutum, margine revoluto. Leguntina et

semina matura non visa.

Type. — Transvaal, 2530 (Lydenburg): Witklip

Forest Research Station (-BD), Stirton 1482 (PRE,

holo.!).

Perennial herb, 10-15 cm tall, flowering in spring.

Stems erect, densely covered with straw-coloured

hairs. Leaves mostly 1-foliolate, 6-9 cm long,

2. 5- 3,0 cm wide, narrow-elliptic to narrow-ovate to

lanceolate, apex acute, base cuneate, both sides

sericeous especially along veins, glandular; margin

somewhat revolute. Stipules 15-20 mm long,

lanceolate, fused, glandular and hairy. Petioles

10-20 mm long. Racemes axillary, 10-15-flowered,

equalling leaves, peduncle 6-9 cm long. Flowers

yellow, 8-10 mm long, bracts ± 8 mm long persis-

tent. Calyx 5-7 mm long, lobes equal, triangular,

acuminate, keel lobe longest, laterals falcate, vexillar

lobes free. Standard 9-10 mm long, 4, 5-5, 5 mm

wide, clawed, reflexed, obovate, glandular and hairy,

longer than keel and wings; appendages well devel-

oped, fused and hooded, free from prominent auri-

cles. Wings 8, 5-9,0 mm long, 2, 0-3, 2 mm wide at

maximum, oblong, auriculate, longer than keel. Keel

blades 6, 0-7,0 rnrn long, 3 mm wide at maximum,

glandular and hairy, pocket present. Stamina l sheath

5. 5- 6,0 mm long, free stamen geniculate, 5, 4-6,0

322

STUDIES IN THE LEGUMINOSAE — PAPILIONOIDEAE OF SOUTHERN AFRICA

Fig. 4 .—Eriosema gunniae. 1, habit; 2, stem with flowers, x 0,5; 3, flower bract, x 7,5; 4, flower, x 2,3; 5 calyx opened

otit, x 5,3; 6a, standard opened out, x 3,8; 6b, standard closed, x 3,8; 7, wings, x 3,8; 8, keel, x 3,8; 9, vexillar stamen,

x 5,3; 10, staminal sheath, x 5,3; 1 1 , staminal sheath closed with stigma and portion of style exserted, x 5 3- 12 discoid

floral nectary, x 15; 1 3, gynoecium, x 5,3; 14, stigma, x 33. ....

C. H. STIRTON

323

mm long; anthers uniform. Gynoecium 5 mm long;

ovary 2,5 mm long with 0,5 mm gynophore, densely

covered with long hairs, curvature 2,0 mm high;

stigma capitate, exerted beyond stamens. Nectary

present; margin revolute. Mature fruits and seeds not

seen. Fig. 4.

Restricted to the eastern Transvaal between Pil-

grims Rest, Graskop, Witklip, Sudwala and Nels-

hoogte (Fig. 2). This species is found on undisturbed

grassy plateaux. It grows commonly in association

with another rare legume, Rhynchosia villosa

(Meisn.) Druce.

Transvaal.— 2430 (Pilgrims Rest): Pilgrims Rest (-DD),

Rogers 14908; Graskop (-DD), Galpin 14580 Holland s.n. 2530

(Lydenburg): Witklip Forest Research Station (-BD), Stirlon

1482; pass above Sudwala Caves (-DB), Grobbelaar 1439; Nels-

hoogte (-DB), Mutter 2157.

Eriosema gunniae was first collected near Pilgrims

Rest in 1915. It has been commonly referred to E.

cordatuin, but can be distinguished by its stigma ex-

serted from the staminal sheath and by its persistent

flower bract that approximates the length of the

flower. Four collections are included temporarily in

E. gunniae: Young A72 from Middelburg, Young

A214 from Susterstroom, Rudatis 2513 from Tautes-

berg and Grobelaar 1675 from Steenkampbergen.

This composite group occurs to the west of the range

of E. gunniae sensu stricto. Further collections from

the intervening areas may enable their correct status

to be decided.

This distinctive Eriosema is named in honour of

Miss Mary Gunn, who was a recipient of the 1976

Bolus Medal for outstanding achievement in the field

of botany by an amateur botanist (see Veld & Flora

62: 30-31, 1 976). Her knowledge of botanical litera-

ture and plant collectors has been of great import-

ance to both South African plant taxonomy and tax-

onomists alike.

2. Eriosema preptum C. H. Stirton, sp. nov., af-

finitate incerta.

Herba perenna erecta 20-60 cm alta, flore vernale.

Caules 1-15, pilis albis brevibus trichomatibusque

longioribus interspersis vestiti. Caudex stylopodio

longo, gracilis moniliformis iuventute, aetate un-

dulescens vel constringescens sed dauciformescens.

Eolia trifoliolata raro infima unifoliolata (nunc vulgo

obovata), 4, 5-6,0 cm longa, 2, 0-3,0 cm lata,

lateralia minora asymmetrica, elliptica vel anguste

obovata, apice subacuto, basin versus cuneata,

sparse pubescentia, infra dense lanata, venibus pro-

minentibus ob indumentum densum trichomatibus

longioribus appressis, glandularia, marginibus

revolutis. Stipulae 8-14 mm longae, libri. Racemi ax-

illares, (8-) 25-35 florati ut pseudospicae congestae,

folia subtendentia superantes. Flores 6-7 mm longi,

ad 3 mm lati, aurantiaci venis rubribus vel croceo-

aurantiacis, bractei 6 mm longi, cito caduci. Lobi

calycis aequales, triangulares, tubam aequantes. V ex-

ilium 6-7 mm longum, obovatum, appendix prae-

sens connata, per summum unquem de auriculo in

auriculum extensa; dorsum pubescens, glandulare.

Petala carinae breviora auam alae. Gynoecium dense

pubescens. Fructi lu-12 mm longi, 8 mm lati,

molliter sericei. Semini grisei vel pallide brunnei, gut-

tati vel maculati.

Type. — Natal, 2930 (Pietermaritzburg): Scott-

sville, Pietermaritzburg (-CB), Stirton 1242 (PRE,

holo.; K, iso.).

Perennial herb, erect, 20-60 cm tall, spring flower-

ing. Stems 1-15, clothed in short white hairs with

longer hairs interspersed. Rootstock with long stylo-

podium, thin and beaded when young but becoming

wavy or constricted but carrot-like when mature.

Leaves 3-foliolate, rarely the lowest leaves 1-foliolate

(then mostly obovate), 4, 5-6,0 cm long, 2, 0-3,0 cm

wide, laterals smaller and asymmetrical, elliptic to

narrowly obovate, apex subacute, base cuneate,

sparsely pubescent above, densely woolly below with

veins prominent due to dense covering of longer ap-

pressed hairs, glandular, margins revolute. Stipules

8-14 mm long, free. Racemes axillary, (8-) 25-35

flowered in congested pseudo-spikes, overtopping

the subtending leaves. Flowers 6-7 mm long, up to 3

mm wide, orange with red veins or yellow-orange;

bracts 6 mm long, rapidly caducous. Calyx teeth

equal, triangular, ± equal to calyx tube. Standard

6-7 mm long, obovate, appendage present, fused,

extending across top of the claw from auricle to auri-

cle, back hairy and glandular. Keel petals shorter

than wing petals. Gynoecium densely hairy. Fruits

10-12 mm long, 8 mm wide, softly sericeous. Seeds

grey or light brown, with speckles or blotches. Fig. 5.

Eriosema preptum is endemic to Natal (Fig. 2) and

extends some 100 km inland from the coastal belt. In

the past it has been consistently called E. squar-

rosum, an unrelated Cape species. The nature of this

confusion will be dealt with in detail in a subsequent

paper on the Eriosema squarrosum complex. This

species favours sandy sites along roadsides and

ditches but is also commonly found in grassland.

Natal. — 2831 (Nkandla): Nkwaleni River Valley (-CB), Codd

1839; 18 km from Eshowe to Ginginghlovu (-DC), Stirton 1297.

2832 (Mtubatuba): Hluhluwe Game Reserve (-AA), Scon-Smith

10. 2930 (Pietermaritzburg): Pietermaritzburg (-CB), Stirton 368,

1139, 1242, 1410; 5 km from Table Mountain to Pietermaritzburg

(-DA), Stirton 1032; near mid-IUovo (-DC), Stirton 1114. 2931

(Stanger): 43 km from Stanger to Mtunzini (-AB), Stirton 407,

1001, 1002; near Compensation (-BA), Stirton 1160. 3030 (Port

Shepstone): 8 km from Eston to Winkelspruit (-BB), Stirton 1122.

3. Eriosema transvaalense C. H. Stirton, sp.

nov., E. cordato E. Mey. affinis, sed floribus

minoribus, pubescentia, florum colore seminibus dif-

fert.

Herba perenna ad 15 cm alta, flore vernale. Caules

multi, implexi, prostrati vel decumbentes, basi

ramificantes, subtiliter pubescentes pilis flavescen-

tibus reflexis. Folia trifoliolata, infima semper

unifoliolata, 3, 5-6, 5 cm longa, 2, 5-3, 5 cm lata

rotundata vel ovata, ellipticescentia, ambo super-

ficies virides, subtiliter pubescentes; ima venatione

prominente elevata. Stipulae semiconnatae. Rhachis

3-4 mm longa. Racemi 6-8 florati, foliolos

superantes. Flort laeti rosei flavique, 9-10 mm longi,

3 mm lati, bracteo ad 5 mm longo. Calyx 6 mm

longa, lobi tubam subaequitantes. Vexillum 9 mm

longum, 6 mm latum, obovatum, subcucullatum;

dorsum tomentosum glandulare; appendices

praesentes, supra unquem, connatae et ad auriculos

leniter evolutos extensae. Petala carinae dense glan-

dulares. Gynoecium 6 mm longum, ovarium dense

pubescens. Fructus 15-16 mm longus, 10 mm latus,

oblique oblongus rostro 2 mm longo, molliter flavo-

pubescens, glandularis. Semen 5-6 mm longum, 3

mm latum, castaneum purpureomaculatum.

Type.— Transvaal, 2329 (Pietersburg): near

Ebenezer Dam (-DD), Stirton 1438 (PRE, holo.; K,

iso.).

Perennial herb, up to 15 cm tall, spring flowering.

Stems many, matted, prostrate or decumbent, bran-

ching at the base, finely pubescent with reflexed

yellowish hairs. Leaves 3-foliolate, with lower leaves

324

STUDIES IN THE LEGUMINOSAE — PAPILIONOIDEAE OF SOUTHERN AFRICA

Fig. 5 —Eriosema preptum. I, habit; 2, stem with fruits and flowers, x 0,5; 2b, stem vesture, x 4; 3, node showing free

stipules, x 3,3; 4, stipule, x 2,3; 5, flower bract, x 7,5; 6, calyx opened out, x 5,3; 7a, standard opened out, x 3,3; 7b,

standard closed, x 3,3; 8, wing, x 3,3; 9, keel, x 3,3; 10, vexillar stamen, x 5,3; 11, staminal sheath, x 5,3; 12, discoid

floral nectary, x 15; 1 3, gynoecium, x 5,3; 14, stigma, x 33; 15, fruit, x 2; 16a, seed with strophiole, face view, x 5,3;

16b, seed with strophiole, marginal view showing hilum, x 5,3.

C. H. STIRTON

325

always 1-foliolate, 3, 5-6, 5 cm long, 2, 5-3, 5 cm

wide, rounded to ovate, becoming elliptic, both sur-

faces green, finely pubescent above and below, lower

surface with prominent raised venation. Stipules

semi-connate. Rhachis 3-4 mm long. Racemes 6-8

flowered, overtopping leaflets. Flowers pale pink and

yellow, 9-10 mm long, 3 mm wide, bract up to 5 mm

long. Calyx 6 mm long, teeth ± equal to tube. Stan-

dard 9 mm long, 6 mm wide, obovate, somewhat

hooded, back tomentose, glandular; appendages pre-

sent, above the claw, fused and extending to weakly

developed auricles. Wing petals 9 mm long, longer

than keel, prominent peg present which fits tightly in-

to the pocketed keel, sparsely glandular and hairy.

Keel petals densely glandular. Gynoeeium 6 mm

long; ovary densely pubescent. Fruit 15-16 mm long,

10 mm wide, obliquely oblong with 2 mm long beak,

softly yellow pubescent, glandular. Seed 5-6 mm

long, 3 mm wide, chestnut brown with purple flecks.

Fig. 6.

This species was collected for the first time as

recently as five years ago and is endemic to isolated

populations in the Magoebaskloof-Haenertsberg

region (Fig. 2). E. transvaalense hybridizes with E.

angustifolium Schinz. It grows in open grassland and

is particularly evident along firebreaks.

Transvaal. — 2329 (Pietersburg): near Ebenezer Dam (-DD),

Stirton 1438. 2330 (Tzaneen): Magoebaskloof Hotel (-CB), Stir-

ton 1445.

Fig. 6. — Eriosema transvaalense. Stirton 1438, holotype in PRF.

ACKNOWLEDGMENTS

I should like to express my thanks to the following

people. Messrs H. K. Airy-Shaw and R. D. Meikle

and Dr R. K. Brummitt, Royal Botanic Gardens,

Kew, for their indispensable advice and assistance

with nomenclatural problems and Dr. H. F. Glen for

two of the Latin translations.

The Director, Royal Botanic Gardens, Kew, for

use of the facilities of the Royal Botanic Gardens,

Kew, for permission to use Kew negatives 19290 and

19292 and to Mr M. Svanderlik for taking the

photographs.

Dr J. H. Ross (Royal Botanic Gardens and Na-

tional Herbarium, Victoria); Prof. D. Isely and Dr

N. Lersten (Iowa State University), Mr R. Barneby

(New York Botanical Garden); Prof. R. Dahlgren

(Copenhagen); Mr K. H. Mattisson (Lund) and Dr

R. Polhill, Dr B. Verdcourt and Mr G. LI. Lewis

(Royal Botanic Gardens, Kew) for discussions and

helpful criticism of the work leading up to this paper.

Finally, I would like to thank Dr J. P. Rourke

(Compton Herbarium) for drawing my attention to

the Lebeekia waltersii problem; Dr I. B. Walters

(Worcester) for his patient and constructive corres-

pondence and Dr A. Schreiber (Botanische Staats-

samlung, Munchen) for her opinions about

Lebeekia.

UITTREKSEL

Ses Psoralea-s/?es/es van Afrika word na Cullen

Medik. oorgedra: C. biflora ( Flarv .) C. FI. Stirton,

C. holubii (Burtt Davy) C. H. Stirton, C. drupacea

(Bunge) C. H. Stirton, C. jaubertiana (Fenzl) C. H.

Stirton, C. obtusifolia ( D.C .) C. H. Stirton en C.

plicata (Del.) C. FI. Stirton. Psoralea patersoniae

Schonl., gebaseer op ’n ingevoerde tuinplant, word

as ’n sinoniem onder Cullen corylifolia (L.) Medik.

geplaas. Die volgende nuwe name word gepubliseer:

Lebeekia waltersii C. H. Stirton van die subgenus

Plecolobium C. H. Stirton; Bituminaria bituminosa

(L.) C. H. Stirton van die subgenus Bituminaria en

B. acaulis (Stev.) C. H. Stirton van die subgenus

Christevenia Barneby ex C. H. Stirton; Rhynchosia

arida C. H. Stirton; Eriosema gunniae C. H. Stirton,

E. preptum C. H. Stirton en E. transvaalense C. H.

Stirton. Eriosema capitatum E. Mey. word as ’n

sinoniem onder Psoralea tomentosa Thunb. geplaas,

maar aangesien P. tomentosa Thunb. ’n latere homo-

niem van P. tomentosa Cav. is, moet dit na P. sericea

Poir. verwvs word.

REFERENCES

Dandy, J. E., 1967. Index of generic names of vascular plants

1753-1774. Regn. Veg. 51: 1-130.

Forbes. H. M. L., 1930. The genus Psoralea Linn. Bothalia 3:

116-136.

Holub, J. 1970. Lamiastrum versus Galeobdolen and comments

on problems of unitary designations in Fabricius’s work

‘Enumerati methodica plantarum horti medici helmstadien-

sis’ . Folia geobol. phytotax. 5: 61-88.

Meikle, R. D., 1977. Flora of Cyprus. Kew: Bentham-Moxon

Trust.

Rauschert. S., 1968. Zur frage der Gattungsnamen bei Fabricius.

Taxon 17: 153-156.

Stafleu, F. A. & Cowan, A. S., 1976. Taxonomic Literature. 1:

A — G. Utrecht: Bohn, Scheltema & Holkema.

Stirton, C. H., 1980. Psoraleeae. In R. Polhill & P. Raven,

Advances in legume systematics. 329-336. Kew: Royal

Botanic Gardens.

■

Bothalia 13, 3 & 4: 327-330 (1981)

The genus Dipogon (Leguminosae— Papilionoideae)

C. H. STIRTON*

ABSTRACT

A taxonomic revision of the genus Dipogon Liebm. is presented. Only one variable species, D. lignosus (L.)

Verde., is recognized.

RESUME

LE GENRE DIPOGON (LEGUMINOSAE— CAESALP1NIOIDEAE)

Une revision taxonomique du genre Dipogon Liebm. est presentee. Seule, une espece variable, D. lignosus ( L .)

Verde., est reconnue.

DIPOGON

Dipogon Liebm. in Index Sem. Hort. Acad.

Hauniensi: 27 (1854) and in Annls Sci. nat., ser. 4, 2:

374 (1854); Verde, in Kew Bull. 24: 406-409 (1971);

R. A. Dyer, Gen. 1: 275 (1975).

Dolichos L. sect. Eudolichos Taub. subsect. Bar-

batae Taub. in Pflanzenfam. 3: 383 (1894).

Dolichos L. sect. Pogonodolichos Harms subsect.

Gibbosi Harms in Pflanzenw. Afr. 3, 1: 679 (1915).

Verdcourtia Wilczek in Bull. Jard. bot. Etat Brux.

36: 250 (1966).

Perennial twiner, becoming woody below, thinly

pubescent, glabrescent. Leaves pinnately trifoliolate,

petiolate, stipulate; leaflets ovate-acuminate to

oblong-lanceolate, up to 7 cm long. Flowers purple,

in short dense racemes on peduncles longer than

leaves. Corolla 1,0-1, 5 cm long. Style channelled,

bearded along its upper margin, strongly curved near

base and apex in same direction, the middle part

being gently curved in opposite direction. Legume

straight to falcate, 3-5 cm long, 4-5-seeded, style

persistent; seeds black.

A monotypic genus found in the Cape Province

from the Cape Peninsula to Grahamstown.

Dipogon lignosus (L.) Verde, in Taxon 17: 537

(1968); Verde, in Kew Bull. 24: 406 (1971).

Dolichos lignosus L., Sp. PI. 726 (1753); Hort. Cliff. 360, t. 20

(1737); Aiton, Hort. Kew 3: 33 (1789); Smith, Specileg. Bot. 19, t.

21 (1792); Curtis’s bot. Mag. 1 1 : t .380 (1797); G. Don, Gen. Syst.

2: 237 (1832); Freeman in Bot. Gaz. 66: 512, f.3, 5, 6 & 7 (1918);

Burkardt in Rev. Fac. Agron. Vet. Buenos Aires 6: 306 (1929);

Burkardt, Las Leguminosas Argentinas, ed. 2: 422, f. 128, j-k

(1952): Marechal & Otoul in Bull. Jard. bot. Etat Brux. 35: 73,

f. 14. t. F-H, photo. 4 (1965); Verde, in Regnum veg. 40: 26 Adnot.

(1965); Bronkers & de Keyser in Bull. Jard. bot. Etat Brux. 36: 57

(1966); Marechal & Otoul in Bull. Rich. Agron Gembloux, nov.

ser. 1: 63, 1/2A (1966), non sensu Roxb. nec Prain et al. Verdcour-

tia lignosa (L.) Wilczek in Bull. Jard. bot. Etat Brux. 36: 254, f.

5-6 (1966). Neotype: t. 21 in Smith, Spicileg. Bot. (1792). For

photo, see Bot. Gaz. 66: 520 (1918).

Dolichos gibbosus Thunb., Prodr. FI. Cap.: 130 (1800); FI.

Cap. 590 (1823); Harv. in FI. Cap. 2: 244 (1862); Marloth, FI. S.

Afr. 2: t.26 (1925); Levyns, Guide FI. Cape Penin.: 155 (1929);

Phill. in Flower. PI. Afr. 11: t.402 (1931); Anon., S. Afr. Gdng

Country Life 22: 87 (1932); Kidd, Wild Flow. Cap. Penins. t.41, 4

(1950). Lectotype: Cape, ‘in collibus montium urbei Cap. b. Spei’,

Thunberg s.n (UPS-16755, microfiche!).

D. capensis sensu Thunb., Prodr. FI. Cap.: 130 (1800), non L.,

‘Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

D. benthamii Meisn. in Hook., J. Bot. 2: 95 (1843). Type: Cape,

‘in planitie capensi’ (III.E.b.), Krauss 861.

D. gibbosus Thunb. var. uniflorus Harv. in FI. Cap. 2: 244

(1862). Type: Cape, ‘in collibus montium urbi Cap. b. Spei’,

Thunberg s.n. (UPS-16747, microfiche!).

D. jacquinii sensu Piper in Bull. U.S.D.A. 318: 5 (1915), non

DC.

The typification of Dolichos lignosus L. has

caused considerable difficulty and despite attempts

by a number of authors to typify it, it has always

been left unresolved (Freeman, 1918; Verdcourt

1971; Hutchinson, unpublished note; Dandy, un-

published note). It seems appropriate now that the

plant has become quite widely cultivated and has

even begun to assume weedy habits in Australia, that

its correct status be established. Freeman (1918) and

Verdcourt (1971) have given exhaustive accounts of

the origin, usage and misusage of this name. I will

therefore not repeat their arguments here, but briefly

outline why I have chosen t. 21 in Smith’s Spicileg.

Bot. as the neotype even though the key diagnostic

character in Sp. PI. 726 (1753) ‘leguminibus strictis

linearibus’ is not depicted in t. 21.

The protologue in Sp. PI. 726 (1753) is as follows:

‘lignosus. 9. Dolichos caule perenni,

pedunculis capitatis,

leguminibus strictis linearibus.

Dolichos caule perenni

lignoso. Hort. cliff. 360 t. 20.

Phaseolus indicus perennis,

floribus purpurascentibus.

Eichr. carol. 36.

Habitat- — 1? ’

The effective diagnostic part of Linnaeus’s defini-

tion lies in the words ‘pedunculis capitatis,

leguminibus strictis linearibus’ as the first phrase

‘Dolichos caule perenni’ is repeated for D.

polystachios the following species. It is also clear that

the effective diagnostic part is ommitted from Hort.

Cliff. 360. There is no specimen of D. lignosus in the

Linnaean herbarium, so one is forced to look else-

where for a type. The first possibility is the cited plate

in the Hortus Cliffortianus. This cannot be regarded

as a lectotype, hoewever, because the figured plant

lacks fruit and does not have capitate flowers. Lin-

naeus actually stated in the text that his specimen did

not produce fruit (‘Absoluta florescentia absque

fructu periit’). There is, however, a specimen in the

Hort. Cliff which one might consider as a lectotype,

since it is quite sterile. But it shows no sign of having

328

THE GENUS DIPOGON (LEGUMINOSAE — PAPILIONOIDEAE)

10

Fig. I .—Dipogon lignosus. 1, habit; 2, flowering branch, x 1; 3, flower bract; 4, flower; 5, calyx opened out; 6a, standard

opened out; 6b, standard closed; 7, wing; 8, keel; 9, vexillar stamen; 10, staminal sheath; 1 1, discoid floral nectary;

12, gynoecium; 13, stigma; 14a, seed, side view; 14b, seed, marginal view showing hilum.

C. H. STIRTON

329

borne an inflorescence and consists mostly of newly

flushed growth; neither does it match the figured

plate accompanying the Hort. Cliff, description of

D. lignosus. The Hort. Cliff, specimen could be a

Dipogon, but it is very difficult to be sure. For these

reasons it is an unsuitable choice for a lectotype. The

remaining possibility is Linnaeus’s reference to

Eichrodt’s Hortus Carolsruhensis. However, he does

not mention the fruit and, as the phrase name is

vague, it also does not satisfactorily resolve the

problem.

There are two noticeable changes in Linnaeus’s Sp.

PL and Hort. Cliff, accounts. In Hort. Cliff, he

states ‘Crescit in America’, whereas in the Sp. PI. he

omits the origin of the plant altogether. This may in-

dicate that his view of the species had changed, as it

certainly did in the second edition of the Sp. PL,

where he included in synonomy ‘cacara. s. phaseolus

perennis’ (a form of Lablab purpureus) and stated

for the species as a whole ‘Habitat in India’. Verd-

court (1971) has described the subsequent confusion

which thence accompanied the names Dolichos lab-

lab and D. lignosus. The second difference in the

Hort. Cliff and Sp. PI. accounts is the inclusion in

the latter of a description of the fruits.

The whole problem of typification of this species

hinges, I believe, on Linnaeus’s uncertainty about the

nature of fruits in Phaseolus and Dolichos ( Vigna

was not known at the time as such and was treated by

him under D. lablab and D. lignosus; see note by Lin-

naeus on p. 1015, Sp. PL 2.). It began when Lin-

naeus’s concept of his species D. lignosus changed

between his Hort. Cliff, treatment and his 1st edition

of the Sp. PI. I am fairly certain that when he added

‘leguminibus strictis linearibus’ he must have seen a

fruit of a Vigna and not one of Dolichos lignosus. It

is not known whence Linnaeus obtained his informa-

tion about the fruit, neither did it help matters when,

in his second edition, he included in the synonomy of

D. gibbosus the name ‘cacara. s. phaseolus perennis’

now known to be a variant of Lablab purpureus and

yet again with a different fruit.

One can either select a neotype and preserve the

name Dolichos lignosus or abandon it. An accep-

tance that Linnaeus erred in his addition of the fruit

character to his original phrase name in the Hort.

Cliff, would allow one to accept one of the two dis-

tinctive and unambiguous plates published by Smith

(Spicileg. Bot. t. 21, 1792) and by Curtis (Curtis’s

bot. Mag. 11: t. 380, 1797) shortly thereafter. It was

only much later that the identity of Dolichos lignosus

really became confused. The most parsimonious so-

lution is, therefore, I think, to select Smith’s t. 21 as

the neotype of Dolichos lignosus, the basionym of

Dipogon lignosus (L.) Verde.

Voluble perennial arising from underground, verti-

cal, deeply lenticelled rootstock with dichotomously

branching laterals. Shoots spirally twisted, up to 3 m

long, weak, glabrescent. Leaves trifoliolate; stipules

up to 6 mm long, basifixed, oblong-lanceolate, per-

sistent, clasping but patent when old; petioles up to 5

cm long; leaflets 2-7 cm long, 1-4,5 cm wide, late-

rals smaller, stipellate, eglandular, paler beneath,

glabrescent; terminal leaflet ovate-acuminate, late-

rals gibbous on lower margin. Petiolules up to 3 mm

long. Rhachis 1-2 cm long, puberulent, armed with

two persistent acrorhachial stipels. Racemes 5-10

(-33)-flowered, axillary, up to 25 cm long, longer

than leaves, shortly and densely racemose towards

apex, sometimes twisted; pedicels 5-10 mm long,

somewhat flattened, becoming purplish with age,

armed with small caducous bracteoles near apex;

bracts small, green, persistent until flower abscises.

Flowers purple, turning pale mauve, 10-15 mm long,

reflexed; bracts persistent; calyx campanulate, lobes

5, short and broad, tube twice longer than lobes, 2

horn (upper) lobes rounded, lateral and keel lobes

triangular, ciliate. Standard 10-15 cm long and

broad; apex emarginate, base auricled, with promi-

nent appendages (callus lobes, callosities) situated

low down extending from auricles to apex of claw,

raised to form an entrance against which insects can

thrust their thoracic region. Wings longer than keel

blade, hanging slightly flared, with upper auricles in-

serted between appendages, pinkish. Keel blades ros-

trate, apex incurved, purple tipped. Stamens diadel-

phous, stamens held at two levels. Pistil sessile; ovary

linear, with hairs along upper ridge; ovules 4-5; style

channelled, bearded along upper inner margin,

strongly curved near the base and apex in same direc-

tion, the middle part being gently curved in opposite

direction; stigma capitate, fringed with hairs. Discoid

floral nectary present. Legume 40-60 mm long, 8-10

mm wide, oblong, attenuate at base and apex, tipped

with persistent style. Seeds 4-5, subglobose, 3, 5-4, 5

mm wide; hilum 2,5-3 mm long, black or speckled.

Germination hypogeal, epicotyl absent; primordial

leaves ovate, base cordate, opposite, petiole up to 1,8

mm long; stipules 2, undivided, oblong-lanceolate,

persistent; acrorhachial stipels present. Chromosome

number 2n = 22. Fig. 1.

D. lignosus is endemic to the Cape Province and

extends from the Cape Peninsula as far east as Gra-

hamstown (Fig. 2). Its overall distribution falls with-

in the fynbos. Flowering begins in July, reaches a

peak in October then declines rapidly after Decem-

ber. This species grows commonly in scrub forests,

along the perimeter of high forests ( Galpin 3988

reports one plant overtopping a 6 m tree), and more

recently it has been reported from waste places and

gardens.

Despite a number of recent investigations into

generic affinities in the Phaseoleae, there is still no

agreement about the affinity of Dipogon to other

genera. Lackey (1977a) placed Dipogon in the sub-

tribe Phaseolinae Benth. between Alistilus N. E. Br.

and Dolichos L., having suggested earlier (1977b)

that Dipogon was closely related to Lablab Adans.

and Alistilus and should perhaps be united with

them. Baudet (1978), in contrast, placed Dipogon in

the subtribe Phaseolinae, section Phaseolastrae, but

Alistilus and Dolichos in section Dolichastrae. Mare-

chal, Mascherpa & Stainier (1978) suggested links

with Lablab. They included both Dipogon and Lab-

lab in their numerical analysis of the Phaseolus —

Vigna complex. The peripheral affinity of Dipogon

and Lablab to Phaseolus and Vigna (Marechal et al.,

1978) would rather suggest, as already alluded to by

Lackey (1977a) and Verdcourt (1970), that Dipogon,

Alistilus and Lablab may be better considered allies

of Dolichos. However, until the African represen-

tative of Dolichos are better known, this problem will

have to be deferred.

The following herbarium material is recognized as D.

lignosus:

South Africa.— A lexander s.n. (2 sheets), 46 (K); Archibald

4836 (PRE); Alherstone s.n. (GRA); Barker 1673 (GRA, PRE);

Bayliss 108 (K, PRE), 2963 (PRE); Boucher 821, 1652 (PRE); Brit-

ten 78 (GRA), s.n. (Oct. 1946, 2 sheets, (GRA); Burchell 437,

6024, 7012 (K); Cummings 67 (RUH); Dahlstrand 837 (GRA,

STE); Drege 278 (GRA), s.n. (K); Dver 444 (GRA, PRE); Ecklon

1683 (K); Esterhuysen 709 (PRE), 23245 (K); Forest Department

330

THE GENUS DIPOGON (LEGUM1NOSAE — PAPILIONOIDEAE)

Fig. 2. — Known distribution of Dipogon lignosus in southern Africa.

Port Elizabeth 72 (GRA); Fourcade 1630 (GRA), 5739 (STE);

Fries, Norlindh & Weimarck 567 (K); Gamble 22052 (K); Galpin

s.n. (22.9.1897, PRE), 3988 (GRA); Garside 62 (K); Gerber s.n.

(RUH); Gillett 103 (STE), 3376 (PRE, STE); Godfrey s.n.

(11.12.1952, PRE); Grobbelaar 333 (PRE); Hafstrom & Acocks

2306 (PRE); Heeg 92 (RUH); Henry 16 (PRE); Herbarium Harvey

772 (BM, K); Hilner 86 (GRA); Hops 199 (GRA); Hooker 533 (K);

Hutchinson 645 (BM, K), 1164 (K, PRE); Jordaan 3910 (STE),

s.n. (STE 18592); Joubert 487 (STE); Keet 410 (GRA, STE), s.n.

(STE 13499); Kerfoot K5508 (STE); Kies s.n. (5.9.1940, PRE);

Kruger 343 (STE); Levvns 40 (BM); Lynes s.n. (BM); MacGillvray

504 (K); Mar loth 13047, 13448 (PRE); Marsh 611 (PRE, STE);

Miles s.n. (RUH); Morgan 12 (RUH); Muir 91 A (PRE), s.n. (STE

10554); Nature Conservation Cape 208 (PRE); Nelson s.n. (BM);

Noel 153 (RUH); Oliver 3627 (PRE); Pappe s.n. (K, 3 sheets); Pat-

terson 361 (GRA); Phillips 357 (K); Pienaar K58 (STE); Pillans

3054 (PRE); Rodin 1030 (K, PRE); Rogers 1060, 2052 (BM), 4523

(GRA), s.n. (K), 26436 (PRE); Salisbury 206 (PRE); Salter 2058,

9313 (BM); Scharf 1401 (PRE); Schelpe 4166 (BM); Schlechter

2659 (GRA), 4704 (K), 9321 (BM, GRA, K); Schlieben & Ellis

12399 (K, PRE, STE); Schonland 533 (GRA); Shaw 43 (RUH);

Smith 4832, 4930 (K); Story 317, 2593 (PRE); Stirton 6330 (PRE);

Strey 797 (PRE); Taylor 4370 (STE); Theron 640 (PRE); Thode

A917 (K, PRE), 8358 (STE); Thompson J. B. 56 (PRE); Thomp-

son, M. 854 (K, PRE); Tyson s.n. (9.1916, PRE); Van Breda 17,

351 (PRE), 589 (K); Van Dam s.n. (12.1918, PRE); Van Rensburg

487 (PRE); Wallich s.n. (BM); White 5162 (PRE); Worsdel s.n

(K); Wolley-Dod 12 (BM, K); Zeyher 2413 (GRA, PRE), s.n

(GRA).

Verdcourt (1971) has already listed the cultivated

material of D. lignosus housed in K. To this can be

added Thomson s.n. (K) from Ceylon: Gamble 16999

(K) from India; Symon 9541 (ADW; K), Rodd 1493

(K; NSW) and Constable 7148 (K; NSW) from Aus-

tralia; and Bangerter 5189 (AK; K) from New Zea-

land. D. lignosus has become naturalized in Australia

and seems to be increasing its range there.

ACKNOWLEDGEMENTS

For loan of herbarium material I sincerely thank

the Directors of the K, GRA, PRE, RUH, STE, and

BM. I am especially thankful to Mr D. E. Meikle and

Dr B. Verdcourt for the many hours of robust argu-

ment with which the the typification of Dolichos

lignosus was pursued. Linda Cowan kindly drew the

plate of Dipogon lignosus.

UITTREKSEL

’n Taksonomiese hersiening van die geslag

Dipogon Liebm. is onderneem. Net een veranderlike

spesie D. lignosus ( L .) Verde., word erken.

REFERENCES

Baudet. J. C., 1973. Interet taxonomique des caracteres epider-

miques dans le complexe Phaseolus — Vigna. Bull. Soc. bot.

Belg. 106: 53-59.

Freeman, G. F., 1918. The purple hyacinth bean. Bot. Gaz. 66:

512-523.

Lackey, J., 1977a. A revised classification of the tribe Phaseoleae

(Leguminosae: Papilionoideae), and its relation to canavanine

distribution. Bot. J. Linn. Soc. 74: 163-178.

Lackey, J., 1977b. A synopsis of Phaseoleae (Leguminosae,

Papilionoideae). Ph. D. thesis, Iowa State University, Ames.

Marechal, R., Mascherpa, J., & Stainier, F., 1978. Etude

taxonomique d’un groupe complexe d’especes des genres

Phaseolus et Vigna (Papilionaceae) sur la base de donnees

morphologiques et polliniques, traitees par l’analyse infor-

matique. Boissiera 28: 1-273.

Verdcourt, B., 1970. Notes on Dipogon and Psophocarpus

(Leguminosae). Kew Bull. 17: 537-538.

Verdcourt, B., 1971. Studies in the Leguminosae — Papilionoid-

eae. 3. Kew Bull. 24: 406-409.

Bothalia 13, 3 & 4: 331-332 (1981)

Notes on the taxonomy of Rubus in southern Africa

C. H. STIRTON*

ABSTRACT

The taxonomy of Rubus in South Africa is beset with problems. These include the introduction of extra-African

species as ornamentals and crops, the apparent segregation of new forms and finally hybridization with indigenous

species These problems are compounded by poor and incomplete collecting of Rubus in South Africa, and by the

tries'1 ^ ° re atin® in,rocluced taxa t0 the many and varied species, varieties and ecotypes occurring in other coun-

RESUME

NOTES SUR LA TAXONOMIE DU RUBUS EN AFRIQUE A USTRALE

La taxonomie du Rubus en Afrique du Sud est assiegee de problemes. Ceux-ci incluent T introduction d’especes

extra africaines comme plantes ornementales el de cultures, Vapparente segregation de nouvelles formes et finale-

ment /’hybridation avec des especes indigenes. Ces problemes se compliquent par des collections incompletes et

pauvres de Rubus en Afrique du Sud et parla difficulty d’ apparenter les taxa introduits aux especes nombreuses et

variees, aux varietes et aux ecotypes des autres pays.

Prior to the cytogenetic studies of Gustafsson

(1942, 1943) there had been only a few studies such as

those of Sudre (1908-1913), Focke (1911-1914) and

Bailey (1941), that had attempted to produce an

overall taxonomy of Rubus. By 1913 there were

already some 3 350 Latin names in existence, many

of which were applied to primary hybrids or very

localized varieties (Gustafsson, 1943). As Newton

(1975) has suggested, this may have been due to the

vague species concepts prevailing at that time or it

may have resulted from parochial attitudes adopted

towards plants of widespread distribution. Never-

theless, from the work done by Gustafsson and other

geneticists, we now know that much of the mor-

phological diversity was a result of natural hybridiza-

tion, polyploidy and apomixis.

Pseudogamy, a process whereby heterozygous

segregates and hybrid derivatives can be maintained

in nature, has played a particularly important role in

the development of many agamic complexes in

Rubus (Grant, 1971). A batologist not only has to

contend with these basic genetic difficulties, but also

has to deal with considerable phenotypic plasticity

(Beijerinck, 1953; Heslop-Harrison, 1963). It is little

wonder then that this remarkable genus had defied

taxonomists for over three centuries and that there is

still no consensus on supraspecific categories.

The problems of Rubus taxonomy in South Africa

are aggravated by the introduction and naturaliza-

tion of exotic species, the apparent segregation of

new forms in areas surrounding cultivated black-

berries, the role of hybridization among local, as well

as between local and exotic species, and finally by

inadequate herbarium material.

Harvey (1862) recognized five species of Rubus in

South Africa: Rubus fruticosus, R. ludwigii, R. pin -

natus, R. rigidus and R. rosaefolius. The last overall

revision was by C. E. Gustafsson (1933) who added a

further six species: R. adolfi-friederici, R. chrysocar-

pus, R. ecklonii, R. immixtus, R. intercurrens and R.

transvaalensis. Also mentioned by Gustafsson were

two species of introduced brambles: R. affinis and R.

argutus, of which only the former was noted as a

naturalized weed. Not included in any of these

* Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

studies are the now naturalized weeds R. niveus

Thunb. (Java bramble), R. cuneifolius Pursh

(American bramble) and R. phoenicolasius Maxim,

(wine berry).

It is often difficult to decide whether certain

species are indigenous or introduced. A case in point

is R. immixtus C. E. Gust. The type locality of this

species is Hogsback in the eastern Cape, an area

which has an extremely variable Rubus flora and well

noted for its large number of naturalized European

plants such as gorse (Ulex europaeus), roses (Rosa

spp.) and hawthorns (Crataegus). It seems quite pro-

bable, therefore, that R. immixtus may be either a

European species or a hybrid with R. rigidus Sm. in

its ancestry.

Gustafsson (1933) only catalogues one hybrid, R.

affinis x rigidus, in his account Rubi Africani. If

correct, this is a hybrid between an indigenous spe-

cies and an introduced extra African species. What of

the role of hybridization among indigenous species?

Harvey (1862), under his doubtful species, recorded

that an Ecklon and Zeyher specimen was ‘almost in-

termediate between R. pinnatus and R. rigidus’.

Adamson & Salter (1950) state that R. pinnatus

hybridizes freely with R. fruticosus in the Cape

Peninsula area. Focke (1911) reported two hybrids:

R. plicatus x pinnatus and R. pinnatus x rigidus.

The tremendous variability of R. rigidus and R. pin-

natus in South Africa may well be explained by their

apparent ability to hybridize with other species. It is

interesting to note that Amor & Miles (1975) could

find no trace of hybridization having occurred in

Victoria, Australia, although they did not preclude

this as a future possibility once the introduced species

had increased their ranges.

These preliminary observations suggest that a tax-

onomic revision of Rubus in South Africa is highly

desirable, but such a study would obviously be a

long-term project. Although many modern revisions

of Rubus are based largely on cytological investiga-

tions, it is nevertheless still necessary, in many parts

of the world, to rely on herbarium material. If this

material is inadequate or incomplete then the task

becomes very difficult. The importance of complete

herbarium collections has been stressed by a number

of workers (Amor & Miles, 1974; Beijerinck, 1953;

Edees, 1959; Watson, 1958). My impression, after

having seen much of the available herbarium mate-

332

NOTES ON THE TAXONOMY OF RUBUS IN SOUTHERN AFRICA

rial in South Africa, is that few collectors have been

aware of what would constitute adequate pressed

material for the naming of a specimen of Rubus. For

this reason, and because so little material is available

for study and because some areas are undercollected,

I have outlined the four components which make up

a useful herbarium voucher (see also Amor & Miles,

1974; Beijerinck, 1953) — this in the hope that it will

encourage collectors to collect more material.

It is important to collect:

1. One 10 cm section, with leaves, selected from the

middle of a first-year cane of vegetative growth

(primocane). Rubus usually, but not always, flowers

in the second season. The first season or primocanes

are easily recognized by their lush and robust growth.

2. One 10 cm section, with leaves, from the middle

of a flowering cane (floricane). This is necessary as in

most species of Rubus in South Africa the floricanes

differ markedly in shape, size, leaf-shape and

number, and presence or absence of a white bloom.

3. A complete inflorescence with flowers, and fruits

if these are available.

4. A few petals dried separately. The petals of some

species, if not collected separately, become lost dur-

ing drying as they abscise rapidly after collection.

This dried material should be accompanied by full

descriptive notes. Of great taxonomic value are the

colours of petals, young and old fruits, primocanes

and floricanes. The relative length and colour of

stamens and styles, as well, as the relative lengths of

calyx lobes and petals which are often very diagnos-

tic. It should also be noted whether canes are erect,

arching or looping. Looping canes may tip-root dur-

ing the Autumn equinox. Odd forms or unusual

plants should also be collected with a note to that ef-

fect. It is important to accurately record the localities

of oddities, particularly if they are collected near

blackberry orchards or in areas known to be heavily

treated with herbicides, as these chemicals are known

to cause chimeras and unusual phenotypes.

ACKNOWLEDGEMENTS

I am indebted to Professor K. D. Gordon-Gray,

Drs R. L. Amor, O. A. Leistner and Mr M. J. Wells

for their valuable comments and criticism.

UITTREKSEL

Die taksonomie van Rubus in Suid Afrika is om-

ring van probleme. Dit sluit in die invoer van spesies

van buite Afrika as sierplante en gewasse, die oen-

skynlike segregasie van nuwe vorms en uiteindelike

verbastering met inheemse spesies. Hierdie probleme

word vererger deur swak en onvolledige versameling

van Rubus in Suid Afrika en omdat dit moeilik is om

die verwantskap tussen ingevoerde taksons en die

vele en uiteenlopende spesies, varieteite en ekotipes

wat in ander lande voorkom, te toon.

REFERENCES

Adamson, R. S. & Salter, T. M., 1950. Flora of the Cape Penin-

sula. Cape Town: Juta.

Amor, R. L. & Miles, B. A., 1974. Taxonomy and distribution of

Rubus fruticosus L. aggr. (Rosaceae) naturalized in Victoria

Muelleria 3, 1: 37-62.

Bailey, L. H., 1941. Species Batorum. Genies Herb. 5, 1: 1-19.

Beijerinck, W. 1953. On the habit, ecology and taxonomy of the

brambles of the Netherlands. Acta bot. neerl. 1: 523-545.

Edees, E. S., 1959. The difficulties of a Rubus referee. Proc. bot.

soc. Br. I si. 3: 406-419.

Focke, W. O., 1911-1914. Species Ruborum. Monographiae

generis Rubi Prodromus 1-3. Biblthca bot. 17, 72: 1-223; 19,

83: 1-274.

Grant, V., 1971. Plant speciation. New York: Columbia Univer-

sity Press.

Gustafsson, A., 1942. The origin and properties of the European

blackberry flora. Hereditas 28: 249-277.

Gustafsson, A., 1943. The genesis of the European blackberry

flora. Acta Univ. lund. 39, 6: 1-200.

Gustafsson, C. E., 1933. Rubi africani. Ark. Bot. 26A, 7: 1-68.

Harvey, W. H., 1862. Rosaceae, In W. H. Harvey & O. Sonder

(eds), Flora Capensis 2. Kent: Reeve.

Heslop-Harrison, J., 1963. New concepts in flowering plant tax-

onomy. London: Heineman.

Newton, A., 1975. Rubus L. In C. A. Stace (ed.), Hybridization

and the flora of the British Isles. London: Academic Press.

Sudre, H., 1908-1913. Rubi Europae. Paris: Libraire des

Sciences Naturelles.

Watson, W. C., 1958. Handbook of the Rubi of Great Britain

and Ireland. Cambridge: University Press.

Bothalia 13, 3 & 4: 333-337 (1981)

New records of naturalized Rubus in southern Africa

C. H. STIRTON*

ABSTRACT

Rubus niveus Thunb. and R. phoenicolasius Maxim, are recorded for the first time in southern Africa Notes are

given on their morphology, present distribution and weed status.

RESUME

NOUVELLES ENREGISTREMENTS DU RUBUS NATURALISE EN AFRIQUE AUSTRALE

Rubus niveus Thunb. et R. phoenicolasius Maxim, sont enregislres pour la premiere fois en Afrique australe. Des

notes sur leur morphologie leur distribution actuelle et leur statut en tant que mauvaise herbs sont donnees.

INTRODUCTION

As pointed out in a previous paper (Stirton, 1981),

the taxonomic problems of Rubus in southern Africa

will only be solved after years of intensive field work

and genetic and taxonomic evaluation. However,

because an accurate taxonomy is economically im-

portant to silviculture and agriculture, results will be

published as they become available. This paper

records for the first time in southern Africa two in-

troduced Rubus spp., which have become naturaliz-

ed, namely Rubus niveus, the Java bramble and R.

phoenicolasius, the wineberry.

TAXONOMY

1. Rubus niveus Thunb., Dissert. Rubi 9

(1813); Focke in Biblthca bot. 72: 182 (1911);

Graham in FI. Trop. E. Afr., Rosaceae 40 (I960);

Hanizah, Toha & Van Steenis, Mountain Flora of

Java, t. 45, 5 (1972). Type: Thunberg s.n. herb. no.

12275 (UPS, holo., photo.!).

Primocanes up to 2 m, tip-rooting, bright green

but covered with dense white bloom, eglandular,

glabrous, round, prickles either patent or slightly

deflexed (Fig. 3.1). Floricanes reddish, glabrous,

mostly without bloom. Leaves 3-5-partite on flori-

canes, but mostly 7-partite on primocanes; upper sur-

face of leaflets thinly hairy, sulcinervate, dark yel-

low-green, lower surface white woolly with yellowish

venation, occasionally armed with small recurved

prickles; terminal leaflet broadly ovate or elliptic,

apex acute to acuminate, base truncate to rounded,

larger than laterals; petioles and petiolules tomen-

tose, armed with falcate or deflexed prickles. Stipules

free, ascending, becoming patent, winged at base

(Fig. 3.2). Inflorescence broadly pyramidal or round-

ed, terminal, ± 12-flowered on floricanes, 2 to

4-flowered on primocanes, mostly axillary, leafy,

rhachis tomentose. Sepals up to 6 mm long, ex-

ceeding the petals, densely pilose, tips green and

glabrous. Petals bright pink, shorter than calyx

lobes, 5 mm long and wide, suborbicular with

crinkled margins, apices incurved, touching. Anthers

black, exceeding pink stigmas; filaments pink.

Carpels densely tomentose (Fig. 4.1); fruits orange

when young, turning red and finally greyish purple.

Fig. 1.

Rubus niveus has a wide Asian distribution: Hima-

laya (Kashmir to Sikkim), S. India. Sri Lanka. Bur-

*Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X 101 , Pretoria, 0001.

ma, Thailand, Laos, Vietnam, Malaysia, Java, Bali,

Flores, Timor, Luzon and Celebes (Kalkman, 1976).

In Africa it has been recorded from Kenya, Tan-

zania, Zimbabwe, South Africa and Swaziland (Fig.

2). Van Steenis (1976) considers the African plants to

be conspecific with the Asian plants.

Transvaal. — 2531 (Komatipoort): Ngomangoma Waterfall,

10 km south of Barberton (-CC), Stirton 1766; 12 km from

Barberton on road to Havelock (-CC), Stirton 6857.

Swaziland — 2531 (Komatipoort): near Havelock Mine (-CC),

Codd 7831. 2631 (Mbabane): Mbabane (-AC), Miller 5/200; The

Taven (-AC), Pritchard s.n.; Usutu Forest (-CA), Mott 471;

Ukutula (?), Compton 25060.

Java bramble has been found in grassland, along

wooded stream-banks and among Focks in tall herb

communities. It easily tolerates the semi-shaded con-

ditions prevailing in Eucalyptus plantations, but

tends to favour higher altitude areas between 1 000

and 1 600 metres. It flowers between January and

April.

The origin of Java bramble in Africa is still

speculative. Graham (1960) has suggested that this

species may have been introduced into east Africa

from India or Malaysia, but he cited no evidence. Is

this species a recent introduction or has it occurred in

Africa for a very long time? This question is difficult

to answer and one can only surmise from what little

indirect evidence there is. It is not known whether the

species occurs in Malawi and its occurrence in Zim-

babwe is difficult to explain. One of the Swaziland

specimens ( Miller 5/200) has a note that suggests R.

niveus was introduced into Swaziland from Kenya.

As regards South Africa, there seems little doubt that

Java bramble is a recent arrival in the south-eastern

Transvaal. Firstly, it is unlikely that early collectors

such as Galpin and Thorncroft would have missed

collecting such a distinctive species and secondly, my

own observations are that Java bramble has rapidly

expanded its range over the last five years. This ex-

tension has been particularly noticeable from

Havelock in Swaziland into the Transvaal towards

Barberton. Compton remarks in a note on his collec-

tion 25060 that the plant is a weed in the Ukutula

area. During an excursion in 1977 to the Barberton-

Havelock area, 1 gained the impression that if not

checked Java bramble could become a menace in

forests of the eastern Transvaal. Its eradication

should begin now, while infestations are still small

and economically controllable.

Previously in South Africa R. niveus was referred

to either R. intercurrens C. E. Gust, or R. immixtus

334

NEW RECORDS OF NATURALIZED RUBUS IN SOUTHERN AFRICA

Fig. 1 .—Rubus niveus. 1, mature second year plant with autumn primocanes; 2, 3-foliolate leaf from

floricane; 4, 7-foliolate leaf from primocane; 5, autumn-induced morphological change in stem apices

showing stem tip rooting (Stirton 7146).

335

Figs 2—5. — Rubus niveus. 2, Distribution in southern Africa. 3, part of primocane showing: 1, patent and slightly detlexed

prickles; 2, ascending paired stipules. 4, tip of a 5-foliolate floricane showing: 1, densely tomentose carpels,

2, glabrous, shiny sulcinervate upper surface of leaflets; 3, recurved prickles. 5, under surface ot a -toholate

primocane leaf showing strong nervation and under surface pubescence.

336

NEW RECORDS OF NATURALIZED RUBUS IN SOUTHERN AFRICA

C. E. Gust. There is, however, only a superficial

resemblance to these species. R. niveus is easily

separated from these and most other Rubus spp. in

South Africa by its distinctive concolorous leaves,

almost paralleled secondary sulcinate nervation (Fig.

5) together with its small pink flowers, tomentose

fruits and white bloom on the primocanes. The

nearest species which could be confused with it is R.

ludwigii Eckl. & Zeyh., which also has similar

flowers, hairy fruits, concolorous leaves and a white

bloom. It differs from R. niveus, however, in its

deeply incised leaflets and few secondary veins,

regardless of the size of the leaflets.

R. niveus, like many European brambles of the

subgenus Eubatus Focke, produces positively

geotropic cane tips during the Autumn equinox when

its fruits are ripening. By May the cane tips have

entered the ground and have proliferated a mass of

adventitious roots (Fig. 1.5).

2. Rubus phoenicolasius Maxim, in Izv. imp.

Akad. Nauk. 17: 160 (1872); Hooker in Curtis’s bot.

Mag. t. 6479 (1880); Bailey in Gentes Herb. 5: 902

(1945).

Primocanes up to 2 m tall, robust, arching, tip-

rooting, axis terete with scattered, straight or falcate

prickles, densely covered by red acicles, stalked

glands and a fine wispy white pubescence. Floricane

axis similarly covered. Leaves 3-partite, or upper-

most simple; petioles and petiolules with a thin wispy

pubescence, short acicles and reddish glandular

hairs, armed with falcate pricklets; upper surface of

leaflets thinly-hairy, green or greyish green, lower

surface densely white tomentose with straight prick-

lets scattered along the nerves; terminal leaflet broad-

ly ovate, 7-10 cm long, 3-6 cm wide, cordate or sub-

truncate at base, apex abruptly short pointed, mar-

gins coarsely biserrate, apices of teeth apiculate;

lateral leaflets smaller, oval, assymmetrical, 3-5 cm

long. Stipules linear, persistent, adnate to base of

petiole. Inflorescence a short 8 to 14-flowered ter-

minal raceme. Sepals large, exceeding petals, glan-

dular inside, glandular hairy outside, lanceolate,

spreading in flower, closed during fruit maturation

but open in ripe fruit. Petals white, shorter than

sepals, erect, curved inwards, spathulate, apex cre-

nate, outside ciliate. Stamens short. Carpels hairy;

fruits ovoid, 1-2 cm long, composed of ± red, ellip-

soid glabrous drupes; seeds strongly reticulate. Fig.

6.

C. H. STIRTON

337

Fig. 7.— Distribution of Rubus phoenicolasius in southern Africa.

R. phoenicolasius is an introduced ornamental and

berry plant that has escaped from gardens and is now

locally naturalized in parts of the Natal Midlands

(Fig. 7). It has been found in wasteland and in natur-

al veld and is fairly rare. The earliest record of escape

is 1950.

Natal. — 2930 (Pietermaritzburg): Balgowan (-AC), Marr &

Scotney 8\ Highlands, Richmond area (-CD), Beald 719.

Wineberry is native in Japan, North Korea, South

Korea and northern China (Bailey, 1923). In Japan it

occurs in plagioseral grassland of the subarctic

macroclimatic zones and is one of the principal

species in the fourth (shrubby stage) of the six succes-

sional stages to climax (Numata, 1974). It was in-

troduced from Japan into Europe by Maximovicx in

the 1870’s (Hooker, 1880) and is today still cultivated

for its ornamental value and its edible fruit (Heslop-

Harrison, 1968). In 1890 it was introduced into the

United States by John Childs from seeds obtained in

Japan by G. Georgeson and sent to J. T. Lovett of

New Jersey (Bailey, 1941). It has since escaped culti-

vation and is now established along roadsides, in

thickets and in open woods in several parts of the

north-eastern United States (Ferrald, 1950). It is also

naturalized in parts of Europe.

Wineberry undergoes normal meiosis, with n = 7

(Chomisury, 1927) and 2n = 14 (Darrow, 1937; Jin-

no, 1951; Britton & Hull, 1957). A number of at-

tempts have been made to hybridize wineberry with

commercial crops such as the European raspberry, R.

idaeus L., especially to develop insect and disease-

resistant strains. Wineberry is rather susceptible to

raspberry mosaic virus (Zeffer, 1923; Giddings &

Wood, 1925), a and (3 — leaf-curl viruses (Converse,

1962) and strawberry necrotic shock virus (Frazier,

1966). The raspberry mosaic virus is transmitted

either by aphid vectors such as Amphoraphora rubi

or by grafting (Converse, 1962). The leaf-curl viruses

are also transmitted by an aphid vector, Aphis idaei

(Stace-Smith, 1962). As this plant hosts a number of

important raspberry diseases, it should be eradicated.

Marr & Scotney 8 collected in Richmond was severely

affected by a necrotic viral disease.

ACKNOWLEDGEMENTS

1 am indebted to Professors C. Kalkman and C. J.

van Steenis of Leiden for their useful comments; to

Mr T. H. Arnold, at the time South African Liaison

Officer in Kew and to Dr F. Kupicha for their careful

investigation of the Kew material; to Mrs R. Weber

for the artwork; to Mrs Romanowski for the photo-

graphs; to Dr L. E. Codd for his encouragement and

critical interest; and finally to Prof. K. D. Gordon-

Gray (University of Natal), Dr. R. L. Amor (Keith

Turnbull Research Institute, Australia), Dr O. A.

Leistner and Mr M. J. Wells for editorial comment.

UITTREKSEL

Rubus niveus Thunb. and R. phoenicolasius

Maxim, word hier vir die eerste keer in Suid-Afrika

erken. Die morfologie, verspreiding en onkruidstatus

word bespreek.

REFERENCES

Bailey, L. H., 1923. Various cultivated Rubi aside from Eubatus.

Gent. Herb. 1,4: 147.

Bailey, L. H., 1941. Species Batorum. Gent. Herb. 5, 1: 1-19.

Britton, D. M. & Hull. J. W., 1957. Chromosome numbers of

Rubus. Fruit var. hort. Dig. 11: 58-60.

ChOmisbury, N., 1927. [The germinability and cytology of pollen

of some varieties of Primus and Rubus.] Angew. Bot. 9:

626-636. Biol. Abstr. 4: 1517.

Converse, R. H., 1962. Insect and graft transmission of alpha-

and beta-curl viruses of raspberries. Phytopathology 52: 728.

Rev. appl. Mycol. 42: 538.

Darrow, G. M., 1937. Blackberry and raspberry improvement.

Yearbook USDA. 1937: 496-533.

Fernald, M. L., 1950. Gray’s manual of botany 8th edn.

Frazier, N. W., 1966. Natural and experimental infection of

Rubus with strawberry necrotic shock virus. Phytopathology

56: 568-569.

Giddings, N. J., & Wood. J. I., 1925. Diseases of fruits and nut

crops in the United States in 1924. PI. Dis. Reptr. Suppl. 39.

105 pp.

Graham, R. A., 1960. Rubus. In FI. Trop. E. Afr., Rosaceae,

21-40. London: Crown Agents for Overseas Governments

and Administrations.

Heslop-Harrison, Y., 1968. Rubus DC. In T. G. Tutin, V. H.

Heywood, N. A. Burgers, D. M. Moore, D. H. Valentine, S.

M. Walters, & D. A. Webb (eds), Flora Europaea. Cam-

bridge: University Press

Hooker. J. D., 1880. Rubus phoenicolasius. Curtis’s bot. Mag. t.

6479.

Jinno, T., 1951. [Chromosomes in Rubus. 2.]. Jap. J. Genet. 26:

133-135. PI. Breed. Abstr. 29: 585.

Kalkman, C., 1976. Pers. Comm.

Numata, M., 1974. Grassland Vegetation. In M. Numata (ed.).

The flora and vegetation of Japan. Tokyo: Kodansha Scienti-

fic Books.

Stace-Smith, R., 1962. Studies on Rubus virus diseases in British

Columbia. 8. Raspberry leaf-curl. Can. J. Bot. 40: 651-657.

Stirton, C. H., 1981. The taxonomy of Rubus in South Africa.

Bothalia 13: 331-332.

Van Steenis, C. J., 1976. Pers. Comm.

Zeller, S. M., 1923. Mosaic and other systemic diseases of

brambles in Oregon. Circ. Oregon, agric. Exp. Stn 491,

15 pp.

Bothalia 13, 3 & 4: 339-363 (1981)

Taxonomic studies in the Disinae. V. A revision of the genus

Monadenia

H. P. LINDER*

ABSTRACT

The genus Monadenia (Disinae, Orchidaceae) is revised. Sixteen species arranged in four sections are recognized.

The distribution of each species is plotted, and 12 species are illustrated. A hypothesis on the phylogenetic relation-

ships of the species is presented.

RESUME

ETUDES TAXONOMIQUES DES DISINAE. V. UNE REVISION DU GENRE MONADENIA

Le genre Monadenia ( Disinae , Orchidaceae) est revise. Seize especes arrangees en quatre sections sont reconnues.

La distribution de chaque espece est situee, et 12 especes sont illustrees. Une hvpothese sur les relations phvlogeneti-

ques des especes est presentee.

INTRODUCTION

Monadenia Lindl. is a small genus of 16 species,

which is closely related to Disa Berg. (Orchidoideae,

Orchidaceae). Fifteen of the 16 species are restricted

to the Cape Flora Floral Region (Goldblatt, 1979),

whereas the remaining species is rather widespread in

the montane grasslands (White, 1978) of southern

Africa (Fig. 1). The plants are generally less than 300

mm tall, and have rather inconspicuous green to

brownish flowers.

The first species of Monadenia were described by

Thunberg in 1794 as Satyrium, and were transferred

to Disa by Swartz (1800). The genus Monadenia was

erected by Lindley in 1838. He included eight species

in this genus, of which five were new species. How-

ever, he sowed the seeds for later nomenclatural con-

fusion by misapplying M. rufescens (Thunb.) Lindl.

to another taxon, later to be named M. coinosa by

Reichb. f. (1847). The true M. rufescens he named

M. macrocera. He also gave the new names M.

inicrantha and M. prasinata to what should have

been M. bracteata and M. cernua, respectively.

Sonder (1847) and Reichenbach (1847) added four

new names and one new species to the genus. Our

knowledge of the genus was greatly extended by

Harry Bolus and Rudolf Schlechter, who described

four new and rather rare species in the genus. The

last species was added to the genus in 1948 (Lewis,

1948).

To date there have been four taxonomic revisions

of the group: Lindley (1838), Kraenzlin (1900),

Schlechter (1901) and Rolfe (1913). Lindley only had

material of half the presently known species before

him. By 1900 the majority of species were known, or

at least available in the European herbaria. However,

Kraenzlin missed several species, and his work shows

that he did not have access to sufficient material.

Schlechter’s and Rolfe’s work differ in approach.

Whereas Schlechter employed a broad species con-

cept (combining M. physodes with M. cernua and

M. reticulata with M. macrostachya), Rolfe raised

almost every possible taxon to specific rank. Two

species, M. basutorum (Schltr.) Rolfe and M. leyden-

burgensis Kraenzl., are here transferred to Disa, and

have not been discussed above, and are excluded

from Table 1 .

*Bolus Herbarium, University of Cape Town, Rondebosch, 7700.

The rank of the group has been controversial.

Lindley (1838) recognized the group as a distinct

genus, allied to Disa. This treatment was followed by

Reichenbach (1847) and Sonder (1847), Bolus (1888,

1889, 1893, 1911, 1913) and Schlechter (1898, 1901)

recognized Monadenia as a section of Disa. How-

ever, Rolfe (1913) maintained Monadenia as a dis-

tinct genus, as did Phillips (1926), Schelpe (1966) and

Dyer (1976). The reasons for treating Monadenia as a

distinct genus in this study will be dealt with in detail

in another publication.

MORPHOLOGY

All the species in Monadenia may be shown to be

rather minor variations on a simple pattern: erect

herbs with linear to ovate cauline leaves, inflores-

cence a spike, dorsal sepal erect or curved forwards

and shallowly galeate, oblong to obovate, spur

straight, cylindrical, pendent from the base of the

galea, lateral sepals somewhat smaller than the dorsal

sepal, patent or reflexed, petals erect, obliquely

ovate-oblong, retuse or truncate, lip linear to elliptic,

pendent, lip and petals fleshy, rostellum simple with

a deep central notch for the single large viscidium.

As in the rest of the subtribe Disinae, there are two

Fig. 1.— Distribution of the genus Monadenia. The number of

species occurring in each grid square is indicated. The collec-

tion from Madagascar is not included.

340

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

TABLE 1. — Summary of the history of Monaclenia

Species excluded: M. basutorum (Schltr.) Rolfe = Disa basutorum Schltr.

M. leydenbergensis Kraenzl. = Disa stachyoides Reichb. f.

testicular tubers. There may be some variation in the

tuber shape, but this variation is difficult to quantify

due to lack of material.

The variation in leaf shape may be used to some

extent to subdivide the genus: in sect. Densiflora the

leaves tend to be linear-lanceolate, as compared to

the more lanceolate leaves of the other sections. M.

comosa, the only species that often occurs in half-

shaded conditions, has rather distinctive spreading

ovate leaves, sharply differentiated from the upper

sheathing cauline leaves.

Inflorescence shape also more or less follows the

sectional classification. The small-flowered sect.

Densiflora has compact cylindrical spikes, the large-

flowered sect. Monaclenia semi-dense cylindrical

spikes, and in sect. Tenuicornes the inflorescence

tends to be lax. M. sabulosa and to a lesser extent M.

pygmaea have an obovate inflorescence. This varia-

tion may also be expressed in the angle between the

stem and the ovaries: in sect. Densiflorae the ovaries

spread away from the axis, whereas they are almost

erect and adpressed to the axis in sect. Tenuicornes.

Only in sect. Tenuicornes do the floral bracts show

any sharp discontinuities: in M. rufescence the bracts

are almost of a leathery texture, partially obscuring

the flowers and imbricate, whereas in M. reticulata,

M. comosa and M. bolusiana they are of a more

membranous texture, with the venation clearly visible

on dried specimens, and the bracts not imbricate.

There is little variation in the shape of the dorsal

sepal. In the majority of taxa the galea is best

described as oblong. In some, i.e. M. sabulosa, the

galea is obovate. There is some variation in the angle

at which the galea is held: from vertical to falcately

curved forwards, but this is difficult to quantify.

The lateral sepals are almost invariably oblong. In

some taxa the sepals are reduced in size relative to the

dorsal sepal. This is often associated with being

reflexed (e.g. M. sabulosa). Generally, the apical

parts of the sepals are recurved (e.g. M. bracteata).

This is considered to be the primitive state.

Presumably derived from this state are the sharply

reflexed sepals of M. sabulosa and M. atrorubens or

the spreading sepals of M. ophrydea.

The petal structure is remarkably uniform

throughout the group, and the variation is often

rather subtle. This is shown rather clearly by M.

rufescens, in which the petal shape differs marginally

from M. reticulata, but the angle and the juxtaposi-

tion to the galea is somewhat changed, with a

markedly different effect. M. sabulosa and M. pyg-

maea are somewhat distinct by virtue of the bifid

petals.

Lip shape varies from linear to elliptic in the genus,

and the shape is constant for each species. In sect.

Tenuicornes it may be used as a good differentiating

character for the different species.

Rostellum structure is dominated by the small

anther, often partially pendent, and the single large

viscidium. The rostellum is comparatively simple

with a deep notch in which the viscidium is held. The

rostellum is often flanked by two erect flanges of

tissue, which may also flank the anterior part of the

anther. The size of these flanges varies among the

species. The origin of these flanges is not known.

They may be derived from the staminodes.

The viscidium is generally an almost square struc-

ture. However, in M. rufescens it is elongated to the

back, and has a deep dorsal groove. The two pollen-

masses may be as long as or shorter than the caudi-

cles.

The stigma is variable in shape. In general, the two

lateral lobes are larger than the posterior lobe. In

some cases the stigma is sessile at the base of the

rostellum, in others it may be on a tall, often curved,

stipe as in sect. Tenuicornes.

The spur length and shape provides a very valuable

character and has been used as the diagnostic charac-

ter for the sectional delimitation. In sect. Tenui-

cornes the spur is slender, acute and longer than the

dorsal sepal. In sect. Monadenia it is much inflated,

shorter or longer than the galea, obtuse, rounded or

acute. In sect. Densiflorae the spur is as long as or

shorter than the galea, slender or subclavate, acute to

retuse, occasionally constricted at the base.

H. P. UNDER

341

Kami es-

bergensi s

summer rainfall forms

spur acute

west coast form, spur short

south coast, spur long

M. brevicornis

M. cernua

Fig 2. — Postulated phylogenetic relationships among the species

in Monadenia.

The morphological relationships in the genus are

summarized in the cladogram in Fig. 2. The sectional

classification is based on flower size, inflorescence

shape and spur shape and length. The first two

characters are related to each other, and also to

overall plant size. Other characters that vary within

the genus (sepal reflexion, stigma size, lip shape and

leaf structure) have been found to vary between

otherwise obviously closely related taxa (e.g. M.

atrorubens and M. ophrydea for sepal reflexion, lip

shape, stigma size, and M. comosa and M. bolusiana

for leaf shape), and are therefore, unlikely to reflect

evolutionary tendencies.

PHYLOGENY AND EVOLUTION

The postulated phylogeny of the genus is derived

from the relationships among the taxa as indicated by

the morphological data. However, to arrange the

taxa into a sequence, ancestral and derived character

states have to be suggested. The relative age of the ex-

tant taxa may be estimated from eco-geographical

data.

Ancestral character states are probably those states

found in other taxa in the Disinae, and also simpler

states from which more complex states may be deriv-

ed. In Monadenia the ancestral form was probably a

slender herb with narrowly lanceolate, subimbricate

cauline leaves, an inflorescence with numerous

medium-sized flowers with the sepals about 6 mm

long, spreading to slightly reflexed, subequal, the

dorsal sepal shallowly galeate with a slender spur as

long as or slightly shorter than the sepal, subpendent,

petals falcate, erect next to the rostellum, wider

towards the base, lip lorate, obtuse, pendent, anther

subpendent, rostellum low with a single viscidium,

stigma sessile. Such a plant would be comparable to

the suggested ancestral forms found in several sec-

tions of Disa (Disa polygonoides Lindl., D. vaginata

Lindl. and D. obtusa Lindl.). In Monadenia, M.

bracteata has the greatest similarity to this set of

character states.

Relatively older species may be expected to be mor-

phologically isolated, often with relic distributions in

the form of widely scattered populations over large

distribution areas, and possibly with wide ecological

amplitudes. Morphologocally isolated species are

found in sect. Densiflorae (M. sabulosa, M.

pygmaea, M. ecalcarata and M. conferta), whereas in

the rest of the genus only M. rufescens is mor-

phologically clearly isolated.

Several species in the genus show wide distribution

ranges (M. bracteata, M. densiflora, M. comosa and

M. reticulata), but only M. densiflora and M.

reticulata have disjunct distributions. The majority

of the species show a restricted distribution range,

either on the west coast, or the south coast, or the

area bounded by Caledon, Worcester and Cape

Town. Two species are only known from very

restricted localities: M. ecalcarata from a single

specimen on the Cape Peninsula and M.

macrostachya from a single locality in the

Ramies berg.

All the species in sect. Densiflorae have at least

some morphologically or distributionally primitive

features: M. bracteata and M. densiflora with

widespread distributions and being morphologically

ancestral, and the remainder of the species in the sec-

tion being morphologically isolated.

The remaining species in the genus are grouped in-

to interlinking species groups, in which the individual

species are often difficult to separate rigorously.

They present the aspect of a rapidly evolving group.

Sect. Tenuicornes has diversified mainly in the

mountains of the Cape, with one group adapted for

the immediately post-fire temporal niche (M.

ophrydea and M. atrorubens), with suitable cryptic

coloration, and the other group adapted for unburnt

vegetation, with one species found on rock ledges

and the other at high altitudes in high-altitude low

heathlands (Taylor, 1978) (M. comosa and M. bolu-

siana). One isolated species in the section appears to

have become specialized for a particular pollination

syndrome ( M . rufescens). These five species

therefore show a certain degree of specialization, and

may all be derived from M. reticulata, the

widespread and presumably simplest member of the

section.

Sect. Monadenia has a rather specialized spui-

shape, and is a lowlands group, except in the summer

rainfall area. The three species are geographically

separated.

The single member of sect Kamiesbergenses is best

understood as the link between the older sect. Densi-

florae and the derived remainder of the genus.

These postulated phylogenetic relationships are

simplistically displayed in Fig. 2.

MONADENIA

Monadenia Lindl., Gen. Sp. Orch. 356 (1838);

Benth. & Hook, f., Gen. PI. 3: 630(1883); Pfitzer in

342

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

Natiirl. PflFam. 2, 6: 98 (1889); Kraenzl., Orch.

Gen. Sp. 1: 808 (1900); Rolfe in FI. Cap. 5, 3: 186

(1913); Senghas in Schltr., Die Orchideen 1: 274

(1972); R. A. Dyer, Gen. 2: 995 (1976). Type species:

Monadenia brevicornis Lindl., Gen. Sp. Orch. 357

(1838) (lectotype).

Disa Berg. sect. Monadenia (Lindl.) H. Bol. in

Trans. S. Afr. phil. Soc. 5: 137 (1888); Schltr. in Bot.

Jb. 31: 202 (1901).

Monadenia brevicornis is here selected as the lecto-

type, since it is the only correct name Lindley placed

in the genus in 1838.

The generic name refers to the single viscidium and

is derived from the greek ‘monos’ = one and ‘ade-

nos’ = gland.

Leaves all cauline; dorsal sepal shallowly galeate,

spur pendent from the base of the galea; petals obli-

quely ovate-oblong, subcarnose, partially exserted

from the galea; lip linear to elliptic, subcarnose;

anther usually semi-pendent, with a single large visci-

dium.

Plants terrestrial, herbaceous, usually erect, 40-600

mm tall; tubers 2, testicular or cylindrical; basal

sheaths hyaline, obtuse; leaves all cauline, usually

imbricate, the leaf bases sheathing, the blades elliptic

to linear-lanceolate, usually erect, the lowermost the

largest, the upper grading into the floral bracts; in-

florescence cylindrical to secund, dense to lax;

ovaries usually twisted, 5-25 mm long; bracts ovate

to lanceolate, acute to acuminate, slightly shorter

than the ovaries to longer than the flower. Flowers

resupinate (except M. ecalcarata Lewis), purplish,

brown or green; dorsal sepal shallowly galeate, usuaL

ly oblong, obtuse, 2,5-15 mm long; spur pendent

from the base of the galea, slender or clavate, ob-

solete or longer than the galea; lateral sepals patent

or reflexed, usually oblong, often shorter than the

galea; petals generally obliquely narrowly ovate-

oblong, acute to bifid, the broad base enclosing the

anther and the apex erect in the galea, partially ex-

serted from the galea; lip patent to pendent, linear to

elliptic, subfleshy; anther horizontal to semipendent

with a single large concrete viscidium and two cells;

rostellum simple with a deep notch containing the

viscidium and with two often well developed lateral

flanges flanking the anterior part of the anther;

stigma equally or unequally tripulvinate, shortly

stipitate to as tall as the rostellum.

Sect. Densiflora Linder, sect, nov., calcari sepalo

dorsalo plerumque breviore, sepalis 2,5-5 (-7) mm

longis, inflorescentia densa dignoscenda.

Spur usually shorter than the dorsal sepal, rarely as

long as the dorsal sepal; flowers small, sepal 2,5-5

(-7) mm long; inflorescence dense, cylindrical.

Type species: Monadenia bracteata (Swartz) Dur.

& Schinz.

This section is linked to the rest of the genus by M.

densiflora. In the M. densiflora/ M. bracteata group

the spur is straight or slightly curved, and about as

long as the dorsal sepal. From this group two lines of

development may be postulated: a line in which the

spur is highly reduced, and flower size is decreased

KEY TO SPECIES

la Spur as long as or shorter than the dorsal sepal:

2a Spur saccate or obsolete:

3a Inflorescence slender, 1 0 mm in diameter; lateral sepals 2,5 mm long 2. AT. conferia

3b Inflorescence stout, 1 5 mm in diameter; lateral sepals 4 mm long I . AT. ecalcarata

2b Spur well developed, more than 1 mm long:

4a Spur clavate, rounded 14. AT. physodes

4b Spur slender:

5a Spur about as long as the dorsal sepal:

6a Spur with a sharp bend; petals deeply bilobed 3. AT. sabutosa

6b Spur straight or gently curved; petals rounded to obtuse:

7a Lateral sepals c. 7 mm long 7. AT. macrostachya

7b Lateral sepals less than 4 mm long 5. AT. bracteata

5b Spur about Vi as long as the dorsal sepal:

8a Spur constricted at the base and triangular in cross-section 6. AT. densiflora

8b Spur cylindrical; not constricted at the base 4. AT. pygmaea

lb Spur longer than the dorsal sepal:

9a Spur clavate:

10a Spur rounded 15. AT. cernua

10b Spur acute 16. AT. brevicornis

9b Spur slender:

1 la Plants suffused beetroot-red when fresh, reddish brown when dry:

12a Lip 8-10 mm long, narrowly elliptic to lorate; lateral sepals spreading 12. AT. ophrydea

12b Lip 5-7 (-8) mm long, oblong to narrowly oblong; lateral sepals reflexed 11. AT. atrorubens

1 lb Plants green when fresh, pale to dark brown when dry:

13a Bracts leathery, venation not visible; petals purple 13. AT. rufescens

13b Bracts membranous to thin-tissued, venation visible; petals lime-green:

14a Lip lorate; leaves linear-lanceolate 8. AT. reticulata

14b Lip elliptic; leaves elliptic to lanceolate:

15a Lowest 2 (-3) leaves elliptic, spreading, sharply differentiated from the remaining leaves, flowe-

ring mostly in October 10. AT. comosa

1 5b Leaves gradually smaller from the base to the apex of the stem; flowering mostly in December and

January 10. AT. bolusiana

H. P. LINDER

343

(M. ecalcarata, M. conferta), and a line in which the

spur becomes sharply curved, the inflorescence

becomes relatively massive, and the flowers face

downwards (M. pyginaea, M. sabulosa).

The M. bracteata/M. densiflora group is here con-

sidered as being ancestral. These two species are

closely related, and may be separated by the shape of

the spur and the galea. Both taxa are ecologically

diverse and widespread in the Cape Flora. M.

bracteata is ubiquitous in the area, and M. densiflora

shows quite a remarkable degree of variation in the

floral morphology (see below).

Both Monadenia ecalcarata and M. conferta are

rare M. ecalcata is known from a single specimen,

whereas the latter taxon is more widespread, but

populations are very sparse. As both taxa show

specializations not found in the other (the non-

resupinate flowers in M. ecalcarata and the beetroot-

red colouring in M. conferta ), they cannot be placed

in an ancestor-descendant relationship.

Monadenia sabulosa and M. pyginaea are

restricted to the Cape Peninsula and the Caledon

District, showing the False Bay disjunction in their

distributions as recorded by Rourke (1972) for

Leucospermum hypophyllocarpodendron subsp.

hypophyllocarpoclenclron and Linder (1981b) for

Herschelia purpurascens. However, as orchids are

rather easily distributed by seed, it is suggested that

this distribution pattern is the result of the distribu-

tion of available habitats, rather than of geological

history, as suggested by Rourke (1972) for Leuco-

spermum hypophyllocarpodendron . Although the

group is quite distinctive, the species are clearly

distinct. Although there is some overlap in the

habitats of the species, M. pygmaea generally occurs

on more rocky and more mountanous localities,

whereas M. sabulosa is restricted to lowland sand-

flats. Both taxa are generally recorded as flowering

after fire.

1. Monadenia ecalcarata Lewis in J1 S. Afr.

Bot. 14: 31 (1948). Type: Cape Province, Cape

Peninsula, on damp rocky northern slopes of Con-

stantiaberg, Lewis 1487 (SAM, holo.!).

Icon: .11 S. Afr. Bot. 14: 32, Fig. 2 (1948).

Plant 130 mm tall; tubers testicular, 10 mm in dia-

meter; basal sheaths 2, hyaline, obtuse, up to 20 mm

long; leaves linear-lanceolate, acute, erect and curved

towards the stem, conduplicate, the longest at the

base of the stem, 70 mm long, the upper grading

rapidly into the floral bracts, densely imbricate; in-

florescence a dense cylinder, 60 mm long and 15 mm

in diameter with numerous imbricate flowers; ovaries

slender, c. 10 mm long; bracts as tall as the flowers,

lanceolate, acuminate. Flowers not resupinate, lime-

green; dorsal sepal shallowly galeate at the base, nar-

rowly oblong, obtuse, 4 mm long; petals obliquely

narrowly oblong, obtuse, fleshy, 3 mm long; lip pen-

dent, lorate, obtuse, 3-4 mm long; antb about 1

mm long; rostellum small, erect; stir a nearly

square, pulvinate.

Diagnostic features. Flowers with lateral sepals 4 mm

long, spur obsolete, flowers not resupinate.

Flowering time: October.

This species is known from a single collection. It

differs in too many characters from its nearest rela-

tions, M. conferta and M. micrantha, to be con-

sidered as an aberrant form. It is puzzling that an or-

chid should be so rare (Fig. 3).

2. Monadenia conferta (H. Bol .) KraenzL, Orch.

Gen. Sp. 1: 810 (1900); Rolfe in FI. Cap. 5,3: 187

(1913). Type: Cape Province, Cape Peninsula,

Raapenburg, Guthrie 725 ( = Bolus 7097) (BOL, lec-

to.l; K!).

Disa conferta H . Bol., leones Orch. Austro- Afr. 1 : t. 28 ( 1 893);

Schltr. in Bot. Jb. 31: 212 (1901).

Icon: H. Bol., leones Orch. Austro-Afr. 1: t. 28

(1893).

Plant slender, suffused with beetroot-red, 80-220

mm tall; tubers c. 10 mm long; leaves linear, acute,

numerous, densely imbricate, the largest generally

near the base of the stem, up to 70 mm long, erect,

conduplicate, grading apically into the floral bracts;

inflorescence dense, cylindrical, slender 30-130 mm

long and 10 mm in diameter; ovaries 5-10 mm long;

bracts as tall as or taller than the flowers, lanceolate,

subacuminate, longer towards the base of the inflo-

rescence. Flowers lime-green at anthesis, sepals soon

turning beetroot-red; dorsal sepal shallowly galeate,

rounded to obtuse, oblong, 2,5 mm long; spur sac-

cate, 0,1 -0,2 mm long; lateral sepals reflexed,

oblong to narrowly oblong, rounded to obtuse, 2 mm

long, shallowly concave, erect next to the rostellum;

lip pendent, narrowdy oblong, obtuse 2-2,5 mm

long; another horizontal, 0,5-7 mm long, connective

longer than the pollen masses, viscidium minute;

rostellum 1 mm tall with large lateral lobes flanking

the anther; stigma fused to the base of the rostellum.

Fig. 4.

Diagnostic features. Flowers minute, lateral sepal

2,5 mm long; inflorescence dense, slender, cylin-

drical; spur obsolete.

Flowering time: (September-) October

(-December).

344

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

Fig 4. — Monadenia conferta. 1, flower seen from behind, x 10.

2, flower in front view, x 10. 3, dissected flower: a, rostellum;

b, petal; c, anther; d, stigma, x 20. All drawn from Williams

s.n. (BOl ).

Locally rare to common in the western Cape Pro-

vince, occurring on seasonally damp to dry sand,

rarely on rocky or gravelly slopes (Fig. 5).

Cape *-3419 (Caledon): Roue Rivier Berge near Elim (-AD),

Dec. 1896, Schlechter 9618 (BM; BOL; G; K; P; PRE). 3418

(Simonstown): Buffels River Valley at Hangklip, on a rocky

hillside, (-BD), Nov. 1945, Barker 3925 (NBG); Cape Point

Nature Reserve, Olit'antsbos (-AD), Oct. 1955, Lewis 4455

(SAM).

This species may readily be identified by the

slender beetroot-red plant with minute spurless

flowers.

Only one population has been studied in the field

( Linder 1528, Caledon District, Babylon’s Tower).

Occasional to rare individuals occurred on dry scree

slopes, flowering one year after a fire. Several collec-

tors noted that the species flowered after fire. The

altitude range of the species is from sea level to about

600 m.

Bolus (1893) described Disa conferta from two liv-

ing specimens from different localities. These he cites

as Bodkin in BOL 6231 from near Houw Hoek and

Guthrie in BOL 7097 from Raapenburg on the Cape

Peninsula. The latter collection is also in BOL under

the number Guthrie 725. The Guthrie collection is

also duplicated in Kew and is here selected as the

lectotype.

3. Monadenia sabulosa (H. Bol.) Kraenzl-, Orch.

Gen. Sp. 1: 814 (1900); Rolfe in FI. Cap. 5,3: 195

(1913). Type: Cape Province, Cape Peninsula,

Kenilworth, Bolus 1374 (= 7104) (BOL, holo . ! ;

BM!; G!; K!; P!; SAM!; UPS!; W!; ZT!).

Disa sabulosa H. Bol. in leones Orch. Austro-Afr. 1 t. 27

(1893); Schltr. in Bot. Jb. 31: 207 (1901).

Icon: FI. Bol., leones Orch. Austro-Afr. 1: t. 27

(1893).

Plants 80-200 mm tall; leaves lanceolate, acute,

sheathing at the base, semi-erect, the margins un-

dulate, imbricate, the largest leaves near the base of

the stem, up to 50 mm long, the apical leaves grading

into the inflorescence; inflorescence dense, obovate,

30—70 mm long and up to 40 mm in diameter; ovaries

c. 15 mm long; bracts as tall as the flowers,

lanceolate, acute, green. Flowers with lime-green

sepals and yellow petals and with a faint scent; dorsal

sepal subspathulate to broadly obovate, falcate in

sideview, the upper half shallowly galeate, obtuse,

10-15 mm long; spur pendent, flexuose near the

base, slender, acute to retuse, 10-15 mm long; lateral

sepals reflexed, oblong, rounded, 7-9 mm long,

veins prominent; petals subobliquely oblong, the

apex deeply and equally bifid, the lobes 2 mm long,

petal 7-8 mm long, erect, twisted to face forwards;

lip oblanceolate, obtuse, pendent, subfleshy, c. 6 mm

long; anther pendent, 2 mm long, viscidium large,

square, caudicles 1,5-2 mm long; rostellum lateral

lobes partially flanking the anther, c. 2 mm tall;

stigma Hat, horizontal, 1 mm tall. Fig. 6.

Diagnostic features. Dorsal sepal subspathulate, fac-

ing downwards; spur geniculate; petals deeply bifid.

Flowering time: October (-November).

A rare, stout, greenish orchid, occurring on damp

to dry sand, usually flowering after fires, in the

western Cape Province (Fig. 7).

♦In the treatment of each taxon only representative specimens have

been cited. A full list of all specimens studied is given in the

Appendix, pp. 361-363.

H. P. LINDER

345

Fig 6. — Monadenia pygmaea (1-5) and M. sabulosa (6-10).

1, plant of M. pygmaea, x 1, from Stokoe s.n. (BOL). 2,

flower in side view. 3, flower in front view. 4, dissection of

flower. 5, dorsal sepal, (2-5 all drawn from Linder 1507, x

6.) 6, plant of M. sabulosa, X 0,5, from Guthrie 1080. 7,

flower in side view. 8, dorsal sepal in rear view. 9, flower in

front view. 10, dissection of flower: a, rostellum; b, petal; c,

anther; d, stigma. (7-10 from Linder 1508; 7-9 x 3 and 10 x

5.)

Cape. — 3318 (Cape Town): sandflats at Kenilworth near

Wynberg (-CD), Oct. 1891, MacOwan & Bolus U74 (BOL;

SAM). 3418 (Simonstown): near Hangklip on road to Betty’s Bay,

on burnt ground (-BD), Oct. 1961, Thomas in NBG 56374 (NBG).

The type collection of this species was distributed

under two different numbers — as A. Bolus in BOL

7104 (selected as lectotype) and as A. Bolus sub

MacOwan & Bolus 1374.

Monadenia sabulosa may readily be distinguished

from its allies by the deeply bifid petals, and from all

the species in the genus by the sharp knee-bend in the

spur and the subspathulate dorsal sepal.

A population studied at Betty’s Bay ( Lincler 1508)

occurred in sand among boulders on the lower slopes

of the mountains, in a fire-break. Individuals occur-

red frequently, mostly in seasonally damp sand. Col-

lectors’ notes indicate that this species is most

frequently collected after the climax vegetation has

been removed (i.e. by fire or mechanical means) from

sandy areas. The species ranges from near sea level to

about 1 200 m altitude.

The conservation status of the species ot the Cape

Peninsula is not clear. The majority of collections are

from areas now heavily disturbed or built over.

However, the status in the Betty’s Bay area appears

to be satisfactory. It must be noted that the species is

restricted to the lower slopes of the mountains and

the coastal Hats, which are prone to invasion by

various introduced Acacia species. In the long term,

these localities will require active conservation.

4. Monadenia pygmaea (H. Bol.) Dur. & Schinz,

Consp. FI. Afr. 5:111 (1894); Kraenzl. in Orch. Gen.

SP. 1: 813 (1900); Rolte in FI. Cap. 5,3: 190 (1913).

Type: Cape Province, Cape Peninsula, Muizenberg

Mountain, Bodkin in BOL 4970 (BOL, holo.!; G!;

K!; SAM!).

Disa pygmaea H. Bol. in J. Linn. Soc., Bot. 20: 72 (1885);

Schltr. in Bot. .lb. 31: 213 (1901).

Icon: H. Bol., leones Orch. Austro-Afr. 3: t 37

(1913).

Plants 45-150 mm tall; leaves narrowly ovate,

acute, imbricate, semi-erect, all equal in size, 15-20

mm long; inflorescence cylindrical, dense, 15-100

mm long, longer than the leafy part of the stem;

ovaries c. 5 mm long; bracts taller than the flowers,

lanceolate, acute, green. Flowers with lime-green lip

and petals, green lateral sepals and rusty brown dor-

sal sepal and spur; dorsal sepal shallowly galeate,

oblong, obtuse, 5-6 mm long and up to 2 mm deep;

spur c. 2,5 mm long, cylindrical, retuse, falcately

curved towards the ovary; lateral sepals patent with

the apices reflexed, oblong, obtuse, 4 mm long;

petals obliquely narrowly oblong with a small tooth

to the rear, the apex shortly and acutely bifid, erect,

twisted to face partially forwards, 4 mm long; lip

subpendent, the apex reflexed, lorate, 3-4 mm long;

anther pendent, 1 mm long; viscidium elongated with

a dorsal groove; rostellum with small side lobes, 1

mm tall; stigma 0,3 mm tall, somewhat angled for-

wards. Fig. 6

346

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

Diagnostic features. Flowers small, lateral sepals 4

mm long; plants less than 150 mm tall, inflorescence

longer than the leafy stem.

Flowering time: October and November.

Monadenia pygmaea occurs occasionally in small

populations in sandy areas, often flowering after fire

(Fig. 8).

Cape. — 3418 (Simonstown): Steenberg Plateau, burnt sandy

slope near reservoir (-AB), Oct. 1945, Lewis 1095 (SAM). 3419

(Caledon): Klein Riversbergen near Onrust (-AC), Nov. 1896,

Sch techier 9501 (BOL).

This species is allied to M. sabulosa by the bifid

petals and the slight knee-bend in the spur, but may

be distinguished from that species by the much

smaller flowers and plants (sepals about 4 mm long).

From the rest of the genus it may be separated on the

inflorescence that is always longer than the leafy

stem, the flowers that face downwards, and the short

spur.

A population studied near Betty’s Bay (Linder

1 507) was found near M. sabulosa in sand in a recent-

ly cleared fire-break at the base of the mountain. In-

dividuals occurred frequently in seasonally damp

sandy areas. Other collectors also generally indicate a

sandy substrate, except two collections from the

Lebanon Forest Reserve, which are from dry rocky

slopes, one with loose rocky sand. Several collections

were made after fire. The altitude range of the species

is from sea level to almost 1 000 m. The rainfall

ranges from about 800 mm to over 1 500 mm in some

localities, mostly restricted to the winter.

Although the majority of collections are from

localities suffering from severe disturbance, the

species has also been recorded from localities likely

to have some long-term protection, such as the

Lebanon Forest Reserve.

5. Monadenia bracteata ( Swartz ) Dur. & Schinz,

Consp. FI. Afr. 5:111 (1894). Type: Cape of Good

Hope, Sparrman s.n. (W, holo.!).

Disa bracteata Swartz in Vet. Acad. Handl. 21: 211 (1800).

Monadenia micrantha Lindl., Gen. Sp. Orch. 357 (1838);

Kraenzl., Orch. Gen. Sp. 1: 818 (1900); Rolfe in FI. Cap. 5, 3: 190

(1913). Disa micrantha (Lindl.) H. Bol. in Trans. S. Afr. phil. Soc.

5: 142 (1888); Schltr. in Bot. Jb. 31: 213 (1901). Type: Knysna,

Ruigtevallei Drege 1261 (K, holo.!; G! ; P ! ; W!).

Disa praetermissa Schltr. in Ann. Transv. Mus. 10: 246(1924).

Type: Humansdorp, sandy grassland near Stormsrivier, Schlechter

s.n. (B, holo.!).

Monadenia australiensis Rupp in Austr. Orch. Rev. 11: 70

(1946). Type: West Australia, Stirling Range, Youngs Siding,

Southerland s.n. (PERTH, holo.).

Plants up to 300 (-500) mm tall; leaves linear-

lanceolate, acute, imbricate, numerous, erect and

usually curved inwards, the largest near the base,

40-120 mm long, grading apically into the floral

bracts; basal sheaths hyaline, obtuse, 10-20 mm

long; inflorescence cylindrical, imbricate, flowers

numerous, 20-120 mm long; ovaries erect, 6-10 mm

long; bracts shortly overtopping the flowers, rarely

much taller than the flowers, narrowly ovate,

acuminate. Flowers green with the sepals usually

tinted maroon; dorsal sepal shallowly galeate, broad-

ly oblong, obtuse to acute, 3-4 mm long and c. 1 mm

deep, falcate in side view; spur shallowly triangular,

pendent from the base of the galea, rounded to trun-

cate, 3-4,5 mm long; lateral sepals patent with the

apices usually reflexed, oblong, obliquely acute,

2, 5-3, 5 mm long; petals erect, partially included in

the galea, broadly obtriangulate in outline, obliquely

acute, 2-2,5 mm long, concave, venation strongly

falcate; lip pendent narrowly oblong to lorate, often

oblanceolate, 2-2,5 mm long; anther pendent, c. 1

mm long; rostellum with a large central notch,

staminodes of variable size; stigma pulvinate. Fig.

10.

Diagnostic features. Flowers small, lateral sepals

2, 5-3, 5 mm long, spur pendent, 3-4,5 mm long,

slender, rounded or truncate, inflorescence dense,

cylindrical.

Flowering time: September-November.

Monadenia bracteata is a widespread and common

little orchid (Fig. 9), that occurs frequently in

wasteland areas, especially in areas of mild distur-

bance. In undisturbed vegetation it is widespread,

but somewhat rarer. In Australia it has been recorded

as a roadside weed.

Cape. — 3318 (Cape Town): Rondebosch, University of Cape

Town football field (-CD), Sept. 1946, Leighton 2079 (BOL).

3418 (Simonstown): Fish Hoek mountain, in shade under trees

(-AB), Nov. 1944, Lewis 823 (SAM). 3322 (Oudtshoorn):

Saasveld, near George, in grassland at the forest edge (-DC), Oct.

1965, Morze 2026 (BOL). 3323 (Willowmore): near Stormsriver

mouth (-DD), Nov. 1894, Schlechter 5965 (BM; G; K; W; Z).

The identity of Disa bracteata has led to much con-

fusion. N. E. Brown applied the name to an illustra-

tion of what is clearly Monadenia multiflora. Lindley

(1838) cited a plate of Brown (1818) as being Disa

bracteata. Bolus (1888) discussed the problem of the

identity of D. bracteata, indicating that it could not

be confused with Disa cylindrica and that it is likely

to be a Monadenia. He could not, however, trace a

type specimen either in the Swartz or Thunberg her-

barium. Rolfe (1913) was of the opinion that D.

bracteata is M. multiflora, as ‘It does not describe

the cylindrical spur and linear lip of M. micrantha

Lindl’. In Vienna I located a capsule containing

several Bowers and an inflorescence of Monadenia

micrantha. The capsule was annotated ‘ Disa

bracteata Swartz scripsit’. The words ‘Disa’ and

‘Swartz scripsit’ were in a hand that I do not know,

but ‘bracteata’ is written in a hand closely resembling

H. P. LINDER

347

Fig 9. — Distribution of Monadenia bracteata.

that of Olof Swartz (vide Burdet, 1978). As there are

several sheets of Swartz types at Vienna (Linder,

1981a), this is not exceptional, and might well be the

type of Disa bracteata.

The type of Disa praetermissa has unfortunately

been lost. However, from the protologue this name

has to be included in the synonymy here, as the

distinguishing characters mentioned by Schlechter

(1924) (Spur length and leaf shape) are quite variable

in M. bracteata.

Monadenia bracteata is closely allied to M. den-

siflora, but may be distinguished from this species by

the terete spur, which is generally as long as the dor-

sal sepal. In M. densiflora, the galea is often

obovate, a state that never occurs in M. bracteata.

From the rest of the genus these two species are

separated by the small flowers (sepals less than 8 mm

long) and spur as long as or shorter than the dorsal

sepal.

This is by far the most common species of Monad-

enia, and occurs in a wide range of habitats, from dry

sandy areas to gravelly mountain slopes, on damp

clayey soils and in black turf sand. The altitude range

of the species is from sea level to 1 500 m, and speci-

mens have been collected in full sunlight and in

shade. The rainfall regime over the distribution range

varies from less than 800 mm p.a. to well over 1 600

mm, and from an all-year rainfall in the Knysna-

Humansdorp area to several months of summer

drought in the Clanwilliam area.

In undisturbed vegetation populations have a low

density and individuals occur scattered over a large

area. In disturbed conditions (road-cuttings) and

fills, old fields, gravel pits, sport fields, etc.) large

populations may be found. Over its whole distribu-

tion range, this species is well established as a

pioneer.

M. bracteata has also been recorded from Austra-

lia, initially under the name M. australiensis, as it

was thought to be indigenous to Australia (Rupp,

1946). Erickson ( 1965) noted that Tt was found to be

most common on the roadsides, the usual lodging

place for immigrant plants ...’ Pocock (1972)

recorded this species only from the south-west of

Australia, flowering in November. It was probably

introduced accidentally to that continent.

Fig 10. — Monadenia bracteata (1-3) and M. densiflora

(4-6). 1, plant of M. bracteata, x 0,5, from Esterhuysen

12074. 2, flower in side view. 3 petal. (2-3 from Watters 508,

x 5.) 4, flower of M. densiflora in front view. 5, flower in

rear view, showing the spur. 6, dissection of flower: a

rostellum; b, petal; c, anther; d, stigma. (4-6 from Linder

1513, x 5.)

348

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONA DEN I A

Fig. 11. — Distribution of Mona-

denia densiflora.

6. Monadenia densiflora Lindt. , Gen. Sp. Orch.

357 (1838); Kraenzl., Orch. Gen. Sp. 1: 814 (1900).

Lectotype: Cape of Good Hope, Thom s.n. (K,

lecto.!).

Disa densiflora (Lindl.) H. Bol. in J. Linn. Soc., Bot 25: 197

(1889).

Monadenia multiflora Sond. in Linnaea 19: 101 (1847);

Kraenzl., Orch. Gen. Sp. 1:811 (1900); Rolfe in FI. Cap. 5, 3: 188

(1913). Disa multiflora (Sond.) H . Bol in Trans. S. Afr. phil. Soc.

5: 140 (1888); Schltr. in Bot . Jb. 31: 215 (1901). Type: Cape Prov-

ince, Cape Peninsula, Doornhoogte, Ecklon & Zevher s.n. (S,

holo.!; K!; P!; W!).

Disa auriculata H. Bol., leones Orch. Austro-Afr. 1: t. 77

(1896); Schltr. in Bot. Jb. 31: 214 (1901). Monadenia auriculata

(H. Bol.) Rolfe in FI. Cap. 5, 3: 189(1913). Syntypes: Cape Prov-

ince, Cape Peninsula, Guthrie s.n. (BOL!); Humansdorp, near

Storms River, Schlechter 5958 (BOL!; BM!; G!; W!; Z ! ).

leones: H. Bol., leones Orch. Austro-Afr. 1: t. 77

(1896), as Disa auriculata: 3: t. 39 (1913), as D.

multiflora.

Plants slender to robust, 75 — 1 95—350 mm tall;

tubers up to 20 mm long; basal sheaths hyaline,

obtuse to apieulate, 2-3, up to 40 mm long; leaves

linear-lanceolate, acute, conduplicate, numerous,

imbricate, 50-90-140 (-220) mm long, generally

reaching to the base of the inflorescence and grading

apically into the floral bracts; inflorescence slender

cylindrical, 25-95-180 mm long, usually longer than

the leafy shoot, flowers numerous, imbricate; ovaries

5-10 mm long, twisted; bracts as tall as or over-

topping the flowers, generally longer towards the

base of the inflorescence, marrowly ovate, acuminate.

Flowers with dull green petals and lip, and rusty red

to green sepals; dorsal sepal shallowly galeate,

oblong, obtuse, erect or falcately curved forwards,

3-5 (-7) mm long and 0,5-1 mm deep; spur pendent

from the base of the galea, triangular in cross-

section, constricted at the base, apex obtuse to

retuse, (1-) 2-3 (-4) mm long, shorter than the

galea; lateral sepals . oblong-ovate, sub-obtuse to

rounded, spreading, apices reflexed, 3-5 (-7) mm

long; petals erect, narrowly oblong, subfalcate,

2,5-5 (-7) mm long, rounded to rarely acute, the

apical 1/3 fleshy; lip pendent, narrowly oblong to

oblanceolate, rounded, 2,5-5 (-7) mm long, apex

fleshy; anther pendent, c. 0,8 mm long; rostellum

with a deep notch for the single viscidium, viscidium

flanked by two flat projections; stigma horizontal,

unequally three-lobed. Fig. 10.

Diagnostic features. Flowers small, lateral sepals

3-5 (-7) mm long, spur shorter than the galea, 2-4

mm long, triangular in cross-section, somewhat con-

stricted at the base; plants 75-350 mm tall.

Flowering time: (September-) October (-December).

Monadenia densiflora is a widespread, but usually

not common, species, occurring usually in sandy

places, but also in a wide range of other habitats

(Fig. 11).

Cape. — 3318 (CapeTown): Table Mountain, Groene Kloof, 600

m (-CD), Nov. 1897, Galpin 4612 (PRE), 3418 (Simonstown):

Constantiaberg, slopes above Baviaans Kloof (-AB), Oct. 1947,

Lewis 1833 (SAM); Krom River (-AB), Oct. 1945, Compton

17486 (NBG). 3319 (Worcester): Wemmershoek Peak, 1 650 m

(— CC), Dec. 1944, Lewis 850 (SAM). 3323 (Willowmore): at

Stormsriver Mouth (-DD), Nov. 1978, Linder 1903 (BOL).

Monadenia densiflora Lindl. was based on two

syntypes at Kew, namely Thom s.n and Drege 8288.

According to Article 9.2 of the I.C.B.N. (1978), the

element most closely resembling the protologue has

to be selected as the lectotype. Thom s.n. most close-

ly resembles the protologue. ‘Calcare pendulo fili-

forme antice alato’ and ‘labello lineari-spathulato

carnoso’ can only refer to this element, whereas

‘petalis galea duplo brevioribus’ refers to

Drege 8288. N. E. Brown in a note on Thom s.n.

wrote that the description of M. densiflora was based

on this specimen, but Rolfe (1913) effectively lecto-

typified the name M. densiflora by attaching the

Thom specimen (lThom 732’) to that species and

referring Drege 8288 to M. physodes. This lectotypi-

fication should be followed.

This highly variable species has small flowers

(sepals 3-5 (-7) mm long) with a spur shorter than

the galea. It may be separated from the closely

related M. bracteata by the relatively short spur

which is triangular and constricted at the base.

This species has a peculiarly disjunct distribution,

with the vast bulk of the collections known from the

western Cape Province, and a few rare collections

from the Tsitsikamma coastline between Humans-

dorp and Knysna. As discussed above, there is some

morphological differentiation between these two cen-

tres.

In the western Cape, this species occupies a wide

range of habitats. It has been recorded from sand

near sea level, at Olifantsbos in the Cape Point

Nature Reserve, from the summit of Table Moun-

tain, and from rock ledges in the Wemmershoek

Mountains. The altitude range of the species is from

sea level to over 1 600 m. In general, it appears to be

restricted to slightly damp areas, and has often been

collected after fires. It does not appear to be common

in any of these habitats. The macro-climatic condi-

tions vary from all year rainfall in the Knysna area,

to summer drought conditions on the Cape Flats. In

H. P. LINDER

349

the mountains of the western Cape, it often occurs in

the south-eastern cloud zone.

Monadenia multiflora has generally been recogniz-

ed as a distinct species. Sonder (1847) clearly distin-

guished it from M. bracteata by the larger flowers

and shorter spur. However, M. auriculata has never

been satisfactorily separated from M. multiflora.

Bolus (1896) stated ‘It comes near to D. multiflora,

tnihi, in its flowers, but the spike is always slenderer,

and the inflorescence usually less dense’. Schlechter

(1901) claimed that D. auriculata may easily be

recognized by ‘the more slender habit, smaller

flowers, more helm-shaped dorsal sepal and the ros-

tellum’. Rolfe (1913) used the relative lengths of the

spike and the leafy shoot to separate the two putative

taxa. Lewis (1950) separated the taxa on the width of

the spike and the shape of the spur. A careful

analysis of all the available material, as well as field

studies in the various localities where the types were

collected, showed that the two taxa could not be

maintained as distinct from each other and M. densi-

flora.

Monadenia densiflora is rather variable with re-

spect to flower size, and the shapes of the petals and

the spur. Although the inflorescence is generally

longer than the leafy shoot, the ratio is often re-

versed. Spur length varies from 1 mm (Barker 3893)

to 4 mm (Page in BOL 16232) and spur shape from a

rounded apex to an emarginate apex. There is some

geographical pattern to flower size. Collections from

the Tsitsikamma coastal flats have the smallest

flowers (lateral sepal c. 3 mm long, Linder 1903),

whereas collections from the summit of Table Moun-

tain on the Cape Peninsula show the largest flowers

(Lewis 1107). One of the syntypes of D. auriculata

(Schlechter 5958) is representative of the small-

flowered Tsitsikamma form, whereas the lectotype of

M. densiflora is typical of the large-flowered form

from Table Mountain. The type of M. multiflora,

Ecklon & Zeyher s.n., is from the more common in-

termediate size range (lateral sepals c. 3,4 mm long).

Sect. Kamiesbergenses Linder, sect, nov., calcari

sepalo dorsalo aequanti, sepalis c. 7 mm longis, in-

florescentia laxa dignoscenda.

Spur about as long as the dorsal sepal; flowers

medium sized, sepals about 7 mm long; inflorescence

lax.

Type species: Monadenia macrostachya Lindl.

The sole species in this section, M. macrostachya,

could be placed into any of the other three sections,

as it is closely related to M. densiflora, M. reticulata

and M. brevicornis. However, in the cladistic analy-

sis followed here, it would appear to be best placed in

a position linking the other groups, and possibly

ancestral to them. If the other three groups were to

be recognized as being distinct sections, this species

would also have to be placed into a section by itself.

This analysis would imply that the present distribu-

tion of M. macrostachya is only a relic of a more ex-

tensive older distribution. If this is correct, it is cer-

tainly interesting that the relic should be in an outlier

of the Cape Flora, where the species richness is rela-

tively low.

7. Monadenia macrostachya Lindt., Gen. Sp.

Orch. 357 (1838); Kraenzl., Orch. Gen. Sp. 1: 812

(1900), excl. syn.; Rolfe in FI. Cap . 5, 3: 189 (1913).

Type: Cape Province, Namaqualand, Rooiberg.

Drege 8289 (K, hoi o.!; G! ; K!).

Disc/ macrostachya (Lindl.) H. Bol. in J. Linn. Soc., Bot 25: 197

(1889); Schltr. in Bot. Jb. 31: 209 (1901), excl. syn.

Plant slender, 300 mm tall; leaves imbricate, the

lower 4-8 lorate, acute, erect, conduplicate, up to

130 mm long and 20 mm wide, remaining leaves

mostly sheathing, grading into the floral bracts; in-

florescence lax with several flowers, up to 200 mm

long; ovaries c. 10 mm long; bracts as tall as the

flowers, ovate, acuminate, probably dry. Dorsal

sepal erect, narrowly oblong, acute, 8 mm tall and

galea 1 ,5 mm deep; spur pendent from the base of the

galea, somewhat inflated, obtuse to retuse, 6-8

mm long and about 2 mm in diameter, straight;

lateral sepals reflexed, oblong, obtuse, c. 7 mm long;

petals obliquely narrowly ovate, very acute, 6 mm

long, erect; lip pendent, lorate, obtuse, 6 mm long;

anther semipendent, 1,5 mm long; rostellum erect,

tall; stigma on a 2 mm tall pedicel.

Diagnostic features. Dorsal sepal 8 mm long, spur

subclavate, retuse, as long as the dorsal sepal, bracts

dry, basal leaves lorate, erect, acute, cauline leaves

mostly sheathing.

Flowering time: ? September.

This species is only known from two collections:

the type collection probably made in about 1835 by

Drege; and from a plant past flowering, collected by

me in 1977 (Fig. 12). The above description was

prepared from the meagre material available, and

upon receipt of further material may well be found to

be wanting.

Although M. macrostachya is clearly distinct, it is

related to several of the major groups in Monadenia,

and may be regarded phylogenetically as a relic that

links these three groups together (Fig. 2).

Following Bolus (1911) and Schlechter (1901) in

most herbaria Monadenia macrostachya and M.

reticulata are regarded as synonymous. In this treat-

ment M. reticulata is regarded as a distinct species.

350

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONA DEN I A

A single small popolution of this species was

located by the author in the Kamiesberg on the

Rooiberg. The plants occurred in dry stony soil near

a stream, and had been heavily grazed by sheep.

There might well be more populations, as the moun-

tains have not been properly explored.

Sect. Tenuicornes Linder, sect, nov., calcari tenui,

sepalis dorsalibus longiore, sepalis 6-10 mm longis,

inflorescentia laxa dignoscenda.

Spur slender, longer than the dorsal sepal; flowers

medium sized, sepal 6-10 mm long; inflorescence

lax.

Type species: Monadenia ophrydea Lindl.

This rather distinct section includes six species: M.

reticulata, which is closely related to M. macros ta-

chya, in which it has been included by some botan-

ists; M. rufescens, with a relatively short spur and an

elongated viscidium; M. bolusiana and M. comosa,

with green petals and the stigma on a tall stipe; and

M. ophrydea and M. atrorubens in which the plants

are suffused with beetroot-red. The section may be

distinguished from sect. Densiflorae and sect.

Kamiesbergenses by the spur always being longer

than the dorsal sepal, and from sect. Monadenia by

the slender spur.

Both the M. comosa and the M. ophrydea groups

may be derived from M. reticulata. Consequently,

this species is treated as being ancestral in the section.

Other evidence in favour of the ancestral position of

this species is its wide distribution and large range of

habitats. It is also closely related to M. macrosta-

chya.

M. rufescens is clearly related to M. reticulata by

the slender spur and the shape of the petals, but may

readily be distinguished by the three-dimensional

shape of the flowers, the dense tissue of the bracts

which do not show the venation, and the elongated

viscidium. It appears as if this species may have evol-

ved a peculiar pollination syndrome. Although M.

rufescens is essentially restricted to the western Cape

Province, it has been collected from a wide range of

habitats.

Within the M. comosa complex three groups might

conveniently be recognized. The differences between

these groups are given in Table 2. The characters

listed in the table are approximations, as several are

rather difficult to quantify e.g. variation in leaf

shape and the habitats. No difference in the floral

structures among the groups has been detected.

The three groups behave like ecotypes. Group one

occurs almost invariably in rock crevices, usually in

shady places. Group three generally grows in full

sunlight, usually on the upper ridges of mountains.

Various populations occur in a wide range of

habitats: deep sand ( Ebersohn 151), footpaths

( Linder 1672) and in rock crevices and ledges ( Linder

1748). These two groups occur in the Langeberg and

in the mountains between Caledon and Clanwilliam.

Group two occurs on the summit of the Swartberg

near Oudtshoorn, at a high altitude, in a zone receiv-

ing the bulk of the rain in the summer months. It

grows in a habitat similar to that of Group three.

Groups one and three behave like distinct species.

The ranges of these two groups are sympatric but,

although populations of the two groups occur on the

same mountain, the plants occupy different habitats,

and interbreeding is prevented by allochronic flower-

ing. Morphologically, individuals can be placed into

these two groups on the basis of the leaf structure

and the colour of the vegetative parts of the plants,

both in the fresh and dried state.

Group two does not occur sympatrically with

groups two and three, and its biological role can

therefore not be determined on the basis of its in-

teraction with closely related populations. Mor-

phologically, the populations are quite variable, with

the majority of individuals rather similar to group

one, but with the variation ranging almost to that

typical of group three.

Taxonomically the M. comosa complex may be

treated in several ways:

1 . Asa variable species with three infraspecific taxa.

However, groups one and three behave as distinct

species.

2. As two distinct species, with group two recogniz-

ed at infraspecific level in either of the two species.

However, group two is intermediate between the two

species, and such a treatment implies that either

group two arose by secondary convergent evolution,

or that group one arose by secondary divergence

from group two.

3. As two distinct species, with group two recog-

nized as an ancient hybrid complex between the two

species. This would account for the extraordinary

range of variation in the populations assigned to

group two. Hybridization on the Swartberg could be

accounted for by the delay of flowering of group

one, leading to synchronous flowering. This would

lead to the present situation, where parents, hybrids

and the results of introgression can no longer be

separated. This interpretation of group two would

also strengthen the notion that groups one and three

represent biological species.

TABLE 2. — Groups recognized within M. comosa s.l.

H. P. LINDER

351

The third treatment is followed in this study.

Although the M. comosa complex is closely related

to M. reticulata, it may be separated from this species

by the higher stigma and somewhat longer spur. Both

species of the group are best regarded as specialized

derivatives from M. reticulata.

Monadenia ophrydea and M. atrorubens are also

very closely allied and have frequently been con-

fused. There are numerous small differences between

the species, most of which are rather difficult to

observe on dried material and have at least some

exceptions. Only the rostellum height/stigma height

ratio appears to be a constant differential character.

The differences between the two species are listed in

Table 3.

TABLE 3. — Morphological differences between M. alrorubens and

M. ophrvdea (note that few collections have all these characters)

The morphological differences between the species

are correlated with ecogeographical differences. M.

atrorubens occurs on the west coast of the western

Cape Province, and extends southwards to Houw

Hoek in the Caledon District, whereas M. ophrydea

extends from Table Mountain on the Cape Peninsula

eastwards to Humansdorp. M. atrorubens grows

mainly on deep sand, in areas in which there is almost

no precipitation in the summer months, whereas M.

ophrydea is generally found on mountain slopes in

shallower stony sand, in areas where there is at least

some rainfall in the summer months. Flowering in M.

atrorubens peaks in September, while in M.

ophrydea it peaks in October.

This group appears to be derived from M.

reticulata, and is morphologically very similar to it.

The beetroot-red colour of the plants appear to be a

single-gene factor, as several individuals have been

collected in which the red coloration is absent. Such

individuals are difficult to distinguish from M.

reticulata when they are dried.

8. Monadenia reticulata (H. Bol . ) Dur. &

Schinz, Consp. FI. Afr. 5,3: 111 (1894); Kraenzl.,

Orch. Gen. Sp. 1: 816 (1900); Rolfe in FI. Cap 5,3:

193 (1913). Type: Cape Province, Cape Peninsula,

Constantiaberg, Bodkin in BOL 4988 (BOL, holo.!;

K!).

Disa reticulata H. Bol. in J. Linn. Soc., Bot. 22: 75 (1885);

Schltr. in Bot. Jb. 31: 209 (1901), excl. syn.

leones: H. Bol., leones Orch. Austro-Afr. 2: t. 88

(1911); 3: t. 41 (1913), as Disa macrostachya.

Plants 80-400 mm tall; leaves generally linear-

lanceolate, rarely narrowly oblong, acute, imbricate,

erect, the lower the longest, up to 150 mm long,

gradually grading into the upper leaves that are

almost completely sheathing, grading into the floral

bracts; inflorescence cylindrical, generally dense with

numerous flowers, 30-150 mm long; ovaries c. 10

mm long; bracts reaching to the top of the flowers,

narrowly ovate, acuminate, the reticulate venation

clearly visible on dried material. Flowers lime-green,

occasionally with some maroon tinting or mottling

on the petals or sepals, scent strong, soapy; dorsal

sepal shallowly galeate, oblong, obtuse, somewhat

curved forwards, 7-8 mm long; spur pendent from

the base of the galea, slender, acute, longer than the

galea, 10-20 mm long and c. 1 mm in diameter;

lateral sepals oblong, rounded, 6-7 mm long; petals

obliquely narrowly ovate to rarely narrowly oblong,

apically obliquely retuse to rounded, the broad base

of the petals enclosing the anther, the apex twisted to

face forwards, 5-6 mm long; lip pendent, lorate to

rarely narrowly oblanceolate, acute to rarely round-

ed, 4-6,5 mm long; anther semipendent, 1,5 mm

long; rostellum simple, 1 mm tall; stigma simple, not

as tall as the rostellum.

Diagnostic features. Flowers with the lateral sepals

6-7 mm long; spur 10-20 mm long, slender, acute;

lip lorate; bracts with an obvious reticulate venation;

leaves generally linear-lanceolate, erect, about 8.

Flowering time: November and December.

Monadenia reticulata occurs occasionally in seas-

onally damp places in the western Cape Province,

often after fire, between sea level and 1 700 m. It also

extends eastwards along the mountains to George

(Fig. 13).

Cape. — 3318 (Cape Town): Table Mountain, 700 m (-CD),

Dec. 1879, Bolus 4542 (BOL). 3320 (Montagu): Groot-

vadersbosch, Heidelberg (-DD), Dec. 1958, Barker 8838 (NBG).

3322 (Oudtshoorn): Montagu Pass, 600 m (-CD), Nov. 1894,

Schlechter 5791 (BOL; Z).

This species has for some time been confused with M.

macrostachya (Schlechter, 1901; Bolus, 1913), but it

may be separated by the spur which is longer than the

dorsal sepal, the laxer inflorescence and the narrower

leaves. From the rest of the Monadenia species with

slender spurs it is distinct by the papery bracts, which

clearly display the venation in the dried state, the

greenish Bowers and the erect linear-lanceolate

leaves.

The collections all appear to be from sandy soil

derived from Table Mountain Sandstone. Most of

the collections are from areas where at least some

precipitation occurs in the summer months, even if

only from cloud derived from south-east winds.

9. Monadenia comosa Reichb. f. in I.innaea 20:

687 (1847); Kraenzl., Orch. Gen. Sp. 1: 813 (1900);

Rolfe in FI. Cap. 5, 3: 194 (1913). Type: Cape of

Good Hope, Gueinzius s.n. (W, holo.!).

Disa comosa (Reichb. f.) Schltr. in Bot. Jb. 31: 206 (1901); H.

Bol., leones Orch. Austro-Afr. 3: t. 43 (1913).

Monadenia rufescens Lindl., Gen. Sp. Orch. 356 (1838), non

Thunb. Disa affinis N. E. Br. in Gdnrs’ Chron. 24: 402 (1835),

nom. nov. Lectotype: Cape Province, near Genadendal, Drege

1252 (K, holo.!).

Icon: H. Bol., leones Orch. Austro-Afr. 3: t. 43

(1913), as Disa comosa.

Plants erect or subflexuose, 80-300 (-600 mm tall;

tubers slender cylindrical, up to 50 mm long, basal

sheaths hyaline, obtuse, 1-2; leaves variable the

basal 2 (-3) elliptic to rarely narrowly elliptic, ob-

352

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

Fig 13. — Distribution of

Monadenia reticulata.

tuse, semi-erect, 30-70-120 (-170) mm long, the re-

maining leaves closely sheathing the stem, acute, lax

to imbricate, similar to the floral bracts; in-

florescence slender, usually lax, 40-150 mm long,

with 1-20 flowers; ovaries erect, 15-20 mm long;

bracts slightly shorter than the ovaries, narrowly

ovate to rarely ovate, acute to acuminate, venation

reticulate and visible. Flowers lime-green, occa-

sionally tinted red; dorsal sepal erect to somewhat

curved forwards, shallowly galeate, oblong, obtuse

to rounded, 9-11 mm tall and c. 1 mm deep; spur

pendent from the base of the galea, slender cylin-

drical, acute, adpressed to the ovary, 17-24 mm long

and c. 1,5 mm in diameter; lateral sepals patent to

reflexed, obliquely oblong-ovate, obtuse, 6-7 mm

long; petals obliquely ovate-narrowly oblong, obli-

quely retuse to emarginate, 6-8 mm long, the broad

basal part enclosing the anther and the narrower

apical part curved forwards; lip elliptic, decurved,

obtuse to acute, subfleshy, 6-8 mm long; anther

semipendent, 2 mm long; rostellum simple, 2 mm

tall; stigma as tall as the rostellum. Fig. 14.

Diagnostic features. Spur slender, acute, 17-24 mm

long: bracts with the venation visible; stigma as tall

as the rostellum; petals lime-green; basal two leaves

spreading, the remaining leves closely sheathing the

stem.

Flowering time: (September-) October (-November).

A widespread and often common green flowered

orchid in rock crevices and on ledges (Fig. 15).

Cape. — 3319 (Worcester): Slab Peak, Michell’s Pass (-BA),

Oct. 1941, Compton 11955 (NBG). 3318 (Cape Town): rock crev-

ices on Table Mountain (-CD), Oct 1883, MacOwan & Bolus 170

(BM: BOL; G; K; P; W; ZT). 3419 (Caledon): Kanonkop, above

Greyton, 1500 m (-BA), Jan. 1964), Oliver s.n. (BOL). 3322

(Oudtshoorn): summit of Swartbergpas (-AC), Dec. 1977, Linder

1 740 (BOL).

Reichenbach described M. comosa from a Guein-

zius collection. In the Reichenbach collection at

Vienna there is a single specimen of M. comosa col-

lected by Gueinzius, but it is labelled ‘Monadenia

rufescens’. However, this annotation is not in

Reichenbach’s hand, and M. rufescens is the name

that Lindley ( 1 838) applied to this taxon. This sheet is

likely to be the holotype.

Lindley mistook the identity of Thunberg’s Satyr-

ium rufescens, and described it as being: ‘ foliis radi-

calibus binis oblongis, caule vaginato’, a character

unique to M. comosa. Of the two collections cited by

Lindley, one, Drege 1252, fits his description, and

the other is referable to Monadenia rufescens

(Thunb.) Lindl. in the correct sense.

Reichenbach proposed the name M. comosa for

the species, but he did not refer to Lindley’s work.

The name is therefore not a nomen novum, and the

type is the Gueinzius collection. In 1885 N. E. Brown

proposed Disa affinis as a nomen novum for M.

rufescens Lindl. non Thunb., apparently unaware of

the identity of M. comosa Reichb. f. This confusion

was resolved by Durand & Schinz (1894).

Monadenia comosa belongs to the group of species

in Monadenia with long slender spurs. It may be

separated from the other species in this group by the

two or three large spreading narrowly ovate basal

leaves, the dry papery floral bracts and the pale green

Bowers.

Fig. 14. — Monadenia comosa, from Esterhuysen 16200, x 0, 5.

H. P. LINDER

353

Fig. 15. — Distribution of

Monadenia comosa

This species is widespread in the mountains of the

western and southern Cape Province, where it occurs

almost exclusively in crevices and on ledges on rocks

and cliffs, often in half-shade. A few collections are

from talus slopes e.g. Linder 1601. In general,

populations consist of numerous individuals, often

forming almost monospecific associations on rock

ledges. On the summit of the Swartberg at Oudts-

hoorn, plants occur in gravelly soil ( Linder 1740).

The altitude range of the species is from near sea

level to 1 600 m, although the majority of the collec-

tions are from below 1 000 m. Rainfall over the dis-

tribution range varies considerably from an all-the-

year rainfall in the Knysna area, to summer drought

conditions in the western Cape area.

10. Monadenia bolusiana ( Schltr .) Rolfe in FI.

Cap. 5,3: 194 (1913). Type: Cape Province, Cape Pe-

ninsula, Table Mountain, Bodkin in BOL 4903

(BOL, lecto.!; K!).

Disa bolusiana Schltr. in Bot. Jb. 24: 426 (1898); Schltr. in Bot.

Jb. 31: 207 (1901).

Icon: H. Bol., leones Orch. Austro-Afr. 2: t. 90

(1911).

Plants erect, ( 1 00—) 200-300 mm tall; basal

sheaths hyaline, obtuse; leaves imbricate, the lower

2-5 narrowly oblong, acute, semi-erect, flat, the

largest 50 (-70) mm long, grading gradually into the

sheathing, narrowly ovate, acute upper leaves; in-

florescence subimbricate, occasionally secund,

30-130 mm long and with 2-25 flowers; ovaries

15-20 mm long, erect; bracts about as long as the

ovaries, narrowly ovate, acute, the reticulate vena-

tion usually visible. Flowers lime-green, occasionally

tinted red; dorsal sepal erect, curved forwards,

shallowly galeate, oblong, obtuse to rounded, 8-11

mm long and c. 1 mm deep; spur pendent from the

base of the galea, slender, acute, 16-22 mm long and

about 1 mm in diameter; lateral sepals reflexed,

oblong, rounded, 6-8 mm long; petals obliquely

ovate-narrowly oblong, obliquely emarginate, 6-8

mm long, the broad basal part enclosing the anther,

the narrower apical part erect; lip patent or recurved,

elliptic, obtuse to acute, subfleshy, 5-7 mm long; an-

ther semipendent, 1,5-2 mm long; rostellum erect,

1,5 mm tall; stigma as tall as rostellum. Fig. 16.

Diagnostic features. Spur slender, acute, 16-22 mm

long; petals lime-green, stigma as tall as the

rostellum; lower cauline leaves larger than the upper,

but gradually grading into them; plants flowering in

December and January.

Flowering time: (October-) December-January.

Fig. i6.— Monadenia bolusiana. 1, whole plant, from Esterhuysen

9780, x 0,5. 2, flower in front view, x 3. 3, flower in side view,

x2. 4, dissection of flower; a, rostellum; b, petal; c, anther;

d, stigma; e, viscidium, x 5. 2-4 from Linder 1748.

354

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

Monadenia bolusiana is a small dark green to even

purplish herb that occurs occasionally to frequently

in exposed gravelly soil on the summit of the moun-

tains in the western Cape Province (Fig. 17).

Cape. — 3419 (Caledon). Betty’s Bay, Harold Porter Nature

Reserve, east slope of Voorberg, 200 m (-BD), Nov. 1970, Eber-

sohn 151 (NBG). 3320 (Montague): Strawberry Hill (-DD), Dec.

1957, Stokoe in CH 9904 (NBG).

Schlechter (1898) cites three collections after his

description of Diso bolusiana: Zeyher 1 570 and Bolus

4903 from Maclear’s Beacon on Table Mountain and

Schlechter s.n. from the Langeberg above Zuurbraak

near Swellendam. It is likely that these collections

were in Schlechter’s herbarium in Berlin that was de-

stroyed in World War II. In BOL there is an excellent

duplicate of Bolus 4903, and that is selected here as

the lectotype. However, there are two collections

under that number. The first one is a Bodkin collec-

tion, made in January 1883, while the second is a

Scully collection, made in January 1884. The illustra-

tion for the leones was made from the second collec-

tion, while Schlechter specified the first collection as

the type (‘Bolus n. 4903, bluehend im Januar 1883’).

This species is closely allied to Monadenia coinosa

by virtue of the long slender spur, green flowers and

papery bracts. It may be separated from this species

by the numerous cauline leaves, the purplish colour

of the plants and the flowering season (November to

January).

Monadenia bolusiana occurs mostly in the full sun-

light on the summit ridges of mountains, in shallow

gravelly soils, often in footpaths and other disturbed

areas. At present there are rather few herbarium

records of this species, but this might be due to

undercollecting in these more or less inaccessible

localities. Fieldwork has revealed it to occur at least

occasionally, and possibly even frequently and wide-

spread, in the summit zones of the ridges. The alti-

tudinal range of the species is from 200 m in the Bet-

ty’s Bay area to over 1 200 m in the Langeberg and

Hottentots Holland Mountains. Over the whole

range there is a certain amount of moisture available

in the summer months from condensation from the

‘south-easter’ clouds.

1 1. Monadenia atrorubens ( Schltr .) Rolfe in FI.

Cap. 5, 3: 196 (1913). Type: Cape Province, Clan-

william, Zwartboschkraal, Schlechter 5167 (B,

holo. t ; BOL!; Z!).

Disa atrorubens Schltr. in Bot. Jb. 24: 427 (1898); Schltr. in Bot.

Jb. 31: 205 (1901).

Monadenia ophrydea Lindl. sensu Kraenzl., Orch. Gen. Sp. 1:

817 (1900), pro parte.

Plants 100—400 mm tall, suffused with beetroot-

red; tubers 15-20 mm long, acute; basal sheaths

hyaline, acute or obtuse, about 20 mm long, occa-

sionally much longer; cauline leaves green, narrowly

lanceolate to lanceolate, acute, the longest at the

base, up to 90 mm long, semi-erect and sheathing at

the base, upper leaves grading into the floral bracts,

imbricate to subimbricate, acute; inflorescence

slender, lax, (40-) 80-150 (-250) mm long and with

(3-) 10-180 (-250) flowers; bracts green, lanceolate

to narrowly ovate, acute, reaching the flowers or

overtopping them; ovary slender, 15-20 mm long.

Flowers horizontal, sepals and spur beetroot-red, lip

and petals almost black; dorsal sepal narrowly

oblong, obtuse, shallowly concave, 7-10 mm long,

3—4 mm wide and c. 1 ,5 mm deep; spur pendent from

the base of the sepal, cylindrical, tapering to an acute

apex, parallel to the ovary, 15-30 mm long; lateral

sepals reflexed at anthesis, narrowly oblong, obtuse,

6-8 mm long; petals erect inside the galea, obliquely

ovate, subacute, often shallow bidentate, fleshy

apically and along the margins with papillae on the

inside margin surface, 5-7 mm long and 3-5 mm

wide; lip oblong to narrowly oblong, rounded,

fleshy, horizontal basally but soon decurved, 5-7

(-8) mm long and 2-4 mm wide; rostellum erect, c.

1,5 mm tall with short lateral lobes flanking the

caudicles; anther semipendent, 1-1,5 mm long, the

single viscidium a large disc; stigma c. 2 mm in

diameter, on a pedicel as tall as the rostellum and

angled forwards. Fig. 18

Diagnostic features. Plants suffused with beetroot-

red in fresh state, reddish brown when dry, spur

slender, up to 30 mm long, lip oblong to narrowly

oblong, lateral sepals purple to beetroot-red, re-

flexed.

Flowering time: August-October.

The plants, suffused with beetroot-red, form small

populations in sandy areas, where they may be seen

in flower the first year after a fire. (Fig. 19).

Cape. -31 19 (Calvinia): between Oorlogs Kloof and

Papkuilsfontein. (-AC), Sept. 1939, Leipoldt 3236 (BOL). 3318

(Cape Town): Pella Mission, common on sandflats after fire

(-DA), Sept. 1977, Linder //40(BOL).

This species appears to be restricted to sandy areas

in the western Cape Province between Houw Hoek

and Calvinia, from near sea level to about 1 000 m in

the northern end of the distribution area. Two popu-

lations were studied in the field ( Linder 1 140& 1129).

Both flowered the first year after the fire, and occur-

red on slightly seasonally damp deep sand. The popu-

lations were clearly defined, and within the area the

species occurred frequently. Among the burnt twigs

and branches the purplish red plants were scarcely

visible. Despite extensive field work in the western

Cape, no plants were found in unburnt vegetation.

Over the whole distribution range the summers are

dry, and in the winter 500-800 mm rain falls.

The differences between this species and M.

ophrydea are discussed above.

12. Monadenia ophrydea Lindt., Gen. Sp. Orch.

358 (1838); Kraenzl. Orch. Gen. Sp. 1: 817 (1900);

Rolfe in FI. Cap. 5, 3: 195 (1913). Type: Cape Prov-

ince, Paarl, Drakenstein Mountains, Drege 8290 (K,

holo.!).

H. P. LINDER

355

Fig. 18. — Monadenia ophrydea (1-3) and M. alrorubens (4-6). 1 ,

plant of M. ophrydea, from Esterhuysen 10573, x0, 5. 2,

front view of flower, x3. 3, dissection of flower, x5. (2-3

from Linder 1596.) 4, front view of flower of M. alrorubens,

x 3. 5, side view of flower, x 3. 6, dissection of flower, x 5:

a, rostellum; b, petal; c, anther; d, stigma. (4-6 from Linder

1129.)

Disa ophrydea (Lindl.) H. Bol. in Trans. S. Afr. phil. Soc. 5:

142 (1888); Schltr. in Bot. Jb. 31: 204 (1901).

Brownleea penlheriana Kraenzl. ex Zahlbr. in Ann. Nat. Hof-

mus. Wien 20: 6 (1905). Type: Cape Province, Montagu Pass,

Penther 189 (W, holo.!).

Plants slender, up to 400 mm tall, suffused with

beetroot-red tubers 10-20 mm long; basal sheaths

1-2, hyaline, obtuse or acute, 1 (-3) mm long; leaves

green, narrowly lanceolate, acute, rarely obtuse, the

largest near the base of the stem, up to 100 mm long

and 20 mm wide, usually imbricate; inflorescence

lax, up to 300 mm long with (1-) 5-15 (-30) flowers;

bracts purplish-green, ovate to shortly acuminate,

20-40 mm long and 15-25 mm wide, generally as tall

as the flowers; ovaries 20-25 mm long. Flowers

horizontal, purple-red, the lateral sepals paler and

often white; dorsal sepal oblong-obovate, shallowly

galeate, rounded, 9-11 mm long, 5-7 mm wide and

1 -2 mm deep; spur pendent from the base of the dor-

sal sepal, slender, tapering to an acute apex, slightly

curved towards the ovary, 20-24 mm long; lateral

sepals spreading, narrowly ovate, acute, rarely some-

what reflexed, much paler than the rest of the flower,

8-10 mm long; petals erect inside the galea, very

obliquely ovate, acute with a shallowly emarginate

apex, falcate in side view, somewhat curved around

the rear of the anther, the apex and anterior margins

fleshy; lip narrowly elliptical to almost lorate,

subacute to acute, fleshy, pendent, 8-10 mm long;

rostellum simple, flanking the caudicles, c. 3 mm tall;

anther semipendent, c. 4 mm long, pollen-masses

almost globular, viscidium almost globular, 0,8 mm

in diameter; stigma equally tripulvinate, c. 1, 5 mm

in diameter, about Vi as tall as the rostellum. Fig. 18.

Diagnostic features. Plant suffused with a beetroot-

red when fresh, reddish brown when dry, spur

slender, up to 24 mm long, lateral sepals reddish and

white, spreading, lip narrowly elliptical to lorate.

Flowering time: October and November.

A slender herb, suffused with beetroot-red, occur-

ring usually in extensive populations in recently burnt

veld on mountain sides, in damp conditions, in the

western and southern Cape Province (Fig. 19).

Fig. 19. — Distribution of

Monadenia alrorubens

(solid circles) and

Monadenia ophrydea

(open circles).

356

TAXONOMIC STUDIES IN THE D1SINAE. V. A REVISION OF THE GENUS MONADENIA

Cape. -3318 (Cape Town): damp grassland on Table Mountain,

700 m (-CD), Oct. 1879, Bolus 4538 (BOL). 3323 (Willowmore):

Outeniqua Mountains near Joubertina (-DD), Dec. 1946,

Esterhuysen in BOL 23635 (BOL).

Several populations of this distinct species have

been studied over the length of the distribution

range. Generally the populations are extensive with

numerous individuals, occurring on damp peaty

mountain slopes facing the coast and consequently

the rain-bearing winds. This species, as in M.

atrorubens, appears to flower only after fire, when

the beetroot-red plant colour is cryptic in the burnt-

out vegetation. The altitude range is from near sea

level in the Betty’s Bay (Caledon) area to more com-

monly 300 - 1 000 m in the Langeberg and eastwards

to Humansdorp. Over the whole distribution range,

the species receives some precipitation throughout

the year: in the east from the all-the-year rains and in

the west from condensation or occasionally rain from

the ‘south-easter’ clouds.

For the differences from M. atrorubens, see above.

13. Monadenia rufescens (Thunb.) Lindl., Gen.

Sp. Orch. 356 (1838); Kraenzl., Orch. Gen. Sp. 1:

818 (1900). Type: Cape of Good Hope, Thunberg in

herb. Thunberg 21456 (UPS; holo.!; W!).

Satyrium rufescens Thunb., Prod. 5 (1794). Disa rufescens

(Thunb.) Swartz in Vet. Acad. Handl. 21: 210 (1800); Schltr. in

Bot. Jb. 3 1 : 209 ( 1901 ). Monadenia lancifolia Sond. in Linnaea 19:

100 (1847), nom. nov.

Monadenia macrocera Lindl., Gen. Sp. Orch. 358 (1838); Rolfe

in FI. Cap. 5, 3: 193 (.1913). Type: Cape of Good Hope, Thom s.n.

(K, holo.!).

Monadenia leptostachya Sond. in Linnaea 19: 101 (1847). Type:

Cape Province, Cape Peninsula near Wynberg, Ecklon & Zeyher

s.n. (S, holo.!; P!; W!).

Icon: H. Bol., leones Orch. Austro-Afr. 2: t. 89

(1911).

Plants erect, 140-235-400 mm tall; tubers testicu-

lar, 20 mm in diameter; base of the stem often with

the remnants of old leaf fibres; basal sheaths hyaline,

grading into the leaves; leaves imbricate, the largest

leaves near the base of the stem, narrowly lanceolate,

to linear, acute, 50-70 mm long, conduplicate,

gradually smaller and more sheathing towards the

apex of the stem, grading into the floral bracts;

inflorescence a slender spike, 30-150 mm long and

15 mm in diameter, with 2-25 flowers; ovaries 10-15

mm long; bracts usually slightly longer than the

ovaries, narrowly ovate, acuminate, dark green,

imbricate and partially obscuring the flowers.

Flowers lime-green with dark purple petals and lip;

dorsal sepal angled forwards, shallowly galeate,

oblong, obtuse to rounded, 9—11 mm long and c. 1

mm deep; spur pendent from the base of the galea,

slender cylindrical, acute, 10-16 mm long; lateral

sepals erect or patent, oblong, obtuse, 6-8 mm long;

petals obliquely narrowly oblong, obliquely truncate,

wider at the base, subfleshy, concave, 6-7 mm long,

the broad basal part enclosing the anther, the nar-

rowly oblong apical part forming a tube with the dor-

sal sepal; lip decurved, narrowly oblong, obtuse, 6-7

mm long; anther semipendent, 1,5 mm long, visci-

dium elongated backwards with a longitudinal dorsal

groove; rostellum c. 1 mm tall; stigma pedicellate,

not as tall as the rostellum. Fig. 20.

Diagnostic features. Flowers with the lateral sepals

6-8 mm long, spur slender acute, 10-16 mm long;

floral bracts prominent, green; leaves gradually grad-

ing from large at the base to sheathing at the apex;

stigma lower than the rostellum.

Flowering time: (September-) October (-November).

Monadenia rufescens is a slender herb that occurs

occasionally in seasonally damp localities in the west-

ern Cape Province, between sea level and 1 000 m

(Fig. 21).

Cape. -3318 (Cape Town): Lower Plateau of Table Mountain,

750 mm (-CD), Oct. 1884, Bolus 4969 (BOL). 3418 (Simonstown):

Vlakkenberg vlei, 600 m (-AB), Oct. 1947, Compton 20180

(NBG). 3419 (Caledon): flats east of Viljoenspass (-AA), Sept.

1 949, Stokoe in SAM 61250 (SAM).

The nomenclatural history of this species is com-

plex. The type of Satyrium rufescens, although not in

a good condition, is clearly recognizable. However,

Lindley (1838) misidentified Satyrium rufescens

Thunb. (‘I know no plant to which the definitions of

Swartz and Thunberg exactly apply’). His description

of Monadenia rufescens (Thunb.) Lindl. applies to

Drege 1252, and must be referred to M. comosa

Reichb. f. However, the combination Monadenia

rufescens (Thunb.) Lindl. is valid, and has to be used

for M. rufescens sensu Thunb. Ironically, the other

Fig. 20. — Monadenia rufescens. 1, whole plant, from Salter 8479,

x0,5. 2, flower in angled front view, x3. 3, flower in side

view with the lateral sepal removed, x 3. 4, column, x 10: a,

rostellum; c, anther; d, stigma; e, viscidium. (2-4 from

Linder 1610.)

H. P. LINDER

357

collection cited by Lindley is clearly Satyrium

rufescens Thunb. Sonder (1847) noted that Mona-

denia rufescens Lindl. is not the same as Disci rufes-

cens (Thunb.) Swartz. He upheld the former, and

proposed Monadenia lancifolia as a nomen novum

for the latter. Lindley (1838) had also received a

specimen of the true Satyrium rufescens Thunb., and

named it Monadenia macrocera (1838). This name

was used by Rolfe (1913), who followed the ‘Kew

Rule’. Sonder (1847) also proposed another name,

Monadenia leptostachya for the species which he

separated from M. lancifolia on several characters.

Examination of the types indicated that these names

all apply to the same taxon.

Monadenia rufescens is systematically relatively

isolated within the group of slender long-spurred spe-

cies of Monadenia. The dorsal sepal and petals are

angled forwards and are applied to each other to

form a tube leading over the rostellum into the spur.

The rostellum is higher than the stigma and the single

large viscidium is elongated backwards. Consequent-

ly it has a large surface exposed to the proboscis of an

insect probing into the spur. The viscidium also has a

longitudinal dorsal groove, which may also be related

to this pollination syndrome. Further the species may

be distinguished by the leathery bracts and the purple

petals and lip.

Although M. rufescens is relatively widespread in

the western Cape Province, it does not appear to be

common at any locality and plants frequently appear

to occur singly. The majority of collections are from

slightly damp localities in a variety of habitats, rang-

ing from dunes to rocky mountain sides at about

1 000 m. Although the distribution range covers an

area of summer drought, many of the sampled

localities must receive at least some moisture con-

densed from the ‘south-easter’ clouds.

This species is not well known, and more field data

could cast some light on its ecological requirements.

Sect. Monadenia

Spur longer or shorter than the dorsal sepal,

clavate; flowers relatively large, sepals 7-14 mm

long; inflorescence generally dense.

Type species: Monadenia brevicornis Lindl.

This section of three species is clearly defined by

the swollen or clavate spurs, the larger flow'ers and

the more robust plants. Only M. brevicornis ap-

proaches M. reticulata in some populations.

The three species replace each other geographical-

ly, with narrow overlap zones. M. physodes occurs

between the Kamiesberg in Namaqualand and the

Cape Peninsula. M. cernua extends from the Cape

Peninsula to Humansdorp, whereas M. brevicornis is

widespread in montane grasslands in southern

Africa, reaching as far south-west as Knysna.

Although the vast majority of the collections are

readily distinguished on morphological features,

there are some intermediate collections between M.

physodes and M. cernua and between M. brevicornis

and M. reticulata.

M. physodes and M. cernua are readily

distinguished by the relative spur length (Fig. 22).

However, some of the collections from the overlap

zone on the Cape Peninsula are rather difficult to

assign to one of the two species, whereas others do

not present any difficulty (i.e. Le Sueur in BOL

4973). One of the intermediate collections is the type

collection of M. inf lata Sond. An analysis of several

flowers from this collections indicated that it is best

placed under M. cernua.

16

15

14

13

l12

x

o 11

z

10

9

8

7

O

• •

o o

o o o

o o

8

10 11 12 13

LATERAL SEPAL LENGTH (mm)

FiG- 22. — Variation in the relative spur length in Monadenia

physodes and M. cernua. The solid circles represent M.

cernua and the open circles M. physodes. The half-solid

circles represent two collections from the Cape Peninsula that

are intermediate between the two taxa, one of which is the

type collection of Monadenia inflaia.

358

TAXONOMIC STUDIES IN THE D1SINAE. V. A REVISION OF THE GENUS MONADENIA

There are no intermediate forms in the overlap

zone between M. cernua and M. brevicornis — as

though the morphological differences between the

two taxa are accentuated in this area, with collections

of M. cernua with longer spurs than is usually found

in this species.

However, the southernmost populations of M.

brevicornis approach M. reticulata. Generally these

two taxa are separated by the wider and more robust

cauline leaves and more inflated spurs of M. brevi-

cornis. In collections from the Uniondale area (Ester-

huysen 10692) the leaves are narrower, more clus-

tered towards the base of the stem, and the spurs are

more slender. A possible explanation is hybridization

between these two taxa, with introgression into M.

brevicornis.

14. Monadenia physodes ( Swartz ) Reichb. f. in

Flora 66: 461 (1883); Rolfe in FI. Cap. 5, 3: 191

(1913). Type: Cape of Good Hope, Thunberg in herb

Thunberg 21455 (UPS, holo.!).

ID is a physodes Swartz in Vet. Acad. Handl. 21: 211 (1800);

Thunb., FI. Cap. 12 (1823); Lindl., Gen. Sp. Orch. 356 (1838);

Kraenzl., Orch. Gen. Sp. 1: 788 (1900).

D. cernua (Thunb.) Swartz, Schltr. in Bot. Jb. 31: 210 (1901),

pro parte.

Plants robust, 250-600 mm tall; tubers up to 40

mm long; basal sheaths hyaline, obtuse, 1-2; leaves

linear-lanceolate, acute, conduplicate, falcately

erect, the longest at the base, 140-200 mm long and

c. 20 mm wide, grading apically into the foral bracts;

inflorescence cylindrical, 50-300 mm long and 30-40

mm in diameter, flowers subimbricate; ovaries c. 15

mm long; bracts as long as the flowers, narrowly

ovate, acuminate. Flowers with lime-green sepals,

often mottled or suffused maroon, petals lime-green;

dorsal sepal shallowly galeate, oblong, obtuse erect,

9-11 mm tall and c. 1 mm deep; spur pendent from

the base of the galea, clavate, rounded, 7-9 mm long

and 3-6 mm wide, adpressed to the ovary; lateral

sepals reflexed, oblong, obtuse to rarely acute, 7-10

mm long; petals obliquely narrowly ovate-oblong,

bluntly acuminate or obliquely retuse, 7-9 mm long,

the broad basal part enclosing the anther and the nar-

rower apical part twisted to face out of the galea; lip

pendent, lorate to narrowly oblanceolate, rounded,

7-10 mm long; anther semipendent, 2 mm long;

rostellum lateral lobes large, flanking the anther, 2

mm tall, stigma on a 1 mm tall pedicel, horizontal,

lateral lobes much larger than the central lobe.

Diagnostic features. Flowers with the lateral sepals

7-10 mm long; spur clavate, rounded, 7-9 mm long,

rostellum up to 2 mm tall.

Flowering time: September-October.

This robust herb occurs in the western Cape Prov-

ince, mostly in swampy localities, but at least one

record is from a dry slope after a fire (Fig. 23).

Cape. — 3018 (Kamiesberg): Leliefontein, Little Namaqualand

(-AB), Oct. 1940, Leipoldl 3809 (BOL). 3318 (Cape Town): about

6 km north of Malmesbury (-BC), Oct. 1964, Rabinowitz in NBG

77143 (NBG); lower slopes of Lions Head near Sea Point, 80 m

(-CD), Sept. 1884, Le Sueur in BOL 4973 (BOL).

Thunberg (1794) regarded this species as synony-

mous with M. cernua, as he cites both in his proto-

logue. M. physodes was only separated by Swartz in

1800. In general, the distinction between the two taxa

has been recognized, except for Schlechter (1901),

who regarded M. physodes merely as a smaller form

of M. cernua.

This species is unique because of its strongly

clavate spur which is shorter than the dorsal sepal. It

is clearly allied to M. cernua, but may readily be

distinguished by the relatively short spur and the

peculiar rostellum with massive lateral lobes flanking

the anther.

