MARY GUNN LIBRARY

NATIONAL BOTANICAL INSTITUTE

PRIVATE BAG X 101

PRETORIA 0001

REPUBLIC OF SOUTH AFRICA

w\af*Y,

GUNN,.4ffY

e.eesoiiaaji i

Soulh ^-riosiiiuU

BiodiversiW

Digitized by the Internet Archive

in 2016

https://archive.org/details/bothaliavolume606unse

JSotbalia

A RECORD OF

CONTRIBUTIONS

FROM THE

NATIONAL HERBARIUM

|V

UNION OF SOUTH AFRIC^..^™-—

library

PRETORIA ,, • • *, if IOTMTAN*

( vr hol««y

* cats?*?**

( 2 2*10'

1 B- ‘ JTHEK

- . » * ;<T» Y L?** ^

EDITED BY

R. A. DYER, M.Sc., D.Sc., F.R.S.S.Af.,

CHIEF, DIVISION OF BOTANY, DEPARTMENT OF AGRICULTURE, PRETORIA AND

DIRECTOR OF THE BOTANICAL SURVEY OF THE UNION OF SOUTH AFRICA

THE GOVERNMENT PRINTER, PRETORIA

G.P.-S 1339563-1957-8—1,700.

)

CONTENTS.

Parti. Published 1951. Page

Studies of Some South African Resupinate

Hymenom.ycetes, by P. H. B. Talbot. ... 1

A Morphological, Anatomical and Cyto-

logical Study of Potamophila prehensilis

(Nees) Benth., by B. de Winter 117

The Taxonomy, Anatomy and Cytology of

Pseudobromus africanus (Hack.) Stapf, by

B. de Winter 139

South African Species of Acacia with

Glandular Glutinous Pods, by I. C.

Verdoorn 153

Revision of the Genus Cyclopia and Notes

on Some other Sources of Bush Tea.

by P. Kies 161

Notes cn Asparagus africanus and Related

Species, by P Kies 177

New and Interesting Records of South

African Fungi, by P. H. B. Talbot 183

Encephalartos natalensis , by R. A. Dyer and

I. C. Verdoorn 205

Newly Described Species, by E. A. Bruce,

A. Breuckner, R. A. Dyer, P. Kies and

I. C. Verdoorn 213

Part II. Published 1954.

Micromorphology of the Lower Hymeno-

mycetes, by P. H. B. Talbot 249

A Note on Sterigmata in General, by M. A.

Donk 301

The Genus Stereum in South Africa, by

P. H. B. Talbot 303

On the Genus Lopharia Kalchbrenner &

MacOwan, by P. H. B. Talbot 339

Some Graminicolous Species of Helmin-

thosporium and Curvularia Occurring in

South Africa, by K. M. Putterill 347

Page

Concerning Some South African Pestalo-

tiopsis , by R. L. Steyaert 379

The Morphology and Anatomy of Utricu-

laria transrugosa Stapf, by J. Slinger. . 385

The Nomenclature of Species of Allophylus

in South Africa, by B. de Winter 407

The Nomenclature of the Cape Acacia, by

I. C. Verdoorn 409

Newly Described Species, by L. Britten,

A. Breuckner, H. B. Christian, R. A.

Dyer, M. D. Henderson, D. J. B. Killick,

R. Story, I. C. Verdoorn and H. Wild.. . 415

Part III. Published 1956.

Entorrhiza calospora sp. nov., and Some

Other Parasitic Fungi in Limeum Roots,

by P. H. B. Talbot 453

The Cyphelloid Fungi of South Africa, by

P. H. B. Talbot 465

New and Interesting Records of South

African Fungi. Part ii, by P. H. B. Talbot 489

Three Species of Curvularia from Pinas, by

G. C. A. van der Westhuizen 501

Notes on Certain South African Erythrina

Species, by L. E. Codd 507

The Schotia Species of Southern Africa, by

L. E. Codd 515

New and Interesting Records of South

African Flowering Plants, by E. E. A.

Archibald, L. E. Codd, R. A. Dyer,

A. D. J. Meeuse and D. van Druten. . . . 535

The Oleaceae of Southern Africa, by I. C.

Verdoorn 549

Part IV. Published 1957.

The South African Convolvulaceae, by

A. D. J. Meeuse 641

3

INDEX TO VOLUME VI.

Page

Acacia 153, 184, 409

Acacia (nomenclature of Cape) 409

Acacia (spp. with glandular glutinous pods) 153

Acacia barbertoniensis Schweickerdt 154

borleae Burtt Davy. .. 154, 155, 157 , 160

capensis (Burm.) Burch ... .409, 412, 413

capensis Colla 413

exuvialis Verdoorn 153, 154-5,

156, 157, 160

glandulifera Schinz non Wats. ... 153, 156

glandulifera Wats 153

horrida Willd 409, 411-2

karroo Hayne 183, 409, 413

mollissima 185-6, 471, 492

nebrownii Burtt Davy 153-5, 156,

158, 160

rnibica Bth 412

orfota (Forsk.) Schweinf. 412

permixta Burtt Davy 153-4, 155,

156, 158, 160

permixta var glabra Burtt Davy. . 153, 156

rogersii Burtt Davy 153, 156

swazica Burtt Davy 154, 155,

156, 159, 160

tenuispina Verdoorn 153-5, 156,

159, 160

ACANTHACEAE 535, 602

Acanthopsis 538

Acanthus 538

Acia Karst 5, 59, 264

Acia conferta Talbot 59, 64, 111

fusco-atra (Fr.) Pat 59

subceracea Wakef 59, 111, 285

Acremoniella atra (Corda) Sacc 191, 202

Acrothecium 369

AGROSTIDEAE 139, 142, 143

Agrostis 368-9

AIZOACEAE 454

Akrophyton tuberculatum Lebert 495

Albuca 544

Albuca crudenii Archibald 542-4

glandulosa Bak 542-544

Aleurocystus Lloyd 465-6

Aleurocystus capensis Lloyd 465-6, 477, 479

Aleurodiscus Rabenh. ex J. Schroet..5, 26, 262,

269, 276, 282, 283,

284, 290, 465, 466

Aleurodiscus acerinus (Pers. ex Fr.) H. & L. 19,

26-7, 62, 266,

466, 467, 482

acerinus var alliaceus (Quel.)

B. &G 26-7, 466-7, 482-3

acerinus var longisporus H. &

L 19, 26, 27-8, 62,

86, 466-7, 483

amorphus 268, 271, 282-3, 288

Page

Aleurodiscus

apiculatus Burt 469

aurantius 271, 283

botryosus Burt 467-8, 482-3

capensis Lloyd. 468, 477, 479, 486

cerussatus (Bres.) H. & L.. 28, 468

corneas (Lloyd) Lloyd 468, All

disciformis (DC ex Fr.) Pat. . 28,

268, 468-9, 482, 483

farlowii : . . 283

ionides 470

mirabilis (B. & C.) Hohn. . . . 271,

282-3, 469, 482-3

oakesii 282-3

polygonioides (Karst.) Pilat. . 470-1,

482-3

roseus (Pers. ex Fr.) H. & L. 468,

470-1, 482, 483

weirii 283

Allophylus 407

Allophylus decipiens (Sond.) Radik 407

dregeanus (Sond.) de Winter. . . 407

erosus Radik 408

gazensis Bak 408

melanocarpus (Sond.) Radik. 407, 408

monophyllus Radik 407

natalensis (Sond.) de Winter.. . . 407

spicatus (Poir.) Radik 407

spicatus (Thunb.) Fourcade 407

transvaalensis Burtt Davy 408

Aloe 545

Aloe (series Latebracteatae) 442

Aloe (series Procrassa) 440, 442

Aloe alooides (Bolus) van Druten 544-5

babatiensis Christian & Verdoorn.. . . 440-2,

451

cryptopoda 442

percrassa 440, 442

pretoriensis Pole Evans 222

recurvifolia Groenewald 544-5

rubrolutea 440

Amauroderma 279, 280, 288

Ampelocissus 545

ANACARDIACEAE 539

Aniseia Choisy 729, 730

Aniseia biflora (L.) Choisy 723

calycina (Roxb.) Choisy 728

calystegioides Choisy 730

hacke liana Schinz 726

Anisomyxa plantaginis Nemec 459

Anthoxanthum 369

Anthoxanthum odoratum 368

Archibald et alia: New and Interesting

Records of S.A. Flowering Plants 535

Argyreia Choisy 772

Aristida 190

4

Page

Armillaria mellea 266, 278-9

Arthrobotrys 489

Arthrobotrys oligospora Fres 489, 496

Arundinaria tesselata 125

ASCLEPIADACEAE 213, 415, 540

Aspalathus 161

Aspalathus contaminatus (Thb.) Druce . .170, 1 72,

176

corymbosus E. Mey 170

lebeckia contaminata Thb 170

tenuifolia DC 170, 176

Asparagopsis juniperina Kunth 177

lamarkii Kunth 177

niveniana Kunth 177

Asparagus africanus Lam 177, 178, 179

africanus var B. Bak 177

africanus var brevifolia Kies. . . 177

africanus var concinnus Bak. ... 178

africanus var dependens Bak. ... 177

africanus var pubescens Bak. ... 177

africanus var Y. Bak 178

africanus var wrightii Bak 179

asiaticus Bak 179

asparagoides (L.) Wight 442-3

compactus Salter 179

concinnus (Bak.) Kies 178, 181

dependens Bak 177

dependens Thb 177

glaucus Kies 229, 230

laricinus 178

medeoloides Thunb 443

microphyllus Burch, ex Bak.. . . 178

multiflorus Bak 177, 179

multituberosus R. A. Dyer 442-3

ovatus Salter 443

pearsonii Kies 179

retrofractus 179

rivalis Burch 179

rivalis (Burch.) Kies 179

suaveolens Burch 229

Aspergillus 39

Asterocystis radicis De Willd 459

Asterodon 295

Asterodon ferruginosus 295, 297

Asterostroma Massee 5, 54, 295

Asterostroma albido-carneum Massee 54

cervicolor (B. & C.) Massee. 54,

55, 107, 289, 297

corticola Massee 54

pallidum Morgan 54

rhizomorpha 266

Asterostromella H. & L 5, 51, 54, 283,

284, 290

Asterostromella dura B. & G 51, 53, 105

duriuscula (B. & Br.)

Talbot 19, 45, 51, 54.

104 , 313

investiens 283

rumpiana Talbot 53, 54, 106

Asthenochloa 122

Astripomoea A. Meeuse 644, 109

Astripomoea lachnosperma (Choisy) A.

Meeuse 709, 710

malvacea (Klotzsch) A.

Meeuse 709, 710

rotundata(Pilger) A. Meeuse 709, 77/

Asmbchlaena Hall. f. 709

Page

Astrochlaena lachnosperma (Choisy) Hall. f. 709, 710

malvacea (Klotzsch) Hall. f. . . 710

rot undata Pilger 711

ATICH1ACEAE 185

Atichia glomerulosa (Ach.) Stein 184-5, 195

Atractobasidium 260

Atroctobasidium corticioides 262

Atriplex 418, 420

Atriplex erosa Brueckner & Verdoorn .... 416-7,

420, 421

muelleri Bth 418, 420-1

rosea L 420

semibaccata R. Br 416, 420-1

suberecta Verdoorn 418-9, 420-1

tartarica L 418

Auricularia Bull, ex Fr 5, 6, 29, 251, 253-4,

259, 262, 264, 302

Auricularia auricula-judae 271

butleri Massee 29

corrugata Sow 6

mesenterica (Dicks, ex Fr.) Fr. 6, 69

mesenterica var lobata

(Sommerf.) Quel 6

papyrina Bull 30

phosphorea (L.) Sow 15

reflexa (Berk.) Bres 28

reflexa Bull 28

rugosissima (Lev.) Bres 28

sordescens Ces 29

strigoso-zonata { Schw.) Lloyd.. 28

tremeUoides 6

BAMBUSEAE 117, 120, 124, 125, 127

Basisporium Molliard 189

Basisporium gallarum sensu Dale 189

gallarum Molliard 189

Batatas (Choisy) Griseb. (sect, of Ipomoea) 713

Batatas edulis (Thunb.) Choisy 746

Biovularia 385

Blackiella 418, 420

Blaeria L 431

Blepharis 538

Boletus pttrpurascens DC 30

BOLIVARIACEAE 602

Bolivaria Cham. & Schltr 601

Bonamia Thouars 643, 662

Bonamia capensis (E. Mey. ex Choisy) Burtt

Davy 663

madagascariensis Poir 664

poranoides Hall, f 665

schizantha (Hall, f.) A. Meeuse.. . 665

suffruticosa (Schinz) Burtt Davy. . 663

Bonamya Neck 664

Bonordeniella aspera Linder 187

Borbonia pinifolia Marl 170, 172

Botryobasidium Donk 257, 261, 302

Botryobasidium coronation . . 262

Botryosporium longibrachiatum (Oud.)

Maire 490, 497

Botrytis longibrachiata Oud 490

Bourdotia 291

Bouruotia cinerella var trachyspora 289

eyrei 289

Brachybasidium 251

Brachycladium 369

Brachycladium ramosum Bainier 370

spiciferum Bainier 369

5

Page

Brachvelytrum africanmn Hackel 140, 142

Brachysporium 369

Brachysporium gracile (Wallr.) Sacc. var

graminewn Rabenh 353

Brachystelma 541, 542

Brachystelma crispum Grah 540

galpinii N. E. Br 541

montanum R. A. Dyer 540

pilosum R. A. Dyer 541-2

Breweria R. Br 662

Breweria bacehariodes Baker 663

capensis (E. Mey. ex Choisy)

Baker 663

malvacea Klotzsch 710

sessili flora Baker 663

suffruticosa Schinz 663

Britten, L. et alia: Newly described species 415

Briza maxima 363

minor 363

Bromus L 139, 142, 143

Bromus asper Murray 142

Bruce, E. A. et alia: Newly described species 213

Brueckner, A. et alia: Newly described

species 213, 415

Buddelja salvifolia 470

BURSERACEAE 214

Byssus phosphorea L 15

Calonyction (Choisy) Griseb. (sect, of

Ipomoea) 711, 712, 713

Calonyction aculeatum (L.) House 765

bona-nox Boj 765

muricatum 765

speciosum Choisy 765

Calycanthemum Klotzsch (sect, of Ipomoea) 713

Calycanthemum leucanthemum Klotzsch 722

Calyptrospermum A. Diet 601

Calystegia R. Br 643, 696

Calystegia sepium (L.) R. Br 697

soldanella) L.) R. Br. ex R. et S. 697, 698

Campanolea Gilg. & Schellenb 592, 594

Campanolea mildbraedii Gilg. & Schellenb. 594

Cannabis sativa 227

Cantharellus 251

Capnodium citricolum 185

Caralluma lutea N. E. Br 222

rogersii Bruce & Dyer 416

Cephalanthae (Choisy) Hall. f. (subsect, of

Ipomoea) 713

Cephalosporiwn dendroides Ell. & Kellerm. 490

Ceranthus (section Linociera ) 592, 593

Ceiatobasidium 257-261, 263, 302

Ceratobasidium anceps 261, 268

fibrillosum 262

sterigmaticum 258

Ceratocystis adiposa (Butler) Moreau. . .489, 496

Ceratostomella adiposa (Butler) Sartoris 490

Ceratotheca 233

Ceratotheca saxicola E. A. Bruce 233

triloba E. Mey. ex Bernh 233

Cercospora 347, 367

Cercospora phaeocarpa Mitter 347

sacchari Breda de Haan 354

Ceriomyces venulosus (B. & C.) Torrend. . . 25

Ceropegia 213

Ceropegia caffrorum Schltr 213-4

decidua E. A. Bruce 213-4, 239

Chaetomium 457

Chaetoplea 349

Chaetopsis graminicola B. & Br 190

CHENOPODIACEAE 215, 416

Chorisanthae Hall f. (subsect, of Ipomoea) 713

Christian, H. B. et alia: Newly described

species 415

Chytridium brassicae (Wor.) Dang 458

Cintractia leucoderma (Berk.) P. Henn 188

Cissus connivens Lam 545

connivens Lam. var meyeriana Planch 545

glabra E. Mey 545

natalitia (Szysz.) L. E. Codd 545-6

oriental is Harv 545

orientalis Lam 545

sulcatus C. A. Smith 545-6

Citrus 185

Cladoderris 285, 325, 340

Cladoderris australica 287

infundibuliformis 287

pritzelii P. Henn 57

Clavaria 251, 261, 269, 285, 290

Clavulina 261

Clerodendron glabrum 185

Cochliobolus 347, 351, 368

Cochliobolus heterostrophus Drechsl. ... 351 , 361

kusanoi (Nisik.) Drechsl 351

miyabeanus (Ito & Kurib.)

Drechsl 351

sativus (Ito & Kurib.)

Drechsl 351, 355

setariae (Ito) Drechsl 351

stenospilus (Carpenter)

Drechsl 351

Codd, L. E. : Notes on Certain S. African

Erythrina species 507

Codd, L. E. : The Schotia species of

Southern Africa 515

Codd, L. E. et alia : New and Interesting

Records of S.A. Flowering plants 535

Coleus pentheri Gtirke 227, 229

vagatus E. A. Bruce 227-9

Collybia 278

Colophospermum mopane (Kirk ex Benth.)

Leonard 416

Combretum 470

Commiphora betschuanica 214-5

calciicola Engl 214-5

neglecta Verdoorn 214-5, 240

rehmanni Engl 215

COMPOS1TAE 220, 421

Coniophora DC. ex Pers 5, 33, 36, 285

Coniophora arida 36, 266-7, 27!

atrocinerea Karst 35-6

betulae 26-7, 297

brunneola (B. & C.) Cooke 35

cerebella (Pers.) Duby 36

ellisii (B. & C.) Cooke 35

fodinarum Talbot. . .34, 36, 64, 91

fumosa Karst 34-6

leucothrix B. & C 35

olivacea (Fr.) Karst.. 34, 35, 36, 92

papillosa Talbot 33. 91

puteana (Schum. ex Fr.) Karst. 34,

36, 267

sistotremoides (Schw.) Massee 34

sordulenta (Cooke & Massee)

Sacc 53, 105

6

Page Page

7

Page

Corticium fumosum Fr 35

galzini 261, 267, 270, 271

giganteum (Fr.) Fr 24

gloeosporum Talbot. .. 12, 14, 15, 74,

275, 284, 285, 287

grammicum P. Henn 52

habgallae B. & Br 477

hinnuleum Bres 54

javanicum Zimm 17

karstenii 271

lactescens 279, 288

lacteum Fr 19, 270

laetum (Karst.) Bres 18, 271, 272

laevePers 14, 15, 267, 271, 285

latitans Karst 21

leucothrix B. & C 35

luteocystidiatum Talbot. . . 12, 13, 289

mauritianum Berk 16

microsporum 270

mo lie B. & C 16, 58

molle Fr 16

mollerianum Bres 22

myxosporum Karst 21

niveocremeum 270

nudum Fr 19, 25

ochraceofulvum B. & G 59, 285

ochrofarctum Burt 13

olivaceum Fr 35

pallidum Bres.. 12, 13, 14, 73, 288-9

pelliculare Karst 19

petersii B. & C 23

polygonioides Karst 470

polygonium 267, 271, 279, 288-9

polyporoideum 271

portentosum 19, 51-3

praetennissum 289, 290

radicatum P. Henn 23

roseocremeum Bres 21, 289

roseolum Massee 18

roseum (Pers. ex Fr).. . . 18, 283 — 4, 470

roumeguerii Bres 22

salmonicolor B. & BR 13, 17, 18,

78, 271

sambuci 285

scutellare B. & C 12, 16, 77

seriale Fr 13

trigonospermum 271

tristiculum B. & Br 46

tuberculosum Pat 25

tulasnelloideum 270

tumulosum Talbot 12, 17

vagum B. & C 19

vellereum 271

zimmermanni Sacc. & Syd 17

Craterellus sinuosus 262

Ceratostigma 235

Crossandra 538-9

v-ui vuiaiia li jjj, JVJV7, JVJJ,

368-70, 378, 501-3

Curvularia (Some graminicolous species) 347, 369

Curvularia (Three species from Pinus) 501

Curvularia brachyspora Boedijn. . . .501-2, 504-5

lunata 370

maculans (Bancr.) Boedijn 501-2,

504-5

pallescens Boedijn 502-5

ramosa (Bain) Boedijn 357, 370

Page

Curvularia spicifera (Bain.) Boedijn 357, 360,

369-70, 378

Cuscuta L 642, 643, 644

Cuscuta abyssinica 655

africana Willd 646, 652

alpestris Fourc 652

angulata Engelm 646, 655

appendiculata Engelm 645, 649

australis R. Br 645, 641

balansae Boiss. et Reutt. var

mossamedensis (Welw.) Yunck. . 655

bifurcata Yunck 645, 648

campestris L 645, 648

capensis Choisy 652

cassytoides Nees ex Engelm. . . .645, 651

cordofana (Engelm.) Yunck 647

cucullata Yunck 652

epithymum Murr. ex L 647, 656

epitribulum Schinz 650

europaea L 644

gerrardii Baker 646, 652

hyalina Roth 646, 650

kilimanjari Oliv 645, 650

madagascariensis Y unck 656

madagascariensis var schlechteri

Yunck 656

medicaginis C. H. Wright 648

natalensis Baker 646, 653

nitida E. Mey 646, 654

obtusiflora H. B. et K 647

obtusiflora var cordofana Engelm. . 647

planiflora Ten 655

planiflora var madagascariensis

(Yunck.) A. Meeuse 646, 656

planiflora var mossamedensis

Welw. et Hiern 647, 65 5

suaveolens Ser 646, 648

timorensis Decne 651

trifolii Babingt 656

CUSCUTACEAE 645

CYCADACEAE 220, 426

Cyclopia Vent 162

Cyclopia (Revision of the genus) 161

Cyclopia asbtonii Hofm. & Phill 171

aurea Fourc 169, 172

aurescens Kies 162, 164 , 114, 176

aurescens var glauca Kies 162, 163,

164, 174, 176

bolusii Hofm. & Phill. . .162, 163, 164,

174, 176

brachypoda Benth 169

brachypoda Hofm. & Phill. . . 168-9, 171

brachypoda var intermedia Hofm.

& Phill 169

buitonii Hofm. & Phill. . 161, 163, 167,

buxifolia (Burm. f.) Kies. 163, 168, 169,

175, 176

capensis Salter 162, 165, 174, 176

cordifolia Benth 168

dregeana Kies 163, 169, 175, 176

falcata (Harv.) Kies 163, 168, 169,

170, 175, 176

falcata var ovata Kies 163, 168,

175, 176

Page

Cyclopia filiformis Kies 163, 167, 174 , 176

galioides (Berg.) DC 161, 162, 164,

174, 176

galioides E. Mey 166

galioides E. & Z 164

genistoides E. Mey 166

genistoides R. Br... 161, 162, 166, 171,

114, 176

genistoides Walp 164

genistoides var heterophylla Harv. 163,

166, 174, 176

genistoides var ovalifolia Kies.. 162, 163,

166, 174, 176

Cyclopia genistoides var teretifolia (E. & Z.)

Kies 162, 166, 176

grandi flora A. DC 170

grandifolia Benth 170

heterophylla E. & Z 166

heterophylla var linear ifolia E. & Z. 166

intermedia E. Mey 163, 169, 172,

175, 176

laricina E. Mey 166

latifolia Benth 168

latifolia DC 168

latifolia E. Mey 169

latifolia E. & Z 170

latifolia Walp 170

laxiflora Benth 170

longiflora 168

longifolia Vog 163, 167, 175, 176

maculata (Andr.) Kies 162-3, 166,

171, 174, 176

meyeriana Walp 161, 162, 165,

174, 176

montana Hofm. & Phill 162, 165,

174, 176

montana var glabra Hofm. &

Phill 162, 165, 174, 176

plicata Kies. . 161, 162, 167, 175, 176

pubescens E. & Z. . 162, 165, 174, 176

sessiliflora E. & Z 163, 169, 170,

171, 175, 176

sessiliflora Mey 165

snbteretifolia 171

subternata Hofm. & Phill 169, 171

subternata Vogel. . 163, 110, 171, 175,

176

subternata var laxiflora (Bth.)

Kies 163, 170, 175, 176

tenuifolia Lehm 161, 166, 171

teretifolia E. & Z 166

vogelii var brachypoda Harv 169

vogelii var falcata Harv 168

vogelii var intermedia Harv 169

vogelii var laxiflora Harv 170

vogelii var subternata Harv 169, 170

Cylindro-helminthosporium 347, 349-50, 367

Cynodon 369, 370

Cynodon bradleyi 362, 368

dactylon 362, 368

transvaalensis 368, 370

CYPERACEAE 457

Cyperus flavescens 456

Cyphella Fr 465, 471, 486

Cyphella alboviolascens (A. & S. ex FR.)

Karst 271, 471-2, 477, 482, 484

ampla Lev 476

Cyphella applanata Talbot 472, 482, 485

bloxami 472

cheesmani Massee 472, 482

citrispora 271

curreyi B. & Br 471-2

disciformis 472

friesii Crouan 472

friesii Quel 472, 476, 486

friesii Weinmann 472

fidvodisca Cooke & Massee. ...473, 475,

484

pelargonii Kalchbr 271, 473, 476,

482, 485

punctiformis (Fr.) Karst 271, 474

punctiformis var strigosa Kalchbr.

& Cooke 474

t abac ina Cooke & Phil 474, 484

variolosa Kalchbr 473, 474, 476,

482, 484

CYPHELLACEAE 465

Cystobasidium sebaceum 262

Cytidia Quel 465, 476

Cytidia cornea Lloyd 468, 476, 477, 486

flocculenta (Fr.) H. & L... . 15, 270, 472,

476, 478, 482, 486

habgallae (B. & Br.) Martin. . . .466, 468,

476, 477, 482, 486

simulans Lloyd 478, 482, 486

Dacryomyces 263, 272

Dacryomyces deliquescens 262, 271

Dacryopinax spathularia 251, 271

Dactylis glomerata 193

Daedalea 275

Daedalea biennis 280

quercina 287

Danthonia 426

Dekindtia Gilg 592, 601

Dekindtia africana Gilg 594, 600, 601, 636

Dendrothele H. & L 276, 465, 478

Dendrothele duthieae Talbot 478-9, 487

Dendryphion atrum Corda 489

Dendryphiopsis atra (Corda) Hughes. . .489, 496

de Winter, B., Morphology, Anatomy and

Cytology of Potamophila prehensilis

(Nees) Benth 117

de Winter, B., Nomenclature of species of

Allophylus in S.A 407

de Winter, B., Taxonomy, Anatomy and

Cytology of Pseudobromus africanus

(Hack.) Stapf 139

Deyeuxia Beauv 142

Dichaelia Harv 542

Dichondra J. G. et R. Forst 657

Dichondra repens J. G. et R. Forst 657

Dichostereum Pilat 53, 284, 313

Digitaria sanguinalis 494—5

Dinochloa 124

Diplodia pinea 501-2

Distichlis spicata (L.) Greene 348, 359

Donk, M. A., A Note on Sterigmata in

General 301

Drechsleia 349, 350

Drechslera arundinis 350

catenaria 350

gigantea 350

tritici-vulgaris 350

Duportella Pat 5. 43, 46, 50, 285

9

Page

Duportella tristicula (B. & Br.) Reinking.45, 46-7 >

58, 99-100

264, 287-9, 293, 295

tristiuscula (Berk.) Pat 46

velutina Pat 46

Dyer, R. A. et alia : New and interesting

records of S.A. Flowering Plants 535

Dyer, R. A. et alia: Newly described species 213,

415

Dyer, R. A. & Verdoorn, I. C.: Encepha-

larlos natalensis 205

Ecchyna faginea 271

Eichleriella 262

Eichleriella incarnata 262

spinulosa 271

Eleusine aegyptica 362

coracana 362

indica 362

Encephalartos altensteinii Lehm 205-6

cycadifolius Lehm 427-8

eximius Verdoorn 426-8, 448-9

friderici-guiliemi 428

ghellinckii Lem 426-8

humilis Verdoorn 220-1 , 241

laevifolius Stapf & Burtt

Davy 220-1

lanatus Stapf & Burtt Davy 220-1

lehmannii 428

natalensis Dver & Verdoorn 205-

211

transvenosus 206

woodii 205-6

Endoconidiophora 490

Endoconidiophora ad’posa ( Butler) Davids. . 490

Entorrhiza C. Weber 456-7

Entorrhiza aschersoniana (Magn.) Lagerh.. 456-7

calospora Talbot 453-62

caricicola Ferd. & Winge 457

casparyana (Magn.) Lagerh.. . . 457

cellulicola (Naeg.) de Toni 457

cypericola (Magn.) Weber 456-7

digitata Lagerh 457

raunkiaeriana Ferd. & Winge. . 457

scirpicola (Correns) Sacc. & Syd. 457

solani Fautrey 457

Epithele 276

Epithele typhae 271, 276

Eragrostis 368

Eragrostis curvula 368

major 360, 362

pilosa 368

Erica L 431

Erica eylesii Bolus 431

eylesii var blaeriodes H. Wild 429-31

gazensis H. Wild 428-9

lanceolifera S. Moore 428-9

milaniiana Bolus 428-9

thryptomenoides S. Moore 431

ERICACEAE 428

Erineum 72, 72

Eriospermum (sect, of fpomoea) 713

Eriospermum 443-4

Eriospermum bellendini 444

bifidum R. A. Dyer 443

zeyheri R. A. Dyer 444

Erythrina (Notes on certain S. African

species) 507

Page

Erythrina abyssinica Lam 507, 510-511

caffra Auct. plur. non Thunb... . 509

caffra Thun b.. .507, 508-9, 510, 513-4

constantiana Micheli 508-9

gibbsae Bak 511

insignis Tod 509

latissima E. Mey 507, 510-1

lysistemon Hutch. . 507, 508, 509, 514

tomentosa R. Br. ex A. Rich. .507, 51 1

viarum Tod 508-9

Erythrophysa alata (E. & Z.) Hutch 515

Eucalyptus 187, 493

Eu-Helminthosporium 347, 349, 367

Eupatorium retrofractum Thb 220

Euphorbia barnardii W. D. & S 225

clavarioides Boiss 222

clivicola R. A. Dyer 227-2, 242

confinalis R. A. Dyer 222-3, 243

corniculata R. A. Dyer 225-6

enormis N. E. Br 224-5

excelsa W. D. & S 222-3

keithii R. A. Dyer 223-4, 244-6

restricta R. A. Dyer 224-5, 247

schinzii Pax 221-2, 225-6

sekukuniensis R. A. Dyer 224

triangularis Desf 222-3

unicornis R. A. Dyer 225, 248

zoutpansbergensis R. A. Dyer. . 223-4

EUPHORBIACEAE 221

Evolvulus L 643, 661

Evolvulus alsinoides L 661

eapensis E. Mey. ex Choisy 663

hederaceus Burm. f 700

mummularius L 661

Exidia 255, 291, 301

Exidia thuretiana 267

Exobasidium Woronin 5, 11, 272

Exobasidium andromedae Peck 11

arctostaphyli Harkn 11

azaleae Peck 11

cassandrae Peck 11

discoideum Ellis 11

giseckiae Allescher 77, 77

karstenii Lind 11

karst enii Sacc 11

inyrtilli Siegm 11

oxycocci Rostr 11

peckii Halst 11

rlrododendri Cramer 11

vaccinii (Fuckel) Wor 11, 71

vaccinii-myrtilli (Fuckel) Juel. 11

Fagara 234

Fagara eapensis Thunb 234

humilis E. A. Bruce 234- 5

magalismontana Engl 234

Falkia Linn, f 643, 658

Falkia abyssinica Engl 660

dichondroides Baker 659, 660

diffusa (Choisy) Hall, f 65°, 660

oblonga Berhn 659, 660

repens Linn, f 658, 659

villosa Hall, f 65?

Favolaschia 283, 290

Favolaschia peziziodea 283

Favolus 275

Favolus megaloporus 293

10

Page

Femsjonia luteoalba 262, 271

natalensis Cooke 476, 486

Festuca L 139, 142, 143, 426

FESTUCEAE 139

Fomes 55, 275, 291

Fomes laevigatus 291

Fusarium 356

Fusarium avenaceum 357, 360

culmorum 357

Fusidium vaccinii Fuckel 11

Galega genistoides Thunb 166

genistoides Thb. var B 164

Galzinia cymosa 262

Genista buxifolia Burm. f 168

Genlisea 385

GENTIANACEAE 602

GERANIACEAE 432

GESNERIACEAE 433

Gloeocystidiellum 291

Gloeocystidium argillaceum H. & L 13, 14

pallidum Bres 13

pallidum subsp. argillaceum 1 3

Gloeopeniophora 291

Gloeosoma 465

Gloeosoma capensis Lloyd 477, 479

Gloeotulasnella 291, 301

Gloiothele 277

Glomerularia ramosa Karst 191

Gloniopsis levantica Rehm 185, 196

Glonium parvulum (Gerard) Sacc 185-6, 196

Gompholobium maculatum Andr 161, 166

Gonatobotrys ramosa (Karst.) Reiss. . . .191, 201

GRAM1NEAE 120, 122, 123, 126

Grammothele B. & C 5, 59, 276-7, 295

Grammothele cineracea Bres 60, 277

liniata B. & C 60, 277

mappa B. & C 59, 60, 113,

277 287—8

polygramma B. & C 60, 277

pseudomappa Talbot .59, 60, 112,

277, 287

Grandinia Fr 5, 60

Grandinia bicolor Talbot 60

farinacea 277

mucida 279

rosea P. Henn 61

Guadua 125

Guajacum L 516

Guajacum afrum L 517

Gynmosporium cinctum B. & Br 190

Gyrophana himantioides (Fr.) B. & G 31

Helichrysum argentissimum J. M. Wood. . . 423

coactum M. D. Henderson.. . 421

fulgidum Willd 423

grandibracteatum M. D. Hen-

derson 422

randii S. Moore 422-3

scopulosum M. D. Henderson 423

setosum Harv 424

tenax M. D. Henderson 423

tenuifolium Killick 424-5

triliniatum DC 424-5

umbraculigerum Less 421, 422

Helicobasidium Pat 5, 7

Helicobasidium compactum (Boedijn)

Boedijn 7, 70, 266, 271

longisporum Wakef 7, 8

Helicobasidium mompa Tanaka 8

mompa forma macrospo-

rum Hara 7

purpureum (Tul.) Pat.. .8, 262

purpureum var orientate

Pat 7

Helicobasis purpureas (Tul.) Clem. & Shear 8

Helicogloea 251-3, 255, 256

Helicogloea lagerheimi 261, 262

Helicoma ambiens Morgan 492

curtisii Berk. 187, 188, 198, 492, 493

interveniens Talbot 492, 498

larvale Morgan 491

mulleri Corda 492

velutinum Ellis 492

Helicomyces gracilis Morgan 492

Helicosporium gracile (Morgan) Linder — 492,

498

lumbricoides Sacc. em.

Matruchot 492-3, 498

lumbricopsis Linder 493

ramosum Talbot 493, 498

vegetum Nees 492

Helminthosporium (Some graminicolous

species) ....347, 369

Helminthosporium acrothecioides Lindfors 355

arundinis 350

avenae 350

bicolor Mitra 348, 357,

358-9, 360, 374

brizae Nisikado 363, 376

bromi 350

catenarium 350

crustaceum P. Henn 366

cynodontis Marign. .362-3, 368,

376

dermatioideum Bubak &

Wrobl.,351, 357, 360, 363,

368-9, 378

dictioides Drechsl 350

erythrospilum Drechsl. . 350

euchlaenae Zimm 358

euchlaenae strain sideris 358

giganteum Heald & Wolf 350,

351

gramineum Rabenh 348,

350, 352, 353-4, 373

halodes Drechsl 348, 359,

360, 374

halodes var tritici Mitra 357,

359

hoffmanni B. & C 366

hordei Eid 351

inconspicuum Cooke &

Ell 360

kusanoi 351

leersii Atk 348, 365, 376

leucostylum Drechsl 355,

361-2, 375

maydis Nisikado &

Miyake 351, 361

mediocre K. M. Putterill 354-5.

373

miyakei Nisikado. . . . 347, 35 1 .

367-8, 377

N. Henry 357, 360, 374

nodulosum B. & C 355

1]

Page

Helminthosporium olisponense 350

oryzae Breda de Haan.351, 366

penicillosum Speg 350

ravenelii Curt .366-7, 368, 377

rostratum Drechsl. . .348, 357,

359-60 , 367, 375

sacchari ( B. de H.) Butler 348,

364, 365, 376

sativum Pam., King &

355-7, 359- 60, 364, 374

setariae 351

siccans Drechsl 350

sigmoideum Cav 351

stenospilum 351

teres Sacc... .349, 350, 351-3,

354, 373

tetramera McKinney.. . .369-70

tonkinense Karst. &

Roum 366

tritici-repentis 350

tritici-vulgaris 350

turcicum Passer. . .359, 360-1,

375

urochloae K. M. Putterill 365-6,

377

vagans Drechsl 350

Helvetia mesenterica Dicks 6

Henderson, M. D. et alia: Newly described

species 415

Herpotrichia pezizu/a (B. & C.) Ell. &

Everh 187

Heterochaete 276, 277

Heterochaete andina Pat 10

Heterochaetella 285

Hewittia W. et A 643, 698

Hewittia bicolor (Vahl) Wight 698

sublobata (Linn, f.) O. Ktze 698

Hoehnelomyces Weese 302

Hordeum 352, 354, 356

Hordeum distichum 352

vulgare 352, 354

Hormodendrum 351

Hyaloria A. Moll 251, 302

Hydnum L. ex Fi 5, 62

Hydnum alnicolum Velenov 55

ambiguum B. & Br 62

argutum Fr 61

daviesii Sow 55

erinaceum 279, 288

microdon Pers 55

mucidum Fr 62

ochraceum Pers. ex Fr 55, 56, 62

pudorinum Fr 55, 62

sclerodontium Mont. & Berk 62

stipitatum Fr 61

Hygrophorus 251

Hymenochaete Lev... .5, 43, 48, 264, 277, 291

Hymenochaete bonariensis Speg 41

castanea Wakef 46-7

cervina B. & C 49

cinerascens (Schw.) Lev 40

corrugata (Fr.) Lev 48

crassa (Lev.) Berk 42

dregeana (Berk.) Massee. . . 57, 59,

344

ellisii Berk. & Cooke 35

Page

Hymenochaete fasciculata Talbot .48-9, 101, 292

fuliginosa (Pers.) Bres 48-9

fuscoviolascens (Mont.)

Sacc 50, 62

griseocervina P. Henn 43

kalchbrenneri Massee 42, 329

lictor Petch 49

luteobadia (Fr.) H. & L... .45, 50,

320, 330

multispinulosa Peck 42

nigricans (Lev.) Bres 45, 292,

322, 330

ochromarginata Talbot. . . .48, 49,

50. 103

olivaceum Cooke 43, 315

pellicula 329

purpurea Cooke & Morgan. 42, 329

rigidula B. & C 50

rubiginosa Dicks, ex Lev. . . 45, 50,

326

scabriseta Cooke 42, 329

semistupposa Petch. . .48, 49, 102

tabacina (Sow. ex Fr.) Lev.. 50.

327

tenuissima Berk 50

tristicula (B. & Br.) Massee. 46-7,

50

tristiuscula ( B. & Br.) Massee 46

umbrina B. & C 42

vinosa (Berk.) Cooke 42, 329

Hymenogramme 276, 277

HYMENOMYCETES (Micromorphology

of lower) 249

HYMENOMYCETES (Some S. African

resupinate) 1

Hypericum 469

Hypochnella violacea 271, 285

Hypochnus 309

acerinus (Pers.) Pat 28

bagliettoanus 10

eylesii v. d. Byl 63

olivaceus Fr 35

pallescens (Schw.) Burt 53, 105

puniceus A. & S 63

purpureas Tul 8

thelephoroides (Ell. & Everh.)

Burt 53

Hysterium insidens Schw 186, 187, 197

parvulum Gerard 185

Hysterographium insidens (Schw.) Sacc 186

Ibbet sonia genistoides Sims 161, 166

Ilex aequifolium L 536, 539

llysanthes 235

Ilysanthes welwitschii Engl 235

Iphonia polygalaefolia Benth. & Hook. . . . 220

Ipomoea L 644, 711

Ipomoea adenioides Schinz 720. 770

adumbrala Rendle & Britten 730

aitoni Choisy 738

alba L 720, 165

albivenia (Lindl.) Sweet 719, 768

angulata E. Mey. ex Drege 738

angustifolia Jacq 706

angustisecta Engl 758

aqjatica Forsk 716, 153

araclmoidea Boj. ex Choisy 738

arachnasperma Welw 714, 715, 136

12

Page

Ipomoea argyreoides Choisy 775

aspericaitlis Baker 758

atherstonei Baker 718, 741

baclei Choisy 765

barret tii Rendle 775

batatas (L.) Lam 715, 721, 746

bathycolpos Hall f 716, 749

beUecomans Rendle 730

biflora (L.) Pers 723

biloba Forsk 754

bipinnipartita Engl 705

blepharosepala Hochst. ex A. Rich 729

bolusiana Schinz 715, 721, 758

bowieana (Rendle) Baker 692, 693

cairica (L.) Sweet 720, 761

calcarata N.E. Br 760

calophylla Fenzl 736

calycina Benth. ex C. B. Clarke. . 728

calycina Meissn 728

cana E. Mey. ex Drege 775

ca/ystegioides (Choisy) Hall f 730

cardiosepala Hochst. ex FI. Cap.. 729

cardiosepala Meissn 729

cecilae N.E. Br 727

chloroneura Hall f 714, 739

clappertonii .R Br 753

eonceiroi Rendle 769

congesta R. Br 717, 135

contorta C hoisy 751

contorta Engl 776

convolvuloides Hall, f 748

convolvuloides Schinz 706

coptica (L.) Roth 720, 760

coscinosperma Hochst. ex Choisy 714,

121

crassipes Hook 715, 717, 718, 730

crispa (Thunb.) Hall, f 719, 151

cnrtoi Rendle 784

cynanchifolia C. B. Clarke. 723, 724, 725

cynanchifolia Meissn 723, 724

dammarana Rendle 769

dasysperma Jacq 720, 760

demissa Hall, f 747

dicliroa (R. & S.) Hochst. ex

Choisy 736

digitata L 719, 767

dissect a Willd 760

eenii Rendle 743

eriocarpa R. Br 714, 722

ficifolia Lindl 715, 738

fragilis Choisy 747

fragrans Boj 766

geminiflora Welw 723, 724, 725

gerrardi Baker 768

gerrardiana Rendle 734

gracilisepala Rendle 714, 725

greenstockii Rendle 730

hackeliana (Schinz) Hall, f 713, 726

liardwickii (Spreng.) Hemsl 728

hederacea Jacq 733

hederacea Auct. non Jacq 733

hellebarda Schweinf. ex Hall. f. . . 752

heptaphylla (Rottl. & Willd.)

Voigt 720, 764

hewittioides Hall, f 731

hispida (Vahl) R. & S 722

hispida Zucc 722

Page

Ipomoea hochstetteri House 720, 762

holosericea E. Mey. ex Drege. ... 738

holubii Baker 780

huillensis Baker 683

hystrix Hall, f 756

inamoena Pilger 756

involucrata P. Beauv 713, 744

k en trocaulos C. B. Clarke 704

kwebensis N. E. Br 762

lachnosperma Choisy 709

lambtoniana Rendle 778

lapathifolia Hall, f 720, 752

leucanthemum (Klotzsch) Hall. f. 714,

122

linosepala Hall, f 726

lilacina B1 766

longipes Garcke 747

longipes Engl 747

lugardii N. E. Br 742-3

magnusiana Schinz 720, 742

malvaefolia (Rendle) Baker 706

marlothii Engl 770

megalochlamys Baker 784

mesenteroides Hall, f 758

natans Dinter & Suesseng 753

nil (L.) Roth 717, 733

oblongata E. Mey. ex Choisy 778

obscura (L.) Ker 716, 746

odontosepala Baker 760

oenotherae (Vatke) Hall, f 7 1 5 . 720, 727

oenotheroides (Linn, f.) Rafin. ex

Hall, f 775

ommaneyi Rendle 718, 740

otjikangensis Dinter & Pilger 742

ovata E. Mey. ex Rendle 732

pachypus Pilger 727

pahnata Forsk 761

paniculata ( L.) R. Br 767

papilio Hall, f 716, 717, 750

parkeri Choisy 766

pellita Hall. f. 718, 732

pes-caprae (L.) Sweet 714, 754

pes-tigridis L .......711, 720, 744

petunioides Baker 727

pileata Roxb 713, 745

pilosa Cav 736

pilosa Houtt 736

pilosa (Roxb.) Sweet 736

pinnata Hochst. ex Choisy 707

plantaginea (Choisy) Hall. f. 755

plebeia R. Br 714, 723

plebeia R. Br. subsp. africana A.

Meeuse 723

praetermissa Rendle 758

pterygocaulos Steud. ex Choisy. . 702

pulchella Roth 764

purpurea (L.) Roth 717, 734, 735

pyramidalfs Hall .f 784

quinquefolia Hochst 762

quinquefolia L 762

quinquefolia var pubescens Baker . 703

quinquefolia var purpurea Hall. f.. 762

randii Rendle 778

replans Poir 753

rhodesiana Rendle 780

riparia G. Don 717, 766

13

Page

Ipomoea robertsiana Rendle Ill

rubro-coerulea Hook 759

saccata Hall, f 760

sagittaefolia Burm. f 753

sagittaefolia Hochr 753

sarmentacea Rendle 730

saundersiana Baker 708

seineri Pilger 769

semisecta Merxm 756

shirambensis Baker 719, 770

shirensis Oliv 782

simplex Hook 758

simple<Thunb 715, 755

sinensis (Desr.) Choisy 714, 128

sinensis subsp. blepharosepala

(Hochst. ex A. Rich.) Verdcourt

ex A. Meeuse 729

stenosiphon Hall, f 783

sublucens Rendle 776

suffruticosa Burch 776

tetraptera Baker 702

transvaalensis A. Meeuse.. 7 16, 718, 748

tricolor Cav 716, 759

turpethum (L.) R. Br 708

uncinata Hutch Ill

undulata Baker 751

verbascoidea Choisy 719, 769

verecunda (Rendle) N. E. Br 703

violacea L 759

vitifolia E. Mey. ex Drege 738

welwitschii Vatke 715, 718, 756

wightii ( Wall.) Choisy 714, 737

woodii N. E. Br 773

xipliosepala Baker 725

zambesiea Baker 752

IRIDACEAE 436

Iris 456-7

Irpex 11

Irpex dregeanus (Berk.) Talbot. ... 287. 339, 344

Irpex vellereus B. & Br 339, 344

Jaapia argillacea 271-3

Jacquemontia Choisy 643, 699

Jacquemontia capitata (Desr.) G. Don 700

ferruginea (Steud.) Choisy 699

tamnifolia (L.) Griseb 700

thomensis Henriq 722

JASM1NEAE 602

Jasminum Linn 550, 556

Jasminum abyssinicum (Hochst. ex) DC. . . 557,

angulare Vahl.. 556-7, 559, 560, 561,

563, 566, 614

angulare var glabratum E. Mey. 549,

560, 561, 562

breviflorum Harv 549, 557, 564 ,

565-6, 614

capense Thb 560, 561

flexile Jacq 559

fluminense Veil.. .557, 560, 561, 562,

563, 564, 615

gerrardi Harv 549, 564, 565,

566, 617

glaucum (L. f.) Ait 557, 566, 567,

569, 618

glaucum var lanceolatum E. Mey 566

Page

Jasminum glaucum var parvifolium E. Mey. 567,

568

goetzianum Gilg 562

hirsutum Linn 556

humile Linn. Cult 557

ligustrifolium Lam 566

lupinifolium Gilg. & Schellenb. . 558-9

mauritianum Boj 562-3

meyeri-johannis Engl 569

multiflorum (Burm.) Andr. Cult. 556

multipartitum Hochst 556-7, 567,

569, 570, 619

natalense Gilg. & Schellenb 560-1,

563

officinale Linn 556

pauciflorum Bth 572

pubeseens Willd 556

quintum Schinz. . . 557, 558, 559, 611

schroeterianum Schinz 562-3

stenolobum Rolfe.557, 569, 570, 620

stolzeanum Knobl 566

streptopusE. Mey. 549, 558, 570, 572

streptopus subsp. streptopus.,558, 570,

571, 621

streptopus subsp. transvaalense

(Sp. Moore) Verdoorn. . . .558, 570

572, 622

swynnertonii Sp. Moore 572

tortuosum Willd 557-8, 559, 560

612, 613

transvaalense Sp. Moore 572

wyliei N.E. Br 563^1

Jatropha (subgenus Adenoropium) 227

Jatropha messinica E. A. Bruce 226-7

pseudoglandulifera Pax 226-7

pseudo-glandulosa 227

Jola 251

JUNCACEAE 457

Juncus articulatus 459

bufonius 456, 459

lamprocarpus 459

Kies, P. Notes on Asparagus africanus and

related species 177

Kies, P. Revision of the genus Cyclopia and

notes on some other sources of Bush Tea 161

Kies, P. et alia: Newly described species.. . 213

Killick, D. J. B. el alia: Newly described

species 415

Kneiffia 285

Kneiffia purpurea (Cooke & Morgan) Bres. 42

roumeguerii (Bres.) Bres 22

setigera Fr 21, 285

Kniphofia ensifolia Bak 231

ensifolia var albiflora E. A. Bruce 231

rivularis Berg 231

tuckii Bak 231

Koeleria 426

Kordyana 258

LABIATAE 227, 439

Lacellina graminicola (B. & Br.) Petch. .190, 200

libyca Sacc & Trott 191

Lacbnella flammea (A. & S.) Fr 184

Lachnocladium 295

Lactarius 278

Lasiosphaeria pezizula (B. & C.) Sacc. .187, 198,

493

Lecanidion 183

14

Page

Lecanidion atratum (Hedw.) Rab 183

Leers ia 120

Leersia hexandra 365

virginica 365

Lcgendrea Webb, et Berth 773

Leiocalyx Hall. f. (sect, of Ipomoea) 712

LENTIBULARIACEAE 385

Lenzites 275

Leptodon ochraceum (Pers. ex Fr.) Quel.. . . 55

Leptosphaeria 351

Leptosphaeria doliolum (Pers.) de Not 351

salvinii Catt 351

LEPTOTACEAE 465

Licentia yao-chanica Pilat 56-7, 340

Lichen atratus Hedw 183

Ligneira 458

Ligniera isoetes Palm 458

junci (Schwartz) Maire & Tison. . 453,

457, 458-9, 460, 463

pilorum Fron & Gaillat 458

verrucosa Maire & Tison 458

LILTACEAE 229, 440, 542

Limeum (parasitic fungi in) 453

Limeum glomeratum 453-4

viscosum 453-4

Linderomyces 280

Lindernia 235

Lindernia dongolensis E. A. Bruce 235-6

stuhlmanii Engl 235

Linociera Sw 550, 573, 592, 593-4, 601

Linociera africana Knobl 600, 601

africana (Welw.) Gilg. & Schel-

lenb 600

battiscomei Hutch 593, 594, 597,

599, 600, 601. 636

congesta Bkr 601

foveolata (E. Mey.) Knobl. . .594, 595,

598

foveolata subsp. foveolata .. 594, 595,

596-7, 600, 631, 632

foveolata subsp. majora Verdoorn 594,

596-7, 598, 599, 600, 634

foveolata subsp. tomentella

Verdoorn.. 590, 597, 598, 601, 633

latipetala M. R. F. Taylor 594

ligustrina Sw 593

marloihii Knobl 595-6, 632

milbraedii Gilg. & Schellenb 594

peglerae (C. H. Wr.) Gilg. &

Schellenb 594, 597, 599, 635

Lloydella Bres 57, 285, 297, 343

Lloydella beyrichii (Fr.) Bres 40

bicolor (Pers. ex Fr.) Bres 39

cinerascens (Schw.) Bres 40

involuta (Klotzsch) Bres 318

occidentalis (Ell. & Everh.) H. &

L 41

scabriseta (Cooke) H. & L 42

schomburgkii (Berk.) Bres. var

brimnea Bres 43

Lonicera 470

Lopharia (the genus) 339, 343

Lopharia Kalchbr. & MacOwan 5, 56

Lopharia dregeana (Berk.) Talbot 19, 47, 57,

58, 110, 339, 344

javanica P. Henn. & E. Nym 57, 58,

62, 339, 342, 343

Lopharia lirellosa Kalchbr. & McO.. . .56-7, 309,

339-40, 343

mirabilis (B. & Br.) Pat. . .41, 56, 57-9,

62, 109, 287, 309, 310, 339^46

Macrosporium 350

Maltebrunia Kunth 117, 121

Maltebrunia leersioides Kunth 121

prehensilis Nees 117, 118, 121

Marasmius 266, 283

Marasmius epiphylloides 283

Matula poroniaeformis (B. & Br.) Massee. 478

rompelii (Rick) Lloyd 478

Meeuse, A. D. J. : The South African Con-

volvulaceae 641

Meeuse, A. D. J. et alia: New and Interesting

records of S. African Flowering Plants. . 535

Meliola 347, 351

Melocaninae 124

Menodora Humb. & Bompl 550, 601, 603

Menodora africana Hook.. 602, 604,

605-7, 638

helianthemoides Humb. &

Bonpl 602

heterophylla Mor. ex DC 607

heterophylla var australis

Steyerm 602, 605, 606,

607, 639

heterophylla Oliver non Mor... 606

juncea Harv 549, 602, 603-5,

607, 637

Merremia Dennst. ex Hall f 643 , 700

Merremia angustifolia (Jacq.) Hall, f 706

bipinnipartita (Engl.) Hall, f.,,701, 705

bowieana Rendle 692, 693

convolvulacea Dennst 700

hederacea (Burm. f.) Hall, f 700

kentrocaulos (C. B. Clarke)

Rendle 701, 704

malvaefolia Rendle 702

palmata Hall, f 701, 704

pinnata (Hochst. ex Choisy)

Hall, f 701, 707

quercifolia Hall, f 701

tet rapt era (Baker) Hall. f. 702

tridentata (L.) Hall. f. ssp. angus-

tifolia (Jacq.) Ooststr 702, 706

tuberosa (L.) Rendle 708

turpethum (L.) Rendle 708

verecunda Rendle 701, 703

Merulius Haller ex Fr 5, 29

Meridius confluens Schw 30, 31, 33

corium (Pers. ex Fr.). . .29, 30, 31, 33,

88

gelatinosus Lloyd 29, 32, 33, 90

himantioides Fr 29, 31, 32-3, 89

lacrymans Wulf. ex Fr. . .32, 33, 59,271

papyrinus (Bull.) Quel 30

serpens Fr 33

squalidus Fr 31-3

strigoso-zonata Schw 28

tenui 'i Peck 31

umbrinus Fr 31-3

Mesobotrys graminicola (B. & Br.) Sacc. . . . 190

Metaporana N. E. Br 664

Metaporana ango/ensis N. E. Br 665

densiflora (Hall, f.) N. E. Br... 665

Micromeria 440

15

Page

Necator decretus Massee 17, 18

Neotyphula guinensis 262

Nigrospora Zimm 189

Nigrospora oryzae (B. & Br.) Petch 189, 199

sphaerica (Sacc.) Mason 189

Notosceptrum 545

Notosceptrum alooides Benth 544

Nyctalis asterophora 272

parasitica 272

Nyctanthus glauca Linn, f 566

OCHNACEAE 232

Ochna glauca Verdoorn 232

pretoriensis Phillips 232-3

Odontia Pers. ex S. F. Gray 5, 61, 285

Odontia arguta (Fr.) Quel. . . .61, 114, 275, 287

bicolor 279, 287, 295

crustosa (Pers.) Quel 62, 285, 287

knysnana v. d. Byl 61, 62, 115,

papillosa

setigera (Fr.) Miller.

285, 287

287

21

Oedemansiella 285, 290

Oedogonium 388

OleaLinn 550, 572, 573

Olea africana Mill 573, 574, 576-80, 632

buxifolia Mill 582

capensis Linn 573, 581, 582, 584, 626

capensis subsp. capensis. 573, 581-2, 584-7,

592, 626-8

capensis subsp. enervis (Harv.) Ver-

doorn 573, 581, 588-9, 629

undulata Jacq 582

undulata Jacq. var planifolia E. Mey . . 582,

584

urophylla (Gilg.) Gilg. & Schellenb... 581

verrucosa Link. ... 573, 574, 576, 578, 623

welwitschii (Knobl.) Gilg. & Schellenb 581 ,

593

woodiana Knobl. 573, 574, 577, 578, 579,

592, 624

OLEACEAE (of Southern Africa) 549

Olpidiaster radicis (De Willd.) Pascher 459

Olpidium borzii de Willd 459

brassicae (Wor.) Dang 453, 457,

458-9, 460, 463

radicicolum de Willd 459

Omphalobium Jacq. f. ex DC 516

Omphalobium schotia Jacq. f. ex DC 523, 525

Operculina S. Manso 643 , 708

Operculina turpethum (L.) S. Manso 708

Ophiobolus 349-50

Ophiobolus heterostrophus Drechsl 351

sativus 355

Ophioglossum nudicaule L. f 544

Oryza 120, 122, 124

Oryza australiensis Domin 122

prehensilis (Nees) Steud 118

ridleyi Hook f 122

sativa 121, 494

ORYZEAE 117, 120-5, 127, 136

Oryzinae 121

Oryzoideae 123

Panicum frumentaceum 362

16

Page

Papulaspora aspergiliiformis 192

Passalora graminis (Fuckel) Hohn. . . . 192-3, 204

Patellaria atrata (Hedw.) Fr 183, 194

Patouillardina 260

Patouillardina cinerea 262

Paxillus lateritius 280 1

PEDAL1ACEAE 233

Pegolettia polvgalaefolia Less 220

retrofracta (Thb.) Kies 220

Pelargonium carnosum L 432

jacobii R. A. Dyer 432, 450

mirabile Dinter 432

Pellicularia 257-9, 261, 266

Pellicularia filamentosa 268

flavescens 258-9, 261, 271

pruinata 262, 273

subcoronata 275

Peniophora Cooke 5, 20, 263-4, 266,

285, 297

Peniophora aegerita 287

affinis 287

arenata Talbot 20, 22

argillacea Bres 13, 14

aspera (Pers.) Sacc. .20, 21, 36, 81,

267. 285, 287

atrocinerea Mas see. 43^1, 306, 315-6

aurantiaca 269

berkeleyi Cooke 40, 343

byssoidea (Pers.. ex Fr.) H. &

L 21, 267, 270, 287-8

carneola (Bres.) H. & L 13

cinerascens (Schw.) Sacc 40

cinerea (Pers. ex Fr.) Cooke. 19, 24-5,

42, 267, 269, 287, 329

cremea Bres 20, 25, 285, 287

dissita (Berk.) Cooke 41

ephebia (B. & C.) Massee 41

filamentosa (B. & C.) Burt. . . .20, 23,

24, 83, 297

gigantea (Fr.) Massee.. . .20, 23, 24,

84, 264, 267, 288

glebulosa Bres 25, 287-8

heterobasidioides 260

incarnata 290

intermedia Massee 42

laevigata (Fr.) Massee 47

laevis 287

leprosa B. & G 22

livida 290

longispora 287-8

macro Karst 22

molleriana (Bres.) Sacc 22

moricola (Berk.) Massee 40

neglecta (Peck) Peck 41

nuda (Fr.) Bres 19, 25

occidentalis EU. & Everh 41, 343

pallidula (Bres.) B. & G....21, 287-8

pelliculosa Talbot. . .20, 25, 63, 80,

287

proxima 267, 287-8

pubera (Fr.) Sacc. . . .23, 267, 287-8

quercina 269

radicata (P. Henn.) H. & L... . 23

roumcguerii (Bres.) Burt.. 20, 22,

23, 41, 82, 287-8

sanguinea 285

ichweinitzii Massee 40

Peniophora setigera (Fr.) Bres 21, 267, 287

sphaerospora 290

stratosa Burt 22

subalutacea 270, 287-8

subiculosa Burt 20

tenuis 289, 290

tomentella Bres 21, 287

unicolor Peck 23

velutina 267, 287

Pennisetum 364

Pennisetum clandestinum 255

typhoideum 355, 362

Periconia 351

Perrotia flammea (A. & S.) Boud 184, 194

Pestalotia Auct. non de Not 379

Pestalotia burchelliae Laughton. .. .379, 381, 383

caffra Syd 379, 381, 383

cassinis 382

encephalartos Laughton . 379, 381, 383

laughtonae Doidge 379, 381, 383

laurophylli 381, 383

milletiae Laughton 379, 381, 383

ocoteae 381, 383

pelargonii Laughton 379, 381, 383

podocarpi Laughton non Dennis. 379,

381, 383

pterocelastri Laughton 382, 383

rapaneae Laughton non Viegas. . 379,

381, 383

trichocladi Laughton. .. .381. 382, 383

watsoniae Verw. & Dipp 382, 383

Pestalotiopsis Steyaert 379

Pestalotiopsis (Some South African) 379

Pestalotiopsis glandicola (Cast.) Steyaert. . 381,

382, 383

ilicis (West) Steyaert 382

planimi (Vize) Steyaert 381

trichocladi (Laughton) Stey-

aert 381, 382

Peziza albovioiascens A. & S. ex Fr 471

flammea A. & S 184

punctiformis Fr 474

Phaeocyphella galeata 271

Pharbitis (Choisy) Griseb. (sect, of lpomoea) 711,

713

Pharbitis fragrans Boj. ex Choisy 766

Phellinus Quel 291

Phlebia Fr 5, 28, 29, 264

Phlebia hispidula Berk 29

mesenterica (Dicks, ex Fr.) Fr 6

pileata Peck 28

reflexa Berk 28

rubigiiiosa Berk. & Rav 28

rugosissima Lev 28

strigoso-zonata (Schw.) Lloyd 28, 87

zonata B. & C 28

Phleogena Link 251, 302

Phleogena faginea 260

Phoma 350

Phycopsis africana Syd 185

Phymatot' ichum 495

Phymatotrichum silvicolum Tabenh. &

Watk 495, 500

PHYTOLACCACEAE 445, 454

Phytophthora 192

Pinguicula 385

Pinus 501, 503

17

Page

Pinus patula 501-2, 505

taeda 327

Piricularia grisea (Cooke) Sacc 494-5

oryzae Cav 494, 499

setariae Nishikado 494

zingiberi Nishikado 494

Pistillaria cardiospora 271

Platygloea effusa 271

Pleospora 350

Pleospora herbarum (Pers.) Rab 350

teres Died 352

trichostroma (Fr.) Wint 353

Podalyria genistoides Wi 1 Id 164, 166

Podocarpus 469

Podosporiella verticillata O’Gara 357

Polydesmus 187

Polylepidea 423-4

Polyothamnus (sect, of Rivea sensu Hall. f.).780, 784

Polypompholyx 385

Polyporus 275, 291

Polyporus angolensis 274

carneofulvus 293

cuticularis 293

dictyopus 295

dryadeus 293

gilvus 292

grammocephalus 292

nidulans Fr 24

ochroporus 293, 295

patouillardi 293 , 295

purpurascens (DC.) Pers 30

radiatus 292

rutilans (Pers.) Fr 24, 297

subiculoides 279, 280

sulphureus 274, 278

tabacinus 292, 293

Polystictus 275

Polystictus pinsitus 274

subiculoides Lloyd 287

tabacinus Mont 50

Populus 477

Populusdeltoides 475, 489

Porana Burm. f. 665

Porana densiflora Hall, f 665

subrotundifolia De Wild 782

Poria 11. 57, 63. 275, 291

Poria contigua 292

eupora 287

lamellosa P. Henn 277

mucida 279, 280

obducens 287

ravenalae B. & Br 60

uncinata 292

versipora (Pers.) Fr 63

viridans 270

Porogramme 276, 277

Porotheleum (Fr. ex Fr.) Fr 63, 465, 479-80

Porotheleum incanum (Kalchbr.) Sacc 479-80,

481, 486

Porothelium Eschweiler 480, 481

Potamophila R. Br 117, 121, 122, 126

Potamophila parviflora R. Br 117, 121, 126

prehensilis (Nees) Benth. 117-8, 120,

121-2, 126, 128-135, 137

Psammotropha alternifolia Killick 445-6

androsacea Fenzl 445-6

Pseudoblepharis 539

1339563-2

Page

Pseudobromus K. Schum 139, 142-3

Pseudobromus africanus (Hack.) Stapf. .131, 140,

142, 145-151

silvaticus K. Schum 142, 144

Psilopeziza mirabilis B. & C 469

Ptychogaster 26

Puccinia caricis 262

Punctularia Pat 5, 25

Punctularia affinis (B. & C.) Talbot.. 25, 85, 324

atropurpurascens (B. & Br.)

Petch 25, 26

tuberculosa (Pat.) Pat 25, 26

Putterill, K. M.: Some Graminicolous

species of Helminthosporum and Curvu-

laria occurring in S. Afri/ca 347

Pyrenophora 347, 349, 350, 351, 368

Pvrenophora avenae 350

chaetomioides Speg 350

graminea Ito 350, 353

japonica Ito 350

phaeocomes (Reb.) Sacc 350

polytricha da Cam 350

relicina (Fuckel) Sacc 350

teres 349-50, 352

tritici-repentis (Died.) Drechsl. 350

Quamoclit Moench 712

Quercus 477

Radaisiella elegans Bain 490

Radulum Fr 5, 57, 62

Radulum javanicum (P. Henn.) Lloyd 62

membranaceum (Bull.) Bres 19

mirabile B. & Br 56, 62, 309,

339^40, 343

orbiculare Fr 62

Raphnia retroflexa B. Thb 169

Reticidaria affinis B. & C 25, 26

atrorufa B. & C 25, 26

pyrrhocreas 26

venulosa B. & C 25, 26

Rhizoctonia 8, 363, 368

Rhizoctonia crocorum (Pers.) DC 8

medicaginis DC 8

violacea Tul 8

Rhizomyxa hypogaea Borzi 459

Rhus 407

Rhus (section Tomentosa) 540

Rhus batopjhylla L. E. Codd 539-40, 547

decipiens E. Mey 407

erosus E. Mey. non Thunb 407-8

erosus Thunb 407-8

lancea 481

leitcocarpa E. Mey 407

melanocarpa E. Mey 407

monophylla (E. Mey.) Presl 407

spicatum Thunb 407

tomentosa L 539

undulata Jacq 407

Rhynchospora aurea 188

cyperiodes 188

dolichostyla 188

Rivea Choisy 711, 712, 773, 774

Rivea adenioides (Schinz) Hall f. 770

holubii (Baker) Hall, f 780

oenotheroides (Choisy) Hall, f 775

pyramidalis (Hall, f.) Hall. f. 784

shirensis (“ schirensis ”) (Oliv.) Hall. f. 782

stenosiphon (Hall, f.) Hall, f 783

18

Page

Rivea suffruticosa (Burch.) Hall, f 776

Rubus 540

RUTACEAE 234

Saccharum officinarum 364

Saccoblastia sebacea 262

Salsola glabrescens 218

humifusa Breuckner 215-1

rabieana Verdoom 218-9

tuberculata 218

SAPINDACEAE 407

Schinzia Denstatt 456

Schinzia Naegeli 456

Schinzia aschersoniana Magn 456-7

casparyana Magn 457

cellulicola Naegeli 457

cypericola Magn 456-7

scirpicola Correns 457

Schmidelia 407

Schmidelia decipiens (E. Mey.) Presl 407

dregeana Sond 407

leucocarpa Sond 407

melanocarpa Sond 407

monophylla (E. Mey.) Presl 407

monophylla var natalitia Szyszyl 407

rehmanniana Szyszyl 407

natalensis Sond 408

Schotia 515, 516, 532

Schotia (of Southern Africa) 515

Schotia afra (L.) Bodin 5 1 5, 516, 529, 530, 532

afra (L.) Bodin var afra.. .516, 517, 518,

519, 520

afra var angustifolia 516, 517. 518,

519-21

afra var ovalifolia 517

afra var tamarindifolia 517

alata Thunb 515

angustifolia E. Mey 520

brachypetala Sond 515-6, 518, 519,

522-3, 525, 526-9, 532

brachypetala var pubescens Burtt

Davy 528

capitata Bolle . . 5 1 5-6, 518, 519, 521-3,

532

cuneifo/ia Gandoger 523, 525

diversifolia Walp 523, 525

latifolia Jacq... .515-6, 518, 519, 523-6,

529, 530, 532

latifolia Jacq. X S. afra (L.) Bodin

Form A 516-7, 529, 531, 532

latifolia Jacq. X S. afra (L.) Bodin

Form B 516, 523, 530, 532

macrophylla 530

parvifolia Jacq 520-1

rogersii Burtt Davy 526, 528

semireducta Merxm 526, 528-9

speciosa Jacq 518

speciosa var ovalifolia Harv 529, 530

speciosa var tamarindifolia 518, 520

stipulata Ait 529, 530

tamarindifolia Afz. ex Sims 518, 520

tamarindifolia var forbesiana Baill . . 521-2

transvaalensis Rolfe 521-2, 528

venusta Mason 520

Schrebera Roxb 550

alata (Hochst.) Welw..55/, 552-3, 556

argyrotricha Gilg 553, 555, 556

Page

Schrebera gilgiana Lingelsh 554, 556

greenwayi Turrill 556

latialata Gilg 551-3

mazoensis Sp. Moore 553-4

merkeri Lingelsh 553

nyassae Lingelsh 553

obliquifoliolata Gilg 553

saundersiae Harv 551-3

schinoides L 652, 653

swietenioides Roxb 551

tomentella (Welw.) Gilg 556

trichocalda Welw 551

Sclerochiton 539

Sclerochiton harveyanus Nees 539

ilicifolius A. Meeuse 535-9

Sclerotium crocorum Pers 8

oryzae 351

Scolecotrichum graminis Fuckel 192

SCROPHULARIACEAE 235

Sebacina Tul 5, 10, 251, 284-5, 287, 290-1

Sebacina africana Burt. 10

epigaea (B. & Br.) B. & G 10, 272

fugacissima 267

globospora 261

megaspora 262

obscura G. W. Mart 302

sublilacina 284, 287

Seddera Hochst 643, 662

Seddera capensis (E. Mey. ex Choisy) Hall. f. 662,

663

Seddera mucronata (Engl.) Hall, f 663

schizantha Hall. f. 665

suffruticosa (Schinz) Hall. f. 662, 663

virgata Hochst. et Steud 662

Seismosarca 289, 291

Senecio brevidentatus M. D. Henderson. . . 425

decurrens DC 426

digitatifolius DC 426

macroalatus M. D. Henderson. . . . 426

tugelensis Wood & Evans 425

Septobasidium Pat 5, 8, 63, 250, 252, 259

Septobasidium bagliettoanum (Fr.) Bres. . . 9, 10

266

bogoriense Pat 9, 10

burtii 262

carestianum Bres. var natalense

Couch 9, 10

compaction Boedijn 7

curtisii (B. & D.) Boedijn & Steinm. 9, 10

grandispinosum Couch 9, 10

griseopurpureum Couch 9

mompa (Tanaka) Rac 7, 10

natalense Couch 9

philippinense Couch 9, 10

protractum Syd 9

pseudopedicellatum Burt 9

schweinitzii Burt 9

Septonema spilomeum Berk 185, 186-7, 197

toruloides Berl 190

Setaria 365

Setaria italica 494

viridis 494

Shutereia Choisy 698

Shutereia bicolor (Vahl.) Choisy 698

sublobata (Linn, f.) House 698

Shuteria W. et A 698

19

Page

Sirobasidium Lagerh. & Pat 302

Sirobasidium brefeldianum 262

Skinneria Choisy non Forst 701

Slinger, J. Morphology and Anatomy of

Utricularia transrugosa Stapf 385

Solanum tuberosum 457

Solenia Pers. ex Fr 465, 480

Solenia Candida Pers 480-1

minima Cooke & Phil 480, 482, 485

natalensis W. B. Cooke. 480-1, 482, 485

rhoina W. B. Cooke 481, 482

Sophora galioides Berg 164

genistoides L 166

henistoides Thb. var 3, 164

Sorghum halipense 359, 360, 361

sudanense 360—1

verticilliflorum 359-61

vulgare 360

Sorodiscus radicicolus I. Cook 455

Sorolpidium betae Nemec 459

Sorosphaera junci Schwartz 458-9

Sphaeria pezizula B. & C 187

Sphaeronema adiposum Butler 490

Spiranthera Boj. non St. Hil 701

Sporobolus 367

Sporobolus capensis 367

fimbriatus 367

pyramidalis 367

Stapelia getleffi Pott 416

gigantea N. E. Br 416

STAPELIEAE 541

Steccherinum ochraceum (Pers.) S. F. Gray. 55

Stereum Pers. ex S. F. Gray 5, 37, 279, 283,

290, 297, 303-4

Stereum (The genus in S. Africa) 303

Stereum acerinum (Pers. ex Fr.) Fr 28, 466

acerinum var longisporum H. & L. 26

adnatum Lloyd.. 38, 39, 45, 304, 325-6

affiine Lev. .289, 304, 316, 330-1, 334

albobadium (Schw. ex Fr.) Fr. . .45, 284,

305

albocinctum B. & Br 52, 53, 105

amoenum Kalchbr. & MacOwan . . 305,

316-7, 319

amoenum Lev 316

atrocinereum (Massee) v. d. Byl. . . 43^45,

306, 315-6

australe Lloyd 306, 314, 320, 327,

328, 330-1, 336

ayresii 284

balsameum Peck 37

balsameum f. reflexion Peck 37

bellum (Kunze) Sacc 307, 315, 331,

334

bellum sensu Bresadola 308

bicolor (Pers. ex Fr.) Fr.37, 39, 40, 45,

62, 94, 275, 289, 290,

308, 316, 319, 332, 338

bresadoleanum Lloyd 308, 309, 318

caperatum (Berk. & Mont.)

Massee 309, 328

caperatum Lloyd 309, 328, 340

cinerascens (Schw.) Massee 37, 40, 41,

57-9, 95, 287, 309, 329,

331, 337, 339-43, 345-6

cinereum Lev 310

coffeatum B. & C 39, 308

Stereum concolor (Jungh.) Sacc. .. .310, 319, 322

crassum Fr 42

crispum Quel 37

cyphelloides B. & C 311, 331, 334

diaphanum (Schw.) Cooke 289, 290,

311, 324, 330, 333

dissitum Berk 41, 343

durbanense v. d. Byl 312, 313, 326,

327, 328, 331, 335

duriusculum B. & Br 45, 51-3, 289,

313 332 338

elegans Mey 45, 313-4, 321, 328

fasciatum (Schw.) Fr.45, 50, 307, 314,

319, 320, 322

friesii Lev. . . 306, 308, 314-5, 331, 335

fulvum (Lev.) Sacc 306, 315-6, 320,

325 327 331-2 337

fuscum (Schrad.) Quel. .39, 45, 308-9,

316

glabrescens B. & C 305, 316

hirsutum (Willd.) Pers. ex S. F.

Gray. .. .279-80, 293, 303-6, 316-7,

319, 322, 327, 329, 331, 335

hirsutum forma kalchbrenneri 317

induratum Berk 52, 53, 105

involutum (Klotzsch) Fr 287, 289,

308-9, 317-8, 322, 331, 334

kalchbrenneri Sacc. .303, 305, 317, 319

karstenii 287-8

laxum Lloyd 45, 308, 309, 319

lobatum (Kunze ex Fr.) Fr 310, 314,

319-20, 322, 330-1, 335

lobatum var cinereum Lloyd ex

Doidge 306, 320

lugubris Cooke 29

luteobadium Fr 45, 314, 320

membranaceum Fr..43-5, 316, 320, 329

midas 284

moricolum Berk 40, 343

murraii (B. & C.) Burt 279-80, 320,

321, 323, 331, 336

murrayi Auct 316

neglect um Peck 41

nitidulum Berk 314, 321, 325,

328, 330, 333

notatum B. & Br 322

ochraceoflavum Schw. ex Peck 322

ostrea (Blume & Nees) Fr 322

pannosum Cooke 39, 40, 308

papyrinum Mont 43, 316, 329

percome B. & Br 45, 47, 322

perlatum Berk 322

proximum Lloyd 318, 322

pruinatum B. & C 45, 322-3

pulverulentum Lev 323

pulverulentum Peck 323

purpurascens Lloyd 41

purpureum (Pers. ex Fr.) Fr 279-80,

284-5, 287, 323-4, 330-1, 337, 470

pusillum Berk 312, 324, 328

radicans (Berk.) Burt 324-5, 338

ravenelii B. & C.. 321-2, 325, 330, 333

retirugum Cooke 43, 315, 325

rigens Karst 3 8

rimosum Berk 39, 45, 313, 325

1339563-3

20

Page

Stereum rimosum var africanum Talbot ... 37, 38 >

39, 45, 304, 313,

325-6, 328, 331, 336

rubiginosum Fr 45, 50, 326

rugosiusculum B. & C 285, 323—4

rugosum (Pers.) Fr 313, 326

sanguinolentum (A. & S. ex Fr.)

Fr 37, 38, 93, 326-7,

331, 336

schomburgkii Berk 19, 37, 40, 43,

44-5, 47, 58, 97-8, 275, 293,

295, 306, 315-6, 320, 325, 327

spadiceum Fr 279, 317, 327

subpileatum Berk 312-3, 327

tabacinum Sow. ex Fr. var.

australis Mont 327

tenebrosum Lloyd 306, 327

thozetii Berk 45, 289, 313^4, 321,

324, 327-8, 330, 333

tomentosum v. d. Byl 312-3, 328

transvaalium v. d. Byl 306, 326, 328

ti/rgidum Lloyd 309, 310, 328,

339-43, 345-6

umbrinum B. & C. .25, 37, 41, 43, 47,

96, 288, 291, 293, 295, 316,

329, 330, 331-2, 337

vellereum Berk 306, 3 1 7, 329-30

versicolor (Swartz ex Fr.) Fr. . .319, 330

villosutn Lev 330

vitile Fr 330

vorticosum Fr 324, 330

Sterigmata (A Note on) 301

Steyaert, R. L. Concerning Some S. African

Pestalotiopsis 379

Stictocardia Hall. f. 644, 772

Stictocardia beraviensis (Vatke) Hall, f 773

tiliifolia (Desr.) Hall. f. 772

woodii (N. E. Br.) Hall, f 773

Stigmatolemma incanum 479, 481

Stigmatolemma Kalchbr 465, 479, 481

Stigmatolemma incanum Kalchbr 479-80, 481

Story, R. et alia: Newly Described species. 415

Strelitzia 480

Streptandra Hall. f. (Sect, of Merremia). . . 701

Streptocarpus kentaniensis Britten & Story 433-6

lutea 433

meyeri Burtt 433

Stromatoscypha Donk 480

Stypinella purpurea Schroet 8

Stysanus 192

Stysanus ulmariae McWeeney 493

Talbot P. H. B. Cyphelloid Fungi of S.

Africa 465

Talbot, P. H. B. Entorrhiza calospora sp.

nov. and some other parasitic fungi in

Limeum roots 453

Talbot, P. H. B. Micromorphology of the

lower Hymenomycetes 249

Talbot, P. H. B. New and Interesting

records of S. African fungi 183, 489

Talbot, P. H. B. On the genus Lopharia

Kalchbr. & MacOwan 339

Talbot, P. H. B. The genus Stereum in

S. Africa 303

Talbot, P. H. B. Studies of some S. African

resupinate Hymenomycetes 1

Tavaresia 416

Tavaresia angolensis Welw 415-6

meintjesii R. A. Dyer 415, 447

Thanatophyton crocorum Nees 8

Thelephora Erh. ex Fr 5, 62

Thelephora acerina (Pers.) Pers. . . .27-8, 62, 466

albido-carnea Schw 54

atro-coerulea Trog 15

atropurpurascens B. & Br 25

bellum Kunze 307

beyrichii Fr 40

bicolor Pers 39, 308

biennis Fr 62, 308-9, 316

caerulea Schrad 15

cinerascens Schw 40, 309, 343

confluens Fr 18

corium Pers 30

crassa Lev 42

diaphana Schw 311

disciformis DC. ex Fr 468

fimbriata Roth 15

flocculenta Fr 476

fulva Lev 315-6

fusca Schrad 39, 40

fuscoviolascens Mont 62

gigantea Fr 24

hirsuta Willd 316

hirsuta B. Pers 37

incarnata var. B. Pers 30

indigo Schw 15

insinuans Schw 53

involuta Klotzsch 318

lobata Kunze 319

mesenterica (Dicks.) Pers 6

murraii B. & C 320

olivacea Fr 36

pallescens Schw 53

pannosa Sow. ex Fr 39

pedicellata Schw 63

pergamenea Pers 24

pulverulenta Lev 323

punicea A. & S 63

purpurea Pers 323

radicans Berk 324

rosea Pers. ex Fr 470

sanguinolenta A. & S. ex Fr. . .37, 326

sericea B. Pers 37

vinosa Pers 39

Thelepora cretacea Fr 63. 116

grisea Rick 63

Theleporus cretaceus Fr 63

Themeda triandra 437, 440

Theodora Medik 516

Theodora capitata (Bolle) Taub 521

speciosa Medik 518

Tlnvaitesiella Massee 339, 340

Thwaitesiella mirabilis (B. & Br.) Massee. . 56,

339, 340

TILLETIACEAE 453

Tilletiopsis 272

Tomentella Pat 5, 63, 271

Tomentella echinospora 271

eylesii (v. d. Byl) 63

punicea (A. & S.) 63

Torenia 235

Torrubiella ochracea 495

Torula herbarum (Pers.) Link ex Fr 189, 200

21

Page

Trametes meyenii 275, 276

serialis 272

serpens 274, 275

Trechispora 271

Trechispora raduloides 262

Tremella 250, 252-5, 259, 264, 301-2

Tremella epigaea B. & Br 10

Trianthema pentandra 453-4

Trichoderma lignorum 188

viride Pers. ex Fr 188, 199

Trichosporium curtisii Massee 25

sphaericum Sacc 189

Trichothecium griseum Cooke 495

Trichurus 192

Trichurus gorgonifer Bain 192, 203

Triticum 359-60, 368, 370-1

Triticum vulgare 356

Truncatella Steyaert 382

Truncatella conorum-piceae (Tubeuf) Steyaert 382

ramulosa (v. Beyma) Steyaert.. 382

Tuber croci Duby 8

Tulasnella 250-1, 253, 256, 257-9,

261, 291, 301, 302

Tulasnella tulasnei 255-6

violacea 262

violea 267

Turbina Rafin 644, 773

Turbina corymbosa (L.) Rafin 773

curtoi (Rendle) A. Meeuse 784

holubii (Baker) A. Meeuse 775, 780

oblongata (E. Mey. ex Choisy) A.

Meeuse 775, 778

oenotheroides (Linn, f.) A. Meeuse 775

pyramidalis (Hall, f.) A. Meeuse. . 784

robertsiana (Rendle) A. Meeuse. 775, 777

shirensis (Oliv.) A. Meeuse 775, 782

stenosiphon (Hall, f.) A. Meeuse. 775, 783

suffruticosa (Burch.) A. Meeuse. 775, 776

Urginea 545

Urginea alooides Bolus 544

Urochloa helopus 366

Ustilago leucoderma Berk 188

Utricularia 385

Utricularia (distribution in Africa) 386

Utricularia andongensis Welw 387

baumii Kam 387

brachyceras 404-5

bryophila Ridley 385, 387

capensis Spreng. .385, 387, 404, 405

charoides Stapf 387

cymbantha Oliv 385, 387

diploglossa Welw 387

ecklonii Spreng. .385, 387, 404, 405

exilis Oliv 387, 405

exoleta R. Br 385, 387

firmula Welw 387

foliosa Linn 385, 387

globulariaefolia 403

kirkii Stapf 385, 387, 400,

404, 405, 406

linarioides Welw 387

lividaE. Mey 385, 387, 388,

404, 405

livida var iransrugosa Stapf. .388, 405

manii Oliv 387

micropetala Smith 387

obtusa Swartz 385, 387

Page

Utricularia odontospernta Stapf 387

papillosa Stapf 387

platyptera Stapf 387

prehensilis E. Mey 387

reflexa Oliv 387

rigida Benj 387

sandersonii Oliv 385, 387

sanguinea Oliv 387

schweinfurthii Bak 387

spiralis Smith 387

stellaris Linn 385, 387, 403

striatula Smith 385, 387

subulata Linn 385, 387

thoningii Schumach 387

tortilis Welw 387

transrugosa Stapf (Morphology

& Anatomy) 385

transrugosa Stapf 385, 387-406

tribracteata Hochst 387

trichoschiza Stapf 387

villulosa Stapf 387

volubilis 403

welwitschii Oliv 387

van der Westhuizen, G. C. A. Three species

of Curvularia from Finns 501

van Druten, D., et alia: New and Interesting

Records of S. African Flowering Plants . . 535

Vararia Karst 54, 284, 313

Vararia pallescens (Schw.) Rogers & Jack- ’

son 53

Veluticeps 276-7

Veluticeps heimii 277

VERBENACEAE " 237

Verdoorn, I. C. Nomenclature of the Cape

Acacia 409

Verdoorn, I. C. Oleaceae of Southern

Africa 549

Verdoorn, I. C. S. African Species of Acacia

with glandular glutinous pods 153

Verdoorn, I. C. & R. A. Dyer: Encephalar-

tos natalensis 205

Verdoorn, I. C. et alia: Newly described

species 213, 415

Vernonia polygalaefolia Licht. ex Less ’ 220

VITACEAE 545

Vitex 237

Vitex (subgen. Euvitex) 237

Vitex amboniensis Giirke 237-8

madiensis Oliv 237-8

patula E. A. Bruce ! 237

Vitis 471

Vitis natalitia Szysz 545

Vuilleminia 25 i , 254, 256

Vuilleminia comedens 262, 271

Wild, H. et alia: Newly described species. . 415

Xanthips (Griseb.) Hall. f. (sect, of Mer-

remia) 70 1

Xanthochrous Pat 291

Xenolachne D. P. Rogers ' ' . 255, 302

Xenospoiella larvalis (Morgan) Linder. .491, 497

rosea Talbot 491, 497

thaxteri Linder 491

Zeamays 189, 190, 191, 360-1

Zingiber mioga 494

officinalis 494

Zizaniinae 121

Zygnema !!!!!!!! 388

Studies of some Soutli African Resupinate

Hymenomycetes. *

By

P. H. B. Talbot.

Resupinate Hymenomycetes do not form a closely integrated group, and it must

be borne in mind that the term “ resupinate ” is ambiguous, being used both morpholo-

gically and as a colloquial name for the large, non-taxonomic group of lower Basidio-

mycetes whose members are characterised by a fructification lying flat on the sub-

stratum with the hymenium on the outer side. While the taxonomic work which

follows has been limited to South African species, fungi from other regions are men-

tioned in comparison.

Collectors in South Africa, with the notable exception of Mr. W. G. Rump of the

Natal Museum, have paid little attention to resupinate Basidiomycetes, as the collec-

tions cited in the subsequent pages will testify. These fungi are seldom of striking

appearance, which may be one reason why they were so poorly represented in the

famous early collections. The data in Table I are presented in illustration of this

point. The table also summarises the more important early literature references to

South African fungi. These records mostly lack descriptions, or if descriptions are

given then they are generally totally inadequate by present standards.

TABLE 1.

Table indicating how very few Resupinate Hymenomycetes were listed or described from collections

of fungi made by famous early collectors in South Africa.

Even in recent years the tendency has been to overlook resupinates. For example,

Phillips (1931) lists 127 species of fungi from Knysna forests, of which only two are

resupinate Hymenomycetes. None at all are found in Miss Pegler’s collection of

seventy-nine species from Kentani (Pole Evans and Bottomley, 1917). Miss Wakefield’s

* Most of this work was prepared while the author occupied the post of South African Assistant

at the Herbarium of the Royal Botanic Gardens, Kew, and he is deeply indebted to the several authorities

responsible for this arrangement. At Kew, he was singularly fortunate in receiving the guidance of

Miss E. M. Wakefield and Dr. B. Barnes of Chelsea Polytechnic, whose interest and unstinted help is

gratefully acknowledged.

2

examination of Hoeg’s collections of African fungi resulted in the listing of five resu-

pinates among a total of fifty-five species (Wakefield, 1936). Simpson and Talbot (1946)

list only five resupinates out of a total of fifty-four species from Qudeni forest.

Without adequate collections, and facilities for comparison of material with types

in overseas herbaria, taxonomic work in this group has never progressed beyond the

pioneer works of van der Byl (1929, 1934) where about forty species are described.

The present paper is an attempt to extend and modify van der Byl’s work on the basis

of new collections which have accumulated in the herbaria. However, there are still

insufficient good collections of certain species, for this country is strangely lacking in

those amateur naturalists on whom so much has depended in other countries for the

advance of mycology.

DISTRIBUTION, NATURAL HABITATS, AND SUBSTRATA.

The relative paucity of collections makes it impossible to obtain a clear picture

of the distribution of the different species within South Africa, but it is likely that they

are widespread throughout all parts of the country where the same type of natural

habitat occurs. The physical configuration of the Union is responsible for a great

variety of natural habitats and conditions of localised occurrence, only a few of which

are suitable for the growth and development of resupinate Hymenomycetes. Un-

doubtedly the optimum conditions of moisture, temperature and rotting vegetation

are found in the indigenous forests and bush, and to a lesser extent in the artificial

plantations of Eucalyptus spp. or Acacia mollissima and A. decurrens. In the Thorn-

veld, subjected to periodic hot winds from the north, a limited water supply and recurring

grass fires which prevent the tree formation from closing in so that moist, shady condi-

tions do not obtain (Bews, 1912 p. 309; 1913, p. 497), resupinates are rare except near

stream banks. The ravages of termites on fallen sticks and branches in the Thornveld

without doubt limits the amount of rotting vegetation available for the support of

saprophytes.

It is suggested that the same species of fungi are likely to be widespread in scattered

indigenous forest formations which owe their existence to the Great Escarpment (A. W.

Rogers, 1922) of South Africa. Such forests occur chiefly in the botanical region known

as the “ Steppe and Forest Province ” (Pole Evans, 1922), that is, on the south-east

slopes of the escarpment and subsidiary mountains facing moisture-laden winds from

the Indian Ocean, and sheltered from the hot north winds. Forest formations dependent

in this way on topography are fully discussed by Bews (1912, 1913), Henkel et alia

(1936), and Phillips (1931), who conclude that their phanerogamic constitution varies

only slightly in different localities. The coastal forests of Natal are closely allied to

tropical East African bush (Bews), 1912, while those of Knysna further south (Phillips,

1931) are composed of species derived from tropical ones; they are, however, climati-

cally and constitutionally more temperate in nature than those of Natal, the Eastern

Cape and the Transvaal.

Marloth (1903, 1905) has shown how effectively the vegetation is able to precipitate

the moisture carried by the S.E. winds at the Cape, in the absence of actual rain. On

the other hand, Bayer and Coutts (1938) draw attention to the remarkably frequent

occurrence of very low relative humidity in the Thornveld (less than 50 per cent, at

midday on 174 days in the year, and as low as 7 per cent, at 8.30 a.m. when the “ Berg ”

winds are blowing); this in itself would almost exclude saprophytic fungi from flourishing

in such a habitat at most seasons.

The above ecological evidence supports the suggestion that the same species of

resupinate fungi might well be widespread in all natural forests in the Union, and that

our fungus flora is likely to include many elements derived from, or at least common to,

the more central African territories. Here an analysis of the distribution of forty-five

3

species, described in this paper, is of interest. The analysis is based on collections

which the author examined in Kew Herbarium. Although the number is small it at

least forms a random sample of the known South African resupinates. Of the forty-

five species-

13 are recorded only from South Africa;

5 are truly cosmopolitan ;

7 are of mixed tropical and temperate affinity, but not fully cosmopolitan;

9 are associated only with Europe and/or North America;

1 1 are associated only with the tropics of the Far East, Africa or America.

These figures will change with the inclusion of new records in the future. The

likelihood is that the fungi now appearing as peculiar to South Africa will eventually

be linked with tropical ones rather than temperate species, for until now the tropics,

especially in Africa, have received less attention from collectors than the temperate

regions. It cannot be too strongly emphasised that resupinate Hymenomycetes are

far more abundant in South Africa than might be supposed from a list of the species

described in this paper. But much work is necessary before the value of new collections

which the author is making can be assessed.

A highly specialised habitat is the gold mines of the Transvaal. They are rich in

fungi growing on underground timbers and textiles (Brown, 1937), but such fungi have

largely been excluded from this study. They frequently vary astonishingly from the

normal in this unnatural habitat, and their place of origin is often uncertain, for much

timber is imported.

It is seldom possible to record in this paper the substratum with which the described

fungi are associated. This aspect of collecting has almost always been neglected, and

it is not always possible to be sure of the identity of a fallen stick or branch lying under

such a mixed canopy as is usually found in our forests.

MATERIALS AND METHODS OF EXAMINATION

Most material used in these studies consisted of dried herbarium specimens.

Wherever possible, these were implemented by the author’s own collections of fresh

material of the same species, and by specimens from other sources in Kew Herbarium.

But in describing and citing collections, reference is made only to South African material,

unless otherwise stated.

The techniques used in microscopic work were based on those of Overholts (1929),

Martin (1934) and Burt (1914-1926; 1929). In interpreting the nature of the various

accessory organs and minute hyphal characters, the work of Overholts (1929) and

Pilat (1926) have been closely followed, and their definitions apply here. The terms

“ generative ” and “ skeletal ” hyphae are used in the sense of Corner (1932 a; 1932 b).

The principal reagents used in mounting and staining sections or crush preparations

were: (1) 5 per cent, potassium hydroxide solution, with or without the addition of a

very small arpount of aqueous 1 per cent, phloxine solution; (2) lactophenol for species

which darken too much in KOH, or whose cystidia are soluble in KOH ; (3) Melzer’s

iodine-chloral hydrate for staining conducting vessels and for testing the amyloid

reaction of spores; (4) alcohol for removing the air bubbles from tissues; (5) sepia

solution, to give a dark-ground effect for detecting a mucilaginous investment of spores;

(6) dilute hydrochloric acid for dissolving obstructive mineral inclusions. In general

it was not attempted to make permanent preparations. Semi-permanent preparations

are made fairly satisfactorily by ringing lactophenol mounts with gold size, Noyer or

Kronig cements. These cements all have some or other disadvantage. Polyvinyl

alcohol-lactophenol is a good mountant for all but the most delicate resupinates;

4

this medium may also be used as a ringing agent for lactophenol mounts, but the ring

slowly becomes dissolved from the inside and needs renewing after about eight months.

When solution of the ring occurs, no obstructive suspension is formed in the mounting

fluid. The advantage of this medium as a ringer is that it is fluid, easy to apply, sets

clean and transparent within a few hours, and does not chip or crack.

LITERATURE CITED.

Bayer, A. W., & (1938) Morning and midday relative humidities at Pietermaritzburg, South

Coutts, J. R. H. Africa. S.A. Journ. Sci. 35 (1938) 154.

Berkeley, M. J (1843. a) An enumeration of the fungi collected by Zeyher in Uitenhage.

Hooker’s Lond. Journ. Bot. 2 (1843) 507.

(1843. b). On two Hymenomycetous fungi belonging to the Lycoperdaceous

group. Hooker’s Lond. Journ. Bot. 2 (1843) 200.

(1876). An enumeration of the fungi collected at the Cape of Good Hope

during the stay of the English Transit of Venus Expedition in 1874. Journ.

of Bot. 14 (1876) 173.

Bews, J. W (1912). The vegetation of Natal. Ann. Natal Mus. 2 (1912) 253-331.

— (1913). An oecological survey of the Midland of Natal with special reference

to the Pietermaritzburg district. Ann. Natal Mus. 2 (1913) 485-545.

Bottomley, A. M.. . . (1917). An account of the Natal fungi collected by J. Medley Wood. S.A.

Journ. Sci. 13 (1917) 424-446.

Brown, Mrs. M (R. Lurie) (1937). Mine timber preservation — Mine Fungi. S.A. Journ.

Sci. 33 (1937) 383-389.

Burt, E. A (1914-1926). The Thelephoraceae of North America. Parts 1-15 in Ann.

Mo. Bot. Gard. vols. 1-13.

— (1929). Classification of the species of Corticium by the tissues of the fructi-

fication. Proc. Internat. Congr. Plant Sci. 1926, Ithaca. 2 (1929) 1598.

Cooke, M. C (1879). Natal Fungi. Grevillea 8 (1879) 69 et. seq.

— (1881). Fungi Exotici. Grevillea 10 (1881) 123.

Corner, E. J. H (1932. a). The fruit body of Polystictus xanthopus Fries. Anns. Bot. 46 (1932)

71-111.

— (1932. b). A Fomes with two systems of hyphae. Trans. Brit. Myc. Soc. 17

(1932) 51-81.

De ThUmen (1875). Fungi Austro-Africani. Flora. 58 (1875) 378.

— (1876). Ibid. Flora. 59 (1876) 362, 423, 568.

— (1877). Ibid. Flora. 60 (1877) 407.

— (1878). Ibid. Flora. 61 (1878) 353.

Evans, I. B. Pole (1922). The Main Botanical regions of South Africa. Bot. Survey Menv

S.A. No. 4 (1922) Pretoria.

Evans, I. B. Pole, and (1917). An enumeration of the fungi collected at Kentani in the Cape Province

Bottomley, A. M. by Miss Alice Pegler, A.L.S., in 1911-1914. Ann. Bolus Herb. 2 (1917)

185-193.

Fries, E (1848). Fungi Natalenses. K. Vet. Akad. Handl. Stockholm (1848) 121-154.

Henkel, J. S., (1936). An account of the Plant Ecology of the Dukuduku Forest Reserve

Ballenden, S. and and adjoining areas of the Zululand coast belt. Ann. Natal Mus. 8 (1936)

Bayer, A. W. 95-125.

Hennings, P (1895). Fungi Austro-Africani. Hedwigia 34 (1895) 325.

(1898). Ibid. Hedwigia 37 (1898) 293.

Kalchbrenner, C. . . . (1881. a). Fungi MacOwaniani. Grevillea 9 (1881) 107.

— (1881. b). Ibid. Grevillea 10 (1881) 57-59.

(1882). Ibid. Grevillea 10 (1882) 104-106.

Kalchbrenner, C., (1880). South African Fungi. Grevillea 9 (1880) 7.

and Cooke, M. C.

(1881). Natal Fungi. Grevillea 10 (1881) 27.

Leveille, J. H (1846). Descriptions des champignons de l’herbier du M,useum de Paris.

Ann. Sci. Nat. ser. 3, 5 (1846) 111.

Marloth (1903, 1905). Results of Experiments for ascertaining the amount of moisture

deposited from the south-eastern clouds. Trans. Phil. Soc. S.A. 14 (1903)

403 and 16 (1905) 97.

Martin, G. W (1934). Three new Heterobasidiomycetes. Mycologia 26 (1934) 263.

Montagne, J. F. C. . . . (1847). Enumeratia et descriptio fungorum quos cl. Drdge in Africa, etc.

Ann. Sci. Nat. ser. 3, 7 (1847) 166.

Overholts, L. O (1929). Research methods in the taxonomy of the Hymenomycetes. Proc.

Internat. Congr. Plant Sci. 1926, Ithaca. 2 (1929) 1688-1712.

Phillips,.! F. V H931). Forest succession and ecology in the Knysna region. Bot. Survev

S.A. Mem. No. 14 (1931) 159-160.

5

Pilat, A (1926). Monographic der mitteleuropaischen Aleurodiscineen. Ann. Myc.

24 (1926) 203-230.

Rogers, A. W (1922). Physical features of South Africa. In Bot. Survey S.A. Mem. No. 4

(1922) 9-23.

Simpson, M. W., and (1946). An enumeration of fungi collected at Qudeni Forest Reserve, Zulu-

Talbot, P. H. B. land, in February, 1945. S.A. Journ. Sci. 42 (1946) 131-134.

Sydow, P (1899). Fungi Natalenses. Hedwigia 38 (1899) 130-134.

Van der Byl, P. A . . . (1929). Die Suid-Afrikaanse Thelephoraceae. Ann. Univer. Stellenbosch.

7:A: 3 (1929) 1-52.

— (1934). Die Suid-Afrikaanse Hydnaceae of Stekelswamme. Ann. Univer*

Stellenbosch. 72; A: 1 (1934) 1-9.

Wakefield, E. M (1936). African fungi collected by O. A. Hoeg, in 1929-1930. Kong.

Norske Vidensk. Selsk. Forhandl. 9 (1936) 42-53.

Winter, G (1884). Repertorium Rabenhorstii Fungi Europaei et extra-Europaei. Cent*

XXXI-XXXII. Hedwigia 23 (1884) 164.

- — (1885). Exotische Pilze. Hedwigia 24 (1885) 21.

TAXONOMIC

In essentials the classification adopted here is that of Bourdot and Galzin in

“ Hymenomycetes de France ” (1928). The genera are grouped thus: —

HETEROBASIDIOMYCETAE

Auriculariaceae (Auricularia', Helicobasidium ; Septobasidium).

Tremellaceae (Sebacina\ Heterochaete).

HOMOBASIDIOMYCETAE-APHYLLOPHORALES

Exob asidiace ae ( Exobasidium) .

POROHYDNACEAE

1) Corticieae

Corticinae ( Corticium ; Peniophora\ Punctularia; Aleurodiscus).

Merulinae ( Phlebia ; Merulius ; Coniophora).

Stereinae ( Stereum ; Duportella; Hymenochaete).

Asterostromellinae ( Asterostromella ).

2) Asterostromeae ( Asterostroma ).

3) Hydneae (Mycoleptodon; Lopharia; Acia; Grammothele; Grandinia; Odontia\ Hydnum ; Radulum).

4) Phylacterieae ( Thelephora ; Tomentella ).

5) Poreae ( Thelepora ).

Note. — In the taxonomic section, an exclamation mark (!) after a specific epithet indicates that the

type, or authentic material, of that species has been examined.

Except when otherwise stated, herbarium numbers of specimens quoted in this work refer to the

collections housed in the Mycological Herbarium of the Union Department of Agriculture in Pretoria,

designated in abbreviation as Herb. Pretoriae.

HETEROBASIDIOMYCETAE

Basidia transversely, longitudinally or vertically septate, or deeply divided, or

differentiated into probasidia or hypo- and epibasidia; producing spores which on

germination develop secondary spores from which the mycelium arises ; or if the basidium

is undivided, then the characters of spore germination still subsist.

There are relatively few heterobasidiomycetes recorded from South Africa, and of

these extremely few are resupinate in habit. This is partly due to the lack of adequate

collections, but may eventually prove to be a characteristic of the flora dependent on

the climate. Further observation can alone decide. Descriptions and notes on the

resupinate species so far recorded follow on the succeeding pages. Where there are

several species in one genus, a key to species is provided.

6

AURICULARIA Bull, ex Fries.

1. Auricularia mesenterica (Dicks, ex Fries) Fries, Epicr. (1838) 555, Hym. Eur. (1874)

646; Berkeley, Outlines of Brit. Fung. (1860) 272; Saccardo Syll. Fung. 6 (1888)

762; Lloyd, Myc. Writ. 5 (1919) 873; Bourdot and Galzin, Hym. de Fr. (1928) 15;

Rea, Brit. Basid. (1922) 727.

Helvetia mesenterica Dicks., Crypt. Fasc. 1 (1785) 20; Bolton, Hist, of Fung.

Growing about Halifax 3 (1791) Tab. 172.

Auricularia corrugata Sowerby, Eng. Fung. 3 (1803) Tab. 290.

Thelephora mesenterica (Dicks.) Pers., Syn. Fung. (1801) 571.

Ph/ebia mesenterica (Dicks, ex Fr.) Fries, Elench. 1 (1828) 154.

Auricularia tremelloides Bulliard, Champ, de la Fr. (? 1780) Tab. 290; Quelet,

FI. Myc. de Fr. (1888) 24.

Illustrations'. Lloyd Myc. Writ. 5 (1919) f. 1490; Rolland, Atlas des Champ, de Fr.

(1910) 92, PI. 106, f. 242; Hussey, Illustr. Brit. Myc. 2 (1855) PI. 6; Ramsbottom,

Handbk. of Larger Brit. Fungi (1923) 156, f. 95.

Plate 1.

Effused over a wide area, resupinate with narrow reflexed margin, thick, gelatinous-

coriaceous, in section up to 1,000 g thick excluding the tomentum; drying thin and

horny, brittle. Hymenium covered by reticulate, plicate ridges up to 1 mm. in height,

light reddish brown when fresh, drying livid purplish (dark vinaceous drab to deep

quaker drab — Ridgway) with a greyish pruina. Margin thickly tomentose on abhy-

menial surface, imperfectly zoned greyish and brownish.

Basidia: 4-6-6 X 60-73 g, divided by three transverse septa into four cells, each capable

of bearing a lateral sterigma, 2-2-5 X 21 g.

Spores: hyaline, smooth, cylindric-curved, or ellipsoid with one side depressed, 10-15

X 4-5-5 g.

Paraphyses: filamentous, simple or branched, separating the basidia.

Hyphae: gelatinised, 1-2 g diam., loosely woven in centre of trama but compacted

into a darker bordering layer adjacent to the tomentum. Probably only the

lumen is seen.

Hairs of the abhymenium : pale coloured, 2-3 g diam., thickly intertwined, with a very

narrow lumen.

Mineral inclusions: found in the hymenium between and beneath the basidia.

Specimens examined: 28495, Rump (251), Town Bush, Maritzburg, Oct. 1934.

The specimen on which this description is based occupied an area of 7 x 33 inches

on an old stump, and the reflexed portion was in most parts only a few millimetres

wide. Most European specimens have a widely reflexed part on which the zonation

of the hairy surface is quite distinct. Probably when more specimens are found in

South Africa it will be discovered that they are not always almost entirely resupinate.

The spore size in this specimen is at the lower end of the range recorded for France by

Bourdot and Galzin.

2. Auricularia mesenterica var. lobata (Sommerf.) Quel., Sommerf. in Mag. Nat. Vidensk.

(1827); Fries, Elench. 2 (1828) 34, Hym. Eur. (1874) 646; Berkeley, Outl. of Brit.