I have not seen any populations of this peculiar

species in the wild, and collectors notes do not indi-

cate the habitat. Oliver ( Oliver 4761) reports that

near Ceres the species flowered in sandy stony flats

after a fire.

The altitude range of the species is from near sea

level at Sea Point, to over 1 000 m in some inland

localities. Many of the localities of herbarium collec-

tions are now disturbed, and the number of popula-

tions of the species must be much reduced from that

of a century ago. The distribution range of the

species falls in an area of summer drought and total

rainfall values of 500-800 mm p.a.

There are two records of an unpleasant mouse-like

scent.

Fig. 23. — Distribution of Monadenia

physodes (solid circles) Mona-

denia cernua (open circles).

H. P. LINDER

359

15. Monadenia cernua ( Thunb .) Dur & Schinz,

Consp. FI. Afr. 5: 111 (1894); Kraenzl., Orch. Gen.

Sp. 1: 815 (1900); Rolfe in FI. Cap. 5, 3: 192 (1913).

Type: Cape of Good Hope, Thunberg in herb. Thun-

berg 21431, 21432 (lecto.), 21433 (all UPS.!).

Satyrium cernuum Thunb., Prodr. PI. Cap. 5 (1794). Disa cer-

nua (Thunb.) Swartz in Vet. Acad. Handl. 21:211 (1800); Thunb.,

FI. Cap. 12 (1823); Lindl., Gen. Sp. Orch. 356 (1838); Schltr. in

Bot Jb. 31: 210 (1901), excl. M. physodes.

Disa prasinata Ker-Gawl. in Edwards’s bot. Reg. t. 210 (1817).

Monadenia prasinata (Ker-Gawl.) Lindl., Gen. Sp. Orch. 358

(1838). Type: Edwards’s bot. Reg. t. 210 (1817), iconotype.

Monadenia inflata Sond. in Linnaea 19: 102 ( 1847). Type: Cape

Province, Cape Peninsula near Wynberg, Ecklon & Zevher s.n. (S,

holo.l; BOL1; Kl; P!; SAM!; W!; Z!).

leones: Edwards’s bot. Reg. t. 210 (1817); H. Bol.,

leones Orch. Austro- Afr. 2: t. 91 (1911), as Disa cer-

nua.

Plants robust, 200-600 mm tall; basal sheaths hya-

line, obtuse, 1-2; leaves linear-lanceolate, acute,

conduplicate, imbricate, the largest at the base,

( 1 00—) 140-200 mm long and 15-20 mm wide, grad-

ing apically into the floral bracts; inflorescence

100-250 mm long and 30-40 mm in diameter,

flowers subimbricate; ovaries c. 15 mm long; bracts

as tall as the flowers, narrowly ovate, subacuminate.

Flowers with cream-green sepals, mottled maroon,

and lime-green petals and lip; dorsal sepal shallowly

galeate, slightly angled forwards, oblong, obtuse,

10-14 mm long and c. 1 mm deep; spur pendent

from the base of the galea, clavate, rounded, 11-17

mm long and 3-5 mm wide, adpressed to the ovary;

lateral sepals reflexed, narrowly oblong to oblong,

acute to obtuse, 9-13 mm long; petals subobliquely

narrowly ovate-oblong, bluntly acuminate to obli-

quely retuse, 7-10 mm long, the broad basal part

flanking the anther and the narrow apical part

twisted to face out of the galea; lip pendent lorate,

rounded, subfleshy, 8-12 mm long; anther subhori-

zontal, 3 mm long; rostellum with 2-3 mm tall lateral

lobes flanking the anterior part of the anther; stigma

on a 1 mm tall pedicel, slightly angled forwards, flat,

lateral lobes larger than the posterior lobe. Fig. 24.

Diagnostic features. Flowers with the lateral sepals

9-13 mm long; spur clavate, rounded, 11-17 mm

long; rostellum with 2-3 mm tall side lobes.

Flowering time: (September-) October (-November).

Monadenia cernua is a tall robust herb that occurs

in damp to swampy habitats in the western and

southern Cape Province, (Fig. 23) on the flats bet-

ween the mountains and the sea, from the Cape

Peninsula to Humansdorp.

Cape. — 3318 (Cape Town): Rietvalley (-DC), Oct. Zeyher 1569

(SAM). 3322 (Oudtshoorn): Montagu Pass, George, 400 m (-CD),

Oct. 1880, Young in BOL 5534 (BOL). 3323 (Willowmore): be-

tween Keurbooms River and Storms River (-CD), Oct. 1938,

Gil lett 4565 (BOL).

Satyrium cernuum Thunb. was based on a mixed

type; The discordant element was removed by Swartz

and described as Disa physodes (1800).

In 1817 Ker-Gawler described Disa prasinata from

a plant imported from South Africa by a Mr Griffin,

who successfully flowered it. The plate in Edwards’s

Botanical Register has to serve as an iconotype, as no

other type material is available. From the illustration

it is difficult to decide whether the plant belongs to

M. physodes or M. cernua, but the spur appears to be

slightly longer than the dorsal sepal, a character also

found in a Ecklon & Zeyher collection from Rietvlei

Fig. 24. — Monadenia cernua. 1 , flower in side view, x 3. 2, flower

in front view, x 3. 3, column in side view, x 6: a rostellum; c,

anther; d, stigma; e, viscidium. All from Walters 536.

near Cape Town. This collection is grouped with M.

cernua. Disa prasinata is therefore regarded as a

synonym of M. cernua. The type of Monadenia in-

flata Sond. agrees in all characters with M. cernua.

This species is closely related to both M. brevicor-

nis and M. physodes, and to some extent present a

morphocline between these two taxa. It may be dis-

tinguished from the former by the rounded to obtuse

spur, which is straight and not curved towards the

ovary, and from the latter by the spur which is longer

than the lateral sepal (Fig. 22).

The single population of this species which 1

studied in the field occurred on the Tsitsikamma

coast, in a marshy area along the National Road. In

this part of the country there are very few natural

habitats still extant. Collectors’ notes from the

Bredasdorp area indicate that M. cernua occurs in

damp sandy habitats. No further habitat data are

available. One collection from Tsitsikamma ( Bower

600) was found after fire.

The altitude range of the species is from near sea

level at Cape Town, to more commonly between 100

and 300 m above sea level on the coastal flats. The

majority of the populations must have been des-

troyed, as most of the area is under fairly intense cul-

tivation, and in the Knysna area, afforested. Over the

whole distribution range at least some summer rain-

fall occurs. The total rainfall varies from about 600

mm to over 1 000 mm p.a.

360

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

16. Monadenia brevicornis Lindl., Gen. Sp.

Orch. 357 (1838); Kraenzl., Orch Gen. Sp. 1: 816

(1900); Rolfe in FI. Cap 5, 3: 192 (1913). Type: Cape

of Good Hope, Mund s.n. (K, holo.!).

Disa brevicornis (Lindl.) H. Bol. in J. Linn. Soc., Bot. 25: 196

(1889); Schltr. in Bot. Jb. 31: 211 (1901).

Icon: H. Bol., leones Orch. Austro-Afr. 3: t. 40b

(1911).

Plants 200-500 mm tall; tubers up to 50 mm long;

leaves narrowly lanceolate to rarely narrowly ovate,

acute, numerous, imbricate, the largest near the base,

up to 150 mm long and grading apically into the

floral bracts; inflorescence a lax or cylindrical spike,

40-300 mm tall; ovaries 10-15 mm long, more or

less erect; bracts as tall as or overtopping the flowers,

narrowly ovate to ovate, acuminate. Flowers with

lime-green petals and lip, lip with a maroon base,

lateral sepals green, dorsal sepal rust-coloured to

maroon; dorsal sepal erect, shallowly galeate, obtuse

to rounded, apiculate, narrowly obovate to oblong,

7-10 mm tall and c. 1 mm deep; spur pendent with

the apex curved towards the ovary, cylindrical, 2-3

mm in diameter and 7-11 mm long; lateral sepals

reflexed, oblong, obtuse, apiculate, 5-9 mm long;

petals obliquely narrowly ovate to oblong, obliquely

retuse, erect and twisted to face forwards, 5-9 mm

tall; lip pendent, narrowly oblong, obtuse, 6-8 mm

long; anther semipendent, 1,5-2 mm long; rostellum

partially flanking the anther with a deep notch to the

front containing the viscidium, 1-2 mm tall; stigma

with the rear lobe smaller than the lateral lobes short-

ly pedicellate, horizontal. Fig. 25.

Diagnostic features. Flowers large, lateral sepals 5-9

mm long; spur cylindrical, acute, the apex curved

towards the ovary, 7-11 mm long and 2-3 mm in

diameter.

Flowering time: November— February.

Monadenia brevicornis is a slender herbaceous or-

chid, that occurs frequently in montane grassland

from the southern Cape Province to southern Malawi

and Madagascar (Fig. 26).

Cape. — 3423 (Kynsna): The Crags, Knysna (-AA), Nov. 1949,

Compton 21729 (NBG).

Transkei. — 3129 (Port St Johns): Port St Johns (-DA), Oct.

1939, McLoughlin 403 (BOL).

Natal. — 2930 (Pietermaritzburg): escarpment above the Byrne

valley (-CC), Nov. 1975, Hilliard 5583 (NU).

Zimbabwe. — Inyanga, Bidford Estate, 1 800 m, March 1958,

Beasley 64 (K; SRGH).

This species is closely allied to M. cernua, from

which it may be distinguished by the acute spur.

From the rest of the genus, the species is distinct,

because of its inflated spur, longer than the dorsal

sepal, and its distribution in the summer rainfall/

winter drought region.

M. brevicornis occurs very widespread in the mon-

tane grassland regions of southern Africa. Generally

plants are found in slightly damp areas in the grass-

land. The altitudinal range varies from near the coast

in the southern Cape Province (where some popula-

tions occur in Cape ‘Fynbos’ in the Humansdorp,

Grahamstown and King William’s Town areas), to

between 1 300 and 2 700 m in Natal and northwards

to Malawi. In this region the rainfall of about

800- 1 200 mm p.a. occurs almost totally in the sum-

mer months, whereas the higher altitudes receive

snow in the winter months (White, 1978). Popula-

tions tend to be extensive and sparse.

Although the species appears to be relatively com-

mon in South Africa and Zimbabwe, there is only a

single collection known from Malawi (Mt Mlanje)

and from Madagascar.

As at least Mt Mlanje in Malawi has been floristic-

ally well investigated by various botanists, the species

has to be accepted as being rare in that country.

A sweet scent has occasionally been recorded.

PUTATIVE HYBRIDS

Monadenia atrorubens x sabulosa

Linder 1510 consists of a single plant found at Betty’s

Bay in the western Cape, growing between popula-

tions of both putative parents. The specimen is inter-

mediate for all characters between the two parents.

EXCLUDED SPECIES

Monadenia leydenburgensis Kraenzl., Orch. Gen.

Sp. 1: 811 (1900); Rolfe in FI. Cap. 5, 3: 189(1913).

Type: Transvaal, Lydenburg, along the Crocodile

Fig. 25. — Monadenia brevicornis. I , whole plant, x 0,5. 2, flower

in side view with one sepal removed, X3. 3, flower in front

view, x 3. 4, dissection of flower, x 6: a, rostellum; b, petal;

c, anther; d, stigma. All from McLoughlin 42.

H. P. LINDER

361

River, Wilms 1864 (Z, holo.!; BM!; K!) = Disa

stachyoides Reichb. f. in Flora 64: 328 (1881).

Monadenia basutorum (Schltr.) Rolfe in FI. Cap 5,

3: 196 (1913), based on Disa basutorum Schltr. in

Bot. Jb. 20, 50: 17 (1895). Type: Natal, Drakensberg

summit, Thode s.n. (K, iso.!) = Disa basutorum

Schltr.

Monadenia junodiana Kraenzl. in Vierteljahrschr.

Nat. Ges. Zurich 74: 108 (1929). Type: Transvaal,

Mamotsuri, Junod 1208 (Z, holo.!) = Disa fragrans

Schltr. in Bot. Jb. 20, 50: 40 (1895).

ACKNOWLEDGEMENTS

1 would like to thank the numerous people with

whom I discussed this work, especially my super-

visor, Prof. E. A. Schelpe. I am grateful to the direc-

tors and curators of the various herbaria where 1 was

provided with working facilities, or who loaned

material for study purposes. This research was done

while holding a Smuts Fellowship and a CSIR post-

graduate bursary.

UITTREKSEL

Die genus Monadenia ( Disinae , Orchidaceae) word

hersien. Sestien spesies, in vier seksies verdeel, word

erken. Die verspreiding van elke spesie word afgeba-

ken en twee spesies word geillustreer. ’n Hipotese oor

die filogenetiese verwantskappe van die spesies word

aangebied.

REFERENCES

Bentham, G. & Hooker, J. D., 1883. Genera Plantarum III.

London.

Bolls, H., 1888. The orchids of the Cape Peninsula. Trans. S.

Afr. phit. Soc. 5: 1-201.

Bolls, H., 1889. Contributions to South African botany, part IV.

J. Linn. Soc., Bot. 25: 156-209.

Bolus, H., 1893. leones Orchidearwn Austro- Africanarum I,

part 1. London: Wesley.

Bolus H., 1896. leones Orchidearwn Austro- Africanarum 1,

part 2. London: Wesley.

Bolus, H., 1911. leones Orchidearwn Austro-Africanarwn II.

London: Wesley.

Bolus, H., 1913. leones Orchidearwn Austro-Africanarwn 111.

London: Wesley.

Burdet, H.M., 1978. Cartulae ad botanicorum graphicem XIII.

Candollea 33: 365—405.

Brown, N. E., 1815. Terrestrial orchids of South Africa. Gdnrs’

Chron. 24: 402-404.

Brown, N. E., 1818. Disa bracteata. Edwards’s bot. Reg. 4: t . 324.

Durand, T. & Schinz, H., 1894. Conspectus Florae Africae V.

Bruxelles: Charles van de Weghe.

Dyer. R. A., 1976. The genera of southern African flowering

plants. 11. Gytnnospertns and monocotyledons. Pretoria:

Botanical Research Institute.

Erikson, R., 1965. Orchids of the West.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships, and origins. Ann. Mo. bot.

Gdn 65: 369-436.

Holmgren, P. K. & Keuken, W., 1974. Index Herbario rum, part I .

The herbaria of the world. Utrecht: Oosthoek, Scheltema &

Holkema.

Ker, J. B., 1817. Of the three species of the natural order

Orchideae. Q. Jl Sci. Arts 4: 199-206.

Kraenzlin, F., 1900. Orchidacearum Genera el Species. Vol. 1,

part 13. Berlin: Mayer & Mueller.

Lewis, G. J., 1948. Plantae novae africanae. Jl S. Afr. Bot. 14: 31

Lewis, G. J., 1950. Orchidaceae. In Adamson, R. S. & Salter,

T. M. (eds) Flora of the Cape Peninsula. Cape Town: Juta.

Linder, H. P., 1981a. Taxonomic studies in the Disinae. IV. A

revision of Disa Berg. sect. Micranthae Lindl. In preparation.

Linder, H. P., 1981b. Taxonomic studies in the Disinae. VI.

A revision of Herschelia Lindl. Bothalia 13: 000-000.

Lindley, J., 1838. The genera and species of orchidaceous plants.

London: Ridgways.

Pfitzer, E., 1889. Orchidaceae. In Engler, A & Prantl, K. (eds)

Die Natiirlichen Pflanzenfamilien 2.6 Leipzig: Engelmann.

Phillips, E. P., 1926. The genera of South African flowering

plants. Mem. bot. Surv. S. Afr. 10.

POCOCK, M. R., 1972. Ground orchids of Australia.

Reichenbach. H. G., 1847. Orchidiographische Beitrage. Linnaea

20: 673-696.

Reichenbach, H. G., 1883. Die Orchideen des Herbars

Thunbergs. Flora 66: 459-463.

Rolfe, R. A., 1913. Orchidaceae. In Thiselton-Dyer, W. T. (ed.)

Flora Capensis 5, 3. London: Reeve.

Rourke, J. P., 1972. Taxonomic studies on Leucospermum R. Br.

Jl S. Afr. Bot. Suppl. Vol. 8.

Rupp. H. M. R., 1946. Two new orchids from western Australia.

Australian Orchid Review 1 1 : 70.

Schelpe, E. A., 1966. An introduction to the South African

orchids. Cape Town: Purnell.

Schlechter, R., 1898. Orchidaceae africanae novae vel minus

cognitae. Bot. Jb. 24: 418-433.

Schlechter, R., 1901. Monographic der Diseae. Bot. Jb. 31:

134-313.

Schlechter, R., 1924. Contributions to South African Orchid-

ology. Ann. Transv. Mus. 10: 233-252.

Senghas, K., 1973. Orchidoideae. In R. Schlechter (ed.), Die

Orchideen, ed. 3. Paul Parey.

Sonder, G., 1847. Enumeratio Orchideatum, quas in Africa

Australi Extra-tropicum collegerunt C. F. Ecklon, Dr, et C.

Zeyher. Linnaea 19: 71 — 1 12.

Stafleu, F. A., et al., 1978. International code of botanical

nomenclature. Utrecht: Bohn, Scheltema & Holkema.

Swartz, O., 1800. Orchidemes slagter och arter upstallde.

Kongliga Vetenskaps Academia Handlinger, Stockholm 31:

202-254.

Taylor, H. C., 1978. Capensis. In Werger, M. J. A. (ed.)

Biogeography and ecology of southern Africa. The Hague:

Junk.

Thunberg, C. P., 1794. Prodromus Plantarum Capensium.

Upsala.

Thunberg, C. P., 1823. Flora Capensis. ed. J. A. Schultes. Stutt-

gardt.

White, F., 1978. The afro-montane Region. In Werger, M. J. A.

(ed.) Biogeographv and ecology of southern Africa. The

Hague: Junk.

APPENDIX: SPECIMENS STUDIED

The specimens are listed alphabetically according

to the name of the collector. The figures in brackets

refer to the number of the taxon in the text. Herbaria

from which each collection has been studied are indi-

cated by the letter codes of Holmgren & Keuken

(1974). Two taxon numbers connected by a dash, e.g.

(3-4), indicates that the collection is a hybrid be-

tween the two species.

A cocks 1030 (5) S; 5318 (5) S; 19866 (5) PRE; 22797 (15) PRE;

23499 (5) PRE. Adamson 438 (16) K; in SAM 52937 (12) SAM.

Alexander s.n. (5) K. Andreae 27 (6) PRE; sub Marloth 645 (9)

PRE. Atherstone 25 (16) K.

Ball 550 (16) K; 568 (16) K. Barber 445 (16) K. Barker 328 (14)

NBG; 1622 (5) NBG; 3398 (13) NBG; 3892 (5) NBG; 3893 (6)

NBG; 3925 (2) NBG; 4172 (5) NBG; 8838 (8) NBG. Becrs/ey 64(16)

K. Begley 8 (10) SAM. Boardman 49 (16) BOL; 262 (16) PRE.

Bodkin in BOL 4903 (10) BOL; 4970 (4) BOL, G. PRE, SAM, Z;

4988 (8) BOL, K; 623 1 (2) BOL. Bohnen 1066 (5) PRE. Bolus 3859

(5) BM, BOL, K, Z; 4336 (14) K; 4538 (12) BM, BOL, K, Z; 4542

(8) BOL; 4551 (13) BM, BOL, K, PRE, Z; 4555 (9) BM, BOL, K,

SAM, Z; 4885 (6) BOL, K, PRE, Z; 4897 (8) BOL, K; 4969 (13)

362

TAXONOMIC STUDIES IN THE DISINAE. V. A REVISION OF THE GENUS MONADENIA

BOL, K, Z; 4973 (14) PRE; 6862 (6) Z; 7104(3) BOL, SAM; 7317

(16) BOL, G, Z; 9938 (11) BOL; 1 1646 (9) BOL; 13504 (4) BOL;

13505 (6) BOL; sub Guthrie 1080 (3) BOL; s.n. (5) BOL; s.n. (5)

BOL; s.n. (8) BOL; s.n. (9) BOL; s.n. (11) BOL; s.n. (13) BOL.

Bond 1 15 (9) NBG; 1206 (9) NBG. Boucher 699 (5) STE; 829 (12)

STE; 1642 (12) STE; 1946 (9) STE; 2422 (5) PRE, STE. Bower 600

(15) PRE. Bovleli (16) K. Buchanan s.n. (16) K, W. Burchell 40\5

(5) K; 6139 (5) K; 7321 (9) K.

Colder in SRGH 46276 (16) K. Cassidv 22 (5) NBG. Cloete in CH

66620(13) NBG. Codd 2716 ( 16) K, PRE. Coleman 521 (16) PRE.

Compton 4154 (9) BOL, NBG; 4507 (12) BOL, NBG; 4700 (13)

BOL; 9756 (9) NBG; 9759 (9) NBG; 11928 (5) NBG; 11955 (9)

NBG; 11993 (5) NBG; 16244(9) NBG; 16446(12) NBG; 17486 (6)

NBG; 17496 (9) NBG; 18545 (9) NBG; 18609 (9) NBG; 18572 (5)

NBG; 19411 (6) NBG; 20090 (5) NBG; 20180 (13) NBG; 20221 (5)

NBG; 21729 (16) BOL, NBG; 23670 (5) NBG. Cutting s.n. (5)

BOL.

Davis & Stokoe in SAM 49539 (9) PRE. Dalv s.n. (5) PRE.

Dekenal in CH 57315 (12) NBG. Devenish 813 (16) PRE; 1376(16)

PRE. De Villiers in NBG 1910/30(12) BOL; s.n. (5) STE. Doidge

4802 (16) K. Drege 1252a (9) BM, G. K, P, S; 1252b (9) BM, K, P;

1261b (5) BM, G, K, P, W; 8290 (12) K; in SAM 21998 (13) SAM.

Drege 94 (5) Z. Diimmer 547 (13) BM; 554a (6) BM; 934 (9) BM;

s.n. (5) BM.

Ebersohn 151 (10) NBG; 1 52 (12) NBG. Ecklon 247 (5) M. Emdon

30 (12) STE. Esterhuysen 400 (5) NBG; 493 (5) NBG; 6175 (9)

BOL, PRE; 6396 (9) BOL, PRE; 6543 (9) BOL; 6882 (16) BOL;

7086 (16) BOL; 8206 (9) BOL; 9025 (5) BOL; 9780 (10) BOL;

10573 (12) BOL; 10692 (16) BOL, PRE; 10712(9) BOL; 10999(9)

BOL; 1 1 194 (6) BOL; 1 1693 (5) NBG; 12074 (5) BOL; 12200 (9)

BOL, PRE; 13313 (16) BOL; 16200 (9) BOL; 18979a (15) BOL;

21978 (5) BOL; 23635 (12) BOL; 31166 (9) BOL; in SAM 54324 (2)

SAM.

Fannin 33 (16) K. Farnham s.n. (5) BOL. Flanagan 1687 (16) BOL,

PRE; 1807 (16) BOL, PRE, SAM. Fourcade 519a (6) K; 1458 (15)

K, STE; 1626(9) BOL; 2840(12) K, STE; 2848 (16) K; 3446(5) K,

STE, Franklin 40 (16) NU. Frowien in PRE 15649 (5) PRE. Fry

sub Galpin 2719 (16) PRE.

Galpin 308 (16) PRE; 2719(16) PRE; 4605 (9) GRA, K, PRE; 4606

(5) K, PRE; 4607 (5) K, PRE; 4608 (12) PRE; 4609 (9) BOL, K;

4610 (12) BOL, PRE; 4612 (6) PRE. Garside 210 (9) K; 1058 (5) K;

1696 (5) K. Gillett 982 (12) STE; 1352 (5) STE; 1783 (9) STE; 1842

(6) STE; 1856 (5) STE; 2091 (5) STE; 2094 (12) STE; 4565 (15)

BOL; 4573 (5) BOL. Glass 617 (16) Z. Goatcher in BOL 6862 (6)

Z. Goldblatt 324 (II) BOL. Gordon s.n. (16) PRE. Gueinzius 264

(16) P. Guthrie 725 (2) BOL; 1082 (9) NBG; 2729 (14) BOL; in

BOL 7096 (6) BOL; in BOL 7097 (2) BOL, K.

Haf strom & Acocks 2089 (9) PRE. Hall 419 (16) BOL; 736 (5)

BOL; 737 (5) BOL; 1068 (5) BOL; 1073 (5) BOL; 1081 (5) BOL;

1104 (16) BOL; 1116 (16) BOL; 1122 (5) BOL; 1194 (9) BOL.

Hallack in BOL 6093 (16) BOL, K. Hanekom 1264 (9) PRE; 2176

(13) PRE. Harvey 141 (5) K. Havgarth in PRE 22340 (16) K, PRE,

Z. Havnes Palmer in CH 57305 (4) NBG. Hilliard 1695 (16) NU;

5583 (16) NU. Hilliard & Burn 8028 (16) NU. Holland 3739 (16)

BOL. Horrocks 28 (5) NBG. Humbert 13642 (16) P. Hutchinson

575 (9) BM, K, PRE; 672(6) K; 1225 (5) K; 1412a (15) K; 1432(16)

BOL, K.

Jackson in CH 85840 (12) NBG; in CH 86427 (3) NBG. Jacobsen

3720 (16) PRE; 3815 (16) K, PRE. Jacobsz 109a (16) PRE; 2/62

(16) PRE. Jacot Guillarmod, Getliffe and Mzamane 184 (16) K,

PRE. Jeppe in PRE 33400 (6) PRE; 33401 (15) PRE; 33402 (16)

PRE; 33403 (5) PRE; 33404 (12). PRE. Jordaan (5) STE.

Kassner 1 374 ( 16) P; 1479 (5) P. Keet 1024 (12) PRE. Kellerman 26

(13) STE. Kies in CH 57316 ( 10) NBG. Killick 3858 (16) PRE; 3873

(16) PRE. Killick & Vahrmeijer 3631 (16) PRE. Krige in BOL

1 3491 (6) BOL. Kruger 29 (5) STE; 508 (4) STE; 556 (5) STE; 1001

(9) STE; 1072 (12) STE.

Lambs. n. ( 13) BOL. Laughton 77 (5) BOL; 79(5) BOL. Lavranos

9380 (16) PRE; 15232 (16) PRE. Leighton 840 (8) SAM; 1344 (9)

BOL, PRE; 1441 (6) BOL; 1499 (5) BOL; 2079 (5) BOL; 2151 (9)

BOL; 3146 (1 1) BOL. Leipoldt 3236(1 1) BOL; 3237 (5) BOL; 3809

(14) BOL, K; s.n. (5) BOL. Le Sueur in BOL 4973 (14) BOL. Lewis

93 (5) SAM; 648 (5) SAM; 721 (13) SAM; 722 (12) SAM; 787 (5)

SAM; 788 (13) SAM; 816 (5) SAM; 823 (5) SAM; 850 (6) SAM;

851 (8) SAM; 1094 (13) SAM; 1095 (4) SAM; 1096 (5) SAM; 1097

(5) SAM; 1098 (6) SAM; 1099 (6) SAM; 1100 (6) SAM; 1101 (11)

SAM; 1107(6) SAM; 1108(2) SAM; 1487 (1) SAM; 1490(8) SAM;

1832 (13) SAM; 1833 (6) SAM; 2403 (14) SAM; 3546 (6) SAM;

3547 (5) SAM; 3548 (6) SAM; 3549 ( 1 3) SAM; 4389 (5) SAM; 4455

(2) SAM; 4756 (5) SAM; 4757 (8) SAM; 4758 (10) SAM; 4759 (6)

SAM; 4760 (5) SAM; 5020 (9) SAM; 5542 (II) NBG; 5645 (9)

NBG; 6161 (5) NBG; s.n. (II) BOL. Linder 752 (8) BOL; 763 (5)

BOL; 811 (16) BOL; 831 (16) BOL; 841 (16) BOL; 934 (16) BR,

BOL; 943 (16) BOL; 973 (16) BOL; 996 (16) BR; 1000 (16) BOL;

1 129 (11) BOL; 1140(11) BOL; 1149(9) BOL; 1243 (5) BOL; 1479

(9) BOL; 1507 (4) BOL; 1508 (3) BOL; 1509 (11) BOL; 1510

(3-11) BOL; 1512 (11) BOL; 1513 (6) BOL; 1519 (5) BOL; 1524

(12) BOL; 1528 (2) BOL; 1536 (6) BOL; 1537 (5) BOL; 1551 (5)

BOL; 1552 (16) BOL; 1563 (5) BOL; 1564 (15) BOL; 1570 (16)

BOL; 1571 (12) BOL; 1578 (12) BOL; 1580 (5) BOL; 1583 (5)

BOL; 1596 (12) BOL; 1599 (5) BOL; 1601 (9) BR, BOL; 1608 (12)

BOL; 1610 (13) BOL; 1617 (7) BOL; 1694 (10) BOL; 1706 (12)

BOL; 1742 (9) BR, BOL; 1745 (5) BOL; 1748 (10) BOL; 1807 (2)

BOL; 1988 (16) BOL; 2078 (16) BOL. Linlev in SAM 49536 (12)

PRE, SAM; in SAM 49537 (9) SAM; 49538 (9) SAM; in SAM

49541 (9) SAM; in SAM 56094 (8) SAM; in SAM 56898 (13) SAM;

in SAM 56902 (5) SAM; in SAM 56903 (5) SAM; in SAM 56904 (6)

SAM; in SAM 56908 (6) SAM; in SAM 61249 (10) SAM. Liver-

sidge 232 (16) NBG. Long 163 (5) K; 164 (5) K; 827 (5) K, PRE;

835 (16) K, PRE.

MacOwan 381 (5) K, SAM; 679 (16) BM, K, SAM, W, Z, ZT.

MacOwan & Bolus 170 (9) BM, BOL, G, K, P, SAM, W, ZT; 171

(12) BM, BOL, G, K, P, W, ZT; 1374 (3) BOL, SAM. Mann 55 (5)

K. Manning in CH 87243 (12) NBG. Marais 55 (16) PRE. Marloth

61 (9) PRE; 230 (12) PRE; 664 (9) PRE; 1768 (12) PRE; 1854 (1 1)

PRE; 1864 (9) PRE; 4972 (6) BOL; 8916 (5) PRE. Matheson in

SAM 59669(16) SAM. Mauve 4761 (12) PRE. McLoughlin 12(16)

BOL; 42 (16) BOL; 164(16) BOL; 403 (16) BOL; 488 (16) BOL; in

PRE 26260 (16) BOL, K. PRE. Meebold 11895 (5) M. Minicki in

SAM 59671 (2) SAM. Moll 2218 (16) PRE. Moore s.n. (16) BOL.

Morze 2026 (5) BOL; 2027 (5) BOL. Moss 4141 (5) K; 19257 (5) K.

Muir 664 (12) PRE; 746 (5) PRE; 748 (5) PRE; 1124 (12) PRE;

1125 (9) PRE; 1795 (6) PRE; 2330(9) PRE; 2331 (9) PRE; 2332(9)

PRE.

O’Brien s.n. (16) K. O’Connor 2\7> (16) NU; 214 (16) NU; 296 (16)

NU; 334 (16) NU. Oliver 3942 (9) STE; 4202 (5) STE; 4761 (14) K,

PRE, STE; 5088 (9) STE; 5370 (9) PRE, STE; 5438 (5) STE; 5447

(9) PRE, STE; 5591 (9) STE; 6044 (9) STE; s.n. (5) BOL; s.n. (9)

BOL.

Page in BOL. 16232 (6) BOL. Pappe in SAM 21989 (5) SAM; in

SAM 21990 (5) K, SAM; in SAM 21985 (6) SAM; in SAM 21993

(14) SAM; s.n. (12) K. Paterson s.n. (5) BOL, PRE. Penfold 163

(5) NBG; in CH 5731 1 (1 1) NBG. Penther 46(5) M; 53 (16) W; 57

(16) W; 60(12) W; 65 (16) S, W; 81 (16) W; 99(12) W; 104 (5) M;

179 (12) W; 186 (8) W; 215 (16) W; 236 (5) W; 277 (5) W; 329 (5)

W. Phillips 1335 (5) SAM; 1337 (9) SAM; 1341 (12) SAM; 1852(5)

SAM. Phillipson s.n. (5) BOL. Phipps 666 (16) K. Pillans 2748 (5)

PRE; 2748b (6) PRE; 8483 (5) BOL; 9176 (5) BOL; 9324 (5) BOL.

Pocock 503 (9) STE. Porter in CH 57313 (12) NBG. Prentice in

SAM 10789 (14) SAM; in SAM 10790 (9) SAM; in SAM 10791 (9)

SAM. Primos sub Marloth 1 1705 (9) PRE. Purcell 424 (9) STE; in

SAM 91219 (5) SAM; in SAM 91220 (5) SAM.

Rabinowitz in CH 57307 (14) NBG; in CH 77143 (14) NBG. Rat-

tray in BOL 15784 (16) BOL. Rehmann 581 (8) Z; 582 (5) Z; 1950

(9) M. Rogers 2824 (16) Z; 17735 (5) Z; 23633 (5) PRE; 26494 (5)

Z; 26569 (5) Z; 27166 (5) G; 27970(16) Z; 29057 (8) K; s.n. (13) K.

Rudatis 565 (16) STE; 784 (16) BM, K.

Salter 8290 (12) K; 8473 (12) BOL; 8479 (13) BOL; 8554 (3) SAM;

9374 (5) BM; 322/15 (6) BM; 322/17 (2) BM; 323/2 (12) BM;

323/5 (13) BM; in SAM 55850 (5) SAM. Sanderson 894(16) BOL;

938 (5) K; s.n. (16) K. Saunders s.n. (16) BOL. Schelpe 132 (16)

NU; 4227 (5) BM; 4882 (11) BOL; 4895 (5) BOL; 4979 (5) BOL;

6328 (16) BOL, K; 7114(16) BOL; 7118 (5) BOL; 7155 (16) BOL;

7162 (16) BOL. Schlechter 1479 (6) G, K, M, P, W, Z; 1550 (13)

BM, G, GRA, K, S, W, Z; 4713 (16) BOL, K, P, PRE, W, Z; 5167

(11) BOL; 5791 (8) BOL, Z; 5958 (6) BOL, K, PRE; 5965 (5) BM,

G, K, P, PRE, W, Z; 5974 (16) BM, K, Z; 9501 (4) BOL; 9502 (8)

BOL; 9618 (2) BM, BOL, G, K, P, PRE, Z. Schmidt 606 (5) M;

607 (5) M. Schonland 610 (5) PRE. Schur in CH 57290 (6) NBG.

Scullv in BOL 4903 (10) BOL. Seltzer in CH 57304 (5) NBG. Sim

856 (16) NU; 864 (16) NU. Smith 4871 (5) PRE. South 617 (16)

PRE. Stokoe 241 (12) PRE; 1151 (9) PRE; 6764 (9) BOL; 7382(15)

BOL; 9065 (9) BOL; in BOL 16649 (12) BOL; in BOL 17531 (4)

BOL; in BOL 18388 (9) BOL; in CH 9904 (10) NBG; in SAM

36760 ( 1 3) SAM; in SAM 49534 (5) SAM; in SAM 49539 (9) SAM;

in SAM 49540 (9) SAM; in SAM 55884 (9) SAM, in SAM 57747

(12) SAM; in SAM 57748 (9) SAM; in SAM 57749 (9) SAM; in

SAM 57750 (5) SAM; in SAM 59668 (3) SAM; in SAM 61248 (5)

SAM; in SAM 61250 (13) SAM; in SAM 63158 (5) SAM; in SAM

631 59 (9) SAM; in SAM 63759 ( 12) SAM; in SAM 65620 (5) SAM;

in SAM 68251 (9) SAM; sub Marloth 10569(9) PRE; s.n. (4); s.n.

(9) BOL; s.n. (9) BOL; s.n. (10) BOL; s.n. (10) BOL. Strauss 33

(11) NBG.

Tavlor 247 (6) BOL; 252 (10) BOL; 258 (12) BOL; 6462 (9) STE.

Thode A 1 02 1 (5) PRE; A2285 (5) PRE; in STE 3836 (16) STE; in

STE 5427 (5) STE; in STE 6108 (5) STE; in STE 61 10(12) STE; in

STE 8138 (16) STE. Thomas in CH 56374 (3) NBG. Thoms s.n. (5)

H. P. LINDER

363

M. Thompson 3220 (5) STE. Trauseld 569 (16) PRE; 983 (16)

PRE. Trimen s.n. (13) BM. Truiers.n. (13) STE.

Vahrmeijer 1062 (16) PRE. Van Niekerk 191 (5) BOL. Verreaux

s.n. (5) G.

Wall 2228 (8) S; s.n. (5) S; s.n. ( 16) S. Walters 508 (5) BOL; 534 (8)

BOL; 536 (15) BOL. Wasserfall 972 (5) k, PRE. Werdermann &

Oberdieck 139 (5) PRE. Wild 938 (16) k. Wilson in SAM 21999

(13) SAM. Winkler 30 (5) NBG; 5049 (5) NBG. White 610 (5) Z.

Wolley-Dod 394 (5) BM, BOL; 1788 (9) k; 1845 (5) k; 1846 (13)

BM; 2179 (8) BOL; 2992 (12) BOL, k; 3066 (13) BM, BOL; 3067

(12) BM, BOL, K; 3212 (6) k; 3506 (9) BOL; 3587 (6) BOL, k;

3601 (6) BM; 3602 (5) BM; 3635 (15) BOL, k; 5636(4) BM, BOL.

Wood 12254 (16) PRE; 12257 (16) SAM, Z. Worsdell 54b (5) k.

Wright 135 (5) k; 2377 (16) NU. Wurts 436 (5) NBG; 2039 (5)

NBG.

Zeyher 242 (12) k; 1564 (6) G, k, PRE, SAM, W; 1569(15) BOL,

k, SAM; 1570 (13) BOL, G, k, SAM, W; 3924 (12) BM, k, P,

SAM, W; 3925 (9) BM, k, S, SAM, W; 4680 BOL, k, P, PRE,

SAM.

INDEX

Page

Brownleea pentheriana kraenzl. ex Zahlbr 355

Disa Berg., pro parte 342

Disa affinis N. E. Br 351

atrorubens Sehltr 354

Disa auriculata H. Bol 348

basutorum Sehltr 361

bolusiana Sehltr 353

bracteata Swartz 346

brevicornis (Lindl.) H. Bol 360

cernua (Thunb.) Swartz 358

comosa (Reichb. f.) Sehltr 351

conferta H. Bol 343

densiflora (Lindl.) H. Bol 348

fragrans Sehltr 361

macrostaehya (Lindl.) H. Bol 349

mierantha (Lindl.) H. Bol 346

multiflora (Sond.) H. Bol 348

ophrydea (Lindl.) H. Bol 355

ph vsodes Swartz 358

praetermissa Sehltr 346

prasinala ker-Gawl 359

pygmaea H. Bol 345

reticulata H. Bol 351

rufescens (Thunb.) Swartz 356

sabulosa H. Bol 344

stachyoides Reichb. f. 361

Monadenia Lindl 341

atrorubens (Sehltr.) Rolfe 354

auriculata (H. Bol.) Rolfe 348

Page

australiensis Rupp 346

basutorum (Sehltr.) Rolfe 361

bolusiana (Sehltr.) Rolfe 353

bracteata (Swartz) Dur. & Sehinz 346

brevicornis Lindl 360

cernua ( Thunb. ) Dur. & Sehinz 359

comosa Reichb. f. 35]

conferta (H. Bol) Kraenzl 343

densiflora Lindl 348

ecalcarata Lewis 343

inf lata Sond 359

junodiana kraenzl 361

lancifolia Sond 356

leptostachya Sond 356

leydenburgensis kraenzl 360

macrocera Lindl 356

macrostaehya Lindl 349

mierantha Lindl 346

multiflora Sond 348

ophrydea Lindl 354

physodes (Swartz) Reichb. f. 358

prasinala (ker-Gawl.) Lindl 359

pygmaea ( H . Bol.) Dur. & Sehinz 345

reticulata (H. Bol.) Dur. & Sehinz 351

rufescens Lindl. non Thunb 351

rufescens (Thunb.) Lindl 356

sabulosa (H. Bol.) Kraenzl 344

Satyrium cernuum Thunb 359

rufescens Thunb 356

Bothalia 13, 3 & 4: 365-388 (1981)

Taxonomic studies in the Disinae. VI. A revision of the genus

Herschelia

H. P. LINDER*

ABSTRACT

The genus Herschelia (Disinae, Orchidaceae) is revised. Sixteen species, one subspecies and one variety are

recognized. Two new species from tropical Africa (H. chimanimaniensis Linder and H. praecox Linder) and a new

variety from the Cape Province H. lugens (H. Bol.) Kraenzl. var. nigrescens Linder are described. Three new com-

binations are made by transferring the two species of Forficaria and Disa sect. Mieroperistera (one species) to

Herschelia. Thirteen species are illustrated, and the nomenclature and the available information about the habitats

of the taxa are discussed. The species are grouped into two subgenera, one of which is further divided into two sec-

tions and four series. This classification is based on the putative phylogeny, as determined by the method devised by

Wagner (1962).

RESUME

ETUDES TAXONOMIQUES DES DISINAE. VI. UNE REVISION DU GENRE HERSCHELIA

Le genre Herschelia ( Disinae , Orchidaceae) est revise. Seize especes, une sous-espece el une variete sont reconnues.

Deux nouvelles especes d’Afrique tropicale (H. chimanimaniensis Linder et H. praecox Linder) et une nouvelle

variete originaire de la province du Cap, H. lugens, (H. Bol.) Kraenzl. var. nigrescens Linder sont decrites. Trois

nouvelles combinaisons sont fai'es en transferant les deux especes de Forficaria et Disa sect. Mieroperistera (une

espece) a Herschelia. Treize especes sont illustrees, et la nomenclature ainsi que /’information disponible quant a

leurs habitats et taxa sont discutees. Les especes sont groupees en deux sous-genres, une desquelles est de plus divisee

en deux sections et quatre series. Cette classification est basee sur la phylogenie putative, determinee selon la

methode decrite par Wagner (1962).

INTRODUCTION AND HISTORICAL OVERVIEW

Herschelia is one of the ‘minor’ genera in the sub-

tribe Disinae (Orchidoideae, Orchidaceae). The

genus is centred in the Cape Floral Region

(Goldblatt, 1978), where 12 of the 16 species occur.

The remaining four species occur in the montane

grasslands (White, 1978) of southern and south-

central Africa, extending marginally into East Africa

(Robyns & Tournay, 1955).

There has been a conspicuous lack of consensus

about the delimitation of the genus and the species.

Lindley, who describe the genus in 1838, included

only H. graminifolia in it, and placed the other

known species into Disa sect. Trichochila. This treat-

ment was followed by Bentham & Hooker (1883) and

Pfitzer (1889). Harry Bolus transferred all the then

known species to Herschelia, which he treated as a

section of Disa. Schlechter (1901) followed a similar

approach. Bolus provided descriptions and illustra-

tions for the majority of the species, and did excellent

work on the nomenclature and morphology of vari-

ous species in the group (1882, 1889, 1893, et seq.).

Although Rolfe (1913), Schelpe (1966) and Dyer

(1976) essentially follow the generic delimitations of

Bolus and Schlechter, they treat Herschelia as a

genus. Forficaria has generally been regarded as

allied to Herschelia (or to Disa sect. Trichochila), but

has never been included in the same group. Kraenzlin

(1900) produced a rather artificial treatment, where

H. spathulata s.l., and Disa lacera are removed from

Herschelia, and grouped with Disa cooperi, D.

scullyi and D. thodei in Disa sect. Spathulatae.

Several species of this genus have from time to

time been introduced into cultivation in Europe, but

there is little evidence that the plants lasted more than

a few seasons. According to Hooker (1886, 1889), H.

hians and H. spathulata subsp. spathulata flowered

at Kew. In 1905 it was noted that H. graminifolia, H.

spathulata and an unknown species of Herschelia

*Bolus Herbarium, University of Cape Town, Rondebosch, 7700.

were in cultivation (Anon., 1905). In 1912 Rolfe

remarked: ‘Though the species ( H . lugens) has been

repeatedly introduced to cultivation in this country,

it is by no means easy to maintain in good condition,

owing to its tendency to dwindle away after flower-

ing’. In 1955 Dyer remarked that attempts to intro-

duce H. graminifolia were still unsuccessful and that

all tuberous orchids fared badly in cultivation. At

present several species ( H . spathulata subsp.

spathulata, H. lugens, H. purpurascens, H. barbata

and H. graminifolia) are in cultivation in South

Africa. It is to be hoped that methods of propagating

these species may soon be perfected, as several of the

more striking species are already rare in nature, and

may soon, if present trends continue, become extinct.

MORPHOLOGY

In general, the habit of all the species of Herschelia

is grass-like with a radical tuft of linear, usually

rolled, erect leaves, and a slender erect stem. The

vegetative structures in this genus appear to be well

adapted to the various ecological preferences and

phenology of the species, within the framework of

the general grass-like structure.

The tubers are rather variable in size and are often

remarkably large for the size of the plants. Large

tubers are often found on plants collected from well-

drained sandy areas.

The radical leaves display three patterns:

(a) In the winter-rainfall and all-the-year rainfall

regions, the majority of the species have linear,

rolled, erect leaves that are produced before the

flowers, and that may be dry or green at anthesis.

The leaves do not reach above the base of the inflor-

escence, presumably as this would obscure the lower

flowers from any pollinators.

(b) In the winter-rainfall region, the H. spathulata

group has flat leaves. The plant is early flowering,

before the summer drought starts, and the flat leaves

may reflect the absence of xeric adaptations.

366

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

(c) In the summer rainfall regions the plants appear

shortly after the start of the summer rains. This could

be a mechanism to avoid competition with the gras-

ses, which are usually burnt in the winter months.

The leaves are produced after the flowers and are

often up to three times as long as the inflorescence.

Where fires frequently occur, a sheath of old leaf

fibres often accumulates around the base of the stem.

This indicates that individual tubers must reach a

considerable age, as only 5-15 leaves are produced

each season. It is also evidence of their remarkable

fire resistance.

The inflorescences do not show any exceptional

variation. There is a reduction in the number of flo-

wers with increase in flower size — perhaps an index

of the degree of advancement of the species. The flo-

ral bracts are of little taxonomic value. In the majo-

rity of cases the flowers are resupinate. Only in sub-

gen. Forficaria is the lip upper-most, in the one spe-

cies due to the flowers being twisted through 360°, in

the other due to a lack of resupination. In the majo-

rity of taxa the flowers are of various shades of blue.

In sect. Microperistera the flowers are white, whereas

they are more or less maroon in subgen. Forficaria.

In all cases, the flowers are the only brightly coloured

element on the plants.

The sepals are strongly differentiated into the

patent, shallowly concave, lanceolate to ovate lateral

sepals and the galeate or deeply concave dorsal sepal.

The latter varies from orbicular to ovate, and in some

taxa may be shortly spathulate: probably an

advanced character. The galea is usually deep, but it

may also be very shallow. Generally, the galea bears

a spur, but this is absent in three species. The spur is

horizontal at the base, cylindrical, usually obtuse,

and only in one case longer than the dorsal sepal. In

the bulk of the species the spur is a small conical or

cylindrical appendage on the back of the galea. It is

not clear whether the absence of a spur should be

considered primitive or a derived characteristic, or

both, but it clearly arose twice. The presence of a

long spur, found only in H. schlechterana, is almost

certainly a derived state, within the genus.

The inner perianth whorl in Herschelia is highly

specialized. The petals generally show a very distinct

structure, consisting of a basal anticous lobe flanking

the stigma, with the basal part of the limb reflexed

parallel to the anther, and the apex curved upwards.

In subgen. Forficaria this structure is peculiar; the

basal anticous lobe is rudimentary, the apical half of

the petals is villous. Among the remaining taxa,

variation is restricted to the width of the limb of the

petal, the angle of the upward curvature and the

shape of the apex. A lanceolate apex is considered

primitive and a bifid or flabellate expanded apex

derived.

Three basic types of lip structure occur in the

genus:

(1) a patent, reniform lip (subgen. Forficaria)

(2) an ovate, acute lip with entire margins (sect.

Microperistera)

(3) an ovate lip, variously modified or dissected

(sect. Herschelia)

It is probable that an ovate, entire lip is primitive, as

it approaches the undifferentiated form of the organ.

The reniform lip structure of subgen. Forficaria is

unique in the subtribe. With sect. Herschelia exten-

sive variation in lip shape occurs. Occasional speci-

mens of H. hians have entire lips, and the whole

range of variation from them to deeply lacerate lips

may be found in the one species. From this basic

form both the ovate lips with crisped margins, cha-

racteristic of ser. Herschelia, and the beard-like lips

characteristic of the more advanced members of ser.

Lacerae may be derived. This bearded state appears

to have evolved twice (in H. lugens and in H. multi-

TABLE 1 .-Occurrence of taxonomically useful characters in the genus

o

5'

3

n

Z

3

fu

3

8)

3

T5'

3

Spur absent

short

long

Sepals blue

white

red

Petals acute

bifid

lacerate

expanded apex

aciculate

Lip sessile

stalked

entire

crenulate

bearded

trilobed

ovate

reniform

Rostellum

lateral lobes

three horns

Flowers resupinate

not

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

X X

X

H. P. LINDER

367

fida). Spathulate lips also evolved twice: in H. multi-

fida and in ser. Spathulatae.

The column in Hersche/ia is, to a certain extent,

characteristic of the genus. Typically the anther is

horizontal, usually with two large viscidia (these

may, however, be fused), the rostellum erect, with

three equal lanceolate, acute lobes, holding the visci-

dia between them (Bolus, 1882) and a horizontal,

shortly pedicellate stigma with three lobes, the rear or

odd lobe much smaller than the lateral lobes. The

rostellum and stigma structures are variable, but

unfortunately it is rather difficult to get a clear pic-

ture of these structures from dried material. Fre-

quently the rostellum consists of two lateral horns,

which are canaliculate and hold the viscidia at their

apices. The central lobe appears to be highly reduced

or lost. It is possible that the central lobe could be

formed by the fusion of the inner walls of the canali-

culate lateral rostellum horns. This suggestion is

made more likely by the fact that the inner rostellum

lobe is frequently bilobed or bifid. This would indi-

cate that the trilobed rostellum is derived from the

bilobed canaliculate rostellum. Variation in the

stigma structure affects the size of the odd lobe,

which varies from large as the lateral lobes to much

smaller.

The occurrence of taxonomically important cha-

racters in the genus is indicated in Table 1.

PHYLOGENY

The construction of a putative phylogeny for the

genus is basic to the production of a phylogenetic

classification of the species (Funk & Stuessy, 1978).

Objective methods for the construction of phylogen-

ies have been proposed by Hennig (1966), Wagner

(1962) and several others, and are ably reviewed by

Fund & Stuessy (1978) and Bremer & Wanntrop

(1978). Essentially, the derived and generalized

character states for the taxonomically important

characters are postulated, and groups of species are

formed on the number of shared derived character

states. By using the Wagner method (Wagner, 1962)

the more specialized taxa in the genus may be iden-

tified.

The determination of derived character states is

based on two processes (Bremer & Wanntrop, 1978):

detection of transformation series in a structure and

determining the distribution of the character state in

related groups. Generalized character states are likely

to be wide-spread in related groups (Judd, 1979).

This analysis was applied to the variation in the lip

shape, petal apex shape, spur shape and flower col-

our. In lip and petal apex shape there are clear

transformation series from simple entire structures to

variously specialized structures. The simple entire

structures, especially in the lip shape, are also

widespread within the Disinae. The spur shape is

more difficult to rank. It is likely that a short

semipendent spur is typical for the subtribe, but there

is no such structure in Herschelia. Consequently, the

spurless state is considered to be primitive or

generalized. However, this implies that there must

have been a reversal in the evolutionary sequence for

this structure, as two clearly unrelated taxa (subgen.

Forficaria and H. goetzeana) are spurless. In the

flower colour, blue is assumed to be the generalized

state. This implies that white flowers evolved twice:

in H. barbata and H. schlechterana. As these two

taxa are not closely related, this is thought to be like-

ly. The characters used for the analysis are listed in

Table 2. Generalized characters are scored zero, spe-

cialized characters one, and intermediate states are

scored 0,5. For several characters there are several

specialized states.

TABLE 2. — Characters used for the phylogenetic analysis

The summed values for each taxon are used to

place the taxon on the ‘Wagner Tree’. Species shar-

ing the largest number of derived or specialized cha-

racter states are linked first, whereas species sharing

no specialized characters are linked to the putative

ancestor.

The Wagner Tree (Fig. 1) clearly demonstrates the

major groups among the species:

(a) the H. forficaria group is highly specialized

and quite isolated. The basic structure of the petal

and the specialization of the lip, as well as the vege-

tative characters, indicate that it has to be included in

Fferschelia.

(b) H. schlechterana also shares the vegetative

and petal structures typical of the genus, but has a

Fig 1 . — Wagner Tree for Hersche-

lia. The circles indicate the

degree of divergence.

368

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

unique spur, white flowers and an entire lip. It may

also be derived directly from the ancestral stock of

the genus.

(c) The remainder of the genus may be divided

into two groups, mainly on the lip and petal apex

shapes. H. spathulata has a spathulate three-lobed

lip, whereas in H. graminifolia and H. purpurascens

the lip is sessile and the lip margins undulate. Both

these groups are restricted to the Cape Floral Region

(sensu Goldblatt, 1978), and show signs of recent, if

not ongoing, speciation. The H. hians group is cha-

racterized by bearded lips. Within the group various

lines of development have used different petal

shapes. The group may be understood to form a

remarkable sequence of geographical and ecologi-

cally replacing species. The position of H. forcipata

is not clear, the species is known only from a single

collection. The available data indicate that the taxon

may well be ancestral to the group. H. goetzeana

emerges on the ‘Wagner Tree’ as being rather primi-

tive. However, it may well be highly specialized, as a

reduced form derived from H. baurii. At present it is

an enigma and is known from a single collection

from the summit of Mt Mbeya in southern Tanzania.

HERSCHELIA

Herschelia Lindl., Gen. Sp. Orch. 362 (1838);

Benth. & Hook, f., Gen. PI. 3: 630 (1883); Pfitzer in

Naturl. Pflfam. 2,6: 98 (1889); Kraenzl., Orch. Gen.

Sp. 1: 801 (1900); Rolfe in FI. Cap. 5,3: 199 (1913);

Senghas in Schltr., Die Orchideen 1 : 275 (1973), excl.

syn.; Dyer, R.A. Gen. 2: 995 (1976). Type species:

Herschelia coelestis Lindl. [ = //. graminifolia

(Spreng.) Dur. & Schinz]

Disa Berg. sect. Herschelia (Lindl.) H. Bol. in

Trans S. Afr. phil. Soc. 5: 168 (1888); Schltr. in Bot.

Jb. 31: 282 (1901), excl. Disa sect. Amphigena in

synonymy; Summerh. in FI. Trop. E. Afr. 156: 154

(1968).

Disa Berg. sect. Trichochila Lindl., Gen. Sp. Orch.

353 (1838); Pfitzer in Naturl. PflFam. 2,6: 98 (1889).

Type species: Disa barbata (L. f.) Swartz. [ = H. bar-

bata (L. f.) H. Bol.], lectotype.

Forficaria Lindl., Gen. Sp. Orch. 362 (1838);

Benth. & Hook, f., Gen. PI. 3: 631 (1883), Pfitzer in

Naturl. PflFam. 2,6:97 (1889); Kraenzl., Orch. Gen.

Sp. 1: 722 (1900); Rolfe in FI. Cap. 5,3: 207 (1913);

Senghas in Schltr. Die Orchideen, 1: 271 (1973); R.

A. Dyer, Gen. 2: 995 (1976). Type species: Forficaria

graminifolia Lindl. [ = Herschelia forficaria (H. Bol.)

Linder],

Disa Berg. sect. Forficaria (Lindl.) Schltr. in Bot.

Jb. 31: 297 (1901).

Disa Berg. sect. Spathulatae Kraenzl., Orch. Gen.

Sp. 1: 793 (1900). Type species: Disa spathulata (L.

f.) Swartz [Herschelia spathulata (L. f.) Rolfe], lec-

totype.

Disa Berg. sect. Microperistera H. Bol. in Trans.

S. Afr. phil. Soc. 16: 149 (1907). Type species: Disa

schlechterana H. Bol. [ = Herschelia schlechterana

H. Bol.) Linder],

Herschelia is named after Sir John F. W. Herschel

(1792-1871), an astronomer who spent some years at

the Cape.

Characteristic of this genus are the radical, linear,

subsclerophyllous leaves, dry floral bracts, lax inflor-

escences, the rarely entire lip, usually variously dis-

sected or stalked, petals with a basal anticous lobe,

and the limb initially horizontally reflexed, soon fal-

cately or geniculately curved upwards behind the

anther, anther horizontal with one or two viscidia,

rostellum generally with three equal erect, lanceolate

lobes, stigma horizontal, the odd lobe smaller than

the lateral lobes.

Plants slender, grass-like, erect 100-1 000 mm tall;

tubers testicular, rarely three present, very variable in

size, from 10-60 mm long, hirsute; roots unbranch-

ed, thick; base of the stem usually with a sheath of

old leaf fibres; radical leaves 5-20, linear, flat or

rolled, subsclerophyllous, shorter than or longer than

the base of the inflorescence, produced during,

before or after flowering; cauline leaves lax or im-

bricate, brown, acuminate, usually longer and

overlapping to the base of the stem; inflorescence lax

with 1-30 flowers; bracts usually broadly ovate,

acuminate to setaceous, dry, varying from half as

long as to slightly longer than the ovary; ovaries

usually spreading from the stem, slender, 10—30 mm

long. Flowers resupinate, usually blue or shades of

blue, to white with pale blue veins, rarely purplish

red or with green parts; dorsal sepal erect or angled

forwards, generally galeate, rounded to acuminate,

usually ovate in front view with the galea about half

as deep as tall; spur horizontal from the base of the

galea, at length straight, decurved or curved up-

wards, rarely longer than the sepals, cylindrical or

conical; lateral sepals usually patent, lanceolate to

ovate, obtuse to acute, 6-30 mm long; petals with a

basal anticous lobe flanking the stigma, oblong to

ovate, and a limb which is linear or lorate, the basal

part of which is reflexed to the horizontal, flanking

the anther, the apical part of which is curved falcate-

ly or geniculately upwards behind or near the apex of

the anther, the apex of which may be lanceolate, ex-

panded fan-like or more or less bifid; lip generally

more or less dissected, rarely entire, usually ovate in

outline, rarely spathulate, always specialized in some

way; anther horizontal or somewhat pendent, the

two cells parallel, with two viscidia which may be

separate or fused; rostellum generally with the three

lobes equal, erect, lanceolate, acute, rarely with the

lateral lobes canaliculate and the central lobe not pre-

sent; stigma horizontal, shortly pedicellate, the odd

lobe smaller than the lateral lobes, the whole struc-

ture usually wider than the rostellum.

Subgen. Forficaria {Lindl.) Linder, stat. nov.

Forficaria Lindl., Gen. Sp. Orch. 362 (1838). Type

species: Forficaria graminifolia Lindl.

Flowers with the lip facing towards the axis, dorsal

sepal without a spur, lip reniform.

This subgenus contains two closely related species:

H. forficaria and H. newdigateae, that appear to be

vicariants as defined by Davis & Heywood (1963).

H. P. LINDER

369

KEY TO THE SPECIES

la Lip entire:

2a Lip kidney-shaped; petals ciliate:

3a Petals obscurely bilobed, flattened; from the southern Cape Province 2. H. newcligateae

3b Petals aciculate; from the western Cape Province 1 , H.forficaria

2b Lip ovate to lanceolate; petals glabrous:

4a Spur 30-35 mm long 3. //. schlechlerana

4b Spur less than 15 mm long:

5a Apex of the petals obtriangulate, flabellate, truncate:

6a Spur conical, tapering; lip margins curved upwards 6. H. purpurascens

6b Spur subclavate; lip margins curved downwards 5. H. graminifolia

5b Apex of the petals acute, entire or bifid:

7a Spur bifid 12 .H.forcipata

7b Spur obtuse 11.//. hians

lb Lip more or less lacerate or bearded:

8a Lip stalked:

9a Lip blade deeply lacerate 10. H. multifida

9b Lip blade entire, 3-lobed or ovate:

10a Central lobe of lip (1 2-) 16-22 mm long; from east of Swellendam 4b. H. spathulata

subsp. tripartita

10b Centra lobe of lip less than 14 mm long; from west of Swellendam 4a. H. spathulata

subsp. spathulata

8b Lip sessile:

11a Petals deeply bilobed; from north of the Limpopo River:

12a Lateral sepals 6-8 mm long; from the Chimanimani Mountains in Zimbabwe ... 15 . H. chimanimaniensis

12b Lateral sepals 8-25 mm long;

13a Petals as long as the galea; lip almost entire; from the Nyika Plateau in Malawi 13. Ft. praecox

13b Petals about Zi as long as the galea; lip deeply lacerate:

14a Spur present 14. //. baurii

14b Spur absent 16. //. goetzeana

lib Petals obtriangulate or obscurely bilobed; from south of the Limpopo River:

15a Lip longer than the lateral sepals, green or almost black:

16a Flower blue to green 9a. H. lugens vat. lugens

16b Flower almost black 9b. H. lugens var. nigrescens

15a Lip shorter than the lateral sepals, more or less blue:

17a Lateral sepals longer than 15 mm; galea acuminate 7. H. barbata

17b Lateral sepals less than 16 mm long; galea rarely acuminate:

18a Limb of the petal linear, apex somewhat expanded; lateral sepals 12-16 mm long 8 .H. venusta

18b Limb of the petals lorate, apex lanceolate or obtriangulate; lateral sepals less than 13 mm long:

19a Petals lanceolate; lip shallowly dissected; from west of Grahamstown 11 H. hians

19b Petals obtriangulate; lip deeply dissected; from east of Grahamstown 14 . H. baurii

1. Herschelia forficaria ( H . Bol.) Linder,

comb. nov.

Forficaria graminifolia Lindl., Gen. Sp. Orch. 362 (1838);

Kraenzl., Orch. Gen. Sp. 1: 723 (1900); Rolfe in FI. Cap. 5, 3: 207

(1913). Disa forficaria H. Bol., leones Orch. Austro-Afr. 1: t. 87

(1896), nom. nov. Type: Cape Province, Paarl Division, Du Toit’s

Kloof, Drege 2211b (K, holo.l).

Icon: Flower. PI. Afr. 11: t. 415 (1931)

Plants up to 500 mm tall; tubers c. 50 mm long;

stems often with a sheath of old leaf fibres around

the base; radical leaves linear, erect, reaching to the

base of the inflorescence, acute; cauline leaves lax,

narrowly ovate, acuminate, 20-40 mm long; inflor-

escence lax, c. 100 mm long; bracts shorter than the

ovaries, ovate, acuminate, dry; ovaries c. 15 mm

long. Flowers twisted through 360°, sepals greenish

red, petals and lip deep maroon; dorsal sepal erect to

reflexed at anthesis, deeply concave, orbicular,

apiculate, 8-12 mm in diameter; lateral sepals patent

*In the treatment of each taxon only representative specimens have

been cited. A full list of all specimens studied is given in the Ap-

pendix, pp. 387-388.

narrowly ovate, acute to shortly acuminate, shal-

lowly concave, 8-12 mm long; petals with the basal

half lorate, reflexed parallel to the anther, 4-6 mm

long, the apical half aciculate, geniculately bent for-

wards, terete, acute, tomentose; lip reniform, patent,

4-5 mm long and c. 6 mm wide, the distant margin

somewhat swollen and shortly villose; anther

horizontal, 4-5 mm long, caudicles short; rostellum

small with well-developed staminodes; stigma pedi-

cellate, almost equally trilobed. Fig. 2.

Diagnostic features. Dorsal sepal spurless, lip

reniform, petal apex almost setaceous and tomen-

tose, flowers twisted through 360°.

Flowering time: January and February.

A rare reed-like plant that occurs on well-drained

gravelly slopes in the Caledon and Cape Peninsula

Division (Fig. 3).

Cape.* — 3418 (Simonstown): Klaver Valley (-AB), Jan. 1922,

Pillans 4125 (BOL). 3419 (Caledon): Viljoenspass (-AA), Feb.

1933, McGillett 718 (BOL).

370

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

Fig. 2. — Elerscnelia forficaria (1-4) and Et. newdigateae (5-6).

1, habit, xO,25, from Gillettll8. 2, flower, x3. 3, lip, x5.

4, petal, x5 (2-5 from Rosenbruck s.n.) 5, lip of H. new-

digateae, x 5. 6, petal, x 5. (5 & 6 from Bolus 6327.)

Very little is known about this rather peculiar

species. It appears to grow on well-drained gravelly

mountain slopes, but I have not succeeded in finding

any populations in the field. Several collections were

made shortly after veld fires, but this may be due

merely to the greater ease of finding these rather

cryptic plants in the restioid vegetation that

dominates the habitat. The plants occur singly and

widely scattered. The altitude range of the species is

from about 100 m to 600 m. Rainfall is concentrated

in the winter months, and totals about 800 mm p.a.

(Jackson, 1961).

The specific epithet ‘ graminifolia ’, under which

the species is commonly known, cannot be transfer-

red to Disa nor to Herschelia, as Sprengel described a

Disa graminifolia in 1 826, which Durand and Schinz

transferred to Herschelia (1894). When Bolus (1896)

transferred Forficarin graminifolia to Disa. he pro-

posed D. forficaria as a nomen novum. This epithet

is here transferred to Herschelia.

2. Herschelia newdigateae (L. Bol.) Linder,

comb. nov.

Disa newdigateae L. Bol. in Flower. PI. Afr. 11: t. 415 (1931).

Type: Cape Province, Knysna, Forest Hall, Newdigate in BOL

6327 (BOL, holo. !).

Icon: H. Bol., leones Orch. Austro-Afr. 1: t. 87

(1896), as Disa forficaria.

Plants up to 500 mm tall, radical leaves linear,

acute up to 300 mm long; cauline leaves lax to subim-

bricate, acuminate, 30-50 mm long, completely

sheathing; inflorescence lax, c. 100 mm long; bracts

about as long as the ovaries, acuminate, narrowly

ovate; ovaries c. 15 mm long. Flowers not resupinate;

sepals greenish red, petals and lip deep maroon; dor-

sal sepal somewhat spathulate with a very short limb,

the blade deeply concave, orbicular, apiculate, c. 8

mm in diameter, the margin somewhat dentate; later-

al sepals ovate, acute, concave, c. 8 mm long; petals

with the basal 4 mm narrowly ovate, parallel to the

anther, the apical 'A curved up behind the anther,

somewhat expanded apically and shallowly bilobed,

tomentose; lip reniform, 4 mm long and 6 mm wide

with the front margin ciliate; anther horizontal, c. 2

mm long; rostellum with the lateral lobes canalicu-

late, very small, central lobe apparently obsolete,

viscidia big; stigma horizontal, flat. Fig. 2.

Diagnostic features. Dorsal sepal spurless, lip reni-

form, petal apex somewhat flattened and shallowly

bilobed, tomentose, ovary not twisted.

Flowering time: March-April.

Very rare in the area between Nature’s Valley and

Plettenberg Bay (Fig. 3), from where it is only known

from two collections. It grows on dry slopes in short

macchia vegetation facing the sea. Superficially this

species resembles H. forficaria, but a study of the

flower soon reveals several distinguishing characters

(shape of petal and galea, apex of lip and relative

length of anther. Fig. 2). These two taxa are clearly

eco-geographic vicariants.

The differences between the two taxa were not ob-

served by the several taxonomists, who had studied

the available material. Bolus (1896) published an il-

lustration of H. newdigateae under the name H. for-

ficaria. It was only when more fresh material of H.

forficaria became available that the differences be-

tween the taxa were detected.

Subgen. Herschelia

Flowers resupinate, dorsal sepals (with one excep-

tion) spurred, lip more or less ovate and generally

lacerate.

Fig. 3. — Distribution of El. forfi-

caria (open circles) and Et.

newdigateae (closed circles).

H. P. LINDER

371

Type species: H. graminifolia (Spreng.) Dur. &

Schinz.

This subgenus contains those species which have

traditionally been placed in Herschelia and which are

often popularly known as ‘Blue Disas’. The group

contains a wide range of forms and is here further

subdivided.

Sect. Microperistera H. Bol. in Trans. S. Afr. phil.

Soc. 16: 149 (1907).

Type species: Herschelia schlechterana (H. Bol.)

Linder.

Lip entire, sessile, ovate, spur longer than sepals.

3. Herschelia schlechterana ( H . Bol.) Linder,

comb. nov.

Disa schlechterana H. Bol. in Trans. S. Afr. phil. Soc. 16: 149

(1907); Rolfe in FI. Cap. 5,3: 250 (1913). Type: Cape Province,

Riversdale District, Garcias Pass Luyt in BOL 10571 (BOL,

holo . ! ; BM!; BR!; K!; W!).

Icon: H. Bolus, leones Orch. Austro-Afr. 2: t. 75

(1911).

Plants about 600 mm tall; tubers c. 4 mm long and

15 mm in diameter; base of the stem often with the

fibrous remains of old leaves; radical leaves about

10, 300-400 mm long and about 2 mm wide, sulcate,

the inner leaf surface smooth and the outer ridged

longitudinally; cauline leaves about 9, lax or subim-

bricate, completely sheathing, dry, acute, 30-50 mm

long; inflorescence lax, 100-200 mm long and with

3-12 flowers; bracts about 2A as long as the ovaries,

lanceolate, acuminate, dry; ovaries about 30 mm

long at anthesis, slightly curved. Flowers cream with

mauve veins; dorsal sepal erect, galea obtuse, 22-25

mm tall, c. 16 mm wide and 8 mm deep, the margins

curved outwards; spur from a shortly conical base,

horizontal at the base and at length gradually decurv-

ed, slender cylindrical, subacute, 30-50 mm long;

lateral sepals patent, lanceolate to narrowly oblong,

suboblique acute 20-25 mm long, aciculus 0,5-1 mm

long; petals with the basal anticous lobe orbicular, c.

3 mm in diameter, decurrent with the limb of the

petal, limb lorate, 14 mm long, the basal part

horizontal, soon geniculately curved through 135° to

face forwards, the apex lanceolate, acute; lip patent,

flat, narrowly oblong to lorate, acute, 15-20 mm

long; anther somewhat pendent, 4,5 mm long with

two globular viscidia; rostellum equally 3-lobed, 4

mm tall; stigma unequally 3-lobed, the odd lobe

smaller than the lateral lobes, horizontal and c. 3 mm

in diameter.

Diagnostic features. Flowers large, lateral sepals

20-25 mm long, spur 30-35 mm long; lip entire, nar-

rowly oblong to lorate.

Flowering time: December.

This striking species has only been recorded from

the dry north-facing slopes of the Langeberg (Fig. 4)

in the vicinity of Riversdale, where it appears to grow

amongst the sclerophyll bush in moister places.

Cape. — 3321 (Ladismith): Garcias Pass (-CC), Dec. 1930,

Ferguson s.n. (BOL).

Sect. Herschelia

Lip sessile or stalked, usually lacerate, spur shorter

than the sepals.

Type species: H. graminifolia (Spreng.) Dur. &

Schinz.

Within this section four clear groups may be

recognized, mainly on the basis of petal and lip mor-

phology, and secondarily on the shape of the spur.

The exact relationships among these series is not

clear, but it is likely that ser. Spathulatae and ser.

Herschelia are closely related rather than to ser.

Hians and ser. Ecalcaratae. The former two series

both occur in the western Cape Province and have ex-

panded apical petal lobes and non-lacerate lips,

whereas the latter two sections have lacerated lips

and extend into East Africa.

Ser. Spathulatae ( Kraenzl .) Linder, stat. nov.

Disa sect. Spathulatae Kraenzl., Gen. Sp. Orch. 1:

793 (1900).

Type species: H. spathulata (L. f.) Rolfe, lecto-

type.

Lip spathulate, with a short or long stalk and the

apical lobe obtusely trilobed to deeply trifid, petals

expanded apically.

This series contains a single species complex and is

restricted to the western and southern Cape Pro-

vince.

4. Herschelia spathulata (L. /.) Rolfe in FI.

Cap. 5,3: 205 (1913). Type: Cape of Good Hope,

Thunherg s.n. (LINN, holo.!; UPS!; W!).

Plants 120-300 mm tall, tubers 15-30 mm long;

base of the stems often with a sheath of fibrous leaf

remains; basal sheaths 2-3, hyaline, acute radical

leaves 5-20, linear, 50-150 mm long and 2-4 mm

wide, narrower towards the base, semi-erect, curved

falcately; cauline leaves (2-) 3, dry, 20-30 mm long,

completely sheathing, acuminate, grading to the

floral bracts; inflorescence laxly 1-5-flowered, up to

100 mm long; bracts varying from half as long as the

ovary to longer than the ovary, broadly ovate,

acuminate to setaceous, dry; ovaries straight or

slightly curved, 15-30 mm long. Flowers variable in

colour from maroon to pale lime or green and blue;

dorsal sepal erect, subspathulate, the limb horizon-

tal, 1-3 mm long, the blade usually galeate, rarely

flat, erect, 9-17-20 mm long, broadly ovate, obtuse,

usually 5 mm deep; spur usually clavate, rarely cylin-

drical, obtuse, straight or strongly curved down-

wards, 1,5-3 mm long; lateral sepals patent or curv-

ed up in front of the flowers, narrowly ovate to

ovate, oblique, acute the apical part conduplicate,

6-16 mm long; petals with the basal anticous lobe

oblong, 3 mm long, decurrent with the limb; petal

limb linear to lorate, 7-12 mm long, the basal part

parallel to the anther, the apical part curved upwards

behind the anther, the apex dilated, unequally

bilobed, up to 4 mm wide; lip spathulate, the claw

linear, 5-35 mm long and 1-2 mm wide, the blade

obscurely trilobed to deeply trifid, the margins gene-

372

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

rally undulate, the central lobe usually longer than

the lateral lobes, lip may be held horizontally with

the blade pendent, or the whole structure may be

pendent; anther more or less horizontal, 1-3 mm

long, viscidia separate; rostellum equally trifid, erect,

the lobes acute, 1-2 mm tall; stigma with the lateral

lobes much larger than the odd lobe, horizontal, 2-3

mm wide and 1 mm tall.

Diagnostic features. Petal spathulate, claw linear,

5-35 mm long, blade obscure trilobed to deeply tri-

partite; leaves linear, flat, green at anthesis.

Two subspecies are recognized in this species.

There are no absolute differentiating characters be-

tween the two postulated taxa, but the overlap in va-

riation is minimal (see Fig. 5). The taxa are allopat-

ric, with about 500 km between the two distribution

areas. They may therefore be recognized as

geographical subspecies.

22

(S

a

v

0)

o

20

18

16

C 12

10 1 ‘ — * ‘ * < < . . .

4 6 8 10 12 14 16 18 20 22 24

Length of Median Lip Lobe

Fig. 5. — Variation in flower size (measured by the length of the

dorsal sepal) and lip shape (measured by the length of the

median lip lobe) in H. spathulata.

(a) subsp. spathulata

Orchis spathulata L. f., Suppl. PI. 398 (1781). Satyrium

spathulatum (L. f.) Thunb., Prod. 5 (1794). Disa spathulata (L. f.)

Swartz in Vet. Acad. Handl. 21: 213 (1800); Lindl., Gen. Sp.

Orch. 353 (1838); Kraenzl., Orch. Gen. Sp. 1: 794(1900); Schltr. in

Bot. Jb. 31: 283 (1901), pro parte. Herschelia spathulata (L. f.)

Rolfe in FI. Cap. 5, 3: 205 (1913). Type: Cape of Good Hope,

Thunberg s.n. (LINN, holo.l; UPS!; W.!).

Disa propinqua Sond. in Linnaea 19: 95 (1847). Type: Cape

Province, Clanwilliam Division, Brakfontein, Ecklon & Zeyher

s. n. (S, holo.!; K!; W!).

D. propinqua Sond. var. trifida Sond. in Linnaea 19; 96 (1847).

Type: not seen.

D. atropurpurea Sond. in Linnaea 19: 96 (1847); Kraenzl.,

Orch. Gen. Sp. 1: 794 (1900). D. spathulata var atropurpurea

(Sond.) Schltr. Bot. Jb. 31: 284 (1901). Herschelia atropurpurea

(Sond.) Rolfe in FI. Cap. 5, 3: 205 (1913). Type: Cape Province,

Tulbagh District, Tulbagh Waterfall, Ecklon & Zevher s.n. (S,

holo.!; K!).

leones: Curtis’s bot Mag. t. 6891 (1886), as Disa

atropurpurea; H. Bol., leones Orch. Austro- Afr. 3:

t. 53 (1913), as D. spathulata; 3: t. 54, as D.

spathulata var. atropurpurea; Rice, Wild. Flow.

Cape G. H. 163.2 (1950).

Central lobe of lip 5—14 mm long, inflorescence

with 1 - 5 flowers (Fig. 6). Distributed in the western

Cape Province from Caledon to Nieuwoudtville.

Flowering time: September and October.

Widespread in the western Cape Province (Fig. 7)

on both sandstone and shale, in a range of habitats,

usually occurring in small populations.

Cape. — 3119 (Calvinia): in mountains near Twakfontein,

Nieuwoudtville (-AC), Leipoldt 601 (BOL). 3218 (Clanwilliam):

in sand near Swartbosch kraal (-BC), Sept. 1894, Schlechter 5 165

Fig. 6. — Herschelia spathulata subsp. spathulata. 1, habit, x0,5,

from Linder 1244. 2, flower, xl,5, from Linder 1453. 3,

petal, x 3, from Linder 1453. 4, lip, x 1,5, from Linder s.n.

5, lip, x 1,5, from Linder 1245.

(BOL; P; PRE; W; Z). 3319 (Worcester): Tulbagh (-AC), Oct.,

Pappe s.n. (BOL; SAM). 3419 (Caledon): Boontjieskraal (-AA),

Sept. 1977, Linder 1458, 1459 ( BOL).

The altitude range of this species is from 150 to

1 000 m and the precipitation ranges from 200 to 800

mm p.a., mostly occurring in the winter months. The

species has been recorded from both gravelly and

deep alluvial sands derived from Table Mountain

Sandstone, as well as clayey soils derived from

Malmesbury shales. Populations growing on shales

occurred most frequently on the cooler south-facing

slopes, whereas those located on sandstone derived

soils ranged from well-drained sites to the margins of

temporary vleis.

The variation patterns in this subspecies are com-

plex, with four characters varying extensively (flower

size, lip lobe shape, flower colour and the three-

dimensional position of the lateral sepals). The

characters are to some extent correlated and a small-

flowered form (probably Disa atropurpurea) and a

large-flowered form may be recognized. The small-

flowered form has dark red flowers, the lip lobe is

more or less ovate and the lateral sepals appear to

close at night, whereas the large-flowered form has

pale lime or green or pale red flowers, the lip lobe is

deeply three-lobed, and the lateral sepals are always

spreading. Both forms may be found over the whole

distribution range of the subspecies. However, a

survey of the available herbarium material indicated

that the variation range of each character is con-

tinuous, and the characters are not perfectly cor-

related. No formal taxa are therefore recognized.

The subspecies was studied in the field in several

localities. Only at two localities, Wolseley and

Caledon, did both forms occur together. At Wolseley

(near Worcester) the small-flowered form occurred

on a well-drained slope, whereas the large-flowered

form occurred about 3 km distant on the margin of a

temporary vlei ( Linder 1244, 1245). At Caledon both

forms occurred on the same slope, and the in-

dividuals of the two forms occurred intermixed. The

H. P. LINDER

373

Fig. 7. — Distribution of H. spat-

hulata subsp. spathulata

(solid circles) and subsp. tri-

partita (open circles).

small-flowered form was much rarer than the large-

flowered form. An analysis of this population for

two characters (flower size and the lip shape) (Fig. 8)

clearly shows a bimodal distribution, with a few rare

intermediate forms.

i/>

c

41

Sepal (mm)

Length of Lateral

Lip Lobe (mm)

Fig. 8. — Variation in the lip shape, length of the lateral lip lobe

and flower size in one population of H. spathulata subsp.

spathulata.

The treatment here is provisional. A detailed study

is required to elucidate the mechanism by which this

complex polymorphism is maintained. Could this be

a case of incipient sympatric speciation?

Sonder (1847) distinguished his Disa propinqua

from H. spathulata by its generally more robust

nature. I have studied the type specimen: it is not

much different from the type of H. spathulata and is

clearly included in the range of variation of the latter

taxon. Disa atropurpurea was distinguished from its

congeners by the shorter lip claw and the maroon col-

our of the flowers (Bolus, 1913; Rolfe, 1913). Lip

claw length varies continuously in H. spathulata

from shorter than that of the type of D. atropurpurea

to 35 mm. Flower colour and the lip shape indicate,

however, that it may well belong to the cryptic small-

flowered form discussed above. It appears as if

Hooker (1886) and subsequent botanists have

misunderstood the complex nature of the so-called

H. atropurpurea.

(b) subsp. tripartita ( Lindl .) Linder, stat. et

comb. nov.

Disa tripartita Lindl., Gen. Sp. Orch. 353 (1838); Kraenzl.,

Orch. Gen. Sp. 1: 797 (1900). Type: Eastern Cape Province,

Albany, Geelhoutboom, Drege 3577a (K, holo.!; P ! ; S!).

D. spathulata (L. f.) Swartz, Schltr. in Bot. Jb. 31: 284 (1901),

pro parte.

Herschelia tripartita (Lindl.) Rolfe in FI. Cap. 5, 3: 204 (1913).

Central lobe of the lip (12-) 16-22 mm long, in-

florescence with 1-2 flowers. Recorded from the

Uniondale area in the Langkloof.

Flowering time: October.

Distribution: Fig. 7.

Cape. — 3323 (Willowmore): Haarlem (-CB), Oct. 1930, Four-

cade 4344 (BOL); Louterwater (-DC), Sept. 1969, Marsh 1408

(PRE).

This subspecies is rather local in the renoster-

bosveld near the summit of the Langkloof, near

Uniondale. The rainfall in the area is about 600 mm

p.a. which is distributed over the whole year.

The type locality for this subspecies is on the

Bushmans River, near Grahamstown. This is about

250 km to the east of other known distribution

records. It is difficult to decide whether Drege’s

locality is incorrect, or whether this species is more

widely distributed than the present records indicate.

Ser. Herschelia

Lip sessile, narrowly elliptical to elliptical, margins

crenulate, petals with apices expanded into flabellate

structures.

Type species: H. graminifolia (Spreng.) Dur. &

Schinz.

Two closely related species are included in this

series: H. graminifolia and H. purpurascens. The

former species is widespread from the Cape Penin-

sula to Port Elizabeth, and the latter is restricted to

the southern Cape Peninsula and the coastal areas of

the Caledon Division. The habitat differences be-

tween the two taxa are summarized in Table 3. Al-

though populations of the two species occur within a

few kilometres of each other, it is clear that the spe-

cies occupy quite different habitats. These differen-

ces are maintained by the prevention of gene flow

between the two species by allochronic flowering

(Table 4). The earlier flowering of H. purpurascens

might be the result of this species occupying a drier

habitat but, as the flowering time remains the same

under cultivation, it appears to be genetically fixed.

374

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HER SCHEL I A

TABLE 3. — A comparison of the habitats of H. graminifolia

and H. purpurascens in the western Cape

*Measured at Simonstown and Kleinmond for H. purpurascens,

and at Table Mountain (Maclears Beacon) and Steenbras Dam

(Caledon Division) for H. graminifolia.

TABLE 4. — Flowering times of H. graminifolia and

H. purpurascens

5. Herschelia graminifolia ( Spreng .) Dur. &

Schinz, Consp. FI. Afr. 5: 111 (1894); Kraenzl.,

Orch. Gen. Sp. 1: 802 (1900).

Disa graminifolia Ker-Gawl. ex Spreng. in Linn. Syst. Veg. 3:

699(1828); Schltr. in Bot. Jb. 31 : 290 ( 1901 ). Type: Cape of Good

Hope, Masson s.n. (BM, holo.).

Herschelia coelestis Lindl., Gen. Sp. Orch. 363 (1838); Rolfe in

FI. Cap. 5,3: 201 (1913). Type: Cape of Good Hope, Burchell 7801

(K, holo.!).

leones: Ker-Gawler in Q. J1 Sci. Arts t. 1, fig. 2

(1819); H.Bol. leones Orch. Austro-Afr. 1: t. 37

(1893); Rice, Wild Flow. Cape G. H. 167.2 (1950);

Flower PI. Afr. 30: t. 1172 (1955).

Plants 500-1 000 mm tall; the base of the stem

often with a sheath of old leaf fibres; radical leaves

usually 5, 200-500 mm long and up to 5 mm wide,

frequently rolled acute, semi-erect; cauline leaves lax,

6-9, 20-40 mm long, acuminate, closely sheathing;

inflorescence lax, 40-120 mm long and with 2-6

flowers; bracts about 2A as long as the ovaries, dry,

broadly ovate, acuminate to setaceous; ovaries 1 5—25

mm long, straight or slightly curved. Flowers blue to

violet-purple, apices of the petals green and the lip

more purple than the sepals; dorsal sepal erect, galea

obtuse, 15-20 mm tall and 5—10 mm deep; spur from

the base of the galea, usually straight, 2-4 mm long,

clavate, rounded rarely cylindrical and obtuse; lateral

sepals narrowly oblong to oblong, obtuse, apiculate,

patent, 13-18 mm long and 6-10 mm wide; petals

with the basal anticous lobe orbicular to rarely

oblong, margins usually entire, 3-4 mm in diameter,

limb of the petal lorate, 11-16 mm long, geniculately

upeurved through 90° about 2A down the length, the

apex expanded into a flabellate structure with entire

or dentate margins and a diameter of 4-6 mm,

twisted to stand erect behind the anther; lip narrowly

elliptical to elliptical, margins usually denticulate and

down-curved, obtuse, patent, 11-16 mm long; an-

ther c. 5 mm long, viscidia separate or fused;

rostellum with three equal lanceolate lobes; stigma

flat, unequally three-lobed. Fig. 10.

Diagnostic features. Lip entire, spur 2-4 mm long,

clavate, lip flat or the margins somewhat down-

curved.

Flowering time: (December-) January to March.

Widespread and local along the coastal mountains

from the Cape Peninsula to Port Elizabeth on soils

derived from Table Mountain Sandstone, usually in

well-drained stony habitats (Fig. 9).

Cape. — 3319 (Worcester): Slanghoek Ridge Peak, 1 350 m

(-CA), March 1952, Esterhuysen 19990 (BOL). 3318 (Cape

Town): Table Mountain, 750 m (-CD), February 1884, MacOwan

& Bolus 167 (BOL; BM; K; P; SAM; W; ZT). 3320 (Montague):

Langeberg near Swellendam (-CD), January 1893, Schlechter

2061 (BOL; Z). 3323 (Willowmore): Louterwater (-DC), January

1941, Compton 10499 (NBG).

Although this species is widespread, there appear

to be sizeable gaps in the distribution range. This

could reflect the lack of botanical exploration of the

Langerberg. Distinct populations may be readily

recognized, although individuals are quite far apart.

There appears to be rather little variation among the

populations. Occasional, presumably recessive,

white-flowered plants have been found.

The altitude range of the species is from 300 m to

1 500 m, and the majority of collections are from

areas receiving in rainfall, an excess of 1 000 mm

p.a., usually with no long dry season.

The species was first referred to in the literature by

Ker-Gawler, who published a plate of it in 1819. The

Fig. 9. — Distribution of Herschelia

graminifolia.

H. P. LINDER

375

plate was prepared by a ‘Dutch’ soldier, ‘an artist of

great skill as a designer of the objects of natural

history’, whom Francis Masson met at the Cape.

Although Ker-Gawler annotated the plate as ‘ Disci

graminifolia’, he provided neither a description nor a

diagnosis. According to Article 32.1 (c) of the

I.C.B.N. (1978), the name has to be regarded as a

nomen nudum. In 1828 Sprengler lists Disa

graminifolia, providing it with a Latin description

and refers to the Masson collection. Sprengler

therefore validly published the name.

Lindley (1838) refers the name Disa graminifolia to

the synonymy of his Herschelia eoelestis. Rolfe

(1913), following the Kew Rule, upheld Herschelia

eoelestis in preference to H. graminifolia. H.

graminifolia, as the oldest available name, has to be

upheld as correct.

Fig 10. — Herschelia purpurascens (1) and H. graminifolia (2-4).

1, flower of H. purpurascens with one lateral sepal removed,

x 1,5, from Linder 759. 2, flower of H. graminifolia, x 1,5.

3, lip, column and petal of H. graminifolia, x 1,5. 4, petal,

x3. (2-4 from Linder 1763.)

6. Herschelia purpurascens (H. Bol.) Kraenzl-,

Orch. Gen. Sp. 1: 803 (1900); Rolfe in FI. Cap. 5,3:

200 (1913). Type: Cape Province, Cape Peninsula,

Muizenberg Mountain, Bolus 4893 (BOL, holo . ! ;

K!).

Disa purpurascens H. Bol. in J. Linn. Soc., Bot. 20: 482 ( 1884);

Schltr. in Bot. Jb. 31: 291 (1901).

leones: H. Bol., leones Orch. Austro-Afr. 1: t. 86

(1896); Rice, Wild Flow. Cape G. H. 146.3 (1950).

Plants 250—500 mm tall; tubers c. 30 mm long and

10 mm in diameter; base of the stem frequently with

a sheath of old leaf fibres; radical leaves about 10,

from half as long as the stem to as long, up to 1 mm

wide, rigid and erect, the midrib sclerenchymatous

and prominent; cauline leaves completely sheathing,

5-7, acuminate, dry, 20-40 mm long, grading

apically into the floral bracts; inflorescence lax, up to

15 mm long and with 1-2-7 flowers; bracts Vi to 2/j

as long as the ovaries broadly ovate, acuminate to

setaceous, dry; ovaries 15-20 mm long, straight or

slightly curved. Flowers blue, the lip more purplish

than the sepals, the rear lobes of the petals yellow or

green; dorsal sepal erect, galea subacuminate, 15-25

mm tall and 10-15 mm deep, ovate; spur from the

base of the galea, horizontal or slightly curved up-

wards, conical obtuse, 1-4 mm long; lateral sepals

oblong, acute, patent, 15-18 mm long; petals with

the basal anticous lobe oblong to semicircular, 3-4

mm in diameter, the margin entire or crenulate, the

limb of the petal lorate, 8-10 mm long, falcately

curved upwards inside the galea, the apex expanded

into a 4-5 mm wide fan with a crenulate margin; lip

broadly ovate, obtuse, with a short limb, margins

crisped and curved upwards, 12-18 mm long; anther

horizontal, 3 mm long, with two viscidia that may be

partially fused; rostellum equally 3-lobed, 2,5 mm

tall; stigma sub-equally 3-lobed, horizontal, 4 mm in

diameter. Fig. 10

Diagnostic features. Lip margin undulate, curved

upwards; spur conical, tapering to an obtuse point,

lateral sepals 15-18 mm long.

Flowering time: October and November.

Local near the coast at low altitudes in the

southern Cape Peninsula and the Caledon Division,

(Fig. 11), growing in well-drained localities.

Cape. — 3418 (Simonstown): Simonstown (-AB), November

1892, Fair in BOL 7992 (BOL); Cape Point (-AD), November

1947, Compton 20236 (NBG). 3419 (Caledon): Betty’s Bay (-BD),

November 1977, Linder 759 (BOL).

This species has a rather restricted distribution

along the coast line along the southern Cape Penin-

sula and between Cape Hangklip and Cape Agulhas.

The altitude range of the species is from sea level to

about 100 m. The summers tend to be dry, with only

occasional rain. The rainfall in this region varies

from 800 to 1 200 mm p.a. Although this species is

not common, populations are not as scattered as in

H. graminifolia. There is not much variation in the

species, but occasional plants with white flowers have

been recorded (possibly the recessive condition).

H. purpurascens is very close to H. graminifolia,

from which it may be distinguished by the conical

spur, the upeurved lip margins and the much earlier

flowering time.

Ser. Lacerae Linder, ser. nov., labello crenato vel

lacero, petalis bidentatis laceris vel acutis. dignoscen-

da.

Type species: H. lugens (H. Bol.) Kraenzl.

376

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

Lip sessile or spathulate, the apical blade generally

lacerate, rarely entire, petals bidentate, lacerate or

acute.

This series contains nine species, all closely related.

If a species such as H. forcipata or H. hians is

selected as the starting point, several clear lines of

development of the petal and lip structures may be

detected. The lines all lead from an entire to a beard-

ed lip, and from a simple acute petal to a lacerate or

bifid petal apex. These postulated relationships are

graphically shown in Fig. 12.

I?1 Spur longer

g lip bearded

2 ^

in winter drought areas

Lip bearded

Flowers white

sepals acute

Coastal sand flats

North of False Bay

Lip lacerate

Petal apex expanded

From Agulhas to Cape Town

Fig. 12. — Relationships and putative evolution of Herschelia ser.

Lacerae.

The relationships of H. hians, H. venusta and H.

forcipata are not clear, and Hall (1973b) included the

former two in the same species. H. hians is well

established in the southern Cape Province, and the

majority of the collections do not show much varia-

tion; however, in some populations the lip margins

may vary from almost entire to deeply lacerate, as

opposed to the normal crenulate condition. H. for-

cipata is clearly very closely related to H. hians, but

differs in the shallowly bidentate spur and the long

floral bracts. It is quite possible that further

fieldwork may show that these forms also lie within

the range of variation of H. hians. The position of H.

venusta is somewhat clearer. The lip is irregularly

lacerated, and the petal somewhat expanded towards

the apex (Fig. 14). The distribution of this species is

puzzling, with records from the Caledon and Cape

Peninsula Divisions, and the Hogsback Mountains

near King William’s Town. Either the species has had

a polytopic origin, or it is merely an extreme form of

H. hians.

From this postulated basal group three lines of

development may be suggested:

(a) to H. barbata. This species is quite closely related

to H. venusta, especially to the forms of that species

occurring on the Cape Peninsula. Compared to H.

venusta, H. barbata shows many of the

characteristics of a more recently derived species: an

almost unique flower colour (white) in the genus, an

inflorescence with few large flowers (cf. Linder,

1981), and a small, probably neo-endemic distribu-

tion range. Until a few decades ago, the two species

overlapped on the Cape Peninsula, but with the ex-

tinction of the Cape Peninsula populations of H.

barbata, the interaction between the two taxa in the

overlap zone can no longer be investigated.

(b) to H. multifida. H. lugens and H. multifida are

closely related, sharing several peculiar characters:

lacerate petals, green bearded lips, a shortish conical

spur. Morphologically they can only be separated by

the length of the lip stalk. Within H. multifida there

is extensive variation in the length of the lip stalk, but

the great majority of the populations possess a lip

stalk, longer than 20 mm, whereas only a few popula-

tions on the fringes of the distribution range of the

species have shorter lip stalks. One collection from

the Cape flats, Schelpe 6313, possesses some flowers

typical of H. lugens (Fig. 14); others have a lip stalk

up to 10 mm long, approaching the type collection of

H. multifida. With the exception of the populations

around Grahamstown, H. lugens occurs on coastal

sandy flats, and H. multifida occurs in the montane

areas. H. lugens occasionally grows mixed with H.

hians ( Hall 664), but the relationship to this species is

not clear.

(c) to H. baurii. H. baurii is linked to H. hians via

H. venusta, and again the differences between the

species are not very clearcut. Morphologically the

South African populations of H. baurii may be

separated from H. venusta by the longer spur and the

more deeply dissected lip, whereas the northern

populations of H. baurii are distinguished by the

equally bidentate petals. Ecologically, H. baurii is

the tropical extension of a Cape genus, and it occurs

in areas of winter drought. H. baurii shows extensive

geographical variation, but in only two places does

this result in speciation: in the Chimanimani Moun-

tains (H. chimanimaniensis) and the Nyika Plateau

(H. praecox). Detailed analysis of populations may

result in the recognition of geographical subspecies

within the species, but at present the data for this are

inadequate.

7. Herschelia barbata (L./.) H. Bol. in J. Linn.

Soc., Bot. 19: 236 (1882); Kraenzl. in Orch. Gen. Sp.

1 : 804 (1900); Rolfe in FI. Cap. 5,3: 201 (1913). Type:

Cape of Good Hope, Sparrman s.n. (LINN, holo.!;

S!).

Orchis barbata L. f., Suppl. PI. 399 (1781). Satyrium barbatum

(L. f.) Thunb., Prod. 5 (1794). Disa barbata (L. f.) Swartz in Vet.

Acad. Handl. 21: 212 (1800); Lindl., Gen. Sp. Orch. 354 (1838);

Schltr. in Bot. Jb. 31: 286 (1901).

leones: H. Bol. leones Orch, Austro-Afr. 3: t. 51

(1913), as Disa barbata; Mason, Western Cape Sand-

veld Bowers, PL 34.2 (1972).

Plants 250-500 mm tall, the base of the stem often

with a sheath of fibrous leaf remains; tubers oval to

cylindrical, 20-40 mm long; radical leaves often

reaching the base of the inflorescence but never over-

topping the flowers, c. 1 mm wide, the base expanded

to clasp the stem, conduplicate, 4-7, sclerenchyma-

tous; cauline leaves completely sheathing, lanceolate,

H. P. LINDER

377

acuminate, 20-40 mm long, dry, grading into the

floral bracts; inflorescence lax, about 60 mm long

and with 2-6 flowers; bracts dry, half as long to as

long as the ovary, lanceolate, acuminate; ovaries

15-20 mm long, straight or slightly curved. Flowers

white to very pale blue, veins and lip more or less

blue, spur often green, no scent detected; dorsal sepal

angled forwards, galea acuminate, 15-25 mm long,

13-18 mm wide and 8-12 mm deep, ovate; spur con-

ical, not clearly distinct from the galea, usually ob-

tuse to shallowly bifid, rarely acute, straight or rarely

upcurved, 1-5 mm long; lateral sepals patent, nar-

rowly oblong, acute, suboblique, 15-25 mm long;

petals with the basal anticous lobe oblong, c. 2 mm in

diameter, flanking the stigma, usually decurrent with

the rest of the petal, the limb linear, straight or sub-

falcate, 5-6 mm long, apex obtriangular, incised to

bifid, 3—4 mm long and c. 3 mm wide; lip horizontal

at the base and soon decurved, ovate, deeply

lacerate, c. 15 mm long, the entire central part

lanceolate, c. 3 mm wide; anther 3 mm long with 2

distinct viscidia, often only partially separated, ellip-

tical to ovate, taller than wide; rostellum almost

equally trilobed, lobes awn-shaped, 1,5 mm long,

erect; stigma unequally tripulvinate, the odd lobe

much smaller than the lateral lobes, horizontal, fused

to the base of the rostellum. Fig. 14.

Diagnostic features. Lip deeply lacerated or bearded;

flowers white to pale blue with pale blue veins; inflor-

Fig. 13.— Distribution of Herschelia barbata.

Fig 14. — Herschelia barbata (from Linder 1806): 1, tlower, x 1,5;

2, column and petal, x5. Herschelia hians (from Linder

1731): 3, flower, x 1,5; 4, column and petal, x 6; 5, lip, x 3.

Herschelia venusta (from Bolus 17494): 6, lip, x3; 7, petal,

x3. Herschelia lugens: 8, flower with one lateral sepal re-

moved, x 1,5, from Schelpe 6313; 9, column and petal, x3,

from Jacot-Guillarinod s.n.

escence with 2-6 flowers; dorsal sepal acuminate,

15-25 mm long.

Flowering time: October and November.

Very local and rare in damp sandy localities on the

Cape Flats and sandy coastal flats to the north of the

Cape Peninsula, mostly extinct (Fig. 13).

Cape— 3318 (Cape Town): Mamre Road, 150 mm (-BC),

November 1978, Linder 1806 (BOL); sand dunes on the Cape

Peninsula, 30 m, (-CD), October 1883, MacOwan & Bolus 166

(BM; BOL; K; P; W; ZT).

This species is at present known from a single

locality near Darling. Formerly it appears to have

been relatively common on the Cape Flats, where it

was frequently collected: the populations in these

localities now all appear to be extinct. It occurs in

damp to wet areas on Recent Sand, near sea level.

The rainfall in these areas occurs almost wholly in the

winter months, and is approximately 400-600 mm

p.a.

This species is closely related to H. venusta, from

which it may be distinguished by the more acuminate

sepals and the white flowers. Formerly the two taxa

overlapped on the Cape Peninsula, and both were

recorded from the Kenilworth Race Course. Due to

human activities during the last decades, the taxa are

now allopatric.

378

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

8. Herschelia venusta ( H . Bol.) Kraenzl. , Orch.

Gen. Sp. 1: 805 (1900); Rolfe in FI. Cap. 5,3: 202

(1913), excl. syn. Type: Cape Province, Cape Flats,

Bolus 4556 (BOL, lecto.!; K! ; PRE!).

Disa venusta H. Bol. in J. Linn. Soc., Bot. 20: 482 (1884).

D. lacera Swartz, Schltr. in Bot. Jb. 31: 287 (1901), pro parte.

leones: H. Bol., leones Orch. Austro-Afr. 3: t. 52

(1913), as D. lacera; Flower. PI. Afr. 6: t. 234 (1926),

as D. lacera.

Plants 300-600 mm tall; tubers up to 40 mm long;

base of the stems often with a sheath of old leaf

fibres; radical leaves 6-12, reaching to the base of

the inflorescence or shorter, less than 2 mm wide,

sclerophyllous with the veins prominent, rigid-erect;

cauline leaves lax, acuminate, 15-40 mm long,

brown; inflorescence lax with 2-5-12 flowers; bracts

dry, as long as the ovary or half as long as the ovary,

ovate, acuminate; ovaries 15-20 mm long, straight

or curved. Flowers blue; dorsal sepal erect,

acuminate, galea 10-18 mm tall and 4-8 mm deep;

spur horizontal from the base of the galea, conical,

acute, 1,5-3 mm long; lateral sepals patent, oblong

to narrowly oblong, obtuse to acute, 12-16 mm long;

petals with small basal anticous lobes, oblong, 1,5

mm long, the limb linear, falcate, 6 mm long, apic-

ally expanded to form a fan up to 3 mm wide or un-

equally and often obliquely bilobed; lip more or less

ovate in outline, variably lacerate, shorter than the

lateral sepals, curved downwards, 7-12 mm long;

anther horizontal, 2 mm long, possibly with the two

viscidia fused; rostellum with 3 equal narrowly

lanceolate lobes, 1 mm tall; stigma unequally 3-

lobed, 2 mm wide and 1 mm tall, horizontal. Fig. 14.

Diagnostic features. Flowers with the lateral sepals

12-16 mm long; spur 1,5-3 mm long; lip variably

lacerate; petals with the basal anticous lobe less than

2 mm long, the limb linear, falcate with an expanded

apex or strongly unequally bifid.

Flowering time: October — January.

Local along the coast of the Cape Peninsula and

the Caledon Divisions, growing probably in sandy or

damp localities, rare in the Hogsback Mountains

near King William’s Town (Fig. 15).

Cape — 318 (Cape Town): Kenilworth, 30 m (-CD), Bolus s.n.

(BOL). 3419 (Caledon): Hermanus Flower Show exhibit (-AD),

Specimen in BOL 17494 (BOL); Vogelgat (-AD), Schlechter 9544

(BOL; BR; PRE). 3227 (Stutterheim): Hogsback Mountain

(-CA), Rattray in BOL 15 770 (BOL).

1 have not seen this species in the wild. It occurs

from the Cape Peninsula to Hermanus on the Cale-

don Coast, and near King William’s Town. Collec-

tor’s notes indicate damp habitats, generally under

sandy conditions. The taxon is possibly extinct on the

Cape Peninsula, where it has only been recorded

18 20 22 1 26 28

Fig 15.— Distribution of Herschelia venusta (open circles) and

H. hians (closed circles).

from the Kenilworth Race Course and from Steen-

berg. However, it appears to be locally common in

the Betty’s Bay and Hermanus areas, only flowering

after fire.

There are two collections from the Hogsback

Mountains near King William’s Town which have to

be referred to this species. However, there are numer-

ous slight differences, difficult to quantify, which

raise the possibility that the Hogsback population

may have been derived independently from H. hians.

This species has previously been included in H.

hians, as the lip shape was used as the sole differen-

tiating character. Although the lip shape ip H.

venusta is somewhat different from that in H. hians,

the extensive variation that may be found in a single

population in both species is convincing evidence that

this character may not be sufficiently reliable for

specific delimitation. H. venusta may also be dif-

ferentiated from H. hians by the shorter spur, the

much more slender petal with a more slender limb, a

smaller basal anticous lobe and the enlarged bifid

apex, and by the shallower galea. On this evidence, it

is here regarded as distinct.

Bolus (1884) cited two syntypes. Bolus 4566 occurs

in three herbaria, and the material is in better condi-

tion, than Harvey 140. The former was therefore se-

lected as lectotype.

9. Herschelia lugens (H. Bol.) Kraenzl., Orch.

Gen. Sp. 1; 806 (1900); Rolfe in FL Cap. 5,3: 203

(1913). Types: Cape Province, near Cape Town,

Bolus 3810 (BOL, lecto.!; K!); near Kuils River,

Pappe 377 (BOL!; SAM!), Pappe 39 (BOL!; K!;

SAM!), Ecklon 1566 (G!; S!; SAM!; W!).

leones: H. Bol., leones Orch. Austro-Afr. 2: t. 76

(1911); Curtis’s bot. Mag. t. 8415 (1912); Rice, Wild

Flow. Cape G. H. 148.2 (1950).

Plants 450-1000 mm tall; tubers 2-3, c. 50 mm

long and 20 mm wide: base of the stem often with a

sheath of old leaf fibres; radical leaves 8-15, from Vi

to 2A as long as the shoot, never overtopping the

lowest flower in the spike, c. 2 mm wide, rigid erect,

the veins sclerified and prominent; cauline leaves lax,

acuminate, 60-20 mm long with the longest at the

base of the stem, dry, grading into the floral bracts;

inflorescence lax, 70-150-400 mm long and with

5-10-25 flowers; bracts about half as long as the

ovaries broadly ovate, acuminate to setaceous, dry;

ovaries slender, slightly curved, 15-25 mm long.

Flowers with a cream-green galea, mauve lateral

sepals, a green to grey-green lip and white petals, but

with some variation in the colour rarely almost black;

dorsal sepal erect, galea obtuse, 12-16 mm tall and

about 10 mm deep, narrowly ovate; spur from the

base of the galea, generally slender, cylindrical,

straight or curved upwards, 1-5 mm long, sometimes

conical; lateral sepals patent, narrowly oblong, ob-

tuse or acute, 8-13 mm long, subconduplicate in the

apical half; petals with the basal anticous lobe

oblong, c. 4 mm long, parallel to the stigma; limb

linear, 10-15 mm long, with a geniculate bend

through c. 90° near the middle, the distal half of the

petal varying from obliquely obtriangulate to lorate,

acute or deeply lacerate; lip ovate, c. 13-19 mm long,

deeply dissected and beard-like, horizontal at the

base but soon decurved, longer than the lateral

sepals; anther horizontal or semi-pendent, 3 mm

long, viscidia separate; rostellum equally 3-lobed, c.

1,5 mm tall; stigma unequally 3-lobed with the odd

lobe smaller than the lateral lobes, horizontal, 3 mm

in diameter. Fig. 14.

H. P. LINDER

379

Diagnostic features. Flowers medium-sized, lateral

sepals 8-13 mm long; lip bearded, longer than the

lateral sepals, greenish or almost black.

Flowering time: October to November (-March).

This species occurs locally along the coastal flats

between the Cape Peninsula and Port Elizabeth,

usually in sandy well-drained conditions, and around

Grahamstown in macchia on sandstone derived soils

(Fig. 16).

(a) var. lugens

Disa lugens H. Bol. in J. Linn. Soc. 20: 483 (1884); Schltr. in

Bot. Jb. 31: 288 (1901).

Flowers cream-green, lip usually much longer than

the lateral sepals, widespread from the Cape Penin-

sula to Grahamstown.

Cape. — 3318 (Cape Town): sand dunes near Cape Town, 30 m

(-CD), Nov. 1886, MacOwan & Bolus 494 (BOL; P; Z; ZT). 3421

(Riversdale): Yzervarkfontein (- BC), May 1913, Muir 908 (BOL;

PRE). 3325 (Port Elizabeth): marshy flats near Van Staadens

Railway Station, 300 m (-CC), Oct. 1958, Hall 664 (BOL). 3326

(Grahamstown): mountains near Grahamstown (-BC), MacOwan

700 ( GRA; SAM).

This variety has a wide distribution, occurring

rather locally on the coastal flats. The only inland

locality is in macchia on the hills surrounding

Grahamstown. In most localities and all localities

south-west of Knysna, it occurs on deep sand, and

usually grows inside restioid tussocks. A population

studied near Grahamstown occurred on soils derived

from quartzite. The majority of the populations

occur in well-drained conditions, under a rainfall

regime that varies from 500 to 1 000 mm p.a., in

some areas evenly distributed over the whole year,

and in others with a period of summer drought.

The association between plants of this species and

restioid tussocks might be a result of predation, as

plants under cultivation do not require restioid

tussocks for normal growth.

This variety is becoming rare in the western Cape,

owing to the spread of urban development in the

coastal areas, and the spread of alien vegetation into

the remaining habitats. It may already be extinct on

the Cape Flats.

It was for some time confused with Herschelia bar-

bata, because of the superficial similarity of the two

taxa, especially when the plants are dried. Bolus

(1884) cited several syntypes in his protologue, in-

dicating that the material was known for some time

without being recognized.

(b) var. nigrescens Linder, var. nov., a var.

lugenti floribus nigrescentibus differt.

Flowers purplish black, recorded only from the

coast at Cape St Francis, near Humansdorp.

Type. — Cape, 3424 (Humansdorp): near Oyster

Bay in the vicinity of White Point (-BA), Muller s.n.

(NBG, holo.!).

This almost black variety of H. lugens was only

discovered in 1979 in the Humansdorp area, and as

yet not much information is available about it. It

appears to grow in the vicinity of the common H.

lugens var. lugens, but may be distinguished by the

almost black flowers, which also appear to be some-

what larger than is common for the typical variety. It

is desirable to recognize this form formally, to pre-

vent any confusion should the taxon be introduced

into cultivation.

Information on the genetic status of the variety

would doubtlessly give greater certainty about the

rank of this taxon.

The two varieties grow sympatrically. Leaves and

underground parts of the new variety were not seen.

10. Herschelia multifida ( Lindl .) Rolfe in FI.

Cap. 5, 3: 206 (1913). Type: Cape Province, Clan-

william District, Blouberg, Drege 3577b (K, holo.).

Disa multifida Lindl., Gen. Sp. Orch. 353 (1838); Schltr. in Bot.

Jb. 31: 285 (1901).

D. charpentieriana Reichb. f. in Linnaea 20: 668 (1847); Schltr.

in Bot. Jb. 31: 285 (1901). Herschelia charpentieriana (Reichb. f.)

Kraenzl., Orch. Gen. Sd. 1 : 807 (1900); Rolfe in FI. Cap. 5, 3: 206

(1913); Hall in Flower. PI. Afr. 42: t. 1673 (1973). Type: Cape of

Good Hope, Gueinzius s.n. (W, holo.!).

D. macroglottis Sond. ex Drege in Linnaea 20: 219(1 847), nom.

nud.

leones: H. Bol., leones Orch. Austro-Afr. 2: t. 77

(1911), as Disa charpentieriana; Rice, Wild Flow.

Cape G. H. 1771 (1950), as Herschelia charpentier-

iana; Flower. PI. Afr. 42: t. 1673 (1973), as H. char-

pentieriana.

Plants 400-600 mm tall; tubers up to 40 mm long;

base of the stem often with a sheath of old leaf fibres;

radical leaves 10-20, usually about half as long as the

stem and rarely reaching to the base of the in-

florescence, c. 1 mm wide, rigid with scleren-

chymatous veins; cauline leaves completely sheath-

ing, 20-50 mm long, lanceolate, acuminate, lax,

grading into the floral bracts; inflorescence lax,

40-100 mm long and with 3-8 flowers; bracts

Fig 16. — Distribution of Hersche-

lia lugens var. lugens (closed

circles) and var. nigrescens

(open circle).

380

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

usually 3A as long as the ovaries, broadly ovate, very

acuminate to setaceous, dry; ovaries 15-20 mm long;

straight or slightly curved. Flowers blue; with a green

lip, often with greenish veins and a brownish spur;

dorsal sepal erect, galea ovate, acuminate, 10-20 mm

tall and c. 6 mm deep; spur conical, subacute, rarely

cylindrical, obtuse or almost absent, generally curved

downwards, 1-4-6 mm long; lateral sepals patent,

lanceolate to narrowly ovate, acute, 10-15 mm long,

the apical halves deeply concave; petals with the

basal anticous lobe oblong, 3-4 mm long, the

margins entire or crenulate, rarely decurrent with the

limb, the limb falcate or rarely geniculate, linear,

7-10 mm long, the apex acute or narrowly ob-

triangular, occasionally lacerate; lip with a linear

(10—) 30-65-100 mm long limb, the blade narrowly

ovate, deeply lacerate; anther 3-4 mm long, viscidia

narrowly ovate, c. 1 mm long; rostellum equally

3-lobed, lobes lanceolate, 1 mm long; stigma with the

lateral lobes better developed than the odd lobe, on a

1 mm tall pedicel, 3,5 mm in diameter. Fig. 17.

Diagnostic features. Lip dissected, with a claw at

least twice as long as the dissected blade and usually

much longer.

Flowering time: November and December.

Herschelia multifida is a very distinct species that

may always be recognized by its peculiar lip with a

Fig 17. — Herschelia multifida, from Linder 1642. 1, flower,

x 1,5. 2, petal, x3. 3, column, x3.

long claw. It is rather widespread in the mountains of

the Cape Floral Region, from Prince Albert in the

south-east to the Vanrynsdorp Bokkeveld in the

north-west (Fig. 18).

Cape. — 3219 (Wuppertal): Cedarberg, Nov. 1913, Pattison in

BOL 14455 (BOL). 3319 (Worcester): Agterwitzenberg Vlakte

(-AA), Nov. 1967, Powrie 168 (BOL); Darling Bridge (-CA),

Nov. 1940, Esterhuysen 3802 (BOL). 3419 (Caledon): base of

mountains at Highlands Forest, Elgin (-AA), Nov. 1965, Oliver in

STE 29974 (PRE; STE). 3322 (Oudtshoorn): Swartberg Pass,

slopes of Krevasberg, 1 200 m (-AC), Dec. 1942, Stokoe 8679

(BOL).

This species occurs on slightly damp mountain

sides: ‘swampy slope’ (Esterhuysen 20907) on the

Roodeberg near Ceres; well-drained soil near a

stream on the Piketberg ( Linder 1642), a slight

seepage on a mountain side in the Skurfteberge near

Ceres ( Linder 1656) and a dry, well-drained stony

mountain side, in the mist belt, on the Swartberg at

Prince Albert ( Linder 1743). Many of the popula-

tions receive snow almost every winter. Rainfall is

also concentrated in the winter months, and is about

1 000 mm p.a. The altitude range of the species is

from 300 to 1 500 m.

There is considerable variation in the lip length in

this species: this is the basis on which H. charpen-

tieriana and FI. multifida had been separated

(Schlechter, 1901; Rolfe, 1913). An analysis of the

distribution of the short-spurred forms indicated that

these collections all exist at the margin of the

distribution area of the species as defined here, i.e.

on the Swartberg above Prince Albert and in the

Cedarberg and northwards to the Vanrhynsdorp

Bokkeveld. It is suggested that the short-spurred

form is a variation that occurs in the marginal areas

of the species and cannot be formally separated.

Although the type collection of H. multifida

should be at Kew, I have not seen it. There is a

photograph of the type at the Bolus Herbarium and

from this the nature of Lindley’s Disa multifida is

clear. There is some uncertainty about the type locali-

ty, which Drege indicates as ‘Blaauwberg’. It appears

to be between Boskloof and Heuningvlei, near Clan-

william in the northern Cedarberg.

H. P. LINDER

381

Disa macroglottis Sond. ex Drege has to be treated

as a nomen nudum, as the name, listed with a

reference to an Ecklon & Zeyher collection, was

published without a description.

11. Herschelia hians ( L . /.) Hall in Flower. PI.

Afr. 42: t. 1674 (1973). Type: Cape of Good Hope,

Thunberg s.n. (LINN, holo.!; UPS!).

Satyrium hians L. f., Suppl. PI. 401 (1781). Limodorum hians

(L. f.) Thunb., Prod. 3 (1793). Disa hians (L. f.) Spreng., Linn.

Syst. Veg. 3: 698 (1826). Eulophia hians (L. f.) Spreng., Linn.

Syst. Veg. 3: 720 (1826); Rolfe in FI. Cap. 5,3: 32 (1912).

Disa excelsa sensu Lindl., Gen. Sp. Orch. 356 (1838), non

(Thunb.) Swartz; H. Bol. in J. Linn. Soc., Bot. 25: 203 (1889);

Schltr. in Bot. Jb. 31: 292 (1901). Herschelia excelsa sensu Rolfe,

in FI. Cap. 5,3: 200 (1913), non (Thunb.) Rolfe.

Disa lacera Swartz in Vet. Acad. Handl. 31: 212 (1800); Lindl.,

Gen. Sp. Orch.: 354 (1838); H. Bolus in J. Linn. Soc., Bot 25: 202

(1889); Kraenzl., Orch. Gen. Sp. 1: 797 (1900); Schltr. in Bot. Jb.

31: 287 (1901). Herschelia lacera (Swartz) Fourc. in Trans. R. Soc.

S. Afr. 21: 81 (1932). Type: Cape of Good Hope, Sparrman s.n.

(W, holo.!).

Disa outeniquensis Schltr. in Ann. Transv. Mus. 10: 246 (1924).

Type: Cape Province, Mossel Bay District, Robinson Pass, Bolus

12327 (BOL, lecto.!).

D. lacera Swartz var. multifida N. E. Br. in Gdnrs’ Chron. 1 888

2: 664 (1888). Iconotype: Gdnrs’ Chron. 1888, 2, fig. 93 (1888).

leones: Curtis’s bot. Mag. 115: t. 7066 (1889);

Flower. PI. Afr. 42: t. 1674 (1973).

Plants 400-600 mm tall; tubers up to 20 mm long;

base of the stem often with a sheath of old leaf fibres;

radical leaves 8-13, often hysteranthous, reaching up

to the base of the inflorescence or shorter, less than 2

mm wide, sclerophyllous with prominent veins, semi-

rigid erect; cauline leaves completely sheathing, lax,

acuminate, brown, 20-40 mm long; inflorescence

lax, up to 200 mm long and with 3-16 flowers; bracts

from half as long to as long as the ovaries, ovate to

broadly ovate, acuminate to setaceous, dry; ovaries

15-20 mm long, slender, usually curved. Flowers

varying in colour from very pale blue to purplish

blue, the lip frequently darker coloured than the

sepals; dorsal sepal erect, obtuse to more commonly

acute, galea 10-15 mm long and 8-10 mm deep,

ovate to broadly ovate; spur horizontally from the

base of the galea, at length decurved, conical, taper-

ing to a subacute apex, 4-6 mm long; lateral sepals

patent, oblong-narrowly ovate to rarely narrowly

oblong, usually rounded, rarely acute, the apical part

subconduplicate, 8—12 mm long; petals with the

basal anticous lobe broadly oblong, rounded or trun-

cate, the margins entire or shallowly serrulate,

subdecurrent with the limb, 3-4 mm long, the limb

lorate, 7-10 mm long, the apex acute or rarely with

the small tooth to the rear, curved falcately or

geniculately upwards behind the anther; lip patent

with the margins curved upwards, narrowly oblong,

oblong or broadly elliptical, rarely broadly auriculate

at the base, the margins rarely entire, the distal

margin usually more or less crenulate to rarely

lacerate, 7—12 long; anther horizontal, 2-3 mm long,

viscidia separate; rostellum with 3 equal narrowly

lanceolate lobes, 1 mm tall; stigma unequally 3 lobed

with the odd lobe smaller than the lateral lobes,

horizontal, 2-2,5 mm in diameter and on a 1 mm tall

pedicel. Fig. 14.

Diagnostic features. Flowers with the lateral sepals

8-12 mm long; lip narrowly oblong to broadly ellip-

tical, the margins entire or more commonly with the

distal margin crenulate; petals with the limb lorate,

the apex acute; spur 4—6 mm long.

Flowering time: December to January (-February).

Widespread and fairly common in the southern

Cape Province on the coastal flats and the first range

of mountains, occurring mostly on well-drained and

often stony sandstone mountain slopes and flats (Fig.

15).

Cape. — 3419 (Caledon): Hermanus (-AC), Jan. 1920, Burn

Davy 18483 (BOL). 3321 (Ladismith): Garcias Pass (— CC), Dec.

1977, Linder 1714 (BOL). 3322 (Oudtshoorn): mountains near

George (-CD), Jan. 1897, Bolus 13514 (BOL). 3422 (Mosselbay):

near Sedgefield, near Knysna, 30 m (-BB), Oct. 1963, Chater s.n.

(BOL; PRE; STE). 3323 (Willowmore): Outeniqua mountains

near Joubertina (-DD), Jan. 1947, Esterhuysen 13599 (BOL; K;

PRE). 3325 (Port Elizabeth): Port Elizabeth (-DC), Feb. 1889,

Galpin 391 (PRE).

This species forms open scattered populations

throughout the southern Cape Province, generally on

well-drained and often stony soils. The rainfall in the

area is almost evenly distributed over the whole year,

and the total rainfall ranges from 600 to over 1 000

mm p.a. The altitude range of the species if from 80

m to over 1 000 m.

Extensive variation in flower colour and lip lacera-

tion occurs. The majority of populations investigated

showed extraordinary colour variation, from pale

sky-blue to deep purple-blue flowers. Colour varia-

tion commonly occurs within populations, but lip

variation tends to be rather between populations. In

the Knysna area the lip is occasionally entire. Despite

extensive searching, I have failed to locate any such

populations, but occasional herbarium specimens

with entire lips are known from the Knysna area.

Over the rest of the range the distal margin of the lip

tends to be shallowly crenulate. In the Langeberg

near Riversdale the lip frequently shows two broad

auricles at the base, and the apex is variably shallow-

ly lancerate (type of D. lacera Swartz agrees with this

lip shape) ( Linder 1714). Some collections from near

Port Elizabeth also show lips that are variably

lacerate (e.g. Hall 1160).

Hall (1973b) partially resolved the nomenclatural

confusion surrounding the correct name for this tax-

on. Satyrium hians L. f. [ = Limodorum hians (L. f.)

Thunb.] was transferred to Eulophia by Sprengel and

was mistaken for Eulophia clavicornis by subsequent

botanists (Lindley, 1838; Bolus, 1889; Rolfe, 1912).

Sprengel cited Thunberg’s Limodorum hians as

basionym. However, Sprengel had also transferred

the epithet to Disa, citing Satyrium hians L. f. as

basionym.

Thunberg (1794) proposed Satyrium excelsum as a

nomen novum for Orchis tripetaloides L. f. As

Orchis tripetaloides is a perfectly good name, S.

excelsum has to be regarded as a superfluous name.

Swartz (1800) transferred the epithet to Disa. Lindley

(1838) noted that Thunberg’s descriptions of D.

excelsa (Thunb.) Swartz did not agree with the type

of O. tripetaloides and maintained it as a distinct

species. However, since the name is derived from a

superfluous name, Disa excelsa sensu Lindley has to

be regarded as a illegitimate name. The same then

applies to Herschelia excelsa (Thunb.) Rolfe.

Disa lacera has generally been regarded as the same

species as D. venusta. However, a study of the type

found in Vienna showed that D. lacera agrees more

with the form of H. hians found in the Langeberg at

Riversdale than with D. venusta. Swartz’s pro-

tologue, describing the lip as oblong, substantiates

this finding, and D. lacera is included as a synonym

under H. hians. The var. multifida which N. E.

Brown described appears to be a more lacerate ver-

sion of D. hians, of which a few collections from the

Knysna area had been recorded. It does not show the

characters of H. venusta.

382

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

Disa outeniquensis of Schlechter agrees in all

characters with the most common form of H. hians.

12. Herschelia forcipata ( Schltr .) Kraenzl.,

Orch. Gen. Sp. 1: 807 (1900); Rolfe in FI. Cap. 5, 3:

200 (1913). Type: Cape, Trimen s.n. (BM, holo.!).

Disa forcipata Schltr. in Bot. Jb. 24: 428 (1897); in Bot. Jb. 31 :

292 (1901).

Plant 600 mm tall; radical leaves and tubers not

known; cauline leaves closely sheathing, acuminate,

subimbricate, 40-50 mm long, dry; inflorescence

laxly 25-flowered, 200 mm tall, cylindrical; bracts as

long as the ovaries, lanceolate, acuminate, dry;

ovaries slender, generally straight, c. 20 mm long.

Flowers greenish yellow; dorsal sepal acute, galea

broadly ovate, c. 13 mm long and 6 mm deep; spur

horizontal from the base of the galea, slender cylin-

drical, straight, 3-4 mm long, the apex bifid; petals

with a smaller orbicular basal anticous lobe, decur-

rent with the limb, 2 mm in diameter, the limb lorate,

geniculately curved up through 90° behind the

anther, c. 10 mm long, the apex unequally bifid, the

lobes acute, the anterior lobe longer than the

posterior lobe; lip lanceolate, acute, the margins

entire, c. 10 mm long; anther reflexed width the con-

nective shorter than the cells; rostellum erect, trifid,

the lobes acute; stigma pulvinate.

Diagnostic features. Lip lanceolate, acute, entire;

petals with the apices bilobed, spur horizontal,

slender, bifid.

This species is known from the single plant in

Trimen’s collection, made somewhere in the Cape in

1870, probably in the Knysna area (Schelpe pers.

com). It is a very distinct taxon by virtue of the

bilobed petals, the entire lip, the peculiar spur and

rather dense inflorescence. It appears to be related to

FI. hians, rather than to the H. graminifolia group,

as Schlechter (1897, 1901) and Rolfe (1913) sug-

gested. The greenish yellow colour indicated by the

collector, however, is reminiscent of H. lugens.

13. Herschelia praecox Linder, sp. nov., a H.

baurii (H. Bol.) Kraenzl. labello fere integro, petalis

pro ratione majoribus differt. Crescit in graminosis

montis Nyikae Malaworum.

Type. — Zambia, Northern Province, Nyika

Plateau, Sept. 1967, Williamson 312 (K, holo.!).

Plants 200-400 mm tall; tubers ovate, about 30

mm long; stems usually with a basal sheath of fibrous

leaf remains; radical leaves produced after flowering,

about 6, c. 300 mm long and 1-2 mm wide, semi-

erect, subsclerophyllous; cauline leaves lax, acumi-

nate, completely sheathing, c. 20 mm long; inflores-

cence laxly 2-10 flowered and 40-130 mm long;

bracts ovate, acuminate, dry, about 10 mm long;

ovaries 10-15 mm long, usually curved. Flowers

white to blue or dark mauve, occasionally the apices

of the petals green; dorsal sepal erect, galea

acuminate with the apex reflexed, ovate, 10-12 mm

tall, c. 8 mm wide and 4-6 mm deep; spur horizontal

from the base of the galea, often gradually ascend-

ing, cylindrical to laterally flattened, rounded, c. 2

mm in diameter and 3-5 mm long; lateral sepals

patent, narrowly ovate to lanceolate, subacuminate,

10-12 mm long, shallowly concave; petals with an

ovate basal anticous lobe, c. 2 mm in diameter, the

limb lorate, falcately curved up next to the anther in-

side the galea, c. 1 2 mm long and 1 ,5—2 mm wide, the

Fig. 19. — Distribution of Herschelia praecox (open circle), H.

goetzeana (solid circle) and H. chimanimaniensis (half-solid

circle).

apex widened and unequally acutely bilobed with the

anterior lobe longer than the posterior lobe; lip nar-

rowly ovate to lanceolate, 11-13 mm long, margins

varying from almost entire to shallowly fimbriate,

curved upwards; anther horizontal, c. 3 mm long, the

connective longer than the anther; rostellum with 1,5

mm tall canaliculate erect lateral lobes; stigma

subsessile, somewhat angled forwards. Fig. 20.

Diagnostic features. A short plant, flowering in

September, shortly after the winter grass fires, re-

Fig 20. — Herschelia praecox. 1, habit, x0,5. 2, flower with the

lateral sepals and the^lip removed, x6. 3, column, x8. 4,

lateral sepal, x6, 5, petal, x6. 6, lip, x6. I from Tyrerllb,

2-6 from Williamson 312.

H. P. LINDER

383

stricted to the montane grasslands on the Nyika

Plateau, Malawi and Zambia.

Malawi — Northern Province, Rumphi District, Nyika Plateau,

September 1962, Tyrer 966 (BM; SRGH).

Zambia. — Northern Province, Lundazi District, Nyika Plateau,

September 1964, Robinson 6259; September 1968, Williamson

1023 (K; SRGH).

The epithet ‘praecox’ refers to the early flowering

time, shortly after the winter fires.

H. praecox is very similar to H. baurii, and if it

were compared with all the variations of the latter

taxon over its whole range, it might be found dif-

ficult to maintain as a distinct taxon. The entire

distribution area of H. praecox is included in that of

H. baurii. In the overlap region, on the Nyika

Plateau, the flowers of H. baurii are larger than

anywhere in the rest of its range (Fig. 22), therefore

creating a size differential from H. praecox. There is

also a shift in flowering time: on the plateau, H.

baurii flowers in October and November, whereas H.

praecox flowers in September. This difference was

also noted by Williamson (1977). Morphologically,

the lack of deep dissection of the lip and the large

petals provide substantiation for what is clearly a

biological micro-species.

H. praecox grows in well-drained short montane

grassland on the Nyika Plateau (Vesey-Fitsgerald,

1963; Chapman & White, 1970), under a rainfall

regime of 1 000-2 000 mm p.a., mostly concen-

trated during the summer months. The winter

months are dry, with occasional frost and frequent

grass fires. H. praecox appears to flower after the

fires, possibly before the first rains and before the

grasses grow tall, therefore the flower spikes do not

have to compete with the grasses and are consequent-

ly short. Leaves are only produced when flowering is

completed, possibly after the rains start. As such,

this species occupies the temporal niche before that

of H. baurii.

14. Herschelia baurii (H. Bol.) Kraenzl., Orch.

Gen. Sp. 1: 804 (1900); Rolfe in FI. Cap. 5, 3: 204

(1913. Type: Mt Baziya, Baur 814 (K, holo.!).

Disa baurii H. Bol. in J. Linn. Soc., Bot. 25: 174(1889); Schltr.

in Bot. Jb. 31: 289 (1901).

D. hamatopetala Rendle in Trans. Linn. Soc., 2, 4: 47 (1894); N.

E. Br. in FI. Trop. Afr. 7: 286 (1898); Summerh. in FI. Trop. E.

Afr. 156: 177 (1968). Herschelia hamatopetala (Rendle) Kraenzl.,

Orch. Gen. Sp. 1: 803 (1900). Type: Malawi, Mt Mlanje, Whyte

s.n. (K, lecto.l).

Herschelia bachmanniana Kraenzl., Orch. Gen. Sp. L q0S

(1900). Type: Transkei, East Pondoland, Bachmann 414 (?Bt).

Disa longilabris Schltr. in Bot. Jb. 38: 150 (1907). Herschelia

longilabris (Schltr.) Rolfe in Orch. Rev. 27: 9 (1919). Type:

Tanzania, Kinga Mountains, slopes of Mt Buongwe, Goetze 1222,

1226 (B, holo. t: Z!).

leones: H. M. L. Bol. in Ann. Bolus Herb. 4: PI. 11

(1926); Williamson, The orchids of south-central

Africa, PI. 71 (1977).

Plants 200-400 mm tall; tubers 20-30 mm long;

base of the stem frequently with a thick sheath of old

leaf fibres; radical leaves 5-10, produced after

flowering, frequently overtopping the spike, up to

300 mm long and less than 2 mm wide, semirigid and

subsclerophyllous; cauline leaves lax, completely

sheathing, acuminate, 15-25 mm long, larger

towards the base of the stem; inflorescence lax, rarely

subsecund, with 2-14 flowers; bracts usually about

Vi as long as the ovaries, rarely as long as the ovaries,

ovate, acuminate, dry; ovaries straight or curved,

10-15 mm long. Flowers varying from pale sky-blue

to deep purple-blue, the lip frequently a darker blue

than the sepals; dorsal sepal erect, galea ovate, acute,

(8-) 10-20 mm tall, 6-12 mm wide and 5-10 mm

deep; spur horizontal from the base of the galea,

often somewhat ascending, cylindrical, rounded, 4-6

mm long; lateral sepals patent, oblong to rarely

lanceolate, acute to rounded, (8-) 10—18 mm long,

shallowly concave; petals with the basal anticous lobe

oblong to ovate, 1-2,5 mm in diameter, rounded, the

margin rarely crenulate, limb linear, 8-13 mm long,

the apex variously expanded, deeply bifid to lacerate

to acute; lip horizontal, at least at the base, broadly

to narrowly ovate, 10-25 mm long, more or less

deeply dissected; anther horizontal, 2-5 mm long,

viscidia separate, ovate; rostellum erect, 2-3 mm

tall, equally 3-lobed; stigma horizontal, 1 mm tall

and 1,5-2, 5 mm in diameter. Fig. 22.

Diagnostic features. Flowers with the lateral sepals

10-18 mm long; lip deeply and unevenly lacerate,

generally longer than the lateral sepals; petals with

usually a bilobed apex; spur generally cylindrical and

obtuse,

Flowering time: mostly September and October, but

with occasional collections from almost every month

of the year.

A variable plant with hysteranthous leaves which

occurs widespread in the montane grassland in

southern and south-central Africa (Fig. 21), flower-

ing at the beginning of the rainy season.

Transkei. — 3129 (Port St Johns): Msimkaba River mouth

(-BA), Aug. 1976, Venter & Vorster 195 (PRE).

Natal. — 2929 (Underberg): Kamberg (-BD), Oct. 1941,

Schelpe 013 (NU).

Fig. 21. — Distribution of Herschelia baurii.

384

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

Swaziland. — Sept. 1910, Steward 8875 (PRE).

Transvaal. — 2630 (Carolina): Brereton P.S. (-CD), Sept.

1945, Acocks 11731 (PRE). 2531 (Komatipoort): Saddleback

Mountains, Barberton (-CC), Sept. 1889, Galpin 427 (BOL,

PRE). 2330 (Tzaneen): Woodbush Mountains (-CC), Sept. 1927,

Moss 15432 (PRE).

Zimbabwe. — Melsetter District, Mt Musapa, 2100 m, Wild 3556

(K; SRGH). Inyanga District, World’s View, Sept. 1956, Davies

2120 (K; SRGH). Inyanga District, head on Nyamaziwa River,

1800 m Sept. 1965, Biegel 257 (SRGH).

Malawi. — Southern Province, Mulanje District, Mt Mlanje,

2 000 m, Oct. 1941, Greenway 6318 (K). Northern Province,

Rumphi District, Nyika Plateau, Oct. 1958, Robson 297 (K).

Tanzania. — Southern Highlands, Njombe District, Elton

Plateau, Nov. 1963, Richards 18500 (K); Mbeya District, summit

of Mt Mbeya, Oct. 1957, Watermeyer 167 (K).

Fig. 22. — Herschelia baurii, from Williamson 119. 1, flower with

the lateral sepal and the lip removed, x 1,5. 2, front view of

flower, x 1,5. 3, front view of column and petals, x3. 4,

petal, x3.

This species occurs in the grasslands of the Mon-

tane Region (White, 1978), between Grahamstown in

the eastern Cape Province and Kigoma in western

Tanzania. Robyns & Tournay (1955) also recorded

this species from the Ruwenzori Mountains in Zaire.

In general, the altitude range of these grasslands is

from 1 000 to 2 400 m, and they receive approx-

imately 1 000 to 1 500 m rainfall p.a., concentrated

in the summer months, usually with about four dry

months in winter (Jackson, 1961; Phipps & Goodier,

1962; Chapman & White, 1970). Vesey-Fitzgerald

(1963) notes that the grasslands are generally burnt in

the winter months, after which the geophytes flower.

During the rainy season the grasses grow rapidly, to

flower at the end of the rainy season. H. baurii

exploits the temporal niche at the beginning of the

rainy season, before the grasses grow tall. This lack

of height competition is manifested by the rather

short flowering spikes. The leaves are produced after

flowering is completed.

As might be expected from such a wide-ranging

species, there is extensive geographical variation.

Plants from Malawi tend to have much larger

flowers, and relatively shorter spurs, than plants

from Zimbabwe or South Africa (Fig. 23). In South

Africa the petals usually do not have deeply bifid

apices, but are narrowly obtriangulate towards the

apex, often obliquely bifid or lacerate, whilst over

the rest of the range the petals are deeply bifid. The

geographical variation in flower size is shown in Fig.

23. This variation has been used to maintain separate

taxa for South Africa and the areas north of South

Africa (Summerhayes, 1968). However, a detailed

study has shown that there is extensive overlap for all

characters, with collections from Zimbabwe fre-

quently intermediate between those from South

Africa and those from Malawi.

In south Tanzania a smaller-flowered form occurs,

which has been kept distinct as Disa longilabris by

Summerhayes (1968). However, it only appears to be

smaller in all its parts than H. baurii, and the

characters mentioned by Summerhayes (1968) were

found to be variable. However, there is little material

of D. longilabris and the available evidence indicates

that it is best treated as a depauperate form of H.

baurii.

The type material of H. bachmanniana could not

be traced. Kraenzlin (1900) notes that it is very

similar to H. baurii, but that it possesses a much

shorter spur, rounded sepals, dissected petals and

flattened lip fimbriae. However, the type specimen of

D. baurii is peculiar in that the spur is about twice as

long as is typical for the species and this probably

confused Kraenzlin. The other characters mentioned

by Kraenzlin occur in H. baurii.

15. Herschelia chimanimaniensis Linder, sp.

nov., a H. baurii (H. Bol.) Kraenzl. floribus

minoribus, sepalis lateralibus 6-8 mm longis differt.

7

T

1

6

5

4

3

a

(/)

3

2

8

• • • •

• 99

e

o

o o

o o o o

9

O

9 10 11 12 13 14 15 16 17 18

Lateral Sepal Length (mm)

Fig. 23. — Geographical variation

in flower size and spur length

in Herschelia baurii. Material

from South Africa plotted in

solid circles, from Zimbabwe

in half-solid circles and from

Malawi and Tanzania in open

circles.

H. P. LINDER

385

Crescit in montibus CFiimanimanorum Zimbabwe.

Type. — Zimbabwe, Chimanimani, Ball 577 (K,

holo. ! ; SRGH!).

Plants slender, 200 400 mm tall; base of the stem

occasionally with a sheath of the fibrous leaf re-

mains; radical leaves apparently produced after

flowering, 3-6, 1 50-200 mm long and 1 -2 mm wide,

subsclerophyllous; cauline leaves closely sheathing,

4- 8, 15-20 mm long, acuminate, the lower 2-3

much larger and imbricate; inflorescence laxly

2-8-flowered, 30-80 mm long; bracts ovate,

acuminate, about half as long as the ovary, dry;

ovaries 5-15 mm long. Flowers pink-mauve to white,

rarely blue; dorsal sepal erect, galea ovate, acute,

5- 7 mm tall, 4 mm wide and 4 mm deep; spur

horizontal from the base of the galea, often gradually

ascending, straight, cylindrical or somewhat laterally

flattened, rounded, 3-4 mm long; lateral sepals pa-

tent, narrowly oblong - ovate, acute, 6—8 mm long,

shallowly concave; petals with the basal anticous lobe

ovate, 1-1,5 mm in diameter, the limb lorate,

falcately curved up next to the anther, included in the

galea, c. 5,5 mm long and 1 mm wide, the apex acute-

ly bifid; lip ovate, 8-10 mm long, deeply dissected,

the margins curved upwards; anther horizontal, c.

1,5 mm long; rostellum lateral lobes canaliculate,

erect, 1 — 1,5 mm tall; stigma horizontal, c. 0,6 mm

tall, flat, c. 1 mm in diameter. Fig. 24.

Diagnostic features. Plants slender, flowers small

with the lateral sepals 6-8 mm long. Occurs in the

Chimanimani Mountains of Zimbabwe.

Flowering time: September and October.

A small-flowered and very slender plant which oc-

curs between about 1 500 and 1 800 m in the

Chimanimani Mountains in eastern Zimbabwe and

the neighbouring areas in Mozambique (Fig. 19).

Zimbabwe. — Melsetter District, Chimanimani Mountains, Oct.

1950, Munch 327 (K; SRGH); Grosvenor 178 (K; SRGH).

Mozambique. — Manica e Sofala District, Chimanimani Moun-

tains, August 1964, Whetlan 2145 (SRGH).

Fig. 24. — Herschelia chimanimaniensis, from Bull 577. 1, habit,

x 0,5. 2, flower with the lateral sepal removed, x 4. 3, flower,

x4. 4, front view of column and petals, x8. 5, lip, x4. 6,

lateral sepal, x4, petal, x4.

From collectors’ notes it appears as if the species

has a wide ecological range, as it has been recorded

from montane grassland, cliff faces, rocky ground,

seasonally damp ground and bogs. Phipps & Goodier

(1962) describe the climate as cool, with frequent

mist in the summer and frequent frost in the winter

months. Available records indicate that rainfall is

highly dependent on local topography and varies

from 800 to 3 000 mm p.a.

This new species is clearly closely related to H.

baurii. The ecological relationships of the two taxa in

the Chimanimani Mountains are not clear. The maj-

ority of collectors recognized the two taxa as being

distinct and collected them under different numbers.

It also appears as if H. baurii occurs at higher alti-

tudes than H. chimanimaniensis, but no explicit data

are available. H. chimanimaniensis may be distin-

guished by the smaller flowers and more slender

habit.

Ser. Ecalcaratae Linder, ser. nov., labello lacero,

calcari obsoleto dignoscenda. Flowers with a bearded

lip, spur obsolete.

Type species: H. goetzeana Kraenzl.

The single species in this series is most peculiar.

Although it is clearly related to ser. Lacerae in the

bearded lip and petal shape, it differs from H. baurii

in the absence of a spur and the acute entire petal. It

is difficult to explain the origin of this species. If it

cannot be derived from H. baurii, it must be seen as a

relic of a previous expansion of Herschelia from the

south.

16. Herschelia goetzeana Kraenzl. in Bot. Jb.

30: 286 (1901). Type: Tanzania, Njombe District,

Bulongwa, Goetze 925 (B, holo. t ).

Disa goetzeana (Kraenzl.) Schltr. in Bot. Jb. 38: 150 (1906), in

obs. non Kraenzl. (1900), nom. illegit. D. waiter i Schltr. in Bot.

Jb. 53: 544 (1915), nom. nov.; Summerh. in FI. Trop. E. Afr. 156:

176 (1968).

Plants 200-600 mm tall; tubers about 20 mm long;

base of the stem with a sheath of old leaf remains;

radical leaves 5-10, 150-250 mm long and 1-3 mm

wide, suberect, veins prominent, subsclerophyllous;

cauline leaves lax, acuminate, about 25 mm long,

completely sheathing; inflorescence closely

1-9-flowered; bracts ovate, acuminate, as long as the

ovary, dry; ovaries about 10 mm long. Flowers foam-

pink to crimson-cherry-coloured; dorsal sepal erect,

galea hemispherical, almost orbicular from the front,

8-14 mm tall and 10-16 mm wide; spur obsolete;

lateral sepal oblong, obtuse, spreading downwards,

10 mm long; petals with the basal anticous lobe

oblong, 2 mm long, limb narrowly oblong to lorate,

falcate, subacute, c. 7 mm long; lip patent, 20-25

mm long and up to 12 mm wide, entire at the base,

obscurely 3-lobed, deeply and finely lacerate; anther

horizontal, 7,5 mm long and with two distinct

viscidia; rostellum with 3 linear erect lobes; stigma

pulvinate, c. 2,5 mm in diameter.

Diagnostic features. Flowers with the lateral sepals c.

1 1 mm long; lip pectinate or lacerate; spur obsolete.

Flowering time: March.

Known from a single collection from southern

Tanzania (Fig. 18), where it is said to be restricted to

rock crevices on Mt Mbeya between 2 700 m and

2 900 m.

Tanzania — Southern Highlands, Mbeya District, Mbeya Peak,

2 700 m, March 1960, Kerfoot 1632 (K).

386

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHEL1A

Kerfoot (1964) describes the vegetation of the

Mbeya Range, and notes for the high-altitude litho-

phytic communities that cloud and mist occur fre-

quently. Growth is highly seasonal, with flowering

occurring between February and April, and the flora

is dominated by Orchidaceae.

Morphologically, this species is quite distinct, but

clearly related to the species in Ser. Lacerae It is

therefore, other than H. praecox and H. chimanima-

niensis, unlikely to be a neo-endemic. It is difficult to

understand where this species fits in with the rest of

the genus and in the classification proposed here it is

anomalous. It would be most interesting to see more

material of this species and to confirm its limited dis-

tribution.

The type collection has been lost, but from the des-

cription there cannot be any doubt about the identity

of the species. The name ‘goetzeana’ cannot be trans-

ferred to Disa, as there is already a Disa goetzeana

Kraenzl. Schlechter (1915) proposed D. walteri as a

nomen novum for the species. The name is derived

from Walter Goetze, the original collector of the

species.

ACKNOWLEDGEMENTS

I am indebted to the various people with whom I

have discussed this genus, and who criticized some of

my ideas. I would especially like to thank my super-

visor, Prof. E. A. Schelpe, for numerous comments

and much patience. I am grateful to the Directors

and Curators of the various herbaria where I was

provided with working facilities, or who loaned

material for study purposes. This research was done

while holding a Smuts Fellowship and a CSIR Post-

graduate bursary.

UITTREKSEL

Die genus Herschelia (Disinae, Orchidaceae ) is her-

sien. Sestien spesies, een subspesie en een varieteit

word erken. twee nuwe spesies uit tropiese Afrika

(H. chimanimaniensis Linder en H. praecox Linder

sowel as ’n nuwe varieteit ui die Kaap [H. lugens ( H .

Bolus) Kraenzl var. nigrescens Linder ] word hier be-

skryf. Drie nuwe kombinasies word gemaak deur die

oorplasing van die twee spesies van Forficaria en

Disa seksie Microperistera ( een spesie ) na Herschelia.

Dertien spesies word gei'l/ustreer en die nomenklatuur

en die beskikbare gegewens in verband met die groei-

plekke van die taksons word bespreek. Die spesies is

in twee subgenera gegroepeer waarvan een in twee

verdere seksies en vier series verdeel is. Hierdie klassi-

fikasie is gebaseer op die veronderstelde filogenie

soos bepaal deur die metode opgestel deur Wagner

(1962).

REFERENCES

Anonymous. 1905. Horticultural notes. J. Hort. 3, 51: 123.

Bentham, G. & Hooker, J. D., 1883. Genera Plantarum. London.

Bolus, H., 1882. Notes on some Cape orchids. J. Linn. Soc., Bot.

19: 233-238

Bolus, H., 1884. Contributions to South African botany. J. Linn.

Soc.. Bot, 20: 467-488,

Bolus, H., 1888. The orchids of the Cape Peninsula. Trans.

S. Afr. phil. Soc. 5: 1-201.

Bolus, H., 1889. Contributions to South African botany. Part IV.

J. Linn. Soc., Bot. 25: 156-209.

Bolus, H., 1893. leones Orchidearum Austro-Africanarum 1 (1).

London: Wesley.

Bolus, H., 1896. leones Orchidearum Austro-Africanarum 1(1).

London: Wesley.

Bolus, H., 1907. Contributions to the African flora. Trans. S.

Afr. phil. Soc. 16: 135-152.

Bolus, H., 1911. leones Orchidearum Austro-Africanarum 2.

London: Wesley.

Bolus, H., 1913. leones Orchidearum Austro-Africanarum 3.

London: Wesley.

Bolus, H. M. L., 1926a. Disa lacera. Flower. PI. Afr. 6: t. 234.

Bolus, H. M. L., 1926b. South African orchids. Ann. Bolus Herb.

4: 62-67.

Bolus, H. M. L., 1931. Disa forficaria. Flower. PI. Afr. 1 1: t. 415.