Fung. (1860) 272, PI. 18, f. 1; Bourdot and Galzin, Hym. de Fr. (1928) 15; Rea,

Brit. Basid. (1922) 727; v. d. Byl in Ann. Univ. Stellenbosch 1: A: 3 (1923) 10, f. 7.

Differs from the species A. mesenterica by the possession of a lobed margin to the

pileus, and zones which are glabrous, velutinous or hispid, v. d. Byl (l.c.) records

A. lobata from the Victoria Falls and I have also seen a specimen, 23373, Eyles (1312),

Apr. 1981, from the same locality in Rhodesia. This specimen was small and sterile,

not resupinate, showing a distinctly lobate margin, but otherwise would be taken for

A. mesenterica with a rather browner hymenium than usual. In practice is is difficult

to maintain A. lobata even as a variety.

7

HELICOBASIDIUM Patouillard.

1. Helicobasidium compactum (Boedijn) Boedijn ! in Archief Theecultuur 4 (1930) 41;

Boedijn and Steinmann in Bull. Jard. Bot. Buitenzorg ser. 3, 11 (1931) 169.

Septobasidium compactum Boedijn in Comm. Gen. Expt. St. A.V.R.O.S. 26

(1926) 7.

Helicobasidium mompa Tanaka forma macrosporum Hara in Journ. Seric. Assoc.

Japan, Tokyo, 6 (1917) 725. [non Septobasidium mompa (Tanaka) Rac.]

Helicobasidium purpureum (Tul.) Pat. var. orientate Pat. in Bull. Soc. Myc. de Fr.

36 (1920) 176.

Above Synonymy after Boedijn.

Illustrations'. Boedijn (1930 and 1931, l.c.).

Plate 2.

Resupinate, investing plant parts or soil with a thick, soft, spongy cushion, fawn

to chocolate coloured with violet tints, up to 2 cms. in thickness, or occasionally rather

thin. Hymenium smooth, dry, membranous, pruinose with spores, coloured fawn to

violaceous-fawn. Context concolorous, spongy. Mycelial cords frequently creep

over the substratum, according to Boedijn.

Basidia: arising direct from hyphae, without probasidia; cylindrical, at first straight,

later curling over like a crozier and 1-4 celled, hyaline, 40-80 x 5-6 g. (Sterig-

mata 2, lateral from the basidial cells, when mature 1-2 x 25 g, fide Boedijn.)

Spores: hyaline, smooth, elliptic with one end round and the basal end attenuate and

often abruptly bent at the attachment, 4-5-5 x 18-23 g.

Hyphae: subhymenial hyphae hyaline or very dilutely coloured, much bent and twisted,

profusely branched; tramal hyphae brown, thin walled, lacking clamps, much

branched, septate, 4-6-(7) g wide, chiefly vertically arranged, sometimes forming

zones of different densities. Fine granular mineral matter is often scattered among

the hyphae.

Specimens examined: On Pinus luchuensis, E. M. Laughton, Bergplaats, George,

C. P., Aug. 1937 (in Herb. I.M.I.); on Pinus longifolia, 27347 and 27346,

D. F.O., Timbadola, Louis Trichardt, Oct. 1933; on Pinus longifolia, 26855,

D.F.O., Hangklip, Louis Trichardt, Oct. 1931 and Sept. 1932; On Acacia

me/anoxylon, 9049, D.F.O., Katberg Main Forest, 1/6/1915 (this specimen is

sterile but otherwise agrees with H. compactum ); on Pinus longifolia, 27345,

Forester, Hangklip, Louis Trichardt, Jan. 1934; on Pinus sp. 27701, Conserva-

tor of Forests, Hangklip, Louis Trichardt, Sept. 1934; on base of dying

Pinus longifolia, 26329, Forester, Hangklip, Louis Trichardt, Oct. 1930;

on Pinus insignis, 21054, D.F.O., Karatara, Knysna, 13/12/1926 (sterile);

on Acacia melanoxy/on, 9154, J. D. Keet, Katberg (basidia present, but no

spores; as Corticium laetum in Herb. Pretoriae); on Camellia theifolia, A.C.

Tunstall, Assam, India, det. K. B. Boedijn, 22/4/1926.

The pathogenicity of this species on Pinus trees in the Transvaal is discussed by

Miss Bottomley [in S.A. Jour. Sci. 33 (1936) 374-5]. A note on one of the herbarium

sheets, probably copied from a letter from Miss Wakefield, states: “The very thick

spongy texture and the stratified hymenium are unusual in the genus Helicobasidium

and in this respect, as well as in spore characters, the fungus agrees exactly with Helico-

basidium compactum Boedijn, which is the cause of a serious root disease of coffee in the

Dutch East Indies, and has been recorded on various other plants in that region. The

species of true Helicobasidium are all suspect as root parasites, and this case is therefore

particularly interesting .... The fruiting stage can usually be recognised by the

presence of a pale lilac ‘ bloom ’ on the surface, due to the abundance of spores. The

species is very close to H. longisporum described from roots of cacao from Uganda

(Kew Bull., 1917). H. longisporum forms a delicate, pulverulent layer, and the spores

are more deeply coloured and average longer, 20-26-(30)

Other species of Helicobasidium with lilac, vinous or purple hymenia are: H. pur-

pureum (Tul.) Pat. [spores reniform, 10— 12— (1 5) X 6-7 /i], H. longisporum Wakef. !

(spores purple, 25-26 X 4-5 p) and H. mompa Tanaka (spores 10-12 X 5-7 p).

2. Helicobasidium purpureum (Tulasne) Patouillard in Bull. Soc. Bot. Fr. 32 (1885)

171; Saccardo, Syll. Fung. 6 (1888) 666; Boudot and Galzin in Bull. Soc. Myc. de

Fr. 25 (1909) 17, Hym. de Fr. (1928) 9; Buddin and Wakefield in Trans Brit. Myc.

Soc. 12 (1927) 116-140, cum icones; Patouillard in Bull. Soc. Bot. Fr. 33 (1886)

335-337, Essai Taxon, sur les Hym. (1900) 12; Ware in Trans. Brit. Myc. Soc. 14

(1929) 94; Sampson and Western, Diseases of Brit. Grasses and Herbage Legumes

(1941) 42.

Hypochnus purpureus Tulasne in Ann. Sci. Nat. ser. 5, 4 (1865) 295, in Journ.

Linn. Soc. Bot. 13 (1871) 37 and Ann. Sci. Nat. ser. 5, 15 (1872), 227.

Helicobasis purpureus (Tul.) Clements and Shear, Genera of Fungi (1931) 341.

Rhizoctonia crocorum (Pers.) DC. — Duggar in Ann. Mo. Bot. Gard. 2 (1915)

403; Buddin and Wakefield in Trans. Brit. Myc. Soc. 12 (1927) 116-140, in

Ann. Appl. Biol. 11 (1924) 292, in Trans. Brit. Myc. Soc. 14 (1929) 97; Ware

in Trans. Brit. Myc. Soc. 14 (1929) 94; Watson in Trans. Brit. Myc. Soc. 14

(1929) 95; Doidge and Bottomley, Revised List of Plant Diseases occurring in

South Africa (1931) 35.

Rhizoctonia medicaginis DC. — Eriksson, Die Pilzkrankheiten der Landwirt-

schaftlichen Kulturgewasche 1 (1926) 197; van der Byl, Plantsiektes, hul oorsaak

en bestryding (1928) 343.

Rhizoctonia violacea Tulasne, Fungi Hypogaei pg. 188; Saccardo, Syll. Fung. 14

(1899) 1175; van der Lek, Meded. Rijks Hoogere Land-, Tuin-, en Boschbouw-

school, Wageningen 12 (1917) 94; Ware in Trans. Brit. Myc. Soc. 14 (1929) 94;

Board of Agric. and Fisheries Leaflet 171, London (1906); Eriksson, Fungoid

diseases of Agric. Plants (Transl.) London (1912); Eriksson, Die Pilzkrank-

heiten der Landwirtschaftlichen Kulturgewasche 1 (1926) 243.

Thanatophyton crocorum Nees; Tuber croci Duby, Sclerotium crocorum Pers.,

Stypinella purpurea Schroeter (fide Saccardo).

Illustrations : Buddin and Wakefield (1927) loc. cit. ; See list in Saccardo, Syll.

Fung. 20 (1911) 678.

Only the Rhizoctonia stage of this species has so far been found in South Africa.

It occurs as a mycelial mat, byssoid to compressed, covering the base of stems, roots of

small herbs, and potato tubers. It is a distinctive violet colour, often drying to a cin-

namon-drab colour, and forming mycelial strands.

Hyphae: (in KOH) reddish-brown with slightly darker thin walls, very even in diameter

over long distances, regularly septate, branching almost at right angles, lacking

clamps. When old, the cells have thicker walls. Short, swollen cells associated

with sclerotia do not occur in the Rhizoctonia stage. The hyphae are (2 - 5)— 5— ( 11)

p wide.

In colour and occurrence this forms a very distinctive sterile mycelium.

Specimens examined: As R. crocorum : 35572, Baragwanath, on Solarium tuberosum,

Tzaneen, Transvaal, 2/7/1946; 31833, G/asse (N.H.584), on Medicago,

n’Karini, Weenen, Natal, 25/4/191 7. As R. violacea: 2540, G/asse, on

Medicago sativa, Weenen, Natal, 9/7/1912.

SEPTOBASIDIUM Patouillard.

The genus Septobasidium is well represented in the Union, but it is felt that it lies

somewhat outside the scope of this paper. However, as the author has worked through

all its representatives in the Pretoria Herbarium, he appends the following key to species

for the purposes of identification. He has found it impossible to improve on the key,

9

descriptions and figures of these species given by J. N. Couch in his monograph “ The

genus Septobasidium,'" 1938, and therefore quotes this work except in the citation of

some of the literature references and herbarium numbers of the specimens which he has

examined and identified.

I. Basidia 2-celled, with persisting probasidium.

A. — Context with distinct pillars, 2-3 mm. thick, blackish; surface grey, much cracked, texture

firm and hard throughout.

1 . S. protractum.

B. — Context without pillars; surface smooth and glabrous in places, colour blueish-grey.

2. S. griseopurpureum.

II. Basidia 4-celled, curved, without persisting probasidium.

Context with distinct pillars: with two horizontal layers, subiculum and top layer, without a

winged margin.

A. — Surface smooth or minutely warted, velvety, colour near benzo-brown, usually with a

purplish tint; margin with tent-like structures.

3. 5. Schweinitzii.

B. — Surface smooth at first, becoming slightly wrinkled or cracked, clay colour to drab.

4. S. natalense.

III. Basidia 4-celled. straight, without persisting probasidia.

A. — Context with pillars, usually tall and distinct. Surface smooth, glabrous, buff-coloured ;

pillars dark brown, unbranched, subiculum whitish.

5. S. pseudopedicellatum.

B. — Context with pillars, usually short and stubby.

(a) Surface with mottled brownish and whitish areas, with large conical spines,

otherwise smooth; probasidia irregular.

6. S. grandispinosum.

( b ) Surface smooth but top layer incompletely formed, leaving pinholes and cracks;

cream colour to cinnamon brown; probasidia pyriform, borne in clusters (S.

earestianum). With surface even less compact, and basidia and spores larger.

7. S. Carestianum var. natalense.

(c) Surface nearly smooth or with holes, cracks or spines, or mound-like insect

houses; purplish-black throughout.

8. S. Curtisii.

IV. Basidia 4-celled, curved, with persisting probasidia.

A. — Context with pillars, tall or stubby, distinct, mostly unbranched; three- or four-layered;

surface some shade of grey, usually pale smoke-grey, smooth.

9. S. bogoriense.

B. — Context without pillars; in section less than 500 p thick; forming small anastomosing

patches intermingled with mosses and liverworts; fruiting surface smooth with upturned

margins, brownish; rhizomorphs extending between cracks of bark.

10. S. Bagliettoanum.

V. Basidia 4-celled, straight or slightly curved, usually without a probasidial cell.

Context with distinct pillars; with closely packed parallel upright threads in the hymenium;

context 400-700 p thick; hymenium conspicuously cracked.

11. S. philippinense.

1. Septobasidium protractum Sydow! in Ann. Myc. 10 (1912) 33; Couch, l.c., p. 91;

v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 22.

Specimens examined: 2002, 22552, 30280, 34588.

2. Septobasidium griseopurpureum Couch l.c., p. 105.

No specimens seen.

3. Septobasidium Schweinitzii Burt in Ann. Mo. Bot. Gard. 3 (1916) 324; Couch,

l.c., p. 112. (= Thel. pedicellata Schw., non S. pedicellatum Pat., fide Miss Wake-

field in MS. note.)

No specimens seen.

4. Septobasidium natalense Couch! loc. cit., p. 118.

Specimens examined: 2107, 17276, 21011, 25477, 27690.

5. Septobasidium pseudopedicellatum Burt in Ann. Mo. Bot. Gard. 3 (1916) 327;

Couch, l.c., p. 132; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 22; Coker in

Journ. Elisha Mitchell Soc. 35 (1920) 125.

No specimens seen.

10

6. Septobasidium grandispinosum Couch! loc. cit., p. 154.

Specimens examined: 12463, 20350.

7. Septobasidium Carestianum Bres. var. natalense Couch! loc. cit., p. 157.

Specimens examined: 26321, 28322, 33555, 22086, 27688, 30824, 34403.

8. Septobasidium Curtisii (B. & D.) Boedijn and Steinmann in Bull. Jard. Bot. Buiten-

zorg iii: 2 (1931) 181; Couch, loc. cit., p. 164.

Specimens examined: 1890, 2381, 22338, 28522, 17820.

9. Septobasidium bogoriense Patouillard in Otto Warburg, Monsunia 1 (1900) 138;

Couch, loc. cit., p. 213; Boedijn and Steinmann in Bull. Jard. Bot. Buitenzorg iii:

2 (1931) 205. [= Septobasidium mompa (Tanaka) Rac. non Helicobasidium

mompa Tanaka.]

Specimens examined: 30276.

Van der Byl’s description of S. mompa [in Ann. Univ. Stellenbosch 7 (1929) 23]

is not in accordance with Couch’s, nor with the specimens I have seen. Van der Byl’s

plant differs in purple-brown to blackish colour, lack of pillars, and ovate spores.

10. Septobasidium Bagliettoanum (Fr.) Bresadola in Ann. Myc. 3 (1905) 164; Couch,

loc. cit., p. 241; v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 23. [=Hypochnus

Bagliettoanus Fries, Hym. Eur. (1874) 705.]

This species is noted on the record of v. d. Byl cited above, but he himself does not

appear to be sure of the diagnosis and his description, taken from Saccardo, Syll. Fung.

6 (1888) 661, differs in important points from the description in Couch’s monograph.

11. Septobasidium philippinense Couch !, loc. cit., p. 247.

Specimens examined: 2382, 28962, 34125.

SEBACINA Tulasne.

1. Sebacina africana Burt [in Ann. Mo. Bot. Gard. 13 (1926) 338], described from a

collection of van der Byl (No. 1 342) at Knysna, is stated by McGuire in his mono-

graph of the genus Sebacina [Lloydia 4 (1941) 43] to be “ a Corticium with typical

clavate undivided basidia at the surface and possessing hyaline subulate gloeocys-

tidia.” The present writer has not yet seen specimens of this. The type is in the

Farlow Herbarium and there is an isotype in Herb. v. d. Byl, Stellenbosch.

2. Sebacina epigaea (B. & Br.) Bourd. and Galz. [in Hym. de Fr. (1928) 39] was

collected by Eaton at the Cape and identified by Berkeley as Tremella epigaea

B. & Br. [Jour, of Bot. 14 (1876) 175]. There is not a specimen in the Berkeley

Herbarium at Kew, and apparently the species has not since been collected in

South Africa.

3. In Herb. Pretoriae No. 11520 (as Sebacina sp.) is a specimen which is sterile and

consists of a few hyphal strands and much detritus of an amorphous nature resem-

bling white paint. The specimen is certainly indeterminable, and not referable to

Sebacina.

HETEROCHAETE Patouillard.

1. Heterochaete andina Pat. apud Pat. and Lagerheim in Bull. Soc. Myc. de Fr. 8

(1892) 120; Patouillard, Essai sur les Hym. (1900) 25, f. 18.

This species is described and recorded for South Africa by van der Byl [in Ann.

Univ. Stellenbosch 1 (1923) 5, f. 4] but neither of his two collections has been available

to me for study.

HOMOB ASIDIOMY CET AE.

Basidia always non-septate, claviform, cylindric, ovoid or urniform, bearing at the

apex a number of sterigmata (2 or 4, rarely 1-3 or 5-8 or an indefinite number) at whose

extremities the spores develop. Probasidia lacking. Spores on germination do not

produce secondary spores but give rise to a definite mycelium directly.

11

There follow descriptions and notes on the resupinate species of the Aphyllophora-

ceae recorded for South Africa, excluding the genera Irpex and Poria. Where there are

several species in one genus, a key to the species is provided.

EXOBASIDIUM Woronin.

1. Exobasidium Vaccinii (Fuckel) Woronin in Verh. der naturf. Ges. zu Freiburg 1,

B. 4, Heft 4 (1867) 397; Sorauer, Handbuch der Pflanzenkrankh. Ed. 3 : 2 (1906)

379; Saccardo SyLl. Fung. 6 (1888) 664; Patouillard, Essai taxon, sur les Hym.

(1900) 35; Ross, Die Pflanzengallen Mittel und Nord Europas 2 Ed. (1927) 29,

f. 30; Burt in Ann. Mo. Bot. Gard. 2 (1915) 649; Rea, Brit. Basid. (1922) 725;

Bourdot and Galzin, Hym. de Fr. (1928) 75; Donk, Rev. Ned. Hetero- en Homobas.

1 (1931) 124.

Fusidium Vaccinii Fuckel, Fung. Rhen. Exsicc. No. 221.

Exobasidium Andromedae Peck, 26 Rept. of Botanist Univ. N.Y. State (1874) 73;

Saccardo, Syll. Fung. 6 (1888) 665.

Exobasidium Azaleae Peck, loc. cit., p. 72; Saccardo, Syll. Fung. 6 (1888) 665;

Bourdot and Galzin, Hym. de Fr. (1928) 76.

Exobasidium Rhododendri Cramer in Rabenh. Fung. Eur. Exsicc. No. 1910;

Saccardo, Syll. Fung. 6 (1888) 664; Bourdot and Galzin, Hym. de Fr. (1928) 75.

E. Arctostaphyli Harkness; E. cassandrae Peck; E. discoideum Ellis; E. Karstenii

Sacc. ; E. karstenii Lind.; E. Myrtilli Siegm.; E. oxycocci Rostr. ; E. Peckii

Halst. ; E. Vaccinii- Myrtilli (Fuckel) Juel.

Above synonyms after Donk.

Illustrations : Sorauer (1906) l.c. ; Patouillard (1900) l.c., f. 26; Ross (1927) l.c.,

f. 30; Clements and Shear, Genera of Fungi (1931) PI. 42, f. 14.

Plate 3, fig. 1.

Fungus parasitic on leaves of higher plants, producing hypertrophy and galls.

Mycelium hyaline, inter- and intracellular.

Basidia: produced subepidermally, unilaterally or amphigenously on leaves, in groups,

eventually breaking through and projecting beyond the epidermis, clavate, with

(2)-4-6 straight thin sterigmata; Basidia 26-40 x 4-6-6 p.

Spores: spindle-shaped or cylindric, narrowing at one end, straight or curved, hyaline,

smooth, with granular contents, 7-15 X 2-4 p, becoming septate and budding off

sessile secondary conidia.

Conidial Hyphae: accompanying the basidia, branched, vertical, projecting beyond the

basidia and bearing conidia which are cylindric, narrowed at each end, 1-5 x 8 p,

according to Donk.

Specimens examined: 17819, Davies, on Azalea, Cape Town, 3/11/23; 30645,

Anderssen, on Azalea, Pretoria (nursery plants imported from Belgium);

34151, Wager, on Azalea, Gillets, Natal, 28/8/1944; 361 10, Sutton, on Azalea,

Villerust, Durban, 21/1 1/1946.

The secondary spores are produced as buds from the mature basidiospores, not

by a process of germination or renovation. This fact, together with the fact that the

basidia are undivided, indicates that Exobasidium is a homobasidiomycete and not a

heterobasidiomycete. It is likely that the species was introduced into South Africa

with Azalea plants and is not indigenous to the country.

2. Exobasidium Giseckiae Allescher apud Hennings, Pilze Ostafrika, in A. Engler,

Die Pflanzenwelt Ostafrikas und der Nachbargebiete (1895) 54; Saccardo, Syll.

Fung. 14 (1899) 230; v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 9.

Illustrations'. Plate 3, fig. 2.

Fungus parasitic on the leaves and flowers of Giseckia pharnacioides L., producing

slight hypertrophy of the infected areas in the form of pale-rosy to distinctly red galls

which are amphigenous. Mycelium hyaline, inter- and intracellular.

12

Basidia: produced subepidermally in groups, eventually breaking through and projecting

beyond the epidermis, clavate-cylindric, about 30 x 3-5 u, with four sterigmata.

Accompanied by very narrow paraphyses, unbranched.

Spores: hyaline, at first ellipsoid then later ovate, 8-9 x 6-6-5 ,u (“ 6-9 x 4-5 /*” —

Engler).

Specimens examined: Host plant in each case is Giseckia pharnacioides. 2248,

J. Burtt-Davy (12824), Kaffraria nr. Christiana, Bloemhof distr., 15/3/1912;

7380, Pott, Magaliesberg, 29/1/1914; 9685, Moran, Kimberley, 31/3/1916;

10033, Potts, Bloemfontein, Feb. 1917; 10054, Fuller, Pienaar’s River,

Pretoria, 16/3/1917; 21100, Wager, Tuinplaats, Springbok flats, 5/2/1927;

25919, Paynter, Pretoria North, 26/3/1931; 26073, Liebenberg (2848), Nel-

spruit. May 1931; 26685, Phillips, Magaliesberg, 1/12/1 932 ; 32670, Hean,

Taungs, C.P., 8/4/1940; 32804, du Plessis, nr. Vaalwater, N. Transvaal;

26101, Smith (6051), Wonderboom Poort, Pretoria, 22/2/1932; 26629, Wager,

Goldin, Nelspruit, 14/12/1932; 550, Pole Evans, Pienaar’s Poort, Pretoria,

21/1/1909.

“ Erineum ” Galls on Gymnosporia buxifolia leaves.

Illustrations'. Plate 4.

Two collections in Herb. Pretoriae, Nos. 11684 and 13087, under the genus Exo-

basidium prove to be non-fungal galls of the sort which were formerly given the generic

name Erineum. The under-surface of the leaves is covered with closely arranged small,

golden-yellow pustules from which fascicles of hairs arise, the hairs being large and

clavate and bearing a slight resemblance to basidia. However, they are thicker-walled

than a basidium, have occasionally up to two transverse septa and always lack any

suggestion of sterigmata and spores. No hyphae are present in sections of the galls

and there is every indication that the hypertrophy and production of hairs is caused

by insect stimulation, possibly by a mite. Most of the cells within the galls are clearly

neoplasms in which the nuclei are still plainly visible. An illustration of this structure

is given in Plate 4. It is similar to Erineum aureum illustrated in Greville’s Scot. Crypt.

Flora 1 (1823) PI. 33.

The specimens referred to above are: 11684, Junod, Rikatle, P.E. Africa, 31/5/1918;

13087, Doidge, Zilikats Nek, Pretoria, 10/4/1920.

CORTItIUM Pers. ex Fries.

Key to species described : —

1) With gloeocystidia:

a. Gloeocystidia numerous, immersed, with bright yellow contents; spores 4-5 x 9—1 1 - 5 p.

C. luteocystidiatum (1).

aa. Gloeocystidia immersed, hyaline, often capped by amber coloured resinous globules;

spores 3-4 X 7-10 p C. pallidum (2).

2) With cystidioles:

Cystidioles rare, immersed or slightly emergent, hyaline, sometimes with a subapical septum;

spores pip-shaped, usually agglutinated, 3-4 x 5 -5-6 -5 p C. gloeosporum (3).

3) Without gloeocystidia or cystidioles:

a. Fungus bright blue throughout C. caendeum (4).

aa. Fungus not at all blue.

v. Hymenium waxy, smooth, apricot to cinnamon-buff colour, spores 6-6-7 x 10-16 ja

C. armeniacum (5).

w. Hymenium cracking into flaky areoles with whitish subiculum; dirty white to light tan

colour; spores 5— 7— (9) X 3 • 5— (5) p C. scutellare (6).

x. Hymenium tuberculate, reddish-ochre with liver-brown tubercles and pale margin,

context dingy yellow; spores 3-4-2 x 4 -5-5 -5 p C. tumulosum (7).

13

y. Hymenium rosy, paling to flesh colour, smooth or cracked into areoles connected by

whitish subiculum; hyphae up to 10 /x wide; spores 6-5-8-3 x 10-13 /x C. salmoni-

color (8).

z. Hymenium waxy, whitish, like candle grease when fresh, smooth, seldom cracked,

drying buff to light pinkish-buff; spores oblong-subspherical, 7-8 x 10-11 -5 /x, with

prominent apiculi C. eonfluens (9).

1. Corticium luteocystidiatum Talbot apud Wakefield and Talbot in Bothalia 4 (1948)

941, f. 2.

Resupinate, not adnate, membranous, cracking into large separated areas. Hymen-

ium ochraceous-salmon drying to pale yellow-orange, almost smooth or papillate or

with fine raduloid ridges. Margin lifting from substratum on drying, sometimes with

small indentations, rarely distinctly fibrillose-radiate. Context concolorous, about

300 p thick. A bright yellow colour develops on contact with KOH.

Basidia: clavate, very variable size, up to 40 p long.

Spores: 4-5-(6) x 9—1 1 • 5— (12) p, cylindric or ellipsoid, frequently with one side depres-

sed or flattened, hyaline, smooth.

Gloeocystidia : Numerous, conspicuous, with bright yellow oily globular contents,

scattered throughout the subhymenium and trama except next to the substratum,

often penetrating the hymenium, irregular elongated outline, about 60-80 x 6-14

p, irregular or clavate to cylindrical.

Hyphae : very fine, mostly indistinct, basally horizontal, the superior hyphae erect.

Specimens examined: 28307, Rump (241), Town Bush, Maritzburg, Oct. 1934;

28557, Rump (167), Impolweni, Natal, Sept. 1934; 28690, Rump (283) Town

Bush, Maritzburg, Nov. 1934; 28679, Rump (319. b) Town Bush, Maritzburg,

Nov. 1934; 28947, Doidge, Xumeni, Donnybrook, Natal, July 1935; 34378,

Rump (644) Krantzkop, Natal, 1935; 27759, Rump (104), Town Bush, Maritz-

burg, Aug. 1934; Belgian Congo: (Specimens in Herb. Kew) Mission Agrono-

mique, P. Hyac. Vanderyst (16309 and 16325), 1925.

This species is one of the relatively few with bright yellow gloeocystidia and is

distinguished especially by its hymenium colour and by its spores. The colour reaction

with KOH is a help in diagnosis. C. abeuns Burt, recorded from South Africa, has

almost colourless gloeocystidia and spores of a different size and shape (6-7 x 4-6 p).

C. seriate Fr., another species with bright yellow gloeocystidia has larger spores,

6 x 10-13 p and is otherwise different, particularly in colour.

2. Corticium pallidum Bresadola, Fung. Trident, 2 (1898) 59; Rogers and Jackson

in Farlowia 1 (1943) 296.

G/oeocystidium pallidum (Bres.) Hohnel and Litsch. in K. Akad. Wiss. Wien

Sitzungsb. 116 (1907) 838; Bourdot and Galzin, Hym. de Fr. (1928) 258.

G/oeocystidium argillaceum von Hohnel and Litschauer in K. Akad. Wiss. Wien

Sitzungsb. 117 (1908) 1094 ( nomen nudum); Ibid in Osterr. Corticeen (1908)

67 (with description); Bourdot and Galzin Hym. de Fr. (1928) 258.

Gloeocystidium pallidum subsp. argillaceum sensu Bourd. and Galz. in Bull. Soc.

Myc. de Fr. 28 (1913) 263, Hym. de Fr. (1928) 258.

Corticium ochrofarctum Burt in Ann. Mo. Bot. Gard. 13 (1926) 275. [Non Corti-

cium ( Peniophora ) argillaceum Bresadola, Fung. Trident. 2 (1898) 63, nec

Peniophora carneola (Bres.) Hohnel and Litsch., Osterr. Corticeen (1908) 70].

Above synonymy after Rogers and Jackson, loc. cit.

Illustrations: Hohnel and Litschauer (1907) loc. cit. (As Gloeocystidium pallidum.)

Plate 5.

Resupinate, indeterminate, granular, very thin, up to 140 p in thickness, crustose,

whitish becoming spotted with dark resinoid granules giving the whole hymenium a

beige colour, adnate, non-continuous under the lens, subfarinaceous.

14

Basidia: “ 20-45 X 5-8 p” (Bourd. and Galz.), with 2 or 4 long, straight sterigmata,

4-6 p in length.

Spores: Oblong or subcylindric, often with one side depressed, hyaline, smooth, often

unigu'ttulate, 7-10 x 3-4 p.

Gloeocystidia: fusiform, or obtuse at upper end, hyaline, thin walled, with homogeneous

contents, 48-60 x 4-7 p, capped by resinous granules, immersed throughout the

trama.

Resinous granules: amber to reddish-brown, irregular in size and shape but commonly

13-20 fi diam., often aggregated, capping gloeocystidia and hymenial hyphae,

deposited in all parts of the trama, partly soluble in hot lactic acid.

Hyphae: hyaline, completely indistinct in the specimen seen from South Africa. “ Often

little distinct, 2 3-4-5 p diam., with rare clamps, the mycelial hyphae more regular,

2-5 fi.” — Bourd. and Galz.

Specimens examined: Exsicc. Brinkmann Westfalische Pilze iv. No. 157; 28939,

Doidge and Morgan, Donnybrook, Natal, February, 1935.

Peniophora argd/acea Bres., a species which produces amber coloured resinous

granules, was at first confused with the present species by von Hohnel and Litschauer,

but it differs in having emergent, thin-walled cystidia up to 100 //. long, and considerably

wider elliptic-depressed spores. These differences were first observed by Bourdot and

Galzin. Recently Rogers and Jackson (1943, l.c.) have shown that Gloeocystidium

argillaceum Hohnel and Litsch. should be regarded as a synonym of Corticium pallidum

Bres., raising it from the subspecific status suggested by Bourdot and Galzin (1913,

l.c.).

3. Corticium gloeosporum Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 940.

Illustrations'. Plate 6.

Resupinate, orbicular when young becoming widely effused and confluent, adnate,

later areolately cracked and separable, the edges of the cracks sometimes lifting, often

showing a silky fibrillose whitish subiculum between the cracks. Hymenium smooth,

somewhat waxy, undulate or tubercular in places, cream coloured, becoming pale buff

tinged incarnate, or rosy buff paling in the herbarium. Margin abrupt or very narrowly

radiating in white adnate hyphae. Context creamy, 200-500 p thick.

Basidia: cylindric-clavate, in close fascicles forming a very compact palisade, 4-5-5

X 24-30 p, usually with a clamp connection at the base of the basidium, arising

from deep seated hymenial hyphae.

Spores: pip-shaped, 3-4 x 5 -5-6 -5 p, smooth, hyaline, usually agglutinated in groups

of up to 4 spores in number, with a basal apiculus, often somewhat curved, not

amyloid.

Cystidioles: rather rare usually, very slightly emergent, or immersed, fusiform, hyaline,

thin walled, with a pointed apex, occasionally with a subapical septum, 4-5 x

32-45 p, with homogeneous contents.

Hyphae: subhymenial hyphae more or less erect, thin walled, rather indistinct; basal

hyphae very distinct, thin walled, much branched, septate, with abundant clamp

connections and occasional H-anastomoses, 2-5 p wide, closely interwoven in a

subhorizontal direction.

Specimens examined: Type, 28288, Rump (212), Town Bush, Maritzburg, Oct.

1934, on indigenous wood possibly Fagara rapensis; 27603, Rump (26),

Bulwer, Natal, 1934, on bark of wattle; 28712, Joubert, Windy Hill Estate,

Maritzburg, on bark of Acacia mollissima, 30/10/1 936.

This species differs from Corticium laeve Pers. principally in having basidia and

spores which are very considerably smaller. (C. laeve basidia 25-40-90 x 4-5-9 p;

spores 7-9 12 x 4-5-7 p, in European specimens.) C. laeve has not been collected in

15

South Africa. Corticium gloeosporum corresponds almost exactly with C. laeve in

external features but is perhaps more areolately cracked and thinner at the margin;

the spores of both species have the highly characteristic pip-shape and the property of

adhering to one another; their hyphae correspond almost exactly; the presence of similar

cystidioles and of basal clamps to the basidia are other points of similarity. C. gloeo-

sporum is almost certainly the South African analogue of C. laeve. In one of the

specimens (28712) some of the hyphae have a finely granular incrustation which is

soluble in potassium hydrate. Large cystidioles may be interpreted by some workers

as cystidia, with consequent change of genus, but in assigning this species to Corticium

the author has followed the practice observed with C. laeve. If mounts of the spores

are made in weak sepia solution it can readily be observed that each spore has a very

narrow mucilaginous investment which may account for their property of adhering

together in groups. [This property does not seem to be very common but it has also

been observed in Corticium centrifugum (Lev.) Bres. and Cytidia flocculenta (Fr.) Hohnel

and Litschauer.]

4. Corticium caeruleum (Schrad. ex Fr.) Fries, Epicrisis Syst. Myc. (1838) 562, Hym.

Eur. (1874) 651, Fung. Natalenses (1848) No. 38; Berkeley, Outl. Brit. Fung.

(1860) 274; Kalchbrenner in Grev. 10 (1881) 59; Massee in Journ. Linn. Soc. Bot.

27 (1890) 151; Saccardo, Syll. Fung. 6 (1888) 614; Quelet, Flor. Myc. de Fr. (1888)

10; Wakefield in Trans. Brit. Myc. Soc. 4 (1913) 119; Rea, Brit. Basid. (1922) 673;

Burt in Ann. Mo. Bot. Gard. 13 (1926) 301; Bourdot and Galzin, Hym. de Fr.

(1928) 183; Petch in Ann. Roy. Bot. Gard. Peradeniya 9 (1925) 287; van der Byl

in Ann. Univ. Stellenbosch 7 (1929) 27.

Thelephora caerulea Schrader in De Candole FI. Gall. Fr. 2 (1815) 107; Persoon,

Myc. Eur. 1 (1822) 147; Fries, Elenchus Fung. (1828) 202; Hussey, Illustr.

/Brit. Myc. 1 (1847) t. 20 b.

Byssus phosphorea Linn., Spec. Plantarum Ed. 3, 2 (1764) 1638.

Auricularia phosphorea (L.) Sowerby, Eng. Fung. (1803) t. 350.

Mycinema phosphoreum (L.) Agardh., Systema Alg. (1824) 33.

Thelephora indigo Schweinitz in Naturforsch. Ges. Leipzig Schrift 1 (1822) 107

(fide Burt).

Thaelaephora fimbriata Roth, Catalecta Botan. 2 (1800) 257, tab. 9, f. 2.

Thelephora atro-coerulea Trog. in All. Bot. Zeit. 2 (1832) 560.

Illustrations : Wakefield (1913) loc. cit., t. 3, f. 26; Sowerby l.c., as Auricularia

phosphorea', Hussey l.c., as Thelephora caerulea.

Plate 7.

Resupinate, effused, velutinous-floccose becoming membranous, separable when

moist, bright indigo to Prussian blue throughout. Margin concolorous or paler,

thinning out, satiny. Colour changing immediately to greenish in potassium hydrate.

Context loose, 200-400 g thick.

Basidia: 5-5-8 x 30-40 g, with 2-4 sterigmata.

Spores: not seen in S. African specimens, “hyaline, ovate-elliptical, 7-9 x 4-6 u” —

Wakefield.

Hyphae: loosely intertexed, 4-5 g diam., with numerous septa and clamp connections,

thick walled, coloured blueish especially near the hymenium. A few of the hyphae

are thinwalled.

Specimens examined: 26847, Ackerman (2), Maritzburg, 1933 (in Herb. Kew);

28888, Bottom/ey, The Cavern, Drakensberg, Natal, 1927; 27801 Craik,

Caverns Valley, Mont-aux-Sources, Natal, Feb. 1935; 34320, Doidge, Xumeni,

Donnybrook, July 1944; 35323, Rump, (590), Karkloof Bush, March 1944;

16

33206, Doidge and Bottomley, Wolhuterskop, Rustenburg Distr., 5/5/1939;

van der Byl (1051), East Africa, 1923; Maitland (38) 1929, (110) Jan. 1931,

(106) Dec. 1930, Cameroon Mountains.

5. Corticium armeniacum Saccardo, Syll. Fung. 6 (1888) 637 (non C. armeniacum

B. & C.! nomen nudum, in Herbarium, which is an Aster ostromella).

Corticium ceraceum Berk, and Ravenel! in Ravenel Fung. Car. Exsicc. 3 (1855)

29, nom. nudum', Massee in Jour. Linn. Soc. Bot. 27 (1890) 150; Hohnel and

Litschauer in K. Akad. Wiss. Wien Sitzungsb. 1 16 (1907) 785, f. 6; Burt in Ann.

Mo. Bot. Gard. 13 (1926) 216; v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 30.

Corticium molle B. and C. ! in Jour. Linn. Soc. Bot. 10 (1868) 336, Grevillea 1

(1873) 180. (non C. molle Fries.)

Corticium mauritianum Berk. ! in Herb.

Illustrations'. Hohnel and Litschauer l.c., text fig. 6.

Plate 8.

Resupinate, widely effused, fleshy then becoming very membranous-ceraceous,

somewhat shiny, very smooth, with very occasional cracks, not adnate. Hymenium

apricot colour becoming cinnamon-buff in the herbarium. Margin lighter in colour.

Context white, 100-500 y thick.

Basidia: flexuous, narrow-clavate, (30)-56 x (5)-7 y, sterigmata and spores not seen in

South African material; sterigmata 4, thick, curved, 10-14 x 1 -5-3 y fide Hohnel

and Litschauer, loc. cit.

Spores: (from type) long ellipsoid to cylindrical, flattened unilaterally, usually with a

prominent apiculus, 6-6-7 x 10-16 y, hyaline, smooth.

Hyphae: in a generally erect position; those below the hymenium very compact and

indistinct; basal hyphae erect or suberect, 3-3 -5-4 y wide, at first distinct but

later cleaving together and modified by waxy coating; clamp connections, when

present, are quite numerous and frequently in whorls, sometimes absent.

Minerals: quite large concretions sometimes found in the upper layers of the trama.

Specimens examined: Ravenel Fung. Car. 3, No. 29, in Herb. Kew; v. d. Byl (2568),

Natal, 1930; 27802, Conservator of Forests, Manubi, Butterworth, Cape,

23/10/1934, on Ekebergia capensis', 30227, Rump (445), Krantzkop, Nov.

1935; 30219, Rump (429), Compensation Beach, Natal, Dec. 1934; 34326,

Rump, Town Bush, Maritzburg, 1935; 36708, Talbot, Kloof Falls, Natal,

May 1948; Diimmer (2398), on Ficus sp., Kipayo, Uganda, Apr. 1915.

The type of the species, from North America, has very variable spores which at

their narrowest measure 5-5 y, depending in the same spore on the side from which it

is viewed. The Uganda specimen cited above is undoubtedly this species, but its spores

are 4-5-5 -5 x 9-13 y, oblong-elliptic, and lacking a prominent apiculus.

6. Corticium scutellare Berk, and Curtis ! in Grevillea 2 (1873) 4; Saccardo, Syll.

Fung. 6 (1888) 634; Massee in Jour. Linn. Soc. Bot. 27 (1890) 128; Burt in Ann.

Mo. Bot. Gard. 13 (1926) 192; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 27.

Illustrations : van der Byl loc. cit., PI. 2, f. 9.

Plate 9.

Resupinate, effused, dirty-white becoming light tan, waxy, adnate, thin, 100-250 y

in section, cracking into small irregular areoles in great profusion, sometimes flaking

away and leaving a little of the white subiculum behind. Context whitish.

Basidia: clavate or obconical, circa 24-33 x 4-7 y, with 2 or 4 short, curved sterigmata.

17

Spores: In the type specimen 7-9 x 4-5 / 1 , elliptical; fide Burt loc. cit., hyaline, smooth,

4-6 X 2-3 p; as observed in South African specimen, 5-7 x 3-5 p, elliptical,

flattened unilaterally, with a distinct apiculus.

Hyphae : narrow, 2 • 5-3 • 5 p wide, more or less erect below the basidia, but so mixed

with mineral inclusions in the deeper tissues as to be almost indistinguishable.

Specimens examined: Type, 2473 Car. Inf., in Kew Herbarium; 27680, Rump

(53), Winterskloof, Natal, 1934.

It will be noted that the spores in the South African specimen are smaller than

those in the type, but not as small as those quoted for the species by Burt. Further

collections are required to establish the range of variation in spore size; in all other

characters the South African specimen is a good match with the type of the species.

7. Corticium tumulosum Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 941, f.3.

Resupinate, closely adnate, orbicular becoming effused, cracking deeply crosswise

in one direction. Margin thinning out to a definite, pale coloured edge. Hymenium

tuberculate, waxy, papillate or ridged, the raised portions being liver-brown in colour

and the remainder reddish-ochre. Context dingy yellow, showing through the cracks.

In section 60-100-(200) p including basal layers which separate the periderm cells of

the substrate and also contain much crystalline matter.

Basidia: cylindric-clavate, 28-34 x 5-5 /» with 4 short, curved sterigmata.

Spores: hyaline, smooth, broadly ellipsoid, 3-4-2 x 4 -5-5 -5 p.

Hyphae : distinct, pale yellowish horizontal weft : the few hyphae which curve abruptly

into the hymenium are colourless; septation frequent, without clamp connections,

3-4 p wide, thin walled.

Minerals: A great deal of amorphous or crystalline mineral matter at the base of the

trama.

Specimens examined: 28897, Lansdell, The Willows, Pretoria distr., 27/4/1937;

36857, Talbot, Garstfontein, Pretoria distr., 21/8/48.

In external appearance, and especially in colour, C. tumulosum is very much like

C. cacaoicolor Petch ! [in Ann. Roy. Bot. Gard. Perad. 9 (1925) 288], The latter,

however, has spores which are oblong-ovate, 3x7 p, and hyphae which are much

finer and less distinct than those of C. tumulosum.

8. Corticium salmonicolor Berk, and Broome ! in Jour. Linn. Soc. Bot. 14 (1873) 71;

Saccardo, Syll. Fung. 6 (1888) 620; Burt in Ann. Mo. Bot. Gard. 13 (1926) 227;

Massee in Jour. Linn. Soc. Bot. 27 (1890) 122; Petch, Phys. and Diseases of Hevea

brasiliensis (1911) 209, Diseases and Pests of the Rubber Tree (1921) 134, Ann.

Roy. Bot. Gard. Perad. 9 (1925) 281; Brooks in Agric. Bull. F.M.S. 2 (1914) 238;

Brooks and Sharpies in Dept. Agric. F.M.S. Bull. No. 21 (1914) figs. 1-19; Rorer,

Trinidad Dept. Agric. Bull. 15 (1916) 86, f. 1-2; Fawcett, Citrus Diseases and their

Control (1936) 324; Lee and Yates in Philippine Jour. Sci. 14 (1919) 657, PL 1-7.

Necator decretus Massee ! in Kew Bull. (1898) 119; Saccardo, Syll. Fung. 16

(1902) 1094; Rant in Bull. Jard. Bot. Buitenzorg 4 (1912) 1, f. 1-14.

Corticium javanicum Zimmermann in Centralbl. fiir Bakt. 7 (1901) 103, f. 3 [non

C. javanicum (P. Henn.) Sacc. and Syd.]; Rant, (1912) loc. cit.

Corticium Zimmermanni Sacc. and Syd. In Saccardo, Syll. Fung. 16 (1902) 1117;

ibid 17 (1905) 169.

Illustrations : Brooks and Sharpies, 1914 l.c.; Rant, 1912 l.c. (as C. javanicum) ;

Zimmermann, 1901 l.c. (as C. javanicum).

Plate 10.

Type: in Herb. Kew, from Ceylon.

18

Resupinate, effused, membranous-pellicular, hymenium rosy when fresh, drying

fleshy-buff or creamy-yellowish colour in the herbarium, smooth, cracked into many

small areoles connected by whitish subicular strands. Context 250-350 y thick.

Basidia : irregularly cylindric-clavate, 6-8 X 26-40 y.

Spores : hyaline (pale rosy tint in a mass), broadly elliptic to ovate or subglobose, often

unilaterally depressed, 6 -6-7-8 -3 X 10-13-2 y, with a prominent basal apiculus.

Hyphae: hymenial hyphae densely aggregated and indistinct; basal hyphae strand-like,

hyaline, thickwalled, seldom branched, with frequent septa ; superior hyphae much

branched, witrf frequent septa, thin walled. All hyphae take stain readily except

in their walls; rare H-anastomoses but no clamps seen, hyphae loosely intertexed,

the basal ones horizontally, the superior ones ascending. Hyphal width 6—10—

11-5 y, their width and thickness of walls progressively diminishing from the basal

tissues upwards.

Conidial stage: ( Necator decretus). Not seen in South African specimens.

Specimens examined: 31998, Ballenden, on Pyrus malus, Maritzburg, 8/1/1918;

12829, Maritzburg, Natal; 34168, Forest Officer, on dead Quercus reticulata,

Tweefontein nr. Sabie 21/4/1944; 20669, Dungan, on Pyrus malus. Town

Bush, Maritzburg, 19/2/1926 (as C. laetum ); 12177, Marshall, on Pyrus malus,

Hilton Road, Natal, May 1919 (as C. laetum)', 14479, Govt. Horticulturist,

on Pyrus malus, Hilton Road, Natal, Feb. 1921 (as C. laetum)', Brooks, on

Ipomaea carnea, Malay States, 16/10/1914 (inch Necator stage) in Herb. Kew;

Jacobson, ex Herb. Hort. Bot. Bog. (6019), Sumatra, 1924, in Herb. Kew;

Rorer, on Cacao, Trinidad, July 10, 1915, in Herb. Kew; MacDonald, on

leaves of Coffea, Kenya, Dec. 1923 and March, 1924, .in Herb. Kew; Type

from Ceylon in Herb. Berkeley at Kew; Petch, on Hevea brasiliensis, Ceylon,

1913.

Apart from its parasitic occurrence on woody plants, a thinner more areolately

cracked hymenium, and possibly somewhat thicker walls to the basal hyphae, this species

is so near to C. laetum (Karst) Bres. in microscopic structure that a very close relation-

ship must be implied. C. laetum is more membranous, smooth, lacking areoles con-

nected by subicular strands. It has been reported by Stevens [The Fungi which cause

Plant Disease (1913) 408] as parasitic on apple and fig trees.

C. roseum Pers. (= C. roseolum Massee !) bears a resemblance to these species but

its spores are smaller (6 x 9-10 y) and its thick-walled, narrow (2-5-5 y) hyphae have

numerous clamp connections.

C. aurora Berk. ! differs from C. laetum in texture and thickness, but more especially

in its spores which are subclavate, slightly curved, basally attenuated, 13-16 x 3-4 y

[see Bourdot and Galzin, Hym. de Fr. (1928) fig. 62]. Massee [in Jour. Linn. Soc.

Bot. 27 (1890) 141] recorded the spores as “ ellipsoid, basally apiculate, 10-11 x 7-8

which led Hohnel and Litschauer to suggest its identity with C. laetum. In fact, it is

not confusable with this or with C. sa/monico/or.

9. Corticium confluens (Fries) Fries, Epicr. Syst. Myc. (1838) 564, Hym. Eur. (1874)

655; Berkeley, Outl. Brit. Fung. (1860) 276; Saccardo, Syll. Fung. 6 (1888) 626;

Massee in Jour. Linn. Soc. Bot. 27 (1890) 133; Bresadola in I.R. Accad. Agiati

Atti, iii, 3 (1897) 112; Bourdot and Galzin in Bulk Soc. Myc. de Fr. 27 (1911) 252,

Hym. de Fr. (1928) 212; Rea, Brit. Basid. (1922) 679; Burt in Ann. Mo. Bot. Gard.

13 (1926) 220; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 29.

Thelephora confluens Fries, Syst. Myc. 1 (1821) 447.

Corticium confluens var. subcalceum Karsten in Rev. Myc. 10 (1888) 74.

Above synonymy after Burt.

Illustrations : Plate 1 1 .

19

Resupinate, widely effused, when fresh waxy-membranous, whitish, sometimes

glistening with raised hairs at the surface, appearing like a patch of candle grease;

rather adnate when dry, and drying membranous, smooth, seldom cracked, buff to

light pinkish buff. Margin paler, thinner and either finely radiate-byssoid or ending

abruptly. Context pale buff, 200-500 /z thick in section.

Basidia : cylindric-clavate, somewhat contorted, large, 6 • 5-1 1-5 x 50-80 \i, often

guttulate, with sterigmata 4-5 g long, or longer.

Spores: 7-8 x 10-11-5 g; hyaline, oblong-subspherical or broadly ovoid, with a

prominent lateral apiculus, smooth, with finely granular content (stippled appear-

ance) or with large guttules, abundant.

Paraphyses : simple, smooth, about 2 g wide, interspersed among the basidia, sometimes

with up to four short lateral branches near the apex. Occasionally these para-

physes are lacking.

Hyphae: superior hyphae more or less vertical, agglutinated, rather indistinct, thin-

walled, hyaline, with few clamps, 2-3 g wide. Basal hyphae usually more or less

horizontal and entirely collapsed or very indistinct.

Specimens examined: 35328, Rump (687), on Ouercus, Maritzburg, 1943; 36850,

Talbot, on Acacia sp., Buffelspoort, Rustenburg distr., 12/8/1948; 36914,

Talbot, Fountains Valley, Pretoria, 24/10/48.

Bourdot and Galzin (1928) point out the interesting fact that Corticium confluens

and Radulum membranaceum (Bull.) Bres. have the same internal structure and spore

characters, and that the corticioid form may pass into the raduloid in some specimens.

The large oblong spores with prominent apiculi are unmistakable, as also is the “ candle

grease ” appearance of fresh specimens.

Synonyms and excluded species in the genus CORTICIUM :

atrocinereum Kalchbr., nomen nudum, = Stereum Schomburgkii Berk, (q.v.) if

the MacOwan specimen in Herb. Kew is authentic,

calceum Fries sensu Romell and Burt. Nomen confusum fide Rogers and Jackson

in Farlowia 1 (1943) 284. The MacOwan specimen in Herb. Kew. is Aleurodis-

cus acerinus (Pers.) Hohn and Litsch. (? var. longisporus Hohn. and Litsch) q.v.

calceum Fries var. lacteum Fries. Nomen confusum. MacOwan (1074), Somerset

East, is the type specimen of Aleurodiscus acerinus var. longisporus Hohn. and

Litsch.

ceraceum Berk, and Ravenel. Nomen nudum = Corticium armeniacum Sacc.

cinereum Pers. ex Fries. = Peniophora cinerea (Pers. ex Fr.) Cooke.

Dregeanum Berk. = Lopharia Dregeana (Berk.) comb. nov. (q.v.).

nudum Fries. = Peniophora nuda (Fr.) Bres. (q.v.).

CORTICIUM SPECIES RECORDED FOR SOUTH AFRICA, BUT NOT AVAILABLE FOR STUDY,

abeuns Burt — v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 29; in S.A. Journ.

Sci. 23 (1926) 288.

bombycinum (Sommerf.) Bres. — Miss E. M. Wakefield in Litt.

lacteum Fries — v. d. Byl (1929) loc. cit., p. 30 [fide Rogers and Jackson in Farlowia

1 (1943) 294, this is a nomen dubium ].

pelliculare Karst — v. d. Byl (1929) loc. cit., p. 29.

portentosum B. and C. ! — v. d. Byl (1929) loc. cit., p. 28. South African specimens

of this species which the author has so far seen are all considered to be Astero-

stromella duriuscula (B. and Br.) comb. nov. (q.v.).

vagum B. and C. — Phillips in Bot. Survey of S.A. Mem. 14 (1931) 159.

20

PENIOPHORA Cooke.

Key to species described : —

1. Context yellow-buff or isabelline colour, turning strong violet colour instantly on application of a

drop of potassium hydroxide solution. P. filamentosa (5).

2. No violet colour reaction with potassium hydroxide solution :

A. — Cystidia occupying a relatively narrow zone confined to the hymenium and subhymenium:

a) Cystidia occupying a zone 70-100 p thick above a byssoid subiculum of thick-walled

hyphae; cystidia encrusted, 20-36-(40) x 10-13 p, fairly thin-walled

P. pelliculosa (1).

aa) Cystidia thick-walled, heavily encrusted, 36-80 X 10-11 -5 p; subicular hyphae not

byssoid, indistinct P. gigantea (6).

AA. — Cystidia not in a distinct zone above the subiculum, staged throughout the trama or at least

originating in mid-trama:

a) Cystidia septate, thin-walled, often with clamp connections, encrusted

P. aspera (2).

aa) Cystidia not septate or clamped, thick-walled, densely encrusted :

b) Cystidia 6-10 x 50-120 p ; spores 2-5-3 x 6-8 /x; hyphae distinct, 3-6 p

wide P. arenata (3).

bb) Cystidia 10-16 x 50-70 p; spores 3 x 5-5 p; hyphae agglutinated, 3 p wide

P. Roumeguerii (4).

1. Peniophora pelliculosa sp. nov.

Illustrations'. Plate 12.

Type : 30231, Rump (450), on indigenous wood, Krantzkop, Natal, Dec. 1935.

In Herb. Kew with isotypes in Pretoria Herbarium and Herb. University of

Toronto.

Resupinate, widely effused, thick, soft, pellicular, separable when moist. Hymen-

ium very smooth, usually not cracked, coloured “ pinkish-buff ” (Ridgway), or a more

fleshy-yellow, able to flake off like plaster leaving the byssoid white subiculum showing

beneath. Margin indeterminate, byssoid to arachnoid. Context floccose, white,

width in section 550-1,000 p.

Basidia: cylindric-clavate, compact in fascicles, 5-6 x 16-23 p (immature).

Spores : ellipsoid or oblong, 3-3 • 3 x 5-7 p, smooth, hyaline.

Cystidia: in a zone 70-100 p. thick above the subiculum, mostly immersed, a few pro-

jecting up to 12 p, relatively thin-walled, some thickened at the apex, heavily

encrusted with mineral matter which dissolves in KOH, 20-36-(50) x (8)— 10— 1 3 p,

more or less cylindric-clavate to obconic, frequently capitate-encrusted only.

Hyphae: indistinct in cystidial layer; in context very distinct, branched, hyaline, septate,

without clamps, loosely interwoven, rather thick-walled, with an incrustation of

large detersile crystals especially under the cystidial zone, 3 -5-5-6 -5 p wide,

somewhat narrower beneath the hymenium.

Specimens examined: Type, 30231, Rump (450), Krantzkop; 34383, Rump (653),

Krantzkop, Natal, 1935; 34359, Rump (559), Krantzkop, Natal, 1935.

This species falls in the group containing P. cremea Bres. (for which it has been

mistaken in South Africa) and P. subiculosa Burt. The latter species has not been seen,

but it is evidently thinner and has a chamois-coloured subiculum and spores which are

smaller, 3-3-5 x 2-5 p [Burt in Ann. Mo. Bot. Gard. 12 (1925) 259], P. eremea is not

so thick and byssoid-pellicular as the new species, and has larger cystidia which project

further. The mineral encrustation is much more marked in the cystidia and hyphae

of the new species. I have much pleasure in acknowledging the assistance of Dr. H. S.

Jackson of Toronto University, who examined collections of this species and reported:

“ The combination of rather short incrusted cystidia with the byssoid subiculum of

thick-walled hyphae makes this unique among the species I am familiar with.”

21

2. Peniophora aspera (Pers.) Saccardo, Flor. Ital. Crypt. Hymen. (1916) 1182; Rogers

and Jackson in Farlowia 1 (1943) 282.

Peniophora setigera (Fr.) Hohnel and Litschauer in K. Akad. Wiss. Wien Sitzungsb.

115 (1906) 7.

Peniophora setigera (Fr.) Bresadola in Litt. ; Rea, Brit. Basid. (1922) 692; Bourdo

and Galzin, Hym. de Fr. (1928) 309.

Kneiffia setigera Fries, Epicrisis Syst. Myc. (1838) 529; Bresadola, Hym. Hung.

Kmetiani (1897) 40, in Ann. Myc. 1 (1903) 103; Saccardo, Syll. Fung. 6 (1888)

510 and ibid. 9 (1891) 218; Karsten in Hedwigia 28 (1889) 195.

Odontia setigera (Fr.) Miller in Mycologia 26 (1934) 19, PI. 2, f. 3.

Corticium latitans Karsten in Rev. Myc. 10 (1888) 74, fide Rogers and Jackson

(l.c.).

Corticium Berkeleyi Cooke and Massee in Journ. Linn. Soc. Bot. 27 (1890) 133;

Burt in Ann. Mo. Bot. Gard. 13 (1926) 183, fide Rogers and Jackson (l.c.).

Corticium myxosporum Karsten, Symb. Myc. Fenn. 9, p. 53; Saccardo, Syll. Fung.

6 (1888) 635; Massee in Journ. Linn. Soc. Bot. 27 (1890) 130; Hohnel and

Litschauer in K. Akad. Wiss. Wien Sitzungsb. 1 15 (1906) 7, [non C. myxosporum

sensu Bresdola in Ann. Myc. 1 (1903) 94.].

Illustrations'. Miller (1934) loc. cit., as Odontia setigera.

Plate 13.

Resupinate, effused, membranous-byssoid, firmly adnate. Hymenium white then

creamy, later alutaceous and cracking on drying, rather poroid or finely papillate and

silky-setose. Context greatly variable in thickness.

Basidia: cylindric, 33 x 5-6 g, somewhat fasciculate.

Spores: hyaline, smooth, cylindric, one side often flattened or curved, 3-5 X 9-11 g,

with granular or guttulate contents.

Cystidia: cylindrical, 8-11 x 100-150 g (or longer), included or emergent up to about

40 g, septate, often with clamps at the septa, fairly thin-walled, hyaline, frequently

encrusted with detersile mineral matter, arising deep in the subhymenium or mid-

trama as lateral branches of the hyphae.

Hyphae : subhymenial hyphae 2 ■ 5-3 • 5 g wide, hyaline, much branched, septate, with

scanty clamp connections, frequently collapsed or indistinct. Basal hyphae similar

but seldom collapsed and up to 6-8 g wide, thin-walled.

Specimens examined: van der By/ (1461), on dry Eucalyptus poles, Tzaneen, Trans-

vaal, July, 1925.

The species is very variable in colour, thickness, and convolution of the hymenium .

The South African collection cited above is a small one in which the fungus is thin

and of a uniform creamy colour. Otherwise it corresponds very well with British

material of which I have made numerous collections.

The collection, v. d. Byl (1496), Woodbush, Transvaal, cited as P. setigera in Nel,

Ann. Univ. Stellenbosch 20 (1942) 91 has been examined in Herb. Kew. and proves to be

a species of Grandinia.

The septate cystidia of P. aspera are highly distinctive. Other British species

which have septate cystidia are P. pallidula (Bres.) Bourd. and Galz., P. tomentella

Bres. and P. byssoidea (Pers. ex. Fr.) Hohn. and Litsch., but in all these the cystidia are

much smaller and not above about 6 g in width.

In a young condition, where some of the cystidia are still unseptate and with dense

contents, P. aspera approaches very closely the form of Corticium roseo-cremeum Bres.

22

3. Peniophora arenata Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 944,

fig. 4.

Resupinate, widely effused, membranous or pellicular, easily separable when moist,

cracking on drying revealing a whitish fibrillose subiculum, orbicular when young.

Margin whitish, fibrillose. Hymenium light sandy-buff colour, pruinose to farinaceous

and finely granular. Context white, up to 300 p in thickness.

Basidia: clavate, 3-5 X 23-30 p.

Spores: hyaline, smooth, cylindric, 2-5-3 x 6-8 fx, apiculate-attenuate at one end,

usually slightly curved.

Cystidia: immersed, rarely projecting beyond the hymenium, cylindrical, thick walled,

very heavily encrusted, attenuated gradually at the base into a long hypha-like

cystidiophore originating in mid-trama, sometimes forking briefly at the apex,

sometimes closely grouped in fascicles, 6-10 x 50-120 //.

Hyphae: distinct, hyaline, rigid, thick-walled, with very sparse clamp connections, in a

dense basal horizontal layer and a rather looser semi-vertical layer near the hymen-

ium, frequently septate, sometimes encrusted, 3-6 p wide. The basal hyphae are

stouter and thicker walled than the superior hyphae.

Specimens examined: 27645, Type, Rump (34), Cato Ridge, Natal, 1934; 28916,

Doidge and Morgan, nr. Donnybrook, Feb. 1935; 28891, Bottomley, The

Cavern, Drakensberg, Natal, 20/7/37; 35314, Rump (733), Champagne Castle,

Natal, Feb. 1945; 33209, Doidge and Bottomley, Wolhuterskop, Boschfon-

tein Kloof, Transvaal, 5/5/39; Rump (625), Table Mountain, Natal, 1935.

The hymenium of this species, with its sandy appearance both in texture and

colour, is distinctive. Microscopically, P. arenata should be grouped with P. leprosa

Bourd. and Galz. The latter differs especially in having cystidia which are more fre-

quently emergent and often tapering towards the apex, and also in having a distinct

basal layer of thick-walled, heavily encrusted hyphae. In both species the cystidia are

long and roughly cylindrical, heavily encrusted, showing a tendency to fork or fragment,

and arise from a hypha-like base but are sharply differentiated from hyphae.

4. Peniophora Roumeguerii (Bres.) Burt in Ann. Mo. Bot. Gard. 12 (1925) 270;

Bourdot and Galzin, Hym. de Fr. (1928) 316.

Corticium Roumeguerii Bresadola, Fung. Trident. 2 (1892) 36; Sacc. Syll. Fung. II

(1895) 125.

Kneiffia Roumeguerii (Bres.) Bresadola in Ann. Myc. 1 (1903) 103.

Corticium Mollerianum Bresadola in Saccardo, Soc. Brot. Bol. 11 (1892) 13.

Peniophora Molleriana (Bres.) Saccardo, Soc. Brot. Bol. 11 (1892) 13; Saccardo,

Syll. Fung. 11 (1895) 128; Bourdot and Galzin in Bull. Soc. Myc. de Fr. 28

(1912) 401; Wakefield in Trans. Brit. Myc. Soc. 5 (1915) 132; Rea, Brit. Basid.

(1922) 693.

Peniophora macra Karsten. [Fide Bourd. and Galz. l.c. (1928)].

Peniophora stratosa Burt in Ann. Mo. Bot. Gard. 12 (1925) 333, nec. P. stratosa

Petch in Ann. Roy. Bot. Gard. Peradeniya 9 (1925) 293 [fide Rogers and Jackson

in Farlowia 1 (1943) 324],

Illustrations-. Bresadola, Fung. Trident. 2 (1892) PI. 144, f, 1.

Plate 14.

Resupinate, widely effused, very adnate, membranous, brittle when dry, cream to

chamois or biscuit colour, much cracked when dry, smooth or sometimes a little tuber-

culate. Hymenium at first rather pruinose, later with a waxy aspect. Margin narrow,,

whitish, pruinose to finely pubescent. In section up to 700 ,« thick.

23

Basidia: narrow, cylindric-clavate, 3-5 X 12-20-30 /x.

Spores: 3 X 5-5 g, ovoid-oblong, sometimes more elliptical, briefly and obliquely

attenuated at the base.

Cystidia: very numerous, scattered throughout the trama, only a few emergent, thick-

walled, encrusted, subconical or fusiform with acute apex, occasionally much

fragmented, 10-16 x 50-70 g .

Hyphae : erectly arranged, thin-walled, hyaline, with frequent septa but no clamp con-

nections, 3 [x wide, usually agglutinated and rather indistinct.

Specimens examined: 27764, Rump (115), Town Bush, Maritzburg, Aug. 1934;

27761, Rump (107), Town Bush, Maritzburg, Aug. 1934. (The latter without

spores and in poor condition.)

In hyphal characters and in the possession of abundant, thick-walled, acutely

pointed cystidia, P. Roumeguerii is closely allied to P. gigantea (Fr.) Massee and P.

pubera (Fr.) Sacc., both of which differ from P. Roumeguerii in producing their cystidia

only in the superior part of the trama and hymenium. P. gigantea, moreover, becomes

horny on drying while the other two species remain brittle.

5. Peniophora filamentosa (B. and C.) Burt apud Coker in Elisha Mitchell Sci. Soc.

Journ. 36 (1921) 162, PI. 32, f. 5, 6; Burt in Ann. Mo. Bot. Gard. 12(1925) 320;

Bourdot and Galzin, Hym. de Fr. (1928) 311.

Corticium filamentosum Berk, and Curt, in Grevillea 1 (1873) 178; Saccardo, Syll.

Fung. 6 (1888) 619; Massee in Journ. Linn. Soc. Bot. 27 (1890) 154.

Corticium Petersii Berk, and Curt, pro parte, in Grevillea 1 (1873) 177.

Peniophora unicolor Peck in N.Y. State Mus. Rept. 43 (1890) 66; Saccardo, Syll.

Fung. 9 (1891) 239.

Corticium radicatum P. Hennings, Piize Ostafrikas (1895) 54; Saccardo, Syll. Fung.

14 (1899) 222 ? (see Hohnel and Litschauer in K. Acad. Wiss. Wien. Sitzungsb.

117 (1908) 1093.

Peniophora radicata (P. Henn.) Hohnel and Litschauer, loc. cit., p. 1092 — see

Saccardo, Syll. Fung. 21 (1912) 411.

Above synonymy after Burt (1925).

Illustrations : Burt (1921) loc. cit. PI. 32, f. 5, 6.

Plate 15.

Resupinate, membranous, easily separable from the substratum when moist,

smooth, soft, typically yellow-buff colour (buffy citrine, Ridgway) or isabelline.

Margin concolorous or paler, composed of radiating adpressed fibrils, often forming

branched mycelial strands. Context concolorous, 150-400 / x thick. Reacting in KOH

to produce a strong violet colour.

Basidia: cylindrical, 3-4-5 X 28-38 fx.

Spores: 3 -5-4 -5 x 2-2-8 fx, ellipsoid, hyaline, smooth.

Cystidia: arising as branches of the hymenial hyphae only, fusoid, cylindrical or subulate,

fairly thick-walled, encrusted with granules which rapidly dissolve in KOH but

not in lactic acid, immersed or projecting up to 40 g beyond the basidia, 5 • 7-9 • 3

X 35-65 u.

Hyphae: hyaline, but encrusted with KOH-soluble coloured granules, with rare clamp

connections, loosely intertexed, 2 -8-5 -6 /x, with appreciably thick walls but wide

lumen.

Specimens examined: 35306, Rump (581), Botanical Gardens, Maritzburg, 1943;

12029, Diimmer (636), Kyagwe, Uganda, May 1915 (as P. radicata).

24

These specimens agree very closely in all particulars with Burt’s description of the

species. The vinous colour reaction with KOH is especially striking, as also are the

soluble granular encrustation of the cystidia, and the small colourless spores. Mycelial

strands, when present, are a useful pointer to the classification of the species in the group

Radicatae of Bourdot and Galzin.

A vinous colour reaction with KOH is also a characteristic of Polyporus rutilans

(Pers.) Fr. (= P. nidulans Fr.), a fungus with the same type of context, colour, hyphae

and encrustation of hyphae, and spores, as Peniophora filamentosa. In P. rutilans

the hyphae are 2-5 g diam., with thin to thickish walls and sparse clamps; the spores

are ellipsoid, smooth, hyaline to faintly coloured, 3-5 X 2-2-7 g. There is surely a

very close relationship between these two species, which in present classifications are

held so far apart.

6. Peniophora gigantea (Fries) Massee in Journ. Linn. Soc. Bot. 25 (1889) 142; Karsten

in Finska. Vet. -Soc. Bidr. Nat. och Folk 48 (1889) 422; Bresadola in I.R. Accad.

Agiati Atti, iii, 3 (1897) 113; Bourdot and Galzin, Hym. de Fr. (1928) 318,

Bull. Soc. Myc. de Fr. 28 (1913) 401; Rea, Brit. Basid. (1922) 693; Burt in Ann.

Mo. Bot. Gard. 12 (1925) 216.

Thelephora gigantea Fries, Obs. Myc. 1 (1815) 152, Syst. Myc. 1 (1821) 448.

Corticium giganteum (Fr.) Fries, Epicr. Syst. Myc. (1838) 559, Hym. Eur. (1874)

648; Peck in N.Y. State Mus. Rept. 28 (1876) 52; Saccardo, Syll. Fung. 6

(1888) 610.

Thelephora pergamenea Pers., Myc. Eur. 1, p. 150 (fide Bourd. and Galz.).

Illustrations: Fries, leones Hym. 2, PI. 197, fig. 3.

Plate 16.

Resupinate, widely effused on coniferous wood and bark, white or hyaline when

fresh, smooth, waxy, easily separable from the substratum ; contracting on drying to an

easily detachable, horny or parchment-like layer, coloured whitish to buff or pink-buff.

Margin white, radiately fibrillose. Context pale coloured when dry.

Basidia: cylindrical, 27-30 x 3 -5-4 -3 g.

Spores: (not seen with certainty in South African material), 2-5-4 x 4-6-5 g, oblong-

subcylindric, attenuated at the base briefly and obliquely.

Cystidia: occupying a narrow zone confined to the hymenium and subhymenium,

colourless, thick-walled, heavily encrusted with minerals, fusiform or conical,

10-11-5 x 36-80 g.

Hyphae: hyaline, vertically arranged and indistinct except near the substratum where

they are more horizontal, gelatinously modified thus usually only the lumen is

visible, thick walled, with rare clamps, (3)— 6— (8) g in width.

Subhymenial hyphae are thin-walled, about 2-7 g wide.

Specimens examined: 30854, Hillman Bros. Ltd., on pitch pine wood, Lourenco

Marques, P.E. Africa, 5/10/1939; 32176, Weintroub and Simpson (T.R.L.

1313), on bark of Pinus sp., Exchange Yard, Johannesburg, 10/9/1940.

Particularly characteristic of this species are its occurrence on conifers, its waxy

texture becoming corneous when dry, its fibrillose margin which lifts and curls on drying,

and the cystidia confined to the outer part of the fructification. The two specimens

cited above were both collected in timber yards and were most probably imported to

Southern Africa.

PENIOPHORA species recorded for South Africa, but not available for study.

cinerea (Fr.) Cooke; v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 16. [As

Corticium cinereum Fr. recorded in: Kalchbrenner in Grev. 10 (1881) 59; De

Thiimen in Flora 61 (1878) 354; Wood in Rept. Natal Bot. Gard. (1898) 17.]

*

25

The author has examined several South African specimens misdetermined as

P. cinerea, including 22044, P. MacOwan (1054), which is actually Stereum

umbrinum B. and C.

cremea Bresadola; v. d. Byl (1929) l.c., p. 16 and in S.A. Journ. Sci. 23 (1926)

288; Burt in Ann. Mo. Bot. Gard. 12 (1925) 263. Several collections of P.

pelliculosa sp. nov., in Herb. Pretoriae, were misdetermined as P. cremea.

glebulosa Bresadola; v. d. Byl (1929) l.c., p. 15 and in S.A. Journ. Sci. 23 (1926) 288.

nuda (Fr.) Bresadola; v. d. Byl (1929) l.c., p. 17. [As Corticium nudum Fr. recorded

in: Kalchbrenner in Grev. 10 (1881) 59; Wood in Rept. Natal Bot. Gard.

(1898) 17.]

PUNCTULARIA Patouillard.

1. Punctularia affinis (Berk, and Curt.) comb. nov.

Reticularia affinis B. and C. ! in Journ. Linn. Soc. Bot. 10 (1869) 347, Saccardo,

Syll. Fung. 7 (1888) 418.

Reticularia venulosa B. and C. ! in Journ. Linn. Soc. Bot. 10 (1869) 347.

Reticularia atro-rufa B. and C. ! in Journ. Linn. Soc. Bot. 10 (1869) 347; Saccardo,

Syll. Fung. 7 (1888) 419.

Thelephora atropurpurascens B. and Br. in Jour. Linn. Soc. Bot. 14 (1875) 64;

Saccardo, Syll. Fung. 6 (1888) 546.

Punctularia atropurpurascens (B. and Br.) Petch in Ann. Roy. Bot. Gard. Pera-

deniya 6 (1916) 160.

Trichosporium Curtisii Massee in Jour. Myc. 5 (1889) 185, t. 14, f. 3; Saccardo,

Syll. Fung. 10 (1892) 583.

Ceriomyces venulosus (B. and C.) Torrend in Bull. Soc. Portug. Sci. Nat. 4, p. 9.

Corticium ? tuberculosum Pat. in Bull. Soc. Myc. de Fr. 8 (1892) 118; Saccardo,

Syll. Fung. 11 (1895) 126.

Punctularia tuberculosa (Pat.) Pat. in Bull. Herb. Boiss. (1895) 57; Saccardo,

Syll. Fung. 14 (1899) 223; Patouillard, Essai sur les Hym. (1900) 57, f. 40.

Illustrations : Massee (1889) loc. cit. (as Trichosporium Curtisii)', Patouillard

(1900) loc. cit. (as Punctularia tuberculosa).

(conoidial stage); fig. 2 (perfect stage).

Plate 17, fig. 1 (conidial stage); fig. 2 (perfect stage).

(a) Conidial Stage: Loose, floccose, pulvinate, or irregular tufts of hyphae.

Coloured Dauphin’s violet and greyish violet-blue to dark plum-purple or bluish violet-

black (Ridgway). “ At first lavender with a white margin, then lavender to greyish

blue in the centre and reddish purple outwards. They finally collapse into a purple

brown mass of matted hyphae and spores.” — (Petch, l.c.).

Hyphae: frequently adherent in easily seen strands, 2-2-5-(3) jx diam., lightly coloured

in the microscope, purplish with the naked eye, thin-walled, with clamp connections

and septa. Some hyphae appear to be minutely sculptured.

Conidia: purple-brown, globose, ovate or somewhat elliptical, much variation in shape,

4 fi diam., 5-6-6 x 4 p, or 8 x 4 p; smooth. (Petch, loc. cit., states that the

conidia are minutely verrucose, at first borne terminally, then laterally.)

( b ) Perfect Stage: Not yet seen in South African material. The description

below is drawn from Petch’s Ceylon specimens in Herb. Kew.

Thin, effused, resupinate, occasionally radially grooved; centre subgelatinous;

margin byssoid; drying horny and crustose. “ General colour vinous, purplish in the

centre and reddish elsewhere; margin white; centre covered with close-set pulvinate

elevations, sometimes radially elongated.” — (Petch, l.c.). In section the abhymenial

26

layer is seen to consist of dark, fuscous hairs, arising from a dark basal seam. Hairs

thin-walled, fuscous, with clamps, 2-3-(4) g wide, densely interwoven. Middle layer

hyaline, with hyphae which are indistinct through gelatinisation. Upper layer light

brown, subgelatinous. No hymenial elements can be clearly distinguished. There is an

abundance of mineral matter in the troughs surrounding the papillae of the “ hymenium,”

and in concretions throughout the trama.

Specimens examined: 35015 (T.R.L. 2052), Simpson and Talbot, on Olea laurifolia,

Qudeni, Zululand, Feb. 1945; 30068, Bower, on Copaifera mopane, Louis

Trichardt, Tvl. ; T.R.L. 359, details missing; 33429, Phillips, Krugersdorp,

21/3/1942 (as Trichosporium ? purpureum Massee); 36862, Talbot, Fountains

Valley, Pretoria, 29/8/48.

As Punctularia atropurpurascens (B. and Br.) Petch: Conidial: Peradeniya,

Ceylon, Petch Oct. 1917; Petch (4328) 1914. Immature perfect stage:

Herb. Perad. (4676) 1915.

As Reticularia affinis B. and C. ! 3012, Car. Inf.; 454, Cuba, C. Wright', 1043,

Ravenel, S.C.

As Reticularia atro-rufa B. and C. ! Herb. Hook. 1867, Cuba; Brazil, sine loc.;

Glaziou, Rio de Janeiro, 1876 (8540); Cuba, C. Wright (534); Peradeniya,

Thwaites (17), Nov. 1867.

As Reticularia venulosa B. and Br. ! Cuba, C. Wright (675); Fungi Cubenses

Wrightiani (524).

As Punctularia tuberculosa (Pat.) Pat. on Olea Europea, R. Maire Myc. Boreali

Africani (112), 1913.

In S.A. Journ. Sci. 42 (1946) 133, Simpson and Talbot listed No. 35015 incorrectly

as Ptychogaster sp.

Reticularia pyrrhocreas ! is a fungus with a similar conidial structure, but with

larger, red-brown conidia. It is said to have been used by Red Indians as a face powder.

ALEURODISCUS Rabenhorst.

1. Aleurodiscus acerinus (Pers.) Hohnel and Litschauer var. longisporus Hohnel and

Litschauer ! in K. Akad. Wiss. Wien. Sitzungsb. 116 (1907) 805, PL 2, f. 4; van der

Byl in Ann. Univ. Stellenbosch 7 (1929) 25.

Aleurodiscus acerinus var. alliaceus (Quel.) Bourd. and Galz. in Bull. Soc. Myc. de

Fr. 28 (1912) 352; Pilat in Ann. Myc. 24 (1926) 217.

Corticium alliaceum Quelet, FI. Myc. de Fr. (1888) 5; Saccardo, Syll. Fung. 6

(1888) 629.

As “ Stereum acerinum Pers. var. longisporum Hohnel and Litschauer ” in Saccardo,

Syll. Fung. 21 (1912) 388.

Illustrations'. Hohnel and Litschauer (1907) loc. cit.

Plate 18.

Type : MacOwan (1074), Promont. bonae spei, Somerset East, in cortice arborum

varium, 1875.

Resupinate, crustose, adnate, chalk-white becoming pallid ivory-white, finely and

closely cracked in herbarium specimens; in section 80-100 g thick.

Basidia: clavate with a wavy outline, 35-45 x 5-7 g, not in a continuous palisade,

exceeded by the branched hyphae from beneath.

Spores: not seen (see note below).

Hyphae: 2 g or less in width, hyaline, erect, much branched and closely arranged in the

basal part, emerging as racemosely branched paraphyses between and beyond the

basidia; ^fery heavily encrusted with amorphous mineral matter.

27

Specimens examined: 21947, De Thumen Myc. Univ. (807) as Corticium calceum

Fr. var. lacteum Fr., MacOwan (1074); 20840, Fungi MacOwaniani No.

1074, as C. calceum Fr., In cortice vivo arborum frondosarum in sylvis ad

pedem montis Boschberg prope Somerset East. Febr. LXXV; 27566,

Louwrens, Fountains Valley, Pretoria, 16/4/1934; MacOwan, Cap. B. Sp.

9/83 ex Herb. Kalchbrenner in Herb. Kew.

The entire fructification is so heavily impregnated with minerals that treatment of

sections with strong HC1 is necessary in order to make out the structure. Sections

treated with HC1 and then rinsed with water and mounted on KOH-phloxine were

most satisfactory. Lactophenol will not dissolve the mineral matter.

The author was unable to find spores in the South African material which he

examined. The assignation of these specimens to the variety longisporus is therefore

in some doubt, but it seems reasonable in view of the fact that the Cape Province is the

type locality of the variety. The variety A. acerinus var. longisporus was founded on

de Thumen Mycotheca Universalis Exsicc. No. 807 (as Corticium calceum var. lacteum

Fr.), a collection of MacOwan from Somerset East, the distinction being purely a small

difference in spore measurement. (See Table 2.)

TABLE 2.

Spore Measurements of A. acerinus and its Varieties.

On the above observations of spore size and shape there appears to be an inter-

grading between A. acerinus and A. qcerinus var. longisporus which indicates that it

may not be practicable to hold them apart when more collections come to light.

A MacOwan collection from Somerset East also reached Kalchbrenner and was

recorded in “Fungi MacOwaniani” [Grev. 10 (1881) 59] as Corticium calceum Fr.

That collection, or part of it, is in Herb. Kew now. Montagne [in Ann. Sci. Nat.

3 ser., 7 (1847) 175] and de Thumen [in Flora 61 (1878) 353; ibid. 59 (1876) 363] also

record Corticium calceum or its variety lacteum from MacOwan’s Somerset East col-

lections. These collections may be summarised thus: — -

(a) MacOwan, Cap. B. sp. 9/83, determined as Corticium calceum Fr. by Kalch-

brenner. (Specimen seen, and is Aleurodiscus acerinus or var. longisporus.)

( b ) MacOwan (9452, b), Samteesvlakte prope Enon; published by Montagne as

C. calceum Fr.

(c) MacOwan (1032), Somerset East; published by de Thumen as C. calceum Fr.

( d ) MacOwan (1074), Somerset East; published by de Thumen as C. calceum

Fr. var. lacteum Fr., and later made the type of Aleurodiscus acerinus var.

longisporus by Hohnel and Litschauer.

From the Cape Province also, is a collection by Drege (Mus. Paris 9452) which

Leveille published as Thelephora ( leijostroma ) acerina Pers. [in Ann. Sci. Nat. ser. 3,

5 (1846) 150].

28

Corticium calceum Fr., even in the restricted sense of Romell and Burt [Burt in

Ann. Mo. Bot. Gard. 13 (1926) 203] has been shown by Rogers and Jackson [in Farlowia

1 (1943) 284] to be a nomen confusum which should be rejected. Most records of

“ Corticium calceum ” in South Africa probably refer to A/eurodiscus acerinus var.

longisporus. In van der Byl’s account of the South African Thelephoraceae [Ann.

Univ. Stellenbosch 7 (1929)], C. calceum and A. acerinus var. longisporus are dealt with

separately, and it is therefore not clear what fungus is represented by v. d. Byl’s concept

of C. calceum.

For future convenience in deciding whether A. acerinus var. longisporus should be

recognised as a distinct variety, the most important literature references to the species

A. acerinus are appended: —

Aleurodiscus acerinus (Pers.) Hohnel and Litschauer in K. Akad. Wiss. Wien

Sitzungsb. 116 (1907) 804; Bourdot and Galzin, Hym. de Fr. (1928) 334; Burt

in Ann. Mo. Bot. Gard. 5 (1918) 196; Pilat in Ann. Myc. 24 (1926) 216.

Corticium acerinum Persoon, Obs. Myc. 1 (1796) 37; Romell, Bot. Not. (1895) 71.

Thelephora acerina (Pers.) Pers., Syn. Fung. (1801) 581, Myc. Eur. 1 (1822) 152;

Fries, Syst. Myc. 1 (1821) 453, Hym. Eur. (1874) 648; Leveille in Ann. Sci. Nat.

ser. 3, 5 (1846) 150.

Stereum acerinum (Pers. ex Fr.) Fr., Epicrisis (1838) 554; Saccardo, Syll. Fung. 6

(1888) 587; Massee in Journ. Linn. Soc. Bot. 27 (1890) 202; Petch in Ann.

Roy. Bot. Gard. Perad. 6 (1916) 157.

Hypochnus acerinus (Pers.) Pat. in Rev. Myc. (1889) 166, Bull. Soc. Myc. Fr. 5

(1889) 30.

Illustrations'. Hhonel and Litschauer (1907) loc. cit., Tab. 2, f. 6.

ALEURODISCUS species recorded from South Africa but not available for

study :

cerussatus (Bres.) Hohnel and Litschauer; Van der Byl in Ann. Univ. Stellen-

bosch 7 (1929) 25.

disciformis (DC) Patouillard; Van der Byl (1929) loc. cit., p. 24.

PHLEBIA Fries.

1. Phlebia strigoso-zonata (Schwein.) Lloyd in Lloyd Myc. Writ. 4 (1914) L. 53: 15;

Burt in Ann. Mo. Bot. Gard. 8 (1921) 394.

Merulius strigoso-zonatus Schweinitz in Trans. Amer. Phil. Soc. n.s. 4 (1834) 160.

Auricularia strigoso-zonata (Schwein.) Lloyd (as “ McGinty ”) in Lloyd Myc.

Writ. 4 (1913) L. 46 : 6.

Phlebia rugosissima Lev., Champ. Exot. in Ann. Sci. Nat. ser. 2, 3 (1844) 214;

Saccardo, Syll. Fung. 6 (1888) 499.

Auricularia rugosissima (Lev.) Bres. in Ann. Myc. 14 (1916) 231.

Phlebia reflexa Berk, in Hook. Jour. Bot. 3 (1851) 168; Saccardo, Syll. Fung. 6

(1888) 500.

Auricularia reflexa (Berk.) Bres. (non A. reflexa Bulliard) in Ann. Myc. 9 (1911)

551 ; Lloyds Myc. Writ. 5 (1918) L.67: \2.

Phlebia rubiginosa Berk, and Rav. in Ravenel Fasc. iii: 23, and in Grevillea 1

(1873) 146; Saccardo, Syll. Fung. 6 (1888) 499.

Phlebia zonata Berk, and Curt, in Grevillea 1 (1873) 146; Saccardo, Syll. Fung. 6

(1888) 499.

Phlebia pileata Peck in 29th Ann. Rept. N.Y. State Mus. (1877) 45; Saccardo,

Syll. Fung. 6 (1888) 499.

29

Phlebia hispidula Berk, in Journ. Linn. Soc. Bot. 16: p. 167; Saccardo, Syll. Fung. 6

(1888) 499.

Auricularia sordescens Cesati, Myc. Born. (1879) 10; Saccardo, Syll. Fung. 6

(1888) 764.

Stereum lugubris Cooke in Grevillea 12 (1884) 85; Saccardo, Syll. Fung. 6 (1888)

574.

Auricularia Butleri Massee in Kew Bull. (1906) 94; Saccardo, Syll. Fung. 21

(1912) 441; Banerjee in Bull. Bot. Soc. Bengal 1 (1947).

Illustrations'. Plate 19.

Resupinate-reflexed, becoming dimidiate, sessile, with imbricate pileoli. Margin

strongly involute and of a lighter colour, yellow-ochre when freshly collected and

moistened. Abhymenial surface densely tomentose, dark reddish-brown, deeply con-

centrically furrowed. Hymenium black when dry, reddish-black and somewhat

gelatinous when moist, radially raised in minute ridges or pleats, concentrically furrowed

in formation corresponding to the upper surface. Context dark. Pileate parts 3-6 cms.

in length. Reported by one collector to have a faint cocoa smell when fresh.

Basidia: cylindric-clavate, (3)-3-5-4-3 x ( 1 6)— 3 1—36 g, in very tight palisade. None

seen with sterigmata.

Spores: not seen [“hyaline, obovoid, 9-10 x 5 g,” fide Bresadola in Ann. Myc. (1911)

loc. cit.]

Tissue differentiation: 600 g thick, excluding the much thicker tomentum ; Hymenium

composed of a thin dark layer of hyphae covered externally with minute dark

globose granules. Next to the tomentum is a thick, black zone. Middle layers

composed of fine filamentous hyphae, rather indistinct and gelatinised. (Bresadola,

loc. cit., states that the hyphae are 2-4-5 g wide, hyaline, with clamp connections.)

Surface hairs: fuscous; 3-5 g wide, rather thick-walled, with abundant clamp con-

nections.

Specimens examined: 27646, Rump, Town Bush, Maritzburg, 1934; Ledeboer,

on wattle stump, nr. Maritzburg, Jan. 1945; 36700, Talbot, on Acacia mollis-

sima stump, Lions River, Natal, 17/5/1948 ; 36695, Talbot, on Acacia mollissima,

Byrne, Natal, 24/5/1948.

As the basidia in this species are difficult to see and, when seen, are very rarely

mature, the genus has always been in doubt. Burt (1921) states that the species is a

Eubasidiomycete and cannot be placed in Auricularia. Bresadola described the basidia

as “ cylindric-subclavate, 30-35 x 4-5 g ” but placed the species in Auricularia. In

several specimens which the author examined in Kew and in Pretoria, no septate basidia

were seen, but admittedly all were immature. On account of the holobasidia he places

the species in Phlebia, while realising that the eflfuso-reflexed, subgelatinous pileus does

not fit in well with this genus.

The species is widely distributed in the Far East, Australia, New Zealand and

America, but is not found in Europe.

MERULIUS Haller ex Fries.

Key to species described : —

1. Hymenium light buff with fleshy tint when old; spores colourless, subcylindrical, about 2 -5-3 -5

X 6-7 • 5 p. M. corium ( 1 ).

2. Hymenium dark coloured when old, reddish brown; spores coloured, elliptical, about 5-6-5 x

8-11 -6 m-

a) Young hyphae often with pale lilaceous tint; hyphae with infrequent clamp connections and

frequent ampoule-swellings; no rapid colour change when moistened

M. himantioides (2).

aa) Young parts without lilaceous tint; hyphae with frequent clamps but no marked ampoules;

when moistened there is a rapid colour change from dresden brown to bright russet, also

becoming slightly gelatinous; hyphae frequently finely encrusted M. gelatinosusi 3).

30

1. Merulius corium (Pers. ex Fries) Fr., Elenchus Fung. 1 (1828) 58, Epicr. (1838)

500, Hym. Eur. (1874) 591; Rea, Brit. Basid. (1922) 620; Saccardo, Syll. Fung. 6

(1888) 413; Burt in Ann. Mo. Bot. Gard. 4 (1917) 322; Donk, Rev. Nederl. Hetero-

bas. en Homobas. — Aphyllophoraceae 1 (1931) 155; Kalchbrenner in Grev. 10

(1881) 57; Lloyd, Myc. Writ. 6 (1920) 952; Nel in Ann. Univ. Stellenbosch 20

(1942) 77.

Thelephora corium Persoon, Syn. Meth. Fung. (1801) 574; Greville, Scot. Crypt.

FI. 3 (1825) tab. 147.

Thelephora incarnata var. j3 Persoon, Myc. Eur. 1 (1822) 131.

Auricularia papyrina Bulliard, Champ, de la Fr. (1780-1784) PI. 402; Sowerby,

Eng. Fung. 3 (1803) tab. 349.

Merulius papyrinus (Bull.) Quelet, FI. Myc. de Fr. (1888) 32; Bourdot and Galzin,

Hym. de Fr. (1928) 347.

Boletus purpurascens De Candolle, FI. Franc. 6 (1815) 41.

Polyporus purpurascens (DC.) Persoon, Myc. Eur. 2 (1825) 60.

Illustrations'. Greville (1825) l.c., tab. 147; Burt (1917) l.c., f. 7.

Plate 20.

Resupinate, effused, at first orbicular, membranous-ceraceous, soft, thin, in section

300-500 p. Margin eventually lifting, narrow, showing a narrow, whitish, villose,

sometimes sulcate, reflexed surface. Hymenium in early stages and near margin

smooth, becoming poroid with small shallow reticulations, about 2 per mm., drying

buff with often a fleshy pink tint.

Basidia: cylindrical, narrow, densely aggregated, 2-5-4 x 24-33 p; sterigmata 2-4,

straight, narrow, up to 4 p long.

Spores: not seen in South African specimens; hyaline, subcylindric, smooth, 2 -5-3 -5

X 6-7 • 5 p (see note on spore size, below).

Hyphae: subhymenial hyphae thin walled, hyaline, branched, with numerous septa,

2-4 p wide, with or without accompanying mineral matter in the form of small

granules. Basal hyphae thicker — walled, hyaline, branched, septate, 3-6 / wide.

All hyphae very distinct and forming a fairly loosely interwoven network, lacking

clamp connections. The thick-walled hyphae appear to be merely older hyphae,

not a specialised skeletal system.

Specimens examined: 28885, Bottomley, The Cavern, Drakensberg, Natal, 20/7/37;

31725 (N.H. 391), Klapmuts, Cape, Dec. 1916; 26654, Stephens (158), Kuils

River, Cape, Sept. 1932; 1350, Doidge, Garstfontein, Pretoria Dist., 11/4/1911 ;

Rump (625), Table Mountain, Natal, 1935; Dr. Holub , “Interior of South

Africa,” (in Herb. Kew).

Merulius confluens Schweinitz is recorded for South Africa by Nel [in Ann. Univ.

Stellenbosch 20 (1942) 77], but the author has not yet been able to consult this specimen

in the van der Byl Herbarium. M. confluens is extremely near to M. corium and is held

as synonymous by Lloyd [in Lloyd, Myc. Writ. 3 (1909) 422 and ibid. 4 (1914) L. 52: 8]

to which opinion I subscribe after examining some of the American specimens of

M. confluens in Herb. Kew, not, however, including the Type, and numerous of my own

collections of M. corium. Burt (1917, l.c., p. 319) writes, “ M. confluens has the general

aspect of M. corium but is distinguished from that species by frequently a more broadly

reflexed margin, which is shallowly, concentrically sulcate when broadly reflexed, by

larger and usually deeper pores, by the incrusted hyphae of the subhymenial region

and by the small spores.”

As there is some variation in spore size between European and American specimens

of M. corium, Burt (l.c., p. 323) emphasises the incrustation of hyphae in M. confluens

as a clear-cut diagnostic character. This is a most uncertain character: in specimens

31

of M. corium collected in England, and those in Herb. Kew, there is often much mineral

matter in the subhymenium so that the hyphae appear encrusted. Spore size is not a

reliable character, for in M. corium it is very variable and is generally greater than that

of M. confluens except in America, where, according to Burt’s descriptions the spores

of the two have almost identical measurements. Spore sizes in these two species are

indicated in the accompanying table (see Table 3).

Macroscopic variations in M. corium are sufficient to embrace all the supposed

points in which M. confluens is held to differ. The author would suggest that typically

the two species represent extremes of a graded series, but without seeing authentic

material of them one cannot formally propose M. confluens as a synonym of M. corium.

TABLE 3.

Spore Sizes of Some Merulius Species.

2. Merulius himantioides Fries !, Syst. Myc. 1 (1821) 329, Epicr. (1838) 501, Hym.

Eur. (1874) 592; Saccardo, Syll. Fung. 6 (1888) 415; Burt in Ann. Mo. Bot. Gard.

4 (1917) 349; Rea, Brit. Basid (1922) 623.

Gyrophana himantioides (Fr.) Bourd. and Galz. in Bull. Soc. Myc. de Fr. 39 (1923)

'13, Hym. de Fr. (1928) 354.

Merulius tenuis Peck in N.Y. State Mus. Rept. 47 (1894) 147; Saccardo, Syll.

Fung. 11 (1895) 105 (fide Burt).

Merulius umbrinus Fries, Elench. Fung. 1 (1828) 61 [fide Lundell and Nannfeldt,

Fungi Exsicc. Suecici (1941) No. 1014].

Merulius squalidus Fries, Elench. Fung. 1 (1828) 62 [fide Lundell and Nannfeldt,

Fungi Exsicc. Suecici (1941) No. 1014]; Fries, Hym. Eur. (1874) 594; Saccardo,

Syll. Fung. 6 (1888) 420, ibid 23 (1925) 465; Rea, Brit. Basid. (1922) 623;

Bresadola in Ann. Myc. 18 (1920) 69; Lloyd, Myc. Writ. 4 (1914) L. 52: 26.

Illustrations'. Fries, leones Hym. (1877-1884) PI. 192, f. 1; Romell in Arkiv. for

Bot. II2 3 (1911) 28, P). 2, f. 19 (as M. himantioides).

Plate 21, fig. 1 .

Effused, resupinate, occasionally a little reflexed, dry, brittle. Hymenium spread

over gyrose plicate folds which are almost poroid in places, or elsewhere bluntly toothed,

irpiciform, or labyrinthiform; when dry, umber to warm Brussels Brown colour, when

moistened turning a warmer, darker colour. Margin thin and dingy white, pale buff,

or with a distinct lilac tinge, soft, sometimes produced in places into short rhizoidal

strands. Context pale coloured, somewhat floccose, in section 300-500 • thick.

Basidia: cylindric-clavate, (5 - 3)— 8—1 1 X (33)-44-(60) /i, with four sterigmata up to

5 p long.

1599-2

32

Spores: 5-6-5 x 8-10-6 g, light yellow colour, broad elliptic, often with one side

flattened, smooth, often 1-2 guttulate.

Hyphae: of two kinds, hyaline or coloured, the types differing only in colour. Hyphae

thin-walled, not encrusted, with frequent large clamp connections and occasional

H-anastomoses and ampoullar swellings, very frequently collapsed, closely inter-

texed, branched, very variable in width, 2-12 g wide.

Specimens examined: Timber Research Lab. (2090) on wattle pole near mouth of

an adit in Elands Drift Mine, Sabie, E. Tvl. (Dry bulb 60° F., wet bulb

47° F.), 1945; T.R.L. (361), Henderson, on old beam of wood on the ground

at Exchange Yard, Johannesburg, 1938; 36730, Louwrens, Wynberg Park,

Wynberg, 19/6/48.

As M. squalidus : Exsicc. W. Brinkmann, Westfalische Pilze iii: No. 121; 24861,

Stephens, Westbrooke, Newlands C.P., June, 1929.

Bresadola [in Ann. Myc. 18 (1920) 69] reduces M. squalidus to a synonym of M.

umbrinus which has page priority in the Elenchus Fungorum 1 (1828) 61. This treat-

ment is substantiated by Lundell and Nannfeldt (loc. cit.) who include both these species

as synonyms of M. himantioides, with the following notes: “The fresh colour of the

subiculum was almost lilacino-roseus. It seems certain that this species reaches full

development only in years with exceptionally large precipitation. Fries’ original water

colour drawing shows excellently a juvenile stage with margin still coloured, but the

reproduction in Icon. Sel. Tab. 193 : 1 has lost the colour and shows a white margin.

It is only with increasing age that the margin turns whitish. Authentic material in

Upsala Herb, shows M. squalidus Fr. and M. umbrinus Fr. to be synonyms. The former

species represents a young stage, and the latter was described from the old over-ripe

stage.”

The South African specimen No. 24861 (as M. squalidus) is a close match with the

Brinkmann Exsicc. No. iii: 121. Specimen T.R.L. 361 is a close match in all respects

with M. himantioides as represented in the British Collection of Herb. Kew. The

microscopic characters of T.R.L. 2090 are also the same, but macroscopically it has a

more robust growth form and a hymenium which is more distinctly toothed than usual.

M. himantioides differs from the better-known M. lacrymans most notably in

microscopic characters and habitat, but also in not being rather fleshy and rusty brown

in colour. The spores of M. lacrymans are paler, narrower [(4-3)-5-(6) x 1 0 //) and

flatter, and its coloured hyphae are always thick-walled or even without a visible lumen,

while its hyaline hyphae may be either thin- or thick-walled (see Plate 21, fig. 2).

M. lacrymans Wulf. ex Fr. has once been recorded in South Africa [v. d. Byl in

Trans. Roy. Soc. S.A. 10 (1922) 285] occurring under the linoleum of an old house in

Stellenbosch. This is an interesting record, for M. lacrymans is a fungus sensitive to

high temperatures and dry conditions, and is usually only to be found in temperate

regions. It is unlikely that M. lacrymans would flourish in the gold mines, for although

the relative humidity is high, it is accompanied by high temperatures [cfr. Brown in

S.A. Journ. Sci. 33 (1936) 383],

3. Merulius gelatinosus Lloyd ! in Lloyd, Myc. Writ. 7 (1922) 1 158, f. 2293, published

by typographic error in part as “ Merulius gelatinous ”; non M. gelatinosus Petch !

in Ann. Roy. Bot. Gard. Perad. 9 (1925) 315, published by typographic error as

“ Mesulius gelatinosus.”

Illustrations : Lloyd (1922) loc. cit., f. 2293.

Plate 22.

Resupinate, effused over old logs and forest debris. When dry, brittle, hymenium

Dresden brown, almost smooth with only the faint outline of reticulations. When

wetted, changing colour immediately to a bright warm russet and swelling into semi-

33

gelatinous convolutions of somewhat darker colour. Margin smooth, pale, yellowish

buff. Thickness in section up to 2 mm. when dry; context pale coloured.

Basidia: 6-5-8 x (36)-40-48 g, with 4 curved sterigmata 4-5 g long, clavate, hyaline.

Spores: (6) 6-5 x 10— ( 11-6) g, smooth, elliptical, bright yellow.

Hyphae: Subhymenial tissue pale coloured in mass but composed of hyaline and coloured

hyphae; hyphae thin-walled, irregular in outline, frequently septate, branched, with

frequent large clamp connections, 2-3-5 g wide. Nearer the base, the hyphae are

hyaline or quite deeply coloured, frequently much encrusted, septate, branched,

(2)-4-10 g wide, with clamp connections. The tissues merge into one another and

the three distinct layers of hyphae described by Lloyd (l.c.) were not seen.

Specimens examined: 31481, Duthie, on old logs and debris, Belvidere, Knysna,

C.P., July 1921. (Probably authentic material.)

The distinctive points about this species are (1) the almost smooth hymenium when

dry; (2) change of colour and texture on moistening (though the gelatinous consistency

should not be over emphasised); (3) thin-walled, coloured or hyaline hyphae, with

clamps, and encrusted in parts.

Except for the greater irregularity of outline, and encrustation, the hyphal characters

are like those found in M. himantioides, though in the latter the clamp connections are

usually rather infrequent. In M. lacrymans the hyphae lack clamps, are thick-walled

and coloured.

M. gelatinosus Petch ! is very different from M. gelatinosus Lloyd ! , and being a

later homonym the former species should be renamed.

Synonyms and excluded species in MERULIUS.

confluens Schweinitz — M. corium Pers. ex Fries, fide Lloyd in Myc. Writ. 3

(1909) 422 and ibid. 4 (1914) L. 52: 8.

squalidus Fries = M. himantioides Fries, fide Lundell and Nannfeldt, Fungi

Exsicc. Suecici (1941) No. 1014.

umbrinus Fries = M. himantioides Fries, fide Lundell and Nannfeldt, loc. cit.

MERULIUS species recorded for South Africa but not available for study:

confluens Schweinitz; Nel in Ann. Univ. Stellenbosch 20 (1942) 77.

lacrymans Wulf ex Fr. ; van der Byl in Trans. Roy. Soc. S.A. 10 (1922) 285; Nel

in Ann. Univ. Stellenbosch 20 (1942) 77.

serpens Fries; Kalchbrenner in Grevillea 10 (1881) 57.

CONIOPHORA DC. ex Persoon.

1. Coniophora papillosa Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 939, f. 1.

Illustrations : Plate 23, fig. 1.

Type : Timber Research Lab. (1852), on timber, Venterspost Gold Mining Co.,

' Ltd., 23/10/42. (In Herb. Kew.)

Resupinate, widely effused, adnate, circa 500 g thick, brittle, occasionally cracked

when dry. Hymenium dark drab colour, covered with small, hemispherical, discrete,

fertile papillae of superficial origin. Context fuscous.

Basidia: very rarely seen, not in palisade, clavate-cylindric, hyaline, 23-30 x 4-5 g,

at maturity projecting almost full length above the mass of tissue; sterigmata 2

or 4, short.

Spores: Elliptical or ovate, occasionally subglobose, coloured yellowish brown, smooth,

4-6 x 7-5-10 g, sometimes unilaterally flattened, rather thick-walled, sometimes

guttulate, very numerous and embedded in the upper parts of the trama, hyaline

when immature.

34

Tissue differentiation : Basal layers of trama composed of much interwoven, branched,

very dark coloured hyphae, 2 p wide, some dendroid but not markedly so. The

remainder of trama composed of lightly coloured to hyaline tissue which is indistinct

or in parts almost pseudoparenchymatous and is progressively darker from hymen-

ium to substratum.

Specimens examined: Type, T.R.L. (1852), cited above; 30194, Rump (393), on indigen-

ous wood in swamp bush, Compensation Beach, Natal, 1935.

Despite their very different habitats these two specimens are identical except that

no basidia could be demonstrated in the Natal specimen. This Coniophora is very

distinctive, particularly in the arrangement of the tissues. The hymenium has a

characteristic colour and is more markedly papillate than is seen in other species.

C. sistrotremoid.es (Schw.) Massee is described as papillate but is entirely different in

other respects.

2. Coniophora fodinarum sp. nov. (Etym. fodina = a mine).

Illustrations'. Plate 23, fig. 2.

Type; 27558, R. Lurie (T.R.L. 81) on underground timbers. New Modder Mine,

Johannesburg, 3/4/1934. Type in Herb. Kew, with Isotypes in Herb. Pretoriae

and Herb. University of Toronto.

Resupinate, widely effused, readily separable from the substratum, brittle when

dry, not cracking, up to 2 mm. in thickness. Hymenium very tubercular, coloured

olivaceous umber, yellowish or raw umber. Abhymenial surface soft, velvety, dark

brown. Context pale coloured. Rhizoidal strands may develop.

Basidia: not seen except in collapsed condition.

Spores: very variable in size, 5-6-6 -6 x 6 -6-7 -5-8 -3-10 p, and varying in shape from

subglobose to elliptical with one side often flattened, usually the latter shape;

coloured, smooth, relatively thin-walled, free or embedded among hyaline much

branched, filamentous hyphae, in a layer up to 260 p in thickness.

Hyphae: (1) of the spore-bearing layer: filamentous, hyaline, much branched, less than

1 p diam. (2) Of the middle layers (500-660 p thick), lightly coloured, 2 p diam.,

densely intertexed with more filamentous hyphae which form a vague tissue.

(3) Abhymenial hyphae, thin-walled, dark coloured, septate, not profusely branched,

3-(6) p diam., forming a spongy layer over 250 p in thickness.

Specimens examined: Type, 27558, cited above; 28857, R. Lurie, Randfontein

Estates, Johannesburg, Jan. 1937; 24871, Wi/man, on Acacia mollissima caps,

River View Mine, Wedburg, C.P., 25/7/1928; 28767, R. Brown (38), Jubilee Mine,

T.G.M.E., Pilgrims Rest, E. Tvl., 19/10/1936.