Bremer, K. & Wanntrop, H., 1978. Phylogenetic systematics in

botany. Taxon 27: 317-329.

Brown, N. E., 1888. Disa lacera Swartz and var. multifida N. E.

Br. (n. var.). Gdnrs’ Chron. 3: 664

Brown, N. E., 1889. Orchideae. In W. T. Thiselton-Dyer (ed.),

Flora of Tropical Africa 7. London: Reeve.

Davis, P. H. & FIeywood, V. H., 1963. Principles of angiosperm

taxonomy. Edinburgh: Oliver & Boyd.

Drege, J. F., 1847. Vergleichungen der von Ecklon und Zeyher

und von Drege gesammelten suedafrikanischen Pflanzen. Lin-

naea 20: 183-258.

Dyer, R. A., 1955. Herschelia graminifolia. Flower. PI. Afr. 30: t.

1172.

Dyer, R. A., 1976. The genera of Southern African flowering

plants. 2. Gymnosperms and monocotyledons. Pretoria:

Botanical Research Institute.

Durand, T. & Schinz, H., 1894. Conspectus florae Africae 5.

Brussels: Charles van de Weghe.

Fourcade, H. G., 1932, Contributions to the flora of the Knysna

and neighbouring divisions. Trans. R Soc. S. Afr. 21:

75-102.

Funk, V. A. & Stuessy, T. F., 1978. Cladistics for the practising

plant taxonomist. Syst. Bot. 3: 159-178.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships, and origins. Ann. Miss. Bot.

Card. 65: 369-436.

Hall, A. V., 1973a. Disa charpentieriana. Flower. PI. Afr. 42: t.

1673.

Hall, A. V., 1973b. Herschelia hians. Flower. PL Afr. 42: t. 1674.

Hennig, W., 1966. Phylogenetic systematics. Urbana: University

of Illinois Press.

Holmgren, P. K. & Keuken, W., 1974. Index Herbariorum, Part 1.

The herbaria of the world. Utrecht: Oosthoek, Scheltema &

Holkema.

Hooker, J, D., 1886. Disa atropurpurea. Curtis’s bot. Mag. t.

6891.

Hooker, J, D., 1889. Disa lacera var. multifida. Curtis’s bot.

Mag. t. 7066.

Jackson, S. P. (ed.), 1961. Climatological atlas of Africa. Lagos:

CCTA/CSA.

Judd, W. S., 1979. Generic relationships in the Andromedeae

(Ericaceae). J. Arnold Arbor. 60: 477-503.

Kerfoot, O., 1964. A preliminary account of the vegetation of the

Mbeya Range, Tanganyika. Kirkia 4: 191-205.

Kraenzlin, F., 1900. Orchidacearum Genera et Species. Berlin:

Mayer & Mueller.

Kraenzlin, F, 1901. Beitrage zur Flora von Africa. Bot. Jb. 30:

280-289.

Linder, H. P., 1981. Taxonomic studies in the Disinae. IV. A

revision of Disa Berg. sect. Micranthae Lindl. In Prepara-

tion.

Lindley, J., 1830-1840. The genera and species of orchidaceous

plants. London.

Linnaeus, C. (fil . ), 1781. Supplementum Plantarum. Brunsvigae.

Mason, H., 1972. Western Cape Sandveld flowers. Cape Town:

Struik.

Pfitzer. E., 1889. Orchidaceae. In A. Engler & K. Prantl (eds),

Die natiirlichen Pflanzenfamilien. Leipzig: Wilhelm

Engelmann.

Phipps, J. B. & Goodier, R., 1962. A preliminary account of the

plant ecology of the Chimanimani Mountains. J. Ecol. 50:

291-391.

Reichenbach, H. G., 1847. Orchidiographische Beitrage. Linnaea

20: 673-696.

Rendle, A. B., 1894. The plants of Mlanji, Nyasaland. Trans.

Linn. Soc. 2, 4: 1-68.

Rice, E. G. & Compton. R. H., 1950. Wild flowers of the Cape of

Good Hope. Cape Town: Botanical Society of South Africa.

Robyns, W. & Tournay, R., 1955. Flore des Spermatophytes du

Parc National Albert. III. Monocotylees. Bruxelles.

Rolfe, R. A., 1912. Disa lugens. Curtis’s bot. Mag. t. 8415.

Rolfe, R. A., 1912- 1913. Orchidaceae. In W. T. Thiselton-Dyer

(ed.) Flora Capensis 5,3. London: Reeve.

Schelpe, E. A., 1966. An introduction to the South African

orchids. Cape Town: Purnell.

Schlechter, R., 1897. Orchidaceae africanae novae vel minus

cognitae. Bot. Jb. 24: 418-433.

Schlechter, R., 1901. Monographic der Diseae. Bot. Jb. 31:

134-313.

Schlechter, R., 1906. Orchidaceae africanae. IV. Bot. Jb. 38:

144-165.

Schlechter, R., 1915. Orchidaceae Stolzianae, ein Beitrag zur

Orchideenkunde des Nyassalandes. Bot. Jb. 53: 477-605.

Schlechter, R., 1924. Contributions to South African orchid-

ology. Ann. Transv. Mus. 10: 233-252.

H. P. LINDER

387

Senghas, K., 1973. Orchidoideae. In R. Schlechter, Die Orchid-

een, ed. 3. Paul Parey.

Sonder, G., 1847. Enumeratio Orchidearum, quas in Africa

Australi Extratropicum collegerunt C. F. Ecklon, Dr., et C.

Zeyher. Linnaea 19: 71-112.

Sprengel, C., 1826. Linnaea Systema Vegetabilum 3. Gottingen.

Stafleu, F. A., et a!., 1978. International code of botanical

nomenclature. Utrecht: Bohn, Scheltema & Holtema.

Summerhayes, V. S., 1968. Orchidaceae, Part 1. In E. Milne-

Redhead & R. M. Polhill (eds). Flora of Tropical East Africa

156. London.

Swartz, O., 1800. Orchidemes slagter och arter upstallde. Kong.

Vet. Acad. Hand!., Stockholm 31: 202-254.

Thunberg, C. P., 1794. Prodromus Plantarum Capensium.

Uppsala.

Vesey-fitzgerald, D. F., 1963. Central African grasslands. J.

Ecol. 51: 243-274.

Wagner, W. H., 1962. A graphic method for expressing relation-

ships based upon group correlations of indexes of divergence.

In L. Benson, Plant taxonomy, methods and principles

415-417. New York: Ronald Press Co.

White, F., 1978. The afro-montane Region. In M. J. A. Werger

(ed.), Biogeography and ecology of southern Africa. The

Hague: Junk.

Williamson, G., 1977. The orchids of south central Africa.

London: Dent.

APPENDIX: SPECIMENS STUDIED

The specimens are listed alphabetically according

to the name of the collector. The figures in brackets

refer to the number of the taxon in the text. Herbaria

from which each collection has been studied are indi-

cated by the letter codes of Holmgren and Keuken

(1974).

Acocks 2087 (6) S; 3678 (10) S; 1 1731 (14) PRE. Alchenck 549 (5)

Z. Andraea 278 (5) STE. Atherstone 29 (9a) K.

Baker 37 (6) K. Ball 577 (15) K, SRGH. Balsinhas-Kersberg 2054

(14) PRE. Barker 56 (4a) NBG; 620(11) NBG; 625 (11) NBG; 3042

(4a) NBG; 444 (6) NBG; 4785 (4a) NBG; 5760 (4a) NBG; 6043 (11)

(NBG; 8476 (6) NBG; 8828 (1 1) NBG. Barkley in BOL 4884 (5)

BOL. Bayliss 3067 (11) NBG; 4162 (11) NBG. Benson 149 (13)

BM; 189 (13) BM. Bergius s.n. (5) P. Biegel 257 (14) BOL, SRGH.

Bodkin 494 (9a) BOL, K, SAM. Bolus 494 (9a) BOL, K, SAM;

1552 (11) BOL; 3810 (9a) BM, BOL, GRA, PRE, Z; 4393 (4a)

PRE; 4566 (8) BM, BOL, K, PRE; 4566b (7) BM; 4857 (7) BOL,

GRA, K; 4893 (6) BOL, GRA, K; 5278 (10) BOL, K; 7992 (6) Z;

9788 (14) BOL, K; 11379 (5) GRA; 11645 (10) BOL, PRE; 12327

(11) BOL, BR; 13514 (11) BOL; 17494 (8) BOL; s.n. (46) BOL;

s.n. (5) BOL; (6) SAM; s.n. (8) BOL; s.n. ( 1 1) BOL. Bond 878 (1 1)

NBG. Bowie 12 (5) BM; 13 (4a) BM; s.n. (11) BM. Breach s.n. (10)

BOL. Britten 1058 (11) GRA, PRE; 6575 (9a) GRA. Brook s.n.

(4a) PRE. Buchanan 1016 (14) K. Buchholz s.n. (5) STE. Buhr 5

(11) BOL. Burchell 747 (11) K; 808a (5) K; 7001 (5) K; 7801 (5) K;

4572 (1 1) K; 4592 (1 1) K; 7182 (1 1) K; 7372 (5) W. Burn Daw 1915

(14) K; 4245 (5) PRE; 12004 (1 1) PRE; 18483 (1 1) BOL; 22070(14)

K; s.n. (14) BOL.

Carter & Barres s.n. (4a) BOL. Chapman 148 (14) BM; 330 (14)

BM, SRGH. Chase 2949 (14) BM, BOL, K; 4059 (14) SRGH.

Chater s.n. (11) BOL, PRE. Clarke in NBG 2184/31 (9a) BOL.

Codd 523 (14) BM, PRE; 9714 (14) GRA, K, PRE, UPS. Collins

1538 (4a) K. Compton 4150 (4a) BOL, NBG; 4515 (11) BOL,

NBG; 10499 (5) NBG; 12322 (6) NBG; 16644 (8) NBG; 20236 (6)

NBG; 24487 (14) NBG; 25150 (14) NBG; s.n. (5) BOL. Cookson 3

(14) PRE, SRGH. Cooper 1464 (11) BM, K, W; sub Eyles 2743

(14) PRE, SRGH. Cresswell s.n. (9a) SAM. Crook 136 (14)

SRGH; 409 (14) K, SRGH. Crundall s.n. (14) PRE Culver 20(14)

BM, BOL, BR, K. Cuthbert 5 (14) K. Cutting s.n. (9a) BOL; s.n.

(11) BOL.

Dahlstrand 2586 (11) GRA. Daly & Sole 527 (9a) GRA. Davidson

s.n. (4a) SAM. Da vies 2120 (14) PRE, SRGH; s.n. (4a) SAM. Dart

s.n. (9a) BOL, STE. Delhaye s.n. (14) K. Denman 237 (1 1) GRA.

De Villiers s.n. (1 1) STE. De Vos 738 (8) PRE. Doe s.n. (8) BOL.

Drege 1234 (4a) K, P; 2211a (11) K, W; s.n. (5) SAM; 63 (11)

GRA. Drummond 8956 (14) SRGH; 8981 (14) SRGH; 8982 (15)

SRGH. Dummer 756 (9a) BM; 1502 (14) BM. Duthie 1027 (11)

STE. Du Toit s.n. (9a) BOL. Dyer 229 (9a) PRE.

Ecklon 1565 (9a) W. Edwards 2157 (14) NU; s.n. (14) BOL.

Esterhuysen 3802 (10) BOL; 4628 (11) BOL; 7298 (10) BOL; 8207

(4a) BOL; 13599 (1 1) BOL, K, PRE; 13664 (5) BOL; 15154 (5) K,

PRE; 17441 (4a) BOL, PRE; 19990 (5) BOL; 20353 (4a) BOL;

20907 (10) BOL; 22712 (5) BOL, K. Eyles 2743 (14) PRE, SRGH.

Fair in BOL 3810 (9a) PRE; in BOL 7992 (6) BOL, GRA, PRE.

Ferguson s.n. (3) BOL; s.n. (11) BOL. Flugge-De Smit s.n. (5)

BOL. Fourcade 542 (11) GRA; 548 (5) BOL, GRA; 3167 (9a) K;

4309 (11) K, STE; 4344 (4b) BOL; 6443 (4b) PRE. Franks 9810

(14) PRE, Froemling s.n. (9a) NBG. Frowein 16131 (8) PRE.

Galpin 294 (9a) PRE; 427 (14) BOL, GRA, K, PRE, SAM; 428

(14) GRA; 4917 (5) PRE; 12545 (5) PRE. Carley 167 (14) SRGH.

Garside 46 (9a) K, PRE. Gemmell 5037 ( 1 1) PRE. Gerloin 249 ( 1 1 )

GRA. Germain 1563 (14) BR. Gillen 326 (1 1) K; 718(1) BOL; 1602

(4b) BOL; 17648 ( 14) K, PRE; 17799 (14) K; s.n. (5) STE; s.n. (9a)

K; s.n. (11) BOL. Gillett & Bolus in BOL 18506 (11) BOL. Gillies

107 (14) NU. Glass s.n. (9a) GRA, PRE. Goatcher s.n. (4a) BOL.

Goetze 1222 (14) BM. Grant 2464 (4a) BR, M, PRE; s.n. (14) BM.

Greenway 6318 (14) K, PRE. Grice s.n. (14) NU. Grosvenor 178

(15) K, SRGH; 257 (14) K, SRGH. Gueinzius s.n. (10) W. Guthrie

871 (4a) BOL; 4383 (11) NBG; 8384 (11) NBG; in BOL 6861 (10)

GRA.

Hafstrom & Acocks 2087 (6) PRE, Z, S. Hal! 664 (9a) BOL;

1043 (4a) BOL; 1160 (11) BOL. Hallack in BOL 6210 (11) BOL,

GRA. Hall-Martin 429 (14) PRE. Handel Hamer in BOL 16964

(11) BOL. Harvey 116 (9a) K; 148(7) K; s.n. (5) BM, K, W. Har-

wood s.n. (14) PRE. Hayes Palmer s.n. (1 1) NBG. Haynes 304 (5)

STE; 528 (5) STE. Hendrickx 3465 (14) PRE. Hermann 871 (4a)

NBG. Hill 2 (14) K. Hilliard & Burn 10404 ( 14) NU. Huysteen s.n.

(6) STE.

Immelmann 246 (4a) BOL.

James in BOL 23178 (1 1) BOL. Jameson s.n. (6) K. Jenkins 8228

(14) PRE. Jeppe in PRE 3383 (5) PRE; in PRE 33384 (4b) PRE; in

PRE 33385 (1 1) PRE. Johnson s.n (14) K. Joubert s.n. (5) K.

Karsten s.n. (14) NBG. Keet 1154 (11) GRA. Kennedy s.n. (5)

PRE. Kensley 280 (1 1) GRA. Kensit 9339 (8) BOL. Kerfoot 1632

(16) K. Kettle 18 (13) PRE. Kettlewell in BOL 25392 (6) BOL.

Keulder s.n. (9a) STE. Kirk s.n. (14) K. Kolbe 2412 (9a) GRA.

Krauss s.n. (5) M.

Lavis s.n. (4a) BOL. Leach 4121 (14) K, SRGH; 14941 (14) BOL,

SRGH; 21121 (14) S. Lees 99(14) K. Leighton 402 (4a) BOL; 3067

(11) BOL, PRE. Leipoldt 601 (4a) BOL; 3233 (4a) BOL; 3234 (4a)

BOL; 3810 (4a) BOL; 4243 (11) BOL; in BOL 11379 (5) PRE; s.n.

(4a) BOL. Lewis 828 (8) SAM; 1837 (4a) SAM; 2402 (6) SAM;

2404 (7) SAM; 2680 (4a) SAM; 4737 (4a) SAM; 5202 (4a) NBG.

Liebenberg 7805 (1 1) K, PRE. Linder 759 (6) BR, BOL; 1244 (4a)

BOL; 1245 (4a) BOL; 1453 (4a) BOL; 1458 (4a) BOL; 1460 (4a)

BOL; 1471 (4a) BOL; 1549(4a)BOL, BR, K; 1656(10) BOL; 1714

(11) BOL; 1729(11) BOL; 1731 (11) BOL; 1743 (10) BOL; 1763 (5)

BOL; 1 806 (7) BOL. Long 1 65 (9a) K; 494 (9a) K; 507 ( 1 1 ) K, PRE;

520 (11) K, PRE. Luyt in BOL 10571 (3) BOL, PRE; 11379 (5)

BOL.

MacOwan 700 (9a) BM, GRA, K, SAM; 1045 (11) BM, GRA, K,

SAM; 1045b (5) SAM, W; 2587 (4a) SAM; 2690 (9a) SAM; s.n. (5)

NBG. MacOwan & Bolus 166(7) BM, BOL, K, P, W, ZT; 167(5)

BM, BOL, K, P, SAM, UPS, W, ZT; 494 (9a) BOL, P, PRE, W,

Z, ZT. MacNicol s.n. (11) NBG. Magennis s.n. (4a) BOL, PRE.

Mahon s.n. (14) K. Marloth 332 (6) BOL, PRE; 425 (5) PRE; 483

(10) PRE; 1601 (9a) BOL; 2310 (4a) PRE; 6678 (10) BOL, PRE;

7273 (7) PRE; 7941 (4a) PRE; 8337 (10) PRE; 8435 (4a) PRE; 8847

(9a) PRE; 10061 (11) PRE; 1 1008 (9a) PRE; s.n. (5) BOL; s.n. (6)

SAM. Marsh 1408 (4b) PRE, SRGH. Matthews 28 (14) SRGH.

Mauve 5005 (14) PRE. McClounie 10 (14) K; 92/3 (14) BM.

McLoughlin 362 (14) BOL; 92/3 (14) K, P, PRE, S, UPS; s.n. (11)

BOL. Mgaza 488 (14) K. Michael et al. 971 (13) SRGH. Mid-

dlemost 1721 (4a) NBG; 1954 (11) NBG. Minicki s.n. (1) BOL.

Moorshead sub Moss 17594 (11) BM. Morris 52 (14) K; in BOL

13478 (4a) BOL. Moss 15432 (14) K, PRE; 17594 (11) BM, 18237

(11) BM. Muir 62 1 (11) PRE; 908 (9a) BOL, PRE, SAM; in PRE

16266 (9a) PRE; s.n. (4a) PRE. Munch 274 (14) SRGH; 327 (15)

K, SRGH. Myburg s.n. (9a) NBG.

Newbould & Jefford 1849 (14) K. Newdigate in BOL 6327 (2)

BOL; s.n. (11) BOL.

O’Brien s.n. (9a) BM; s.n. (11) K , Z. O’Connor 216 (14) NU; 368

(14) NU. Oldevig-Roberts s.n. (11) S. Oliver 3181 (9a) PRE, STE;

3006 (6) PRE; in STE 29974 (10) PRE.

Pappe 38 (4a) SAM; 39 (9a) BOL, SAM; 377 (9a) BOL, GRA,

SAM; in BOL 4393 (4a) BOL; s.n. (4a) K; s.n. (11) K, SAM.

Parker s.n. (6) BOL. Paterson 106 (5) GRA; 488 (11) GRA, Z;

1277 (11) BOL, GRA; s.n. (6) BOL. Pattison in BOL 14455 (10)

BOL. Pawek 1409 (13) SRGH; 3800 (14) K; 10275 (14) SRGH.

Penther 50 (11) BM, M, W; 154 (5) M, W; 251 (4a) W; in PRE

10575 (1 1) PRE. Peters s.n. (4a) SAM. Pillans 3530 (6) PRE; 4056

(5) PRE; 4125 (1) BOL; 8275 (8) BOL; 9723 (10) BOL. Plowes

2807 (15) K, SRGH; 2808 (14) K, SRGH. Pott 1278 (9a) PRE.

Powrie 168 (10) BOL. Prentice s.n. (4a) PRE.

Rattrav in BOL 15770 (8) BOL. Rauh & Schlieben 9761 (14) M,

PRE. Rehmen 529(11) BM, Z; 571 (5) Z. Reynolds 4200 (14) PRE.

388

TAXONOMIC STUDIES IN THE DISINAE. VI. A REVISION OF THE GENUS HERSCHELIA

Richards 6804 ( 14) K; 18500 (14) K; 22561 (14) K; 22574 (14) K, P.

Robinson 6259 (13) K, M, SRGH. Robson 297 (14) K; 358 (14)

BM, K, SRGH. Rogers 10550 (5) Z; 16554 (1 1) Z; 19079(14) PRE;

19767 (14) GRA, K, P, S; 21404 (14) BOU, K, Z; 26487 (11) Z;

26865 (1 1) Z; 27643 ( 1 1 ) Z; 30155 (14) BR, K, P, Z. Rosenbruck in

BOL 27817 (1) BOL. Rvcroft 2559 (4a) NBG. Ryder in NBG 40/28

(11) BOL, K; s.n. (3) K.

Salter 8703 (8) NBG; 325/1 (5) BM; 325/2 (6) BM; 325/3 (7) BM;

325/4 (8) BM; in SAM 53195 (8) SAM. Schelpe 013 (14) NU; 826

(14) NU; 4267 (11) BOL; 4994 (11) BOL; 5006a (11) BOL; 6313

(9a) BOL. Sch lech ter 481 (5) P, Z; 2061 (5) BOL, Z; 4997 (4a)

PRE, Z; 5165 (4a) BOL, GRA, P, W, Z; 5928 ( 1 1) Z; 9544 (8) BM,

BR, GRA, K, PRE; s.n. (4a) BR, K. Schlieben 1366 (14) BM, G,

M, P, S, Z. Schmidt 3 (5) M. Schnisterhol 213 (5) S. Schonland

1519 (9a) GRA, Z; 2410 (9a) PRE; 3662 (11) GRA; sub Galpin

4917 (5) PRE. Seltzer s.n. (9a) NBG. Seltzer & Parke s.n. (9a)

BOL. Shirley 234 (14) NU. Sidey 490 (14) PRE; 1498 (14) PRE, S;

4057 (11) PRE. Sim 2457 (14) BOL, PRE. Skead 210 (14) NU.

South 127 (9a) GRA; s.n. (1 1) GRA, PRE. Stander s.n. (11) STE.

Starke 127/27 (11) BOL; NBG 60/27 (11) BOL. Stewart 8875 (14)

GRA, PRE. Stokoe 2560 (1) BOL; 7324 (5) BOL, SAM; 7386 (5)

BOL; 8679 (10) BOL; in SAM 54389 (1 1) SAM; in SAM 54714 (1 1)

SAM; s.n. (5) SAM. Stolz 127 (14) BM, G, K, M, W, Z; 2192 (14)

C, G, M, S, W, Z; 2193 (14) C, G, M, Z. Strey 9853 (14) PRE.

Sturgeon in SRGH 30524 (14) K, SRGH. Symons 22 (14) SAM; in

PRE 14847 (14) PRE.

Taylor 191 (14) NU; 634 (11) NBG; 1713 (14) NU; 1755 (14) NU;

1786 (14) NU, SRGH; 1791 (15) NU; 5455 (9a) STE; 6220 (5) STE.

Tennant 5 (5) NBG. Thode A48 (5) PRE; A387 (14) K, PRE;

A 1022 (11) K, PRE; in STE 5247 (5) STE; in STE 5307 (1 1) STE;

in STE 6104 (5) STE; in STE 6526 (1 1) STE; in STE 8128 (5) STE.

Thomas s.n. (4a) NBG; s.n. (6) NBG. Thorncroft 2478 (14) K,

PRE. Thornton s.n. (14) PRE. Thulin & Mhoro 1201 (14) K, UPS;

1207 (14) K. Thunberg 21429 (9) UPS; 21443 (11) UPS. Trauseld

412 (14) PRE. Trimen s.n. (12) BM; s.n. (9a) BM. Tvrer 726 (13)

BM, SRGH; 966 (13) BM, SRGH; 978 (13) BM. Tyson 1537 (14)

BOL.

Wahlberg s.n. (5) S. Wall s.n. (3) S; s.n. (4a) S; s.n. (5) S; s.n. (9a)

S; s.n. (11) S. Wallich 113 (11) BM. Watermeyer 167 (13) K. West

256(11) W. Westwood 694 (14) SRGH. Whellan 1493 (11) SRGH;

2137 (14) SRGH; 2145 (15) SRGH. Whyte 345 (14) K; s.n. (14) K;

s.n. (14) K. Wild 1366(14) K, SRGH; 3556 (14) PRE, SRGH; 4668

(14) SRGH; 4669 (14) K. Willan 176 (14) K; in BOL 24915 (14)

BOL. Williamson 119 (14) K; 312 (13) K; 1023 (13) K, SRGH.

Wilms 1406 (14) BM. Wolley-Dodd 358 (9a) BM; 359 (7) K; 391 (8)

BM, K; 840 (5) BM; 885 (5) K; 1798 (9a) K; 2005 (6) BM, K. Wood

9290 (14) BOL, K; 10599 (14) K, PRE. Worsdell s.n. (4a) K; s.n.

(9a) K; s.n. (14) K . Wright 2412(14) NU. Wurts55S (5) NBG; 2136

(11) NBG. Wvliesub Wood 10599 (14) GRA, PRE, SAM; in Diim-

mer 1502 (14) BM.

York 34 (9a) K.

Van Zinderen Bakker 56(1 1) NBG. Venter 848 ( 1 4) PRE. Venter &

Vorster 195 (14) PRE; 196 (14) PRE. Vogelpoel s.n. (14) BOL.

Voigt in PRE 13199 (14) PRE.

Zeyher 504 ( 1 1 ) P ; 628 ( 1 1 ) K; 1 566 (9a) K, S, SAM; 1 567 (7) K, P ,

SAM, W; 3917 (11) S; 3918 (10) BM, K, P, S, W; s.n. (5) SAM;

s.n. (6) K. Zinn s.n. (5) SAM.

INDEX

Page

Disa Berg.

sect. Forficaria (Lindl.) Schltr 368

sect. Herschelia (Lindl.) H. Bol 368

sect. Microperistera H. Bol 368

sect. Spathulatae Kraenzl 368

sect. Trichochila Lindl 368

atropurpurea Sond 372

barbata (L. f.) Swartz 376

baurii H. Bol 383

charpentieriana Reichb. f 379

excelsa sensu Lindl 381

forcipata Schltr 382

forficaria H. Bol 368

graminifolia Ker-Gawl. ex Spreng 374

hamatopetala Rendle 383

hians( L. f.) Spreng 381

lacera Swartz 381

longilabris Schltr 383

lugens H. Bol 379

macroglottis Sond. ex Drege 379

multifida Lindl 379

newdigateae L. Bol 370

outeniquensis Schltr 381

propinqua Sond 372

purpurascens H. Bol 375

schlechterana H . Bol 371

spathulata (L. f.) Swartz 372

var. atropurpurea (Sond.) Schltr 372

venusta H. Bol 378

tripartita Lindl 373

walteri Schltr 385

Eulophia hians (L. f.) Spreng 381

Forficaria Lindl 368

graminifolia Lindl 368

Page

Herschelia Lindl 368

atropurpurea (Sond.) Rolfe 372

bachmanniana Kraenzl 383

barbata ( L.f .) H. Bo! 376

baurii (H. Bol.) Kraenzl. ... 383

charpentierana (Reichb. f.) Kraenzl 379

chimanimaniensis Linder 384

coelestis Lindl 374

excelsa sensu Rolfe 381

forcipata (Schltr.) Kraenzl 382

forficaria (H. Bol.) Linder 368

goetzeana Kraenzl 385

graminifolia (Ker-Gawl. ex Spreng.) Dur. & Schinz 374

hamatopetala (Rendle) Kraenzl 383

hians (L. /.) Hall 381

lacera (Swartz) Fourc 381

longilabris (Schltr.) Rolfe 383

lugens (H. Bol. ) Kraenzl 378

multifida (Lindl.) Rolfe 379

newdigateae (L. Bol.) Linder 370

praecox Linder 382

purpurascens (H. Bol.) Kraenzl 375

schlechterana)//. Bol.) Linder 371

spathulata (L. f.) Rolfe 371

subsp. spathulata 372

subsp. tripartita (Lindl.) Linder 373

venusta (H. Bol.) Kraenzl 378

tripartita (Lindl.) Rolfe 373

Limodorum hians (L. f.) Thunb 381

Orchis barbata L.f 376

spathulata L.f 372

Satyrium barbatum (L. f .) Thunb 376

excelsum Thunb 381

hians L. f 381

spathulata (L. f.) Thunb 372

Bothalia 13, 3 & 4: 389-413 (1981)

An analysis of the African Acacia species: their

distribution, possible origins and relationships

J. H. ROSS*

ABSTRACT

The three subgenera recognized within the genus Acacia are outlined and the global distribution of each is indi-

cated. The differences between the subgenera and the degree of relationship and levels of specialization are dis-

cussed briefly. It is suggested that the ancestral members of the genus were climbers or lianes. Past geological events

considered likely to have influenced the distribution of the Acacia species in Africa are outlined. The number of

species recorded from each African country is tabulated and the distribution and concentration of species within the

genus Acacia as a whole and within each subgenus in Africa are illustrated. The highest concentrations of species

within each subgenus occur in tropical east and south-east Africa. The distribution of species within some of the

individual African countries and possible affinities are discussed and attention is drawn to the main centres of

endemism. The distribution of the African species is correlated with the major phytogeographical regions recog-

nized on the continent. The relationships between the African and the American, Madagascan, Indian and

Australian Acacia species are discussed briefly.

RESUME

UNE ANAL YSE DES ESPECES D’ ACACIA AFRICAINS: LEUR DISTRIBUTION , ORIGINES

EVENTUELLES ET RELA TIONS

Les trois sous-genres reconnus dans le genre Acacia sont esquisse et la distribution globale de chacun est indiquee.

Les differences entre les sous-genres et le degre de relation et niveaux de specialisation sont brievement discutes. II

est suggere que les membres ancestraux du genre etaient des grimpants ou des lianes. Des evenements geologiques du

passe consideres comme avoir vraisemblablement influencer la distribution des especes d’ Acacia en Afrique sont

decrits. Le nombre des especes enregistrees dans chaque pays d' Afrique a ete presente sous forme de tables et la

distribution ainsi que la concentration des especes dans le genre Acacia dans son ensemble et dans chaque sous-genre

d’ Afrique sont illustres. Les concentrations les plus elevees d ’especes dans chaque sous-genre surviennent en Afri-

que tropicale orientate et sud-orientale. La distribution des especes dans certains pays africains individuels et les

affinites possibles sont discutees et /’attention est attiree sur les centres principaux d ’endemisme. La distribution des

especes africaines et des especes d’ Acacia americaines, malgaches, indiennes et australiennes est brievement

discutee.

INTRODUCTION

The genus Acacia at present consists of about

1 100 species (perhaps as many as 1 200) which are

widely dispersed in the Americas, the Carribbean and

Pacific Islands, Africa, Madagascar and the Mas-

carenes, Asia, the Indo-Malesian region and Austra-

lia. Europe is the only large geographical area devoid

of indigenous Acacia species, and there are no indige-

nous species in New Zealand despite its relative

proximity to Australia. The fossil record indicates

that the genus was previously more widely distributed

having been present formerly in the Ukraine (Shche-

kina, 1965) and in New Zealand (Mildenhall, 1972,

1975). Most species of Acacia occur in regions where

the rainfall is markedly seasonal or low, relatively

few inhabiting rainforest areas, but even then the

rainfall is usually unevenly distributed throughout

the year and even in the wettest parts there is usually

a short dry season. This does not imply that the genus

originated in an arid or semi-arid region. On the con-

trary, it is considered probable that Acacia originated

in the tropical lowlands and that most of the xero-

phytic features within the genus are secondary.

Much evidence has accumulated in recent years to

support the contention that there have been large

scale movements of continents during geological time

and Raven & Axelrod (1974) summarized the biogeo-

graphic support for the theory of plate tectonics.

Raven & Axelrod (l.c.) postulated that West Gond-

wanaland, combined Africa and South America, was

a primary area of evolution for many major orders of

*National Herbarium of Victoria, Birdwood Avenue, South

Yarra, 3141, Victoria, Australia.

angiosperms and perhaps the earliest angiosperms

themselves, and that the initial radiation of the

angiosperms occurred when direct migration was

possible between South America, Africa, India, Ant-

arctica and Australia, and via Africa to Laurasia.

West Gondwanaland included vast arid to subhumid

areas in tropical latitudes where the terrain and

edaphic conditions were diverse and provided every

opportunity for rapid evolution. Thorne (1977), on

the other hand, favoured south-eastern Asia and

associated archipelagoes as the primary centre of

origin of the most primitive angiosperms and main-

tained that prior to its fragmentation West Gond-

wanaland was merely one of a number of important

centres for the development of the tropical angio-

sperm flora.

In support of their contention Raven & Axelrod

(l.c.) argued that West Gondwanaland was centrally

located on routes of dispersal at the time when the

primary evolutionary radiation of the angiosperms

was taking place. The climatic changes associated

with the fragmentation of Gondwanaland, which, ac-

cording to Raven & Axelrod, commenced in the mid-

Cretaceous approximately when the first angio-

sperms are encountered in the fossil record, probably

had a profound influence on the evolution of the

angiosperms and on the establishment of major lines

within it. There is some disagreement over the timing

of the fragmentation of Gondwanaland, an event of

primary importance from a phytogeographic point of

view. Melville (1975), for example, was of the opi-

nion that Gondwanaland started to fragment earlier

than indicated by Raven & Axelrod. In this paper I

have followed Raven & Axelrod in assuming that

390

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Gondwanaland started to fragment in the mid-

Cretaceous and that the angiosperms were already

fairly widely distributed.

The opening of the South Atlantic 125-130 million

years ago heralded the spread of more mesic climates

over much of West Gondwanaland and is thought to

have started the sur^e of angiosperms into the mesic

lowland record abolit 1 10 million years ago (Raven &

Axelrod, l.c.). The pattern of appearance of the

angiosperms in the lowland record suggested to

Raven & Axelrod that the primitive members of

several extant orders and perhaps even a few families

were already in existence by the close of the early

Cretaceous 110 million years ago. Thorne (1978),

however, maintained that few, if any, extant

families, and certainly no extant genera, had evolved

in West Gondwanaland before the final sundering of

South America from Africa 100 million years ago.

Raven & Axelrod (l.c.) expressed the opinion that

most modern angiosperm families were in existence

in the Paleocene about 65 million years ago before

the connection between Africa and Eurasia was

severed. The family Leguminosae is thought to have

originated or at least have undergone its primary

radiation and differentiation into three subfamilies in

West Gondwanaland and Raven & Axelrod expressed

the view that Mimosoideae, Caesalpinioideae and

perhaps Papilionoideae were in existence by the

Paleocene. There are no reliable Cretaceous records

of Caesalpinioideae or of the other two subfamilies.

From what can be inferred about its history and pre-

sent distribution patterns, Raven & Axelrod presume

that Mimosoideae migrated between Africa and

South America during or prior to the Paleocene.

After the Paleocene (54 million years ago) the evi-

dence suggests only limited migration between these

two continents.

Vassal (1972) recognized three subgenera within

Acacia, namely, subgenera Acacia, Aculeiferum and

Heterophyllum*, chiefly on the basis of characters of

seeds and seedlings and on the occurrence of stipular

spines and pollen characters. The three subgenera

recognized by Vassal broadly correspond to group-

ings of Bentham’s six series which is fortunate as

most of the characters on which Vassal’s classifica-

tion are based are not obvious from the gross mor-

phology of conventional herbarium specimens. The

relationship between Bentham’s series and Vassal’s

subgenera is as follows:

1. Subgenus Aculeiferum Vassal ( = series Vul-

gares Benth. and series Filicinae Benth.)

2. Subgenus Heterophyllum Vassal ( = series

Phyllodineae Benth., series Botryocephalae

Benth. and series Pulchellae Benth.)

3. Subgenus Acacia ( = series Gummiferae Benth.)

The following sections within each subgenus were

proposed by Vassal (1972) and Guinet & Vassal

(1978) and the names are used in this paper:

1. Subgenus Aculeiferum (Sections Aculeiferum,

Monacanthea and Filicinae)

2. Subgenus Heterophyllum (Sections Hetero-

phyllum, Uninervea and Pulchelloidea)

3. Subgenus Acacia (Section Acacia)

*The subgeneric name Phyllodineae (DC.) Seringe has priority and

will have to be adopted in place of Heterophyllum Vassal, but the

name Heterophyllum is retained for the purpose of this paper.

GLOBAL DISTRIBUTION OF THE GENUS ACACIA

An indication of the global distribution of each

subgenus is given in Figs 1-3.

The distributions of subgenera Acacia and Aculei-

ferum are very similar but subgenus Acacia apparent-

ly enjoys a slightly wider distributional range than

subgenus Aculeiferum. Subgenus Aculeiferum has a

more restricted distribution in Africa than subgenus

Acacia, is present in New Guinea while subgenus

Acacia is absent, and only just reaches Australia (in

the vicinity of Coen in northern Queensland) where it

is represented by a solitary species (A. albizioides

Pedley) in contrast to subgenus Acacia which is

widely distributed in northern Australia although

represented by fewer than ten species. The vast maj-

ority of species in the genus belong to subgenus Het-

erophyllum which is fundamentally an Australian

group (including Tasmania), while a further eighteen

species (Pedley, 1975) occur in Madagascar and the

Mascarenes, New Guinea, Formosa, the Philippines

and the Pacific Islands to Hawaii. The genus reaches

its southern limit of distribution in Tasmania. The

position of A. willardiana Rose, which occurs on the

west coast of North America (Mexico), is not clear

(see later discussion) but if it is placed in subgenus

Heterophyllum as advocated by Vassal & Guinet

(1972) then the distribution of the subgenus shown in

Fig. 3 should be extended eastwards from Hawaii to

Mexico.

Guinet & Vassal (1978) are of the opinion that the

three subgenera were differentiated by the Oligo-

Miocene period (4= 27 million years ago), and that no

fundamental difference seems to exist "between their

geographical distribution then and the present. They

pointed out that the apparent absence of the genus in

the fossil record during the Paleocene is surprising,

particularly if the genus is held to have had a mono-

phyletic origin and if one considers that its distribu-

tion during the Neogene was what it is now.

ORIGIN AND POSSIBLE IDENTITY OF THE ANCESTRAL

MEMBERS OF THE GENUS ACACIA

Like the origin of the angiosperms, the identity of

the ancestral form of Acacia has been the subject of

much speculation and disagreement, but recently a

broad consensus appears to have been reached which

contradicts the earlier views of Andrews (1914) and

Atchison (1948). Andrews and Atchison considered

the Gummiferae (subgenus Acacia) to be the

ancestral form as its members contained the morpho-

logical characters of the genus that they considered to

be primitive, namely, bipinnate leaves and persistent

spinescent stipules. Atchison maintained that

chromosome number variation, morphological uni-

formity and geographical distribution contributed

toward establishing Gummiferae as the ancestral

form of the genus pointing out, in support of this

contention, that Gummiferae is the only section of

the genus with a cosmopolitan distribution (this is

not strictly correct). The other sections of the genus

were held to have developed from the original forms

in secondary centres where isolation through climatic

or edaphic change was favourable to the survival of

new types. Tindale & Roux (1975), on the basis of a

limited sample, suggested that the chemical content

of the South African species with non-spinescent

stipules (subgenus Aculeiferum) is generally more ad-

vanced than that of the species with spinescent

stipules (subgenus Acacia), a suggestion that sup-

ports the above view. However, most of the Gummi-

ferae are polyploids and polyploidy is now held to

J. H. ROSS

39!

Fig. 1. — An indication of the

global distribution of sub-

genus Acacia (excluding the

distribution of Acacia farne-

siana in Australia).

Fig. 2. — An indication of the

global distribution of sub-

genus Aculeiferum.

Fig. 3. — An indication of the

global distribution of sub-

genus Heterophyllum (ex-

cluding A. willardiana which

probably belongs in subgenus

Aculeiferum).

correspond to a high degree of differentiation in the

genus Acacia.

Robbertse (1974) outlined the possible evolution of

the inflorescence and flowering system in the South

African acacias and considered subgenus Acacia

( Gummiferae ) to be more advanced than subgenus

Aculeiferum ( Vulgares), a view which is supported by

the detailed studies of Guinet & Vassal (1978). Rob-

bertse considered it probable that the paniculate

flowering system, spicate inflorescence, pedicellate

flowers, presence of a cup-shaped disc and a pedicel-

late ovary, all of which are found in subgenus Aculei-

ferum, are primitive characters.

Guinet & Vassal (1978) evaluated the degree of

relationship and specialization of the major sub-

divisions recognized within the genus on the basis of

pollen, chromosome, seed, pod, inflorescence and

vegetative characters. Each character was divided

into three states, namely, unspecialized, specialized

and highly specialized, and that section of the genus

displaying the characters considered to be highly

specialized was itself considered to be highly special-

392

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

ized. The division of some characters, for example

seed size, into somewhat arbitrary size classes is ques-

tionable, especially as it was not disclosed why the

range of continuous variation in seed size was divided

in such a manner or what criteria were employed to

established the degree of specialization of each size

class. For example, seed were divided into the follow-

ing size classes, small (less than 5 mm long), medium

(5-10 mm long) and large (more than 10 mm long).

Small seed were regarded as unspecialized but it is

quite conceivable that in some instances small seed

may be specialized. In addition, the range of varia-

tion in seed size in many species obscures the limits of

Guinet & Vassal’s size classes.

Guinet & Vassal (/.c.) concluded that:

1. On the basis of pollen morphology subgenus

Acacia is the most specialized of the three subgenera

and subgenus Aculeiferwn the least specialized.

Although the pollen of subgenus Heterophyllwn is

generally more specialized than that of subgenus

Aculeiferum, the two subgenera share important

characters, for example, the absence of columellae

and the presence of simple apertures.

2. Subgenus Acacia is clearly distinguished from the

other two subgenera by the high level of specializa-

tion of the chromosome characters (chromosome

numbers and the degree of homogeneity of the

karyotype). Subgenera Aculeiferum and Heterophyl-

lum are more homogeneous and have similar levels of

differentiation of chromosome characters. On the

basis of chromosome characters, subgenus

Aculeiferwn is the least specialized and subgenus

Acacia the most specialized.

3. The characters of the seeds of subgenus Acacia

are often highly specialized, and the levels of special-

ization in the series Filicinae and Monacanthea of

subgenus Aculeiferum are fairly close.

4. The cotyledonary and adult foliar characters

selected did not appear to be specialized in subgenera

Aculeiferum and Acacia except in a few rare cases,

while many specializations occurred in subgenus Het-

erophyllum except in the development of spines-

cence. Section Filicinae of subgenus Aculeiferum

shows no specialized characters.

5. The characters of the inflorescence and pod selec-

ted indicated that certain characters in subgenus

Acacia are infrequently encountered in the other sub-

genera.

Guinet & Vassal attempted to estimate the total

levels of specialization within each subgenus (and

within the sections within each subgenus) and con-

cluded that:

1 . On the basis of their average level of specializa-

tion subgenus Acacia is the most specialized sub-

genus and Aculeiferum the least specialized.

2. Section Filicinae of subgenus Aculeiferum is

characterized by a preponderance of unspecialized

characters and shows the least diversity of all series in

the genus.

3. Section Monacanthea of subgenus Aculeiferum

contains more possible primitive states than section

Aculeiferum and appears to be less advanced than

the latter.

4. The Phyllodineae are more specialized than the

Botryocephalae and the Pulchellae in subgenus

Heterophyllwn but the persistence of unspecialized

states occurs to a similar degree in the three series.

Guinet & Vassal (/.c.) stressed that if the correla-

tion of the characters selected reflects a true relation-

ship between the major subdivisions of the genus,

then subgenera Aculeiferum and Heterophyllum are

more closely related to one another despite the fact

that they occupy basically different geographical

areas which show relatively little overlap, than are

subgenera Aculeiferum and Acacia which share a

common geographical area.

Guinet & Vassal favoured the concept that Acacia

originated in West Gondwanaland in an area that

approximates to the area presently occupied by Cen-

tral America (from Mexico to Bolivia). In support of

this contention Guinet & Vassal pointed out that

characters which are absent in the genus Acacia itself

in America are nevertheless found in indirectly

related genera. For example, the fundamentally

Australian extraporate pollen type (subgenus

Heterophyllum) exists in some South American

genera closely related to Piptadenia, and phyllodes

are present in some South American species of

Mimosa. These occurrences were regarded by Guinet

& Vassal as evidence that the American continent

contains most of the evolutionary potential for the

characters now found in the genus Acacia and

accords with Guinet’s (1969) earlier suggestion that

phyllodes and the pollen type commonly found in

subgenus Heterophyllum may have originated in

South America and that Australia was a secondary

centre of development and differentiation. However,

because of the occurrence in Australia of phyllodes

and the pollen type alluded to, it can equally well be

argued that Australia also contains most of the

evolutionary potential for the characters now found

in the genus Acacia and that these characters

originated in Australia. Guinet & Vassal are of the

opinion that section Filicinae of subgenus

Aculeiferum preserves the morphological characters

closest to those postulated as being ancestral in the

genus. Section Filicinae is poorly known and much

more information is required.

The rainforest areas of the world were previously

much more extensive and during the Paleogene

humid forests stretched through much of America,

Africa, Arabia, India, Malaysia and Australia.

Acacia species are not well represented in rainforest

areas at the present time and it is thought that this is

probably due to their general intolerance of low light

intensities. In the absence of any indications to the

contrary, it seems reasonable to assume that mem-

bers of the genus have always been similarly intoler-

ant of low light intensities. The Acacia species which

are currently the most successful in rainforest areas

are the climbers and it appears as though the climb-

ing habit has enabled species to exploit situations in

forests where light penetrates to the ground, for ex-

ample in clearings, on the banks of streams or on

forests margins, and reach and maintain an emergent

position in the canopy. If, as is considered likely,

Acacia originated in lowland forests, it is suggested

that the ancestral members were climbers or lianes

and, this being the case, members of proto- Aculei-

ferum which were similar in some respects to some

members of subgenus Aculeiferum.

Taking the African species as an example, a

number of the members of section Monacanthea sub-

genus Aculeiferum appear to be obligate climbers (A.

lujae De Wild., A. kraussiana Meisn. ex Benth.)

while others (A. brevispica Harms, A. ataxacantha

DC.) occur as climbers in forested areas or as scan-

.1. H. ROSS

393

dent shrubs or even large spreading shrubs in neigh-

bouring woodland or grassland areas. A. ataxacan-

tha occurs as a climber in forests and on forest

margins, as a scandent shrub or non-scandent

spreading shrub in woodland or grassland, but on oc-

casions it grows as a substantial single-stemmed tree

up to 10 m high in southern Africa. A. ataxacantha

apparently exhibits the evolutionary potential that

would have been necessary for a forest-dwelling

climber to adapt and exploit the new habitats created

in surrounding grassland and woodland areas as the

forests retreated.

A. ataxacantha and the other climbers are in-

variably armed with scattered recurved non-stipular

prickles but occasional plants are entirely or almost

entirely unarmed. Some species [A. caffra Thunb.)

Willd., A. galpinii Burt Davy, A. polyacantha Willd.]

in section Aculeiferum of subgenus Aculei-

feruin which are typically armed with prickles in

pairs at the nodes are likewise sometimes unarmed.

In some species (A. caffra) in section Aculeiferum

occasional specimens are found where a few irregu-

larly scattered prickles occur in addition to the paired

prickles at the nodes. This illustrates the apparent

ease with which scattered or paired prickles can be

lost and how scattered prickles could give rise to

paired or solitary prickles at the nodes or vice versa.

Once again, the evolutionary potential for such

change is still apparently present. Members of the

American section Filicinae which Guinet & Vassal

consider to be ancestral are typically unarmed but

prickles could have been lost as indicated.

Although subgenera Acacia and Aculeiferum share

a common geographical area there are fundamental

differences between them, as indicated by Guinet &

Vassal, and they do not appear to be closely related

which suggests that subgenus Acacia did not arise di-

rectly from subgenus Aculeiferum or vice versa. For

example, the colporate pollen of subgenus Acacia

with columellae is considered unlikely to have devel-

oped from the porate type without columellae. It

seems more likely, therefore, that subgenus Acacia

was derived from proto -Aculeiferum rather than

from subgenus Aculeiferum itself. Subgenus Hetero-

phyllum was possibly derived directly from subgenus

Aculeiferum or, failing that, from proto-Aculei-

ferum.

It is difficult to speculate on the identity of the an-

cestral proto- Aculeiferum except very generally. It is

suggested that the ancestral members were climbers

or lianes, either unarmed or armed with prickles,

with many-jugate bipinnate leaves. Robbertse’s (1974)

Findings suggest that they would have possessed a

paniculate flowering system and that the flowers

were pedicellate. The transition from capitate to

spicate inflorescences and vice versa appears to have

occurred several times during the development of the

genus and there is no certainty as to which condition

might be considered unspecialized.

OUTLINE OF THE CRETACEOUS-QUATERNARY

HISTORY OF AFRICA

The present distribution patterns in Acacia in

Africa have been determined by events that lie deep

in the past but it is difficult to assess, except very

generally, the effects of past geological events on the

flora of a continent. Unfortunately the plant fossil

record in Africa is generally poor from the Jurassic

until the Quaternary (Plumstead, 1969), largely

because of the uplift of the continent following the

break-up of Gondwanaland and the limited extent ot

lowland basins in which fossils were preserved. Con-

sequently little information is available during the

time that the angiosperms evolved and became domi-

nant during the Cretaceous. According to Axelrod &

Raven (1978), who provided an excellent survey of

the late Cretaceous and Tertiary in Africa, the

vegetation of Africa since the middle Cretaceous has

been shaped by diverse physical factors, and in its

broadest features the vegetation history in Africa has

paralleled that of other austral continents which have

remained relatively stable in latitude since the

Cretaceous. A summary follows of the more impor-

tant features of the Cretaceous-Quaternary history

of Africa, largely as outlined by Axelrod & Raven.

During the late Cretaceous and Paleocene (75-55

million years ago) Africa lay 15-18° further south

than at present. Relief was relatively low and a

lowland rainforest stretched from coast to coast

clothing nearly all of Africa except perhaps for the

southern tip; the late Cretaceous and Paleocene rain-

forest covered much of North Africa which was then

situated near the equator and what is now the Sahara

desert. This was a period of benevolence during

which rains were reliable and the widespread forests

flourished, although even at this time isolated

pockets of aridity are likely to have existed at the

edge of the tropics because of high pressure cells and

in edaphically dry sites in both tropical and

temperate zones (Axelrod, 1972). The southern part

of the continent, which came under the influence of

the westerlies, probably had a cool wet climate

(Goldblatt, 1979). At this time all of the temperate

austral lands were covered with a dense Podocar-

pus-Nothofagus-eve rgreen dicot temperate forest

and consequently it is inferred that a forest flora of

the Podocarpus-Nothofagus type (Nothofagus was

not necessarily present) probably covered the

southern tip of southern Africa. The inferred

distribution of vegetation in Africa during the late

Cretaceous-Paleocene is illustrated by Axelrod &

Raven, Fig. 6 (1978) along with that during the late

Oligocene-early Miocene, middle-late Miocene and

Recent.

By the close of the Cretaceous, Africa was isolated

from South America and India-Madagascar and was

surrounded by ocean. Although direct migration of

plants to and from Africa was restricted after the

mid-Cretaceous, direct interchange with South

America was much easier than at present as the

Atlantic was relatively shallow and numerous islands

provided stepping stones between the two continents

(Raven & Axelrod, 1974).

By the close of the Oligocene, the African plate

had moved north to virtually its present position.

During the late Oligocene-early Miocene (30-25

million years ago) the low relief in Africa was altered

by uplift accompanied by warping (King, 1967),

especially along the east coast, and the present land-

scape of the continent started to take shape. Volcanic

activity started on a major scale and the East African

rift valleys were initiated. Uplift brought a cooler

drier climate and the development and spread of dry

climate over tropical Africa probably began near the

close of the Oligocene about 27 million years ago

(Axelrod, 1972) and has continued to the present as

the rift valleys continue to grow (Raven & Axelrod,

1974). The formation of a volcanic field from Ethio-

pia southwards down the rift valleys during the Mio-

cene increased the development of rainshadows

which in turn brought greater drought and tempera-

ture extremes.

394

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

As a result of the Neogene trend to a drier climate

brought on by the general uplift of the continent,

changes in circulation, and the resultant decrease in

moisture, savanna started to spread at the expense of

rainforest and the African rainforest was progres-

sively impoverished. The development of rain-

shadows in the rift valleys favoured the spread of

savanna and then grassland, at first locally in small

patches but then more extensively as the rainshadow

effect increased. By the mid Miocene lowland rain-

forest is thought to have had only a patchy distribu-

tion along the northern parts of the east coast, and it

seems probable that a temporary dry season during

which little or no rain fell was already evident in the

Miocene.

By the close of the Oligocene-early Miocene the

vegetation of Africa had assumed a near-modern

aspect although the composition and distribution of

vegetation differed in many important respects from

that of today (see Axelrod & Raven, Fig. 6, 1978).

A further major factor that affected the African

flora was the development of the cold Benguella cur-

rent. By the early Miocene Antarctica had moved to

its present position and glaciation had been initiated.

When glaciation commenced in Antarctica cold

water started to bathe the west coast of Africa bring-

ing to it a drier colder climate. A full ice sheet did not

appear until the Pliocene about 5 million years ago

and it waxed and waned for 2-3 million years. As the

major ice sheet spread the Benguella current increas-

ed in strength and became progressively colder bring-

ing with it increased drought to the west coast of

tropical Africa. The extensive Pliocene ice sheet

would have brought a much drier climate not only to

the coast of west tropical Africa but it may possibly

also account for the dry global climate in the mid

Pliocene (Raven & Axelrod, 1974).

As aridity spread and a seasonally dry climate

became established, particularly during the Pliocene

as the Antarctic ice cap developed, the African rain-

forest continued to be more and more restricted in

distribution and impoverished and the taxa compris-

ing the forests became progressively more discon-

tinuous. The strengthening high pressure systems

brought a drier climate to the interior of Africa and

the spreading drought tended to disrupt and im-

poverish the African flora, the recurring aridity in

the tropics during successive periods of ‘ice-age aridi-

ty’ resulting in increased selection pressure for

drought resistant taxa. As a consequence, rainforest

areas were replaced by savanna and grassland, savan-

na and grassland by dry thorn scrub and dry thorn

scrub by semidesert and desert vegetation.

The later phases of this trend in the Pliocene pro-

bably resulted in the appearance of local areas of

semidesert, but widespread regional semideserts and

deserts are apparently the consequence of later

phases of ‘ice-age aridity’. According to Quezel

(1979), a desert climate was probably initiated in the

major part of the Sahara, at least in the lower

altitudinal zones, during the Pliocene.

Throughout the Tertiary there was a symmetrical

distribution of climate and vegetation in the central

tropical belt. The present African vegetation shows

much greater asymmetry than that of the early to late

Tertiary (see Axelrod & Raven, Fig. 6, 1978). White

(1965) discussed the marked differences that exist at

present between the Sudanian and Zambezian floris-

tic domains. The Sudanian Domain is much drier

than the Zambezian and its greater aridity has been

largely responsible for the impoverishment of its

flora.

The close of the Pliocene and the early Pleistocene

were characterized by major uplifting which raised

the interior plateaux by over 1 700 metres. The alti-

tude of parts of eastern Africa has increased by over

2 000 metres above that of the Miocene and has

brought to it a drier climate. The Pliocene-Pleisto-

cene uplifting and associated climatic fluctuations

favoured local speciation.

Fluctuations in the Quaternary climate also had a

significant effect on the tropical African rainforest

flora, with the drier periods being times of extinction

of taxa requiring more or less continuously wet con-

ditions. Rainforest expanded during the humid inter-

glacial periods and contracted again during the dry

glacial periods. Wild (1968) reconstructed tentative

vegetation maps of Zimbabwe showing how the vege-

tation would have differed from that of today if rain-

fall decreased by 50% or increased by about 150%

above present levels. Wild demonstrated that if rain-

fall increased by 150% above present levels, Zimbab-

wean forests that are now isolated would have been

sufficiently widespread to have been in contact with

the main forest areas of Zaire and West Africa which

would explain why some species in isolated Zimbab-

wean forests have west African affinities. Wild sug-

gested that Quaternary pluvials of only 50% higher

rainfall would probably have resulted in more or less

continuous forest at lower altitudes through much of

tropical Africa, but Axelrod & Raven (1978) con-

sidered this unlikely unless a considerable amount of

rain fell during the present dry winter season so that

the rainfall was fairly evenly distributed throughout

the year, a situation which was itself considered un-

likely because of the strength of the then prevailing

anticyclonic circulatory systems. Axelrod & Raven

suggested that the present links in forest taxa between

the Zaire-West Africa and the relic forest patches in

Zimbabwe may date from the early Miocene.

Even during the past 20 000 years there have been

major climatic changes over much of Africa (Van

Zinderen Bakker, 1974) emphasizing that continued

existence is not possible without continuous change.

The tropical African rainforests continue to contract

rapidly as a consequence of human activities and pro-

bably to a lesser extent because of climatic changes.

It is against this background of continuous change

that the present distribution of the Acacia species

must be seen. Just as the present distributions differ

from those of the past, so too will those of the future

differ from those of the present. Indeed, the present

conservation status of a number of species is uncer-

tain, especially of some of the endemic species with

restricted distributions in the Horn of Africa.

ANALYSIS OF THE AFRICAN ACACIA SPECIES

The number of species recorded from each country

in Africa is indicated in Table 1 , the countries corres-

ponding with the usual political boundaries except

that, for the sake of convenience, the territory of the

Afars and Issars has been included with Somalia.

Table 1 was compiled from data contained in a con-

spectus of the African species (Ross, 1979) which was

itself based on an examination of specimens in

several African, British and European herbaria and

on information contained in the regional African

floras. The African Acacia species remain incom-

pletely known and numerous taxonomic problems

J. H. ROSS

395

await elucidation, especially in north-east tropical

Africa. For the purpose of Table 1 and in the discus-

sion which follows the 115 species accepted by Ross

(1979) have been taken as the number of species for

the continent (this figure excludes A. macrothyrsa

Harms which is now considered (Hunde, 1979) to be

a synonym of A. amythethophylla A. Steud. ex A.

Rich.). Taxa such as A. farnesiana (L.) Willd., which

is not thought to be indigenous, and A. schlechteri

Harms and A. andongensis Welw. ex Hiern, about

whose precise taxonomic status there is some doubt,

have been excluded as have the A. erioloba E. Mey x

A. haematoxylon Willd. hybrid and other hybrids,

A. purpurea Bolle, A. mauroceana DC. and A. calli-

coina Meisn., which are names of uncertain applica-

tion, and A. sp. near Senegal, A. sp. near somalensis

and others which are insufficiently known. The 115

species recognized by no means represent the final

total number of species for the continent but this

figure does serve as a basis, imperfect as it is, for an

analysis of the African species. Because of the varia-

tion in the size of individual countries and because no

country has species evenly distributed throughout it,

the number of species per country is of somewhat

limited value alone. Furthermore, the distribution of

species within countries in tropical east, south-east

and southern Africa is far better documented than in

a number of countries in other areas of the continent.

However, despite these limitations and, although

perhaps the figures provided in Table 1 are incorrect

in some details and likely to need alteration in the

light of additional information, it is believed that the

overall patterns that emerge are sufficiently accurate

to be of value.

Examination of Table 1 reveals, not surprisingly,

that the highest concentrations of species occur in

countries in tropical north-east, east and south-east

Africa. The figures suggest that there is a tendency in

most countries in tropical north-east, east, south-

east, southern and south-west Africa for subgenus

Acacia to be proportionately better represented than

subgenus Aculeiferum, although Mozambique is an

obvious exception to this generalization, while in

Zaire and in countries to the north-west subgenus

Aculeiferum is often numerically as important as or

TABLE 1. — The representation of the Acacia species within each African country

396

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Fig 4. — The general distribution

of the Acacia species in Africa

and an indication of the con-

centration of species over the

distributional range of the

genus.

even proportionately more important than subgenus

Acacia although once again there are a number of ex-

ceptions. This apparent proportional preponderance

of subgenus Aculeiferum in Zaire and parts of

tropical west Africa may possibly be due to the

presence of extensive forested areas which climbing

species in subgenus Aculeiferum have been able to

colonize but from which subgenus Acacia has been

largely excluded.

The general distribution of the genus Acacia in

Africa and an indication of the concentration of

species over the distributional range is shown in Fig.

4. As in the case of the information in Table 1 , Fig. 4

is more accurate in some areas than in others and this

uneven treatment is a reflection of the uneven

knowledge of the genus over its range of distribution.

The genus is widely distributed over the continent

being absent only from the extreme northern portion

of north Africa, part of Mauritania and western

Sahara in West Africa, the vicinity of Cape Town in

the extreme south-western tip of the continent and

from parts of the west coast in South West Africa.*

Tunisia is the only political entity on the continent in

which no indigenous Acacia species are found. The

greatest concentration of species occurs in tropical

east and south-east Africa and, as one would expect,

fewest species occur in desert regions to the north and

south and in the rainforest areas, particularly in

‘Namibia

Zaire and in parts of tropical west Africa. Although

the genus is so widespread and forms such a con-

spicuous feature of the landscape over much of the

continent, the number of species found in Africa is

lower than one might expect and represents less then

one-sixth of the number of species found in

Australia.

Flaving noted the distribution of the genus as a

whole in Africa, the distributions of subgenus Acacia

and of subgenus Aculeiferum are now examined (see

Figs 5 & 6). Of the 115 species accepted for Africa, 52

belong to subgenus Aculeiferum and 63 to subgenus

Acacia.

It is at once apparent that there are significant dif-

ferences in the distributional ranges of the two sub-

genera. Subgenus Acacia extends far further north

than subgenus Aculeiferum being found for the most

part as far north as 30° N latitude except in Algeria

and Morocco where the subgenus occurs even further

north and in the Nile valley in Egypt where a species

occurs in the Nile delta. Subgenus Aculeiferum, on

the other hand, does not occur much north of 20° N

latitude except in Egypt and more particularly in the

Nile Valley. In the extreme south of the continent

subgenus Acacia occurs to within 100 km of Cape

Town while subgenus Aculeiferum has not succeeded

in penetrating the south-west tip of the Cape Pro-

vince and is also absent from the high country in

Lesotho and the eastern Orange Free State. Both

subgenera are absent from parts of the west coast of

J. H. ROSS

397

Fig 5. — The general distribution

of subgenus Acacia in Africa

and an indication of the con-

centration of species over the

distributional range of the

subgenus.

South West Africa bordering the Atlantic Ocean and

from part of western Mauritania. Although the range

of distribution of some individual species in subgenus

Aculeiferum, for example A. ataxacantha, is almost

as extensive as that of some of the more widespread

species in subgenus Acacia such as A. sieberana DC.

(see Fig. 7), the total distributional range of subgenus

Acacia greatly exceeds that of subgenus Aculeiferum.

The fact that subgenus Acacia enjoys a wider range

of distribution on the continent than does subgenus

Aculeiferum suggests that the former has been able

to occupy habitats, particularly the harsh habitats in

the central Sahara and the high country of Lesotho,

from which subgenus Aculeiferum has been excluded.

Most (? all) members of subgenus Acacia are poly-

ploid while those of subgenus Aculeiferum are dip-

loid and the possibility exists that polyploidy has con-

ferred greater genetic plasticity on subgenus Acacia

which has enabled members of the subgenus to suc-

cessfully occupy a greater diversity of habitats. On

the other hand, however, some of the climbing spe-

cies in subgenus Aculeiferum have been successful in

the forested areas of the continent, particularly in

Zaire and in parts of tropical west Africa, in which

subgenus Acacia is not represented. It would appear

that the climbing habit, which is unknown in sub-

genus Acacia, has enabled members of subgenus

Aculeiferum to exploit situations in forested areas

where light penetrates to the ground. Subgenus

Acacia is apparently absent from Liberia, Sierra

Leone, Gabon, Equatorial Guinea and the densely

forested areas in Zaire.

Apart from the exceptions noted above, the distri-

bution of species in subgenera Acacia and Aculei-

ferum over the remainder of the continent is roughly

similar although the number of species in individual

areas within each subgenus varies. The highest con-

centration of species in both subgenera occurs in

tropical east and south-east Africa.

The pattern of distribution exhibited by A. sieber-

ana (see Fig. 7) is fairly representative of that shown

by a number of widespread species in both sub-

genera, extending from Senegal in the west to the

Sudan or Ethiopia in the north-east and down tropi-

cal east Africa through Kenya, Uganda and Tanzania

skirting around the forested areas of central Zaire

with one arm swinging westwards through Zambia,

south-east Zaire, Zimbabwe and Botswana to Angola

and northern South West Africa and another arm

continuing southwards through Mozambique, the

Transvaal and Swaziland into Natal. The two arms

of distribution approximate roughly to the tempera-

ture and rainfall belts.

The distribution of species within some of the in-

dividual countries is now considered in more detail.

Two species, both members of subgenus Acacia,

have been recorded from southern Morocco in the

extreme north-west of the continent, namely, A.

398

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Fig 6. — The general distribution

of subgenus Aculeiferum in

Africa and an indication of

the concentration of species

over the distributional range

of the subgenus.

seval Del. and A. gummifera Willd. While A. seyal is

widely distributed in Africa, A. gummifera is en-

demic in Morocco and has a very restricted distribu-

tion within the country. Morocco is the only country

in north or west Africa, which has an endemic Acacia

species. Although A. gummifera is adapted to a

Mediterranean climate Quezel (1979) considered it to

be of tropical origin and represent a ‘differentiated

vestige’ which has been ‘in situ’ since the Pleistocene.

A. gummifera is most closely allied to the tropical

and widespread A. nilotica (L.) Willd. ex Del. and

possibly arose from A. nilotica stock that became

isolated as a result of climatic vicissitudes which have

largely eliminated the tropical element from Mediter-

ranean Africa.

The northern limit of Acacia in Libya and Algeria

corresponds with the northern limit of distribution of

A. tortilis (Forssk.) Hayne subsp. raddiana (Savi)

Brenan. The Sahara is inhabited by very few Acacia

species and, according to Quezel (1979), the evidence

indicates that they penetrated into the Sahara only at

the end of the last pluvial, probably after several

previous phases of extension, but did not reach the

Mediterranean regions because of their thermal

demands.

In Egypt most species occur in a zone along the

Nile River with fewer species occurring in the adja-

cent desert areas. A. nilotica follows the course of the

Nile into the delta itself.

Thirty-one species, none of which is endemic,

occur in the Sudan. The highest concentration of

species occurs in the south-east where the territory

adjoins Ethiopia, Kenya and Uganda, and the

numbers decrease quite sharply in the south-west,

central and northern areas. The genus is very poorly

represented in the north-western portion of the coun-

try.

Ethiopia, with its diversified topography, has 43

species, the second highest number among the Afri-

can countries. In addition to being rich in species,

Ethiopia is an important area of speciation for

Acacia, each subgenus having three endemic species.

Among the endemics are A. walwalensis Gilliland

and A. pseudonigrescens Brenan & J. H. Ross, two

very distinctive species confined to the Ogaden. The

highest number of species occurs in the north-east

adjacent to the Red Sea, but the Harar Province and

Ogaden are also rich in species. A. reficiens Wawra

has a very disjunct distribution (see Fig. 8) : subsp.

misera (Vatke) Brenan occurring in north-eastern

Uganda, south-eastern Sudan, Kenya, eastern

Ethiopia and Somalia and subsp. reficiens occurring

in south-western Angola and northern South West

Africa, the species providing a good example of the

well known distributional discontinuity between the

more arid areas of South West Africa and the north-

east Horn of Africa. Verdcourt (1969) and De Winter

(1971) provided examples of disjunctions in other

J. H. ROSS

399

genera and families which furnish evidence of a

former arid corridor across the continent from the

north-east to the south-west. Conditions for direct

migration were probably best when arid phases of the

Pleistocene were at a maximum and distances bet-

ween the arid zones were least or possibly when a

continuous arid belt extended across Africa from the

north-east to the south-east. The distribution of A.

stuhlmannii Taub. provides a less extreme example

of the disjunct distribution illustrated by A. reficiens,

while some species have a more or less continuous

distribution from the north-east to the south-west.

Somalia has a remarkable flora and the Acacia

species are no exception. Although only 32 Acacia

species have been recorded from Somalia, nine are

endemic (seven in subgenus Aculeiferum and two in

subgenus Acacia). Somalia has been an important

centre of speciation in the ‘A. Senegal (L.) Willd.

complex’ where six very distinctive endemic taxa

have arisen, namely, A. ankokib Chiov., A. cara-

niana Chiov., A. cheilanthifolia Chiov., A. ogaden-

sis Chiov., A. somalensis Vatke and A. sp. near A.

somalensis. A. zizyphispina Chiov., another member

of the A. Senegal complex’, although not endemic in

Somalia itself, has a very restricted distribution in

southern Somalia and in the Ogaden in Ethiopia. A.

bricchettiana Chiov. (subgenus Acacia) is similarly

confined to the Ogaden region in Somalia and Ethio-

pia. Several other species, for example A. condvlo-

clada Chiov., yet another member of the ‘A. Senegal

complex’, and A. edgeworthii T . Anders, (subgenus

Acacia) are confined to Ethiopia, Somalia and the

Northern Frontier Province of Kenya (A. edge-

worthii also occurs on Socotra), while A. turnbul-

liana Brenan is confined to Somalia and the Northern

Frontier Province of Kenya. A. leucospira Brenan,

another endemic, is a very distinctive species with

minute laterally compressed leaflets which are remi-

niscent of those of A. haematoxylon in southern

Africa (see Fig. 9). Small laterally compressed leaf-

lets are unknown elsewhere amongst the African

species but A. leucospira and A. haematoxylon are

Fig 8. — The known distribution of Acacia reficiens.

not closely related. The Horn of Africa has been an

important centre of speciation in Acacia.

Kenya, with 42 species, has only one endemic spe-

cies, namely, A. thomasii Harms, a member of the

‘A. Senegal complex’, while A. paolii Chiov. subsp.

paucijuga Brenan is endemic in the north-west of the

country. The poverty of endemism in Kenya is in

marked contrast to Tanzania. The distribution of

Acacia in Kenya according to the provinces recog-

nized in the Flora of Tropical East Africa is as fol-

lows: Kl, Northern Frontier : 26 species; K2,

Fig 9.— The known distributions of Acacia leucospira and A.

haematoxylon.

400

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Turkana : 18 species; K3, Rift Valley : 15 species; K4,

Central : 23 species; K5, Nyanza : 12 species; K6,

Masai : 19 species; K7, Coast : 23 species. These

figures illustrate the richness of the Northern Fron-

tier Province, mainly because of the presence there of

the Somali element, the central and Coastal Pro-

vinces, and the relative poverty of western Kenya.

Uganda, with 27 species, has far fewer species than

either Kenya or Tanzania and the relative poverty

evident in western Kenya is again evident over much

of Uganda. The distribution of Acacia in Uganda

according to the provinces recognized in the Flora of

Tropical East Africa is as follows: Ul, Northern : 24

species; U2, Western : 11 species; U3, Eastern : 16

species; U4, Buganda : 13 species. The Northern

Province which is well watered in the west is by far

the richest province.

Tanzania, with 50 species, has significantly more

species than any other country and both subgenera

Acacia and Aculeiferum are best represented in Tan-

zania. In addition, ten of the species, three in sub-

genus Aculeiferum and seven in subgenus Acacia, are

endemic in Tanzania. The distribution of Acacia in

Tanzania according to the provinces in the Flora of

Tropical East Africa is as follows: Tl, Lake : 23 spe-

cies; T2, Northern : 28 species; T3, Tanga : 28 spe-

cies; T4, Western : 29 species; T5, Central : 27 spe-

cies; T6, Eastern : 21 species; T7, Southern High-

lands : 14 species; T8, Southern : 16 species. These

figures illustrate the relative poverty of Acacia

species in the Southern Highlands and Southern

Province in contrast to the remainder of the country

over which the species are fairly evenly spread. Tan-

zania has been an important centre of speciation in

the ‘ A . drepatiolobium Harms ex Sjostedt complex’,

a complex of species with characteristically enlarged

stipular spines (‘ant-galls’). In addition to the wide-

spread A. drepanolobium, six species with enlarged

stipular spines, namely A. bullockii Brenan, A. burt-

tii Bak. f A. erythrophloea Brenan, A. malacoce-

phala Harms, A. mbuluensis Brenan and A. pseudo-

fistula Harms, are endemic in Tanzania and two of

them, A. bullockii and A. erythrophloea, are

endemic in the Western Province. The Western Pro-

vince is outstanding on account of the number of

endemics found within it: seven of the ten endemic

Acacia species in Tanzania are found within the

Western Province although only the above two

species are confined to it. A. taylorii Brenan & Exell,

a member of the ‘A. pennata complex’ with scattered

recurved prickles, is endemic along the coast in the

Southern Province towards the Mozambique border,

A. tephrodermis Brenan in the same complex is

endemic in the Eastern Province and A. latistipulata

Harms, yet another member of the complex, occurs

from central Tanzania to central Mozambique. A.

ancistroclada Brenan, although not endemic in Tan-

zanzia itself, has a restricted distribution in north-

east Tanzania and south-east Kenya. A. stuhlmannii

has a discontinuous distribution being recorded from

Ethiopia, Somalia, Kenya and northern and central

Tanzania in the north and from Zimbabwe,

Botswana and the Transvaal in the south.

Zaire illustrates very well the point made earlier

that Acacia species are not spread uniformly

throughout a country. Twenty-four species are

recorded from Zaire but the great majority are con-

fined to the mainly wooded grasslands of the Ubangi-

Uele, Lac Albert, Lacs Edouard et Kivu, Bas-

Katanga and Haut-Katanga regions (phytogeograph-

icai regions used in Flore du Congo, du Rwanda et du

Fig 10. — The known distributions of Acacia lujae and A. kraus-

siana.

Burundi) in the north, north-east, east and south-east

respectively. Only four species, A. kamerunensis

Gandoger, A. ciliolata Brenan & Exell, A. pentagona

(Schumach.) Hook. f. and A. lujae, all climbing

species in subgenus Aculeiferum, occur in the Fores-

tier Central region. A. lujae is the only endemic

Acacia species in Zaire. A. lujae seems more closely

related to A. kraussiana, which is endemic along the

coast in southern Mozambique and Natal and from

which it is separated by a wide interval, than it is to

any members of the ‘A. pennata complex’ (see Fig.

10).

Acacia is very poorly represented in Congo Brazza-

ville, Cabinda, Gabon and Equatorial Guinea, the

sole member of the genus in each of the last three

countries being a climbing member of subgenus Acu-

leiferum. In contrast to tropical east Africa, the

genus is generally poorly represented in tropical west

Africa.

Mozambique, with 40 species, has only one endem-

ic species, namely A. torrei Brenan, a member of the

complex with glandular glutinous pods which is

centred on the Transvaal in South Africa. Although

not endemic in Mozambique itself, a number of other

species occurring in Mozambique have restricted dis-

tributions outside the country. A. latistipulata and

A. kraussiana have already been alluded to but

another example is A. eriocarpa Brenan whose distri-

bution extends from Mozambique into eastern Zam-

bia and central and western Zimbabwe. It is not clear

whether or not A. rovumae Oliv. occurs in Mozam-

bique as there are no definite records of the species

from the territory. The type locality of A. rovumae

was given as ‘Rovuma Bay’ without any indication of

whether it was collected from the north or south side,

crucial information as the Rovuma river forms the

boundary between Mozambique and Tanzania. The

absence of any definite records of the species from

Mozambique suggests that it does not occur in the

territory. The distribution of Acacia in Mozambique

according to the provinces recognized in Flora

J. H. ROSS

401

Zambesiaca is as follows: Niassa : 22 species;

Zambezia : 16 species; Tete : 21 species; Manica e

Sofala : 20 species; Sul do Save : 17 species; Lourenco

Marques : 23 species. The highest numbers of species

occur in the extreme north and south of the country

and in the Tete province.

Of the 38 species occurring in Zimbabwe only A.

chariessa Milne-Redhead, which is almost always

found on serpentine soils, is endemic. Vegetation

growing on serpentine soils often shows stunting or

depauperation (Wild, 1974) and A. chariessa often

looks superficially like a diminutive form of the

widespread A. ataxacantha, although differing from

it in several significant respects. The distribution of

Acacia species in Zimbabwe according to the prov-

inces recognized in Flora Zambesiaca is as follows:

North : 19 species; East : 17 species; Central : 15 spe-

cies; West : 27 species; South : 27 species. The great-

est numbers of species occur in the south and west,

the numbers there being increased by the presence of

several members of the complex with glandular

glutinous pods, namely A. borleae Burtt Davy, A.

exuvia/is Verdoorn, A. nebrownii Burtt Davy and A.

permixta Burtt Davy, and species such as A. erioloba

E. Mey. and A. luederitzii Engl, which favour dry

woodland and often occur on the Kalahari sands.

No endemic species occur in Zambia, Malawi or

Botswana. The southern province of Zambia, which

to some extent represents a northern extension of the

western province of Zimbabwe, is by far the richest

province, while the western province with only six

species is the poorest. The north and south-east pro-

vinces of Botswana each have 23 species but the

genus is poorly represented in the arid south-west

province where only 9 species occur.

The highest number of species in South West

Africa occurs in the north-west and the highest

number in Angola in the south-west in the Huila,

Mossamedes and Benguela Districts. Two of the 23

species in South West Africa, namely A. montis-usti

Merxm. & Schreiber and A. robynsiana Merxm. &

Schreiber, are endemic in the north-west and two of

the 25 species in Angola, A. antunesii Harms and A.

quintanilhae Torre, are endemic in the south-west.

While both endemics in South West Africa are

members of subgenus Aculeiferum, the endemics in

Angola both belong to subgenus Acacia. In addition

to these endemics with narrow distributional ranges,

A. hebeclada DC. subsp. tristis Schreiber is confined

to much the same area in north-western South West

Africa and south-western Angola. Reference has

already been made to the discontinuous distribution

shown by A. reficiens but A. mellifera (Vahl) Benth.

subsp. mellifera also has a disjunct distribution

occurring in northern South West Africa and south-

western Angola (Mossamedes and Benguela Dis-

tricts) in the south and in Tanzania and territories to

the north. The number of Acacia species in Angola

falls away from the south-west to the north and east

particularly towards the forested areas in the north

although A. welwitschii Oliv. subsp. welwitschii is

endemic in northern Angola. South-western Angola

and north-western South West Africa (with the

exception of a narrow strip along the coast from

which the genus is absent) appear to be one of the

more important centres of speciation in Acacia in

southern tropical Africa.

The southern limit of distribution of the genus in

Africa occurs in the Cape Province in South Africa

and, as one would expect, the number of species in

South Africa declines rapidly to the south particular-

ly along the east coast. Twenty-one species, all of

which occur in Zululand and Tongaland in the north,

are recorded from Natal. However, impoverishment

to the south in Natal is fairly rapid as a number of

species reach their southern limit of distribution in

Natal. Of the twenty-one species which occur in

Tongaland and Zululand in the north, only thirteen

occur south of the Tugela River, and of these thirteen

only five species extend south of the Umtamvuna

river into Transkei. The number of species in Natal

also falls away fairly rapidly with increasing altitude

towards the interior and the widespread A. karroo

Hayne is the only species in Lesotho. The majority of

species in Swaziland occur in the lowveld in the east

with fewer species in the higher areas. The Transvaal,

with 35 species, has the highest number of species of

all of the provinces in South Africa. Once again, the

majority of species occur in the lowveld in the east

and in the north with fewer species occurring in the

highveld region with its colder winters. A number of

species favouring sandy soils occur in the western

portion of the province. The complex of species with

glandular glutinous pods appears to be centred in the

Transvaal where six of the seven species within the

complex are found. Only A. torrei, which is endemic

in Mozambique, is absent. Eleven species occur in the

Cape Province one of which, A. redacta J. H. Ross,

is endemic. A. redacta is an extremely interesting

species in many respects showing no close affinity

with any other species and the possibility exists that it

is not an Acacia at all but is referable to a new mono-

typic genus. Although not confined to the northern

Cape, A. haematoxylon has a restricted distribution

outside the province occurring in the extreme south-

west of Botswana and in the eastern portion of South

West Africa, while the A. erioloba x A. haematoxy-

lon hybrid is confined to the northern Cape. Acacia

is best represented in the northern Cape on the

Kalahari sands and only A. karroo is common over

much of the southern and western portion of the

province.

TABLE 2.— The countries within which endemic Acacia species occur and the number of endemic species within each country

402

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Fig 11. — The distribution of the narrow endemics within

subgenus Acacia referred to in Table 2.

The countries in which endemic species occur and

the number of endemics recorded within each are

given in Table 2.

Because of the variation in the size of individual

countries and because no country has an even spread

of endemism, the number of endemic species per

country is of limited value alone. For example, Tan-

zania with ten endemic species out of a total of fifty

species has a greater number of endemics than

Somalia with nine endemics out of a total of 32

species, but the proportion of endemics in Somalia is

greater. However, despite the limitations, the infor-

mation in Table 2 is nevertheless fairly instructive.

Table 2 shows that the highest numbers of endemic

species occur in Tanzania, Somalia and Ethiopia in

tropical east and north-east Africa. Interestingly, the

highest number of endemic species in subgenus

Acacia occur in Tanzania while the highest number

of endemics in subgenus Aculeiferum occur in

Somalia. Morocco is the only country in north or

west Africa which has an endemic species. Although

the Sudan has one more species than Somalia, the

Sudan has no endemic species while Somalia has

nine, and Ethiopia with one more species than Kenya

has six endemics and Kenya only one. The distribu-

tion of the narrow endemics within each subgenus

referred to in Table 2 are illustrated in Figs 1 1 and 12.

The endemism in the genus Acacia in Africa is

shown in Table 3.

Fig 12.— The distribution of the narrow endemics within

subgenus Aculeiferum referred to in Table 2.

The high proportion of species endemic in one

country in contrast to the much lower figures for

those species endemic in two or three countries

emphasizes the prevalence of species with narrow

ranges.

It is instructive to briefly compare the distribution

of the African Acacia species in Fig. 4 with the

phytogeographical regions of Africa (see Fig. 13).

Fig. 13 is a slightly simplified version adapted from

the scheme accepted by Brenan (1979) which was

itself based on the earlier works of Wickens (1976),

White (1965) and Chapman & White (1970). Wickens

(/.c.) recognized eight Regions in Africa, three of

which were themselves divided into domains.

The Sudano-Zambezian Region, which is charac-

terized by a strong seasonal climate, corresponds to

the tropical savanna and is by far the largest Region

in Africa extending north and south of the equator

but physically continuous by a relatively narrow con-

nection in east Africa. To the north it is bounded by

the deserts and semideserts of the Sahara, in the cen-

tre it circumscribes the limits of the forests of the

Guineo-Congo Region, and in the south it extends to

the deserts and semideserts of the Karoo Namib

Region and the Cape Region. As is to be expected,

the majority of Acacia species occur within this

region.

1 . The Sahelian Domain extends from Mauritania

and Senegal on the Atlantic in the west to the Red Sea

TABLE 3. — Endemism in the genus Acacia in Africa

J. H. ROSS

403

Fig. 13. — The phytogeographical

regions and domains of Africa

(After Brenan, 1979).

coast of the Sudan in the east forming a narrow belt

bounded on the north by the Saharo-Sindian Region.

The domain thus separates the desert flora of the

Sahara from the deciduous woodlands and other

communities of the Sudanian domain. The southern

boundary of the Sahelian domain largely corre-

sponds to the southernmost extent of the Pleistocene

sand invasion (Wickens, 1976). A number of wide-

spread species such as A. tortilis, A. seyal, A. ehren-

bergiana Hayne and A. nubica Benth. are character-

istic of this domain but no endemic species occur

within it.

2. The Sudanian Domain extends from Senegal to

the eastern border of the Sudan and forms a wide belt

bounded to the north by the former domain and to

the south by the Guineo-Congo Region. Acacia is

well represented within the deciduous woodlands and

savannas of this domain, most of the species occur-

ring within it being fairly widespread in tropical

Africa and some extending to Asia. The Sudanian

domain is much drier than the Zambezian domain

and its greater aridity has been largely responsible for

the marked differences between the two domains, the

impoverishment of the flora of the former, and the

asymmetrical distribution of savanna-type vegetation

north and south of the equator (White, 1965).

3. The Afro-oriental Domain, which includes the

‘Horn of Africa’ and extends southwards into Tan-

zania, contains by far the greatest number of Acacia

species and the greatest number of endemics illustra-

ting the importance of this area for speciation in

Acacia.

4. The Zambezian Domain consists of the remainder

of the Sudano-Zambezian Region lying to the south

of the Afro-oriental domain and is also rich in species

and endemics. The main centre of endemism within

this domain is north-western Tanzania while a secon-

dary centre occurs in the area occupied by Zim-

babwe, Botswana, south-western Mozambique and

the northern Transvaal.

The Guineo-Congo Region represents the main

evergreen or partly evergreen forests of Africa

extending from Senegal in the west, southwards to

Angola and eastwards to Ruwenzori. The climate is

characteristically one of high even temperatures and

rainfall more or less throughout the year. As is to be

expected of a genus which is intolerant of low light

intensities, Acacia species are not very well represen-

ted within this region tending to occupy forest

margins or clearings where openings in the canopy

permit light to penetrate. The climbing species in

subgenus Aculeiferum tend to be better represented

in the forested region than the trees and shrubs of

subgenus Acacia. Three domains are recognized

within the region, namely, the Guinea, Congo and

Usambaro-Zululand domains. The Usambaro-Zulu-

land domain, which consists of the scattered relics of

rainforest along or near the east coast of Africa, has

the highest number of species of the three domains

404

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

and five endemics, all members of subgenus Aculei-

ferum, occur within it. One of the endemics, A.

kraussiana occurring in Natal and southern Mozam-

bique, seems closely related to A. lujae, the only

endemic species in the Congo domain. Neither the

Congo nor the Guinea domain is very rich in Acacia

species which is not surprising as these are the best

forested domains in the region.

Acacia is absent from the Afromontane Region

which corresponds to the Montane Forest and Erica-

ceous belts of the tropical mountain regions and

from the Afroalpine Region which is the zone above

the Ericaceous belt of the former Region.

The Saharo-Sindian Region is a region of desert

and semidesert to the north of the Sahelian domain

extending from western Sahara, Mauritania and

southern Morocco eastwards across Sinai to India.

The climate is characterized by high temperatures,

often with extreme diurnal fluctuations, and low and

often irregular rainfall. A few widespread species

such as A. tortilis, A. seyal, A. ehrenbergiana and A.

nilotica occur in parts of the region and there is a

local concentration of species in the Nile valley.

According to Quezel (1979), the Sahelian species

penetrated into the Sahara only at the end of the last

pluvial but did not reach the Mediterranean Region

because of their thermal demands.

The Mediterranean Region, characterized by hot

dry summers with mild winters during which much of

the rain falls, is the northernmost fringe of Africa

bounded in the south by the Saharo-Sindian Region

and in the north by the Mediterranean. A. gummi-

fera is endemic in Morocco and a few species occur in

the Nile valley but, apart from these, A. tortilis is

possibly the only other species to occur in parts of the

Region.

The Cape Region, dealt with in detail by Goldblatt

(1979), occupies the extreme south-west of the conti-

nent and has a Mediterranean type climate in which

the genus Acacia is unimportant and represented by

only a few species. A. karroo and A. caffra are the

most widespread species within the Region.

The Karoo-Namib Region occupies the interior of

the Cape Province, the western portion of South

West Africa and extends into south-western Angola.

Except for the southern portion of the Cape Province

and isolated pockets along the Atlantic coast, Acacia

is fairly well represented throughout the area, par-

ticularly in northern South West Africa and south-

western Angola. A. erioloba, A. hereroensis Engl.,

A. karroo, A. tortilis and A. mellifera are

characteristic of many areas. Four endemic species

occur in northern South West Africa and south-

western Angola, an important local centre of specia-

tion in Acacia.

RELATIONSHIP BETWEEN THE AFRICAN AND

AMERICAN SPECIES

The currently accepted sequence of events which

marked the begining of the fragmentation of West

Gondwanaland and the separation of the African

and South American plates was outlined by Raven &

Axelrod (1974) and is summarized briefly here.

Separation of the African and South American plates

started about 127 million years ago with the final

marine connection between the North and South

Atlantic occurring about 100 million years ago. The

two continents remained in close proximity with only

a narrow strait separating the present Gabon from

north-eastern Brazil for another 10 million years. By

the end of the Cretaceous, about 65 million years

ago, the two continents are thought to have been

separated at their closest points by about 800 km

although they are said to have been linked by

numerous volcanic islands. South America moved

away from Africa and gradually converged with

North America becoming equidistant between the

two early in the Eocene about 50 million years ago.

South America subsequently became more isolated

from Africa and closer to North America until a

direct land connection between North and South

America occurred in the Pliocene 7-5 million years

ago. This sets the background against which the

African and American species must be seen.

There is no evidence and it seems unlikely that

Acacia or its prototype had differentiated when

Africa separated from South America about 100

million years ago, and it is not clear whether the

genus crossed the Atlantic in the late Cretaceous

about 65 million years ago (by which time the three

subfamilies of Leguminosae are known to have been

differentiated) when the two continents were

separated but thought to have been linked by

numerous volcanic islands, or whether the genus was

carried over the Atlantic in Paleocene or Eocene

time. From what is known or can be inferred about

its history and present distribution patterns Raven &

Axelrod (1974) presume that Mimosoideae migrated

between Africa and South America during or prior to

the Paleocene when the Atlantic was much narrower

than at present, while Thorne ( 1 978) is of the opinion

that migration could have occurred in the late

Cretaceous, Paleocene or Eocene. The evidence sug-

gests that only limited migration occurred between

these continents after the Paleocene.

Both subgenera Acacia and Aculeiferum are

present in Africa and in America. However, the

representation of subgenus Aculeiferum in each con-

tinent is different: sections Monacanthea and

Aculeiferum occurring in Africa and sections

Monacanthea and Filicinae in America, i.e. Filicinae

is absent from Africa and Aculeiferum is absent from

America. The position of A. willardiana Rose, which

has a horizontally flattened petiole and is endemic to

the west coast of North America (Mexico), is uncer-

tain and opinions differ as to whether it should be

referred to subgenus Aculeiferum or subgenus

Heterophyllum. Vassal & Guinet (1972) included A.

willardiana in subgenus Heterophyllum while Pedley

(1975) argued that the species could be accommo-

dated without difficulty in subgenus Aculeiferum as

it showed relationships to American species of the

subgenus. In support of their earlier contention that

A. willardiana belonged to subgenus Heterophyllum

Guinet & Vassal (1978) pointed out that the seed of

A. willardiana contains the amino-acid ‘willardine’

(Gmelin, 1959) which has also been recorded from

the Australian species A. podalyriifolia A. Cunn. ex

G. Don and A. dealbata Link both of which are

members of subgenus Heterophyllum. However,

‘willardine’ is also present in the Asian A. modesta

Wall., a member of subgenus Aculeiferum (Evans et

al, 1977). The available evidence suggests that A.

willardiana is in fact referable to subgenus

Aculeiferum.

No species in section Monacanthea is common to

Africa and the Americas although some tropical

species on each continent show close similarities, for

example, A. brevispica Harms, A. schweinfurthii

Brenan & Exell, A. pentagona (Schumach.) Hook. f.

and allies in Africa and A. riparia H. B. K. and A.

J. H. ROSS

405

paniculata Willd. in tropical America and the West

Indies.

Although subgenus Acacia is well represented in

both Africa and America no species is common to

Africa and America; A. farnesiana is not indigenous

in Africa having been introduced and subsequently

become naturalized in some areas. A. sieberana in

Africa shows some relationship to the American A.

macrantha Humb. & Bonpl. ex Willd., an observa-

tion noted by Bentham (1875).

A feature shared by some species of subgenus

Acacia in central America and in Africa is the pro-

duction of swollen stipular spines and a mutualistic

association of these spines with several species of

ants, but the degree of mutualism in each continent

differs. Some of the neotropical acacias with swollen

spines, commonly referred to as swollen thorn

acacias or ‘bulls-horn acacias’, have spicate inflores-

cences and others capitate inflorescences and the

species do not appear to constitute a close phyletic

unit, yet they share many adaptive ecological and

morphological traits and, according to Janzen

(1974), provide outstanding examples of evolution-

ary convergence. Janzen (/.c.) observed that the

species of obligate acacia-ants in the New World are

not specific to a swollen-thorn Acacia species, but

rather to its life form. As an example he cited A. col-

linsii Safford, which has at least eight species of obli-

gate acacia-ants living in it over its range from

Mexico to Columbia, all of which also live in other

swollen-thorn acacias.

The acacias with swollen spines in central America

do not appear to occur south of Venezuela or Colum-

bia (Janzen, 1966). The mutualistic interaction be-

tween ants and acacias in central America has been

detailed in a series of papers by Janzen (1966, 1967a,

1967b) and subsequently summarized (Janzen,

1969a). Janzen (1969a) noted that the central Ameri-

can acacias with swollen stipular spines differ from

the other Acacia species in the area in having:

1. partially hollow spines which are occupied by

ants.

2. modified leaflet tips called Beltian bodies which

constitute the primary source of protein and oil for

the ant colony.

3. greatly enlarged foliar nectaries which supply the

sugar requirements for the ant colony.

4. all-year-round leaf production on most individ-

uals which provides a relatively constant source of

food for the ants.

5. an absence of chemical and structural traits that

protect other acacias from most herbivores in the

environment. The ants are functionally analagous to

the chemicals released by some plants in their com-

petitive interactions with other plants; like these

chemicals the ants are ‘produced’at a metabolic cost

to the plant (Janzen 1969b). Janzen (1966) expressed

the view that the swollen-thorn acacias of central

America have lost, apparently through evolutionary

change, their ability to withstand the phytophagous

insect damage and competitive pressure of neigh-

bouring plants without the protection of the obligate

acacia-ants.

Janzen concluded that those Acacia species with

ants do not normally duplicate their defence systems

and thus do not make toxic compounds such as

cyanogenic glycosides in quantity. One exception he

noted was A. chiapensis Safford which possesses

both types of defence systems. Janzen (1974) con-

cluded that A. chiapensis is a marginal host for obli-

gate acacia-ants and in many features of growth and

habit resembles non-ant acacias (Rehr et al., 1973).

However, Siegler et al. (1978) found specimens of A.

hindsii Benth., a species inhabited by an obligate

acacia-ant, in Oaxaca and Jalisco to be strongly

cyanogenic which is another exception to Janzen’s

earlier observation.

The swollen-thorn acacias occur in the wettest

areas of tropical central America. As the drier areas

are approached the acacia cannot retain its leaves

long enough to keep the ant colony alive and the

unoccupied shoot does not survive to maturity

because of insect damage. In cooler areas the growth

of the acacia is slower and the ants are apparently

insufficiently active in cool weather to deter the

phytophagous insects and vertebrate browsers adap-

ted to cool weather and thus the acacia receives more

damage than it can tolerate and the ant colony

starves to death owing to a lack of leaf products.

All of the central American swollen-thorn acacias

have a sweet pulp around the seeds and the seeds are

dispersed by birds. Janzen (1969a) noted that there

appeared to be ‘a selective pressure acting on all the

swollen-thorn acacias that favours bird-dispersal of

seeds’. The species of swollen-thorn acacia with the

widest distributions are those whose seeds are most

readily removed by birds while those with seeds that

are less easily removed have more restricted distribu-

tions. The birds begin dispersal of seeds as soon as

the pods are ripe which is important to plants that

lose 60-100% of a particular seed crop through the

predations of the larvae of Bruchidae. As the initial

infestation usually destroys 40-80% of the seeds and

all seeds remaining on a tree have usually been killed

within two months of seed maturation, the rapid

removal of the seeds by birds is possibly critical to the

survival of the Acacia species.

Like the central American species, the African spe-

cies with swollen stipular spines, commonly called

‘ant-galls’, do not consist of a group of closely

related species. Most of the African ‘ant-gall’ acacias

have white or pale yellowish white flowers in capitate

heads but some have deep yellow flowers and two

species have spicate inflorescences. The African

Acacia species with swollen spines vary from those

which are apparently partially ant-dependent to those

which have no regular mutualistic association with

ants.

Hocking (1970) investigated the East African

swollen-thorn acacias and, although he worked on

several different species, concentrated on A.

drepanolobium Harms ex Sjostedt which is probably

the most ant-dependent African species. Hocking

found that while A. drepanolobium can be grown to

at least flowering stage in the absence of ants and

probably the ants can be raised without the Acacia,

in nature the association is essentially an obligate one

as the ants and A. drepanolobium seldom persist in-

dependently. No more than 1% of the A. drepanolo-

bium plants in the study area were found to be

without ant associates. It follows that advantage

must accrue to both parties so that these associations

are also mutualistic as in the case of the New World

A. cornigera L. (Janzen 1966, 1967b). However,

although converging on the central American system,

the association between A. drepanolobium and the

ants has not reached the same degree of development

as that in the neotropics.

Several species in the ‘A. drepanolobium complex’

are associated with ants to a lesser extent. At the

other extreme are a number of species with swollen

spines, for example A. luederitzii Engl. var. retinens

406

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

(Sim) J. H. Ross & Brenan, which have no mutualis-

tic association with ants, the ants and other insects

merely taking advantage of the hollow spines as suit-

able domatia. The hollow spines in A. luederitzii var.

retinens and in several other species are frequently

unoccupied and often entire plants lack any enlarged

spines.

The African acacias with swollen spines differ

from the American swollen-thorn species in the fol-

lowing respects:

1 . they lack Beltian bodies at the tips of the leaflets.

Hocking (l.c.) suggested that the occurrence of

Beltian bodies at the tips of the leaflets in the New

World Acacia species may ensure a more uniform

distribution of ants on the foliage and be an adapta-

tion to an environment in which phytophagous

insects are a relatively greater threat than browsing

herbivores.

2. they lack all-year-round leaf production.

3. they are not confined to the wettest areas. On the

contrary, in Africa they occupy areas which experi-

ence a pronounced dry season. Hocking (l.c.) sugges-

ted that the establishment of ants on the African

acacias may improve their adaptation to a dry

environment through the pruning out, by the ants, of

the axillary buds of the swollen stipules.

4. the extra-floral nectaries do not appear to be

developed to the same extent.

5. the seeds are not surrounded by sweet pulp.

6. the seeds are not distributed by birds. In Africa

some of the large herbivorous mammals rapidly dis-

perse the seed of certain Acacia species, for example

A. tortilis (Forssk.) Hayne, away from the parent

plant and in so doing play a similar role to that

played by birds in central America.

It is not known whether or not the African species

lack the chemical and structural traits that protect the

American acacias from most herbivores in the en-

vironment and the matter needs investigation.

The genera of ants involved in the mutualistic asso-

ciation with species of Acacia in Africa and in

America differs as one might expect. Pseudomyrmex

is the important ant genus in America while Crema-

togaster is the most important genus in Africa. The

development of swollen spines and the mutualistic

association with ants in Africa and in America

appears to have taken place independently in each

continent and represent an example of convergent

evolution.

RELATIONSHIP BETWEEN THE AFRICAN AND THE

MADAGASCAN AND MASCARENE SPECIES

Madagascar was connected with Africa into the

mid-Cretaceous when it was situated against Tan-

zania-Kenya about 15° N of its presenty position

(Axelrod & Raven, 1978). Madagascar then formed

part of the now largely submerged Mascarene

Plateau which joined India in the east into the late

Cretaceous. Precisely when Madagascar-India separ-

ated from Africa is still not certain but it could have

occurred at any time between the mid- and late Creta-

ceous. India separated from the Madagascar-Mascar-

ene subcontinent early in the Paleocene about 65

million years ago and moved north to meet the Asian

land mass by the middle Eocene about 45 million

years ago (Axelrod & Raven, 1978).

Both subgenera Acacia and Aculeiferum are pre-

sent in Africa and in Madagascar but, in addition,

subgenus Heterophyllum occurs in Madagascar and

the Mascarenes. Subgenus Heterophyllum is essen-

tially Australian so the occurrence of A. xiphoclada

Bak. in Madagascar and A. heterophylla (Lam.)

Willd. in the Mascarenes is of considerable

phytogeographic interest. As species with phyllodes

do not occur on the mainland of any other continent

it is probably reasonable to assume that any species

of Acacia with phyllodes now occurring outside of

Australia must either have come from the Australian

region or have been derived from species which have

(Pedley, 1975). Bell & Evans (1978) found that the

seed of A. heterophylla and all of the Australian

species analysed showed a single characteristic amino

acid pattern which led them to suggest that Australia

and the Mascarene Islands once formed part of the

same land mass and that the seed chemistry of

subgenus Heterophyllum is the seed chemistry that

characterized the ancestral species of Gondwana-

land. A. heterophylla is superficially very similar to

A. koa A. Gray which is endemic in the Hawaiian

Islands some 15 000 kilometres away but differs in

characters of the corolla, pod, seed and seedlings

(Vassal, 1969). Carlquist (1965) postulated that A.

koa and A. heterophylla are probably descendants of

seeds which floated from Australia into the Pacific

and Indian Oceans respectively. While this may be

true, both species are, however, tetraploid and pro-

bably not primitive.

Despite the close proximity of Madagascar to

Africa, as far as is known only one indigenous spe-

cies in subgenus Aculeiferum, namely A. rovumae

Oliv., is common to both Africa and Madagascar. In

tropical east Africa A rovumae occurs on or near the

coast and the appearance of the pods suggests that

they are indehiscent and water-borne. If the pods are

indeed dispersed by water this may possibly account

for the occurrence of the species in Africa and in

Madagascar and it seems reasonable to assume that

migration of the species between the two areas took

place in geologically recent time. There is a very

doubtful and unlikely record of A. pervillei Benth., a

Madagascan species, from Delagoa Bay, Mozambi-

que, but the most likely explanation is that the label

does not belong with the specimen (Ross, 1973). Ben-

tham (1875) was of the opinion that A. pervillei was

more closely allied to the South American A.

lacerans Benth. than it is to any other Old World

species. There is no recent taxonomic revision of the

Madagascan species and several of the species are in-

sufficiently known.

RELATIONSHIP BETWEEN THE AFRICAN AND INDIAN

SPECIES

There are far fewer Acacia species in India than in

Africa but both subgenera Acacia and Aculeiferum

are present in Africa and in India and, moreover, two

species in each subgenus are common to each land

mass which provides clear evidence of a close rela-

tionship between the Indian and African acacias. Al-

though migration between Africa and India is now

extremely difficult or perhaps impossible because of

the intervening arid areas, in former times the two

areas were connected by a belt of tropical forest and

savanna and direct migration was possible. The fact

that a number of the acacias in India and in Africa,

in areas that are now widely separated, cannot be

distinguished at specific level suggests that their

separation is geologically relatively recent and was

attained when direct migration was possible between

India and Africa and not when India ‘rafted’ north-

J. H. ROSS

407

TABLE 4. — Species common to Africa and India and the infraspecific taxa recorded within each species

Africa

India

A. horrida (L.) Willd. subsp.

benadirensis (Chiov.) Hillcoat

& Brenan

A. nilotica (L.) Willd. ex Del.

subsp. nilotica

subsp. tomentosa (Benth.)

Brenan

subsp. adstringens (Schumach.

& Thonn.) Roberty

subsp. subalata (Vatke)

Brenan

subsp. kraussiana (Benth.)

Brenan

subsp. leiocarpa Brenan

A. polyacantha Willd. subsp.

campylacantha (Hochst. ex

A. Rich.) Brenan

A. Senegal (L.) Willd. var.

senega!

var. kerensis Schweinf.

var. leiorhachis Brenan

var. rostrata Brenan

A. horrida subsp. horrida

A. nilotica subsp. indica (Benth.) Brenan

[subsp. hemispherica Ali & Faruqi and subsp. cupressif ormis

(J. L. Stewart) Ali & Faruqi occur in Pakistan]

A. polyacantha subsp. polyacantha

A. senega I var. senega!

wards after breaking from Africa with Madagascar

about 100 million years ago (Raven & Axelrod, 1974)

or perhaps earlier (Schuster, 1976) otherwise one

would have expected greater morphological diver-

sification to have occurred. The morphological dif-

ferences between the species common to Africa and

India are mostly slight but the differences, taken

together with the geographical discontinuity between

the African and Indian populations, have been con-

sidered sufficiently significant to warrant the popula-

tions being regarded as subspecifically or varietally

distinct (see Table 4).

In addition to these species which are common to

both Africa and India, the Indian A. pennata (L.)

Willd., one of the climbers with scattered recurved

prickles in subgenus Aculeiferum, is extremely close-

ly related to a number of African species such as A.

brevispica, A. schweinj urthii, A. pentagona and

allies. Once again, the degree of similarity between

A. pennata and the African species suggests that their

separation is geologically relatively recent. It is clear

that there is a much closer affinity between the

African and Indian species than there is between the

African and South American species.

Although not present in India, A. tortilis, a

member of subgenus Acacia which is widespread in

Africa where it is represented by a number of sub-

species, extends into Arabia as do several other Afri-

can species. A. gerrardii Benth. is represented in the

Negev Desert by subsp. negevensis Zohary, this sub-

species being separated from all of the other variants

in Africa by a wide geographical discontinuity.

The Indian species A. ferruginea (Roxb.) DC., a

member of subgenus Aculeiferum, has some pollen

characteristics which are specific to the Australian

subgenus Heterophyllum, while at least twenty-six

species in subgenus Heterophyllum have a porate

type of pollen with simple appertures which is char-

acteristic of subgenus Aculeiferum (Guinet & Vassal,

1978) illustrating the apparent close relationship be-

tween subgenera Aculeiferum and Heterophyllum.

RELATIONSHIP BETWEEN THE AFRICAN AND

AUSTRALIAN SPECIES

The vast majority of the Acacia species (830 spe-

cies fide Hopper & Maslin, 1978) occur in Australia

where considerable morphological diversity has oc-

curred. Acacia is represented in Africa by subgenera

Acacia and Aculeiferum and in Australia by sub-

genera Acacia, Aculeiferum and Heterophyllum.

Most of the Australian species (over 800) belong to

subgenus Heterophyllum, which is basically Austra-

lian (including Tasmania and New Guinea) although

a few representatives also occur in Madagascar and

the Mascarenes, the Pacific Islands and Hawaii.

Although subgenera Acacia and Aculeiferum occur

in Australia they are poorly represented on the conti-

nent, subgenus Acacia being represented by fewer

than ten indigenous species mainly in the more north-

ern parts of the continent (see Fig. 14) and subgenus

Aculeiferum by the recently described endemic

species (A. albizioicles Pedley) which occurs in the

vicinity of Coen in northern Queensland (see Fig.

15). The only species in subgenus Acacia common to

Africa and Australia is A. farnesiana (apart from A.

nilotica, A. karroo and A. albida Del. which are fair-

ly recent introductions into Australia) but, as already

indicated, A. farnesiana was almost certainly in-

troduced into Africa. Bentham (1864) stated that A.

farnesiana had every appearance of being indigenous

in Australia but doubt has been expressed from time

to time (Pedley, 1979) as to whether or not the species

is in fact indigenous. A. farnesiana has an extensive

distribution in Australia covering over 5 million

square kilometres (N. Hall, pers. comm.) which

greatly exceeds that of any other member of sub-

genus Acacia in Australia or the distribution of all of

the other members of the subgenus collectively (see

Fig. 16). If A. farnesiana is not indigenous in Austra-

lia then presumably it must have been introduced

prior to European settlement as the species was en-

countered in the inland areas of Australia by early

explorers.

The Australian members of subgenera Acacia and

Aculeiferum are not particularly well known; some

members of subgenus Acacia are not yet described

although the subgenus is currently being revised. Ex-

cept for the pantropical A. farnesiana which has a

diploid chromosome number of 52 and the natural-

ized A. nilotica subsp. indica (Benth.) Brenan in

which 2n = 44, 52, 104 (Vassal, 1974), there are no

chromosome data available for other Australian

408

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Fig. 14. — An indication of the

general distribution of sub-

genus Acacia in Australia

(excluding A. farnesiana).

Fig. 15. — An indication of the

distribution of subgenus

Aculeiferum in Australia.

Fig. 16. — An indication of the

general distribution of Acacia

farnesiana in Australia.

J. H. ROSS

409

members of subgenus Acacia and none for the soli-

tary representative of subgenus Aculeiferum.

The contemporary Australian Acacia species

appear to present a contradiction. On the one hand

the majority of the Australian species (members of

subgenus Heterophyllum) are unique and differ sig-

nificantly from all of the other species in other

regions (except for the few outliers discussed by Ped-

ley, 1975) which suggests that they have evolved in

isolation, while on the other hand this uniqueness is

contradicted by the presence of a handful of species

in subgenus Acacia and the solitary representative of

subgenus Aculeiferum which share a number of the

characters found in members of the respective

subgenera elsewhere. This would seem to suggest that

Acacia may have entered Australia on more than one

occasion. For example, it would appear that the

solitary member of subgenus Aculeiferum has

entered Australia from New Guinea in geologically

recent time, perhaps before the separation of

Australia and New Guinea in the late Pleistocene.

The other alternative is that the species is a relic. The

important questions to which answers will have to be

found are, firstly, when did Acacia first enter the

Australian region and, secondly, from where.

There would appear to be two possibilities,

namely, either the ancestors of Acacia must have

been present in the Australian region at the time of

the fragmentation of Gondwanaland although there

is no evidence that the ancestors had differentiated at

this stage or else the genus entered the region during

the Tertiary. According to Raven & Axelrod (1974)

geological evidence suggests the possibility of more

or less direct migration between West Gondwana-

land, combined Africa and South America, and East

Gondwanaland, combined Australasia and Antarc-

tica, up until the close of the early Cretaceous about

1 10 million years ago ( ± 10 million years). When the

connection between Africa and Australasia was

severed is not clear, but it is possible that Madagascar

and India served as a subtropical route of migration,

perhaps somewhat interrupted, to Australasia into

the late Cretaceous. Once the migration of tropical

alliances by this route was no longer possible,

Australasia was connected with the rest of the world

by a cool-temperate pathway to South America via

Antarctica up until about 45 million years ago.

However, the present mainly tropical and subtropical

global distribution of Acacia, the absence of species

from the southern portion of South America and the

few species in the southern tip of South Africa and in

Tasmania, suggests that Acacia did not enter

Australia from the south via this cool-temperate

route as some other genera such as Nothofagus

appear to have done. In general, the ease of migra-

tion to Australasia is believed to have decreased dur-

ing the late Cretaceous to Paleogene only to increase

again in the Neogene as the Australian plate neared

Asia. Nevertheless, according to Raven & Axelrod

(l.c.), migration between Australasia and Africa via

India and Madagascar was probably relatively direct,

but with fairly long steps over water, after the start of

the Tertiary 65 million years ago.

The first recorded appearance of Acacia pollen in

the Australian fossil record is in mid-Miocene

deposits (+ 16 million years ago) in southern Vic-

toria (Cookson, 1954). To explain the apparent

absence of Acacia pollen from Paleogene beds

Cookson concluded that either the genus Acacia was

not represented in the Nothofagus- conifer forests

which are known to have covered large areas of

Australia during the early Tertiary period or else the

genus did not become an integral part of the

Australian flora as a whole until after the Lower

Miocene period (25 million years ago). Although the

evidence was not conclusive, Cookson favoured the

latter possibility.

If Cookson’s view that Acacia did not occur in

Australia prior to the Lower Miocene is correct, then

either subgenus Heterophyllum developed in

Australia after the Lower Miocene, probably from

subgenus Aculeiferum and possibly from a species

such as the Indian A. ferruginea or its ancestors, or

else subgenus Heterophyllum entered Australia after

the Lower Miocene after developing elsewhere in an

area where it has since become extinct except for

isolated species such as A. confusa Merr., A. koa, A.

heterophylla and A. xiphoclada. The occurrence of

the same seed amino acid pattern in A. confusa,

which occurs in the Pacific Islands and Taiwan, as

that found in the seeds of African and Asian species

suggested to Evans et al. (1977) that this species

originated in Asia rather than in Australia.

Another possibility, however, is that the genus

may have been established in Australia prior to the

Lower Miocene but confined to the northern part of

the continent. The suggestion that Acacia first

became established on the northern part of the

Australian plate and later spread to other parts of the

continent when suitable conditions prevailed for it to

do so was advanced by Andrews (1914) and sup-

ported by Pedley (1975). Vassal (1972), on the other

hand, considered that the primitive section of

subgenus Heterophyllum probably occupied the

whole of the Australian continent at the beginning of

the Tertiary.

Although the centre of present-day development of

species in a subgenus need not necessarily reflect the

centre of past development, the great diversity exhi-

bited by subgenus Heterophyllum in Australia and

the lack of close relatives on other continents sug-

gests that subgenus Heterophyllum developed in

Australia from which source a few species such as A.

confusa, A. koa and A. heterophylla were subse-

quently dispersed to other areas and that the sub-

genus did not enter Australia after developing else-

where. It is tempting to speculate that subgenus

Acacia entered northern Australia where it has since

largely remained in the mid-Tertiary (or perhaps even

later) when Australia came into contact with the

south-east Asian plate but it is difficult to account

for the poor representation of subgenus Aculeiferum

in Australia.

As Australia was separated from Africa and India

by a considerable marine gap and was distant from

any other tropical land mass for millions of years un-

til late in the Oligocene when contact was made with

the south-east Asian plate, Melville (1975) concluded

that ‘the characteristic Australian flora — excluding

the Indo-Malaysian element of relatively recent

origin — must have evolved in situ from ancestors of

Permian age’. Melville (l.c.) continued that ‘Accep-

tance of this conclusion implies that evolutionary

trends in many families such as Proteaceae, Res-

tionaceae, Leguminosae and Compositae, must have

been initiated already in the Permian for the ob-

served parallel evolution to have taken place subse-

quently on separated Gondwanic fragments’.

Irrespective of when the genus first became estab-

lished in Australia and from where it came its devel-

opment and subsequent spread over the continent

have been influenced by past geological and climatic

changes. During the period from the Triassic, until

410

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

northward drift began in the Eocene, Australia was

situated about 15° south of its present latitude

(Jones, 1971).

The Australian Tertiary pollen record is largely

that of rainforest which must have been widespread,

though not necessarily continuous, over the southern

part of the continent (Martin, 1978). The high con-

tent of gymnosperms in the Australian Paleocene

assemblages is thought to indicate a cool temperate

climate similar to present-day Tasmania (Martin,

/.<:.). The subsequent increase in Myrtaceae and other

angiosperms is taken to indicate a relative increase in

temperature to a warm temperate or subtropical

climate. Temperatures reached the maximum for the

Tertiary in the early Eocene and the dramatic in-

crease in Nothofagus in the mid-Eocene marks the

onset of a cooling trend although further fluctuations

were experienced. The Oligocene to early Miocene

was a period of an equable climate with very high

rainfall and stable temperatures while the mid-late

Miocene was a time of profound change when

Nothofagus and many other taxa disappeared from

the fossil record in south-eastern Australia (Martin,

1977). During the late Miocene and Pliocene a

moderate rainfall and drier type of vegetation existed

and it is thought that remnants of lower Tertiary

flora existed in small refuge areas in the eastern

highlands and migrated westwards when the climate

became wetter for a relatively brief period in the late

Pliocene (Martin, 1977). There is every indication

that aridity increased subsequent to Pliocene time

and the climate became more seasonal with a well-

marked dry period and it is not until the Pleistocene

that grasslands and/or savanna woodlands became

prominent in south-eastern Australia. According to

Gill (1975) it was only at this stage about two million

years ago that the full opportunity for speciation in

Acacia and Eucalyptus occurred. As aridity inten-

sified closed forest would have been eliminated from

all but locally favourable sites in north-eastern

Australia.

In rainforest areas Acacia species are confined to

marginal areas and clearings and it is only when the

canopy is disturbed that the light requiring elements

of the Australian flora become established (Bur-

bidge, 1960). The retreat of closed forest with in-

creasing aridity would have favoured the dispersal of

Acacia species and the fossil finds discussed by

Cookson (1954) possibly indicate an expansion in the

distributional range of Acacia which coincided with

the retreat of Nothofagus. Hopper & Maslin (1978)

suggested that the recent speciation in Acacia in

Western Australia has been promoted by recurrent

migration, extinction and isolation of populations as

a result of Pleistocene climatic fluctuations and their

erosional consequences in climatically transitional

areas. The possible dispersal of Acacia in Australia

as a whole is dealt with by Pedley (1980).

Apart from the obvious differences such as the de-

velopment of phyllodes in many members of sub-

genus Heterophyllum in Australia and the differen-

ces between the three subgenera in pollen, chromo-

somes, seeds, seedlings, inflorescences and pods

alluded to by Guinet & Vassal (1978), there are other

differential tendencies between subgenus Hetero-

phyllum and the African representatives of sub-

genera Acacia and Aculeiferum which are briefly dis-

cussed here.

I . The African species (except A. albida) are invari-

ably deciduous during the dry season, either regularly

or irregularly so, while the Australian members of

subgenus Heterophyllum are invariably evergreen.

The only Australian species thought to be deciduous

is A. ditricha Pedley, a member of subgenus Acacia,

but it is possible that some other members of this

subgenus are also deciduous.

2. The flattened spreading crowns which are so

characteristic of some of the African species such as

A. tortilis subsp. heteracantha (Burch.) Brenan and

subsp. spirocarpa (Hochst. ex A. Rich.) Brenan, A.

sieberana var. woodii (Burtt Davy) Keay & Brenan,

A. lahai Steud. & Hochst. ex Benth. and A.

abyssinica Hochst. ex Benth. subsp. calophylla

Brenan are absent amongst the Australian species.

One explanation advanced for the flattened crowns

in Africa is that it is an adaptive response to browsing

(Brown, 1960). Another suggestion that has been

offered is that it is the result of insolation damaging

the apical growing buds, but if this was the case it is

strange that none of the Australian species has

developed the same adaptive response where the

effects of insolation are as great as they are in Africa.

The African species in general ‘look’ different to

most of the Australian members of subgenus Hetero-

phyllum.

3. The stipules in all African members of subgenus

Acacia are spinescent and invariably very prominent

and in the Australian members of the subgenus the

stipules are typically spinescent at least when young

although they are usually small or occasionally

absent. The stipules in subgenus Heterophyllum are,

with few exceptions, small, inconspicuous and often

deciduous. Many of the species are entirely unarmed

but in others spinescence has arisen in various ways

through the modification of phyllodes, branchlets

and peduncles. A. paradoxa DC. and A. victoriae

Benth. are exceptional in having stipular spines

although those of the latter are sometimes reduced to

blunt outgrowths, and stipular spines also occur in

some of the Western Australian species (Pedley,

1978). Stipular spines are far better developed in the

African species of subgenus Acacia than they are in

the Australian members of subgenus Acacia or in

subgenus Heterophyllum. Brown (1960) suggested

that ‘ Acacia in Australia passed through an earlier

period in which spininess had little adaptive value,

followed by a time in which selective pressures again

arose favouring the development of spines de novo in

shrubs of the forested country and in the lower

shrubs of the open country’.

Brown (/.c. ) attributed the lack of spinescence in

many of the Australian Acacia species and in other

dominant genera like Eucalyptus to the ‘long-

continued absence or scarcity of effective large brow-

sers’ until the recent introduction of domesticated

animals. Large browsing herbivores are now or have

recently been abundant in Africa and in tropical

America in areas where spinescent acacias occur and

certainly in Africa the acacias constitute an import-

ant source of food for many browsing mammals.

However, the argument loses some validity as many

species in other genera have survived browsing pres-

sure without the aid of spines although admittedly

they may have developed other deterrents such as of-

fensive chemical attributes to discourage large

browsers. No similar radiation of large browsers

appears to have occurred in Australia, the endemic

Australian mammal fauna consisting of marsupials,

monotremes, rodents and bats. Of these, the kanga-

roo is the largest survivor but kangaroos are chiefly

J. H. ROSS

411

grazers and it seems reasonable to assume that the

extinct giant kangaroos were also grazers. Brown

( l.c .) considered the development of spininess in

many of the smaller shrubby acacias, grasses and

other shrubs of the closed forest understandable see-

ing that kangaroos, smaller marsupials and rodents

feed on them.

None of the Australian members of subgenus

Acacia has swollen stipular spines which are so

characteristic of some of the African and American

species and none of the Australian species has form-

ed any mutualistic association with ants although the

seeds of many Australian plants, including some

Acacia species, are dispersed short distances by ants

(Berg, 1975). The Australian continent is richly

endowed with ants so the lack of a mutualistic

association between acacias and ants cannot be due

to a scarcity of ants. Hocking (1970) suggested that if

the Australian members of subgenus Acacia ever had

the tendency to produce swollen spines then

presumably the expression of the character has been

lost under reduced selection pressure from her-

bivorous animals. The extra-floral nectaries in the

Australian members of subgenus Acacia are small.

4. A number of the Australian members of subgenus

Heterophyllum flourish in a cold and wet environ-

ment in southern Victoria and Tasmania far further

south than on any other continent and have occupied

a habitat that is generally avoided by the indigenous

African species. The African species tolerate hot and

dry, hot and wet, and cold and dry habitats but

where cold and wet conditions persist for any great

period of the year acacias are usually infrequent in

their occurrence or absent.

5. The African species invariably flower in spring or

early summer and the inflorescences are usually pro-

duced with or before the young leaves. In contrast,

the Australian members of subgenus Heterophyllum

are evergreen and a number of species flower in

winter.

6. The anthers of nearly all of the African species

(except A. albida and A. reclacta, neither of which it

has been suggested is referable to Acacia, Robbertse

& Von Teichman, 1979) are adorned with a small

apical deciduous gland but none of the Australian

members of subgenus Heterophyllum appear to have

glands on the anthers. The function of the glands in

the African species is not clear but Hocking (1970)

suggested that the tissue filling the glands may con-

tain useful amounts of nitrogenous material which

may be utilised by phytophagous insects. A. biclwillii

Benth. and A. sutherlandii (F. Muell.) F. Muell.,

both Australian members of subgenus Acacia, have

anthers with small deciduous apical glands but I have

not seen suitable material of other Australian

members of the subgenus, except of A. farnesiana

which lacks the glands, to establish whether they also

have the deciduous glands. This needs further in-

vestigation.

7. Some of the Australian members of subgenus

Heterophyllum have seed with conspicuous brightly

coloured arils, whereas none of the African species

does.

8. Bruchids are responsible for the destruction of

vast quantities of seeds of some American and Afri-

can species in subgenera Acacia and AculeiJ'erum

whereas seeds of the Australian species are relatively

unaffected by bruchids, probably because few indige-

nous species of bruchid occur in Australasia and the

Pacific Islands (Southgate, 1978). A survey of the

free non-protein amino acids in the seeds of 106

Acacia species by Evans el al. (1977) revealed that the

genus can be divided into four biochemically dif-

ferent groups on the basis of their seed chemistry and

Southgate {l.c.) suggested that one of the factors in-

fluencing the ability of bruchid larvae to survive

within a seed may be the level of certain amino acids,

notably pipecolic acid and some heteropolysac-

charides. The amino acid composition of the seed of

the Australian members of subgenus Heterophyllum

differs from that of the seed of members of other

subgenera, and a possible explanation for this may be

that members of subgenus Heterophyllum have

apparently evolved without the selection pressure of

bruchid predation.

DISCUSSION

As a result of the multidisciplinary approach to

Acacia in recent years much evidence has accumu-

lated which indicates that fundamental differences

exist between subgenera Acacia and Aculeiferum,

and that subgenus Aculeiferum is more closely

related to subgenus Heterophyllum despite the fact

that they occupy basically different geographical

areas which show relatively little overlap, than are

subgenera Acacia and Aculeiferum which share a

common geographical area. The differences between

subgenera Acacia and Aculeiferum are such that it is

considered unlikely that the one gave rise to the other

directly but rather that they arose from a common or

similar prototype. Many questions concerning the

origin, evolution and dispersal of the genus and of

the relationships within it remain to be answered to

enable a better understanding to emerge.

Although the African Acacia species have received

a considerable amount of attention during the last

few decades they remain inadequately known and nu-

merous taxonomic problems await elucidation. Des-

pite the incompleteness of the information on the dis-

tribution of the African species the overall patterns

that emerge are probably sufficiently accurate to be

of value. Further collecting, especially in tropical

north-east Africa and in west tropical Africa, will

resolve some of the taxonomic problems and provide

more accurate information on the distributions of

many species.

Some species in subgenus Aculeiferum are almost

as widespread in Africa as the most widespread mem-

bers of subgenus Acacia, but the distribution of sub-

genus Acacia as a whole in Africa exceeds that of

subgenus Aculeiferum. That subgenus Acacia enjoys

a wider range of distribution than subgenus Aculei-

ferum suggests that the former has been able to

occupy habitats from which the latter has been ex-

cluded and the possibility exists that it has been

assisted in this by the greater genetic plasticity con-

ferred on it as a consequence of its members being

polyploid. On the other hand, climbing members of

subgenus Aculeiferum have been successful in

forested areas of the continent in which subgenus

Acacia is not represented, the climbing habit, which

is not known in subgenus Acacia in Africa, enabling

species to take advantage of suitable sites in forested

areas. The highest concentration of species in each

subgenus occurs in tropical north-east, east and

south-east Africa but different parts of the continent

have been important areas of local speciation for

each subgenus, the highest concentration of endemic

species in subgenus A cacia occurring in Tanzania and

the highest concentration of endemics in subgenus

Aculeiferum in Somalia.

412

AN ANALYSIS OF THE AFRICAN ACACIA SPECIES: THEIR DISTRIBUTION, POSSIBLE ORIGINS

AND RELATIONSHIPS

Despite the advances in our knowledge of the Afri-

can species in recent decades, detailed population

studies are required and information is needed on

their biology and autecology. Only when such a

reservoir of information is available will a better

understanding of the African species emerge. And,

what is true for the African species applies equally to

those in other continents. It would be highly benefic-

ial as a first step to have a conspectus of the Acacia

species occurring on each continent reflecting the

current state of taxonomic knowledge along the lines

of that produced for the African species (Ross, 1979)

or the more detailed revision of Queensland species

(Pedley, 1978, 1979), and ultimately a conspectus of

the genus as a whole. This is, of course, a fairly for-

midable task especially when one considers the Aus-

tralian species but much valuable work has already

been done on the Australian species. Hopefully, a

conspectus of the Australian species will be prepared

before too long. Acacia is a fascinating genus, which

commends itself to further study.

ACKNOWLEDGEMENTS

I am most grateful to Mr L. Pedley, Assistant Di-

rector, Queensland Herbarium, for answering several

queries, for information concerning the distribution

of subgenus Acacia in Australia and for providing a

draft account of his manuscript on the derivation and

dispersal of Acacia, to Mr J. R. Maconochie, North-

ern Territory Herbarium, and Mr B. R. Maslin,

Western Australian Herbarium, for information on

the occurrence of subgenus Acacia in the Northern

Territory and Western Australia respectively, to Mr

N. Hall, Castle Hill, New South Wales, for informa-

tion on the distribution of Acacia farnesiana in

Australia, to Mr C. H. Stirton, South African Liai-

son Officer, Royal Botanic Gardens, Kew, England,

for providing information on some Madagascan

Acacia species, and to Mrs G. Bray, National Herb-

arium of Victoria, for typing the manuscript.

UITTREKSEL

Die drie subgenera wat in die genus Acacia erken

word, word in hooftrekke beskrywe en die globale

verspreiding van elk word aangedui. Die verskille tus-

sen die subgenera en die graad en verwantskap en

vlakke van spesialisasie word kortliks bespreek. Die

voorstel word gemaak dat die voorouers van die

genus klim- of slingerplante was. Geologiese gebeure

in die verlede wat ’n moontlik invloed op die ver-

spreiding van Acacia -spesies in Afrika kon gehad

het, word geskets. Die aantal spesies wat vir elke land

in Afrika aangeteken is word getabuleer en die ver-

spreiding en konsentrasie van spesies binne die genus

Acacia as geheel en binne elke subgenus in Afrika

word geillustreer. Die hoogste konsentrasie van

spesies binne elke subgenus koin in oos en suid-oos

tropiese A frika voor. Die verspreiding van spesies in

sommige van die afsonderlike Afrika-lande en

moontlike verwantskappe word bespreek en die aan-

dag word op die hoofsentra van endemisme gevestig.

Die verspreiding van die spesies van A frika word met

die hoof fito-geografiese streke wat op die vasteland

erken word, in verband gebring. Die verwantskappe

tussen die Acacia spesies van Afrika, Amerika,

Malgassie, Indie en Austral ie word kortliks bespreek.

REFERENCES

Andrews, E. C., 1914. Development and distribution of the

natural order Leguminosae. J. Proc. R. Soc. New South

Wales 48: 333-407.

Atchison, E., 1948. Studies in the Leguminosae. II.

Cytogeography of Acacia (Tourn.) L. Am. J. Bot. 35:

651-655.

Axelrod, D. I., 1972a. Ocean-floor spreading in relation to

ecosystematic problems. Univ. Arkansas Mus. Occas. Paper

4: 15-76.

Axelrod, D. L, 1972b. Edaphic aridity as a factor in angiosperm

evolution. Am. Nat. 106: 311-320.

Axelrod, D. I. & Raven, P. H., 1978. Late Cretaceous and Ter-

tiary vegetation history of Africa. In M. J. A. Werger (ed.),

Biogeographv and ecology of southern Africa 1 : 77-130. The

Hague: W. Junk.

Bell, E. A., 1971. Comparative biochemistry of non-protein

amino-acids. In J. B. Harborne, D. Boulter & B. L. Turner

(eds), Chemotaxonomy of the Leguminosae 179-206. New

York: Academic Press.

Bell, E. A. & Evans, C. S., 1978. Biochemical evidence of a

former link between Australia and the Mascarene Islands.

Nature 2D: 295-296.

Bentham, G., 1864. Flora Australiensis Vol. 2. London: Lovell

Reeve & Co.

Bentham, G., 1875. Revision of the suborder Mimoseae. Trans.

Linn. Soc. Lond. 30: 335-664.

Berg, R. Y., 1975. Myrmecochorous plants in Australia and their

dispersal by ants. Aust. J. Bot. 23: 475-508.

Brenan, J. P. M., 1979. Some aspects of the phytogeography of

tropical Africa. Ann. Mo. bot. Gdn 65: 437-478.

Brown, W. L., 1960. Ants, Acacias and browsing mammals.

Ecology 41: 587-592.

Burbidge, N. T., 1960. The phytogeography of the Australian

region. Aust. J. Bot. 8: 75—211.

Burtt, B. L., 1971 . From the South: an African view of the floras

of western Asia. In P. H. Davis, P. C. Harper & 1. C. Hedge

(eds), Plant life of South-West Asia 134-149. Edinburgh:

Edinburgh Botanical Society.

Carlquist, S., 1965. Island life. New York: Nat. Hist. Press.

Chapman, J. D. & White, F., 1970. The evergreen forests of

Malawi. Oxford: Commonwealth Forestry Institute, Univer-

sity of Oxford.

Cookson, I. C., 1954. The Cainozoic occurrence of Acacia in

Australia. Aust. J. Bot. 2: 52-59.

Crocker, R. L., 1959. Past climatic fluctuations and their in-

fluence upon Australian vegetation. In A. Keast, R. L.

Crocker & C. S. Christian (eds), Biogeography and ecology in

Australia 283-290. The Hague: W. Junk.

De Winter, B., 1971. Floristic relationships between the northern

and southern arid areas of Africa. Mitt. bot. Stsamml.,

Munch. 10: 424-437.

Evans, C. S., Qureshi, M. Y. & Bell, E. A., 1977. Free amino

acids in the seeds of Acacia species. Phytochemistry 16:

565-570.

Frakes, L. A. & Kemp, E. M., 1972. The influence of continental

positions on early Tertiary climates. Nature 240: 97-100.

Gill, E. D., 1975. Evolution of Australia’s unique flora and fauna

in relation to the plate tectonics theory. Proc. R. Soc. Viet.

87: 215-234.

Gmelin, R., 1959. Die freien aminosauren der samen von Acacia

wiUardiana (Mimosaceae). Hoppe-Seyl. Z. 316: 164-169.

Goldblatt, P., 1979. An analysis of the Flora of Southern Africa:

its characteristics, relationships, and origins. Ann. Mo. bot.

Gdn 65: 369-436.

Guinet, Ph., 1967. Parallelisme entre les caracteres du pollen dans

les groupes Piptadeniees et Acaciees des Mimosacees. Rev.

Paleobot. Palynol. 3: 151-153.

Guinet, Ph., 1969. Les Mimosacees. Etude de palynologie fon-

damentale, correlations, evolution, Trav. Sect. Sci. Tech.

Inst. Fr. Pondichery 9: 1-203.

Guinet, Ph. & Vassal, J., 1978. Hypotheses on the differentiation

of the major groups in the genus Acacia (Leguminosae). Kew

Bull. 32: 509-527.

Hepper, F. N., 1965. Preliminary account of the phytogeographi-

cal affinities of the Flora of West Tropical Africa. Webbia 19:

593-617.

Hickey, L. J. & Doyle, J. A., 1977. Early Cretaceous fossil

evidence for angiosperm evolution. Bot. Rev. 43: 3-104.

Hocking, B., 1970. Insect associations with the swollen thorn

acacias. Trans. R. ent. Soc. Lond. 122: 211-255.

Hopper, S. D. & Maslin, B. R., 1978. Phytogeography of Acacia

in Western Australia. Aust. J. Bot. 26: 63-78.

Hunde, A., 1979. Acacia amythethophylla, the correct name of a

widespread African tree. Bot. Notiser 132: 393-395.

Janzen, D. H., 1966. Coevolution of mutualism between ants and

Acacias in Central America. Evolution 20: 249-275.

Janzen, D. H., 1967a. Fire, vegetation structure, and the ant x

acacia interaction in Central America. Ecology 48: 26-35.

J. H. ROSS

413

Janzen, D. H., 1967b. Interaction of the bull’s-horn acacia

(Acacia cornigera L.) and an ant inhabitant ( Pseudomyrmex

ferruginea F. Smith) in eastern Mexico. Univ. Kans. Sci. Bull.

47: 315-558.

Janzen, D. H. 1969a. Birds and the ant x acacia interaction in

Central America, with notes on birds and other myrmeco-

phytes. Condor 71: 240-256.

Janzen, D. H., 1969b. Allelopathy by myrmecophytes: the ant

Azteca as an allelopathic agent of Cecropia. Ecology 50:

147-153.

Janzen, D. H., 1974. Swollen-Thorn Acacias of Central America.

Smithson. Contr. Bot. 13: 1 — 131.

Johnson, L. A. S. & Briggs, B. G., 1975. On the Proteaceae — the

evolution and classification of a southern family. Bot. J.

Linn. Soc. 70: 83-182.

Jones, J. G., 1971. Australia’s Caenozoic drift. Nature 230:

237-239.

Kemp, E. M., 1978. Tertiary climatic evolution and vegetation

history in the southeast Indian Ocean region. Palaeogeog.

Palaeoclim. Palaeoecol. 24: 169-208.

King, L. C., 1967. The morphology of the Earth. Edn 2. New

York: Hafner.

Leroy, J. F., 1979. Composition, origin, and affinities of the

Madagascan vascular flora. Ann. Mo. bot. Gdn 65:

535-589.

Martin, H. A., 1977. The history of Ilex (Aquifoliaceae) with

special reference to Australia : evidence from pollen. Aust. J.

Bot. 25: 655-673.

Martin, H. A., 1978. Evolution of the Australian flora and

vegetation through the Tertiary: evidence from pollen.

A Icheringa 2: 181 -202 .

Melville, R., 1975. The distribution of Australian relict plants

and its bearing on angiosperm evolution. Bot. J. Linn. Soc.

71: 67-88.

Mildenhall. D. C., 1972. Fossil pollen of Acacia-type from New

Zealand. N. Z. J. Bot. 10: 485-494.

Mildenhall. D. C., 1975. Palynology of the Acacia-bearing beds

in the Komako district, Pohangina Valley, North Island, New

Zealand. N. Z. J. Geol. Geophys. 18: 209-228.

Moll, E. J. & White, F., 1978. The Indian Ocean coastal belt. In

M. J. A. Werger (ed.). Biogeography and ecology of southern

Africa 1: 561-598. The Plague: W. Junk.

Moreau, R. E., 1952. Africa since the Mesozoic: with particular

reference to certain biological problems. Proc. zool. Soc.

Lond. 121: 869-913.

Muller, J., 1970. Palynological evidence on early differentiation

of angiosperms. Biol. Rev. 45: 417-450.

Pedley, L., 1975. Revision of the extra-Australian species of

Acacia subgenus Heterophvllum. Contr. Queensland Herb.

18: 1-24.

Pedley, L., 1978. A revision of Acacia Mill, in Queensland.

Part 1. Austrobaileya 1, 2: 75-234.

Pedley, L., 1979. A revision of Acacia Mill, in Queensland. Part

2 Austrobaileya 1, 3: 235-337.

Pedley, L., 1980. Derivation and dispersal of Acacia (Legumino-

sae) with particular reference to Australia (in press).

Plumstead, E., 1969. Three thousand million years of plant life in

Africa. Geol. Soc. S. Afr., Annexure to Vol. 72: 1-72.

Quezel, P., 1979. Analysis of the Flora of Mediterranean and

Saharan Africa. Ann. Mo. bot. Gdn 65: 479-534.

Raven, P. H. & Axelrod, D. 1., 1972. Plate tectonics and

Australasian paleobiogeography. Science 176: 1379-1386.

Raven, P. H. & Axelrod, D. I., 1974. Angiosperm biogeography

and past continental movements. Ann. Mo. bot. Gdn 61:

539-673.

Rehr, S. S., Feeny. P. & Janzen, D. H., 1973. Chemical defenses

in Central American non-ant-Acacias. J. Anim. Ecol. 42:

405-416.

Richards, P. W., 1973. Africa, the ‘Odd Man Out’. In B. J.

Meggers, E. S. Ayensu and W. D. Duckworth (eds), Tropica!

forest ecosystems in Africa and South America : A compara-

tive review 21-26. Washington, D.C.: Smithsonian Inst.

Press.

Robbertse, P. J., 1974. The genus Acacia in South Africa, 11, with

special reference to the morphology of the flower and in-

florescence. Phytomorphology 24: 1-15.

Robbertse, P. J. & Von Teichman, I., 1979. The morphology of

Acacia redacta J. H. Ross. Jl S. Afr. Bot. 45: 11-23.

Ross. J . H , 1973. Notes on Acacia species in Southern Africa: III.

Bothalia 11: 127-131.

Ross, J . H., 1974. Notes on Acacia species in Southern Africa: IV.

Bothalia 11: 231-234.

Ross, J. H., 1979. A conspectus of the African Acacia species.

Mem. bot. Surv. S. Afr. 44: 1-155.

Schuster, R. M., 1972. Continental movements, ‘Wallace’s line’

and Indomalayan-Australasian dispersal of land plants: some

eclectic concepts. Bot. Rev. 38: 3-86.

Schuster, R. M., 1976. Plate tectonics and its bearing on the

geological origin and dispersal of angiosperms. In C. B. Beck

(ed.), Origin and early evolution of Angiosperms. New York

and London: Columbia Univ. Press.

Seigler, D. S., Dunn, J. E., Conn, E. E. & Holstein, G. L., 1978.

Acacipetalin from six species of Acacia of Mexico and Texas.

Phytochemistry 17: 445-446.

Shchekina, N. A., 1965. On a find of polyads of the genus Acacia

L. in the Neogene deposits of the Ukraine. Ukr. Bot. Zhurn.

22: 83-89.

Southgate, B. J., 1978. Variation in the susceptibility of African

Acacia (Leguminosae) to seed beetle attack. Kew Bull. 32:

541-544.

Stebbins, G. L. Jr., 1952. Aridity as a stimulus to plant evolution.

Am. Nat. 86: 35-44.

Thorne, R. F., 1972. Major disjunctions in the geographic ranges

of seed plants. Quart. Rev. Biot. 47: 365-411.

Thorne, R. F., 1973. Floristic relationships between tropical

Africa and tropical America. In B. J. Meggers, E. S. Ayensu

& W. D. Duckworth (eds), Tropical forest ecosystems in

Africa and South America: A comparative review 27-47.

Washington, D.C.: Smithsonian Inst. Press.

Thorne, R. F., 1977. Where and when might the tropical angio-

spermous flora have originated? Gdns' Bull., Singapore 29:

183-189.

Thorne, R. F., 1978. Plate tectonics and angiosperm distribution.

Notes R. bot. Gdn Edinb. 36: 297-315.

Tindale, M. D. & Roux, D. G., 1974. An extended phytochemical

survey of Australian species of Acacia: chemotaxonomic and

phylogenetic aspects. Phytochemistry 13: 829-839.

Tindale, M. D. & Roux, D. G., 1975. Phytochemical studies on

the heartwoods and barks of African and Australian species

of Acacia. Boissiera 24a: 299-305.

Van Steenis, C. G. G. J., 1962. The land-bridge theory in botany.

Blumea 1 1 : 235-372.

Van Zinderen Barker. E. M., 1974. The evolution of late-

quaternary palaeoclimates of southern Africa. Palaeoecology

of Africa 9: 160-202.

Van Zinderen Barker, E. M., 1978. Quaternary vegetation

changes in southern Africa. In M. J. A. Werger(ed.). Biogeo-

graphy and ecology in southern Africa 1: 131-143. The

Hague: W. Junk.

Vassal, J., 1969. A propos des Acacias heterophylla et koa. Bull.

Soc. Hist. nat. Toulouse 105: 443-447.

Vassal, J., 1972. Apport des recherches ontogeniques et

seminologiques a l’etude morphologique, taxonomique et

phylogenique du genre Acacia. Trav. Lab. Forest. Toulouse,

Tome 1, Vol. 8, Art. 17: 1-128. Reprinted from Bull. Soc.

Hist. nat. Toulouse 108: 125-247.

Vassal, J., 1974. Reports on chromosome numbers in the genus

Acacia. Bull. International Group for Study of Mimosoideae

2: 21-29.

Verdcourt, B., 1969. The arid corridor between the northeast and

southwest areas of Africa. Palaeoecology of Africa 4:

140-144.

White. F., 1965. The savanna woodlands of the Zambezian and

Sudanian Domains. Webbia 19: 651-681.

White, F., 1979. The Guineo-Congolian Region and its relation-

ships on other phytochoria. Bull. Jard. bot. nat. Belg. 49:

11-55.

Wicrens, G. E., 1976. The flora of Jebel Marra (Sudan Republic)

and its geographical affinities. Kew Bull. Add. Ser. 5: 1-368.

Wild, H., 1968. Phytogeography in South Central Africa. Kirkia

6: 197-222.

Wild. H., 1974. Variations in the serpentine floras of Rhodesia.

Kirkia 9: 209-232.

Wood, J. G., 1959. The phytogeography of Australia (in relation

to radiation of Eucalyptus, Acacia, etc.). In A. Keast, R. L.

Crocker & C. S. Christian (eds), Biogeography and ecology in

Australia 291-302. The Hague: W. Junk.

■>

Bothalia 13, 3 & 4: 415-429 (1981)

Sexual nuclear division in Neocosmospora*

K. T. VAN WARMELO**

ABSTRACT

The process of sexual nuclear division in authentic isolates of Neocosmospora vasinfecta E. F. Smith, N. africana

Von Arx and a related species isolated from soybean was studied. No significant differences were found. Six

chromosomes were counted in each isolate. There was no evidence of aneuploidy or irregular reconstitution of

daughter nuclei. The divisions leading to the formation of binucleate, homokaryotic ascospores were typically

meiotic.

RESUME

DIVISION NUCLEAIRE SEXUELLE CHEZ LE NEOCOSMOSPORA

Le processus de division nucleaire sexuelle dans des isolats authentiques de Neocosmospora vasinfecta E. F.

Smith, N. africana Von Arx et une espece apparentee isolee du soya a ete etudie. Aucune difference significative n ’a

ete trouvee. Six chromosomes ont ete comptes dans chaque isolat. II n’y eut aucune evidence d’aneuploidie ou de

reconstitution irreguliere des noyaux filles. Les divisions conduisant a la formation du binucleate, d ’ascospores

homokariotiques furent typiquement meiotiques.

INTRODUCTION

A fungus isolated from soybean stem material

showed great similarity in morphology and dimen-

sions to two existing species of Neocosmospora, i.e.

N. vasinfecta and TV. africana. In fact, the local

culture could be placed almost equally well in either

species. It was further noticed that there were very

few morphological differences between the two exist-

ing species. A study was, therefore, undertaken to

examine the differences and similarities in the various

processes and structures of the local isolate and

authentic cultures of the two similar species. This

paper reports on the nuclear divisions leading to

ascosporogenesis.

REVIEW

The genus Neocosmospora was described in 1899

(Ferry, 1900). The type, and only species at that time,

was TV. vasinfecta (Atk.) Smith, derived from

Fusarium vasinfecta Atk. (Seaver, 1909), which had

been isolated from soil and had caused damping off

of several types of cultivated plants.

Neocosmospora remained monotypic until Von

Arx (1955) described a new species, TV. africana Von

Arx, which had been isolated from soil under grass-

land plots near Johannesburg. TV. africana has not

yet been implicated as a plant pathogen (Udagawa,

1963).

According to Von Arx (1955) TV. africana differed

from TV. vasinfecta only in possessing a smooth

epispore which was fairly regular in thickness. Von

Arx proposed that, despite the differences in external

spore texture, his new species should be included in

the genus Neocosmospora. The surface texture of

ascospores from authentic culture of both species of

Neocosmospora have been reported to show very

little differences under scanning microscopic exam-

ination (Van Warmelo, 1976).

Doguet (1956) made a detailed examination of the

development of the perithecia of both TV. vasinfecta

*Based on a Ph.D. (Agric.) thesis submitted to the University of

Stellenbosch, 1973. Supervisor: Prof. P. S. Knox-Davies.

**Rand Afrikaans University, P.O. Box 524, Johannesburg, 2000.

and TV. africana and found that it followed the same

sequence in both species. This supported the state-

ment by Von Arx (1955) that TV. africana was similar

to TV. vasinfecta except for spore surface.

The process of meiosis in the fungi has received

much attention and the mechanisms of division have

been well described in a large variety of fungi (Olive,

1953 & 1965). Further significant contributions were

made by Rogers (1964, 1965, 1967, 1968a, 1968b,

1968c & 1971), Lu (1966, 1967a & 1967b), Aldrich

(1967), Barry (1967), Furtado (1970), Huguenin &

Boccas (1970) and Wells (1970).

Division of the diploid nucleus follows a pattern

basically similar to that established for the higher

plants (Lu, 1966; Swanson, 1968; Woo & Partridge,

1969; Wells, 1970). There are, therefore, the same

stages and processes during prophase 1 which lead to

a recombination and segregation of chromatin after

completion of the meiotic division. There are,

however, some differences between the meiotic divi-

sions in fungi and higher plants.

The nuclear membrane remains intact throughout

the division (Huffman, 1968) or becomes discon-

tinuous at the poles only (Aldrich, 1967). In these

cases the spindles are fully intranuclear. Degenera-

tion of the membrane also occurs, however, during

the division, for example during anaphase (Wells,

1970), during metaphase (Olive, 1965; Lu, 1967b)

and even as early as pachytene (Uecker, 1967).

The formation of a metaphase plate, which is in-

frequent in the fungi (Olive, 1965; Wells, 1970), was

reported by Furtado (1970) and Rao & Mukerji

(1970).

Centriolar plaques (also described as centrioles and

centrosomes) occur in some organisms (Knox-Davies

& Dickson, 1960; Lu, 1967a & 1967b; Uecker, 1967;

Wells, 1970), but are absent in others (Aldrich,

1967).

Two types of fibres are found in the spindle, viz.

chromosomal fibres which are attached to the

chromosomes, and spindle fibres which are con-

tinuous between the centriolar plaques (Aldrich,

1967; Lu, 1967b; Wells, 1970). It is important to note

416

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

that the spindle fibres may form in the absence of the

centriolar plaques (Aldrich, 1967), in which case the

fibres converge at the discontinuity of the nuclear

membrane.

Asynchronous disjunction of chromosomes was

frequently reported (Rogers, 1964, 1965, 1967; Olive,

1965; Uecker, 1967; Huffman, 1968; Furtado, 1969;

Wells, 1970) and seems to be general in the fungi.

In the Pyrenomycetes the second meiotic division

is commonly followed by a third division to give rise

to eight haploid nuclei around which the ascospores

then form (Singleton, 1953; Olive, 1965). Further

division may take place within the delimited

ascospores to give rise to multinucleate ascospores

(Van Warmelo, 1966). Ascospores may, however, be

delimited at the 16-nucleate stage (Furtado, 1970;

Rao & Mukerji, 1970), or delimitation may occa-

sionally be highly irregular (Rogers, 1967).

As a result of the frequency with which the regular

processes of ascosporogenesis have been found in the

Pyrenomycetes, this pattern of development is

generally applied to all Ascomycetes, even to species

in which few details are available. However, ex-

trapolated data can be misleading, as was shown by

Rajendren (1967), who found that in Hemileia

vastatrix the nuclear cycle did not agree entirely with

that expected from earlier work on the Uredinales.

According to Doguet (1956) the ascogonium of

Neocosmospora vasinfecta arises by somatogamy

and gives rise to numerous ascogenous hyphae.

Moreau & Moreau (1950), however, stated that

dikaryotization occurs by means of a ‘short an-

theridial filament’, which associates closely with the

coiled ascogonium. Doguet (1956) did not observe

karyogamy in the penultimate cell of the crozier. The

presence of nuclei with what appeared to be two

nucleoli was, however, taken as an indication that

karyogamy had occurred.

According to Doguet (1956) nuclear divisions in

the ascus follow the normal pattern leading to the

formation of eight binucleate ascospores. Unfor-

tunately only very few drawings of the various divi-

sions were published. Doguet also stated that the

mechanism of ascosporogenesis in N. africana is

identical to that found in N. vasinfecta, implying that

an ascogonium is formed by somatogamy and that

the ascospores are binucleate. None of these stages

was, however, illustrated.

MATERIALS AND METHODS

Three cultures were examined in this investigation:

1. Neocosmospora vasinfecta E. F. Smith; CBS

237.55.

2. Neocosmospora africana Von Arx; CBS 237.

3. Neocosmospora sp. isolated from soybean stem,

Pietermaritzburg, South Africa. Referred to in

the following text as ‘isolate P’.

Cultures were maintained on 1,5% Difco malt agar

and yeast-starch agar (Cooney & Emerson, 1964) at

25°C under intermittent illumination, 12 h/d, by

near ultra violet light (NUV) fluorescent tubes (GEC

F15T8/BLB). Isolate P was also maintained on Difco

oatmeal agar.

The nuclear divisions leading to ascosporogenesis

were examined in all three isolates studied. Identical

techniques were used for similar stages in the dif-

ferent isolates to ensure absolute comparability.

Nuclei were stained using the HCL-Giemsa techni-

que (Van Warmelo, 1971) and examined and photo-

graphed under bright field illumination with a yellow

green filter.

RESULTS

Unless otherwise stated, all the following observa-

tions apply equally to all three examined specimens.

The earlier sexual stage seen was the multinucleate

ascogonium (Figs 1, 2, 40, 83 & 84). The ascogonia

varied in size and form from the highly nucleated,

simple spherical ascogonium (Fig. 1) to the elon-

gated, cylindrical ascogonium (Figs 83 & 84) and the

complex convoluted ascogonium (Figs 2 & 40). The

nuclei were small, densely stained and associated in

pairs. At no stage was the mechanism of dikaryoti-

zation observed. Ascogonia were found in very

young perithecia and were identified as ascogonia on

the basis of the usually multinucleate condition with

commonly, associated pairs of nuclei.

The only ascogenous hyphae seen were found in

Neocosmospora africana (Fig. 41). These were short,

indistinct and typically binucleate. After a very re-

stricted growth they showed signs of crozier forma-

tion.

Crozier and young ascus formation

Binucleate pre-croziers were found in N. africana

(Fig. 42) and isolate P (Fig. 85). These were rounded

and did not yet show the characteristic curvature of

the crozier. The binucleate condition was, however,

an indication of their origin.

Typical binucleate croziers were found (Figs 3, 43

& 86). Their development into young asci followed

the typical pattern (Figs 4, 44, 45, 46, 47, 87 & 88).

Occasional large croziers were found (Fig. 5), but

were not of unusual significance.

The young ascus enlarged and became cylindrical

after karyogamy (Figs 48, 49 & 89) which gave rise to

the diploid nucleus (Figs 6, 50 & 90).

The nuclei showed marked variations in appear-

ance during karyogamy and the subsequent diplo-

phase. At the conclusion of the conjugate division in

the binucleate crozier the nuclei were small and in-

tensely and uniformly stained. During karyogamy

the nuclei no longer stained uniformly and showed

fragments of linear bodies interpreted as partially

condensed chromosomes. The nuclei in the terminal

and basal cells remained unchanged. The diploid

nuclei were less varied in appearance. A featureless

uneven mass of chromatin, apparently surrounded

by a nuclear membrane, was sometimes observed

(Fig. 90). Older diploid nuclei showed differentiation

(Figs 6 & 50).

Division 1

Nuclear divisions in the asci are numbered accord-

ing to the system used by Singleton (1953).

During prophase of the first meiotic division the

nuclei enlarged and were usually centrally placed in

the ascus. There was very little differentiation of

chromosomes at this stage (Figs 7 & 91) and the uni-

nucleate basal and terminal cells of the crozier were

often still present at the base of the ascus (Figs 50, 91

& 92).

At leptotene the chromosomes were readily distin-

guishable as long thin strands (Figs 8, 51 & 92). In

Fig. 92 there is evidence of a nuclear membrane

around the chromosome mass. Nucleoli were demon-

strated in some leptotene nuclei (Fig. 51), but not in

others (Fig. 8).

K. T. VAN WARMELO

417

At zygotene the chromosomes were much more

distinct, and thickenings and other details became

visible. The nuclei also appeared to be larger than

during leptotene, possibly due to a better spread of a

less cohesive chromosomal mass during squashing of

the asci.

Zygotene showed chromosomes which were identi-

fied as homologues by the similarities in the terminal

chromomere patterns (Figs 9, 10, 52, 53 & 93).

Chromosomes pairs were seen in which synapsis was

apparently completed while other chromosomes were

only partially associated (Figs 52 & 93).

At pachytene the synapsed homologues were close-

ly associated and were no longer distinguishable as

chromosome pairs (Figs 11, 54 & 94). Occasionally

fully synapsed pairs were found in which the two

component chromosomes of the bivalent could be

seen (Fig. 14). Nucleoli were only occasionally pre-

sent during pachytene (Fig. 15).

Pachytene stages in which the chromosomes could

be counted were found in TV. vasinfecta (Figs 12 & 13)

and isolate P (Figs 95 & 96). No countable stages

were found in TV. africana.

In TV. vasinfecta (Figs 12 & 13) six chromosomes

were counted and their lengths estimated. Centro-

meres could not be distinguished.

The estimated lengths of the six chromosomes,

numbered in descending order of length (Singleton,

1953), were:

1 = 13,5 /xm 2 = 7,5 /xm 3 = 6,0 /tun

4 = 5,25 /xm 5 = 1,75 /xm 6 = 1,5 /xm

The lengths of chromosomes 5 and 6 were prob-

ably underestimated as the chromosomes appeared to

have been compressed during squashing of the pre-

paration.

Six chromosomes were also counted in isolate P

(Figs 95 & 96). As in TV. vasinfecta, no centromeres

were distinguishable and the nucleolus had dis-

appeared. The estimated lengths of the six chromo-

somes were:

1 = 25 /xm 2=15 fxm 3=12 /xm

4=11 /xm 5=8 /xm 6=7 /xm

Although there was little agreement between the

estimated chromosome lengths of TV. vasinfecta and

isolate P, the relative lengths as given below agreed

closely.

Relative chromosome lengths

Only in the case of the ratio of chromosomes 4 and

5 is there a marked difference between the two fungi.

It was pointed out above, however, that the lengths

of chromosomes 5 and 6 in TV. vasinfecta were prob-

ably underestimated.

During diplotene the bivalents shortened and

became fuzzy in outline (Fig. 97). Crossover points

were seen as small loops in the contracted bivalents

(Figs 16, 55 & 98). Late diplotene chromosomes were

noticeably shorter than pachytene chromosomes

(Fig. 99) and were also thinner.

In both TV. vasinfecta and TV. africana what

appeared to be a ring chromosome was seen during

diplotene (Figs 16 & 55). There was however, no evi-

dence of a ring chromosome during pachytene.

Diakinesis showed no abnormal or unexpected

features (Figs 17, 56 & 100). The chromosomes were

short, highly condensed and almost spherical. They

could not be counted.

Metaphase I chromosomes were highly condensed,

almost spherical and very short (Figs 18, 57 & 101).

They were not arranged in a metaphase plate across

the equatorial plane of the spindle, but appeared as a

dense central aggregation of chromatin in the

spindle. Spindles were sometimes clearly seen (Fig.

101) , but were more frequently indistinct, their

presence and orientation being indicated by diffrac-

tion lines in the cytoplasm of the ascus (Figs 57 & 58).

Spindles were orientated along the long axis of the

ascus (Figs 18 & 58), or at right angles to it (Figs 57 &

102) , or obliquely (Fig. 101).

Chromosome counts at metaphase I were un-

reliable due to the close association of the chromo-

somes. Occasionally, however, fairly reliable counts

could be made. In polar view (Fig. 102) six chromo-

somes were distinguished in isolate P, while the same

number was seen in TV. africana (Fig. 58).

In isolate P the chromosome count at metaphase I

agreed with that made at pachytene. Although no

countable pachytene stages were found in TV. afri-

cana, the chromosome number at metaphase I was

the same as that found in both TV. vasinfecta and iso-

late P. Important also was the observation that, in TV.

africana, there appeared to be four large chromo-

somes and two smaller ones (Fig. 58), a situation

similar to that found at pachytene in TV. vasinfecta.

In isolate P (Fig. 102) there were also size differences

between the chromosomes but they were less marked.

It was impossible to estimate the chromosome sizes

from metaphase I figures.

Separation of the chromosomes during anaphase I

appeared to be sequential rather than synchronous

and a number of different chromosome arrange-

ments were seen. In TV. africana a series of stages

illustrating sequential separation were found. A stage

was seen of the transition from metaphase I to

anaphase I (Fig. 59) in which the arrangement of the

chromosomes was virtually identical with one at

metaphase I (Fig. 58). The only differences was that

one or two bivalents had separated before the other

chromosomes. A further stage was seen (Fig. 60) in

which separation of the homologues had occurred

with only slight movement of the chromosomes on

the spindle. An even later stage was seen (Fig. 61) in

which movement of the chromosomes had begun

whilst maintaining a similar chromosomal arrange-

ment.

Sequential rather than synchronous separation of

bivalents was also seen in TV. vasinfecta (Fig. 19) and

isolate P (Fig. 103). Precocious chromosomes were

linked to the median unseparated bivalents by thin

spindle strands.

At mid anaphase I the condensed chromosomes

were spread in a band along the spindle (Fig. 62) and

individual chromosomes could not be distinguished.

At late anaphase I the chromosomes had started to

form terminal groups (Figs 20 & 104).

Telophase I nuclei were usually small, spherical

and homogeneous. At early telophase I there were

occasional lagging chromosomes (Figs 63 & 106).

However, no case was seen in which any lagging

chromosomes failed to be incorporated in the

daughter nuclei. The reconstituted daughter nuclei

rounded off at the conclusion of the division, con-

nected only by the remains of the spindle (Figs 64 &

418

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

105). Discarded nucleoli were occasionally seen (Fie.

107).

At the conclusion of telophase I the nuclei entered

interphase I. Generally this process was asynchro-

nous and asci were found in which one nucleus was in

interphase 1 while the sister nucleus was still in telo-

phase I (Fig. 21).

Interphase 1 nuclei were larger than at telophase,

approximately spherical and heterogeneous (Figs 22,

65 & 108). Individual chromosomes could not be dis-

tinguished, although nucleolus-like bodies were

sometimes seen (Fig. 22).

Division II

Prophase II nuclei were less diffuse than at inter-

phase I and stained intensely (Figs 23, 66 & 109). The

chromosomes of N. vasinfecta and N. africana were

short, whereas those of isolate P were indistinct. In N.

vasinfecta nucleoli were seen in the nuclei (Fig. 23).

Nucleoli were not seen in N. africana and isolate P.

Metaphase II was indistinct. At early metaphase

the deeply-staining chromosomes were arranged cen-

trally between the two centriolar plaques (Fig. 67).

The chromosomes were so closely associated that

counts were impossible (Figs 24, 68 & 110). Cen-

triolar plaques were not always visible although

spindles were sometimes seen (Fig. 110).

Anaphase II followed the same general pattern as

anaphase I. Spindles were either parallel with (Fig.

69), or across (Figs 111 & 112) the long axis of the

ascus. Sequential disjunction of the chromatids was

once again suggested by the linear arrangement of

chromosomes with interconnecting interzonal or

spindle fibres.

Occasionally chromosomes were clearly delimited

at anaphase II (Fig. 1 13). Six chromosomes were seen

at this stage.

Telophase II nuclei (Figs 25, 70, 114 & 115) were

identical with the telophase I nuclei. No lagging

chromosomes or spindle bridges were seen at this

stage.

Interphase II nuclei (Figs 26, 71 & 116) resembled

the interphase I nuclei described. The four interphase

II nuclei were commonly arranged as two adjacent

pairs arranged symmetrically around the centre of

the ascus (Figs 26 & 1 16).

Division III

Prophase III nuclei resembled the prophase II

nuclei described. Initially long and diffuse, the chro-

mosomes shortened and became intensely stained

(Figs 27, 72, 117 & 118).

Prometaphase III stages showed short dense chro-

mosomes (Figs 28 & 73) and occasional centriolar

plaques (Fig. 73).

Metaphase III nuclei (Figs 29 & 119) resembled the

metaphase II nuclei described. The arrangement of

the four nuclei followed no regular pattern.

Anaphase III was seen only in isolate P (Figs 120 &

121) and was similar to the previously described ana-

phase stages. Divisions of the nuclei were not fully

synchronous (Figs 120 & 121) and alignment of the

spindles was variable.

Lagging chromosomes at late anaphase 1 1 I/early

telophase 111 were seen (Fig. 30) but these became in-

corporated into the reconstituted telophase III

nuclei. Evidence for this was the absence of lagging

chromosomes at late telophase. These nuclei (Figs 74

& 122) were similar to the telophase stages described,

but could also be larger (Fig. 123).

Interphase III nuclei (Figs 31, 32, 75 & 124) re-

sembled the nuclei seen at interphase I and II.

The earliest signs of spore delimitation were seen at

interphase III (Fig. 31). There were, however, inter-

phase III stages with no signs of incipient cytoplas-

mic cleavage (Figs 75 & 124), whereas in some asci

well developed cytoplasmic cleavages were seen (Figs

32 & 76).

Division IV

Prophase IV nuclei (Figs 33, 77 & 125) were similar

to the prophase II and III nuclei described. In all

isolates asci in prophase IV without any signs of inci-

pient cytoplasmic cleavage were found (Figs 33, 77 &

125), whereas other asci also in prophase IV were

found in which well defined spore initials were clearly

visible (Figs 78 & 126).

Prometaphase IV (Fig. 127) was found in isolate P

only and was similar to prometaphase stages of the

preceding divisions. Metaphase IV and anaphase IV

stages were not found in any of the isolates.

Telophase IV nuclei (Figs 34, 35, 79, 128, 129 &

130) were identical with the previously described

telophase stages. Interphase IV nuclei (Figs 37 & 131)

also resembled similar stages in the preceding divi-

sions.

Neocosmospora vasinfecta differed from the other

isolates with respect to spore delimitation. Whereas

the spores in N. africana and isolate P were in-

variably delimited before telophase IV (Figs 79 &

129), occasional asci were found in N. vasinfecta

which contained sixteen nuclei not separated by

cytoplasmic cleavage planes (Figs 34 & 35). Such asci

were large and commonly globose. Normal cylindric

asci with cytoplasmic cleavage planes at this stage

were common (Fig. 36). As the mature spores of N.

vasinfecta were always found in asci of normal size,

it is postulated that the large uncleaved 16-nucleate

asci degenerated without delimiting spores. How,

why and when this occurred could not be determined.

Following division IV the eight ascospores in an

ascus were binucleate. Nuclei in spores of increasing

age, as determined from the increasing thickness of

the spore wall, showed a progressive condensation

(Figs 38, 39, 80, 81, 82, 132, L33 & 134). Nuclei in

mature spores were thus condensed and intensely

stained.

Ascus proliferation

Fusion of the terminal cell of the crozier with the

basal cell was commonly seen in N. africana and

isolate P, but was less common in N. vasinfecta. A

single ascogenous hypha could therefore give rise to

further asci.

DISCUSSION

The mechanism of dikaryotization was not eluci-

dated. As no signs of receptive hyphae or sper-

matium formation were seen, the most likely mech-

anism is random somatogamy. As all three isolates

are homothallic this mechanism is not unlikely.

Doguet’s (1956) observation of defined ascogonial

structures in Neocosmospora vasinfecta and N. afri-

cana was thus not confirmed.

It is well established that the sexual nuclear divi-

sion in the ascus is an ordered process corresponding

to gametogenesis in higher organisms. The division

can therefore correctly be called meiotic.

K. T. VAN WARMELO

419

Sequence of nuclear divisions in the ascus

The sequence of nuclear divisions leading to the

formation of mature ascospores was normal in all

three isolates. This confirms the statement by Doguet

(1956) that ascosporogenesis in TV. vasinfecta and TV.

africana follows the established pattern.

Homology of meiotic prophase chromosomes

Zygotene chromosome pairs were seen in all iso-

lates in which synapsis was apparently completed

while other chromosome pairs were only partially

associated. Homology of certain chromosomes was

deduced from the similarities in chromomere pattern.

Synaptic failure could either indicate slow pairing or

non-homology of chromosome pairs or areas. Only

in N. vasinfecta was one configuration seen (Figs 12

& 13) which suggested incomplete synapsis. This

could, however, have been due to a chromosome

abnormality not necessarily indicative of non-homo-

logy.

Furthermore, non-homologous chromosomes in

the karyotype would result in pachytene associations

showing incomplete bivalent formation. No post-

pachytene stages showing abnormalities were, how-

ever, observed. All the available evidence therefore

indicates that, although synapsis during zygotene is

rather slow, there is complete homology of all

chromosomes pairs.

The chromosome number of each of the three iso-

lates was determined at several stages during asco-

sporogenesis. Six chromosomes were counted in all

cases. This number agrees with the counts made dur-

ing mitotic divisions (Van Warmelo, 1977).

The lengths of the pachytene chromosomes were

estimated in N. vasinfecta and isolate P. Although

chromosome lengths at pachytene vary considerably

with the structure of the bivalent and the degree of

squashing during preparation of the material, it is

reasonable to assume that all chromosomes in a

single nucleus will be squashed to very much the same

extent. Relative chromosome lengths are therefore

probably more important than absolute lengths.

The relative lengths of the chromosomes of N. vas-

infecta and isolate P are almost identical, even

though their estimated absolute lengths are markedly

dissimilar. If it is assumed that chromosomes 5 and 6

in N. vasinfecta (Fig. 13) became distorted during

squashing, there is a close similarity between the

karyotypes of these isolates.

Although no countable pachytene stages were seen

in N. africana, it is significant that four long and two

short chromosomes were present at metaphase I (Fig.

58). As the chromosomes were highly condensed the

relative lengths could not be estimated. It is, there-

fore, postulated that the chromosome ratios of N.

africana at least approximate the ratios determined

for N. vasinfecta and isolate P.

Metaphase / anaphase

Even though spindles were not always seen, evi-

dence for their presence was provided by the presence

of centriolar plaques and also diffraction patterns in

the cytoplasm around the chromosomes. When

spindles were seen, their size and orientation agreed

with the observed diffraction patterns. Spindles are

presumably present during nuclear divisions,

although they cannot always be demonstrated.

Metaphase separation is essentially similar in all

the isolates. It is suggested that separation of

chromosome pairs/chromatids is sequential rather

than synchronous. These results are in accordance

with the statement by Olive (1965) that asynchronous

anaphase separation in the ascus appears to be the

rule in the fungi.

Although anaphase separation was asynchronous,

there was no evidence of irregular chromosome dis-

junction. This was deduced from the absence of

lagging chromosomes at late telophase or interphase,

indicating complete reconstruction of the daughter

nuclei, the constancy of chromosome numbers and

the phenotypic constancy of the cultures. It can,

therefore, be assumed that the regularity of karyo-

type and genome replication is high.

Ascopore delimitation

The occasional absence of cytoplasmic cleavage

planes at telophase IV and interphase IV might be

regarded as an important feature of Neocosmospora

vasinfecta, and an important difference between it

and N. africana and isolate P. However, the fact that

the 16-nucleate uncleaved asci were not seen to form

ascospores and apparently degenerated, reduces the

emphasis that should be placed on this characteristic

as taxonomic criterion. Although it is a difference be-

tween N. vasinfecta and the other isolates, it does not

seem to be an important one.

There are, therefore, no significant differences be-

tween TV. vasinfecta E. F. Smith, N. africana Von

Arx and ‘isolate P’ and the former two species can-

not De distinguished on the basis of sexual

chromosomal data.

ACKNOWLEDGEMENTS

This study was supported by research grants from

the Rand Afrikaans University, Johannesburg and

the Council for Scientific and Industrial Research,

Pretoria.

UITTREKSEL

’n Swamsoort afkomstig van sojaboonweefsel het

duidelike ooreenkomste getoon met twee bestaande

Neocosmospora soorte, t.w. N. vasinfecta E. F.

Smith en N. africana Von Arx. Die geslagtelike kern-

delings in outentieke kulture van hierdie twee soorte

en die plaaslike isolaat is ondersoek om vas te stel of

daar verskille tussen die kulture was. Geen betekenis-

volle verskille is waargeneem nie. Ses chromosome is

in alle gevalle aangetref en indie relatiewe lengtes was

blykbaar dieselfde. Daar was geen tekens van aneu-

ploiedie of onreelmatige vorming van dogterkerne

nie. Delings is dus tipies meioties en lei tot vorming

van tweekernige homokariotiese askospore.

REFERENCES

Aldrich, H. C., 1967. The ultrastructure of meiosis in three

species of Physarum. Mycologia 59: 127-148.

Barry, E. G., 1967. Chromosome aberrations in Neurospora, and

the correlation of chromosomes and linkage groups. Genetics,

Princeton 55: 21-32.

Cooney, D. G. & Emerson, R., 1964, Thermophilic fungi.

London: W. H. Freeman. 188 pp.

Doguet, G., 1956. Morphologie et organogenie du Neocosmos-

pora vasinfecta E. F. Smith et du Neocosmospora africana

Von Arx. Annls Sci. nat. Bot. ser. 11, 17: 353-370.

Ferry, R., 1900. Un nouveau genre de champignon (Neocosmos-

pora) qui constitue un redoutable fleau pour le cotonnier, la

citrouille et la Vigna sinensis. Revue My cot. 22: 121-124.

Furtado, J. S., 1970. Ascal cytology of Sordaria brevicollis.

Mycologia 62: 453-461.

Huffman, D. M., 1968. Meiotic behavior in the mushroom

Colly bia maculata var. scorzonerea. Mycologia 60: 451-456.

HUGUEN1N, B. & Boccas. B., 1970. Etude de la caryocinese chez le

Phytophthora parasitica Dastur. C. R. Acad. Sci. Paris, ser.

D. 271: 660-663.

420

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

Knox-Davies, P. S. & Dickson. J. G., 1960. Cytology of Helmin-

thosporium turcicum and its ascigerous stage, Trichometa-

sphaeria turcica. Am. J. Bot. 47: 328-339.

Lu, B.C., 1966. Fine structure of meiotic chromosomes of the

Basidiomycete Coprinus lagopus. E.xpl Cell Res. 43: 224-227.

Lu, B. C., 1967a. The course of meiosis and centriole behaviour

during the ascus development of the Ascomycete Gelasino-

spora calospora. Chromosoma 22: 210-226.

Lu, B. C., 1967b. Meiosis in Coprinus lagopus: a comparative

study with light and electron microscopy. J. Cell Sci. 2:

529-536.

Moreau, C. & Moreau, Mireille., 1950. Neocosmospora

vasinfecta E. F. Smith. Faux wilt du cotonnier. Fiches de

Phytopathologie. Revue My col. 15. Suppl. 2. 5 p.

Olive, L. S., 1953. The structure and behavior of fungus nuclei.

Bot. Rev. 19: 439-586.

Olive, L. S., 1965. Nuclear behavior during meiosis. Chapter 7.

In G. C. Ainsworth & Sussman, A. S. (eds), The fungi. An

advanced treatise. Vol. 1. London: Academic Press.

Rajendren, R. B., 1967. A new type of nuclear life cycle in

Hemileia vastatrix. Mycologia 59: 279-285.

Rao, V. R. & Mukerji, K. G., 1970. Cytology of the ascus in

Ascotricha guamensis. Mycologia 62: 301-306.

Rogers, J. D., 1964. Hvpoxylon pruinatum: the chromosome

number. Mycologia 56: 369-373.

Rogers, J. D., 1965. Hvpoxylon fuscum. 1. Cytology of the ascus.

Mycologia 57: 789-803.

Rogers, J. D., 1967. Hvpoxylon multiforme: cytology of the

ascus. Mycologia 59: 295-305.

Rogers, J. D., 1968a. Nuclear phenomena in the ascospores of

Hypoxylon punctulatum. Can. J. Bot. 46: 865-866.

Rogers, J. D., 1968b. Xylaria curta: cylo\ogy of the ascus. Can. J.

Bot. 46: 1337-1340.

Rogers, J. D., 1968c. Hvpoxylon deustum: the chromosome

number. Mycopath. Mycol. appl. 35: 249-255.

Rogers, J. D., 1971. Observations on the ascogenous system of

Hvpoxylon microplacum. Can. J. Bot. 49: 1075-1077.

Seaver, F. J., 1909. The Hypocreales of North America — I.

Mycologia 1: 41—76.

Singleton, .1. R., 1953. Chromosome morphology and the chro-

mosome cycle in the ascus of Neurospora crassa. Am. J. Bot.

40: 124-144.

Swanson, C. P., 1968. Cytology and cytogenetics. London: Mac-

millan 596 pp.

Udagawa.S., 1963. Neocosmospora i n Japan. Trans. Mvcol. Soc.

Jap. 4: 121-125.

Uecker, F. A., 1967. Stephensia shanori. 1. Cytology of the ascus

and other observations. Mycologia 59: 819-832.

Van Warmelo, K. T., 1966. The cytology of Mvcosphaerella

pinodes. Bothalia 9: 195-202.

Van Warmelo, K. T., 1971. Somatic nuclear division in Stem-

phylium botryosum. Bothalia 10: 329-334.

Van Warmelo, k. T., 1976. Scanning electron microscopy of

Neocosmospora ascospores. Mycologia 68: 1181-1187.

Van Warmelo, K. T., 1977. Asexual nuclear division in Neocos-

mospora. Bothalia 12: 247-250.

Von Arx, J. A., 1955. Ein neuer Ascomycet aus Afrika. Antonie

van Leeuwenhoek 21: 161-165.

Wells, K., 1970. Light and electron microscopic studies of Asco-

bolus stercorarius. I. Nuclear divisions in the ascus.

Mycologia 62: 761-790.

Woo. .1. Y. & Partridge, A. D., 1969. The life history and

cytology of Rhvtisma punctatum on big leaf maple. Mycologia

61: 1085-1095.

Zickler, Denise., 1970. Division spindle and centrosomal plaques

during mitosis and meiosis in some Ascomycetes. Chromo-

soma 30: 287-304.

K. T. VAN WARMELO

421

Figs 1-20. — Meiosis in Neocosmospora vasinfecta: x 2 000 unless otherwise stated. 1, spherical multinucleate

ascogonium; 2, morphologically complex multinucleate ascogonium; 3, binucleate crozier; 4, tetranucleate

crozier; 5, tetranucleate crozier; 6, crozier after karyogamy; 7, beginning of prophase I; 8, leptotene; 9,

zygotene; 10, zygotene/pachytene; 11, pachytene; 12, pachytene; 13, pachytene with numbered

chromosomes, x 4 000; 14, pachytene showing bivalent structure, x 4 000; 15, late pachytene; 16,

diplotene; 17, diakinesis; 18, metaphase 1; 19, early anaphase 1; 20, mid anaphase 1.

422

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

*

f

i

I

, * *1

Figs 2 1 -32. — VIeiosis in Neocosmospora vasinfecta: x 2 000. 2 1 , telophase I ; 22, interphase I; 23, prophase II;

24, metaphase II; 25, telophase II; 26, interphase II; 27, prophase 111; 28, prometaphase III; 29, metaphase

III; 30, anaphase/telophase III; 31, interphase III; 32, interphase III in delimited ascospores.

K. T. VAN WARMELO

423

4 ^ *»*»..'*

Figs 33-39. — Meiosis in Neocosmospora vasinfecta: x 2 000. 33, prophase IV; 34, prophase IV without spore

delimitation; 35, telophase IV; 36, telophase IV; 37, interphase IV; 38, mature binueleate ascospores; 39,

mature binueleate ascospores.

424

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

Figs 40-59. — Meiosis in Neocosmospora africana: x 2 000. 40, ascogonium with paired nuclei; 41 , ascogenous

hypha; 42, binucleate pre-crozier; 43, binucleate crozier; 44, nuclear division in the binucleate crozier; 45,

tetranucleate crozier after nuclear division; 46 tetranucleate crozier; 47, crozier after septum formation; 48,

karyogamy in the young ascus; 49, diploid nucleus in the young ascus; 50, beginning of prophase I; 51, lep-

totene; 52, zygotene; 53, zygotene/pachytene; 54, pachytene; 55, diplotene; 56, diakinesis; 57, metaphase I

with transverse spindle; 58, metaphase 1 with longitudinal spindle; 59, metaphase/anaphase I.

K. T. VAN WARMELO

425

Figs 60-74. — Meiosis in Neocosmospora africana: x 2 000. 60, early anaphase I; 61, anaphase 1; 62, anaphase

1; 63, anaphase/telophase I; 64, telophase 1; 65, interphase 1; 66, prophase II; 67, early metaphase II; 68,

late metaphase II; 69, anaphase II; 70, telophase II; 71, interphase II; 72, prophase III; 73, prometaphase

III; 74, telophase III.

426

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

Ligs 75-82. — Meiosis in Neocosmospora africana: x 2 000. 75, interphase III before spore delimitation;

76, interphase III before spore delimitation; 77, early prophase IV: 78, prophase IV; 79, telophase IV; 80,

interphase in binucleate ascospores; 81, interphase in binucleate ascospores; 82, interphase in binucleate

ascospores.

*■% .

K. T. VAN WARMELO

427

Figs 83-105. — Meiosis in Neocosmospora isolate P: x 2 000 unless otherwise stated. 83, ascogonium with two

nuclear pairs; 84, ascogonium with three nuclear pairs; 85, binucleate pre-crozier; 86, binucleate crozier; 87,

tetranucleate crozier after nuclear division; 88, tetranucleate crozier; 89, karyogamy in the young ascus; 90,

diploid nucleus in the young ascus; 91 , beginning of prophase I; 92, leptotene; 93, zygotene; 94, pachytene;

95, pachytene, x 4 000; 96, pachytene with numbered chromosomes, x 4 000; 97, pachytene/diplotene;

98, diplotene; 99, late diplotene; 100, diakinesis; 101, metaphase I; 102, metaphase I in polar view; 103,

anaphase I; 104, late anaphase 1; 105, telophase I.

428

SEXUAL NUCLEAR DIVISION IN NEOCOSMOSPORA

Figs 106-125. — Meiosis in Neocosmospora isolate P: x 2 000. 106, telophase I with slow chromosome; 107,

telophase I with residual nucleolus; 108, interphase 1; 109, prophase II; 110, metaphase II; 111, metaphase

II; 112, anaphase II; 113, metaphase/anaphase II; 1 14, telophase II; 1 15, late telophase II; 1 16, interphase

II; 117, early prophase III; 118, prophase 111; 119, metaphase III; 120, anaphase III; 121, anaphase III; 122,

telophase III; 123, large telophase III nuclei; 124, interphase III; 125, prophase IV.

• s»

K. T. VAN WARMELO

429

Figs 126-134. — Meiosis in Neocosmospora isolate P: x 2 000. 126, prophase IV with spore delimitation; 127,

prometaphase IV; 128, telophase IV with residual median nucleoli; 129, telophase IV: 130, late telophase

IV; 131, interphase IV; 132, interphase in binucleate ascospores; 133, interphase in binucleate ascospores;

134, interphase in binucleate ascospores.

Bolhalia 13, 3 & 4: 431-433 (1981)

Miscellaneous notes on the genus Pelargonium

J. J. A. VAN DER WALT* and P. J. VORSTER*

ABSTRACT

It is pointed out that the name sect. Pelargium DC. must be replaced by sect. Pelargonium , and Pelargonium car-

diophyllum Harv. (1860) by P. setulosum Turcz. (1858); P. mossambicense Engl, is reported as a new record for the

Transvaal; a new name P. ternifolium Vorster is given to P. Irifolialum Harv., nom. illeg.; and the typification of

P. multifidum Harv., P. plurisectum Salter, P. dolomiticum Knuth and P. dasyphyUum E. Mey. ex Knuth is

discussed.

NOTES DIVERSES SUR LE GENRE PELARGONIUM

// est souligne que le nom sect. Pelargium DC. doit etre remplace par sect. Pelargonium, et Pelargonium cardio-

phyllum Harv. (I860) par P. setulosum Turcz. (1858); P. mossambicense Engl, est rapporte comme un nouvel

enregistrement pour le Transvaal; un nouveau nom, P. ternifolium Vorster est donne a P. trifoliatum Harv. nom.

illeg.; et I’establissement de types de P. multifidum Harv., P. plurisectum Salter, P. dolomiticum Knuth et P. dasy-

phyllum E. Mey. ex Knuth est discute.

CORRECT NAME OF THE TYPE SECTION OF THE GENUS

PELARGONIUM

The genus Pelargonium L’Herit. was typified** by

Van der Walt in J1 S. Afr. Bot. 45: 377 (1979), who

chose P. cucullatum (L.) L’Herit. ( Geranium

cucullatum L.) as the lectotype species. The section

containing this species was placed as sect. Pelargium

DC., Prodr. 1: 658 (1824) and this was followed by

Harvey in FI. Cap. 1 : 301 (1860), Knuth in Pflanzenr.

4, 129: 455 (1912) and Van der Walt, Pelargoniums

5. Afr. xiv (1977). According to Art. 22 of the Code,

this section, containing the type species, must be

referred to as Pelargonium L’Herit. sect.

Pelargonium.

PELARGONIUM MOSSAMBICENSE ENGL., NEW TO THE

FLORA OF SOUTHERN AFRICA REGION

Pelargonium mossambicense Engl., Pflanzenw.

Ost.-Afr. C: 225 (1895) has been known up to now

from only a few collections on the Zimbabwe side of

the mountains bordering Zimbabwe and Mozambi-

que, and from the type collection made at Gorongosa

in Mozambique in 1884-1885. It can now be

announced that it also occurs in the Transvaal, hav-

ing been found ca. 1,6 km east of Pilgrim’s Rest on

the road to Graskop (Cockl 9797 in K!) and at

Pilgrim’s Rest (Rogers 23257 in BOL!). These

records extend the known distribution of this species

southwards by about 700 km. This distribution pat-

tern, with populations widely separated by the broad

Limpopo Valley, with small relic populations in the

south, is shared by several species, for example the

ferns Pvrrosia schimperana (Mett.) Alston [cf.

Vorster in Bothalia 11: 287 (1974)] and Selaginella

imbricata (Forssk.) Spring ex Decne. [op cit. 12: 259

(1977)], Leucospermum saxosum S. Moore [cf.

Rourke in Flower. PI. Afr. 41: t. 1633 (1971)] and

*Department of Botany, University of Stellenbosch, Stellenbosch,

7600.

**Hanks & Small, N. Amer FI. 25: 23 (1907) also designated a

lectotype species for the genus Pelargonium, but for reasons ex-

plained elsewhere (in press) this is not acceptable.

Aloe swynnertonii Rendle [cf. Reynolds, Aloes S.

Afr. 220 — 222 (1950) as A. chimanimaniensis ].

NEW NAME FOR PELARGONIUM TRIFOLIATUM HARV.

Pelargonium ternifolium Vorster, nom, nov.

P. trifoliatum Harv. in FI. Cap. 1: 271 (1860); Knuth in

Pflanzenr. 4,129: 351 (1912); non Sweet (1926); nec P. trifolio-

latum (Eckl. & Zeyh.) Steud. (1841). Type: ‘Cape, Klein Draken-

stein, Stell.’ (Stellenbosch), Drege 7497 ( K ! ; L!; P!; S!; W!).

P. CARDIOPHYLLUM HARV. VERSUS P. SETULOSUM

TURCZ.

P. cardiophyllum Harv. was described in FI. Cap.

1: 284 (1860), based on Ecklon & Zeyher 601 and

Zeyher 2084. The name was subsequently wrongly

upheld by Knuth in Pflanzenr. 4, 129: 389 (1912),

who placed the earlier name, P. setulosum Turcz. in

Bull. Soc. nat. Moscow 31: 422 (1858), based on

Zeyher 2084, in synonymy. The latter is the correct

name for this species.

Earlier, Ecklon & Zeyher, Enum. 1 : 77 ( 1 835), had

wrongly associated their specimen Ecklon & Zeyher

601 with Geranium elegans Andr., Bot. Rep. 1: t. 28

(1799), when they effected the combination Eumor-

pha elegans (Andr.) Eckl. & Zeyh.

Harvey ( l.c .) also described a variety laciniatum,

having the ‘leaves 3-lobed, the lobes laciniate and

toothed’. However, both undivided and deeply lobed

leaves may occur on the same plant and, with our

present knowledge of the species, it does not appear

justified to recognize the variety as a separate entity.

Pelargonium setulosum Turcz. in Bull. Soc. nat.

Moscow 31: 422 (1858). Type: Cape, ‘C, b. spei’,

Zevher 2084 (? MW, holo.; G!; K!; PRE, 2 sheets!;

S!; W!; WU!; Z!).

Eumorpha elegans sensu Eckl. & Zeyh., Enum. 1:77 (1835), as

to Ecklon & Zeyher 601, non (Andr.) Eckl. & Zeyh. l.c. [Geranium

elegans Andr., Bot. Rep. 1: t. 28 (1799)], nec Pelargonium elegans

(Andr.) Willd., Sp. PI. ed. 4, 3: 655 (1800).

P. cardiophyllum Harv. in FI. Cap. I: 284 (1860); Knuth in

Pflanzenr. 4, 129: 389 (1912), syn. nov. Syntypes: Cape, ‘Rocky

mountain sides, above the Baths, and Baviansberg, GnadenthaT,

Ecklon & Zeyher 601 (SI; SAM!); ‘River Zonderende’, Zeyher

2084 (G!; K!; ? MW; PRE, 2 sheets!; S!; W!; WU!; Z!).

432

MISCELLANEOUS NOTES ON THE GENUS PELARGONIUM

Geraniospermum cardiophyllum (Harv.) Kuntze, Rev. Gen. 1 :

94 (1891). P. cardiophyllum Harv. var. laciniatum Harv. l.c., syn.

nov. Type : Cape, without precise locality, Zeyher s.n. (S ! ).

TYPIFICATION OF PELARGONIUM MULT I FID UM HARV.

AND P. PLURISECTUM SALTER

With the original description of Pelargonium mul-

tifidum Harv. in FI. Cap. 1: 282 (1860) two speci-

mens were cited, viz. Drege 9460 and Zeyher 2054

('Herb. Sond.’). From this citation the impression is

gained that two syntypes are involved. However,

both numbers appear on single sheets in S, G and P,

on two sheets in G and P respectively on the same

label, whereas only the number Zeyher 2054 appears

on sheets in P and W. The probable solution is provi-

ded by one of the three sheets in P, which bears both

numbers as well as the inscription ‘Legit Zeyher.

Communicavit Drege . ..’. It therefore seems as if

only one collection is involved, namely Zeyher 2054

which was distributed under the number Drege 9460.

The material of this species in Harvey’s collection

(TCD) consists of three small branches of Zeyher

2054 and a tiny fragment marked Drege 9460, but

without any indication of locality of origin so that it

is unlikely that this sheet was used solely or mainly

for compiling the original description. Sonder’s main

collection, referred to by Harvey, was acquired by

Stockholm in 1875 and there is a sheet in S containing

both numbers as well as the locality and name written

on in Sonder’s handwriting. We therefore designate

this sheet as lectotype of Pelargonium multifidum

Harv. (1860), non Salisb. (1796), as well as of P.

plurisectum Salter (1941) which replaces the former

name.

NEOTYPE FOR PELARGONIUM DOLOMITICUM KNUTH

With the original description of P. dolomiticum

Knuth in Bot. Jb. 40: 71 (1907) it was stated that the

type ( Engler 2889) was housed in B (‘Typus in herb.

Berol.!’). This specimen can no longer be traced in B,

and it is presumed to have been destroyed during

World War II. Neither could any duplicates be

traced.

Although herbarium material of this species was

borrowed from all the major herbaria in South

Africa and Europe, it transpired that these collec-

tions do not contain any specimens with duplicates

suitable for typification purposes. It appears as if the

petals are easily shed during the preparation process,

so that complete specimens with ample flowers prov-

ed to be very scarce. We finally decided to designate

the following specimen as neotype : Orange Free

State, (2926): Bloemfontein (-AA), J. W. Mostert

1661 (PRE!).

This is not a very neatly prepared specimen, but it

has more flowers than most seen by us, and it is com-

plete with mature fruits and part of the perennial

basal part of the stem. It conforms with Knuth’s

original description, it is representative of the species

and provides a good impression of the species in

general.

Even though no type material survives, there is no

doubt about the identity of P. dolomiticum. The

original description did not mention that there are

four petals only, but it did mention that the posterior

petals are up to 20 mm long. This serves to distin-

guish P. dolomiticum from P. senecioides L’Herit.

which has five petals not longer than ca. 9 mm, (with

which P. dolomiticum is otherwise likely to be con-

fused). In fact, even Knuth in Pflanzenr. 4, 129:

401-402 (1912) cited under P. senecioides a number

of specimens which undoubtedly are P. dolomiticum

[ Dinter 976 (SAM!); 662; Fleck 488a; 571a; 821 (all

in Z!)].

THE DELIMITATION AND TYPE COLLECTION OF PELAR-

GONIUM DASYPHYLLUM E. MEY. EX KNUTH

The name Pelargonium dasyphyllum first appeared

as a nomen nudum in Drege ’s Zwei pflanzengeo-

graphische Dokumente, pp. 60, 90, 91 and 209

(1843). Harvey, in FI. Cap. 1: 279 (1860) listed it in

the synonymy of P. crithmifolium J. E. Sm. Knuth,

in Ptlanzenr. 4, 129: 373 (1912) acknowledged it as a

separate entity and validated the name by providing a

description and indicating Drege 3245 as a type. In

addition to the type, Knuth cited several other collec-

tions.

Of the material cited by Knuth, we have only been

able to trace Rehmann 2776 and 2823 (both in Z) and

two type sheets. The Rehmann specimens are very

poor fragments impossible to identify positively, and

were collected in the Hex River Valley, which is far

outside the distribution area of P. dasyphyllum as in-

dicated by other collections. Of the type sheets, the

first, in P, consists of a fragment which appears to be

the upper portion of a branch which branched rather

profusely, with a dense mass of leaves and a single

unbranched peduncle with a three-flowered pseudo-

umbel. The second sheet, in PRE, consists of a single

short, branched inflorescence branch, and came

originally from P. It is not evident that it is a frag-

ment of the specimen in P, so that it must be con-

sidered to be a duplicate. While these two fragments

are hardly sufficent to provide an image of the

species, Knuth’s description could have been com-

piled from them, except for the fruit which is not evi-

dent on either of the sheets.

During our investigations of live material, both in

the field and in cultivation, we separated a number of

specimens of what we consider to be a separate spe-

cies from P. alternans Wendl. These specimens seem

to match the description and type of P. dasyphyllum.

According to our observations, P. dasyphyllum at

first glance is morphologically very similar to P.

alternans and only remotely resembles P. crithmi-

folium. It is, however, readily distinguishable from

P. alternans, to such an extent that we have no hesi-

tation in recognizing it as a separate species, as set

out in Table 1. The branched inflorescence branches

are nevertheless reminiscent of P. crithmifolium, and

probably signify a close relationship to that species.

The length of the pedicel and the hairiness of the

leaves are characters which usually hold good, but

these cannot be used exclusively to distinguish the

two species. P. alternans occasionally resembles P.

dasyphyllum in respect of these characters and it is

recommended that a combination of characters be

used for this purpose.

The type locality of P. dasyphyllum was stated by

Knuth (l.c.) to be ‘near the Copper Mountains’.

However, neither of the two sheets of the type collec-

tion which we have seen, bears that inscription. The

sheet in P is labelled ‘Camdeboosberg’ on a label dif-

ferent from that bearing the number. This is pro-

bably wrong, as Camdeboosberg is far outside the

J. J. A. VAN DER WALT AND P. J. VORSTER

433

TABLE 1.— Comparison of diagnostic characters of Pelargonium dasyphyUum and P. allernans

known geographical range of P. dasyphytlum. The

sheet in PRE is labelled ‘Zilverfontein’, which is well

within the known distribution range of this species.

Drege ( op cit.) cited three localities for P. dasyphyl-

lum:

1. ‘Camdeboosberg’ (p. 60), which, as explained

above, could not have been this species. The numbers

cited also did not include the type number.

2. ‘Zilverfontein’ (p. 90) This is well within the

known geographical range of this species, but again

the cited numbers did not include the type number.

3. ‘Zwischen Zilverfontein, Kooperbergen und

Kaus’ (p. 91). No collection numbers were cited, but

the description of the locality is not dissimilar to that

given by Knuth (l.c.). We therefore conclude that,

despite the labels on the type sheets, the correct type

locality is probably ‘Between Silwerfontein, Copper

Mountain and Kaus’ (about 29'/2°S, 18°E). We can

only guess whether Knuth (l.c.) obtained his informa-

tion from a now-lost sheet of Drege 3245 or deduced

it from Drege (op. cit.).

Attention is drawn to a sheet in S which has the

words ‘Zwischen Zilverfontein, Kooperbergen and

Kaus’ typed on the label, and the name ‘Pelargon.

dasyphyllum’ written on in Meyer’s handwriting.

This may well be a duplicate of the type collection,

but it lacks the number cited by Knuth. The specimen

is a small fragment of poor quality.

ACKNOWLEDGEMENTS

We are indebted to the following instances and

persons : the C.S.l.R. and the Research Fund of the

University of Stellenbosch for financial support; Mr

C. H. Stirton for determining for us the precise date

of publication (7 September 1858) of the name P.

setulosum, while stationed at Kew as liaison officer

of the Botanical Research Institute; Mr P. Drijfhout,

horticulturalist in charge of our live collection of

Pelargonium, who collected information on the sec-

tion Otidia of Pelargonium both in the field and in

the greenhouse over several years, and who drew our

attention to a number of characters separating P.

dasyphyUum from P. allernans.

UITTREKSEL

Daar word op gewys dat die naam seksie Pelar-

gium DC. vervang moet word dear seksie Pelar-

gonium, en Pelargonium cardiophyllum Harv. (I860)

dear P. setulosum Turcz. (1858); P. mossambicense

Engl, word gemeld as ’n nuwe rekord vir die Trans-

vaal; ’n nuwe naam P. ternifolium Vorster word

gegee aan P. trifoliatum Harv., nom. illeg.; en die

tipifikasie van P. multifidum Harv., P. plurisectum

Salter, P. dolomiticum Knuth en P. dasyphyUum E.

Mey. ex Knuth word bespreek.

Bothalia 13, 3 & 4: 435-461 (1981)

Notes on African plants

VARIOUS AUTHORS

AMARYLLIDACEAE

A NEW SPECIES OF STRUMAR/A

Strumaria barbariae Oberm., sp. nov., S. pho-

nolithica affinis, sed planta majora perianthio infun-

dibuliforme differt.

Bulbus oblongo-ovoideus, c. 40 mm altus tunicis

coriaceis in collo productis. Folia 2(— 4), opposita,

erecta, linearia, c. 120(-200) x 8 mm, apice breve

acuto, Umbella pedunculo c. 400 mm longo; flores

6-12, patentes vel nutantes, pedicellis 15 mm longis.

Perianthium infundibuliforme, c. 40 mm longum,

album aliquantum roseolum. Tepala cohaerentia,

lobis anguste obovatis, 10 mm longis. Stamina subae-

qualia, inclusa, fundo tepalorum vel styli connata;

anthera alba, versatiiis, dorsifixa. Stylus teretus, ex-

sertus, 25 mm longus; stigmata 3, minuta; ovarium

ovulis 6 in quoque loculo.

Type. — South West Africa, 2616 (Aus): farm ‘De

Aar’, c. 30 km ESE of Aus (-CB), Lavranos &

Pehlemann 17153 (PRE, holo.).

Bulb oblong-ovoid, c. 40 mm tall, somewhat com-

pressed laterally, the old leaf-bases forming a neck c.

50-90 mm long. Leaves subhysteranthous, 2-4 per

shoot, erect, opposite, produced from a separate

lateral bud, linear, up to 200 mm long when fully

developed, 8 mm wide, shiny, dark green. Umbel c.

Fig. 1. — Strumaria barbariae. Photo of holotype specimen,

Lavranos & Pehlemann 17153, x 0,6.

6(-12)-flowered; spathe valves 2, small, mem-

branous; bracteoles small, filiform; pedicels 20-40

mm long, terete, firm. Flowers sweetly scented, pa-

tent to nutant, white but tinted light pink near base.

Tepals 6, cohering to form a funnel-shaped tube c. 30

mm long and 7 mm in diam. above; lobes spreading,

shortly acute. Stamens 6, subequal, the inner some-

what shorter; the filaments of the outer stamens

fused to base of tepals, those of inner stamens fused

to base of style forming a short triangular column;

anthers yellow, versatile, introrse with parallel

locules. Ovary 3-locular, ovules c. 6 in each locule;

style terete, shorter than stamens, stigma obtuse,

minutely 3-lobed. Capsule globose, c. 10 mm diam.,

thin-walled; seeds globose, c. 4 mm. Fig. 1.

S.W.A. — 2616 (Aus): Farm De Aar, c. 30 km ESE of Aus,

abundant but local on Schwarzkalk dolomite (-CB), Lavranos &

Pehlemann 17153, Erni 1038, Giess 13683.

This species from the farm ‘De Aar’ in South West

Africa was at first thought to represent Dinter’s

species S. phonolithica. Dinter’s type was destroyed

in Berlin in 1943 and his description is somewhat

vague. However, bulbs collected by Merxmuller and

Giess (28384) in the Klinghardt Mountains, Dinter’s

type locality, flowered in the Munich Botanic Garden

in 1973-74 and disproved its identity with the De Aar

species. Bulbs collected on farm De Aar in 1979

flowered at the Johannesburg Municipal Nurseries in

February 1980. Mrs Barbara Jeppe painted them for

a forthcoming book. It gave me this opportunity to

dedicate the species to her in appreciation of the

valuable contributions she has made to the know-

ledge of our flora.

Strumaria barbariae differs from S. phonolithica

Dinter in its larger size and in the shape of the

perianth, which is funnel-shaped, not salver-shaped

as in the Dinter species. Both species lack the

characteristic, often liver-coloured, ringed sheath

surrounding the base of the leaf-cluster so often pre-

sent in the Cape species. Most of these southern

species form no perianth-tube or only a short one,

the free tepals forming a loose bell-shaped perianth.

In most of them the style-base is swollen (strumose),

which suggested the name Strumaria. The genus is

not well known.

A. A. Obermeyer

ASCLEPIADACEAE

THE RIOCREUXIA ELANAGANII COMPLEX: A REASSESSMENT

The Riocreuxia flanaganii complex [see R. A. Dyer

in Bothalia 12,4: 632 (1979) and in FI. S. Afr. 27,4;

87 (1980)] is reassessed in this note. Of the three sub-

species, subsp. woodii is restored to specific rank,

subsp. alexandrina is raised to specific rank and

subsp. segregata is reduced to synonymy under R.

burchellii.

The revised relationship of the four previous

subspecies is reflected in the following key. In all taxa

the outer corona arises from above the base of the

staminal column and the inner lobes reach as high as

or higher than the staminal column.

436

NOTES ON AFRICAN PLANTS

Outer corona-lobes small, not spreading, 2-toothed; peduncles

0-20 mm long; corolla 10-12(13) mm long; stems bifar-

iously pubescent 1. R.flanaganii

Outer corona-lobes spreading or spreading-erect; peduncles

usually 20 mm or more long; stems usually unifariously

pubescent:

Outer corona-lobes slender, spreading-erect, bifid, equal to or

overtopping staminal column; corolla 8-10(12) mm long

2. R. alexandrina

Outer corona-lobes spreading, subtruncate of bifid; corolla

(14)15 - 18 mm long:

Inflorescence of (1)2 umbellate clusters of flowers; corolla

± 15 mm long; tube with globose base; outer corona-

lobes ± truncate 3 . R. woodii

Inflorescence usually elongate, several-noded, rarely only (5)

20 mm long, few to several flowers from a node; corolla

± 17 mm long with slightly inflated base; outer corona

lobes spreading, bifid 4. R. burchellii

1. Riocreuxia flanaganii Schltr. in Bot. Jb. 18,

Beibl. 45: 13 (1894); R.A. Dyer in FI. S. Afr. 27,4: 87

(1980) in part excluding subspp. woodii, segregata

and alexandrina.

It is recommended that the concept of this taxon

should revert to that afforded it by the original

author, Schlechter, and substantiated by N. E. Br. in

FI. Cap. 4,1: 804 (1908), with a distribution in the

eastern Cape and southern Transkei.

2. Riocreuxia alexandrina (Huber) R.A. Dyer,

stat. nov.

Ceropegia flanaganii Schltr. var. alexandrina Huber in Mems

Soc. broteriana 12: 169 (1958). Riocreuxia flanaganii Schltr.

subsp. alexandrina (Huber) R. A. Dyer in Bothalia 12: 632 (1979);

in FI. S. Afr. 27, 4: 88 (1980).

Perennial herb with a cluster of subfleshy roots.

Stems twining, with scattered hairs or hairs in single

row. Leaves broadly cordate at base, ovate-lanceo-

late, up to about 75 mm long, 50 mm broad, more or

less hairy. Peduncles short or up to about 35 mm

long, with rather dense, more or less terminal clusters

of flowers. Corolla 8-10(12) mm long, Vi — Vi divid-

ed; tube slightly inflated near base, 3-5 mm long;

lobes 5-6 mm long, filiform, united at tips. Staminal

column about 1 mm high. Corona arising about

halfway up staminal column; outer lobes slender,

spreading-erect, as high as or slightly higher than

staminal column; inner lobes oblong-erect, mem-

branous, slightly higher than the staminal column.

Our knowledge of this species is based on speci-

mens collected by Rudatis 1540 in southern Natal, in-

land from Port Shepstone, near Moyeni at about 700

m. Some specimens are associated with tall grass-

land. The species is notable for its small flowers, the

corolla of which is divided to at least half to two-

thirds of its length.

3. Riocreuxia woodii N. E. Br. in FI. Cap. 4,1:

803 (1908).

Riocreuxia flanaganii Schltr. subsp. woodii (N. E. Br.) R. A.

Dyer in Bothalia 12,4: 632 (1979); in FI. S. Afr. 27,4: 87 (1980).

Ceropegia flanaganii Schltr. var. fallax Huber in Mems Soc.

broteriana 12: 169 (1958).

This species is known only from the type specimen

collected over 100 years ago by Medley Wood, near

Inanda, inland from Durban. The locality falls with-

in the distribution area of the variable species Rio-

creuxia torulosa Decne. and it requires careful obser-

vation of the coronal structure to distinguish the two

species. Its association with R. flanaganii seems un-

warranted and its resuscitation to specific rank is ad-

vocated.

4. Riocreuxia burchellii K. Schum. in Nattirl.

PflFam. 4,2: 273 (1895); R. A. Dyer in FI. S. Afr.

27,4: 85 (1980).

Riocreuxia flanaganii Schltr. subsp. segregata R. A. Dyer in

Bothalia 12,4: 632 (1979); in FI. S. Afr. 27,4: 87 (1980).

Of the two known collections placed under R.

flanaganii Schltr. subsp. segregata, the one, Fitz-

Simons & Van Dam in TRV 25981 (PRE), was origin-

ally identified as R. torulosa Decne. and the other,

Pole Evans 19656 was originally identified as R. picta

Schltr. The abnormally short peduncles of these two

cited specimens distracted attention from the third

species with an overlapping distribution near Wak-

kerstroom, namely R. burchellii (-R. polyantha).

The reassessment is that the two specimens in ques-

tion represent an unusual growth form of R. bur-

chellii.

R. A. Dyer

COMMELINACEAE

TWO NEW SPECIES OF COMMELINA

Commelina bella Oberm., sp. nov. C. living-

stonii affinis, sed planta majora, robustiora, foliis ad

marginem crenulatis, albis, racemo inferiore,

1-3-floribus, longe pedunculato valde differt.

Herba perennis (chamaephyta), erecta, compacta,

setulosa, ad 350 mm alta. Caudex compactus,

radicibus lignosis longis. Folia linearia 40-50 x 10

mm, margine undulato incrassato albo. Spathae 1-2,

ad apicem ramorum sessiles. Cymae 2; inferiora

1 -2-flora; stipes c. 25 mm longus, setulosus; cyma

superioria c. 3-4-flora. Flores grandes, pallide

caerulei, lilacini vel albi. Petala c. 20 mm. Stamina

typica; anthera caerulea; staminodia antherodiis

bulbosis flavidis. Capsula oblongo-globosa, dura.

Type. — Transvaal, 2428, (Nylstroom): 20 km S of

Warmbaths on Great North Road (-CD), along

roadside, Smook 1494 (PRE, holo.).

Sturdy compact bushes (chamaephytes) c. 0,35 m

tall, setulose with short white setae; with a hard

gnarled rootcrown and long firm woody roots. Stems

erect, firm, with internodes c. 50 mm long. Leaves

linear, c. 40-50 x 10 mm, apex sharply acuminate-

recurved, base merging into a short open sheath,

light greyish-green, margin undulate, forming a thick

white rim. Spathes terminal, opposite upper leaves,

sessile, funnel-form, triangular, apex acute to acu-

minate, recurved, c. 25 mm long. Cymes 2; lower

cyme much exserted on a hairy peduncle c. 25 mm

long, 1 -2-flowered, male; upper cyme c.

3-4-flowered, bisexual. Flowers pale blue, lilac or

white, large. Petals c. 20 mm. Stamens 3, two with

normal blue locules, the central curled with a large

connective; filiaments white; staminodes with purple

filaments and orange-yellow antherodes consisting of

4 bulbous bodies and 2 smaller ones. Capsule oblong-

globose, hard, c. 10 mm, 3-seeded; seeds smooth.

(No well developed capsule seen.) Fig. 2.

Recorded from the Springbok Flats in the

Transvaal in turf soil. Flowering November-

January.

Transvaal. — 2428 (Nylstroom): 20 km S of Warmbaths on

Great North Road (-CB), Smook 1494; Warmbaths (-CD), Burtt

Davy 2232; Singlewood Halt (-DB), Mauve 4279. 2527 (Rusten-

burg); near Brits (-DB), Emmenis in PRE 38063. 2528 (Pretoria):

Rust de Winter (-BA), Pole Evans 3852.

VARIOUS AUTHORS

437

Fig. 2. — Commelina bella. Smook 1494, holotype in PRE.

Commelina modesta Oberm., sp. nov. C. liv-

ingstonii affinis, sed planta minora, laxa, glabra vel

glabrescens; spathis floriferis solitariis ad apicem

ramorum; laminaque folii basin versus sensim in

vaginam attenuata differt.

Herba perennis (chamaephyta), erecta, laxe

ramosa. Caulex compactus, radicibus longis lignosis.

Rami tenues, internodiis elongatis. Folia linearia, c.

80(-100) x4 (-10) mm, plana, glabra vel pauce

glabriuscula, lamina basin versus sensim in vaginam

attenuata. Spatha singularia, ad apicem ramorum,

sessilis, infundibuliformis, triangularis, apice acuta.

Cyma solitaria. Flores parvi, albi vel caerulei. Cap-

sula 3-locularis; loculi globosi, laevigati; semina

globosa, laevigata.

Type. — Transvaal, 2531 (Komatipoort): Barber-

ton, lower hill slopes (-CC), Galpin 808 (PRE, holo.).

Small spreading, diffusely branched, glabrescent

bushes (chamaephytes) up to c. 0,3 m tall. Rootstock

woody, gnarled, knobby (the knobs presenting re-

mains of swollen bases of annual stems); roots

woody, long, initially covered by a velamen of

roothairs. Stems several, erect, with long internodes

up to 60-100 mm long. 1 -2 mm in diam. Leaves with

lamina linear, flat, 8Q-100 x 4-10 mm, attenuated

at the base into a pseudo-petiole, white-punctulate;

sheath membranous, subauriculate. Flowering

spathe solitary (rarely 2), apical, sessile or nearly so,

funnel-form, shortly triangular, 15 mm long, 10 mm

broad, apex short, acute, minutely puberulous and

with scattered white setae. Cyme solitary. Flowers

small, blue or white (“pink” fide Galpin 808). Sepals

ovate, c. 5 mm, membranous, upper minute. Petals:

upper one rounded, c. 15 mm ventral ones minute.

Stamens typical, the anthers occasionally with dark

margins, the central semicircular; staminodes with

yellow bulbous antherodes. Capsule with globose,

shiny, cream-coloured locules; seeds globose, 5 mm,

smooth, farinaceous, dorsal seed occasionally

aborted.

Widespread in Transvaal, Natal and Swaziland, in

rocky habitats. Flowering November-March.

Transvaal. — 2230 (Messina): Sand River workings (-AC),

Wild 7624 (SRGH, PRE). 2330 Tzaneen: Mbayimbayi (-BB),

Lang in TRV 32230 (PRE). 2428 (Nylstroom): Nylsvley Ecosystem

Research Station (-DA), Smook 1491. 2527 (Rustenburg): Rusten-

burg (-CA), Van Dam in TRV 19714 (PRE). 2528 (Pretoria):

Pretoria, Schanskop (-CA), Leemann 51 (PRE). 2530 (Lyden-

burg): Lowveld Botanic Garden, Nelspruit (-BD), Buitendag 371

(NBG, PRE).

Natal. — 2632 (Bella Vista): Nkonjane-Aberkorn Drift (-CC),

Moll & Pooley 4197 (NH). 2731 (Louwsburg): Itala Nature

Reserve (-CB), McDonald 332 (PRE, NH). 2732 (Ubombo):

mountain pass near Josini Dam, Stirton 498 (PRE). 3030 (Port

Shepstone): Gibraltar (-CB), Strey 10343 (PRE, NH).

This species was usually placed under C. liv-

ingstonii C. B. Cl., but is a more slender bush found

in rocky habitats. The leaves narrow gradually below

into a pseudo-petiole, whereas in C. livingstonii they

widen below and then narrow abruptly into the

sheath.

A. A. Obermeyer

A NEW COMBINATION IN COMMELINA

Commelina diffusa Burm. f. subsp. scandens

( C.B . Cl.) Oberm., comb, et stat. nov.

Commelina scandens Welw. ex C. B. Cl. in A.

DC., Monogr. Phan. 3: 146 (1881); F.T.A. 8: 37

(1901). Type: Angola, Pungo Andongo, banks of

River Cuanza near Nbilla, Welwitsch 6642 (BM,

holo.).

This subspecies is common along watercourses in

northern Botswana and it also occurs in the warmer

parts of the Transvaal and Natal in wet habitats. It is

more robust than the typical variety and has longer

leaves.

A. A. Obermeyer

TWO NEW SPECIES OF CYANOTIS

Cyanotis pachyrrhiza Oberm., sp. nov., species

distincta.

Herba colonifera, repens, diffusa. Caudex com-

pactus. radicibus numerosis longis, robustis, crassis.

Folia 5—7, aggregata, linearia, c. 70-150 x 5-8 mm,

supra glabra nitida, subtus violacea, pilis erectis

albis. Spicae axillares, saepe pedunculatae,

multiflorae, compactae. Spathae cordatae, saepe

elongatae, spicas longe superantes. Bracteolae par-

vae. Flores et fructus typici.

438

NOTES ON AFRICAN PLANTS

Type. — Transvaal, 2430 (Pilgrims Rest); Ohrig-

stad Dam, on rocky hill slopes (-DC), Mauve &

Retief 5245 (PRE, holo.).

Perennials forming colonies, with a purple colour-

ing and soft white pubescence. Caudex compact,

small. Roots many, long, hard, tapered below, thick,

c. 3 mm in diam., white (black when dry). Leaf-

cluster with 5—7 erect, linear leaves, c. 70-150 x 5—8

mm, glabrous and shiny above, purple below and

with erect white hairs. Spikes axillary, on erect or

semi-decumbent annual stems; flowers many in a

compact cluster, often pedunculate, the subtending

spathe long and narrow, recurved, not enveloping

flowers, or in some short; bracteoles small, ovate.

Calyx with lobes fused basally, c. 7 mm, setose. Cor-

olla just exserted from calyx, maroon; lobes triangu-

lar. Stamens with white filaments, fusiform below

apex where the maroon, beaded, patent hairs are

situated; anthers yellow. Style fusiform below

stigma. Capsule subquadrate, sparsely setose above;

seeds oblong-globose, c. 1,5 mm, wrinkled. Fig. 3.

So far only recorded from the eastern Transvaal,

but with one record from the north-western Trans-

vaal, montane, on rock ledges (quartzite). Flowering

December.

Transvaal. — 2428 (Nylstroom): Geelhoutkop (-AD), Breyerin

TRV 17813. 2430 (Pilgrims Rest): Mariepskop (-DB), Van Dam in

TRV 26303; Lissabon Waterfall (-DD), De Feyter 65. 2530 (Ly-

denburg); top of Steenkampsberg (-AA), Codd 8052 (PRE); Dull-

stroom (-AA), Galpin 13126; Mount Anderson (-BA), Smuts &

Gillett 2434.

Cyanotis robusta Oberm., sp. nov., C. lapidosa

Phill. affinis, sed planta majora, robustiora, radici-

bus crassibus lignosis, lobis calycis c. 10 mm longis

differt.

Herba colonifera, robusta. Radices crassae,

longae, lignosae. Folia inequalia, 5-7, rosulata,

linearia, 100(-300)x 15-20 mm, supra glabra, subtus

lanata, pilis albis longissimis tenuis appressis. Spicae

axillares, multiflorae, compactae. Calycis lobi

lineari, c. 10 mm longi. Flores coerulei vel lilacini,

typici.

Type. — Natal, 2830 (Dundee): along road between

Mooi River and Weenen, c. 10 km past Middelrus

(-CC), on rock ledges, under bushes, Arnold 1372

(PRE, holo.).

Spreading rhizomatous perennials forming col-

onies, saxicolous. Roots many, long, c. 4 mm thick

(outer tissues apparently somewhat spongy).

Rhizome compact, small, covered by lanate remains

of leaf-bases. Leaves spreading, unequal in size,

linear, 80(-300)x 15-20 mm, smooth and shiny

above, lanate below with long thin white appressed

hairs. Flowering stems annual, basal, spreading, c.

200-300 mm long; cymes axillary, compact,

enveloped by long canaliculate spathes, c. 20-50

mm, becoming progressively smaller towards apex.

Flowers c. 5-10 in each axil, pale sky-blue to mauve.

Calyx with linear lobes c. 10 mm long, glabrescent to

lanate. Stamens etc. typical of genus. Fig. 4.

Fig. 3. — Cyanotis pachyrrhiza. a

whole plant, x 0,5; b, flower

with bract and bracteoles, x

1,5 c, stamen, x 3,5; gynoe-

cium, x 3,5. Mauve & Retief

5245.

VARIOUS AUTHORS

439

Fig. 4. — Cyanotis robusta. Arnold 1372, holotype in PRE.

Fig 5. — Ficinia x lucida C. B. Cl., Bolus 6082 (K, holotype), an

interspecific hybrid between Ficinia cedarbergensis and F.

ixioides subsp. glabra.

Transvaal.— 2430 (Pilgrims Rest): Wolkberg, NW face of

Serala Peak (-AA), Muller & Scheepers 165 (PRE).

Natal. — 2830 (Dundee): Umkhumba Mountain near Weenen

(-CC), West 1441 (NH); between Mooi Rjver and Weenen (-CC),

Arnold 1372. 2929 (Underberg); Joneskloof near Estcourt (-BB),

West 1538 (NH).

The long thick roots and more robust habit distin-

guish it from C. lapidosa. There appears to be varia-

tion in size and degree of pubescence.

A. A. Obermeyer

CYPERACEAE

FICINIA LUCIDA: AN INTERSPECIFIC HYBRID BETWEEN F. CEDARBERGENSIS AND F. IXIOIDES SUBSP. GLABRA

The type of F. lucida C. B. Cl. ( Bolus 6023, Fig. 5)

was collected in the Cedarberg mountains of the

western Cape Province in 1878, a little over one hun-

dred years ago. It most closely resembles F.

cedarbergensis Arnold & Gordon-Gray and F.

ixioides Nees subsp. glabra Arnold & Gordon-Gray,

possessing characters of both these taxa, yet it cannot

satisfactorily be placed in either of them.

F. lucida resembles F. cedarbergensis mainly in

vegetative form, and in the form of the style and in-

florescence axis. Its inflorescence bracts and glumes

and the overall form of the inflorescence are more

like those of F. ixioides. In F. lucida: (i) the style is

thick, red and granular (Fig. 6b) like that of F.

cedarbergensis (Fig. 6a) and unlike that exhibited by

F. ixioides, which is long, slender, light brown and

papillate (Fig. 6c); (ii) the 2 or 3 outermost in-

florescence bracts have papery lobes along their

margins, a condition not found elsewhere in Ficinia,

except in F. ixioides, and unlike that in F.

cedarbergensis (see Figs 7a-c); (iii) the glumes have

obtuse, apiculate apices — a feature absent in F.

cedarbergensis, yet closely resembling the situation

found in F. ixioides (see Figs 8a-c); (iv) the sheathing

bases of the lower 2-3 inflorescence bracts enclose

the head more or less completely, as in F. ixioides,

and unlike F. cedarbergensis in which only !/$ of the

head is covered (see Figs 9a-c); and (v) the in-

florescence axes, although somewhat intermediate in

form, are more typical of the type characterizing F.

cedarbergensis (see Figs lOa-c).

Two of the characters that F. lucida shares with its

putative parents, are rare in the genus Ficinia. The

species (mentioned below), which do share these

features, all have distributions disjunct from F.

lucida and its putative parents: (i) the only other

species to possess a red, granular style is F. levynsiae

Arnold & Gordon-Gray; (ii) the only other species to

possess the same form of inflorescence axis is F.

petrophylla Arnold & Gordon-Gray.

440

NOTES ON AFRICAN PLANTS

F. cedarbergensis

F. lucida

F. ixioides

Styles

Lower most

Inflorescence bracts

Glumes

Inflorescence heads

Inflorescence axes

Figs 6 - 10. — Comparative morphology of styles ( x 8), inflorescence bracts ( x 2), glumes ( x 6,5), inflorescence heads ( x 2),

inflorescence axes ( x 4) between Ficinia cedarbergensis, F. ixioides subsp. glabra and F. lucida.

VARIOUS AUTHORS

44!

It is here postulated that F. lucida C. B. CL,

known only from a single specimen ( Bolus 6023, Fig.

5), is of hybrid origin. This theory cannot be proved

at present, since no breeding, chromosome or

pollination studies have been carried out in the

genus. The evidence for a hybrid origin of F. lucida

rests on its intermediacy between two distinct species

that are very different morphologically; the lack of

other specimens that even remotely resemble it; the

likelihood of cross-pollination (outbreeding) in the

group, as indicated by protandry and probable wind

pollination; and the occurrence of the putative parent

species in the type locality (see Fig. 11).

The idea of interspecific hybridation within Ficinia

is not new. Levyns (1950) in her treatment of the

genus for the Flora of the Cape Peninsula stated that

hybrids between certain species appeared to exist.

Hybridization is certainly suspected to be responsible

for much of the variation within the section

Bracteosae of Ficinia. It appears not only to have

brought about the blurring of specific boundaries in

the F. indica- complex but has possibly been an im-

portant factor in the origin of some species such as F.

grandiflora Arnold & Gordon-Gray.

The status of F. lucida C.B.C1. is therefore altered

as follows:

Ficinia x lucida C. B. Cl. in Dur. & Schinz, Con-

sp. FI. Afr. 5: 640 (1895). Type: South Africa, Cape,

Cedarberg mountains, Clanwilliam Div., Bolus 6023

(k, holo!).

T. H. Arnold

ECOLOGICAL ADAPTATIONS AND POSSIBLE CONVERGENCE IN FICINIA ARENICOLA VAR.

ERECT A AND MARISCUS DURUS

An obvious, although superficial similarity exists

between two fairly remotely related members of the

family Cyperaceae, namely Ficinia arenicola Arnold

& Gordon-Gray var. erecta Arnold & Gordon-Gray

and Mariscus durus (Kunth) C. B. Cl. Morphologi-

cal, anatomical and ecological comparison has

revealed additional shared characters which may be

interpreted as an example of convergence in these

two taxa.

The geographical distributions of these taxa (Fig.

12. la & 2a) are almost identical: extending along the

southern Cape coast from Mossel Bay to just east of

Port Elizabeth (22°E-27°E) i.e. in an intermediate

zone between the summer and winter rainfall regions.

Both taxa are normally restricted to a 5—10 km wide

belt of grassland, close to the sea, but extend slightly

further inland (up to 30 km) between 26°E and 27°E.

In M. durus the inflorescence takes the form of a

simple, more or less contracted umbel (Fig. 12. lb)

and in F. arenicola var. erecta (Fig. 12. 2b) it is a

compact head-like spike. In both taxa the in-

florescence is terminally situated at the apex of the

inflorescence stem, but appears to be lateral or sub-

lateral due to the presence of the lowermost in-

florescence bract which closely resembles and out-

wardly appears to be a continuation of the culm (par-

ticularly in M. durus). This condition is rare in both

genera.

The resemblance in habit (Fig. 12. lb & 2b) bet-

ween these taxa is enhanced by the similarity in plant

size. The culms of M. durus average 260 mm in

height, whereas in F. arenicola var. erecta the

average height is 240 mm.

In both taxa the stem bases are woody and are

covered by old persistent leaf bases. In M. durus they

are horizontally or obliquely displaced and are

notably thickened, whereas in F. arenicola var. erecta

they are erect and slender. Both taxa have slender

horizontal, wiry subterranean stolons, 2 or 3 mm in

diameter. These extends to a length of 2-12 cm and

are clothed in light brown lanceolate, papery scales.

Each stolon originates laterally from the base of the

stem and forms a single new plant at its extremity.

The leaves of both taxa generally exceed the culms

in height. They are olive-green in colour, glabrous,

erect and rigid, lack a midrib, are subcylindrical to

reniform in cross section (sometimes with the apex

dorsiventrally flattened) and superficially closely

resemble the culms. In Ficinia this leaf type is not un-

common, also being found in F. ixioides Nees subsp.

glabra Arnold & Gordon-Gray, F. pinquiorC. B. Cl.

and F. deusta (Berg.) Levyns. Among the South

African species of Mariscus, however, M. durus is

the only species without expanded leaf blades. This

leaf shape, resulting from the permanent infolding of

the leaf margins, is usually regarded as a direct adap-

tation to water stress, with the degree of infolding of

the leaf margins having been correlated with dry en-

vironmental conditions as it effectively reduces the

leaf area from which transpiration can take place

(Metcalfe, 1970).

The taxa grow in almost identical habitats and

have on occasion been seen to grow side by side.

They are confined almost exclusively to the margins

of seasonally inundated depressions. The soil of these

depressions is usually a heavy, dark clay which binds

residual water and reduces its availability during dry

periods (Russell, 1950; Hillel, 1971). Protective adap-

tation against desiccation may also be necessary as

these plants often grow close to the sea where the soil

and air have a high saline content. According to Dyer

442

NOTES ON AFRICAN PLANTS

Sclerenchyma

Chlorenchyma

Vascular bundles

Air canals

Ground tissue

I 0,75mm |

24° 25°

26° 27°

28°

2b

Fig 12. — Comparison of distribution, morphology and leaf and culm anatomy, la-d, Mariscus durus; 2a -d, Ficinia

arenicola var . erecta. la & 2a, sympatric distribution, lb & 2b, habit (lb, Forster 2332, x 0,25; 2b, Arnold 603, x 0,5) il-

lustrating erect growth form with leaves over-topping the inflorescence and well-developed subterranean stolons. I c & 2c,

transverse sections of leaf blade (lc, Forster 2358; 2c Arnold 603) showing large air canals, reduced abaxial surface and

similar terete outline. Id & 2d, transverse sections of culm (Id, Vorster 2342; 2d, Arnold 603) showing circular distribu-

tion of air canals.

VARIOUS AUTHORS

443

(1937) salt vleis and saline marshes near the coast,

without direct communication with the sea, are

characterized by a rich cyperaceous flora, including

M. durus.

A study of the cross-sectional leaf anatomy of

these taxa (Fig. 12. lc & 2c) shows that both have

well-developed chlorenchymatous tissue limited to a

band directly adjacent to the abaxial epidermis. The

vascular bundles in both taxa are typically collateral,

extending in an arc through the centre of the leaf bet-

ween the chlorenchyma and ground tissue (hypoder-

mis), and decreasing in size towards the infolded,

somewhat obscure, leaf margins.

An arc of large, well-developed lacunae or air

canals is a typical feature of the anatomy of both

taxa. In M. durus these are surrounded by chloren-

chyma and are located between radial rows of 1-4

vascular bundles which increase progressively in size

towards the ground tissue. In F. arenicola var. erecta

the air canals are irregularly dispersed between the

vascular bundles without a regular pattern of ar-

rangement.

In cross-section the culms of M. durus (Fig. 12. Id)

are circular, and in F. arenicola var. erecta (Fig.

12. 2d) they are slightly elliptical in shape. In both

taxa a band of chlorenchyma occurs immediately ad-

jacent to the epidermis. In M. durus the air canals are

smaller and more regular in shape and are located in

a circle along the margin between the chlorenchyma

and ground tissue alternating with the outer vascular

bundles. In F. arenicola var. erecta the irregularly

shaped air canals are located between the smaller

outer vascular bundles and the larger inner bundles

and ground tissue perimeter.

The presence of large well-developed air canals in

the leaves and culms of these taxa is seen as a signifi-

cant and necessary adaptation to water-logging of the

soil and periodic partial submergence of the plants

during the wet season. The air canals are essential for

aeration of the subterranean parts of the plant during

flooding (Sifton, 1957; Esau, 1965; Fahn, 1969). In

contrast, in F. arenicola var. arenicola which grows

in well-drained, well-aerated sandy soil (often in

close proximity to var. erecta), the air canals if pre-

sent at all, are poorly developed and very small, be-

ing less than ‘/4 the size of the largest vascular

bundles. A detailed account of the anatomical dif-

ferences exhibited by the two varieties of F. arenicola

will be presented in a later paper. M. durus lacks air

canals in the leafy extensions to the culms, represen-

ting the lowermost inflorescence bract, notwithstan-

ding the presence of well-developed canals in the

culms. No information on this aspect is available for

F. arenicola var. erecta.

The morphological and anatomical similarities ex-

hibited by these two taxa could be interpreted as a

possible example of convergent evolution resulting

from their apparent adaptation of the same ecologi-

cal niche, where they cope with extremes in water

availability at various times of the year. On the one

hand the need exists for water conservation during

dry periods, while conversely, good aeration becomes

critical during excessively wet periods.

REFERENCES

Dyer, R.A., 1937. The vegetation of the divisions Albany and

Bathurst. Mem. bot. Surv. S. Afr. 17: 60.

Esau, K., 1965. Plant anatomy. 767 pp., New York: John Wiley.

Fahn, A., 1969. Plant anatomy. 534 pp., London: Pergamon

Press.

Hillel, D., 1971. Soil and water, physical principles and pro-

cesses. 288 pp., London: Academic Press.

Metcalfe, C. R., 1971. Anatomy of the monocotyledons. 5,

Cyperaceae. 597 pp., London: Oxford University Press.

Russell, E. W., 1950. Soil conditions and plant growth. 635 pp.,

London: Longmans, Green.

Sifton, H. B., 1957. Air-space tissue in plants. Bot. Rev. 23:

300-312.

T. H. Arnold & P. J. Vorster

THE CORRECT AUTHOR CITATION FOR MARISCUS MACROPUS

The combination Mariscus macropus is usually at-

tributed to C. B. Clarke with Bockeler as author of

the basionym Cyperus macropus Bock. (1870). But

Cyperus macropus (Bock, is illegitimate, being a later

homonym of C. macropus Miq. (1860). However, if,

following Article 72 of ICBN (1978), the epithet

macropus on transfer to Mariscus is treated as new,

dating from 1895, the resulting combination is

treated as new, and is attributed solely to C. B.

Clarke, thus: Mariscus macropus C. B. Clarke.

P. J. Vorster

THE CORRECT AUTHOR CITATION FOR MARISCUS DUBIUS

Most recent authors, e.g. Kiikenthal in Pflanzenr.

101, 4, 20: 563 (1936); Podlech in Mitt. bot.

Stsamml., Munch. 3: 524 (1960); in Prodr. FI. S. W.

Afr. 165: 37 (1967); Napper in J1 E. Africa nat. Hist.

Soc. 28: 17 (1971); Gordon-Gray in Ross, Flora of

Natal 106 (1972) attribute the combination Mariscus

dubius to Hutchinson in FI. W. Trop. Afr. ed. 1,2:

485 (1936) with Rottbfll as author of the basionym.

Hutchinson’s combination is, however, antedated by

that which appeared in Gamble, FI. Madras 1644

(1931). In FI. Madras the treatment of the

Cyperaceae is attributed to G. E. C. Fischer, but in

the text the combination M. dubius is ascribed to

Kiikenthal. Five years later in Pflanzenr. l.c.,

Kiikenthal attributed the combination to Hutchin-

son, presumably unaware of Fischer’s earlier action.

Following Rec. 46C of the ICBN (1978) the combina-

tion should be written thus: M. dubius (Rottb.)

Kiikenth. ex G. E. C. Fischer.

P. J. Vorster

ON THE IDENTITY AND GEOGRAPHICAL DISTRIBUTION OF MARISCUS ANGULARIS, M. CHERSINUS,

CYPERUS BULLATUS AND C. CAPENSIS VAR. POLYANTHEMUS

In the course of a revision of the taxonomy of the

South African species of Mariscus, the following

type specimens were examined and found to repre-

sent one and the same species:

1. Lugard 142 (GRA!); Harbor sub Rogers 6310

(BOL!; J!): syntypes of Mariscus chersinus N. E. Br.

in Kew Bull. '1921: 300 (1921).

444

NOTES ON AFRICAN PLANTS

2. Dinter 2354 (B!): type of Cyperus bullatus

Kiikenth. in Reprium nov. Spec. Regni veg. 29: 198

(1931).

3. Schlechter 11683 (B!; G ! ; Z!): type of Cyperus

capensis (Steud.) Endl. var. polyant hemus Kiikenth.

in Pflanzenr. 101, 4, 20: 540 (1936).

The earliest name is Mariscus chersinus N. E. Br

(1921) and therefore it is the correct name to use for

this taxon. The synonymy is as follows:

Mariscus chersinus N. E. Br.

Cvperus bullatus Kiikenth. in Reprium nov. Spec. Regni veg.

29: 198 (1931).

C. chersinus (N. E. Br.) Kiikenth. in Pflanzenr. l.c. p. 525

(1936).

C. capensis (Steud). Endl. var. polyanthemus Kiikenth. l.c. p.

540.

Mariscus bullatus (Kiikenth.) Podlech in Mitt. bot. Stsamml.,

Munch. 3: 523 (1960).

It is surprising that Kukenthal maintained C. cher-

sinus, C. bullatus and C. capensis var. polyanthemus

as three distinct taxa. His recognition of spurious dif-

Fig 13. — Mariscus chersinus. Portion of a mature umbel of

Vorster 2637, from Sihangwane in northern Natal. Scale =

50 mm.

Fig. 14. — Mariscus chersinus. Known geographical distribution.

ferences was clearly influenced by the apparently dis-

junct geographical distribution of these taxa. Kuken-

thal placed M. chersinus (as Cyperus chersinus) in

Cyperus sect . Umbellati, whereas C. bullatus and C.

capensis var. polyanthemus were placed (correctly) in

sect. Bulbocaules. The type collection of C. capensis

var. polyanthemus is immature, so that it is not sur-

prising that he grouped it under C. capensis rather

than under C. bullatus. However, he distinguished C.

bullatus from C. capensis on the grounds of subterete

versus trigonous mature spikelets, but this distinction

breaks down as the type of C. bullatus has distinctly

trigonous spikelets. In fact, this species shows

remarkably little morphological variation through its

wide distribution range.

Recent collections in northern Transvaal have fill-

ed in gaps in the known distribution range, and it is

now clear that M. chersinus is distributed from the

Caprivi area, throughout central Botswana, southern

Zimbabwe, northern Transvaal, southern Mozambi-

que to northern Natal (Fig. 13), practically reaching

the Indian Ocean coast (Fig. 14). Up to now it has

only been found in tall, open savanna on deep, loose

sand.

The type specimen of M. angularis Turrill

[Schonland 3848 (GRA, holo. ! )] is a typical example

of M. macrocarpus Kunth with several flowers to a

spikelet. It falls in the subgenus Umbellati on ac-

count of its spherical, woody base. It is difficult to

understand how Kukenthal could associate this

specimen with M. chersinus (as Cyperus chersinus

var. angularis), as that species has 1 -flowered

spikelets and a lanceolate pseudo-bulb, and was at

that stage known only from hundreds of kilometres

away in Botswana.

P. J. Vorster

THE IDENTITY AND TYPIF1CATION OF MARISCUS DREGEANUS

Mariscus dregeanus Kunth, Enum. 2: 120 (1837)

has for many years been regarded as a synonym of

M. dubius (Rottb.) Kiikenth., e.g. by C. B. Clarke in

FI. Brit. Ind. 6: 620(1893); in FI. Cap. 7: 187 (1897);

in FI. Trop. Afr. 8: 380 (1902); Kukenthal in

Pflanzenr. 101, 4, 20: 563 (1936); Hutchinson in FI.

W. Trop. Afr. ed. 1,2: 485 (1936); Chermezon in FI.

Madag. 29: 24 (1937); Hooper in FI. W. Trop. Afr.

ed. 2,3: 295 (1972); Gordon-Gray in Ross, Flora of

Natal 106 (1972) and Kern in Van Steenis, FI. Males.

ser. 1,7: 643 (1974). Examination of the original

description and type material has, however, revealed

that M. dregeanus differs from M. dubius. The

misconception could have arisen in two ways:

1 . The two species have a superficial resemblance to

each other in the dried condition, and could

mistakenly have been considered to be mere variants

of the same variable species. It is known that material

of the two species is mixed in numerous herbaria.

VARIOUS AUTHORS

445

2. The wrong specimens could have been interpreted

to be the type collection of M. dregeanus. It is con-

ceivable that any Drege specimen from eastern South

Africa conforming more or less to the description of

M. dubius could be interpreted as being the material

Kunth referred to as ‘Africae australis ora orientalis’,

Drege s.n. I have seen two such specimens: one of the

two sheets of Drege 4383 in B contains two specimens

from ‘Basche’, of which the two right-hand in-

florescences represent M. dubius (the two left-hand

inflorescences are M. dregeanus). A second sheet of

M. dubius in B is annotated ‘no. 4383 intermixt ....

Ora orientalis Africae australis . . . Draege’. Unfor-

tunately I have not had the opportunity to search

through the collections at K, L and P for specimens

which could have influenced the various authors

cited above to arrive at their conclusions, but both

sheets of Drege 4383 in B as well as the sheet of M.

dubius (‘no. 4383 intermixt’) bear labels written up

by Clarke as M. dregeanus. It would seem, therefore,

as if Clarke at that time (1893) was unable to

distinguish between M. dregeanus and M. dubius.

Fig. 15. — Mariscus dregeanus. Lectotype, Drege 4383, in B, with

enlargement of the inflorescence. Scale adjacent inflorescence

= 20 mm.

The original description of M. dregeanus contains

the phrase ‘spicis compositis subquinis, fasciculato-

congestis, sessilibus’, which clearly characterizes M.

dregeanus. The specimen of Drege 4383 in B which is

accepted here as lectotype (Fig. 15), has bee chosen

for the following reasons:

1. It has the multifid spike so typical of this species,

and which is mentioned in the original description.

2. The sheet bears an old inscription ‘Mariscus

Dregeanus Kth.’ and ‘Ora orientalis Africae australis

. . . ’ which is almost identical to the locality note ac-

companying the original description.

Fig. 16. — Inflorescence of Mariscus dregeanus (left, from Vorster

2526 from near Warmbad, Transvaal) and M. dubius (right,

from Vorster 2492 from St Lucia Bay). Scale bars = 10 mm.

SPIKELET LENGTH (mm)

Fig. 17. — Scatter diagram comparing spikelet dimensions in

Mariscus dregeanus (A) and M. dubius (%).

TABLE 1. — Comparison of some characters of Mariscus

dregeanus and M. dubius

446

NOTES ON AFRICAN PLANTS

3. The sheet is labelled ‘Hb. Kunth’, and is thus like-

ly to have been used by Kunth when he described M.

dregeanus.

Inflorescences of both species dry to a straw-

colour, but the two species can readily be distinguish-

ed from each by the characters given in Table 1 (see

also Figs 16 & 17).

P. J. VORSTER

THE CORRECT AUTHOR CITATION FOR CYPERUS ELEPHANTINUS

Both Kiikenthal in Pflanzenr. 101, 4, 20: 51 (1936)

and Gordon-Gray in Ross, Flora of Natal 103 (1972)

used the name Cy perns elephantinus C. B. Cl. (1895).

Flowever C. elephantinus C. B. Cl. is a nomen

nudum and therefore illegitimate. The first validly

published name given to the species was Mariscus

elephantinus C. B. Cl. in Flora Capensis 7: 195

(1897), where a full description was provided. In

1909 C. B. Clarke in lllustr. Cyper. t. 20, figs. 1 & 2

again used the name Cyperus elephantinus, but this

cannot be accepted as valid publication of the name

of a new combination under Cyperus, since there is

no description or a reference to the basionym,

Mariscus elephantinus. It was not until 1936 that the

combination was validly published: Kiikenthal (l.c.)

published the combination Cyperus elephantinus,

providing a full Latin description and citing Mariscus

elephantinus. Although Kiikenthal did not

deliberately make the combination Cyperus elephan-

tinus, it was nevertheless effected and should be ac-

cepted. The correct author citation is therefore

Cyperus elephantinus (C. B. Cl.) Kiikenth.

P. J. VORSTER

ERICACEAE

TWO NEW SPECIES OF ERICOIDEAE

Erica insignis£’. G. H. Oliver, sp. nov. in genere

singularis et facile recognita combination trium

characterum: corolla redactissima aspectu aliquot

coronae parvae circum ovarium, segmentis calycis

maximis et staminibus bene exsertis 6-8plo corolla

longioribus; ab Erica nabea Guthrie & Bol., specie

altera generis exhibente corollam redactam,

staminibus bene exsertis, floribus multo majoribus et

habitu statim dignoscenda.

Fruticulus ad 400 mm altus, caudice lignoso. Rami

erecti pubescentes. Folia 3nata erecta imbricata

parum falcata, 8-12 mm longa, linearia ad angustis-

sime ovata, acuta ad obtusa, parum alata ad basim,

sulcata carinataque, statu iuvenili omnino puberula,

et ciliata glandulis mini^tissimus subsessilibus, minus

pubescentia vetustate; petiolo 1,5-2 mm longo,

pubescente glandulo-ciliatoque. Flores 2-3nati ad

extrema ramorum et/vel l-3nati ad extrema

ramulorum lateralium; pedicellis 1,3-3 mm longis,

dense pubescentibus; bracteolis 3 approximatis

subaequalibus aut aequalibus, 8,5-9 x 3-4 mm,

duabus lateralibus asymmetricis, omnibus ovato-

attenuatis acutis, sulcatis carinatisque in parte 2A

superiore, sparse et minutissime puberulis, densiore

puberulis ad basim, sparse ciliatis glandulis minutis

subsessilibus. Calyx 4partitus bicyclicus, segmentis

exterioribus ab- et adaxialibus, 18-24x6-7 mm,

segmentis interioribus lateralibus 16-20 x 5 mm, om-

nibus anguste ovato-acutis, carinatis sulcatisque in

parte 14 superiore, naviculatis, sparse et minutissime

puberulis utrinque, sparse ciliatis glandulis minutis

subsessilibus. Corolla 41obata 4-5plo calyce breviora

4,5 x 4,5 mm oblato-urceolata, 4saccata et profunde

canaliculata sub interstitiis, glabra; lobis filamenta

amplectentibus, emarginatis, breviter ciliatis.

Stamina 8 libera exserta; filamentis elongato-

linearibus, 25-32x0,5 mm, glabris; antheris 2,7-3

mm longis, oblongis obtusis, parum prognathis

rotundisque vel parum bilobatis ad basim, muticis

glabris; poro 1,8 mm longo; polline in tetradis.

Ovarium 41oculare, ovulis multis, breviter cylin-

draceum, 2,5-3 x 2-2,5 mm, manifestum ad inter-

stitia corollae, 8sulcatum glabrum, crista conspicua

similis coronae apice immerso; stylo 30-37 mm

longo tereti, longe exserto glabro; stigmate

capitellato. Fructus ad 5 mm longus, capsula

loculicidalis, valvis quattuor cucullatis.

Type. — Cape, Prince Albert District, northern

slopes of Kangoberg in the Great Swartberg Range,

1 400 m, December 1979, Oliver 7469 (STE, holo.;

K; NBG; PRE).

Dwarf woody shrub up to 40 cm often small and

bonsai-like with a woody rootstock. Branches erect,

pubescent with spreading white hairs. Leaves 3-nate,

erect imbricate slightly falcate 8-12 mm long, linear

to very narrowly ovate, with acute to obtuse red apex

and slight wings towards the base, sulcate and cari-

nate, puberulous all over when young and ciliate with

minute subsessile glands becoming less so with age;

petiole 1,5-2 mm long, pubescent and gland-ciliate.

Flowers 2-3-nate at the ends of branches and/or

1 -3-nate at the ends of short lateral branchlets

sometimes becoming crowded towards the ends of

branches; pedicels 1,3-3 mm long, densely pubes-

cent; bracteoles 3 approximate, subequal to equal

8,5-9 x 3-4 mm with the two laterals asymmetrical,

all ovate-attenuate to broadly ovate and more

markedly attenuate, acute, sulcate and carinate in the

upper 2A, sparsely and very minutely puberulous,

denser towards the base, sparsely edged with minute

subsessile glands, pale green turning bright red.

Calyx 4-partite, 2-ranked the outer two ad- and abax-

ial, 18-24x 6-7 mm, the inner two lateral 16-20X 5

mm, all narrowly ovate-acute, carinate and sulcate in

the upper 'A, boat-shaped, sparsely and very minute-

ly puberulous on both surfaces, sparsely edged with

minute subsessile glands, pale green turning bright

red. Corolla 4-lobed 4-5 x shorter than the calyx

and only just exceeding the ovary, 4,5 x 4,5 mm

broadly oblate urceolate, 4-pouched and deeply

channelled below the interstices, glabrous, pale

green; lobes clasping the relatively large filaments,

emarginate, shortly ciliate-edged. Stamens 8, free,

exserted; filaments elongate linear 25-32x0,5 mm,

glabrous, pale green; anthers 2, 7-3,0 mm long,

oblong obtuse, slightly prognathous and rounded or

slightly bilobed at the base, muticous, glabrous, pale

brown; pore 1,8 mm; pollen in tetrads. Ovary

4-celled with numerous ovules, 2,5-3 x 2-2,5 mm

shortly cylindric just visible above the corolla in-

terstices, 8-grooved, with conspicuous corona-like

ridge with 8 bumps, apex and stigma attachment

sunken, glabrous, pale green; style 30-37 mm terete,

far exserted, glabrous, pale green; stigma capitellate,

Fig. 18. — Erica insignis. 1, flower, abaxial view; 2, flower, lat. view; both x 3; 3, medium bracteole, outer and side views;

4, lateral bracteole; 5, abaxial sepal; 6, corolla; 7, anther, rear, front and side views; 8, ovary; all x 6; 9, young fruit in

longitudinal section; 10, mature fruit seen from above; 1 1 , base of leaf, x 12. All drawn from the holotype, Oliver 7469

(STE).

black. Fruit up to 5 mm long, loculicidal capsule with

4 hooded valves. Fig. 18.

Cape. — 3321 (Ladismith): Swartberg, upper north slopes of

Kangoberg (-BD), Taylor 1102 (BOL); Oliver 7469 (K; NBG;

PRE; STE). 3322 (Oudtshoorn); Swartberg Pass, lower northern

slopes (-AC), Bond 1676 (PRE): Oliver 7463 (NBG; STE); Swart-

berg Pass Mountains, upper north slopes without precise locality,

Stokoe 6408 (BM; BOL).

This remarkable species is very distinct in the genus

for its very reduced corolla which appears almost like

a corona around the ovary, its extremely large calyx

segments and for its far exserted stamens four to five

times the length of the corolla.

The extremely reduced corolla is not unique in the

genus, as it is shared with only one other species.

Erica nabea Guth. & Bol., which is a species of

uncertain position in the Ericoideae. It was originally

placed in a monotypic genus as Macnabia montana

by Lehman on account of the floral arrangement, but

was included under Erica by Guthrie and Bolus

(1905) in Flora Capensis. Their view may have to be

reversed when my investigations of the generic

delimitation within the Ericoideae, at present in pro-

gress, are finally completed. For the time being, it is

best to accept that Erica nabea belongs to the genus

Erica in the section Adelopetalum and therefore the

new species, Erica insignis, as well.

E. insignis, although possessing a basic floral pat-

tern like E. nabea, can easily be separated from it. E.

nabea occurs on the moist coastal flats and southern

slopes of the George-Knysna-Uitenhage Districts and

is an erect single stemmed shrub up to 1,5 m high

with virgate branches. It has smaller whitish green

flowers (sepals 14-16 mm long), which are usually

arranged in dense pseudoracemes. The stamens are

included, the corolla is subovoid, the ovary oblong

obtuse and the seeds are flat and broadly winged. E.

insignis on the other hand occurs in very dry rocky

habitats on the northern slopes of the Great Swart-

berg and is often a gnarled bonsai-like shrublet up to

0,4 m high. It has larger bright red flowers (sepals

1 8—24 mm long), which are aggregated towards the

ends of branches, but never in dense long

pseudoracemes. The stamens in E. insignis are well

exserted, the corolla is broadly oblate-urceolate and

4-channelled, the ovary is shortly cylindrical with a

sunken apex and the seeds are not winged.

Despite the above clear differences between the

two species there is an overall similarity, which would

448

NOTES ON AFRICAN PLANTS

suggest some closer relationship somewhere during

their evolution. But their widely separated distribu-

tion ranges and habitat preferences will remain dif-

ficult to explain.

E. insignis has for many years remained an elusive

subject for me. It was discovered as early as 1935 by

that veteran mountaineer and collector, T. P.

Stokoe. A postcard written by him to N. S. Pillans

and attached to his collection in the Bolus Herbarium

is worth quoting ‘My Swartberg trip produced

nothing exciting, a few phylica and Erica of the 7

weeks Poort Mountain Type. I found an Erica,

Petiveri like in Type but I do not think it is it. Rock

Erica, faces the blazing north sun pale greeny yellow

when young but vivid scarlet as it reaches fruiting

stage alt. 5 000 ft. flowers are not pendent but at

right angles to stem’. Little did he realize what pro-

blems this species would create. Then in 1954 H. C.

Taylor recollected the species at 5 000 ft on a high

altitude traverse of a number of peaks.

Since the early 1960s, I have visited the Great

Swartberg in the vicinity of the pass about ten times,

twice with Taylor, and have covered most of the nor-

thern slopes on high altitude traverses from

Gamkaskloof in the west to Tierberg in the east

without finding the species.

Just before leaving in 1979 on another attempt to

find it, I received a recent collection of Ericaceae

from W. Bond of Saasveld and among them was the

species collected not at 5 000 ft, but at 3 000 ft at the

base of the Swartberg. En route to the upper slopes

of the Kangoberg, I recollected the species at Bond’s

locality at the foot of the pass. This immediately gave

the clue as to its habitat preferences — bare open

rocks facing north to west at the extreme lower limit

of the fynbos and Table Mountain Sandstone (TMS).

Here, it was growing with such unlikely species as

Portulacaria afra, Cotyledon orbiculata and

Passerina sp.

The visit to Kangoberg revealed a similar situation.

The species was found not at high altitude but on the

steep very rocky lower slopes at the lowest limit of

the fynbos and TMS formation just above the

renosterveld and uppermost karroid elements on the

shales. Again it was growing in rock cervices with

very little or apparently no soil facing north to west

and associated with dry elements of fynbos vegeta-

tion. The species was definitely not a member of the

high montane fynbos vegetation.

The shrublets were very much ‘bonsaied’ with the

rather few branches arising from a gnarled woody

rootstock which was pressed into the rock cervices or

spread over the rock surface. Occasionally plants

grew on flat rock surfaces where soil could ac-

cumulate and were more robust and floriferous. It is

remarkable that the plants were able to survive in this

environment facing the blazing hot sun for much of

the day. The flowers when young are pale green, but

soon turn bright red when fully exposed to the sun

therefore adding to the striking appearance of the

plants.

There appears to be little variation of significance

in the material available. The hairiness is extremely

fine and sometimes sparse and may easily be over-

looked. The fruit, like all ericoid fruits, is loculicidal.

It exhibits xerochastic movements connected with

seed dispersal. The hooded valves open in a dry

atmosphere and close on being moistened and

therefore the seeds will only be released in dry sunny

weather when they can easily be dispersed by the

wind.

Scyphogyne calcicola E. G. H. Oliver, sp. nov.

in genere distincta et sejuncta propter locum et

ecologiam sed quasi accedens S. divaricatam Benth.,

speciem montium Caledonensium et Worcesteren-

sium a qua absentia glandularum, floribus ter-

minaliter aggregatis subglabris et ovario glabro dif-

fert.

Fruticulus compactus erectus ad 300 mm altus

valide lignosus. Rami fastigiati virgati foliis solum

versus apices instructi, minute pubescentes, trigoni,

vetiores notabile articulati. Folia 3nata erecta incurva

subimbricata, 1,4-2 mm longa, elliptica ad ovata,

acuta ad subacuminata, complanata pubescentia ad

basim adaxialem, convexa sulcata glabraque in facie

abaxiali, marginibus minute serratis hyalinis. Flores

(3) 6 (8)nati congesti versus apices ramorum vel

2-3nati in ramulis brevibus lateralibus inferioribus,

apparente sessiles. Calyx 41obatus inaequalis, in-

terstitia corollae contingens; lobis latissime ovatis ob-

tusis, omnibus subcucullatis in parte superiore,

glabris, infra ciliatis supra denticulatis. Corolla

41obata, 1,5-1, 7 mm longa obconica ad cyathiformis

glabra; lobis latissimis 0,5 mm longis protuberan-

tibus inter lobos calycis, inaequaliter crenatis.

Stamina 4, connata usque ad medias antheras; tubo

filamentorum ± 0,4 mm longo glabro; antheris 1

mm longis ellipticis obtusis glabris muticis inclusis

aut manifestis; poro fere usque ad mediam thecam;

polline in tetradis. Ovarium 1-loculare, 1-ovulatum,

glabrum 0,5 mm longum ellipsoideum; stylo 0,9 mm

longo glabro, 4porcato; stigmate 1,3 mm lato, late

infundibuliformi, exserto, relative giganteo et con-

spicuissimo glabro rubro. Fructus incognitus.

Type. — Cape, Bredasdorp District, limestone hills

near Heidehof (Awila) just north of Pearly Beach,

Feb. 1980. Oliver 7604 (STE, holo.; BM; BOL; K;

MO; NBG; PRE; S).

Shrublet, compact erect up to 30 cm high, woody.

Branches numerous fastigiate virgate with leaves only

towards apices, minutely pubescent with spreading

hairs, trigonous, the older noticeably jointed with

conspicuous leaf scars and grey remnants of inter-

nodal ridges. Leaves 3-nate, erect incurved subim-

bricate 1 ,4-2 mm long elliptic to ovate to broadly so,

acute or subacuminate, flat and pubescent at base

adaxially convex, sulcate and glabrous abaxially,

with minutely serrate hyaline margins; petiole

0,5-0, 6 mm long, adpressed, shortly ciliate,

puberulous adaxially, glabrous abaxially. Flowers (3)

6 (8)-nate crowded at ends of branchlets or lower

down 2-3-nate on short lateral shoots, apparently

sessile. Calyx 4-lobed to halfway, unequal, reaching

to corolla interstices, the larger one 1 ,2-1,5 mm long

broadly deltoid acute to subacuminate, the smaller

ones 1-1,4 mm long very broadly ovate obtuse, all

subcucullate in upper half, glabrous, ciliate towards

the lower edges, denticulate in upper part, carinate

above, green. Corolla 4-lobed, 1,5-1, 7 mm long ob-

conical to cyathiform, glabrous, green with lobes

tinged red; lobes very broad 0,5 mm long, bulging

out between calyx lobes, unevenly crenate. Stamens A

connate to halfway up the anthers; filament tube ±

0,4 mm long glabrous; anthers 1 mm long elliptic ob-

tuse, glabrous, muticous, included to manifest; pore

almost Vi length of cell. Ovary 1 -celled with a single

pendulous ovule, glabrous 0,5 mm long ellipsoid;

style 0,9 mm long glabrous 4-ridged; stigma 1,3 mm

wide broadly funnel-shaped, exserted, relatively very

large and conspicuous, glabrous, red. Fruit not

known. Figs 19 & 20.

Cape. — 3419 (Caledon): Spitzkop, hills near Heidehof (Awila)

(-DA), May 1973, Oliver 4272 (STE); April 1975 Oliver 5818 (E;

VARIOUS AUTHORS

449

Fig. 19. — Distribution of • Erica

insignis and of ▲ Scyphogyne

calcicola.

1500r

NBG; PRE; STE; W; Z); Feb 1980, Oliver 7604 (BM; BOL; K;

MO; NBG; PRE; S; STE).

The genus Scyphogyne at present consists of 18

species which occur on the mountains mostly in the

region from the Caledon District in the south to the

Cedarberg in the north. Only the very common S.

muscosa (Ait.) Druce is recorded in the Bredasdrop

District from a few localities on the coastal hills

where it is associated with fynbos vegetation on acid

sandstone. Therefore the discovery of this new

species growing on the limestone formations, is of

particular interest and hence the specific epithet.

The species was discovered by accident while a

survey was being made of the unique series of

limestone hills between the Uilenkraal River mouth

and Quoin Point. On two subsequent visits to the

area to collect better material a thorough search

revealed only the single population first discovered.

The plants were growing on a fairly steep south-

facing slope below small cliff faces. The soil was the

Fig. 20. — Scyphogyne calcicola. 1, flower; 2, corolla; 3, abaxial

sepal; 4, lateral sepal; 5, gynoecium; 6, anther, front, side and

back views; all x 25; 7, leaf, x 12,5. Drawn from the holo-

type, Oliver 7604 (STE).

typical light brown sandy soil associated with coastal

limestone in the south-western Cape. The surround-

ing vegetation was very short restiads and equally siz-

ed Rutaceae and Fabaceae.

The calcified dunes of the coastal areas of the

south-western and southern Cape have only in recent

geological times become available for colonization by

plants. Being poor in nutrients and well drained,

their colonization has been by fynbos elements from

the nearby sandstone hills. In many cases speciation

has accompanied this colonization. There are some

notable examples of this among the typical Cape

families, Proteaceae, Restionaceae and Ericaceae. In

the Ericaceae there are a number of endemic

limestone species. This is particularly so in the genus

Syndesmanthus which I regard as the most recently

evolved genus in the tribe Ericeae. In the same range

of hills there occur several distinct and endemic

species of Erica — E. saxicola Guth. & Bol., E.

gracilipes Guth. & Bob, E. calcareophila E. G. H.

Oliver and the remarkable E. occulta E. G. H. Oliver

which grows out of cracks in the limestone cliffs like

a bonsai cypress. The problem of endemism on these

limestone formations among typically acid sandstone

loving genera is enigmatic. Very little research has

been done on this intriguing problem and it is hoped

that the recently established Fynbos Biome Project

will tackle such problems.

On first examination the material did not appear to

belong to the genus Scyphogyne, but dissection of

numerous flowers in spirit of the more recent collec-

tion showed that the species bears all the characters

characteristic of the genus namely no bracteoles, an

unequal 4-lobed calyx, a 4-lobed corolla, 4 stamens

and a 1 -celled, 1-ovuled ovary. Only one flower was

found to vary from this — it had an ovary with 2 cells

and 1 ovule per cell which is characteristic of the

genus Salaxis.

In the description it is stated that the flowers are

‘apparently sessile’. This has been used because it is

difficult to decide whether the large swollen region

below the calyx is a swollen receptacle or a very short

pedicel swollen and merged with the base of the

calyx. This condition has been noticed in a number of

ebracteate genera with unequal calyces.

The relationship of 5. calcicola is difficult to deter-

mine. As already mentioned, 5. muscosa is the only

other species anywhere near to the limestone hills,

but it is a completely different species with a different

facies with its scattered almost hidden flowers and its

free stamens. In the key S. calcicola runs down to a

position near S. divaricata (Klotzsch) Benth., a

species from the high mountains of the Caledon and

Worcester Districts. „ _ tI _

E. G. H. Oliver

450

NOTES ON AFRICAN PLANTS

ERIOCAULACEAE

A NEW COMBINATION IN ERIOCAULON

Eriocaulon dregei Hochst. var. sonderanum

( Koern .) Oberm. comb, et stat. nov.

E. sonderanum Koern. in Linnaea 27: 669 (1854);

N. E. Br. in F. C. 7: 55 (1897). Type; Transvaal,

Magaliesberg, Zevher 1731 (SAM, iso.!), Burke s.n.

(K!).

Mr R. D. Meikle of Kew has suggested in a per-

sonal communication that E. sonderanum could be a

variety of E. dregei. The present-day ample collec-

tions and the types of the names involved support

this view. E. dregei var. dregei appears to be a

polyploid. Var. sonderanum is the common variety

in the Transvaal; in Natal both varieties occur, but in

the eastern Cape the larger var. dregei becomes com-

mon, whereas var. sonderanum is rare.

A. A. Obermeyer

IRIDACEAE

A NEW SPECIES OF CROCOSMIA

Crocosmia pearsei Oberm., sp. nov., C. paniculata

affinis, sed planta solitaria, rhachidi recta non

distincte flexuosa (modo dicto anglice ‘zig-zag’) et

floribus majoribus differt.

Cormi tunicae fibrosae. Planta ad 1 m alta,

decidua, solitaria. Eolia c. 5 subdisticha, anguste

elliptica, apicibus et basibus longe attenuata, c. 500

x 50-70 mm, plicata, basi obliqua, decidua. Spicus

distichus, erectus simplex vel pauce ramosus, densus.

Spathae minorae ovatae, c. 10 mm longae, apicula-

tae, brunneae. Perianthium aurantiacum, c. 90 mm

longum; tubus longus, incurvatus, base angustus,

sensim versus faucem ampliatus; limbus obliquus,

lobi inaequales, lobo superiore grandiore et erecte,

25-30 mm longo. Stamina tubo prope basin affixa;

antherae lineari-sagittatae, dorsifixae, versatiles.

Ovarium 3-loculare, ovulis pluris. Capsula globosa,

c. 7 mm longa (immatura).

Type. — Natal, 2829 (Ladysmith): summit of

Mnweni Pass in Drakensberg, c. 2900 m (-CC), R.

O. Pearse 34 (PRE, holo.).

Corm tunicated, fibrous. Plants up to 1 m tall,

solitary, deciduous. Leaves c. 5, distichous, narrowly

elliptical, attenuated above and below, c. 500 x

Fig. 21. — Crocosmia pearsei. Pearse 34, sheet I of holotype in

PRE.

Fig. 22. — Crocosmia pearsei. Pearse 34, sheet II of holotype in

PRE.

VARIOUS AUTHORS

451

50-70 mm, plicate, thin but tough, the main nerve

placed to one side at the base, deciduous. Spike erect,

simple or with a short basal side branch; flowers

distichously arranged, close together; rhachis straight

or showing a faint zig-zag pattern. Spathe-valves

small, ovate c. 10 mm long, apiculate, brown.

Perianth orange, c. 90 mm long, with a long curved

tube narrow at the base, gradually widening towards

the mouth; limb with the upper lobe larger and erect,

elliptic, c. 25-30 mm long. Stamens attached to the

tube near base; anthers linear, sagittate, dorsifixed,

versatile. Ovary 3-locular, ovules many. Capsule

globose, c. 7 mm long (immature). Figs 21 & 22.

Natal. — 2829 (Ladysmith): summit of Mnweni Pass in

Drakensberg, c. 2900 m (-CC), Pearse 34 (PRE, holo.). 2828

(Bethlehem): Icidi Pass (-DD), Gray s.n. (PRE).

It gives me much pleasure to name this rare

Crocosmia after Mr R. O. Pearse, the discoverer of

this species and the author of ‘Mountain Splendour’,

A NEW SUBSPECIES OF

Gladiolus microcarpus Lewis subsp. italaensis

Oberm., subsp. nov., a subsp. typica foliis glabris et

rhachidi dense pubescenti differt.

Type. — Natal, 2731 (Louwsburg): Itala Nature

Reserve (-AC), Mauve et al. 5266 (PRE, holo.).

When Lewis in 1966 examined material of

Gladiolus at PRE, she identified J. C. Smuts 2371

from Wakkerstroom, collected in December 1944, as

near G. microcarpus cf . J1 S. Afr. Bot., Suppl. 10: 87

(1972). In 1976 Messrs Brown and Shapiro under-

took a survey of the flora of the Itala Nature Reserve

and collected specimens of this species (their No. 411)

at the junction of the Thalu and Pongolo Rivers, in a

Lowveld type of vegetation. This was the first record

of a precise locality and it seemed worthwhile to ex-

amine it more closely as the typical material of G.

microcarpus comes from the Drakensberg at high

altitudes, some 200 km to the south or south-west of

the localities referred to above. The species therefore

appears to have a disjunct distribution. The reserve

was visited in January 1980 and more material was

obtained from the same area. It was found on a N.

W. -facing, steep slope, growing in rock crevices

shaded by small trees and shrubs. The flowers were

delicate in texture and light pink in colour; the lower

lobes had a dark pink midrib flanked by a white area.

When fading the pink turns to pale- and later to dark

lilac. Although closely related to the typical form,

some differences in the pattern of pubescence and the

distribution suggest that we are dealing with a

subspecies (Fig. 23).

Key to subspecies

Rhachis and spathes glabrous; leaves with short, patent hairs on

the nerves; growing in the Drakensberg at high altitudes, c.

1 800 m above sea-level; plants usually pendulous

subsp. microcarpus

Rhachis and spathes shortly and densely pubescent; leaves

glabrous; growing in Lowveld vegetation, c. 250 m above sea-

level; plants erect subsp. italaensis

a lavishly illustrated book on the flora of the

Drakensberg. There is a photograph of this

Crocosmia (unfortunately captioned No. 3 instead of

No. 4 on p. 89 of his book). The species is related to

C. paniculata (Klatt) Goldbl., but its leaves are an-

nual, not evergreen, the rhachis is more or less

straight or only faintly zig-zag and it bears bigger

flowers. C. paniculata is the common species on the

mountain slopes, forming large evergreen colonies. It

is therefore understandable that collectors overlook-

ed this elusive solitary plant, probably dismissing it

as a young stage of C. paniculata. So far it has only

been seen twice by Mr Pearse at high altitudes

(2 200 - 3 000 m), where it was flowering in January.

Other mountaineers, for instance Mr H. G. Gray,

who collected this species on Icidi Pass, also report it

as rare.

A. A. Obermeyer

GLADIOLUS MICROCARPUS

Fig 23. — Gladiolus microcarpus subsp. italaensis. Painting of

holotype plant. Mauve et al. 5266, by R. Holcroft.

The Natal Parks Board and their staff at the Itala

Nature Reserve are thanked most sincerely for their

kind co-operation.

A. A. Obermeyer

LILIACEAE

A NEW VARIETY OF ALOE FROM THE VRYHEID DISTRICT

Aloe reitzii Reynolds var. vernalis Hardy, var.

nov., a varietate typica bracteis floralibus deltoideo-

acuminatis, 6x 4-5 mm, non lanceolato-acutis 14x7

mm, capsula matura minima 15 mm longa 10 mm

diametro, non 25 mm longa et 14 mm diametro dif-

fert.

Type.— Natal, Vryheid District, farm ‘Oor-

spring’, 1971-09-10, Hardy 3589 (PRE. holo.).

452

NOTES ON AFRICAN PLANTS

Plant succulent, acaulescent or rarely with a stem

up to 50 cm, the stem procumbent, simple. Leaves

lanceolate-ensiform, arcuate-erect, densely rosulate,

dull green, up to 64 cm long, 5-9 cm broad at the

base, the apex armed with a pungent thorn; upper

surface flat to slightly concave, smooth, devoid of

markings; lower surface convex, smooth or

sometimes with thorns in the median line near the

apex, the thorns pungent, brownish, about 3 mm

long, 5 mm apart, the teeth arising from white bases,

the interspaces straight, the same colour as the leaf.

Inflorescence 70-75 cm high, two to four branched

from below the middle. Peduncle 3,5-4 cm diam.

near the base, naked below the first branch, branches

below the racemes clothed with several sterile bracts.

Racemes cylindric, slightly acuminate, densely multi-

flowered, 30-40 cm long, 5-6 cm diam., the buds

and flowers pendulous. Bracts deltoid-acuminate, 6

mm long by 4-5 mm broad at the base, scarious,

brownish, many-nerved, reflexed. Pedicels green, 3

mm long. Perianth curved-cylindrical, up to 40 mm

long, the base rounded, tube 5 mm diam. enlarging

to about 7 mm above the middle, narrowing slightly

towards the mouth; outer segments connate into a

tube for about 25 mm, the free portion 15 mm, the

two upper segments bright orange with five red-

brown nerves, margins paler, the apices subacute,

slightly falcately connivent, the lower segment

lemon-yellow with three to five darker coloured

nerves, apex subacute, straight; inner segments free

but dorsally adnate to the outer to about the middle,

with narrow white margins, the upper segment with a

15 mm keel the colour of the perianth becoming

brownish at the apex, the two lower segments with a

less pronounced keel formed by three greenish

nerves. Filaments distinctly flattened, the three inner

narrower and lengthening in advance of the three

outer, pale lemon-yellow within the perianth, the ex-

serted portion brownish-orange. Anthers exserted by

8 mm. Style a deeper yellow than the filaments, the

exserted portion the same brownish-orange colour.

Stigma becoming exserted by 10 mm. Ovary olive-

green, 6mm long, 3 mm diam. at base. Capsule 15

mm long, 10 mm diam. Fig. '24.

Natal. — 2730 (Vryheid): farm ‘Oorspring* between Vryheid

and Natal Spa (-DB), 10 Sept. 1971, Hardy 3589 (PRE).

This interesting variety differs from the typical

variety in the following respects: the floral bracts are

deltoid-acuminate in shape, 6 mm long by 4-5 mm

broad as opposed to lanceolate-acute and 14 mm

long by 7 mm broad. The mature capsules are also

very much smaller, 15 mm long and 10 mm in dia-

meter compared with 25 mm long and 14 mm dia-

meter in the typical variety.

There are also differences in flowering time and in

distribution: var. reitzii flowers during February and

Fig. 24 — Aloe reitzii var. vernalis. Plant in natural habitat on

steep, well-drained granitic slopes.

March, whereas var. vernalis flowers during the

spring months of August and September, hence the

varietal epithet. Var. reitzii has been recorded from

three or four localities in the Belfast District of east-

ern Transvaal, whereas var. vernalis is apparently

confined to a very small area in the Vryheid District

of Natal.

A closely related species, A. gerstneri Reynolds,

occurs near the Nondweni River south of where A.

reitzii var. vernalis occurs and can, when not in

flower, be easily confused with the latter; in both the

slender leaf marginal teeth arise from distinct white

bases, a character not occurring in var. reitzii. When

in flower, however, A. gerstneri is readily distin-

guished by its cylindric-ventricose flowers which are

up to 30 mm long.

D. S. Hardy & C. Reid

A REAPPRAISAL OF URGINEA ALTISSIMA

When discussing the Urginea epigea-Drimia altis-

sima problem in Bothalia 13: 139 (1980), I had unfor-

tunately omitted to examine the Thunberg type speci-

men of Ornithogalum altissimum L.f. I have since

established that it was collected in the eastern Cape at

the Zondags River and that it bears its flowers on

long patent pedicels Fig. 25. 1 now realize that it is

the widespread summer rainfall region species which

is found as far north as tropical Africa. Jessop in J1

S. Afr. Bot. 43: 288 (1977) was correct in his interpre-

tation, whereas 1 was wrong and I apologize. But,

like Baker in FI. Cap. 6: 470 (1897), I prefer to place

it in Urginea for it produces white stellate flowers on

long patent pedicels.

However, the reference by Jessop to the plate in

Curtis’s bot. Mag. 27: t. 1074 (1808) is misleading as

this depicts a different species, namely the well-

known ‘Maerman’ from the winter rainfall region

Fig. 26. This species is recognized by its long, dense

spike bearing sessile or nearly sessile flowers. Unfor-

tunately it was wrongly identified as Urginea

altissima (L.f.) Baker by Adamson & Salter in the

Flora of the Cape Peninsula p. 194 (1950) and other

VARIOUS AUTHORS

453

Fig. 26 — Drimia forsteri (Bak) Oberm. Curtis’s bot. Mag 27: t.

1074 (1808), as ‘Drimia altissima’.

Fig. 27. — Types of Drimia (Urginea) forsteri (Bak.) Oberm.:

Forster (K, holo.!); Cooper 3275 (K, iso.).

The Director of the Thunberg Herbarium is

thanked for the photograph of the type of Urginea

(Ornithogalum) altissima (L. f.) Bak. and to the

Director of the Royal Botanic Gardens, Kew, thanks

are due for the photographs of Drimia ( Urginea )

fosteri (Bak.) Oberm.

Note. Drimia altissima Hook, in Curtis’s bot. Mag.

t . 5522 ( 1 865), non Ker-Gawl. , is an entirely different

species. It was given the new name, D. alta R. A.

Dyer in 1943 (Flower. PI. S. Afr. 23: t. 890). There is

however, the older name Drimia robusta Bak. in

Saund. Ref. bot. t. 190 (1870), which should be

adopted for this eastern inland species. It is not the

same as D. elata Jacq. (cf. Jessop in J1 S. Afr. Bot.

43: 285).

A. A. Obermeyer

Fig. 25. — Holotype of Urginea (Ornithogalum) altissima (L. f.)

Bak. (UPS 8275), from eastern Cape.

Cape botanists and it must receive another name.

The flowers are somewhat fleshy with tepals that

form a shallow cup and recurve above and the

stamens are usually coherent at first. Because of this,

1 prefer to place it in Drimia.

Drimia forsteri (Bak.) Oberm., comb. nov.

Urginea forsteri Bak. in FI. Cap. 6:469(1897). Types: Forster

s.n. (K, holo.!); Cooper 3275 (K, iso.). Fig. 27.

Drimia altissima sensu Ker-Gawl. in Curtis’s bot. Mag. 27:t.

1074 (1808), non Ornithogalum altissimum L.f.

Selected specimens:

Cape — 3218 (Clanwilliam): Piketberg (-DA), Marloth 7692.

3318 (Cape Town): Cape Peninsula, Table Mtn, Apostle Ravine

(-CD), Marloth 13029; Llandudno (-CD), Leighton 458; De

Grendel, western slope of Tygerberg hills (-DC), Boucher 3516.

3319 (Worcester): Worcester Veld Reserve (-CB), Olivier 67. 3422

(Mossel Bay): Mossel Bay (-AA), Marloth 7329.

454

NOTES ON AFRICAN PLANTS

MESEMBRYANTHEMACEAE

NOMENCLATURE IN THE GENUS MESTOKLEMA

The original description of the genus Mestoklema

was published on 29 August 1936 in English only

(Brown, 1936). This means that the genus and the

four new species which Brown described in the same

posthumous paper, are all invalidly published in

terms of the International Code of Botanical

Nomenclature, Article 35, which requires that for all

names of new taxa published after 1 January 1935, a

Latin diagnosis or description be given. And, since

the genus was not validly published. Brown’s three

new combinations were also not validly published

(Article 43).

Latin translations of Brown’s original descriptions

are given here in order to validate the five names

which were not published validly. These descriptions

have not been changed except for the measurements,

which have been metricated. In the case of the new

combinations, Brown’s literature citations have been

taken over unchanged.

Mestoklema TV. E. Br. ex Glen

Mestoklema N. E. Br., Gdnrs’ Chron. 100: 164 (1936) descr.

anglice.

Frutices vel fruticuli, una specie cum trunco ar-

boriforme, dumaliter, saepe intricate ramificati, ali-

quando (semper?) radicibus tuberosis; rami juvenes

minute papulosi, tactu asperiusculi ubi exsiccatis,

pallidi; cymi perdurantes, indurescentes, subspines-

centes neque pungescentes. Folia opposita, basin non

connata, desciscentia saepe prominentiam denticuloi-

deam relinquentia, trigona vel subteretia, minute vel

microscopice papulosa micantiaque, cum caespibus

foliarum in axillis. Flores parvi, pedicellati, in cymis

terminalibus bracteatis dichotome 2-5-plo divisis.

Bracteae parvae, foliiformes. Calyx subaequanter

5-lobatus, nonnulli lobi marginis angustis mem-

branaceis. Petala uniseriata, linearia, saepe vix

longiora quam lobi calycis. Stamina multa, conico-

collecta, intima barbata; staminodia nulla. Stigmata

5, erecta, breviora quam stamina, subulata, fortasse

papillata. Ovarium semisuperius, conicum vel con-

vexum, superne cristatum, 5-valvatum

-loculatumque, valvuli recurvati dum expansi,

carinae expansoriae inferne contiguae, superne

divergentes, vulgo sordide aurantiacae vel

aurantiaco-brunneae, alibus angustulis, acutis, mem-

branaceis, sine tuberculo placentale. Semini ovoidei,

brunnei, laeves. Type species: M. tuberosum (L.) N.

E. Br. ex Glen.

Mestoklema tuberosum (L.) TV. E. Br. ex Glen,

comb. nov.

Mesembryanthemum tuberosum L., Sp. PL 484 (1753); Haw.,

Obs. Gen. Mesemb. 271 (1795); Misc. Nat. 89 (1803); Syn. PI.

Succ. 252 (1812); Rev. PI. Succ. 179 (1821); Salm Dyck, Monogr.

Gen. Aloes Mesemb. 49, t. 2 (1854); Berger, Mesemb. Portulac.

100, fig. 13. 11I-IV (1908). Delosperma tuberosum (L.) Schwant.,

MOllers dt Gartn.-Ztg 42: 258 (1927). Iconotype: Dill., Hort. Elth.

275, t. 207, fig. 264 (1732).

M. spinosum O. Kuntze, Rev. Gen. PI. 3: 109 (1893), non L.

var. macrorhizum (Haw.) TV. E. Br. ex Glen,

comb, et stat. nov.

Mesembryanthemum macrorhizum Haw. in Till. Phil. Mag.

1826: 331 (1826); Salm Dyck, Monogr. Gen. Aloes Mesemb. 49, t.

3 (1842); N. E. Br. in J. Linn. Soc., Bot. 45: 121 (1920). Lecto-

iconotype: Duncanson watercolour (K ! ).

M. megarhizum Don, Gen. Syst. 3: 145 (1834); Sond., FI. Cap.

2: 441 (1862); Berger, Mesemb. Portulac. 101 (1908).

Brown states that the type is a plant of which

Bowie sent tubers to Kew in 1820 from South Africa.

The plant itself has disappeared, but there is a Dun-

canson watercolour of one branch of it still extant at

Kew; this is now designated as the lecto-iconotype.

Mestoklema arboriforme (Burch.) N. E. Br. ex

Glen, comb. nov.

Mesembryanthemum arboriforme Burch., Trav. 1: 343 (1822);

N. E. Br. in J. Linn. Soc., Bot. 45: 124 (1920). Type: Cape, be-

tween Wittewater and Rietfontein, Burchett 2004 (K, holo.l).

Mestoklema copiosum TV. E. Br. ex Glen

Fortasse frutex parvus vel fruticulus, rami visi

[Brownio — HFG] 150-200 mm longi, inferne 3-4

mm lati, internodiis 4-12 mm longis, juvenili minute

papulosi albidi, primo cinerascentes, deinde laeves-

centes brunnescentesque aetate. Folia adscendentia-

patentia, 10-17 mm longa, ca. 1,5-2 mm crassa,

trigona, obtusa, haud vel aegre apiculata, ut videtur

superne subcanaliculata, inferne obtusissime

carinata, microscopice papulosa. Cymi 40-90 mm

diametro, 4-5-plo furcati, subdense pluri- (15-50)

-florati. Bracteae 2-4 mm longae, crassae, apice ob-

tusissimae subrecurvatae. Pedicelli 2-3 mm longi.

Receptaculum 2,5-3 mm diametro, perbreve ob-

conicum; lobi calycis 2-3 mm longi, deltoidei. Cor-

olla ut videtur ca. 8 mm diametro, petala ca. 3 mm

longa, 0,5-0, 7 mm lata, obtusa vel acuta. Stamina

ca. 1,5 mm longa. Stigmata 5-7, ca. 0,6-1 mm

longa, crassa, acuminata. Ovarium superne convex-

urn. Capsula 2,5 mm diametro, obconica; cetera ut in

descriptione generico.

Type. — Cape. Mazelsfontein, Anderson 720 (K,

holo.l).

Mestoklema elatum TV. E. Br. ex Glen

Frutex 60-120 cm altus, rami primi adscendentes,

inferne 4-6 mm lati, internodiis 10-25 mm longis,

juvenili perminute papulosi cinerei; laevescentes

brunnescentes aetate; cortex exsiccatus longitudine

rugosus. Folia adscendentia-patentia vel patentia,

8-15 mm longa, 1,3- 1,5 mm crassa, compressa-

trigona, superne plana vel subcanaliculata, inferne

carinata, apice recurvata, lateraliter visa inferne

subacuta vel obtusa, microscopice papulosa. Cymi

25-40 mm diametro, dense 15-20 -florati, pedicelli

2-6 mm longi. Bracteae pusillae, ca. 1-1,5 mm

longae, ut videtur mox desciscentes. Receptaculum

2,5-3 mm diametro, lobi calycis ca. 1,5 mm longi,

deltoidei, obtusi. Corolla 6-7 mm diametro; petala

vix 2 mm longa, obtusa, ‘purpurea’ — fortasse color

aliquis magenteae. Stamina ca, 1,5 mm longa.

Stigmata vix 1 mm longa. Capsula 2,5-3 mm

diametro, mox desciscens; cetera ut in descriptione

generico.

Type. — Cape, between Committees and Hunt’s

Drift, Dyer 890 (K, holo.l).

Mestoklema illepidum TV. E. Br. ex Glen

Planta ca. 230-250 mm alta, glabra, multiramosa.

Rami primi inferne 5-8 mm crassi, internodiis 6-13

mm longis, juvenili minute papulosi, exsiccati albidi,

aetate brunnescentes. Folia parva, 3-8 mm longa,

1-1,5 mm crassa, compresso-subteretia, obtusa,

appice subrecurvata, microscopice papulosa, viridia.

VARIOUS AUTHORS

455

Flores perparvi, in cymis terminalibus bracteatis,

bracteis 1-2 mm longis. Pedicelli 2-4 mm longi.

Receptaculwn 2-2,5 mm diametro; lobi calycis 1,5

mm longi, deltoidei, obtusi. Corolla ut videtur ca. 6

mm diametro; petala 2 mm longa, 0,4-0, 5 mm lata,

obtusa, ‘submagentea’. Stamina ca. 1 mm longa, ut

videtur alba. Stigmata ca. 0,5 mm longa subulata.

Capsula ca. 3 mm diametro; cetera ut in descriptione

generico.

Type. — Cape, on the road from Grahamstown to

Bedford, Dyer 2336 (K, holo.!).

Mestoklema albanicum N. E. Br. ex Glen

Planta ut videtur ca. 230-380 mm alta, glabra.

Rami primi ut videtur basin decumbentes, deinde

erecti vel adscendenti, ramosi, basin 2-3,5 mm lati,

internodiis pro parte maxima 13-19 mm longis,

juvenili minute papulosi cinerei, laeviscentes brun-

nescentesque aetate. Folia patentia, apice plus

minusve recurvata, 6-10 mm longa, 1,5-3 mm

crassa, supra ut videtur plana vel subcarinata, infra

obtuse carinata, fortasse compressa, minute

papulosa. Cymi 25-55 mm diametro, laxi, pedicelli

8-15 mm longi, Calyx subaequaliter 5-lobatus, lobi

2-2,5 m longi, deltoidei, obutsi; receptaculum ca. 3

mm diametro, breve obconicum. Corolla ut videtur

ca. 7 mm diametro, petala ca. 2,5-3 mm longa,

linearia, obtusa, ‘submagentea’. Stamina ca. 2 mm

longa. Stigmata ca. 1 mm longa, subulata; ovarium

supra convexum. Capsula 4-5 mm diametro; cetera

ut in descriptione generico.

Type. — Cape, 12 km from Grahamstown towards

Cradock, Dyer 1308 (K, holo.!; PRE, iso.!).

I should like to thank Drs O. A. Leistner and D. J.

B. Killick for reading through this manuscript and

for their helpful suggestions.

REFERENCE

Brown, N. E., 1936. Mesembryanthemum. Gdnrs’ Chron. 100:

164-166.

H. F. Glen

ORCHIDACEAE

NOTES ON SOUTH AFRICAN SPECIES OF HOLOTHRIX

Holothrix filicornis Immelman & Schelpe, sp.

nov., FI. scopularia (Lindl.) Reichb. f. affinis, calcari

longissimo, foliis scapisque glabris, spica non secun-

dissima differt.

Folia glabra. Scapus ebracteatus, glaber, 65-260

mm longus. Sepala glabra, ovata, acuta, 1, 5-2,0 mm

longa, 0, 5-1,0 mm lata, viridi-rubra. Petala triloba,

3, 0-8,0 mm longa, 0, 5—1,0 mm lata, viridi-alba,

lobis filiformibus, paulum carnosis. Labellum quin-

quilobum, 4, 5-9,0 mm longum, 1,5-2, 5 mm latum,

viridi-album, lobis filiformibus, paulum carnosis.

Calcar conicum, dependens vel leviter incurvatum,

7-1 1 mm longum.

Type. — Cape, 2917 (Springbok): 23 km west of

Springbok (-DB), Acocks 19269 (BOL, holo.; PRE;

K).

Leaves glabrous. Scape glabrous, without bracts,

65-260 mm long. Ovary twisted, green tinged with

red. Sepals glabrous, ovate, acute, 1,5-2, Ox 0, 5-1,0

mm, green tinged with red. Petals with three filiform,

slightly carnose lobes comprising Vi-Vi of the total

petal length, 3, 0-8, Ox 0,5-1, 0 mm, white tinged

with green. Lip slightly carnose, divided into five

filiform lobes, lobes comprising Vi-V* of the total lip

length, lip 4, 5-9, Ox 1,5-2, 5 mm, greenish white.

Spur conical, curved, pendulous, 1,25-2,0 times the

length of the lip, 7-1 1 mm long.

This species appears to be confined to northern

Namaqualand. It grows in rock crevices and on stony

slopes; flowering from June to September.

Cape. — 2816 (Oranjemund): south-western Richtersveld (-BB)

Williamson 2562 (BOL); Williamson 2565 (BOL). 2817

(Vioolsdrif): north-east end of Paradysberg (-AC), Tolken 3293

(BOL); Richtersveld, Stinkfontein, south-west of Zylsrus, upper

eastern slopes on summit of ridge (-CA), Oliver, Tolken & Venter

652; Richtersveld, Cornelsberg in Stinkfontein mountains, in main

large kloof, on east side below beacon (-CA), Oliver, Tolken &

Venter 702. 2917 (Springbok): Richtersveld, Karoegapoort, 9 km

south of Lekkersing, main hill north of poort, on north side

(-AA), Oliver, Tolken & Venter 823; 23 km west of Springbok

(-DB), Acocks 19269 (BOL; PRE; K).

H. filicornis appears to be most closely related to

H. scopularia among the Holothrix species in South

Africa, both having 3-lobed petals and a bractless

scape. The petals in all other species with divided

petals have a variable number of lobes. However, H.

scopularia is a high-montane species from the Dra-

kensberg, and the two are also easily distinguished on

the basis of the long spur, the glabrous leaves and

scape, and the less strongly secund spike of H. fili-

cornis.

The most striking character of this species is the

relatively very long spur. Unfortunately the specific

epithet longicornu has already been used in the

genus, and so filicornis was chosen instead.

Holothrix brevipetala Immelman & Schelpe, sp.

nov., H. cernua (Burm. f.) Schelpe affinis, petalis

brevibus, lobis labii brevioribus et latioribus differt.

H. hispiclula sensu Bolus, Ic. Orch. Austro-Afr. 3: t. 17 (1913)

partly, excl. syn. L. f., Thunb.; Orch. Cape Penins. ed. 2: 20, t. 13

(1918) partly, excl. syn. L. f., Thunb.; Schltr. in Ost. bot. Z.

(1898) 443, partly, excl. syn. L. f., Thunb.; non (L. f.) Dur. &

Schinz.

H. parvifolia sensu Rolfe in FI. Cap. 5,3: 103 (1913) partly, excl.

syn. Thunb.; Bolus, Orch. Cape Penins. ed. 1: 115, t. 24 (1888);

non Lindl.

Folia squamellis vel pilis crassis dense vestita.

Scapus ebracteatus, crassus, 60-300 mm longus, pilis

hispidis reflexis. Sepala dense pilosa, ovata, acuta,

1,0-2, 5 mm longa, 0,5-1, 5 mm lata, luteo-viridia,

ad apicem carnosa. Labellum trilobum vel quin-

quelobum, 2, 5-3, 5 mm longum et latum, luteo-

viride, interdum verrucosum (necnon petalum), lobis

brevibus latis carnosis. Calcar conicum, leviter incur-

vatum, 1 -2 mm longum.

Type. — Cape, 3324 (Steytlerville): Humansdorp,

koppie above Oudebos (— CC), Jeppe in PRE 33391;

holo., K).

Leaves densely covered with small squamules or

stout hairs, sometimes withered at flowering. Scape

with short, hispid, reflexed hairs, without bracts,

stout, 60-310 mm. Sepals densely pilose, 1,0-2, 5

mm. Petals undivided, carnose at the apices, 2, 5-3, 5

x 0,5-1, 5 mm, yellow-green. Lip carnose, lip and

petals sometimes warty, lip oval in shape with 3-5

short, broad lobes, 2, 5-3, 5 x 2, 0-3, 5 mm, yellow-

green. Spur conical, slightly curved, 1, 0-2,0 mm.

H. brevipetala occurs in the south-western Cape

and in the southern Cape areas towards Port Eliza-

456

NOTES ON AFRICAN PLANTS

beth. It grows in sand and in shallow soil in rock-

crevices and, though it can be found in flower from

August to April, it usually flowers between October

and January.

Cape. — 3318 (Cape Town): Waai Vlei, Wolley-Dod 2339

(BOL); Table Mountain summit, near Klaassenbosch (-CD),

Bolus & Scully s.n. (BOL; SAM); Top of Nursery Buttress (-CD),

Bruvns 141/75 (NBG); Table Mountain, Echo Valley (-CD),

Compton 8284 (NBG); Table Mountain, between Skeleton and

Window Gorge (-CD), Leighton 750 (BOL); Table Mountain,

summit (-CD), Schlechter 464 (BOL). 3319 (Worcester): near

Elgin, between Grabouw and Paardeberg, (-AC), Stokoe s.n.

(SAM); Paarl, Slanghoek mountains, Witteberg, northern-western

side (-CA), Wasserfall 605 (NBG). 3324 (Steytlerville): (-CC),

Jeppe in PRE 33391. 3326 (Grahamstown): Coldstream (-BC),

Glass s.n. (PRE). 3418 (Simonstown): Constantiaberg (-AB),

Compton 7546<5(NBG); Cape Point, Smith’s Farm (-AD), Comp-

ton 6067 (NBG); mountain above Smits-winkelbaai (-AD),

Schlechter 667 (BOL). 3419 (Caledon): Palmiet River Mountains,

Stokoe s.n. (SAM). 3420 (Bredasdorp): Heidelberg, Naauwpoort

Peak (-BB), Thorn s.n. (SAM). 3423 (Knysna): Knysna (-AA),

Duthie s.n. (BOL.) 3424 (Humansdorp); Tsitsikama, Ratel’s

Bosch, flats below road, Fourcade 583 (GRA).

This species has been misidentified as H. hispidula

(L. f.) Dur. & Schinz which is based on Sparnnann

34 (LINN!); and as H. parvifolia Lindl., a superflous

name for H. hispidula. Both these names must be in-

cluded as synonyms of the species H. cernua (Burm.

f.) Schelpe, and a new name chosen for the species

referred to by Bolus, Rolfe, Schlechter and other

authors. H. brevipetala has been chosen because the

species has short, broad petals (and lip-lobes) which

characterize it and distinguish it from its close ally H.

cernua.

H. villosa Lindl. var. condensata (Sond.) Im-

melman, stat. et comb. nov.

H. condensata Sond. in Linnaea 19:76(1848). Type: Cape, 3420

(Bredasdorp): Swellendam, in sand dunes (-AB), Mund s.n. (S;

K).

H. lithophila Schltr. in Ost. bot. Z. (1898)446; Engl, in Bot. Jb.

26: 331 (1898). Type: Cape, 3419 (Caledon): on mountain above

Vogelgat lagoon (-AD), Schlechter 9556 (Bf ; one Bower at K).

A number of characters appeared at first to

separate H. condensata from H. villosa, i.e. denser

spike, larger Bowers and leaves, stouter scape,

broader lip-lobes and longer, broader petals.

However, on closer examination, these characters

were all found to intergrade or at least to overlap. In

their ‘typical’ forms the two taxa are separable, but

numerous intermediate specimens make it difficult to

uphold H. condensata as a separate species.

The distribution ranges of the two varieties

overlap, with H. condensata tending to occupy

damper habitats, and occurring almost exclusively

(but not quite: see type) in rock crevices and on cliffs.

H. villosa is also found in rock crevices but does oc-

cur frequently in other habitats, in fynbos, grassland,

semi-arid areas, sandy areas under young wattles,

and on roadsides. There is no difference in flowering

time. H. condensata is therefore reduced to a variety

of H. villosa.

Holothrix parviflora (Lindl.) Reichb. f. in Otia

Bot. Hamb. 2: 119 (1881). Type: Cape, 3323

(Willowmore): Swanepoelpoort Mountains (-BB),

Drege 8276a (K).

Tryphia parviflora Lindl. in Hook. Comp. bot. Mag. 2: 209

(1836).

T. secunda Lindl. in Hook. Comp. bot. Mag. 2: 209(1836), non

Orchis secunda Thunb. (1823). Type: Cape, 73322 (Oudtshoorn);

Koratra (Karatara?) (-DD), Drege s;m; (K).

Holothrix lindleyana Reichb. f. in Otia Bot. Hamb. 2: 119

(1881). Type: as for T. secunda Lindl.

H. G. Reichenbach considered Tryphia parviflora

Lindl. to be a separate species from T. secunda when

he transferred both to Holothrix. However, the

epithet secunda could not be used because of the ex-

istence of H. secunda (Thunb.) Reichb. f. (1823), and

so he gave the species a new name, H. lindleyana

Reichb. f.

T. parviflora was described by Lindley from a

specimen said to differ from T. secunda Lindl. in be-

ing half the height and having flowers one quarter the

size of T. secunda. The greater range of material

available today shows that the two are linked by in-

termediate specimens and they are considered to be

conspecific. The name H. lindleyana is therefore

superfluous.

H. incurva Lindl. in Hook. Comp bot. Mag. 2:

207 (1836).

H. rupicola Schltr. in Bot. Jb. 24: 419 (1897) syn. nov. Syn-

types: summit of Mont-aux-Sources, Thode 6 (not seen), slopes of

Mont-aux-Sources, Flanagan 11981 (BOL!).

H. scopularia (Lindl.) Reichb. f. in Otia Bot.

Hamb. 2: 119 (1881).

H. multisecta H. Bol. in J. Linn. Soc. Bot. 25: 170, 190, fig. 7

(1890), syn. nov. Type: Stockenstrom, Elandsberg summit, Scullv

391 (K; BOL!).

H. aspera (Lindl.) Reichb. f. in Otia Bot. Hamb.

2: 119 (1881).

H. confusa Rolfe in F. C. 5,3: 105 (1913), syn. nov. Syntypes:

mountain sides about Clanwilliam, Leipoldt sub MacOwan & Bolus

s.n. (Herb. Norm. Aust. 1757); Blaauwberg, Schlechter 8465 (K;

BOL!); near Oliphant’s River Mountains, Schlechter 5036 (K); near

Piekenierskloof and near Modderfontein, Schlechter 5077 (not

seen); Hex River Valley, Wolley-Dod 4054 (K; BOL!).

H. exilis Lindl., Gen. et Sp. Orch. 283 (1835).

H. exilis var. brachylabris (Sond.) H. Bol., Ic. Orch. Austro-

Afr. 1: t . 1 4, fig. A (1896), syn. nov. Type: Uitenhage, Zevhers.n.

(K).

H. cernua (Burm. /.) Schelpe in Orchid Rev. 74:

•394 (1966).

H. squamulosa Lind. var. scabra Bol. in Trans. S. Afr. phil.

Soc. 5,1: 1 14, t.23a (1888), syn. nov. Type: Cape Flats, sometimes

on old thatched roofs near Rondebosch, Bolus 7022a (BOL!).

H. squamulosa Lindl. var. hirsuta H. Bol. ibid., syn. nov. Type:

locality as above, Bolus 7022b (K; BOL!).

H. squamulosa Lindl. var. glabra H. Bol. ibid., syn nov. Type:

locality as above, Bolus 7022c (K; BOL!).

H. culveri H. Bol. in Trans. S. Afr. phil. Soc.

16: 147 (1905).

H. culveri var. Integra H. Bol. ibid., syn nov. Type: Barberton,

Culver 84a (K; BOL!).

This note is based on a thesis presented for an

M.Sc. at the University of Cape Town.

I am grateful for guidance and assistance in this

project by Prof. E. A. Schelpe, who also had the op-

portunity of examining types at Kew. His sketches of

these types have been seen by me.

Kathleen Immelman

VARIOUS AUTHORS

457

POACEAE

A NEW COMBINATION IN ER/OCHLOA

Eriochloa meyerana (Nees) Pilg. subsp. gran-

diglumis ( Stent & Rattray) Gibbs Russell, comb, et

stat. nov.

Panicum meyeranum Nees, FI. Afr. Austr. 32 (1841), as

meyerianum. — var. grandiglume Stent & Rattray in Proc. Trans.

Rhod. scient. Ass. 32: 28 (1933), as grandeglume.

Eriochloa meyerana (Nees) Pilg. in Natiirl PflFam. ed. 2, 143:

56 (1940), as meyeriana.

In 1841 Nees described Panicum meyeranum

which Pilger transferred to Eriochloa in 1940. 1 am in

agreement with this decision, which is also supported

by Clayton [Kew Bull. 30,1 : 107 (1975)], since the

species is obviously more closely allied to Eriochloa

than to Panicum, in spite of the presence of a lower

glume, a character not occurring in any other

Eriochloa.

In Eriochloa there is a bead-like swelling at the

base of the spikelet which is the result of the fusion of

the lower glume and the lowest rhachilla internode.

In Eriochloa meyerana the glume projects above this

bead as a truncate, inward-turned cuff surrounding

the base of the spikelet. However, specimens from

Zimbabwe, the Soutpansberg and the Kruger Na-

tional Park in the Transvaal and from Zululand,

have an ovate flat lower glume surmounting the bead

that may be Vi to Vi the length of the spikelet. These

specimens are readily recognized, are confined to a

discrete region within the range of the species and

constitute a taxon which in our opinion merits

subspecific rank. This taxon was described as var.

grandiglume by Stent and Rattray from Zimbabwe in

1933. Clayton (1975) cites this variety in synonymy

under Eriochloa meyerana. A new combination rais-

ing this taxon to subspecific rank is made here.