This species differs from C.puteana (Schum. ex Fr.) Karst, in being easily separable

from the substratum when dry, in its greater thickness, and particularly in the organisa-

tion of the hyphae and their deep colour in the basal layer. The spores are also very

much smaller than those of C. puteana [(5 - 3) 6 • 5—8 x 9-12-14-7 p in British collec-

tions]. The author has not seen examples of C.fumosa Karst., but the present speci-

mens do not agree with the plate of this in Fries, leones Tab. 198, f. 3, nor with the

description in Saccardo, Syl 1 . Fung. 6 (1888) 651. The especial differences are the

arachnoid-membranous, adherent, smooth features of C. fumosa. Rogers and Jackson

[in Farlowia 1 (1943) 273] reduce C. fumosa to synonymy with C. olivacea (Pers. ex Fr.)

Karst, which is cystidiate and quite different from the new species proposed here.

Dr. Jackson, of Toronto University, very kindly examined specimens of C. fodinarum

at my request, and confirmed that it was probably undescribed, and unlike any species

of Coniophora known to him. The species was formerly confused with C. puteana or

C. fumosa in South African Herbaria.

35

3. Coniophora olivacea (Fries) Karsten in Bidr. Kanned Finl. nat. o. Folk 37 (1882)

162; Saccardo, Syll. Fung. 6 (1888) 649; Massee in Journ. Linn. Soc. Bot. 25 (1889)

129; Bresadola in I.R. Accad. Agiati Atti iii, 3 (1897) 116; Burt in Ann. Mo. Bot.

Gard. 4 (1917) 257; Rogers and Jackson in Farlowia 1 (1943) 273.

Hypochnus olivaceus Fries, Obs. Myc. 2 (1818) 282, pro parte.

Corticium (Hypochnus) olivaceum Fries, Hym. Eur. (1874) 660 pro parte.

Corticium ( Coniophora ) olivaceum (Fr.) Cooke in Grevillea 8 (1880) 89.

Coniophorella olivacea (Fr.) Karsten in Bidr. kanned Finl. nat. o. Folk 48 (1889)

438; Bresadola in Ann. Myc. 1 (1903) 110; Bourdot and Galzin, Hym. de Fr.

(1928) 362; Rea, Brit. Basid. (1922) 628; Lundell and Nannfeldt, Fungi Exsicc.

Suecici. Uppsala. Fasc. XV-XVI (1939) No. 750.

Corticium leucothrix Berk, and Curt, in Grevillea 2 (1873) 4.

Corticium ( Coniophora ) leucothrix (B. and C.) Cooke in Grev. 8 (1880) 89.

Coniophora leucothrix (B. and C.) Cooke in Saccardo, Syll. Fung. 6 (1888) 648;

Massee in Journ. Linn. Soc. Bot. 25 (1889) 133.

Corticium brunneolum B. and C. in Grevillea 2 (1873) 4.

Corticium ( Coniophora ) brunneolum (B. and C.) Cooke in Grevillea 8 (1880) 88.

Coniophora brunneola (B. and C.) Cooke in Saccardo Syll. Fung. 6 (1888) 648;

Massee in Journ. Linn. Soc. Bot. 25 (1889) 134.

Hymenochaete Ellisii Berk, and Cooke in Grev. 4 (1876) 162.

Corticium ( Coniophora ) Ellisii (Berk, and Cooke) Cooke in Grevillea 8 (1880) 89.

Coniophora Ellisii (Berk, and Cooke) Cooke in Saccardo, Syll. Fung. 6 (1888)

648; Massee in Journ. Linn. Soc. Bot. 25 (1889) 129.

Coniophora fu/vo-o/ivacea Massee in Journ. Linn. Soc. Bot. 25 (1889) 134; Saccardo,

Syll. Fung. 9 (1891) 241.

Coniophora atrocinerea Karsten in de Thiimen Myc. Univ. No. 1806 (1881);

Soc. pro Faune et Flora Fennica Meddel. 6 (1881) 12; Finska Vet. -Soc. Bidrag

Natur. och Folk 37 (1882) 162; Saccardo, Syll. Fung. 6 (1888) 650; Massee in

Journ. Linn. Soc. Bot. 25 (1889) 132; Burt in Ann. Mo. Bot. Gard. 4 (1917)

260; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 17; Rogers and Jackson

in Farlowia 1 (1943) 273 (fide Rogers and Jackson loc. cit.).

Coniophorella atrocinerea Karsten, Finl. Basidsv. (1889) 438; Bourdot and Galzin,

Hym. de Fr. (1928) 364.

Coniophora fumosa Karsten Symb. 8 pg. 13; Saccardo Syll. Fung. 6 (1888) 651;

Bourdot and Galzin, Hym. de Fr. (1928) 360; Hohnel and Litschauer, Beitr. der

kennt. der Corticeen (1908) 16. (fide Rogers and Jackson, loc. cit.).

Corticium fumosum Fries pro parte. Fries, leones Fung. tab. 198, f. 3.

Above synonymy after Burt, and Rogers and Jackson.

Illustrations'. Trans. Brit. Myc. Soc. 6 pg. 73; Burt in Ann. Mo. Bot. Gard. 4

(1917) 258, f. 14.

Plate 24.

Resupinate, widely effused, byssoid to membranous, cracking or remaining entire

on drying, fragile, separable when fresh, somewhat adnate when dry, olivaceous, drying

isabelline or brownish. Margin paler and byssoid or arachnoid when distinct. In

section 100-150 p thick (up to 1,300 p thick according to v. d. Byl, l.c.).

Basidia: hyaline, cylindric, 6-5-10 x 39-45 p, with 2-4 sterigmata up to 8-5 p long.

A few hyaline, simple paraphyses present, 3 -5-4 -2 p wide.

Spores: yellow-brown, smooth, with a distinct apiculus, often 1-guttulate, elliptical with

one side often depressed and narrowing towards the apiculus, (5)-6-7-(8) x (10)-

1 2—1 3— ( 14) p.

36

Cystidia: numerous, cylindric with rounded apex and gradually tapered base, arising

as lateral branch of subhymenial or mid-tramal hypha, brownish with paler apex,

septate (almost hyaline when young and non-septate), relatively thin-walled,

covered with large detersile crystals partly soluble in KOH, projecting 52-110 /x,

in size 12-13 x 1 00— 1 50— (300) /x, usually about 140 fi long.

Hyphae: densely intertexed, clamps very rare, septate, not encrusted, much branched,

(2 • 8)— 4— (5 • 6) fi wide. Inferior hyphae brown, rather lax; superior hyphae brown,

compact, mixed with some hyaline hyphae in the subhymenium. No marked

differentiation of tissues.

Occurrence: usually on pine wood.

Specimens examined: 27665, Stephens (367), on base of pine stump near University

Course, Rondebosch, C.P., June 1934; 32086, Weintroub (T.R.L. 1158), on

badly decayed joist of flooring board, Exchange Yard, Johannesburg, 1940;

36823, A. A. Pearson ( Stephens 650) on stumps of Pinus sp., Groot Con-

stants Woods, 18/6/1948.

Burt (1917, l.c., pg. 239) indicates that there are many species connecting Conio-

phorella with Coniophora, and that the presence or absence of cystidia is not a good

character for segregating these genera. For this reason his rejection of the genus

Coniophorella is followed here, though the cystidia of C. olivacea are an extremely

distinctive aid to specific diagnosis. In this species the cystidia are modified lateral

branches of ordinary hyphae, and in form and origin resemble those of Peniophora

aspera q.v.

The South African collections cited above appear to be darker than British ones,

and their spores average slightly larger and slightly darker, especially in the epispore,

than in the British specimens. The writer does not consider that the differences are

sufficiently marked to warrant a new name, especially as he has not been able to compare

with the type of C. atrocinerea, which Rogers and Jackson reduce to synonymy with

C. olivacea. The spores in South African specimens are mummy-brown colour, 6 x

12 fi; those of British specimens are Saccardo’s umber, 5 x 11-4 ;±.

Note on Synonymy:

Burt (1917 l.c.) notes that an authentic specimen of Thelephora olivacea Fries

[Fries, Elenchus Fung. 1 (1828) 197; Epicr. Syst. Myc. (1836-38) 543; Berkeley, Outl.

Brit. Fung. (1860) 269] in Herb. Kew is Coniophora Betulae, and not the present species.

Bourdot and Galzin [in Hym. de Fr. (1928) 360] note: “ C. fumosa Karst., after

an authentic specimen, is, according to Hohnel and Litschauer (Beitr. 1906, p. 26), a

form of C. arida with much smaller spores, 9-10 x 6-7 //. The original of the same

C. fumosa does not differ from C. olivacea, according to the same authors (Beitr. 1908,

P- 16).”

Note on other CONIOPHORA species recorded for South Africa.

(1) C. puteana (Schum. ex Fr.) Karst. [= C. cerebella (Pers.) Duby], Mrs. R.

Brown (R. Lurie) records in S.A. Journ. Sci. 33 (1936) 388 that several collections of this

species have been made in the gold mines of the Central and East Witwatersrand, at

temperatures between 68°-72° F. These collections have not all been seen by the

writer, but he feels that they may prove to be C. fodinarum, which comes from mines

and has frequently been mistaken for C. puteana.

(2) In Herb. Pretoriae No. 30633 is a Coniophora collected from the wall of a dairy

by K. Morgan, Hopevale, Donnybrook, Natal, 6/5/1939. This fungus is characterised

by a very heavy encrustation of minerals on the hyphae which is suggestive of Coniophora

Betulae (Schum.) Karst, but as good material of that species is lacking Kew I am unable

to make a satisfactory comparison. Superficial comparison suggests that the hymenium

37

of No. 30633 is much too dark (dark red-brown) and that the trama is too thick and

membranous. There is also some doubt whether C. Betulae is not merely a variation

from C. suffocata (Peck) Massee. [See Rogers and Jackson in Farlowia 1 (1943) 280.]

(3) Miss D. Weintroub has submitted to us several specimens of Coniophora

collected from mires, which at present cannot be named.

STEREUM Persoon ex S. F. Gray.

Key to species described: —

1. Fresh plant exuding reddish juice when wounded: conducting organs seen microscopically:

a) Thin; reflexed surface clothed with short silky hairs; spores (7)— 8— (9) X 3-3 -5 on conifers

S. sanguinolentum (1).

aa) Thick; conspicuously rimose; reflexed surface clothed with a thick pad-like tomentum;

spores 3 -5-5 -5 x 2-3 /<.; on indigenous woods S. rimosum var. africanum (2).

2. Fresh plant not “ bleeding ”; conducting organs absent.

a) Gloeocystidia abundant, hyaline; hymenium pallid; abhymenial surface snuff-brown; spores

3-4 -5 X 2-3 /x S. bicolor (3).

aa) Gloeocystidia absent:

x. Large encrusted (rarely smooth) peniophoroid cystidia present:

y. Cystidia hyaline, or faintly yellowish, not arising from skeletal hyphae;

spores 10—13 x 6-7 /x S. cinerascens (4).

yy. Cystidia dark coloured (sometimes nearly smooth), arising as apical modifi-

cations of skeletal hyphae; spores 6-8 X 3-4 ix S. umbrinum (5).

xx. Peniophoroid cystidia absent:

Cylindrical, rugose, coloured “ false setae ” arising from skeletal hyphae present,

darkening in potash; spores 6-7 X 3-4 fx S. Schomburgkii (6).

1. Stereum sanguinolentum (A. and S.) Fries, Epicrisis (1838) 549, Hym. Eur. (1874)

540; Albertini and Schweinitz, Consp. Fung. (1805) 274 (under /?. Sterea of Thele-

phora)-, Schweinitz in Naturforsch. Ges. Leipzig Schrift 1 (1822) 106; Berkeley,

Outl. Brit. Fung. (1860) 271; Saccardo, Syll. Fung. 6 (1888) 564; Quelet, FI. Myc.

de Fr. (1888) 14; Karsten, Myc. Fenn. 3 (1876) 306, Finnlands Basidsv. (1887)

396; Cooke, Brit. Fung. Handbook 1 (1871) 317; Massee in Journ. Linn. Soc. Bot.

27 (1890) 189; Greville, Scot. Crypt. Flora 4 (1826) t. 225; Bresadola in Ann. Myc.l

(1903) 92; Hohnel and Litschauer, Weisner Festschr. (1908) 60; Herter, Pilze in

Krypt. — FI. der Mark Brand. 6 (1910) 125; Brinkmann, 44 Jahresber. Westf. Prov.

Ver. f. Wiss. (1916) 34; Burt in Ann. Mo. Bot. Gard. 7 (1920) 145; Bourdot and

Galzin in Bull. Soc. Myc. de Fr. 37 (1921) 109, Hym. de Fr. (1928) 373; Pilat in

Hedwigia 70 (1931) 59; Velenovsky, Ceske houby (1922) 761; Rea, Brit. Basid.

(1922) 663.

Thelephora sanguinolenta Alb. and Schwein., Consp. Fung. (1805) 274; Fries,

Syst. Myc. 1 (1821) 440; Elenchus Fung. 1 (1828) 178; Greville, Scott. Crypt.

FI. 4 (1826) t. 225.

Thelephora hirsuta Persoon, Synopsis (1801) 570, (fide Pildt).

Thelephora sericea p Persoon, Myc. Eur. 1 (1822) 117, (fide Pilat).

Stereum crispum Quelet, Assoc. Franc. 18 Suppl. (1891) 2; Schroeter, Pilze Schl.

in Kohns Krypt. — FI. v. Schl. 3 (1889) 427; Herter in Krypt. — FI. der Mark

Brand. 6 (1910) 124, (fide Pilat).

Stereum balsameum Peck in N.Y. State Mus. Rept. 27 (1875) 99, ibid. 30 (1879)

75; Saccardo, Syll. Fung. 6 (1888) 584; Massee in Journ. Linn. Soc. Bot. 27

(1890) 196, (fide Pilat).

Stereum balsameum f. reflexum Peck, N.Y. State Mus. Rept. 47 (1894) 152, (fide

Pilat).

38

Stereum rigens Karsten in Finska Vet.-Soc. Bidrag Natur och Folk 37 (1882)

243, ibid. 48, p. 396; Saccardo, Syll. Fung. 1 1 (1895) 121 ; Pilat in Bull. Soc. Myc.

de Fr. 42 (1926) 109, Hedwigia 70 (1931) 62 (as f. rigens Karst, pro sp.); fide Burt.

Illustrations : Greville (1826) l.c., PI. 225; Burt in Ann. Mo. Bot. Gard. 7 (1920)

145, f. 20.

Plate 25.

Flesh exuding reddish juice when wounded in fresh state. Coriaceous, thin,

resupinate, effused, becoming narrowly reflexed, orbicular-confluent; margin acute,

pallid. Hymenium wood-brown becoming darker, smooth or cracking rimosely,

often zonate. Surface villose with short, adpressed, silky hairs, zonate and striate,

some tint of buff. In section 400-500 g thick, excluding surface hairs.

Basidia: Clavate, 25-40 x 4 -5-6 -5 g.

Spores: hyaline, smooth, cylindric, unilaterally depressed, (7)-8-(9) X 3-3-5 g.

Conducting organs: conspicuous, reddish-brown, numerous, in the intermediate

tissues and curving upwards to the hymenium, 3-4 g wide, very occasionally forked.

Surface hairs: simple, thick-walled, agglutinated, short, adpressed, 4-5 g wide.

Specimens examined: 28933, Doidge and Morgan, on dead coniferous wood,

Donnybrook Forest, Natal, Feb. 1935.

This specimen is more reflexed than is usual in British material, but the writer feels

no doubt as to its identity. Stereum rimosum var. africanum, the common “ bleeding ”

Stereum of South Africa, is much more reflexed, thicker, has yellow conducting vessels,

and spores about half the size of those in S. sanguino/entum. The thin, almost papery

texture of the latter is at once a characteristic difference. S', sanguinolentum is usually

a North Temperate species, and was possibly introduced to South Africa with conifers

imported for plantations.

2. Stereum rimosum Berk. var. africanum Talbot apud Wakefield and Talbot in

Bothalia 4 (1948) 945, fig. 5.

Stereum adnatum Lloyd in Myc. Writ. 7 (1925) 1336, f. 3093.

Effused, resupinate-reflexed, sometimes sessile, attached by a broad umbo, or

composed of several connate, resupinate-umbonate pilei, each roughly circular in out-

line, coriaceous. Surface cinnamon-buff colour, concentrically furrowed, covered

with a thick, felty, pad-like tomentum. Margin even or lobate. Hymenium rimose,

showing pallid, silky context, often concentrically zoned, warm buff or pinkish buff,

when older becoming a darker vinaceous or cinereous colour. Almost certainly a

“ bleeder ” when fresh.

Basidia: closely aggregated, 4-4-5 g wide at apex.

Spores: hyaline, smooth, elliptic-ovate, with one side frequently depressed, and a small

attenuated apiculus, 2-3 x 3-5 5-5 g.

Conducting organs: yellow, 5 -5-8 -5 g wide, with rigid walls, in a layer about 200 g

wide, distributed in the subhymenium and curving upwards into the hymenium,

not emergent.

Hyphae: thin-walled, hyaline, frequently septate, 3-5 g wide. Trama bordered next

to the tomentum by a narrow orange coloured zone. Width in section, excluding

tomentum, 700-1,000 g.

Surface hairs: thick-walled, very much intertwined, almost hyaline to pale yellowish,

4-2 g in width.

Specimens examined: Type, 30233, Rump (486), Karkloof, Natal; 30268, Morgan

and Doidge, Donnybrook, Natal, 1936; 30777, Bower, Entabene, Louis

Trichardt, Tvl. 24/8/38 (bleached and insect-eaten); 28285, Rump (205),

Town Bush, Maritzburg, Oct. 1934; Uganda; T. D. Maitland (460) ,1919 and

(19. A) 1915; The following under Stereum adnatum in Herb. Kew: 27755,

Rump (98), Hilton Road, Natal, Aug. 1934; 28296, Rump (222) Town Bush,

Maritzburg, Oct. 1934; 28303, Rump (232), Town Bush, Maritzburg, Oct. 1934.

39

Massee [in Journ. Linn. Soc. Bot. 27 (1890) 187] quotes the spores of the species

S. rimosum as “ globose, 6-7 g diam.” I have examined the type specimen in Herb.

Berkeley (Darjeeling, 7,500 ft.) at Kew, and was unable to confirm the presence of such

basidiospores although there were large numbers of globose hyphomycete spores,

4-6-5 g diam., echinate and faintly coloured, which were conclusively traced to conidio-

phores of an Aspergillus sp. Should basidiospores of the sort described by Massee

be found in the species S. rimosum, it will be necessary to alter the variety africanum

to specific rank. Apart from the above consideration of spores, the species and the new

variety are identical in microscopic structure but differ macroscopically. Whereas the

species is usually more or less pileate or sessile-umbonate, the variety is more resupinate-

reflexed and has a tomentum of much greater thickness. The variety has a more

markedly zoned, lighter coloured hymenium than the almost livid hymenium of the

species.

Lloyd (loc. cit.) described Stereum adnatum from South Africa, but stated that it

lacks distinctive cystidia, ducts and vesicular cells. The writer has recently seen a

co-type of Lloyd’s species (No. 1708, Doidge, on dead branch, Zoutpansberg, Tvl.,

4/8/191 1) which possesses ducts and is identical with S. rimosum var. africanum. It

is considered that the fungus only merits varietal rank, and consequently S. adnatum

is reduced to synonymy. C. H. Humphrey (in Litt., Division of Botany File No.

M. 31/86/1) considers that S. rimosum and S. adnatum are identical.

3. Stereum bicolor (Pers. ex Fries) Fries, Epicrisis (1838) 549, Hym. Eur. (1874) 640;

Saccardo, Syll. Fung. 6 (1888) 565; Massee in Journ. Linn. Soc. Bot. 27 (1890)

177; Rea, Brit. Basid. (1922) 665; Kalchbrenner in Grev. 10 (1881) 58; van der Byl

in Trans. Roy. Soc. S.A. 10 (1922) 153; Wakefield in Det. Kong. Norske Vidensk.

Selsk. Forh. 9 (1936) 52.

Thelephora bicolor Persoon, Syn. Meth. Fung. (1801) 568, Myc. Eur. 1 (1822) 122;

Fries, Syst. Myc. 1 (1821) 438.

Lloydella bicolor (Pers. ex Fr.) Bresadola in Lloyd Myc. Writ. 1 (1898) 51.

Stereum fuscum (Schrader) Quelet, Flor. Myc. de Fr. (1888) 14; Karsten, Finn.

Basidsv. (1889) 397; Bresadola, Fung. Kmet. (1897) 106; Burt, in Ann. Mo. Bot.

Gard. 7 (1920) 117, PL 4, f. 26; Bourdot and Galzin, Hym. de Fr. (1928) 382;

van der Byl in Ann. Univ. Stellenbosch 7 (1929) 41; Pilat in Hedwigia 70

(1930) 92.

Thelephora fusca Schrader, Spic. FI. Germ. (1794) 184; Persoon, Syn. Meth. Fung.

(1801) 568, Myc. Eur. 1 (1822) 122; Fries, Syst. Myc. 1 (1821) 438.

Stereum coffeatum Berk, and Curt. ! in Grevillea 1 (1873) 164; Saccardo, Syll.

Fung. 6 (1888) 568; Massee in Journ. Linn. Soc. Bot. 27 (1890) 190.

Thelephora vinosa Persoon, Syn. Meth. Fung. (1801) 578, (fide Pilat).

Stereum pannosum Cooke ! in Grevillea 8 (1879) 56; Massee in Journ. Linn. Soc.

Bot. 27 (1890) 185 (non Thelephora pannosa Sow. ex Fr.).

Illustrations : Burt (1920) l.c., PI. 4, f. 26; Fries, leones Hym. PI. 197, f. 2; Karsten,

leones Hym. PL 2, f. 9.

Plate 26.

Occasionally entirely resupinate, usually resupinate-reflexed, sometimes pileate,

imbricate, soft spongy texture. Abhymenial surface snuff-brown, concentrically

furrowed, floccose, becoming smooth. Margin paler. Hymenium smooth, whitish

to creamy, not furrowed, sometimes rimose when dry, rather velvety. In section up

to 1,000 n thick; hymenial layer hyaline, about 100 g thick, the rest of the tissue coloured

brownish. Not adnate.

40

Basidia: 3 -5-5 -5 X 22-27 p, cylindric, with 2 or 4 sterigmata.

Spores: hyaline, smooth, elliptical, unilaterally depressed, or oblong, frequently guttu-

late, 3-4-5 X 2-3 p.

Gloeocystidia : very abundant in the hymenium, not usually emergent, hyaline, very

refractile, thin walled, cylindric, fusoid or somewhat ventricose, averaging 4-11

X 90 p, borne on very narrow hyaline hyphae about 2 p wide, frequently fragmented

when old.

Hyphae: all smooth, thin-walled, with occasional to numerous clamp connections,

much branched, septate; subhymenial hyphae colourless to pale brownish, 1 -5-3 p

wide; tramal hyphae brown, 3-6 pi wide, loosely intertexed.

Specimens examined: v. d. Byl (2239), South Africa, sine loc. ; 27596, Rump (12),

Maritzburg, 1934; Read (T.R.L. 200), Rooikoppies, Duiwelskloof, N.Tvl.,

Apr. 1937; MacOwan (1244), C.B.S. 9/83, in Herb. Kew; Thorold (77), Njoro,

Kenya, 1923; Farquharson (48), Eket Distr., S. Nigeria; Maitland (451),

Uganda, March 1919; Maitland, Victoria Nyanza, Uganda, 14/7/1914.

Note on Nomenclature:

In the works of Persoon and Fries prior to the “ Epicrisis ” (1838), the species was

placed in the tribe Stereum of the genus Thelephora, and T.fuscum Schrad. was quoted

as a synonym, indicating that the authors recognised it as the same plant but had re-

named it, no doubt because the epithet bicolor is so much more descriptive of the species.

Art. 40 of the International Rules of Botanical Nomenclature states: “ The name of a

taxonomic group is not validly published when it is merely cited as a synonym.” Further,

as nomenclature of this group starts with Fries’ (1821-32) “ Systema ”, the name of

the species in question must be taken as 5. bicolor (Pers. ex Fr.) Fr., and not as S.

fuscum (Schrad.) Quel, as is most frequently done.

Stereum pannosum Cooke ! appears to be the same as S. bicolor in all respects.

Bresadola [in Ann. Myc. 14 (1916) 232] gives the former as a synonym of Lloydella

Beyrichii (Fr.) Bres., which was published as Thelephora Beyrichii by Fries in Linnea

5 (1830) 529. Bresadola’a definition of the genus Lloydella [in Lloyd, Myc. Writ. 1

(1899) 51] was, “ Stereum possessing cystidia,” but he included in it species such as

S. cinerascens (with large encrusted cystidia), S. bicolor (with gloeocystidia) and later

S. Schomburgkii (with “ false setae ”), and it is thus a highly artificial genus not worthy

of retention.

4. Stereum cinerascens (Schw.) Massee ! in Journ. Linn. Soc. Bot. 27 (1890) 179;

Burt in Ann. Mo. Bot. Gard. 7 (1920) 203; Bourdot and Galzin, Bull. Soc. Myc.

de Fr. 37 (1920) 203, Hym. de Fr. (1928) 376; Pilat in Hedwigia 70 (1931) 83;

van der Byl in Ann. Univ. Stellenbosch 7 (1929) 43.

Thelephora cinerascens Schweinitz in Amer. Phil. Soc. Trans. N.S. 4 (1832) 167.

Hymenochaete cinerascens (Schw.) Leveille in Ann. Sci. Nat. Bot. ser. 3: 5 (1846)

152.

Peniophora cinerascens (Schw.) Saccardo, Syll. Fung. 6 (1888) 646.

Lloydella cinerascens (Schw.) Bresadola in Lloyd Myc. Writ. 1 (1901) 51.

Peniophora Schweinitzii Massee in Journ. Linn. Soc. Bot. 25 (1889) 145.

Corticium aschistum Berk, and Curt. ! in Amer. Acad. Arts and Sci. Proc. 4 (1858)

123; Berkeley in Grevillea 2 (1873) 3.

Peniophora Berkeley i Cooke ! in Grevillea 8 (1879) 20; Saccardo, Syll. Fung. 6

(1888) 642; Massee in Journ. Linn. Soc. Bot. 25 (1889) 144.

Stereum moricola Berk. ! in Grevillea 1 (1873) 162; Saccardo, Syll. Fung. 6 (1888)

567.

Peniophora moricola (Berk.) Massee in Journ. Linn. Soc. Bot. 25 (1889) 141.

41

Stereum dissitum Berk. ! in Grevillea 1 (1873) 64.

Peniophora dissita (Berk.) Cooke in Grevillea 8 (1880) 150; Saccardo, Syll. Fung.

6 (1888) 645; Massee in Journ. Linn. Soc. Bot. 25 (1889) 143.

Corticium ephebium Berk, and Curt. ! in Grevillea 1 (1873) 178; Saccardo, Syll.

Fung. 6 (1888) 618.

Peniophora ephebia (B. and C.) Massee in Journ. Linn. Soc. Bot. 25 (1889) 131.

Stereum neglectum Peck, N.Y. State Mus. Rept. 33 (1880) 22.

Peniophora neglecta (Peck) Peck, N.Y. State Mus. Rep. 40 (1887) 76.

Peniophora occidentalis Ellis and Everh. ! in Bull. Torrey Bot. Club 24 (1897)

277 ; Saccardo, Syll. Fung. 14 (1900) 224.

Lloydella occidentalis (Ellis and Everh.) Hohnel and Litschauer in K. Akad. Wiss.

Wien. Sitzungsb. 116 (1907) 791.

Stereum purpurascens Lloyd, Myc. Writ. 4 (1914) L. 53: 15.

Hymenochaete bonariensis Speg. (fide Bresadola, Sel. Myc. ii. Trento. 1916).

Above Synonymy after Pilat, loc. cit.

Illustrations : Cooke in Grev. 8 (1879) PI. 122, f. 4; Burt in Ann. Mo. Bot. Gard.

7 (1920) f. 36 and PI. 6, f. 64.

Plate 27.

Coriaceous, resupinate, not adnate, effused, sometimes with a narrowly reflexed

margin. Abhymenial surface ochraceous, ashen or warm buff, with sharp hairs, con-

centrically furrowed. Flymenium drab or pinkish buff, smooth, sometimes slightly

scabrid to touch. In section 250 -400-800 g thick. Cystidia plainly visible with a

hand lens.

Basidia: 40 50-(60) x 9-10 g, clavate.

Spores: hyaline, smooth, oblong-cylindric, often with one side depressed, 6-7 x 10-13/z.

Cystidia: embedded or occasionally projecting, heavily encrusted with large crystals,

thick-walled, often faintly coloured at the base, conical, 1 00— ( 1 50) x 12-24 g.

Hyphae: 3 -5-4-4 -5 g wide, hyaline to very faintly coloured, rather opaque and stout-

walled, those adjacent to the substratum forming a narrow, dense, yellow-brown

layer.

Abhymenial hairs: much intertwined, like the hyphae but dark coloured.

Specimens examined: v. d. Byl (2732), Table Mountain, Natal, June 1930; 28688,

Rump (280), Town Bush, Maritzburg, Nov. 1934; 28498, Rump (255), Town

Bush, Maritzburg, Oct. 1934; 34377, Rump, Town Bush, Maritzburg, 1936;

35421, Rump (700), Isipingo Beach, Natal, 1944; 31368, Duthie, Knysna;

28926, no details of collection; Maitland (114), Uganda (as Peniophora

occidentalis in Herb. Kew.).

The marked affinity of S. cinerascens to Lopharia mirabilis ! is discussed elsewhere

(page 57). On casual examination S. cinerascens might be taken for a Peniophora,

but it is distinguished even when wholly resupinate by a Stereum distribution of tissues,

and by its loose attachment to the substratum by distinct hairs.

Peniophora flavido-alba Cooke !, an American species, is very near to S. cinerascens

but its spores are only 4-5-6 x 2 -5-3 -5 g, and also it cracks a lot on drying. The

small spores of this species come within the range of Peniophora Roumeguerii which is

somewhat allied in structure, but is nevertheless quite distinct.

5. Stereum umbrinum Berk, and Curtis ! in Grevillea 1 (1873) 164; Wakefield in

Kew Bull. (1915) 369; Burt in Ann. Mo. Bot. Gard. 7 (1920) 191 ; Pilat in Hedwigia

70 (1930) 94; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 42; Bourdot and

Galzin in Bull. Soc. Myc. de Fr. 37 (1921) 123, Hym. de Fr. (1928) 382.

42

Thelephora crassa Leveille in Gaudichand, Voyage Bonite Bot. 1 (1846) 190,

PI. 139, f. 1 [non Stereum crassum Fries, R. Soc. Sci. Upsal. Actis. 3 (1851) 111].

Hymenochaete crassa (Lev.) Berk, apud Cooke in Grevillea 8 (1880) 148; Saccardo,

Syll. Fung. 6 (1888) 597; Massee in Journ. Linn. Soc. Bot. 27 (1890) 114.

Hymenochaete umbrina B. and C. apud Cooke in Grev. 8 (1880) 148; Saccardo,

Syll. Fung. 6 (1888) 598; Massee in Journ. Linn. Soc. Bot. 27 (1890) 113.

Hymenochaete vinosa (Berk.) Cooke ! in Grev. 8 (1880) 149; Saccardo, Syll.

Fung. 6 (1888) 600.

Hymenochaete multispinulosa Peck in Bot. Gaz. 7 (1882) 54; Saccardo, Syll.

Fung. 6 (1888) 600; Massee in Journ. Linn. Soc. Bot. 27 (1890) 108.

Hymenochaete scabriseta Cooke ! in Ravenel, Fung. Amer. (1882) 717; Massee

in Journ. Linn. Soc. Bot. 27 (1890) 113, PI. 5, f. 7.

Lloydella scabriseta (Cooke) Hohnel and Litschauer in K. Akad. Wiss. Wien.

Sitzungzb. 115 (1906) 1580.

Hymenochaete purpurea Cooke and Morgan ! apud Cooke in Grev. 1 1 (1883)

106; Morgan in Cicinnati Soc. Nat. Hist. Journ. 10 (1888) 198; Saccardo, Syll.

Fung. 6 (1888) 597; Massee in Journ. Linn. Soc. Bot. 27 (1890) 115.

Kneiffia purpurea (Cooke and Morgan) Bresadola in Ann. Myc. 1 (1903) 100.

Peniophora intermedia Massee in Journ. Linn. Soc. Bot. 25 (1889) 143; Saccardo,

Syll. Fung. 9 (1891) 238.

Hymenochaete Kalchbrenneri Massee ! in Journ. Linn. Soc. Bot. 27 (1890) 116;

Saccardo, Syll. Fung. 9 (1891) 230.

Above synonymy after Pilat.

Illustrations : Burt (1920) l.c., f. 31; Massee in Journ. Linn. Soc. Bot. 27 (1890)

PI. 5, f. 7.

Plate 28.

Resupinate, effused, sometimes narrowly reflexed, never pileate; margin shortly

villose. Context soft, spongy. Hymenium velutinous, cracking but little in drying,

sometimes pitted, umber, vinaceous purple, purple-brown, light sandy-brown, or

snuff-brown in colour.

Basidia: hyaline or very faintly coloured, about 6 X 30 p.

Spores: cylindrical to ellipsoid, hyaline, smooth, 3-4 x 6-8 p.

Cystidia: originating in basal or middle part of the trama, curving upwards into hymen-

ium and frequently projecting 10-20 p beyond; dark yellow-brown, in young

stages lighter colour, not very thick-walled, encrusted or rugose especially near

the apex, rarely quite smooth, 100-250 X 7-9 p, cylindric-clavate or fusoid, arising

as apical modifications of skeletal hyphae.

Hyphae: lightly coloured, 3-4-5 p wide, in a loose network, lacking clamp connections,

not forming a distinct intermediate layer of tissue.

Specimens examined: Type in Herb. Berk, at Kew; 22044, P. MacOwan (1054),

Boschberg, Somerset East, 1876 (as Peniophora cinerea in Herb. Pretoriae);

van der By 1 (2737), Izotsha, Natal, Feb. 1930; 30220, Rump (433) Umgeni

Bush, Durban; 27626, Rump , Cato Ridge, Natal, 1934; 34357, Rump, Table

Mountain, Natal, 1935; 34393, Rump (613), Table Mountain, Natal, July

1935; 35419, Rump (679), on Quercus, Maritzburg, 1943; 34381, Rump ,

Town Bush, Maritzburg, May 1936; 33392, 33400, Scott, on Popu/us, Pretoria,

March 1942; 36839, Talbot, Garstfontein, Pretoria, 21/8/48; 36710, Talbot,

Kloof Falls, Natal, 31/5/48; the following from Town Bush, Maritzburg,

Aug.-Nov., 1934 viz. 28294, Rump (218); 28702, Rump (305); 28276, Rump

(208); 28277, Rump (209); 21161, Rump (122).

43

The cystidia of this species are apical modifications of skeletal hyphae, reminiscent

of the origin of the “ false setae ” in the genus Duportella. They do not darken in

potassium hydrate and are paler and differ in origin and morphology from the true

setae of an Hymenochaete, but in some specimens they lack conspicuous encrustation

or roughness. This is one of the few species of Stereum lacking a distinct layer of

horizontal hyphae as an intermediate or basal tissue. Its closest affinity is with Stereum

papyrinum Mont. ! (= S. membranaceum Fr.) which may be distinguished by frequently

being pileate and always possessing wider, more encrusted, more peniophoroid cystidia

with thicker walls.

Most Natal specimens of S. umbrinum have a purplish tinge in place of the umber

colour, and their tissues are rather more loosely interwoven than in typical North

American specimens. But the fungus does vary greatly in texture and colour, as is

beautifully demonstrated in a set of E. A. Burt's specimens in Herb. Kew., illustrating

the synonymy of the species.

No. 34393 has hyaline cystidia, but is otherwise indistinguishable from the purple

forms of this species.

6. Stereum Schomburgkii Berkeley ! in Journ. Linn. Soc. Bot. 13 (1873) 168; Saccardo,

Syll. Fung. 6 (1888) 568; Wakefield in Kew Bull. (1914) 259; Lloyd, Myc. Writ.

6 (1919) 960; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 43; Bannerjee in

Journ. Ind. Bot. Soc. 14 (1935) 33.

Lloydella Schomburgkii (Berk.) Bres. var. brunnea Bresadola in Ann. Myc. 18

(1920) 45; Saccardo, Syll. Fung. 23 (1925) 534.

Stereum atrocinereum (Massee) v. d. Byl in Ann. Univ. Stellenbosch 7 (1929) 44.

Peniophora atrocinerea Massee ! in Journ. Linn. Soc. Bot. 25 (1889) 141.

Stereum retirugum Cooke ! in Proc. Roy. Soc. Edinb. (1882) 456; Massee in

Journ. Linn. Soc. Bot. 27 (1890) 186; Saccardo, Syll. Fung. 23 (1925) 510;

Bresadola in Ann. Myc. 14 (1916) 232.

Hymenochaete olivaceum Cooke ! in Grevillea 14 (1885) 11; Massee in Journ.

Linn. Soc. Bot. 27 (1890) 116.

Hymenochaete griseocervina P. Henn., fide Bresadola in Ann. Myc. 18 (1920) 70.

Illustrations: Plate 29, mature stage; Plate 30, young stage.

Resupinate or resupinate-reflexed, or conchiform attached by a small umbo,

orbicular-confluent, thin, coriaceous, or papyraceous, loosely attached to the sub-

stratum. Reflexed surface tomentose, glabrescent when old, buffy-brown, concentri-

cally furrowed in pileate specimens, the troughs being paler. Hymenium smooth, or

more usually furrowed concentrically and cracking radially. Colour very variable

through umber, brown, olive brown, light yellow brown, brownish slate, depending

on the state of development. Margin narrow, yellowish, finely fibrillose, usually free.

Context concolorous, 200-(500) g thick.

Basidia : clavate, 4-5 • 5 x 20-25 g, in young stages in a regular palisade, later interrupted

and exceeded by the false setae.

Spores: not seen (“ 6-7 x 3-4 g, colourless ” — v. d. Byl).

False setae: 3-5-8 g wide, cylindrical, often with a fusoid apex, thick-walled with a very

narrow lumen sometimes expanded at the apex, immersed or emergent up to 16 g,

originating as horizontal skeletal hyphae which curve upwards into the hymenium,

therefore the total length is variable and indeterminable, very densely arranged,

darkening in 5 per cent. KOH.

Hyphae: (1) Skeletal hyphae, brown, thick-walled, smooth, without clamps, of the same

dimensions as the false setae. (2) Colourless hyphae intermeshing the skeletal,

3 -5-4 -5 g wide, often indistinct, thin to thicker-walled, with occasional clamp

connections. Hyphae forming a more or less horizontal weft without a darker or

more compact basal layer.

44

Surface hairs: pallid to light yellow-brown, 5 p diam., fairly thick-walled, septate, with

occasional clamp connections.

Specimens examined: Type, Schomburg, Port Darwin, Australia; Cheesman, New

S. Wales, Australia, 1914; Crawford , New England, Australia; 27644, Rump

(33), Cato Ridge, Natal, 1934; 27544, Rump, Winterskloof, Natal, 1934;

27552, Rump (20), Table Mountain, Natal, 1934; 35327, Rump (683), on

Quercus, Maritzburg, 1943; 28942, Doidge, Xumeni Forest, Donnybrook,

Natal, June 1935; 20944, MacOwan (1210 ?), in Herb. Pretoriae as Corticium

a/bocinereum Kalch., possibly a pen error of MacOwan for C. atrocinereum;

Rump (273) Town Bush, Maritzburg, Oct. 1934; 34951, Simpson and Talbot,

Qudeni, Zululand, 1945; O. A. Hoeg (F. 67), Eshowe, Zululand, Aug. 1929;

Maitland (542), Mombasa, Kenya, Sept. 1920; Diimmer (1133), Kipayo,

Uganda, Oct. 1914; Diimmer (945), Kipayo, Uganda, Aug. 1914; Maitland

(34 a.), Uganda, Jan. 1915; Farquharson (3), S. Nigeria, 1914; van der Byl

(1049), East Africa, 1923; MacOwan, Cape Province (in Herb. Kew

under Corticium atrocinereum Kalchbr.); Pegler (1234), Kentani, C.P., (in

Herb. Kew under Stereum membranaceum Fr.); the following specimens

from Town Bush, Maritzburg, Oct. 1934, viz. 28496, Rump (252); 28499,

Rump (258); 28304, Rump (235); 28504, Rump (274); 28687, Rump (278).

Stereum atrocinereum (Massee) v. d. Byl is reduced to synonymy for the following

reasons: On the only sheet of “ Corticium atrocinereum Kalchbr.” in Herb. Kew there

is a note from the collector, P. MacOwan, which reads “ I take this to be Corticium

atrocinereum Klch. described from the single set of sps. sent in ’66. 1 have not yet

received from the author an authentic type, but the description agrees. Authentication

hereafter. P. MO.” As far as can be ascertained this specimen was never authenticated

and the name Corticium atrocinereum Kalchbr. is a nomen nudum and must be dis-

regarded. This specimen is, however, the type of Peniophora atrocinerea Massee.

Van der Byl (l.c.) bases his description on a collection by MacOwan, Crypt. Aust.

Africa 1197, Boschberg, Somerset East, in Herb. S.A. Musuem No. 34248, which he

assumes is part of the same collection as reached Kew under the name C. atrocinereum.

Massee gives the spore size as 10 x 4 5 //; van der Byl as 8 x 4 p; the present writer

has only found spores in No. 35327, where they measured (5)-6-(7) x (3)-4 p.

The MacOwan specimens differ from the other specimens of S. Schomburgkii only

in the colour of the hymenium, but this slight difference is found to be embraced in

variations in colour which form a natural graded series over a large number of collec-

tions. The reason for this variation in colour is discovered microscopically, and is of

considerable interest. The depth of colour is dependent on the number of skeletal

hyphae which curve upwards and emerge as false setae. Thus in the pale stage (Plate

30) (S. atrocinereum) there is an almost continuous palisade of basidia at the surface,,

and only a few of the horizontal dark hyphae curve up through the hymenium. There

is also a larger proportion of hyaline generative hyphae than is found in dark forms.

This too is the case in the pale margin of an otherwise characteristically dark-coloured

S. Schomburgkii and it suggests that the pale form is simply a young stage in the develop-

ment of the plant. In dark specimens the basidia are separated to a considerable extent

by the intrusive false setae, and do not form a continuous palisade. The greatest

number of emergent false setae is always found towards the centre away from the

growing margin. The pale condition is undoubtedly associated with young stages of

growth (e.g. at the margin), but it seems likely that it is also associated with rapidity of

centrifugal growth, for some pale specimens are luxuriantly developed and occupy a

large area. It would seem that the false setae are differentiated some time after the

formation of the hymenium and that they remain smooth and more or less horizontal

beneath the hymenium until a certain age or slackening in the growth rate, then there

is some stimulus which causes them to curve upwards and become minutely rugose

45

false setae. Petch [in Ann. Roy. Bot. Gard. Perad. 9 (1925) 259] observes that in

many species of Stereum the fungus may attain full stature before the hymenium is

formed, and in such herbarium specimens there is likely to be a lack of accessory hy-

menial organs, e.g., cystidia. Present work shows that in Stereum Schomburgkii the

reverse appears to be true, namely, that the basidia are precocious. This is of particular

interest in view of the very late development of basidia above the false setae in Duportella

tristicula (B. & Br.) Reinking, a species which the auther considers related to S. Schom-

burgkii in the same way as the genera Gloeocystidium and Peniophora are to Corticium.

(See notes on D. tristicula, page 46.)

The only record of Stereum membranaceum Fr. in South Africa [Pole Evans and

Bottomley in Ann. Bolus Herb. 2 (1918) 192] is based on a specimen from Kentani,

Pegler (1234). This is not S. membranaceum but appears to be a pale form of S. Schom-

burgkii in which the false setae are more encrusted apically than is usual in this species.

Synonyms and Excluded Species in STEREUM.

adnatum Lloyd —Stereum rimosum Berk. var. africanum Talbot q.v.

atrocinereum (Massee) v. d. Byl \— Stereum Schomburgkii Berk. q.v.

duriusculum B. and Br. \=Asterostromella duriuscula (B. and Br.) comb. nov. q.v.

fuscum Schrad. ex Quelet=S7<?/Y,wm bicolor Pers. ex Fr. q.v.

luteobadium Fries =Hymenochaete luteobadia (Fr.) Hohnel and Litsch.

ru\)ig\nomm=Hymenochaete rubiginosa q.v.

Incorrect Records of STEREUM spp. in South Africa.

albobadium (Schw. ex Fr.) Fr. Recorded in Grevillea 10 (1881) 58. The speci-

men referred to lacks cystidia and branched paraphyses and cannot be S. albo-

badium. The material is so scanty that it probably cannot be named with

certainty.

elegans Meyer. Recorded in Fungi MacOwani, Grev. 10 (1881) 58. The speci-

men referred to is Stereum Thozetii Berk.

luteobadium Fries. Recorded in Grev. 10 (1881) 58. The specimen referred to is

Stereum fasciatum Schw.

membranaceum Fries. Recorded in Ann. Bolus Herb. 2(1918) 192. The specimen

referred to is a pale form of Stereum Schomburgkii Berk.

rimosum Berk. ! Specimens ex Herb. Pretoriae Nos. 28285 and 30777 in Herb.

Kew. should be transferred to S. rimosum var. africanum nov. var.

percome B. and Br. ! The MacOwan specimen in Herb. Kew. No. 7355 is Hymeno-

chaete nigricans (Lev.) Bres.

pruinatum B. and C. ! The MacOwan collection (1227), Somerset East, in Herb.

Kew. No. 7396 has been compared with the type of S. pruinatum and is not that

species.

Resupinate Species of STEREUM Recorded for South Africa but not available

FOR STUDY.

laxum Lloyd. Lloyd, Myc. Writ. 4 (1915) L. 60, p. 10.

46

DUPORTELLA Patouillard.

1. Duportella tristicula (B. & Br.) Reinking in Philippine Journ. Sci. 17 (1920) 364,

[in error as Duportella tristiuscula (Berk.) Pat.].

Corticium tristiculum B. & Br. ! in Journ. Linn. Soc. Bot. 14 (1873) 71; Saccardo,

Syll. Fung. 6 (1888) 622; Petch in Ann. Roy. Bot. Gard. Perad. 9 (1925) 282.

Hymenochaete tristiuscula (B. & Br.) Massee in Journ. Linn. Soc. Bot. 27 (1890)

111 (orthographic variant); Teodoro, Enum. Philippine Fung., Dept. Agric.

& Comm. Tech. Bull. 4 (1937) 277.

Hymenochaete tristicula (B. & Br.) Massee, Wakefield in Kew Bull. (1916) 73.

Hymenochaete castanea Wakefield ! in Kew. Bull. (1914) 260; Saccardo, SylL

Fung. 23 (1925) 528.

Duportella ve/utina Patouillard in Philippine Journ. Sci. 10 (1915) 87.

Illustrations'. Wakefield in Kew Bull. (1914) 260 (as H. castanea).

Plate 31, young stage; Plate 32, mature stage.

Resupinate, widely effused, orbicular-confluent, when old sometimes becoming

narrowly reflexed at the margin otherwise closely adnate. Hymenium at first chestnut

brown, smooth and velutinous, later more waxy, fawn brown, purple-brown or pale

ashen, plainly tuberculate and finely reticulately cracked. Margin pale brown rimmed

with a very narrow whitish edge, determinate, not very conspicuous in old specimens.

In section 200-500 p thick.

Basidia: 23-40 x 4-5-7 p, clavate, at first scattered, immersed among the false setae;

later forming a hyaline palisade 5-15 p wide above the false setae and when fertile

sometimes emerging a further 10-13 p above the general level of the basidia.

Sterigmata 4, slightly curved, up to 4 p long.

Spores: hyaline, oblong or subcylindrical, usually slightly curved, 3-3 -5-4 x 7—8 • 5—9—

10 p; Massee (l.c.) cites them as: “ olivaceous elliptical, 10 x 4 /i.”

False setae: deep yellow-brown, darkening in potash, cylindrical, very numerous and

densely arranged, fairly thick-walled; apex rounded and sometimes septate, finely

rugose; originating as skeletal hyphae and differentiated only by a vertical position

and sculptured apex; at first exceeding the hymenial elements, later becoming

embedded under them, 3-3 -5-4 p diam.

Cystidia: thin-walled, fusoid, hyaline, emergent up to 20 p beyond the basidia, encrusted

or almost smooth, 10 x 40-45 p, scanty and only found in old specimens.

Gloeocystidia: ventricose or fusiform, smooth, hyaline, thin-walled, sometimes with a

subapical cross septum, 12-25 x 50-80 p, with granular or homogeneous contents,

arising from the basal layers, present at all ages of the fungus.

Hyphae: (1) Skeletal hyphae, brown, and of the same construction as the false setae

but smooth and apparently not septate, loosely interwoven in a horizontal layer.

(2) Septate, thin-walled amost colourless, branching, form of generative hyphae,

closely arranged between both the skeletal hyphae and the false setae, 2-3 p wide,

conglutinate and indistinct at the junction with the substratum, but not forming a

notably differentiated basal layer.

Specimens examined : Type in Herb. Berkeley, Ceyion 994, at Kew; 1 1533, Leighton,

on wattle, Buccleuch, Natal, 25/5/1918; 28937, Doidge and Morgan, Donny-

brook. Natal, Feb. 1935; 28946, Doidge, Xumeni, Donnybrook, 1935;

27768, Rump (123), on Acacia mollissima. Town Bush, Maritzburg, Aug.

1934; 28492, Rump (272), Town Bush, Maritzburg, Oct. 1934; Hoeg (F. 113),

on “ Syringa ” ( Melia Azedarach ), Maritzburg, (in Herb. Kew); 27553,

Rump, Bot. Gard., Maritzburg, 1934; 27628, Rump, Cato Ridge, Natal,.

47

1934; 36706, Talbot, on Dalbergia armata. Kloof Falls, Natal, 31/5/48;

36705, Talbot, Kloof Falls, Natal, 31/5/48; Rump (625), Table Mountain,

Natal, 1935.

This species is most closely related to Stereum Schomburgkii Berk, (q.v., pg. 43),

in the same way as a Gloeocystidium or Peniophora is related to a Corticium, and there

is also some macroscopic resembla’nce between the two species, but D. tristicula is

thicker, more tuberculate, and hardly reflexed. One wonders nevertheless whether

S. Schomburgkii would not be more appropriately placed in Duportella. Its lack of a

basal layer, colour reaction of the false setae with KOH, organisation of the tissues,

and general appearance all appear notable pointers to close affinity, and more important

than absence of cystidia and gloeocystidia.

The false setae are particularly characteristic of these two species. The writer

has not yet seen them in any other species, though they are to some extent paraded in

the hymenium of Stereum percome B. & Br. ! by hyaline hyphae which arise from the

skeletal hyphae and become thickly encrusted with heavy yellowish crystals, and also

in the cystidia of Lopharia Dregeana (Berk.) comb. nov. In Stereum umbrinum the

cystidia also arise from skeletal hyphae but are clearly differentiated as cystidia and do

not darken in potash. In fungi of the Peniophora laevigata (Fr.) Massee group, the

coloured cystidia are staged throughout the trama, arise as cystidia from the outset,

and are not merely apical modifications of skeletal hyphae since these are entirely

lacking. These cystidia are not homologous with the false setae of D. tristicula and

S. Schomburgkii though they are approximately the same width and colour, but shorter.

A difference between D. tristicula and S. Schomburgkii is that the former delays

the production of a continuous layer of basidia, whereas the basidia are precocious in

the latter species. D. tristicula is a fungus in which the body reaches full stature before

the complete differentiation of the basidia and cystidia, a fact which Petch [in Ann.

Roy. Bot. Gard. Perad. 9 (1925) 259] observes is common in many species of Stereum.

Development in duportella tristicula.

As a result of examination of three old specimens from South Africa, it is now

possible to give a new and fuller concept of the development of this species. The

specimens referred to are Hoeg (F. 113), and Herb. Pretoriae Nos. 27768 and 28492

(in Herb. Kew).

The young stages are orbicular, enlarging by confluence to the effused state found

in mature specimens. When young, the colour is a warm argus brown (Ridgway) all

over, except at the margin which is paler and bordered by a very narrow whitish band,

and the texture and appearance is velvety, smooth and uncracked, the reason for this

being that the hymenium is as yet unformed and the surface is composed of the rounded

tips of the upright false setae only. Sections show the presence of these false setae

arising almost vertically from a narrow, basal, horizontally-woven hyphal layer and

also the presence of gloeocystidia, almost all of which are basal in origin. This is

the stage which received the name Hymenochaete castanea Wakef. Later, on receipt

of older specimens, Miss Wakefield [in Kew Bull. (1916) 73] reduced this species to

synonymy with Hymenochaete tristicula (B. & Br.) Massee.

When the hymenium is formed, the first basidial elements are found spaced out

in loose groups between, and exceeded by the false setae, but later the latter are overlaid

by a continuous hymenium consisting of basidia in close palisade and occasional pro-

jecting, fusoid, rugose or smooth, thin-walled cystidia. Such cystidia are not mentioned

in previous descriptions of the species. The fungus by this time has lost its velvety

texture and its surface has turned from brown to fawn, or purple-brown, or ashen

grey, become rather tuberculate and cracked into innumerable minute reticulate cracks.