Clayton (1975) states that the spikelet and inflores-

cence of E. meyerana are intermediate between Erio-

chloa and Brachiaria mutica (Forssk.) Stapf and pos-

tulates that E. meyerana and related elements may be

the result of hybridization between B. mutica and

different local species of Eriochloa. However, B.

mutica is a tropical species which does not occur

naturally in southern Africa. Typical E. meyerana

may have migrated southwards after it originated in

the tropical regions. The subspecies grandiglumis

probably originated further south and if this taxon is

also of hybrid origin one should look for the other

parent in this area. No obvious parental species,

however, presents itself. Because the subspecies

grandiglumis is consistently more robust than the

typical subspecies, it may be that it is simply a

polyploid of E. meyerana subsp. meyerana.

Cytological investigation may throw more light on

this possibility.

Selected specimens:

Transvaal. — 2231 (Pafuri): Soutpansberg (-CA), Codd 5421

(PRE). 2331 (Komatipoort): Kruger National Park (-DD),

Coetzee 7028 (PRE).

Natal. — 2632 (Bela Vista): Maputaland (-DD), Maputaland

Expedition s.n. (PRE). 2831 (Nkandla): Hlabisa, Zululand (-BB),

Ward 2100 (PRE).

G. E. Gibbs Russell

NOTES ON DIGIT ARIA IN SOUTH AFRICA

Since Henrard’s monograph (1950) on Digitaria

only Chippindall (1955) has contributed to the tax-

onomy of the genus in South Africa. Chippindall’s

contribution, however, was limited to comments and

suggestions regarding the taxonomy and nomencla-

ture, but few definite decisions were made. Although

Henrard’s monograph is a taxonomic milestone and

will have to be considered in any subsequent work on

the genus, some authors have criticized various

aspects of the monograph, especially the unsatisfac-

tory delimitation of species and infraspecific taxa. In

southern Africa it is mainly taxa of the section Erian-

thae, that prove difficult to delimit. Most authors

agree that too many taxa are recognized and that

some of the presently accepted species might more

correctly be assigned to taxa of lower rank.

A recent taxonomic investigation of the South

African species of the genus by the present author

(1978) has confirmed this view. Thus, in the enume-

ration that follows, several taxa recognized by Hen-

rard are reduced to synonymy, one species is reduced

to subspecific rank and transferred to another

species, while one new subspecies is described.

1. Digitaria monodactyla (Nees) Stapf in FI. Cap. 7:

373 (1898).

D. monodactyla (Nees) Stapf var. explicata Stapf in FI. Trop.

Afr. 9: 442 (1919).

2. Digitaria diversinervis (Nees) Stapf in FI. Cap. 7:

379 (1898).

D. diversinervis (Nees) Stapf var. woodiana Henr. in Mon. Dig.

200 (1950).

3. Digitaria natalensis Stent in Bothalia 3: 152

(1930).

D. natalensis Stent subsp. stentiana Henr. in Blumea 1 : 93

(1934).

D. natalensis Stent subsp. stentiana Henr. var. paludicola Henr.

in Mon. Dig. 483 (1950).

D. macroglossa Henr. in Mon. Dig. 404 (1950).

D. macroglossa Henr. var. prostrata (Stent) Henr. in Mon. Dig.

406 (1950).

D. rigida Stent in Bothalia 3: 151 (1930).

4. Digitaria seriata Stapf in FI. Trop. Afr. 9: 432

(1919).

D. polevansii Stent in Bothalia 3: 149 (1950).

5 . Digitaria setivalva Stent in Bothalia 1 : 268 ( 1 924).

D. valida Stent subsp. burchelliana Henr. in Mon. Dig. 975

(1950).

D. pentzii Stent subsp. dregeana Henr. in Mon. Dig. 976(1950).

6. Digitaria eriantha Steud. in Flora 12: 448 (1829).

6.1 subsp. eriantha

D. smutsii Stent in Bothalia 1: 268 (1924).

D. geniculata Stent in Bothalia 3: 154 (1930).

D. stentiana Henr. in Blumea 1: 97 (1934).

D. bechuanica (Stent) Henr. in Mon. Dig. 295 (1950).

D. hiascens Mez in Bot. Jb. 57: 193 (1921).

6.2 subsp. pentzii (Stent) Kok, comb, et stat. nov.

D. pentzii Stent in Bothalia 3: 147 (1930).

D. valida Stent in Bothalia 3: 148 (1930).

D. valida Stent var. glauca Stent in Bothalia 3: 149 (1930).

D. decumbens Stent in Bothalia 3: 150 (1930).

6.3 subsp. stolonifera (Stapf) Kok, stat. nov.

D. pentzii Stent var. stolonifera (Stapf) Henr. in Mon. Dig. 544

(1950).

6.4 subsp. transvaalensis Kok, subsp. nov.

A typo subspeciei longitudine spicularum differt.

Spiculae 2,2 — 2,6 mm longae.

Type. — Transvaal, 2527 (Rustenburg): Silkaats-

nek, 1974-01-17, Kok 577 (PRE, holo.; PRU).

P. D. F. Kok

458

NOTES ON AFRICAN PLANTS

POLYGALACEAE

A CONSPICUOUS NEW SPECIES OF MURALTIA

Recently, while working at Kew, 1 found Esterhuy-

sen 24754 among the South African material of Poly-

gala. It certainly looked like a Polygala but, after

careful study, I concluded that it belonged rather to

the genus Mural tia, which is confined to South

Africa, with the exception of M. flanaganii which ex-

tends northwards to Tanzania. After consulting

Levyns’s (1954) excellent monograph, it became clear

that the specimen represented a new species.

Muraltia is closely allied to Polygala L., but differs

in having an ericoid habit (rare in Polygala); flowers

solitary and axillary (in terminal or lateral racemes,

rarely solitary in Polygala); the carina is differen-

tiated into claw and limb, with a cushion-like swell-

ing (this differentiation is absent in Polygala), a

2-lobed expanded, leaf-like crest (fimbriate or pluri-

lobed, rarely absent in Polygala); 7 stamens (Poly-

gala has 8 stamens, rarely 9, 4 or 5, sometimes only 6

fertile with 2 staminodes); the filaments are united

almost to the base of the anthers (in Polygala they are

usually free for a considerable distance); and lastly,

in Muraltia the anthers dehisce through a longitudi-

nal slit (in Polygala dehiscence takes place through a

large oblique pore). The pollen morphology

(zonocolporate) is similar to that of Polygala, but

more homogeneous.

Turczaninow (1855) divided Muraltia into two sec-

tins, which were considered as subgenera by Harvey

(1860) and Levyns (1954). After an extensive study of

herbarium material of Polygala (all African species

and representatives of every main division of the

genus), I cannot accept the division of Muraltia into

two subgenera. The genus is morphologically much

more homogeneous than Polygala and its pollen

grain is always zonocolporate and isopolar in con-

trast to Polygala, which has isopolar and heteropolar

pollen. If one accepts the two divisions of Muraltia as

subgenera, one would have to divide Polygala, not

into subgenera, but into separate genera, which is

unacceptable. Polygala is a world-wide, very

heterogenous, genus, but a good one and it is doubt-

ful whether anyone would suggest or accept the split-

ting of Polygala into separate genera. There are no

valid geographical or distributional grounds for ac-

cepting that Muraltia should be divided into two

subgenera. Therefore, Muraltia must be regarded as

having two sections, namely sect. Muraltia and sect.

Psiloclada Turzc in Bull. Soc. Imp. Nat. Moscou 27,

4: 353 (1855). These two sections may be easily distin-

guished by applying Levyns’s subgeneric key:

Leaves usually fascicled, sometimes solitary; flowers sessile or with

short pedicels; inner sepals usually slightly longer than the

outer; crest attached in various ways; capsule never cernuous,

often with long horn-like processes at the apex, occasionally

without horns sect. Muraltia

Leaves solitary; flowers with distinct, often long pedicels; inner

sepals much longer than the outer; attachment of the crest

long, almost vertical; capsule often cernuous, with or without

4 short horn-like processes at the apex, the horns never long

and slender sect. Psiloclada

Section Psiloclada appears to be transitional be-

tween Muraltia sect. Muraltia and Polygala in the

following characters: in having the two inner sepals

coloured and sometimes almost wing-like, much

longer than the outer ones, and a usually hornless

capsule indistinguishable from that of Polygala. The

new species, M. elsieae, which belongs to sect. Psilo-

clada, has the largest wings (inner sepals) in the

genus, resembling a Polygala, hence the initial identi-

fication of the plant as a Polygala. In spite of this

superficial resemblance in flower and in fruit (cap-

sule hornless and broadly winged), it is nevertheless a

true Muraltia (ericoid habit, solitary flowers; carina

with distinct claw and limb, a cushion-like swelling,

and an expanded leaf-like crest; stamens 7).

The new species, M. elsieae, is named after Miss

Elsie Esterhuysen of the Bolus Herbarium, whose

collections in South Africa have made a valuable

contribution to the knowledge of the flora of that

country. It can easily be distinguished from the other

species in the section by adapting the first dichotomy

in Levyns’s key to subgen. Psiloclada. On the

characters, ‘Inner sepals about twice as long as the

outer, concealing the carina, leaves rough; capsule

without apical horns or teeth’, she keys out only M.

polyphylla (DC.) Levyns. The key can be amended as

follows:

1. Inner sepals at least twice as long as the outer, concealing the

carina; capsule without apical horns or teeth:

Flowers 4-5 mm long, inner sepals about twice as long as the

outer; leaves scabrous, 5 — 10 mm long . . . 9. M. polyphylla

Flowers 11 - 13,5 mm long, inner sepals more than twice as long

as the outer; leaves not scabrous, 15-25 mm long

9a M. elsieae

1 . Inner sepals much less than twice as long as the outer, not con-

cealing the tip of the carina; capsule with apical horns or

teeth : Remaining species

Muraltia elsieae J. Paiva, sp. nov.

Frutex ramosus circa 1,5 m altus; rami teretes

glabri. Folia alterna petiolata, petiolo 0,5 mm longo

glabro; lamina 15-25 x 1—1,5 mm, lanceolato-linea-

ris, apice subacuta, breviter mucronata, glabra.

Flores purpurei, solitarii pedicello 2-3 mm longo

glabro; bracteae bracteolaeque similes, 1—1,5 mm

longae, cucullatae, ovatae, ciliatae, obtusae. Sepala

inaequalia; sepalum posterius, 2,5 x 1,5 mm, ova-

tum, apice nonnihil apiculatum, ciliatum; alae

11-13,5x4,5 mm, ovato-ellipticae, apice obtusae,

basim versus ciliatae; sepala anteriora libra 2x1,5

mm, ovata, apice nonnihil apiculata, ciliata. Petala

superiora 6-7 x 1 mm, oblonga, sed ad basim majora

obtusa, quam carina paulo longiora; carina 3,5

x 1,5-1, 8 mm, ungue limbum aequanti; crista ampla

1 mm longa, lobis superioribus et inferioribus simili-

bus. Ovarium 1x0,75 mm, applanato-ellipsoideum,

apice bilobatum glabrum; stylus gracilis, 1,5 longus,

teres, lobo posteriore stigmatico bene evoluto

longiore, anteriore breviore truncato glabro. Capsule

6-7 x 5,5-6 mm, applanato-ellipsoidea, apice

bilobata, glabra, margine alata, 1,3—1, 8 mm lata.

Semina 5x1,5 mm, ellipsoidea, sparse albo-

pubescentia, carunculata; caruncula 1,5 mm longa,

pubescens, appendicibus brevibus membranaceis.

Type. — Cape, 3321 (Ladismith): Seven Weeks

Poort, rocky slopes, 990-1 320 m, fl. & fr.

1955-10-13, Esterhuysen 24754 (BOL.; K, holotypus).

Slender single-stemmed shrub c. 1,5 m tall,

branching above, branches cylindric, glabrous.

Leaves alternate, very shortly petiolate (petiole 0,5

mm long, glabrous), 15 -25 x 1-1,5 mm, lanceolate-

linear, subacute and shortly mucronate at the apex,

glabrous. Flowers pale purple, solitary; pedicels 2-3

mm long, glabrous, bracts and bracteoles similar,

1-1,5 mm long, obtuse, cucculate ovate, ciliate.

Sepals unequal; posterior 2,5 x 1,5 mm, ovate-elliptic

somewhat apiculata at the apex, ciliate; wing-sepals

VARIOUS AUTHORS

459

Fig. 28 .—Muraltia elsieae. a, habit ( x 0,7); b, small section of branch ( x 1 ,3); c, leaf ( x 2,7); d, flower ( x 5,3); e, flower

with one wing removed ( x 5,3); f, vertical section of flower ( x 5,3); g, flower without sepals and ovary ( x 6,7); h, ovary,

style and stigma (x 6,7); i, capsule (x 5,3); j, seed (x 13,3). All from Esterhuysen 24754.

460

NOTES ON AFRICAN PLANTS

1 1-13,5 x 4-5 mm, ovate-elliptic, obtuse at the apex,

ciliate towards the base; anterior sepals free, 2x 1,5

mm, ovate, somewhat apiculate at the apex, ciliate.

Petals, the upper 6-7 x 1 mm, oblong but enlarged

towards the base, obtuse, sparsely pubescent towards

the base outside, and pubescent inside, somewhat

apiculate, slightly longer that the carina; carina

3,5-4x 1,5— 1,8 mm, with the claw as long as the

limb; crest 1 mm long with superior and inferior

lobes similar. Ovary broadly ellipsoid, 1 xO,75 mm,

glabrous, shortly bilobed at the apex; style 1,5 mm

long, terete, the anterior branch bruncate, glabrous,

the posterior stigmatic branch well-developed. Cap-

sule 6-7 x 5,5-6 mm, broadly ellipsoid to subcom-

pressed globose, bilobed at apex, glabrous, margin

winged (wing 1,3- 1,8 mm wide). Seed 5x 1,5 mm,

ellipsoid, sparsely white pubescent; caruncle 1,5 mm

long, pubescent and with very short appendage. Fig.

28.

Cape. — 3321 (Ladismith): Seven Weeks Poort, Esterhuysen

24754 (BOL; K, holotypus); Marloth 2946 (BOL).

REFERENCES

Harvey, W. H., 1860. Polygaleae. In W. H. Harvey&O. Sonder,

FI. Cap. 1: 79-113.

Levyns, M. R., 1949. The floral morphology of some South

African members of Polygalaceae. Jl S. Afr. Bot. 15: 79-92.

Levyns, M. R., 1952. Clues to the past in the Cape Flora of to-day.

S. Afr. J. Sci. 49, 3-4: 155-164.

Levyns, M. R., 1954. The genus Muraltia. Jl S. Afr. Bot., Suppl.

2, 247 pp.

Levyns, M. R., 1955. Some geographical features of the family

Polygalaceae in Southern Africa. Trans, r. Soc. S. Afr. 34, 3:

379-386.

Turczaninow, N., 1855. Animadversiones ad primam partem

herbarii Turczaninowiani, nuc Universitatis Caesarea

Charkoviensis. Bull. Soc. fmp. Nat. Moscou 27, 4: 271

(bis)— 372.

J. A. R. Paiva*

VITACEAE

A NEW SPECIES OF CYPHOSTEMMA FROM THE TRANSVAAL

Cyphostemma hardy i Retief, sp. nov., C.

bainesii (Hook, f.) Desc. affinis, a qua fructu glan-

duloso pubescente differt.

Planta succulenta, radici tuberosa, usque 0,5 m

alta. Caulis cylindraceus, aliquando ramosus,

ramulis junioribus tectis trichomatibus longis, albis,

multicellulosis, eglandulosis; cortice brunneo. Folia

spiraliter disposita, simplices vel 3(5) foliolata,

sessilia vel petiolata, pubescentia; lamina obovata,

1 30-320 x 8-20 mm, margine dentata rubella, apice

obtusato vel truncato; stipulis praesentibus.

Inf/orescentia cymosa foliis opposita; pedunculis

cum pedicellis trichomatibus multis multicellulosis

stipitatis glandulosisque. Flores tetrameri, bisex-

uales. Calyx cupulatus, integer. Corolla flavo-virens;

petalis 2 mm longis, cucullatis, caducis. Stamina

filamentis dorsifixis. Ovarium biloculare, disco et

glandibus quattuor cingente, ommibus ovario adna-

tis. Stylus persistens, stigmate bilobato. Fructus

bacca globulosa, 11-15 mm longus, pubescens glan-

dulosus, ruber vividus. Semina cristata.

Type. — Transvaal, 2427 (Thabazimbi): wooded

koppies on the farm Malmanieshoek, Waterberg

District (-BA), Hardy 1256 (PRE, holo.; K; MO).

Succulent up to 0,5 m in height with a tuberous

rootstock. Stem cylindric, sometimes branched,

younger covered with long, whitish multicellular,

non-glandular trichomes and stalked, glandular

trichomes; bark brown, peeling off in pale brown

strips. Leaves spirally arranged, simple or

3(5)-foliolate, base of terminal leaflet symmetrical,

bases of lateral leaflets asymmetrical, main leaf

sessile or petiolate but leaflets sessile, petioles 60-90

mm long if present, greyish green, densely covered

with white, multicellular trichomes lying more or less

in the same direction, trichomes tubular or cells of

the trichome twisted, all laminae obovate, 130-320

mm long and 8-20 mm broad, margin toothed, red-

dish, apex rounded or truncate; stipules lanceolate,

± 15-20 mm long, densely hairy, sometimes denticu-

late, teeth and apex sometimes gland-tipped. Inflor-

escence a leaf-opposed corymbose cyme; peduncles

and pedicels with numerous, multicellular, stipitate

glandular trichomes, 0,7-1 mm long. Flowers tetra-

merous, bisexual. Calyx 0,5-0, 7 mm in height entire.

* Botanical Institute, Coimbra, Portugal.

cupshaped. Corolla 4-partite, yellowish green tinged

with maroon; petals ± 2 mm long, hooded at the

apex, becoming reflexed, caducous. Stamens 4;

filaments dorsifixed; anthers 0,5-0, 7 mm long.

Ovary bilocular, surrounded by a disk and four

glands alternating with the petals and stamens, all

adnate to the ovary. Style 0,8—1 mm, persistent, ter-

Fig 29. — Cyphostemma hardyi. Hardy 1256, holotype in PRE.

VARIOUS AUTHORS

461

ajT

Fig. 30. — Cyphostemma hardyi. Painting of part of holotype

plant by Cythna Letty.

minating in a slightly bilobed stigma. Fruit a globose

to ovoid bright red berry, 11-15 mm long and

12,5-14 mm broad, glandular pubescent, glands

1—1,3 mm long, stalked, multicellular with globose

heads. Seed 8-10 mm, crested, embedded in a firm

bright yellow mass. Figs 29 & 30.

Transvaal. — 2427 (Thabazimbi): wooded koppies on the farm

Malmanieshoek, Waterberg District (-BA); Hardy 1256 (K; MO;

PRE); amongst boulders of Waterberg conglamerate, farm

Malmaniesriver, Waterberg District (-BA), Hardy, Relief &

Herman 5287 (PRE); south-western slopes of the Kransberg

(-BC), Dyer & Verdoorn 4230 (PRE).

C. hardyi is endemic to the Transvaal, where it is

known only from a restricted area in the western part

of the Waterberg mountain range. It occurs on slopes

amongst boulders of Waterberg-conglomerates in the

shelter of trees such as Albizia tanganyicensis Bak. f.

subsp. tanganyicensis, which also has a conspicuous

papery bark. Other succulents like Huernia quinta

(Phill.) White & Sloane, Aloe chabaudii Schoenl., A.

marlothii Berger, A. transvaalensis Kuntze and A.

aculeata Pole Evans are found growing in the same

area.

The peculiar growth form of Cyphostemma hardyi

distinguishes it from all the other known species of

the genus in the Transvaal which are herbaceous or

succulent, prostrate or erect creepers. The stems of

C. hardyi are sometimes very short, because the

plants are often subjected to veld fires causing them

to die back. The leaves fall off during the winter

resting period and leave prominent leaf scars.

C. hardyi has an extraordinary distribution. The

species of the genus Cyphostemma that show the

closest affinities with C. hardyi occur in South West

Africa and in no other part of southern Africa. C.

hardyi, C. bainesii (Hook, f.) Desc. and C. uter (Ex-

ell & Mendon^a) Desc. are all stem-succulents having

caudices more or less 0,3-0, 8 m high whereas C. jut-

tae (Dinter & Gilg) Desc. and C. currorii (Hook, f.)

Desc. [ = C. cramerianus (Schinz) Desc.] are suc-

culent trees 4-7 m in height. Although the leaves of

C. bainesii resemble those of C. hardyi, the leaf

margin is shallowly toothed in the case of C. bainesii,

and the leaves of C. hardyi are much more densely

covered with the long thin multicellular hairs, present

on both species. C. uter has leaves with a conspi-

cuous wavy margin and multicellular, glandular hairs

similar to those on the fruits of C. hardyi, as well as

multicellular non-glandular hairs with peculiar out-

growths.

The first record of C. hardyi in the National

Herbarium was from a steep south-western slope of

the Kransberg where Dr R. A. Dyer and Dr Inez Ver-

doorn collected material of the species during

January 1942. The species has been named after Mr

D. Hardy, who is a well known horticulturist on the

staff of the Botanical Research Institute.

E. Retief

Bothalia 13, 3 & 4: 463-466 (1981)

The phytogeography and ecology of Macrocoma

(Orthotrichaceae, Musci) in Africa

ROBERT E. MAGILL* and DALE H. VITT**

ABSTRACT

Of the 1 1 species presently recognized in the genus Macrocoma (Orthotrichaceae, Musci), four occur in Africa.

These species are xeromorphic and occur in savannas, woodlands, and forests, roughly corresponding to the Afro-

montane Phytogeographic Region. Macrocoma tenue (Hook. & Grev.) Vitt subsp. tenue is widespread in southern

Africa, north to Ethiopia and also occurs in Australasia. Three species are endemic, with M. pulchella (Hornsch.)

Vitt found only in the Cape area; M. abyssinica (C. Mull.) Vitt occurring in eastern Africa, and M. lycopodioides

(Schwaegr.) Vitt distributed sporadically in eastern and southern South Africa.

RESUME

LA PHYTOGEOGRAPHIE ET L’ECOLOGIE DU MACROCOMA (ORTHOTRICHACEAE, MUSCI) EN

AERIQUE

Des II especes actuellement reconnues dans le genre Macrocoma (Orthotrichaceae, Musci) quatre se trouvent en

Afrique. Ces especes sont xeromorphiques et survienment dans les savanes, les terrains boises et les forets correspon-

dant grosso-modo a la Region Phytogeographique A fromontane. Macrocoma tenue (Hook. & Grev. ) Vitt sous-esp.

tenue est repandue en Afrique australe, le nord le I’Ethiopie et survient aussi en Australasie. Trois especes sont

endemiques, avec M. pulchella (Hornsch.) Vitt survenant seulement dans la region du Cap; M. abyssinica (C. Mull.)

Vitt survenant en Afrique orientale et M. lycopodioides (Schwaegr.) Vitt distribute sporadiquement dans I'Est et

dans le Sud de T Afrique du Sud.

The moss genus Macrocoma (Hornsch. ex C.

Mull.) Grout is a member of the family Orthotricha-

ceae, subfamily Macromitrioideae. Within this sub-

family there are about eight genera, of which Macro-

mitrium with approximately 250 species is the largest.

Macrocoma was first recognized as a section by C.

Muller in 1845, and later Brotherus (1901-1909) con-

sidered it a subgenus of Macromitrium. Grout (1944)

was the first author to recognize this taxon at the

generic level, and most authors have accepted it since

that time as a distinct genus.

Briefly, the genus can be characterized by creep-

ing, slender stems and branches with imbricate,

closely set leaves, and by the uniformly rounded to

elliptic leaf cells. The basal cells, particularly at the

margins are more or less similar in shape to the upper

ones. The calyptrae are large, plicate and cover the

entire capsule. From Macromitrium and Grout iella,

the macroscopic appearance of species of Macro-

coma is quite different since in the former two

genera, the leaves are always twisted, contorted, or

incurved giving the individual plants a ‘bushy’ ap-

pearance. Keys to all species are presented in Vitt

(1980a) and the taxonomy and nomenclature are

discussed by Vitt (1973, 1980a, 1980b). Presently

eleven species, one with two subspecies, can be recog-

nized in the genus. Within the subgenus Trachyphyl-

lum (Broth.) Vitt there are two species [M. gracillima

(Besch.) Vitt and M. papillosa (Ther. in Herz.) Vitt],

both restricted in distribution to southern South

America. The subgenus Macrocoma is widespread in

both Southern and Northern Hemispheres, with two

species widespread in distribution. Macrocoma or-

thotrichoides (Raddi) Wijk & Marg. and M. tenue

(Hook. & Grev.) Vitt occur in both the Old and New

World (Fig. 1), M. frigida (C. Mull.) Vitt, M.

^Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

**Department of Botany, University of Alberta, Edmonton,

Alberta, Canada, T6G 2E9.

gastonyi Vitt, and M. brasiliensis (Mitt.) Vitt are

endemic to South and Central America ( + Mexico),

whereas M. iwatsukii Vitt has been collected only in

the Himalayan Mountains.

Three species occur only in southern and eastern

Africa, (Fig. 1) and, along with M. tenue subsp.

tenue comprise the African members of this genus.

These species occur on the branches and trunks of

angiospermous or rarely coniferous trees and occa-

sionally on rocks and boulders. Sim (1926) stated

that species of the M. tenue complex ‘are common

and are present in every bush and on nearly every soft

moist stone in South Africa’. The African species of

Macrocoma are xeromorphic and occur in savannas,

woodlands and forests, roughly corresponding to the

Afromontane Phytogeographic Region (cf. Werger,

1978). The species are frequently encountered along

the afromontane ‘archipelago’ that extends from

coastal areas of the Cape Province, northward along

the mountainous regions of eastern Africa, to

Ethiopia. This part of the Afromontane Region is

mostly surrounded by the Zambezian Domain of the

Sudano-Zambezian Phytogeographic Region, where

typical vegetation types are dry woodlands, savannas

or grasslands. The absence of Macrocoma from

lowland rainforests, typical of the Guinea-Congolian

Phytogeographic Region of central and western

Africa, probably explains the conspicuous absence of

the species from west African outliers of the

Afromontane vegetation in the Cameroon and

Guinea-Ivory Coast areas.

Ecologically the species are highly adapted to such

habitats with high light intensities and low annual

precipitaion as the African savannas and woodlands.

Specimens are, however, frequently collected in

dense, moist, montane or kloof forests, indicating a

rather wide acceptance of habitats throughout the

Afromontane Region.

The maps illustrating the distributions of African

species of Macrocoma (Figs 2-5) were computer

464

THE PHYTOGEOGRAPHY AND ECOLOGY OF MACROCOMA (ORTHOTRICH ACEAE, MUSCI) IN AFRICA

Fig 1. — Approximate world distribution of the nine species in Macrocoma subgenus Macrocoma. Northern South American

patterns are not well known and here hypothetical as are the western Australian and Chinese distributional limits. The In-

dian populations are probably more widespread than indicated. Data taken only from specimens examined (see Vitt,

1980a for herbaria consulted). Abbreviations: a = abyssinica, b = brasiliensis, f=frigida, g=gastonyi, i = iwatsukii,

1 =lycopodioides, o = orthotrichoides, p = pulchella, s = tenue subsp. sullivantii, and t = tenue subsp. tenue.

plotted. In connection with research on the Flora of

Southern Africa, the Data Section of the Botanical

Research Institute, has written programs that plot

distributions for taxa in southern Africa (Figs 3, 5, &

6) or part of Africa and Madagascar as seen in Figs 2

& 4. Distributional data are entered for each taxon

using the Quarter-Degree Square Reference System

(Edwards & Leistner, 1971; Edwards, 1981). The

mapping program compares entries in the distribu-

tion data file of the taxon to be mapped and plots a

quarter-degree square only once. Therefore, a dot on

the map indicates presence of the taxon in the

quarter-degree square and may represent single or

multiple entries in the data file.

(1) Maerocoma abyssinica (C. Mull.) Vitt —

Ethiopia, Kenya, Malawi, Rwanda, Tanzania, Ugan-

da, Zaire. This species is most closely related to M.

tenue and appears to be endemic to the Afromontane

flora of eastern Africa. Restricted in distribution, M.

abyssinica is most frequently collected in the Usam-

bara and Ruwenzori Mountain areas. A few collec-

tions have come from the Uluguru-Mlanje Moun-

tains to the south and the type represents the only

collection seen from the Ethiopian Mountain system

to the north (Fig. 2).

(2) Macrocoma tenue (Hook. & Grev.) Vitt

subsp. tenue — Angola, Ethiopia, Kenya, Reunion,

Lesotho, Madagascas, Malawi, Rwanda, South

Africa, Swaziland, Tanzania, Transkei, Zaire, Zam-

bia, Zimbabwe. Also known from eastern Australia

and New Zealand. The most widely distributed of the

African species, M. tenue is often collected on trees

and occasionally on rocks and boulders. The species

is frequently encountered in Afromontane forest

remnants within the Fynbos Biome of the southern

and south-western Cape Province. Extending to the

northeast, it becomes more common in the high

forests and grasslands of the Drakensberg and es-

carpment of the eastern Transvaal (Fig. 3). The

species has been collected only occasionally in the

Chimanimani Mountains of Zimbabwe and the Ulu-

guru-Mlanje Mountains of Malawi and Tanzania,

but occurs frequently in the Ruwenzori and Usam-

bara Mountains of eastern Africa. Only a single

specimen is presently known from the mountains of

southern Ethiopia, marking the northernmost exten-

sion of its range.

Outside its major north-south distribution axis, M.

tenue is found in the montane vegetation of Mada-

gascar and Reunion, and in isolated western outliers

ROBERT E. MAGILL AND DALE H. VITT

465

Fig. 3. — Southern African distribution of Macrocoma tenue Fig 5. — African distribution of the endemic Macrocoma

subsp. tenue. lycopodioides.

of the Afromontane vegetation and surrounding

Zambezian woodlands of central Zaire and Angola

(Fig. 4).

(3) Macrocoma lycopodioides ( Schwaegr .) Vitt

— South Africa, Swaziland, Transkei. Presently

known only from eastern and southern parts of

southern Africa, M. lycopodioides is less common

than M. tenue. The two species are practically sym-

patric in southern Africa and are often intermixed.

Macrocoma lycopodioides is easily differentiated

from its closest relative, M. tenue, by having fragile

leaf apices. Specimens of M. lycopodioides are occa-

sionally collected in relic montane forest within the

Fynbos Biome of the southern and south-western

Cape, as well as forests and woodlands of the eastern

Cape, Transkei and Natal. Extending northward into

the savannas or forests of the Orange Free State,

Swaziland and the escarpment of the eastern Trans-

vaal, the species reaches the northernmost part of its

range in the woodlands of the Soutpansberg in the

northern Transvaal (Fig. 5).

(4) Macrocoma pulchella ( Hornsch .) Vitt —

South Africa. Endemic to South Africa, this species

is known from only a few collection sites in the

south-western Cape Province. Very little is known

about its distribution or ecology, but it also appears

to be restricted to areas of relic montane forest within

the Fynbos Biome. It is interesting that the areas

where M. pulchella has been collected (Table Moun-

tain, Devil’s Peak and Jonkershoek) receive some of

the highest precipitation in southern Africa. The spe-

cies is probably most closely related to M. tenue,

however, the fused, 2-layered, well-developed peri-

stome and unique gametophytic features (see Vitt,

1980b) make this a very distinctive species (Fig. 6).

Fig 6.— African distribution of the endemic Macrocoma

pulchella.

ACKNOWLEDGEMENTS

We gratefully acknowledge the National Science

and Engineering Research Council of Canada for

grant A6390 to the junior author for partial support

466

THE PHYTOGEOGRAPHY AND ECOLOGY OF MACROCOMA (ORTHOTRICH ACEAE, MUSCI) IN AFRICA

of this research. African distributions are mapped

based on specimens in BOL, BM, BR, C, COLO,

DUKE, EGR, H, L, MO, NAM, PRE, & Herb.

Townsend (KEW), and we wish to thank the curators

of these herbaria for loan of specimens. We also wish

to thank the Data Section (BRI) and especially Dr H.

F. Glen for access to the computer programs.

UITTREKSEL

Van die 11 spesies wat tans in die genus Macro-

coma ( Orthotrichaceae , Musci) erken word, kom vier

in Afrika voor. Hierdie spesies is xeromorfies en

word in savannas, bosse en woude aangetref, min of

meer ooreenkomstig die Afromontane Plantgeorgra-

fiese Stree. Macrocoma tenue (Hook. & Grev.) Vitt

subsp. tenue is wydverspreid in suidelike Afrika,

noordwaarts tot in Ethiopie, en word ook in Austra-

lasia aangetref. Drie spesies is endemies, waarvan M.

pulchella ( Hornsch .) Vitt slegs in die Kaapse gebied

voorkom; M. abyssinnica (C. Mull.) Vitt in oostelike

Afrika aangetref word, en M. lycopodioides

( Schwaegr .) Vitt sporadies versprei is in die ooste en

suide van Suid- Afrika.

REFERENCES

Brotherus, V. F., 1901-1909. Musci. In A. Engler & K. Prantl,

Die Natiirlichen Pflanzenfamilien 1(3, I, II): 277-1246.

Leipzig: W. Engelmann.

Edwards, D. & Leistner, O. A., 1971. A degree reference system

for citing biological records in southern Africa. Mitt. bot.

Stsamml., Munch. 10: 501-509.

Edwards, D., 1981. A note on the extension of the Degree

Reference System for citing biological distribution records to

north of Equator and west of Greenwich meridian. Bothalia

13: 574-575.

Grout, A. J., 1944. Preliminary synopsis of the North American

Macromitriae. Bryologist 47: 1-22.

Muller, C., 1845. Synopsis Macromitriorum hactenus cogni-

torum. Bot. Ztg 3: 521-526, 539-545.

Sim, T. R., 1926. The bryophyta of South Africa. Trans. R. Soc.

S. Afr. 15: 1-475.

Vitt, D. H., 1973. A revisionary study of the genus Macrocoma.

Revue bryol. lichen. 39: 205-220.

Vitt, D. H., 1980a. The genus Macrocoma I. Typification of

names and taxonomy of the species. Bryologist 83: 405-435.

Vitt, D. H., 1980b. The nomenclature and taxonomy of Macro-

coma lycopodioides (Schwaegr.) Vitt. J. Bryol. 1 1: 219-229.

Werger, M. J. A., 1978. Biogeographical division of southern

Africa. In M. J. A. Werger, Biogeography and ecology of

southern Africa 1: 147-170. The Hague: W. Junk.

Bothalia 13, 3 & 4: 467-472 (19b.)

Phytogeography and speciation in the vegetation of

the eastern Cape

G. E. GIBBS RUSSELL* and E. R. ROBINSON**

ABSTRACT

The eastern Cape is a region of variable environmental factors, with a flora estimated at about 3 600-4 000

species and encompassing 21 of Acocks’s (1975) veld types. It lies at the edges of the major phytochoria present in

southern Africa, with many tropical species reaching the southern and western limits of their distribution, and many

south-western Cape and Karoo species reaching the northern and eastern limits of their distribution. The apparently

low incidence of species endemic to the eastern Cape may be the result of selection for ‘generalist’ genotypes and the

close proximity of different phytochoria, which may allow species to migrate between phytochoria to fill niches

resulting from environmental change.

RESUME

PHYTOGEOGRAPH IE ET SPEC! A TION DE LA VEGETA TION DU CAP ORIENTAL

Le Cap oriental est une region de milieux a facteurs variables, avec une flore estimee a environ 2 600-4 000

especes et renfermant 21 des types champetres Acocks ( 1975). II repose aux bords de la phytochoria majeure present

en Afrique australe, avec beaucoup d’especes tropicales atteignant les limites australes et occidentales de leur

distribution, et beaucoup d’especes du Cap sud-occidental et du Karoo atteignant les limites septentrionales et orien-

tates de leur distribution. L ’incidence apparemment faible d ’especes endemiques au Cap oriental peut etre le result at

de selection pour genotypes ‘generaliste’ et la proximite etroite de differents phytochoria, lesquels peuvent permettre

aux especes d’emigrer entre phytochoria pour remplir des niches resultant de changement de milieu.

INTRODUCTION

The eastern Cape, here defined as the area south of

32°S and between 24°E and the Great Kei River, has

long been known as an area rich in species and com-

munities and one of phytogeographical interest and

complexity. To date, however, little systematic work

has been done and the flora remains poorly

understood. This paper presents data illustrating

species and community diversity within the eastern

Cape, and between it and other areas of southern

Africa, and discusses some aspects of speciation in

the eastern Cape flora.

DESCRIPTION OF THE AREA

The region covers an area of about 88 000 knr,

and is climatically and topographically very variable.

Much of the range of climate and many landforms

found elsewhere in southern Africa can be en-

countered in the eastern Cape.

The land rises from sea level in the south and

south-east to about 2 100 m in the north-west and

1 500 m in the north-east (Fig. 1). There are two ma-

jor mountain ranges, the Winterberg roughly in the

centre of the area (with a maximum height of 2 360

m), and the coastal ranges north and west of Port

Elizabeth, which are about 1 500 m high. The terrain

is much dissected by numerous small and a few large

rivers, so that there is little flat country, except in the

west, and this naturally produces marked environ-

mental heterogeneity over short distances.

Rainfall varies in amount and seasonality. As a

general rule, amount decreases from the coast inland

and from east to west (Fig. 1), with the highest pre-

cipitation being recorded from the southern slopes of

the central mountain ranges. An appreciable propor-

tion of the area is semi-arid. Furthermore, over much

*Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

"""Department of Botany, University of the Witwatersrand,

Johannesburg, 2001.

of the area rainfall is unreliable and droughts are not

uncommon. Of greater phytogeographical signifi-

cance, is the seasonal pattern of rainfall. In the

southern third of the eastern Cape, a considerable

proportion of the rain falls in winter, whereas in the

north-eastern part the rainfall has a summer pattern;

therefore over much of this region there can be some

rain at nearly any time of the year. Also, the line

denoting equal probability of spring or autumn rain-

fall runs through the northern part of the area. Some

places receiving summer rains have a bimodal distri-

bution of rain, with peaks in late spring and autumn.

A further aspect of rainfall distribution which adds

to environmental heterogeneity is illustrated in Fig.

2. Here the month-by month movement of the 50 mm

rainfall isohyet is plotted for the summer rainfall

region, and it can be seen that it converges on the

area of Grahamstown, with the result that in the cen-

tre of the eastern Cape there is a concentration of

seasonal rainfall changes into narrow' bands.

(Weather Bureau, 1965).

Cape.

468

PHYTOGEOGRAPHY AND SPECIATION IN THE VEGETATION OF THE EASTERN CAPE

Fig. 2. — Seasonal movement of the 50 mm rainfall isohyet in the

summer rainfall area, showing its convergence on the eastern

Cape (Weather Bureau, 1965).

Temperatures also vary from the coast inland, with

summer mean daily minima of 14-19°C and maxima

of 25-32°C, and winter mean daily minima of

2-10°C and maxima of 16-21 °C. The more extreme

temperatures occur in the north and west of the area.

Frost occurs in winter at higher altitudes, particularly

at sites where temperature inversions occur. Snow

has been recorded occasionally from a few localities

at low altitude (e.g. Grahamstown in 1976) and is

rather regular in mountainous parts of the area

(Weather Bureau, 1965).

Geologically much of the eastern Cape is underlain

by Karoo System rocks (Haughton, 1963). These

consist mainly of shales and sandstones of the Beau-

fort and Ecca Series, with dolerite dykes and intru-

sions common to the whole Karoo System. Silcrete

‘cappings’ produce many of the highest hills. In addi-

tion, a number of outcrops of Witteberg Quartzite of

the Cape System are found and their occurrence is

biogeographically very important. Even small and

isolated outcrops, such as that forming Mount Coke

near King William’s Town, support ‘islands’ of fyn-

bos vegetation of the Cape Floral Kingdom deep

within vegetation of tropical origin.

The soils derived from these rocks differ con-

siderably. The most noticeable fine-scale variations

attributable to the underlying rocks are those produc-

ed by the dolerite intrusions and the quartzites.

Change from one soil type to another can be abrupt,

and because the soil type can affect water availability

there is considerable interaction between climatic and

edaphic parameters, producing a wide range of habi-

tats over short distances.

A further environmental factor which must be

mentioned is the impact of man. Over the past 100

years or so European farming has changed the pat-

tern of utilization of the vegetation and has intro-

duced many alien species. There are records that by

1705 the Xhosa had settled on the banks of the Great

Fish River and were in possession of large herds of

cattle (Soga, 1931). Grazing and overgrazing by

sheep, goats and cattle, and more recently the intro-

duction of alien plants, must have had considerable

influence on species composition and plant specia-

tion in the area.

FLORISTICS, PHYTOGEOGRAPHY AND

PHYTOSOCIOLOGY

Florist ics and phytogeography

To date there is no complete flora for the area. A

preliminary check list is in preparation, and extra-

polation from this list and Martin & Noel’s (1960)

Flora of Albany and Bathurst gives an estimated

number of 3 600 to 4 000 vascular plant species. This

estimate agrees reasonably well with the number of

species in areas of similar size, such as Natal (Gibbs

Russell, 1975; Oliver, 1977). Floristic information

for the present paper has been obtained from mono-

graphs and other works dealing with various groups,

and data collected by the authors in the course of

other studies.

The eastern Cape has long been known to botanists

as an area where major vegetation units meet in

southern Africa. Here the Indian Ocean Coastal Belt,

Zambezian Domain, Afromontane, Karoo-Namib

and Capensis phytochoria of the sub-continent are

adjacent to one another (Werger, 1978; Goldblatt,

1978; U.N.E.S.C.O./A.E.T.F.A.T., 1980). At many

sites the mixing is so intimate that species of different

phytochoria intermingle in a single stand of vegeta-

tion. Detailed examination of distributions of taxa

show clearly that the elements composing the eastern

Cape flora have their centres elsewhere (Tables 1-3).

The distribution of both grasses and trees (Table 1)

reflects their tropical origin. The majority reach their

south-western limits in the eastern Cape and range to

the north and east. A minority of species are of

southern derivation and extend no further north than

our area. Nearly all the species that extend in both

directions from the eastern Cape are those with ex-

tremely wide ranges in southern Africa. Taxa

characteristic of the south-western Cape (Table 2)

and the Karoo (Table 3) also show a drop in numbers

of species across our area, with many taxa that come

as far as the eastern Cape but go no further to the

north or east. In general then, the tropical elements

of the southern African flora extend no further to the

south and west than the eastern Cape, and the south-

western Cape and Karoo species do not extend fur-

ther to the north and east. The eastern Cape,

therefore, is a region where many taxa of diverse

phytogeographical units reach their limits of distribu-

tion. However, the flora of the region has apparently

a rather low proportion of endemics (Table 4), sug-

gesting that there has been little speciation here in the

recent past.

It is not possible to determine the number of

endemic species in a flora whose composition is not

yet well known. However, when the number of

endemics are found for certain plant groups in the

eastern Cape it is shown in Table 4 that the percent of

endemic species in these groups falls far below the

percent of endemic species in areas well known for

high levels of endemism. Furthermore, extracting

data from Goldblatt’s (1978) list of genera endemic

to southern Africa, it is seen (Table 5) that the

numbers of species per genus is much lower for

genera restricted to the eastern Cape than for

endemic genera in all of southern Africa. Moreover,

the high number of species per genus in endemic

genera which reach the eastern Cape, but have the

greater part of their distribution elsewhere, em-

phasizes once again that the eastern Cape is on the

edge of major distribution patterns.

In summary, the diversity of taxa in the eastern

Cape is due to combinations of species from different

phytochoria meeting at the ends of their ranges, and

not a result of speciation taking place in the area.

Phytosociology

Plant community data complement the phytogeo-

graphical picture. Acocks (1975) describes 70 veld

G. E. GIBBS RUSSELL AND E. R. ROBINSON

469

TABLE 1. — Distributions of ‘tropical’ taxa

TABLE 4. — Percentages of endemic taxa

%

Eastern Cape

Southern African endemic genera

(Goldblatt, 1978) 5

Grasses (Chippindall, (1955) 5

Aloe (Reynolds, 1950) 6

Gladiolus (Lewis, Obermeyer & Barnard, 1972) 0

Crassulaceae (Tolken, 1977) 18

Oxalis (Salter, 1944) 3

Trees (Coates Palgrave, 1977) I

All seed plant species in published volumes

of Flora of Southern Africa

(Dyer el a!., 1963; Codd et at., 1966,

1970; Ross 1975, 1976, 1977;

Leistner, 1979, 1980) 5

Other areas, of high endemism

SW Cape (Weimarck, 1941) 83

Cape Floristic Region (Goldblatt, 1 978) 73

Namib Desert (Robinson, 1978) 35

All of southern Africa (Goldblatt, 1978) 80

types, which are plant communities or aggregations

of communities. Of the 70 veld types, 21 (30%) occur

in the eastern Cape. Therefore a region with only

about 6,5% of the area of South Africa has more

veld types than any other single region (the next most

diverse areas of comparable extent are Natal with 16

and the south-western Cape with 10 veld types). Fur-

thermore, no single veld type extends both south-

west and north-east from the eastern Cape. Acocks

only considered six of the 21 veld types to be restric-

ted to the eastern Cape, and four of these he terms

‘false’ (man-induced). Edwards’s (1977) map of

South African biomes shows the same pattern. Eight

of the 1 1 biomes occur in the eastern Cape, not one

of which is restricted to the area. All the biomes ex-

tend into our area either from the west or from the

north and east, and none extends in both directions

from the eastern Cape.

As well as the spatial variation in communities, the

eastern Cape vegetation is unstable through time. It

is all too well known by agriculturalists that many

communities are prone to rapid changes in floristic

470

PHYTOGEOGRAPHY AND SPECIATION IN THE VEGETATION OF THE EASTERN CAPE

TABLE 5. — Numbers of species in genera endemic to southern Africa (Goldblatt, 1978)

TABLE 6. — Numbers of weed species in various areas of South

Africa (Henderson & Anderson, 1966)

TABLE 7. — Weeds, aliens and species of disturbed sites in Albany

composition as soon as they are disturbed. The data

of Henderson & Anderson (1966) and Martin & Noel

(1960) show clearly how readily weeds and alien

species become established over much of our area.

(Tables 6 & 7). This suggests that many communities

may be very sensitive to changes. Of perhaps even

greater interest is the way elements of different

phytochoria (for example, Pteronia incana and

Elytropappus rhinocerotis of the winter rainfall area,

Pentzia incana and Felicia spp. of the Karoo, and the

tropical species of the scrub woodland communities)

can invade communities of a range of phytochoria.

Although at least some of these invasions are in-

itiated or encouraged by man’s activities, the fact

that elements of different regions can replace one

another emphasizes the dynamic nature of the vege-

tation as a whole and gives a clue to the possible

reason for the apparent lack of speciation in the area.

FACTORS PROMOTING SPECIATION

At this point it is worth reviewing briefly the kinds

of environmental and other factors that lead to frag-

mentation of gene pools and subsequent divergence

of sub-populations under various selection pressures.

Data for the eastern Cape flora can then be examined

in the light of prevailing environmental conditions

and the currently accepted theories of the kinds of

environment which favour micro-evolution.

The fragmentation of populations into sub-units as

a factor likely to promote speciation has been widely

accepted since it was proposed by Wright in 1931. If

the sub-units are sufficiently isolated from one

another to become differentiated by selection

pressures but there is still limited gene flow to retain

variability, it may be expected that species complexes

will develop (Stebbins, 1972). This condition will be

found in regions of marked, fine-scale environmental

heterogeneity. Related to this is environmental, par-

ticularly climatic, instability over time. Changes in

climate can serve to restrict populations to favour-

able areas, thereby fragmenting the gene pool. As

pointed out by Stebbins (1952, 1972 & 1974), as far as

higher plants are concerned, water availability is here

of critical importance. Raven (1964) showed the ef-

fect of edaphic variability and the way it will interact

with climatic fluctuation.

The above conditions together underlie the idea

that speciation is likely to be rapid at the limits of

species distribution ranges. Here, marginal popula-

tions are likely to become genetically isolated from

the general gene pool and may undergo differentia-

tion. This idea was elegantly proposed by Valentine

(1967) under the title of the ‘species pump’ hypo-

thesis, and has been discussed at some length by Steb-

bins (1974).

By taking the above points into account, we can

describe one sort of environment in which speciation

should be actively occurring. This would be a place in

which there is considerable variation, in space and

time, of a variety of parameters (climate, soil, land-

scape, etc.) and where many taxa are nearing the

limits of their geographical ranges. In addition, the

balance of evidence suggests that the climate should

be semi-arid (but see Simpson, 1977, for a discussion

of speciation in tropical forests). From the descrip-

tion of the physical environment and the vegetation,

the eastern Cape seems to meet the above conditions

for speciation. However, the flora does not show the

expected pattern of diversity in terms of numbers of

endemic species.

DISCUSSION

When one looks at the species and communities of

the eastern Cape, three features are apparent. First,

physiographic and climatic variability result in

spatial and temporal heterogeneity on a fine scale.

This allows many different species and vegetation

types to exist in close proximity to one another and at

the same time results in instability of the communi-

ties. Second, the distribution ranges of a large

number of taxa end here, confirming the recognition

of the eastern Cape as the boundary for a number of

phytochoria. Third, neither taxa nor syntaxa seem to

G. E. GIBBS RUSSELL AND E. R. ROBINSON

47]

have become differentiated enough to be recognized

as separate despite the operation of factors that else-

where lead to speciation or the development of

distinct plant communities. There are relatively few

plant taxa that are endemic or restricted to the area,

and of the six veld types restricted to the eastern Cape

only two (Alexandria Forest and Eastern Province

Thornveld) are not the result of invasion apparently

initiated by man. It must be stressed that much more

taxonomic and syntaxonomic information is needed

for the area. At present neither the flora nor the com-

munities are adequately treated in any work and the

data presented above are crude. In spite of this, it

may be in order to examine the situation a little more

closely, if only to suggest possible directions for fur-

ther investigation.

At present there is only scanty palaeoclimatic or

floristic data for the eastern Cape, so we have no real

idea of how long the conditions now experienced in

this area have continued. However, when one con-

siders the changes that have occurred over southern

Africa [see for instance Livingstone (1975), Tankard

& Rogers (1978) and Lancaster (1979)], and the

decade to century long fluctuations commented on

by May (1979), there can be little doubt that the

eastern Cape climate must have been unstable for a

very long time. As indicated earlier, the area is the

boundary between the winter and summer rainfall

zones, so any change in macroclimate over southern

Africa would lead to a dramatic change in condi-

tions. However, on the west coast of southern Africa

there is a region which is climatically similar in some

ways to the eastern Cape in that the summer and

winter rainfall zones meet and there is a variety of

habitats. In this area of the north-western Cape and

southern Namib a number of taxa show active speci-

ation (Robinson, 1978, Moffett, 1979).

Why then has the eastern Cape not developed a

distinctive flora as a result of active speciation? It is

possible that the nature of the dominant life forms,

the variability of the environment and the position of

the eastern Cape between the major phytochoria

have all resulted in selection for ‘generalist’ geno-

types rather than producing ‘specialists’. Hamrik et

al. (1979) have shown that woody plants and wind

pollinated plants (grasses) show greater genetic diver-

sity within a species on average than do herbaceous

and entomophilous ones. Hedrik et al. (1976) have

shown that heterogenous environments encourage or

maintain high levels of genetic diversity in taxa. Fur-

thermore, it would be expected that of the species

making up a phytochorion, the generalists would be

more likely to extend to the ends because of their

wider tolerances. Therefore, the species in the eastern

Cape may tend to be the most genetically diverse

representatives of their respective phytochoria.

In other areas where speciation is rapid, it is single

species or communities that are reaching the ends of

their ranges, whereas in our area it is phytochoria

(i.e., assemblages of many species and communities)

that are reaching their distribution limits. It is impor-

tant to remember that the species making up the dif-

ferent phytochoria have evolved under very different

selection pressures. This may mean that in an area

such as the eastern Cape where a number of

phytochoria are adjacent to one another there is a

species already present (perhaps most likely a

‘generalist’) which can occupy virtually any niche

that becomes available. Therefore a change in en-

vironmental factors will simply lead to movements ot

species populations to new sites rather than the

evolution of new genotypes. This idea is supported

by the effects of fire on macchia vegetation reported

by Trollope (1973) and Downing et al. (1978). In the

Amatole Mountains fynbos species ( Cliffortia

linearifolia, C. paucistaminea and Erica brownleeae)

form dense closed communities. When such fynbos is

burned on a regular basis the community changes to

grassland dominated by tropical grasses of the genera

Themeda, Tristachya and Panicum. At the com-

munity level, although Acocks’s treatment is too

broad to allow more detailed discussion, it may be

significant that the ‘false’ veld types of the eastern

Cape are the result of Cape and Karoo phytochoria

invading disturbed areas in phytochoria of tropica!

origin.

This reasoning leads to the conclusion that there is

a saturation of the environment by species in such a

region, which is apparently contradicted by the large

number of weeds and aliens found in most com-

munities in our area. However, most of the weedy

species are annuals, a life form that is not well

represented in Africa generally, and there may be

niches, particularly those produced by man’s ac-

tivities, to which alien taxa are better adapted than

any indigenous species.

CONCLUSION

The eastern Cape is floristically rich, with an esti-

mated 3 600-4 000 species, and is phytogeographic-

ally and phytosociologically complex, and yet there

are few endemic or restricted taxa or vegetation

types. We suggest two hypotheses to explain this

situation. First, selection pressures, particularly

climatic instability, have acted to produce a flora in

which ‘generalist’ genotypes predominate; and se-

cond, close proximity of phytochoria of different

evolutionary histories ensures that somewhere there

is a species already present that can fill, by migration,

any new niche which may result from environmental

change. Studies to confirm these hypotheses would

be of great interest not only to evolutionary biolo-

gists but also to agriculturalists because of the prac-

tical implications. Much of the eastern Cape is only

marginally suitable for agronomy and the natural

vegetation is therefore of great significance for

agricultural production. Management of such inher-

ently unstable ecosystems for sustained productivity

demands detailed knowledge of the environment.

The information given in this paper shows that the

eastern Cape is a very complex place and that our

understanding of its vegetation is far from complete.

Studies of the species and communities making up

the vegetation will greatly increase the chances that

management plans can be made that will take ac-

count of the long-term variability which this area ex-

hibits.

ACKNOWLEDGEMENTS

Many of the ideas expressed here were developed

while both authors were engaged in various research

and teaching posts at the University of Fort Hare and

we therefore make acknowledgement to this institu-

tion; and especially to Mr W. S. W. Trollope, who

has kept us always conscious of the need for practical

application of biological study.

U1TTREKSEL

Die Oos-Kaap is ’n gebied met veranderlike omge-

wingsfaktore, ’n flora wat na beraming 3 600-4 000

spesies bevat en 21 van Acocks (1975) se veldtipes in-

472

PHYTOGEOGRAPHY AND SPECIATION IN THE VEGETATION OF THE EASTERN CAPE

sl mt. Dit le op die grense van die hoof fitochoria wat

in suidelike Afrika verteenwoordig is, verskeie

tropiese spesies wat die said- en westelike grense van

hulle verspreiding bereik kom voor, terwyl baie suid-

westelike Kaap en Karoo spesies wat die noord- en

oostelike grense van hulle verspreiding bereik, daarin

gevind word. Die skynbaar lae voorkomssyfer van

spesies endemies in die Oos-Kaap kan wees as gevolg

van seleksie vir ‘algemene’ genotipes in die teenwoor-

digheid van verskillende fitochoria waar immigrasie

van spesies tussen fitochoria kan plaasvind om so-

doencle nisse te vul wat as gevolg van veranderings in

omgewingstoestande kan ontstaan.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed. Mem.

bot. surv. S. Afr. 40: 1-128.

Adamson. R. S. & Salter, T. M., 1950. Flora of the Cape Penin-

sula. Cape Town: Juta.

Baker, H. A. & Oliver, E. G. H., 1968. Ericas in southern Africa.

Cape Town: Purnell.

Chippindall, L., 1955. A guide to the identification of grasses. In

D. Meredith, The grasses and pastures of South Africa ,

Part 1. Johannesburg: C.N.A.

Coates Palgrave, K., 1977. Trees of southern Africa. Cape

Town: Struik.

Codd, L. E., De Winter, B. & Killick, D. J. B., eds, 1970. Flora

of southern Africa. 13: 1-221.

Codd, L. E., De Winter, B. & Rycroft, H. B., eds, 1966. Flora of

southern Africa. 1: 1-166.

Downing, B. H., Robinson, E. R., Trollope, W. S. W. & Morris,

J. H., 1978. The influence of macchia eradication techniques

on botanical composition of grasses in the Dohne Sourveld of

the Amatole Mountains. Proc. Grassld Soc. sth. Afr. 13:

111-116.

Dyer. R. A., Codd, L. E. & Rycroft, H. B., eds, 1963. Flora of

southern Africa. 26: 1-307.

Edwards, D., 1977. Biomes of South Africa. Unpublished map

prepared at Botanical Research Institute, Pretoria.

Gibbs Russell, G. E., 1975. A comparison of the size of various

African floras. Kirkia 10,1: 123-130.

Goldblatt, P., 1978. An analysis of the flora of southern Africa:

its characteristics, relationships and origins. Ann. Mo. bot.

Gdn 65,2: 360-436.

Hamrik, J. L., Linhart, Y. B. & Mitton, J. B., 1979. Relation-

ships between life history characteristics and electrophoreti-

cally detectable genetic variation in plants. Ann. Rev. Ecol.

Syst. 10: 173-200.

Haughton, S. H., 1963. Stratigraphic history of Africa south of

the Sahara. Edinburgh: Oliver & Boyd.

Hedrik, P. W., Ginevan, M. E. & Ewing, M. P., 1976. Genetic

polymorphism in heterogenous environments. Ann. Rev.

Ecol. Syst. 7: 1-32.

Henderson, M. & Anderson, J. G., 1966. Common weeds in

South Africa. Mem. bot. Surv. S. Afr. 37: 1-440.

Herre, H., 1971. The genera of the Mesembryanthemaceae. Cape

Town: Tafelberg.

Jacot Guillarmod. A. M. F., 1971. Flora of Lesotho. Lehre:

Cramer.

Lancaster, 1, N., 1979. Evidence for a widespread late

Pleistocene humid period in the Kalahari. Nature, Lond.

279, 5709: 145-146.

Leistner, O. A., ed. 1979. Flora of southern Africa 10,1: 1-59.

Leistner. O. A., ed. 1980. Flora of southern Africa 27,4: 1-91.

Lewis, G. J., Obermeyer, A. A. & Barnard, T. T., 1972. Gladiolus

a revision of the South African species. Jl S. Afr. Bot. Suppl.

to Vol. 10: 1-315.

Livingstone, D. A., 1975. Late Quarternary climatic change in

Africa. Ann. Rev. Ecol. Syst. 6: 249-280.

Martin, A. R. H. & Noel, A. R. A., 1960. The flora of Albany

and Bathurst. Grahamstown: Rhodes University.

May, R. M., 1979. Arctic animals and climatic changes. Nature

Lond. 281, 5728: 177-178.

Moffett, R. O., 1979. Sarcocaulon and the Gariep: products of a

hostile environment. Paper presented to the 5th Annual Con-

gress of the South African Association of Botanists, 24-27

January, 1979, Stellenbosch.

Oliver, E. G. H., 1977. An analysis of the Cape flora. Proceed-

ings of Second National Weeds Conference of South Africa,

1-18. Cape Town: Balkema.

Raven, P. H., 1964. Catastrophic selection and edaphic endem-

ism. Evol. 18(2): 336-338.

Reynolds, G. W., 1950. Aloes of South Africa. Johannesburg:

The Aloes of South Africa Book Fund.

Robinson, E. R., 1978. Phytogeography of the Namib Desert of

South West Africa (Namibia) and its significance to discus-

sions of the age and uniqueness of this desert. In E. M. van

Zinderen Bakker Snr & J. A. Coetzee, eds, Palaeoecology of

Africa and the surrounding islands Vol. 10/11: 67-74. Rot-

terdam: Balkema.

Ross, J. H., 1972. Flora of Natal. Mem. bot. Surv. S. Afr.

37: 1-418.

Ross, J. H., ed., 1975. Flora of southern Africa 16,1: 1-159.

Ross, J. H., ed., 1976. Flora of southern Africa 22: 1-161.

Ross, J. H., ed., 1977. Flora of southern Africa 16,2: 1-142.

Salter, T. M., 1944. The genus Oxalis in South Africa. Jl S. Afr.

Bot. Suppl. to Vol. 1: 1-355.

Simpson, B. B., 1977. Biosystematics and geography. In J. A.

Romberger, ed., Biosystematics in agriculture. Beltsville Sym-

posia in Agricultural Research 2: 151-172.

Soga, J. H., 1931. The Ama-Xosa: life and customs. Alice:

Lovedale Press.

Stebbins, G. L., 1952. Aridity as a stimulus to plant evolution.

Amer. Natur. 86: 33-44.

Stebbins, G. L., 1972. Ecological distribution of centers of major

adaptive radiation in angiosperms. In D. Valentine, ed., Tax-

onomy, phytogeography and evolution. New York: Academic

Press.

Stebbins, G. L., 1974. Flowering plants, evolution above the

species level. London: Edward Arnold.

Tankard, A. J. & Rogers, J., 1978. Late Cenozoic palaeoenviron-

ments on the west coast of southern Africa. J. Biogeogr.

5: 319-337.

Tolken, H. R., 1977. A revision of the genus Crassula in southern

Africa. Contr. Bolus Herb. 8, 1 & 2: 1-595.

Trollope, W. S. W., 1973. Fire as a method of controlling

macchia (fynbos) vegetation on the Amatole Mountains of the

eastern Cape. Proc. Grassld Soc. sth. Afr. 8: 35-42.

U.N.E.S.C.O./A.E.T.F.A.T., 1980. Vegetation map of Africa

(sheet 4). (In press).

Valentine, J. W., 1967. The influence of climatic fluctuations on

species diversity within the tethyan provincial system. In

G. C. Adams & D. V. Ager, eds, Aspects of tethyan bio-

geography. Systematics Assn Publ. 7: 153-166.

Weather bureau, 1965. Climate of South Africa, part 8,

general survey. Weather Bureau WB 28.

Weimarck, H., 1941. Phytogeographical groups, centres and in-

tervals within the Cape Flora. Lunds Universitets Arsskrift,

N. F. Avd. 2 Bd. 37, Nr. 5: 1-143.

Werger, M. J. A., 1978. Biogeographical division of southern

Africa. In M. J. A. Werger, ed., Biogeography and ecology

of southern Africa. The Hague: Junk.

Wright, S., 1931. Evolution in Mendelian populations. Genetics

16: 97-159.

Bothalia 13, 3 & 4: 473-485 (1981)

Information available within the PRECIS data bank of the

National Herbarium, Pretoria, with examples of uses to which

it may be put

J. W. MORRIS* and R. MANDERS**

ABSTRACT

The contents of the computerized information storage and retrieval system (PRECIS) of the National Herbarium,

Pretoria (PRE) are described at length mainly by means of frequency histograms of descriptor codes. The frequency

distributions found are discussed in the light of the history of the herbarium, the geography of the area and the

habits of plant collectors. Two uses of PRECIS are illustrated by example. Firstly, the flowering phenology of

Eragrostis capensis, Themeda triandra and Heteropogon contortus is plotted and, secondly, the route followed by

Dinter in South West Africa/Namibia from December 1933 until March 1935 is described. It is concluded that the

system should be of particular use in revisionary studies, regional floras and biogeographic research.

RESUME

INFORMA TION DISPONIBLE DANS LE DEPOT DE DONNEES PRECIS DE L 'HERBARIUM NA TIONAL

DE PRETORIA, A VEC DES EXEMPLES D’UTILISA TION A VEC LAQUELLE IL PEUT ETRE EMPLOYE

Les contenus d 'un depot d 'informations sur ordinateur et un systeme de recouvrement (PRECIS) de /’ Herbarium

National de Pretoria (PRE) sont decrits tout au tong, principalement au moyen d’histogramme de frequence de

codes de description. Les distributions de frequence trouvees sont discutees a la lumiere de Vhistoire de I’Her-

barium, la geographie de la region et les habitudes des collectionneurs de plantes. Deux utilisations du PRECIS sont

illustrees par des examples. Premierement la phenologie de la fleur cTEragrostis capensis, Themeda triandra et

Heteropogon contortus est situee et, deuxiemement, la route suivie par Dinter au Sud-Ouest africain/Namibia, de

Decembre 1933 jusqu 'a Mars 1935 est decrite. On tire la conclusion que le systeme devrait etre particulierement utile

pour des etudes de revision, les flores regionales et la recherche biogeographique.

INTRODUCTION

Over the past eight years a computerized data

banking system for the National Herbarium (PRE)

has been designed, programmed and data have been

loaded. The entire project has been completed recent-

ly and this communication is intended to describe the

scope of the information contained in the data base,

attempt to account for its shortcomings and to give

examples of results that can be derived from it. A

knowledge of what is available will enable prospec-

tive users to design their enquiries efficiently and will

indicate what can and cannot be extracted from the

data bank. It is not our intention to present new tax-

onomic, floristic or ecological results, based on this

information, but merely to summarize it for the

benefit of future users and to highlight a few of the

possible uses of the system through examples.

The aim of the project was to encode all the infor-

mation from the approximately half million her-

barium specimens collected in the Flora of Southern

Africa Area and housed in the National Herbarium,

Pretoria, and to produce a flexible information

retrieval system by means of which the data could be

made available in a number of useful forms. A

systems analysis (Morris, 1974), progress reports

(Morris & Leistner, 1975; Morris & Glen, 1978) and a

description of the backlog encoding task (Morris,

1980) have traced the development of the system.

The frequency of occurrence of each descriptor on

herbarium specimens, i.e. how many specimens had

an indication of flower colour or notes on habitat,

was unknown when this data bank was designed. The

net was thrown extremely wide with the intention

*Datametrical Services, Department of Agriculture and Fisheries,

Private Bag XI 16, Pretoria, 0001, formerly as**.

**Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

that as much data as possible would be captured in

retrievable form. Of great use to the designers of

future systems will be knowledge of the frequency of

occurrence of descriptors in a herbarium of the size

and scope of ours. It is a great waste of computer

disk storage space if provision is made on every speci-

men record for a particular descriptor which has a

low frequency, i.e. is missing or absent most times,

and more efficient storage procedures can be used in

those cases. Frequency of occurrence of descriptors

for specimens housed in the National Herbarium are

summarised below, together with brief comments on

the distributions found.

Many uses of this and other such computerized

herbarium data banks have been proposed (see e.g.

Crovello, 1972). Two uses which are particularly

suited to this approach are countrywide phenological

studies and the reconstruction of routes followed by

early collectors. We present a preliminary account of

the flowering times of three widespread grass species

and notes on the probable route followed by Dinter

in South West Africa/Namibia on his last visit to the

territory as examples of the kind of information that

is now available from PRECIS.

CONTENTS OF THE DATA BANK

General

At the time when the statistics given below were ex-

tracted, 496 909 backlog specimens had been loaded

into the data bank. Although the system makes pro-

vision for the continuous addition of new collections

(see Morris & Glen, 1978) and about 10 000 are being

accessioned annually, those that had been added by

means of the new plant collection form were

deliberately excluded from this analysis. Some 4 100

additional backlog specimens, still to be loaded at the

47 4 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OFTHE NATIONAL HERBARIUM, PRETORIA

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

time, have since been accessioned. We consider it

unlikely that these omitted specimens will signifi-

cantly alter the results presented.

A complete list of descriptors and codes is given in

Morris (1980). This list should be consulted in con-

junction with this paper.

Geographical distribution of specimens

(a) Region

The distribution of specimens by region is given

in Fig. 1. All specimens from the Flora of Southern

REGIONS

5 1 9 |

8853 F2]

1967421

91590

32 1 |

697021

188561 I

1064 1

376451

86680

1383391

Rngo 1 a

Botswana

Cape Province

Lesotho

Natal

0. F. S.

Rhodes i a

S. w. n.

Swaz i 1 and

Transvaal

Fig. 1. — Distribution of specimens by region. Numbers in the left-

hand column indicate numbers of specimens in each region

and bars represent proportion, scaled by the maximum, of

specimens in each region. Similar conventions are used in all

following figures of this kind.

Africa area as well as all type specimens in the her-

barium were encoded, explaining the relatively small

number of specimens from Angola, Mozambique

and Rhodesia/Zimbabwe in the herbarium. It is sur-

prising that there are more Cape specimens housed in

the herbarium than there are Transvaal ones as the

herbarium is situated in the Transvaal and another

three large herbaria serve the Cape Province.

Although the Cape has a longer botanical history

than the Transvaal, it is only during the latter parts

of this century that large numbers of specimens have

been collected so that this cannot be advanced as a

reason for this disproportionate distribution of speci-

mens. Natal and South West Africa/Namibia are

better represented than the Orange Free State and

neighbouring independent countries (Swaziland,

Lesotho and Botswana).

An indication of the intensity of collection may be

obtained from a study of the number of specimens

collected per square km of each region. In Fig. 2 pie

charts are used to indicate the number of specimens

Fig. 2. — Pie charts of: a, number of specimens per region; b, areas

of regions; c, number of specimens per square km of each

region. (A = Botswana, B = Cape Province, C = Lesotho,

D = Natal, E = Orange Free State, F = South West Africa,

G = Swaziland and H = Transvaal.)

per country and province of South Africa falling

within the Flora of Southern Africa Area (Ross et ai,

1977), the area of that geographic unit and the

number of specimens collected per unit area of the

region. The largest regions, Botswana, Cape Prov-

ince and South West Africa, are less collected on the

basis of number of specimens per unit area than

Transvaal, Natal and Swaziland, the last two being

very small in area compared with most other regions.

Botswana, South West Africa and the Orange Free

State are particularly undersampled by this criterion.

(b) Grid references

Grid references have only been given to specimens

on a regular basis since the development of the

quarter degree grid reference system by Edwards &

Leistner (1971) and thus we anticipated that a large

proportion of the collection would be without this

useful geographic identifier. Of the total, 10,5% had

codes when they were accessioned. Many more

specimens with grids have been collected in the

Transvaal than in any other region. An attempt is be-

ing made at present to automatically assign grids to

the specimens without grids on the basis of region

and major and minor locations.

(c) Major and minor locations

A total of over 41 300 major and minor locations

have been used by collectors. A page of examples of

localities is given in Table 1. It is our intention to

remove synonymous names and correct spelling

mistakes in the collector and locations files in due

course.

Curatorial information

(a) Year collected

The dramatic, almost linear, increase in the

number of specimens accessioned annually after 1880

is illustrated in Fig. 3. Factors contributing to the

rapid increase include the appointment of J. Burtt

Davy as Government Botanist in Pretoria in May

1903 and the establishment of the Botanical Survey

Advisory Committee by I. B. Pole Evans during 1918

(Dyer, 1977). The former lead to the foundation of

what is now known as the National Herbarium

(PRE) and the latter to the appointment of ecologists

who have collected widely as part of their brief. The

effect of World War II on collections is clearly shown

by the drop in accessions from 1936 to 1945. During

the five-year period before encoding began (1971-

1975) about 9 600 specimens were collected and

accessioned annually. The decrease in the last period

plotted (1976-1980) is attributed to curatorial activi-

ties being concentrated on the preparation of the her-

barium for encoding rather than on the accessioning

of new material.

The oldest specimens housed in the herbarium in-

clude about 150 collected by Burchell between 1810

and 1815. Ecklon, Zeyher and Drege are responsible

for the relatively high peak in collecting from 1825 to

1830 and the continuation of activity until 1840 (Fig.

3). These three collectors are represented by a total of

8 000 specimens in the herbarium, spanning the years

1 826 to 1 850. Alexander, who later changed his name

to Prior, collected 1 100 specimens (housed in PRE)

from 1846 to 1850. The collections of Wahlberg ac-

count, in part, for the peak during the five-year

period 1856-1860.

J.W. MORRIS AND R. MANDERS

475

TABLE 1. Examples ot localities trom PRECIS. The columns on the right give the number of times the name is used

as a major and minor location, respectively

(b) Contributions from the Cape and Transvaal to

the herbarium

It has been shown above that there are more Cape

Province specimens in the herbarium than Transvaal

ones, even though the herbarium is situated in, and

primarily serves, the Transvaal. In Fig. 3, the propor-

tions of Cape and Transvaal specimens to the total

over five-year periods are plotted. Until 1915 there

are proportionally far more Cape specimens than

Transvaal ones in the herbarium. At that time the

proportions from both provinces stabilize at about

0,4 each and then decrease slightly to between 0,2

and 0,4 at the present time. For nearly all periods

after 1915 the proportion of Cape specimens still re-

mains higher than that of Transvaal specimens. Ex-

ceptions are from 1916-1920 and 1941-1945, both

times of world war. The gradual decreases in both

proportions in recent years suggest the increasing

accession of specimens from other regions. The

relatively low proportion of Cape specimens during

1871-1875 is partly due to active collection by

Rehmann in Natal during that period. It is not

known who was active in the Transvaal during the

period 1886-1890 and responsible for the large pro-

portional contribution of Transvaal specimens dur-

ing that period, but the collections of Galpin from

Barberton (Phillips, 1930) will undoubtedly con-

tribute to the total.

(c) Month of collection

A clear summer peak in collecting activity is seen

from the distribution of specimens by month of col-

476 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OF THE NATIONAL HERBARIUM, PRETORIA

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

1 . 00

, 90

. 80

. 70

. G 0

. 50

.40

. 30

. 20

. 1 0

0 . 00

"0

->

o

T)

O

Fig. 3. — Histogram of number of

specimens accessioned during

successive five-year periods

from 1806. The proportions

of Cape (C) and Transvaal

(T) specimens to the total are

also indicated for each period.

lection (Fig. 4a). Slightly fewer collections than ex-

pected during February may be caused by there being

fewer days for collecting during that month.

Holidays during December may have influenced the

number of collections during that month. The pat-

tern in the Transvaal is similar with a maximum in

January and very little collecting activity from May

until September (Fig. 4b). In the Cape, on the other

hand, where most collecting has been done in the

winter-rainfall part of the Province, a more equable

distribution of specimens by month of collection is

found with a clear peak in September (Fig. 4c).

58624 I

42203 1 I

45229 I

35504 I I

25198 I 1

16320 I I

20662 I I

24671 I 1

41001 1 I

48014 1

52487 I

49287 I

January

February

March

Rpr i 1

May

June

July

August

Sep t embe r

October

No vembe r

December

(d) Distribution of types

A type code is assigned to 2% of the specimens in

the data bank. Nearly half of these specimens are

types (46%), while 26% are isotypes, 15% holotypes

and the other 12% are syntypes, paratypes and lecto-

types (Fig. 5). Only 36 specimens received a neotype

code. It is likely that more type specimens are housed

in the herbarium. Those reflected here are those iden-

tified as such on labels and exclude those that have

not yet been annotated.

(e) Duplicates

When more than one sheet of a specimen (i.e. of

the same collection) was found together by the en-

coders, the number of duplicates was recorded. This

information is required by the curator so that he will

know how much space in cabinets is taken up by

duplicates. Two percent of the collection have more

than one sheet filed. Eighty-one percent of these

specimens have two duplicates and very few speci-

mens are represented by more than three duplicates.

(f) State of specimen

Ninety-seven percent of the specimens received a

flower code for state of specimen. Seventy-three per-

cent of these specimens had mature flowers, 15%

were coded with flowers present, 11% with flowers

absent and only 1% of the specimens had immature

flowers (Fig. 6a). A surprisingly large proportion of

specimens have flowers (89% including both imma-

ture or mature flowers). Ninety-five percent of the

specimens have a code for fruit (Fig. 6b). Immature

fruit occur on 57% of the specimens, 15% of the

specimens are coded with fruit present and 20% of

the specimens have mature fruit. Only 8% of the

21885 I I

16364 I I

13994 I 1

964 1 I ~l

5674 I ~~l

3725 I I

3363 I I

3413 I I

6074 I I

12354 I I

18518 I 1

16806 I I

b

16498 I 1 January

10539 1 I February

12397 I I March

12494 I I April

9400 I I May

6648 I I June

9962 I I July

14751 I I Rugust

25881 I I September

22616 I I October

19729 I I November

17047 I I December

c

Fig. 4. — Distribution of specimens by month of collection: a, all

regions; b, Transvaal; c, Cape Province.

1693 I I Holotype

1028 I ~1 Syn/Cotype

176 □ Paratype

36 [ Neotype

108 [] Lectotype

5094 I I Type

2921 I I Isotype

Fig. 5. — Distribution of specimens by type status.

January

February

March

April

May

June

Ju 1 y

August

September

Oc tobe r

No vembe r

Decembe r

J.W. MORRIS AND R. MANDERS

477

5304? I I

3846 D

352075 I

70576 I I

a

Rbsent

Immature

Mature

Present

37268 [

268371 [

36758 [

70031 [

12693 □

475 |

456859 C

589 |

Rbsent

Immature

Mature

Present

c

343044 1 I Absent

139 | Immature

135500 I I Mature

1005 | Present

d

Fig. 6. — Distribution of specimens by state of specimen: a, flower;

b, fruit; c, leaf; d, root.

specimens have fruit absent. A leaf code is assigned

to 97% of the specimens, of which, again 97% have

mature leaves. Leaves are absent from only 3% of

the specimens (Fig. 6c). Ninety-seven percent of the

specimens have a code for presence of roots. Seventy-

two percent of these specimens have no roots and

28% have mature roots.

(g) Language used on label

Eighty-six percent of all the collections have labels

in English in comparison with the 6% which are in

Afrikaans (Fig. 7). Even higher percentages of the

(b) Flower and fruit colour

Thirty percent of the specimens have a flower

colour code and nearly 30% of these have yellow

flowers (Fig. 9a). Twenty-seven percent of the speci-

mens have white or cream flowers, 13% mauve or

purple flowers and 10% have pink flowers. There are

very few specimens with grey, orange or brown

flowers and only 138 specimens are recorded with

black flowers. A flower code is given to over 50% of

the specimens collected in Lesotho (52%) and 50% of

the Swaziland specimens have a flower colour code.

Both these percentages are significantly higher than

the overall average.

A fruit colour code is assigned to only 1% of the

total number of specimens. Thirty percent of these

specimens have red fruits while 24% have green fruits

(Fig. 9b). Only 80 specimens were recorded with blue

fruits and 57 specimens have grey fruits. The fact

that red fruit occurs more frequently than any other

colour fruit can probably be explained by the way

that many seeds and fruit are distributed. Birds and

small mammals are attracted by brightly coloured

fruits, red in particular. Another consideration in

connection with fruit colour is that collectors are apt

not to record usual colours, such as green and brown,

and only note unusual colours.

39529 I

458 D

43278 I

14696 I

2662 □

10401 I

20089 I

8685 I

6101 I

3015 □

138 |

a

Wh i te/c re a

Grey

Ye 11 ou

Pink

Orange

Re d

Mauve

Blue

Green

Broun

B 1 ack

Fig. 7. — Distribution of specimens by label language.

collectors use English in Natal (95%), Botswana

(96%), Lesotho (98%) and Swaziland (98%). Thirty-

two percent of the collectors in South West Africa

use German.

Plant descriptors

(a) Abundance and distribution

Twenty percent of the specimens have an abun-

dance code. Over 50% of these are common, 17%

are very rare and rare, 18% occur occasionally and

12% of the specimens are abundant (Fig. 8). A distri-

bution code (i.e. widespread/local) is assigned to

20% of all specimens collected. Eighty-six percent of

these specimens are coded as widespread and 14%

have a local occurrence.

16850 I I

17410 I I

51131 I

1 1601 I 1

Rare

Occasional

Common

Rbund ant

336 I I

57 □

711 I I

146 □

526 I I

2252 I

344 | |

80 □

1745 I

613 I 1

610 1 1

Nh i te/c re a

Grey

Ye 1 1 ou

Pink

Orange

Red

Mauve

Green

Broun

Black

b

Fig 9. — Distribution of specimens by: a, flower colour code;

b, fruit colour.

(c) Life form

A life form code is given on 25% of the specimens.

Of these, 36% are shrubs (Fig. 10a), 27% herbs and

19% trees. Distribution by life form for specimens

from the Transvaal is similar to that of the total

sample but that for the Cape suggests, as one would

expect, the presence of fewer trees and many more

shrubs in the Cape than in the country as a whole

(Fig. 10b).

(d) Annual/perennial code

Four percent of the specimens have an annual/

perennial code. Over 60% of the specimens are

perennial while 37% are indicated as being annual.

Very few specimens had either ephemeral or biennial

codes. Relatively high percentages of the specimens

are perennial in Natal (86%), the Orange Free State

(82%), Lesotho (81%) and Swaziland (92%). In

Fig 8. — Distribution of specimens by abundance code.

478 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OF THE NATIONAL HERBARIUM, PRETORIA,

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

a

3393 I I

191GB I

2225 I I

5324 I

12 IS I I

1 S3 D

2015 I I

374 □

557 □

110 1

b

Fig 10. — Distribution of specimens by life form code: a, entire

collection; b, Cape specimens only.

comparison with this, 64% of the specimens in South

West Africa are annual, compared with 36% which

are coded as perennial.

(e) Woody / herbaceous code

A woody/herbaceous code is assigned to only 1%

of the specimens. Seventy-eight percent of these

specimens are woody and 22% are herbaceous. No

specimens are coded as half-woody although there is

a code assigned for such a descriptor.

(f) Evergreen /deciduous code

An evergreen deciduous code is given to only 1 214

of the specimens. Of these, 53% are evergreen, 44%

deciduous and only 2% are coded as semi-deciduous.

In Lesotho, 100% of the specimens with this descrip-

tor are evergreen while 100% of the specimens in

Botswana are deciduous.

(g) Weed status code

Only 1% of the specimens have a weed code.

Ninety-eight percent of these specimens are en-

croachers and only 13 specimens are coded as exotic

plants. This latter information is usually obvious

from the name of the plant, accounting for the very

low frequency with which the information was

recorded separately on the label.

(h) Height and diameter at breast height

A height code is assigned to 18% of all the speci-

mens. Forty-four percent of these specimens are in

the range of 0-0,5 m (0-1,6 ft). Another 45% (Fig.

11a) of the specimens occur with a steep and then

gradual decrease in numbers to 3,5 m (11,5 ft). The

remaining 1 1% are higher than 3,5 m. An interesting

fact is that specimens are clustered at certain heights,

indicating rounding off by the collectors of the

heights of the plants that they collect. Peaks in the

number of specimens occur at 4,5 m (about 14 ft),

6 m (20 ft), 7,5 m (25 ft), 9,0 m (30 ft), 10,5 m (35 ft),

12,25 m (40 ft), 13,75 m (45 ft), 15,25 m (50 ft),

17,0 m (55 ft), 18,5 m (60 ft) and at 20,0 m (65 ft). As

most specimens were collected before metrication,

the heights in feet given above possibly indicate the

midpoints of common height estimation ranges used

by collectors.

Nineteen percent of all plants coded as being

‘woody’ have a height code and 22% are in the range

0-0,25 m (0-0,8 ft) and 25% are in the range of

0,26-0,5 m (0,8-1, 6 ft). Once again, there is a sharp

drop in the number of specimens higher than 0,50 m

up to about 3 m (Fig. lib). In the case of woody

plants, even sharper peaks occur at certain heights

than with all specimens together.

In Lesotho, where 17% of the specimens have a

code for height, 73% are in the range 0-0,5 m and

19% in the range 0,51-1,00 m. Only 28 specimens of

height 1 ,01-1,5 m, 31 of height 1 ,51-2,0 m and 13 of

height 2,01-2,5 m have been collected (Fig. 11c).

Almost no specimens had a height code greater than

2,5 m in Lesotho where, anyway, only 60 specimens

are given a code of ‘tree’ for life form.

Diameter at breast height (DBH) is recorded on

only 1 901 specimens, all of which are also coded as

woody. Most specimens have a DBH of less than

0,5 m (Fig. 12) with additional peaks at 1,2-1, 3 m

(about 4 ft) and 1,8-1, 9 m (6 ft). There are 158

specimens with a DBH greater than 2,0 m.

Habitat and vegetation features

(a) Substrate

Only 24% of the specimens have an indication of

substrate type and 88% of these specimens are col-

lected from either soil or stony soil. Bare rock, cliff

faces and beach dunes are equally well represented

(3% each) and in water is recorded for 2 040

specimens (2%). In South West Africa, desert dunes

are recorded on 5% of the specimens and in Natal

cliff face is recorded on 6% of specimens and beach

dunes on 7%.

(b) Moisture regime

Only 12% of the specimens have a moisture regime

code. From the distribution of codes, it is clear that

Tree

Shrub

Dwarf shrub

Herb/grass

Geophyt e

Ep i phy t e

C 1 i mbe r

Paras i te

Succu 1 ent

Hyd rophyte

4892 1 1

4002 | |

3S3I □

3505 | 1

2215 □

329 fl

2052 □

192 |

□

910 0

37 |

91 j

1267 □

39 |

319 0

32 |

570 0

100 |

200 j

35 |

205 |

a

4221 □

4008 □

4220 □

321 |

2221 □

2050 □

80 |

620 D

2202 □

25 |

114 |

909 0

122 |

1203 □

39 |

0.5

1 .0

1 .5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

0.0

0.5

7.0

7.5

28 □

31 □

13 0

9.5

0.0

1 .0

I .5

15.5

17.0

10.5

20.0

] 0.5

1.5

2.0

2.5

3.0

3.5

4 . 0

4.5

5.0

5.5

0.0

Fig. 11. — Distribution of speci-

mens by height class. Num-

bers on right indicate ends of

0,5 m class intervals, a, all

specimens; b, only specimens

with a code for woody; c, for

all specimens from Lesotho.

J.W. MORRIS AND R. MANDERS

479

Fig. 12. — Distribution of specimens by diameter at breast height

(DBH). Numbers on right indicate ends of 0,05 m class inter-

vals.

only unusual situations, e.g. stream banks, marsh/

swamp and pan/depression are recorded by collec-

tors. Dry river beds in South West Africa are well

sampled in this category.

ris, 1980) and that ‘collected near road’ or a similar

statement was needed before code 06 could be assign-

ed and it can therefore be concluded from these

statistics that roadside collecting is a major activity

of plant collectors. Other favourite collecting sites

appear to be gardens, recently burnt veld and

disturbed areas (Fig. 13).

(e) Aspect

It was expected that compass points would be

equally represented. The results, however, show that

many more specimens are collected from south-

facing habitats than any others (Fig. 14) and that the

four cardinal points are far better represented than

the four intermediate points. The distribution is even

more exaggerated in the Orange Free State (Fig. 14b).

The results indicate guessing by collectors to the

nearest 90°, and often to the nearest 180°. It is not

known why south is the most commonly-collected

aspect but it is thought that south-facing slopes, with

more mesic conditions and often forest patches, will

have higher species diversities and therefore have

more species to be collected from them. The bias is

particularly marked in the Cape (Fig. 14c). Most even

distribution by aspect is shown by specimens from

Natal and South West Africa. Relatively few speci-

mens are, however, collected from NE- and NW-

facing slopes in South West Africa.

(0 Slope

Four percent of the specimens have a code for

slope, over 60% of which are level. A further 25%

are recorded from steep slopes and 10% from gentle

slopes. Specimens from South West Africa (96%),

Botswana (93%) and the Orange Free State (75%)

are virtually all collected from level sites, whereas

about 50% of the Natal specimens, 55% of Lesotho

specimens and 55% of Swaziland specimens are from

steep slope sites, as would be expected from a know-

ledge of the topography of these areas.

(c) Soil type and soil colour

A soil type code is assigned to only 4% of the

specimens. Of these, over half (56%) are collected

from sandy soil. Other categories with over 5%

representation are gravel (7%), loam (13%), clay

(9%) and calcrete (8%).

A disappointing 3% of specimens have soil colour

of the A-horizon codes and only 99 specimens have

B-horizon soil colour codes. Fifty percent of the

A-horizon soil colours are red.

(d) Biotic effects

Biotic effects codes are assigned to only 2% of the

specimens. The distribution gives an insight into the

plant collecting habits of botanists. Nearly half the

specimens with codes (44%) are collected alongside

roads and railway lines while 19% are from gardens

(Fig. 13). Eleven percent of the specimens are from

recently burnt areas. Encoders were specifically told

not to encode ‘on road to . . .’ as a biotic effect (Mor-

2239 I

122? CD

142 D

1095 □□

6570 I

15176 I

1413 I I

366 1 I

2903 I

229 '□

Cultivated land

Abandoned land

Planted pasture

PI ant at 1 on

Garden

Ro ad/Ra i 1 s i de

Heavily grazed

Recent 1 y burnt

Di sturbed -other

No effect seen

Fig. 13. — Distribution of specimens by biotic effects code.

6523 I 1 N

1146 I I NE

3568 I ~l E

1610 [ I SE

8118 I 1 S

1466 I I SW

3215 I 1 W

982 I I NW

a

234

34

34 1

40

310

36

178

53

b

]

N

NE

E

SE

S

SW

W

NW

2568 I I N

418 I I NE

1337 1 I E

777 1 I SE

4363 1 1 S

854 I 1 SW

1423 I I w

361 I I NW

c

Fig. 14.— Distribution of specimens by aspect: a, all specimens;

b, specimens from the Orange Free State; c, from the Cape

Province.

480 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OF THE NATIONAL HERBARIUM PRETORIA

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

(g) Altitude

Nearly 40% of the specimens have an altitude, in-

dicating the importance attached to altitude by plant

taxonomists. As expected, close agreement between

known geographic altitude ranges of the provinces

and altitude at which plants are collected is found.

Overall, there is evenly-spread collecting from sea

level to 1 500 m (Fig. 15a) and then a sharp decrease

in collecting intensity. In the Cape the decrease is

more gradual but starts at 600 m (Fig. 15b). In the

Transvaal, collecting decreases on either side of the

1 400 m contour (Fig. 15c), a pattern also shown by

the Orange Free State, only in a more extreme form.

In Natal, 52% of specimens have altitudes. A

number of small peaks are shown, possibly corres-

ponding with the altitudes of botanical survey areas

(Fig. 15d) or towns with active collectors. Collecting

continues in the Drakensberg to at least 3 200 m al-

though the distribution is only plotted as far as

2 300 m. Collecting at high altitudes is a feature of

mountainous Lesotho (Fig. 15e).

(h) Vegetation and veld type

A vegetation code is given to 21% of the speci-

mens. Of these specimens, nearly half are coded as

grassland. Taken together, woodland, savanna and

scrub/thicket formations account for a further 30%

and forest for 20% (Fig. 16). Desert, with only 363

specimens, karoo (including semi-desert) and fynbos

are particularly undersampled. Bearing in mind the

large area covered by karoo vegetation in South

Africa, it has been poorly collected. On the other

hand, as habitat and species diversities are lower in

the drier parts of the country, a lower sampling inten-

sity is to be expected there.

In Table 2 a representation of specimens by veld

types (Acocks, 1975) for the four provinces of South

Africa is given. The sample is small as only 3,1% of

the specimens are assigned veld types. In the table the

number of specimens from each province and the

total for the country are given for each veld type and

then the number of specimens collected per 100 km:

of that veld type is given as a measure of collecting

intensity. Veld type areas were obtained from Ed-

wards (1974). Eleven veld types (Kalahari Thornveld,

Mixed Bushveld, Sourish Mixed Bushveld, Karroid

Broken Veld, Arid Karoo, Succulent Karoo, Orange

River Broken Veld, False Upper Karoo, Highland

Sourveld, Cymbopogon — Theineda Veld and Dry

Cymbopogon — Themeda Veld) cover half the area of

South Africa and yet only 14% of the specimens with

veld types were collected from them indicating again

the uneven distribution of collection effort. Five of

the large veld types listed above are either Karoo or

False Karoo types, whereas the largest, Kalahari

Thornveld, is an arid region tropical bush and savan-

na type. Therefore the under-collection of the more

arid parts of South Africa, shown from an analysis

of codes assigned to vegetation, is confirmed by this

analysis of specimen distribution by veld type.

Macchia and Coastal Macchia have the greatest

number of specimens collected from them and also

the highest collection frequency. The high collection

13056 I I 0

1 1589 | ~~1 100

1 1349 | | 200

8874 I ~) 300

7324 1 | 400

9298 | 1 500

13166 | | 600

5072 | ~| 700

8484 1 1 800

1 1647 | ~| 900

5133 1 ~| 1000

8993 | | 1100

13295 | 1 1200

6656 1 1 1300

15014 | | 1400

13037 | | 1500

4526 | 1 1600

7481 | 1 1700

5858 | | 1800

1735 | | 1900

3353 | | 2000

2177 | [ 2100

639 □ 2200

946 □ 2300

a

8131

6 138

8510

6049

4688

5158

7884

33 10

4893

5199

2277

3647

4720

1763

2899

3069

807

1201

1208

395

606

305

□

169 D

209 D

b

a

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

2100

2200

2300

247 □

1087 1 |

200

300

400

500

600

700

800

900

1000

1 100

1200

1300

1400

1500

1600

1700

1800

1900

2000

2 100

2200

2300

305 □

0

100

200

300

400

500

600

700

800

900

i0Be

1 100

1200

1300

1100

1500

>800

1700

1800

1900

2000

2100

2200

2300

71 □ 1500

203 I I 1600

831 I | 1700

107 I I 1000

17 □ 1 900

93 I I 2000

231 I 1 2100

90 I I 2200

223 I I 2300

629 I I 2100

1 H I I 2500

298 I I 2600

281 I ~~1 2700

105 | | 2000

308 I I 2900

182 I I 3000

I 15 I I 3100

65 □ 3200

29 D 3300

37 □ 3100 p|G

13 □ 3500

e

15.— Distribution of specimens by alti-

tude. Numbers in right-hand column

indicate ends of 100 m class intervals,

a, all specimens; b, Cape Province;

c, Transvaal, d, Natal; e, Lesotho.

J.W. MORRIS AND R. MANDERS

481

3E3 D Desert

1752 □ Karoo

45536 I I Grassland

9801 I 1 Savanna

14304 I I Woodland

7806 I I Scrub/thicket

20838 I I Forest

5790 I I Fynbos

Fig. 16.— Distribution of specimens by vegetation type.

frequency for Themeda Veld to Highland Sourveld

Transition in Natal is due to the small area covered

by this veld type (52 km2) and the relatively large

number of specimens collected from it (25). The high

collection frequencies for Pondoland Sourveld in

Natal (14 specimens/104 km2), Themeda Veld to

Cymbopogon— Themeda Veld transition in Trans-

vaal (54/419) and Highland Sourveld to Cymbopo-

gon— Themeda Veld in Natal (20/104) can be

similarly explained. In the Cape, other well-represen-

ted veld types are from the smaller Karoo, Karroid

Bushveld, False Karoo and temperate and transition-

al forest and scrub types. In the Transvaal, bushveld

and the tropical bush and savanna types are well col-

lected. There are few specimens with veld types from

the Orange Free State and most are from False Upper

Karoo, Highland Sourveld and pure grassveld types.

In Natal, well-represented veld types include Coastal

Tropical Forest types, Valley Bushveld, Highland

Sourveld and Pure grassveld types.

A small amount of mis-coding of specimens is

bound to occur in a project of this magnitude. Thus

seven specimens of Coastal Forest and Thornveld,

coded as being from the Orange Free State, were ob-

viously given either the wrong Region code or the

wrong veld type code. Similarly, Mopani Veld (4

specimens) does not occur in Natal. Where a veld

type does not occur in a province, specimens record-

ed for that veld type in the province were omitted in

Table 2. A total of 602 specimens (3,9% of specimens

with veld type codes) was excluded in this way.

Presence of such errors in the data bank should not

prevent users from deriving valuable information

from it. General trends should be sought and not ex-

ceptions or obvious errors.

Encoding rules and collector’s biased knowledge

can explain some of the results presented above.

Some veld types are inherently easier to identify than

others (e.g. Mopani Veld, Macchia) and they will,

therefore, tend to be noted on collecting labels more

often than small, obscure ones, identifiable by

specialists alone. The close correspondence between

the number of specimens coded as vegetation: fynbos

and veld type: Macchia is not surprising because the

encoders were taught that the terms are synonymous,

and where one code was used, the other was normally

also encoded.

Economic botany data

(a) Actual-potential use

An actual-potential use code is given to only 2% of

the specimens. Ninety-three percent of these speci-

mens are actually used, whereas 7% of them have a

potential use.

(b) Utilized by

A disappointing 8 702 specimens have a utilized by

code. Forty-six percent of these specimens are util-

ized by man while stock utilizes 27% of the specimens

(Fig. 17). Relatively high percentages of specimens

are utilized by man in Swaziland (63%), Botswana

(68%) and Lesotho (68%). Stock utilizes 41% of the

TABLE 2. — Distribution of specimens by veld type and province. See text for further explanation

Number of specimens Specimens/ 100 square km

(Continued overleaO

482 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OF THE NATIONAL HERBARIUM, PRETORIA

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

specimens in the Cape Province and 61% in the

Orange Free State in comparison with the 29% and

13% which are utilized by man in the Cape and

Orange Free State respectively.

(c) Economic property

Only 1% of the specimens have an economic pro-

perty code, of which 59% are stated as being eaten

(Fig. 18). Other economic properties which occurred

with a frequency of more than 5% were poison (7%),

general medicine (13%) and structural purposes

(6%). Another 78 specimens are used for magic or

ritual purposes, 121 specimens are used as beverage,

80 as fuel, 74 as sand binders, 61 as ground covers

and 216 as ornamental plants in gardens.

Thirty-six percent of the specimens with economic

property codes in Lesotho are used as medicine com-

pared with 34% which are used as food. In Swaziland

a relatively high 14% of the specimens are used for

structural purposes.

Examples of the specimens which have magic as an

economic property are listed in Table 3. From an ex-

amination of the entire set of specimens with this

code it is clear that such attributes are given to plants

from diverse taxa but that relatively few collectors

regularly noted such attributes. Most specimens with

this property were collected by J. Gerstner (Natal)

and W. G. Barnard (Transvaal).

(d) Special record

A special record code was used to indicate plants

collected for a specific purpose or survey. A total of

9 288 specimens have such a code. Of these, 34% are

for SKF alkaloid studies, 25% for anatomy studies,

19% for a cancer research project, 9% for pollen

studies (UOFS) and 5% for the ‘Stijfsiekte’ survey.

J.W. MORRIS AND R. MANDERS

483

Fig 18. — Distribution of specimens by economic property code.

PHENOLOGY

It is likely that phenological studies, i.e. the rela-

tionships between climatic factors and periodic

phenomena in organisms (Daubenmire, 1959), will be

facilitated in future through the use of PRECIS.

Many plants, as an example of phenology, are known

to flower in response to some change in the environ-

ment (Salisbury, 1963). It is possible to find all the

specimens of a particular species that were either

flowering or fruiting at the time of collection (from

the state of the specimen codes) and to sort these by

month of collection (and day, if necessary) in order

to study the march of phenological development over

a large geographic range.

To illustrate our contention that PRECIS can be of

use in this way, the flowering patterns of three wide-

spread indigenous grass species were obtained. Print-

outs were made of all specimens, from South Africa

only, with flowers present or mature, sorted by

month and day of collection. With the aid of these

printouts as well as computer-generated distribution

TABLE 3. — Examples of specimens coded with ‘magic’ economic property code

484 INFORMATION AVAILABLE WITHIN THE PRECIS DATA BANK OF THE NATIONAL HERBARIUM, PRETORIA,

WITH EXAMPLES OF USES TO WHICH IT MAY BE PUT

maps of those specimens with quarter degree grid ref-

erences, the distributions of Eragrostis capensis (Fig.

19), Themeda triandra (Fig. 20) and Heteropogon

contortus (Fig. 21) were drawn. All three species are

widespread in the Transvaal, Orange Free State and

Natal and occur along the southern Cape coast as far

as Cape Town. The general distribution patterns

based on the computer data base are remarkably

similar to those given by Chippindall (1955) for the

same species even though only about two-thirds of

the specimens in the data base could be plotted (those

specimens flowering at other times of year, without

date of collection, collected without flowers or with

localities that could not be converted to grid

references were excluded). In all three species, early

flowering (August to October) is clearly restricted to

the eastern and southern coastlines and to a less

marked extent to the midlands of Natal. Later

flowering (October to February, depending on the

species concerned) is found inland in the Cape and

throughout the distribution ranges of the species.

Where three flowering periods are plotted there is a

clear progression of flowering from the coast to the

interior.

It is not our intention to provide a detailed ex-

planation for the phenomenon described above. We

consider that photoperiodism (Salisbury, 1963) is

unlikely to be accounting for the observed pattern, as

the pattern is not directly related to latitude, and that

temperature is likely to be playing a major role. It has

been shown by Schulze & McGee (1978) that iso-

therms run parallel to the coasts of South Africa with

decreasing values with distance inland, reflecting the

effects of continentality. Thus we expect that the

coast is warmer earlier in the growing season than the

interior at a given latitude, accounting for a more

rapid phenological development on the coast.

Although temperatures in the far northern Transvaal

are as high, if not higher, than those on, say, the

Natal coast, soil moisture as an interacting factor is

probably limiting growth early in the season in the

Transvaal.

With the aid of PRECIS, the phenology of a large

number of species can be investigated rapidly. In this

way, it would be possible to ascertain how general the

pattern found with these three grasses is, as a first

step towards explaining the phenomenon in detail.

From preliminary studies of this kind with tree

species, Acacia karroo Flayne, Erythrina lysistemon

Hutch, and Apodytes dimidiata E. Mey. ex Arn., it

appears that not all widespread species respond to the

same environmental triggers in the same way as these

species appear to have synchronized flowering and

fruiting throughout their range within South Africa.

DINTER’S COLLECTING ROUTE

According to Dr L. E. Codd (pers. comm.) the

last route followed by Dinter, an important collector

in South West Africa, is not as well known as those

of his earlier visits, accounts of which he published.

As an example of the use to which PRECIS could be

put for this purpose, a printout of Dinter’s specimens

collected from December 1933 until he finally left the

country was made with specimens sorted according

to date of collection. The following reconstruction is

based on that printout of specimens housed in the

National Herbarium.

From December 1933 until mid-March 1934 Dinter

collected extensively in the Karibib district. His speci-

men numbers ran from 6721 to about 7474 and com-

Fig. 19. — Distribution of flowering in Eragrostis capensis within

South Africa.

Fig. 20.— Distribution of flowering in rhemeda triandra.

Fig. 21. — Distribution of flowering in Heteropogon contortus.

J.W. MORRIS AND R. MANDERS

485

mon collecting localities included Kalkhugel,

Kalkbuschsteppe, Okongava, Arnusema, Ameib,

Erongo and Unduas. In mid-March he moved north-

east to Grootfontein, collecting on the way, and on

the 18th March started collecting in the Tsumeb

district. He remained in that area until mid-July, col-

lecting specimens 7497 to 7703. Localities included

Heidelberg, Bobos, Neitsas and Guchab. During this

period he also collected occasionally at Karakowisa,

east of Grootfontein. During August he returned to

the Karibib district where he collected a few speci-

mens near Okahandja. Dinter then travelled south to

Windhoek where he collected specimens on the 10th

October. A few days later he collected south-west of

Windhoek at Lichstenstein, Friedental, Weissenfels

and in the Hakos Mountains. He stayed in this area

until about the 20th December, collecting numbers

7889 to 8054. During late December he travelled fur-

ther south to Buchholzbrunn, near Aus, where he

collected numbers 8264 to 8307. From this area he

travelled to Swakopmund, collecting specimens at

Helmeringhausen and Nudaus on the way. In March

1935 he collected at Swakopmund, with the last

specimen to be collected in the territory apparently

being Hypertelis caespitosa Friedr. (no. 8471) on the

15th March 1935.

We have shown above that a fairly detailed ac-

count of a collector’s route can be obtained from a

study of his specimens in such a data bank. Without

selection and sorting by computer, of course, and a

good collection of his material in the herbarium, it

would not have been feasible to undertake such a

task. Other factors contributing to the success of this

particular exercise were the presence of dates, se-

quential collecting numbers and localities that could

be identified on most labels. Absence of one or more

of these items would have made reconstruction far

more difficult, if not impossible.

CONCLUSION

In this paper we have presented an account of what

information is contained within the PRECIS data

bank and have given two examples of the kinds of in-

formation that may be extracted from it. In some

areas, e.g. economic botany and habitat infor-

mation, there is far less information available on her-

barium sheets than had been expected. Also, it has

become clear that collectors note abnormal charac-

teristics and situations far more regularly than nor-

mal ones, leading to an apparent bias in some cate-

gories. On the other hand, we consider that particu-

larly valuable information has already been extracted

and should continue to be extracted from PRECIS

for the benefit of the botanical community in South

Africa. Obviously, the information presented in this

paper has limited value in itself but the data bank

should be of particular use in revisionary studies,

regional floras and biogeographic research.

ACKNOWLEDGEMENTS

This project could not have been a success without

the wholehearted support of many people. We would

like to single our Messrs G. J. Smit, J. P. D. van

Wyk, J. Nel and Miss B. Young and all the computer

programmers who have worked on this project

through the Department’s Computer Centre; Dr H.

F. Glen and Mrs J. Jooste who were closely associa-

ted with the project during its development; Dr L. E.

Codd and Miss M. D. Gunn for discussions on

historical aspects; and, finally, Dr B. de Winter for

continued faith in and encouragement with this pro-

ject.

UITTREKSEL

Die inhoud van die gerekenariseerde inligting ber-

ging en onttrekking stelsel (PRECIS) van die Nasio-

nale Herbarium, Pretoria (PRE) word breedvoerig

bespreek, hoofsaaklik by wyse van frekwensie

histogramme van beskrywingskodes. Die frekwensie

verspreidings wat gevind is, word in die lig van die

geskiedenis van die herbarium, die geografie van die

gebied en die gewoontes van plantversamelaars

bespreek. Twee gebruike van PRECIS word dear

voorbeelde geillustreer. Eerstens, word die blom-

fenologie van Eragrostis capensis, Themeda triandra

en Heteropogon contortus geplot en, tweedens, word

die roete wat dear Dinter in Suidwes-Afrika vanaf

Desember 1933 tot Maart 1935 gevolg is, beskryf.

Daar is tot die gevolgtrekking gekom dat die stelsel

van besondere nut in hersienings studies, streeks

floras en biogeografiese navorsing kan wees.

REFERENCES

Acocks, J. P. EL, 1975. Veld types of South Africa. 2nd edn.

Mem. bot. Surv. S. Afr. No. 40.

Chippindall, L. K. A., 1955. A guide to the identification of

grasses in South Africa. In D. Meredith, The grasses and pas-

tures of South Africa. Johannesburg: CNA.

Crovello, T. J. 1972. Computerization of specimen data from the

Edward Lee Greene Herbarium (ND-G) at Notre Dame. Brit-

tonia 24: 131-141 .

Daubenmjre, R. F., 1959. Plants and environment. New York:

John Wiley.

Dyer, R. A., 1977. Botanical research in South Africa in the

twentieth century. In A. C. Brown, A history of scientific

endeavour in South Africa 240-264. Cape Town: Royal

Society of South Africa.

Edwards, D. 1974. Survey to determine the adequacy of existing

conserved areas in relation to vegetation types. A preliminary

report. Koedoe 17: 2-37.

Edwards, D. & Leistner, O. A., 1971 . A degree reference system

for citing biological records in Southern Africa. Mitt. bot.

StSamml., Munch. 10: 501-509.

Morris, J. W., 1974. Progress in the computerization of

herbarium procedures. Bothalia 11: 349-353.

Morris, J. W., 1 980. Encoding the National Herbarium (PRE) for

computerised information retrieval. Bothalia 13: 149-160.

Morris, J. W. & Glen, H. F., 1978. PRECIS, the National

Herbarium of South Africa (PRE) computerised information

system. Taxon 27: 449-462.

Morris, J. W. & Leistner, O. A., 1975. Progress with com-

puterization of the National Herbarium, Pretoria. Boissiera

24: 411-413.

Phillips, E. P., 1930. A brief historical sketch of the development

of botanical science in South Africa and the contribution of

South Africa to botany. S. Afr. J. Sci. 27: 39-80.

Ross, J. H., Leistner, O. A. & De Winter, B., 1977. A guide to

contributors to the Flora of Southern Africa (F.S.A.) Pre-

toria: Department of Agricultural Technical Services.

Salisbury, F. B., 1963. The flowering process. Oxford: Pergamon

Press.

Schulze, R. E. & McGee, O. S., 1978. Climatic indices and classi-

fication in relation to the biogeography of Southern Africa.

In M. J. A. Werger, Biogeography and ecology of Southern

Africa 19-52. The Hague: Dr W. Junk.

Bothalia 13, 3 & 4: 487-491 (1981)

Leaf anatomy of the South African Danthonieae (Poaceae).

IV. Merxmuellera drakensbergensis and M. stereophylla

R. P. ELLIS*

ABSTRACT

The leaf blade anatomy of Merxmuellera drakensbergensis (Schweick.) Conert and M. stereophylla (J. G.

Anders.) Conert is described and illustrated. These two closely related species have virtually identical leaf

anatomy — both the leaf in section and the abaxial epidermis. The close anatomical resemblance between these two

species raises doubts about their specific status. This is especially significant when compared with the considerable

differences observed between the anatomical ‘forms’ recognized in M. disticha (Nees) Conert (Ellis, 1980) and

M. stricta (Schrad.) Conert (Ellis, 1980a).

RESUME

L'ANATOMIE DE LA FEUILLE DU DANTHONIEAE (POACEAE) SUD AFRICAIN.

IV. MERXMUELLERA DRAKENSBERGENSIS ET M. STEREOPHYLLA

L ’anatomie de la feuille de Merxmuellera drakensbergensis (Schweick.) Conert et M . stereophylla (J. G. Anders.)

Conert est decrite et illustree. Ces deux especes etroitement apparentees ont une anatomie de la feuille virtuellement

identique-tant la section de la feuille que Tepidermis abaxial. La ressemblance anatomique etroite entre ces deux

especes souleve des doutes au sujet de leur staut specif ique. Ceci est specialement significatif quand on compare les

differences considerables observees entre les “formes” anatomiques reconnues dans M. disticha (Nees) Conert

(Ellis, 1980) et M. stricta (Schrad.) Conert (Ellis, 1980a).

INTRODUCTION

Merxmuellera drakensbergensis (Schweick.)

Conert (1970) ( = Danthonia drakensbergensis

Schweick.) and M. stereophylla (J. G. Anders.)

Conert (1970) ( = D. stereophylla J. G. Anders.) are

wiry-leaved perennials forming rigid, erect tussocks.

The unbranched culms grow vertically and the leaves

are rigid and taper to a pungent apex. In both species

these leaves are setaceous and tightly involute or

canaliculate.

These two species are conspicuous components of

the alpine vegetation of the Drakensberg mountains

to which they are restricted. M. drakensbergensis

occurs in the Barkly East and Maclear Districts of the

north-eastern Cape, along the Drakensberg moun-

tains of Natal and Lesotho and at Mariepskop in the

Drakensberg of the north-eastern Transvaal. M.

stereophylla has a more limited distribution, being

found only in the Drakensberg areas of Natal and

Lesotho at altitudes above 2 000 m. Although the

distribution of these two species overlaps in the Natal

and Lesotho alpine areas, they can, nevertheless, be

distinguished both ecologically and morphologically.

M. drakensbergensis occupies mesic situations in

the streambank and mud patch communities (Killick,

1963; Edwards, 1967) of the alpine belt along the

summit of the high Drakensberg. The habitat of M.

stereophylla, on the other hand, is essentially xeric

and this species is common in the alpine grassland of

the basalt cliffs as a crevice and ledge plant. In the

Danthonia Tussock Grassland (Edwards, 1967) M.

drakensbergensis is dominant around sponges and

mud patches, but on rocky areas M. stereophylla is

the principal grass. Although these two closely

related species (Anderson, 1960) have long been con-

fused, they are distinct ecologically and in the field

can readily be distinguished by their differing habitat

requirements.**

Vegetatively, these species can also be easily recog-

nized. M. stereophylla has rigid, erect, grey-green

*Botanical Research Institute, Department of Agriculture

and Fisheries, Private Bag X101, Pretoria, 0001.

leaves, whereas M. drakensbergensis has softer leaves

which are olive-green in colour. M. drakensbergensis

plants are up to 100 cm tall and M. stereophylla is a

slightly smaller plant up to 80 cm high. A characteris-

tic feature of M. drakensbergensis, which is not evi-

dent in M. stereophylla, concerns the behaviour of

old leaf blades (Anderson, 1960). These normally

break off above the ligule and the remaining portion

of blade splits along the median nerve and the resul-

tant halves recurve outwards in opposite directions.

This useful field diagnostic character appears to be

consistent and it is only in recently burnt plants that

this character is not evident. However, as these

species are badly injured by fire (Edwards, 1 967) they

are largely confined to fire-protected moist or rocky

habitats. M. macowanii (Stapf) Conert also exhibits

this vegetative characteristic and can, therefore, be

confused with M. drakensbergensis. Both species are

also streambank plants, but M. macowanii appears

to be limited to the montane and sub-alpine belt

below the summit of the Drakensberg.

The relatively recent description and recognition of

M. drakensbergensis and M. stereophylla is some-

what surprising in view of these distinct ecological

and vegetative differences. M. drakensbergensis was

only described in 1938 (Schweikerdt, 1938), prior to

which it was referred to M. macowanii. M. stereo-

phylla received recognition as recently as 1960

(Anderson, 1960), although Chippindall (1955) men-

tioned an undescribed species from the high Drakens-

berg and was undoubtedly referring to this species. A

probable reason for these species remaining undes-

cribed for so long is the relative inaccessibility of the

area in which they occur, as well as the fact that

♦♦According to Killick (1978) M. drakensbergensis is an ubiquitous

species in the alpine belt of the Sani Pass area of the southern

Drakensberg: it often occurs in flushes or along streambanks but is

also found on rock outcrops and in Alpine Grassland, sometimes

covering fairly large areas. M. stereophylla, on the other hand,

appears to be restricted to dry outcrops at higher altitudes in this

part of the Drakensberg. M. drakensbergensis is, therefore, not

restricted to semi-aquatic communities in this area but displays a

wider ecological tolerance. It, nevertheless, prefers deeper and

moister soils than M. stereophylla.

488

LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). IV. MERXMUELLERA

DRAKENSBERG ENSIS AND M. STEREOPHYLLA

spikelet differences are slight and only a matter of

degree. Thus the arrangement of the hairs on, and

the length of, the lemmas (including lobes and awns)

differ slightly (Anderson, 1960, 1962).

That these two species are very closely related

(Anderson, 1960) is confirmed by the anatomy of

their leaf blades which is almost identical. No consis-

tent and measurable structural differences are evi-

dent and evidence from leaf anatomy, therefore,

casts some doubt on the validity of granting these

species specific status. This is especially significant

when compared with the situation in M. stricta

(Schrad.) Conert (Ellis, 1980a) and M. disticha

(Nees) Conert (Ellis, 1980) where disjunct, relatively

important anatomical differences were found to

occur within each of these species. These anatomical

differences were found to be consistently correlated

with ecological and morphological differences and,

therefore, these ‘forms’ of M. stricta and M. disticha

appear to warrant similar taxonomic treatment to

M. drakensbergensis and M. stereophy/la. The grant-

ing of specific status to M. drakensbergensis and

M. stereophylla, therefore, requires reassessment.

The only anatomical differences observed between

these two species are slight differences in size and

hence a combined description will suffice for both.

Leaf anatomy and epidermal structure will be des-

cribed following the terminology of Ellis (1976, 1979)

and the following abbreviations will be used in the

COMBINED ANATOMICAL DESCRIPTION OF

MERXMUELLERA DRAKENSBERGENSIS AND

M. STEREOPHYLLA

Leaf in transverse section

Leaf outline-, permanently and tightly infolded

with elliptical outline. Laminae slightly assymmetri-

cal about the median vb such that adaxial furrows of

one half of lamina align with adaxial ribs on other

half of lamina. As a result margins overlap slightly.

Adaxial channel always a deep, narrow cleft with a

slight opening where margins overlap. Leaf size:

setaceous; leaves narrow (0,56 mm-3,60 mm wide)

Figs I -8.— Leaf blade outline of Merxmuellera drakensbergensis and M. stereophylla in transverse section. I -3, M. drakens-

bergensis, all x 160. (I, Du Toil 669; 2, Liebenberg 5707; 3, Du Toil 2313.) 4-6, M. stereophylla, all X 160. (4, Jacol

Guillarmod 3733; 5, Roberts 3152; 6, Edwards 2284.) 7, M. drakensbergensis, x400. (Ellis 3191.) 8, M. stereophylla,

x 400. (Ellis 3139.)

R. P. ELLIS

489

but tend to be thinner in M. stereophylla (0,56

mm-1,13 mm wide) than in M. drakensbergensis.

This tendency reflected in number of vbs in leaf sec-

tion. 11-15 vbs present with 11 vbs always in M.

stereophylla (Figs 4-6) and M. drakensbergensis

usually with 13 or 15 vbs (Figs 1 & 2) but sometimes

only 11 vbs present (Fig. 3). Ribs and furrows :

medium to deep cleft-like adaxial furrows between all

vbs; rounded (Fig. 8) or slightly flat-topped (Fig. 7)

ribs over all vbs; one vb per rib. Abaxial surface

smooth or with very slight undulations associated

with vbs (Fig. 3). Median vascular bundle : present

but indistinguishable structurally from l’vbs.

Vascular bundle arrangement : no 2’vbs; 3’vbs ab-

sent between consecutive lateral l’vbs. All bundles

centrally located between upper and lower epider-

mides. Vascular bundle structure : vbs circular or

elliptical in shape. Xylem and phloem distinguishable

in all vbs; phloem adjoins ibs; phloem divided ver-

tically into two equal groups by intrusion of fibres

(Figs 7 & 8). Metaxylem vessel diameter narrow being

only slightly greater than the diameter of the obs

cells; slightly thickened. Vascular bundle sheaths :

obs of all vbs horse-shoe shaped with wide abaxial in-

terruptions. Adaxial interruptions usually fairly nar-

row (Fig. 8) but may be wide (Fig. 7) especially in M.

drakensbergensis. No bundle sheath extensions pre-

sent. Obs cells round or elliptical, sometimes with

straight radial walls (Fig. 8); all obs cells similar in

shape but small, being only slightly larger than the

mesophyll cells in cross-sectional area; cell walls

slightly but distinctly thickened; without chloro-

plasts. Inner sheath complete around all vbs; ibs cells

similar in size to the obs cells but with considerably

thicker walls, especially inner tangential wall (Fig. 8).

Sclerenchyma : adaxial girders inversely anchor-

shaped with narrow (Fig. 8) or sturdy (Fig. 7) stem;

fibres interrupt obs. Abaxial sclerenchyma in form of

continuous subepidermal band (especially well

developed adjacent lateral l’vbs) of varying thickness

with large trapezoidal girders extending to, and inter-

rupting, the obs; girders comprised of thick-walled

fibres usually lignified although fibres near margin

may be of cellulose (Fig. 1). Leaf margin-, very small,

pointed, poorly developed cap. Mesophyll'. non-

radiate; chlorenchyma of regular, small, isodiametric

cells; tightly packed with air spaces not visible; in

Y-shaped groups occupying sides and bases of adax-

ial furrows (Figs 7 & 8). No colourless cells present.

Adaxial epidermis : poorly developed bulliform cells

at bases of furrows. Macro-hairs absent. Pointed

prickles with broad, but not bulbous bases; present

throughout costal zones. Outer walls of epidermal

cells arched and somewhat inflated and appear to be

papillate. These may, however, represent sections,

through varying planes, of the prickles. Abaxial

epidermis: bulliform cells absent. Outer cell walls

markedly thickened with a continuous, thick cuticle.

No macro-hairs, prickles or papillae.

Abaxial epidermis in surface view

Intercostal zone : undifferentiated and entire

abaxial epidermis similar in structure (Fig. 9 & 12)

and essentially a costal zone due to development of

continuous sub-epidermal fibrous layer. Stomata :

absent from abaxial surface. Prickle-hairs : hooks

and prickles not present. Micro-hairs : not seen on

any of the specimens examined. Macro-hairs : absent.

Silica bodies : equidimensional in surface view; either

cuboid (Fig. 14), round (Fig. 10) or somewhat ellip-

tical (Fig. 1 1); usually fitting into concavity in closely

associated cork cell. Granules present; cracks some-

times present (Fig. 14). Width of silica bodies slightly

narrower than costal long cells and cork cells. Similar

silica bodies present throughout abaxial epidermis.

Costal cells : silica cells and cork cells alternate with

single costal long cells throughout abaxial epidermis;

all files of similar cell arrangement. Silica may be

relatively sparse (Fig. 10) with many silico-suberose

couples actually consisting of a pair of cells (a cork

and a silica cell) or even only a single cork or silica

cell. Costal long cells elongated horizontally; at least

3 x longer than wide; sides parallel; anticlinal walls

heavily thickened and usually pitted (Fig. 10 & 13);

undulations moderate to deep but difficult to distin-

guish clearly due to excessive cuticle thickening.

Specimens examined:

M. drakensbergensis

O.F.S. — 2828 (Bethlehem): Witsieshoek-Mont-aux-Sources area

(-DB), Ellis 3137, 3138, 3154, 3155, 3156.

Natal. — 2828 (Bethlehem): Sentinel (-DB), Du Toit 669. 2829

(Harrismith): Cathedral Peak Forest Reserve (-CC), Ellis 1398,

Figs 9-14.— Abaxial epidermis of Merxinuellera drakensbergensis and M. stereophylla. 9-11, M. drakensbergensis. (9, Kil-

lick & Marais 2183 x 160; 10, Ellis 1398, x 640, note number of cork cells associated with silica bodies; 1 1, Staples 242,

x 160, phase contrast.) 12-14, M. stereophylla. (12, Ellis 3139, x 160; 13, Ellis 3139, x400; 14, Killick & Vahrmeijer

4019, x 640.)

490

LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). IV. MERXMUELLERA

DRAKENSBERGENSIS AND M. STEREOPHYLLA

3189, 3190, 3191, 3304. 2929 (Underberg): Sani Pass (-CB), Du

Toil 2313.

Lesotho. — 2828 (Bethlehem): Maluti Mts (-CC), Staples 242.

2929 (Underberg): Pone Valley (-AC), Coetzee 824; Mokhotlong,

Liebenberg 5707; Sani Pass summit (-CA), Du Toil 2209.

Cape. — 3028 (Matatiele): Naudes Nek (-CA), Werdermarm &

Oberdieck 1139.

M. stereophylla

O.F.S. — 2828 (Bethlehem): Golden Gate National Park (-DA),

Roberts 3152; Witsieshoek-Mont-aux-Sources area (-DB), Ellis

3139.

Natal. — 2828 (Bethlehem): Mont-aux-Sources (-DD), Schelpe

1390. 2829 (Harrismith): Cathedral Peak Forest Reserve (-CC),

Ellis 1397, 1408, 3180, 3186, 3298, 3307, Killick 1317. 2929

(Underberg): Giants Castle Game Reserve (-AB), Killick &

Vahrmeijer 4019, Edwards 2284.

Lesotho. — 2828 (Bethlehem): Letseng-la-Terai (-CA), Lo.xton

& Ellis 997. 2829 (Harrismith): Cleft Peak area (-CC), Killick &

Marais 2183. 2929 (Underberg): Sani Pass summit (-CA), Du Toil

2208; Tschlanyane Valley (-CD), Jacot Guillarmod 3733.

DISCUSSION AND CONCLUSIONS

From the above anatomical description of the leaf

blades of M. drakensbergensis and M. stereophylla,

and from the accompanying photomicrographs (Figs

1-14), it is clearly evident that the structure of these

two species is remarkably similar — both the leaf sec-

tions and the epidermis. The only difference detected

in this study is a small variation in the cross-sectional

area and width of the leaf sections. The magnifica-

tions of Figs. 1-6 are identical and a comparison of

M. drakensbergensis leaf blades (Figs 1 -3) with those

of M. stereophylla (Figs 4-6) reveals that the leaves

of M. drakensbergensis have a tendency to be larger.

This difference is not consistent as numerous in-

termediates occur. Nevertheless this size difference is

correlated with the number of vascular bundles pre-

sent in the leaf blade. M. stereophylla has 1 1 vascular

bundles, whereas M. drakensbergensis has 13 or 15.

However, even this distinction is not consistent e.g.

Fig. 4 which is intermediate with 12 vascular bundles

and Fig. 3, M. drakensbergensis, with only 10

vascular bundles.

These slight quantitative differences are, neverthe-

less, characteristic and consistent for the majority of

the specimens examined in this study. However, in

this representative and extensive sample there were

several specimens which proved difficult to identify

satisfactorily using Anderson’s (1962) key. For exam-

ple, both Ellis 1397 and Loxton & Ellis 997 were in-

itially determined as being M. strieta but, on check-

ing, were identified as M. drakensbergensis and final-

ly placed in M. stereophylla, mainly because they do

not exhibit the outward recurving of the old leaf

blades. In these specimens the arrangement of the

hairs on the lemmas is apparently atypical. Edwards

2284 is another specimen which has caused problems

with identification. Anderson and Conert both deter-

mined this specimen as being M. aureoeephala (J. G.

Anders.) Conert, but in this study it has been classi-

fied as being M. stereophylla.

In these morphologically atypical specimens the

number of vascular bundels in the leaf section, and

the corresponding width of the leaf blade, appear to

be of no assistance in identification. Thus, Ellis 1397,

collected in saturated, spongy ground is probably

ecologically best placed in M. drakensbergensis,

although it would be a very small, atypical specimen.

It has only 11 vascular bundles in the leaf section,

which is characteristic of M. stereophylla. The

ecological, morphological and anatomical indica-

tions are, therefore, in conflict.

Ellis 1397 and 1408 are specimens collected near or

at the summit of the Drakensberg escarpment at

Cathedral Peak. The Merxmuellera populations at

this particular locality were characterized by being

extremely variable morphologically. Anatomical

studies of the M. strieta and A/, disticha (Ellis, 1980,

1980a) specimens from this locality revealed that the

various specimens actually belonged to different

‘forms’ with distinct anatomical, morphological

and ecological characteristics and that it was mere co-

incidence that they were found growing in such close

proximity. The M. drakensbergensis and M. stereo-

phylla specimens from these populations, on the

other hand, show no correlation between morpho-

logical, anatomical and ecological characteristics

and Ellis 1397 and 1408 in fact represent true inter-

mediates.

This observation, together with the close resem-

blance of the leaf anatomy of these two species, in-

dicates that M. stereophylla and M. drakensbergensis

are very closely related, more so, in fact, than the

anatomical forms of M. strieta and M. disticha. The

specific status of M. stereophylla and M. drakens-

bergensis is, therefore, questioned and indications

are that these two taxa warrant similar taxonomic

treatment to the anatomical forms of M. strieta and

M. disticha.

Both M. drakensbergensis and M. stereophylla ex-

hibit the same type of arrangement of vascular

bundles along the width of the lamina, there being no

third order vascular bundles between the lateral first

order bundles. This is essentially similar to the posi-

tion described in M. strieta (Ellis, 1980a) and in-

dicates the relationship of M. strieta to M. stereo-

phylla and M. drakensbergensis. In addition, the

anatomical evidence indicates that M. stereophylla

and M. drakensbergensis exhibit a similar degree of

divergence from the typical M. strieta anatomical

form as do the other three anatomical forms of M.

strieta (Ellis, 1980a). This implies that consideration

should be given to the granting of equivalent taxo-

nomic status to M. drakensbergensis, M. stereo-

phylla, typical M. strieta and to the three other forms

of M. strieta. The evidence gained in the present

study indicates that subspecific rank is probably jus-

tified for each of the above entities of the M. strieta

group.

M. drakensbergensis and M. stereophylla are

thought to be related to three other species of Merx-

muellera, all of which also occur in the Drakensberg

mountains. These are M. macowanii (Stapf) Conert,

M. davyi (C. E. Hubb.) Conert and M. aureoce-

phala (J. G. Anders.) Conert and together these five

species form a more or less distinct and closely rela-

ted group within the genus (Anderson, 1962). This is

not confirmed by anatomy, however, which indicates

a relationship between M. disticha and M. macowanii

and M. davyi. This is based on the alternating

arrangement of lateral first order bundles with third

order vascular bundles common to the latter three

species (and probably to M. aureoeephala) (Ellis, in

press). A similar relationship exists between M.

disticha and M. macowanii and M. davyii to that

demonstrated between M. strieta and M.

drakensbergensis and M. stereophylla. Indications

are, once again, that similar patterns of adaptive

radiation have occurred in the Drakensberg area

from parental stock of both M. strieta and M.

disticha. Thus, in both groups, cave sandstone and

R. P. ELLIS

491

basalt, alpine bog, and streambank ‘forms’ appear to

have evolved in response to the environmental condi-

tions presently prevailing in the Drakensberg moun-

tains. This unique situation will undoubtedly reward

further population and cytogenetical studies.

ACKNOWLEDGEMENTS

The author is grateful to the Department of Forest-

ry, the National Parks Board and the Natal Parks,

Game and Fish Preservation Board for assistance

with the collection of fresh material. Miss R.

Manders and Miss L. Breytenbach are thanked for

efficient technical assistance.

U/TTREKSEL

Die anatomiese struktuur, van die blaar in dwars-

snee en die abaksiale epidermis, van Merxmuellera

drakensbergensis ( Schweick .) Conert en M. stereo-

phylla (7. G. Anders.) Conert word beskryf en ge'tl-

lustreer. Hierdie twee naverwante spesies is feitlik on-

uitkenbaar op anatomiese kenmerke alhoewel /mile

’n sekere mate van ekologiese en morfologiese skei-

ding toon. Die sterk anatomiese ooreenkoms tussen

hierdie twee spesies bevraagteken hulle spesifieke

status veral wanneer vergelykings get ref word met die

aansienlike anatomiese verskille wat in M. stricta

( Schrad .) Conert (Ellis, 1980a) en M. disticha ( Nees )

Conert (Ellis, 1980) waargeneem is.

REFERENCES

Anderson, J. G., 1960. Notes and new records of African plants:

Gramineae. Bolhalia 7: 419-420.

Anderson, J. G., 1962. Notes and new records of African plants:

Gramineae. Bolhalia 8: 170-172.

Chippindall, L. K. A., 1955. In D. Meredith, The grasses and

pastures of South Africa. Johannesburg: CNA.

Conert, H. J., 1970. Merxmuellera, eine neue Gattung der

Gramineen. Senckenberg. Biol. 51: 129-133.

Edwards, D., 1967. A plant ecology survey of the Tugela River

Basin, Natal. Mem. bot. Surv. S. Afr. 36: 1-285.

Ellis, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. 1. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in sur-

face view. Bothalia 12: 641-672.

Ellis, R. P., 1980. Leaf anatomy of the South African Dantho-

nieae (Poaceae). II. Merxmuellera disticha. Bothalia 13:

185-189.

Ellis, R. P., 1980a. Leaf anatomy of the South African Dantho-

nieae (Poaceae). 111. Merxmuellera stricta. Bothalia 13:

191-198.

Killick, D. J. B., 1963. An account of the plant ecology of the

Cathedral Peak area of the Natal Drakensberg. Mem. bot.

Surv. S. Afr. 34: 1-178.

Killick. D. J. B., 1978. Notes on the vegetation of the Sani Pass

area of the southern Drakensberg. Bothalia 12: 537-542.

Sch weikerdt. H. G., 1938. Descriptions of, and notes on, South

African grasses. Feddes Reprium 43: 88-92.

Bothalia 13, 3 & 4: 493-500 (1981)

Leaf anatomy of the South African Danthonieae (Poaceae).

V. Merxmuellera macowanii, M. davyi and M. aureocephala

R. P. ELLIS*

ABSTRACT

Transverse sections and abaxial epidermal scrapes, of herbarium and freshly fixed leaf blade material, of Merx-

muellera macowanii (Stapf) Conert, M. davyi (C. E. Hubb.) Conert and M. aureocephala (J. G. Anders.) Conert,

were examined using light microscopy. The leaf anatomy of these three species is very similar in all respects with the

exception of certain M. aureocephala specimens. In addition, the anatomy indicates a relationship between these

three species and M. disticha (Nees) Conert. This group of species differs anatomically from M. stricta (Schrad.)

Nees, and related species such as M. drakensbergensis (Schweick.) Conert and M. stereophylla (J. G. Anders.)

Conert, in the sequence of vascular bundles along the width of the leaf blade and associated characters. However,

the M. aureocephala specimens, not having the M. disticha type of vascular bundle arrangement, anatomically

resemble the M. stricta group of species, and M. aureocephala appears to be intermediate between these two species

groups.

RESUME

L’ANATOMIE DE LA FEUILLE DU DANTHONIEAE (POACEAE) SUD AFRICAIN. V. MERXMUELLERA

MACOWANII, M. DAVYI ET M. AUREOCEPHALA

Des sections transversales et des grattages epidermaux abaxiaux, d ’herbarium etde materiel de feuille fraichement

fixee, de Merxmuellera macowanii (Stapf) Conert, M. davyi (C. E. Hubb) Conert et M. aureocephala (J. G.

Anders.) Conert, ont ete examinees en utilisant la microscopie lumineuse. L’anatomie de la feuille de ces trois

especes est tres similaire dans tous les domaines a /’exception de certains specimens de M. aureocephala. De plus,

I’anatomie indique une relation entre ces trois especes et M. disticha (Nees) Conert. Ce groupe d’especes differe

anatomiquement deM. stricta (Schrad.) Nees, et des especes apparentees telles que M. drakensbergensis (Schweick.)

Conert et M. stereophylla (J. G. Anders.) Conert, dans la succession des faisceaux vasculaires le long de la largeur

de la feuille et des caracteres associes. Cependant, les specimens de M. aureocephala, n ’ayant pas le type d’arrange-

ment de faisceaux vasculaires de M. disticha ressemblent anatomiquement au groupe d’especes M. disticha et

M. aureocephala apparait etre intermediaire entre ces deux groupes d’especes.

INTRODUCTION

Merxmuellera macowanii (Stapf) Conert ( = Dan-

thonia macowanii Stapf), M. davyi (C. E. Hubb.)

Conert ( = D . davyi C. E. Hubb.), and M. aureoce-

phala (J. G. Anders.) Conert ( = D . aureocephala J.

G. Anders.) (Conert, 1970) are all wiry, tufted, tus-

sock-forming, perennial grasses. M. macowanii, in

particular, forms large, lax tussocks up to 60 cm in

diameter with leaves up to 100 cm long arching out-

wards from the tuft base. M. macowanii and M.

davyi are summer-flowering, whereas M. aureoce-

phala is a winter-flowing species.

These three species occur in mountain vegetation

along the eastern escarpment of southern Africa. M.

davyi is found at altitudes above 2 000 m on Mt

Mlanje in Malawi, the Inyanga mountains of Zim-

babwe and Mariepskop in the eastern Transvaal Dra-

kensberg (Conert, 1975). M. macowanii occurs from

the Transvaal Drakensberg southwards as far as the

Witteberge, Stormberge and Amatole Mountains of

the eastern Cape. It occurs between 1 500 and 3 000

m and is also found in the midlands of Natal. M.

aureocephala appears to be localized and restricted to

the high Drakensberg of Natal in the Cathedral and

Cathkin Peak areas.

M. macowanii is frequently dominant along

streambanks and in marshy areas of the montane and

subalpine belts of the Drakensberg (Killick, 1963;

Edwards, 1967) but is, nevertheless, a xeromorphic

grass with sclerophyllous leaves. M. davyi and M.

aureocephala, on the other hand, prefer more xeric

habitats and occur on steep grassy slopes and in

rocky situations in mountain grassveld (Anderson,

1962).

*Botanical Research Institute, Department of Agriculture

and Fisheries, Private Bag X101, Pretoria, 0001.

These habitat preferences bear striking resemblan-

ces to the niches occupied by M. drakensbergensis

(Schweick.) Conert and M. stereophylla (J. G.

Anders.) Conert (Ellis, 1981). Furthermore, M.

macowanii and M. drakensbergensis, both of which

occupy mesic streambank and seepage habitats, dis-

play vegetative similarities in that the old leaf blades

break off a short distance above the ligule, split along

the mid-vein and then recurve outwards (Chippin-

dall, 1955; Anderson, 1960). The above five species

are considered to form a more or less closely related

group within the genus (Anderson, 1962) and, there-

fore these ecological and morphological parallels are

not unexpected.

The present study examined these relationships

anatomically and indications are that two groups ac-

tually exist within these five species. M. drakensberg-

ensis and M. stereophylla, therefore, display more

anatomical similarities with each other than with

either M. macowanii or M. davyi. It is significant

that this anatomical sub-division separates species

occupying similar niches and exhibiting similar old

leaf blade behaviour. M. aureocephala specimens ap-

pear to be somewhat intermediate anatomically and

possibly form a link between these two groups.

M. macowanii, M. davyi and M. aureocephala

resemble one another anatomically, and, therefore, a

combined description of their leaf blade anatomy will

suffice. The terminology of Ellis (1976, 1979) will be

used in the description with the following abbrevia-

494 LEAF ANATOMY OF THE SOUTH AFRICAN D ANTHONIE AE (POACEAE). V. MERXMUELLERA MACOWANII

M. DAVYI AND M. AUREOCEPHALA

COMBINED ANATOMICAL DESCRIPTION OF

MERXMUELLERA MACOWANII, M. DA VYI AND

M. AUREOCEPHALA

Leaf in transverse section

Leaf outline : infolded with reduced U- or

V-shaped outline; opening to at least 45° possible

(Fig. 2) except in certain M. aureocephala specimens

which are permanently infolded with elliptical out-

lines (Figs 7 & 8). Adaxial channel deep and either

cleft-like or variable depending on degree of infold-

ing prevailing. Lamina always assymmetrical about

the median vb; l’vbs of opposite halves of lamina

alternate and an extra 3’vb is usually present in one

half e.g. four 3’vbs in the lower half and three in the

upper half of Fig. 1. This assymmetry occurs in all

specimens except typical M. davyi specimens (Figs 4

& 5). Leaf size : the total number of vbs in the leaf

section varies from 13-17 in M. aureocephala, 15-17

in M. davyi and 17—27 in M. macowanii. Leaf

thickness varies between 0,35-0,45 mm in M.

aureocephala and M. davyi but up to 0,55 mm in

M. macowanii. Ribs and furrows : massive adaxial

ribs with rounded to triangular apices associated with

l’vbs and small triangular ribs with 3’vbs. Medium

depth adaxial furrows between all vbs; cleft-like

depending on degree of infolding of leaf; Y-shaped

as massive ribs almost meet laterally and then furrow

diverges on either side of rib over 3’vb (Fig. 11);

found in all specimens except certain M. aureoce-

phala specimens where lateral l’vbs are not inter-

spaced by 3’vbs (Figs 7, 8 & 10). Abaxial surface

smooth. Median vascular bundle: present, character-

istically smaller than lateral l’vbs. Vascular bundle

arrangement : no 2’vbs present; l’vbs and 3’vbs alter-

Figs I - 1 1 . — Anatomy of the leaf blade in transverse section of Merxmuellera macowanii, M. davyi and M. aureocephala.

I 3, 6, M. macowanii, all x 100, considerable variation in leaf width evident. (I , Story 476; 2, Ellis 2394; 3, Ellis 3282;

6, Codd & De Winter 3239.) 4-5, M. davyi, x 100. (4, Davidson & Mogg 33315; 5, Van der Schijff 5832.) 7-9,

M. aureocephala, all X 100, note sequence of vascular bundles. (7, Killick 3540; 8, Ellis 3179; 9, Edwards 843.) 10, M.

aureocephala, X 400. ( Ellis 3179.) 1 1, AT. macowanii, x 400, phloem divided into two groups by intrusive fibres. ( Ellis

3282.)

R. P. ELLIS

495

nate along width of lamina except near margin where

two or more consecutive l’vbs may be present; at

least two 3’vbs occur between the median vb and

these successive l’vbs near the margin (Table 1) ex-

cept in certain specimens of M. aureocephala (Figs 7

& 8) where only a single 3’vb may be present on either

side of the median vb followed by four or five l’vbs.

All vbs located in centre of blade. Vascular bundle

structure : vbs elliptical (Fig. 1 1) or round (Fig. 10) in

outline; xylem and phloem distinguishable in 3’vbs;

phloem of l’vbs divided into two similar groups by

intrusion of fibres (Figs 10 & 11). Protoxylem vessel

and lysigenous cavity present; metaxylem vessels cir-

cular, of slightly greater diameter than obs cells.

Vascular bundle sheaths', obs elliptical or horseshoe-

shaped with wide adaxial interruptions; interruption

especially pronounced in some M. aureocephala

specimens (Fig. 10) such that obs only present oppo-

site xylem; these specimens without adaxial interrup-

tions or extensions (Fig. 10). In all other specimens

adaxial extensions present; of colourless cells

gradually decreasing in size as walls increase in thick-

ness until they merge into sclerenchyma strand (Fig.

11). Obs cells slightly larger in diameter than meso-

phyll cells; all similar in shape; rounded; without

chloroplasts. Ibs entire with uniformly thickened

walls (Fig. 1 1) or with inner tangential walls thicken-

ed (Fig. 10) in certain M. aureocephala specimens.

Sclerenchyma: adaxial girders inversely anchor-

shaped with long, wide stem on all bundles. Abaxial

sclerenchyma continuous sub-epidermal band of

varying thickness, with large trapezoidal girders ex-

tending to, and interrupting, the obs. Fibres either

heavily lignified (Fig. 10) or resemble collenchyma in

section (Fig. 11). Marginal sclerenchyma cap small

and pointed. Mesophyll: arrangement non-radiate;

cells uniform, small, isodiametric and tightly packed.

Restricted to Y-shaped groups on sides and bases of

furrows. Arms of Y uneven due to difference in size

of adaxial ribs associated with l’vbs and 3’vbs (Fig.

1 1) except in certain M. aureocephala specimens (Fig.

10). Colourless cells : absent. Adaxial epidermis:

restricted groups of 3-4 bulliform cells present at

base of furrows; better developed in M. macowanii

(Fig. 11) than M. aureocephala (Fig. 10) and M.

davyi. In M. davyi prickles with straight, broad barbs

and without bulbous bases common and well-devel-

oped (Figs 4 & 5); present in M. macowanii to a

slightly lesser degree (Figs 1-3) but absent in certain

M. aureocephala specimens where adaxial epidermis

consists of papillate cells. Abaxial epidermis: no

bulliform cells; outer periclinal wall thickened and

covered by continuous, thickened cuticle. No macro-

hairs, prickles or papillae occur.

Abaxial epidermis

Intercostal zone : undifferentiated; entire abaxial

epidermis essentially costal in structure (Figs 14 & 16)

due to hypodermal sclerenchyma development.

Stomata: absent. Prickle hairs: not present. Micro-

hairs: not developed on abaxial surface. Macro-

hairs: absent. Silica bodies: elliptical (Fig. 13) to tall

and narrow (Fig. 15); outlines smooth. Closely asso-

ciated with cork cell or pair of short cells. Width of

silica bodies narrower than adjacent costal short and

long cells (Fig. 13). Silica bodies sparsely developed

or even absent (Fig. 17). Costal cells: silica cells and

cork cells, either singly or in pairs, alternate with cos-

tal long cells throughout abaxial epidermis. Long

cells elongated; at least 3 x longer than wide; sides

parallel; anticlinal walls heavily thickened and slight-

ly undulating (Fig. 13) to strongly corrugated (Fig.

17).

Specimens examined:

M. macowanii

Transvaal. — 2530 (Lydenburg): Dullstroom (-AC), Codcl &

De Winter 3239 ; De Winter & Codcl 183. 2730 (Vryheid): Wakker-

stroom (-AD), Devenish 1152.

O.F.S. — 2828 (Bethlehem): Golden Gate Highlands National

Park (-DA), Ellis 2394.

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(-CC), Ellis 1455, 3282; Killick 1090. 2929 (Underberg): Estcourt

(-BB), Acocks 10659. 2930 (Pietermaritzburg): Pietermaritzburg

(-AC), Edwards 2673; Greytown (-BA), Ellis 3372.

Cape — 3027 (Lady Grey): Barkly East (-DC), Joubert s.n.

(Matatiele): Naude’s Nek (— CA), Story 476. 3126 (Queenstown):

Buffelsfontein (-BC), Stretton 182.

M. davyi

Transvaal. — 2430 (Pilgrim’s Rest): Mariepskop (-DB), Van

der Schijff 5832; Wedermann & Oberdieck 1908; God’s Window

(-DD), Davidson & Mogg 33315.

M. aureocephala

Natal. — 2829 (Harrismith): Cathedral Peak Forest Reserve

(— CC), Ellis 3179, Killick 3450, 1727; Mweni Pass, Edwards 843.

2929 (Underberg): Cathkin Peak area (-AB), Edwards 2453.

DISCUSSION AND CONCLUSIONS

The leaf anatomy of M. macowanii and M. davyi

is remarkably similar — both the leaf in transverse

section (Figs 1-6) and the abaxial epidermis (Figs

12-15). From the limited number of M. davyi speci-

mens available for examination in this study (none of

which was collected and fixed in the field), the only

difference detected was a tendency for M. davyi

leaves to be narrower with fewer vascular bundles

per section. This is not a distinct difference, however,

and several specimens overlap in this characteristic

(Table 1). M. macowanii and M. davyi are considered

to be closely allied (Anderson, 1962) and their leaf

anatomy supports this close relationship. However,

this anatomical evidence casts some doubt on the

specific status accorded these two taxa and a closer

comparison of these two species is necessary.

In contrast to the leaf anatomy, M. macowanii and

M. davyi seem well separated ecologically and oc-

cupy different habitats — mesic streambank and

seepage areas (Killick, 1963; Edwards, 1967) as op-

posed to drier rocky situations (Anderson, 1962). In

addition, these two species are almost entirely

separated geographically with only a small area of

possible sympatry in the eastern Transvaal at Mariep-

skop and God’s Window (Fig. 18). M. davyi extends

northwards into central Africa along the eastern

mountains, whereas M. macowanii occurs south-

wards as far as the north-eastern Cape.

Spikelet differences also appear to adequately dif-

ferentiate these two species and the degree of fusion

and the awned nature of the lemma lobes appear to

be distinctive. In this respect M. macowanii and M.

davyi apparently differ considerably and M. davyi

actually bears a stronger resemblance to M.

aureocephala than to M. macowanii which has

characteristic adnate, awnless lemma lobes (Ander-

son, 1962).

Theoretically these ecological and morphological

differences appear to be diagnostic yet in practice

their application seems to have been inconsistent.

Thus the specimens collected at God’s Window' and

Mariepskop ( Davidson & Mogg 33315 and Van der

Schijff 5832) were initially identified as M. maco-

496 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). V. MERXMUELLERA MACOWANII

M. DA VYI AND M. AUREOCEPHALA

Figs 12-17. — The abaxial epidermis of Merxmuellera macowanii, M. davyi and M. aureocephala as seen in surface view.

12-13, M. macowanii. (12, De Winter & Codd 183, X 250, large number of single costal short cells without silica bodies;

13, Stretton 182, X 1 000, irregular short cell arrangement.) 14-15, M. davyi. (14, Davison & Mogg 33315, X 250, only

single short cells occur; 15, Wedermann & Oberdieck 1908, x 1000.) 16-17, M. aureocephala. (16, Killick 3450, x 250;

1 7 , Ellis 3179, x 640.)

wanii and have only recently been assigned to

M. davyi (Conert, 1975) — notwithstanding the fact

that a key was published in 1962 specifically to facili-

tate the identification of M. davyi, M. macowanii

and other closely related species (Anderson, 1962). In

addition, the specimen Codd & De Winter 3239 has

similarly proved difficult to identify satisfactorily. In

1947 it was named M. macowanii, changed to

M. davyi in 1975 but again placed in M. macowanii

during the present study. Anatomical indications are

that this specimen is better placed in M. macowanii

(Fig. 6).

Therefore, although anatomical evidence appears

to be in conflict with morphological and ecological

indications, closer analysis shows that M. macowanii

and M. davyi are, in fact, not consistently separable

and are probably very closely related. In the light of

the above evidence, a reassessment of their specific

status, therefore, appears justified.

An additional consideration, which must be borne

in mind when assessing the taxonomic status to be ac-

corded these two taxa, is the almost identical situa-

tion observed in M. drakensbergensis and M. stereo-

phylla (Ellis, 1981). These two species are also in-

separable on anatomical grounds and a gradation in

leaf size and vascular bundle number was also noted.

Several other parallels exist between these two pairs

of species. M. drakensbergensis and M. macowanii

both occupy mesic, damp habitats and both display

characteristic behaviour of the old leaf blades. M.

stereophylla and M. davyi occur in drier, rocky situa-

tions and tend to have narrower leaves.

M. macowanii and M. davyi have purposely been

considered separately from M. drakensbergensis and

R. P. ELLIS

497

Fig 18. — Distribution of Merxmuellera macowanii (A), M. davyi

( ♦) and M. aureocephala (▼) in southern Africa. Shaded

symbols represent localities of specimens studied anatomical-

ly. Compiled from specimens at the National Herbarium, Pre-

toria (PRE).

M. stereophylla, even although they share so many

common characteristics. This is because a distinct

anatomical attribute consistently separates these two

species pairs— the arrangement of the different

orders of vascular bundles along the width of the

lamina. In M. macowanii and M. davyi at least two

alternating pairs of first and third order bundles,

commencing with the median bundle, are present

before consecutive lateral first order bundles are en-

countered (Table 1). In M. drakensbergensis and M.

stereophylla, on the other hand, only a single third

order bundle is present between the median bundle

and successive lateral first order bundles. The alter-

nating sequence of first and third order bundles is

correlated with several other anatomical characters

such as form of the adaxial furrows and shape of the

mesophyll cell groups.

It is possible that position along the length of the

leaf blade may affect this arrangement of the first

and third order vascular bundles. However, it has

been shown that, towards the apex of the lamina, the

lateral veins disappear one by one, commencing with

the marginal pair, until the median bundle remains to

form the pungent tip (Burbidge, 1946). It is unlikely,

therefore, that the bundle sequence, as noted here,

will be affected by the position of the sections ex-

amined as all material was taken from the central

third of the leaf blade between ligule and apex (Ellis,

1976) and, in addition, the relevant bundle sequence

is not marginal but adjacent to the median bundle or

midrib (Table 1).

Up to this point M. aureocephala has been exclu-

ded from this discussion because, in respect of this

anatomical difference, it does not conform with

either of the vascular bundle sequences described

above but exhibits mixed and intermediate condi-

tions. Fig. 7 illustrates the condition typical of M.

drakensbergensis and M. stereophylla where a single

third order bundle is followed by successive first

order bundles. Fig. 9, on the other hand, shows the

alternating pattern of first and third order bundles

characteristic of M. macowanii and M. davyi.

M. aureocephala, therefore, exhibits both types of

bundle arrangement that consistently separate M.

macowanii and M. davyi from M. drakensbergensis

and M. stereophylla. In addition, similar differences

are found in different leaves of the same plant and

even in single leaves. Thus leaf samples were taken

from duplicate specimens of Killick 3450 and one

showed the M. drakensbergensis/ M. stereophylla

type of bundle sequence (Fig. 7) whereas the other

showed the M. macowanii/M. davyi type (Table 1).

An exceptional case is illustrated in Fig. 8 where each

of these two types of vascular bundle arrangement

occur within a single leaf — one type in each half of

the lamina.

The presence of all these intermediates in M.

aureocephala casts some doubt on the importance at-

tached to this difference in bundle arrangement in the

present study. M. aureocephala, therefore, appears

to hold the key to the understanding of relationships

within this group of closely related species. Unfortu-

nately, however, the taxonomic status of M. aureoce-

TABLE 1. — The arrangement of first (1 ') and third (3') order vascular bundles along the leaf blade from the median vascular bundle

(midrib) to the margin

Specimens examined

Midrib

Vascular bundle sequence

498 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). V. MERXMUELLERA MACOWANII

M. DA VYI AND M. AUREOCEPHALA

phala itself appears somewhat tenuous. To date this

species is only known from six collections, all from a

restricted area, of less than 20 km in diameter, in the

subalpine belt of the Cathedral and Cathkin Peak

areas of the Drakensberg (Fig. 18). Morphologically

it is very similar to M. davyi except that the spikelets

are larger in all parts, and the glumes are lanceolate

instead of narrowly lanceolate (Anderson, 1962). Its

winter- flowering habit, in fact, is the single diagnos-

tic character separating M. aureocephala from its

four close relatives. However, the specimens assigned

to M. aureocephala may actually represent examples

of late or early flowering in the other Merxmuellera

species, e.g. Edwards 2284 has been determined by

Conert 1973 as being M. aureocephala, but it is now

considered as being M. stereophylla. Winter visits to

these inhospitable mountains, to study fertile field

populations, seem essential to a better understanding

of the taxonomic status of all the summer-rainfall

area Merxmuellera species.

Until these field studies have been undertaken, the

true significance of the different vascular bundle ar-

rangement sequences cannot be assessed. However,

assuming that these patterns are a phylogenetically

important difference, the available evidence indicates

that M. aureocephala occupies a basic systematic

position in this group from which each of the two

distinct types have been derived. This implies a close

relationship for these five species as postulated by

Anderson (1962) as well as an origin in the Drakens-

berg mountains and not in the temperate, winter-

rainfall areas of the Cape.

If the distribution of these two types of bundle ar-

rangement is examined in all the summer-rainfall

Merxmuellera species, however, a different origin

seems likely. M. disticha (Nees) Conert, including

each of its anatomical forms, exhibits the alternating

sequence of first and third order vascular bundles

(Ellis, 1980). M. disticha, therefore, shares this

character with M. macowanii, M. davyi and some M.

aureocephala specimens (Table 2). All the four M.

stricta (Schrad.) Conert anatomical forms (including

M. guillarmodiae Conert), on the other hand, have

similar bundle arrangement to M. drakensbergensis

and M. stereophylla as well as other M. aureocephala

specimens (Ellis, 1980a) (Table 2). Thus, within this

group of 12 summer-rainfall Merxmuellera taxa

(Table 3), M. aureocephala remains the only taxon

intermediate for this anatomical character.

TABLE 2. — The arrangement of first (1') and third (3') order

vascular bundles along the leaf blade from median vascular bundle

to margin in the summer-rainfall Merxmuellera species

Table 3 diagramatically illustrates each of these

twelve Merxmuellera taxa arranged according to

vascular bundle sequence and grouped into the

various habitats occupied by these various taxa. It is

immediately evident from Table 3, that in each of the

niches occupied by Merxmuellera spp. a taxon

displaying each of the bundle sequence types occurs.

A representative of each type of bundle arrangement

occurs in the cave sandstone, basaltic soils, alpine

bogs, alpine xeric sites and mesic sites. In addition,

morphological and anatomical similarities often exist

between these pairs of taxa inhabiting similar niches

e.g. the old leaf blade behaviour in M. macowanii

and M. drakensbergensis and the mesophyll and epi-

dermal structure in the alpine bog forms of M. stricta

and M. disticha. Once again, M. aureocephala is the

exception.

Typical M. stricta and M. disticha forms are both

widespread in the Cape and extend to lower altitudes

in the Drakensberg. Throughout this wide distribu-

tional range both species are very uniform in both

morphology and anatomy. It is only at higher alti-

tudes, above the cave sandstone, that anatomical and

morphological diversification is prevalent. An alter-

native hypothesis is, therefore, that M. stricta and

M. disticha have independently colonized the wide

variety of microhabitats present at higher altitudes

in the Drakensberg by evolving locally adapted eco-

types for each of the various niches. This adaptive ra-

diation, in response to identical environmental condi-

tions, has resulted in very similar phenotypic expres-

sions by the ecotypic forms of each species. As M.

stricta and M. disticha occur sympatrically through-

out most of their ranges it must be assumed that they

originally possessed distinct but diverse genotypes

which somehow were capable of responding in

similar ways to the different environmental condi-

tions encountered in the Drakensberg. This appears

to explain the unique morphological and anatomical

convergence observed in all the habitats occupied.

The bundle sequence may, therefore, reflect a basic

genetic different between M. stricta and M. disticha

ancestral forms that has been retained in all the

ecotypic forms.

This hypothesis does not explain the position of M.

aureocephala. If the origin of all these taxa is to be

sought in putative ancestors of M. stricta and M. dis-

ticha then the only explanation for M. aureocephala

lies in a hybrid origin. Once again population and cy-

togenetical studies seem necessary to elucidate this

question.

From these antomical studies on the summer-rain-

fall Merxmuellera species (Ellis, 1980; 1980a; 1981) it

is nevertheless clear that, at this stage, at least 12 enti-

ties can be recognized. A further two, presently plac-

ed in Pentaschistis, also merit consideration (Ellis,

1980a). All these taxa are undoubtedly interrelated

and the most practical systematic treatment, at this

stage, appears to be the upholding of only two basic

species (M. stricta and M. disticha) with numerous

infraspecific taxa, possibly of subspecific rank, in-

cluded in each. M. guillarmodiae, M. macowanii, M.

davyi, M. drakensbergensis and M. stereophylla

should be reduced to subspecific rank, whereas, the

anatomical forms of M. stricta and M. disticha

justify taxonomic recognition with subspecific status

as well. M. aureocephala is the one entity on which

the present studies have shed very little light and a

taxonomic recommendation at this stage would be

unwise.

R. P. ELLIS

499

TABLE 3. — Diagrammatic representations of the leaf anatomy of the summer-rainfall Merxmuellera taxa according to

habitat and vascular bundle arrangement

500 LEAF ANATOMY OF THE SOUTH AFRICAN DANTHONIEAE (POACEAE). V. MERXMUELLERA MACOWANII,

M. DA VYI AND M. AUREOCEPHALA

ACKNOWLEDGEMENTS

Miss R. Manders is thanked for excellent technical

assistance as are Dr B. de Winter and Miss L. Smook

for identifying the voucher specimens. The Depart-

ment of Forestry, the National Parks Board and the

Natal Parks Board kindly gave permission to collect

material on their property.

UITTREKSEL

Dwarssnitte en abaksiale epidermale skrapings,

van herbarium, asook vars gefikseerde blaarmate-

riaal, van Merxmuellera macowanii ( Stapf) Conert,

M. davyi (C. E. Hubb.) Conert en M. aureocephala

( J . G. Anders.) Conert is met behulp van ’n ligmikro-

skoop ondersoek. Die blaaranatomie van hierdie drie

spesies is in alle opsigte dieselfde, met die uitsonde-

ring van sekere M. aureocephala eksemplare. Verder

toon die anatomie ’n verwantskap tussen hierdie drie

spesies en M. disticha (Nees) Conert. Hierdie spesies-

groep verskil anatomies van M. stricta ( Schrad .)

Conert, en verwante soorte soosM. drakensbergensis

( Schweick .) Conert en M. stereophylla (J. G.

Anders.) Conert, in die volgorde van die vaatbondels

langs die blaarwydte en geassosieerde kenmerke. Die

M. aureocephala eksemplare wat nie die M. disticha

tipe vaatbondelvolgorde toon nie, lyk anatomies na

die M. stricta groep van spesies, en M. aureocephala

is dus intermedier tussen hierdie twee spesiegroepe.

REFERENCES

Anderson, J. G., 1960. Notes and new records of African plants:

Gramineae. Bothalia 7: 419-420.

Anderson, J. G., 1962. Notes and new records of African plants:

Gramineae. Bothalia 8: 170-172.

Burbidge, N. T., 1946. Morphology and anatomy of the western

Australian species of Triodia R. Br. II. Internal anatomy.

Trans. R. Soc. S. Aust. 70: 221-234.

Chippindall, L. K. A., 1955. In D. Meredith, The grasses and

pastures of South Africa. Johannesburg: CNA.

Conert, H. J., 1970. Merxmuellera, eine neue Gattung der

Gramineen. Senckenberg. Biol. 51: 129-133.

Conert, H. J., 1975. Merxmuellera guillarmodae Conert n.sp.

Senckenberg. Biol. 56: 145-152.

Edwards, D., 1967. A plant ecology survey of the Tugela River

Basin, Natal. Mem. bot. Surv. S. Afr. 36: 1-285.

Ellis, R. P., 1976. A procedure for standardizing comparative

leaf anatomy in the Poaceae. I. The leaf blade as viewed in

transverse section. Bothalia 12: 65-109.

Ellis, R. P., 1979. A procedure for standardizing comparative

leaf anatomy in the Poaceae. II. The epidermis as seen in sur-

face view. Bothalia 12: 641-672.

Ellis, R. P., 1980. Leaf anatomy of the South African Dantho-

nieae (Poaceae). II. Merxmuellera disticha. Bothalia 13:

185-189.

Ellis, R. P., 1980a. Leaf anatomy of the South African Dantho-

nieae (Poaceae). III. Merxmuellera stricta. Bothalia 13:

191-198.

Ellis, R. P., 1981. Leaf anatomy of the South African Dantho-

nieae (Poaceae). IV. Merxmuellera drakensbergensis and M.

stereophylla. Bothalia 13: 487-491.

Killick, D. J. B., 1963. An account of the plant ecology of the

Cathedral Peak area of the Natal Drakensberg. Mem. bot.

Surv. S. Afr. 34: 1-178.

Bolhalia 13, 3 & 4: 501-518 (1981 )

Plants used by the Tsonga people of Gazankulu

C. A. LIENGME*

ABSTRACT

A study was undertaken in part of the Tsonga homeland, Gazankulu, to identify plants used by these people. A

list of Tsonga plant names was extracted from a Tsonga-English dictionary and this was used as a basis for the study.

The uses of almost 200 plants were recorded, including medicine, food, building materials, firewood, household

utensils, implements, implement handles and toys. This information is presented in the form of an annotated list of

the plants. This is followed by a discussion of some of the more important uses. Some of the aspects of Tsonga

taxonomy are briefly discussed and illustrated with examples.

RESUME

PLANTES UT1LISEES PAR LE PEUPLE TSONGA DE GAZANKULU

Une etude a ete entreprise dans une partie du pays Tsonga, Gazankulu, afin d ’itentifier les plantes utilisees par ces

personnes. Une liste de noms Tsonga de plantes a ete extraite d’un dictionnaire Tsonga-Anglais et ceci fut utilise

comme base pour I’etude. Les utilisations de presque 200 plantes frent enregistrees, incluant les medicaments,

l' alimentation, les materiaux de construction, le bois de chauffage, les ustensils menagers, I'outillage, les manches

d ’outils et les jouets. Cette information est presentee sous forme d ’une liste annotee de plantes. Ceci est suivi d 'line

discussion de certaines des plus importantes utilisations. Certains aspects de la taxonomie Tsonga sont brievement

discutes et illustres avec des exemples.

INTRODUCTION

Much of the traditional culture and knowledge of

the tribal people in southern Africa is in danger of

being lost unless it is recorded. Tribal plant uses and

botanical knowledge is of more than academic or his-

torical importance and may be linked directly to

plant utilization and conservation. The Botanical Re-

search Institute has set iself the task of recording this

information, commencing with a project in the

northern Transvaal. The main aim of this project was

to obtain a record of plants used by the Tsonga peo-

ple of Gazankulu. A second aim was to test approa-

ches to the gathering of ethnobotanical information

prior to embarking on more extensive studies of the

impact of tribal peoples on their environment.

The reasons for choosing Gazankulu as the study

area were threefold. Firstly, it is a tribal area where

traditional uses of plants have persisted. Secondly, it

is a homeland area where a thorough knowledge of

the plant-based tribal economy will be useful in

future land-use and environmental planning. Third-

ly, the author spoke the language, could communi-

cate easily with the people and knew the area.

Study Area

The Gazankulu homeland consists of four separate

units situated in the Transvaal Lowveld (see Fig. 1).

The main block of the homeland, comprising the dis-

tricts of Giyani and Malamulele, was chosen as the

study area for this project. It is approximately

450 000 hectares in extent, lying west of the Kruger

National Park, between the Levubu River to the

north and the Letaba River to the south, with an arm

in the middle stretching westwards to Elim Hospital.

Topography

The study area includes plateau areas (about 1 000

m altitude) and steep slopes and valleys at the plateau

edge; but most of the area consists of low-lying plains

(approx. 300-650 m altitude) with scattered hills.

*Botanical Research Institute, Department of Agriculture

and Fisheries, Private Bag X101, Pretoria, 0001.

In this preliminary study no attempt was made to

define these areas or the plant uses peculiar to them.

Climate

The area has a hot, wet summer and a cool, dry

winter. Spring is generally hot and dry and autumn

warm and moist. The mean annual rainfall varies

from 500 mm in the south-east to 1 000 mm in the

502

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

Fig. 2. — Acocks’s veld types oc-

curring in the study area. Re-

drawn from Acocks (1975).

west. Eighty to ninety percent of the rain falls in sum-

mer. Temperatures range from a mean winter mini-

mum of 8°C to a mean summer maximum of 30°C.

(Department of Co-operation and Development,

pers. comm.).

Vegetation

The vegetation of the area includes three veld-

types: Lowveld Sour Bushveld, Arid Lowveld and

Mopane Veld (Acocks, 1975) (See Fig. 2).

The People

The life of the Tsonga people has been covered in

great detail by Junod (1962). The Tsongas are basic-

ally agriculturalists, cultivating crops and keeping

cattle and goats. They are dependent on the indige-

nous vegetation for many things, ranging from struc-

tural timber to supplementary food and medicine.

Historically, the Tsongas were an east coast people

who occupied the southern half of Mozambique, the

adjacent eastern edge of the Transvaal and north-

eastern Natal (Tongaland). Many Tsongas migrated

to the eastern Transvaal lowveld and adjoining es-

carpment areas on a number of occasions during the

nineteenth century; firstly, during 1835-1840 when

their country was conquered by Nguni who left

Natal; secondly, during the ‘War of Succession’ be-

tween 1856 and 1862; and thirdly, during the war be-

tween the Portuguese and the Tsongas in 1894 and

1895 (Junod, l.c.). Many also migrated to the mining

towns of the Transvaal in the early twentieth century.

Today a number of Tsonga settlements are found as

far west as the Waterberg District and Rustenburg

(Van Warmelo, 1974).

Demography

The growth of the Tsonga population in

Gazankulu as a whole has been rapid, from an

estimated 37 000 in 1904 to 75 570 in 1951 and

350 245 in 1976. A further rapid increase is expected

with populations of about 600 000 and 1 000 000 be-

ing projected for the years 2000 and 2020 respectively

(Department of Co-operation and Development,

pers. comm.).

In the study area the position is similar: following

population influxes and with an estimated annual

growth rate of about 5%, the area supported a

population of just under 200 000 in 1976. (Depart-

ment of Co-operation and Development, pers.

comm.). This represents an average population den-

sity of 44 persons per square kilometre, with the

western parts being the most heavily populated.

RESEARCH PROCEDURE

A list of approximately 550 Tsonga plant names

was extracted from a dictionary (Cuenod, 1976) and

was used as a starting point for the project. Field

data sheets were devised for recording the use and the

botanical name of each plant on the list as well as the

locality and the informant.

As it turned out, the inhabitants of the study area

often did not know some of the names incorporated

in the dictionary. These were names used in other

areas or referring to plants not occurring in the study

area. Many of the names were synonymous, so that

the final list of plants for which there were Tsonga

names contained just under 400 plants. Information

C. A. LIENGME

503

on uses was obtained for almost half of these plants

(190 species) and 170 specimens were collected.

The field work for the project was carried out dur-

ing three trips to the study area between April and

September 1977. An attempt was made to cover as

much of the area as possible, visiting many of the not

too inaccessible villages. The procedures used for

gathering information were either:

1. to approach people whom the author knew to

be helpful (who in turn often recommended others),

or

2. to approach people at random along roads or in

villages. These people were often engaged in par-

ticular activities that were worth recording, or reveal-

ed where to find knowledgeable people locally.

In all these contacts a knowledge of the language

and way of life of the people was invaluable.

Both men and women were approached. Women

were generally asked about plants used as food,

thatch and mats and men about plants used in

building and the making of utensils. Both men and

women were asked about basketry and medicinal

plants.

Most of the information on medicinal plants came

from herbalists and information on plants from

whose wood objects and utensils are carved from

specialist woodcarvers, although some of these plant

uses are common knowledge.

There were three approaches to questioning:

1. asking what a particular plant was used for;

either giving its Tsonga name or pointing the plant

out;

2. asking what plants in the locality were em-

ployed for various uses; and

3. on seeing a particular object or observing an ac-

tivity (such as thatching), enquiring as to the plants

used. Wherever possible, a specimen of the plant/

plants was collected for identification. Colour slides

and black and white photographs were taken, record-

ing the utilization of plants.

The approach used depended largely on the situa-

tion. The first of the three approaches is only really

effective if one points out a plant and asks about its

uses. This approach is necessary if one wants to get

both sides of the story (the plant side and the product

side) in order to cross-check on information received.

The second approach is useful for getting an idea of

which plants have localized or widespread distribu-

tion and uses.

The most rewarding approach is the third since the

informant is thoroughly familiar with the plant use.

The object (finished or unfinished) is available and it

is possible to record how the plants are prepared for

use, how they are used, what other plants are used in

conjunction with them, what part of the plant is us-

ed, how much wastage there is and what damage oc-

curs to other plants in the process of collecting.

It soon emerged that many village and household

activities are highly seasonal. Thus, continuous

recording or more regular short visits to the area

would be essential for any accurate quantitative

assessment of the use of plant material to be made.

The information on plants used by the Tsonga of

Gazankulu presented here does not cover all seasons

and should be regarded as a preliminary list only.

Verification of information on plant uses comes

from two sources; literature and cross-checking in

the field, using information from a number of infor-

mants. The latter method works quite well in the case

of plants with common, widespread uses. Specialist

uses, for example medicinal uses, are not easy to

check, since not all herbalists use the same plants.

Verification of information on plant uses comes

from two sources; literature and cross-checking in

medicinal plants. It was possible to verify the use of

only 3 of the 35 Tsonga medicinal plants collected

during this project. Fourteen other species were

either used by other tribal groups for the same pur-

pose as the Tsongas, or had related species with the

same or similar medicinal use. In many other cases

use of a plant was verified in the literature, but not

for the Tsonga specifically. The most useful publica-

tions were Van Wyk (1972), Palmer & Pitman (1972),

Watt & Breyer-Brandwijk (1962) and Shaw (1974).

Junod (1962) provided verification of a few plant

uses, but was most valuable for checking on tradi-

tional methods of thatching, building and basket-

making.

THE PLANTS AND THEIR USES

The plants have been arranged in alphabetical

order of their genus and species names. Tsonga

names appear below the botanical names. Only

original information on plant uses, gathered during

the survey, has been included. Where a voucher

specimen was collected, the collector’s number is in-

dicated in brackets after the plant name. Exotic

species are indicated by *. Indexes of the uses and of

the Tsonga names are appended to the paper.

Acacia ataxacantha DC. (Liengme 22)

muluwa

This is a fairly common tree in the Lowveld Sour

Bushveld. It is often used as fuel, providing a very

hot fire. Axe-handles are made from the wood and

several types of baskets are made from thin strips of

the wood. These strips are woven into mats and then

shaped into baskets. The most commonly-seen

baskets made in this way are: 1. a shallow saucer-

shaped basket for winnowing, called a ‘rihlelo’, and

2. a fairly deep basket with a wide round opening.

The latter normally has a winnowing basket as a lid.

Acacia caffra (Thunb.) Willd.

mbvinyaxihloka; nkaya/nkayi; nkhayani

A common tree in some parts, whose wood is hard

and durable and is used for building, fence posts and

firewood.

Acacia davyi N.E. Br. ( Liengme 57)

xisidane

This small tree is another source of firewood.

Acacia gerrardii Benth. var. gerrardii

nsasani

This tree is not very common and seems to be more

or less limited to the Lowveld Sour Bushveld areas. It

is a source of firewood.

Acacia karroo Hayne ( Liengme 9, 158)

munga

Its wood is an excellent firewood. The bark is used

as cord and is sometimes made into ropes.

Acacia nigrescens Oliv.

nkaya/nkayi

504

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

The wood is hard and heavy and is used predomin-

antly for building. It is also often used for making

pestles. This is one of the characteristic trees of the

Arid Lowveld and is plentiful.

Acacia sieberana DC. var. woodii (Burtt Davy)

Keay & Brenan

nkowankowa

This common bushveld tree is used as firewood.

Acacia tortilis (Forssk.) Hayne subsp. heteracantha

(Burch.) Brenan

ngoka; nsasani

This is another source of firewood.

Adansonia digitata L.

mowu/muwu; ximowu/ximuwu

The dry pulp of the fruit is eaten. This tree is

almost entirely confined to the Mopane Veld.

Adina microcephala (Del.) Hiern var. galpinii (Oliv.)

Hiern

muhlome/nhlume/muthuma

The wood of this tree is strong and is sometimes

used in building. It is also left standing in villages as a

shade tree. Young branches make natural stirring

sticks, used for stirring mealie meal ( Zea mays meal)

while it is cooking.

Afzelia quanzensis Welw. (Liengme 204)

nxenhe

The wood is used for building. This species occurs

on hills and ridges in the Arid Lowveld and the

Mopane Veld.

Agave sp.*

xikwenga

The fibre extracted from the leaves of this exotic is

used for making string, ropes, table mats and other

items. The fibre is still extracted manually at home,

but most people now buy the string. Prior to the in-

troduction of Agave the fibre of Sansevieria species

was probably used. In the vicinity of the large Agave

plantations scattered around Gazankulu the inflores-

cense stalks are used as fencing material.

Albizia harveyi Fourn.

mola/molani; molela

This tree is often found in villages as a shade tree.

It is also used as firewood.

Albizia versicolor Welw. ex Oliv. ( Liengme 108)

mbhesu/mbheswi; muvambangoma; mvhangazi wo

basa; mucece

The Tsonga name ‘muvambangoma’ can be trans-

lated as meaning ‘stretched out like a drum skin’

(‘-vamba’-stretch out; ‘ngoma’-drum). ‘Muvha-

ngazi wo basa’ can be translated as ‘white kiaat’

(‘muvhangazi’ is the Tsonga name for kiaat and

‘basa’ means white). The roots and bark of the tree

are apparently used to make a medicine for driving

out demons. The wood is hard and quite beautiful

and is used for carving mortars and other objects.

The tree is also regarded as a useful shade tree in

villages.

Allophyllus decipiens Radik. ( Liengme 51)

muzuzugwane xihlahla

This is a fairly common small tree used as fire-

wood.

Aloe davyana Schoenl. var. davyana

mhangani

The leaves of this aloe are used in a game played by

children. The nature of the game is uncertain.

Amaranthus cruentus L. ( Liengme 100)

nhlaba ya foie

Snuff is made from the flowering tops of the plant.

Amaranthus thunbergii Moq. ( Liengme 175)

thyeka/thyeke

This is a common ruderal herb near dwellings and

in old fields. The leaves are cooked, usually as a con-

stituent of relishes.

Annona senegalensis Pers. ( Liengme 12)

murhompfa; muyembe; ndzompfa/ndzopfa/

ndzhopfa

The pulp of the fruit is eaten, but the plant is not

common enough for it to be important in the diet.

The empty fruit is used by children to make a ‘pop-

gun’. Openings at each end of the fruit are plugged

with pieces of mealie ( Zea mays) cob. If the plug at

one end is hit hard the plug at the other end pops out.

Antidesma venosum E. Mey. ex Tul. ( Liengme 24)

mpfalambati; mphatakhwari

The fruit is eaten and the plant is one of those

whose young flexible branches are used as wattles in

building. This is a species of the Lowveld Sour Bush-

veld.

Aptosimum lineare Marloth & Engl. ( Liengme 92)

ximahlomahlwane

The juice of the leaves of this small herbaceous

plant is used by herbalists as eye-drops.

Arachis hypogea L.*

manga

Peanuts are cultivated in some parts of Gazankulu.

Artabotrys brachypetalus Benth. ( Liengme 219)

ntita/ntiti/ntinta; xivudzi

A strong fibre is obtained from this scrambling

plant, which occurs in the Lowveld vegetation types.

The fruit is eaten and the plant also has medicinal

uses.

Asclepias burchelli Schltr. (Liengme 84)

kotoni

A decoction of the roots is used for the treatment

of intestinal worms.

Asparagus virgatus Bak. ( Liengme 141)

nkungulantila/nkwangulantilo

This plant is thought to have magical properties.

When unwanted rain threatens, a plant is cut and set

alight. The youngest child in the family waves this

around to chase the rain away.

Athrixia phylicoides DC. ( Liengme 21)

kofi ya nhova

The leaves of this plant are used to make a hot

drink, like tea.

C. A. LIENGME

505

Balanites maughamii Sprague ( Liengme 107)

nulu

This small tree occurs in the vicinity of hills on the

lowveld plains. The roots are pounded and made into

a medicine for apparently driving out demons. A

musical bow called ‘xipendane’ is made from this

wood.

Bambusa sp.*

musengele

This exotic species is found growing in some

places. Stems are split and used for fences and pali-

sades.

Ban hi ni a galpinii N.E. Br.

ntshwiriri/ntshiriri

The seeds of this rambling shrub are used as beads

for necklaces. This is a distinctive species of the Low-

veld Sour Bushveld.

Bequaertiodendron magalismontanum (Sond.) Heine

& J. H. Hemsl. ( Liengme 208)

nombela

The fruit of this tree is very pleasant-tasting. The

tree occurs in the Lowveld Sour Bushveld.

Berchemia discolor (Klotzsch) Hemsl. (Liengme 118)

nyiri/ nyiyi/muwe

Pestles, axe-handles and other objects are carved

from the wood. The fruit is eaten, especially by

children.

Bide ns pilosa L.*

muxiji

This is a common ruderal around dwellings and in

fields. The leaves are cooked in relishes or as

‘spinach’.

Blumea aurita (L.f.) DC. ( Liengme 221)

munywane

This plant is placed in hot water to make a

medicine and applied to sore places on the body. It is

considered to be the male form of ‘munywane’; the

female form being a plant identified as Epaltes

gariepina.

Bolusanthus speciosus (H. Bol.) Harms

mpfimbahongonyi; nkamba/nkhamba; nkohlwane

The wood of this tree is recognized as being good

for making furniture. It is often used for implement

handles and walking sticks. There is a substance in

the roots which is reputed to have a sleep-inducing

effect.

Brachylaena discolor DC. subsp. transvaalensis

(Phill. & Schweick.) J. Paiva ( Liengme 14)

mphata

Young stems and branches are strong and pliable

and are used in making the rims of winnowing and

other baskets made from strips of Acacia ataxacan-

tha wood. The outer of the two wooden rings of the

rim is made from this plant. Young stems and bran-

ches are also used for tying together roof frame-

works. The tree is also a source of firewood. It occurs

in the Lowveld Sour Bushveld.

Bridelia micrantha (Hochst.) Baill. ( Liengme 18, 53)

mindzere/mundzere/ndzerhe

This is another species occurring in the Lowveld

Sour Bushveld. The bark is used by herbalists and

witch-doctors to make a stomach medicine. The fruit

(small berries) are eaten, particularly by children.

Bridelia mollis Hutch. ( Liengme 214, 224)

kumbekumbe; swatima; swimbyambya

The small fruit is eaten.

Burkea africana Hook,

mpulu; nkenga

This tree is rare in the study area, being found only

occasionally in the Lowveld Sour Bushveld. The

wood is recognized as being good for furniture, hav-

ing a good texture and colour and not splitting and

twisting as it dries.

Cajanus cajan (L.) Millsp.* ( Liengme 162)

ndodzi

The pigeon pea is cultivated in some of the moister

western parts of the study area.

Calodendrum capense (L.f.) Thunb.

mbhovu

The dried fruits of this tree are used to make ankle-

rattles which are worn by dancers at celebrations and

feasts.

Cap paris fassicu laris DC. var. fassicularis

muhobadale

The roots of this creeper are used for medicinal

purposes, of uncertain nature.

Capparis tomentosa Lam. ( Liengme 232, 242)

khawa; mukorongwe

The roots are used to make a stomach medicine.

The plant is fairly common in the Arid Lowveld.

Capsicum frutescens L.*

viriviri

Peppers are occasionally found in gardens.

Carissa edulis Vahl ( Liengme 11, 223)

nchungulu/nchuguru/ntshuguru

This is a common shrub of the Lowveld Sour

Bushveld, producing large quantities of fruit in mid-

to late summer. The fruit is delicious and is eaten

fresh by all.

Cassia abbreviata Oliv. subsp. beareana (Holmes)

Brenan

lumanyama/numanyama

This tree is sometimes left standing in villages as a

shade tree.

Cassia Occident a l is L. ( Liengme 44)

nembenembe

The seeds are cooked and eaten, but its importance

as a foodstuff is not known. The plant is common in

the moister areas and is often found along roads.

Cassine aethiopica Thunb. (Liengme 212)

nqayi

The wood of this tree is popular for making walk-

ing sticks.

Cassine transvaalensis (Burtt Davy) Codd (Liengme

138)

ximapana; nkubatsebi

This tree has strong wood which is used for imple-

ment handles, yokes and spoons.

506

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

Catha edulis (Vahl) Forssk. ex Endl.

rithadzi

Young branches of this tree are used as withies for

roof frameworks. The stimulant properties of the

leaves are apparently not known.

Cephalanthus natalensis Oliv.

muthondwa/ntondo

This is a fairly abundant climber in the Lowveld

Sour Bushveld. Its somewhat bitter fruit is eaten.

Clematis brachiata Thunb. ( Liengme 63)

maamba; mikoka

The leaves of this common herbaceous climber are

boiled in water and a person suffering from a head-

ache inhales the vapours while seated under a

blanket. The roots are crushed and placed in the nose

to clear up colds.

Cocculus hirsutus (L.) Diels. ( Liengme 81, 189, 237)

risotse/rixoto; xotse/xotso

This is a common climber in the Arid Lowveld and

Mopane Veld whose stem is used for making baskets,

especially conical baskets called ‘xirundzu’. During

the dry season the leaves remain green and are

sometimes cooked and eaten as a vegetable.

Colocasia antiquorum Schott * ( Liengme 34)

raboda; rupi

This exotic is sometimes found growing along

streams in the moister western parts of Gazankulu.

The leaves are cooked and eaten.

Co/ophospermum mopane (Kirk ex Benth.) Kirk ex

J. Leonard {Liengme 91)

nxanatsi

The mopane tree is the dominant woody plant of

the Mopane veld which covers a considerable part of

the study area. The wood is hard and durable and is

the major building material where it occurs, being us-

ed for roof supports, roof frameworks, fence-posts

and stockades. The wood is widely used as firewood

and is recognised as one of the best firewoods. It is

also used for making pestles, yokes and sand-sledges.

The bark is very fibrous and is used as cord. The tree

is an indirect source of food to the people, being the

food-plant of the mopane worm (the larva of the

moth Gonimbrasia belina). This worm occurs in

summer. It is collected, dried and stored for eating.

Combretum apiculatum Sond. subsp. apiculatum

{Liengme 11, 126)

mpotsa/mpoza; mugarasaka

The wood of this common tree is hard, making it

useful material for building and fence-posts.

Implement-handles and spoons are also made from

the wood.

Combretum erythrophyllum (Burch.) Sond.

{Liengme 36)

mbvuvu/mvuvu; mgupa

The wood is used as fuel, but this is not common

practice. The tree is more or less confined to stream

banks.

Combretum hereroense Schinz. subsp. hereroense

var. hereroense {Liengme 76)

mpotsa/mpoza; xikhavi

The wood is cut for building and is also used for

axe-handles and pick-handles. Various wooden items

are carved from it. The young branches are strong

and flexible and are used as withies.

Combretum imberbe Wawra {Liengme 98)

mbimba; mondzo

The wood is very hard and durable and is frequent-

ly used in building, usually for the main supporting

poles of a hut. Mortars are carved from the wood

and sometimes pestles as well. This species occurs in

the Arid Lowveld.

Combretum paniculatum Vert, subsp. microphyllum

(Klotzsch.) Wickens {Liengme 199)

chochelamandleni yanstongo; mpfunta; mpfunte-

mpfunte

The Tsonga word ‘chochelmandleni’ means ‘tap

out into the hand’, referring to the use of the plant by

children: they shake the nectar from the flowers into

their hands and then lick it up. A diminutive is form-

ed by the addition of ‘ya ntsongo’ to the name. The

tree Schotia brachypetala is the actual

‘chochelamandleni’. The plant is a straggly, scrambl-

ing shrub occurring along river banks, flowering pro-

fusely in spring.

Commiphora pyracanthoides Engl,

xifata/xifati

The wood is extremely useful and is used for

bowls, plates, axe-handles and a musical instrument

called ‘mbila’.

Corchorus confusus Wild {Liengme 193)

guxe

The leaves of this ruderal, common in villages and

along roads, are cooked in a relish or gravy and eaten

with mealie meal {Zea mays meal).

Corchorus tridens L. {Liengme 93)

guxe

This species is used in the same manner as Cor-

chorus confusus.

Cordia grandicalyx Oberm.

ntogwe; tshaman’hwati

The dried fruit is used to make ankle-rattles for

dancing.

Cordia ovalis R. Br. ex DC. {Liengme 75, 85, 227)

mpon’wana; mtlele

The young branches are flexible and are used as

withies as well as for securing thatch to roofs. The

fruit is eaten, generally by children.

Crossopteryx febrifugia (Afzel. ex G. Don) Benth.

nkombekwa/nkombelwa

Bowls and other utensils are carved from the

wood.

Croton megalobotrys Muell. Arg. {Liengme 95)

nxunguxungu/nxungwexungwe

A purgative medicine is made from the pounded

bark added to a few other ingredients. The plant is

recognized by most people as being poisonous and

dangerous. It seems to occur in the riverine vegeta-

tion.

Crytolepis capensis Schltr. (Liengme 96)

nyokani

The roots of this forb are used to make a worm

remedy. The use of the plant is indicated by its name;

‘nyoka’ means ‘snake or worm’.

C. A. LIENGME

507

Cucumis melo L.* ( Liengme 121)

rhanga

The fruit of this small melon is cooked and eaten.

It is not certain whether this plant is cultivated or

not, but specimens were found growing wild. The

fruit of the collected specimen was about 100 mm

long, oval and yellow.

Cucurbita maxima Duch.*

gawana; rhanga; xilutana

Squashes are commonly cultivated in Gazankulu,

probably more so than Cucurbita pepo.

Cucurbita pepo L.*

Pumpkins are also cultivated.

Cussonia spicata Thunb.

musenje; xipokota

Planks are sometimes cut from the wood.

Cymbopogon validus (Stapf) Stapf ex Burtt Davy

{Liengme 29, 181)

mgejo; deke

This grass is common in the Lowveld Sour Bush-

veld. It is used as thatch, particularly as the first layer

of the thatch (the ‘ceiling’). The name ‘deke’ is most

probably derived from the Afrikaans word ‘dak’.

Cyperus latifolius Poir. ( Liengme 5)

njekejeke (leaves); xigoya (culm)

Both the leaves and the culms are used to make

mats. The plant is abundant along streams in the

Lowveld Sour Bushveld.

Cyperus sexangularis Nees {Liengme 186, 246)

nhlahle; risama

This occurs along rivers and streams in the Arid

Lowveld and Mopane Veld and is used there to make

mats.

Cyphostemma humile (N.E. Br.) Desc. ex Wild &

Drum, subsp. humile {Liengme 172)

ndlejane

The leaves of this succulent are crushed and the li-

quid used as ear-drops.

Dalbergia melanoxylon Guill. & Perr. (Liengme 120)

xilutsi; xipalatsi

The wood is much sought-after by wood-carvers,

because of its beautiful black heartwood. The wood

is used for ornaments, walking sticks, knobkieries

and also for headrests. Reasonably large specimens

of the species seem to be rare.

Dicerocaryum zanguebarium (Lour.) Merr. subsp.

zanguebarium (Liengme 88)

dinda/dindza; hlwehlwe (seed)

The plant’s juice is used as a shampoo. It is com-

mon in the Arid Lowveld and the Mopane Veld.

Diochrostachys cinerea (L.) Wight & Arn. subsp.

africana Brenan & Brumitt (Liengme 25)

ncenga; ndhenga/ndzhenga

This small common tree/shrub is used as firewood.

Dioscorea cotinifolia Kunth (Liengme 28)

nsidwa; risidwa

The tough stems of this climber are used in basket-

making. They are used for binding the rims and

waists of baskets to the body of the basket.

Diospyros mespiliformis Hochst. ex A. DC.

mgula; ntoma

The fruit of this tree, common in the Lowveld Sour

Bushveld and Arid Lowveld regions, ripens in late

winter. Cultivated food can be quite scarce at this

time of the year and the fruit is thus an important

food resource. The wood is used to make mortars,

door frames, hut poles and yokes and occasionally

pestles.

Diospyros natalensis (Harv.) Brenan subsp.

natalensis

xintomantomane

The Tsonga name can be taken to mean ‘little

Diospyros’. Sticks are cut from this shrub and used

to build palisades (walls around or between a group

of huts), as well as being used as withies in hut

building. The fruit is eaten.

Dodonea viscosa Jacq. var. viscosa (Liengme 216)

mudodivisa

This is grown as a hedge in many villages.

Dombeya burgessiae Gerr. ex. Harv. (Liengme 3)

mukurhu

The bark of this small shrub is strong and is used

as cord. The plant occurs as an understorey plant in

the forest along the escarpment in the extreme west

of the study area.

Dombeya rotundifolia (Hochst.) Planch, var.

rotundifolia (Liengme 78)

mbikanyaka; nsihapukuma; xiluvarhi

The wood is used for many purposes, from

building to carving spoons for stirring food. The

flowers are believed to have magico-medicinal pro-

perties. They are sprinkled over a hen’s eggs to pre-

vent the chickens from dying once they have hatched.

Dovyalis zeyheri (Sond.) Warb. (Liengme 50)

chipachipachane;

This is another source of firewood.

Ekebergia capensis Sparrm.

nyamarhi

This is one of the trees left standing in villages to

provide shade.

Epaltes gariepina (DC.) Steetz (Liengme 220)

munywane

A medicine is made from this herbaceous plant for

relieving pain. The medicine is made by placing the

plant in hot water. The lotion is then applied to the

affected part of the body. This species is considered

to be the female form of the plant ‘munywane’.

Blumea aurita is the male form. Both are used for the

same kind of medicine.

Erythrina lysistemon Hutch,

muvale; nsisimbana

The wood is not strong and is only used occa-

sionally as fence-posts. Truncheons are planted to

provide a living fence.

Eucalyptus spp.*

ndlulamithi

Gum poles can be bought from timber producers

near the homeland and are often used instead of in-

digenous timber in the construction of roofs.

508

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

Euclea crispa (Thunb.) Guerke var. crispa (Liengme

194, 203)

xintomantomane

This species has the same Tsonga name as

Diospvros natalensis, indicating that the Tsonga con-

sider the two plants as the same or similar. The bran-

ches of this shrub are flexible and are used as withies

in hut construction. It is common in the Mopane

Veld.

Euclea divinorum Hiern (Liengme 111, 226)

nhlangula/nhlohlangula

The fruits are eaten, mostly by children.

Eaurea saligna Harv.

muthango; n’wamidzumba

This is recognized as a good timber tree and is used

for building and furniture. It occurs occasionally on

the western edge of the study area, in the Lowveld

Sour Bushveld. The wood is also used as firewood,

but it is said that a fire made with this wood needs to

be well-tended or it goes out. The nectar is sucked out

of the flowers by children.

Eaurea speciosa (Welw.) Welw. (Liengme 17)

muthango; n’wamidzumba

This species is generally found at higher altitudes

than Eaurea saligna. It is not distinguished from that

species by the Tsonga, having the same names and

uses.

Eicus burkei (Miq.) Miq. (Liengme 157)

xirhombe; xirhomberhombe; nhluhlawumbe

The bark is easily stripped off young branches and

is used as cord.

Eicus capensis Thunb. (Liengme 151)

nkuwa

This tree is found along streams in the Lowveld

Sour Bushveld regions. The ‘fruit’ is edible but is

usually infested with insects. The wood is sometimes

used as fuel.

Ficus capreaefolia Del. (Liengme 104)

phalavurha; xinkuwana

This species occurs along river banks in the Arid

Lowveld and Mopane Veld. The young branches are

very flexible and are used as withies. The ‘fruit’ is

eaten.

Eicus soldanella Warb. (Liengme 109)

nkuwa ya tintsava

This species of Eicus is found on rocky hills in the

lowveld, often actually growing on rocks. Its Tsonga

name is indicative of its habitat, meaning ‘the fig of

the hills’. The ‘fruit’ is tasty and is eaten.

Eicus sonderi Miq. (Liengme 1 17, 222)

xirhomberhombe xa tintsava

The Tsonga name for this tree is also indicative of

its habitat, meaning the ‘xirhomberhombe’ of the

hills. The ‘fruit’ is sometimes eaten.

Eicus stuhlmannii Warb.

nhlulawumbe; xirhombe; xirhomberhombe

The ‘fruit’ of this fig is sometimes eaten, but it is

apparently not as pleasant as some of the other

species.

Ficus sycamorus L. (Liengme 139, 148, 217)

nkuwa

This large tree occurs mainly along river banks,

but in the Lowveld Sour Bushveld it is also found

away from the rivers. Here it is often one of the few

trees remaining in the field and in villages. The ‘fruit’

is edible, but, as with many wild figs, is usually in-

fested with insects.

Flacourtia indica (Burm. f.) Merr.

muqokolo/nqokolo; xivambula

The fruit is eaten.

Garcinia livingstonei T. Anders,

mbhimbi/mhimbi

This tree is not very common in the study area. The

fruit is eaten and is also used to make an alcoholic

beverage. Whether this practice is common or not is

uncertain.

Gardenia spatulifolia Stapf & Hutch,

ntsalala; xitsalala

The heartwood of this tree/shrub is dark and

spoons are carved from it. It is thought that a stick

cut from this species and placed in the ground inside

a witch-doctor’s hut will protect him, presumably

from evil forces.

Gnidia rubescens B. Peterson (Liengme 228)

xinyokanyokane

A worm medicine is prepared from the roots of

this herb.

Gossypium herbaceum L. var. africanum (Watt)

J. B. Hutch. & Ghose (Liengme 80)

miseha; ricinda

The boll provides a fibre which is little used today.

It was used previously in much the same way that cot-

tonwool is used today.

Grewia species

nsihana

The name ‘nsihana’ appears to be a generic name

for the Grewia species.

Grewia flavescens Juss. var. flavescens (Liengme

244)

nciwana; nsihana

The wood of this common shrub is used for mak-

ing musical instruments and rims and basal rings for

baskets. The fruit is eaten.

Grewia flavescens Juss. var. olukundae (Schinz)

Wild (Liengme 97, 131)

nsihana yo kulu

The Tsonga name means ‘large Grewia' . The fruit

is edible, being particularly favoured by children.

Grewia occidentalis L. (Liengme 142)

ntsepukane; ntswukelane

This is a species of the Lowveld Sour Bushveld.

The leaves are cooked and eaten as a vegetable.

Heteropogon con tortus (L.) Beauv. ex Roem. &

Schult. (Liengme 83, 177)

xilungwa

This grass is used for thatching roofs by the

method of tying the grass into mats first and then at-

taching these to the roof.

C. A. LIENGME

509

Heteropyxis natalensis Harv. ( Liengme 20)

nthathasani

Spoons are carved from the wood of this small

tree. It occurs in the Lowveld Sour Bushveld.

Hexalobus monopetalus (A. Rich.) Engl. & Diels

( Liengme 72)

mbomu; nxakama

The fruit of this small tree is eaten.

Hibiscus cannabinus L.

ntsembyana/ntswembyane

The bark of this plant is very fibrous and it is used

for making twine and rope. It is a weed, often seen

along roads and in old fields.

Hvparrhenia dichroa (Steud.) Stapf ( Liengme 31,

'160, 164, 184, 192)

ntsenga; tlongwe

This is one of the several species of this genus that

are used as thatch. Hyparrhenia species are common

in the Lowveld Sour Bushveld, but not in the other

two veld types. Most of the species have the same

Tsonga names.

Hyparrhenia dregeana (Nees) Stapf ex Stent

( Liengme 30)

ntsenga

A thatch grass.

Hyparrhenia hirta (L.) Stapf (Liengme 165, 183, 191)

ntsenga; tlongwe; deke

Another thatch grass. The last of the Tsonga

names is probably derived from the word ‘dak’.

Hyparrhenia rudis Stapf ( Liengme 6)

ntsenga

Also a thatch grass.

Hyparrhenia tamba (Hochst. ex. Steud.) Anders, ex

Stapf (Liengme 130)

tlongwe

A fifth Hyparrhenia species which is used as

thatch.

Hyperthelia dissoluta (Nees) Clayton (Liengme 7,

'167, 179, 180, 185, 188)

tlongwe

This is a common grass of the Lowveld Sour Bush-

veld, which is used as thatch.

Hyphaene natalensis Kunze

vucema

This palm is rare in the study area and is protected.

The leaves were used to make certain types of bags

and baskets as well a beer-strainers. These objects are

not very common any more because of the scarcity of

material.

Ipomoea batatas (L.) Lam.*

gapi (tuber); ritiyi

The sweet potato is cultivated in many parts of the

study area.

Jatropha curcas L.* (Liengme 154)

nhlamfura ya valungu

This exotic is found in many villages. The people

know that the seed contains oil but they don’t seem

to use it. It is very poisonous and there have been a

number of cases of poisoning of children by the fruit.

Kigelia africana (Lam.) Benth. (Liengme 94)

mpfungu/mpfungurhu; muveve

This large tree occurs predominantly near rivers.

The large fruit is burnt when an unwanted storm

threatens. Dense smoke is produced and this is be-

lieved to chase the clouds away. The pulp of the fruit

is known to have been ground into flour and used as

food during famines.

Lagenaria siceraria (Molina) Standi,

rindanga

The calabash is widely cultivated in Gazankulu.

Besides being used as a vegetable, it also provides

bowls and scoops.

Landolphia kirkii Dyer (Liengme 207)

muvungwa; mungu

This scrambling plant of the rocky hills has edible

fruit.

Lannea discolor (Sond.) Engl. (Liengme 70)

ximutswani; ximombyana; xinkanyana; mumbumu-

mbu; ximpupyani

The Tsonga name ‘xinkanyana’ can be taken to

mean ‘the little marula’: the marula is called

‘nkanye’. The fruit is eaten and the bark is used as

cord.

Lannea edulis (Sond.) Engl,

nchuchungwa

The small fruit is eaten.

Lannea stuhlmannii (Engl) Engl. (Liengme 245)

ndivata/ndzivata; ximbukanyi; ximombonkanyi

The bark provides a purplish brown dye which is

used for drying the materials used in basket-making

in order to give the baskets coloured stripes or pat-

terns. The bark is also used as cord. The wood of the

roots is sometimes used in the rims of baskets. The

fruit is eaten. The Tsonga name ‘ximombonkanye’

means ‘it has the face of the marula’: ‘mombo’

means ‘face’ and ‘nkanye’ is the marula tree. This

corresponds to the English common name of the spe-

cies— bastard marula.

Lantana rugosa Thunb. ( Liengme 87)

tihove ta valungu

The fruit is eaten by children.

Leonotis sp. (Liengme 102)

mahlanganiso

The Tsonga name of this plant is derived from the

verb ‘hlanganisa’ which means ‘to unite or join’. If a

person has a broken bone, the witch-doctor pricks

the skin in the region of the break and applies a

lotion made from the tops of the plant. This is

thought to heal the bones quickly.

Lippia javanica (Burm. f.) Spreng. (Liengme 27)

musuzwane; ntungufana

Plants are cut and tied together to make rough

brooms. The plant also has a number of medicinal

uses. The leaves are boiled in water to make a cough

medicine and a bleeding nose is cured by plugging it

with leaves. The plant occurs commonly in the Low-

veld Sour Bushveld, especially in areas where the

bush has been cleared.

510

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

Lonchocarpus capassa Rolfe

mbhandzu/mbhandzwa

This tree is fairly common in the Arid Lowveld and

Mopane Veld areas, particularly along rivers. A de-

coction of the bark is used as a cure for colds.

Maerua angolensis DC. ( Liengme 174)

xiyimanamurhi

A purgative medicine is made from the bark.

Maerua parvifolia Pax ( Liengme 229)

nongonongo

The roots are used for making medicine, the pur-

pose of which is uncertain.

Manihot utilissima Pohl*

ntusumbulu

This is occasionally cultivated.

Maytenus heterophylla (Eckl. & Zeyh.) N. Robson

f Liengme 48)

xihlangwa/xilangwa

Spoons and stirrers, called ‘rifeto’, are sometimes

made from the wood. It is also used as firewood.

Melia azedaraeh L.*

xifiringoma

This exotic provides roof poles and fence-posts. It

is only common in the moister western parts of the

study area, where it has become naturalized in dis-

turbed areas.

Mimusops zeyheri Sond. ( Liengme 64)

mibubulu; mpfuxane; nhlantswa

The fruit is eaten.

Oncoba spinosa Forssk.

mbhovu; nchowana; tongwaan

The dried fruits of this tree are used in making

ankle-rattles worn at dances.

Opuntia ficus-indica (L.) Mill.*

mudoro

The fruit of this naturalized exotic is often eaten.

The plant is widespread in the Lowveld Sour Bush-

veld.

Ozoroa engleri R. & A. Fernandes

xinungumafi

The wood burns well and it is specially selected

when a hot fire is required.

Ozoroa reticulata (Bak. f.) R. & A. Fernandes subsp.

reticulata var. reticulata ( Liengme 169)

xinungu; mfute

Besides being used as firewood, this species also

has a medicinal use: the roots and bark are used to

make a purgative.

Pappea capensis Eckl. & Zeyh. ( Liengme 171)

gulaswimbi; guvaswivi; xikwakwaxu

This tree sometimes produces large crops of its

small fruit. The pulp of the fruit, which is quite sour,

is eaten.

Parinari curatellifolia Planch, ex Benth. subsp.

mobola (Oliv.) R. Grah.

mbulwa

The fruit tastes pleasant and is eaten in fairly large

quantities. It generally ripens in late winter or early

spring. Dried, the fruit keeps for several months.

Peltophorum africanum Sond.

ndzedwe; ndzhuva; nhlanhlanu

The wood is used for fence-posts and for carving

bowls. It occurs in the Lowveld Sour Bushveld.

Pennisetum americanum (L.) Leeke subsp.

americanum*

mahoba

Millet is cultivated in most parts of Gazankulu.

The grain is mostly used in the brewing of traditional

beer.

Phaseolus sp.*

nyawa

Beans are cultivated in some areas.

Phoenix reclinata Jacq.

mbovu; ncindzu; nchindu-lisundu

This palm occurs scattered along streams in the

Lowveld parts of the study area. An alcoholic bever-

age used to be made from the fruit, but it is uncertain

if this is still done. Mats are made from the rachis of

the leaf. The rachis is split in four and the pieces are

tied or threaded together with twine.

Phragmites mauritianus Kunth ( Liengme 163)

rihlanga

This reed is common along most rivers. It is used

to build palisades, in roof construction and as an

underlayer in thatching.

Phyllanthus reticulatus Poir.

swatima lowutsongo; nthethenya

The fruit is eaten.

Phyllanthus verrucosus Thunb. ( Liengme 218)

nsangasa

The fruit is eaten.

Physalis peruviana L.* (Liengme 2)

malanguti (fruit (pi.)

This exotic has become naturalized in the moister

areas of Gazankulu (Lowveld Sour Bushveld). Its

edible fruit is popular with all.

Piliostigma thonningii (Schumach.) Milne-Redh.

(Liengme 40)

nkokotso/nkolokotso; xidengana

This tree occurs occasionally in the Lowveld Sour

Bushveld. The large pods are sometimes fed to cattle.

Pittosporum viridiflorum Sims

mphatakhamelo

Spoons are carved from the soft white wood.

Plectranthus esculenthus N.E. Br.

nthada/ntheda v

The tubers of this herbaceous plant are eaten.

Pluchea dioscorides (L.) DC.

bvimba

This is another plant whose name indicates its use.

The leave were used as a cork or stopper to close

bottles, pots or jar-s. This practice is seldom heard of

C. A. LIENGME

511

today. The Tsonga name is derived from the verb

‘-bvimba’, meaning ‘seal with a lid’.

Pseudolachnostylis maprouneaefolia Pax (Liengme

140)

nxojowa

The wood is used for carving spoons.

Ptaeroxylon obliquum (Thunb.) Radik,

ndazi; ndzari

Various utensils and other items are carved from

the wood. The tree occurs occasionally, usually near

rivers.

Pterocarpus angolensis DC. (Liengme 110)

mvhangazi; murotso; ntsonde

This tree, commonly known as kiaat, has beauti-

ful, hard but workable wood from which furniture,

bowls, plates, spoons and carved objects of art are

made. It is a favoured wood for making items for

sale as curios. The tree is protected in the homeland

and the wood is not easily obtainable. The bark is

used in veterinary medicine. It is believed to cure

limping if it is placed on the injured part of the beast.

Pterocarpus rotundifolius (Sond.) Druce subsp.

rotundifolius (Liengme 179)

mpyalelangala; muhadamba; muyataha; ndleve ya

ndlopfu; nxelana/nxelela

Hoe handles are made from wood, which is strong

but not very durable. The tree occurs in the Arid

Lowveld.

Pterolobium Stella turn (Forssk.) Gmel.

rikatsi/rikatsu

This is sometimes planted as a living fence, pro-

viding an impenetrable barrier.

Rhoieissus digitata (L.f.) Gilg & Brandt (Liengme

234)

mbhesa/mphesa; mbhezana

The roots of this climber are used in the prepara-

tion of a stomach medicine.

Ricinus communis L.*

nhlampfurha

Oil is extracted from the seeds and used for a

number of purposes. It is rubbed on the skin and is

also used as ear-drops. The plant is a weedy exotic,

common in the Lowveld Sour Bushveld. The Tsonga

name is derived from the word ‘mafurha’ which

means ‘fat, oil or grease’.

Saccharum officinale L.*

mova

Sugar-cane is cultivated in gardens in some parts of

the study area.

Sansevieria hyacinthoides (L.) Druce (Liengme 187)

xikwenga

A fibre is obtained from the leaf, which was once

used as a thread for sewing and for making string.

Agave fibre has replaced this in most areas.

Sarcostemma viminale (L.) R. Br. (Liengme 119)

neta

This is a fairly common plant in the Arid Lowveld

and the Mopane Veld, generally growing over trees.

The milky sap is added to food given to cows in the

belief that it will make them produce more milk.

Scirpus inclinatus (Del.) Aschers. & Schweinf.

(Liengme 8, 198)

mixaka; vungwane

This plant occurs along rivers and streams in the

Lowveld Sour Bushveld and is used to make mats of

inferior quality to those made from Cyperus

latifolius.

Sclerocarya caffra Sond.

nkanye

This widely distributed, common tree has many

uses. The wood is not very hard when fresh and is

fairly easily workable, being made into spoons, mor-

tars, pestles, bowls and plates, drums and cattle

yokes. The wood hardens as it dries and is quite

durable. It seems that only male trees are cut. Female

trees are spared because of the edible fruit they bear.

The fruit falls off the tree and ripens on the ground

during February/March. It is collected and taken

home, eaten or used to make a beer called ‘bukanye’.

The pips are removed from the fruit by boiling it in

water. The skin comes off and is discarded and the

flesh can be separated from the pip. The pips are

stored for use later and the flesh is usually eaten or

allowed to ferment to make beer. The pips are stored

until they dry, when they are cracked and the kernels

removed. These are very tasty when roasted and are

eaten by themselves or added to relishes. The marula

tree is probably the most important wild fruit tree of

the Tsonga. The bark and roots are used to make a

cleansing medicine for the stomach.

Securinega virosa (Rosb. ex Willd.) Rax & K. Hoffm.

nhlangawume/nxangawume; nsangasa

The fruit is eaten by children. The flexible young

branches are used as withies in building.

Setaria sp. (Liengme 168)

xihovane; xichakala; pundze

This is one of the grasses used for thatching in the

traditional way.

Solanum incanum L. (Liengme 176)

ndzhulwane; nthuma yo kulu

This ruderal is often found around dwellings. The

fruit is crushed and placed on the umbilical wound of

a new-born baby until it heals.

Solanum panduraeforme E. Mey. (Liengme 173)

ndulwane/ndzhulwane; nthuma; rirhulwane

This ruderal has the same use as Solanum in-

canum.

Spilanthes mauritiana (Pers.) DC. (Liengme 147)

xixwene; xixwenelamhofu

The leaves are rubbed on mouth-ulcers in order to

ease the pain.

Spirostachys africana Sond.

ndzopfuri; xilangamahlo

Bowls, ornaments and stools are carved from the

wood. The tree is not very common in the study area.

Sporobolus africanus (Poir.) Robyns & Tournay

(Liengme 159)

muxikijane

This grass is sometimes used as thatch, but it is

more often used to lash the thatch down onto the

roof frame-work. It is also used to make items such

as table-mats, sold as curios.

512

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

Strychrtos madagascariensis Poir. (Liengme 113)

nkwakwa

This tree occurs occasionally in the area and is

usually left standing in fields. The pulp of the fruit is

edible. It can be eaten raw but is normally stamped

into a flour and kneaded into cakes or made into a

food called ‘mpfuma’. The young branches are used

as withies in building.

Strychnos spinosa Lam. (Liengme 15)

nhlala/nsala

The pulp of the fruit is eaten. Fresh pulp is

sometimes added to milk to make it sour. The pulp is

also added to mealies (Zea mays) that have been

stamped and cooked.

Syzygium cordatum Hochst. (Liengme 49)

muhlwa/muthwa

This tree occurs along streams and rivers in the

Lowveld Sour Bushveld. The fruit is eaten.

Tabernaemontana elegans Stapf (Liengme 206)

nkahla/nkahlwane

The wood is soft and is used for carving spoons.

The pulp of the fruit is edible, but it is not certain to

what extent it is eaten. The tree occurs in the north-

western parts of the study area.

Tecomaria eapensis (Thunb.) Spach.

khujana

The young branches of this shrub are used in the

manufacture of baskets. The rims of baskets made

from Acacia ataxacantha wood strips consist of two

rings of wood bound to the body of the basket. The

inner of these two rings is made from the wood of

this shrub.

Terminalia sericea Burch, ex DC. (Liengme 32)

mugosi; nkonolo/nkonola/nkohono

The wood of this tree is used for making door-

frames, pestles, spoons and a number of other ar-

ticles. The root is used to make an emetic.

Themeda triandra Forssk.

kondze; mbvume; xivopfu

This is one of the grasses used for thatching in the

traditional way.

TrichiUa emetica Vahl (Liengme 101, 209)

nkuhlu

Oil is extracted from the seeds and used for

medicinal purposes. The pulp of the fruit is said to be

edible. The wood is one of those from which a certain

type of musical bow is made.

Turraea obtusifolia Hochst. (Liengme 125)

mbhovane

Herbalists use the crushed leaves and fruit of thi

shrub to make an enema.

Tylosema fassoglensis (Schweinf.) Torre & Hillc.

(Liengme 47)

nthamula

The roots of this creeper yield a brown dye when

crushed in water. The twigs are used in basket-

making and the seeds are roasted and eaten.

Vangueria infausta Burch. (Liengme 99)

mpfilwa; ntswila

This small tree occurs occasionally in the Lowveld

Sour Bushveld. The fruit is eaten and the pulp of the

fruit is sometimes added to milk to make it sour.

Xeromphis obovata (Hochst.) Keay

mutungababara

The fruit of this Arid Lowveld shrub is used to

make an emetic. It is believed that the crushed roots,

if given to a patient, drive out demons.

Xerophyta retinervis Bak.

sirudzu

This plant occurs on rocky hills and ridges in the

lowveld areas. The plants are tied together to make

brooms and are also used as fuel for firing pots.

Ximenia americana L. var. microphylla Welw.

(Liengme 231, 225)

musomuwu; ntsengele; ntsengele yantsongo

The fruits are eaten. The plant occurs on hills in

the Arid Lowveld and Mopane Veld.

Ximenia caffra Sond. var. caffra (Liengme 1 14, 230)

This species also has fruit which are eaten.

Zanthoxylum capense (Thunb.) Harv. (Liengme 58)

khinungumorupa; manhungwane; nugani; xirho-

mbehleta

The roots are cooked in the water to make a medi-

cine for colds. The leaves are crushed and put in the

nose as a decongestant.

Zea mays L.*

mavele

This is the most important crop cultivated by the

Tsonga. It forms the basis of their diet in the form of

meal and on the cob. The crop is cultivated all over

Gazankulu, even in dry areas. The Tsonga name

given here is actually the word used to describe any

grain crop. There are several other names given to

parts of the plant or referring to different stages of its

growth. For example, ‘nan’wanyi’ is a very young

plant; ‘mphovo’ is an immature ear of maize.

Ziziphus mucronata Willd. subsp. mucronata

ncecenyi; mphasamhala

This is a widespread tree, quite common in the

Arid Lowveld. The fruit is eaten, mostly by children.

DISCUSSION

Food plants

Of the wild plants listed as being sources of edible

fruit or leaves it would seem that the most important

are:

Selerocarya caffra (fruit)

Strychnos species (fruit)

Diospyros mespiliformis (fruit)

Parinari euratellifolia subsp. mobola (fruit)

Bidens pilosa (leaves)

Corehorus species (leaves), and

Amaranthus thunbergii (leaves).

These are widely and commonly used. The last

three are used in relishes/sauces which are eaten with

maize meal (Zea mays), the staple food.

The fruit of Selerocarya is highly prized, being

used to make a beer called ‘bukanye’. Diospryos,

c. A. LIENGME

513

Strychnos and Parinari fruits are important as they

ripen in late winter and early spring, when cultivated

foods are scarce. Strychnos madagascariensis fruit is

especially important in times of drought, as the dried

pulp can keep for months.

Most of the other edible fruits of the study area are

small and mostly eaten by children, sometimes by

adults.

The Tsonga diet consists mainly of cultivated food

plants. Zea mays is the staple food, being widely

cultivated. Pennisetum americanum subsp. ameri-

canum (millet) is also cultivated, mainly to provide

malt for brewing beer.

Common vegetable crops include:

sweet potato — Ipomoea batatas

pumpkin — Cucurbita pepo

squash — Cucurbita maxima

beans — Phaseolus sp.

peanuts — Arachis hypogea

calabash — Lagenaria siceraria

The pigeon pea, Cajanus cajan, cassava, Manihot

utillisima and tomatoes, onions, spinach and leeks

are less frequently cultivated. Sugar-cane, Saccharum

officinale, is also grown in some areas.

The Tsonga generally cook once a day and the

main meal consists of mealie meal ( Zea mays meal),

‘vuswa’, and savoury, ‘xixevo’. The latter can be

vegetables or sauce (consisting of wild leaves,

spinach, peanuts, marula nuts, beans or any other

cultivated vegetable) or meat. Flying-ants, grass-

hoppers and mopane worms are also eaten.

Tsonga alcoholic beverages are made from marula

fruit, sugar-cane, maize meal, sorghum and millet,

some being more potent than others.

Building ( including thatching)

The traditional Tsonga hut of recent times consists

of wood and mud walls and a conical roof consisting

of a wooden framework covered with thatch. The

roof is often supported by poles outside the perimeter

of the walls. A hut requires a fair amount of both

wood, in the form of poles, sticks and wattles, and

grass. Fences and stockades are often made entirely

out of wood and large quantities of wood are used in

these structures. In the remoter areas all building

timber is obtained from the indigenous vegetation,

whilst in less remote areas with access to plantations,

Eucalyptus is often used.

The traditional method of thatching roofs is to tie

grass into mats called ‘makenye’. Grass species such

as Themeda triandra, Heteropogon con tort us or

Setaria spp. are used. The mats are rolled up and

stacked until the roof is ready to be thatched. Then

the mats are laid out on the roof, starting at the lower

edge of the roof, and secured.

This thatching is of inferior quality to that produc-

ed by the method of reverse-thatching, a less com-

mon method of thatching in Gazankulu. The reason

for this latter type of thatching not being common is

that it requires grass species like Hyperthelia disso-

luta and Hyparrhenia spp. and these hardly occur in

the Arid Lowveld and Mopane Veld vegetation types

which cover the greater part of Gazankulu. They do

occur in the Lowveld Sour Bushveld and reverse-

thatched roofs are somewhat commoner here. Good

thatch-grass can be bought from other areas by

whose who can afford it.

Reverse-thatched roofs are far more permanent

than the others, lasting up to 30 years and longer.

The traditional thatched roofs need replacing every

few years.

Much grass is required to thatch a single roof and

the women may walk many kilometres to collect suf-

ficient grass. The women collect and clean the thatch,

but it is the men who do the thatching.

Medicinal plants

Much of the information collected on medicinal

uses of plants was obtained from a herbalist, Mr

Mondlane. Some of his cures are commonly known

amongst the Blacks; for example, a stomach medi-

cine made from the roots of Capparis tomentosa.

Many of his medicines appear not to have been pre-

viously recorded. Eye-drops made from Aptosium

lineare and a cure for broken bones made from a

Leonotis sp. are two examples.

This herbalist had a wide knowledge of the medici-

nal uses of plants and knew as well that some plants

only grew in certain parts of Gazankulu. To collect

all the plants he needed he travelled quite long dis-

tances. Besides the specialist knowledge of herbalists,

the people also have a knowledge of medicinal

plants. For example, Clematis brachiata is commonly

used to cure colds and headaches. Often the people

know that a plant is used by the ‘doctors’ for curing a

condition, but do not know how it is used because

that is the ‘doctor’s’ secret.

Utensils and basketry

Although some of the traditional utensils have

been replaced by modern goods available at any

trading store, many are still in very wide use. Mortars

and pestles as well as winnowing baskets are to be

seen everywhere, as are wooden spoons and stirrers

and conical baskets. Calabashes still have their tradi-

tional use in beer drinking. Utensils which are bought

from a store are usually tin/enamel mugs, basins and

plates, buckets and 3-legged cast-iron pots. Basket-

making is one craft which has not died out complete-

ly, having been stimulated in some parts by demand

from tourists. Baskets are of various types and are

made from several materials. Conical baskets are us-

ed for storing mealie meal, beans and other food.

These are made from Cocculus hirsutus stems or

plaited grass culms, wound around in circles one

above the other, each bound to the one below by

Cocculus stems split in half. Twilled baskets are

made from thin strips of Acacia ataxacantha wood or

Hyphaene natalensis fronds. The latter plant is,

however, scarce and baskets made from it are now

rare. Twilled baskets include the shallow, saucer-like

winnowing baskets, spherical baskets with or without

lids and ‘wallets’. Beer strainers are also twilled. The

winnowing baskets and the sphercial baskets are

common. The body of these baskets is made from

Acacia ataxacantha wood strips. The rim is made

from wood of Brachylaena discolor, Tecomaria

capensis, Grewia spp. or Lannea stuhlmannii, and is

bound to the body with stems of Dioscorea cotini-

folia. A third type of basket is made predominantly

from a plant called ‘staf’ in Tsonga; a climber iden-

tified as Secamone alpinii.

Tsonga botany

Junod (1962), in a brief discussion on Tsonga

botany, noted the following:

(1) that the notion of ‘genus’ is present in Tsonga

taxonomy;

(2) that species are distinguished by mentioning

their habitat or sex; and

514

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

(3) diminutives are used to distinguish species.

All three of these aspects were also noted by the re-

searcher. For example, Faurea saligna and F.

speciosa, easily distinguishable by their different

leaves, have the same Tsonga name, indicating that

they are in the same ‘genus’ in Tsonga taxonomy.

There are several Grewia spp. which fall under the

name ‘nsihana’. Some of the species have other

names too, but ‘nsihana’ seems to be a generic name.

Tsonga genera do not necessarily correspond with

ours. With regard to the second of the points made

by Junod, the fig, Ficus sycamorus, is called ‘nkuwa’

and is distinguished from F. soldanella, which is call-

ed ‘nkuwa ya ntsava’, (‘ya ntsava’ meaning ‘of the

hill’) by the addition of the habitat to the name.

Epaltes gariepina and Blumea aurita have the same

name ‘munywane’ but are distinguished as being

female and male forms respectively. Diminutives

were noted as well. For example, Berchemia discolor

is ‘nyiri’ and B. zeyheri is ‘nyiyani’, ‘the little ‘nyiri’.

Diminutives are also indicated by the suffix

‘-ntsongo’ to the plant name, as in the case of

Phyllanthus reticulatus which is called ‘swatima

lowutsonga’, ‘the little ‘swatima’. Bridelia mollis is

‘swatima’.

A much more complete list of Tsonga plant names

and their uses is necessary in order to make a more

detailed analysis of Tsonga taxonomy.

UITTREKSEL

’n Studie is in ’n deel van die Tsonga-tuisland,

Gazankulu, onderneem om plante wat deur hierdie

mense gebruik word te identifiseer. ’n Lys van

Tsonga-plantname is met behulp van ’n Tsonga-

Engelse woordeboek opgestel as basis vir die studie.

Die gebruike van bykans 200 plante vir, onder

andere, medisinale doeleindes, voedsel, boumateri-

aal, vuurmaakhout, huishoudelike gereedskap, im-

plemente, implements tele en speelgoed is aangeteken.

Hierdie inligting word in die vorm van ’n verklarende

lys van die plante aangebied. Dit word gevolg deur ’n

bespreking van sommige van die meer belangrike ge-

bruike. Sommige aspekte van Tsonga taksonomie

word kortliks bespreek en met voorbeelde toegelig.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. Mem. bot.

Surv. S. Afr. No. 40.

Cuenod, R., 1976. Tsonga-English Dictionary. Johannesburg:

Sasavona.

Junod. H. A., 1962. The life of a South African tribe. New York:

University Books Inc.

Palmer, E. & Pitman, N., 1972. Trees of southern Africa. Cape

Town: A. A. Balkema.

Shaw, E. M., 1974. Material culture. In W. D. Hammond-Tooke,

Bantu-speaking peoples of southern Africa. Ed. 2, pp. 85-

126. London: Routledge & Keagen Paul.

Van Warmelo, N. J., 1974. The classification of cultural groups.

In W. D. Hammond-Tooke, Bantu-speaking peoples of

southern Africa. Ed. 2, pp. 58-84. London: Routledge &

Keagen Paul.

Van Wyk, P., 1972. Trees of the Kruger National Park. Cape

Town: Purnell.

Watt, J. M., & Breyer Brandwijk, M., 1962. Medicinal and

poisonous plants of southern and eastern Africa. Ed. 2. Lon-

don: E. &. S. Livingstone.

APPENDIX A

INDEX OF PLANT USES

Baskets: Acacia ataxacantha, Brachylaena discolor subsp. trans-

vaalensis, Cocculus hirsutus, Dioscorea cotinifolia, Grewia fla-

vescensvar. flavescens, Hyphaene natalensis, Lannea stuhlmannii,

Tecomaria capensis, Tylosema fassoglensis.

Beads : Bauhinia galpinii.

Beverage (Alcoholic): Garcinia livingstonei, Pennisetum america-

num subsp. americanum, Phoenix reclinata, Sclerocarya caffra.

Beverage (Non-Alcoholic): Athrixia phylicoides.

Buildings (Timber): Acacia caffra, A. nigrescens, Adina micro-

cephala var. galpinii, Afzelia quanzensis, Colophospermum

mopane, Combretum apiculatum subsp. apiculatum, C. hereroen-

se subsp. hereroense var. hereroense, C. imberbe, Cussonia

spicata, Diospyros mespiliformis, Dombeya rotundifolia var.

rotundifolia, Eucalyptus spp., Faurea saligna, F. speciosa, Melia

azedarach, Terminalia sericea.

Building (Withies): Antidesma venosum, Brachylaena discolor

subsp. transvaalensis, Catha edulis, Combretum hereroense subsp.

hereroense var. hereroense, Cordia ovalis, Diospyros natalensis

subsp. natalensis, Euclea crispa var. crispa, Ficus capreaefolia,

Securinega virosa, Strychnos madagascariensis.

Cordage: Acacia karroo, Colophospermum mopane, Dombeya

burgessiae. Ficus burkei, Hibiscus cannabinus, Lannea discolor,

L. stuhlmannii.

Dye: Lannea stuhlmannii, Tylosema fassoglensis.

Fencing: (including palisades) Acacia caffra, Bambusa sp., Colo-

phospermum mopane, Combretum apiculatum subsp. apicula-

tum, Diospyros natalensis subsp. natalensis, Erythrina lysistemon,

Peltophorum africanum, Phragmites mauritianus.

Fibre: Agave spp., Artabotrys brachypetalus, Gossypium her-

baceum var. africanum, Sansevieria hyacinthoides.

Fodder: Piliostigma thonningii.

Food: Adansonia digitata, Amaranthus thunbergii, Annona

senegalensis, Antidesma venosum, Arachis hypogea, Artabotrys

brachypetalus, Bequaertiodendron magalismontanum, Berchemia

discolor, Bidens pilosa, Bridelia micrantha, B. mollis, Cajanus ca-

jan, Capsicum frutescens, Carissa edulis, Cassia occidentalis,

Cephalanthus natalensis, Colocasia antiquorum, Combretum

paniculatum subsp. microphyllum, Corchorus confusus, C.

tridens, Cordia ovalis, Cucumis melo, Cucurbita maxima, C.

pepo, Diospyros mespiliformis, D. natalensis subsp. natalensis,

Euclea divinorum, Ficus capensis, F. capreaefolia, F. soldanella,

F. sonderi, F. stuhlmannii, F. sycamorus, Flacourtia indica, Gar-

cinia livingstonei, Grewia flavescens var. flavescens, G. flavescens

var. olukondae, G. occidentalis, Hexalobus monopetalus,

Ipomoea batatas, Lagenaria siceraria, Landolphia kirkii, Lannea

discolor, L. edulis, L. stuhlmannii, Lantana rugosa, Manihot

utilissima, Mimusops zeyheri, Opuntia ficus-indica, Pappea

capensis, Parinari curatellifolia subsp. mobola, Pennisetum

americanum subsp. americanum, Phaseolus spp., Phyllanthus

reticulatus, P. verrucosus, Physalis peruviana, Plectranthus

esculenthus, Saccarum officinale, Sclerocarya caffra, Securinega

virosa, Strychnos madagascariensis, S. spinosa, Syzygium cor-

datum, Tabernaemontana elegans, Trichilia emetica, Tylosema

fassoglensis, Vangueria infausta, Ximenia americana var.

microphylla, X. caffra var. caffra, Zea mays, Zizyphus mucronata

subsp. mucronata.

Food (Famine): Kigelia africana.

Fuel: Acacia ataxacantha, A. caffra, A. davyi, A. gerrardii var.

gerrardii, A. karroo, A. sieberana var. woodii, A. tortilis subsp.

heteracantha, Albizia harveyi, Allophyllus decipiens, Brachylaena

discolor subsp. transvaalensis, Colophospermum mopane, Com-

bretum erythrophyllum, Dichrostachys cinerea subsp. africana,

Dovyalis zeyheri, Faurea saligna, F. speciosa, Ficus capensis,

Maytenus heterophylla, Ozoroa engleri, O. reticulata subsp. reti-

culata var. reticulata, Xerophyta retinervis.

Furniture: Bolusanthus speciosus, Burkea africana, Faurea

saligna, F. speciosa.

Games and Toys: Aloe davyana, Annona senegalensis.

Implements and utensils: Acacia nigrescens, Adina microcephala

var. galpinii, Albizia versicolor, Berchemia discolor, Cassine

transvaalensis, Colophospermum mopane, Combretum apicula-

C. A. LIENGME

515

turn subsp. apiculatum, C. imberbe. Commiphora pyracan- guvazwivi

thoides, Diospyros mespiliformis, Dombeya rotundifolia var. guxe

rotundifolia. Heteropyxis natalensis, Lagenaria siceraria, Lippia guxe

javanica, Maytenus heterophylla, Peltophorum africanum, Pittos-

porum viridiflorum, Pseudolachnostylis maprouneaefolia, Ptaer- K

oxylon obliquum, Pterocarpus angolensis, P. rotundifolius subsp. khalavatla

rotundifolius, Sclerocarya caffra, Spirostachys africanus, Taber- khawa

naemontana elegans, Terminalia sericea, Xerophyta retinervis. khinungumorupa

Implement handles: Acacia ataxacantha, Berchemia discolor, Bo-

lusanthus speciosus, Cassine transvaalensis, Combretum apicula-

tum subsp. apiculatum, C. hereroense subsp. hereroense var.

hereroense, Commiphora pyracanthoides.

Living fences and hedges: Dodonea viscosa var. viscosa, Erythrina

lysistemon, Pterolobium stellatum.

khujana

kofi yanhova

kondze

kotoni

kumbekumbe

Magic: Asparagus virgatus, Gardenia spatulifolia, Kigelia afri-

cana, Sarcostemma viminale.

L

lumanyama

Mats: Cyperus latifolius, C. sexangularis, Phoenix reclinata, Scir-

pus inclinatus, Sporobolus africanus.

Medicine: Albizia versicolor, Aptosimum lineare, Artabotrys bra-

chypetalus, Asclepias burchellii. Balanites maughamii, Blumea

aurita, B. gariepina, Bridelia micrantha, Capparis fassicularis var.

fassicularis, C. tomentosa, Clematis brachiata, Croton megalobot-

rys, Cryptolepis capensis, Cyphostemma humile subsp. humile,

Epaltes gariepina, Gnidia rubescens, Leonotis sp., Lippia javan-

ica, Lonchocarpus capassa, Maerua angolensis, M. parvifolia,

Rhoicissus digitata, Ricinus communis, Sclerocarya caffra, Sola-

num incanum, S. panduraeforme, Spilanthes mauritiana, Termi-

nalia sericea, Trichilia emetica, Turraea obtusifolia, Xeromphis

obovata, Zanthoxylum capense.

Musical instruments (including drums): Balanites maughamii,

Calodendron capense, Commiphora pyracanthoides, Cordia

grandicalyx, Grewia flavescens var. flavescens, Oncoba spinosa,

Sclerocarya caffra, Trichilia emetica.

Oil: Jatropha curcas, Ricinus communis, Trichilia emetica.

Plug: Pluchea dioscorides.

Shade trees: Adina microcephala var. galpinii, Albizia harveyi, A.

versicolor. Cassia abbreviata subsp. beareana, Ekebergia capensis.

Snuff: Amaranthus cruentus.

Soap (including shampoo): Dicerocaryum zanguebarium subsp.

zanguebarium.

Thatch: Cymbopogon validus, Heteropogon contortus, Hypar-

rhenia dichroa, H. dregeana, H. hirta, H. rudis, H. tamba, Hyper-

thelia dissoluta, Phragmites mauritianus, Setaria sp., Sporobolus

africanus, Themeda triandra.

Veterinary medicine: Pterocarpus angolensis.

Woodcarving: Albizia versicolor, Berchemia discolor, Bolusan-

thus speciosus, Cassine aethiopica, Combretum hereroense subsp.

hereroense var. hereroense, Crossopteryx febrifugia, Dalbergia

melanoxylon, Gardenia spatulifolia, Ptaeroxylon obliquum,

Pterocarpus angolensis, Spirostachys africanus.

APPENDIX B

INDEX OF TSONGA PLANT NAMES

Tsonga name Botanical name

B

bvimba Pleuchea dioscorides DC.

C

chipachipachane

chochelamandleni

yantsongo

Dovyalis zeyheri (Sond.) Warb.

Combretum panicalutum Vent,

subsp. microphyllum (Klotzsch)

Wickens

D

deke

dinda/dindza

Cymbopogon validus (Stapf)

Stapf ex Burtt Davy

Dicerocaryum zanguebarium

(Lour.) Merr.

G

gapi

gawana

gulaswimbi

Ipomoea batatas (L.) Lam.

Cucurbit a maxima Duch.

Pappea capensis Eckl. & Zeyh.

M (Ma-)

maamba

mahlanganiso

mahoba

malanguti

manga

manhungwane

mavele

(Mb — Mo-)

mbhandzu/mbhandzwa

mbhesa/mphesa

mbhesu/mbheswi

mbhesana

mbhimbi/mhimbi

mbhovane

mbhovhu

mbhovu

mbikanyaka

mbimba

mbomu

mbovu

mbulwa

mbvinyaxihloka

mbvume

mbvuvu/mvuvu

mfute

mgula

mgupa

mhangani

mibululu

mikoka

mindzere/mundzere/ndzerhe

miseha

mixaka

mola/molani

molela

mondzo

mova

mowu/muwu

(Mp — Mt-)

mpfalambati

mpfilwa

mpfimbahongoni

mpfungu/mpfungurhu

mpfunta

mpfunte-mpfunte

Pappea capensis Eckl. & Zeyh.

Corchorus confusus Wild

Corchorus tridens L.

Cucurbita maxima Duch.

Capparis tomentosa Lam.

Zanthoxylum capense (Thunb.)

Harv.

Tecomaria capensis (Thunb.)

Spach

Athrixia phylicoides DC.

Themeda triandra Forssk.

Asclepias burchellii Schltr.

Bridelia mollis Hutch.

Cassia abbreviata Oliv. subsp.

beareana (Holmes) Brenan

Clematis brachiata Thunb.

Leonotis sp.

Pennisetum americanum (L.)

Leeke subsp. americanum

Phy sal is peruviana L.

Arachis hypogea L.

Zanthoxylum capense (Thunb.)

Harv.

Zea mays L.

Lonchocarpus capassa Rolfe

Rhoicissus digitata Gilg & Brandt

Albizia versicolor Welw. ex Oliv.

Rhoicissus digitata Gilg & Brandt

Garcinia livingstonei T. Anders.

Turraea obtusifolia Hochst.

Calodendrum capense (L.f.)

Thunb.

Oncoba spinosa Forssk.

Dombeya rotundifolia (Hochst.)

Planch, var. rotundifolia

Combretum imberbe Wawra

Hexalobus monopetalus (A.

Rich.) Engl. & Diels

Phoenix reclinata Jacq.

Parinari curatellifolia Planch, ex

Benth. subsp. mobola (Oliv.)

R. A. Grah.

Acacia caffra (Thunb.) Willd.

Themeda triandra Forssk.

Combretum erythrophyllum

(Burch.) Sond.

Ozoroa reticulata (Bak. f.) R. &

A. Fernandes subsp. reticulata

var. reticulata

Diospyros mespiliformis Hochst.

ex A. DC.

Combretum erythrophyllum

(Burch.) Sond.

Aloe davyana Schoenl. var.

davyana

Mimusops zeyheri Sond.

Clematis brachiata Thunb.

Bridelia micrantha (Hochst.)

Baill.

Gossypium herbaceum L. var.

africanum (Watt) J. B. Hutch.

& Ghose

Scirpus inclinatus (Del.) Aschers.

& Schweinf. ex Boiss.

Albizia harveyi Fourn.

Combretum imberbe Wawra

Saccharum officinale L.

Adansonia digitata L.

Antidesma venosum E. Mey. ex

Tul.

Vangueria infausta Burch.

Bolusanthus speciosus (H. Bol.)

Harms

Kigelia africana (Lam.) Benth.

Combretum paniculatum Vent,

subsp. microphyllum (Klotzsch)

Wickens

Combretum paniculatum Vent,

subsp. microphyllum (Klotzsch)

Wickens

516

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

mpfuxane

mphasamhala

mphata

mphatakhamelo

mphatakhwari

mphovo

mpon’wana

mpotsa/mpoza

mpulu

mpyalelangala

mtlele

(Mu — Mv-)

mucece

mudodivisa

mudoro

muganukomu

mugarasaka

mugejo

mugosi

muhadamba

muhlome/nhlume

muhlwa

muhobadala

mukurhu

mukorongwe

muluwa

mumbumumbu

munga

mungu

munywane

muono

muqokolo

murhompfa

murorongwe

murhotso

musengele

musenje

musuzwane

muthango

muthondwa/ntondo

muthuma

muthwa

mutungababara

muvale

muvambangoma

muveve

muvungwa

rauwe

muxiji

muxikijane

muyataha

muyembe

muzuzugwane xihlahla

mvhangazi

mvhangazi wobasa

N (Na— Nd-)

riala

ncecenyi

ncenga

nchindu-lisundu

nchowana

Mimusops zevheri Sond.

Zizyphus mucronata Willd.

subsp. mucronata

Brachylaena discolor DC. subsp.

transvaalensis ( Ph ill . &

Schweick.) J. Paiva

Pittosporum viridiflorum Sim

Antidesma venosum E. Mey. ex

Tul.

Zea mays L. (immature plant)

Cordia ovalis R. Br. ex DC.

Combretum apiculatum Sond.

subsp. apiculatum

Combretum hereroense Schinz

subsp hereroense var. hereroen-

se

Burkea africana Hook.

Pterocarpus rotundifolius (Sond.)

Druce subsp. rotundifolius

Cordia ovalis R. Br. ex DC.

Albizia versicolour Welw. ex

Oliv.

Dodonaea viscosa Jacq. var.

viscosa

Opuntia ficus-indica (L.) Mill.

Lannea stuhlmannii (Engl.) Engl.

Combretum apiculatum Sond.

subsp. apiculatum

Cymbopogon validus (Stapf)

Stapf ex Burtt Davy

Terminalia sericea Burch, ex DC.

Pterocarpus rotundifolius (Sond.)

Druce subsp. rotundifolius

Adina microcepliala (Del.) Hiern

var. galpinii (Oliv.) Hiern

Syzygium cordatum Hochst.

Capparis fassicularis DC. var.

fassicularis

Dombeya burgessiae Gerr. ex

Harv.

Capparis tomentosa Lam.

Acacia ataxacantha DC.

Lannea discolor Engl.

Acacia karroo Hayne

Landolphia kirkii T.-Dyer

Blumea aurita (L.f.) DC.

Epaltes gariepina (DC.) Steetz

Strychnos cocculoides Bak.

Flacourtia indica (Burm. f.)

Merr.

Annona senegalensis Pers.

Garcinia livingstonei T. Anders.

Pterocarpus angolensis DC.

Bambusa sp.

Cussonia spicata Thunb.

Lippia javanica L.

Faurea saligna Harv.

Faurea speciosa (Welw.) Welw.

Cephalanthus natalensis Oliv.

Adina microcephala (Del.) Hiern

var. galpinii (Oliv.) Hiern

Syzygium cordatum Hochst.

Xeromphis obovata (Hochst.)

Keay

Erythrina lysistemon Hutch.

Albizia versicolor Welw. ex Oliv.

Kigelia africana (Lam.) Benth.

Landolphia kirkii T.-Dyer

Berchemia discolor (Klotzsch)

Hemsl.

Bidens pilosa L.

Sporobolus africanus (Poir.)

Robyns & Tournay

Pterocarpus rotundifolius (Sond.)

Druce subsp. rotundifolius

Annona senegalensis Pers.

Allophvlus decipiens Radik.

Pterocarpus angolensis DC.

Albizia versicolor Welw. ex Oliv.

Hyphaene natalensis Kunze

Zizyphus mucronata Willd.

subsp. mucronata

Dichrostachys cinerea (L.) Wight

& Arn. subsp. africana Brenan

& Brummitt

Phoenix reclinata Jacq.

Oncoba spinosa Forssk.

nchuchungwa

nchugulu

nchuguru/ntshuguru

ncindzu

ndazi

ndhenga/ndzhenga

ndivata/ndzivata

ndlejane

ndleve yandlopfu

ndlulamithi

ndodzi

ndulwane/ndzhulwane

ndzari

ndzedze

ndzhopfa/ndzompfa/ndzopfa

ndzhulwane

ndzhuva

ndzololwane/nsululwani

ndzopfori

ndzundzuluka/nhundzuluka

Lannea edulis (Sond.) Engl.

Carissa edulis Vahl

Phoenix-reclinata Jacq.

Grewia flavescens Juss. var.

flavescens

Ptaeroxylon obliquum (Thunb.)

Radik.

Dichrostachys cinerea (L.) Wight

& Arn. subsp. africana Brenan

& Brummitt

Lannea stuhlmannii (Engl.) Engl.

Cyphostemma humile (N.E. Br.)

Desc. ex Wild & Drumm.

subsp. humile

Pterocarpus rotundifolius (Sond.)

Druce subsp. rotundifolius

Eucalyptus spp.

Cajanus cajan (L.) Millsp.

Solanum panduraeforme E. Mey.

Ptaeroxylon obliquum (Thunb.)

Radik.

Peltophorum africanum Sond.

Annona senegalensis Pers.

Solanum incanum L.

Peltophorum africanum Sond.

Albizia harveyi Fourn.

Spirostachys africanus Sond.

Ximenia caffra Sond. var. caffra

(Ne — Nj-)

nembenembe

neta

ngoka

nhlaba yafole

nhlahle

nhlala

nhlampfurha

nhlampfurha yavalungu

nhlangawume/nxangawume

nhlanhlanu

nhlangula/nhlohlangule

nhlantswa

nhlata

nhlulawumbe

nhlwehlwe

njekejeke

Cassia occidentalis L.

Sarcostemma viminale (L.) R. Br.

Acacia tortilis (Forssk.) Hayne

subsp. heteracantha (Burch.)

Brenan

Amaranthus cruentus L.

Cyperus sexangularis Nees

Strychnos cocculoides Bak.

Strychnos spinosa Lam.

Ricinus communis L.

Jatropha curcas L.

Securinega virosa (Roxb. ex

Willd.) Pax & K. Hoffm.

Peltophorum africanum Sond.

Euclea divinorum Hiern

Mimusops zevheri Sond.

Ipomoea batatas (L.) Lam.

Ficus burkei (Miq.) Miq.

Ficus stuhlmannii Warb.

Dicerocaryum zanguebarium

(Law.) Merr. (seed)

Cyperus la tif olius Poir.

(Ni-)

nkahlwa/nkahlwane

nkamba/nkhamba

nkanye/nkanyi

nkaya/nkayi

nkenge

nkhayani

nkhohlwane

nkototso/nkolokotso

nkombekwa/nkombelwa

nkonola/nkonolo/nkonono

nkowakowa

nkubatsebi

nkuhlu

nkungulatilo/nkwangulatilo

nkuwa

nkuwa yatintsava

nkwakwa

Tabernaemontana elegans Stapf

Bolusanthus speciosus (H. Bol. )

Harms

Sclerocarya caffra Sond.

Acacia nigrescens Oliv.

Acacia caffra (Thunb.) Willd.

Burkea africana Hook.

Acacia caffra (Thunb.) Willd.

Bolusanthus speciosus

(H. Bol.) Harms

Piliostigma thonnongii

(Schumach.) M.-Redh.

Crossopteryx febrifugia (Afzel. ex

G. Don) Benth.

Terminalia sericea Burch, ex DC.

Acacia sieberana DC. var. woodii

(Burtt Davy) Keay & Brenan

Cassine transvaalensis (Burtt

Davy) Codd

Trichilia emetica Vahl

Asparagus virgatus Bak.

Ficus capensis Thunb.

Ficus sycamorus L.

Ficus soldanella Warb.

Strychnos madagascariensis Poir.

(Nn— N(-)

nombela

nongonongo

nqokolo

nsala

nsangasa

Bequaertiodendron magalismon-

tanum (Sond.) Heine & J. H.

Hemsl.

Maerua parvifolia Pax

Flacourtia indica (Burm. f.)

Merr.

Strychnos spinosa Lam.

Securinega virosa (Roxb. ex

Willd.) Pax & K. Hoffm.

C. A. LIENGME

517

nsasani

nsidwa

nsihana

nsihana yokulu

nsihaphukuma

nsisimbana

ntamungu

nthada/ntheda

nthamula

nthathasani

nthethenya

nthuma

nthuma nkulu

ntinta/ntita/ntiti

ntogwe

ntoma

ntomantomane

ntsalala

ntsenga

ntsembyane/ntswembyane

ntsengele

ntsengele yantsongo

ntsephukane

ntshiriri/ntshwiriri

nt sonde

ntsumbulu

ntswila

ntswukelane

ntungufana

(Nu — Nz-)

nugani

nulu

numanyama

nxakama

nxanatsi

nxelana/nxelela

nxenhe

nxojowa

nxunguxungu/nxungwezu-

ngwe

nyamarhi

nyawa

nyiri/nyiyi

nyokani

N’

n’wamidzumba

P

phalavurha

pundze

R

raboda

rhanga

ricinda

Acacia tortilis (Forssk.) Hayne

subsp. heleracanlha (Burch.)

Brenan

Acacia gerrardii Benth. var. ger-

radii

Dioscorea cotinifolia Kunth

Grewia flavescens Juss. var.

flavescens

Grewia flavescens Juss. var.

olukondae (Schinz) Wild

Dombeya rotundij olia ( Hochst.)

Planch, var. rotundij olia

Erythrina lysistemon Hutch.

Carissa edulis Vahl

Plectranthus esculenthus N.E. Br.

Tvloseina fassoglensis (Schweinf.)

Torre & Hillc.

Heteropyxis natalensis Harv.

Phyllanthus reticula tus Poir.

Solanum panduraeforme E. Mey.

Solanum incanum L.

Artabotrys brachypetalus Benth.

Cordia grandicalyx Oberm.

Diospvros mespilif ormis Hochst.

ex A. DC.

Diospvros natalensis (Harv.)

Brenan subsp. natalensis

Euclea crispa (Thunb.) Guerke

var. crispa

Gardenia spatulifolia Stapf &

Hutch.

Hyparrhenia dichroa (Steud.)

Stapf

Hyparrhenia dregeana (Nees)

Stapf ex Stent

Hyparrhenia hirta (L.) Stapf

Hyparrhenia rudis Stapf

Hvperthelia dissoluta (Nees)

Clayton

Hibiscus cannabinus L.

Ximenia americana L. var.

microphylla Welw. ex Oliv.

Ximenia caffra Sond. var. caffra

Ximenia americana L. var.

microphylla Welw. ex Oliv.

Grewia occidentalis L.

Bauhinia galpinii N. E. Br.

Pterocarpus angolensis DC.

Manihot utilissima Pohl

Vangueria infausta Burch.

Grewia occidentalis L.

Lippia javanica L.

Zanthoxylum capense (Thunb.)

Harv.

Balanites maughamii Sprague

Cassia abbreviata Oliv. subsp.

beareana (Holmes) Brenan

Hexalobus monopetalus (A.

Rich.) Engl. & Diels

Colophospermum mopane (Kirk

ex Benth.) Kirk ex J. Leonard

Pterocarpus rotundifolius (Sond.)

Druce subsp. rotundifolius

Afzelia quanzensis Welw.

Pseudolachnostylis maprouneae-

folia Pax

Croton megalobotrys Mull. Arg.

Ekebergia capensis Sparrm.

Phaseolus sp.

Berchemia discolor (Klotzsch)

Hemsl.

Crypto lepis capensis Schltr.

Fa urea saligna Harv.

Faurea speciosa (Welw.) Welw.

Ficus capreaefolia Del.

Setaria sp.

Colocasia antiquorum Schott

Cucurbita maxima Duch.

Cucumis melo L.

Gossypium herbaceum L. var.

africanum (Watt) J. B. Hutch

& Ghose

rihlanga

rikatsi/rikatsu

rindanga

rirhulwane

risama

risidwa

risot se/rixoto

rithadzi

ritiyi

rupi

S

sirudzu

sitole

swatima

swatima lowutsongo

swimbyambya

T

thyeka/thyeke

tihove tavalungu

tlongwe

tongwaan

tshaman’hwati

V

viriviri

vucema

vungwane

X

xicakala

xicindzu

xidengana

xifata/xifati

xifiringoma

xigoya

xihlangwa/xilangwa

xihovane

xikhavi

xikukutsu

xikwakwaxu

xikwenga

xilangamahlo

xilungwa

xilutana

xilutsi

xiluvarhi

ximahlomahlwane

ximapana

ximbukanyi

ximombonkanyi

ximombyana/ximupyani

ximowu/ximuwu

ximutswani

xinkanyana

xinkuwana

xinungu

xinungumafi

xinyokanyokane

xipalatsi

xipokota

xiputu

xirhombe/xirhomberhombe

Phragmites mauritianus Kunth

Pterolobium Stella turn (Forssk.)

Brenan

Lagenaria siceraria (Molina)

Standi.

Solanum panduraeforme E. Mey.

Cyperus sexangularis Nees

Dioscorea cotinifolia Kunth

Cocculus hirsutus (L.) Diels

Calha edulis (Vahl) Forsk. ex

Endl.

Ipomoea batatas (L.) Lam.

Colocasia antiquorum Schott.

Xerophyta retinervis Bak.

Xeromphis obovata (Hochst.)

Keay

Bridelia mollis Hutch.

Phyllanthus reticula tus Poir.

Bridelia mollis Hutch.

Amaranthus thunbergii Moq.

Lantana rugosa Thumb.

Hyparrhenia dichroa (Steud.)

Stapf.

Hyparrhenia hirta (L.) Stapf

Hyparrhenia tamba (Hochst. ex

Steud.) Anders, ex Stapf

Oncoba spinosa Forssk.

Cordia grandicalyx Oberm.

Capsicum frutescens L.

Hyphaene natalensis Kunze

Scirpus inclinatus (Del.) Aschers.

& Schweinf. ex Boiss.

Setaria sp.

Phoenix reclinata Jacq.

Piliostigma thonningii (Schu-

mach.) M.-Red.

Commiphora pyracanthoides

Engl.

Melia azedarach L.

Cyperus latifolius Poir.

Maytenus heterophylla (Eckl. &

Zeyh.) N. Robson

Setaria sp.

Combretum hereroense Schinz

subsp. hereroense var. here-

roense

Combretum apiculatum Sond.

subsp. apiculatum

Pappea capensis Eckl. & Zeyh.

Sansevieria hyacinthoides (L.)

Druce

Agave sp.

Spirostachys africanus Sond.

Heteropogon contortus (L.)

Beauv. ex Roem. & Schult.

Cucurbita maxima Duch.

Dalbergia melanoxylon Guill. &

Perr.

Dombeya rotundifolia (Hochst.)

Planch, var. rotundifolia

Aptosimum lineare Marloth &

Engl.

Cassine transvaalensis (Burtt

Davy) Codd

Lannea stuhlmannii (Engl.) Engl.

Lannea discolor (Sond.) Engl.

Adansonia digitata L.

Lannea discolor (Sond.) Engl.

Ficus capreaefolia Del.

Ozoroa reticulata (Bak. f.) R & A

Fernandes subsp. reticulata var.

reticulata

Ozoroa engleri R. &. A. Fernan-

des

Gnidia rubescens B. Peterson

Dalbergia melanoxylon Guill. &

Perr.

Cussonia spicata Thunb.

Blumea gariepina DC.

Ficus burkei (Miq.) Miq.

Ficus stuhlmannii Warb.

518

PLANTS USED BY THE TSONGA PEOPLE OF GAZANKULU

xirhombehleta Zanthoxylum capense (Thunb.) xivopfu Themeda triandra Forssk.

Harv. xivudzi Artabotrys brachvpetalus Benth.

xirhomberhombe xantsava Ficus sonderi Miq. xixwene/xixwenelamhofu Spilanthes mauritiania (Pers.)

xisidani Acacia davyi N.E. Br. DC.

xitsalala Gardenia spatulifolia Stapf & xiyimanamurhi Maerua angolensis DC.

Hutch.

Flacourtia indica (Burm. f.)

xivambula

xotse/xotso

Merr.

Cocculus hirsutus (L.) Diels

■

'

-

■ - 4 ’

Bothalia 13, 3 & 4: 519-525 (1981)

A brief account of coast vegetation near Port Elizabeth

H. C. TAYLOR* and J. W. MORRIS**

ABSTRACT

The environment and vegetation of an area of coast north-east of Port Elizabeth, South Africa, are briefly de-

scribed. Five major vegetation categories are recognized. These are: 1. Scaevola thunbergii Pioneer Vegetation of

littoral dunes and Ficinia lateralis Sedgeland of calcrete gravel; 2. Olea exasperata Bush, Pterocelastrus

tricuspidatus Bushclumps and Dune Woodland; 3. Sundays River Scrub; 4. Fynbos of calcrete areas; and 5. The-

meda triandra Grassland. Invasion of the area by Acacia cyclops (Australian Wattle) is described. A recommenda-

tion for the conservation of the area as a nature reserve is made.

RESUME

UN BREF COMPTE-RENDU DE LA VEGETA TION COHERE PRES DE PORT ELIZABETH

Le milieu et la vegetation d’une region cohere au nord-est de Port Elizabeth en Afrique du Sud, sont brievement

decrits. Cinq categories principals de vegetation sont reconnues. Elies sont ; 1. La vegetation Scaevola thunbergii

pionniere des dunes du littoral et Ficinia lateralis Sedgeland de gravier calcaire; 2. Olea exasperata Bush,

Pterocelastrus tricuspidatus Bushclumps et Dune Woodland; 3. Les buissons de la riviere Sundays; 4. Buissons de la

region calcareuse; et 5. Le Themeda triandra du Grassland. L’invasion de la region par /’Acacia cyclops (Wattle

australien) est decrite. Une recommandation pour la conservation de la region comme reserve naturelle est fade.

INTRODUCTION

Some years ago the Botanical Research Institute

was asked to report on the plant communities and

rare and endangered species, if any, to be found

within an area proposed for the development of an

iron-ore loading berth on the coast opposite St Croix

Island, approximately 25 km north-east of Port

Elizabeth. An account of the results of the investiga-

tion, based on a number of visits to the area by the

authors of this paper, was given by Taylor (1976). As

no prior detailed ecological studies had been under-

taken in the area, a relatively intensive study had to

be made for reporting purposes. It is intended that

this brief account will make the interesting informa-

tion collected more readily available than in an un-

published Departmental report.

The area of the investigation, from the Coega

River in the west to the Sundays River in the east, is

bounded by the sea to the south-east and the national

road from Port Elizabeth to Grahamstown to the

north-west. The area extends about 18 km along the

coast and about 4 km inland (Fig. 1).

Regional accounts by Schonland (1919) and Dyer

(1937) have included brief descriptions of vegetation

similar to that found in the study area. Olivier (1977)

compiled a checklist from a nearby area while Penz-

horn et al. (1974) described the vegetation of the Ad-

do Elephant National Park which contains some of

the same veld types although it is not on the coast.

Edwards (1971) briefly described the terrestrial

vegetation of the Swartkops Estuary area, a few km

west of the study area.

PHYSICAL FEATURES

For approximately 300 m inland from the high

water mark the area consists of relatively flat calcrete

beds with occasional sand dunes. The ground then

’Botanical Research Unit, Botanical Research Institute, Depart-

ment of Agriculture and Fisheries, P.O. Box 471, Stellenbosch,

7600.

“Datametrical Services, Department of Agriculture and

Fisheries, Private Bag XI 16, Pretoria, 0001; formerly Botanical

Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

rises, gently at first and later sharply, until it forms a

low ridge which runs roughly parallel to the beach,

approximately 1 km from the sea and 60 m above sea

level. The seaward slope of the ridge consists of deep

sand, partly exposed as open dunes and partly bush-

covered. Inland of the ridge the land slopes gently

downward toward the national road and is relatively

flat with occasional depressions and high points.

There are no permanent water bodies except for the

two rivers bounding the area to the east and west.

Geologically, the area is underlain by shelly and

chalky limestones, sandstones and conglomerates of

the Alexandria Beds. These are exposed along parts

of the ridge, inland of which the Beds are covered by

a layer of sandy soil 10 to 40 cm deep.

The area is classified as semi-arid, with a tempe-

rate, oceanic climate. The average maximum tempe-

rature at Port Elizabeth, 25 km to the south-west, is

25,5°C, the average minimum is 7,1°C, the extreme

520

A BRIEF ACCOUNT OF COAST VEGETATION NEAR PORT ELIZABETH

Pori Elizabeth (62m)

E60-1

w so.

Fig. 2. — Climate diagram for Port Elizabeth, a, duration of rain-

fall record in years; b, duration of temperature record in

years; c, altitude; d, mean annual temperature; e, mean

annual precipitation; f, absolute maximum temperature;

g, mean daily maximum of hottest month; h, mean daily

range of temperature; i, mean daily minimum of coldest

month; j, absolute minimum; k, trace of monthly means

of temperature; and 1, trace of monthly means of precipita-

tion. Data from Anon. (1954) and Anon. (1965).

July

| 3-8 |9-«jl«-25|26-38|m.p.h

Arcs represent 5% Intervals

Fig. 3. — Wind roses from Anon (1960) for Port Elizabeth

(1 m.p.h. = 1,6 k.p.h.).

maximum is 34°C and the extreme minimum is 0°C

(Fig. 2). The prevailing winds are south-west, east

and south-east in January and west in July (Fig. 3).

The average annual rainfall over a recent five-year

period was 456 mm and the long-term average for

Port Elizabeth is 614 mm (Anon, 1965).

NATURAL VEGETATION

Floristically, the area is by no means poor in

species, except for the pioneer vegetation of the lit-

toral dunes. Phytogeographically, this area of coast

is surprisingly interesting, containing elements of the

fynbos flora of the south-western Cape, the grass-

lands of the interior plateaux, the subtropical to

temperate coastal forests, and the possibly forest-

derived woody flora peculiar to the semi-arid valleys

of the eastern Cape. Although these elements inter-

mingle to some extent, especially the grassland and

fynbos, a single element usually predominates in each

of the major vegetation types.

Five major vegetation categories are each briefly

described below in terms of habitat, formation class,

cover, physiognomy, species numbers in releves,

dominance (if any), phytogeography and probable

floristic affinities (where known).

'Specimens have been deposited in STE with duplicates in PRE.

1. Pioneer Vegetation

Two communities occur on the coast edge. Des-

criptions of both are given below.

(a) Scaevola thunbergii Pioneer Vegetation

This community occurs on the littoral dunes. It is a

semi-succulent herbland/dwarf shrubland with 20 to

65% cover, dominated by Scaevola thunbergii (Fig.

4). It is a very simple, pioneer community with an

average height of 30 cm and rarely exceeds 50 cm. It

consists mostly of individual erect stems of 5. thun-

bergii with its spreading, fleshy leaves. The only

other species are, in order of decreasing abundance,

Tetragonia decumbens, Chrysanthemoides monili-

fera, Gazania rigens and the intrusive plant invader.

Acacia cy clops.

(b) Ficinia lateralis Sedgeland of calcrete gravel

This community occurs inland of the littoral dunes

and is developed on calcrete. It is a low (15 cm),

closed community dominated by the sedge Ficinia

lateralis, with a few emergent shrubs of Passerina

vulgaris, which are 60 to 90 cm tall. In places

Helichrysum sp. ( Taylor 9139)* takes the place of F.

lateralis. This is another simple community and is not

extensive.

2. Forest Precursor Communities

Three communities are recognized, all on recent

sands, usually old dunes, on top of and on either side

of the ridge. All contain, as dominants, species found

in the early successional stages of Coast Forest.

(a) Olea exasperata Bush

Total cover in this woody community is 65 to 85%.

It is dominated by Olea exasperata which forms an

almost closed, woody layer 0,5 to 1,2 m tall (Fig. 5).

An open, lower layer, characterized by Restio eleo-

charis, with spreading, succulent and ericoid sub-

shrubs and about 20 cm tall, is present. There is a

sparse to very sparse ground layer of small succulents

and annuals, 5 to 10 cm tall. Species numbers in

releves were between 30 and 40. The Bush occurs on

shallow sand over calcrete on the ridge and is thought

to be similar to Olea exasperata Bush on the Cape

Flats.

(b) Pterocelastrus tricuspidatus Bushclumps

Within the general area of Olea Bush, scattered

Bushclumps occur where the blown sand is slightly

deeper (25 to 50 cm) over the calcrete of the ridge

(Fig. 6). Only a small number of clumps, each from

30 to 50 m2 in area, are present. Total cover is from

95 to 100%, dominated by Pterocelastrus tricuspida-

tus, which is 3 to 4 m tall, with other woody species

including Olea exasperata, Rhus cf. glauca, Euclea

undulata, E. natalensis and E. racemosa. Lower lay-

ers are sparse and generally poorly defined. Species

numbers in releves were between 20 and 30.

(c) Dune Woodland

This closed woodland formation, probably a

higher successional stage beyond Pterocelastrus tri-

cuspidatus Bushclumps, occurs on deep dune sand

sheltered from on-shore winds. The only sample ex-

amined was on the lee side of the dune just west of

the Sundays River bridge on the national road (Fig.

7). Total cover is 95% and average height from 4 to

5 m. The Woodland is dominated by Rhus crenata,

with Sideroxylon inerme and Brachylaena discolor

emergent to over 6 m. Lower layers are virtually ab-

sent. Species number in a releve was 23. Dune

Woodland is clearly related to similar scrub in the

H. C. TAYLOR AND J. W. MORRIS

521

Fig. 4. — Scaevola thunbergii in-

vading First dunes.

Fig 5. — Olea exasperata is domi-

nant woody sub-shrub in fore-

ground; here less than 0,5 m

tall.

Fig 6 .—Pterocelastrus tricuspida-

tus Bushel ump with sea and St

Croix island in background.

522

A BRIEF ACCOUNT OF COAST VEGETATION NEAR PORT ELIZABETH

Fig. 7. — Dune woodland on bank

of Sundays River. Dominants

include Brachylaena discolor,

Sideroxylon inerme and Rhus

crenata.

Knysna region, with a few additional Natal elements,

including Brachylaena discolor and Sanseviera

thyrsiflora.

3. Sundays River Scrub

Although containing a few elements also found in

the Forest Precursor Communities (e.g. Sideroxylon

inerme, Pterocelastrus tricuspidatus and Rhus

crenata), this Scrub is nevertheless physiognomically

and floristically distinct and probably has a different

origin and history. Acocks (1975), in his broad scale

treatment, mapped the whole of the study area as

Valley Bush veld, a specialized and complex group of

vegetation units. In his Memoir, the description of

Valley Bushveld Variation (d) (ii), Sundays River

Scrub, fits this community very well. Practically all

the trees, shrubs and climbers that he lists as of

general occurrence in the Sundays River Scrub were

found in at least one of the four samples studied, and

many of the species of less general occurrence were

also present. Acocks considered that the Sundays

River Scrub Variation was derived directly from the

Alexandria Forest, which he described as a decidedly

xerophytic variation of coastal tropical forest. This

explains the lack of close relationship with the Forest

Precursor Communities, which appear to represent

the beginnings of the more typical coastal subtropi-

cal-temperate forest in this region. Acocks also con-

sidered that the Sundays River Scrub was closely

related to the Gourit7 River Scrub of tne valleys west

of me Knysna forest region. A brief investigation of

Gouritz River Scrub in the Klein Brak River Valley

confirmed Acocks’s surmise, but further com-

parative study is needed to clarify the relationships.

In the area studied, Sundays River Scrub is a bush

formation occurring chiefly on light brown, shallow

sandy soils in the Coega River Valley, spreading in-

land onto the flats near the national road. West of

the Coega River it occurs extensively on similar flats

as far as the Swartkops River. Where relatively un-

disturbed, it consists of a thicket of stout, often

thorny shrubs, with a 2 to 3 m canopy and a total

cover of 95 to 100%. In parts where browsing cattle

have penetrated in numbers, as in the lower part of

the Coega Valley, the canopy has been reduced to 75

to 80% cover and to 1 to 2 m in height. On the

borders of openings and along edges is a sparse,

lower layer of soft-leaved herbs and where openings

are extensive, smaller (20 to 30 cm), predominantly

ericoid dwarf shrubs occur which, in a more detailed

treatment, might be considered a separate communi-

ty. There is also a high proportion of succulents,

especially along the margins and in small openings.

Many climbing species are present, often sprawling

over the top of the canopy.

The community is floristically the richest encoun-

tered, with 50 to 60 species being recorded in 300 m2

plots (Fig. 8). Euclea undulata was present in every

releve and was usually dominant. Other species with

high frequency, and sometimes attaining dominance

or sub-dominance, are Rhus longispina, Cassine

tetragona, Schotia afra, Sideroxylon inerme, Azima

tetracantha and a climber Rhoicissus tridentata. Aloe

ferox is a prominent emergent, sometimes being up

to 4 m tall. Portulacaria afra, though not in every

releve, appears to be associated with high succes-

sional development and high species diversity.

Panicum maximum was found in every releve and is a

characteristic margin species.

4. Calcrete Fynbos

This is a dwarf shrubland formation occurring on

very shallow soil (about 10 cm deep) over calcrete on

the ridge, or its inland slopes (Fig. 9). Total cover is

from 65 to 70%. An upper layer (barely 20 to 30 cm

tall) of dwarf or sprawling shrubs (including ericoids)

constitutes over 50% of the cover. A few emergents

(40 to 50 cm tall) and a lower layer of creeping succu-

lents and very short grasses, 10 cm tall, are present.

Species numbers in releves were between 25 and 35

and characteristically there is no single species domi-

nant in the less disturbed stands, but the family Ruta-

ceae is prominent and a Restio sp. ( Taylor 9132) is

occasionally present. No close affinities are known,

but Ficinia truncata, a species found in all releves,

relates this community, at a high phytosociological

level, to communities in similar habitats along the

southern Cape coast, e.g. in the vicinity of Bredas-

dorp.

H. C. TAYLOR AND J. W. MORRIS

523

Fig. 8. — Open, short phase of Sun-

days River Scrub.

Fig. 9. — Calcrete Fynbos in fore-

ground (note white pebbles).

Sundays River Scrub visible in

background where calcrete is

replaced by deeper, brown,

sandy soils.

Fig 10. — Themeda triandra Grass-

land with isolated pocket of

Sundays River Scrub on right.

Note windspray ‘pruning’ of

canopy.

524

A BRIEF ACCOUNT OF COAST VEGETATION NEAR PORT ELIZABETH

5. Grassland

Themeda triandra Grassland is found on the flat

plateau inland of Calcrete Fynbos, where the soil is

not quite as shallow but is still underlain by calcrete

(Fig. 10). Its upper layer is similar in height to that of

Calcrete Fynbos (10 to 30 cm), but denser with over

50% of the 75 to 95% total cover being contributed

by Themeda triandra, which is clearly dominant.

Ehrharta calycina is sometimes subdominant. The

ten other grass species found in the three samples

studied do not contribute greatly to the cover, but in-

dicate the mixed origin of this community. Elionurus

muticus, Eustachys paspaloides, Setaria sphacelata

and Digitaria sp. ( Taylor 9091) are mainly distributed

to the north and east, whereas Cymbopogon marg-

inatus and Plagiochloa uniolae are mainly concentra-

ted to the south-west. Among the forbs, too, there is

a mixture of tropical-derived genera, like Barleria,

Blepharis, Euphorbia and Tephrosia, with genera

typical of the fynbos of the south-western Cape, e.g.

Diosma, Ficinia, Stoebe and Passerina. Due, perhaps

to its mixed nature, this community had the second

highest species numbers in releves in the area,

namely, 30 to 45.

On overgrazed sites, Themeda triandra assumes a

minor role and dominance is taken over by

Aspalathus lactea ssp. adelphea, Osteospermum im-

bricatum, Selago corymbosa, Psoralea fruticans and

Euphorbia rhombifolia, amongst others. On pro-

tected sites, by contrast, individual pioneer elements

and isolated small thickets of Sundays River Scrub

were noted which, from the nature of their margins,

suggest that the Scrub may be spreading at the ex-

pense of the Grassland (Fig. 10). According to one

local landowner, this is substantiated by accounts of

early settlers who referred to the whole area as a

grassy plain.

The balance between Grassland and Calcrete Fyn-

bos also appears to be delicate, being influenced by

depth of soil and by degree of grazing and trampling.

Local farmers consider that fire is of minor impor-

tance as an ecological factor because the evenly-

spread rainfall (Fig. 2) creates conditions that limit

the occurrence and spread of wildfires, and the

palatable grazing makes the use of intentional burn-

ing to produce new growth unnecessary.

EXOTIC PLANT INVADERS

About thirty-five years ago a landowner intro-

duced the Australian Wattle, Acacia cyclops (also

known as Rooikrans), to stabilize the driftsands on

his property. This species now forms a thicket on the

seaward escarpment and is actively invading the

Forest Initial Communities on the ridge and the

Grassland and Fynbos beyond (Fig. 11). The

Sundays River Scrub, due to the density of its

canopy, is as yet largely free of Wattle, but where

openings are being enlarged by cattle trampling, the

Australian invader is gradually penetrating.

The farmer maintains that the infestation has not

noticeably increased over the years, but judging from

the abundant young growth of Wattle, especially

along tracks, and its active suppression of the grass

cover, colonization by this invader is proceeding

apace. The tempo may, however, be less than in the

Mediterranean-type climate of the south-western

Cape where frequent dry-season fires encourage very

dense regeneration from the heavy seed load in the

soil.

Infestation in the Forest Precursor Communities,

Grassland and Fynbos has now reached the critical

stage where further spread of the invader will destroy

the structure of the natural vegetation over one

quarter to one third of the study area. If these plant

communities are worth saving, both for their value as

grazing and for their scientific interest and natural

beauty, prompt action is necessary.

CONSERVATION

As far as is known, none of the communities

described in this report are conserved at present, ex-

cept for a few areas of Sundays River Scrub in the

Tippers Creek Aloe Reserve at Swartkops, an area

too small for viable ecosystem conservation. The

Addo National Park consists of Addo Bush which,

according to Acocks (1975), is a different variety of

Valley Bushveld. The plant communities of the study

area are of great variety, scientific interest and

economic and recreational value. Some of them, such

as Calcrete Fynbos, are not extensive elsewhere. It is

our recommendation that the entire study area,

bounded on three sides by natural features (Sundays

River, Indian Ocean and Coega River) and on the

fourth by a major road, should be conserved. A

nature reserve of this size would be a worthy asset to

the country as a sample of relatively well-maintained

vegetation and a place for recreation and scientific

study within 25 km of the fast-growing city and third

largest port in South Africa, Port Elizabeth.

UITTREKSEL

Die omgewing en plantegroei van ’n gedeelte van

die kus oos van Port Elizabeth, word kortliks be-

skryf. Die volgende vyf hoof-vegetasiekategoriee

word onderskei: /. Scaevola thunbergii -pioniervege-

Fig. 11. — Invasion of grassland by

Acacia cyclops (Rooikrans).

525

H. C. TAYLOR AND J. W. MORRIS

tasie van strandduine en Ficinia lateralis -biesieveld

van kalkgruis; 2. Olea exasperata-bos, Pterocelastrus

tricuspidatus-6o5A:/om/?e en duinewoud; 3. Sondags-

rivierstruikgewas; 4. Fynbos van kalkgebiede; en

5. Themeda triandra -grasveld. Indringing dear

Acacia cyclops ( Rooikrans ) in die gebied word

beskryf. ’n Voorstel vir die bewaring van die gebied

as ’n natuurreservaat word gemaak.

REFERENCES

Acocks, J. P. H., 1975. Veld types of South Africa. 2nd ed. Mem.

bot. Surv. S. Afr. No. 40.

Anon., 1954. Climate of South Africa. Part I. Climate statistics.

Pretoria: Weather Bureau.

Anon., 1960. Climate of South Africa. Part 6. Surface winds. Pre-

toria: Weather Bureau.

Anon , 1965. Climate of South Africa. Part 9. Average monthly

and annual rainfall and number of rain-days up to the end of

I960. Pretoria: Weather Bureau.

Dyer, R. A., 1937. The vegetation of the Divisions of Albany and

Bathurst. Mem. bot. Surv. S. Afr. No. 17.

Edwards, K. Z., 1971. Vegetation. In Anon., The Swartkops

Estuary. Port Elizabeth: The Zwartkops Trust.

Olivier, Maria C. 1977. A systematic check list of the sperma-

tophyta of the Baakens River valley, Port Elizabeth. Jl S.

Afr. Bot. 43: 145-159.

Penzhorn, B. L., Robbertse, P. J. & Olivier, Maria C., 1974.

The influence of the African elephant on the vegetation of the

Addo Elephant National Park. Koedoe 17: 137-158.

Schonland, S., 1919. Phanerogamic flora of the Divisions

of Uitenhage and Port Elizabeth. Mem. bot. Surv. S. Afr.

No. 1.

Taylor, H. C., 1976. Report on the plant ecology of the St Croix

coast. Unpubl. Report, Pretoria: Department of Agricultural

Technical Services.

Bothalia 13, 3 & 4: 527-530 (1981)

Changes in the herb layer of the riverine woodland in the

Sengwa Wildlife Research Area, Zimbabwe

P. R. GUY*

ABSTRACT

The changes in the dominant species of the herb layer, in particular Blumea gariepina DC., of riverine woodland

in the Sengwa Wildlife Research Area in Zimbabwe, were monitored over a period of five years. Factors contribu-

ting to these changes appear to have been rainfall, utilization and interspecific competition.

RESUME

CHANGEMENTS DANS LA COUCHE D'HERBE DU WOODLAND RIVERAIN DANS LA REGION DE

RECHERCHE SUR LA VIE SAUVAGE DE SENGWA EN ZIMBABWE

Les changements dans les especes dominantes de la couche d’herbe, en particulier Blumea gariepina DC., du

woodland riverain dans !a region de Recherche sur la Vie Sauvage de Sengwa en Zimbabwe ont ete observes pendant

une periode de cinq ans. Les facteurs contribuant a ces changements apparaissent avoir ete les chutes de pluie,

^utilisation et la competition interspecifique.

INTRODUCTION

Following a period of below average rainfall, a

perennial, unpalatable, woody herb Blumea garie-

pina DC., dominated an area heavily utilized by a

variety of game species. The three year period of low

rainfall together with the heavy game pressure may

have caused the death of other species through over-

utilization, allowing B. gariepina, an acknowledged

pioneer of bare ground (Wild, 1969; Hilliard, 1977),

to become established in the absence of competition

(Goodman, 1975). As a result of its effect in decreas-

ing the amount of palatable food in the herb layer, it

was considered worthwhile to monitor the post inva-

sion period of the riparian herbaceous layer by this

unpalatable plant species.

STUDY AREA

The work was carried out in the Sengwa Wildlife

Research Area (18° 10* S; 28° 14* E), an area of 373

km2 lying at the southern end of the Chirisa Safari

Area, Zimbabwe. The average rainfall for the area

for the past thirteen years is 662 mm. The mean an-

nual temperature is 22,2°C (Torrance, 1965). The

vegetation is generally described as deciduous miom-

bo savanna woodland on the sandy soils, and dry

early deciduous savanna woodland, which is

dominated by Colophospermum mopane on the

heavier lower lying clay soils (Wild & Grandvaux-

Barbosa, 1965). The major rivers of the area, the

Lutope, Manyoni and Sengwa, have well-developed

fringes of riparian woodland dominated by Acacia

albida, A. tortilis subsp. heteracantha, Kigelia pin-

nata, Lonchocarpus capassa, and Trichila emetica.

The shrub layer is generally well developed being

composed of, in particular Combretum mossam-

bicense, Diospyros senensis, Grewia flavescens and

Securinega virosa (Cumming, 1975). The establish-

ment of B. gariepina occurred mainly in the riverine

woodlands of the Sengwa and Lutope Rivers, areas

heavily utilized by wildlife, but isolated groups of

this species were found throughout the area in all

vegetation types.

♦Formerly of the Department of National Parks and Wildlife

Management, Zimbabwe. Present address: 72-5 Castlebury Cres-

cent, Willowdale, Ontario, M2H 1W8, Canada.

METHODS

Changes in the herb layer were recorded beginning

in September 1974 using fifty quadrats each one

metre square placed one metre apart along a random-

ly laid line in the central portion of the study area.

The data collected from each species in the quadrats

were density, maximum height to the nearest centi-

metre and above-ground biomass. The last-men-

tioned was measured by harvesting all the plants of

each species at ground level within the quadrat.

These plants were subsequently oven dried at 95°C

for three days or longer immediately after harvesting.

The recordings were made in September of each year

along newly laid randomly placed lines.

The rate of colonization was studied by grading an

open patch of about 1 000 m2 free of all vegetation

except for deeply rooted, small ( < 50 cm high),

woody plants such as Acacia tortilis subsp. hetera-

cantha and Combretum mossambicense. The grader

blade was angled to remove the top 7,5 cm of soil to

leave about 5,0 cm of the A horizon soil. In this way

all seeds that may have been lying on the surface were

also effectively removed. The frequency of the plants

was determined in 40 permanently placed quadrats.

The quadrat size used was 0,25 m2. The quadrats

were randomly placed within the graded area. These

determinations were begun in January 1976 and con-

tinued each month for a year. The graded plot was

allowed a ‘settling down’ period of three months

before determinations were begun to allow the plants

to grow to a size at which they could be identified

with certainty. Determinations were discontinued

after a year, because no differences between the fre-

quency of plants within the plot and the same species

outside of the plot could be detected.

In order to obtain an idea of the age to which B.

gariepina will live, ten seedlings were tagged with

aluminium labels in September 1974.

RESULTS

The results from the first part of the study in which

the changes in density, frequency and biomass were

measured, are presented in Table 1 and Fig. 1. The

monthly frequencies for the most important species

in the graded study area are shown in Fig. 2.

CHANGES IN THE HERB LAYER OF THE RIVERINE WOODLAND IN THE SENGWA WILDLIFE RESEARCH AREA

ZIMBABWE

1974

1975

1976

1977

1978

_ — No of rain days /month

Blumea gariepina

Diet i ptera verticillata

Ocimum urticifolium

Solanum panduriforme

Grass

Others

Fig I . — Changes in biomass, density and frequency of the most important species in the study area. The rainfall for the

year preceding recording is given. The annual total rainfall and annual total biomass are indicated on each respec-

tive histogram.

P. R. GUY

529

TABLE 1. The changes in average density, frequency, average maximum height and total above ground biomass of the most important

species in the five years of recording

Average density (per rn2) Frequency ( % ) Average max. height (cm)

Total above ground biomass

(gm/m2)

tr = trace (< 0,01)

Fig. 2. — Changes in the frequency

of Acacia tortilis subsp.

heteracantha and herbs by

month in the graded study

area in 1976. a, Acacia tortilis

subsp. heteracantha ; b,

Amyranthus hybridus\ c,

Blumea gariepina\ d, Boer-

havia diffusa ; e. Cassia sp. ; f,

Conyza aegyptiaca; g, Cor-

chorus tridens\ h, Dactylo-

ctenium giganteunv, i, Diclip-

tera verticillata\ j, Eragrostis

cilianensis ; k, Indigofera spp.;

I, Ocimum urticifoliunr, m,

Oldenlandia herbacea\ n,

Pterocaulon decurrens ; o,

Sid a alba\ p. Sida cordifolia;

q, Solanum panduriforme\ r,

Urochloa trichopus.

530 CHANGES IN THE HERB LAYER OF THE RIVERINE WOODLAND IN THE SENGWA WILDLIFE RESEARCH AREA,

ZIMBABWE

In 1974, immediately following the establishment

of Blumea gariepina, neither Dic/iptera verticillata

nor Ocimum urticifolium were well represented in the

herb layer (Table 1, Fig. 1). Blumea gariepina was

well established as the dominant species. In 1975,

although there were large numbers of Dicliptera ver-

ticillata, Blumea gariepina was still the most impor-

tant species, but by 1976 some three years after its es-

tablishment, B. gariepina had declined in dominance

and only represented 0,2% of the total biomass

(Table 1). At this time Oci urticifolium contributed

the most to the total biomass and Dicliptera verticil-

lata contributed the most to density. In the following

year, D. verticillata was the dominant species in

terms of both density and biomass. In 1977, the in-

dividuals of D. verticillata consisted of well formed

plants in contrast to the numerous spindly seedlings

of the previous year. This would account for its in-

crease in biomass despite the considerable decrease in

density (Table 1, Fig. 1). No single species was domi-

nant in 1978, as both Dicliptera verticillata and

Ocimum urticifolium had biomass figures of about

43 gm m-2 (Table 1). The increasing importance of

perennial grasses is of interest.

There appears to be a definite successional process

taking place in the herb layer. All four important

species show the same phenomenon. Blumea

gariepina was observed on the downward part of the

process, Ocimum urticifolium and Dicliptera verticil-

lata through the full process with peaks in 1976 and

1977 respectively, and it appears that the perennial

grasses are on the upward part of the succession. Of

the 19 species (except grasses) recorded in the study

area only five were present in all five years of study

(Table 1).

The effectiveness of Blumea gariepina as a coloni-

zer is clearly illustrated in Fig. 2. It rapidly increased

in frequency and having reached a high frequency re-

mained at that level. Other species reacted different-

ly. The annual grasses Urochloa trichopus and Dac-

tyloctenium giganteum decreased in frequency in

May, at the end of the rainy season. Other species,

Boerhavia diffusa and Dicliptera verticillata main-

tained their frequency albeit at a lower level, and

some species, Oldenlandia herbacea and Pterocaulon

decurrens, were present for only part of the year.

Most species, however, had specific patterns of col-

onization, some were late colonizers particularly

Conyza aegyptiaca and Ocimum urticifolium, and

others Corchorus tridens and Sida alba increased in

frequency only to decrease rapidly thereafter.

DISCUSSION AND CONCLUSIONS

Some of the success of Blumea gariepina as a

pioneer can be attributed to its production of large

amounts of wind dispersed seeds and its tolerance of

wide soil and moisture conditions (Wild, 1969;

Hilliard, 1977). In areas of high animal concentra-

tions such as the study area, it success may be further

attributed to its unpalatability. The leaf of B. garie-

pina on steam distillation yields a volatile oil which

consists of 66% of cineol, 10% of d-fenchone and

about 6% of citral (Watt & Breyer-Brandwijk, 1962).

Cineol, the major constituent of oil of eucalyptus, is

poisonous to man, and citral is the major constituent

of lemon grass oil (Watt & Breyer-Brandwijk, 1962).

No animals have been observed to feed on B. garie-

pina despite its abundance. This may be related to the

presence of cineol and citral making the plant unplea-

sant smelling. It may afford some protection to other

species growing in its vicinity as Muller, Muller &

Haines (1964) observed that dew caused deposition

of volatile oils to take place on plants growing near

aromatic shrubs. A sheen was observed on some

herbs and grasses not normally glossy, growing at the

base of B. gariepina plants. Perhaps the protection

afforded to the grasses by these oils has encouraged

their growth despite the continued heavy utilization

of the area by game.

It is apparent that the initial fears that the area

would become dominated by the unpalatable B.

gariepina to the exclusion of the other species were

unfounded. Seedlings of B. gariepina tagged in 1974

survived for three years, indicating that the normal

life of the plant is probably not more than about four

years. High densities of B. gariepina can be expected

to be maintained for three or four years and possibly

longer if conditions are ideal. The periods of higher

rainfall may have caused the decline of B. gariepina,

but this may have also been affected by the superior

competitive abilities of the other species.

ACKNOWLEDGEMENTS

I would like to thank Dr D. H. M. Cumming and

Mr K. Dunham who read and criticized the manu-

script. Messrs H. Charidza and Z. Mhalangu pro-

vided help in the field. This paper is published with

the approval of the Director of National Parks and

Wildlife Management, Zimbabwe.

UITTREKSEL

Die veranderinge in die oorheersende spesies van

die kruidlaag, veral van Blumea gariepina DC., in

rivieroewerboomveld in die Sengwa Wildnavorsings-

gebied, Zimbabwe, is waargeneem gedurende ’n

periode wat oor vyf jaar gestrek het. Faktore wat

bygedra het tot hierdie veranderinge is veral die reen-

val, verbruik en wedywering tussen die verskillende

soorte plant e.

REFERENCES

Cumming. D. H. M., 1975. A field study of the ecology and be-

haviour of warthog. Museum Memoir No. 7. Salisbury: The

Trustees of the National Museums and Monuments of Rho-

desia.

Goodman, P. S., 1975. The relation between vegetation structure

and its use by wild herbivores in a riverine habitat. M.Sc.

thesis, University of Rhodesia (unpublished).

Hilliard. O. M., 1977. Compositae of Natal. Pietermaritzburg:

University of Natal Press.

Muller. C. H., Muller, W. H. & Haines, B. C., 1964. Volatile

growth inhibitors produced by aromatic shrubs. Science 143:

471-473.

Torrance, J. D., 1965. The temperature of Rhodesia. In M. O.

Collins, Rhodesia: its natural resources and economic de-

velopment. 28-29. Salisbury: M. O. Collins.

Watt, J. M. & Breyer-Brandwijk, Maria, 1962. Medicinal and

poisonous plants of southern and eastern Africa. 2nd edition,

London: E. & S. Livingstone.

Wild, H., 1969. The compositae of the Flora Zambesiaca Area, 2.

Kirkial: 121—135.

Wild, H. & GrandvaOx-Barbosa, L.A., 1967. Vegetation map of

the Flora Zambesiaca Area. Supplement to Flora Zam-

besiaca. In H. Wild & A. Fernandes. Flora Zambesiaca. Salis-

bury: M. O. Collins & Government Printer.

Bothalia 13, 3 & 4: 531-552 (1981)

Survival, regeneration and leaf biomass changes in woody

plants following spring burns in Burkea africana — Ochna

pulchra Savanna*

M. C. RUTHERFORD**

ABSTRACT

Effects of two intensities of spring burn on various aspects of woody plants of a Burkea africana — Ochna pulchra

Savanna after one growth season are given. Mortality of woody plants was very low with, for example, that of in-

dividuals of Ochna pulchra being between 1 and 5%. Some species where the above-ground parts were often burned

away completely, as in Grewia flavescens, no mortality of individuals occurred. Basal regeneration shoot mass was

found to depend parabolically on plant height while the ratio of leaf to twig mass in basal shoot regeneration varied

inversely with plant height in Ochna pulchra. The ability of Ochna pulchra plants to produce new basal shoots

appeared to not only depend on size of the plant but also on the number of basal shoots present prior to the fire. In

live Ochna pulchra plants basal regeneration shoot biomass per individual was found to increase exponentially with

greater reduction in canopy leaf biomass. This relation was also affected by possible direct heat effects. Basal shoot

regeneration mass was found to vary greatly with species and varied from 0,7 g/individual for Dichapetalum

cymosum to 285,6 g/individual for Euclea natalensis. There was a clear tendency for non-suffrutex shrub species to

have greater mean basal regeneration shoot mass per plant than that of most tree species. There was a compensatory

effect in Ochna pulchra between number and size of basal regeneration shoots. Standing dead woody plant in-

dividuals (before the burn) were either felled by fire or apparently unaffected by fire and there was no selectivity by

species. Results of the present study are generally supported by other work on the effects of fire in savanna and

some other vegetation types.

RESUME

SUR VI VANCE, REGENERATION ET CHANGEMENTS DE BIO-MASSA DE FEUILLE DANS LES

PLANTES LIGNEUSES A LA SUITE DES BRULAGES DE PRINTEMPS DANS LA SA VANNE BURKEA

AFRICANA— OCHNA PULCHRA

Les effets de brulage de printemps de deux intensites sur les aspects vanes de plantes ligneuses de savannes Burkea

africana — Ochna pulchra a pres une saison de croissance, sont donnes. La mortalite des plantes ligneuses fw tres

faible avec, par exemple, cede des individus r/ ’Ochna pulchra situant entre 1 et 5 %. Certaines especes dont la partie

au dessus du sol etait souvent completement brulee, comme Grewia flavescens ne montrerent aucune mortalite des

individus. On trouva que la regeneration de base des masses de rejets dependait parabolicallement de la hauteur de

la plante tandis que le rapport de feuille a la masse de brindille dans la regeneration des re jets de base variait inverse-

ment a vec la hauteur de la plante chez /’Ochna pulchra. L’aptitude des plantes r/ ’Ochna pulchra a produire de

nouveaux rejets de base apparait non seulement dependre sur la taille de la plante mais aussi sur le nombre de rejets

de base anterieurement presents au brulage. Chez les plantes d 'Ochna pulchra en vie, la bio-masse des rejets de

regeneration de base par individu fut trouvee s’accroitre exponentiellement avec une reduction plus grande dans la

bio-masse des feuilles du couvert. Cette relation fut aussi effectee par des effets possibles de chaleur directe. La

masse de regeneration des rejets de base fut trouvee varier considerablement avec les especes ekt elle varia de 0, 7

g/ individu pour le Dichapetalum cymosum a 285, 6 g/ individu pour /Tiuclea natalensis. II y eut une claire tendance

pour les espces de buissons non-sous-arbriseaux a avoir une plus grande moyenne de masse de rejets de regeneration

de base par plante que celle de la plupart des trois especes. II y eut un effet compensatoire dans /’Ochna pulchra

entre le nombre et la taille des rejets de regeneration de base. Les individus morts de plantes ligneuses debouts (avant

le brulage) furent soil abattus par le feu ou apparement pas atteints par le feu et il n ’y avoir pas de selectivity par les

especes. Les resultats de la presente etude sont generalement soutenus par d’autres travaux sur les effets desfeux de

savanne et certaines autres categories de vegetation.

CONTENTS

Page

1. Introduction 532

2. Selective literature review 533

Mortality/survival of plants after fire 533

Effects of fire on plant canopies 534

Fire and basal shoots 534

Interrelationships after fire 534

3. Method 534

The two burns 534

Experimental layout 535

Measurements 536

4. Results 539

Mortality/survival of plants after the

burns 539

Effects of fire on plant canopies 539

Canopy mortality 540

Occurrence of abnormal canopy leaf

growth 541

Changes in canopy height and

canopy volume 541

Changes in canopy leaf biomass 542

Effect of fire on basal shoot numbers 542

Effect of fire on production of new

basal regeneration shoot biomass 544

Relationships between plant height and

woody plant biomass components

after fire 544

Other pyro-allometric biomass rela-

tions 547

Effects of fire on standing dead in-

dividuals 548

5. Discussion and conclusions 549

Mortality/survival of plants after the

burns 549

Effects of fire on plant canopies 549

Fire and basal shoots 550

Biomass and other relations after fire 550

Acknowledgements 551

Uittreksel 551

References 551

* A publication of the South African Savanna Ecosystem Project.

** Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

532 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA—OCHNA PULCHRA SAVANNA

1. INTRODUCTION

Fire has probably occurred regularly in the savan-

nas of southern Africa, initiated first by agencies

such as lightning and later increasingly by man. The

structure of savanna is probably related to this

incidence of fire. In the past fire, once started, could

spread over vast areas of savanna but more recently

fire is often contained by artificial fire breaks such as

roads. Fire is used as a veld management tool for

both livestock farming and wildlife conservation.

The two main objectives of burning savanna for live-

stock farming are to remove moribund grass and to

decrease the woody plant element. In wildlife

management these two objectives are also important,

but additional reasons include encouraging some

form of rotational grazing, reducing the wild-fire

hazard and controlling animal diseases and parasites

(Trollope et al., in press). In the savannas of the

central-northern areas of the Transvaal the current

frequency of fire is extremely variable from annually

to only every few decades in more protected areas

although the more common frequency may be

described as between one and five years. Fires in

savannas are usually surface fires, and crown fires

are very rare although localized burning of canopies

can occur. Controlled fires are almost always applied

in or near the dry season, usually between May and

November in southern African savanna vegetation

areas. In these areas it is generally recognized that a

fire earlier in this period is less effective in controlling

tree growth than a fire later in this period (West,

1965). Effectiveness of late burns as a means of

woody plant control is usually ascribed to the fire

being hotter and the trees being more susceptible to

heat injury at a time when growth has just started.

However, Deeming et al., (1972) point out that this

stage of plant development (the rapid growth stage)

indicates a ‘high moisture content throughout the

plant’ which (as in other living fuels) acts as a heat

sink since it takes considerable amounts of energy to

dessicate this material. Only after dessication can

such material itself act as a heat source. In contrast to

burning in the dry season, it has been found (Anon,

1960) in the Transvaal lowveld that very little damage

is done to trees and shrubs during late summer

(February) veld burning when grass is green.

That fire can fundamentally affect many com-

ponents of an ecosystem has long been recognized.

Interest has been shown in many diverse effects, for

example, the cycling of nutrients (Trapnell et al.,

1976; Christensen, 1977), changes in the water

balance and the possible effects on past carbon dio-

xide changes in the atmosphere (Reiners & Wright,

1976), while work on woody plants has included

classifying savanna woody plant species according to

different degrees of fire tolerance with the study of

the evolution of pyrophytic habits (Jackson, 1974),

derivation of allometric fuel load prediction formu-

lae (Kaul and Jain, 1967), modelling fuel mass

(McNab et al., 1978) and determining fire scar tree

ring chronologies (Zackrisson, 1977).

The scientific Committee on Problems of the

Environment (set up by the International Council of

Scientific Unions) informally co-ordinates as part of

its mid term project 2 a short term (1977-80) interna-

tional programme for the review of the ecological

effects of fire. South African participation in this

programme is co-ordinated by the Working Group

for Fire Ecology of the Committee for Terrestrial

Ecosystems initiated by the National Programme for

Environmental Sciences of the Council for Scientific

and Industrial Research. The present paper forms

part of the contribution by the South African

Savanna Ecosystem Project on the ecological effects

of fire (Anon, 1978). Results of an experimental burn

within the South African Savanna Ecosystem Project

are being prepared for publication as a synthesis

report by M. V. Gandar (in press) where an overview

is given of effects of fire on the abiotic, decomposer,

primary producer and consumer components of the

ecosystem. The present paper is limited to the more

detailed effects of fire on the woody plant compo-

nent.

The South African Savanna Ecosystem Project is

being conducted on a portion of the recently-

established Nylsvley Nature Reserve (3 120 ha in ex-

tent), 10 km south of Naboomspruit in the northern

Transvaal. The basic ecological characteristics of the

study area are described in Huntley and Morris

(1978) while the project’s overall objectives and

research programme are outlined in Huntley (1978).

The study area lies on the edge of the Springbok

flats on a slightly raised plateau at about 1 100 m

above sea level. Most of the Waterberg System sand-

stone bedrock is covered by sandy soils belonging

mainly to the Hutton and Clovelly forms (Harmse,

1977). Mean annual rainfall is about 630 mm and

occurs mainly in summer. The mean annual air

temperature is 18,6°C. The study site’s past manage-

ment has included light summer grazing by cattle

with small populations of impala and fluctuating

populations of kudu present. The main vegetation

type of the study area has been classified as

Eragrostis pallens — Burkea africana Tree Savanna

(Coetzee et al., 1976) with the most extensive varia-

tion of this being the Eragrostis pallens — Dombeya

rotundifolia variation with dominant trees Burkea

africana and Tenninalia serieea and dominant shrubs

Ochna pulchra and Grewia flavescens. Huntley

(1977) has suggested that the broad-leaf savanna of

the study site is related to the mesic and moist broad-

leaf savanna biome of Africa. In the year before the

experimental burning, there had been no grazing by

cattle in the area to be burned. Since the area had

been unburned for some years and had a recent

history of very low grazing pressure the graminoid

fuel load was probably above average for the given

type of vegetation although the nutrient-poor sandy

soils may be expected to result in a generally lower

heat intensity and slower fire than in several other

vegetation types (Anon, 1960). The presence of trees

also tends to result in a reduction in graminoid fuel

load beneath them so that intensity and frequency of

burning may be expected to be reduced by the trees

(West, 1965). Given the tree leaf litter input each year

(from June to August) and the relatively low decom-

position rates (especially for leaves of Ochna

pulchra), thick layers, sometimes up to 10 cm deep,

occurred below certain trees. On the whole, the

ground fuel composition of the experimental site was

mixed with graminoid, tree leaf litter and some wood

pieces of varying dimensions.

The main objectives of the present study were to

determine the short term effects of two intensities of

spring burn on the individuals of each woody plant

species population in a selected area of the Nylsvley

study site, with particular attention being given to

mortality/survival, degree of canopy reduction,

degree of basal regeneration, changes in leaf biomass

and to relationships between these aspects.

M. C. RUTHERFORD

533

2. SELECTIVE LITERATURE REVIEW

Work on the effects of fire on individual woody

plants in African savanna appears to have been main-

ly concentrated on mortality or survival of plants

after fire. Less attention has been given to effects of

fire on plant canopies and the stimulation of basal

shoots. Very little work has apparently been done on

establishing and quantifying the interrelationships

between different plant dimensions and mass com-

ponents after fire (pyro-allometry). Such relations

may be used to predict, for example, the degree of

dependency of leaf mass re-distribution within the

plant on the size of the plant following fire. The

limited effort in this last mentioned respect possibly

relates to the difficulties involved in determining the

fire effects on the different organs of the woody

plant individual and to the variable degree to which

ground fires reach into the canopy stratum of woody

vegetation as well as to the often very heterogeneous

spatial distribution of woody plants in savannas.