The margin now is still brown with a whitish rim but is very narrow or indistinguishable.

In section it is found that gloeocystidia are still present among the basal tissues, and

48

that the fungus has thickened by development of a fairly wide intermediate layer of

skeletal and generative hyphae between the false setae and the substratum. The

hymenial cystidia are evidently formed only in old specimens. The writer has not been

able to trace the manner of their formation. One specimen in Herb. Kew (Petch:

Herb. Perad. No. 5859) shows admirably every stage in development and the corres-

ponding anatomical changes in section.

The structures called “ false setae ” above, have been variously referred to in the

literature as paraphyses, setae or hyphae. Petch pointed out [in Ann. Roy. Bot. Gard.

Perad. 9 (1925) 282] that these structures are merely the tips of hyphae curling upwards,

and not specially differentiated, and that they differ from the setae of an Hymenochaete

in being relatively thin-walled and immersed in later stages. In the present paper

they are considered as continuations of the skeletal hyphae with a slightly rugose and

inflated apex. They are called “ false setae ” here because they darken in potassium

hydrate and because their connection with skeletal hyphae can only be seen in very

thin sections. This fundamental difference in their origin from the setae of Hymeno-

chaete is, apart from all other considerations, sufficient to justify Duportella as a genus

distinct from Hymenochaete, recognition of which has been slow to gain ground in the

past.

The basal embedded gloeocystidia have also been known as cystidia, e.g., in

Patouillard’s diagnosis of the genus, but their structure and position are typical of the

bodies now known as gloeocystidia.

HYMENOCHAETE Leveille.

Key to species described :

Resupinate; setae commonly set in fascicles found throughout the trama 1. H. fasciculata.

Resupinate; setae not fasciculate, found in a layer about 260 p thick above the trama

2. H. semistupposa.

Reflexed to dimidiate and imbricate, or sometimes resupinate; setae not fasciculate, found in a layer

about 160 p thick above the trama. Margin conspicuous ochre colour. Abhymenial surface soft,

rubbing ochraceous 3. H. ochromarginata.

1. Hymenochaete fasciculata Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 943.

Illustrations: Plate 33.

Type : 28279, Rump (220), on indigenous wood. Town Bush, Maritzburg, Oct.

1934. In Herb. Kew with isotype in Herb. Pretoriae.

Resupinate, adnate, brittle, cracking on drying, verona brown colour. Hymenium

uneven, tuberculate or granular. Width in section 250-400 /t.

Basidia and Spores: not seen.

Setae: entirely embedded, or emergent up to 50 p, somewhat encrusted, 70— 80 ( 1 20)

X 6-6-9 p, deep red-brown, acutely pointed, cylindric to subulate, solitary or set

in fascicles (especially at the papillae) of 8-12 setae. Setigerous layer not differen-

tiated; setae found throughout the tissues.

Hyphae: light yellow-brown, 1 p wide, mostly very much branched with short side

branches, dendroid, bordered on the abhymenial side by a narrow dark seam. In

places two such seams are seen as though the fungus has renewed its growth and

become secondarily thickened.

Specimens examined: Type, cited above; 28500, Rump (262) Town Bush, Maritz-

burg, Oct. 1934.

This species falls in a group with H. corrugata (Fr.) Lev. and H. fuliginosa (Pers.)

Bres. on account of the lack of differentiation of hyphal and setigerous layers, but it is

immediately distinguished by the highly characteristic fasciculate setae. The specimen

No. 28500 has a rather smoother hymenium than the type and fasciculation is not

quite so marked.

49

H. Victor Petch ! [in Ann. Roy. Bot. Gard. Perad. 9 (1925) 277], which was at one

time confused with H. fuliginosa, also has fasciculate setae. Its type differs from the

new species as follows: (1) a smoother, more reddish hymenium; (2) fewer setae per

fascicle (up to 6) and the fascicles not very dense; (3) lighter context colour and a light

yellow basal seam; (4) conspicuous cavities in the trama; (5) the hyphae in H. lictor

are not quite so narrow nor so dendroid as those of H. fasciculata.

Hymenochaete cervina B. & C. ! of the same group, differs in being thinner, paler

coloured, and without fasciculate setae, but in other respects it is close. It is found in

East Africa [Uganda: Dummer (4038); Small (227); in Herb. Kew], Possibly H.

fasciculata is an extreme form of H. cervina but no intermediates have yet been discovered.

2. Hymenochaete semistupposa Petch ! in Ann. Roy. Bot Gard. Perad 9 (1925) 278.

Illustrations'. Plate 34.

Type: Herb. Perad. No. 5627, in Herb. Kew.

Resupinate, effused, separable when moist, brittle when dry, vertically fibrillose

when broken across, up to 900 g thick. Hymenium smooth, little cracked, dark verona

brown. Margin cinnamon-rufous, thinner. Forming a pocket rot of the underlying

wood.

Basidia and Spores: not seen.

Setae: arising in all parts of a setigerous layer about 260 g thick, which may show one

or two interrupted strata of darker material. Setae fusoid to subulate, deep yellow-

brown, apex pointed, 45-70 x 5 -6-8 -4 g, emergent up to 30-50 g, or immersed.

Hyphae: yellow-brown, branched, septate, without clamp connections; those in the

setigerous layer vertical, closely arranged, 3 g wide, lightly coloured; intermediate

hyphae loosely interwoven, 2 -8-3 -5 g wide, with appreciably thick walls but a

wide lumen, the walls being straight, wavy or constricted in outline, when con-

stricted having a moniliform appearance. Basal hyphae closely massed in a

dark seam, horizontal, 70-200 g wide.

Specimens examined: 30232, Rump (453), Krantzkop, Natal, Nov. 1935; Maitland

(517), Gongoni Forest, Kenya, Sept. 1 920 ; Petch (5657, 4044, 5107), Hakgala,

Ceylon.

. In microscopic structure this South African collection is a close match with the

Type specimen from Ceylon, though the setae in the former are rather more subulate

than fusoid. In the Type, the intermediate hyphae do show an occasional wavy out-

line but they are never moniliform. The monilioid constrictions are quite plain in

parts of the Kenyan specimen and reach their highest development in the South African

specimen. There is insufficient reason for separating these African examples as a

distinct new variety. The colour of all these specimens is rather variable even in a

single collection. The typical colour is an olivaceous Prout’s brown with a cinnamon

margin, but it is often a warmer shade of reddish-brown or even darker.

3. Hymenochaete ochromarginata Talbot apud Wakefield and Talbot in Bothalia 4

(1948) 944.

Illustrations : Plate 35.

Type: 2133, Pienaar, Tenadu, Tembuland, 26/2/1912. In Herb. Kew with

isotype in Herb. Pretoriae.

Narrowly resupinate, mainly reflexed to dimidiate, tough, not pliable, laterally

connate, imbricate, 4x2 cm. Upper surface concentrically furrowed, with velvety

pubescence, light bay colour, rubbing away to yellow ochre. Hymenium smooth,

warm argus brown, not cracked, somewhat reflecting the surface furrows. Margin

acute, with bright yellow ochre band less than 1 mm. wide. Width in section 500-650 g.

Basidia and Spores: not seen.

50

Setae: emergent, or fully immersed in a layer up to 1 60 // thick, yellow-brown, subulate,

30-45 x 6-8 /j., with narrow lumen, emergent up to 20 g..

Hyphae: yellow-brown, narrow, 1-5-2 -5 g diam., much interwoven and more or less

horizontal in the greater part of the trama. Trama bordered on both sides by a

seam of dark closely-woven hyphae, the seam not very pronounced on the setiger-

ous side, being only conspicuous in rather thick sections.

Tomentum: on the upper surface, confluent with and exactly similar to tramal hyphae.

Specimens examined: Type, cited above; 30260, Morgan and Doidge, Donnybrook

Natal, 1936; 28879, Bottomley, The Cavern, Drakensberg, Natal, 20/7/1937;

Maitland (93), Cameroon Mountains, Cameroons, 1930 (in Herb. Kew,

filed under unnamed Hymenochaete).

In structure this species comes within the small group which includes H. rigidu/a

B. & C. !, H. tabacina (Sow.) Lev., and H. rubiginosa (Dicks.) Lev. There is no likeli-

hood of confusion with H. tabacina. H. rubiginosa may be distinguished by longer,

narrower setae (50-60 x 5-6 g), by its colliculose bistre hymenium, and wider hyphae

(2-5 g), but especially by becoming indurated and blackish on the abhymenial surface.

In H. ochromarginata the surface remains ochraceous and soft. The hyphal width in

H. ochromarginata is- more variable than was supposed at the time of first describing

the species, and should not be over-emphasised. H. rigidula differs in being pre-

dominantly resupinate and having sensibly darker, wider hyphae (3-4 g).

In Herb. Kew are several specimens from Ceylon under H. rigidula B. & C. (Nos.

3042, 6670, 3866, 6058, 4031) which differ from the type of H. ochromarginata only in

hymenial colour. Petch [in Ann. Roy. Bot. Gard. Perad. 9 (1925) 272] gives notes on

H. rigidula and concludes: “ It would seem probable that the Ceylon species is not the

same as that from Cuba,” (type locality). As the hymenial colour of H. ochromarginata

in South Africa varies from warm argus brown to almost the bistre of H. rubiginosa,

Petch’s specimens lie within this colour range and should be regarded as H. ochro-

marginata.

Synonyms and Excluded Species in the genus HYMENOCHAETE.

tristicula B. & Bv. = Duportella tristicula (B. & Br.) Reinking, q.v.

Incorrect Records of HYMENOCHAETE Species in South Africa.

luteobadia (Fr.) Hohnel and Litsch. Recorded in Grevillea 10 (1881) 58 sub

Stereum. The specimen referred to is in Herb. Kew under Hymenochaete,

but is actually Stereum fasciatum Schw. ( Hymenochaete luteobadia is, how-

ever, found in South and East Africa.)

tenuissima Berk. ! Specimen No. 15596 (Natal Herbarium, 290) is a polypore.

It would seem that the specimen was misplaced at some time, for v. d. Byl

refers to this collection in S.A. Journ. Sci. 18 (1921) 347 and gives a particular

note on its thin, flexible structure, whereas the present specimen is a thick

polypore, namely Polystictus tabacinus Mont.

Resupinate Species of HYMENOCHAETE recorded for South Africa but not

AVAILABLE FOR STUDY.

fusco-violascens (Mont.) Sacc.; Montague in Ann. Sci. Nat. ser. 3, 7 (1847) 174,

sub Thelephora; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 14.

rubiginosa (Schrad.) Lev.; Van der Byl (1929) l.c., p. 14; Massee in Journ. L inn.

Soc. Bot. 27 (1890) 97; Eyles in Trans. Roy. Soc. S.A. 5 (1915) 277, (Rhodesia);

Montagne in Ann. Sci. Nat. ser. 3, 7 (1847) 174, sub Stereum rubiginosum Fr.

There are many unmatched collections of Hymenochaete in Herb. Pretoriae. The

writer could not match them at Kew, and he considers that comparison with North

American collections is necessary. Most collections of Hymenochaete lack basidia

and spores and the genus presents unusually great difficulty.

51

ASTEROSTROMELLA von Hohnel and Litschauer.

1. Asterostromella duriscula (B. & Br.) comb. nov.

Stereum duriusculum Berk, and Broome ! in Journ. Linn. Soc. Bot. 14 (1873) 66;

Saccardo, Syll. Fung. 6 (1888) 585; Bourdot and Maire in Bull. Soc. Myc. de

Fr. 36 (1920) 74; Petch in Ann. Roy. Bot. Gard. Perad. 9 (1925) 267. [non

S. duriusculum sensu Bresadola in Ann. Myc. 6 (1908) 43, nec S. duriusculum

Bres. sensu Overholts in Mycologia 30 (1938) 279.]

Illustrations : Plate 36.

Resupinate effused, up to 1-5 mm. thick, tough, fibrous-suberose. Context pale

creamy white to pale yellowish-brown, markedly stratose. Hymenium very smooth,

creamy to chamois colour, sometimes pruinose with the spores.

Basidia: scattered, not in palisade, exceeded by the hymenial hyphae, about 4 x 24 //.

Spores: globose-subglobose, often with a blunt apiculus, smooth, hyaline, 6-6-8 p diam.

Gloeocystidia: thin-walled, with large irregularly shaped contents staining readily with

phloxine, embedded fairly deeply or in the plane of the basidia, reaching maximum

size of 6-8 x 120 p, usually about 4 x 48 p, rare or abundant depending on the

specimen.

Context: zoned with alternate hyaline and pale yellowish zones, the hyaline part (fide

Petch) being the lower part of each growth zone.

Hyphae: thick-walled or without visible lumen, hyaline, very finely filamentous or up

to 3 /j ! wide, very much branched but not true dichophyses.

Minerals: concretions common in lower layers, especially at the junction of growth zones.

Specimens examined: Type, Thwaites (569 and 329) in Herb. Kew, Ceylon; 27604,

Rump (25), Town Bush, Maritzburg, 1934; 30208, Rump (409), Umhlali,

Natal, 1935; 30209, Rump (410), Compensation Beach, Natal, Dec. 1934;

30210, Rump (411), Umgeni Bush, Durban, March, 1935; 35423, Rump (741),

Richmond, Natal, March, 1945; 34360, Rump, Karkloof, Natal, 1936;

36429, Rump, Maritzburg, May 1947; 35324, Rump (591), Isipingo Beach,

Natal, 1944; 28473, Rump (191), Winterskloof, Natal, 1934; 33902, Rump

(534), Scottburgh, Natal, Feb. 1943; 30823, Rump (479), Port St. Johns,

Pondoland, Aug. 1937; 34543, Scott, Munnik-Tzaneen Road, 27/5/1943;

33993, Rump (675), on Euphorbia pulcherrima (Poinsettia), Maritzburg,

1943; Rump (796), Verulam, Natal, 1946; Glaziou (12333), Rio Janeiro,

Brazil; Alban Stewart (975), Galapagos, 1905; J. Peres Guimarais, Lourenco

Marques, P.E. Africa, 23/3/1940 and 18/3/1940.

Asterostromella duriuscula belongs to a group of stratose resupinates of which

Corticium portentosum B. and C. ! is another example. The taxonomy of the group is

confused. In this group the characters of taxonomic importance are: (1) Context

colour; (2) Nature of the branching of the hyphae; (3) Colour of hyphae, and their

reaction with KOH; (4) Presence or absence of gloeocystidia; (5) Colour and sculpturing

of spores. Of little taxonomic value are: (a) Width, septation, and thickness of the

hyphal walls; ( b ) Hymenium colour (useful for A. dura); (c) Thickness and degree of

stratification of the context; (d) Spore size, for all the spores in this group are almost

globose and about (5 • 5)-6-6 • 5-(8) p in diameter.

For comparative studies a standard mounting technique was used. Sections were

mounted in a drop of concentrated hydrochloric acid (to clear minerals) which was

then drained away and replaced by 10% KOH until bubbling ceased. Two such alka-

line mounts were made, one with the addition of phloxine stain and the other without.

For examining the branching of hyphae the sections were teased out with needles.

The species examined can conveniently be classified in the following manner:

(see also Plate 37) —

52

Group 1. A: Context white or light yellowish; sections not darkening in KOH.

Hyphae hyaline, much branched, dendrophytic not dichophytic.

Group 1. B: Differing from 1. A in having dichophytic hyphae, which are hyaline

or very lightly coloured and darken very slightly in KOH.

Group 2: Context not light coloured; sections darkening in KOH. Hyphae

coloured, markedly dendroid.

Group 3 : Context not light coloured ; sections darkening slightly in KOH. Hyphae

hyaline or very dilutely coloured, dendroid, not dichophytic.

Group 1. A.

(a) Corticium portentosum B. and C. ! ( =C. diminuens B. and C.).

( b ) Stereum duriusculum B. and Br. !

The most significant difference between these two very similar species is one which*

does not appear to have been recorded previously, namely that there are gloeocystidia

in S. duriusculum, which cannot be found in any North American specimens the writer

has examined in this group, including the type of C. portentosum. The gloeocystidia

are difficult to demonstrate but they are present in the type of S. duriusculum, and are

most numerous and conspicuous in some South African specimens which are hied

under both species in Herb. Kew. Two collections from Lourenco Marques are

interesting in that they were both found at about the same time and locality, yet in

one the gloeocystidia are abundant and in the other very rare. Examples with gloeo-

cystidia are also found from Ceylon (Type of S', duriusculum), the Galapagos, and Brazil.

It was first considered that the character of gloeocystidia taken alone did not warrant

the retention of C. portentosum and S. duriusculum as distinct species, it being well

known that such structures are not necessarily found in all specimens of a single species.

But there are indications that the two species should not be merged. The type speci-

mens differ in hymenial colour, C. portentosum being a somewhat lighter chamois,

and in width of hyphae, S. duriusculum having doubly wide hyphae. These characters

are not very helpful because intergrading variations are common. Most specimens

the writer assigns to S’, duriusculum on the grounds of having gloeocystidia, give an

immediate jet black colour if the hymenium is touched with a drop of KOH, while

most assigned to C. portentosum react by developing a reddish-brown colour which

slowly darkens. But again, this chemical test is not invariable. Some of the South

African specimens of 5. duriusculum are rich in mineral concretions at the junction of

the strata. This, too, is a variable feature. The writer has not yet seen an example of

C. portentosum from South Africa, for all those in Herb. Kew he would call S. durius-

culum. It seems probable that C. portentosum is a North Temperate species and pre-

dominantly American.

Certain other fungi have been confused, or merged properly or improperly with

these two: —

Corticium diminuens B. and C. ! is given by Hohnel and Litschauer [in K. Akad.

Wiss. Wien. Sitzungsb. (1907) 743] as a synonym of C. portentosum. I consider

this is substantiated. It is significant that the species is known only from

Alabama, North America.

Corticium grammicum P. Henn. is cited in the same place as another synonym. No

material of this was available for study, but in Saccardo, Syll. Fung. 21 (1912) 395

and in Engler Jahrb. 38 (1905) 106 the spores are given as “ellipsoidal, 2-5-3

X 2 /(,” which indicates that Hohnel’s statement of synonymy should be further

enquired into.

Stereum duriusculum has evidently been confused both by Burt and by Bresadola.

The latter has written on the type sheet in Herb. Kew that C. portentosum,

S. duriusculum and S. induratum Berk. ! are all the same species, and in Ann.

Myc. 14 (1916) 232 he gives S. albo-cinctum B. and Br. ! as a synonym of S.

53

duriusculum. It is plain on examination that S. induratum and 5. albo-cinctum

belong in Group 2 of this treatment, and that S. duriusculum sensu Bresadola is

Asterostromella dura Bourdot and Galzin.

Burt [in Ann. Mo. Bot. Gard. 7 (1920) 236] says that S. duriusculum “ as determined

by Patouillard in Duss, FI. Antilles Fr. 232. 1903, is probably Hypochnus pallescens

(Schw.) Burt, a species common in the West Indies.” Burt draws his description of

the latter species from authentic material in the Curtis Herbarium. The material in

Herb. Kew (7412) sent by Curtis to Berkeley and labelled in the latter’s handwriting as

“ Thel. pallescens (a). Herb. Schwein. !” cannot be considered authentic, for it is actually

a Peniophora bearing very large, thick-walled, encrusted cystidia, conical in shape.

Examination of recent American material which coincides with Burt’s descriptions

shows that this species has nothing to do with S. duriusculum but falls in Group 2 of

this treatment.

Group 1. B.

In Herb. Kew, under C. portentosum are a series of three Swedish collections ex

Mus. Botan. Stockholm, Nos. 1523, 1334 and 4135, which the author considers distinct

from C. portentosum for the reason that their hyphae are truly dichophytic. These

specimens, then, belong to Asterostromella and possibly represent a new species.

Group 2.

(a) Stereum induratum B. and C. ! (-S. albo-cinctum B. and Br. !). This species

lacks the tissue differentiation of a Stereum and possesses dichophytic hyphae.

(. b ) Asterostromella dura Bourd. and Galz. (-S. duriusculum sensu Bresadola).

Differs from the above in greater thickness, darker context, and ochraceous hymenium

colour. The two are obviously very closely related. This species is the type of the

genus Dichostereum proposed by Pilat [in Ann. Myc. 24 (1926) 223] as a segregate from

Asterostromella. Rogers and Jackson [in Farlowia 1 (1943) 310] make this species a

synonym of Vararia pallescens (Schw.) Rogers and Jackson. This may well be so,

but for the present the writer prefers to keep the two species separate, as he believes

that they differ in colour and closeness of branching of the hyphae.

(c) Hypochnus pallescens (Schw.) Burt. [Synonyms: Coniophora sordulenta (Cooke

and Massee) Sacc. !; Hypochnus thelephoroides (Ellis and Everh.) Burt, fide Burt;

Thelephora insinuates Schwein., fide Burt.]

This species is an Asterostromella.

The above three species would all fall into the genus Dichostereum of Pilat (l.c.),

who mentions the species induratum, albocinctum, and duriusculum in this connection;

but it is plain that he has taken duriusculum in the sense of Bresadola. The present

writer includes Dichostereum in the genus Asterostromella.

Group 3.

Asterostromella Rumpiana Talbot, which see (page 54).

There remains to assign C. portentosum and S. duriusculum to a suitable genus,

for it seems that they fit neither Stereum nor Corticium and that their affinities lie wholly

with the rest of this group of stratose resupinates. I propose, therefore, that both

should be referred to the genus Asterostromella despite the fact that they possess hyphae

which are not the dichophyses of Asterostromella, but rather dendrophyses of the kind

found in some species of A/eurodiscus. Aleurodiscus as constituted at present is a

repository for many unrelated species distinguished by possession of pseudophyses,

dendrophyses or acanthophyses but with very little else in common, and it would only

add to the muddle to include two more species on these very slender grounds alone.

Asterostromella itself is a somewhat polyglot genus, but the time does not yet appear

ripe for making segregations of the kind of Dichostereum.

54

Donk [in Rev. Ned. Hetero.-en Homobas.-Aphyll. 1 (1931) 191] points out that

Vararia Karst, is an earlier name for Asterostromella Hohnel and Litsch., and that if

the latter is to be retained, conservation against the former is necessary.

2. Asterostromella Rumpiana Talbot apud Wakefield and Talbot in Bothalia 4 (1948)

939.

Illustrations : Plate 38.

Type: 30200, Rump (399), Nkandhla Forest, Zululand. In Herb. Kew, with

isotype in Herb. Pretoriae.

Resupinate, widely effused, adnate, up to 2 mm. thick. Context tough, corky-

fibrous, markedly stratose, light biscuit-brown colour. Hymenium smooth, with

buffy colour unmatched in Ridgway, turning black instantly on application of KOH.

Basidia: scattered, not in palisade, clavate-cylindric, about 4-5 X 30 p, exceeded by the

hymenial hyphae.

Spores: elliptic-oval (not ovate), hyaline, 5-5 X 8-3 p, with granular contents giving a

finely punctate appearance.

Gloeocystidia : thin-walled, with homogeneous, evenly staining contents, embedded,

6-5 x 53-80 p.

Hyphae: thick-walled or without visible lumen, hyaline or lightly coloured, darkening

slightly in KOH, filamentous or up to 3 p wide, very much branched but not

dichophytic.

Context: zoned with alternate lighter and darker zones.

Minerals: common at the junction of the strata.

Specimens examined: Type, 30200, cited above.

In structure this species is very close to A. duriuscula, from which it differs in darker

colour of the context, spore size and shape, and the more homogeneous contents of the

gloeocystidia. Corticium hinnuleum Bres. [in Hedwigia 56 (1915) 303] from the Philip-

pines may be closely related, but specimens have not been seen.

ASTEROSTROMA Massee.

1. Asterostroma cervicolor (B. and C.) Massee in Journ. Linn. Soc. Bot. 25 (1889) 155;

Bourdot and Galzin in Bull. Soc. Myc. de Fr. 36 (1920) 45, Hym. de Fr. (1928) 400;

Burt in Ann. Mo. Bot. Gard. 11 (1924) 28; Saccardo, Syll. Fung. 9 (1891) 237.

Corticium cervicolor Berk, and Curt. ! in Grevillea 1 (1873) 179; Saccardo, Syll.

Fung. 6 (1888) 621.

Asterostroma corticola Massee ! in Journ. Linn. Soc. Bot. 25 (1889) 155.

Asterostroma albido-carneum Massee ! in Journ. Linn. Soc. Bot. 25 (1889) 155,

non Thelephora albido-carnea Schwein.

Asterostroma pallidum Morgan in Cincinnati Soc. Nat. Hist. Journ. 18 (1895) 38,

(fide Burt).

Illustrations'. Massee (1889) l.c., PI. 46, f. 8, 9, as A. albido-carneum ; Morgan

(1895) l.c. PI. 1, f. 6, as A. pallidum.

Plate 39.

Entirely resupinate, fawn, byssoid when young with a whitish arachnoid margin,

effused, adnate. Hymenium fawn-avellaneous becoming pallid buff when sporing,

pulverulent, smooth.

Context thin, 150-300 p, spongy and dry.

Basidia: 16-23-53 x 6-6 p or narrower, with 2-4 short sterigmata, subcylindric or

clavate, in an uneven palisade.

55

Spores: hyaline, echinulate, subglobose, 4-5 p diam., or oblong, 4 x 6/1.

Asterosetae: throughout the trama, fewer and much smaller in the hymenium and

occasionally projecting slightly beyond it; each star composed of about 5 rays

borne on a thin hypha. Rays 15-70 x 3-3-5 p, light brown, the majority un-

branched, pointed, largest in the basal layers.

Gloeocystidia : demonstrable with phloxine stain, numerous, immersed in subhymenium

or projecting, fusoid or clavate, frequently attenuated into a long neck, 8-12 x

30-45-50 p, contents homogeneous.

Other ground tissues: hyphae hyaline but quite indistinct.

Specimens examined: Type, Peters (4026), Alabama (in Herb. Kew); 30155, Rump

(348. a) on the hymenium of a Pomes sp., Umgeni Bush, Durban, March

1935; 36821, Rump (469), Van Reenen, Natal, Jan. 1937; 34387, Rump (602),

Town Bush, Maritzburg, May 1936.

In the last two specimens cited above, the rays of the asterosetae radiate from

quite a stout central boss, and there is rather more branching of the rays than is typical.

Otherwise these specimens are no different from A. cervicolor.

MYCOLEPTODON Patouillard.

1. Mycoleptodon ochraceum (Pers. ex Fr.) Patouillard, Essai Taxon, sur les Hym. (1900)

116; Rea, Brit. Basid. (1922) 639; Pilat in Bull. Soc. Myc. de Fr. 42 (1926) 102;

Bourdot and Galzin, Hym. de Fr. (1928) 440.

Hydnum ochraceum Pers. ex Fr., Systema Myc. 1 (1821) 414, Hym. Eur. (1874)

612; Persoon apud Gmelin in Linne, Systema Naturae 2 (? 1796) 1440; Persoon,

Syn. Meth. Fung. (1801) 559, T.5, f. 5, Myc. Eur. (1825) 176, Obs. Myc. 1

(1796) 73; Bresadola, Fung. Kmet. (1897) n. 98, in Ann. Myc. 1 (1903) 85;

Saccardo; Syll. Fung. 6 (1888) 457.

Steccherinum ochraceum (Pers.) S. F. Gray, Nat. Arr. Brit. Plants 1 (1821) 651;

Miller in Mycologia 25 (1933) 296.

Leptodon ochraceum (Pers. ex Fr.) Quelet, FI. Myc. de Fr. (1888) 441.

Hydnum pudorinum Fries, Hym. Eur. (1874) 612, Elenchus Fung. 1 (1828) 133;

Saccardo, Syll. Fung. 6 (1888) 456; Quelet, FI. Myc. de Fr. (1888) 440.

Hydnum microdon Persoon, Syn. Meth. Fung. (1801) 561; Bresadola in Ann.

Myc. 1 (1903) 85.

Hydnum Daviesii Sowerby, Engl. Fung. 1 (1797) 15; Saccardo, Syll. Fung. 6

(1888) 457. (Fide Persoon and others.)

Hydnum alnico/um Velenovsky (fide Bourd. and Galzin).

Illustrations : Persoon, Syn. Meth. Fung. (1801) T.5, f.5; Sowerby, Engl. Fung. 1

(1797) T. 15, as H. Daviesii.

Plate 40.

Resupinate, effused, becoming reflexed, or even pileate (not yet seen in South

Africa), readily separable from substratum. Reflexed upper surface pale coloured,

concentrically furrowed, tomentose. Hymenium light ochraceous with a light fleshy

tint; composed of very closely arranged regular subulate small teeth, shorter towards

the margin. Margin whitish, narrow, subfimbriate, in section 250-400u thick excluding

the teeth.

Basidia: 3-5-4 X 1 2— 1 5— ( 1 8) p, clavate, with very short straight sterigmata.

Spores: not yet seen in South African material: 3— 4— (5) x 2-2-5 a, obovate-oblong.

Cystidia: numerous, long-clavate or fusoid, thick-walled, covered at least near the apex

with easily detersile incrustation, 6-5-10 x (40-55)-105 p, arising from ordinary

deep-seated hyphae as terminal modifications near the hymenium.

56

Hyphae: subhymenial hyphae more or less thin-walled; tramal hyphae thick-walled,

(2)-2-5-4 p, colourless, distinct, suberect, densely interwoven.

Specimens examined: 27796, Rump (82), Town Bush, Maritzburg, Aug. 1934;

Simpson and Talbot (T.R.L. 2021), Qudeni Forest, Zululand, Feb. 1945.

In Grevillea 10 (1881) 27 and in S.A. Journ. Sci. 13 (1917) 440, reference is made

to a collection of Medley Wood (479), Inanda, Natal, as H. ochraceum. I have been

unable to find cystidia in this specimen and therefore suggest that it is some other species.

LOPHARIA Kalchbrenner and MacOwan.

1. Lopharia mirabilis (B. and Br.) Patouillard in Bull. Soc. Myc. de Fr. 11 (1895) 14,

PI. 1, Essai Taxon, sur les Hym. (1900) 74; van der Byl in Ann. Univ. Stellenbosch

12 (1934) 2, f. 1 . ; Petch in Ann. Roy. Bot. Gard. Perad. 4(1910) 410.

Radulum mirabi/e B. and Br. ! in Journ. Linn. Soc. Bot. 14 (1873) 61; Saccardo,

Syll. Fung. 6 (1888) 496.

Thwaitesiella mirabilis (B. and Br.) Massee in Grev. 21 (1892) 3.

Lopharia lirellosa Kalchbr. and MacOwan ! in Grev. 10 (1881) 58; Patouillard in

Bull. Soc. Myc. de Fr. 11 (1895) 14, PI. 1, Essai Taxon, sur les Hym. (1900) 74.

Licentia yao-chanica Pilat in Ann. Myc. 38 (1940) 66, text fig. 2, Tab. 1, f. 1, 2, 3.

Illustrations : Patouillard (1895) l.c. ; v. d. Byl (1934) l.c. ; Pilat (1940) l.c.

Plate 41.

Resupinate, effused, sometimes reflexed, slightly waxy when fresh, drying thin and

coriaceous, brittle. Upper surface (when exposed) tomentose and obscurely zoned,

tan colour. Hymenium first smooth and white, then creamy to ochraceous with light

margin, and covered with scattered, erect, laterally compressed, triangular teeth, about

0-5 mm. high, arranged in more or less radial or concentric patterns: teeth coalesce

into short ridges. Thickness of context (excluding ridges) 250-300 /*.

Basidia: large, clavate, 50-65 x 9-11 p, with 4 sterigmata.

Spores: oblong-elliptic or oblong-ovate, occasionally curved, hyaline, smooth, with

rather granular contents, ( 5)— 6 * 6 8 • 3 x 10-13-2 p.

Cystidia: present throughout the hymenium and just below it, conical or subfusiform,

66-130 x 13-21 p, usually emergent up to 50 p, very4hick-w'alled, coarsely encrusted

in the superior half, often smooth and pale brownish at the base, blunt or pointed

at the apex.

Hyphae: closely interwoven, more or less erect, thick-walled, 3-4 p w-ide, subtended

on the side next to the substratum by a thin layer of horizontal faintly-coloured

hyphae.

Surface Hairs: adpressed, faintly coloured, 4 p wide.

Specimens examined: MacOwan , type of Lopharia lirellosa, C. Bon. Spei; Thwaites

(328), type of L. mirabilis, Peradeniya, Ceylon, 1868; ex Herb, von Hohnel,

Peradeniya, Ceylon, 1907; van der Byl (2261), South Africa, 1926; 27797,

Rump, Town Bush, Maritzburg, 1934; 28302, Rump, Town Bush, Maritzburg,

Oct. 1934; 27799, Rump, near Maritzburg, 1934; Rump (v. d. Byl, 2620),

Natal South Coast, Feb. 1931; v. d. Byl (551), on dead P/ectronia, Durban;

v. d. Byl (1429), South Africa; 27769, Rump (125), Town Bush Valley, Maritz-

burg, Aug. 1934; 31911, (Natal. Herb. 714), v. d. Byl, Bluff, Durban, 6/8/191 7 ;

31356, Duthie, Knysna; 31309, Duthie, Knysna; ex Herb. T. Petch (Nos.

4499, 3306, 5354), Peradeniya, Ceylon; Sappan, Flor. Singapore 5119, Singa-

pore Bot. Gard., 1919; F. T. Brooks (113), Fed. Malay States, 1914; M.

Noor, Flor. Selangor 5559, 1920; T. F. Chipp, Flor. Singapore 5910, 1920;

R. E. Holttum (9240), Malay Peninsula, 1922.

57

This species is of peculiar interest as a link between other genera both of the

Hydnaceae and the Thelephoraceae. It has the internal structure of a Stereum of the

“ Lloydella ” group (Bresadola), and in particular a very marked similarity to S. cineras-

cens (see pg. 41). The last-named is thicker and has deeper-embedded cystidia, but

its other microscopic features including spore size and shape are the exact counterpart of

L. mirabilis. But for the fact that this is resupinate, it also connects with Cladoderris

which has a papillate hymenium and, in some species, cystidia. Cladoderris cystidia

are usually small, but Lloyd [Myc. Writ. 4 (1913) 11] says that in C. Pritzelii Hennings

these are large “ metuloids.” The structure of L. mirabilis with a darker compact zone

adjacent to the substratum, and hairs, even when apparently resupinate, is evidence

that the fungus is capable of reflexion, and distinguishes it from Peniophora. Though

Lopharia is grouped next to Radulum in the Hydnaceae on account of its incised ridges,

the writer considers it certainly more nearly related to the Thelephoraceae through

the genus Stereum.

Hymenial configuration in L. mirabilis is variable. One specimen in Herb. Kew

has a distinctly poroid hymenium and could be taken for a Poria. The disposition of

hymenial crests irregularly or in somewhat concentric rings was taken by Patouillard

as the basis for separating L. lirellosa and L. mirabilis, but this is now considered unten-

able. In the type specimens of these, the writer has found that the spores show a distinct

difference in size and shape [L. lirellosa, 6-6-(8-3) x (9 - 9)— 1 3 - 2— ( 1 5) p, mostly elliptical;

L. mirabilis, 5 X 9 ■ 9— ( 13-2) p, mostly almost cylindric with flattened sides], but these

are evidently extremes, for in examination of other collections from South Africa and

Ceylon it is seen that the spore sizes and shapes intergrade. The spores in South

African material are nevertheless slightly larger and wider than those from Ceylon,

but not significantly so. The following summary of spore measurements by different

observers is of interest (Table 4).

TABLE 4.

SPORE MEASUREMENTS OF CERTAIN LOPHARIA SPECIES.

Lopharia javanica P. Henn. has large cystidia too, but spores measuring only 5-6

X 5-7 p, : Lopharia Dregeana (Berk.) comb. nov. (see following) has smaller cystidia

of a different kind, and spores 4-5-5 x 6-5-7 p.

2. Lopharia Dregeana (Berk.) Talbot comb. nov.

Corticium Dregeanum Berkeley ! in Hooker's Lond. Journ. Bot. 5 (1846) 3;

Saccardo, Syll. Fung. 6 (1888) 636; Montagne in Ann. Sci. Nat. ser. 3, 7 (1847)

174.

Hymenochaete Dregeana (Berk.) Massee in Journ. Linn. Soc. Bot. 27 (1890) 114.

Illustrations : Plate 42.

58

Type : Drege (9451. c), South Africa. In Herb. Kew.

Resupinate, effused, not adnate, somewhat flexible, at first orbicular then confluent.

Margin determinate, smooth, somewhat paler than the remainder, becoming a little

reflexed, the reflexed side softly byssoid. Hymenium considerably papillate or with

incised ridges, dry, pruinose, cracked or uncracked, colour varying from rosy-violaceous

or pale ochraceous with a faint vinous tinge to brownish. Context pale, about 400 “ thick.

Basidia: clavate-cylindric, 3-3-5 x 15-20-26 p, in close palisade and difficult to observe.

Spores: hyaline, smooth, elliptical, 4-5-5 X 6-5-7 p.

Cystidia: cylindrical, hyaline to very dilutely coloured, thick-walled, the lumen often

expanding towards the apex, minutely encrusted more usually in the upper third

of the length, very numerous, entirely immersed or emergent up to 27 p, arising as

apical continuations of ordinary hyphae deep in the trama, not darkening in KOH,

(5)-6-8 p wide, up to 170 p long.

Hyphae: hyaline, thick-walled with very narrow lumen, rather vitreous, 5-6 -5-8 p

wide, closely intertwined without showing a well-defined basal or median tissue

in the trama.

Specimens examined: Type, Drege (9451. c), cited above; Drege (4109), C.B.S.,

ex Herb. Montagne; Dunlap, Cameroon Mtns., 5000' alt., 1926. In Herb.

Kew.; Macdonald, Ngong Forest, Nairobi, Kenya, 24/5/29.

Massee (1890 l.c.) wrote of this species that it is a true Hymenochaete, but both in

origin and morphology the accessory organs of this fungus must be taken as cystidia

and not as setae. The cystidia arise as apical modifications of thick-walled, deep-

seated hyphae curving outwards, and in this respect they are of a rare kind whose

origin may be compared with that of the false setae in Duportella tristicula and Stereum

Schomburgkii, though no affinity is to be implied from this.

Microscopically all the specimens of L. Dregeana cited above are identical, but

they differ somewhat in external colour and particularly in the hymenial crests. The

Drege specimens have mere papillae scattered roughly concentrically over the hymenium;

the Macdonald specimen has incised ridges, and in places is almost poroid; there are

closely arranged incised ridges in the Dunlap specimen. In view of these incised ridges

and the presence of cystidia I have referred the species to Lopharia. This choice of

genus is not entirely satisfactory, for the following reasons: —

(1) The cystidia of both other known species of Lopharia ( L . mirabilis and L.

javanica ) are very large, fusoid or subulate, peniophoroid, and not merely

apical modifications of ordinary hyphae.

(2) Lopharia is possibly not a “ good ” genus. Under L. mirabilis it has been

shown that the hymenial configuration is very variable, and that its affinities

lie with the Thelephoraceae, particularly with Stereum cinerascens.

(3) L. Dregeana is composed throughout of uniformly interwoven hyphae without

differentiation into a denser basal layer and hairs as in L. mirabilis.

(4) A fifth specimen in Herb. Kew [Viz. Maitland (143), Cameroon Mtns., March

1931 ] has the same microscopic features as L. Dregeana and the same violaceous

tint, but is widely reflexed with a zonate abhymenial surface. It also has the

long separated teeth of a Hydnum. The writer places this specimen only

tentatively with L. Dregeana, there being a trifling difference in length and

incrustation of the cystidia which causes him uncertainty.

Corticium Dregeanum was said by Saccardo (l.c.) to be near to Corticium mode Fr.,

but from the description this can only be a superficial resemblance and takes no account

of the much more hydnoid collections of the former which have now come to light.

59

A specimen in Herb. Kew (J. B. Ellis, 3723, Iowa) included as Hymenochaete

Dregeana is misdetermined. It is almost certainly Stereum cinerascens (Schw.) Massee,

and is interesting in showing a few small papillae on the hymenium which suggests a

variation towards the form of L. mirabilis.

ACIA Karsten.

1. Acia conferta sp. nov.

Illustrations : Plate 43, fig. 2.

Type: 28494, Rump (275), Town Bush, Maritzburg, Oct. 1934.

Effused, firmly adnate, subceraceous, sometimes crustose, with indeterminate

margin. Spines crowded together, papillate to subulate, or fused in part at the base,

fulvous to chestnut colour when dry, 0 * 3— 1— (1 • 5) mm. in length. Subiculum very

thin, creamy-fawn colour.

Basidia: hyaline, clavate, 16-28 x 4-5-3 g.

Spores: hyaline, elliptic, smooth, often unilaterally depressed, often uniguttulate,

(3 • 5)— 4— 5 • 5— (6) x 2-5-4 g.

Cystidioles: scanty, cylindric, hyaline, with rounded apex, 3-4 x 40^16 g.

Hyphae: of the subiculum, thin-walled, scanty, hyaline, without clamp connections,

2-4 g diam.

Specimens examined: Type, cited above; van der Byl (308); Timber Research Lab.

(9); Timber Research Lab. (14).

Differs from Acia subceracea Wakef. ! [in Trans. Proc. Roy. Soc. S. Australia 54

(1930) 155] in having larger spores, cystidioles with rounded apex, and usually longer

hymenial spines; but otherwise in structure and appearance the two species are very

close. (Cfr. Plate 43, fig. 1.) Except for their hydnoid structure these species are

related in microscopic features to the Ceracea group of Corticium, especially to C.

ochraceo-fulvum Bourd. and Galz. Acia fusco-atra (Fr.) Pat. is darker in colour and

sometimes shows similar but much smaller cystidioles.

GRAMMOTHELE Berkeley and Curtis.

1. Grammothele pseudomappa sp. nov.

Illustrations : van der Byl in Ann. Univ. Stellenbosch 12 (1934) fig. 11 (as G. mappa

B. and C.)

Plate 44.

Type : 34963, Simpson and Talbot, Qudeni Forest, Zululand, Feb. 1945. Type in

Herb. Kew; authentic material in Herb. Pretoriae.

Resupinate, adnate, effused. Margin determinate, about 0-5 mm. wide, light

brown. Hymenium whitish to light buff colour, spread over tooth-like plates which

are poroid at the base and capped with glistening granules. Surface with irregular

anastomosing cracks. Context whitish, 0-5-1 mm. thick.

Basidia: 9-13 X 2-5-4 g, cylindric or clavate, hyaline, very small, with two or four

short straight sterigmata.

Spores: not seen with certainty; possibly broad elliptic or sub-globose, about 1 -5 g diam.

Cystidia: thin-walled, fusoid or cylindric, light yellow-brown, with homogeneous

contents becoming hyaline when old, embedded in hymenium, seldom emergent,

4-6-(8) x 26-36 g. There seem to be similar cystidia embedded in strata especially

towards the base of the trama, but they are colourless and very hard to see clearly.

60

Tissues: hyphae throughout are hyaline, very narrow, very much branched, very densely

intertexed so that individual hyphae cannot be teased out. There are indications

of strata showing that the fungus is perennial. In the latest hymenium the troughs

of the pores are becoming filled with a network of young hyphae.

Minerals: large, often yellowish, irregularly shaped, in concretions up to 120 /i diam.,

scattered throughout the tissues and in the hymenium.

Specimens examined: (All as G. mappa in Herb. Pretoriae.) Type, 34963 (T.R.L.

2018), cited above; 31604, on Gymnosporia, Horseshoe, C.P., 11/8/1915;

11289, v. d. Bvl, on Ilex mitis, Katberg Main Forest, 15/8/1915; 36434,

Rump (766), Bulwer, Natal, Feb. 1947.

Macroscopically this species is very similar to G. mappa B. and C. ! from Cuba,

but the new species has not quite so coarse an appearance. The arrangement of the

teeth and cracks in the two is identical, and a marked contrast to the poroid hymenium

of G. cineracea Bres. ! , and the linear striae of G. lineata B. and C. ! and G. po/ygramma

B. and C. !, which also differ microscopically.

The new species is distinguished from G. mappa by: — -

(1) Smaller basidia. (Basidia of G. mappa are 2-5-4 x 16-19 g);

(2) The presence of coloured or hyaline cystidia embedded in the tissues;

(3) The absence of erect sheaves of brownish hyphae capped by minerals. These

embedded, or rarely emergent, structures are what Patouillard called “ sterile

cylindrical emergences ” [Essai sur les Hym. (1900) 62, f. 44], In G. pseudo-

mappa they are represented merely by large mineral concretions lacking any

hyphal base or core whatsoever. It should be added that Patouillard’s

fig. 44 is exceptionally misleading. For comparison Plate 45, drawn from

type material of G. mappa, is presented.

Lloyd [Myc. Writ. 5 (1916) 581] first cast doubt on the identity of the South African

collections with G. mappa, mentioning the cystidia and the lack of “ sterile emergences.’

Van der Byl [in Ann. Univ. Stellenbosch 12 (1934) 8] describes as G. mappa specimens

which are now referred to G. pseudomappa, while the collection cited by Simpson and

Talbot [in S.A. Journ. Sci. (1946) 133] as G. mappa becomes the type of the new species.

G. mappa is apparently not a South African species.

It may be noted that Lloyd [Myc. Writ. 7 (1923) 1214], in an article headed “ Mis-

cellaneous Notes in Cesati’s Herbarium”, states: “ Grammothe/e mappa is Poria

Ravenelae." The former is certainly not synonymous with P. Ravenalae B. and Br. !.

Possibly this reference is simply a correction of the determination in Cesati’s Herbarium.

GRANDINIA Fries.

1. Grandinia bicolor Talbot apud Wakefield and Talbot in Bothalia 4 (1948) 947, fig. 6.

Type : 27756, Rump (100), Town Bush, Maritzburg; In Herb. Kew, with isotype

in Herb. Pretoriae.

Resupinate, effused, cracking but little on drying, separable from the substratum.

Margin lifting slightly on drying, determinate, light brown, pubescent. Hymenium

pallid, with a pale greenish-yellow, glaucous, tint. Context Prout’s brown colour.

Spines densely crowded, hemispherical or pointed.

Basidia: hyaline, cylindric-clavate, 14 17 x 4-5 //.

Spores: hyaline, smooth, elliplic-cylindric, usually with one side flattened, 7—8 - 5— ( 1 0)

X 2 • 8-3 ■ 5 g.

Hyphae: dark brown, moderately thick-walled, 3 g wide, densely packed adjacent to

the substratum, extending in a looser horizontal network in the mid-trama and

sweeping up in numerous vertical branches into the hymenium. Occasionally

61

several hyphae aggregate into a thick cord in the trama. Between these conspicu-

ous hyphae is a system of branched, nearly hyaline, filamentous hyphae sometimes

seen protruding as branched paraphyses ( ? dendrophyses) from the apex of the

spines.

Minerals: present as crystals or granules.

Specimens examined: Type, cited above; 28502, Rump (270); 28291 Rump (215);

28292, Rump (217); all from type locality, Aug. and Oct. 1934.

A most distinctive species differing from all known species of Grandinia in the colour

contrast of hymenial and abhymenial surfaces, and in the hyphal systems.

GRANDINIA species recorded from south Africa but not available for study:

rosea P. Henn.; Van der Byl in Ann. Univ. Stellenbosch 12 (1934) 8.

ODONTIA Persoon ex S. F. Gray.

1. Odontia arguta (Fr.) Quelet, Flor. Myc. de Fr. (1888) 435; Bresadola, Fung. Kmet.

(1897) n. 114; Bourdot and Galzin, Hym. de Fr. (1928) 427; Rea, Brit. Basid.

(1922) 648; Wakefield in Trans. Roy. Soc. S. Australia 54 (1930) 157; van der Byl

in Ann. Univ. Stellenbosch 12 (1934) 8.

Hydnum argutum Fries, Syst. Myc. 1 (1821) 414, Hym. Eur. (1874) 616.

Hydnum stipitatum Fr. sensu Romell (fide Bourd. and Galz.).

Illustrations'. Bourdot and Galzin (1928) l.c., f. 114.

Plate 46.

Effused, resupinate, rather adnate, dry. Margin indeterminate, whitish. Hy-

menium creamy to ochraceous, with granuliform or subulate spines up to 1 mm. long,

usually much shorter, discrete or basally united, penicillate at the apex. Context

floccose, with a little mineral content.

Basidia: roughly clavate, 3-5 x 13-20 g, with four sterigmata.

Spores: hyaline, broadly ovate, smooth, often uniguttulate, 3-4 x 5 a.

Cystidia: small, subulate, with a capitate encrusted apex, the capitate part 8 x 2 • 5-

3-5 y; or cylindrical with rounded apex and crowned with a yellowish resinous

globule; cystidia borne apically or laterally on the spines.

Hyphae: hyaline, with many clamp connections, 2-(4) ;i diam.

Specimens examined: 28883, Bottom/ey, The Cavern, Drakensberg, Natal, 20/7/

1937; 28313, Doidge and Morgan, nr. Donnybrook, Natal, Feb. 1935;

Acock (v. d. Byl, 2638), Stickland, Feb. 1933; van der Byl (2726), South

Africa, Oct. 1933.

Of the specimens cited above, No. 28883 has both spathulate and globular cystidia.

The two collections of van der Byl have resinous globules but no spathulate cystidia,

and in this they correspond with the Australian forms described by Miss Wakefield

(1930, l.c.). On occasions only spathulate cystidia may be found.

2. Odontia knysnana van der Byl ! in Ann. Univ. Stellenbosch 12 (1934) 9.

Illustrations'. Plate 47.

Resupinate, adnate, confluent, white, later creamy, with byssoid margin. Context

thin, effused. Teeth minute, discrete though close together, 6-8 per mm., granuloid,

microscopically ciliate at the apex.

Basidia: clavate-cylindric, 4 x 23 p.

Spores: hyaline, smooth, elliptical, often with one side flattened, apiculate towards the

side, 2-5-3 x 5 /i.

62

Cystidia: mostly at the apex of the spines, fine pointed, about 35 x 3-4 p., encrusted

with crystals readily soluble in KOH, hyaline.

Hyphae: thin-walled, 3-4 p. diam., with clamp connections, much branched, densely

and suberectly interwoven, encrusted with granules, becoming smooth in KOH

after some time.

Specimens examined: Type, v. d. Byl { 1332), Knysna, Jan. 1924. (Part of Type in

Herb. Kew.)

This species is not far removed from Odontia crustosa (Pers.) Quelet. The latter,

however, has non-encrusted cystidia and hyphae. Its hyphae have only rare clamps

and appear to collapse more readily than those of O. knysnana. As there is only one

collection of O. knysnana in existence, it is not possible to assess whether it is a “ good ”

species or not.

HYDNUM Linn, ex Fries.

RESUPINATE HYDNUM SPECIES RECORDED FOR SOUTH AFRICA BUT NOT AVAILABLE FOR

STUDY.

ambiguum B. and Br.; Kalchbrenner in Grev. 10 (1881) 57, from Kaffraria.

mucidum Fr. ; Kalchbrenner in Grev. 10 (1881) 57, from Somerset East; Saccardo,

Syll. Fung. 6 (1888) 471.

sclerodontium Mont. & Berk.; in Litt. Timber Research Lab. (165), from Rand

gold mines.

SYNONYMS AND EXCLUDED SPECIES IN HYDNUM.

ochraceum Pers. ex Fries —Mycoleptodon ochracenm (Pers. ex Fr.) Pat. The

Medley Wood (479) collection referred to as H. ochraceum in Grev. 10 (1881)

27 and S.A. Journ. Sci. 13 (1917) 440 lacks cystidia and is some other species

of Hydnum.

pudorinum Fries. [Recorded in Grev. 10 (1881) 57.]

—Mycoleptodon ochraceum (Pers. ex Fr.) Pat. q.v.

RADULUM Fries.

SPECIES RECORDED FOR SOUTH AFRICA BUT NOT AVAILABLE FOR STUDY:

javanicum (P. Henn.) Lloyd.; Lloyd, Myc. Writ. 6 (1920) 952.

orbiculare Fries; Kalchbrenner in Grev. 10 (1881) 57, from Somerset East.

SYNONYMY AND EXCLUDED SPECIES IN RADULUM:

javanicum (P. Henn.) Lloyd =Lopharia javanica P. Henn.

mirabile B. and Br. ! =Lopharia mirabilis (B. and Br.) Pat. q.v.

THELEPHORA Ehrhart ex Fries.

NOTES ON RECORDS OF RESUPINATE SPECIES OF THELEPHORA FROM SOUTH AFRICA:

acerina Pers.; Leveille in Ann. Sci. Nat. ser. 3, 5 (1846) \50 ^Aleurodiscus acerinus

(Pers.) Hohnel and Litschauer, or its variety longisporus Hohnel and Litschauer.

q.v.

biennis Fries; Kalchbrenner in Grev. 10 (1881) 58. The MacOwan collections

of this in S.A. Museum are Stereum bicolor fide v. d. Byl in Ann. Univ. Stellen-

bosch 7 (1929) 42.

fusco-violascens Mont.; Montagne in Ann. Sci. Nat. ser. 3, 7 (1847) 174. =

Hymenochaete fuscovio/ascens (Mont.) Sacc. fide v. d. Byl (1929) l.c. Pg. 14.