Published work on the particular fire effects and

plant species considered in the present study has

included various, sometimes conflicting, findings.

a) Mortality /survival of plants after fire

Relatively low mortality of some woody savanna

plants with fire has been indicated by, for example,

the classification in Malawi of Burkea africana as a

‘pyrophytic’ species (Jackson, 1974). In the Kavango

region of South West Africa, Geldenhuys (1977)

tested for mortality in B. africana with fire in an

analysis of co-variance and despite the high pro-

portion of dead trees (35,3% at Rundu), this was not

attributable to treatment. Of the four most common

woody plant species in a Nigerian savanna after

annual (late winter) burns for five years, only B.

africana had all individuals (from 4 to 13 m height)

survive. B. africana increased relative to the other

species with fire and it is suggested that it is through

better adaptation to fires that certain B. africana

savannas may have arisen (Hopkins, 1965). In con-

trast to the above findings, van Rensburg (1971)

reported that some B. africana trees ‘were damaged

and killed’ in Terminalia sericea Woodland on sand

in eastern Botswana after an April burn (toward the

end of the normal rainy season). However, mortality

of Burkea africana trees has been reported to be not

necessarily linked to the effects of the fire in several

areas (Rutherford, in press). Tinley (1966) only refers

to coppices of B. africana being ‘very sensitive to

fire’ in the northern Okavango Swamps of

Botswana.

Effects of long-term application of fire may differ

from the expected short-term effects. Thus, for ex-

ample, in Terminalia sericea — Burkea africana

Savanna at Matopos, Zimbabwe, Kennan (1971)

reports fewer trees present in long-term fire treated

plots than in fire protected plots. There was also a

much smaller proportion of those trees below 0,9 m

in height in the fire treatments than in the protected

plots. In Burkea africana — Terminalia avicennoides

— Detarium microcarpum Savanna Woodland in

Nigeria, Afolayan (1978) found that annual ‘late’

(presumably late in the dry season) burning for four

years decreased tree density particularly for trees less

than 10 cm girth at breast height.

The effect of fire on seedlings has been observed in

Matopos savanna, Zimbabwe, where burning at

three year intervals resulted in many tree seedlings

being killed (West, 1965).

Other woody plant species in the present study

have been characterized on other sites in terms of

their ability to withstand fire. In the Transvaal

lowveld, Van Wyk (1971) reported that Terminalia

sericea and Dichrostachys cinerea were to a ‘certain

extent fire resistant’ and despite burning to ground

level sprouted vigorously after the burn. In Zambia,

Trapnell (1959) found that ‘the fire-tolerance of

Dombeya rotundifolia has been confirmed by

repeated field observation’, and that Strychnos

pungens was semi-tolerant with Lannea discolor pro-

bably also so. Lawton (1978) includes Ochna

pulchra, Strychnos pungens and Burkea africana in a

group that can survive intense dry season fires in

parts of north-eastern Zambia. In Wankie National

Park, Zimbabwe, Rushworth (1978) found that Ter-

minalia sericea and Ochna pulchra were ‘strikingly

frost hardy’ and points out the similar coppicing

reaction of some other species due to both frost and

fire. Geldenhuys (1977), in two regularly burnt areas

of the Kavango region of South West Africa/

Namibia, refers to Ochna pulchra as a ‘fire-sensitive’

species. However, it appears that this finding might

be limited to trees with measurable diameter at breast

height.

Many other woody plant species in African savanna

also appear to be relatively tolerant of fire. Thus in

Acacia thornveld (A. karroo) at Matopos, complete

killing of woody plants by fire of various fire regimes

was extremely rare and nearly all affected plants

regrew vigorously from their undamaged bases after

the fire (Kennan, 1971). In eastern Cape Acacia kar-

roo vegetation, Trollope (1974) found in a spring

head fire that mortality of A. karroo was 9,9% of

which more than three-quarters of the plants were

between 1 and 2Vi m tall. In the Molopo area of the

northern Cape, Donaldson (1966) reported that even

with dry grass artificially packed around the base of

Acacia mellifera subsp. detinens plants, the grass

burns (at various times of the year) only resulted in a

mortality of about 15%. Only when large amounts of

wood and twigs were burned at the bases did a 75%

mortality obtain here due to much longer duration of

heat. In other African savannas, Pterocarpus

angolensis has been classed as ‘fire-tolerant’ (Zam-

bia: Trapnell, 1959) and in east Africa Balanites

aegyptica Savanna no evidence was found of the

death of mature trees being affected by a mainly

annual fire regime (Harrington & Ross, 1974).

Although Trapnell (1959) reported the dominant

canopy species of Brachystegia, Julbernardia and

Isoberlinia (in Zambia) to be ‘fire tender’ (but not

‘fire intolerant’), West (1971) states that some of

these are ‘extremely fire tolerant’ but states that even

these will eventually be eliminated by sufficiently

regular, intense, late dry season burns every year. At

Matopos, Kennan (1971) found ‘that burning had

much the same effect on the sandveld trees as in the

thornveld . . . ’ so that mortality in the sandveld

woody plants was presumably also very low. The

relatively high fire tolerance of woody plants of the

present study and other savanna woody plant species

is not limited to Africa. For example, in a Texan

Pin us taeda — P. echinata forest with a head fire, it

was found (Stransky & Halls, 1979) that of the 10

most important understorey woody plant species,

three (for example, Cornus florida) had zero mor-

tality and all others except one had less than 32%

mortality.

A possible effective adaptation to fire in savannas

is the suffrutex or ‘underground tree’ form such as

that of Dichapetalum cymosum that West (1971) has

suggested as a possible evolutionary adaptation that

resulted in evading fire. The reaction of trees to fire

534 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA— OCHNA PULCHRA SAVANNA

and the possible fire adaptations have resulted in

many savannas being regarded as seriously limiting

the effectiveness of fire alone as a woody plant con-

trolling mechanism for management purposes, for

example, in eastern Cape Acacia karroo Savanna (Du

Toit, 1972a) and Brachystegia spiciformis — Julber-

nardia globiflora Savanna of the Zimbabwe highveld

(Barnes, 1965). In the latter area, burning intervals

longer than annually have been found to be ineffec-

tive in preventing coppice becoming increasingly

vigorous (Barnes, 1965). However, Van Wyk (1971)

stated that in Dichrostachys cinerea — Terminalia

sericea Savanna in the Transvaal lowveld, burning as

little as every three years resulted in plants seldom

escaping the regular damage to reach maturity. It

appears that fire as a tool for the management and

control of undesirable woody plants differs in its

effectiveness according to area and conditions of

application.

b) Effects of fire on plant canopies

Fire often greatly affects the canopy of woody

savanna plants, particularly those of smaller plants.

After a hot November burn in Burkea africana

Savanna in South West Africa, canopy (but not

plant) mortality of the plants less than 2 m tall was

greater than 75% for each of the 4 most common

woody species ( Burkea africana, Terminalia sericea,

Combretum psidioides and Ochna pulchra) with

most (90%) for Ochna pulchra (Rutherford, 1975).

In the Transvaal lowveld Van Wyk (1971) reported

that Terminalia sericea (and Dichrostachys cinerea)

plants of up to about 1 ,2 m were generally burnt back

to ground level although some Terminalia sericea

trees up to 3,7 m height had also been burned back.

In eastern Cape Acacia karroo, 95% of canopy mor-

tality after fire was limited to plants under 2,5 m tall

(Trollope, 1974). In Burkea africana Savanna

(regularly burned at the end of the dry season) at

Makambu, Kavango in South West Africa, Gelden-

huys (1977) found that for shrubs and trees (with

stems less than 5 cm DBH) canopy volume dropped

by two-thirds relative to that of the control. Mean

height of this plant group was 0,9 m compared to 2,2

m of those of the control, that is a decrease of about

60%. That fire generally reduces the canopy heights

of lower woody vegetation in many other regions is

supported by studies such as that in a Texan forest

fire (Stransky & Halls, 1979) where for the ten most

important understorey woody species height decreased

by 41% from 4,4 m to 2,6 m.

After a hot November burn in a Burkea africana

Savanna in South West Africa, canopy mortality of

Securidaca longipedunculata plants greater than 2 m

tall was the lowest of six species, namely zero per cent

(Rutherford, 1975).

In Acacia veld at Matopos, Kennan (1971) states

that in the case of larger trees, burning invariably

caused complete defoliation (‘if they were in leaf

when burning took place’) but seldom did more than

to kill branches up to a height of about 1,8 m.

However, in the eastern Cape, it was found that

canopy mortality of surviving individuals of Acacia

karroo was 79% (Trollope, 1974). Donaldson (1966)

found with burning Acacia mellifera subsp. detinens

in the northern Cape with fuel (grass, wood, dung or

sawdust) at the base of the plants, that generally

there were total ‘top-growth kills’.

Although it is clear from the evidence that fire may

be expected to reduce canopy leaf biomass, it appears

that the effect on radial stem growth may be different

since in the Kavango region of South West Africa it

was found that there was no significant effect of

annual fire treatments on stem basal area increment

over a nine year period in Burkea africana and other

species investigated (Geldenhuys, 1977) while also in

Nigerian savanna it was found that controlled burn-

ing early in the dry season would permit an increase

in established trees’ basal area (Kemp, 1963).

c) Fire and basal shoots

In many studies it has been found that the number

of stems are likely to increase with fire. For example,

in a Texan forest fire mean stem number of the ten

most important understorey woody plant species in-

creased from 1,5 to 2,1 with a maximum increase for

one species ( Ilex vomitoria) from 2,2 to 5,1 (Stransky

& Halls, 1979). A common phenomenon under total

fire protection in savanna, is for woody thickets to

tend to develop. However, Harrington (1974) points

out that in Uganda despite the densest appearance of

Acacia hockii in an unburnt treatment (relative to

that in several burning regimes) it had the lowest

number of stems per bush (and the lowest number of

bushes per hectare).

After a hot November fire in a Burkea africana

savanna in South West Africa, it was found that of

the three species Terminalia sericea, Burkea africana

and Ochna pulchra, the first mentioned had the

greatest percentage of plants with basal regeneration

shoots present (Rutherford, 1975). In scrub sand

veld savanna in Wankie National Park, Zimbabwe,

Rushworth (1975) found that whereas the mean

number of new coppice stems produced on Ter-

minalia sericea in an area unburned for at least eight

years was zero, those burned in just under three

months prior to measurement (an early October

burn) was 20,33 (versus 0,05 in an area with approx-

imately the same burning history as above but

without the October burn). Van Wyk (1971) has

reported vigorous sprouting of Terminalia sericea

after a burn in the Transveld lowveld. Donaldson

(1966) has also commented that the multistemmed T.

sericea of the Molopo area of the northern Cape have

possibly resulted from periodic grassfires in the past.

Interrelationships after fire

After a hot November burn in Burkea africana

Savanna in South West Africa, data showed that a

higher percentage of plants with canopies killed had

basal regeneration shoots present than those with

canopies that survived for all species investigated

(Rutherford, 1975). These included Burkea africana,

Terminalia sericea and Ochna pulchra. In eastern

Cape savanna, Trollope (1974) showed that after a

spring head fire, of those Acacia karroo trees that

survived and had formed basal regeneration shoots

after the fire, 86% had canopies killed leaving only

14% with live canopies. James & Smith (1977) state

that ‘extensive suckering does not usually occur after

low-intensity fires’ while Farmer’s (1962) work on

Populus tremuloides demonstrated that suckering

was related to the reduction of apical dominance by

damage to the above ground parts.

3. METHOD

a) The two burns

Three one hectare square blocks of Camp 2 of the

Nylsvley study area were burned separately before

the remainder of the camp on September 5, 1978. All

work on the woody plant species was done in two of

M. C. RUTHERFORD

535

these hectare blocks which were about 1 km apart.

Both areas were ignited at one side of the hectare

block with flame-throwers directed at the herbaceous

layer which allowed the fires to rapidly attain their

maximum intensities. Plot 1 was ignited at 1 9h0 1 and

plot 2 at 18h00. The mean windspeed at 2,0 m above

ground from 12h00 on September 5, 1978 to 07h00

September 6, 1978 was 1,8 ms-1 (Harrison, 1978).

Both fires were ignited as head fires (burning in the

same direction as the wind) although in plot 1 there

was some degree of backburning (burning in the

opposite direction to the wind) of some islands left

unburned after the main flame front had past. At

19h00 (corresponding to time of burn in plot 1)

screen climatic data gave: air temperature 17,2°C;

relative humidity 33%; vapour pressure 6,4 mb and

saturation vapour pressure deficit 13,1 mb. At 18h00

(corresponding to time of burn in plot 2) screen

climatic data gave: air temperature 19,0°C; relative

humidity 28%; vapour pressure 8,5 mb and satura-

tion vapour pressure deficit 13,3 mb (Harrison,

1978). Mean moisture content of plants (mainly

grasses) of the herbaceous layer one week prior to the

fires was 4,2% (Grunow and Grossman, 1978).

Estimated ground fuel loads (see Section 3c) showed

more frequent higher levels in plot 2 than in plot 1.

Fuel loads were sometimes very localized, for example,

typical individuals of the shrub species Grewia

flavescens (type 1, less than 2,5 m height, — Ruther-

ford, 1979) had 4 800 girr2 of thin finely divided

standing dead wood on the area they covered. Other

areas were sometimes virtually bare, that is, less than

10 g of dry herbaceous material for individual square

metres. Data on mean herbaceous layer dry mass per

unit ground area are not available for the two plots

but from many other clipping studies (Huntley &

Morris, 1978) on the study site, the mean mass of the

standing dead grass lay between 50 and 125 gm“2.

Ladder fuel in larger trees was rare, that is, there was

usually no continuous fuel path from the herbaceous

layer to the tree canopy.

Differences in the behaviour of the burns are given

in Table 1. The burn in plot 2 was more than five

times faster, fire temperatures were higher, flame

heights were greater and burning on an area basis

more complete than in plot 1. Although on the basis

of these data the fire in plot 2 might be regarded as

more intense than in plot 1, because of difficulties in

averaging the great differences in heat intensity at

different levels above ground, the fire in plot 1 is

referred to as the slower burn and that in plot 2 as the

faster burn. This designation may be appropriate

since speed of fire (whether ‘self’ generated or wind

induced) appears to be important to fire behaviour.

That head fires are faster than back fires is com-

monly observable. Trollope (1978) found greater

flame lengths in head fires and a positive correlation

between rate of spread and flame height and maxi-

mum temperatures at grass canopy height in head

fires. In laboratory simulated experiments, Gill

(1974) indicated an increasing flame height with in-

creasing wind speed (up to 0,48 msH). Wind affects

fire behaviour by increasing the flow of oxygen to the

fire and (in a head fire) wind bends the flames over

the unburned fuel and increases the flow of hot gases

from the combustion zone; both processes contribut-

ing to the pre-heating of the unburned fuels (Deem-

ing et a/., 1972) which is particularly important for

realizing the potential of water-conducting woody

material as fuel.

b) Experimental layout

The two plots selected, contained woody and her-

baceous elements that were floristically and structur-

ally typical of the Nylsvley study area vegetation.

Since the method of recording certain aspects of the

plant was different for tree individuals and multi-

stemmed shrubs (see next section) data were pro-

cessed separately for these two groups but so as to

prevent the same species from occurring in both

groups, the groups were defined as the tree species

group, that is, individuals that were trees or normally

have the potential to grow into tree-sized individuals;

and the multistemmed shrub species group whose

members seldom form tree-sized individuals on the

Nylsvley site. In the slower burn area, tree species

were (in order of abundance): Ochna pulchra,

Burkea africana, Terminalia sericea, Strychnos

pungens, Securidaca longipedunculata, Dombeya

rotundifolia and Dichrostachys cinerea. The

multistemmed shrub and suffrutex species were:

Grewia flavescens, Fadogia monticola, Lannea

discolor, Euclea natalensis and Vitex pooara. In the

faster burn area tree species were (in order of abun-

dance): Ochna pulchra, Burkea africana, Terminalia

sericea, Vitex rehmannii, Dombeya rotundifolia,

Strychnos pungens, Securidaca longipedunculata,

Combretum molle, Strychnos cocculoides, Acacia

caffra, Ximenia caffra, Sclerocarya caffra and Pap-

pea capensis. Multistemmed shrub and suffrutex

species were: Grewia flavescens, Dichapetalum

cymosum, Euclea natalensis, Fadogia monticola and

TABLE 1. — Differences in fire behaviour characteristics between two plots

(n = 412)

* B.J. Huntley (pers. comm.)

+ Harrison (1978)

** Plates described section 3b

*** Estimates described section 3c

536 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICAN A— OCHNA PULCHRA SAVANNA

Asparagus suaveolens. Typical structural features of

the vegetation can be seen in the photographs (Figs.

1-6). At the time of the fire many of the woody

plants, particularly those of Ochna pulchra, were

starting to unfurl new leaves. The most advanced

leaves on some individuals of O. pulchra were about

2 weeks old. All the more common species (except

Strychnos pungens) had dropped their old leaves

before the fire so that almost all of the woody plant

leaves formed in the previous growing season were

already added to the fuel load on the ground prior to

the fire. Also present at the time of the fire were a

few woody plant seedlings, mainly those of Burkea

africana. The dominant grass in both burn plots was

Eragrostis pallens.

In each plot a subplot of 30 x 50 m was placed

centrally and demarcated. In the subplots all sizes of

standing individuals of woody plant species were

tagged with numbered aluminium plates (either on 20

cm high stakes at the base of smaller individuals (Fig.

1) or on to the trunk of large individuals) in one half

of the area. All individuals equal to or larger than 2

m height were tagged in the other half to increase the

area. The individuals of this new control set were

geographically close to the burn treatments, probably

had a very similar treatment history in the past and

were also measured in exactly the same way and time

as the burned individuals. It transpired that the

effects of fire were so profound that differences

between treatment and control were usually so great

that statistical analyses of most differences, especially

those concerning basal regeneration, were super-

fluous. The control in many respects only served to

confirm the obvious. References to unburned control

data are thus kept to a minimum with more attention

being given to the differences between the two burns.

c) Measurements

Recordings of plants were made, (i) shortly before

the burns, (ii) just after the burns, (iii) at monthly

intervals after the burns and (iv) after completion of

one season’s growth but before commencement of

leaf fall. Most detailed measurements were made in

periods (i) and (iv).

One to two weeks before the burns, each woody

Fig. 1 . — A view on the 6th Septem-

ber after the fire of part of

one of the burned areas

revealing aluminium tags

marking small woody plants

normally largely concealed in

the herbaceous layer.

sample size of the larger individuals. All standing

dead individuals equal to or larger than 2 m height in

the remainder of each one hectare plot were also tag-

ged and numbered. In the slower burn plot there were

altogether 607 tagged individuals in the subplot and

two additional dead individuals in the rest of the hec-

tare plot. In the faster burn plot 425 individuals were

tagged in the subplot and an additional 12 dead

individuals in the remainder of the plot. Altogether

1 046 individuals were thus tagged for recording the

effects of fire. The day following the burns it was

found that in the slower burn area 37% and in the

faster burn area 3% of tagged individuals had

escaped the fire altogether on large unburned islands

with each of these individuals having no vegetation

burned on their canopy ground projection area.

Since such a relatively high proportion of individuals

was altogether untouched by fire, it was decided to

use these unburned individuals (from both plots) as a

control rather than the originally envisaged tagged

individuals already being monitored as part of a

separate programme in Camp 3 of the Nylsvley study

plant individual was allocated a numbered alumi-

nium tag, and the tag position was also recorded for

relocating the plant later. The species and live or

dead state of the plant was recorded. Also measured

were height of plant above ground level, number of

live basal shoots (these constituted the whole in-

dividual in small non-canopied individuals) and

number of dead basal shoots. Estimates were made

of the proportion of canopy volume that was dead

using a five point scale (0-9,9; 10-34,9; 35-64,9;

65-89,9; 90-100%) based on zones of dead twigs.

Also estimated was the relative amount and composi-

tion of ground layer fuel load under the individual on

a three point scale and with 3 type classes, namely,

woody plant leaf litter, dead standing grass and

pieces of wood material (e.g. Fig. 2). This ground

layer fuel classification only applied well to in-

dividuals up to about 2Vi m height. The area under

large individuals was relatively large with often great

differences in ground fuel load under the same tree.

Photographs of various parts of the vegetation were

taken from reference positions.

M. C. RUTHERFORD

537

Fig. 2. — A pair of Burkea afri-

cana stems to the left of

centre; a stem of a Securidaca

longipedunculata tree on the

right; Ochna pulchra in-

dividuals immediately behind

and to both sides of the

Securidaca stem and dead

fallen branches of Burkea

africana in the right fore-

ground, showing: a, on the

5th September before the burn

the relatively high fuel load;

b, on 6th September after the

burn, the ash production from

burned, fallen branches also

showing the imcomplete burn

of other branches; c, on

6th February at the end of the

growth season and the basal

regeneration from the Burkea

stems and the appearance of

the new grass cover.

538 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA—OCHNA PULCHRA SAVANNA

One and two days after the burns, all labels were

checked for legibility and position and a few replaced

where necessary. Whether the plant was burned or

totally untouched by fire was recorded as was an

estimate of degree of burn namely: burned, for a

broad class including plants that fire had touched at

least at some point of where ground fire occured

under at least part of the canopy; severely burned for

plants where most leaves or other parts of the plant

were at least severely scorched; and completely burn-

ed where the plant’s aboveground parts were com-

pletely burned away or at most only a very short stub

(1 or 2 cm high) remained. Photographs were taken

from the reference positions.

At monthly intervals for six months following the

burns, the photographs of selected components of

vegetation were taken from the reference positions.

Checks of animal (almost entirely insect) browsing of

leaves were made to confirm that this remained at a

low level (less than 5%) so that leaf changes could be

attributed to fire effects.

Six months after the burns (March), the following

were recorded for all tagged individuals: species (a

recording check); live or dead state of plant; canopy

live or dead; height of top parts of canopies;

diameter of stem at 20 cm above ground level (for all

stems >1,0 cm diameter); number of new basal

shoots; number of old basal shoots still present;

number of basal shoots killed by the fire and still pre-

sent; number of old dead basal shoots that survived

the fire; estimation of proportion of original canopy

volume dead (this was difficult to apply in species

Fig 3. — A Grewia flavescens shrub (with pole): a, on 5th September, before the burn; b, on 5th October, one month after

the burn, showing basal regeneration commencing, after virtual total removal of aboveground material by the burn; c, on

7th November, two months after the burn showing basal regeneration at an intermediate stage; d, on 18th December,

more than three months after the burn showing basal regeneration virtually complete; e, on 3rd January; f, on 6th

February demonstrating that virtually all basal regeneration is completed by the December date. The shrubs on the left of

the marked Grewia individual are Euclea natulensis which regenerated more strongly than did the Grewia individuals.

M. C. RUTHERFORD

539

such as Grewia flavescens, where the original outline

of canopies was usually altogether lost through fire

-Fig. 3); biomass of basal twig regeneration (clipped

at ground level); biomass of basal leaf regeneration;

biomass of all leaves in canopy (for individuals <5 m

height). All biomass data were obtained oven dry at

85°C and total sampling was used, that is, no sub-

sampling was employed.

The diameters of stems were taken for use in

already established formulae that predict canopy leaf

mass (Rutherford, 1979). These regression formulae

were first tested by destructive sampling for each of

the more common species in the unburnt control

populations since the formulae were derived for

populations several kilometres distant and three

years previous to the time of fire. If, as was found in

the Ochna pulchra population, there was very good

agreement between predicted canopy leaf mass and

actual destructively sampled canopy leaf mass of the

unburnt population, the regression formula was

applied to the burnt population of the appropriate

species (using original plant height if height was

reduced by fire) to obtain expected leaf mass had the

population not been burned. Therefore, on condition

that the regression formulae still proved suitable, this

procedure provided a more sensitive measure of the

degree of canopy leaf biomass reduction by fire than

that provided by using treatments and controls which

sometimes had greatly differing distributions of

heights within each height class. Canopy leaf biomass

reduction per height class was thus calculated in these

cases by taking the predicted unburned canopy leaf

biomass of the burned individuals and subtracting

the actual canopy leaf biomass obtained by direct

harvest from the burned individuals.

It should be noted that for practical reasons a

small proportion of the individuals included in the

sample for numerical counts of stems, mortality and

so on were not harvested for biomass. Therefore

those numerical non-biomass data concerning differ-

ences before and after the fires are not necessarily

precisely interrelateable with the biomass data set

since there are possible differences in plant size

distribution in the whole sample set and in the

biomass data subset.

Only some data are tabulated since tabulation of

all data for all species, all plant size classes and all the

types of possible fire effects, results in many large

tables with very many empty cells. This difficulty is

inherent in studies such as the present, and arises not

only from variation in the natural vegetation com-

position and structure but also from the inapplica-

bility of some measures of the effects of fire to cer-

tain plant growth forms. The very unbalanced total

data set resulted in no full analysis of variance being

attempted. Instead, for categories where sufficient

data existed, all data from the category were used in

standard statistical tests of significance between

means, often using the non-parametric Wilcoxon test

where appropriate. Most data are indicated graphic-

ally where use was made of data grouped into classes

in the presentation of relations.

4. RESULTS

a) Mortality /survival of plants after the burns

Mortality was defined as the proportion of the

number of plant individuals that were alive before

the fire but dead (above ground) six months after the

fire (‘Root kill’ of Niering et al., 1970). It is possible

that mortality includes individuals that survived or

regenerated after fire but died from other causes

within the six month period following the fire.

That mortality of unburned individuals would be

close to zero was confirmed by the data for the con-

trol that showed a mortality of 1 % of the tree species

group’s individuals and 0% of the multistemmed

shrub species group’s individuals. In the slower burn

area mortality of the tree species individuals was 5%

which was significantly greater (P=0,005) than that

of the unburned control. For the shrub species in-

dividuals mortality was 0%. In the faster burn area

mortality of the tree species individuals was 2%

which was not significantly different from that of the

control. The mortality of the multistemmed shrub

species individuals was significantly greater than that

of the unburned control. This mortality is, however,

subject to further interpretation since one species was

involved, namely, Dichapetalum cymosum, where a

mortality of 64% for its individuals up to 25 cm tall

was indicated. Re-examination in August of most of

these individuals recorded as dead in March showed

that although most had no aboveground parts visible

there were live belowground parts that were in the

process of initiating new shoot growth. In these

cases, therefore, fire possibly only delayed new

growth by one growth season. However, given the ex-

tensive underground branching and interconnections

between ‘individuals’ of this geoxylic suffruticose

species, new regenerative growth possibly occured

after the fire but not in the immediate vicinity of the

labelled ‘individual’. Rushworth (1978) also found

that woody suffrutices such as D. cymosum did, con-

trary to other woody plants, not produce additional

stems per plant unit after fire in Wankie National

Park, Zimbabwe. That the high mortality of ‘in-

dividuals’ of D. cymosum is merely an apparent mor-

tality, means that, in fact, the mortality values for

the multistemmed shrub species individuals was very

close to zero.

Despite individuals of Grewia flavescens having

the highest fuel loads within the plant and that many

plants were completely consumed by the fire (Fig. 3),

there was no mortality within these populations in

either burn. Mortality of Ochna pulchra plants in the

slow burn (5%) was significantly greater (P=0,005)

than that in the faster burn (1%). However, when

grouped into plant height classes (Table 2) there was

no significant difference in mortality for those plants

taller than 0,25 m but only for the group below 0,25

m height. In Burkea africana, mortality was limited

to small plants under 10 cm tall, that is, 54% for

those of both burns. Most of these plants killed were

seedlings.

b) Effects of fire on plant canopies

Plants were defined as canopied where leaves were

carried on stems more than one year old. Non-

canopied individuals were thus made up of only

(young) basal shoots whereas canopied individuals

had an older main stem bearing the canopy leaves

with or without basal shoots present. Because of

some difficulties in ageing basal shoots before the

burns a few non-canopied individuals possibly had

basal shoots slightly older than one year but such

basal shoots were morphologically similar to one-

year old shoots. The effects of the burns on canopies

are of course limited to the canopied plants. The

effects of fire on canopies can be expressed in various

ways, depending upon the canopy attribute con-

sidered. Canopy mortality occurs where the whole

canopy dies, but where the plant still survives in the

form of basal regeneration shoots (‘Stem kill’ of

Niering et al., 1970) (Fig. 4). The occurrence of

abnormal leaf growth refers to the amount of leaves

540 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA— OCHNA PULCHRA SAVANNA

TABLE 2. — Mortality of woody plant individuals following fire according to plant height class

that have grown in a convoluted manner and are

usually produced not from terminal twigs but from

thicker, older wood parts (Fig. 5). Also in terms of

shoot extension in Ugandan savanna, Harrington

(1974) refers to the tops of taller burned bushes

behaving similarly to bushes in an unburnt controlled

treatment. Changes in the height and canopy volume

of plants refers to measurements as described in the

section on methods. Another attribute used is the

change in the total amount of leaf biomass in the

canopy.

i) Canopy Mortality

Only 0,7% of the unburned control tree species

plants’ canopies died. In the faster burn, 43,2%

canopies were killed of which 92,1% were under 2 m

tall with none taller than 5 m. In the faster burn,

these plants up to 1 ,5 m tall had 9,2 canopies killed to

each one that survived. For those over 1,5 m tall

there were only 0,2 canopies killed for each surviving

canopy. In the slower burn, 23,5% canopies were

killed of which 90,1% were under 2,5 m tall and also

with none taller than 5 m. In the slower burn, those

plants up to 1 m tall had 1,6 canopies killed for each

one that survived. For plants between 1 and 1,5 m tall

the ratio was about 1 : 1 while for those more than

1 ,5 m tall there were also only 0,2 canopies killed for

each surviving canopy. Relative to the control tree

species plants, therefore, canopy mortality in both

burns was highly significant particularly for in-

dividuals less than 2 or 2,5 m tall.

In the Ochna pulchra population, canopy mortal-

ity was 32,0% in the slower burn and 44,1% in the

faster burn. Only in one species, namely Vitex

rehmannii was it clear that tall individuals had a

relatively high canopy mortality, that is, 64,3% for

individuals between 2,5 and 5,0 m tall. Despite the

sensitivity of canopies of V. rehmannii to fire, no

plants of this species died after fire. The hollow main

stem of one tree individual, Securidaca longipedun-

culata, was observed to burn vigorously for several

hours after the main flame front had passed, but at

the end of the growth season the canopy showed no

obvious effects of the fire (Fig. 6).

Fig. 4. — A group of Ochna

pulchra individuals with most

canopies killed by the burn

but with prolific basal regen-

eration.

M. C. RUTHERFORD

541

Fig. 5. — a. An Ochna pulchra indi-

vidual showing areas of mis-

sing foliage in lower parts of

former canopy and the pre-

dominantly normally formed

leaves at the extremities of the

highest branches; b, an Ochna

pulchra individual showing

reduction of foliage in former

canopy areas and the pre-

dominantly abnormally form-

ed leaves clustered and pro-

duced directly from areas of

thicker branches and stem and

not at the branch extremities.

Fig 6. — A Securidaca longi-

pedunculata tree: a, one day

after the burn (6th September)

with a hollow stem that was

observed to burn vigorously

for several hours after the

burn; b, five months after the

burn (6th February) indicating

the canopy that shows no

obvious reduction and

appears as unaffected as that

of a normal unburned in-

dividual of this species.

ii) Occurrence of abnormal canopy leaf growth

Apart from observing that the incidence of abnor-

mal canopy leaf growth was generally higher in the

faster burn that in the slower burn, the mere presence

of abnormal leaf growth was found to be less

informative than the actual value of abnormal leaf

biomass compared to that of normal leaf biomass.

Since a clear qualitative recognition of this abnor-

mality is required for expressing such ratios, con-

sideration was limited to the Ochna pulchra popula-

tion where such distinction was most reliable (Fig. 5).

The ratio of normal to abnormal canopy leaf bio-

mass was found to increase exponentially with plant

height (Fig. 7) and is discussed further in Section 4e.

iii) Changes in canopy height and canopy volume

For the unburned tree species control plants, mean

change in height was + 4% with the greatest relative

reduction in any height class being -2%. Although

in unburned plants height appeared to be unimpor-

tant in affecting tree height changes, in both burns

canopied plants of the lowest height classes had the

greatest relative reduction in height (about 100%)

with tallest plants having a zero reduction in height.

The relative decrease in plant height after the faster

burn was greater for each height class than that for

the slower burn, for example in Ochna pulchra (Fig.

8a). In terms of the proportion of plants that

decrease in canopy height it was found, for tree

542 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA — OCHNA PULCHRA SAVANNA

Fig. 7. — Relationship between mean plant height and the ratio of

the mass of normally to abnormally formed canopy leaves for

Ochna pulchra in both burns.

species plants, that 5% of unburnt plants, 28% of the

plants of the slower burn and 50% of the plants of

the faster burn decreased in height. There was

therefore a similar pattern with a greater proportion

of individuals decreasing in height in the faster burn

area (Fig. 8b). A very similar pattern was obtained in

terms of reduction in estimated canopy volume (Fig.

8c). In the multistemmed shrub species group,

changes in height were more variable and less strong-

ly correlated with height.

iv) Changes in canopy leaf biomass

Effects of the burns on the canopy leaf biomass

could be determined as described earlier on the basis

of application of allometric biomass prediction for-

mulae once appropriate tests of the validity of

application had been made. It was found that the

allometric formulae, applied to unburned plants

(stem diameter >1 cm at 20 cm height) of Ochna

pulchra, overestimated the actual harvested canopy

leaf biomass value by only 0,1%. In Terminalia

sericea it was found that canopy leaf mass was

underestimated but remained within 20% of the

actual harvested amount. In Burkea africana it was

found that canopy leaf mass was overestimated by

more than 20% (in the relatively small tested sample

population of the control) and the biomass predic-

tion formulae were thus not applied to the burned B.

africana populations. Since the allometric formulae

cannot be applied to the smallest individuals, canopy

leaf biomass data from unburned control plants were

utilized for the lowest size classes.

In the slower burn area, Ochna pulchra plants

under 1 m tall decreased in canopy leaf biomass by

90%, but this reduction became less marked with

taller plants so that a decrease of only 26% was

found for plants between 2,5 and 5 m tall. In the

faster burn area the plants under 1 m tall decreased in

Mean plant height (grouped data) (cm)

Fig. 8. — Relationships between mean plant height and: a, relative

change in plant height; b, proportion of individuals that

decrease in height (graph inverted for comparison); c, relative

change in canopy volume for Ochna pulchra in (1) the slower

burn and (2) the faster burn.

canopy leaf biomass by 92%, and this also became

less marked with taller plants but with a 50%

decrease for plants between 2,5 and 5 m tall.

In Terminalia sericea of both burns, trees >2,7 m

tall had a canopy leaf biomass of 79% of that

predicted, but for the smaller trees only 6% of the

predicted value. This contrasts with the 83% of that

predicted in the unburned control plants. Therefore,

allowing for the shift in the prediction equation

(given by the control), a reduction of at least 90% in

canopy leaf mass after fire occured for the T. sericea

plants less than or equal to 2,6 m tall. In Burkea

africana of both burns, stem basal area was used as a

measure of plant size to relate to the harvested

canopy leaf biomass. For trees up to a cross-sectional

stem basal area (at 20 cm height) approaching 50 cm2

canopy leaf biomass was less in burned plants than

unburned plants, but for plants with a basal area of

about 50 cm2 or more this difference no longer held

(Fig. 9). For taller plants (approximately greater than

2,5 m) Ochna pulchra had a greater canopy leaf

biomass reduction than Terminalia sericea and

Burkea africana. For smaller plants the differences

were less pronounced between these three main tree

species.

c) Effect of fire on basal shoot numbers

Most basal shoots that were live before the burns

were killed by the fire. Although 14% of the basal

shoots of control tree species plants died without fire,

91% were killed in the slower burn and 100% in the

faster burn. The killing of live basal shoots by fire

was independent of plant height and species. In the

multistemmed shrub species plants killing of live

basal shoots was virtually 100% in both burns.

M. C. RUTHERFORD

543

In stem basal area (cm2)

Fig 9. — Relationship between stem basal area and canopy leaf

mass for: a, unburned individuals; b, burned individuals in

both burns for Burkea africana.

The formation of new post-fire basal shoots was a

common phenomenon in burned individuals (Figs. 10

& 11) and these were virtually completely grown by

December, that is, about three months after the burn

(Fig. 3). Whereas a mean of 0,09 new basal shoots

were formed per unburned control plant, the corres-

ponding values were 2,42 for the slower burn and

4,55 for the faster burn. In the unburned control

plants, however, new basal shoots were not formed

from individuals more than 1 m tall, whereas in

burned individuals these were formed irrespective of

plant height, but with a tendency for fewer basal

shoots to be produced per tall individual. In the

multistemmed shrub species plants approximately

5% of the unburned individuals produced new basal

shoots, whereas corresponding values were 93% in

the slower burn area and 100% in the faster burn

area.

The possible relationship between the number of

live basal shoots before the fire and those formed

after the fire was investigated to determine to what

Fig. 10. — Basal regeneration

shoots of Ochna pulchra on

18th December after above-

ground shoots had been either

killed or removed by the burn.

Fig. 11. — Basal regeneration

shoots of Grewia flavescens

on 3rd January after all

aboveground shoots had been

killed and almost all removed

by the burn.

544 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICAN A— OCHNA PULCHRA SAVANNA

degree live basal shoots were replaced after fire. Data

were expressed as the ratio of the number of new

basal shoots to the number of old basal shoots killed

by the fire. In Ochna pulchra basal shoots killed were

replaced by almost twice (1,88) the number of new

basal shoots. In Burkea africana the ratio was only

1,30 which was, however, still a net increase of basal

shoots per plant.

In Ochna pulchra, although the number of new

basal shoots tended to decrease with increasing plant

height class, it was found that replacement of basal

shoots increased with increasing plant height. No

such trend was discernible in Burkea africana. The

basal shoot replacement relation is given in Fig. 12

for the mean of both burns. The degree of ability of

these plants to produce new basal shoots after fire

appears to depend on not only the size of the plant,

but also on the number of basal shoots prior to the

fire.

Fig 12. — Relationship between mean plant height and the ratio

of number of post burn basal regeneration shoots to number

of pre-burn live basal shoots (basal shoot replacement ratio)

in Ochna pulchra.

d) Effect of fire on production of new basal regene-

ration shoot biomass

Basal shoot production (over the 6 month period

following the burns) averaged per individual from

both burns was 9,59 g for Ochna pulchra, 14,52 g for

Burkea africana and 1,09 g for Terminalia sericea

(Table 3). In each of these populations, basal

regeneration shoot mass of unburned plants was

usually considerably less than 10% of basal mass in

burned plants. This was also valid for the rarer

species, but in some there were very small sized

samples for the control, for example, there was no

Dichapetalum cymosum in the control area. From

Table 3 it can be seen that species with low values

(<10 g) for basal regeneration shoot mass per in-

dividual were Dichapetalum cymosum, Terminalia

sericea, Strychnos pungens and Ochna pulchra. The

species with highest values (>50 g) were Grewia

flavescens, Dombeya rotundifolia and Vitex rehman-

nii with Euclea natalensis having the highest value of

all (285,62 g). Those species with greatest basal

regeneration shoot mass tended to be non-suffrutex

shrub forms whereas tree growth forms usually had

relatively low basal regeneration shoot mass.

In terms of mean biomass of basal shoots per in-

dividual shoot (where these were clearly distinguish-

able), Lannea discolor and Burkea africana had

highest values while lowest values were found in

Dichapetalum cymosum and Terminalia sericea.

Species may be divided into three groups according

to their ratio of basal leaf mass to basal twig mass.

Those with a ratio of less than 1, that is, there was

more twig mass than leaf mass, included Securidaca

longipedunculata and Terminalia sericea. Those with

up to twice the amount of leaf mass to twig mass

included the largest number of species. Those with

more than twice as much leaf mass as twig mass

included the two most important geoxylic suffrutex

species Dichapetalum cymosum and Faclogia mon-

ticola.

Differences in mean basal shoot regeneration mass

between different species were tested (for all species

with more than 3 individuals) often using the non-

parametric Wilcoxon test where appropriate (Table

4). At a level of significance of P = 0,05 more than

half the combinations were significantly different.

Since the reaction of Grewia flavescens was so differ-

ent in two burns, these were treated separately.

In Ochna pulchra in the slower burn, basal shoot

regeneration per individual (4,86 g) was significantly

lower than that (12,28 g) in the faster burn. In

Grewia flavescens the corresponding values were

65,00 g and 160,02 g. This greater production of

basal regeneration shoot mass in the faster burn area

was also confirmed in Burkea africana for each

height class of individuals. In Grewia flavescens, the

live basal shoot mass per unburned individual was

significantly greater (P = 0,005) than the post fire

regeneration basal shoot mass in the slower burn

area, but there was no significant difference to that

produced in the faster burn.

The possible relationship between degree of burn

recorded in both burns and basal regeneration mass

of the woody plants was shown by 63,73 g for those

recorded as burned, 115,12 g for those recorded as

severely burned and 107,68 g for those completely

burned.

The species with the greatest range in basal

regeneration shoot mass was Euclea natalensis (1,74

to 1143,37 g). When the population was divided into

three equal size classes that is up to 0,5 m tall,

0, 5-1,0 m and 1,0-1, 5 m tall, the respective mean

masses were 40,44 g, 248,38 g and 768,72 g and each

value was significantly different from the other.

There was thus a clearly greater basal shoot regenera-

tion mass with increased plant size for this shrub

species.

It was noted during observations that the individual

leaf size, especially of Ochna pulchra, was markedly

larger in basal regeneration shoots than in the canopy.

e) Relationships between plant height and woody

plant biomass components after fire

Earlier reference has been made to effects of the

burns on plant height and canopy volume in Ochna

pulchra and to effects on canopy leaf mass in Burkea

M. C. RUTHERFORD 545

TABLE 3. Basal regeneration shoot mass data tor the more common woody plant species

TABLE 4. — Statistical significance of differences in basal regeneration shoot biomass of different woody plant species after fire

Acacia caffra

africana according to stem basal area. The present

section aims to determine to what extent plant height

governs changes in various plant biomass compo-

nents after fire.

When basal leaf mass is compared to canopy leaf

mass in Ochna pulchra it is found that the ratio of

canopy leaf mass to basal leaf mass increases non-

linearly with increasing tree height (Fig. 13). Plants

under 1 m tall have more than half their leaf mass in

the form of basal leaves; plants around 1 m tall have

basal and canopy leaf mass about equal, while in

plants taller than 1 m canopy leaf mass becomes

rapidly much greater than basal leaf mass.

In Ochna pulchra it was found for both burns

together that the ratio of normally formed to abnor-

mally formed canopy leaf mass increased exponen-

tially with plant height (Fig. 7). For plants about 1 m

tall abnormally formed leaf mass more or less equalled

normally formed leaves. For plants above 1 m tall,

normally produced leaf mass quickly became many

times that of the abnormally produced leaf mass. In

the faster burn plants, abnormally formed leaves

546 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA A ERICA NA — OCHNA PULCHRA SAVANNA

Fig. 13. — Relationship between mean plant height and In the

ratio of canopy leaf mass to basal regeneration leaf mass for

Ochna pulchra in both burns.

contributed relatively more than in the slower burn.

In the faster burn, extreme values for the ratios were

also obtained, that is zero for the lowest height class

and oo for the highest height class.

Differences in leaf mass between burned and un-

burned Ochna pulchra for both burns for each height

class are indicated in Fig. 14. It can be seen that an

increasing proportion of canopy leaf mass was lost

after fire with decreasing plant height to a point

where plants were too small to be canopied. The rela-

tionship approximates an exponential decay curve

where the percentage change in canopy leaf mass

decreases exponentially with lower plant height.

However, in contrast to this monotonic relationship,

an increasing proportion of total leaf mass is lost

with decreasing height only until a height of 1—1,5 m

is reached, thereupon a maximum reduction in total

leaf mass having been attained, less leaf mass is lost

until for smallest plants the mean total leaf mass

becomes a net increase. It is also clear that it is those

plants roughly 0,75—2,5 m tall, that on average have

more than 50% of total leaf mass reduced by fire

whereas for the smallest and largest individuals this

was not so.

Fig. 14. — Relationship between mean plant height and: a, change

in canopy leaf mass; b, change in total plant leaf mass for

Ochna pulchra in both burns. The broken line indicates the

points at which leaf mass is reduced by more than half of the

total amount.

It was found tht in Ochna pulchra, there was an

increase in basal regeneration shoot mass with height

until a maximum was reached for heights about 1 to

1,5 m after which there was a decline in basal

regeneration shoot mass with plant height (Fig. 15).

Fig. 15. — Relationship between mean plant height and basal

regeneration mass showing the relative contributions of leaf

and twig mass for Ochna pulchra in both burns (shoot mass

= leaf mass + twig mass).

A relationship between plant height and the ratio

of leaf mass to twig mass of the basal regeneration

shoots was found for both Ochna pulchra and

Burkea africana (Fig. 16). This inverse relationship

M. C. RUTHERFORD

547

assumed the form of a logarithmically decreasing

ratio with increasing tree height. There was very little

distinction in the relationship between the two burns

in Ochna pulchra. The ratio in Burkea africana was

generally much higher than that for Ochna pulchra

for each height class. This inverse relationship is very

different to that in canopies of unburned Ochna

pulchra where there is no such inverse relationship

with plant height.

Fig 16. — Relationship between plant height and the ratio of leaf

to twig mass in basal regeneration shoots of: a, Ochna

pulchra; b, Burkea africana in both burns.

Fig 17. — Relationships between mean plant height and current

season’s shoot mass showing the relative contribution of leaf

and twig mass for unburnt Ochna pulchra.

Combining the above inverse relationship equation

for Ochna pulchra (Fig. 16) with the parabolic

dependence of basal shoot mass on plant height, the

result is given in Fig. 15. This may be contrasted with

corresponding data in normal canopies (Fig. 17),

Mean reduction of canopy leaf mass

(grouped data) (°„)

(grouped data) ( % )

Fig 18. — Histogram of reduction in Ochna pulchra canopy leaf

mass against basal regeneration shoot mass per plant in: a, the

slower burn; b, the faster burn, n equals the number of in-

dividuals in the sample.

obtained from randomly selected Nylsvley indivi-

duals and from published allometric biomass rela-

tions (Rutherford, 1979). It is clear that (i) the form

of the relations; (ii) the relative proportion of compo-

nents and (iii) changes in this proportion with plant

height all differ radically between O. pulchra basal

regeneration shoots after fire and the normal canopy

shoots.

f) Other pyro-allometric biomass relations

In previous sections, much evidence has been given

to show that the more damage the canopy of, for ex-

548 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA — OCHNA PULCHRA SAVANNA

Fig. 19. — Relationship of reduction in Ochna pulchra canopy

leaf mass and basal regeneration shoot mass per plant in both

burns.

ample, Ochna pulchra, is subjected to, the greater the

amount of basal regeneration. Now using the relative

reduction in leaf biomass of canopies as a measure of

canopy damage this is related to basal shoot regen-

eration mass to provide a more precise relationship.

In deriving these relationships, individuals that were

completely killed by the fire were omitted as were the

very few individuals that actually increased slightly

in canopy leaf mass after fire. Grouping all plants (of

all heights) of O. pulchra into five equal classes of

relative reduction in canopy leaf mass, generally

increased basal shoot mass with classes of increasing

reduction in canopy leaf mass (Fig. 18) was apparent

in both burns. Expressed as a relation (Fig. 19) it was

found that there was an exponential increase in basal

shoot production with an increasing proportion of

canopy leaf mass lost through fire (until a few in-

dividuals pass beyond a certain threshold and die).

Particularly for the uppermost canopy leaf mass

reduction class, the plants of the faster burn area

produced a greater mass of basal regeneration shoots

than those of the slower burn.

Although basal shoot regeneration mass was found

to increase with increased damage to canopies, the

concept should possibly not be extended to very small

(< 0,25m height) uncanopied plants of O. pulchra.

Using the recorded degree of burn as a measure of

damage to these plants, it was found that plants

recorded as burned had a mean shoot mass of 10,35

g, whereas those recorded as severely or completely

burned had a mass of only 5,37 g, which is signifi-

cantly (P = 0,005) lower.

Number of basal regeneration shoots per plant

Fig. 20. — Distribution of plants of Ochna pulchra relative to

both number of basai regeneration shoots and mean in-

dividual basal shoot mass for: a, the slower burn; b, the faster

burn, n is number of plants.

Another feature that emerged for O. pulchra of

both burns was not only an inverse but an approx-

imately rectangular hyperbolic relation between

number of basal regeneration shoots and the mean

mass per basal shoot. In the slower burn area the

data set was more tightly grouped (Fig. 20a) in that

there was no occurrence of individuals with both

more than 6 basal shoots and a mean shoot mass of

more than 2 g. In the faster burn area (Fig. 20b),

although the distribution was still hyperbolic, the

data were more variable. In both these distributions,

plant height appeared not to be important.

g) Effects of fire on standing dead individuals

Of the 45 dead standing woody plants labelled

before the fire (including controls), most belonged to

Ochna pulchra, Burkea africana and Terminalia

M. C. RUTHERFORD

549

sericea. No significant differences in the fire effects

between the two burns were found for these dead

plants. The effects of fire on the height of dead

standing individuals was found to tend strongly to

one of two extremes, namely no reduction in height

(69%) or maximum reduction to ground level (24%).

All those individuals felled by the fire were under 1 ,5

m tall, that is, no individuals over 1 ,5 m tall were fell-

ed by the fire despite the observed active burning at

the base of two of these a day after the fire. Eighty-

six percent of those felled by fire were totally con-

sumed. Although 24% of burned individuals were

felled, this was not much higher than the 19% of in-

dividuals that fell over in the absence of fire in the six

month period. There was no apparent selectivity of

felling of individuals by fire or other agents accor-

ding to plant species. In the main live sample in-

dividuals, basal shoots that were dead before the

burn were almost always totally consumed by fire.

Woody plant leaf litter on the ground, although in-

flammable, sometimes did not ignite fully, for exam-

ple, even in thick layers in Ochna pulchra patches,

burning was often limited to very superficial surface

layers of leaves (Fig. 21).

5. DISCUSSION AND CONCLUSIONS

In keeping with common veld management prac-

tice, the experimental burns were carried out in early

spring when the various deciduous plant populations

were at different stages of bud break development.

Although stage of bud break development at which

the plant is burned may be considered possibly

critical to the subsequent fire effects measured, data

of Hopkins (1963) for work on selected tree species in

south-western Nigerian savanna, appear to indicate

that burning in the period of a few weeks before and

after normal bud break, does not markedly change

the period needed for recovery to start.

a) Mortality /survival of plants after the burns

That mortality was significant only in the slower

burn and was also largely limited to very low tree

species plants, indicates that the effects of a slower

head fire may differ to those of a faster head fire in a

way similar to the different effects between a back

burn and a head burn at ground level in grassland

(Trollope, 1978). The apparently anomalous lower

mortality with greater damage to the canopy in the

faster burn, relates to the compensatory basal

regeneration effect described. Therefore, even in a

faster fire, with a probably more intense overall heat

than in the slower burn, the mortality effects may

relate to more intense heat nearer the ground and not

to heat loads at canopy level. Even where fire was

reported with wind twice the speed of that in the pre-

sent study, mortality remained low (Stransky &

Halls, 1979).

That most species in the present study had mortal-

ity after fire varying from 0 to 5% shows a fire

tolerance that appears common in many other woody

plants in Africa. The limitation of mortality in

Burkea africana to very low plants, mainly seedlings,

agrees well with Jackson’s (1974) classification of it

as a pyrophytic species. The quoted fire ‘tolerance’

or ‘resistance’ (Trapnell, 1959; Van Wyk, 1971;

Lawton, 1978) for Dichrostachys cinerea, Dombeya

rotundifolia, Lannea discolor, Strychnos pungens

and Terminalia sericea is confirmed by the less than

5% mortality for each in the present study. Ochna

pulchra was also shown in this study to be generally

fire tolerant in agreement with indications of Lawton

(1978) and Rushworth (1978) but not with a ‘fire-

sensitive’ description (Geldenhuys, 1977).

The very low mortality rate after fire of multi-

stemmed shrubs (omitting Dichapetalum cymosum),

which includes the 0% mortality of Grewia flaves-

cens despite its relatively high fuel loads, is paralleled

in eastern Cape Acacia karroo Savanna where only

the multistemmed plants of Rhus lucida recovered

fully after a fire (Trollope, 1974).

b) Effects of fire on plant canopies

Data from the present study clearly show the con-

siderable effect of fire on woody plant canopies, par-

Fig. 21.— A typical thick (>5 cm)

layer of leaf litter under a

patch of Ochna pulchra in-

dividuals: a, before the fire

(5th September); b, after the

fire (6th September) showing

the low degree of litter burn

with only a superficial layer of

the litter being burned or

singed. Temperature tempil

plates are visible.

550 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA— OCHNA PULCHRA SAVANNA

ticularly of the smaller plants and these findings

broadly agree with results from other areas.

Whereas a two-thirds decrease in canopy volume

corresponded to trees and shrubs under 5 cm DBH in

Burkea africana Savanna in a fire treatment at

Makambu, South West Africa (Geldenhuys, 1977),

this decrease corresponded to 1 m tall plants of

Ochna pulehra in the present study. The accompany-

ing 60°7o reduction in plant height at Makambu was,

however, not as great as the percentage reduction for

the 1 m tall Ochna pulehra.

Although total plant mortalities may be low in

both Acacia and Burkea-Ochna Savannas, it appears

that in terms of canopy damage and canopy mortality

the Nylsvley study species such as Ochna pulehra and

even Vitex rehmannii may be less susceptible to fire

than some southern African /Icoc/o-dominated vege-

tation.

The observed high resistance of the canopy of a

large individual of Securidaca longipedunculata to

prolonged burning in the present study is in keeping

with the results from the hot November burn in

Burkea africana Savanna in South West Africa.

c) Fire and basal shoots

The present study amply confirmed an expected in-

crease in number of basal stems per plant individual

after fire. However, in contrast to some other cited

findings, regeneration in terms of numbers and mass

of shoots in Terminalia sericea after fire was far less

marked and the only tree species that reacted almost

as vigorously to fire as the Terminalia sericea in the

Wankie study (Rushworth, 1975) was Dombeya

rotundifolia. The present study’s relatively great

basal shoot mass in Lannea discolor agrees with the

report for the Transvaal lowveld where after an

October burn, Lannea discolor was one of the two

species that ‘sprouted well’ on certain plots (Anon,

1960).

That more than half the species combinations

referred to (Table 4) were significantly different with

respect to mean basal shoot regeneration mass after

fire appears to indicate a wide and fairly even range

of values with sometimes limited variance. In the

savanna plant community type studied therefore,

there was no major clustering of woody plant species

in terms of their basal regeneration response to fire.

Although multistemmed shrubs were (in the slower

burn) somewhat more susceptible to killing of live

basal shoots than were tree species, there was a clear

tendency for most multistemmed shrub species, par-

ticularly non-suffrutices, to have higher mean basal

regeneration shoot mass per plant than for most tree

species.

The present study data point to an almost paradox-

ical situation regarding fire induced basal shoot

regeneration of those Ochna pulehra plants with

100% canopy leaf mass reduction through fire. Here

a plant tends either to produce the maximum mass of

basal shoots or none at all (if it dies).

The ratio of number of new basal shoots to the

number of old basal shoots killed by the fire, namely

the basal shoot replacement ratio, is possibly a useful

attribute of vegetation that is subject to recurring

fire, that is where the basal shoots that are produced

after one fire are affected in a subsequent fire. It is

clear that successive fires may not result in a merely

additive process of increasing basal shoot numbers.

If the basal shoot replacement ratio is assumed to be

constant with successive fires, it may be postulated

that basal shoots of Ochna pulehra will increase more

rapidly than those of Burkea africana on the basis of

the existing data. A finding analogous to the present

study result is the determination of a positive correla-

tion indicating that species, for example Burkea

africana, are dependent upon the number of parent

trees for regeneration in regularly burned savanna

vegetation at Makambu, South West Africa (Gelden-

huys, 1977).

d) Biomass and other relations after fire

That there is a positive relation between the

amount of reduction in canopy leaf mass and the

amount of basal regeneration is an underlying reason

for the existence (in Fig. 14) of a point corresponding

to an intermediate height at which plants exhibit a

maximum change in total (basal and canopy) leaf

mass and below which smaller plants have a reduced

loss in total leaf biomass. Although the data only

support a positive change in total leaf mass after fire

for smallest plants, it is likely that very large trees

exhibit no significant change in total leaf biomass

after fire (there is no basis for upward extrapolation

of curve b in Fig. 14).

One of the main findings of the present study may

be described as the difference between basal shoot

mass after fire and unburnt canopy shoot mass as

they depend on plant height (Figs. 15 & 17). Apart

from the different forms of the relation with height

and the relative proportion of the biomass compo-

nents, there are also the differences in this proportion

with plant height (Fig. 16). Du Toit’s (1972b) relation

of stem diameter (mm) and mean coppice regrowth (1

x 10' g) of decapitated trees of Acacia karroo in

the eastern Cape Province

indicates steadily increasingly coppice mass with in-

creasing tree size. This pattern was not attained in the

present fire study since decreased canopy damage

effects in the larger trees of the burn changed this

relationship in the upper size range. Only where a

severe crown fire is conceivable would a Du Toit type

relation be expected to apply to a burned tree popula-

tion.

The present study has clearly demonstrated rela-

tionships between reduction in canopy leaf mass and

basal shoot regeneration mass. But it is important to

note that, although in accordance with expected

effects of reduction in apical dominance, increasing

damage to the canopy increased basal shoot produc-

tion (up to a threshold value), there was an opposite

effect for very small uncanopied O. pulehra plants

where increased damage to the plant reduced new

basal shoot production. That basal shoot produc-

tion, in plants with much canopy leaf mass removed

by fire, was greater in the faster burn area than in the

slower burn area, suggests that the fire heat intensity

also had a more direct effect on stimulation of basal

shoot production possibly through more effective

killing of buds in the canopy.

An indication of the possible limited resources of

plants in their regeneration reaction to fire is

reflected by the inverse hyperbolic distribution

between number of basal shoots per plant and the

mean individual basal shoot mass: a compensatory

effect between number and size of basal regeneration

Shoots in O. pulehra.

M. C. RUTHERFORD

551

It was found that in several respects, the effects of

the faster fire were more variable than in the slower

fire, for example, the relationship between number

of basal regeneration shoots (in O. pulchra) and

mean mass per basal shoot was tighter in the slower

burn than in the faster burn. This possibly parallels

the findings of Trollope (1978) in eastern Cape

grassland where the rate of spread of head fires was

far more variable than that of back fires. In the rela-

tionships between plant height and changes in plant

size (Fig. 8), the faster fire consistently had lower

correlation coefficients than those for the slower fire.

Although most different types of effects of fire on

the woody plants were found to depend on plant

height, a few effects were found to be independent of

plant height namely: the killing of live basal shoots

and the burning away of old dead basal shoots. Plant

height also appeared not to affect the relationship in

O. pulchra between number of basal regeneration

shoots and the mean mass per basal shoot.

The dependency or independency of fire effects

according to plant species varied from the indepen-

dency of the killing of live basal shoots by fire with

(tree) species to the strongest dependency with species

of biomass of basal regeneration shoots per plant.

The consumption by fire of old dead basal shoots

and the effects of fire on dead standing plants were

independent of species.

It is clear that work on specific effects of fire

within the savanna woody plant and plant popula-

tions creates a basis for understanding many features

of short and long term fire response systems in

plants. Much further work is required before an ade-

quate understanding of such systems in populations

of woody savanna plants is obtained. An undoubted

limitation of current work in this field is the neglect

of the role of the below ground component of the

plant in governing the above ground response to fire.

ACKNOWLEDGEMENTS

The author thanks: Messrs P.S. Carr, M.D.

Panagos, B. Curran and T.W. Kgaditsi for their con-

tributions at many stages during the course of this

work; Dr J.W. Morris and Mr P.G.H. Frost for

commenting on the manuscript.

U/TTREKSEL

Uitwerking van twee intensiteite van lentebrand op

verskeie aspekte van houtagtige plante van ’n Burkea

africana — Ochna pulchra savanne na een groeisei-

soen word gegee. Mortaliteit van houtagtige plante

was baie laag, byvoorbeeld in Ochna pulchra was die

mortaliteit tussen 1 en 5%. In sommige soorte waar

die bogrondse dele dikwels heeltemaal weggebrand

het, soos in Grewia flavescens, het geen mortaliteit

van individue voorgekom nie. Daar is gevind dat

basale lootmassa hergroei parabolies afhanklik is van

planthoogte terwyl die verhouding tussen blaar en

takkie massa van basale loot hergroei omgekeerd

verander het met plant hoogte in Ochna pulchra. Dit

wil voorkom dat die vermoe van Ochna pulchra

plante om nuwe basale lote te vorm nie net op plant-

grootte berus nie maar ook op die aantal basale lote

wat voor die brand teenwoordig was. Vir lewendige

Ochna pulchra plante is daar gevind dat basale loot-

massa hergroei per individu eksponensieel toeneem

met groter vermindering in kroonblaar biomassa. Die

verband is ook moontlik deur direkte hitte uitwer-

kings geaffekteer. Basale lootmassa hergroei van

plant soorte het baie gevarieer van 0, 7 g/individu vir

Dichapetalum cymosum tot 285,6 g/individu vir

Euclea natalensis. Daar is ’n duidelike neiging dat

nie-halfstruikvormige struiksoorte ’n groter gemid-

delde basale lootmassa hergroei per plant toon as die

van meeste ander boomsoorte. Daar is ’n kompenser-

ende uitwerking in Ochna pulchra tussen getal en

grootte van basale hergroeide lote. Staande dooie

houtagtige plante (voor die brand) het of omgeval as

gevolg van die vuur, of is waarskynlik nie geaffekteer

deur die vuur nie. Daar is ook geen selektiwiteit op

grond van plantsoort nie. Resultate van die huidige

studie word algemeen gestaaf deur werk op die

uitwerking van vuur in savanne en sommige ander

plantegroeitipes.

REFERENCES

Afolayan, T. A., 1978. Savanna burning in Kainji Lake National

Park, Nigeria. E. Afr. Wildl. J. 16: 245-255.

Anon., 1960. Annual report of the biologist, Kruger National

Park, 1958/1959. Review of activities. Veld Burning. Koedoe

3: 173-194.

Anon., 1978. South African programme for the SCOPE mid-term

project on the ecological effects of fire. South African

National Scientific Programmes Report No. 32. 36 pp. CS1R.

Barnes, D. L., 1965. The effects of frequency of burning and

mattocking on the control of coppice in the Marandellas sand-

veld. Rhod. J. Agric Res. 3: 55-56.

Christensen, M. L., 1977. Fire and soil plant nutrient relations in

a pine wire grass savanna on the coastal plain of North

Carolina. Oecologia (Berl.) 31: 27-44.

Coetzee, B. J., Van Der Meulen, F., Zwanziger, S., Gonsalves,

P. & Weisser, P. J., 1976. A phytosociological classification

of the Nylsvley Nature Reserve. Bothalia 12: 137-160.

Deeming, J. E., Lancaster, J. W., Fosberg, M. A., Furman,

R. W. & Schroeder, M. J., 1972 The national fire-danger

rating system. USDA Forest Service Research Paper RM-84.

165 pp. Rocky Mt Forest and Range Exp. Stn, Fort Collins,

Colorado.

Donaldson, C. H., 1966. Control of blackthorn in the Molopo

area with special reference to fire. Proc. Grassld Soc. sth.

Afr. 1: 57-62.

Du toit, P. F., 1972a. Acacia karroo intrusion: the effect of

burning and sparing. Proc. Grassld Soc. sth. Afr. 7: 23-27.

Du toit, P. F., 1972b. The goat in a bush-grass community. Proc.

Grassld Soc. sth. Afr. 7: 44-50.

Farmer, R. E., 1962. Aspen root sucker formation and apical

dominance. For. Sci. 8: 403-410.

Gandar. M. V., In Press. Description of a fire and its effects in

the Nylsvley nature Reserve. A synthesis report. South

African National Scientific Programmes Report No. 00 (27

pp. typescript in Feb. 1979).

Geldenhuys, C. J., 1977. The effect of different regimes of

annual burning on two woodland communities in Kavango. S.

Afr. For. J. 103: 32-42.

Gill, A. M., 1974. Toward an understanding of fire-scar forma-

tion: field observation and laboratory simulation. For. Sci.

20: 198-205.

Grunow, J. O. & Grossman, D., 1978. Savanna Ecosystem

Project, Nylsvley. Producer Group: grass layer. Summary of

observations on the effect of fire on biomass accumulation

rate (BAR) and biomass regime (BR) in the grass layer of

Burkea Savanna over a 60 day period. Unpublished report to

the National Programme for Environmental Sciences. CS1R

Pretoria. 7 pp. typescript.

Harrington. G. N., 1974. Fire effects on a Ugandan savanna

grassland. Trop. Grassld 8: 87-101.

Harrington, G. N. & Ross, I. C., 1974. The savanna ecology of

Kidepo Valley National Park. 1. The effects of burning and

browsing on the vegetation. E. Afr. Wildl. J. 12: 93—105.

Harrison, T. D., 1978. Report on maximum temperature

measurements during the Nylsvley veld fire of 5th September,

1978 and on post fire micro-meteorological measurements.

Unpublished report to the National Programme for Environ-

mental Sciences. C.S.l.R. Pretoria. 19 pp. typescript.

Hopkins, B., 1963. The role of fire in promoting the sprouting of

some savanna species. J. West Afr. Sci. Assoc. 7: 154-162.

Hopkins, B., 1965. Observations on savanna burning in the

Olokemeji Forest Reserve, Nigeria. J. Appl. Ecol. 2:

367-381.

Huntley, B. J., 1978. Nylsvley— a South African savanna eco-

system project: objectives, organization and research pro-

gramme. South African National Scientific Programmes

Report No. 27. 37 pp. ISBN 0 7988 1311 3.

552 SURVIVAL, REGENERATION AND LEAF BIOMASS CHANGES IN WOODY PLANTS FOLLOWING SPRING

BURNS IN BURKEA AFRICANA—OCHNA PULCHRA SAVANNA

Huntley, B. J. & Morris, J. W., 1978. Savanna ecosystem

project : phase I summary and phase 2 progress. South

African National Scientific Programmes Report No. 29. 52

pp. ISBN 0 7988 1346 6.

Jackson, G., 1974. Cryptogeal germination and other seedling

adaptations to the burning of vegetation in savanna regions,

the origin of the pyrophytic habit. New Phytol. 73: 771-780.

James, T. D. W. & Smith, D. W., 1977. Short-term effects of

surface fire on the biomass and nutrient standing crop of

Populus tremuloides in Southern Ontario. Can. J. For. Res.

7: 666-679.

Kaul, R. N. & Jain, M. B., 1967. Growth attributes — their rela-

tion to fuel yield in Prosopis cineraria (Linn.) McBride (P.

spicigera Linn.) Comm. For. Rev. 46: 155-158.

Kemp, R. H., 1963. Growth and regeneration of open savanna

woodland in northern Nigeria. Comm. For. Rev. 42:

200-206.

Kennan, T. C. D., 1971 . The effect of fire on two vegetation types

of Matopos. Proc. Ann. Tall Timbers Fire Ecologv Conf. 1 1 :

53-98.

Lawton, R. M., 1978. A study of the dynamic ecology of Zambian

vegetation. J. Ecol. 66: 175—198.

McNAB, W. H., Edwards, M. B. & Hough, W. A., 1978.

Estimating fuel weights in slash pine-palmetto stands. For.

Sci. 24: 345-358.

Niering, W. A., Goodwin. R. H. & Taylor, S., 1970. Prescribed

burning in Southern New England: Introduction to long-

range studies. Proc. Ann. Tall Timbers Fire Ecology Conf.

10: 267-286.

Reiners, W, A., & Wright, H. E., 1977. Impact of prehistoric

and present fire patterns on the carbon dioxide content of the

atmosphere. In W. Stumm, Physical and Chemical Sciences

Research Reports, 2. Global chemical cycles and their altera-

tions by man. Report of the Dahlem Workshop, Berlin, West

Germany. November 15-19, 1976. pp. 121-135. Dahlem

Konferenzen: Berlin, West Germany. ISBN 3 8200 1207 9.

Rushworth, J. E., 1975. The floristic, physiognomic and biomass

structure of Kalahari sand shrub vegetation in relation to fire

and frost in Wankie National Park, Rhodesia. M.Sc. thesis,

University of Rhodesia.

Rushworth, J. E., 1978. Kalahari sand scrub — something of

value. The Rhodesia Science News 12: 193-195.

Rutherford, M. C., 1975. Aspects of ecosystem function in a

savanna woodland in South West Africa. Ph.D. thesis,

University of Stellenbosch.

Rutherford, M. C., 1979. Aboveground biomass subdivisions in

woody species of the Savanna Ecosystem Project study area,

Nylsvley. South African National Scientific Programmes

Report No. 36. 33 pp. ISBN 0 7988 1445 4.

Rutherford, M. C., In Press. Woody plant biomass distribution

in Burkea africana savannas. In B. J. Huntley & B.H. Walker,

Ecology of tropical savannas.

Smith, D. W. & James, T. D., 1978. Characteristics of prescribed

burns and resultant short-term environmental changes in

Populus tremuloides woodland in Southern Ontario. Can. J.

Bot. 56: 1782-1791.

Stransky, J. J. & Halls, L. K., 1979. Effect of a winter fire on

fruit yields of woody plants. J. Wild!. Mgmt 43: 1007-1010.

Tinley, K. L., 1966. An ecological reconnaissance of the Moremi

Wildlife Reserve, Northern Okavango Swamps, Botswana.

Johannesburg: Okavango Wildlife Society.

Trapnell, C. G., 1959. Ecological results of woodland burning

experiments in Northern Rhodesia. J. Ecol. 47: 129-168.

Trapnell, C. G., Friend, M. T., Chamberlain, G. T. & Birch,

H. F., 1976. The effects of fire and termites on a Zambian

woodland soil. J. Ecol. 64: 577-588.

Trollope, W. S. W., 1974. Role of fire in preventing bush

encroachment in the eastern Cape. Proc. Grassld Soc. sth.

Afr. 9: 67-72.

Trollope, W. S. W., 1978. Fire behaviour — a preliminary study.

Proc. Grassld Soc. sth. Afr. 13: 123-128.

Trollope, W. S. W., Rutherford, M. C., Noble, R. I.,

Coutinho, L. M., MacDonald, I. A. W., Harrison, T. D.,

Menaut, J. C., Van Wyk, P & Grossman, D., (in press). Fire

as a tool in the management of savanna. In B.J. Huntley,

B.H. Walker, Ecology of tropica! savannas.

Van Rensburg, H. J., 1971. Fire: its effect on grasslands,

including swamps — southern, central and eastern Africa.

Proc. Ann. Tall Timbers Fire Ecology Conf. 1 1: 175-199.

Van Wyk, P., 1971. Veld burnings in the Kruger National Park.

An interim report of some aspects of research. Proc. Ann.

Tall Timbers Fire Ecol. Conf. 1 1 : 9-31 .

West, O., 1965. Eire in vegetation and its use in pasture manage-

ment with special reference to tropical and subtropical A frica.

Mimeographed publication No. 1. Commonwealth Bureau of

Pastures and Field Crops, Hurley, Berkshire. Commonwealth

Agricultural Bureaux. 53 pp.

West, O., 1971. Fire, man and wildlife as interacting factors

limiting the development of climax vegetation in Rhodesia.

Proc. Ann. Tall Timbers Fire Ecology Conf. 11: 121-145.

Zackrisson, CL, 1977. Influence of forest fires on the north

Swedish boreal forest. Oikos 29: 22-32.

Bothalia 13, 3 & 4: 553-556 (1981)

Monitoring Phragmites australis increases from 1937 to 1976

in the Siyai Lagoon (Natal, South Africa) by means of air

photo interpretation

P.J. WEISSER* and R.J. PARSONS*

ABSTRACT

The colonization of the Siyai Lagoon on the north coast of Natal by Phragmites australis was studied by means of

air photo interpretation. It was possible to locate and estimate P. australis areas for 1957 (0,74 ha), 1965 (1,65 ha),

1969 (1,93 ha) and 1976(2,94 ha). Phragmites australis first inhabited the shores of the middle section of the lagoon

followed by rapid expansion in the lower section. The upper section was colonized only at its lower end by expansion

from the middle section. It is suggested that P. australis was unsuccessful in this section because of competition by

the Hibiscus tiliaceus — Barringtonia racemosa Lagoon Fringe Forest. This same community is shading out P.

australis in some places. The notable increase in the rate of advance of land and littoral vegetation into the Siyai

Lagoon was caused by sugar farming activities leading to erosion and sedimentation in the lagoon. A vegetation age

gradient was observed from the upper section to the mouth region. The colonization of most of the Siyai Lagoon

except the immediate mouth zone by P. australis Reedswamp and Hibiscus tiliaceus — Barringtonia racemosa

Lagoon Fringe Forest, can be expected before the turn of the century. Dredging and mechanical control of vegeta-

tion will become necessary if major open water spaces are to be maintained.

RESUME

CONTROLE DES ACCROISSEMENT DE PHRAGMITES AUSTRALIS DE 1937 A 1976 DANS LA LAGUNE

SIYAI (NA TAL, AFRIQUE DU SUD) AU MO YEN DTNTERPRETA TION DE PHOTOGRAPHIES AERIEN-

NES

La colonisation de la lagune de Siyai sur la cote nord du Natal par les Phragmites australis a ete etudiee au moyen

d' interpretation de la photo aerienne. II a ete possible de situer et d’estimer les regions de P. australis pour 1957

( 0 , 74 ha), 1965 ( 1,65 ha) 1969 (1,93 ha) et 1976 (2,94 ha). Phragmites australis habita d’abord les rivages de la sec-

tion mediane de la lagune puis s’etendit rapidement a la section en dessous. La section superieure fut seulement

colonisee dans le dessous par expansion de la section medianne. I! est considere que P. australis ne reussit pas a

s’etendre dans cette section par suite de la concurrence avec /’Hibiscus tiliaceus — Barringtonia racemosa Lagoon

Fringe Forest. Cette meme communaute supplante P. australis dans certains endroits. L’accroissement notable de

progres de la vegetation du littoral et de I’interieur des terres dans la lagune Siyai fut cause par les activites agricoles

sucrieres conduisant a /’ erosion et a la sedimentation dans la lagune. A part P. australis aucun autre macrophyte ne

fut trouve. Une inclinaison d’age de la vegetation fut observee de la section superieure a la region de la bouche. La

colonisation de la plupart de lagune Siyai, a I’exception de lazone immediate de la bouche, par P. australis

Reedswamp et Hibiscus tiliaceus — Barringtonia racemosa. Lagoon Fringe Forest, doit etre attendue avant la fin du

siecle. Le draguage et la lutte mecanique contre la vegetation deviendront necessaires si Ton veut maintenir de

grandes etendues d’eau ouvertes.

INTRODUCTION

An increase of reeds was observed on air photos

when studying the dune advancement at Mtunzini

(Weisser et at., MS). The availability of air photo

coverage for 1937—1979 offered the possibility to

monitor the reed encroachment. Objectives of this

work were to evaluate the adequacy of air photos in

monitoring P. australis expansion; to provide base-

line data on the Siyai Lagoon; to quantify the reed

encroachment; to establish trends and extrapolate

possible developments; and to offer possible manage-

ment suggestions.

The Siyai drainage system is situated in Natal at

latitude 28°58' South and longitude 31°45'45" East.

The Lagoon is situated amidst a well-conserved,

scenically beautiful, forested-dune landscape (Fig.

1). It has a surface of about 8 ha, an axial length of

2,5 km and a catchment of about 18 km2 (Begg,

1978). Two streams are the main tributaries. In this

work only the lagoon was studied and defined as ex-

tending from the confluence of the two tributaries to

the Lagoon Mouth. Four sections were distinguished

from south-west to north-east: upper section (from

confluence to about 650 m north-east); middle sec-

*Botanical Research Institute, Department of Agriculture and

Fisheries, Private Bag X101, Pretoria, 0001.

tion (inflexion area of watercourse); lower section (to

watercourse constriction) and mouth section.

Between 1937 and 1977, the mouth of the Siyai

Lagoon moved about 740 m north-eastwards, at an

average rate of 17,4 m/year (Weisser et al., MS). The

mouth is usually closed, being open for only very

brief periods (of up to a week) after floods. The bar

may be topped by high spring-tides and the mouth is

not opened artificially (Begg, 1978).

Land encroachment into the Lagoon has increased

markedly since sugar farming began in c. 1946.

Changes in run-off and soil exposure caused exten-

sive sedimentation reducing the depth to about 0,25

m in the upper zone of the Lagoon (Begg, 1978).

Maximum depths of 1,5 and 2 m respectively were

measured by the first author in June 1980 at the

crossing of the nature trail (lower zone) and in the

Lagoon mouth. Begg (pers. comm.) recorded a depth

of 2,9 m in the middle section of the Lagoon.

METHODS

Information on the Siyai Lagoon and the reeds was

obtained from air photos through direct inspection,

enlargement and transference onto a base map using

a Bausch & Lomb ZT-4 Zoom Transfer Scope

(=ZTS). This instrument was also used to draw the

1:5 000 base map from the 1977 orthophoto maps

554

MONITORING PHRAGMITES AUSTRALIS INCREASES FROM 1937 TO 1976 IN THE SIYAI LAGOON

(NATAL, SOUTH AFRICA) BY MEANS OF AIR PHOTO INTERPRETATION

Fig. 1. — The middle section of

the Siyai Lagoon covered with

Phragmites australis (June

1980). A photograph taken in

1964 shows only a few patches

of reeds. The foreground

shows leaves of Hibiscus

tiliaceus; this plant and Bar-

ringtonia racemosa are the

main components of the

forest fringe. This community

is encroaching into the P.

australis stands.

2831 DD 21 ‘Mtunzini’ and 2831 DC 25 ‘Ekuhleni’.

Interpretation of the aerial photographs was aided by

a Topcon Stereoscope.

The photographs used were Nos 54692 (Job 117,

1937.05.05), 9426 (Job 400, 1957.05.24), 8082 (Job

499/4, 1965.06.14), 5664 and 5609 (Job 608,

1969.08.18) and 161, 162 and 163 (Job 251,

1976.05.14). Areas were measured with an MOP-

AM02 Image Analyser. Ground truth and additional

field data were gathered on the 21st and 22nd of

March, 21st of May and the 8th and 12th of June

1980.

The potential and limitations of the use of air

photos in vegetation studies have been discussed in

literature, e.g. Edwards (1972) and Weisser (1979).

Concerning this study, the following points should be

borne in mind. The photographs are at different

scales and were not taken at the same time of year.

Another factor is the differing resolution of the air

photos compared, the 1937 photographs having the

poorest resolution. Also, the water levels of the Siyai

Lagoon varied, therefore changing the area occu-

pied. However, the Siyai Lagoon is situated between

steep dune ridges, consequently changes in water

levels have only a minor influence on the area inun-

dated. The reed growth varies seasonally. There is a

time lag between the establishment of reeds and their

detection on air photos. Sometimes P. australis col-

onization may be below the detection limit of the air

photo taken. The density of the P. australis stand

also varies, being at its lowest in winter. This may

affect its detection on the air photos.

RESULTS AND CONCLUSIONS

Field checking confirmed that the Siyai Lagoon

was a good site for studying the colonization and ex-

pansion of P. australis Reedswamp. The period

documented in the air photos was long enough to

enable the reeds to cover all the available habitat in

some sections of the river, whereas in others P.

australis is still actively expanding. The observation

time was also sufficient to allow succession to pro-

ceed, and in some areas P. australis Reedswamp is

being displaced by other riverine vegetation (Fig. 1).

The results deal first with the findings on P. australis,

then with other aquatic or semi-aquatic vegetation,

and finally with the Siyai Lagoon.

Adequacy of air photos to monitor P. australis

colonization

It was possible to locate and estimate P. australis

areas for 1957, 1965, 1969 and 1976 (Table 1 & Fig.

3). The resolution of the 1937 air photos is insuffi-

cient to give conclusive evidence. Difficulties were

encountered in the interpretation of Job 291 (1977)

and Job 329 (1979). Their scale (1:3 000) gives too

few matching points in the optical field of the ZTS.

Colonization pattern

Phragmites australis first colonized the middle and

lower sections of the Lagoon. It failed to establish

itself in the immediate mouth zone and in the upper

region with the exception of a 0,5 m2 patch in the

confluence (June, 1980). Both zones are at present

TABLE 1. — Area changes (in ha) in the Siyai Lagoon (1937-1976) as shown in the air photos

J

P. J. WEISSER AND R. J. PARSONS

555

Fig. 2. — The mouth section of

the Siyai Lagoon showing the

Phragmites australis stand

nearest to the sea. The dunes

to the right are covered with

dune scrub, and the young

dunes on the left are colonized

by the dune pioneer Scaevola

thunbergii.

shallow. It seems that when silting increased drastic-

ally in the upper section, the already present woody

fringe vegetation rapidly invaded the new habitat and

ousted P. australis. In the lower section of the

Lagoon, P. australis was able to establish itself and

to expand in the direction of the Lagoon mouth.

Phragmites australis Reedswamp area changes :

1937-1976

The P. australis surface increases at the Siyai

Lagoon are summarized in Table 1 and represented

graphically in Fig. 3. Whereas from 1937 to 1957

there was an increase of only 0,47 (?) ha in the area

covered by P. australis, the area increased by 2,20 ha

between 1957 and 1976. Therefore expansion of P.

australis in the Lagoon accelerated after 1957 (Fig.

3). The curve will flatten out as the available habitat

diminishes. However, with the north-eastward

advancement of the mouth, new areas suitable for P.

australis colonization are being created (Fig. 2).

Probable causes of reedswamp increase

Reedswamp expansion is often a natural process in

some lagoons. Conspicuous increases of reeds have

been reported in lakes after lowering of water levels

(e.g. Kopf, 1964, in Weisser 1970; Bjork, 1974) or

when siltation increases (Weisser, 1978). Begg (1980)

states that reed encroachment is due to the Lagoon

becoming shallower and less saline. We agree that

sedimentation and consequent reduction in depth are

the chief causes of the remarkable increase in P.

australis. No records are available for assessing the

influence of the salinity factor.

The sedimentation of the Siyai Lagoon has in-

creased strikingly, becoming critical with the advent

of sugar farming in the catchment area (c. 1946 fide

Begg, 1978). This must be considered as the main

reason for the deterioration of the Siyai Lagoon.

Garland (pers. comm.) considers the clearing and

cultivation of Swamp Forest and Cyperus papyrus

Swamp (c. 1955) as being especially detrimental,

because of the consequent elimination of the

‘sponge’ function of the vegetation.

Succession

In some areas of the middle section of the Lagoon,

a replacement of P. australis Reedswamp by the

Hibiscus tiliaceus-Barringtonia racemosa Lagoon

Fringe Forest was observed, probably by shading.

This community seems to be the following phase in

the land reclamation process in this Lagoon. This

corresponds with successional schemes proposed

(Edwards, 1967; see also Ward, 1980).

Reduction of open water by increase of woody shore

vegetation

An expansion of the Lagoon Fringe Forest formed

mainly by Hibiscus tiliaceus and Barringtonia

SIYAI LAGOON

• TOTAL AREA OF LAGOON

A OPEN WATER

X OPEN WATER WITHOUT MOUTH REGION

■ PHRAGMITES AUSTRALIS

1935 1940 1950 1900 1970 I960

Fig. 3. — Area changes in the Siyai Lagoon 1937-1976 as shown

by means of air photo interpretation.

556

MONITORING PHRAGMITES AUSTRALIS INCREASES FROM 1937 TO 1976 IN THE SIYAI LAGOON

(NATAL, SOUTH AFRICA) BY MEANS OF AIR PHOTO INTERPRETATION

racemosa on the upper section of the Lagoon was

observed. This narrowed the Lagoon by occupying

areas that were previously open water.

Age gradient of woody shore vegetation

In the upper section of the Lagoon, the woody

fringe vegetation consists of dense tall trees (about 9

m), whereas the woody vegetation downstream is less

dense, lower and formed by young trees. In the lower

section, the woody shore vegetation is only 2 to 3 m

high, discontinuous and limited to isolated seedlings

of Hibiscus tiliaceus and Barringtonia racemosa.

Therefore, an age gradient for the Lagoon fringe

vegetation from the upper Lagoon section to the

Lagoon mouth can be assumed (see also Weisser et

al., MS).

A few patches of dead woody fringe vegetation

were noted in the lower section of the Lagoon. The

preceding high water levels have been suggested as

the cause by Garland (pers. comm.).

Area changes of the Siyai Lagoon

The surface covered by the Lagoon including

reeds, but excluding the Hibiscus tiliaceus —

Barringtonia racemosa Lagoon Fringe Forest has

increased from about 6,06 ha (1937) to 7,87 ha

(1976). This increase is mainly due to accretionary

processes in the mouth region, which has advanced

about 740 m northwards (1937-1977), therefore oc-

cupying areas previously covered by dunes (Figs 2 &

3).

The area of open water has decreased since 1937,

mainly owing to the reedswamp encroachment and

the expansion of the Lagoon fringe vegetation.

However, the advancement of the Lagoon mouth has

produced new open-water areas, compensating par-

tially for the loss to the reed beds and the Hibiscus

tiliaceus — Barringtonia racemosa Lagoon Fringe

Forest. If the values are corrected by subtracting the

new open-water areas of the mouth section, the real

loss of open-water surface becomes evident (Table 1

& Fig. 3).

Sedimentation

The sedimentation has been heaviest in the upper

zone of the Lagoon. The dense P. australis stand in

the middle zone probably acts as a sediment trap

partly protecting the lower section of the Lagoon

from sedimentation.

If sedimentation is not controlled, the accelerated

filling up of the bed will continue. This could even-

tually lead to the water breaking through and form-

ing another bed.

Prognosis

Because of sedimentation, the colonization by P.

australis Reedswamp and Hibiscus tiliaceus- Barring-

tonia racemosa Lagoon Fringe Forest of most of the

Siyai Lagoon, except the immediate mouth zone, can

be expected probably before the turn of the century.

Any management action tending to diminish sedi-

mentation, e.g. erosion control upstream, should be

encouraged. Reeds could be controlled locally by

repeated underwater cutting. The opening of a new

river course may occur in the long term owing to fill-

ing up of the existing Lagoon bed. Dredging and

mechanical control of vegetation will probably

become necessary if major open-water spaces are to

be maintained at the Siyai Lagoon.

ACKNOWLEDGEMENTS

Acknowledgements are due to Drs J. C. Scheepers,

G. W. Begg and D. Edwards, Messrs W. de Waal and

R. Westfall, Miss B. K. Drews and Mrs J. Weisser

for their valuable comments; Mrs J. Schaap for the

drawing; Messrs I. Garland and C. Buthelezi for their

help during fieldwork; Mesdames J. Mulvenna and

S. Smit for typing the manuscript; Mr R. E. Crofts

(Office of the Director General of Surveys), Prof. D.

Scogings and Mr A. Bikaroo (Survey Department,

University of Natal) for their help in obtaining aerial

photographs; Mrs A. Romanowski for the photo-

graphic work; and the Natal Parks, Game and Fish

Preservation Board for their support during field-

work.

UITTREKSEL

Indringing deur Phragmites australis in die Siyai

Strandmeer aan die Natalse noordkus is bestudeer

deur middel van lugfoto-interpretasie. Dit was

moontlik om Phragmites australis stande uit te wys

en die oppervlaktes daarvan te skat t.o.v. die jare

1957(0, 74 ha), 1965 (1,65 ha), 1969 (1,93 ha) en 1976

(2,94 ha), maar nie vir 1937 nie. Phragmites australis

het eers net op die middelseksie van die strandmeer se

oewers voorgekom, maar het vinnig uitgebrei na die

stroom-af seksie. Die stroom-op seksie is alleenlik

laer af ingedring deur uitbreiding van die middel-

seksie. Dit word veronderstel dat P. australis nie

beter in laasgenoemde seksie kon slaag nie weens die

kompetisie wat die Hibiscus tiliaceus — Barringtonia

racemosa Strandmeer-oewerwoud bied. Laas-

genoemde plantgemeenskap oorskadu P. australis in

sommige plekke. Die merkwaardige toename in die

tempo van verdringing van die Siyai Strandmeer deur

slik en oewerplantegroei word veroorsaak deur

suikerboerdery wat lei tot verspoeling en toeslikking.

’n Plantegroei-ouderdomsgradient vanaf die

stroomopseksie tot by die monding is waargeneem.

Indringing van die grootste gedeelte van die Siyai

Strandmeer deur P. australis Rietmoeras en Hibiscus

tiliaceus — Barringtonia racemosa Strandmeeroewer-

woud kan voor die einde van die eeu verwag word.

Indien daar noemenswaardige oopwaterruimtes in-

stand gehou moet word, sal baggerwerk en meganiese

beheer van die plantegroei noodsaaklik wees.

REFERENCES

Begg, G., 1978. The estuaries of Natal. Natal Town and Regional

Planning Report. Vol. 41.

Begg, G., 1980. Siyai catchment study. Environment RSA. 7, 3:

6-7.

Bjork, S., 1974. The degradation and restoration of Lake Horn-

borga. In I.H.D. Nutson, Nordic case book on inadvertent

effect of man on the Hydrological Cycle. Reprint from the

Institute of Limnology, University of Lund, Lund, Sweden.

Edwards, D., 1967. A plant ecological survey of the Tugela River

Basin. Mem. bot. Surv. S. Afr. No. 36.

Edwards, D., 1972. Remote sensing in the evaluation of the

natural vegetation resources of South Africa. Proc. 5th Symp.

Remote Sensing, Pretoria, CSIR, May 1972, 99-102.

Ward, C. J., 1980. The plant ecology of the Isipingo Beach Area,

Natal, South Africa. Mem. bot. Surv. S. Afr. No. 45.

Weisser, P. J., 1970. Die Vegetationsverhaltnisse des Neusiedler-

sees. Pflanzensoziologische und Okologische Studien. Wiss.

Arbeiten a.d. Burgenland. 45: 1-81.

Weisser, P. J., 1978. A conceptual model of a siltation system in

shallow lakes with littoral vegetation. J. Limnol. Soc. sth.

Afr. 4: 145-149.

Weisser, P. J., 1979. Suitability of air photo interpretation for

monitoring coastal dune vegetation of the Zululand Dunes,

South Africa. In The use of ecological variables in environ-

mental monitoring. The National Swedish Environmental

Protection Board, Report PM 1151, 62-72.

Weisser, P. J., Garland, I. F. & Drews, B. K., MS. Dune ad-

vancement 1937-1977 and preliminary vegetation succession

chronology at Mlalazi Nature Reserve, Natal, South Africa.

Bothalia (in preparation).

J

Bothalia 13, 3 & 4: 557 - 567 ( 1981 )

Structural and floristic classifications of Cape Mountain

Fynbos on Rooiberg, southern Cape

H. C. TAYLOR* and F. VAN DER MEULEN**

ABSTRACT

Structure and floristic composition of the plant cover were used to establish separate classifications of plant com-

munities in Mountain Fynbos on Rooiberg, South Africa. The structural units and floristic associations closely

correlate with each other and their distribution reflects the major environmental influences, aspect and altitude. It is

concluded that, despite the preliminary character of the survey, resource inventories of this type are suitable as a

foundation for park management.

RESUME

LES CLASSIFICA TIONS STRUCTURALES ET FLORIST1QUES DES BUISSONS DE LA MONTAGNE DU

CAP SUR ROOIBERG, CAP AUSTRAL

La composition structural et floristique de la couverture de plantes ont ete utilises pour etablir des classifications

separees de communautes de plantes dans le Fynbos de Montagne sur Rooiberg, en Afrique du Sud. Les unites

structurales et les associations floristiques correspondent les unes avec les autres et leur distribution reflete les in-

fluences du milieu, de Vaspect et de /’altitude. II en est conclu que, malgre le caractere preliminaire de I’etude, la

ressource d’inventaires de ce type convient comme la fondation pour la conduite de pares.

INTRODUCTION

At the request of the Department of Forestry, the

first author carried out a survey of the False Macchia

(Acocks, 1975) or Mountain Fynbos vegetation of

the Rooiberg Mountain Catchment Area to provide

basic data for a management plan (Taylor, in press).

For this purpose, a simple classification, description

and map were needed, dividing the vegetation into

units that are visually homogeneous in physiognomy.

Because these units had to be recognized by non-

botanical personnel, it was decided to adopt a struc-

tural classification, rather than a floristic one. The

structural survey was done by marking units of more

or less uniform vegetation structure and terrain mor-

phology on air photos (scale 1:20 000). These units

were then studied in the field. In about a month’s

field work, sufficient information was gathered to

delineate eight major structural units and 20 varia-

tions, all of which could be related to habitat. These

have been mapped (1:50 000) and described in detail

by Taylor (in press).

Management of Mountain Fynbos reserves such as

Rooiberg aims at maximum production of clear

water from catchments and the maintenance of vege-

tation diversity both in plant species and plant life

forms. Hence, a vegetation study for management

purposes should include both a floristic and a struc-

tural description. Time did not permit an extensive

phytosociological survey on Rooiberg. However, to

determine whether the major structural units could

also be distinguished by their species composition, a

brief Braun-Blanquet survey was carried out.

Twenty-eight releves were laid out subjectively within

stands of vegetation that were regarded as represen-

tative of the main structural units. In the phytosocio-

logical table only vascular plants were taken into

account. This paper summarizes and compares the

results of the structural and floristic surveys. For in-

♦Botanical Research Unit, P.O. Box 471, Stellenbosch 7600.

♦’Formerly, Botanical Research Institute, Pretoria. Present ad-

dress: Dune Waterworks of the Hague, Pompstationsweg 315,

2597 J. V. Scheveningen, Netherlands.

formation on the flora and phytogeography of Rooi-

berg we refer to Taylor (1979).

LOCATION AND HABITAT

About 20 km south-east of Ladismith, Cape Pro-

vince (Fig. 1), lies a west-east trending mountain

range, divided into more or less equal sections by the

Gouritz River. The Rooiberg Mountain Catchment

Area, 25 345 ha in extent, comprises a large part of

the higher western section of this range at approxi-

mately 33°40’S latitude and 21°30’E longitude. It

consists of State and private land, managed by the

Department of Water Affairs, Forestry and Environ-

mental Conservation for the conservation of water,

flora and wild life.

The range is one of a series of isolated mountains

located between Montagu and Uniondale, the fynbos

of each mountain being separated from that of its

neighbours by karroid lowland vegetation of the

Little Karroo.

Fig. 1 .—Map showing location of Rooiberg in southern Cape.

558

STRUCTURAL AND FLORISTIC CLASSIFICATIONS OF CAPE MOUNTAIN FYNBOS ON ROOIBERG,

SOUTHERN CAPE

Geomorphologically, the Rooiberg Range is a

broad anticlinal fold of Table Mountain Sandstone

strata, breached near the top of the arch. The breach

has eroded to form a series of deeply incised kloofs

that run parallel to the main crest before breaking

south in deep gorges that have cut through subsidiary

crests. The Table Mountain Sandstones are inter-

rupted by a shale band that can be traced as a sinuous

line across the north side of the range. Sandy litho-

sols cover the greater part of the reserve. Rock out-

crops and cliffs are common in the interior kloofs but

the main north- and south-facing sides of the range

are not excessively steep.

The climate is hot and dry. Situated on the transi-

tion between the winter- and summer- rainfall areas,

Rooiberg receives only the outer fringe of the winter

cyclonic rains and few summer thunderstorms. The

mean total precipitation for 1977/1978 (the first year

after rain gauges were erected) was about 365 mm,

the northern side of the mountain receiving about

half as much rain as the southern slopes (P. B. Oden-

daal, pers. comm.). Probably the most effective pre-

cipitation is from condensation of the clouds that

form around the high peaks. Considerable tempera-

ture and moisture differences may be expected be-

tween crests (about 1 400 m altitude) and footslopes

of the mountain (about 800 m).

VEGETATION

1. Structural units

The vegetation units identified in the structural

survey depict only gross structural characteristics.

They are based on simple criteria such as estimated

height and canopy cover of the vegetation and esti-

mated proportion of (i) restioid, (ii) narrow-leaved

and (iii) proteoid components. These components

characterize fynbos vegetation. For convenience,

their definitions are repeated here, (i) Restioid refers

to hemicryptophytic plants of the family Res-

tionaceae with sclerophyllous tubular stems and

leaves that have been reduced to membranous, non-

photosynthetic scales or small sheaths arising singly

from each node. They are often tufted but sometimes

rhizomatous. The category includes plants in other

families that have a physiognomy similar to the Res-

tionaceae, for example, most Tetraria and Ficinia

species, some J uncus species and Typha. (ii) Narrow-

leaved means the very narrow leptophylls, usually

sclerophyllous though some may be fleshy or even

succulent. It includes flat, involute and cylindrical

leaves as well as revolute leaves like those of the

Ericaceae. The term had to be substituted for

‘ericoid’ which is now restricted to revolute

sclerophyllous leptophylls. (iii) Proteoid is the term

for plants with leaves and growth form resembling

proteas. They have isobilateral leaves with the shape

and texture of a typical protea e.g. Protea lorifolia,

P. repens or broader leaves like P. nitida and P.

cynaroides. Broad-leaved is used here to denote the

roughly ovate, sclerophyllous leathery leaves of

woody shrubs derived from the northern floras as

against those of typical fynbos plants (e.g. Cassine,

Maytenus, Euclea).

The vegetation units appeared to be suitable map-

ping units. Where local aspect might significantly

affect the vegetation and therefore the management,

variations were distinguished within the major struc-

tural units. In our description the following letters

have been used: N= communities on slopes of

northerly aspects, S= communities on southerly

aspects, C= communities of crests and K= com-

munities of very steep kloofs. Relative positions of

the communities on Rooiberg are shown in Fig. 2.

Communities with northerly aspects

Community Nl. This community, comprising

Variations Nla, Nib, Cl, C2, is found on the

moderately steep upper and middle slopes

(1370-1060 m) to the north of the divide. The vegeta-

tion is a closed restioid and narrow-leaved shrubland

in which the former element is often predominant or

at least conspicuous. Variation Nla of the highlands

east of Rooiberg Peak is mainly restioid and about

0,5-1 m high (Fig. 3). Most of the precipitation in

this area comes from the condensation of clouds

formed by the summer south-east winds. The land is

privately owned and was probably burned fairly fre-

quently in the past.

Variation Cl (Fig. 4) occurs on summits and is

stunted because of its exposed site. It is sometimes

dominated by grasses. Many of the localities, too

small to be mapped separately, have been included in

Variation C2 which occurs along the crest and major

spurs at high altitude. This variation has a higher

species diversity and greater variety in dominance

and height of species due to many local differences in

aspect and degree of slope (Fig. 5). Since the major

fire belts are sited on ridges, the Variations Cl and

C2, like Nla, have been subject to more frequent and

regular burning than other parts of the reserve.

Variation Nib, less frequently burnt, represents

vegetation of northerly aspects west of Rooiberg

Peak and below Variations Cl and C2. Low narrow-

Divide

wv - restioids I- Protea nitida

1500

1000

j

Fig. 2.— Relative positions of

communities on Rooiberg.

H. C. TAYLOR AND F. VAN DER MEULEN

559

Fig. 4. — Variation Cl of Com-

munity N1 on the summit of

Rooiberg Peak. The conspi-

cuous grass is a Pentaschistis

species.

Fig. 5. — Variation C2 of Com-

munity N1 on the spur run-

ning north-west from Rooi-

berg Peak to Taays Rand.

The vegetation is restioid with

Protea eximia in the

right foreground. In and be-

yond the shadow is a patch

of dense proteoid shrubland

on a local southerly slope.

560

STRUCTURAL AND FLORISTIC CLASSIFICATIONS OF CAPE MOUNTAIN FYNBOS ON ROOIBERG,

SOUTHERN CAPE

leaved shrubs prevail, but at middle altitudes the

proteoid element ( Protea repens, P. eximia, P.

lorifolia) increases in height (up to 2 m) and cover (up

to 75%) (Fig. 6). Leucadendron salignum extends in

a broad belt across the middle slopes, giving them a

yellow-green speckled appearance. At the middle

altitudinal limit of this variation is a similar belt of

the grey-leaved small tree, Protea nitida (Waboom)

which, though less widespread than Leucadendron

salignum, often overlaps its range. This ‘Waboom-

veld’ is recognizeable on air photos but is too restric-

ted in area to map separately.

Community N2. Including Variations NE and

NW, this community occurs between about 1 300

and 650 m on steep northerly slopes beneath the

second crest (Fig. 7). These lower slopes receive less

cloud moisture and are therefore warmer and drier

than those of Variation Nla, north of the divide.

Narrow-leaved shrubs exceed restioids in cover while

proteoids are rare and of a more xeromorphic type,

like Protea lorifolia. The community contains a fair

number of succulents and grasses although their

cover is not high. Sites with north-east and north-

west aspects are separated as Variations NE and NW

respectively. The narrow-leaved shrub Passerina

vulgaris tends to dominate on both aspects and

Protea punctata occurs locally on upper slopes in

variation NE.

Community N3. This community is found between

1 060 and 900 m on the plateau or shelf on the nor-

thern side of the mountain below Variation Nib.

Slopes are gentle to almost level and very stony. The

vegetation, about one metre in height, is character-

ized by the predominance of narrow-leaved shrubs in

the upper canopy (Fig. 8). In lower layers, restioids,

grasses and succulents (mainly Mesembryanthema-

ceae) can be quite numerous.

At the western limit of the range or at the outer

edge of the plateau where less rain falls, elements of

Renosterveld, especially the Renosterbos (Elytropap-

pus rhinocerotis), become conspicuous.

Fig 6. — Variation Nib where the

proteoid cover reaches up to

75%: Protea lorifolia in the

foreground, P. eximia behind,

c. 1-2 m.

Fig. 7. — The second crest show-

ing Community N2 between

the skyline and the steep

cliffs. Variation NE the

darker patch at top left, and

Community KN on the preci-

pitous lower slopes.

J

H. C. TAYLOR AND F. VAN DER MEULEN

561

Fig. 8. — At Nicolaaskloof,

typical Community N3 is con-

fined to the sloping shelf in

the foreground. On the

rocky slopes beyond, N3

intergrades with proteoid

Nib above the figure’s head.

Fig. 9. — Community N4 on the

steep north-facing basal slope.

Community N4. This community occurs on steep

north-facing slopes between 900 and 700 m, below

Community N3 down to the base of the mountain.

The soils are very stony, shallow and dry. Like the

previous one, Community N4 is dominated by

narrow-leaved shrubs (Fig. 9) but the vegetation is

more open and poorer in species content, containing

some karroid elements and more succulents than

usual in fynbos because this community borders on

the karroid vegetation of the lowlands.

Community KN (including Community K). The

main drainage of upper Rooiberg comprises four

complex kloof systems that occupy the rift south of

the divide. In their upper reaches the kloofs run

parallel to the divide and then drain south. The north

side of the range is drained by five large kloofs run-

ning east or north-east and a series of shorter, shal-

lower kloofs running north. Community KN is found

at lower levels of all the southward draining kloofs.

The habitat is a very steep to precipitous rock slope

with some soil accumulated in pockets and bands bet-

ween the vertical cliffs (Figs 7 & 10). The kloofs

draining north or north-east have no distinct north or

south aspects, but the vegetation of their steep sides,

Community K, appears structurally similar to that of

Community KN. In both communities the canopy

cover is low (about 25-70%) and consists of scat-

tered shrubs with a rounded growth form 1-2 m in

diameter, interspersed with coarse restioids and

grasses, some broad-leaved bushes of tropical affini-

ty (e.g. Diospyros c/ichrophylla and Cussonia

spicata ) and succulents in the ground layer. Proteoids

are usually absent. The vegetation, like the habitat, is

very similar to N4.

Communities with southerly aspects

Community SI. This community, which includes

Variations SW and SE, is found on moderately steep

562

STRUCTURAL AND FLORISTIC CLASSIFICATIONS OF CAPE MOUNTAIN FYNBOS ON ROOIBERG,

SOUTHERN CAPE

to very steep southerly slopes from 1 490 m down to

about 630 m. The well drained, shallow soil consists

of fine grey humic sand with some leaf litter. At high

elevations, these slopes receive moisture condensed

from the clouds that form in the south-east wind,

therefore reducing the effect of the hot, dry sum-

mers. The vegetation is dense shrubland with a

closed, uniform canopy of proteoids up to 2 m high.

There is at least one lower layer of restioids and

narrow-leaved shrubs similar in structure to the

canopy of Community Nla.

Within Community SI there are several structural

variations, possibly resulting from differing proteoid

dominance: at the highest, steepest sites, Protea

punctata occurs in almost pure stands with some

Leucadendron comosum (Fig. 11); at middle

altitudes Leucadendron eucalyptifolium becomes

either dominant or co-dominant with Protea punc-

tata. At still lower elevations, the number of proteoid

species increases with the addition of Protea eximia

and P. neriifolia but their total cover decreases until

at the lowest parts the proteoid layer contributes

about 50% canopy cover and is only about 1 m high,

while the restioid and narrow-leaved layer becomes

closed. Since these structural variations either form

mosaics or intergrade with each other, they cannot be

mapped separately at the 1: 50 000 scale. Variations

SW and SE, with western and eastern aspects respec-

tively, are mappable units of the open proteoid com-

munity of lower altitudes down to about 400 m,

which include Protea repens, P. lorifolia and

Leucadendron salignum. Grasses are also more fre-

quent in these variations than in the other variations

of Community SI .

Community S2. This community was found only

in a small area on the steep upper south-easterly

slopes of the peak marked by T rigonometrical Survey

beacon No. 149. Stands on similar sites on the

highest peaks, which possibly belong to the same

community, have been affected by a fire belt.

Shallow, black, humic soil, more moist than other

Rooiberg soils, occurs between the bands of outcrop-

Fig. 10. — North-facing krans on

the left with vestiges of Com-

munity KN at its base. Gorge

of the south-flowing Bos-

rivier.

Fig. 11.— Tall dense proteoid

shrubland of Community SI

dominated by Protea punctata

with a fringe of P. eximia in

front, bordering on a fire-belt

in the foreground where

restioids and narrow-leaved

shrubs of the lower layer are

prominent.

TABLE 1.— Floristic classification of communities on Rooiberg (compare with Fig. 2)

x = occurring in close vicinity of releve

Leaf forms: B = broad-leaved sclerophylls; N = narrow-leaved leptophylls; (N) = slightly broader than N; G = graminoid;

P = proteoid; R = restioid (Restionaceae); (R) = restioid (other families); S = succulent; O = other forms.

See text for definitions.

x>

>

— m 04 co CO —

— ^tO(S CO —

— co 3 — i CO —

co-

rn vo "3- m oc

(N r- 3 m —

3 m 3 3 —

3 oo 3 m >n

+

03

>

_ m — — (J —

— 3 — t~~ (Jm

— m r~- O z — 03

— 3 — O Z — 3

m oc

3 m

m 04

vt M

3 O m

3 m 04

4 014

04

x

+

-I- + +

+ +

+ — -I- +

+ + + + +

04 o' m

04 m O

m — 04

40\0

4 VO ON

V, J 04

CJ 04

u-

o-

u —

m oo

m 3

m O

m O'

m 04

3 m O'

3 m vo

04 04 04

4010

3 r-~ 3

oo m oc V m

O' CO O 7 m

1040 Z <3

r- ( — r-~ Z m

oo m m Z <3

O' — 3 Z m

oo o O Z "4-

OOM41 Z 3

oo m m Z 3

•4 0 4MOO

mm 3 04 O'

04 m 4001^

04 O' 4 00 00

04 — 4 00 10

<N O' 4 r-~ m

mm 4mm

— 4 4 oo m

— 4 m

+ +

+ + +

+ + x — + +

+ + X

-I- +

+ + + +

— + — x

+

X

+ + +04

+ + + + + +

+

04 —

04

4/

XI

E

3

C

>.

’£

3

E

o

U

E

o

43

3

<

c/1

.E .2

2 1/5 n. i/i c

D.fc 3 «« 3 5

- * c E « 00

.2 5 .•= = fe s

C 13 >» ,

DC

: Q. 0-2

! Sibg.

i >. 3 co

i N UJ <

ea CO

-75 wo

3 o 2 •

1 3 iC 3 Q

1 " S'^z

3 -X' —

o" 3 •

« U D.

4 4 ”

i « E

4/ XC ,2

DC UJ 2

3 3

•— c/1

"o O

h= 3

-C 43

= CO

CO u-

3

— 0.43 ££

.2 £;S -

c CO C

« <u 3-2

c^o o o

5 E£2Eua,u[-

1 -6 -- 'E 3«

.S-g S>& «

8.2^1

.2 t- <u .x'x:

3 .

On Q-> • —

00 d. a-5

■5 C c« 4 C «

g- O c/1 <■> 3 =

c -a 3 w xC“ 22

- <u ~ 3

4/

««OpgS

3 3 &3-S.2

1 D d 3

O D.-3 ?

c*- _o 3 g

c/l -3 2

o 3 £ o- o

0.43 3 H 43

co-g «i m O

2.2

•r: T3 ^ 3 *p

« « §

3: >'C C<1 o

a °u<i

3 g

c .E

3 43

■Sjg 3 2

o<2r ^

*-• — O 3

3 -2 '-3 «

D, M 33

3 «

00 3

o o E

03^ X

^ ^ « h

ox: <u^

j e. oc z

3 3

o

3 -1-

O-

§■«

<L)

O (L)

<1> s—

a <->

cn

T3

.2 3

C co

O

t E

S 43

fo

.2 o-2.

o >% <u

o- 3 _

>H E

CU — 3 .

O gX 3

00 g O 3

3 — 33 NJ

43 3 3

c^c^xoi

&O

O O

c/icx.2

ZcfiZZZOw

a-OoiZZZazcuc^OcciO

zzoozaozz

ZZZOtflO

+ +

— + + —

— I- + +

— f

-I 1-

+ + +

-I h

+

— X

+ +

+

— +

m — + x

+ — ' 04 04 — 04

+ + X + X

1- +04

+ + + — + + +

m — + + — + +

+ + + - +-

+ + + X X +

04 + x +

— 04 + +

+ +

+ 04

+ + — + X + —

+

+ + + +

+ + +

+ +

04 +

+ 04 +

+ +

+ m +

+ +

04 +

X + + X + + +

+ + + + +

+ — +

m — + — -H — I — I

+ + H — I — I- + + + +

+ x + + + — +

+ 04 + — + +

+ + —

04 + +

+ + +

04 + X X

— + +

+ +04

+ + + X

X 04 X X +

- + +

+ — +

+ + - +

04 + 04 04

04 + + 04

04 + 1-

m m x

+ 04 04

+ X

+ +

— VO

04 -<t

04 m

— m

04 04

O r-

O OO

s

3

C->

-•g.

3 o

2*5

2 ^ c S

- C <D ~

u c o a

2 r S3

- C Q.

£ j- 60 co

~ C. C 3

CO

O

o.S o

-S= C -X

c'u <S

< ii. qj

OZczex

E^

O CO

ex a)

- o

O .0

c/l

o

E

o

CuflU

a u 3 C

- go 2

,2 3^-0

C C D. y

i3-S 3 S

u- 3 o 5

5 0 '

o ,i3

a. iS

o’o O — 1

<z>

T3 '

3 3

3D. -3 .2

?CL J

B. 3 3

3 ??.2.2

o u- 1-

nJ pj uj uj

f 3

CO .2

u- ^

a> ^

cxi:

O, <L>

Dh

O Cd _

=

0.3 o

> O.C—

C 3 ‘2

— O o

.2 3

s- ‘3 3

3 JE 4>

3 3

CO o

03 ^ ^

^2 rrt CO GOc^;

0-2 3. bo

- 3 ‘3 > .

<u

2

hZQ.

te E*2

’ *33 .5 cu

3^S^|.2

o 3 3 2 c 3

0.2.2 3 S E

t-.uun.rafl)

CU UJ UJ C7J U f-

0- D. Cu oC Z Z

£2 QC o CU cu z Z Z OC QC

2 0 go

Silt: ^

aS -

j « ^

: c 3«

[pi

! DC S co

c’S fe g

2 aot *-

3 <U 3 2 3

2 *- S) N - •

2 3 m 3

— <L> .2 ‘3 .O

3 o ??•— '

00 2 .0 O — ;

< Cu W E u I

3 3

w —

c« O

D ^

o D.

4/3 2

x <u g.;

S •£

Z QC Z Z Z CU Qi c?z Z Z Z QCZ

Ec/l

3

3 4/

3 C

— 4/

3 60 CO

CJ * —

D

co

O

.2 ^

c <*

03

c3 _

c3 •§. ^ ^ ^ P cd

) 6p Q, 3 i; t3 c ■

_6X).

>|i!

’o 03

c3

c

03 3 C g

6D 3: P C

Q. O

O 4/

U DC

.2

u ^ 4/ ert

O >>o » a D.-2

0 ^ £ t:

- - - o .O

12

_’o

3 c/i 3 "O

co *3 4/ n

° E E o.

5 0.2

.5 0 _

h u-h sz

•x 3 Z'w

~ D. _ —

E.2 § g CCx:c2

2 5-c

3 CO

3-S.al.g

-C C — co w

jz .2 x: t. 4/

UJ + Cu U DC

Q.JlhIU<U

0CZZZ2EZZOOEZOD:

<u

a

1/3 CO

C4-. '(D

o >

JJ

a3 <u

X) t-

goo

^ C3

C 03

60'S

.2 co

> CO

> (U

O —

2 s

^ 60

<U G

x: *g

w u.

■§ 3

< o

UWUlli

562

STRUCTURAL AND FLORISTIC CLASSIFICATIONS OF CAPE MOUNTAIN FYNBOS ON ROOIBERG,

SOUTHERN CAPE

to very steep southerly slopes from 1 490 m down to

about 630 m. The well drained, shallow soil consists

of fine grey humic sand with some leaf litter. At high

elevations, these slopes receive moisture condensed

from the clouds that form in the south-east wind,

therefore reducing the effect of the hot, dry sum-

mers. The vegetation is dense shrubland with a

closed, uniform canopy of proteoids up to 2 m high.

There is at least one lower layer of restioids and

narrow-leaved shrnhc drvuw = —

at the lowest parts the proteoid layer contributes

about 50% canopy cover and is only about 1 m high,

while the restioid and narrow-leaved layer becomes

closed. Since these structural variations either form

mosaics or intergrade with each other, they cannot be

mapped separately at the 1: 50 000 scale. Variations

SW and SE, with western and eastern aspects respec-

tively, are mappable units of the open proteoid com-

munitv of Inwpr J

front, bordering on a fire-belt

in the foreground where

restioids and narrow-leaved

shrubs of the lower layer are

prominent.

H. C. TAYLOR AND F. VAN DER MEULEN

565

Fig. 12. — The dense, narrow-

leaved shrubland Community

S2 on a cool, moist slope.

Fig. 13. — Scattered Waboom

(Protea nitida) barely dis-

cernible in the narrow-leaved

matrix of Community S3 on

the middle slopes above Buf-

felsfontein.

ping bedrock. The vegetation is a dense, narrow-

leaved shrubland up to 1,5 m tall in which species

indicating moist conditions, like Berzelia intermedia

and Psoralea pinnata, are prominent (Fig. 12). Near

the lower limit some tall proteoid and restioid com-

ponents ( Protea punctata and Cannamois virgata)

occur.

Community S3. This community has a wide distri-

bution from the third crest of the range down to the

foot of the mountain and along its entire length. The

gentle to moderate southerly slopes have dry, sandy

lithosols. In the west, the slope becomes steep, with

bands of kranses (cliffs) interspersed with colluvial

boulders. The vegetation varies in height and cover

but is generally less dense than that of Communities

SI and S2. Floristically and structurally, Community

S3 is probably the most diverse of all the Rooiberg

vegetation types. Leucadendron salignum, the only

constant feature, occurs virtually everywhere. Com-

munity S3a of the upper slopes between 915 and

about 760 m has a fairly dense mixed proteoid

vegetation with the highest concentration of proteoid

species in the reserve; the understorey is chiefly

restioid. A belt of Waboomveld, with Protea nitida

conspicuous, extends along the middle slopes from

about 760 m to 640 m (Community S3b); narrow-

leaved shrubs predominate and grasses are fairly fre-

quent (Fig. 13). Along the lowest slopes, especially to

the west, is a zone of karroid narrow-leaved shrubs

(e.g. Pteronia, Relhania) about 1,5 m high con-

stituting up to 75% canopy cover, in which the coarse

tufted grass Merxmuellera arundinacea is often con-

spicuous (Community S3c .)

2. Florist ic units

We have sampled the vegetation by means of

Braun-Blanquet releves (Table 1) varying in size from

the usual 50 m2 to about 100 nr, 200 nr or 300 m2

depending on differing vegetation structure. Some

units, like SI, are undersampled, the rare ones like

S2, KN & N2 are not sampled at all, while others

have been sampled more intensively because of their

special interest, e.g. the Protea mWo-dominated

Waboomveld which occurs in different floristic sub-

divisions throughout the range of Mountain Fynbos.

566

STRUCTURAL AND FLORISTIC CLASSIFICATIONS OF CAPE MOUNTAIN FYNBOS ON ROOIBERG,

SOUTHERN CAPE

TABLE 2. — Comparison of structural and floristic units

COMPARISON OF STRUCTURAL AND FLORISTIC UNITS

A comparison of the structural and floristic units is

summarized in Fig. 2 & Table 2.

Community /. Of the lower northern slopes this

community comprises all releves in structural units

N3 and N4. Predominant growth forms include

restioids (Restio fruticosus, Thamnochortus

argenteus, Hypodiscus striatus) and narrow-leaved

shrubs {Felicia filifolia, Passerina vulgaris). The first

three releves of this community are located in struc-

tural unit N4 which, it was noted, is poorer in species

than N3. In fact, Releves 475 and 485 have the lowest

species content of any in the survey.

Community II. This ‘Waboomveld’ community

represents stands of middle and lower slopes mainly

with southerly aspects (structural unit S3b)

characterized not only by Protea nitida but also by

Mohria caffrorum and Pelargonium myrrhifolium,

species that are frequent in Waboomveld over a wide

geographical range. Predominant life forms include

proteoids {Protea nitida, Leucadendron salignum),

narrow-leaved shrubs {Elytropappus glandulosus,

Anthospermum ciliare), restioids {Restio cuspidatus,

R. fruticosus) and grasses {Pentaschistis eriostoma,

Ihemeda triandra). At the footslopes of the moun-

tain, below the Waboomveld, karroid elements (e.g.

Pteronia and Relhania spp.) are found.

Releve 448 has no Protea nitida and few other dif-

ferential species. It would seem that within Com-

munity II the species-rich Waboom stands are, in

special situations, part of a wider-ranging vegetation

type that lacks Protea nitida and is in general less rich

in species content. This hypothesis might have impor-

tant management implications and can only be tested

by more sampling. Releve 442 is the only one in

Waboomveld on the northern side of the mountain.

It occurs at a considerably higher altitude than any of

the releves on the south side. Almost two decades ago

Taylor (1963) observed that Protea nitida ‘does seem

to occur at higher altitudes on the northern slopes’,

probably to compensate for the hotter and drier con-

ditions found there.

Community III. This community comprises releves

situated on summits and the crest fire-belt (structural

units Cl & C2), burnt just less than four years before

the survey. Predominant life forms include narrow-

leaved shrubs {Centella virgata, Aspalathus rubens,

Selago brevifolia) and restioids {Restio cuspidatus,

R. fruticosus). The three summit releves, Releves

474, 430 and 222 all show high scores for the grass

Pentaschistis eriostoma; Ehrharta ramosa and Pen-

taschistis colorata also occur in the community.

Releve 439 is particularly poor in differential species.

It may represent a related but undersampled com-

munity. Within Community III, this releve is the only

one not situated on the main crest and that has

vegetation over 1 m in height, suggesting that it may

lie outside the four-year old firebelt.

Community IV. This community comprises two

subcommunities, a and b, mainly distinguished from

each other by the presence in IVb of an overstorey of

proteoids {Protea punctata, Leucadendron spp.) and

j

H. C. TAYLOR AND F. VAN DER MEULEN

567

the fact that IVa has no differential species. Subcom-

munity IVb occurs on upper southern slopes bearing

the proteoid structural unit SI. In Subcommunity

IVa two releves (432, 434) are on upper north-facing

slopes bearing the restioid structural unit Nla; one

(405) is on a spur (C2) and one (472) is on a summit

(Cl). Therefore three structural units appear to be

combined in one floristic group. On the other hand,

one structural unit (Cl) has three releves (474, 430 &

222) in floristic Community III and one (472) in IVa.

The data are insufficient to determine whether these

anomalies are due to fire-age or other habitat factors.

They may merely show that structural units Nla, Cl

and C2 are not distinct entities.

Table 1 indicates that stands of lower and middle

slopes, bearing the structural units N3, N4 and S3

(floristic Communities I, I and II respectively) are

floristically more related to each other than to any of

the other stands. The same appears to apply to stands

of crests and upper slopes bearing the structural units

N1 and SI (floristic Communities III + IVa and IVb

respectively).

The results from the structural and floristic surveys

indicate that structure and species composition of the

vegetation are related to major habitat factors like

altitude and aspect.

CONCLUDING REMARKS

Both in the structural and the floristic surveys, we

have sampled only the major matrix of fynbos com-

munities and not the detailed patterns. Our results

are therefore provisional.

The floristic survey, even a brief one like that on

Rooiberg, supported and supplemented the struc-

tural survey. The latter can be done without extensive

a priori knowledge of the complex fynbos flora, and

has the advantage of being quicker and the data can

be used by personnel with little botanical training.

Emphasis is not on the plant species but rather on its

functional adaptation to the environment. The struc-

turally homogeneous units will be characterized by

particular life and growth forms and can be expected

to react in a reasonably uniform way to treatment

such as burning, and to use such as grazing by game.

The structural classification will therefore help

managers to delineate management units, to deter-

mine rotation lengths and stocking rates, to maintain

balanced, natural ecosystems and to monitor

physiognomic changes.

In the floristic survey, emphasis is on species com-

position of the plant cover, and the ecological

amplitudes of plant species are used to describe plant

communities and their interrelationships. The

floristic classification will indicate which species and

communities are threatened, and will provide infor-

mation on how to conserve them and how to main-

tain species diversity.

In conclusion, we think that a combination of a

structural and floristic survey, as we have described,

is a necessary and suitable basis for fynbos conserva-

tion management.

ACKNOWLEDGMENTS

For help in organizing and carrying out field work

we are grateful to Forester P. Weinberg of Ladismith

and a number of our colleagues and friends. We are

deeply indebted to Mr C. Boucher of the Botanical

Research Unit, Stellenbosch for his assistance in the

computing program for Table 1.

UITTREKSEL

Die struktuur en floristiese samestelling van die

plantegroei is gebruik vir die daarstelling van af-

sonderlike klassifikasies van plantgemeenskappe in

Bergfynbos op Rooiberg, Suid-Afrika. Die struk-

turele eenhede en floristiese assosiasies is sterk met

mekaar gekorreleer en hulle verspreiding weerspieel

die vernaamste omgewingsfaktore, nl. aspek en

hoogte bo seespieel. Ten spyte van die voorlopige

aard van die opname, is die gevolgtrekking dat

hulpbronopnames van hierdie aard ’n geskikte

grondslag vorm vir die bestuur van natuurlike

gebiede.

REFERENCES

Acocks, J. P. FI., 1975. Veld types of South Africa. 2nd edn.

Mem. bol. Surv. S. Afr. No. 40. 128 pp.

Taylor, H. C., 1963. A bird’s-eye view of Cape mountain vegeta-

tion. J. bot. Soc. S. Afr. 49: 17-19.

Taylor, H. C., 1979. Observations on the flora and phytogeo-

graphy of Rooiberg, a dry fynbos mountain in the southern

Cape Province, South Africa. Phytocoenologia 6: 524-531.

Taylor, H. C., in press. A reconnaissance of the vegetation of

Rooiberg State Forest. Department of Forestry, Technical

Bulletin.

■

Bothalia 13, 3 & 4: 569-576 (1981)

Miscellaneous ecological notes

VARIOUS AUTHORS

INVESTIGATION INTO THE SIGNIFICANCE OF PLANT DISPERSION IN ASSESSING PASTURE CONDITION

The description of pasture condition in terms of

percent cover and species composition is undesirable

as these measures are not the sole criteria. It has been

found, for example, at the Rietvlei Agricultural Re-

search Station, near Pretoria, that heavily fertilized

mown veld has a lower basal cover than unfertilized

grazed veld, yet produces more than three times the

amount of herbage than the unfertilized veld. The ul-

timate aim of pasture management is the mainte-

nance of a healthy sward and it is possible that the

measurement of the uniformity of plant cover may be

of value in establishing an objective concept of

pasture condition. It is intended, in this Note, to dis-

cuss various aspects of plant dispersion and their use

in interpreting certain phenomena. In this paper the

nature of the non-randomness is termed ‘overdis-

persed’ or ‘contagious’ when individuals tend to be

clumped together and ‘underdispersed’ or ‘regular’

when individuals are evenly scattered. It is stressed

that the terms overdispersed and underdispersed

refer to the distribution curves of data and not to the

pattern of plant individuals on the ground.

For the purpose of this investigation the Bruce

Levy point quadrat (Levy & Madden, 1933) was

used. Three bridges were made with 20 points each

spaced at 1 cm, 2 cm and 5 cm intervals. The bridges,

referred to in this paper as quadrats, were laid either

100 or 200 times at random in each experimental

plot, all of which were smaller than one hectare. A ‘t’

test showed that sufficient samples had been taken to

ensure an accuracy of 10 percent of the mean number

of strikes per bridge at p = 0,05. In one analysis a

bridge with 30 points at 5 cm intervals was used.

Random distributions may be described by the

Binomial distribution, or by the Normal and Poisson

approximations to that distribution. When the

frequency curves of plants occurring in quadrats are

compared with any one of these curves (depending on

the nature of the sample), it is noticed that there is

often a considerable discrepancy between the observed

and calculated values. Clapham (1936) and Black-

man (1942) compared frequency distributions of

species occurring in quadrats with the Poisson dis-

tribution and they found that many of the species

were not randomly dispersed. They expressed this

deviation from the expected distribution by compar-

ing the observed variance of the samples with the ex-

pected variance. Neyman (1939) and Thomas (1949)

expressed the principle of contagious distribution in

generalized forms of the Poisson distribution. The

various frequency distributions are described below.

The Binomial distribution

The frequencies of 0, 1, 2, . . . successes (strikes) in

N sets are given in terms of the binomial expansion of

N (p + q)n where N is the number of quadrats, p the

probability of an event occurring, q the probability

of the event not occurring and n the number of trials

(points per quadrat). An easy method of calculating

the successive terms in a binomial expansion is by

using the relationship:—

log (probability of x plants

occurring in any 1 quadrat) = log n! - log

(n-x)! - log x!

+ (n-x) log q

+ x log p.

In the Binomial distribution the mean is np and the

variance npq.

The Poisson distribution

This distribution is given by the series

probability of x plants _ mxe~m

occurring in any 1 quadrat x!

where m is the mean number of plants per quadrat.

In the Poisson distribution the mean is equal to the

variance. When m is very small the distribution is an

approximation to a Binomial distribution.

The Neyman contagion

This distribution is a generalization of the Poisson

distribution and is given by:

p(x = 0) = e_mi(l_e m2) and subsequent terms by

p(x = k + l) = m,m2e ^ ^ p (x = k - 1)

k + 1 t = 0 t!

where the parameters m, and m2 are proportional to

the mean number of groups per quadrat and mean

number of individuals per group respectively (Ney-

man, 1939). The parameters can be estimated from

the first and second moments of the distribution as

m2 — (/i 2 Mi)/ Mi and mj — Mi / m2

When m2 becomes very small and m,m2 is finite the

distribution approaches the Poisson where = 1 •

The Thomas distribution

This distribution is given by

p (k = 0) = e~m and subsequent terms by

D(k) = E mre-m (rX)k~r e~rX

r = 1 r! (k-r)!

The parameters m and 1 + X are the mean number of

groups per quadrat and mean number of individuals

per group respectively and are obtained from the first

and second moments of the distribution where

= m (1 + X) and /i2 = m (1 + 3X + X2)

As X becomes very small the distribution approaches

the Poisson (Thomas, 1949).

570

MISCELLANEOUS ECOLOGICAL NOTES

Fig. 1. — Frequency distributions of all plants compared with the expected Binomial distribution in a, plot 1; b, plot 2; c, plot

3; and d, plot 4 with 20 points per quadrat at 5 cm espacement and 100 quadrats; e, plot 1; and f, plot 3 with 10 points

per quadrat at 10 cm espacement and 100 quadrats.

The frequency distribution of all plants occurring

in plot 1 is given in Fig. la. The general appearance

of the veld is that it is uniform and in excellent condi-

tion. The hay cuts from this heavily fertilized plot are

more than three times those from unfertilized veld.

The veld initially was a Themeda triandra, Hypar-

rhenia hirta, Trachypogon spicatus type but, as a

result of the treatment, changed to Setaria

nigrirostris, S. perennis and Eragrostis chloromelas,

with relics of Themeda triandra and Hyparrhenia hir-

ta. An examination of the figure will show that the fit

of the observed values to the expected Binomial curve

is very close (Px2 = 0,8-0, 9). The V/npq ratio

(1,0104) is almost unity, signifying that the observed

and expected parameters show close agreement. In

plot 2, in which the treatment is the same as for plot

1, except that it is limed, the distribution also does

not differ significantly from the expected values (Fig.

lb) although the fit is not quite as good. The V/npq

ratio is 0,9985. Plots 3 and 4 were recently fertilized

and grazed and from Figs lc and Id it is seen that the

distributions of observed data points are distinctly

bimodal and differ significantly from the expected

Binomial distributions. From veld observations it

was evident that vegetation changes were occurring in

these plots. The initial effect of the fertilizer is that

VARIOUS AUTHORS

571

certain climax species are driven out and are replaced

by others lower in the succession, and it is during this

period that the productivity and vigour of certain

species is profoundly affected. The bimodality of the

distributions indicates that one is sampling a

heterogeneous population, consisting of areas of low

cover and others of high cover. The Neyman and

Thomas distributions are built on premises that (a)

the individuals in a cluster are randomly dispersed

and (b) the clusters in a population are randomly

dispersed. Because of the close agreement between

the observed and expected values obtained by Ar-

chibald (1948, 1950), Barnes & Stanbury (1951) and

in this paper, it may be accepted that these premises

are valid. The bimodal curves, which cannot be ade-

quately described by the Neyman and Thomas series,

may therefore be explained as follows: the clusters of

species which react unfavourably to high fertilizer

dressings die out leaving bare areas while those

species, normally randomly dispersed, which have

responded to the fertilizer are showing increased

vigour and a subsequent increase in number and size

and are becoming contagiously dispersed. For exam-

ple, it has been observed that tussocks of Themeda

triandra and Hyparrhenia hirta die out and only after

their death and the dissemination of their litter is the

space which was occupied by those tussocks coloniz-

ed by other species.

Corby (pers. comm.) showed that the further apart

the points are the closer the distribution of species

approached the Poisson values. Tidmarsh & Hav-

enga (1955) showed experimentally that when the in-

dividuals of a population are dispersed at random,

the distance between the points must exceed the aver-

age diameter of the individual to obtain a close fit to

the expected Binomial distribution. The data for Fig.

la-d were obtained from points spaced at 5 cm inter-

vals. The distributions for plots 1 and 3 have been

repeated in Figs le and If to show the effect of point

espacement. The points of the 5 cm bridge were

numbered from 1 to 20, and by recording each strike

against the point number it was later, by taking alter-

nate points, possible to examine and compare the dis-

tributions for the two espacements. The data presen-

ted in Fig. 2 are from a plot which has been moder-

ately grazed by sheep and cattle subsequent to an

accidental fire. Prior to the fire the veld had been

leniently treated and used mainly as a source of hay.

A comparison of the figures given for the 5 cm and

10 cm espacements in Fig. 1 shows that in plot 1 the

closer fit to the expected values is obtained when the

points are spaced at 10 cm, while the non-random

distribution given for 5 cm points in camp 3 becomes

random when the points are spaced 10 cm apart. In

Fig 2a it is seen that the distribution changes from a

highly contagious one for points at 1 cm to a

somewhat doubtful distribution for points at 2 cm to

a random dispersion with points at 5 cm. In the

distributions for the 1 cm espacement (Fig. 2b) there

is close agreement between the observed values and

the Neyman contagion, as one would expect. It is of

interest to note that in the distributions for the 2 cm

espacement (Fig. 2c), the Poisson values do not differ

significantly from the observed values, whereas the

Binomial series does not describe the distribution

adequately. Because the Poisson distribution is an

approximation of the Binomial when the probability

of an event occurring is small, it is felt that in this

case the Poisson values should be discarded and that

it should be accepted that the plants are contagiously

dispersed for that particular espacement. The total

number of strikes for the three espacements show a

fairly close agreement, being 619, 610 and 587 for the

Fig. 2. — a, Comparison of frequency distributions of plants occur-

ring in 200 quadrats, each with 20 variously spaced points,

with the expected Binomial curve for 5 cm espacement; b,

comparison of the Binomial, Poisson and Neyman series with

the actual distribution of plants occurring in 200 quadrats

each with 20 points spaced 1 cm apart; and c, of 200 quadrats

each with 20 points spaced 2 cm apart.

1 cm, 2 cm and 5 cm espacements respectively. A ‘t’

test showed that the differences between these figures

were not significant. It is the writer’s contention that

a cover made up of evenly dispersed small individuals

is more desirable than one of large individuals and

correspondingly large bare areas between the individ-

uals, which causes the soil to be more erodable. It is

probable that the competition for light and water is

less when the dispersion of the plants is uniform.

Where clusters are present, those individuals on the

periphery of the clusters receive the greater light and

water benefits.

Blackman (1935, 1942), Pidgeon & Ashby (1940),

Archibald (1948, 1950) and Clapham (1936) have all

572

MISCELLANEOUS ECOLOGICAL NOTES

Fig. 3. — Frequency distributions of individual species from plot 5 compared with the poisson, Neyman and Thomas distribu-

tions from 100 30-point quadrats with points spaced 5 cm apart; a, Trachypogon spicalus; b, Schyzachyrium sanguineum ;

c, Tristachya !eucothrix\ and d, Eragrostis capensis.

shown that numbers of species in a plant community

are not randomly dispersed and that the more abun-

dant a species is, the greater is the degree of aggrega-

tion present. Ashby (1948) maintains that it is possi-

ble that species, without a biological predisposition

to overdispersion, are randomly dispersed when they

first occupy an area. Barnes & Stanbury (1951) found

that as the vegetation developed, through spreading

of ‘islands’ the distribution becomes highly con-

tagious. If it is accepted that initial colonization is

random and that the randomly distributed species

later become contagiously distributed, it is

reasonable to suppose that subsequent development

of the vegetation will take place in, what may be

termed as a series of waves of randomness, contagion

and a subsequent randomness as the species is re-

placed. This process would continue until some state

of equilibrium is reached where the climax species are

contagiously dispersed while the pioneer relics are

randomly distributed. Should the succession be

driven back, then the reverse process would take

place with the pioneer species tending towards con-

tagion and ultimately the climax species would be

randomly dispersed. From a series of analyses made

at the Rietvlei Research Station, there is evidence

which supports the argument propounded above.

Fig. 3 gives four frequency distributions from plot 5

which is rotationally grazed. From observations, this

plot may be regarded as climax grassland and is in

good ‘health’. A bridge with 30 points spaced at 5 cm

intervals was laid 100 times at random in the camp.

The V/npq ratio is 0,9525 and x2 probability lies be-

tween 0,8 and 0,9. It will be noted that the distribu-

tions for Trachypogon spicatus, Schizachyrium

sangiuneum and Tristachya leucothrix show a closer

agreement to the Neyman and Thomas series than to

the Poisson distribution. (When the probability of a

plant being struck is very small (i.e. when np is small)

then the Poisson series may be used to describe the

distribution, however, when np becomes larger it is

better to use the Binomial distribution). Eragrostis

capensis, a relic pioneer is distributed at random, as

are many other rare species such as Brachiaria serrata

and Eragrostis racemosa (not plotted). The data for

the four species given in Fig. 4 are from plot 3. It will

be remembered that the distribution of all individuals

in this plot was not random for points spaced at 5 cm

intervals (Fig. lc). Heteropogon contortus and

Setaria perennis may be regarded as species near the

climax stage and are overdispersed. E. capensis and

E. racemosa have distributions approaching ran-

domness but are, however, also overdispersed. These

VARIOUS AUTHORS

573

Fig. 4.— Frequency distributions of individual species from plot 3 compared with the Poisson, Neyman and Thomas distribu-

tions from 100 30-point quadrats with points 5 cm apart; a, Setaria perennis ; b, Eragrostis capensis ; c, Eragrostis

racemoscr, and d, Heteropogon contortus.

data suggest that both climax and pioneer species are

contagiously dispersed. It has been shown by the

writer that some species in which the frequency

distribution cannot adequately be described by the

Neyman or Thomas series, a better fit can be obtain-

ed if the points are moved further apart. It is quite

possible that Setaria perennis and Heteropogon con-

tortus would give a better fit to the contagious series

if the points were moved apart. However, with the

bridge used for these analysis, it was not possible to

use 10 points at 10 cm intervals as there would be too

few points to give an adequate representation of the

dispersion. It is nevertheless obvious that the

distribution is not random for these two species.

From Figs 3 & 4 it appears as if the dispersion of the

species can make a considerable contribution to the

interpretation of botanical analyses, and the assess-

ment of the condition of the sward. In plot 5, in

which all plants are randomly dispersed, it is noticed

that certain climax grasses such as Sehizachyrium

sanguineum, Tristachya leucothrix and Themeda

triandra are not randomly distributed, whereas

grasses near the pioneer stage such as Eragrostis

capensis, Brachiaria serrata and Eragrostis racemosa

are randomly distributed. The distributions for these

last two grasses are not given, but the data indicate

that they are randomly dispersed. In plot 3, in which

the distribution of all plants is not random, it is

found that both climax species and pioneer species

are overdispersed.

This paper may be regarded as a preliminary report

of certain points which came to notice during routine

botanical analyses of grazing experiments on the

Rietvlei Agricultural Research Station. It is felt that a

deeper insight into the mathematical distribution of

species and of individuals will throw light on the

mechanism of vegetation changes and the direction in

which these changes are proceeding. It is possible

that the following points will indicate stability: (a)

the frequency distribution of all plants is random

when the points are closely spaced, the criterion of

spacement being determined by the average size of

the plant and the density of the cover; (b) the climax

species are contagiously and the pioneer species ran-

domly dispersed. When these conditions are not ful-

filled, then there are indications that vegetation

changes are occurring and a concept of the magni-

tude of these changes may be obtained by comparing

the observed parameters of the population with

theoretical parameters.

REFERENCES

Archibald, E. E. A., 1948. Plant populations. I. A new applica-

tion of Neyman’s contagious distribution. Ann. Bot. 12:

221-235.

Archibald, E. E. A., 1950. Plant populations. II. The estimation

of the number of individuals per unit area of species in

heterogeneous plant populations. Ann. Bot. 14: 7-21.

Ashby, E., 1948. Statistical ecology. II. A reassessment. Bot. Rev.

14: 222-234.

574

MISCELLANEOUS ECOLOGICAL NOTES

Barnes, H. & Stanbury, F. A., 1951. A statistical study of plant

distribution during the colonization and early development of

vegetation on china clay residues. J. Ecol. 39: 171-181.

Blackman, G. E., 1935. A study by statistical methods of the

distribution of species in grassland associations. Ann. Bot.

49: 749-777.

Blackman, G. E., 1942. Statistical and ecological studies in the

distribution of species in plant communities. I. Dispersion as

a factor in the study of changes in plant populations. Ann.

Bot. 6: 351-370.

Clapham, A. R., 1936. Over-dispersion in grassland and com-

munities and the use of statistical methods in plant ecology.

J. Ecol. 24: 232-251.

Levy, E. B. & Madden, E. A., 1933. The point method of pasture

analysis. N. Z. J. Agric. 46: 267-279.

Neyman, J., 1939. On a new class of ‘contagious’ distributions ap-

plicable in entomology and bacteriology. Ann. Math. Star.

10: 35-57.

Pidgeon, I. M. & Ashby, E., 1940. Studies in applied ecology. I. A

statistical analysis of regeneration following protection from

grazing. Proc. Linn. Soc. N. S. W. 65: 123-143.

Thomas, Marjorie, 1949. A generalization of Poisson’s Binomial

limit for use in ecology. Biometrika 36: 18-25.

Tidmarsh, C. E. M. & Havenga, C. M., 1955. The wheel-point

method of survey and measurement of semi-open grasslands

and karoo vegetation in South Africa. Mem. bot. Surv.

S. Afr. 29.

H. H. von Broembsen*

*This and the following ecological note were written by the late H.

H. von Broembsen of the Botanical Research Institute, Pretoria,

who died in 1966. The notes were prepared for publication by J.

W. Morris of the same Institute. Address: Private Bag. XI 16,

Pretoria, 0001.

A SIMPLE METHOD FOR DETERMINING THE DENSITY OF PLANTS IN A RANDOMLY-DISPERSED POPULATION

It is sometimes necessary to determine the densities

of plants in experimental plots which have been sown

to some crop or other. Where the seed has been sown

in rows, it is a simple matter to count the number of

plants growing in unit lengths of row, and then com-

pute the total number of plants per plot. Where the

seed, however, has been broadcast the method in

which the density of plants, or the total number of

plants per plot, is determined, is different.

When seed is broadcast over a plot, the manner in

which the seed falls to the ground is a random pro-

cess, and the dispersion of the seed over the plot may

be described by the Poisson distribution (Feller,

1957). The Poisson distribution is given by

P(k plants growing on unit area a) = dake~da

k!

where d is the density of plants per unit area. The

probability of zero plants growing in any unit area is

then e~da. Now, if n quadrats are placed at random in

a plot, and it is found that y quadrats contain no

plants then

y = ne_da

and d = '°g n-'°g l (1)

a log e

which means than the density of plants may be com-

puted from a knowledge of the total number of

quadrats which are used for sampling and the

number of quadrats containing no plants. By using

this relationship the necessity for counting the

number of individuals in the quadrats is eliminated.

When using the relationship (1) care must be taken

to choose a quadrat size which will not be, (a) so

large that it always contains plants, or, (b) so small

that the number of empty quadrats is too large. It is

recommended that a rough estimate of the average

density of the individuals first be made over all treat-

ments in the experiment. The size of quadrat which

will give the desired percentage of empty quadrats

may then be computed from

a' = log 100 -log y"

d'log e

where a is the area of the quadrat, y the desired

percentage of empty quadrats and d the estimate of

average density. It is recommended that y equal 50 to

60 percent. With y equal to 50 percent, 34 percent of

the quadrats will contain one plant, 12 percent two

plants, and four percent more than two plants.

In order to determine the number of quadrats

which are required to detect predetermined signifi-

cant differences in densities the following relation-

ship is used:

t - Xl~ x2 Vn~~ (2)

a/v

Since the variance (v) of the Poisson distribution is

equal to the mean (da), equation (2) may be re-

written in the form

n = . 2t_2.da_

(d0a)2

where d0 is the difference in density you wish to

regard as being significant, and t = 1 ,96 for P = 0,05,

or 2,58 for P = 0,01.

The estimates of density should be transformed,

using a square root transformation, before being

subjected to an analysis of variance.

REFERENCE

Feller, W., 1957. An introduction to probability theory and its

applications. New York: Wiley.

H. H. von Broembsen

A NOTE ON THE EXTENSION OF THE DEGREE REFERENCE SYSTEM FOR CITING BIOLOGICAL DISTRIBUTION RECORDS

TO NORTH OF EQUATOR AND WEST OF GREENWICH MERIDIAN

The Degree Reference System proposed by

Edwards & Jessop (1967) and Edwards & Leistner

(1971) has been in general use in South Africa for

well over ten years. Attention has recently been

drawn to a requirement for extending the System for

plotting the distribution of plants for the whole of

Africa and for surrounding islands such as Tristan da

Cunha and Marion Island. This means extension to

both north of the Equator and west of the Greenwich

Meridian, amounting to extension to cover the rest of

the earth’s surface. Although such an extension was

implied by Edwards & Leistner (1971), no formal

conventions for doing so were proposed by them.

This Note gives a simple procedure for differen-

VARIOUS AUTHORS

575

Fig. 5. — Example of numbering

for different quadrants north

or south of Equator and east

or west of Greenwich Meri-

dian.

0°

tiating, when required, between the different

quadrants of the earth’s surface.

As reported by Edwards & Leistner (1971), since

we happen in Africa south of the Equator to be in the

south-eastern quarter of the earth’s surface in rela-

tion to the intersection of 0° latitude and 0°

longitude, the degree square is numbered according

to the degrees of latitude and longitude, in that

order, of the top left hand or north-west corner of

the degree square. By extending this principle, degree

squares in the north-eastern, north-western and

south-western quarters of the earth’s surface may be

numbered in relation to the point of origin in a way

similar to that of the south-eastern quarter, provided

the various quarters are identified by prefixing the

letters NE, NW, SW and SE to the degree numbers.

In other words, the degree squares are numbered just

as the latitude and longitude are numbered on a map,

with the addition of prefixes indicating N or S of the

equator and E or W of Greenwich meridian. This is

illustrated in Fig. 5.

If required, subdivision of the degree square is

then carried out in exactly the same way as previous-

ly: by subdividing into half-degree squares numbered

A, B, C and D, respectively, and each half-degree

square being further subdivided into quarter-degree

squares, again numbered A, B, C, and D, respective-

ly. This is also illustrated in Fig. 1.

In general, use of the Degree Reference System is

therefore similar throughout the world, latitude and

longitude numbers being used as found on a map,

but with the prefixing of the letters N or S and E or

W to indicate, respectively, north or south latitude,

and east or west longitude. Subdivisions of a degree

square are then made by successive division into half-

and quarter-degree squares, which are indicated by

the letters A, B, C, and D.

REFERENCES

Edwards, D. & Jessop, J. P., 1967. Proposed replacement of

magisterial districts by a latitude-longitude grid system for

plant distribution by the Botanical Research Institute. S. Afr.

Forum Botanicum 5: 1-5.

Edwards, D. & Leistner', O. A., 1971. A degree reference system

for citing biological records in southern Africa. Mitt. bot.

StSamml., Munch. 10: 501-509.

D. Edwards

DETERMINATION OF PLOT SIZE

In most vegetation studies a plot technique is used

for sampling. In many studies it is required that each

plot must be large enough to contain the ‘characteris-

tic structure and floristics’ of the phytocoenose. The

determination of plot size has occupied the attention

of many plant ecologists; the general conclusion be-

ing that ‘an objective method of plot size determina-

tion seems impossible’ (Werger, 1972). Werger (1972)

and Moravec (1973) have reviewed these studies.

In this paper a regression equation relating optimal

plot size (Werger, 1972) to easily measured vegeta-

tion characteristics is constructed. If this regression

equation could be improved and tested more widely it

could form the basis of a rule-of thumb method to

estimate optimal plot size in the field.

The methods of Werger (1972) were used to deter-

mine the optimal plot size of 32 phytocoenoses.

Werger’s definition of optimal plot size is that size

which contains a specified percentage of the number

of species calculated to occur in one hectare.

Therefore if there are 80 species in one hectare (100%

information), and if 50% information is required,

then optimal plot size is that area which contains 40

species. Werger (1973) used 50-55% information in

his phytosociological studies in South Africa. The

regression equation

y = a + b loge x

(Gleason, 1925; Goodall, 1952) was used in the pre-

sent study and by Werger (1972), to calculate the

number of species in one hectare of the sampled

vegetation (y is the number of species in area x, and a

and b are constants that are calculated with each set

of x, y values from every phytocoenose). The data to

576

MISCELLANEOUS ECOLOGICAL NOTES

calculate optimal plot size came from nested circular

plots of sizes varying between 0,8 m2 and 1256 m2 in

12 of the 32 phytocoenoses, and from nested rec-

tangular plots of sizes varying between 1 m2 and 256

m2 in the remaining phytocoenoses.

For each of the 32 phytocoenoses the following

vegetation characteristics were recorded:

(1) H = maximum height (m) of the tallest stratum.

(2) T = total projected cover (recorded in 10 classes

1 = 1-10% cover, 2= 11-20% cover, . . . .and so

on 10 = 91-100% cover).

(3) C = projected cover of the tallest stratum

(recorded as for T).

(4) Number of species in 200 m2 (usually determined

by extrapolation of the species-area regression

equation).

(5) Number of strata (for this particular vegetation

characteristic a stratum had to have more than

25% cover to qualify as a stratum).

(6) % species in the tallest stratum.

The phytocoenoses that have been used cover a

range of vegetation types in South Africa, chiefly the

semi-desert vegetation of the Karoo (n = 10) and the

fynbos of the Cape winter-rainfall region (n=17).

Also represented are grassland (n = 2), and woodland

and forest (n = 3). Fifteen of the phytocoenoses used

are from Werger (1972).

Vegetation structure has long been know to affect

plot size (refer to Werger, 1972, for further discus-

sion). Therefore it is not surprising that height of the

tallest stratum (H), total cover (T), and cover of the

tallest stratum (C), are all significant (p< 0,005) in

the regression equation relating optimal plot size to

vegetation characteristics.

Vegetation height is usually positively correlated

with average crown diameter. The larger the

diameter, the larger is the average interplant spacing;

therefore height of vegetation is positively correlated

with optimal plot size.

Because of the greater interplant spacing in vegeta-

tion with low total cover, or in vegetation in which

the tallest strata have low cover, optimal plot size is

negatively correlated with these vegetation

characteristics.

Of the various possible regressions relating optimal

plot size to vegetation characteristics, the most suc-

cessful were the following (various transformations

of the original variables were attempted but proved

less successful):

For 40% information per plot (r2 = 0,51):

P = 32,4 + (2,0) (H) - (1,5) (T) - (1,5) (C)

For 50% information per plot (r2 = 0,50):

P = 80,7 + (4,6) (H) - (3,3) (T) - (3,5) (C)

For 55% information per plot (r2 = 0,48):

P = 127,4 + (6,8) (H) - (4,8) (T) - (5,3) (C)

For 60% information per plot (r2 = 0,46):

P = 199,8 + (10,0) (H) - (6,6) (T) - (8,1) (C)

where P is the optimal plot size (m2) and H, T and C

are as given above.

The r2 values indicate that only half of the varia-

tion in the calculated values of optimal plot size is ac-

counted for by the three vegetation characteristics in

the regressions. The r2 values could probably be im-

proved if more sophisticated vegetation characteris-

tics could be used (e.g. direct measures of crown

diameter, crown overlap, and cover by vegetation

strata). Much of the variation in optimal plot size

which is not accounted for may be due to the errors

that arose when determining the vegetation charac-

teristics of the 15 phytocoenoses reported by Werger

(1972). We had to glean our information for these

phytocoenoses from Werger (1972) and Werger

(1973); we were unable to do any field recordings for

his work. We would also suspect an improvement in

the r2 values if more than one species-area curve had

been constructed in each phytocoenose.

It is difficult to explain the lack of correlation be-

tween optimal plot size and floristic richness (here

measured as number of species in 200 m2-P>0,4).

One would have suspected a positive correlation

(Werger, 1972).

Choice of plot size is often dictated by success or

failure of previously used plot sizes. Therefore, for

example, in regions with a long history of phyto-

sociological research plot sizes that are suitable are

listed by vegetation type (e.g. Mueller-Dombois &

Ellenberg, 1974: p. 48). In the numerous regions

where phytosociological studies are yet to begin, a

regression equation that relates optimal plot size to

vegetation characteristics could be of use. Therefore

if the regressions of the type reported in this paper

could be improved, as we feel they could, a phyto-

sociologist could calculate optimal plot size by deter-

mining a few, easily measured vegetation characteris-

tics.

We thank the C.S.I.R. for funding this project,

and Dr M. J. A. Werger for valuable comments.

REFERENCES

Gleason, H. A., 1925. Species and area. Ecology 6: 66-74.

Goodall, D. W., 1952. Quantitative aspects of plant distribution.

Biol. Rev. 27: 194-245.

Moravec, J., 1973. The determination of the minimal area of

phytocoenoses. Folia Geobot. Phytotax. Praha 8: 23—47.

Mueller-Dombois, D. & Ellenberg, H., 1974. Aims and

methods of vegetation ecology. New York: John Wiley &

Sons.

Werger, M. J. A., 1972. Species-area relationship and plot size:

with some examples from South African vegetation. Bothalia

10: 583-594.

Werger, M. J. A., 1973. Phytosociology of the upper Orange

River Valley, South Africa. Ph.D. thesis, University of

Nijmegen.

B. M. Campbell* and E. J. Moll**

*Botanical Research Unit, P.O. Box 471, Stellenbosch, 7600.

♦♦Department of Botany, University of Cape Town, Rondebosch,

7700.

REVIEW OF THE WORK OF THE BOTANICAL RESEARCH INSTITUTE, 1979/1980

CONTENTS

Introduction 577

Reports of sections 577

Staff list 585

Publications by the staff 588

INTRODUCTION

The need for botanical research to supply answers

to problems in connection with the management,

optimal utilization and conservation of the plants

and vegetation cover of southern Africa is becoming

more evident as environmental problems increase.

For this reason a greater concentration of manpower

on taxonomic and ecologicl research has proved to be

necessary. In the ecological field priority areas

receiving much attention were the fynbos of the

winter rainfall area of the Cape and the coastal

aquatic and adjacent terrestrial habitats. The diffi-

cult field of rootgrowth research yielded some

interesting results but will need a great deal of further

effort. More research on ecophysiology and the

dynamics of vegetation have become essentials for

progress.

The venture by the Department into palaeo-

botanical research, even though a new departure, is a

natural extension of the series of botanical inven-

tories for different plant groups represented by the

Flora of Southern Africa.

A highlight of the year was the declaration of the

Pretoria Botanic Garden as a National Monument,

an indication that it is recognized as a valuable asset

as well as an indispensable part of the research

organization of the Department.

This review covers a nine-month period, from 1

July, 1979 to 31 March, 1980.

REPORTS OF THE SECTIONS

HERBARIUM SERVICES SECTION

The four herbaria of the Institute continued to

identify plants and provide information for a wide

range of people including officers of the Institute,

various State and Provincial Departments, univer-

sities and the public both in South Africa and neigh-

bouring states.

National Herbarium, Pretoria (PRE)

A total of 13 285 specimens was named and 430

visitors dealt with. Accessions to the herbarium

numbered 19 377. During the year 37 loans (7 498

specimens) were sent out to other institutes and 18

loans (1 820 specimens) were received. We received

1 055 specimens on exchange, but did not distribute

any duplicates during the period.

Due to the lack of rain in the western half of the

country, no major expeditions were undertaken.

Several minor collecting trips were made in the

eastern Transvaal, northern Natal and eastern

Orange Free State. Problems were again experienced

with organizing a major expedition to Lesotho. This

was again cancelled at the last moment.

A reorganization of the herbarium (Fig. 1) took

place with the establishment of a separate crypto-

gamic herbarium. The collections were moved to

accommodation vacated by workshops in the base-

ment. The establishment of a fossil herbarium in the

garages under the Library has been necessitated by

the rapidly expanding activities and collections in the

palaeobotanical field. Our modular steel cabinets

were redesigned internally to accept 40 removable

trays to carry the fossil specimens.

Among the numerous visitors who came to consult

the collections and staff were the following: Dr Fred

Hoener (Lesotho: American Peace Corps), Prof. D.

and Dr U. Miiller-Doblies (Berlin), Dr Juliet Prior

(London: Swaziland Archaeological Association),

Mr B.K. Simon (Queensland, Australia), Dr. M.

Schrieber (Regensburg), Mr R.B. Drummond and

Mr L.C. Leach (Salisbury).

Wing A: Mrs E. van Hoepen continues to control the

Wing in a part-time capacity while controlling the

information and identification service.

Miss C. Reid is co-operating with Mr T. H. Arnold

in the preparation of a preliminary revision of Ficinia

(Cyperaceae), which is required for ecological

research in the Fynbos Biome Project.

Miss L. Smook and Dr G. E. Gibbs Russell are

continuing with sorting and re-evaluating the data

bank records for Poaceae.

A fair amount of expansion was possible in the

Wing due to the removal of the cryptogamic collec-

tions.

Wing B: Mr G. Germishuizen is nearing completion

of his revision of the southern African members of

Polygonaceae, which will be submitted as an M.Sc.

thesis to the University of Pretoria. He has com-

pleted the botanical text for a book on medicinal and

edible plants and is currently doing the text for an

illustrated book on the Wild Flowers of the Trans-

vaal with paintings by the artist Anita Fabian.

Mr D. A. Davies was transferred to the regional

herbarium in Stellenbosch to assist with the her-

barium work. His place was taken by Mr B. D.

Schrire who will be transferred to Durban later in

1980.

Wing C : Miss E. Retief is continuing her revision of

Campanulaceae. She has been involed with Mr P.

P. J. Herman in a survey of the seed samples in the

seed collection using the scanning electron micro-

578

scope. She presented a poster paper on the subject in

conjunction with Mr Herman at the S.A.A.B. Con-

gress in Pietermaritzburg in January.

Mr Herman is nearing the end of his work on

Pavetta.

Mr C. Hildyard completed successfully his first

year courses for his B.Sc. degree at the University of

Pretoria.

Wing D: Mr G. J. Goosen left at the beginning of the

year to take up a post with Nature Conservation at

the Etosha Pan Game Reserve.

Miss W. G. Welman continues as the regional

abstracter for Excerpta Botanica (Taxonomica).

Cryptogams : The establishment of the cryptogamic

herbarium in the basement is very welcome. The

change-over to the vertical packeting system and the

roll-out card index cabinets has been a great improve-

ment in respect to space saving and ease of handling.

The repacketing of the mosses and liverworts is now

complete.

A total of 2 971 specimens was identified and

3 604 were added to the collection. Only 8 specimens

were sent out on loan. Exchanges amounted to 3 038

sent out and 484 received.

Dr R. E. Magill attended the International

Bryological Congress in Geneva in September chair-

ing several sessions. Thereafter he visited Paris, Kew

and the British Museum in search of types. He

recently completed his treatment of 200 species for

the Flora.

Mr J. van Rooy is continuing with his revision of

Bryum for the Flora. He should obtain his B.Sc.

degree at the end of 1980.

Mr F. A. Brusse joined the herbarium in February

to work on Lichens. His M.Sc. thesis on Xantho-

parmelia has been submitted to the University of the

Witwatersrand. He is presently checking the whole

collection and supervising its repacketing and re-

organization.

Natal Herbarium, Durban ( NH)

A total of 1 318 specimens was named and 725

visitors dealt with. Among the latter were student

and school groups. Accessions to the herbarium

numbered 1 980. Some 1 667 specimens in 5 batches

were sent out on loan to various institutes.

Mr P. V. C. du Toit left the herbarium at the end

of the period on transfer back to the Pasture

Research Section of Natal Region of the Department

at Cedara. He joined the staff of the Institute in 1972

and took charge of the Natal Herbarium in 1976.

Mrs B. J. Pienaar returned to the herbarium after

a 6 months training period in Pretoria.

Miss A. Wright left in January to continue her

studies at University. She was succeeded by Mrs L.

Nichols who for a number of years had worked in the

Compton Herbarium, Kirstenbosch.

Albany Museum Herbarium, Grahamstown ( GRA )

A total of 2 012 specimens was named and 522

visitors dealt with. Accessions to the herbarium

numbered 1 276. Some 381 specimens in 15 batches

were sent out on loan to various institutes.

Among the visitors who came to the herbarium

were Prof. D. Miiller-Doblies (Berlin), Dr C. Vosa

(Oxford). Dr Sylvia Earle (California) and Dr W.

Farnham (Portsmouth, England) consulted the

Pocock algal collection.

Mrs E. Brink continues to run the herbarium with

Miss G. V. Britten, now in a morning post, to assist

her. No one has been appointed to the vacant techni-

cal assistant post.

A brief training course has been instituted for the

African assistants in the local section of the Provin-

cial Department of Nature Conservation. This has

been done mainly by Alfred Booi, the herbarium’s

African assistant, who produced a training manual in

the vernacular for the purpose.

Government Herbarium, Stellenbosch ( STE)

The number of specimens named totalled 3 210

with 178 visitors requiring information. Accessions

to the herbarium numbered 1 204. In all 21 loans of

2 672 specimens were sent out.

Mrs M. F. Rand left the herbarium at the end of

December for maternity reasons. She started work in

the herbarium in January 1965 and much of the

credit for the building up of the herbarium to its pre-

sent state must go to her. She hopes to continue her

research on Hypoxidaceae at the Compton Her-

barium, Kirstenbosch.

Miss L. Hugo assumed duty as Curator. She will

soon begin revisionary work on Tetraria

(Cyperaceae). Mr D. A. Davies was transferred down

from Pretoria in February to assist with the curating

and identification service. Miss M. Schonken began

work in the herbarium in March. She is currently

completing her research on a group of species in

Pelargonium for her M.Sc. degree at the university of

Stellenbosch. The increase in staff is as a result of the

increasing demands on the identification service

made by the Fynbos Ecosystem Project of the CSIR.

A quick turnover in technical assistants has taken

place during the last few years. The present incum-

bent Mrs R. Wikner is proving invaluable.

S.W.A. Herbarium, Windhoek (WIND)

This herbarium which was a regional herbarium of

the Botanical Research Institute from 1954, was

eventually transferred to the Administration of

S.W.A. /Namibia. It is now an independent her-

barium in southern Africa with Mr M. A. N. Muller

continuing as Curator. We wish him and his staff all

of the best for the future.

FLORA RESEARCH SECTION

Flora of Southern Africa

Two parts of the Flora were published. The first is

Vol. 10, part 1 dealing with 54 species of the families

Loranthaceae and Viscaceae. Line drawings and dis-

tribution maps are provided for most species. The

work was done by Prof. D. Wiens of the University

of Utah, U.S.A. and Dr H. R. Tolken, formerly of

this Institute and now at the State Herbarium,

Adelaide, Australia. The second is Vol. 27, part 4

which treats 130 species of the genera Brachystelma,

Ceropegia and Riocreuxia in the family

Asclepiadaceae. The work was done by Dr R. A.

Dyer who recently retired for the second time from

the Institute (his first retirement was in 1963).

Two volumes in the series on Cryptogams are pro-

gressing well: Vol. 14 (Bryophyta): Fascicle 14,1,1

covering the families Sphagnaceae to Grimmiaceae

has been submitted to the editor by Dr R. E. Magill.

It deals with 190 species, each of which is accom-

panied by a detailed pencil drawing and a distribu-

tion map.

579

Fig. 1. — The main building of the Botanical Research Institute, Pretoria.

Vol. 15 (Pteridophyta): The 250 species of ferns are

being written up by Prof. E. A. Schelpe of the

Bolus Herbarium of the University of Cape

Town. Texts of about 80 species have been sub-

mitted to the editor for criticism and the work

should be completed by the end of 1981.

In the series on Flowering Plants the following

volumes or parts thereof are at a final stage of

preparation or have been submitted to the editor:

Vol. 3: Dr P. J. Vorster of the University of Stellen-

bosch put further finishing touches to his com-

pleted thesis on the 65 species of Mariscus and

related genera.

Vol. 4: Part 2 on 42 species of the families Maya-

caceae to Juneaceae has been submitted to the

editor. Only the genus Aneilema (Com-

melinaceae) has still to be completed. Authors of

this part are Mrs A. A. Mauve, Prof. J. P. M.

Brenan (K), Mr J. Lewis (BM) and Dr R. Faden

(F).

Vol. 7: Prof. M. P. de Vos of the University of

Stellenbosch has offered to adapt her revision of

Romulea to Flora format early in 1981.

Vol. 11: The revisions of Lampranthus and Gib-

baeum (Mesembryanthemaceae), completed by

Dr H. F. Glen, will appear in the Contributions

of the Bolus Herbarium soon.

Vol. 14 (Crassulaceae): As reported before, a revi-

sion of the genus Crassula in southern Africa

has been published in the Contributions of the

Bolus Herbarium and the rest of the text on the

400 species of the volume is with the editor.

Publication is delayed, however, as attempts are

being made to align the work with treatments of

the family presently being prepared for the Flora

Zambesiaca and the Flora of Tropical East

Africa.

Vol. 18,3: This fascicle will comprise Simaroubaceae

to Malpighiaceae, a total of 45 species. Prof. J.

J. A. van der Walt of Stellenbosch University is

adapting his publication on Burseraceae to Flora

format and Mr F. White of the Forestry Her-

barium, Oxford, will do the same for his

publications on Meliaceae and Pteroxylaceae.

The other families are being written up by Miss

K. Immelman.

Vol. 21: Tiliaceae by Prof. H. Wild is being edited.

Most parts of Sterculiaceae prepared by Dr I. C.

Verdoorn are being published in Bothalia. The

major part of Hermannia has already appeared.

Vol. 28: Dr L. E. Codd has completed most of the

work on a fascicle of the Lamiaceae, a family

comprising about 250 species.

Brief mention can be made of some volumes on

which active research is in progress in the Institute.

Vol. 2: Poaceae (Dr G. E. Gibbs Russell).

Vol. 5: Liliaceae (Mrs A. A. Mauve).

Vol. 10: Polygonaceae (Mr G. Germishuizen).

Vol. 11: Mesembryanthemaceae (Dr H. F. Glen).

Vol. 15: Rosaceae — Rubus (Mr C. H. Stirton).

Vol. 16: Fabaceae — Eriosema (Mr C. H. Stirton).

Vol. 25: Ericaceae (Mr E. G. H. Oliver).

Palaeoflora of Southern Africa

A revision of the genus Dicroidium (extinct seed

ferns) comprising some 20 species is being prepared

as Volume 1 of the Palaeoflora series by Drs John

and Heidi Anderson. Material was largely derived

from 45 localities spread throughout the Molteno

Formation. The 50 mosaic photographic plates illus-

trating the reference populations have been prepared

and the text is two-thirds completed.

A Palaeobotanical Herbarium has been established

at the Institute and the fossil collections are now

housed in 20 specially designed cabinets. The collec-

tion has been recatalogued with the prefix PRE/F

and comprises some 5 000 specimens to date.

Register of plant taxonomic projects

A new edition of this register, listing more than 300

current projects on African plants, was completed

and distributed world-wide in microfiche form.

580

Botanical collectors in Southern Africa

The initial write-up of this work, which is being

written by Miss M.D. Gunn and Dr L. E. Codd, is

complete. Part 2 of the work (Dictionary Botanical

Collectors) is being set up for the publisher on an

IBM compositor. The text of Part 1 (Historical intro-

duction) is being revised in places.

Southern African Plants

Brochures on 20 of the most important water

plants are in press. Proofs of the text have been read

and publication of all brochures is expected before

the end of 1980.

Pretoria Flora

Texts of 195 species (including the family

Fabaceae) were typed. Material of a further 120

species is ready for typing. Texts of 291 species were

re-written and are almost ready for typing. Five

artists and twelve researchers of the Institute have

contributed to the work.

Liaison officer, Kew

The present incumbent, Mr C. H. Stirton, con-

tinued with research on the taxonomy of Rubus,

Lantana and various genera of Fabaceae. In addi-

tion, he checked on the nomenclature of the species

to be included in the National Weed List and dealt

with numerous queries from the Institute and from

universities and other institutions, both in the R.S.A.

and overseas.

PLANT STRUCTURE AND FUNCTION SECTION

Until recently the disciplines of plant anatomy and

cytogenetics were the responsibility of the Flora

Research Section. These functions have now been ac-

corded sectional status and, together with the envis-

aged development of plant physiological studies and

facilities at the Botanical Research Institute, should

form the basis of expanding applied taxonomic and

biosystematic studies.

Plant anatomy

The grass leaf anatomy project, the main project

being undertaken by Mr R. P. Ellis, continued to

yield interesting results in the genus Merxmuellera

( = Danthonia) where the comparative anatomy of

the summer-rainfall species has been completed.

These studies have clearly shown that in the Drakens-

berg region there has been extensive diversification of

this genus with the basic M. stricta and M. disticha

lines showing parallel anatomical and ecological

adaptations to the various habitats inhabited by these

species in this area. From the results of this study it

appears as if several taxonomic adjustments are

needed to the current classification of these Merx-

muellera species.

During 1979 Miss Riana Manders was seconded to

the University of Pretoria where she was awarded her

B.Sc. Hons degree. After relieving in the Data Pro-

cessing Section for six months she will concentrate on

the applied aspects of grass leaf anatomy, particular-

ly dietary studies in grazing herbivores using histolo-

gical analysis.

Cytogenetics

The major research conducted during the review

period has been undertaken by Mrs Wilma Gaum of

the Plant Exploration Section as part of the weed

research programme of the Institute. Her cytogeneti-

cal studies have concentrated on Lantana camara and

the genus Rubus and form part of detailed bio-

systematic studies being co-ordinated by Mr C. H.

Stirton. Results have confirmed that Rubus forms a

polyploid series with the basic chromosome number

of x = 7. The diploid species exhibit normal meiotic

chromosome behaviour which indicates that they are

completely fertile — a fact confirmed by embryo sac

analysis. Many of the polyploids, on the other hand,

display relatively abnormal meiosis.

The cytogenetical aspects of the cytotaxonomic

study of the Eragrostis curvula complex have been

completed. Seed of all the specimens examined cyto-

genetically in this project has been lodged with the

Seed Bank of the Plant Protection Research Insti-

tute. This project has been written up in the form of a

thesis for which Mr T. B. Vorster has been awarded

his D.Sc. degree.

ECOLOGY SECTION

The function of the Ecology Section, formerly

known as the Botanical Survey Section, is to study

the vegetation of South Africa and its ecological

relationships. This work covers three main fields of

activity: (1) The identification, description and map-

ping of various vegetation classes; (2) The study of

the ecological relationships between different types

of vegetation — mutually and with the environment

— and also of the various processes and mechanisms

that govern the behaviour of plant communities; (3)

The development of various methods and techniques

required for ecological studies of vegetation.

Veld types of South Africa

This project has been dealt a severe blow by the

passing of the late Mr J. P. H. Acocks. The great

quantity of processed and semiprocessed data as well

as his original observations have been sorted and

safely housed at the Botanical Research Institute,

Pretoria. Here they represent an incomparably com-

prehensive collection of past and present vegetational

and environmental data. The formidable task of fur-

ther processing this data base and exploiting its

potential to the full awaits an improvement in avail-

able manpower.

Transvaal bushveld studies

To complement earlier work on this programme,

Mr R. H. Westfall has started work on a project to

study the ecology of the Sour Bushveld (Veld Type

20). This veld type is starting to undergo greater in-

tensification of agricultural and other uses and ra-

tional planning for multiple-use management must

be set in motion. Increasing intensification of agri-

culture must be reconciled with other sometimes con-

flicting but valid needs such as nature conservation

and recreation. Field work on this project started in

1980 and by the end of March 132 quadrats had been

sampled.

Coastal studies

Dr P. J. Weisser studied the seaward advancement

of the dunes at Mtunzini, Natal by comparing aerial

photographs taken from 1937 to 1977. Dunes ad-

vanced by an average of 95,2 m, at an average rate of

2,4 m per year.

The vegetation of the Wilderness Lakes in the

Cape Midlands and the water-plant encroachment

problem were also studied by Dr Weisser. An in-

crease of emergent water plants and a decrease of

submerged water plants was detected in localized

areas. This is related to the lower water levels owing

581

to the artificial opening of the Wilderness Lagoon

mouth, in addition to natural succession.

Messrs H. C. Taylor and C. Boucher, Dr P. J.

Weisser and co-workers are participating in a world-

wide study of dry coastal ecosystems. The floristic in-

formation gathered from more than 250 releves is

being processed by Braun-Blanquet phytosociolo-

gical methods to identify plant communities of dry

coastal ecosystems of South Africa and to interpret

their environmental relationships. The communities

of the rocky coasts are certainly different from those

of the dunes, but the dune vegetation appears to be

more complex than was previously thought.

Miss B. K. Drews conducted a reconnaissance of

conservationworthy areas of fynbos in the Knysna-

Wilderness-Plettenberg Bay Guide Plan Area for the

former Department of Environmental Planning and

Energy to provide the basis for the protection of

natural areas. A report and map were produced

recommending that 13 areas be identified as natural

areas.

Vegetation survey of the Cape of Good Hope Nature

Reserve

Mr H. C. Taylor has completed the field work for

assessing the rate of infestation of plant invaders

over a ten-year period. This entailed relocating 27

permanent sample plots and re-enumerating the

invader density on them.

A primary survey of Rooiberg Mountain Catchment

Reserve near Ladismith, Cape

The structural units and floristic associations iden-

tified by Mr Taylor during this survey closely cor-

relate with each other, and their distribution reflects

the major environmental influences, aspect and

altitude. He concludes that, despite the preliminary

character of the survey, resource inventories of this

type are suitable as a foundation for park manage-

ment.

In another study, he has shown that the Rooiberg

flora has strong affinities with the eastern fynbos ele-

ment and with the dry fynbos of the inland mountain

ranges.

Hakea project

Mr S. R. Fugler has submitted his findings in the

form of unpublished reports, parts of which may be

published in due course. He concludes that Hakea

encroachment can be controlled subject to availabil-

ity of funds and manpower. Ecological principles are

currently being applied to find practical methods of

effecting control.

Physiognomic classification of mountain fynbos

Mr B. M. Campbell has completed the field work

for the classification. This consisted of 508 plots

located on 20 transects throughout the mountains. In

each plot detailed information on structure and func-

tion (growth-form, height classes etc.), environment

and floristic composition (dominant species etc.) was

collected. All the structural-functional and environ-

mental data have been edited, coded and placed on

computer file ready for analysis. The dominant

species are being identified, and the floristic data are

being coded.

A study of the vegetation along transects through the

western Cape foreland

Mr C. Boucher and Mr P. A. Shepherd have com-

pleted sampling of the natural vegetation along four

transects through the western Cape’s coastal fore-

land. Two hundred floristic and two hundred phys-

iognomic releves and 614 specimens were collected.

All the data were coded for computer analysis. The

preliminary results from the analysis of the first

transect indicate that all three of Acocks’s veld types

found in the area can be clearly distinguished. A

more detailed analysis is not possible until a larger

percentage of the specimens collected have been iden-

tified.

Orothamnus project

The annual monitoring of Orothamnus zeyheri

(‘marsh rose’) populations in the Kogelberg State

forest was undertaken by Mr C. Boucher during Jan-

uary 1980. Phenological growth-rate and population-

size data were recorded. Two new populations were

found. The total number of plants counted compares

favourably with the maximum population size

previously known, if it is considered that between

nine and twelve years have elapsed since regenerative

treatments were originally applied and the O. zeyheri

normally has a lifespan of less than 23 years.

Aquatic ecology

Classification and mapping of aquatic macrophyte

communities of Natal

Mr C. Musil’s preliminary classification of the

water-plant communities of Natal is completed.

Water-plant communities are grouped into five cate-

gories based on their habitat preference and tolerance

of water salinity and its acidity or alkalinity. The

categories are marine, estuarine, brackish water,

moderately fresh to slightly brackish water, slow-

flowing and fast-flowing fresh-water communities.

The distribution of the communities, included within

each category, are mapped and their composition

and environmental tolerances are discussed.

An ecophysiological study of water hyacinth in Natal

Mr Musil’s findings on water hyacinth (Eichhornia

crassipes) have been synthesized and the final report

is being prepared. Growth constants determined in

culture cannot be used to predict growth rates of

plants in the field. Growth rates in the field are high

with an average mass doubling time of 3 days in

eutrophic situations. Growth is highly correlated

with radiation, temperature and relative humidity.

The density of the water hyacinth populations in-

fluences the rate of growth and chemical composition

of plants

National Conservation Plan

The work of Dr J. C. Scheepers and Miss B. K.

Drews on the NAKOR National Plan for Nature

Conservation has mainly entailed the handling,

storage and mapping of data on conserved areas in

South Africa, and the establishment of a computer

data bank. At present, this data bank stores informa-

tion on about 300 existing and 100 proposed conser-

vation areas. These data are available for retrieval

and processing in various ways for planning pur-

poses.

DATA PROCESSING AND ECOSYSTEM STUDIES SECTION

This will probably be the last report by this Section

in its present form as Dr J. W. Morris has been

transferred to Datametrical Services (Head Office)

and an internal re-organization of sections will be

carried out in the near future. The mandate of this

582

Section is the provision of data processing facilities

for research purposes to the rest of the Institute as

well as undertaking plant ecological research at the

ecosystem (function) level. The largest data process-

ing task to be undertaken, the computerization of the

National Herbarium, is now complete. Our contribu-

tion to the Savanna Ecosystem Project at Nylsvley,

determination of biomass relations and seasonal

biomass change in dominant tree and shrub species

has made good progress and a number of reports on

biomass relations have been published. The Savanna

Ecosystem Project data bank is administered by this

Section, under the control of Dr J. W. Morris. He is

also responsible for co-ordination of modelling ac-

tivities as well as research in the Decomposer and

Nutrient Cycling Components of the Project.

The National Herbarium Data Bank (PRECIS) has

changed to the status of a production system and

already useful results are being obtained from it. The

production of maps showing the distribution of

specimens selected from the data bank is a facility

which has been added.

Further work on savanna root growth by Dr M. C.

Rutherford has given valuable results using a

sophisticated root observation chamber. The

chamber incorporates a microscope and travelling

carriage enabling precise observations of the activity

of the belowground plant component that is impor-

tant to the functioning of savanna vegetation as a

whole. Pioneering work on the effects of plant water

status on the radial growth of bushveld trees has also

been carried out by Dr Rutherford. Techniques have

included the severing of tree trunks held in position

while simultaneously monitoring various properties

of the tree trunk and leaves.

PLANT EXPLORATION SECTION

The Economic Botany Section under Mr M. J.

Wells has changed its name to the Plant Exploration

Section, as a result of the responsibility for weed

research having been moved to the new Weed

Research Unit in the Plant Protection Research Insti-

tute. Despite this organizational move it is antici-

pated that botanical aspects of weed research will

continue to be handled by officers seconded to the

Institute for many years to come.

The accent in the past year has been on developing

new fields in plant exploration research under the

team leader Mr T. H. Arnold, whilst concentrating

most of our weed research effort on the compilation

of the National Weed List, which we regard as basic

to future research development.

The origin and evolution of Sorghum

This is a new project being developed by Mr

Arnold. One hundred and two collections of Sorg-

hum have been made during field trips to Natal and

the north and north-eastern Transvaal. Thirty-two

collections from elsewhere in Africa have been ob-

tained from the Department’s seed bank and have

been grown. Herbarium collections, seed (when

available) and photographs have been taken to form

the basis of a study of the origin and evolution of

Sorghum in the sub-continent. Chromosome counts

have begun.

Previous accounts of Sorghum have largely

ignored South African material and already Mr

Arnold has found several variants that were thought

not to exist in South Africa.

The origin and evolution of Pennisetum

A second new project on an indigenous food crop

is that on Pennisetum. It is being carried out by Miss

K. J. Duggan. Ninety-two collections of Pennisetum

were made in Natal and the Transvaal, and seed of 25

collections from elsewhere in Africa was grown, to

provide material for a study of the origin and evolu-

tion of this group. This material already includes a

very wide variety of ‘wild’, cultivated and inter-

mediate plants.

Citrullus studies

Twenty-three collections of Citrullus (Water-

melons) were added, mainly by Mr Arnold and Miss

Duggan, to our already large collection. Recordings

of variations in fruit and seed characteristics of

Citrullus were made and will be used later for a study

of the origin and evolution of this crop plant.

Conservation of germ plasm

This project, quiescent for some time, has

benefited by material gathered in the course of other

studies.

Seventy-eight collections of Sorghum seed and 46

collections of Pennisetum seed were contributed to

this project which aims to conserve germ plasm of

primitive varieties of crop plants. Other collections

included 27 of Citrullus and related taxa, 6 of

Lagenaria and 10 miscellaneous. Information

relating to rare and endangered indigenous species

was compiled by Mrs J. B. Hoffmann who combines

this work with her other duties as Public Relations

Officer.

Ethnobotany

Miss C. A. Liengme has commenced an intensive

investigation of wood utilization by the Tsonga of

Gazankulu. Collections have been made in the study

area to enable firewood and wood used for building

to be identified by gross morphological features. The

quantity and kinds of wood used for different pur-

poses will be monitored over several seasons, and the

effect of timber gathering will be analysed.

Miss Liengme has also continued gathering records

of tribal plant uses for the sub-continent and this will

continue as a long-term project.

Cover and barrier plants

Another new project, a survey of indigenous plants

of potential as barriers or ground covers has been

initiated by Miss L. Henderson. Plants are being col-

lected, grown and screened for use as hedges,

binders, windbreaks, etc., in the various climatic

regions of South Africa. Over 100 plants with poten-

tial have been identified and over 50 have been col-

lected.

Tree distribution in the Transvaal

Dr J. Anderson’s survey of the distribution of

woody plants in the Transvaal has had to take second

place to his palaeoflora work but good progress was

made nevertheless. Two hundred and ten field

listings of woody plants have become available

during the report year. Most of these were provided

by co-operating researchers from outside the Depart-

ment who helped fill distribution gaps in the

Waterberg and in Venda. One thousand nine hun-

dred of the approximately 7 000 1/1 6th degree

squares occurring in the Transvaal have now been

583

sampled. The field work is scheduled for completion

next year.

Information service

Mrs D. M. C. Fourie, our Scientific Information

Officer, handled 261 requests for information about

economic plants and their utilization or control.

These included about 120 identifications. Particular

interest was shown in oil and rubber producing plants

such as Simmondsia (jojoba) and Parthenium

(guayule).

Over two thousand visitors including delegates to

the International Rose Convention, members of the

Wild Life and Tree Societies and many scholars were

taken on tours of the Institute. In this aspect of her

work Mrs Fourie received the assistance of Mrs Hoff-

mann, our new Public Relations Officer.

National weed list

The first draft of the National Weed List which

contained the names of 700 species, has been expan-

ded as a result of suggestions received from many

correspondents. Mr A. A. Balsinhas is responsible

for the expanded list that now includes approximate-

ly 1 500 species and 4 500 common names. The iden-

tities of exotic species have been checked at Kew by

Mr C. H. Stirton and the listed species are being clas-

sified by Miss V. M. Lorentz according to whether

they are indigenous or exotic; and according to the

situations where they cause problems and the kind of

problems caused. The second draft of the list con-

taining this information will be ready by the end of

the year.

Woody invaders

The intensive survey of exotic woody invaders in

the central Transvaal, which was completed last year

by Miss Duggan and Miss Henderson, has been ana-

lysed and the minimum effective sampling level has

been calculated. Sampling at this level is now being

expanded to cover the rest of the Transvaal.

The rate of spread of some invader species has

been calculated from aerial photographs. For exam-

ple, one large infestation of Acacia dealbata (silver

wattle) has increased its area by 20% in 8 years. This

work is continuing.

Book on plant invaders

The book ‘Plant invaders: beautiful but danger-

ous’ compiled and edited by Mr C. H. Stirton for the

Cape Department of Nature and Environmental

Conservation has proved such a success that it is

being re-printed.

Lantana camara

Mr Stirton has completed his review of the vol-

uminous international literature dealing with Lan-

tana, has consulted herbarium material in Europe

and has sorted out most of the nomenclatural prob-

lems of the L. camara complex.

Most of the cytogenetic work on the many varia-

tions occurring in South Africa has also been com-

pleted by Mrs W. G. Gaum. The morphological and

genetic data must now be assessed and explained in

terms of the evolutionary constraints operating on

the important weed group.

Rubus

Many collections of weedy brambles (Rubus spp.)

were made by staff of the Section, and Mrs Gaum

made a start on a study of the complex cytogenetics

of the genus in South Africa. The occurrence of two

species R. niveus (Java bramble) and R. phoenicola-

sinius (wineberry), both with weedy potential, has

been fully documented for the first time, in articles

prepared by Mr Stirton.

Cyperaceae

Several papers dealing with convergent evolution,

an infra-specific hybrid, and morphological variation

within the genus Ficinia have been prepared by Mr

Arnold. These studies are of particular significance

as a number of species of Ficinia are regarded as

forestry weeds in the Cape, and a good understand-

ing of the genus is required.

Nassella tussock

Research on the germination, establishment, and

distribution of the grass weed Nassella tussock in the

winter-rainfall area is being continued by Mr G. Har-

ding, who took over the project from Miss S. Bulley

at the end of 1979. Prior to that Mr Harding com-

pleted his B.Sc. Hons degree whilst seconded to

Natal University.

GARDEN SECTION

There were two major developments in the garden

during 1979/80. Firstly, the garden was declared a

national monument and a plaque was unveiled by Dr

R. A. Dyer, a previous Director, on the 25th of Octo-

ber 1979 to commemorate the event. One of the bene-

fits of national monument status will be protection of

the terrain from expropriation for road-building and

other purposes.

A second major development was the landscaping

of the water garden, karoo and grassland areas. This

landscaping which had long been planned, suddenly

became a reality when large quantities of soil became

available at an adjacent site. Contractors working on

the new police stores were grateful to find a dumping

area for about 35 000 cubic metres of soil, and also

excavated dams for us in order to obtain shale for

road building. This development was made possible

by the quick thinking of Mr H. J. de Villiers and Mr

T. A. Ankiewicz of the garden section. Six koppies

were built and four dams and linking channels were

excavated. In all we obtained material, labour and

machine time to the value of about R100 000, in

return for payment of R9 000. Understandably,

other garden activities had to suffer in order to make

the most of this windfall, and no perennial plants

could be planted out in the nine months under

review. However, progress was made in other areas.

The exacavation of the swamp forest area and the

landscaping of savanna koppies were completed and

the existing irrigation and roads systems were ex-

panded.

The post of curator of the botanic garden has not

been filled and the curator’s duties continue to be

divided between two acting curators, the technicians

in charge of the nursery and garden sub-sections

respectively. Mr D. S. Hardy continued in charge of

the nursery, but health reasons caused Mr J. Erens to

step down as technician in charge of the garden. His

place was taken by Mr H. J. de Villiers, and Mr

Erens moved to the nursery where he is responsible

for the propagation of plants for display purposes.

There were 376 accessions to the garden’s perma-

nent scientific collections and 238 to the new category

of temporary experimental collections, during the

nine months under review. The main experimental

584

collections were of barrier plants, and of Sorghum

and Pennisetum food plants.

Mrs B. C. de Wet and Mrs K. P. Clarke continued

with the task of labelling and record keeping. Mrs de

Wet was particularly active in the computerization of

garden records. Amongst the benefits of com-

puterization, are biome planting lists and directly

typed garden labels on request. Computerization is

also assisting in keeping track of the recent surge of

experimental plantings.

With so much development having taken place in

1979/80, we look forward to several years of con-

solidation, in building up and maintaining biome

plantings, and in utilizing the material already

available.

BOTANICAL RESEARCH INSTITUTE

Scientific, Technical and Administrative Staff

(31st March 1980)

585

Director

Technical Assistant Vacant

B. de Winter, M.Sc., D.SC. (Taxonomy of

Poaceae, especially Eragrostis, and of

Hermannia; plant geography)

Deputy Director

D. J. B. Killick, M.Sc., Ph.D., F.L.S.

(General taxonomy and mountain

ecology)

Assistant Director

D. Edwards, M.Sc., Ph.D. (Ecological

methodology; aquatic plants, remote

sensing and vegetation structure and

physiognomy)

ADMINISTRATION

Administrative Officer. . .

Senior Administrative

Assistant

Senior Administrative

Assistant (Personnel) . .

Assistant Accountant. . . .

Administrative Assistants

Personal Secretary to

Director

Senior Clerical Assistant.

Clerical Assistants

Receptionist

Typists ....

Technician: Assistant to

Secretary-General:

AETFAT

D. F. M. Venter

Mrs G. E. Hussem,

B.A.

J. Conradie

A. Smith (Temporary)

Mrs J. Rautenbach

Mrs M. C. van Niekerk

Mrs M. M. Loots

Mrs T. Creffield

(Registry)

Mrs C. A. Bester

Mrs J. S. A. Strydom

(Personnel)

Mrs I. J. Joubert*

(Registry)

Mrs M. E. M. Venter

Mrs N. Miller*

Mrs S. M. Thiart*

Mrs J. Gerke*

Mrs B. A. Momberg*

Wing A (Cryptogams — Monocotyledons)

Chief Professional Officer Mrs E. van Hoepen

Senior Professional Of-

ficer Miss C. Reid, B.Sc.

Hons

Technician Miss L. Smook, B.Sc.

(Poaceae)

Technical Assistant Mrs A. M. Fourie*

Wing B (Piperaceae — Oxalidaceae)

Senior Professional Of-

ficer G. Germishuizen, B.Sc.

Hons (Polygonaceae)

Professional Officer Vacant

Technical Assistant Mrs I. R. Leistner*

Wing C (Linaceae — Asclepiadaceae)

Chief Professional Officer

Professional Officer

Technician

Technical Assistants

Wing D (Convolvulaceae-

Chief Professional Officer

Professional Officer

Technical Assistant

Miss E. Retief, M.Sc.

(Campanulaceae)

P. P. J. Herman,

B.Sc. Hons

Mrs M. J. A. W.

Crosby, B.Sc.*

Mrs J. I. M. Grobler*

C. Hildyard, B.Sc.

(Elec. Eng.) (part-

time)

•Asteraceae)

Miss W. G. Welman,

M.Sc.

Mrs S. J. Smithies,

M.Sc.

Mrs K. A. Kleynhans*

Cryptogamic Herbarium

Senior Professional Of-

ficer

Professional Officer

Technical Assistants

R. E. Magill, M.S.,

Ph.D. (Musci)

F. A. Brusse, B.Sc.

Hons (Lichens)

Mrs L. R. Filter*

Mrs P.W. van der

Helde

J. van Rooy (Musci)

HERBARIUM SERVICE SECTION

Officer in Charge

Chief Professional Of-

ficers

* Half-day

E. G. H. Oliver, M.Sc.

(Curator; Ericaceae)

Mrs E. van Hoepen,

M.Sc. (Assistant

Curator; supervision

of identifications and

enquiries)

Services

Technical Assistants

Typist

Photographic Room

Photographer

Mrs I. Ebersohn

(Herbarium records,

loans and exchanges,

parcels, etc.)

Mrs S. M. Perold,

B.Sc. (S.E.M. and

laboratory work)

Mrs G. L. Radmacher

(Phytogeography

and data bank)

Mrs A. M. Verhoef

Mrs A. J. Romanowski

National Herbarium, Pretoria (PRE)

586

Natal Herbarium, Durban (NH)

Senior Professional

Officer

Professional Officer

Technician

Technical Assistant.

P. C. V.duToit, M.Sc.

(Curator; Poaceae,

especially Pentas-

chistis, and general

identifications, until

1980/06/30)

B. D. Schrire, B.Sc.

Hons (Curator from

1980/08/01)

Mrs B. J. Pienaar,

B.Sc.*

Mrs L. Nichols

Professional Officer

Senior Technician. .

Graphic Artist

Technical Assistants

Mrs A. A. Obermeyer-

Mauve, M.Sc. (Tax-

onomy, especially

Monocotyledons)

C. H. Stirton, M.Sc.

(Liaison Officer,

Kew; taxonomy,

especially Fabaceae

and weeds)

Miss K. L. Immelman,

M.Sc. (Taxonomy)

Mrs H. M. Anderson,

Ph.D. (Palaeoflora)

Mrs R. C. Holcroft

Mrs C. F. Fourie

Miss N. van der Meulen

Albany Museum Herbarium, Grahamstown

(GRA)

Senior Professional

Officer Mrs E. Brink, B.Sc.

(Curator; general

identifications)

Technical Assistants Vacant

Miss G. V. Britten*

Government Herbarium, Stellenbosch (STE)

Senior Professional

Officer

Professional Officer

Technician

Technical Assistant .

Clerical Assistant . .

Miss L. Hugo, M.Sc.

(Curator; general

identifications)

Miss M. Schonken,

B.Sc. Hons

D. A. Davies, B.Sc.

Mrs R. Wikner

Miss E. N. Pare

FLORA RESEARCH SECTION

Officer in Charge O. A. Leistner, M.Sc.,

D.Sc., F.L.S.

Flora of Southern Africa Team

Chief Professional Officer

Senior Professional Of-

ficers

O. A. Leistner, M.Sc.,

D.Sc., F.L.S. (Tax-

onomy, especially

Malvaceae)

J. M. Anderson, Ph.D.

(Palaeo-ethno-

botany, palaeoflora,

plant geography)

L. E. W. Codd, M.Sc.,

D.Sc. (Taxonomy,

especially Lamia-

ceae; history of plant

collecting)

Mrs G. E. Gibbs

Russell, B.S., Ph.D.

(Taxonomy, espe-

cially grasses)

H. F. Glen, M.Sc.,

Ph.D, F.L.S. (Num-

erical taxonomy,

Mesembryanthema-

ceae)

PLANT STRUCTURE AND FUNCTION

SECTION

Chief Professional Officer

Professional Officer

(Anatomy)

Professional Officer

(Cytogenetics)

Technical Assistant.

R. P. Ellis, M.Sc.

(Anatomy of south-

ern African grasses)

Miss R. Manders,

B.Sc. Hons (Grazing

studies)

Vacant

ECOLOGY SECTION

Officer in Charge J. C. Scheepers, M.Sc.,

D.Sc.

Chief Professional Of-

ficers J. C. Scheepers, M.Sc.,

D.Sc. (Vegetation

ecology, especially of

forest/woodland/

savanna/grassland

relationships; conser-

vation and land-use

planing; phytogeo-

graphy)

H. C. Taylor, M.Sc.

(O/C Botanical Re-

search Unit, Stellen-

bosch; mountain

fynbos and forest

ecology; Braun-

Blanquet approach

and techniques; con-

servation)

C. Boucher, M.Sc

(Lowland fynbos

ecology and phyto-

sociology; conserva-

tion and land-use

planning; Braun-

Blanquet approach

and techniques)

P. J. Weisser, Ph. D.

(Reedswamp eco-

logy; ecological

planning and en-

vironmental impact

studies; Zululand

coast dune vegeta-

tion; conservation)

587

PLANT EXPLORATION SECTION

Officer in Charge

Chief Professional Offi-

cers

Datametricians

Technician

Learner Technician.

Technical Assistants

Senior Administrative

Assistant

Clerical Assistants

J. W. Morris, M.Sc.,

Ph.D.

J. W. Morris, M.Sc.,

Ph.D. (Data process-

ing and quantitative

ecology)

M. C. Rutherford,

M.Sc, Ph.D. Dipl.

Datamet. (Biomass

and production

studies in ecosystems

research, especially

savanna and fynbos)

Miss R. Manders, B.Sc.

Hons

Miss F. M. Thatcher,

Ph.D.

B. Curran B.Sc.

M. D. Panagos

Mrs E. Ewenwell*

Mrs J. H. Jooste*

Mrs N. Nigrini*

Mrs L. E. Oosthuizen

Mrs J. Mulvenna

Mrs S. Smit*

Vacant*

Officer in Charge

Chief Professional Officer

Senior Professional Of-

ficer

Professional Officers. . . .

Research Technicians. . . .

Technical Assistant

Curator

Acting Curator (garden) . .

Acting Curator (nursery).

First Research Techni-

cians

Senior Research Techni-

cians

Research Technician

M. J. Wells, M.Sc.

(Weed research,

palaeo-ethnobotany,

botanical horticul-

ture, fynbos utiliza-

tion and conserva-

tion)

T. H. Arnold, M.Sc

(Conservation of

germ plasm, plant

utilization and Sor-

ghum)

Miss C. A. Liengme,

B.Sc. Hons (Eth-

no-botany)

G. B. Harding, B.Sc.

Hons (Weed research)

Miss K. J. Duggan,

B.Sc. Hons ( Pennise -

turn and weed

research)

Miss L. Henderson,

B.Sc. Hons (Cover

and barrier plants,

and weed research)

Mrs D. M. C. Fourie,

B.Sc. (Scientific in-

formation service

and identification of

exotics)

Mrs W. G. Gaum,

B.Sc. (Cytogenetics)

Miss V. M. Lorentz,

B.Sc. (Weed

research)

Mrs J. B. Hoffmann,

B.Sc. (Public rela-

tions and conserva-

tion of germ plasm)

A. A. Balsinhas

(Plant collecting)

Vacant

H. J. de Villiers

D. S. Hardy

D. S. Hardy (Nursery

supervision, succu-

lents and orchids)

H. J. de Villiers, NTC

III (Hort.), Dipl.

Rec. P.A, (Develop-

ment of savanna

biome)

J. Erens (Propagation

of display material)

T. A. Ankiewicz, Dip.

For. (Administra-

tion, stores and pur-

chases, and develop-

ment of coastal

forest biome)

Mrs M. C. Innes, Nat.

Dip. Tech. (Hort.),

(Nursery and devel-

Senior Professional Of-

ficers F. Musil, M.Sc.

(Aquatic ecology and

survey of aquatic

plants, especially in

Natal; ecophysiologi-

cal studies on Eich-

hornia crassipes and

Salvinia molesta)

R. H. Westfall, B.Sc.

Hons (Ecology and

phytosociology of

Transvaal Bushveld)

B. M. Campbell, M.Sc

(Physiognomic-

structural classifica-

tion of fynbos phyto-

sociology; quantita-

tive methods)

Professional Officers.... Miss B. K. Drews,

B.Sc. Hons (Applied

ecology and environ-

mental planning;

conservation)

Technicians P.A. Shepherd, B.Sc.

(Lowland fynbos

ecology and phyto-

sociology, threatened

species)

Miss A. J. Naude,

B.Sc. (Ecological lit-

erature, air photos

and maps)

Mrs M. Engelbrecht,

B.A. Hons (Fine

Arts) (Draughtman-

ship and carto-

graphy; artwork, lay-

out and design)

DATA PROCESSING AND ECOSYSTEM

STUDIES SECTION

PRETORIA BOTANICAL GARDEN

588

Technical Assistants

opment of fynbos

biome)

Mrs B. C. de Wet*,

B.A. (Garden re-

cords)

Mrs K. P. Clarke

(Garden records)

Learner Technicians

Farm Foremen

TN 4 Supervisor. . .

TN 6

Miss H. D. Eloff

Miss Y. Mennin

G. J. Stolz

H. N. J. de Beer

L. C. Steenkamp

J. P. Booysen

PUBLICATIONS BY THE STAFF (1 July, 1979-31 March, 1980)

Boucher, C., 1980. Vegetation under stress. An introduction to a

study of the western Cape coastal foreland vegetation. Veld &

Flora 66: 14-16.

Campbell, B. M., McKenzie, B. & Moll, E. J., 1979. Should

there be more tree vegetation in the Mediterranean climatic

region of South Africa? Jl S. Afr. Bot. 45: 453-457.

Codd, L. E., 1979. The story of the Barberton Daisy, Gerbera

jamesonii. Veld & Flora 65: 144-115.

Gunn, M. D., & Codd, L. E., 1979. Plant collecting pioneers in

the Barberton area. Veld & Flora 65: 98-101 .

Dyer, R. A., 1980. Brachystelma, Ceropegia and Riocreuxia.

Flora sth. Afr. 27,4: 1-91.

Killick, D. J. B., 1979. African mountain heaths. In R. L.

Specht, Heathlands and related shrublands of the world. A.

Descriptive studies. Amsterdam: Elsevier. 97-116.

Morris, J. W. & Glen, H. F., 1979. PRECIS, the National

Herbarium of South Africa (PRE Computerized Information

System. In G. Kunkel, Taxonomic aspects of African

economic botany. ( Proceedings of the IX Plenary Meeting of

AETFAD ■ 206.' Las Palmas: AETFAT.

Ross, J. H., 1979. A conspectus of the African Acacia species.

Mem. bot. surv. S. Afr. No. 44, 1-155.

Rutherford, M. C., 1979. Plant-based techniques for determin-

ing available browse and browse utilization: A review. Bot.

Rev. 45: 203-228.

Rutherford, M. C., 1980. Annual plant production — precipita-

tion relations in arid and semi-arid regions. S. Afr. J. Sci. 76:

53-56.

Stirton, C. H., 1979. Taxonomic problems associated with

invasive alien trees & shrubs in South Africa. In G. Kunkel,

Taxo-nomic aspects of African economic botany. ( Pro-

ceedings of the IX Plenary Meeting of AETFAT). 218-219.

Las Palmas: AETFAT.

Taylor, H. C., 1979. Observations on the flora and

phytogeography of Rooiberg, a dry fynbos mountain in the

southern Cape Province, South Africa. Phytocoenologia 6:

524-531.

Taylor, H. C., 1979. Phytogeography. In J. Day, W. R.

Siegfried, G. N. Louw & M. L. Jarman, Fynbos ecology: A

preliminary synthesis. 70-81. Pretoria. South African Na-

tional Scientific Programmes. Report No. 40.

Wells, M. J., Duggan, K. J. & Henderson, L., 1980. Woody

plant invaders of the Transvaal. In S. Neser & A. L. P.

Cairns, Proceedings of the Third National Weeds Conference

of South Africa. 11-24. Cape Town: Balkema.

Wells, M. J., Harding, G. B., Balsinhas, A.. & Van Gass,

B., 1980. A preliminary experiment of sampling agronomic

weeds for a national survey. In S. Neser & A. L. P. Cairns,

Proceedings of the Third National Weeds Conference of

South Africa. 62-65. Cape Town: Balkema.

Werger, M. J. A., Morris, J. W. & Louppen, J. M. W., 1979.

Vegetation- soil relationships in the southern Kalahari. Doc.

Phytosociol. (Lille) 4: 967-981.

Wiens, D. & Tolken, H. R., 1979. Loranthaceae and Viscaceae.

Flora sth. Afr. 10,1: 1-59.

589

Book Reviews

Heathlands and Related Shrublands: Descriptive Studies

(Ecosystems of the World, Vol. 9A) edited by R. L. Specht

Amsterdam-. Elsevier Publishing Company. 1979. Pp. xiv + 497.

138 figures, 81 tables. Price R98.

Although this book is numbered 9A in Elsevier’s series, ‘Ecosys-

tems of the World’, it is among the earliest of the thirty volumes to

be published. ‘Heathlands and Related Shrublands’ is to appear in

two volumes of which only the first has been published and is

reviewed here. ‘Part A: Descriptive Studies’ describes the extent,

variation and ecology of heathlands and related shrublands

throughout the world. In Part B the phenological, morphological

and physiological characteristics of the heathland ecosystems will

be dealt with. Volume 9A is an imposing book. The text is set in

clear, double-column print and over half of its 497 pages consists

of tables, diagrams, appendices, references and photographs. In

23 chapters, 30 authors from eight countries describe the

heathlands of the various continents and some adjacent islands.

The presentations vary in length from a two-page compilation to a

comprehensive treatment comprising 83 pages.

In the introductory chapter, ‘Heathlands and related shrublands

of the world’, Specht traces the origin of heathland communities

from the pre-Angiospermous flora of Gondwanaland, and defines

the heathland concept floristically and structurally. In distinguish-

ing between the terms ‘heath’ for a member of the heath families

which he lists in the orders Celastrales, Ericales and Santalales,

‘heathland’ for the vegetation type and ‘heathland species’ for the

many diverse species and families which together comprise the

‘heathland’ community, Specht lays the foundation of uniformity

necessary for the varied treatments that follow.

The next three chapters are of great interest to South African

ecologists. Kruger’s ‘South African heathlands’ is the most com-

prehensive account of the fynbos ecosystem as yet published. His

explicit summary of environmental factors is followed by a

synusial description of vegetation which, though unusual, is most

effective in relating the life cycles of fynbos plants to the dynamics

of regeneration after fire. His account of vegetation zones and

community types is amplified by tables of original data on struc-

ture and floristics and is illustrated by several good photographs.

Bigalke’s account of aspects of vertebrate life in fynbos reveals the

region to be less distinct faunistically than it is floristically. In the

chapter on African mountain heathlands, which are concentrated

mainly along the eastern mountains and highlands, Killick ably

summarizes the information available from literature and cites his

own researches on heathland in the Drakensberg. The floristic

affinity with fynbos is evident in the high proportion of fynbos

taxa throughout the African mountains. The good photographs

show that the physiognomy is also comparable; Fig. 4.3 of the

Ericaceous Belt on Mount Kilimanjaro is especially striking in its

similarity to the Dwarsberg communities in Kruger’s Fig. 2.12.

A brief chapter by Page covers the Macaronesian heathlands.

These are floristically poor but vegetationally luxuriant,

sometimes reaching up to 10 m or more in height. They occur in a

cool temperate climate with high precipitation and show greater

affinity with some European heathy shrublands than with the

heathlands of the African mainland. For this reason the chapter

could have been better placed next to the one on European

heathlands.

The next five chapters deal with Australian heathland

ecosystems. Specht’s excellent account of the heathlands of east

and central Australia is supplemented by two appendices of tables

which can be readily compared with Kruger’s fynbos data. In

structure, life forms, leaf characters and general biology, including

fire-survival strategy, Australian heathlands are strikingly similar

to South African fynbos, with one important difference: the

frequent occurrence in Australia of an overstorey of tall shrubs

and trees. The Australian heath flora, like the Capensis flora in

South Africa, is an ancient one, rich in species, that has evolved in

a changing climate from pre-Angiospermous Gondwana origins.

In both continents, heathlands are confined to nutrient-poor soils

and in both, fire and a high seasonal moisture stress are the chief

ecological factors that mould the physiognomy of the vegetation.

The Australian heathland, however, occurs over a much wider

range of climates than fynbos, and many of the widespread

species, unlike those of fynbos, show great plasticity in growth

form, ‘often the same taxon ranging in habit from a dwarf shrub

to a tree 10 to 30 m in height’ (p. 144). In Chapter 7, George,

Hopkins and Marchant survey the heathlands of Western

Australia. The greater part of this chapter reports the original

results of a study of species richness and biogeography by

Marchant and Hopkins. The vegetational and ecological aspects

are presented in tabular form which, however, seems to me too

compressed to take the place of a concise textual account. Of

special interest are the ‘isoflor’ maps that show a southwestern

concentration of taxa similar to that found in the fynbos flora.

The last three chapters of this section are rather detailed accounts

ot special investigations into the faunal components of some

Australian heathlands.

The vegetation of the island complexes close to the Australian

mainland is next described. The ‘maquis’ of New Caledonia, a

sclerophyllous vegetation on oligotrophic soils, is described in a

brief compilation by Specht. This is followed by a stimulating

account of heathlands and shrublands of Malesia by Specht and

Womersley. As a result of the complex splitting apart and later

merging of Gondwanan tectonic plates, there are major discon-

tinuities in the modern floras of these islands. Heath species are

found as dwarf shrubs above timberline on mountains and at low

altitudes on infertile coastal sands. Very interesting is the fact that

in dipterocarp rain forests on fertile soils, typical heathland species

with coriaceous leaves and lignotubers occur as epiphytes in an

oligotrophic environment in the sunlit branches of the tallest trees.

Some of these epiphytes become terrestrial at high altitudes where

they form a subalpine heathland on infertile soils. The authors

conclude that ‘the emergence of “heath” species from the primary

tropical and lower montane rain forest appears to hold clues to the

development of heathland on oligotrophic soils throughout the

world’ (p. 332). Each of the four authors to Chapter 13 on the

heathlands of New Zealand (Burrow, McQueen, Esler and

Wardle) has, either singly or in collaboration, produced at least

one of the vegetation accounts of the eight heathland regions

described. For a country where ‘the study of heathlands is not far

advanced’, the authors have succeeded in summarizing existing in-

formation sufficiently to enable comparisons to be made with

heathlands elsewhere.

Three British authors, Gimingham, Chapman and Webb, have

written the chapter on European heathlands. The authors’ non-

regional treatment gives a well co-ordinated synoptic picture; good

maps, diagrams and photographs aptly illustrate the well written

text. (Fig. 14.7 showing two complete annual increments on a twig

of Calluna vulgaris is the only close-up photograph of a heathland

plant in the whole book). The temperate oceanic climate of the

European heathland region is quite different from the mediterra-

nean and sub-mediterranean climates of the fynbos biome. Never-

theless, the similarities between European heathland and Cape

fynbos make this chapter of particular interest to South African

readers. Lowland heathland occurs along the Atlantic fringe of the

European continent from Scandinavia to northern Spain;

elsewhere in Europe related vegetation types are found on moun-

tains. Fynbos has a similar pattern of distribution. European

heathlands are usually floristically poor, but their geographical

and climatic ranges are so wide that these heathlands, like fynbos,

‘provide an excellent example of continuous variation in floristic

composition’. Continuum studies and Braun-Blanquet classifica-

tion have both contributed conspicuously to an understanding of

regional variation in heathland composition; both treatments

might well be equally successful in fynbos. Calluna vulgaris is

often the dominant species in European heathlands. The cyclical

dynamic process in Calluna heath has its counterpart in fynbos

succession. The biology and ecology of Calluna, including the pro-

cesses involved in litter conversion and nutrient cycling, have been

closely studied in Europe, especially with reference to the

maintenance of Calluna heathland by fire. The chapter closes with

a fascinating account of the role of fauna in maintaining the

heathland ecosystems. The detailed knowledge accumulated,

especially regarding the invertebrate fauna, makes one only too

aware of our shortcomings in this sphere of fynbos research.

The vegetation of extreme climates is treated in the two follow-

ing chapters. The vegetation of the Artie heathlands beyond the

climatic limit of trees, described in Chapter 15 by Bliss, lias much

in common with the heathlands of the north temperate zone

treated in the preceding chapter. The species grow on shallow soils;

their nutrient availability and dry matter production are low, but

even in the tundra, fires occur and temporarily increase the content

of nitrogen, phosphorus and potassium in plant regrowth.

Specht’s short compilation in Chapter 16 shows that the Japanese

alpine communities are essentially similar to the artic heathlands in

the mountains of the Chukchi Peninsula of Far Eastern Siberia

described by Bliss.

Four of the remaining chapters cover North America and three

South America. To the North American the word ‘heath’ means

simply a plant that is a member of the family Ericaceae or the

order Ericales. The term heath or heathland is applied much less to

plant communities, and it is evident from the text that the North

American ‘heathland’ is different from most of the vegetation

described in previous chapters. True heathland, as defined by

Specht in Chapter 1, consists of shrubs not exceeding 2 m in

height, in the Ericales and other orders, with leaves that are mostly

either ‘ericoid’ (narrow-leptophyll and grooved below) or at least

very small, narrow and sclerophyllous. Apparently this sort of

vegetation does not occur significantly in North America except in

mediterranean-type shrublands which will be dealt with in a later

volume of this series. The North American shrublands described in

the present book are preponderantly broad-sclerophyll and usually

taller than 2 m. In some of the communities, if fire is excluded,

succession proceeds to forest; in others, the shrubland forms a

590

fire-maintained understorey to woodland or open-forest. These

communities are the ‘related shrublands’ referred to in the title of

the book. The only features common to true heathland and related

shrubland seem to be the relatively low stature of the vegetation

and its sclerophylly, oligotrophic soils and the role of fire in com-

munity dynamics. Whereas true heathlands form extensive and

distinct regional biomes in climates which, by and large, are un-

suitable for forest growth, the shrublands described for North

America are confined to ‘islands’ within a forest climate, where a

particular habitat factor, usually edaphic and often re-inforced by

fire, restricts succession to a pre-climax stage. Thus, to the South

African reader seeking knowledge of world equivalents to fynbos

— a heathland sensu stricto — the chapters on North American

shrublands, good as they are, may be confusing.

Most of the above observations apply equally to the South

American vegetation described in the final chapters. The muri and

savannah vegetation of the Guianas described by Cooper, and the

caatingas and campinas of the Amazonas, by Klinge and Medina,

are certainly shrublands. By contrast, Moore’s account of the

southern oceanic wet-heathlands of Tierra del Fuego, southern

Patagonia and islands in the South Atlantic shows that they are

clearly true heathlands, probably similar to the heathlands at

equivalent latitudes in northern Europe. They are dominated by

Empetrum rubrutn, an ericoid-leaved low shrub in the Celastrales

which is allied to the Ericales. Their structure, growth form and

floristic composition are clearly summarized in Table 23.2, a form

of presentation similar to Specht’s tables for the heathlands of

eastern and central Australia and to Kruger’s for fynbos. Com-

parison of these tables makes their physiognomic similarity

immediately clear.

In summary, it seems that the rather different structural

categories, true heathland and related shrubland, were necessarily

combined in one treatment because they sometimes intermix or

interdigitate. I feel, though, that their respective characteristics

needed clarifying by explicit definition and by consistent use of

photographs and standardized tables to illustrate each text. That

said, the editor has shown great skill in bringing so many different

regional accounts together in one volume. South African

ecologists will find many of the chapters, especially those on

Australian and European heathlands, of great value because the

information may be adapted and applied in practical fynbos

management or may suggest lines of future basic research.

Heathlands are so widely and patchily scattered throughout the

world, in such a variety of climates, that their essential unity has

not hitherto been recognized. This book, which provides the first

world-wide treatment of heathland ecosystems, will surely be a

standard reference work for many years, and even at its price of

nearly R100 must find a place in the library of every institution

where vegetation is studied.

H. C. Taylor

The Study of Vegetation edited by M. J. A. Werger. The

Hague : Dr W. Junk bv. 1979. Pp. xi + 316, 73 figures and 9

tables. Price $59.50.

This small book was compiled to celebrate the one-hundredth

meeting of the Commission for the Study of Vegetation of the

Royal Botanical Society of the Netherlands and was intended to be

a review of the developments in the various branches of vegetation

science with special reference to Dutch contributions. Although it

includes chapters on most aspects of vegetation science, the title is

perhaps a little pretentious, or at least misleading, in that coverage

is parochial as far as choice of authors and the emphasis of in-

dividual chapters are concerned. The contents of the book are not

being criticized here as they meet the aim, as elaborated above. My

criticism is directed at the title, from which one may, at first

glance, expect more than the book actually offers.

Dutch contributions to vegetation science through the years

have been far from insignificant and it is gratifying to see chapters

in this book by such respected figures as V. Westhoff and J. J.

Barkman, amongst others. Westhoff, who discusses the history,

present state and future of phytosociology in the Netherlands, was

an excellent choice for the review as he has had decades of first-

hand experience with the subject. Three of the eleven future

developmental trends in phytosociology which he lists are its use as

integrator of several aspects of nature conservation research, the

carrying out of basic ecological surveys and vegetation dynamic

studies in Third World countries and, thirdly, the major impact of

numerical data processing in the discipline. Vegetation texture and

structure are discussed in the contribution by Barkman. He defines

texture as the qualitative and quantitative composition of the

vegetation as to different morphological elements, whereas struc-

ture is concerned with the horizontal and vertical spatial arrange-

ment of these elements. In a clear and concise way he reviews the

extant systems of classification on the basis of texture and struc-

ture, points out their shortcomings and compares them with

selected floristic systems. He concludes with a brief account of his

system which makes use of many parameters already known in the

literature as well as growth forms that he has defined. It appears to

be a detailed, although rather complicated, system and we look

forward to its description by Barkman in extenso in the near

future.

A chapter on multivariate methods in phytosociology, with

reference to the Netherlands was contributed by E. van der

Maarel. The long sections on data transformations, resemblance

functions, classification and ordination add little that is new

except to place Dutch contributions to quantitative ecology, and

the author’s contributions in particular, in a world perspective.

Van der Maarel quotes three recently-published excellent reviews

and his repetition here of the same material appears largely un-

necessary. The review is complete as far as Dutch research is con-

cerned and it is interesting to see how much of relevance in the

field of multivariate methods has been either carried out or

directed by the author. The two other Dutch researchers who, in

my opinion, stand out in this field and whose work is discussed in

the review are W. G. Beeftink and Pauline Hogeweg. Unfortunate-

ly, the subject is advancing at such a rate that this chapter is

already out of date with respect to certain methods — it cannot be

otherwise in a developing field such as this is.

There are a total of eight review papers in the book. Apart from

those already mentioned, the subjects of autecology (W. H. van

Dobben), population biology (S. J. ten Borg), competition (J. P.

van den Bergh), palynology (C. R. Janssen) and nature conserva-

tion (P. A. Bakker) receive adequate treatment in this clear state-

ment on the status of Dutch plant ecology in the broadest sense.

Possibly the most striking feature about the book as a whole, is the

great difference between problems which merit attention in a

small, highly developed and densely populated country like the

Netherlands and those of countries such as ours. While we are still

deeply involved with regional surveys and the autecology of domi-

nant species, they are now investigating subtle interactions be-

tween components of vegetation, dispersal mechanisms in detail

and the biology of individual populations. The biome studies being

mooted at present should contribute materially towards greater

depth coming to our vegetation science. The editor and the Com-

mission for the Study of Vegetation are to be congratulated on the

production of a valuable synthesis on which the future of vegeta-

tion science in the Netherlands can be built.

J. W. Morris

Flowering Trees in Subtropical Gardens by GOnther Kunkel.

The Hague: Dr W. Junk bv. 1978. Pp. 346, line drawings 139,

photographs 6. Price US $36.85.

At the start of this book Gunther Kunkel apologizes for being

what I can only describe as a ‘soul scientist’, for having indulged

himself, including some species and leaving others out, for having

enjoyed his work and made no secret of it. For this there is surely

no need to apologize. His enthusiasm and delight in his subject,

together with the fine illustrative work of Mary Anne Kunkel, lift

this work above the mundane. If an apology is needed, it is for the

overly ambitious and misleading title of the book. The plants dealt

with are flowering trees, the gardens of the Canary Islands, where

they occur, are subtropical and many of the species are widely

cultivated in other subtropical areas, but the same could be said of

South African or Australian gardens, all of which have different,

if overlapping, cultivated floras.

This book has its origin in a previous (1969) publication, volume

1 . of ‘ Arboles Exoticos’ i.e. exotic trees of the Canary Islands, and

that is what it still is — and a very useful tool for botanist, hor-

ticulturahst and layman alike. It deals with 150 species from 98

genera and 48 families. No palms or large monocots have been in-

cluded. Each species is described and illustrated. The descriptions

are both useful and readable (a rare achievement). There are also

brief notes on propagation, cultivation, growth requirements and

uses, with references for further reading. General sections include

an index, bibliography, a list of English and American common

names, a list of Spanish and Canarian common names, a list of

trees native to the Canary Islands, and a key to the identification

of the exotic species treated in the book.

The illustrations are generally superb — both in the way they

portray their subjects, and in their technical execution. Fig. 1,

Magnolia grandiflora is a masterpiece, down to the leathery feel of

the leaves. However, a few of the illustrations do not do justice to

the artist, being: ‘scratchy’ e.g. Fig. 30, Ficus virens; over-inked

e.g. Fig. 34, Casuarina spp.; or not sufficiently representative of

the subject e.g. Fig. 90, Tipuana tipu and Fig. 117, Melicoccus bi-

jugatus. The photographs are poor, but fortunately this does not

detract from the general appeal of the book.

For anyone interested in the Canarian flora, I can recommend

this book with few reservations. For those who consider using it

under South African conditions, the reservations are greater. It

591

would, for example, be extremely misleading to try to identify

trees cultivated in South Africa using a key designed for the

Canary Islands, e.g. our Celt is africana and Erythrina lysistemon

(which are not included) would key out as Celtis australis and

Erythrina caffra, and only an expert would pick out the difference

from the illustrations. If used with discretion, however, this book

can be a distinctly useful means of indentification and reference

source. Of the species treated about 44% are commonly cultivated

here, 20% are reasonably common, 16% occasional, and 20% are

rare, seldom or never seen. It is perhaps in this last category that

this work can be most useful in South Africa, for it brings to mind

a number of potentially useful garden subjects.

M. J. Wells

Medicinal Plants of West Africa by Edward S. Ayensu.

Algonac : Reference Publications. 1978. Pp 330, 77 black and

white photographs and 51 line drawings. Price $29.95.

This book deals with 187 plant species that have been used for

medicinal purposes by the peoples of ‘West Africa’. I can find no

delimitation of the area included other than a map inside the cover

that shows an area roughly bounded by the northern tip of Came-

roon, Nigeria and Niger in the east, and including the western

bulge of Africa as far north as the northern tip of Mauritania in

the west. Nor can I find a firm basis for the inclusion of species,

other than a reference in Prof. Schultes’s foreword to the fact that

of the multitude of drug plants occurring in West Africa, Dr Ayen-

su has chosen ‘a frequently encountered part of the total’.

Therefore ‘Medicinal Plants of West Africa’ cannot be regarded as

a comprehensive reference work to plant medicinal uses in a

specific area or flora.

The author states, ‘The aim of this publication is to bring to the

attention of professional specialists, pharmacologists, bio-

chemists, health administrators and members of W. H. O. the

botanical characteristics of medicinal plants that have been used in

treating various ailments by traditional medical practitioners over

the years .... and it is hoped .... will supply the necessary ethos

to spur the medical community desirous of establishing the linkage

between traditional and modern medicine’. As such it seems to be

a plea not so much for scientific evaluation of traditional

medicines (which has been well under way for the past century) as

for a package acceptance of the role of traditional medicine.

The text on each plant has been handled systematically. Com-

mon names are classified according to country (but not tribe), and

uses are classified according to the part of the plant used (with

some observations on method of application e.g. ‘enema’). The

amount of text per species varies from two lines on Pteris togoensis

to more than three pages on Ricinus communis, and is well

referenced throughout. The book includes a useful glossary to

medical terms; a bibliography of 61 publications; a species index;

and, most useful of all, a medicinal index with plants classified

according to their uses. The line drawings are good, but the

photographs of herbarium sheets although recognizable are unac-

ceptably fuzzy and lacking in contrast and detail.

A comparison of this book with J. O. Kokwaro’s ‘Medicinal

Plants of East Africa’ is inevitable: both are recent publications by

researchers who have grown up in the regions concerned, the titles

are similar and even the covers both sport Erythrina — yet there

are many differences between the two books. Ayensu’s treatment

is the better illustrated and referenced, but the text is more or less

limited to a literature survey and deals with relatively few species

and is almost surgical in style and in the exclusion of tribal

references (which could have added much of interest and value).

Kokwaro’s work on the other hand, is far more comprehensive (in-

cluding over 1 300 species), reads well, is full of new material and

tribal references and includes (in addition to plant and medicinal

indexes) a useful vernacular index — but the text is usually limited

to a few lines on each species, includes only 15 literature references

and is only very sparsely illustrated. A combination of the two

approaches for both East and West Africa would have been

magnificent.

As it stands, South African botanists’ main interest in

‘Medicinal Plants of West Africa’ will be the overlaps with our

own flora and its uses. There are many such overlaps at both

species and generic level, and for this reason alone. Dr Ayensu’s

book is an essential tool for economic botanists in South Africa.

M. J. Wells

Rhodesian Wild Flowers. Illustrations by Margaret H.

Tredgold. Text by H. M. Biegel. Thomas Meikle Series No. 4.

Trustees of the National Museums and Monuments of Rhodesia.

1979. Pp. xvi + 77, 39 black and white drawings and 39 colour

plates. Price ZS10.00.

This book is the successor to the one of the same title by R. A. S.

Martineau and Margaret H. Phear published as the first in this

series as long ago as 1953. In the first ‘edition’ almost all of the il-

lustrations were made from pressed specimens. This ‘edition’ has

18 full colour illustrations made from living plants by Lady Tred-

gold (Margaret Phear), wife of a former Governor of Southern

Rhodesia. There is a brief foreword by R. A. S. Martineau. The

copy I have is soft-covered and a little larger than A4 size.

The first sentence of the text states that the book is for the

layman. There is a brief introduction with an unusually varied con-

tent written by the artist and then a page of notes for users by

H. M. Biegel of the National Herbarium, Salisbury, who compiled

the text. Then comes the main part of the book with alternating

text, which is interspersed with black and white line drawings, and

four colour plates. The book begins with grasses and ends with

composites thus following Engler’s system. A total of 260 species

is covered in the text by a short paragraph on each, giving the

Latin, English and vernacular names, some diagnostic characters

and habitat details, distribution, derivation of names and mention

of related species. The illustrations depict 221 species in colour

with up to nine per plate and 39 as black and white sketches

strategically scattered throughout the text. The book ends with two

pages noting characters as recognition aids to families and genera,

a short glossary and indexes to botanical and common names. The

colour plates are printed on a semi-gloss matt paper.

The book is an improvement on the first and is a welcome addi-

tion to the increasing literature covering the flora of the southern

half of Africa in coloured illustrations. South Africa itself is

becoming well endowed with such works, some alas of rather poor

quality. Further north we have recent works such as that of Plowes

& Drummond for Rhodesia (Zimbabwe) and Moriarty for Malawi.

One could be critical of the occasional rather ‘woolly’ paintings

that occur in the present book (cf. Plate 9). These lack clarity and

give more of an impression of the species than an exact reproduc-

tion. On the whole the colour work, which has a style all of its

own, is pleasing and is sufficient for identification. One would

have expected the artist to erase her pencil signature under the pen

signature as this spoils a number of the plates. The text is very

readable and will, I am sure, appeal to the layman.

I have said in a previous review that a reviewer is invariably

drawn into noting and commenting on errors and the general pro-

duction. The nightmare of any author is the way errors creep into a

book during typing and printing. This book has its share of these

gremlins. Two of the text figs (2 & 39) have their numbers missing,

spelling mistakes are fortunately few, e.g. rhmannii for rehmannii

on p.3, C. bulbisperum for bulbispermum on p.9, one pair of

leaflet on p.30. One or two botanical points need some mention.

Gladiolus natalensis is used instead of G. dalenii. Only ± 450

species are ascribed to the genus Erica in the winter-rainfall area of

the Cape where the true figure is nearer 600! Under Lapeirousia

the ‘tubers’ are noted as being edible when later on corms are cor-

rectly described. There is a curious inconsistency in the use of

family names. In the text one finds the old familiar names like

Gramineae, Leguminosae, Labiatae, Umbelliferae and Com-

positae, but also the modern Clusiaceae (Guttiferae).

There is only one real criticism of the work and that is the cross-

referencing between the drawings and the text which will be

exasperating at times for the layman. The plate and figure

numbers are hidden in the text and are not at all conspicuous. One

has to be sharp-witted about sorting out the plates and figures,

especially as they run close on each other. Matters would have

been so much better if Plate 7: 1 etc. or Fig. 23 had been

strategically placed before each species name. Take, for example,

the text figure on page 30 where there is no name given and the

reference is hidden in the text under Zornia glochidiata.

This book appeared within three years of the work Wild Flowers

of Rhodesia by D. C. H. Plowes & R. B. Drummond in which 193

colour photographs depict 150 species. The styles of the two books

are completely different, but essentially they have one thing in

common — to excite and interest the layman in the wealth of the

Zimbabwean flora. Both should go a long way towards this objec-

tive and both should be possessed and used by all flower en-

thusiasts to whom I would strongly recommend this latest book.

At its price it is good value.

E. G. H. Oliver

A Field Guide to the Natal Drakensberg by Pat Irwin. John

Akhurst & David Irwin. Durban: Natal Branch of the Wildlife

Society of Southern Africa. 1980. Pp. 249, 47 colour photographs,

numerous line drawings and 10 maps. Price R4,50.

Mountaineers, climbers, hikers, wild life enthusiasts and ‘city

slickers’ will all welcome this field guide to the Natal Drakensberg,

the first compact guide to be published on this subject. The guide

covers almost every aspect of the Drakensberg scene, for example

archaeology, early history, climate, weather, geology, geomor-

592

phology, soils, vegetation, birds, reptiles, amphibians, insects and

mammals. However, it is only Chapter 6 on the vegetation that will

be reviewed here.

The authors describe in fairly general terms the zonation of the

vegetation and the main communities pointing out that the en-

vironment becomes more severe with increase of altitude and this

is reflected in the growth forms of the plants. They also refer to the

effect of aspect on vegetation.

Thirteen colour photographs show some of the more familiar

Drakensberg plants, for example Erica cerinthoides, Cussonia

paniculata, Encephalartos ghellinckii, Buddleja salviifolia and

Phygelius aequalis.

Then follows a list of 47 trees and shrubs occurring in the

Drakensberg with their common names. Finally, 70 ‘common

flowers’ are illustrated rather crudely but effectively in black and

white. Unfortunately some of the plant names are incorrectly spelt,

for example Erica cerinthiodes instead of E. cerinthoides,

Harpechloa falx instead of Harpochloa falx and Watsonia den-

siforma instead of W. densiflora. Worse than the misspellings is a

glaring misidentification: the line drawing of ‘Myrica serrata’ is

really of Myrsine africana. The authors have not kept up with re-

cent name changes: Elyonurus argenteus is now Elionurus muticus

and Haemanthus magnificus is Scadoxus puniceus. Since 1978

Buddleia salviifolia must be spelt Buddleja salviifolia. It is a pity

that the authors did not ask a botanist to check the chapter on

vegetation.

In conclusion, Chapter 6 does give the reader a basic impression

of the vegetation and flora of the Natal Drakensberg and, in a

wide-ranging guide of this nature, this is surely all one can expect.

The authors and the Natal Branch of the Wildlife Society of

Southern Africa deserve congratulation for producing this useful

and long overdue guide to the Natal Drakensberg.

D. J. B. Killick

Atlas of the Netherlands Flora. Part 1 , Extinct and very rare

species, edited by J. Mennema, A. J. Quene-Botterenbrood &

C. L. Plate. ’ s-Gravenhage : Dr W. Junk. (English Edition). 1980.

Pp. 226, 332 maps. Dutch Guilders 125.

The publication of plant distribution maps for the Netherlands

was initiated by the Rijksherbarium, Leiden, almost 80 years ago.

Now this Atlas, first in a series of three, appears as the crowning

achievement of all these years of work and is a remarkable piece of

work.

The Atlas is divided into two main sections, the seven introduc-

tory chapters and the chapter containing the distribution maps and

notes on each species. There is a general introduction to the history

of production of the Atlas where an explanation is given of the

complications which arose from having to use records from two

different series of grid maps. This is followed by a chapter on the

physical geography and phytogeography of the Netherlands with

excellent maps and one on the history of plant-geographical in-

vestigations. A short chapter discusses the reliability of the older

literature (1550-1850) which was consulted for locality records.

There is one chapter on the Standard List of the Netherlands Flora

where the inventory code for the 1 399 species of the flora is ex-

plained. Then follow two short chapters on the collection and use

of the data and an explanation of the maps and accompanying

texts. A very useful adjunct is a bibliography to the whole in-

troduction separate from the final one at the end of the work.

The body of the Atlas is formed by the maps and text covering

333 extinct and very rare species. The distribution maps used are

based on the phytogeographical maps of the Institute for In-

vestigation of the Vegetation of the Netherlands (IVON) at the

Rijksherbarium. Herein lay one of the biggest problems en-

countered by the authors. Prior to 1950 the IVON maps used were

divided on a basic grid (40 x 25 km) into 48 hour-squares with

subdivisions into 16 quarter-hour-squares of 2,5 x 2,08 km. After

1950 the maps used the same basic grid divided into 40 squares and

each then into 25 kilometre squares. Prior to 1950 much work on

the distribution of species and their mapping had been done and

could not have been ignored by the authors. Therefore they had to

adopt the odd mapping pattern as seen in the Atlas where black

stars represent post 1950 records and green rectangles the pre-1950

records, the latter not coinciding with the new grid system.

The species are alphabetically arranged with two per page. Each

map is accompanied by a Dutch text and English summary. Each is

given an intriguing reference number, e.g. 0864 lb 20, the first part

(0864) being its registered number in the flora, the middle part (lb)

the ecological category and the last the hour-square frequency

classes for pre (2) and post 1950 (0).

This Atlas is a superb piece of work that has resulted from years

of careful and painstaking work and observation. As the basis for

flora conservation in the Netherland it must be invaluable. As an

example for botanists and conservationists in other parts of the

world it stands as a Fine goal to aspire to. Well-known to most of

us is the classic Atlas of the British Flora by Perring & Walters

(1962). Mention in the introduction is made in passing of the

recently produced Atlas of the Belgian and Luxemburg Flora,

which I have not had the opportunity of examining. Here in South

Africa we must be envious of the achievements of the above but

then we do have a very much richer flora (17 150 spp) and a cor-

respondingly poorer coverage in herbarium collections. Com-

parisons could be made with for example the flora of the Cape

Peninsula with its 2 622 species. To do a similar work for the Cape

Peninsula on the same scale would be a major task especially

seeing that we have so few botanists. One consolation is that we

have recently managed to produce a list of the rare and endangered

species for South Africa. The next step is an atlas!

The presentation and production of this Netherlands Atlas are

excellent. I have no criticisms of real value. There are the occa-

sional printers’ errors or perhaps misspellings. I recommend this

Atlas to all botanists interested in floras and flora conservation.

E. G. H. Oliver

593

INDEX— INDEKS

Acacia ataxacantha DC., 503

Acacia caffra — Aloe marlothii Variant, 207

Acacia caffra — Teucrium capense Savanna, 208

Acacia caffra (Thunb.) Willd., 503

davyi N.E.Br., 503

gerrardii Benth. var. gerrardii, 503

Acacia karroo — Acacia caffra — Teucrium capense Savanna, 211

Acacia karroo — Teucrium capense — Conyza podocephala Savanna, 208

Acacia karroo — Teucrium capense — Felicia muricata Savanna, 211

Acacia karroo — Teucrium capense Savanna Communities, 208

Acacia karroo Hayne, 503

nigrescens Oliv., 503

nilotica (L.) Willd. ex Del., 96

Senegal (L.) Willd., 96

seyal Del., 96

sieberana DC. var. woodii (Burtt Davy) Keay & Brenan, 504

tortilis (Forssk.) Hayne, 96

subsp. heteracantha (Burch.) Brenan, 504

Acacia xanthophloea — Dyschoriste depressa Community, 222

Adansonia digitata L., 504

Adina microcephala (Del.) Hiern var. galpinii (Oliv.) Hiern, 504

Afzelia quanzensis Welw., 504

Agave sp., 504

Albizia harveyi Fourn., 504

versicolor Welw. ex Oliv., 504

Albuca exuviata (Jacq.) Ker-Gawl., 113

filifolia (Jacq.) Ker-Gawl., 113

fugax Ker-Gawl., 113

Allophyllus decipiens Radik., 504

Aloe davyana Schonl. var. davyana, 504

reitzii Reynolds var. vemalis Hardy, 451

Amaranthus cruentus L., 504

thunbergii Moq., 504

Annona senegalensis Pers., 504

Anthericum sensu Jacq., Ill

exuviatum Jacq., 113

filifolium Jacq., 113

fragrans Jacq., 113

spiratum Thunb., 113

Antidesma venosum E. Mey. ex Tul., 504

Aptosimum lineare Marloth & Engl., 504

Arachis hypogea L., 504

Arnold, T. H. Ficinia lucida: an interspecific hybrid between

F. cedarbergensis and F. ixioides subsp. glabra, 439

Arnold, T. H. & Vorster P. Ecological adaptations and possible conver-

gence in Ficinia arenicola var. erecta and Mariscus durus, 441

Artabotrys brachypetalus Benth., 504

Asclepias burchelli Schltr., 504

Aspalthium Medik., 318

acaulis (Stev.) Hutch., 318

frutescens Medik., 318

herbaceum Medik., 318

Asparagus virgatus Bak., 504

Athrixia phylicoides DC., 504

Balanites maughamii Sprague, 505

Bambusa sp., 505

Basilaea coronata Lam., 140

regia sensu Mirb., 140

undulata (Ait.) Mirb., 140

Bauhiiiia galpinii N.E. Br., 505

Bequaertiodendron magalismontanum (Sund.i Heine & J . H. Hemsl., 505

Berchemia discolor (Klotzsch) Hemsl.. 505

Berkheya setifera — Digitaria diagonalis Grassland, 214

Berkheya setifera — Hyparrhenia hirta Grassland, 213

Berkheya setifera — Koeleria cristata — Pentanisia prunelloides Grassland,

213

Berkheya setifera — Rhynchelytrum sctifolium Grassland, 213

Bidens pilosa L., 505

Bituminaria Heist, ex Fabricus, 317, 318

subgen. Bituminaria, 318

subgen. Christevenia Barneby ex C. H. Stirton, 318

acaulis (Stev.) C. H. Stirton. 318

bituminosa (L.) C. H Stirton. 318

Blacria ciliaris, L.f., 77

ciliciiflora (Salisb.) G. Don, 79

incana Barth, 81

ptilota E. Mey. ex Benth.. 73

Blumea aurita (L.f.) DC., 505

Bolusanthus speciosus (H. Bol.) Harms, 505

Book Reviews, 259, 589

Boucher, C. Notes on the use of the term ‘Renosterveld’, 237

Brachylaena discolor DC. subsp. transvaalensis (Phill. & Schweick.)

J. Paiva, 505

Calcrete Fynbos, 522

Bredenkamp, G. J. & Theron, G K. A synecological account of the

Suikerbosrand Nature Reserve. II. The phytosociology of the Venters-

dorp Geological System, 199

Bridelia micrantha (Hochst.) Baill., 505

mollis Hutch., 505

Brownleea pentheriana Kraenzl. ex Zahlbr., 355

Burkea africana Hook., 173, 505

Cajanus cajan (L.) Millsp., 505

Calodendrum capense (L.f.) Thunb., 505

Campbell, B. M. & Moll, E. J. Determination of plot size, 575

Capparis fascicularis DC. var. fascicularis, 505

tomentosa Lam., 505

Capsicum frutescens L., 505

Carissa edulis Vahl, 505

Cassia abbreviata Oliv. subsp. beareana (Holmes) Brenan, 505

occidentals L., 505

Cassine aethiopica Thunb., 505

transvaalensis ( Burtt Davy) Codd, 505

Catha edulis (Vahl) Forssk. ex Endl., 506

Cephalanthus natalensis Oliv., 506

Ceropegia flanaganii Schltr. var. alexandrina Huber, 436

flanaganii Schltr. var. fallax Huber, 436

Clematis brachiata Thunb., 506

Cocculus hirsutus (L.) Diels., 506

Codd, L. E. Obituary: Robert Harold Compton (1886-1979), 244

Codex Bentingiana, 115

Codex Comptoniana, 1 15

Codex Witsenii, 115

Cola Schott & Endl., 277

greenwayi Brenan, 279

mtcrocarpa Brenan, 279

natalensis Oliv., 277

Colocasia antiquorum Schott, 506

Colophosperum mopane (Kirk ex Benth.) Kirk ex J. Leonard, 506

Combretum apiculatum Sond., subsp. apiculatum, 506

erythrophyllum (Burch.) Sond., 506

hereroense Schinz subsp. hereroense, 506

imberbe Wawra, 506

molle R. Br. ex G. Don, 174

paniculatum Vert, subsp. microphyllum (Klotzsch) Wickens, 506

zeyheri Sond., 174

Commelina bella Oberm., 436

diffusa Burm. f. subsp. scandens (C.B. Cl.) Oberm.. 437

modesta Oberm., 437

scandens Welw. ex. C.B. Cl. 437

Commiphora pyracanthoides Engl. 506

Communities with northerly aspects, 558

Communities with southerly aspects, 561

Corchorus confusus Wild, 506

tridens L., 506

Cordia grandicalyx Oberm., 506

ovalis R. Br. ex DC., 506

Crocosmia pearsei Oberm., 450

Crossopteryx febrifugia (Afzel. ex G. Don) Benth., 506

Crotalaria psoraleoides Lam., 293

Croton megalobotrys Muell. Arg., 506

Cryptolepis caper. sis Schltr., 506

Cucumis melo L.. 507

Cullen Medik, 317

bifiora (Har\\) C. H. Stirton, 317

coryiifolia (L.) Medik. 317

dr„oacea (Bunge) C. H. Stirton, 317

hotubii (Burtt Davy) C. H. Stirton, 317

jauoertina (Fenzl) C. H. Stirton, 317

obtusifolia (DC.) C. H. Stirton, 317

plicata (Del.) C. H. Stirton, 317

Curcurbita maxima Duch., 507

pepo L., 507

Cussonia spicata Thunb., 507

Cyanotis pachyrrhiza Oberm., 437

robusta Oberm., 438

Cymbopogon validus (Stapf) Stapf ex Burtt Davy, 507

Cynodon dactylcn Community, 226

Cyperus buliatu ; Kiikenth., 444

capensis (Sieud1. Endl. var. polyanthemus Kiikenth., 444

rhersinus (N.E. Br.) Kiikenth., 444

Cyperus fastigiatus — Echinochloa pyramidalis Community, 226

Cyperus latifolius Pair. 507

sexangularis Nees, 507

Cyphostemma hardyi Retief, 460

humile (N.E. Br.) Desc. ex Wild & Drum, subsp. humile, 507

Cyrtanthus labiatus R. A. Dyer, 135

Dalbergia melanoxylon Guill. <& Perr., 507

Delosperma tuberosum (L.) Schwant., 454

Dicerocaryum zanguebanum (Lour.) Merr. subsp. zanguebarium, 507

Dichapetalum cymosum (Hook.) Engl., 175

Dicharostachys cinerea (L.) Wight & Arn., 175

\

/

594

Digitaria bechuanica (Stent) Henr., 457

decumbens Stent, 457

diversinervis (Nees) Stapf 457

var. woodiana Henr., 457

eriantha Steud., 457

subsp. eriantha, 457

subsp. pentzii (Stent) Kok. 457

subsp. stolonifera (Stapf) Kok, 457

subsp. transvaalensis Kok, 457

geniculata Stent, 457

hiascens Mez, 457

macroglossa Henr., 457

var. prostrata (Stent) Henr., 457

monodactyla (Nees) Stapf var. explicata Stapf, 457

natalensis Stent, 457

subsp. stentiana Henr., 457

subsp. stentiana Henr. var. paludicola Henr., 457

pentzii Stent, 457

subsp. dregeana Henr., 457

var. stolonifera (Stapf) Henr., 457

polevansii Stent, 457

rigida Stent, 457

seriata Stapf, 457

setivalva Stent, 457

smutsii Stent, 457

stentiana Henr., 457

valida Stent, 457

subsp. burchelliana Henr., 457

var. glauca Stent, 457

Dichrostachys cinerea (L.) Wight & Am. subsp. africana Brenan &

Brummitt, 507

Dioscorea cotinifolia Kunth, 507

Diospyros mespiliformis Hochst. ex A. DC., 507

natalensis (Harv.) Brenan subsp. natalensis, 507

Dipogon Liebm., 327

benthamii Meisn., 327

lignosus (L.) Verde., 327

Disa Berg., 368

sect. Forficaria (Lindl.) Schltr., 368

sect. Herschelia (Lindl.) H. Bol., 368

sect. Michroperistera H. Bol., 368

sect. Monadenia (Lindl.) H. Bol., 342

sect. Spathulatae Kraenzl., 368

sect. Trichochila Lindl., 368

affinis N.E. Br., 351

atropurpurea Sond., 372

atrorubens Schltr., 354

auriculata H. Bol., 348

barbata (L.f.) Swartz, 376

baurii H. Bol., 383

bolusiana Schltr., 353

bracteata Swartz, 346

brevicornis (Lindl.) H. Bol., 360

cernua (Thunb.) Swartz, 359

charpentieriana Reichb. f., 379

comosa (Reichb. f.) Schltr., 351

conferta H. Bol., 343

densiflora (Lindl.) H. Bol., 348

excelsa sensu Lindl., 381

forcipata Schltr.. 382

forficaria H. Bol., 368

gramiifolia Ker-Gawl. ex Spreng., 374

hamatopeiula Rendie, 383

hians (L.f.) Spreng., 381

lacera Swartz, 381

longilabris Schltr., 383

lugens H. Bol., 379

macroglottis Sond. ex Drege, 379

micrantha Lindl. (H. Bol.), 346

multifida Lindl., 379

multiflora (Sond.) H. Bol., 348

newdigateae L. Bol., 370

ophrydea (Lindl.) H. Bol., 355

outeniquensis Schltr., 381

physodes Swartz, 358

praetermissa Schltr., 346

prasinata Ker-Gawl.. 359

propinqua Sond., 372

purpurascens H. Bol , 375

pygmaca H. Bol., 345

reticulata H. Bol., 351

rufescens (Thunb.) Swartz, 356

sabulosa H. Bo!., 344

schlechterana H. Bol., 371

spalhulata (L.f.) Swartz, 372

var. atropurpurea (Sond.) Schltr., 372

venusta H. Bol., 378

tripartita Lindl., 373

walteri Schltr., 385

Dodonea viscosa Jacq. var. viscosa, 507

Dolichos L. sect. Eudolichos Taub. subsect. Barbatae Taub., 327

sect. Pogonodolichos Harms subsect. Gibbosi Harms, 327

capensis sensu Thunb., 327

gibbosus Thunb., 327

gibbosus Thunb. var. uniflorus Harv., 327

jacquinii sensu Piper, 327

lignosus L., 327

Dolneu’s Florilegium, 115

Dombeya rotundifolia (Hochst.) Planch., 176

var. rotundifolia, 507

Dovyalis zeyheri (Sond.) Warb., 507

sensu Jessop, 1 1 1

altissima (L.f.) Ker-Gawl., 139

Drimia altissima sensu Ker-Gawl., 453

angustifolia Bak., 139

duthieae (Adamson) Jessop, 113

exuviata (Jacq.) Jessop, 113

exuviata sensu Jessop, 113

forsteri (Bak.) Oberm., 453

multifolia (Lewis) Jessop, 114

Dune Woodland, 520

Dyer, R. A. A new species of Cyrtanthus from Baviaanskloof, south-east-

ern Cape, 135

Dyer, R. A. A new species of Huernia from Owambo, 136

Dyer, R. A. The Riocreuxia flanaganii complex: A reassessment, 435

Dyer, R. A. Obituary: Adolf Joseph Wilhelm Bayer (1900-1978), 244

Edwards, D. A note on the extension of the degree reference system for

citing biological distribution records to north of equator and west of

Greenwich mendan, 574

Ekebergia capensis Sparrm., 507

Ellis, R. P. Leaf anatomy of the South African Danthonieae (Poaceae)

II. Merxmuellera disticha, 185

Ellis, R. P. Leaf anatomy of the South African Danthonieae (Poaceae)

IH. Merxmuellera stricta, 191

Ellis, R. P. Leaf anatomy of the South African Danthonieae (Poaceae)

IV. Merxmuellera drakenbergensis and M. sterophylla, 487

Ellis, R. P. Leaf anatomy of the South African Danthonieae (Poaceae)

V. Merxmuellera macowanii, M. davyi and M. aureocephala, 493

Epaltes gariepina (DC.) Steetz, 507

Eremia parviflora Klotzsch, 86

Erica ciliciiflora Salisb., 79

insignis E. G. H. Oliver, 446

plumosa Thunb., 77

shalliana Hort. Berol. ex Klotzsch, 86

Eriocaulon dregei Hochst. var. sonderanum ( Koern .) Oberm., 450

sonderanum Koern., 450

Eriochloa meyerana (Nees) Pilg. , 457

subsp. grandiglumis (Stent & Rattray ) Gibbs Russell, 457

Eriosema G. Don, 321

buchananii Bak. /., 297

var. richardii (Bak. f. & Haydon) Staner forma ellipticum

(Staner & De Craene) Staner, 299

cajanoides (Guill. & Perr.) Hook, f., 293

capitatum E. Mey., 321

cordatum E. Mey., 288

cordatum E. Mey. x E. kraussiana Meisn., 314

cordatum E. Mey. X E. preptum C. H. Stirton, 314

cordatum E. Mey. x E. salignum £. Mey. (red and yellow flowered

form), 313

cordatum £. Mey. x E. salignum £. Mey. (yellow-flowered form),

312

guenzii Sond., 288

guenzii (Sond.) Harv., 288

gunniae C. H. Stirton, 321

lucipetum C. H. Stirton, 291

nutans Schinz, 299

parviflorum £. Mey., 302

polystachyium E. Mey., 293

polystachyum sensu auett., 299

preptum C. H. Stirton, 323

proschii Briq., 293

psoraleoides (Lam.) G. Don, 293

richardii Bak. f. & Haydon, 299

var. ovatum Staner & De Craene, 297

forma ellipticum Staner & De Craene, 299

Eriosema transvaalense C. H. Stirton X E. angustifolium Burn Davy,

314

transvaalense C. H. Stirton, 323

zuluense C. H. Stirton, 288

Erythrina lysistemon Hutch., 507

Eucalyptus spp., 507

Euclea crispa (Thunb.) Guerke var. crispa, 508

divinorum Hiem, 508

natalensis A. DC., 176

Euclean crispa — Buddleja saligna — Acacia caffra Bush, 207

Euclea crispa — Buddleja saligna Variant, 208

Euclea crispa — Maytenus heterophylla — Setana nigrirostris Savanna, 205

Euclea crispa — Maytenus polyacantha — Canthium gilftllanii Bush and

Savanna Communities, 205

Euclea crispa — Maytenus polyacantha — Scolopia zeyheri Bush, -07

Euclea crispa — Rhoicissus tridentata Bush and Savanna Communities, 200

595

Eucomis albomarginata Barnes, 141

amaryilidifolia Bak., 141

autumnalis (Mill.) Chin., 140

subsp. amaryilidifolia (Bak.) Reyneke, 141

subsp. autumnalis, 140

subsp. clavata (Bak.) Reyneke, 141

clavata Bak., 141

clavata sensu Van der Spuy, 140

robusta Bak., 141

undulata Ait., 140

undulata sensu Lettv, 141

macrophylla hort., 140

Eugenia capensis (Eckl. & Zeyh.) Sond. subsp. natalitia (Sond.)

F. White, 143

natalitia sensu Palmer & Pitman, 143

woodii Diimmer, 143

Eulophia hians (L.f.) Spreng., 381

Eumorpha elegans sensu Eckl. & Zeyh., 431

Fadogia monticola Robyns, 177

Faurea saligna Harv., 508

speciosa ( Welw .) Welw., 508

Ficinia arenicola Arnold & Gordon-Gray var. erecta Arnold & Gordon-

Gray, 441

cedarbergensis Arnold & Gordon-Gray, 439

ixioides Nees subsp. glabra Arnold & Gordon-Gray, 439

Ficinia lateralis Sedgeland of calcrete gravel, 520

Ficus burkei ( Miq .) Miq., 508

capensis Thunb., 508

capreaefolia Del., 508

soldanella Warb., 508

sonderi Miq., 508

stuhimannii Warb., 508

sycomorus L., 508

Ficus sycomorus — Acacia xanthophloea Subcommunity, 222

Ficus sycomorus — Eriochloa meyeriana Subcommunity, 221

Ficus sycomorus — Rauvolfia caffra Community, 221

Flacourtia indica ( Burnt . {.) Merr., 508

Flora Capensis, J. & J. P. Breyne, 115

Forest Precursor Communities, 520

Forficaria Lindl., 368

graminifolia Lindl., 368

Fritillaria autumnalis Mill., 140

longifolia Hill, 140

Fumess, H. D. & Breen, C. M. The vegetation of seasonally flooded areas

of the Pongola River Floodplain, 217

Garcinia livingstonei T. Anders, 508

Gardenia spatulifolia Stapf ex Hutch., 508

Geranium elegans Andr., 431

Gethyllis namaquensis (Schonl.) Oberm., 136

Gibbs Russell, G. E. A new combination in Eriochloa, 457

Gibbs Russell, G. E. & Robinson, E. R. Phytogeography and speciation in

the vegetation of the eastern Cape, 467

Gladiolus microcarpus Lewis subsp. italaensis Oberm., 451

Glen, H. F., Nomenclature in the genus Mestoklema, 455

Gnidia rubescens B. Peterson, 508

Gossypium herbaceum L. var. africanum (Watt)J. B. Hutch & Ghose, 508

Grassland, 524

Grewia spp., 508

flavescens Juss., 177

var. flavescens, 508

var. olukundae (Schinz) Wild, 508

occidentals L., 508

Grisebachia Klotzsch, 69

alba N. E. Br., 81

apiculata N. E. Br., 79

bolusii N. E. Br. (E. G. H. Oliver), 78

ciliaris Benth. non Klotzsch, 84

ciliaris (L.f.) Klotzsch, 74

subsp. bolusii (N. E. Br.) E. G. H. Oliver, 78

subsp. ciliaris, 77

subsp. ciliciiflora (Salisb.) E. G. H. Oliver, 79

subsp. involuta ( Klotzsch ) E. G. H. Oliver, 79

subsp. multiglandulosa E. G. H. Oliver, 80

ciliciiflora (Salisb.) Druce, 79

dregeana Benth., 77

var. vestita, Zahlbr., 73

eremioides MacOwan, 86

var. eglanduia N'.E. Br., 88

var. pubicalyx N.E. Br. , 86

hispida (Klotzsch) E. G. H. Oliver, 72

incana (Bartl.) Klotzsch, 81.

involuta Klotzsch, 79

minutiflora N.E. Br., 89

subsp. minutiflora. 89

subsp. nodiflora (N.E. Br.) E. G. H. Oliver, 90

nivenii N.E. Br., 84

parviflora (Klotzsch) Druce, 85, 86

subsp. eglanduia (N.E. Br.) E. G. H. Oliver, 88

subsp. parviflora, 86

subsp. pubescens E. G. H. Oliver, 89

pentheri Zahlbr., 73

pilifolia N.E. Br., 73

plumosa Klotzsch, 69.

subsp. eciliata E. G. H. Oliver 72

subsp. hirta (Klotzsch) E. G. H. Oliver, 74

subsp. hispida (Klotzsch) E. G. H. Oliver, 72

subsp. irrorata E. G. H. Oliver, 72

subsp. pentheri (Zahlbr.) E. G. H. Oliver, 73

subsp. plumosa, 71

var. scabra N.E. Br., 73

var. serrulata (Benth.) N.E.Br., 71

rigida N.E. Br., 83 __

secundiflora E. G. H. Oliver, 91

serrulata Benth., 71

similis N.E. Br., 86

var. grata N.E. Br., 86

solivaga N.E. Br., 71

thunbergii Rach, 77

velleriflora Klotzsch, 79

zeyheriana Klotzsch, 79

Guy, P. R. Changes in the herb layer of the riverine woodland in the

Sengwa Wildlife Research Area, Zimbabwe, 527

Hardy, D. S. & Reid, C., A new variety of Aloe from the Vryheid District,

451

Hermania L., 1

subgen. Acicarpus Harv., 1

subgen. Euhermannia Harv., 1

subgen. Hermannia, 1

abrotanoides Schrad., 56

affinis K. Schum., 20

alnifolia L., 44

althaeifolia L., 29

althaeoides Link, 28

amabilis Marloth ex K. Schum., 15

amoena Dinter ex M. Holzhammer-Friedrich, 33

angolensis K. Schum., 10

angularis Jacq., 35

angularis sensu Eckl. & Zeyh., 35

arabica Hochst. & Steud. ex Fisch., 16

argentea Sm., 57

argyrata Presl., 22

asbestina Schltr., 21

aspera Wendl., 47

aspericaulis Dinter & Engl., 21

atrosanguinea Dinter, 16

ausana Dinter ex Range, 31

bicornis Eckl. & Zeyh., 20

bipinnata Burch., 56

bolusii Szyszyl., 57

boranginflora Hook., 11

brachypetala Harv., 9

bracteosa Presl, 61

brandtii Engl, ex Dinter, 10

bryonifolia sensu Eckl. & Zeyh., 33

bryonifolia Burch., 26

cana K. Schum., 57

candicans Ait., 28

var. discolor Harv., 31

var. incana Harv., 31

candicar.s sensu Harv., 28

candissina Spreng. f., 32

cavamllesiana Eckl. & Zeyh., 37

chnsanthemifolia E. Mey. ex Harv., 60

chrysophylla Eckl. & Zeyh., 36, 49

collina Eckl. & Zeyh., 22

comosa Burch, ex DC., 31

var. crenata K. Schum., 31

var. minor K. Schum., 31

complicata Engl., 19

concinnifolia Verdoorn, 25

confusa Salter, 58

conglomerata Eckl. <£ Zeyh., 48

cordifolia Harv., 27

coronopifolia sensu Eckl. & Zeyh., 59

cristata H. Bol., 7

var. ge aides Beauv., 7

cuneifolia Jacq., 40

var. cuneifolia, 41

var. glabrescens (Harv.) Verdoorn, 41

cuneifolia sensu Harv., 43

damarana Bak. /., 18

decipiens E. Mey. ex Harv., 49

decumbens Willd. ex Spreng., 22

var. argyrata (Presl) Harv., 22

var. collina (Eckl. & Zeyh.) Harv., 22

var. hispida Harv., 22

denudata L.f., 52

var. denudata, 52

denudata sensu Eckl. & Zeyh., 54

desortorum Eckl. & Zeyh., 42

- !>

(

dinteri Engl., 14

dinteri Schinz, 31

discolor Otto & Dietr., 28

disermifolia Jacq., 33

disermifolia sensu Eckl., & Zeyh., 22

dissecta Harv., 60

disticha Schrad., 48

diversistipula Presl ex Harv., 39

var. diversistipula, 40

var. gracilifolia Verdoorn, 40

dryadiphylla (Eckl. & Zeyh.) Dtuce, 44

dryadiphylla (Eckl. & Zeyh.) Harv., 44

eenii Bak. /., 10

engleri Schinz, 14

ernesti-ruschii Dinter ex M. Holzhammer- Friedrich, 33

exstipulata E. Mey., 18

falcata Eckl. & Zeyh, 53

fasciculata Bak., 17

filifolia L.f ., 51

var. filifolia, 51

var. grandicalyx Verdoorn, 51

var. passerinoides Harv., 51

var. robusta Verdoorn, 52

filipes Harv., 16

var. eliator K. Schum., 16

flammea Jacq., 53

flammula Harv., 53

floribunda Harv., 26

floribunda sensu K Schum., 29

fruticulosa K. Schum., 21

gariepina Ecki & Zeyh., 18

var. dentata Engl., 18

var. integrifolia Engl., 18

gilfillanii N.E. Br., 51

glabripetala Engl., 31

glanduligera K. Schum., 11

glandulosissima Engl., 12

glomerata E. Mey., 48

gracilis Eckl. & Zeyh., 39

grisea Schinz, 13

guerkeana K. Schum., 14

halicacaba DC., 56

helianthemum K. Schum., 18

helicoidea Verdoorn, 46

hereroensis Schinz, 14

hilaris (Eckl. & Zeyh.) Hochr., 20

hirsuta Mill, 29

hirsuta Schrad., non Cav., 47

hirsuta Schrad. & Wendl., 45

hirsuta sensu Eckl. & Zeyh., 23

hispidula Reichb., 27

holosericea Jacq., 37

holubii Bum Davy, 16

hyssopifolia L., 35

var. integerrima Schinz., 36

imbricata Eckl. & Zeyh., 25

incana Cav., 31

incana sensu Thunb., 37

incisa Willd., 61

intricata Adamson, 48

involucrata Cav.. 49

involucrata sense Eckl. & Zeyh., 35

jehannisburgiana Engl., 9

iohanssenii N.E. br.. 30

joubertiana Harv., 54

juttae Dinter & Engl., 60

karakowisensis ined., 10

kirkii Mast., 16

lancifolia Szyszyl., 10

latifolia Jacq., 36

lavandulifolia L., 37

lepidota Buch. ex Krauss, 41

leucanthermoides Presl, 58

leucophylla Presl, 37

lindequistii Engl., 18

linearifolia Harv., 17

lir.ifolia Burm. /., 22

linifolia sensu Eckl. & Zeyh., 51

longiramosa Engl., 10

lugardii N.E. Br., 16

macra Schltr., 62

melissifolia Engl., 26

membraniflora Schltr., 41

merxmuelleri M. Friedrich, 8

micans Schrad., 36

micrantha Adamson, 48

micropetala Harv., 15

mildbraedii Dinter & Engl., 17

minimifolia M. Holzhammer, 14

minutiflora Engl., 29

modesta (Ehrenb.) Mast., 16

var. elatior (K. Schum.) K. Schum., 16

subvar. brevicornis Engl., 16

subvar. macropetala Engl., 16

subvar. mediipetala Engl., 16

subvar. virgatissima Engl., 16

var. tsumebensis Engl., 16

mollis Willd., 31

mollis sensu Eckl. & Zeyh., 28

mucronulata Turcz., 34

multifida DC., 56

multiflora Jacq., 43

muncata Eckl. & Zeyh., 44

muirii Pilians, 25

myrioclada Diels., 61

myrrhifolta Thunb., 58

nemorosa Eckl. & Zeyh., 28

nivea Schinz, 18

odorata Ait., 38

oligantha Salter, 46

orophila Eckl. & Zeyh., 53

orophila sensu Eckl. & Zeyh., 56

paniculata E. Mey., 59

pollens Eckl. & Zeyh. var. glabrescens Harv., 41

passerinaeformis Eckl. & Zeyh., 51

pateillicalyx Engl., 42

patula Harv., 54

paucifolia Turcz., 60

var. chrvsanthemifolia (E. Mey. ex Harv.) Kuntze ex K. Schum.,

60

var. intemedia Kuntze ex K. Scnum., 60

pedunculata Phill., 54

pfeilii K. Schum., 42

phaulochroa K. Schum., 15

pillansii Compton, 50

pinnatisecta Salter, 58

var. auriculata Salter, 58

plicata Ait., 29

plicata sensu Eckl. & Zeyh., 28

polymorpha Eckl. c£ Zeyh., 53, 61

praemorsa Wendl., 31

pratensis Eckl. & Zeyh., 22

presliana Turcz., 54

prismatocarpa E. Mey. ex Harv., 23

procumbens Cav., 57

subsp. procumbens

subsp. myrrhifolia (Thunb.) de Wint., 58

pseudo-mildbraedii Dinder & Engl., 17

pulverata AruJr., 57

racemosa E. Mey., 18

rehmannii Szyszyl., 26

rehobothensis M. Holzhammer-Friedrich, 60

repetenda Verdoorn, 45

rhopalostylys K. Schum. & Schltr., 59

rigida Harv., 46

rotundifolia Jacq., 49

rudis N.E. Br., 56

rudis var. exserta N.E. Br., 56

rugosa Adamson, 28

salvifolia sensu Eckl. & Zeyh., 34

salviifolia L.f., 35

var. salviifolia, 36

var. ovalir Harv., 36

var. grandistipula Harv., 36

var. oblonga Harv., 36

salviifolia sensu Cav., 49

secundiflora Eckl. & Zeyh., 41

sandersonii Harv., 8

scabra Cav., 54

scabra sensu Jacq., 47

scoparia (Eckl. & Zeyh.) Harv., 22

scordifolia Jacq., 24

var. integriuscula Harv., 24

scordifolia sensu Eckl. & Zeyh., 54

seineri Engl., 9

var. latifolia Eng!., 10

seitziana Engl., 14

sideritifoliu Engl., 13

solanillora K. Schum., 13

spinosa E. Mey. ex Harv., 21

spinulosa Engl., 20

squarrosa Dinter ex Range, 20

stipulacea Lehm. ex Eckl. 4 Zeyh., 50

suavis Presl ex Harv., 35

sulcata Harv., 38

tenella Dinter & Schinz, 17

tenuifolia sensu Eckl. & Zeyh.. 59

tenuipes Eng., 15

tephrocarpa K. Schum., 11

ternifolia Presl, 24

I \ I

i

*

I

597

temifolia Presl ex Harv., 24

tigrensis Hochst. ex A. Rich Ctigreensis’), 16

tomentoa (Turcz.) Schinz ex Engl., 9

var. brevifolia Engl., 9

trifoliata L., 24

trifurca L., 20

trifurca sensu Eckl. & Zeyh., 53, 55

iruncata Schinz, 18

urceolata Pillans MS., 61

velutina DC., 38

vestita Thunb., 32

viscosa Hiern, 12

viscosa sensu Burtt Davy, 11

windhukiana Engl., 20

zeyheriana Presl, 58

Mahcrma L., 1

subgen. Mahemia (L.) K. Schum., 1

aryadiphylla Eckl. & Zeyh., 44

hilaris Eckl. & Zeyh., 20

incana Eckl. & Zeyh., 20

odorata Andr., 38

odorata sensu Eckl. & Zeyh., 38

pinnaia sensu Eckl. & Zeyh., 59

scoparia Eckl. & Zeyh., 22

var. glabra Eckl. & Zeyh., 22

spinosa Burch, ex DC., 21

tomentosa Turcz.. 9

Tricanthera modesta Ehrenb., 16

Herschelia Lindl.. 368

atropurpurea (Sond.) Rolfe, 372

bachmanniana Kraenzl., 383

barbata (L.f.) H. Bol., 376

buarii (H. Bol.) Kraenzl., 383

charpentierana (Reichb. f.) Kraenzl., 379

chimanimaniensis Linder, 384

coelestis Lindl., 374

excelsa sensu Rolfe, 381

forcipata (Schltr.) Kraenzl., 382

forficaria (H. Bol.) Linder, 368

goetzeana Kraenzl., 385

graminifolia ( Ker-Gawl . ex Spreng.) Dur. <£• Schinz, 374

hamatopetala (Rendle) Kraenzl., 383

hians (L.f.) Hall, 381

lacera (Swartz) Fourc., 381

longilabris (Schltr.) Rolfe, 383

lugens (H. Bol.) Kraenzl., 378

multifida (Lindl.) Rolfe, 379

newdigateae (L. Bol.) Linder, 370

praecox Linder, 382

purpurascens (H. Bol.) Kraenzl., 375

schlechterana (H. Bol.) Linder, 371

spathulata (L.f.) Rolfe, 371

subsp. spathulata, 372

subsp. tripartita (Lindl.) Linder, 373

venusta (H. Bol.) Kraenzl., 378

tripartita (Lindl.) Rolfe, 373

Heteropogon contortus (L.) Beauv. ex Roem. & Schultr., 508

Heteropyxis natalensis Harv., 509

Hexalobus monopetalus (A. Rich.) Engl. & Diels, 509

Hibuscus cannabinus L., 509

Holothrix aspera (Lindl.) Reichb. {., 456

brevipetala Immelman <£ Scheipe, 455

cemua (Burnt, f.) Scheipe. 456

condensate Sond., 456

confusa Rolfe, 456

culveri H. Bol., 456

var. Integra H. Bol., 456

exilis Lindl.. 456

var. brachylabris (Sond.) H. Bol., 456

filicomis Immelman & Scheipe, 455

hispidula sensu H. Bol., 455

incurva Lindl., 456

lindleyana Reichb. f., 456

lithophila Schltr., 456

muhisecta H. Bol., 456

parviflora (Lindl.) Reichb. /., 456

pan'ifolia sensu Rolfe, 455

rupicola Schltr., 456

scopularia (Lindl.) Reichb. /., 456

squumulosa Lindl. var. glabra H. Bol., 456

var. hirsuta H. Bol., 456

var. scabra H. Bol., 456

var. condensata (Sond.) Immelman, 456

Huemia owamboensis R. A. Dyer, 136

Hyparrhenia dichroa (Steud.) Stapf, 509

dregeana (Nees) Stapf ex Stent, 509

hirta (L.) Stapf, 509

rudis Stapf, 509

tamba (Hochst. ex Steud.) Anders, ex Stapf, 509

Hyperthelia dissoluta (Nees) Clayton, 509

Hyphaene natalensis Kunze, 509

Immelman, Kathleen. Notes on South African Species of Holothrix, 455

Ipomoea batatas (L.) Lam., 509

Jatropha curcas L., 509

Kigelia africana (Lam.) Benth., 509

Killick, D. J. B. Obituary: John Phillip Harison Acocks (191 1-1979), 239

Klingia namaquensis Schonl., 136

Kok, P. D. E. Notes on Digitaria in South Africa, 457

Lagenaria siceraria (Molina) Standi., 509

Landolphia kirkii Dyer, 509

Lannea discolor (Sond.) Engl., 177, 509

edulis (Sond.) Engl., 178, 509

rugosa Thunb., 509 —

Lebeckia Thunb., 318

waltersii C. H. Stirton, 319

Leonotis sp., 509

Ltengme, C. A. Plants used by the Tsonga people of Gazankulu, 501

Liliaceae Subtribus Caudibracteateae Oberm., 137

Limodorum hians (L.f.) Thunb., 381

Linder, H. P. Taxonomic studies in the Disinae. V. A revision of the

genus Monadenia. 339

Linder, H. P. Taxonomic studies in the Disinae. VI. A revision of the

genus Herschelia, 365

Lippia javanica (Burm. f.) Spreng., 509

Lonchocarpus capassa Rolfe, 510

Macrocoma abyssinica (C. Miill.) Vitt, 464

lycopodiodes (Sch-.vaegr.) Vitt, 465

pulchella (Hornsch.) Vitt, 465

tenue (Hook. & Grev.) Vitt subsp. tenue, 464

Maema angohnsis DC., 510

parvifolia Pax, 510

Magill, R. E. Musci austro-africani II. Bryophyte collections in southern

Africa and southern African type specimens in the National Herbarium,

Pretoria, 127

Magill, R. E. & Vitt, D. H. The phytogeography and ecology of Macro-

coma (Orthotricheceae, Musci) in Africa, 463

Manihot utilissima Pohl, 510

Mariscus bullatus (Kiikenth.) Podlech, 444

chersinus N.E. Br., 444

dregeanus Kunth, 444

dubius (Rottb.) Kiikenth. ex G.E.C. Fisher, 443

durus (Knuth) C.B. Cl., 441

macropus C.B. CL, 443

Maytenus polyacantha — Heteropogon contortus Savanna, 205

Maytenus heterophylla (Eckl. & Zeyh.) N. Robson, 510

Melhania Forssk., 263

acuminata Mast., 268

var. acuminata, 268

var. agnosta (K. Schum.) Wild, 269

agnosia K. Schum., 269

albicans Bak. f., 271

bolusii Burtt Davy, 270

burchellii DC., 271

damarana Harv., 272

didyma Eckl. <£ Zeyh., 264

var. linearifolia (Sond.) Szyszyl., 265

dinteri Engl., 271

ferruginea sensu Szyszyl, 269

forbesii Planch, ex Mast., 267

griquensis H. Bol., 270

var. virescens K. Schum., 270

integra Verdoorn, 266

leucar.tha E. Mey., 264

linearifolia Sond., 265

obtusa N.E. Br., 269

ovata var. oblongata K. Schum., 272

polygama Verdoorn, 268

prostrata DC., 265

forma latifolia Bak. f., 265

randii Bak. F., 266

rehmannii Szyszyl.. 270

rupestris Schinz, 270

serrata Schinz, 271

serrulata R.E. Fr., 267

suluensis Gerstner, 272

transvaalensis Szyszyl., 269

veiuiina sensu Exel! & Mendonga, 268

virescens (K. Schum.) K. Schum, 270

Melia azedarach L., 510

Merxmuellera aureocephala ( J . G. Anders.) Conert, 493

davyi (C. E. Hubb.) Conert, 493

disticha (Nees) Conert. 185, 493

drakensbergensis (Schweick.,) Conert, 487

macowanii (Stapf) Conert, 493

stereophylla (J.G. Anders.,) Cories, 487

stricta (J.G. Anders.) Conert, 493

Mestoklema N.E. Br. ex Glen, 454

albanicum N.E. Br. ex Glen, 455

arboriforme (Burch.) N.E. Br. ex Glen, 454

copiosum N.E. Br. ex Glen, 454

■

V

'

datum N.E. Br. ex Glen, 454

iliepidum N.E. Br. ex Glen, 454

tuberosum (L.) N.E. Br. ex Glen, 454

var. macrorhizum (Haw.) N.E. Br. ex Glen, 454

Wesembryanthemum arboriforme Burch., 454

macrorhizum Haw., 454

megarhizum Don, 454

spinosum O. Kuntze, 454

tuberosum L., 454

Mimusops zeyheri Sond., 510

Monadenia Lindl., 341

Sect. Densiflora Under, 342

Sect. Kamiesbergenses Linder, 349

! Sect. Monadenia, 357

Sect. Tenuicomes Linder, 350

atrorubens (Schltr.) Rolfe, 354

auriculata (H. Bob) Rolfe, 348

australiensis Rupp, 346

bolusiana (Schltr.) Rolfe, 353

bracteata (Swartz) Dur. & Schinz, 340

brevicomis Lindl. 360

ceraua (Thunb.) Dur. & Schinz, 359

comosa Reichb. /., 351

conferta (H. Bol.) Kraenzl., 343

densiflora Undl. 348

ecalcarata Lewis, 343

inflata Sond., 359 .

i lancifolia Sond., 356

j leptostachya Sond., 356

macrocera Lindl., 356

tnacrostachya Undl.. 349

micrantha Lindl., 346

i multiflora Sond., 348

ophrydea Undl., 354

ophrydea Lindl. sensu Kraenzl. 354

physodes (Swartz) Reich. /., 358

i prasinata (Ker-Gawl.) Lindl., 359

pygmaea (H. Bol.) Dur. & Schinz, 345

reticulata (H. Bol.) Dur. & Schinz, 351

rufescens (Thunb.) Lindl., 356

rufescens Lindl., non Thunb., 351

! sabulosa (H. Bol.) Kraenzl., 344

Morris, J. W. Encoding the National Herbarium (PRE) for computerized

information retrieval, 149

Morris, J. W. & Manders, R. Information available within the PRECIS

data bank of the National Herbarium, Pretoria, with examples of uses to

which it may be put, 473

Muraltia elsieae J. Paiva, 458

Obermeyer, A. A. A new combination in Commelina, 437

Obermeyer, A. A. A new combination in Eriocaulon, 450

Obermeyer, A. A. A new combination in Gethyllis, 136

Obermeyer, A. A. A new combination in Thuranthos, 139

Obermeyer, A. A. A new species of Crocosmia, 450

Obermeyer, A. A. A new species of Strumaria, 435

Obermeyer, A. A. A new subgenus Rhadamaruhopsis and two new species

of Rhadamanthus, 137

Obermeyer, A. A. A new subspecies of Gladiolus microcarpus, 451

Obermeyer, A. A. A new subtribe in Liliaceae, 137

I Obermeyer, A. A. A reappraisal of Urginea altissima, 452

Obermeyer, A. A. The genus Sypharissa (Liiiaceae), 111

Obermeyer, A. A. The status of Urginea eplgea, 139

Obermeyer, A. A. Two new species of Commelina, 436

Ooermeyer, A. A. Two new species of Cyanotis. 437

Obituaries:

Acocks, J. P. H. by D. J. B. Killick, 239

Bayer, A. W. by R. A. Dyer, 244

Compton, R. H. by L. E. Codd, 244

Ochna pulchra Hook., 179

Olea exasperata Bush, 520

Oliver, E. G. H. Some observations on two early Cape florilegia, 115

Oliver, E. G. H. Studies in the Ericoideae. m. The genus

Griesebachia, 65

Oliver, E. G. H. Two new species of Ericoideae, 446

Oncoba spinosa Forssk., 510

Opuntia ficus-indica (L.) Mill., 510

Orchis barbata L.f., 376

spathulata L.f., 372

Ornithogalum exuviatum (Jacq.) Kunth, 113

filifolium (Jacq.) Kunth, 113

fragrans (Jacq.) Kunth, 113

undulatum (Ait.) Thunb., 140

Ozoroa engleri R. & A. Fernandes, 510

paniculosa (Sond.) R. & A. Fernandes. 179

reticulata (Bak. f.) R. <£ A. Fernandes subsp. reticulata var. reticu-

lata, 510

Paiva, J.A.R. A conspicuous new species of Muraltia, 458

Panicum meyeranum Nees, 457

Pappea capensis Ecld. & Zeyh., 510

Parinari capensis Harv., 180

curatellifoiia Planch, ex Benth., 510

Pelargonium cardiophyllum Harv., 431

dasyphyllum E. Mey. ex Knuth, 432

dolomiticum Knuth, 432

setulosum Turcz.. 431

temifolium Vorster, 431

trifoliatum Harv., 431

Peltophorum africanum Sond., 510

Penmsetum americanum (L.) Leeke subsp. americanum, 510

Phaseolus sp., 510

Phoenix reclinata Jacq., 510

Phragmites mauritianus Kunth, 510

Phyllanthus verrucosus Thunb., 510

Physalis peruviana L., 510 —

Piliostigma thonningii (Schumach.) Milne-Redh., 510

Pioneer vegetation, 520

Pittosporum viridiflorum Sims, 510

Plectranthus esculentus N.E. Br., 510

Pluchea dioscorides (L.) DC., 510

Pseudolachnostylis maprouneaefolia Pax, 511

Psoralea L., 321

Psoralea L., pro parte, 318

acaulis Stev. ap. Hoffm., 318

biflora Harv., 317

bituminosa L., 318

corylifolia L., 317

drupacea Bunge, 317

holubii Burtt Davy, 317

jaubertina Fenzl, 317

obtusifolia, DC., 317

odorata Blatt. 6i Haib., 317

paler so niae Schonl., 317

pedunculata Ker-Gawl., 321

plicata Del., 317

sericea Poir., 321

tomentosa Thunb., 321

Ptaeroxylon obliquum (Thunb.) Radik., 511

Pterocarpus angolensis DC., 511

rotundifolius (Sond.) Druce subsp. rotundifolius, 511

Pterocelastrus tricuspidatus Bushclumps, 520

Pterolobium stellatum (Forssk.) Gmel., 511

Pygmaeothamus zeyheri (Sond.) Robyns, 180

Phalangium exuviatum (Jacq.) Poir., 113

filifolium (Jacq.) Poir., 113

fragrans (Jacq.) Poir., 113

Phragmites australis Community, 225

Phragmites mauritianus Community, 225

Retief, E. A new species of Cyphostemma from the Transvaal, 460

Retief, Elizabeth. A new species of Rhoicissus from Natal, 146

Review of the work of the Botanical Research Institute, 1978/1979, 247,

557

Reyneke, W. F. Three subspecies of Eucomis autumnalis, 140

Rhadamanthus, 137

subgen. Rhadamanthopsis Oberm., 137

karooicus Oberm., 138

namibensis Oberm., 137

Rhoicissus digitata (L.f.) Gilg & Brandt, 511

rhomboidea (E. Mey. ex Harv.) Planch., 146

sessilifolia Retief, 146

Rhus pyroides — Leucosidea sericea Bush and Savanna Communities, 200

Rhus pyroides — Rbamnus prinoides — Acacia karroo Variant, 204

Rhus pvroidec — Rhamnus prinoides — Cassinopsis ilicifolia Variant, 204

Rhus pyroides — Rhamnus prinoides Forest, 200

Rhus pyroioes — Protea caffra — Chrysanthemoiaes monilifera Variant, 204

Rhus pyroides — Protea caffra — Harpochlou falx Variant, 205

Rhynchelytrum setifolium — Monocymbium ceresiiforme Grassland, 211

Rhynchodium cephalotes Steud., 321

sericeum Presl, 321

Rhynchosia Lour., 320

arida C. H. Stirton, 320

cajanoides Guil. & Perr., 293

Ricinus communis L., 511

Rjocreuxia alexandrina (Huber) R. A. Dyer, 436

burchellii K. Schum., 436

flanaganii Schltr., 436

subsp. alexandrina (Huber) R. A. Dyer, 436

subsp. segregata R. A. Dyer, 436

subsp. woodii (N.E Br.) R. A. Dyer, 436

woodi' N.E. Br., 436

Ross, J. H. An analysis of the African Acacia species: their distributions,

possible origins and relationships, 389

Ross, J. H. A survey of some of the pre-Linnean history of the genus

Acacia, 95

Rubus niveus Thunb., 333

phoenicolasius Maxim., 336

Rutherford, M. C. Field identification of roots of woody plants of the

savanna ecosystem study area, Nylsvley, 171

Rutherford, M. C. Survival, regeneration and leaf biomass changes in

woody plants following spring bums in Burkea africana — Ochiui

pulchra Savanna, 531

Saccharum officinale L., 511

.

I

599

Sansevieria hyacinthoides (L.) Druce, 511

Sarcostemma viminale (L.) R. Br., 511

Satyrium barbatum (L.f.) Thunb., 376

cernuum Thunb., 359

excelsum Thunb., 381

hians L.f., 381

rufescens Thunb., 356

spathulata (L.f.) Thunb., 372

Scaevola thunbergii Pioneer Vegetation, 520

Scilla micrantha A. Rich., 139

Scirpus inclinatus (Del.) Aschers. & Schweinf., 511

Sclerocarya caffra Sond., 511

Scyphogyne calcicola E. C. H. Oliver, 448

Securidaca longipedunculata Fresen., 181

Securinega virosa (Roxb. ex Willd.) Pax & K. Hoffm., 511

Setaria rp., 511

Solanum incanum L., 511

panduraeforme E. Mey., 511

Spilanthes mauritiana (Pers.) DC., 511

Spirostachys afficana Sond., 511

Sporobolus africanus (Poir ) Robyns & Tournay, 511

Stirton, C. H. Natural hybridization in the genus Eriosema (Leguminosae)

in South Africa, 307

Stirton, C. H. New records of naturalized Rub us in southern Africa, 333

Stirton, C. H. Notes on the taxonomy of Rubus in southern Africa, 331

Stirton, C. H. Studies in the Leguminosae — Papilionoideae of southern

Africa, 317

Stirton, C. H. The Eriosema cordatum complex. II. The Eriosema corda-

tum and E. nutans groups, 281

Stirton, C. H. The genus Dipogon (Leguminosae — Papilionoideae), 327

Strumaria barbariae Oberm., 435

Strychnos cocculoides Bak., 181

madagascariensis Poir., 512

pungens Soler., 182

spinosa Lam., 512

Sundays River Scrub, 522

Sypharissa Salisb., Ill

exuviata (Jacq.) Salisb. ex. Oberm., 113

filifolia (Jacq.) Salisb. ex Oberm., 113

fragrans (Jacq.) Salisb., ex Oberm., 113

multifolia (Lewis) Oberm., 114

Syzygium cordatum Hochst., 512

Tabemaemontana elegans Stapf, 512

Taylor, H. C. Phytogeography of fynbos, 231

Taylor, H. C. & Morris, J.W. A brief account of coast vegetation near Port

Elizabeth, 519

Taylor, H. C. & Van der Meulen, F. Structural and floristic classifications

of Cape Mountain Fynbos on Rooiberg, southern Cape, 557

Tecomana capensis (Thunb.) Spach., 5T2

Terminalia sericea Burch, ex DC., 512

Themeda triandra — Berkheya setifera — Rhus discolor Grassland Commu-

nities, 212

Themeda triandra Forssk., 512

Thuranthos basuticum (Phill.) Oberm., 139

macranthum (Bak.) C.H. Wr., 139

nocturnale R.A. Dyer, 139

Trachypogon spicatus — Themeda triandra Grassland Communities, 211

Trichilia emetica Vahl, 512

Tryphia parviflora Lindl., 456

secunda Lindl., 456

Turraea obtusifolia Hochst., 512

Tylosema fassogiensis ( Schweinf .1 Torre & Hillc., 512

l Jrginea Steinh., Ill

basutica Phill., 139

duthieae Adamson, 113

ecklonii sensu Duthie, 113

epigea R. A. Dyer, 139

exuviata (Jacq.) Steinh., 113

filifolia (Jacq.) Steinh., 113

flexuosa Adamson, 113

forsteri Bak., 453

fragrans (Jacq.) Steinh., 113

micrantha (A. Rich.) Solms-Laub., 139

multifolia Lewis, 114

unifolia Duthie, 113

tribe Sypharissa, 1 1 1

Valia macrophylla Vis., 264

Van der Walt, J. J. A. & Vorster, P. J. Miscellaneous notes on the genus

Pelargonium, 431

Vangueria infausta Burch., 512

Van Warmelo, K. T. Sexual nuclear division in Neocosmospora, 415

Van Wyk, E. A. The identity of Eugenia woodii, 142

Verdcourtia lignosa (L.) Wilczek, 327

Verdoom, I. C. Revision of Hermannia subgenus Hermannia in southern

Africa, 1

Verdoom, I. C. Revision of Melhania in southern Africa, 263

Verdoom, I. C. The genus Cola in southern Africa, 277

Verdoom, I. C. The genus Waltheria in southern Africa, 275

Viscum anceps E. Mey. ex Sprague, 167

capense L.f. subsp. hoolei Wiens, 167

combreticola Engl., 167

continuum E. Mey. ex Sprague, 167

crassulae Eckl. & Zeyh., 167

menyharthii Engl. & Schinz, 167

minimum Harv., 167

nervosum Hochst. ex A. Rich., 168

obovatum Harv., 168

obscurum Thunb., 168

oreophilum Wiens, 168

pauciflorum, 168

rotundifolium L.f., 168

spragueanum Burn Davy, 168

vermcosum Harv., 168

Vitex rehmannii Guerke, 183

Von Broembsen, H. H. A simple method for determining the density of

plants in a randomly-dispersed population, 574

Von Broembsen, H. H. Investigation into the significance of plant dis-

persion in assessing pasture condition, 569

Vorster, P. On the identity and geographical distribution of Mariscus

angularis, M. chersinus, Cyperus bullatus and C. capensis var.

polyanthemus , 443

Vorster, P. The correct author citation for Cyperus elephantinus, 446

Vorster, P. The correct author citation for Mariscus dubius, 443

Vorster, P. The correct author citation for Mariscus macropus, 443

Vorster, P. The identity and typification of Mariscus dregeanus, 444

Waltheria americana L., 275

var. indica (L.) K. Schum., 275

var. subspicata K. Schum., 275

indica L., 275

Weisser, P. J. & Parsons, R. J. Monitoring Phragmites australis increases

from 1937 to 1976 in the Siyai Lagoon (Natal, South Africa) by means

of air photo interpretation, 553

Wiens, D. & Barlow, B. A. Translocation heterozygosity in southern

African species of Viscum, 161

Xeromp'nis obovata (Hochst.) Keay, 512

Xerophyta retinervis Bak., 512

Ximenia americana L. var. mic/ophylla Welw., 512

caffra Sond., 183

var. caffra, 512

Zanthoxylum capense (Thunb.) Harv., 512

Zea mays L., 512

Ziziphus mucronata Willd. subsp. mucronata, 512

. . ■: V- . ; * '• •

••

-

.

■

*

■

GUIDE FOR AUTHORS

GIDS VIR SKRYWFRS

GENERAL

Bothalia is a medium for the publication of botanical papers

dealing with the flora and vegetation of Southern Africa. Papers

submitted for publication in Bothalia should conform to the

general style and layout of recent issues of the journal (from Vol.

1 1 onwards) and may be written in either English or Afrikaans.

TEXT

Manuscripts should be typed, double-spaced on one side of uni-

formly-sized A4 paper having at least a margin of 3 cm all round.

Latin names of plants should be underlined to indicate italics. All

other marking of the copy should be left to the editor. Metric units

are to be used throughout. Manuscripts should be submitted in

duplicate to the Editor, Bothalia, Private Bag X101, Pretoria.

ABSTRACT

A short abstract of 100-200 words in both English and Afri-

kaans should be provided. In the abstract the names of new species

and new combinations should not be underlined.

FIGURES

Black and white drawings, including graphs, should be in jet-

black Indian ink preferably on bristol board or plastic film. Lines

should be bold enough to stand reduction. Indicate the desired

lettering lightly in pencil: the printer will insert the final lettering.

If authors prefer to do their own lettering, then use some printing

device such as stencilling, letraset, etc. it is recommended that

drawings should be twice the size of the final reduction.

Photographs submitted should be of good quality, glossy, sharp

and of moderate, but not excessive contrast. Photograph mosaics

should be composed by the authors themselves: the component

photographs should be mounted neatly on a white card base

leaving a narrow gap between each print; number the prints using

some printing device.

Figures should be planned to fit, after reduction, into a width of

8 cm, 11 cm or 17 cm with a maximum vertical length of 24 cm.

The number of each figure and the author’s name should be

written on the back of the figure using a soft pencil.

Captions for figures should be collected together and typed on a

separate page headed Captions for Figures. A copy of each caption

should be attached to the base of each figure.

Authors should indicate in pencil in the text where they would

like their illustrations to appear.

TABLES

Tables should be set out on separate sheets and numbered in

Arabic numerals.

CITATION OF SPECIMENS

In citing specimens the grid reference system should be used

(Technical Note: Gen. 4). Provinces/countries should be cited in

the following order: S.W. Africa, Botswana, Transvaal, Orange

Free State, Swaziland, Natal, Lesotho and the Cape. Grid

references should be cited in numerical sequence. Locality records

for specimens should preferably be given to within a quarter-

degree square. Records from the same one-degree square are given

in alphabetical order i.e. (-AC) precedes (-AD), etc. Records

from the same quarter-degree square are arranged alphabetically

according to the collectors’ names; the quarter degree references

must be repeated for each specimen cited. The following example

will explain the procedure:

Natal. — 2731 (Louwsburg): 16 km E. of Nongoma (-DD),

Pelser 354; near Dwarsrand, Van der Merwe 4789. 2829 (Harri-

smith): near Groothoek (-AB), Smith 234; Koffiefontein (-AB),

Taylor 720; Cathedral Peak Forest Station (-CC), Marriott 74;

Wilgerfontein, Roux 426. Grid ref. unknown: Sterkstroom,

Strydom 12.

Records from outside Southern Africa should be cited from

north to south i.e. preceding those from Southern Africa. The

abbreviation ‘distr.’ should be added to all district names e.g.:

Kenya. — Nairobi distr.: Nairobi plains beyond race course,

Napier 485.

REFERENCES

References in the text should be cited as follows: ‘Jones (1955)

stated . . .’ or ‘. . . (Smith, 1956)’ when giving a reference simply

as authority for a statement. The list of references at the end of the

ALGEMEEN

Bothalia is ’n medium vir die publikasie vir plantkundige arti-

kels wat handel oor die flora van Suidelike Afrika. Artikels wat

voorgele word vir publikasie in Bothalia behoort ooreen te stem

met die algemene style en rangskikking van onlangse uitgawes van

die tydskrif (vanaf vol. 11). Dit mag in Engels of in Afrikaans

geskryf word.

TEKS

Manuskripte moet getik wees in dubbelspasiering slegs op een

kant van ewegroot A4-papier, met reg rondom ’n rand van

minstens 3 cm breed. Latynse name van plante moet onderstreep

word om aan te dui dat dit kursief gedruk moet word. Alle ander

merke moet aan die redakteur oorgelaat word. Metrieke eenhede

moet deurgaans gebruik word. Manuskripte moet in tweevoud in-

gedien word by die Redakteur, Bothalia, Privaatsak XI 01,

Pretoria.

UITTREKSEL

’n Kort uittreksel van 100 - 200 woorde moet voorsien word,

beide in Engels en Afrikaans. In die uittreksel moet die name van

nuwe soorte en nuwe kombinasies nie onderstreep word nie.

AFBEELDINGS

Wit en swart tekeninge, insluitende grafieke, moet met pikswart

Indiese ink geteken word, verkieslik op “bristol board” of plas-

tiekfilm. Lyne moet dik genoeg wees om verklein te kan word. Dui

die verlangde byskrifte liggies in potlood aan: die drukker sal die

uiteindelike byskrifte invoeg. Indien skrywers verkies om hulle eie

byskrifte te maak, gebruik dan een of ander hulpmiddel soos letra-

set of ’n sjabloon. Dit is wenslik dat tekeninge tweemaal so groot

as die uiteindelike verkleining sal wees.

Foto’s wat ingedien word, moet van hoe kwaliteit wees — glan-

send, skerp en van matige maar nie oordrewe kontras. Foto-

mosaieke moet deur die skrywer self saamgestel word: die af-

sonderlike foto’s moet netjies monteer word op ’n stuk wit karton

met ’n smal strokie tussen die foto’s; nommer die foto’s met

behulp van een of ander druk-hulpmiddel.

Afbeeldings moet so beplan word dat hulle na verkleining sal

pas in ’n breedte van 8 cm, 1 1 cm of 17 cm met ’n maksimum ver-

tikale lengte van 24 cm.

Die nommer van elke afbeelding sowel as die skrywer se naam

moet op die rugkant van die afbeelding geskryf word met ’n sagte

potlood.

Onderskrifte vir afbeeldings moet bymekaar getik word op ’n

afsonderlike bladsy met die opskrif Onderskrifte vir Afbeeldings.

’n Afskrif van elke onderskrif moet aan die onderkant van elke af-

beelding vasgeheg word.

Skrywers moet met potlood in die teks aandui waar hulle graag

hulle afbeeldings wil he.

TABELLE

Tabelle moet op afsonderlike velle papier kom en genommer

word met Arabiese nommers.

SITERING VAN EKSEMPLARE

Wanneer eksemplare siteer word, moet die ruitverwysing stelsel

gebruik word (Tegniese Nota: Gen. 4). Provinsies/lande moet in

die volgende volgorde siteer word: Suidwes-Afrika, Botswana,

Transvaal, Oranje-Vrystaat, Swaziland, Natal, Lesotho en die

Kaapprovinsie. Ruitverwysings moet in numeriese volgorde siteer

word. Lokaliteitsrekords vir eksemplare moet verkieslik tot binne

kwartgraadvierkante gegee word. Rekords uit dieselfde eengraad-

vierkant word in alfabetiese volgorde aangebied, nl. (-AC) kom

voor (-AD) ens. Rekords uit dieselfde kwartgraadvierkant word

alfabeties gerangskik volgens die versamelaars se name, en die

kwartgraadverwysings moet herhaal word vir elke eksemplaar wat

siteer word. Die volgende voorbeeld sal die metode verduidelik:

Natal. — 2731 (Louwsburg): 16 km O. van Nongoma (-DD),

Pelser 354; naby Dwarsrand, Van der Merwe 4789. 2829 (Harri-

smith): naby Groothoek (-AB), Smith 234, Koffiefontein (-AB),

Taylor 720; Cathedral Peak Bosboustasie (-CC), Marriott 74;

Wilgerfontein, Roux 426. Ruitverwysing onbekend: Sterkstroom,

Strydom 12.

Rekords van buite Suidelike Afrika moet siteer word van noord

na suid, d.w.s. dit gaan die van Suidelike Afrika vooraf. Die af-

korting ‘distr.’ behoort by alle distriksname gevoeg te word, bv:

Kenya. — Nairobi-distr.: Nairobivlakte anderkant die renbaan,

Napier 485.

VERWYSINGS

Verwysings in die teks moet as volg siteer word: ‘Jones (1955)

beweer . . .’ of ‘. . . (Smith, 1956)’ wanneer ’n verwysing slegs as

article should be arranged alphabetically and the literature abbre-

viations used should conform to the World List of Scientific

Periodicals (1965) or the list of Literature Abbreviations (Tech-

nical Note: Tax. 6 AN 1) issued by the Botanical Research In-

stitute, thus:

Brown, N. E., 1909. Asclepiadaceae. In W. T. Thiselton-Dyer,

FI. Cap. 6,2: 518-1036. London: Lovell Reeve.

Hutchinson. J., 1946. A botanist in Southern Africa. London:

Gawthorn.

Kruger, F. J., 1974. The physiography and plant communities of

the Jakkalsrivier Catchment. M.Sc. (Forestry) thesis, Univer-

sity of Stellenbosch (unpublished).

Morris, J. W., 1969. An ordination of the vegetation of

Ntshongweni, Natal. Bothalia 10: 89-120.

If, as in many taxonomic papers, periodicals or books are

mentioned in the text, usually in the species synopsis, they should

be cited as in the following examples: Gilg & Ben. in Bot. Jb. 53:

240 (1915) and Burtt Davy, FI. Transv. 1: 122 (1926).

REPRINTS

Authors receive 75 reprints gratis. If there is more than one

author, this number will have to be shared between them.

outoriteit vir ’n stelling gegee word. Die verwysingslys aan die

einde van die artikel moet alfabeties gerangskik wees en die litera-

tuurafkortings wat gebruik word, moet in ooreenstemming wees

met die World List of Scientific Periodicals (1965) of die lys van

Literatuurafkortings (Tegniese Nota: Tax. 6 AN 1) wat uilgegee is

deur die Navorsingsinstituut vir Plantkunde, as volg:

Brown. N. E., 1909. Asclepiadaceae. In W. T. Thiselton-Dyer, FI.

Cap. 6,2: 518-1036. London: Lovell Reeve.

Hutchinson, J. 1946. A botanist in Southern Africa. London:

Gawthorn.

Kruger, F. J., 1974. The physiography and plant communities of

the Jakkalsrivier Catchment. M.Sc. (Bosbou) tesis, Univer-

siteit van Stellenbosch (ongepubliseerd).

Morris, J. W., 1969 An ordination of the vegetation of Notsho-

ngweni. Natal. Bothalia 10: 89-120.

Wanneer, soos in baie taksonomiese artikels die geval is tyd-

skrifte of boeke in die teks genoem word, gewoonlik in die soort-

sinopsis, behoort hulle siteer te word soos in die volgende

voorbeelde: Gilg & Ben. in Bot. Jb. 53: 240 (1915) en Burtt Davy,

FI. Transv. 1 : 122 (1926)

HERDRUKKE

Skrywers ontvang 75 herdrukke gratis. Wanneer daar meer as

een skrywer is, sal hierdie aantal tussen hulle verdeel moet word.