63

pedicellata Schwein.; Bottomley in S.A. Journ. Sci. 13 (1917) 440, refers to a

collection Medley Wood No. 532. This collection is a species of Septobasidium.

punicea Alb. and Schwein.; Kalchbrenner in Grev. 10 (1881) 58 = Tomentella

punicea (Alb. and Schwein.).

TOMENTELLA Patouillard.

TOMENTELLA species recorded for south Africa but not available for study:

Eylesii (van der Byl); v. d. Byl in S.A. Journ. Sci. 22 (1925) 168, in Ann. Univ.

Stellenbosch 7 (1929) 18, as Hypochnus Eylesii.

puniceus (Alb. and Schwein.); v. d. Byl (1929) l.c. Pg. 19, as Hypochnus ; Kalch-

brenner in Grev. 10 (1881) 58, as Thelephora\ Wood in Rept. Natal Bot. Gard.

(1898) 19, as Thelephora.

THELEPORA Fries.

1. Thelepora cretacea Fries ! in Hornsch. Skand. Beitr. Naturgesch. 2 (1847) 333, in

Bot. Zeit. 6 (1848) 340.

Theleporus cretaceus Fries (orthographic variant) in Actis Acad. Scient. Holm.

(1848) 138, in Summ. Veg. Scand. 2 (1849) 235; Saccardo, Syll. Fung. 6 (1888)

421 ; Lloyd, Myc. Writ. 3 (1910) 479, fig. 374.

Illustrations : Lloyd (1910) l.c., f. 374.

Plate 48.

Resupinate, effused, light creamy yellow colour in herbarium, closely adnate.

Hymenium poroid. Pores shallow, angular, fairly regular, about 3-4 per mm., with

thin dissepiments, each pore with a small basidia-bearing papilla in the centre which

originates in the same manner as the pore walls.

Basidia: not reviving properly on mounting treatment but thought to be irregularly

clavate, about 18 x 3-5 y.

Spores: none seen attached. A few spores seen, globose, hyaline, rather thick-walled,

3-5 y diam., but it is doubtful whether these really belong.

Hyphae: thick-walled, 4 y wide, in a close horizontal network below the pore layer.

Specimens examined : Part of the Friesian Type in Herb. Kew, collected by Wahlberg

in Natal.

The genus Thelepora is closely allied to Poria and may prove to be merely a freak

Poria. Thelepora cretacea comes very near to Poria versipora (Pers.) Fr. in structure.

Fries writes of the papillae in the pores thus : “ This definite formation is not only

peculiar and remarkable, but also serves to distinguish it from PorotheliumP There

is not in fact much resemblance to a Porothelium.

One other species of Thelepora has been described. This is T. griseus Rick [in

Broteria 5 (1906) 15; Saccardo, Syll. Fung. 21 (1912) 361], a Brazilian species of greyish

colour, with round or polygonal, coloured spores, 3-6 y diam. The writer has not

seen any material of this species.

LATIN DIAGNOSES OF NEW SPECIES.

Peniophora pellieulosa sp. nov.

Fungus resupinatus, late effusus, crassus, molliter pelliculosus, udus separabilis.

Color hymenii plus minus roseobubalinus vel stramineus. Subiculum candidum,

floccosum, 550-1,000 y crassum. Basidia cylindraceo-clavata, dense fasciculata,

5-6 x 16-23 y (immatura). Sporae oblongae vel ellipsoideae, laeves, hyalinae, 3-3-3

X 5-7 y. Cystidia in strato superiore 70-100 y crasso immersa, raro usque 12 y

emergentia, dense incrustata vel capitatim incrustata, tenuiter tunicata, interdum apice

1599—3

64

incrassata, clavato-cylindracea, 20-36-(50) x (8) -10-13 \x. Hyphae hyalinae; hyphae

hymeniales et subhymeniales non distinctae; hyphae subiculi distinctissimae, ramosae

laxe intertextae, septatae, basales 3 -5-5-6 /x crassae, subhymenio interdum granulis

incrustato.

Hab. ad lignum, Krantzkop, Natal.

Coniophora fodinarum sp. nov.

Fungus resupinatus, late effusus, facile separabilis, siccitate fragilis haud rimosus,

ad 2 mm. crassus. Hymenium tuberculosum, olivaceo-umbrinum vel umbrinum.

Superficies superior mollis, velutina, fusca. Contextus pallidus. Interdum fila rhizoidea

adsunt. Basidia non visa nisi collapsa. Sporae magnitudine variae, 5-6-6 -6 X

6-6-7-5-8-3-10 /i, subglobosae vel pierumque ellipsoideae, saepe uno latere piano,

coloratae, laeves, liberatae vel inter hyphas hyalinas ramosas filamentas in strato ad

260 n crasso immersae. Hyphae (1) strati sporogenosi hyalinae, ramosae, ad 1 //. diam.

(2) strati intermedii (500-660 fx crassi) pallide coloratae, 2 /x diam., dense intertextae (3)

superiores tenuiter tunicatae, 3-(6) /x diam., fuscae, stratum fuscum spongiosum circiter

250 [i crassum efformantes.

Hab. ad lignum in fodinis Transvaaliae.

Acia conferta sp. nov.

Fungus effusus, arete adnatus, subcereus interdum crustaceus margine indeter-

minato. Aculei conferti, papillosi vel subulati vel ex parte basi coalescentes, fulves-

centes vel castanei, 0 - 3—1—1 - 5 mm. longi. Basidia clavata, hyalina, 16-28 x 4-5-3 fx.

Sporae laeves, ellipticae, hyalinae, saepe uno latere depressae, interdum uniguttulatae,

( 3 • 5)— 4— 5 • 5— (6) X 2-5-4 jx. Cystidiola sparsa, cylindrica, hyalina, apicibus rotundatis,

3-4 x 40-46 ;x. Hyphae basales tenuiter tunicatae, sparsae, hyalinae, haud nodosae,

2-4 [i diametro.

Hab. ad lignum : Pietermaritzburg, Natal.

Grammothele pseudomappa sp. nov.

Fungus resupinatus, effusus. Margo definitus, circiter 0-5 mm. latus, pallide-

brunneus. Hymenium sordide album vel bubalinum, dentibus vel lamellis curtis

confertis, basi porosum, apice corrusco-granulosum. Contextus sordide albus,

0- 5-1 mm. crassus. Basidia 9-13 x 2-5-4 fx, cylindrica vel clavata, hyalina; sterig-

matibus 2-4, rectis. Sporae non certe visae; late ellipsoideae subglobosae, circiter

1- 5 fi diam. Cystidia fusoidea vel cylindrica, tenuiter tunicata, pallide flavo-

brunnea, senectute hyalina, immersa, raro emergentia, 4-6-(8) X 26-36 /x. Hyphae

hyalinae, ramosissimae, tenuissimae, dense intertextae. Crystalla magna, saepe flavida,

irregulariter formata, usque ad 120 tx diam., in trama et hymenio immersa.

Hab. ad lignum: Qudeni, Zululand.

EXPLANATION OF THE PLATES.

The illustrations were drawn with the aid of a Reichert drawing apparatus. In

general, a transverse section through the body of the fungus is shown, and in addition

there are separate drawings of various individual organs at a higher magnification.

The magnifications are indicated by linear scales on the plates, the smaller scale, when

present, invariably referring only to the sectional drawing.

A broken line running horizontally across a sectional drawing indicates that the

full width of the section has not been included. The following scheme of lettering has

been used throughout the illustrations: —

B =basidia. V ^conducting vessels.

S =spores. E = setae.

C =cystidia or cystidioles. FE =false setae.

65

G =g!oeocystidia. AE =asterosetae.

H =hyphae. M ^mineral inclusions.

D =view in transverse section. O =oil globules.

X =substratum tissue. SH =hairs of the abhymenial surface.

INDEX TO GENERA AND SPECIES.

Acia, 5, 59.

conferta, 59, 63, PI. 43.

fusco-atra, 59.

subceracea. 59, PI. 43.

Aleurodiscus, 5, 26.

acerinus, 19, 27, 28, 62.

acerinus var. alliaceus, 26, 27.

acerinus var. longisporus, 19, 26, 27, 28, 62,

PI. 18.

cerussatus, 28.

disciformis, 28.

Aspergillus, 39.

Asterostroma, 5, 54.

albido-carneum, 54.

cervicolor, 54, 55, PI. 39.

corticola, 54.

pallidum, 54.

Asterostromella, 5, 51, 54.

dura, 51, 53, PI. 37.

duriuscula, 19, 45, 51, 54, PI. 36.

Rumpiana, 53, 54, PI. 38.

Auricularia, 5, 6, 29.

Butleri, 29.

corrugata, 6.

mesenterica, 6, PI. 1.

mesenterica var. lobata, 6.

papyrina, 30.

phosphorea, 1.5.

reflexa, 28.

rugosissima, 28.

sordescens, 29.

strigoso-zonata, 28.

tremelloides, 6.

Boletus purpurascens, 30.

Byssus phosphorea, 15.

Ceriomyces venulosus, 25.

Cladoderris Pritzelii, 57.

Coniophora, 5, 33, 36.

arida, 36.

atrocinerea, 35, 36.

Betulae, 36, 37.

brunneola, 35.

cerebella, 36.

Ellisii, 35.

fodinarum, 34, 36, 64, PI. 23, Fig. 2.

fumosa, 34, 35, 36.

leucothrix, 35.

olivacea, 34, 35, 36, PI. 24.

papillosa, 33, PI. 23, Fig. 1.

puteana, 34, 36.

sistotremoides, 34.

sordulenta, 53, PI. 37.

suffocata, 37.

ConiophoreUa, 36.

atrocinerea, 35, 36.

olivacea, 35, 36.

Corticium, 5, 12.

abeuns, 13, 19.

acerinum, 28.

alliaceum, 26.

armeniacum, 12, 16, 19, PI. 8.

aschistum, 40.

atrocinereum, 19, 44.

aurora, 18.

Berkeley i, 21.

bombycinum, 19.

brunneolum, 35.

cacaoieolor, 17.

caeruleum, 12, 15, PI. 7.

calceum, 19, 27, 28.

centrifugum, 15.

ceraceum, 16, 19.

cervicolor, 54.

cinereum, 19, 24.

confluens, 13, 18, 19, PI .11.

confluens var. subcalceum, 18.

diminuens, 52.

Dregeanum, 19, 57, 58.

Ellisii, 35.

ephebium, 41.

filamentosum, 23.

giganteum, 24.

gloeosporum, 12, 14, 15, PI. 6.

grammicum, 52.

hinnuleum, 54.

javanicum, 17.

lacteum, 1 9.

laetum, 18.

laeve, 14, 15.

latitans, 21.

leucothrix, 35.

luteocystidiatum, 12, 13.

mauritianum, 16.

molle, 16, 58.

Mollerianum, 22.

myxosporum, 21.

nudum, 19, 25.

ochraceo-fulvum, 59.

ochrofarctum, 13.

olivaceum, 35.

pallidum, 12, 13, 14, PI. 5.

pelliculare, 19.

Petersii, 23.

portentosum, 19, 51, 52, 53.

radicatum, 23

roseo-cremeum, 21.

roseolum, 18.

roseum, 18.

Roumeguerii, 22.

salmonicolor, 13, 17, 18, PI. 10.

scutellare, 12, 16, PI. 9.

seriale, 13.

tristiculum, 46.

tuberculosum, 25.

tumulosum, 12, 17.

vagum, 19.

Zimmermanni, 17.

66

Cytidia flocculenta, 15.

Dichostereum, 53.

Duportella, 5, 43, 46, 50.

tristicula, 45, 46, 47, 58, PI. 31, PI. 32.

tristiuscula, 46.

velutina, 46.

“ Erineum ” gall., 12, PI. 4.

Exobasidium, 5, 11.

Andromedae, 11.

Arctostaphyli, 11.

Azaleae, 1 1 .

cassandrae, 11.

discoideum, 1 1 .

Giseckiae, 77, PL 3, Fig. 2.

Karstenii, 1 1 .

Myrtilli, 11.

oxycocci, 11.

Peckii, 1 1 .

Rhododendri, 11.

Vaccinii, 77, PI. 3, Fig. 1.

Vaccinii-Myrtilli, 1 1 .

Fomes, 55.

Fusidium Vaccinii, 1 1 .

Gloeocystidium.

argillaceum, 13, 14.

pallidum, 13

pallidum subsp. argillaceum, 13.

Grandinia, 5, 60.

bicolor, 60.

rosea, 61.

Grammothele, 5, 59.

cineracea, 60.

liniata, 60.

mappa, 59, 60, PI. 45.

polygramma, 60.

pseudomappa, 59, 60, 63, PI. 44.

Gyrophana himantioides, 31.

Helicobasidium, 5, 7.

compactum, 7, PI, 2.

longisporum, 7, 8.

mompa, 8.

mompa forma macrosporum, 7.

purpureum, 8.

purpureum var. orientale, 7.

Helicobasis purpureus, 8.

Helvella mesenterica, 6.

Heterochaete andina, 10.

Hydnum, 5.

alnicolum, 55.

ambiguum, 62.

argutum, 61.

Daviesii, 55.

microdon, 55.

mucidum, 62.

ochraceum, 55, 56, 62.

pudorinum, 55, 62.

sclerodontium, 62.

stipitatum, 61.

Hymenochaete, 5, 43, 48.

bonariensis, 41.

castanea, 46, 47.

cervina, 49.

cinerascens, 40.

corrugata, 48.

crassa, 42.

Dregeana, 57, 59.

Ellisii, 35.

fasciculata, 48, 49, PL 33.

fuliginosa, 48, 49.

fuscoviolascens, 50, 62.

griseocervina, 43.

Kalchbrenneri, 42.

lictor, 49.

luteobadia, 45, 50.

multispinulosa, 42.

nigricans, 45.

ochromarginata, 48, 49, 50, PL 35.

olivaceum, 43.

purpurea, 42.

rigidula, 50.

rubiginosa, 45, 50.

scabriseta, 42.

semistupposa, 48, 49, PL 34.

tabacina, 50.

tenuissima, 50.

tristicula, 46, 47, 50.

tristiuscula, 46.

umbrina, 42.

vinosa, 42.

Hypochnus.

acerinus, 28.

Bagliettoanus, 9.

Eylesii, 63.

olivaceus, 35.

pallescens, 53, PL 37.

puniceus, 63.

purpureus, 8.

thelephoroides, 53.

Irpex, 1 1 .

Kneiffia.

purpurea, 42.

Roumeguerii, 22.

setigera, 21.

Leptodon ochraceum, 55.

Licentia yao-chanica, 56, 57.

Lloydella, 57.

Beyrichii, 40.

bicolor, 39.

cinerascens, 40.

occidentalis, 41.

scabriseta, 42.

Schomburgkii var. brunnea, 43.

Lopharia, 5, 56.

Dregeana, 19, 47, 57, 58, PL 42.

javanica, 57, 58, 62.

iirellosa, 56, 57.

mirabilis, 41, 56, 57, 58, 59, 62, PL 41

Merulius, 5, 29.

confluens, 30, 31, 33.

corium, 29, 30, 31, 33, PL 20.

gelatinosus 29, 32, 33, PL 22.

himantioides, 29, 31, 32, 33, PL 21.

lacrymans, 32, 33, PL 21.

papyrinus, 30.

serpens, 33.

squalidus, 31, 32, 33.

strigoso-zonatus, 28.

tenuis, 31.

umbrinus, 31, 32, 33.

67

Mycinema phosphoreum, 1 5.

Mycoleptodon ochraceum, 55, 62, PI. 40.

Necator decretus, 17, 18.

Odontia, 5, 61.

arguta, 61, PI. 46.

crustosa, 62.

knysnana, 61, 62, PI. 47.

setigera, 21.

Peniophora, 5, 20.

arenata, 20, 22.

argillacea, 13, 14.

aspera, 20, 21, 36, PI. 13.

atrocinerea, 43, 44.

Berkeleyi, 40.

byssoidea, 21.

cameola, 13.

cinerascens, 40.

cinerea, 19, 24, 25, 42.

cremea, 20, 25.

dissita, 41.

ephebia, 41.

filamentosa, 20, 23, 24, PI. 15.

gigantea, 20, 23, 24, PI. 16.

glebulosa, 25.

intermedia, 42,

laevigata, 47.

Ieprosa, 22.

macra, 22.

Molleriana, 22.

moricola, 40.

neglecta, 41.

nuda, 19, 25.

occidentalis, 41 .

pallidula, 21.

pelliculosa, 20, 25, 63, PI. 12.

pubera, 23.

radicata, 23.

Roumeguerii, 20, 22, 23, 41, PI. 14.

Schweinitzii, 40.

setigera, 21.

stratosa, 22.

subiculosa, 20.

tomentella, 21.

unicolor, 23.

Phlebia, 5, 28, 29.

mesenterica, 6.

hispidula, 29.

pileata, 28.

reflexa, 28.

rubiginosa, 28.

rugosissima, 28.

strigoso-zonata, 28, PI. 19.

zonata, 28.

Polvporus

nidulans 24.

punmrascens, 30.

rutilans, 24.

Polystictus tabacinus, 50.

Poria, 11, 57, 63.

Ravenalae, 60.

versipora, 63.

Porothelium, 63.

Ptychogaster, 26.

Punctularia, 5, 25.

affinis, 25, PI. 17.

atropurpurascens, 25, 26.

tuberculosa, 25, 26.

Radulum, 5, 57, 62.

javanicum, 62.

membranaceum, 19.

mirabile, 56, 62.

orbiculare, 62.

Reticularia

affinis, 25, 26.

atropurpurascens, 25.

atrorufa, 25, 26.

pyrrhocreas, 26.

venulosa, 25, 26.

Rhizoctonia, 8.

crocorum, 8.

medicaginis, 8.

violacea, 8.

Sclerotium crocorum, 8.

Sebacina, 5.

africana, 10.

epigaea, 10.

Septobasidium, 5, 8, 63.

Bagliettoanum, 9, 10.

bogoriense, 9, 10.

carestianum var. natalense, 9, 10.

compactum, 7.

Curtisii, 9, 10.

grandispinosum. 9, 10.

griseopurpureum, 9.

mompa, 7, 10.

natalense, 9.

philippinense, 9, 10.

protractum, 9.

pseudopedicellatum, 9.

Schweinitzii, 9.

Steccherinum ochraceum, 55.

Stereum, 5, 37.

acerinum, 28.

acerinum var. longisporum, 26

adnatum, 38, 39, 45.

albobadium, 45.

albocinctum, 52, 53, PI. 37.

atrocinereum, 43, 44, 45.

balsameum, 37.

balsameum f. reflexum, 37.

bicolor, 37, 39, 40, 45, 62, PI. 26.

cinerascens, 37, 40, 41, 57, 58, 59, PI. 27.

coffeatum, 39.

crassum, 42.

crispum, 37.

dissitum, 41.

duriusculum, 45, 51, 52, 53.

elegans, 45.

fasciatum, 45, 50. — >

fuscum, 39, 45.

induratum, 52, 53, PI. 37.

laxum, 45.

lugubris, 29.

luteobadium, 45.

membranaceum, 43, 44, 45.

moricolum, 40.

neglectum, 41.

pannosum, 39, 40.

papyrinum, 43.

68

percome, 45, 47.

pruinatum, 45.

purpurascens, 41.

retirugum, 43.

rigens, 38.

rimosum, 39, 45.

rimosum var. africanum, 37, 38, 39, 45.

rubiginosum, 45, 50.

sanguinolentum, 37, 38, PI. 25.

Schomburgkii, 19, 37, 40, 43, 44, 45, 47,

58, PI. 29, PI. 30.

Thozetii, 45.

umbrinum, 25, 37, 41, 43, 47, PI. 28.

Stypinella purpurea, 8.

Thanatophyton crocorum, 8.

Thelephora, 5, 62.

acerina, 27, 28, 62.

albido-carnea, 54.

atro-coerulea, 15.

atropurpurascens, 25.

Beyrichii, 40.

bicolor, 39.

biennis, 62.

caerulea, 15.

cinerascens, 40.

confluens, 18.

corium, 30.

crassa, 42.

fimbriata, 15.

fusca, 39, 40.

fuscoviolascens, 62.

gigantea, 24.

hirsuta B, 37.

incarnata var. B, 30.

indigo, 15.

insinuans, 53.

mesenterica, 6.

olivacea, 36.

pallescens, 53.

pannosa, 39.

pedicellata, 63.

pergamenea, 24.

punicea, 63.

sanguinolenta, 37.

sericea B, 37.

vinosa, 39.

Thelepora cretacea, 63, i*l. 48.

Theleporus

cretaceus, 63.

griseus, 63.

Thwaitesiella mirabilis, 56.

Tomentella, 5.

Eylesii, 63.

punicea, 63.

Tremella

epigaea, 10.

Trichosporium Curtisii, 25.

Tuber croci, 8.

Vararia, 54.

pallescens, 53.

69

Plate 1.

Auricularia mesenterica (Dicks ex Fr.) Fr.

70

Plate 2.

Helicobasidium compactum (Boedijn) Boedijn.

71

Plate 3.

Fig. 1. — Exobasidiumvacciniic (Fuckel) Woronin.

Fig. 2. — Exobasidium Gisekiae /tAllescher.

72

Plate 4.

Transverse section of an “ Erineum ” gall on the leaf of Gymnosporia buxifolia.

73

74

Plate 6.

Corticium gloeosporum Talbot.

75

76

Corticium armeniacum Sacc.

77

Plate 9.

Corticium scutellare B. & C.

Plate 10.

Corticium salmonicolor B. & Br.

79

Plate 11.

Corticium confluens (Fr.) Fr.

80

Plate 12.

Peniophora pelliculosa sp. nov.

81

Plate 13.

Peniophora aspera (Pers.) Sacc.

82

CD

U

*4

oor

Q

Plate 14.

Peniophora Roumeguerii (Bres.) Burt.

83

Plate 15.

Peniophora filamentosa (B. & C.) Burt.

84

Plate 17.

Fig. 1. — Punctularia affinis (B. & C.) comb. nov. conidial stage.

Fig. 2. — Punctularia affinis (B. & C.) comb. nov.

86

Plate 18.

Aleurodiscus acerinus var. /ongisporus Hohnel & Litsch.

87

Plate 19.

Phlebia strigoso-zonatci (Schwein.) Lloyd.

89

Plate 21.

Fig. 1. — Merulius himantioides Fries.

Fig. 2. — Merulius laerymans Wulf. ex Fr.

90

91

^ 1 rw*DOT

,-ao^q

o-' - ^

^ -/"-O-. Q/'Qn,:

0 ^ ..'j > o

^T-o.

- ;^|ff

^y0-. -

cS=^~?^- ;Vvfe

^C'.VnTF

4|#>o4^l

0 ^-eP

0,'CT^ - 0

to

' <?

Plate 23.

Fig. 1. — Coniophora papillosa Talbot.

Fig. 2. — Coniophora fodinarum sp. nov.

g§ru *

92

n r-">

93

Plate 25.

Stereum sanguinolentum (A. & S.) Fries.

94

Plate 26.

Ster'eum bicolor (Pers.) Fr.

95

DDT

1599-4

Plate 27.

Stereum cinerascens (Schw.) Massee.

96

Plate 28.

Stereum umbrinum B. & C.

97

Plate 29.

Stereum Schomburgkii Berk. Mature stage.

IQO

Plate 30.

Stereum Schomburgkii Berk. Young stage.

KJQ

99

Plate 31.

Duportella tristicula (B. & Br.) Reinking. Young stage.

100

101

Plate 33.

Hymenochaete fasciculata Talbot.

102

103

Plate 35.

Hymenochaete ochromarginata Talbot.

104

1<X>

105

mentioned in the text.

Fig. 1. — Swedish collection, Romell (4135), in Herb. Kew. Fig. 4. — Asterostromella dura Bourd. & Galz.

Fig. 2.—Stereum induratum B. & C. Fig. 5. — Hypochnus pallescens (Schw.) Burt.

Fig. 3. — Stereum albo-cinctum B. & Br. Fig. 6. — Coniophora sordulenta (Cooke & Massee) Sacc.

106

UJ

<

Plate 39.

Asterostroma cervicolor (B. & C.) Massee.

108

Plate 40.

Mycoleptodon ochraceum (Pers. ex. Fr.) Pat.

m

109

LQQ

no

Plate 42.

Lopharia Dregeana (Berk.) comb. nov.

Ill

Plate 43.

Fig. 1. — Acia subceracea Wakef.

Fig. 2. — Acia conferta sp. nov.

112

Plate 44.

Grammothele pseudomappa sp. nov.

113

,1

Plate 45.

Grammothe/e mappa B. & C.

114

Plate 46.

Odontia arguta (Fr.) Quelet.

115

116

117

A Morphological, Anatomical and Cytological study

of Potamophila prehensilis (Nees) Benth.

By B. de Winter.

(Division of Botany and Plant Pathology, Pretoria.)

(Being part of the thesis presented in partial fulfilment of the degree of Master of Science

at the University of Pretoria.)

CONTENTS.

1. Introduction 117

2. Description of the Genus 117

3. Description of the South African Species 118

4. Anatomical Features of the South African Species 119

5. Cytological Investigation of the South African Species 121

6. Remarks and Discussion 121

7. Material Examined 125

8. Economic Notes 126

9. Distribution of the South African Species 126

10. Summary 126

1 1 . Literature Cited 127

1. INTRODUCTION.

Potamophila prehensilis (Nees) Benth. was first described by Nees in his FI. Afr.

Austr. (1841) under the name Maltebrunia prehensilis. Bentham in 1881 included the

genus Maltebrunia Kunth under Potamophila R. Br.

Kuntze in his Rev. Gen. Plant. (1893) and Stapf in Dyer, FI. Cap. VII (1900)

described the uppermost florets of Potamophila as hermaphrodite or unisexual with the

organs of one sex reduced. Phillips in his description of this genus in his S. Afr. Grasses

(1931) makes no mention of this reduction. Moreover, when the descriptions of

P. prehensilis and P. parviflora (type species for the genus Potamophila R. Brown) were

compared, it appeared rather doubtful whether P. prehensilis should be included under

this genus.

The present work is an attempt to clear up the doubtful points mentioned above,

as well as to investigate the alleged relationship between the Oryzeae and the Bambuseae,

2. DESCRIPTION OF THE GENUS.

POTAMOPHILA R. Brown.

Potamophila R. Brown, Prodr. FI. Nov. Holl., ed. 2, 1, 67 (1827); Steudel, Syn. Plant.

Glum. 1, 4 (1854); Bentham, FI. Austr. VII, p. 550 (1878); Bentham et Hooker,

Gen. Plant. Ill, 2, p. 1116 (1883); Stapf in Dyer, FI. Cap. VII, 658 (1900); Medley

Wood, Natal, Plants, t. 444 cum descr. (1905); Stent in Bothalia, I, 4, 274 (1924);

Phillips, Gen. S. Afr. FI. PI., 82 (1926); Bews, World’s Grasses, t. 21, p. 75 (1929)

in clavi p. 122; Phillips, S. Afr. Grasses, 169 (1931).

Spikelets laterally slightly compressed on short or rather long pedicels. Inflores-

cence a lax or contracted panicle. Rhachilla disarticulating above the rudimentary,

basal glumes. Florets 3 in number only the uppermost fertile, the lower two reduced to

118

minute empty lemmas. Uppermost hermaphrodite or unisexual with the organs of

one sex reduced. Glumes persistent, reduced to a minute entire or bi-lobed hyaline rim.

Empty lemmas reduced, subulate, elliptic and rounded or lobed. Fertile lemma mem-

branous, 5-nerved, nerves sometimes raised or even slightly winged. Palea more or

less similar to the fertile lemma but slightly longer and 3-nerved. Lodicules 2, usually

large, finely nerved. Stamens 6. Styles 2, distinct; stigmas plumose. Caryopsis

obovate, compressed or terete, crowned by the thickened bases of the styles, enclosed

by the unaltered lemma and palea, but free.

Tall aquatic or hygrophytic grasses; ligules membranous; panicle effuse or con-

tracted.

A genus with four species; two in Africa (East Africa one, South Africa one);

Madagascar one, New South Wales one.

3. DESCRIPTION OF THE SOUTH AFRICAN SPECIES.

Potamophila prehensilis (Nees) Benth.

in Journ. Linn. Soc. 19, 55 (1881); Medley Wood, Natal Plants, t. 444 cum descr.

(1905); Stapf in Dyer, FI. Cap. VII, 659 (1900); Stent in Bothalia 1, 4, 274 (1924)

sphalm. prehensiles; Bews, World’s Grasses, t. 21, 75 (1929); Phillips, S. Afr.

Grasses, t. 48 (1931).

Synonymy. Maltebrunia prehensilis Nees, FI. Afr. Austr., 194 (1841); Dur. et Schinz,

Consp. FI. Afr. V, 788 (1894) sphalm, praehensilis\ Oryza prehensilis (Nees)

Steud., Syn. PI. Glum, 1, 3, (1854).

Perennial, rhizome horizontal, covered with cataphylls, in robust plants more or

less strongly branched. Culms erect, climbing, up to 10 m. high or more, slender,

many noded; innovations produced intravaginally; nodes swollen, annular, pallid to

brownish, glabrous or sometimes minutely hairy; internodes hollow, exserted, terete

to slightly laterally compressed, up to 13-5 cm. long, the lower grooved on one side,

striate, pallid to greenish, sparsely villous below the nodes or glabrous. Leaf-sheaths

tight or more or less lax; retrorsely ciliate along the thin margins, sometimes sparsely

long- ciliate at the top or glabrous, slightly keeled, scabrous on the upper part of the

keel, strongly striate, lighter in colour than the leaves, slightly auricled. Ligule up to

1 mm. long, membranous, usually truncate, at length becoming fimbriate. Leaf-blades

heterophyllous, viz., the usual lanceolate vegitative type and a caudate prehensile linear

type (fig. 2). There exists a gradual transition from the vegetative lanceolate type to

the linear (caudate) prehensile type of blade, which besides possessing a normal assimi-

lative function, also enables the plant to climb. The extremes, however, are very

different in appearance. Lanceolate type, acuminate, up to 14-5 cm. long and 0-6 cm.

broad, dark green in colour, flat, more or less flaccid, sparsely hairy, margins and keel

usually spiny with minute sharp retrorse barbs; primary nerves 3-5 on each side of the

mid-nerve; keel and nerves light yellow or white in colour. Caudate prehensile type,

very narrowly linear and much longer than the usual lanceolate type, up to 23-0 cm.

long and 0-4 cm. broad, tapering to form a capillary pendent apex; margins and keel

very strongly spinous with minute retrorse barbs. Occurring mainly as the first few

leaves on young sterile shoots. Flowering branches usually bearing only the lanceolate

type of leaf. Panicles terminal on the main culm as well as on the lateral branches,

ovate, open and lax, up to 14 cm. in length and 10 cm. wide, the spikelets opening

successively from the apex towards the base of the panicle. Branches up to 10 cm.

long, spreading, fine, stiff, the lower-most longest and sparingly branched; branchlets

3-4 spiculate; pedicels up to 2 cm. long. Spikelets up to 9 mm. long, broadly lanceo-

late, slightly twisted, usually glaucous (pallid when young), scabrous, slightly laterally

compressed, pedicelled; rhachilla not produced, disarticulating above the rudimentary

basal glumes. Florets 3, two represented by empty lemmas, the third hermaphrodite

and well developed. Glumes persistent, much reduced, forming a small entire or

119

bilobed hyaline cup. Sterile lemmas subulate, armed with minute hyaline spines,

variable in length. Fertile floret with a well-developed lemma and palea. Lemma

about 8 mm. long, boat-shaped acuminate, coarsely covered with hyaline spines, coria-

ceous clasping the palea with the tough inflexed margins, nerves raised, set with curved

hyaline hairs, at times nearly winged. Palea similar to the lemma with the exception

that it is 3-nerved, narrower, and possesses hyaline margins. Lodicules 2, broadly ovate,

obtuse, fleshy, up to 1-2 mm. long; nerves fine, indistinct. Stamens 6; anthers linear,

pale yellow, 6 mm. long. Ovary subglobose; styles distinct; stigmas densely plumose

up to 3 mm. long. Caryopsis 5-6 mm. long, brown, spindle-shaped, longitudinally

finely ribbed, terete in transverse section. Embryo 1/6— 1/7 the length of the grain;

hilum linear, as long as the grain; epiblast present, about i the length of the plumule

(small). Starch grains compound, more or less globose, each grain composed of about

14 polyhedral granules. Reaction of KI2 with starch normal.

4. ANATOMICAL FEATURES OF THE SOUTH AFRICAN SPECIES.

The abbreviations used in the text figures designate the following: —

Shoot Characters. (Fig. 3D.)

Shoots circular or elliptic in transverse section. Leaves rolled in the bud, protected

by the sheaths of the lowermost leaves.

Anatomy of the Leaf Blade. (Figs. 4, 5, 6, 7.)

The description of the anatomy is based on observations made from handcut trans-

verse sections. The epidermi were studied by the aid of surface sections, and maceration

of leaf material made it possible to study the shape and size of the cells. Sections were

cut near the base of the leaf to bring out certain characteristics. Drawings were made

with the aid of a Leitz “ Panphot.”

The lanceolate type of leaf has the following characteristics :

Leaf blade flat on both surfaces, extremely thin, about 0-055 mm. in cross-section,

only about 4-6 cell-layers thick, including the epidermi. Keel prominent rounded,

retrorsely barbed. Adaxial epidermis (fig. 6), with 2-3 rows of motor cells between

the bundles, followed by 5-6 rows of ripple- walled cells; the rather narrow and con-

stricted silicified cells adjacent to the bundles alternating with the broader suberized

cells. Stomata occurring on both sides of the bundles; in a double row next to first and

second order bundles, whereas in a single row next to third order bundles. Papillae

present on the motor cells. Ripple-walled cells densely covered with papillae. Narrow

and shorter cells slightly resembling silicified cells occur between the ripple-walled

elements (e.g., long and short elements: Agnes Arber, The Gramineae). Two-celled

bulbous based hairs, long, thin, acuminate, situated here and there above the vascular

bundles (fig. 3B). Epidermis above the vascular bundles (especially the midrib)

retorsely barbed. Shorter two-celled hairs with thin-walled deciduous apices and

occurring between the ripple-walled cells. Motorcells usually three in number occupy-

ing from one-third to one-half the leaf thickness. Abaxial epidermis (fig. 7) resembling

the adaxial except for the absence of the motor cells and more abundant stomata.

120

Vascular bundles'. First order bundles conspicuous, projecting slightly abaxially. Ad-

axial and abaxial stereome strand present; abaxial strand sometimes only consisting of

a few fibres; outer bundle sheath present, consisting of 15-20 parenchyma cells, inter-

rupted by the adaxial stereome strand, inner bundle sheath consisting of 20-22 lignified

cells (mid-vein bundle without outer sheath); phloem surrounded by strongly lignified

cells; lysigenous cavity and xylem surrounded by parenchyma. Second order bundles

occurring infrequently, but constantly submarginally. Third order bundles small;

outer sheath complete, consisting of up to 12 parenchyma cells; inner sheath weakly

developed, not lignified. Midrib prominent, composed of a well-developed abaxial

and a weakly developed adaxial stereome strand together with an abaxial first order

bundle and a superposed adaxial amphivasal bundle of the third order; fundamental

tissue composed for the larger part of colourless parenchymatous cells with a gradual

transition to chlorenchymatous tissue adaxially. Margin rounded, with very little or

no schlerenchyma, set with curved broadbased retrorse hairs 0.3 mm. long in the upper

two-thirds of the leaf (fig. 3A). Chlorenchyma cells more or less regular in transverse

section but with fairly deep invaginations of the cell walls (fig. 4D); in surface section

the chlorenchyma also shows invaginations, thus appearing as rows of separate cells

(elements). Stereome weakly developed; abaxially a few rows of schlerenchyma cells

(fibres) present adjacent to the first and second order bundles; adaxially only a few

fibres present. Third order bundles provided with 1-5 fibres both ad- and abaxially.

Since there is a gradual transition from the lanceolate type of leaf to the prehensile type

(the intermediate types not so frequent), a gradual transition in the anatomical structure

from the one type to the other is also met with. The extreme types, however, show

marked differences. The midrib of the caudate prehensile type (fig. 5) is much larger

and trapezoid in section; except for the superposed amphivasal bundle (third order)

and the median first order bundle, two pairs of amphivasal third order bundles are

situated lateral to the two median bundles. Furthermore the leaf-blade is much narrower

and thicker than the lanceolate type and is set with a greater number of retrorse barbs

along the margins and on the keel. Potamophila prehensilis possesses two types of

leaf-blade which differ both morphologically and also anatomically from each other;

the prehensile leaves predominate on sterile main shoots and aid the plant in its climbing

habit, whereas the lanceolate leaves are most frequently met with on the lateral shoots

arising intravaginally on the main shoot and which eventually terminate in an inflores-

cence. P. prehensilis is therefore heterophyllous, a character worthy of note since,

true heterophylly is of rare occurrence in the Gramineae. The course of the Vascular

bundles in the midrib (fig. 8).

Since superposed vascular bundles are rarely met with in the midrib of the leaves

of the Gramineae (e.g. Bambuseae) and little is therefore known regarding their relative

origin and course, a study in this direction was undertaken. This was done on basis

of a series of handcut sections prepared from a caudate prehensile leaf in the vicinity

just below and above the juncture of blade and sheath. Figure 8 (A-F) depict the origin

and course of the finally superposed bundles of the midrib of the leaf-blade. In (A)

the midnerve and two lateral third order bundles of the sheath are present. At a slightly

higher level (B) the lateral bundles divide, one giving origin to two, the other to three

smaller bundles; the third bundle of the latter group gradually moves inward towards the

midnerve (C) and finally becomes superposed above the midnerve (D). The two remain-

ing small bundles of the two groups also move apart to occupy respective adaxial and

abaxial positions (E); higher up in the leaf adaxial bundles terminate (F).

In the lanceolate type of leaf something similar happens with the exception that

only one of the lateral third order bundles divides into two; one of these branches

gradually moves towards the midnerve finally taking in a superposed position above the

midnerve. The two remaining bundles ultimately take in an abaxial position in the

leaf-blade and hardly are included in the midrib.

Superposed vascular bundles have so far also been met with in the midrib of the

leaf of the remaining South African representatives of the Oryzeae, viz. Leersia and Oryza.

121

5. CYTOLOGICAL INVESTIGATION OF THE SOUTH AFRICAN SPECIES.

Origin of the Matefial.

Material for this study was collected by Dr. H. G. Schweickerdt at Magoebaskloof

in the Northern Transvaal.

Technique.

(cf. “ Technique ” under Pseudobromus p. 142)

Observations.

Potamophila prehensilis 2n=24 (Fig. 9).

P. prehensilis has a complex of 24 somatic chromosomes. Owing to the smallness

of the chromosomes (the length varying from 1 -5-3 p) no morphological study of the

somatic complement was undertaken. No satellites were found as is reported for Oryza

sativa by Nandi (1936) (23). In some metaphase plates an indication of a double struc-

ture was found in the split ends of the chromosomes (fig. 9). The position of the

constrictions could not be determined. Somatic pairing of chromosomes was found

in a few instances. The same phenomenon was reported for the Oryzeae by Ramanujam

(1938) (24).

According to Hubbard in Hutchinson’s Fam. of Flow. Plants (25) the tribe Oryzeae

is composed of two sections, the Oryzinae and Zizaniinae. Hitchcock in his Man. of

Grasses of the U.S. (1935) (7) regards the Zizaniinae as a separate tribe. These facts

are interesting from a cytological point of view.

Ramanujam (24) maintains that the primary basic number of the Oryzeae is five,

the section Zizaniinae having retained the original basic number, while the Oryzinae

developed a secondary basic number, twelve. The cytological investigation of the tribe

therefore seems to support Hitchcock’s classification.

6. REMARKS AND DISCUSSION.

In 1827 the genus Potamophila was created by Robert Brown, based on the species

P. parviflora R. Brown (1). Three years later Kunth in his Rev. Gram. (1830) (2)

created the new genus Maltebrunia on basis of the species M. Leersioides. In 1841 a

second species (South African) was described under this genus by Nees in his FI. Afr.

Austr. (3) namely, M. prehensilis. Bentham (4), however, regarded the two genera as

congeneric and in 1881 included Maltebrunia under Potamophila, the two genera com-

bined forming (according to his views) a natural genus of three species. There remained,

however, certain points of difference between the species. Both Kuntze in his Rev.

Gen. PI. (1893) (5) and Stapf in Dyer, FI. Cap VII (1900) (6) described the uppermost

florets of Potamophila as hermaphrodite or unisexual with the organs of one sex reduced,

so as to include the South African species, which has the fertile floret of the uppermost

spikelets constantly bisexual. A comparison of the characteristics of P. prehensilis

with those of the type species P. parviflora makes it very doubtful whether P. prehensilis

should be included under the genus Potamophila R. Brown. The differences between

the two species are given in the following tabulated form : —

P. parviflora R. Br.

1. Uppermost florets unisexual or bisexual.

2. Lemma and palea membranous, nerves not

raised.

3. Lemma not clasping the palea.

4. Sterile lemmas more or less ovate.

5. A hydrophytic canelike grass 1-1£ m. high,

leaves narrow and erect.

P. prehensilis (Nees) Benth.

Florets all bisexual.

Lemma and palea hardened, nerves raised.

Lemma clasping the palea.

Sterile lemmas subulate.

A hygrophytic climbing grass, up to 10 m.

high, leaves heterophyllous.

122

Taking the above differences into consideration and amplified by further cytological

and anatomical study of the two species in question, it may result in the necessity of

creating a new genus to accommodate the species P. prehensilis (Nees) Benth.

Mention should be made of the totally misleading representation of the habit of

Potamophila prehensilis by Bews (20). Study of the species in its natural habitat has

shown it to be a climber able to reach a height of ± 10 meters, aided in this respect by

the caudate prehensile leaves; the numerous inflorescences are not only borne terminally

on the main shoots but also on relatively short lateral branches.

The interpretation of the spikelet in the Oryzeae has proved difficult mainly on

account of the “ palea ” of the fertile floret not being two-keeled and two-nerved (as

is usually the case in the Gramineae), but being dorsally keeled and three-nerved.

Much difference of opinion, however, still seems to exist in the interpretation of

the lower part of the spikelet as well.

Hitchcock (7) interprets the spikelet of the Oryzeae as follows: the glumes are small

(sometimes much reduced) structures, followed by the three-nerved lemma subtending

the fertile floret. The palea is three-nerved and similar in texture to the lemma. Accord-

ing to this interpretation the spikelet is one-flowered and disarticulates entire, i.e., with

glumes attached.

Most authors, however, interpret the spikelet as being one- to three-flowered

(fig. 10A). This interpretation was originally put forward by Stapf (8) and is sup-

ported by agrostologists of the front rank, i.e., Arber (9) and Hubbard (10A). The

inconspicuous hyaline two-lipped cuplike structure (which may be reduced to a mere

rim), present at the base of the spikelet, is interpreted as representing the reduced glumes ;

the successive structures (regarded as glumes by Hitchcock and others) represent two

sterile lemmas (florets), whereas the third floret is terminal and hermaphrodite. Accord-

ing to this interpretation (fig. 10A) the tip of the rhachis would be situated between

the palea of the third lemma (F) and the sterile lemma (D) of the second floret.

Arber investigated the spikelet anatomically and by study of transverse sections

of the spikelet, arrived at the conclusion that the two scales at the apex of the pedicel

occupy a position corresponding to that of glumes.

Hubbard, in his description of Oryza australiensis Domin, pointed out that the

basal bracts in this species, although small, are not only larger than in most of the other

species of Oryza, but are two quite distinct structures. These represent the much-

reduced glumes. By the discovery of rice spikelets bearing abnormal florets in the

axils of the bracts below and preceding the fertile floret, by Michaud (11), the identity

of these structures (sterile lemmas) has apparently finally become established. Despite

all the arguments forwarded in favour of Stapf’s interpretation of the rice spikelet,

Backer (12) recently reverted to support the older interpretation of Hitchcock.

Chatterjee (13) after giving an outline of the history of the various interpretations

concerning the structure of the rice spikelet, votes in favour of Stapf’s interpretation;

he furthermore proceeds to criticise the arguments forwarded by Backer against Stapf’s

interpretation. Comparing the tip of the pedicle of Oryza (i.e., glumes sensu Stapf),

with the cupular structure occurring at the base of the spikelet in the genus Asthenochloa,

Backer apparently considers these organs homologous. Since the genera Oryza and

Asthenochloa belong to distinct tribes, these structures may equally well be considered

analagous, and need not necessarily be homologous. Chatterjee furthermore points

out that in certain species of Oryza the cupular structure occurring at the apex of the

spikelet is distinctly bi lobed ; in fact these lobes are not attached to the axis on the same

level, but on different levels, viz., one is attached slightly above the other (cf. O. australi-

ensis Domin and O. ridleyi Hook fil.).

In the present investigation it was found that the pedicel tip of Potamophila also is

distinctly bi-lobed, the lobes being attached to the axis on different levels, (cf. fig. IB).

123

Moreover, either of the two sterile lemmas may at times subtend a very delicate

hyaline and much reduced palea (fig. IB). The presence of this palea had already been

observed by Stapf, and was indicated by him in his drawings of a dissection of a spikelet

taken from a Drege gathering, although no mention is made of this in his account of the

species in the FI. Cap. VII. Both drawing and sheet are deposited in the Herb. Benth.

at Kew. The position of the lemma subtending the palea is such that the two lowermost

hyaline bracts must undoubtedly be regarded as representing much reduced glumes

(fig. IB).

Two different hypotheses have been put forward to explain the presence of the

dorsally keeled three-nerved “palea ” of the Oryzeae.

The generally accepted theory is that the fertile floret consists of a lemma subtending

a true palea, which is three-nerved and similar in substance to the lemma. In an effort

to prove this interpretation Rossberg (14) endeavoured to indicate that it is possible

that a true palea may at times be more than two-nerved and dorsally keeled. It must

be admitted that his arguments are quite plausible. Seen in the light of recent dis-

coveries, however, his theory when applied to the Oryzeae, is difficult to maintain.

A totally different interpretation of the upper part of the spikelet was advanced

independently by Peterson (15), Pilger (16) and Parodi (17).

To explain the presence of the six stamens and three-nerved “ palea ” of the rice

spikelet Peterson assumed the spikelet to be two-flowered, the lower flower being

perfect and the upper staminate.^ The structure usually considered to be the “ palea ”

is really the lemma of the staminate floret, while the pales of both are thus wanting.

Parodi’s interpretation is essentially the same as Peterson’s, but the former gives

a more detailed exposition of his views, supported by morphological evidence (18).

Parodi regards the upper part of the spikelet as the result of the fusion of two

florets, of which the pales of both and the pistil of one has disappeared, thus giving rise

to a floret composed of a single ovary and six stamens (cf. fig. 11B). Subsequent

to the publication of this theory a rice plant with abnormal spikelets containing two

grains as well as two bi-nerved pales was actually found. These addorsed pales originat-

ing from the inner basal part of the spikelet, together with the two outer bracts, may thus

be regarded to correspond with the two florets fused to form (together with the glumes

and sterile lemmas) the typical spikelet of the Oryzeae.

The unusual structure of the spikelet and the absence of evident affinity with other

tribes of the Gramineae, induced Parodi to create the independent subfamily the Ory-

zoideae to accommodate the tribe Oryzeae.

It may be mentioned here that a cross-section of a rice spikelet with two florets

was illustrated by Schuster (19) in 1910.

In spite of the observations mentioned above, Chatterjee dismissed Parodi’s theory

as a bold assumption, without commenting on the morphological evidence in favour

of such an interpretation. Moreover he remarks that he has no doubt that the dorsally-

keeled three-nerved structure (occupying the position of a palea) is indeed a true palea,

because in the mature caryopsis the hilum points towards this structure. It might be

pointed out that according to Parodi’s interpretation the hilum would in any case be

turned towards the upper lemma (“ palea ” sensu Chatterjee).

Pilger’s theory (fig. 10B and 11 A) mainly agrees with that of Peterson and Parodi.

Pilger, however, regards the spikelet as the product of the fusion of two florets of which

the pales of both and the pistil stamens and lodicules of one have disappeared. This

implies that Pilger regards the Oryzeae as derived from ancestors with six stamens per

floret, whereas Parodi regards the relatively high number of stamens, as the result of

the combination of two sets of three stamens,* i.e., originating from an ancestral stock

with three stamens per floret.

* One of the characteristics used by Parodi as a basis for his theory, is that the spikelet of the Oryzeae

contains six stamens (not three). From this one can conclude that the six stamens normally found

in this tribe is regarded by Parodi as the result of the fusion of two sets of stamens originating from

different florets.

124

It does seem more logical to accept Pilger’s view of the sequence of reduction in the

floret (Fig. 11 A). According to Pilger the upper floret has disappeared through a

downward reduction, while the consequence of Parodi’s theory is that the reduction has

taken place from above, as well as from below, leaving only the stamens intact (cf.

fig. 11B). If it is assumed that the upper floret is male, one still has to account for

the disappearance of the lodicules, which have not been included in the typical rice

spikelet through fusion of the two florets (2 lodicules are typically present). Apart

from this fact certain characters of the Oryzeae pointing to a phylogenetic relationship

with the Bambuseae, a tribe where 6 stamens per spikelet is of common occurrence,

seem to indicate Pilgers’s interpretation as the most acceptable one.

The Oryzeae on the whole form a well-defined group exhibiting no evident relation-

ship to other tribes. The tribe nevertheless exhibits a number of characters which are

indicative of a possible phylogenetic relationship to the Bambuseae ; these characters

are being put forward in the following tabulated form: —

BAMBUSEAE

1 . Reduction of the florets in the spikelets has taken place in

both an acropetal and basipetal direction.

2. The “ palea ” may be dorsally keeled and similar to the

lemma (Melocanninae).

3. Stamens usually 6 per floret.

4. Styles 2-3.

5. Caryopsis with an elongated linear hilum.

6. Epiblast large and broad.

7. Starch grains compound.

8. Chlorenchyma of the leaf possesses conspicuously invagi-

nated cell walls.

9. Vascular bundles in the mid-rib of the leaf blade are often

superposed.

10. Basic chromosome number x= 12.

ORYZEAE

Likewise.

Styles two, occasionally a third non-

stylar process present.

Likewise.

In the S. African representatives of

the above tribe likewise.

Likewise.

According to Bews (20) Camus furnishes the following information regarding the

spikelet of the Bambuseae : the spikelets are usually several-flowered, but often only one

floret per spikelet is fertile; moreover, the reduction of the floral parts in the spikelet

has taken place both from above as well as below. The structure of the Oryza spikelet

is thus very similar to that met with in the Bambuseae, that is, if Pilgers’s interpretation

is accepted; reduction in the Oryza spikelet has, however, progressed much further.

The keelless “ palea ” and its similarity to the lemma as met with in the Melocan-

ninae suggests a possible relationship to the “ palea ” of the Oryzeae. In Dinochloa this

“ palea ” may be absent and reduction has here progressed to an extreme.

In the discussion of abnormal rice spikelets discovered, Michaud pointed out that

in certain of the spikelets studied, an abnormal flower with an indurate three-nerved

palea was found to be present in the axil of one of the usually reduced sterile lemmas.

Concluding her discussion she remarks that if Parodi’s and Peterson’s hypothesis

is accepted, the presence of a flower in the axil of one of the usually sterile lemmas,

similar in structure to the normal flower found in the Oryzeae, “ is a strong indication

that a branching system is involved,” and that if this is actually the case, it “ would lend

strong support to Parodi’s suggestion that the Oryzeae constitute a separate sub-family.”

It is possible that the somewhat similar development found in certain members

of the Bambuseae where inflorescences are known to develop in the axils of lemmas [as

125

indicated by Pilger (21) in the genus Guadua] may also be regarded as evidence of a

relationship between the Oryzeae and Bambuseae.

Examination of the chlorenchyma of the leaf of all the South African representatives

of the Oryzeae has revealed the presence of conspicuously invaginated cell-walls.

In this respect the Oryzeae strongly resemble various representatives of the Bam-

buseae cf. Haberlandt (22).

Superposed vascular bundles occur in the midrib of the leaf of all South African

representatives of the Oryzeae, a character which is likewise met with in the Bambuseae,

e.g., in the leaves of our native bamboo Arundinaria tesselata.

A more comprehensive anatomical study of the various genera of the two tribes

concerned may possibly furnish additional evidence of a close phylogenetic relationship

between the Bambuseae and the Oryzeae.

7. MATERIAL EXAMINED.

Abbreviations used for the various Herbaria are as follows: —

BOL Bolus Herbarium, Cape Town.

CTM South African Museum, Cape Town.

GRA Albany Museum, Grahamstown.

NH '. . . Natal Herbarium, Durban.

NU Natal University College, Pietermaritzburg.

PRE National Herbarium, Pretoria.

STELL University of Stellenbosch.

UPR University of Pretoria.

The following material was examined : . .

Cape Province. — Willowvale district: Neeyele Forest (also at Monubu), alt. 500',

IV., 1906, Pegler 1338 (BOL, CTM, GRA, PRE). King Williamstown district: Perie

bush, III., 1894, Schonland 856 (GRA). Do., Perie, XI., 1901, Sim 2809 (GRA, PRE).

Lusikisiki district: Igosia Forest, Pondoland, IV., 1918, Mogg 362 (NU). Port St.

Johns district: Port St. Johns River, IV., 1918, Pole Evans H. 18039 (PRE). Do., Port

St. Johns, XII., 1943, Schelpe 367, (NU). Umzimkulu district: Insikini, Nlokolwana

Forest, creeping grass, good fodder grass, IV., 1925, Miller 6045 (PRE).

Natal Province. — Port Shepstone district: Oribi Gorge, IV., 1937, McClean

473 (NH). New Hanover district: Blinkwaterbos, II., 1944, Schweickerdt 1532 (NU,

UPR). . Lions River district: Karkloof Forest, III., 1940, Ryecroft 8 (PRE, NU).

Pietermaritzburg district: Town Bush — Masais Farm, III., 1946, Clarkson 256 (NU).

Do. Town Bush Valley, alt. 3,200', IV., 1946, Fisher 1018 (NU). Do., near Curryspos,

VII., 1945, Acocks 11580 (PRE). Inanda district: Inanda, alt. 1,800'., XII., 1886,

Medley Wood 558 (CTM). Do., Inanda, X., 1881, Medley Wood 1305 (BOL, CTM).

Do., Inanda Heights, III., 1940, Schweickerdt 1432 (NH, PRE). Do., about 10 miles

from Verulam on road to N’dedwe, first record of fruiting material, IV., 1940, Schweick-

erdt 1441 (NH). Do., Verulam, near N’dedwe, IV. 1940, Schweickerdt 1450 (NH,

PRE). Richmond district: in moist forests, climbing 30' or more up trees, IV. 1832,

Galpin 11955 (PRE). Zululand, N’kandla district: Qudeni Forest, alt. 5,000', II., 1940,

Fisher et Schweickerdt 82 (NU, PRE). Do., Qudeni Forest, alt. 5,000', II., 1940,

Fisher et Schweickerdt 117 (NU, PRE). Eshowe district: Eshowe, IV, 1941, McClean

1002 (NH). Do., Eshowe Forest, tolerates a lot of shade, III. 1927, Kotze 248 (PRE).

Between Ubombo Magistr. and Ugaza Mountains — Mkuzi Station, IV., 1944, Gerstner

4502 (PRE).

Transvaal Province. — Petersburg district: The Downs, Junod 4103 (TRV).

Letaba district: De Hoek, Houtbosch, I., 1945, Schweickerdt 1557 (UPR, NU). Do.,

De Hoek, Houtbosch, I., 1945, Schweickerdt 1562 (UPR, NU). Do., De Hoek by

Haenertsburg, VII., 1944, Enslin et Schweickerdt (UPR). Do., near Magoebaskloof

126

Hotel, IV., 1942, Codd and Muller 360 (PRE). Do., Magoebaskloof (Diepkloof),

IV., 1946, de Winter 105 (UPR). Do., Magoebaskloof, Patattabos, IV., 1946, de Winter

107 (UPR). Zoutpansberg district: Louis Trichardt, Hangklip, I., 1945, Schweickerdt

1566 (UPR, NU).

Fragments Ex. Herb. Hort. Bot. Reg. Kew. — Natal: Umpumulo, common in

bush, alt. 2,000' to down to the coast, XI., 1875, Buchanan (PRE). Natal, Dumisa,

Ellesmere, im Unterholz shatt. Walder, VI., 1912, Rudatis 1651 (PRE). Cape Province,

Transkei, Kentani near Manubi, XII., 1911, Saxton 28 (PRE).

Authentic material, fragment of Drege 4352 (PRE) Isotype.

Type Specimen (not seen). — The type specimen Drege s.n. is deposited in the

Botanisches Museum, Berlin-Dahlem and does not bear any locality except that it

comes from Natal. According to Dr. H. G. Schweickerdt this sheet is probably of

the same gathering as Drege 4352, ex Herb. Liibeck, which was seen and identified by

Stapf.

8. ECONOMIC NOTES.

Collector’s notes furnish information to the effect that this grass is heavily browsed

upon by game and thus only reaches the flowering stage in less readily accessible situa-

tions. The species is furthermore reported to be a good fodder grass for stock.

Vernacular names.

Afr. : Klimgras ; Klimopgras. Zulu : uBabe wehlathi.

9. DISTRIBUTION OF THE SOUTH AFRICAN SPECIES (FIG. II).

The species occurs at an altitude from 500 ft. up to 6,000 ft. It is one of our most

characteristic hygrophytic forest grasses and is an important component of the under-

growth of the low-lying riparian forests and at higher altitudes the more temperate

mist belt forests of the eastern escarpment (see map). The species prefers shady,

moist localities. As far as is known at present its northernmost limit appears to be the

southern slopes of the Zoutpansberg.

10. SUMMARY.

1 . A detailed description of the morphology and anatomy of Potamophila prehensilis

(Nees) Benth is given. It is pointed out that true heterophylly occurs in this plant; a

rare occurrence in the Gramineae.

2. The chromosome number of P. prehensilis is 2n=24.

3. A comparison of the characteristics of P. prehensilis with those of the type species

of the genus P. parviflora makes it extremely doubtful whether the former should be

included in the genus Potamophila R. Br. Further study of the two species in question

may result in the necessity of creating a new genus to accommodate the species P.

prehensilis (Nees) Benth.

4. A discussion concerning the interpretation of the rice spikelet in the light of

recent discoveries is given. Pilger’s interpretation is regarded as the most satisfactory

explanation of the anomalies found in the rice spikelet. Pilger interprets the spikelet

as being four-flowered with the two lowermost florets reduced, while the upper part

of the spikelet is the result of the fusion of two florets of which the pales of both and the

lodicules stamens and ovary of the upper have disappeared, thus giving rise to a floret

composed of a single ovary and six stamens enclosed by two lemmas, the lower being

5-nerved and clasping the upper, which is 3-nerved and dorsally keeled.

127

5. The author has endeavoured to enlarge on the alleged phylogenetic relationship

between the Oryzeae and Bambuseae. Several morphological and anatomical simi-

larities hitherto unknown have been recorded.

10. LITERATURE CITED.

1. Brown, R Prodromus Florae Novae Hollandiae, ed. 2, 1, 67 (1827).

2. Kunth Rev. Gram. p. 183, t. 3, En. 1, p. 8 (1830).

3. Nees Florae Africae Austral., 1, p. 194 (1841).

4. Bentham In Journ. Linn. Soc. 19, 55 (1881).

5. Kuntze Rev. Gen. PI. (1893).

6. Stapf, O In Dyer, FI. Cap. VII, 659 (1900).

7. Hitchcock Manual of the Grasses of the U.S. (1935).

8. Stapf, O Ex Hook. f. in Trim. FI. Ceyl., 5, 182 (1900).

9. Arber, A The Gramineae, 184 (1934).

10. Hubbard, C. E Hooker’s leones, PI. 3232 (1934).

11. Michaud, V Bull. Torr. Bot. Club, 71, 624 (1944).

12. Backer, C. A Blumea, Suppl. 3, 45 (1946).

13. Chatterjee, D Nature, 160, 4059, 234 (1947).

14. Rossberg, G Beitrage zur Morphologie des Grasahrchens, 24 (1935).

15. Peterson, N. F Nebraska Acad. Sci., 6 (1935).

16. Pilger, R In Bot. Jahrb., 69, 4 (1939).

17. Parodi, L. R Gramineas Bonariensis, 61 (1939).

18. Parodi, L. R Primera Reuneon Argentina de Agrinomia, 55 (1939).

19. Schuster, J Ueber die Morphologie der Grasbliite. Flora 100 (1910).

20. Bews, J. W World’s Grasses (1929).

21. Pilger, R In Berichte d. Deut. Bot. Gesellsch. XLV. (1927).

22. Haberlandt Physiological Plant Anatomy 278 (1928).

23. Nandi, H. K Journ. Genetics, 33, 315 (1936).

24. Ramanujam, S Annals of Bot. II, V., 107 (1938).

25. Hutchinson . . . . Fam. FI. Plants (1934).

1599—5

128

0-03MM.

0-5 MM.

Fig. 1 . Potamophi/a prehensilis (Nees) Benth.

A, Spikelet. B, Base of the spikelet. A, lower glume; B, upper glume; c, sterile

lemma 1 ; d, sterile lemma 2; e, lemma; F, palea; g, reduced palea; h, pistil; i, lodicules;

J, anther; k, caryopsis, dorsal view; L, caryopsis, ventral view; M, cross-section of the

caryopsis; N, longitudinal section of the embryo; o, compound starch granule. —

(Schweickerdt 1532).

MARY GUNN LIBRARY

NATIONAL BOTANICAL INSTITUTE

PRIVATE BAG X 101

PRETORIA 0001

REPUBLIC OF SOUTH AFRICA

Fig. 2. — A , lanceolate type of leaf. B, caudate prehensile type of leaf.

130

LB

Fig. 3. — A — Broadbased retrorse barb. B — Long thin two-celled hair. C Two-

celled deciduous hair. D — Cross-section of the shoot.

0-5 MM.

131

Fig. 4. Lanceolate type of leaf. A — Diagrammatic cross-section of the leaf-blade.

B — Cross-section of a part of the leaf-blade, including the midrib enlarged. C — Amphi-

vasal superposed bundle enlarged. D — Macerated chlorenchyma tissue.

(Schweickerdt 1532).

I VB

132

midrib. — (Schweickerdt 1532.)

133

134

ST LC 0-1 MM. pp RB

Fig. 7. — Abaxial epidermis of the leaf-blade. — (Schweickerdt 1532.)

135

Fig. 8.— Diagrammatic drawing of the course of the vascular bundles in the midrib

of a caudate prehensile type of leaf. — (Schweickerdt 1532.)

O-OI MM.

*

Fig. 9. — Polar view of a metaphase in a root-tip cell.

Fig. 10. — Diagrammatic drawing to illustrate the different interpretations of the

spikelet of the Oryzeae.

A

Fig. 1 1 . — Diagrammatic drawing to illustrate the difference between the interpre-

tations of the rice spikelet. A — According to Pilger. B — According to Parodi.

A — lower glume, b — upper glume, c — first sterile lemma. D — second sterile lemma.

e — lower lemma of fertile floret. F — upper lemma of fertile floret. G — pales, h —

lower floret. I — upper floret. (The dotted lines indicate reduced structures.)

137

Fig. 12. — Distribution of Potamophila prehensilis in South Africa.

139

The Taxonomy, Anatomy and Cytology of Pseudo-

bromus africanus (Hack.) Stapf .

By B. de Winter

CONTENTS.

1. Introduction 139

2. Description of the genus 139

3. Description of the South African Species 140

4. Anatomical Features of the South African Species 141

5. Cytology of the South African Species 142

6. Remarks 142

7. Material Examined 143

8. Distribution of the Species 144

9. Literature Cited 144

1. INTRODUCTION.

Pseudobromus africanus (Hack.) Stapf, the only indigenous species of the genus in

South Africa, has until recently remained a most elusive and rare plant. For many

years it was known only in herbaria from the type gathering made by Dr. A. Rehmann

in the Northern Transvaal, sometime between 1875-1880. Within recent years, how-

ever, the species has been recorded from a number of additional localities in the south-

eastern parts of the Union.

Stapf (5) in 1900 considered the genus a close relative of the genus Festuca L. and

consequently a representative of the Festuceae; later (6) he placed the genus near Bromus

L. Phillips (7) during 1926 accommodated the genus without comment in the Agros-

tideae, whereas Camus (1) in 1930 supported Stapf’s view both with regard to tribal and

generic affinity.

Hence the present investigation was undertaken to throw further light on both its

tribal and generic affinities.

In the course of this study a wide series of living and dried material was examined;

the living plants were studied both in situ and in cultivation.

The present study is co-ordinated with others on South African Gramineae being

undertaken by Dr. H. G. Schweickerdt (University of Pretoria) and Dr. B. S. Fisher

(Natal University College).

I wish to express my thanks especially to Dr. H. G. Schweikerdt for his guidance

and encouragement in my studies (the present paper comprising part of the thesis

presented in partial fulfilment for the degree of Master of Science at the University of

Pretoria), to Mr. C. E. Hubbard of Kew for his valuable advice, and furthermore to the

Directors and Curators of the various Herbaria who were kind enough to allow me to

examine the relevant material.

2. DESCRIPTION OF THE GENUS.

PSEUDOBROMUS K. Schum.

Pseudobromus K. Schum. in Engler, Pflanzenwelt Ost. Afr., C, p. 108 (1895); Stapf in

Dyer, FI. Cap. VII., 763 (1900); Stent in Bothalia, I, 3, 278 (1924); Bews, World’s

Grasses, 201 (1929); Camus in Bull. Soc. Bot. France, 77 p. 511 (1930); Phillips,

South African Grasses, 174, pi. 55 (1931).

140

Spikelets lanceolate, slightly compressed, peduncled, arranged in a panicle; rhachilla

disarticulating above the glumes, produced and terminated by a reduced awned lemma.

Florets 1 or 2, hermaphrodite, often with a superior (additional) rudimentary floret,

the apex of which may project beyond the glumes. Glumes persistent, subequal, lanceo-

late to oblong, 1-3 nerved. Lemma narrowly lanceolate in profile, sub-herbaceous,

rounded on the back, attenuated at the apex into a long straight awn; 3-5 nerved;

margins incurved; callus very short, obtuse, glabrous. Palea subherbaceous, very

narrow, two-keeled, entire or nearly so. Lodicules 2, hyaline, asymmetrical, unequally

two-lobed. Stamens 3. Ovary oblong, pubescent at the apex; styles 2, usually fairly

short, distinct, ending in slender plumose stigmas; stigmas laterally exserted. Caryopsis

oblong, longitudinally channelled on the ventral face; not adhering to the lemma and

the palea; hairy at the apex. Plants robust. Leaves flat, usually linear or narrow,

herbaceous. Ligule membranous. Panicle large, lax and open, often nodding.

Species five: three in Madagascar, one in South Africa and one in tropical East

Africa.

3. DESCRIPTION OF THE SOUTH AFRICAN SPECIES.

PSEUDOBROMUS AFRICANUS (Hack.) Stapf. (Fig. 1).

P. africanus (Hack.) Stapf in Dyer, FI. Cap. VII, 763 (1900); Stent in Bothalia, 1, 3,

278 (1924); Bews, World’s Grasses, Fig. 25, 201 (1929); Phillips, S. Afr. Grasses,

174, pi. 55 (1931).

Synonymy. — Brachyelytrum africanum Hackel in Bull. Herb. Boiss. Ill, 382 (1896).

Perennial, caespitose, branched from the base, rhizomes short and knotty. Culms

few to many, erect or sharply ascending, up to four feet high, slender, simple. Inter-

nodes up to eight in number, glabrous, shiny, sometimes weakly striate, pallid to greenish.

Nodes constricted, dark green to brown, striate. Sheaths tight, slightly longer than the

internodes, open, strongly striate, nerves whitish, scabrous; margins thin, chartaceous.

Ligule up to 8 mm. long, membranous, acuminate, glabrous; auricles glabrous; collar

inconspicuous, smooth. Leaf-blade dark green, spreading or erect, sometimes more

than 40 cm. long and 1 - 5 cm. wide, tapering into a setaceous point, narrowly lanceolate,

flat on both surfaces, thick, more or less rigid, narrowed at the base; nerves on the

upper surface scaberulous, less so abaxially; midrib prominent, four or five secondary

nerves on each side of the mid-nerve; margins scabrid. Inflorescence a large, fine,

lax, terminal, slightly nodding panicle, up to 40 cm. long; rhachis sometimes flattened,

smooth below, scaberulous upwards, branches fascicled, terete, flexuous, spreading,

fine, up to 20 cm. long, branched. Spikelets 7-9 mm. long, narrowly ovate in outline,

awned, dark green, hairy at the base of the glumes, slightly laterally compressed or

terete, disarticulating above the glumes. Floret 1, hermaphrodite; rhachilla scabrid,

produced to a little more than half the length of the spikelet and bearing a terminal

rudimentary lemma; the reduced empty lemma more or less developed, sometimes

long-awned and the spikelet thus apparently two-flowered. Glumes up to 6 mm. long,

herbaceous membranous, unequal, reaching to about one-half the length of the spikelet;

the lower up to 5 mm. long, lanceolate, glabrous, one-nerved, margins hyaline; the

upper broadly lanceolate, up to 6 mm. long, finely acuminate, margins hyaline three-

nerved, lateral nerves evanescent. Body of the lemma about 8 mm. long, a little shorter

than the palea, narrowly lanceolate, margins hyaline near the base; midnerve produced

into a long capillary awn; awn scabrid, about 20 mm. long. Palea narrowly lanceolate,

up to 9 mm. long, herbaceous, strongly two-nerved, two-keeled with a narrow dorsal

groove in which the produced rhachilla lies appressed. Callus rounded, hairy, gibbous

in lateral view. Lodicules two, oblong, denticulate, swollen at the base, apex hyaline.

Stamens three, more or less 6 mm. long, linear, pale yellow in colour. Ovary globose,

hairy at the apex; styles two, very short; stigmas plumose. Caryopsis tightly enclosed

by the lemma and the palea but not adhering, a little shorter than the palea, about

141

6 mm. long, bearing an apical tuft of hairs, reniform in cross-section; embryo small,

about 1/6 the length of the grain; hilum linear stretching the whole length of the grain;

epiblast present but small; starch grains compound, ovate- globose, consisting of

many polyhedral granules. Reaction of KI2 with starch normal, a deep blue colour.

4. ANATOMICAL FEATURES OF THE SOUTH AFRICAN SPECIES.

The Shoot (Fig. 2).

Shoot slender, long, circular in section, covered at the base with imbricate brown

scales. Leaves arranged on a very short stem, young leaves protected by the sheaths

of the outer leaves.

The anatomy of the stem conforms to the usual type. The stem is distinctly hollow.

Two or three rows of vascular bundles lie near the periphery of the stem. The pith

cavity is lined by thin-walled parenchymatous cells.

The root is typical of the monocotyledonous type. The endodermis cells are very

strongly lignified, especially the inner tangential walls. Outer tangential walls only

slightly thickened. The thickened walls show distinct striations and pits. An interest-

ing point is that the pericycle cells are in some places thickened, and closely resemble

endodermis cells.

The Leaf-blade (Fig. 4). Leaf Anatomy.

Hand cut sections made from the upper part of the lower third of a full-grown leaf

were studied. Surface sections were used to study the epidermis. The study of

macerated material of the blade proved useful in elucidating certain doubtful points.

The drawings were made either with a Leitz Panphot or an Abbe drawing apparatus.

Abbreviations used in the text-figures are as follows:

Leaf-blade tapering towards the margins, near the midrib about 0-39 mm. thick,

near the margins about 0-18 mm. thick. Abaxial surface more or less flat; the adaxial

finely grooved (vascular bundles slightly projecting). Keel prominent, rounded, with

sides sloping. Epidermis (Figs. 3 and 4) : Motor cells 3-5, poorly differentiated near the

margins, otherwise large occupying 1/3-J the leaf thickness. No silicified cells present,

cells very similar, long and narrow, tapering towards the ends. Stomata occurring on

both surfaces; abaxially in 1-2 rows between the bundles, absent between some of the

third order bundles ; adaxially (Fig. 4) in 3-4 rows on both sides of the first and second

order bundles and in 2-3 rows on both sides of the third order bundles. A single row

of short one-celled curved hairs occur adjacent to the bundles ab- and adaxially. Longer

one-celled straight epidermal hairs are found in the region of the stomata. Vascular

bundles irregularly distributed. First order bundles 6-10 excluding the midvein; inner

bundle sheath present and consisting of 16-22 lignified cells; outer bundle sheath con-

sisting of a smaller number of fairly large thin-walled parenchymatous cells which

contain some chloroplasts. Second order bundles up to 12 in number; inner bundle

sheath consisting of 12-14 lignified cells; outer sheath of thin-walled parenchymatous

cells. Third order bundles 14-17 in number; inner sheath consisting of about 10

142

lignified cells; outer of smaller number of parenchymatous cells. Midrib large,

composed of a large first order bundle with adaxial and abaxial stereome strands,

furthermore two second order bundles with their stereome strands; colourless, isodia-

metric parenchyma cells present as well as a few chlorenchyma cells. Margin tapering

to a fairly narrow edge covered with small curved hairs. Chlorenchyma cells large,

regular to irregular in places. Smaller round cells present between the chlorenchyma.

The chlorenchyma cells in longitudinal view possess deeply undulating walls (Fig. 5C).

If the section passes through the constricted part of a cell a small round cell is seen

in the transverse section. Stereome weakly developed; present in the keel and above

the mid-vein as well as abaxial to all bundles, but adaxial to first order bundles and

usually all the third order bundles.

5. CYTOLOGY OF THE SOUTH AFRICAN SPECIES.

Origin of the Material.

The plants studied were gathered at Magoebaskloof, De Hoek. The type locality

is given as Magoebaskloof Houtbosch.

Technique.

Only root tips were used. These were collected at about 10.30 a.m. and fixed in

La Cours 2 B.P. fixative for 24 hours, dehydrated with xylol and embedded in wax

with a melting point of 54° C. Sections were cut at 10-12 jx and stained by the iodine-

gentian violet technique described by La Cour (9).

The chromosome number was determined by making camera lucida drawings of

several equatorial plates. Use was made of a stage micrometer to indicate the scale

of the drawing.

Observations.

Pseudobromus africanus, 2n=28 (Fig. 6).

Basic number (probably) =7.

The chromosomes are of the relatively big festucoid type. The chromosome num-

ber corresponds with that of Festuca and Bromus, where many species with 2n=28

occur.

The length of the chromosomes vary from more or less 4 /i to more or less 10 /x.

In the metaphase plates studied the bigger chromosomes lie on the periphery of the

spindle, while the shorter ones occupy a more central position ( Fig. 6).

6. REMARKS.

The genus Pseudobromus was created by K. Schumann (4) during 1895 on a plant

which he described as Pseudobromus silvaticus.

On the basis of the one-flowered spikelet Schumann considered the genus to be

related to Deyeuxia Beauv. of the Agrostideae, simultaneously pointing out certain

differences. He furthermore mentions the superficial resemblance in facies of P. silvati-

cus and Bromus asper Murray, thereby indicating the source of the generic name

Pseudobromus.

In the same year Hackel (3) described a species Brachyelytrum africanum which,

however, differed from the genus, sensu stricto, in possessing spikelets with relatively

large glumes and a lemma with a rounded, not oblique, callus. During 1900 Stapf (5)

for sound reasons, transferred Hackel’s species to the genus Pseudobromus , thus effecting

the combination P. africanus (Hack.) Stapf.

Stapf (5) in 1900 evidently did not agree with Schumann concerning the generic

relationship of Pseudobromus, since he states that the latter was closely allied to that of

Festuca L. At a later date, however, he (6) places the genus Pseudobromus next to the

143

genus Bromus in his key to the genera of tropical African grasses. It must be assumed,

however, that the position occupied by the genus Pseudobromus in the key referred to

is one of convenience rather than one of natural affinity.

Phillips (7) places the genus under the Agrostideae without comment, in all proba-

bility merely following the arrangement of genera in De Dalle Torre and Harm’s “ Genera

Siphonogamarum .”

A. Camus (1), in her monographic account of the genus, supports Stapf’s view (6)

and furnishes a number of differential characters by which the genera Bromus and

Pseudobromus may be readily distinguished. Her account includes the study of both

African and Madagassian material comprising five species, resulting in an augmented

and amplified conception of the genus, particularly with regard to the number of fertile

florets per spikelet and the nature of the caryopsis.

Despite the above valuable contributions the difficulty still remains, viz., the close

resemblance in facies and structure of members of Pseudobromus K. Schum. with

certain species of Festuca L. The only tangible characters by which these genera may

be distinguished in the case of Pseudobromus, appear to be the one- to two-flowered

nature of the spikelet with an additional rudimentary lemma inserted on the produced

rhachilla and the fewer nerved lemmas. This genus represents an extremely advanced

type, which by the reduction of the upper florets could possibly have originated from a

more primitive type, e.g., Festuca L.

Although the genus Pseudobromus resembles in facies certain species of Bromus,

a number of fundamental differences are exhibited by these two genera, viz., Bromus

possesses laterally compressed spikelets, which are several- to many-flowered ; the keels

of the palea are rigidly ciliate or ciliolate; the ovary is produced into a terminal append-

age on which the styles are laterally inserted ; the caryopsis usually adheres to the lemma

and the palea; an epiblast is lacking in the embryo (2) and the starch grains are simple.

All these characters are, however, not met with in the genus Pseudobromus, and for

these reasons it is evident that the nearest ally of the genus in question appears to be

that of Festuca L.

The anatomical study of the root, stem and leaf did not reveal striking characteristics.

The endodermis of the root here and there appears to be two-layered, probably due to

the pericycle cells with walls thickened in a way similar to those of the endodermis cells.

The chlorenchyma cells of the leaf in macerated condition reveal the presence of

strongly undulating walls suggestive of rather conspicuous intercellular spaces; the

latter are usually overlooked in transverse sections of the leaf.

7. MATERIAL EXAMINED.

All material cited was actually seen. The following abbreviations indicate the

herbaria where the specimens are deposited: —

BOL Bolus Herbarium, Cape Town.

CTM South African Museum, Cape Town.

GRA Albany Museum, Grahamstown.

NH Natal Herbarium, Durban.

NU Natal University College, Pietermaritzburg.

PRE National Herbarium, Pretoria.

STELL University of Stellenbosch.

UPR University of Pretoria.

Cape Province. — Tembuland Forests, V. 1919, R. Baur 491 (BOL, CTM). East

Griqualand, Imsikini near Hoha, III., 1928, van Tonder 5 (PRE).

Natal Province. — New Hanover district : Blinkwaterbos, kom in diep skadu voor,

II., 1944, Schweickerdt 1534 (STELL, UPR). Lions River district: Karkloof Forest,

144

alt. 5,000', III., 1940, Ryecroft 45 (NH, PRE). Do., Karkloof Forest, common, V.,

1940, B. S. Fisher 138 (NH). Do., Karkloof Forest, II., 1940, Ryecroft 3 (NH, NU).

Zululand, N’kandla district: II., 1940, frequent in forest, alt. 5,000', Qudeni, Fisher et

Schweickerdt 115, (NH, NU, PRE). Do., Qudeni, in shade in forest, alt. 5,000', II.,

1940, Fisher et Schweickerdt 83 (NH, NU). Do., Qudeni, frequent in forests, alt.

5,000', II., 1940, Fisher et Schweickerdt 16 (NH, NU). N’Kalhlondla Forest, Cathedral

Park Area, alt. 5,000', II., 1943, Schelpe 203 (NU).

Transvaal Province. — Pietersburg district: Haenertsburg, Houtbos, De Hoek,

meerjarig, volop in bosse, Schweickerdt 1568 (UPR). Do., Haenertsburg, Houtbos,

De Hoek, in die bosse, I., 1945, Schweickerdt 1563 (UPR). Letaba district : Houtbosch,

1875-1880, Rehmann 5732 (GRA) (type number); Do., Magoebaskloof, klam oop

plekke in bos, IV., 1946, de Winter 104 (UPR). Do., Houtbosch, III., 1946, Schweic-

kerdt 1601 (UPR). Do., Maripskop, IV., 1946, boonste bosse, nie volop nie, Schweic-

kerdt 1616 (UPR). De Winter 129 and 271 (cultivated at University of Pretoria, cyto-

logically investigated, PRE, UPR.)

Type Specimen (not examined). — Rehmann 5732, Houtbosch Letaba district, in

Herb. Mus. Bot. Zurich.

8. DISTRIBUTION OF THE SOUTH AFRICAN SPECIES.

The distribution of this species is limited to the higher parts of the mistbelt forests

which occur along the south-eastern and eastern escarpment. According to data

obtained from herbarium sheets it does not occur below an altitude of about 4,000'

above sea level (Fig. 7).

The distribution of the genus is apparently also limited to high altitudes of Africa

and Madagascar. No gatherings have been reported under 4,000' for any of the species.

P. si/vaticus K. Schum. apparently reaches the high altitude of 8,000' in the East African

tropics.

9. LITERATURE CITED.

1. Camus, A In Bull. Soc. Bot. France 77, 512 (1930).

2. Kennedy, P. B The structure of the Caryopsis of Grasses with reference to their Mophology

and Classification in Bull. 19, U.S. Dept, of Agric. (1889).

3. Hackel, E In Bull. Herb. Boiss., 3, 382 (1895).

4. Schumann, K In Engler’s Pflanzenwelt Ost. Afr. C., 180 (1895).

5. Stapf In Dyer, FI. Cap. VII, 763 (1900).

6. Stapf In Prain, FI. Trop. Afr., IX, I, 24 (1917).

7. Phillips, E. P Gen. S. Afr. Flow. Plants (1926).

/

145

Fig. 1. A — Spikelet. B — Fertile floret, dorsal view. C — Base of floret, showing

produced rhachilla and rounded callus. D — Side-view of the floret, showing the gib-

bous base and short rhachilla. E — Sterile reduced lemma inserted at the top of the

produced rhachilla. F — Lower glume. G — Upper glume. H — Lemma, dorsal view.

/ — Palea with produced rhachilla. J — Callus. K — Lodicules. L — Anthers. M —

Pistil. N — Caryopsis, dorsal view. O — Caryopsis, ventral view. P — Cross-section

of the caryopsis. Q — Longitudinal section of the embryo. R — Epiblast. S — Com-

pound starch grain. — (Schweickerdt 1534.)

146

I MM.

Fig. 2. — Cross-section of the shoot. — (Schweickerdt 1534.)

147

Fig. 3. — Abaxial epidermis. — (Schweickerdt 1534.)

148

3 V B

149

Fig. 5. A — Diagrammatic drawing of the leaf. B — Drawing of a part of the leaf

including the midrib. C — Macerated chlorenchyma tissue.

150

Fig. 6. — Polar view of a metaphase of a root tip cell. — (de Winter 129 and 271.)

151

Fig. 7.— Distribution of P. africcmus (Hack.) Stapf in South Africa.

153

South African Species of Acacia with Glandular

Glutinous Pods.

By

I. C. Verdoorn.

During a recent survey of the Kruger National Park a species of Acacia was observed

which is a common constituent of the vegetation in parts around Skukuza and Malelane.

The endeavour to identify this Acacia led to a study of the group of species with glutinous

or glandular pods, for its affinities were obviously with that group. It was found

that it did not exactly match any of the existing species, and is therefore being described

here for the first time and named A. exuvia/is because of the peeling skin on the stem

and branches.

At the time of this investigation the group as found in South Africa was comprised

of four species and one variety. It became obvious, from this study, that as Burtt

Davy suspected, his variety (A. permixta var. glabra ) is as distinct as any of the species

and is therefore here raised to specific rank and named Acacia tenuispina, because of

its generally known common name “ Fyndoring ”, meaning “ slender spines.”

It was also found that A. rogersii Burtt Davy from the northern Transvaal, north

of the Zoutpansberg range, could not be upheld as distinct from A. nebrownii Burtt

Davy, which was the new name given by Burtt Davy in the Kew Bulletin, 1921, for the

S.W. African species A. glandulifera Schinz non Wats. Burtt Davy had only one

specimen from Messina when he described A. rogersii and since then it has been collected

repeatedly along the Limpopo River and in low-lying sites in dry bush country in the

Messina and Dongola areas.

It is characterised in this group of species by the involucel being at the base of the

peduncle and the leaves usually having only one pair of pinnae. Dr. L. E. Codd,

Botanical Survey Officer, who knows the plant at Dongola, came across what seemed

to him to be the same species when recently he travelled in S.W. Africa around Gobabis

and southwards. On examination his specimens were also found to have the involucel

at the base of the peduncle and the leaves with one pair of pinnae. In general his

specimens compared very well with the Transvaal specimens. They also compared

well with other collections from S.W. Africa, including three, Pearson 9256, 9814 and

8945, quoted by Burtt Davy under A. nebrownii in the Kew Bulletin, 1921.

It still remained to compare it with the type of the species. Acacia nebrownii

being a new name for A. glandulifera Shinz (which name is antidated by A. glandulifera

Wats.), the type must be one of the two S.W. African specimens on which Schinz based

his species, Fleck 484a (flowering twig) from great Namaqualand, or Fleck 480a

(fruiting twig) from Tsoaxaub, Hereroland. These specimens are in the Herbarium at

Zurich. A member of our staff, Miss P. Kies, who had the opportunity of visiting

Zurich, kindly examined these specimens and reported that they both had the dis-

tinguishing characters of the involucel and pinnae common to the other material.

The general description of the Zurich material also agrees well with the S.W.

African and Dongola specimens already examined. It therefore appears that Acacia

rogersii Burtt Davy cannot be upheld and must be sunk under Acacia nebrownii Burtt

Davy.

Further study of the group confirmed Burtt Davy’s decision to separate the rest

of the South African material from A. nebrownii. He had at first (Kew Bull. 1921)

cited specimens from Potgietersrust, Leydsdorp and Swaziland under this species, but

in the Kew Bulletin 1922, he described A. permixta and its variety to include those

154

from the first locality and A. swazica those from the last mentioned. He did not account

for the specimen from Leydsdorp which is unknown to us, but from the locality it is

very unlikely that it is A. nebrownii.

The following key and the short account of each species given below summarises

the present concept of the group in South Africa.

Mature pods torulose much curved, glutinous but glands not conspicuous:

Leaflets glandular-punctate with distinct glands usually on lower surface and on margins (giving

margins a crenate appearance), average leaflet under 3 mm. long, rarely longer

l.A. borleae.

Leaflets not glandular-puntate with distinct glands, on the average larger than in above species,

5 mm. long or longer; skin peeling to a marked degree on stem and branches

.2. A. exuvialis.

Mature pods flat only slightly curved, usually with numerous conspicuous, raised glands:

Branchlets and leaf-rachis pubescent with long hairs, branchlets densely so 3. A. permixta.

Branchlets and leaf-rachis glabrous or nearly so :

Involucel at or near base of peduncle; shrubs up to 8 ft., forming thickets; pods 8-10 mm.

broad, leaves usually 1-jugate, sometimes 2-jugate ..4. A. nebrownii.

Involucel midway on peduncle or higher:

Stoloniferous shrubs usually 2-4 ft. tall, sometimes 6 ft. tall, forming colonies in turfy

ground; pods linear, 5-7 mm. broad; leaves 3-6-jugate; leaflets 6-8 pairs, about 3-4 mm.

long, 1-5 mm. broad, veins not conspicuous 5. A. tenuispina.

Small trees 6-8 ft. tall, pods oblong, usually over 7 mm. broad (more or less 1 cm. broad);

leaves 1-2-jugate, rarely more; leaflets 3-6 pairs (or more), usually over 5 mm. and

up to 10 mm. long, veins usually conspicuous 6. A. swazica.

1. A. borleae Burtt Davy in Kew Bull. 1922 ( =A . barbertonensis Schweickerdt in Kew

Bull. 1937).

Fig. 1.

Slender, gregarious trees or shrubs from 3-8 ft. tall, branching from near the base

with slender ascending branches; obviously glutinous in parts, especially the long (up

to 7 cm. by 3 mm.), much curved torulose pods. The leaves have up to six, rarely more,

pairs of pinnae which are short, more or less 2 • 5 cm. long and the leaflets very small,

about 2 mm. long. On the lower surface and margins of the leaflets are comparatively

large glands, those on the margins giving a crenate appearance to the leaflet. As in

other species of the group the straight, slender, white spines are a feature and the flowers

are borne in small yellow, globose heads on axillary, slender peduncles more or less

as long as the leaves. In this species the involucel is in the upper half of the peduncle.

A. borleae occurs in low-lying sites in dry bushveld country from northern Zululand

to Portuguese East Africa and Central Swaziland, with two widely spaced records from

the Kruger National Park, one in the south about 14 miles N.E. of the Pretorius Kop

Camp, in mixed bush on greyish sandy flats, and the other near Punda Maria in the

extreme north of the Park.

2. A. exuvialis sp. nov. ab A. borleae foliolis marjoribus sine glandulis distinctis

et ab A. swazicae leguminibus subtorulosis valde curvatis differt. Arbor parva, non-

nunquam suffruticosa, caulibus ramisque gracilibus glutinosis cuticulis exutis. Spinae

stipulares valde evolutae, strictae, albidae. Folia bipinnata usque ad 7 cm. longa;

pinnae 1-6 jugatae, usque ad 3 cm. longae; foliola 3-6-juga, 3-5-7 mm. longa 2-3 mm.

lata, mucronata, nerviis inconspicuis. Inflorescentia axillaris, capitata; pedunculi

usque ad 3 cm. longi, supra medium involucrati. Flores lutei. Calyx 1 -75 mm. longus,

extus glaber, 5 lobatus; lobis dorso glandulosis. Corollae tubus 2-5 mm. longus,

5 lobatus ; lobi 0 • 75 mm. longi, subacuti, recurvati. Stamina lutea, e tubo 2 mm. exserta.

Legumen stipitatum, maturum valde curvatum, leviter torulosum, subplanum, usque

ad 5 cm. longum, distincte venosum.

Fig. 2.

155

Transvaal. — Nelspruit district: 16m. W. of Skukuza, Codd and Verdoorn 5464

(fruiting specimen) Type !; 4 m. W. of Skukuza Codd and Verdoorn 5467; 9\ miles S.

of Skukuza, Codd and Verdoorn 5502; north banks Sabi River at Skukuza, Letty 30;

near Malelane Camp, Codd 4359. Pilgrims Rest district: 1\ m. S. Olifants Bridge,

Codd 1666. Zoutpansberg district: 3 m. N.W. of Shingwedzi Camp, Codd and de

Winter 5569; 3\ m. N.W. of Shingwedzi Codd and de Winter 5568.

Characteristically this is a slender tree up to 15 ft. tall, the slender ascending branches

becoming more and more branched towards the top, broomlike, and frequently seem

too heavy for the slender stem, the whole tree eventually falling over. It also occurs

in a shrubby form up to 8 ft. tall. The stems and branches are shiny glutinous in parts

and the thin skin peels off in strips or flakes which roll backwards.

The stipular spines are long, straight and ivory white, the leaves 1-6 pinnate and

the leaflets usually in 3 to 6 pairs. The leaflets are from 3-5-7 mm. long and 2-3 mm.

broad, varying in shape from ovate oblong to obovate oblong, midrib oblique, apex

obtuse or subobtuse, mucronate, the surfaces are microscopically grandular with

minute immersed glands and rarely a few scattered stalked glands. The rachis is

slender, flattened and slightly sulcate above and bears a prominent gland between the

pinnae, it is terminated by a 1-4 mm. long very acute bract. The flower heads are

yellow, axillary on slender peduncles up to 3 cm. long, bearing a pair of perfoliate

involucral bracts about midway or higher up. In this character it is distinguished from

A. nebrownii, which has the involucral bracts at or near the base of the peduncle.

The mature pods distinguish A. exuvialis from A. nebronwii, A. swazica and A.

tenuispina in that they are somewhat torulose and strongly curved, without obvious

glands on the surface while the others are flatter, only slightly curved with numerous

raised glands quite obvious on the flat surfaces.

In respect of pods the species described here resembles more nearly A. borleae

which also has torulose strongly curved pods, but they are longer than in our species,

up to 7 cm. long. The smaller leaflets with large glands, especially dorsally and on

the margins, giving a crenate effect, also help to distinguish A. borleae.

A. exuvialis is seen quite frequently in the Kruger National Park. In the southern

part of the Park it occurs plentifully around Skukuza and Malelane in deciduous low-

veld woodland on gritty, granitic slopes. Here it is characteristically a slender tree

but in the northern part of the Park a shrubby form is found along the road from the

Olifants to the Shingwedzi Rivers, growing in open Mopane veld. Here the plants

grow in many-stemmed groups from 6-8 feet tall but, except for having more than

one stem, they agree in all other respects with the typical plant described above.

3. A. permixta Burtt Davy in Kew Bull. 1922.

Fig. 3.

Small trees with a few weakly ascending branches which give it the popular name

of “ slap doring.” It is distinguished in the group by having branches thickly covered

with a long, whitish hairy pubescence. Like the related species it is generously provided

with straight, rather long white spines and the leaves are short with only 2-4 pairs of

short (1-5-2 cm. long) pinnae and up to 8 pairs of leaflets, which are about 3-5 mm.

long with a distinct rnucro. The petiole and rhachides are thinly hairy, and in this

respect, and the leaflets often being ciliate, it also differs from the other species in this

group. The inflorescence is globose, yellow, on axillary peduncles more or less as long

as the leaves and with the involucel midway or higher on the peduncle. The pods

are very flat and slightly curved, up to 4 cm. long and 1 -4 cm. broad with sparsely

scattered raised glands and prominent veins on the flat surfaces.

This species is found usually on sandy or gritty slopes derived from granite forma-

tions, chiefly along the edges of the Pietersburg plateau, where open grassland gives

way to scattered thorn scrub. Here it is showing signs of spreading on overgrazed

veld and is looked upon as a potential weed. Records of the species are noted from

the Brits district and Sibasa area of the Zoutpansberg, both on weathered granite.

156

4. A. nebrownii Burtt Davy in Kew Bull. 1921 (—A. glandulifera Schinz non Wats, and

A. rogersii Burtt Davy).

Fig. 4.

The slender free branching habit of this species and its gregarious nature result

in the production of dense thickets up to 8 ft. high. It is distinguished in the group

by the involucel being at or near the base of the peduncle. For the rest, except that the

peduncles are usually shorter than in the other species and the leaves rarely with more

than one pair of pinnae, it is typical of this group. It was indeed the first of these species

to be described.

In the Transvaal A. nebrownii has been found only north of the Zoutpansberg,

records being mainly from the Salt Pan and along the Limpopo River in the Messina-

Dongola area. In S.W. A. it is quite widely spread and has been recorded from Bechuana-

land. It grows characteristically in low-lying sites in dry bush country usually where

silty soil has accumulated, as at watercourses.

5. Acacia tenuispina sp. nov. stat. nov. =A. permixta var. glabra Burtt Davy in Kew

Bull. 1922, p. 330.

Fig. 5.

Transvaal. — Waterberg district: Naboomspruit, Ga/pin 475m. (type); 541 m.

Potgietersrust district: farm Keron, Galpin 11597; nr. Villa Nora Acocks 8819; farm

Somerset Estate, Pole Evans 3523. Pretoria district: Pretoria North, Smith 6170.

When describing the above variety Burtt Davy wrote: “ This may prove to be a

distinct species when more complete material is available.” Our present knowledge

of the glandular-pod species of Acacia shows that this plant is as distinct from A. permixta

as are any of the other species in the group. The specific epithet was suggested by the

generally used common name “ Fyn doring ” describing the slender thorns with which

the plants bristle. The species is characterised by its stoloniferous habit forming

large clumps or colonies, usually in turfy ground. It is usually low growing, 2-3 ft.

high, but may grow taller, up to 6 ft., according to some records. Besides the differ-

ences of the glabrous branches and narrower pods, this growth habit clearly distinguishes

it from A. permixta which occurs as a small tree about 6 ft. tall. In some respects

it is nearer A. nebrownii and A. swazica but can be distinguished from both in the smaller

leaflets and narrower pods. From A. nebrownii it also differs in having the involucel

about midway on the peduncle and not at or near the base, as in the former. A. exu-

vialis is easily distinguished from it by the habit, being a slender tree, and by the torulose

much curved pods, those of A. tenuispina being linear and only slightly curved.

6. A. swazica Burtt Davy in Kew Bull. 1922.

Fig. 6.

Small trees, rather stout as compared with the others in this group, with short

ascending branches. It is distributed mainly in the Barberton and Nelspruit districts,

where it grows in dry, bush-covered slopes and hillsides, usually in rocky situations.

Specimens from Swaziland are easily distinguished by the leaflets being larger than

in the other species, up to 1 cm. long, and with conspicuous veins. From further north,

however, where the distribution overlaps with that of A. exuvialis as it does near Male-

lane, the leaflets are not quite so large and the veins often not conspicuous. It is,

however, distinguished from A. exuvialis in the pods being flat, up to 1 cm. broad,

somewhat curved and with scattered raised glands on the flat surfaces, as against the

much curved, sub-torulose pods, without raised glands of A. exuvialis.

157

Fig. 1. — Acacia borleae Burtt Davy: a, fruiting inflorescence ; b, portion of flowering

twig; c, leaflet, X 3.

Fig. 2. — Acacia exuvia/is Verdoorn : a, portion of flowering twig; b, fruiting

inflorescence.

158

Fig. 3. — Acacia permixta Burtt Davy: a, portion of fruiting twig; b, leaflet, k 5;

c, portion of flowering twig.

Fig. 4. — Acacia nebrownii Burtt Davy: a, portion of flowering twig; b, leaflet, X 3;

c, portion of fruiting inflorescence.

159

Fig. 5. — Acacia tenuispina Verdoorn: a, portion of flowering twig; b, leaflet, X 3;

c, portion of fruiting inflorescence.

Fig. 6. — Acacia swazica Burtt Davy: a, portion of leafy twig; b, inflorescence; c,

leaflet X 3; d, portion of fruiting twig.

1599—6

f

160

Distribution Map. ( Drawn by J. H. Rose brines').

161

Revision of the Genus Cyclopia and Notes on Some

Other Sources of Bush Tea.

By

P. Kies.

The leaves and flowers of most of the Cyclopias have been used as tea for many

years in the districts where they occur, and lately “ bush tea ” has become a marketable

commodity in wider circles since the great rise in price of Ceylon tea. A revision of

the relative nomenclature has become an urgent necessity. Hofmeyr and Phillips in

Bothalia 106 (1922) give a summary of the commercial varieties of these tea plants,

and, allowing for nomenclature alterations, in general their scheme still holds good.

The genus occurs in the coastal districts of the Cape Province, from Clanwilliam

to Port Elizabeth, being bounded in the north by the mountain ranges — Koue Bokkeveld

Mountains, Klein Swartberg, Groot Swartberg and Kouga Mountains, and the species

seem to be fairly localised.

The earliest mention in botanical literature of one of these plants was an illustration

in Pluk. Phyt. 4: t. 413, fig. 5 (1705) under the title “ Genista rosmarinifolio triphyllos

Promont, Bonae Spei ”. This figure was cited by later authors under various synonyms

until it was eventually placed under Cyclopia genistoides by De Candolle, Prod 2: (1825).

On the same plate of Pluk. Phyt., Fig. 4, is a drawing of a species of Aspalathus which

has repeatedly been misquoted by earlier authors under C. ga/ioides, but Vogel in

Linnaea X (1836) 595, put a stop to this error. A similar error has persisted with

regard to the identity of Gompholobium maculatum Andr. Bot. Rep. pi. 427, VI (1803).

The name was sunk under Ibbetsonia genistoides Sims, in Bot. Mag. t. 1259 (1810),

although the drawings clearly indicate the calyx and bracts in both, and the former is

without doubt identical with C. tenuifolia Lehm ., Linnaea 5: 373 (1830).

The century-old argument about the priority of Ecklon and Zeyher over Ernst

Meyer was settled in Journ. Soc. Bibliogr. Nat. Hist. 1, 4: 101-103 (1937), so that the

nomenclature changes here are made accordingly.

Although the generic characters of Cyclopia are very well defined, the specific

characters show a great variability and in various stages of development the branches

on one individual plant appear, when detached, to be specifically distinct. Systematists

have not been able to agree on the exact limits of each species or variety, and have

regarded species as varieties or vice versa. The present work is an attempt at finding

some stable characters for the classification of the species.

The leaf-shape varies considerably on individual plants, from terete to flattened;

the pubescence is very often entirely lost in the mature state of the branches, or persists

only as axillary tufts of hairs. The length of the pedicel varies with external conditions

or with the age of the flower, and in the fruit it is usually thickened. There are also

minute stipules on some leaves, but they are not constant, being absent from other

leaves on the same branch, or being replaced by additional leaflets as in C. plicata.

Even the number of bracts is not constant in C. burtonii, being either two or three

on the same specimen. The shape of the calyx varies with age; in the bud the actual

shape is more easily distinguished than in an open flower, where the retrofiection of the

calyx has occurred. The lower lobe is usually falcate, keeled and longer than the upper

four lobes, so that it is not of characteristic shape at all. Even in the petals there are

variations. In C. meyeriana the standard may be very clearly apiculate, or just minutely

mucronate, and in C. genistoides there is in some forms a small auricle at the lower

edge of the wing, which may be reduced or lacking on other flowers of the same plant.

162

The actual size and shape of the bracts and calyx seem to be constant, and together

with the texture of the leaves they have made some scheme of classification possible.

CYCLOPIA VENT.

Shrubs, leaves ternate, usually exstipulate, leaflets linear to ovate, margins often

revolute, sessile or shortly stalked. Flowers solitary, axillary, stalked or sessile, with

2 or 3 bracts at base of peduncle. Calyx intruse at base, five fid, lowest lobe carinate,

upper lobes adpressed to vexillum. Vexil/um suborbicular to oblong, plaited at base,

shortly clawed, with or without dark centre. Alae oblong with a transverse fold,

clawed. Carina obtusely rostrate, clawed. Stamens 10, free or slightly connate at

base. Style arcuate. Legume oblong, coriaceous, several seeded.

See figures 1 — 20 (b) and distribution map.

KEY TO CYCLOPIA SPECIES.

A. Stem densely hairy when young, may become glabrous with age, and tufts of hair persist only in

leaf axils; leaves at first hairy becoming glabrous:

I. Flowers pedicelled, bracts not clasping base of calyx :

a. Bracts not plicate:

a) Bracts and calyx villous on outer surface, leaflets terete, calyx lobes lanceolate

1. C. bolusii

b) Bracts and calyx villous on inside, subglabrous outside; leaflets at first hairy,

becoming glabrous, golden, terete to lanceolate; calyx lobes lanceolate

3. C. aurescens

c) Bracts pubescent inside, calyx glabrous both sides, leaflets terete, calyx lobes

lanceolate falcate, acute and apiculate:

i. Leaves stout, straight, dense 4. C. galioides

ii. Leaves slender, curved, lax 5. C. capensis

b. Bracts plicate, apex recurved:

a) Bracts and calyx pubescent outside, calyx ribbed, lobes acuminate

6. C. pubescens

b) Bracts and calyx pubescent inside, calyx lobes obtuse, pedicel pubescent

14. C. plicata

II. Flowers subsessile, bracts large, ovate, clasping calyx :

a) Bracts ciliate or villous on margins, calyx hairy on inside and on margins:

i. Bracts persistently villous on inner surface, calyx lobes ovate, obtuse, glabrous

outside 2. C. bowieana

ii. Bracts glabrous inside, calyx lobes acute, hairy on outer surface

7. C. meyeriana

b) Bracts glabrous on margins and outer surface, calyx becoming glabrous on inner surface :

i. Calyx villous on margin and outer surface, bracts glabrous inside

8. C. montana

ii. Calyx glabrous outside, ciliate on margin, bracts pubescent on keel inside

8. C. montana var. glabra

B. Stem glabrous or minutely pubescent when young, leaves generally glabrous :

I. Leaves terete or subterete by rolling in of margins:

a) Calyx lobes short, straight, deltoid, pubescent within:

i. Bracts obtuse, truncate, becoming coriaceous, leaf slender terete sub-linear or

lanceolate 9. C. maculata.

ii. Bracts acute, keeled, subcoriaceous, leaflets linear-lanceolate to terete, stem and

leaf golden, calyx in bud glaucous 3. C. aurescens var. glauca

b) Calyx lobes lanceolate, falcate, long, pubescent on margins, or glabrous, leaves reflexed

when mature:

i. Leaves slender, terete, curved .10. C. genistoides var. teretifolia

ii. Leaves lanceolate, margins incompletely inrolled 10. C. genistoides var. heterophylla

iii. Leaves subterete, sulcate below 10. C. genistoides

iv. Leaves ovate to terete on same plant 10. C. genistoides var. ovalifolia

163

II. Leaves flat, margins sometimes inrolled :

a) Bracts 2 or 3 on same plant, flowers small, bracts and calyx acute ...11. C. burtonii

b) Bracts always 2:

i. Bracts obtuse with a reflexed mucro :

0 Plant slender, pedicels filiform, leaves linear, lanceolate 12. C. filiformis

00 Plant stouter, leaves elongate, spathulate to linear, bracts ribbed slightly

13. C. longifolia

ii. Bracts acute or obtuse, without reflexed apex :

0 Bracts large, subequal to pedicel, 6-9 mm. long, acuminate, calyx lobes long,

acute, falcate or subfalcate, leaves elliptic, lanceolate or linear, coriaceous,

glaucous 15. C. falcata

00 Bracts medium to small, shorter than pedicel :

X Leaves short, ovate, obtuse at apex, bracts 0-4-0 -6 cm. long:

1 . Leaves cordate or truncate at base, distinctly marked with veins on

undersurface. Calyx lobes acute 16. C. buxifolia

2. Leaves obtuse or acute at base, subcoriaceous, calyx lobes falcate,

lanceolate, acute, leaf margin crisped somewhat

15. C. falcata var. ovata

3. Leaves ovate to terete on same plant

10. C. genistoides var. ovalifolia

XX Leaves linear, lanceolate or spathulate:

1. Bracts sub-truncate, coriaceous, plant with a golden sheen when

dried, calyx lobes short, deltoid, convex on back:

A Leaves thickly coriaceous broadly linear, elliptic, obtuse, calyx

and bracts pubescent inside and on margins

17. C. intermedia

A A Leaves subteretely rolled or narrowly linear — 9. C. maculata

2. Bracts acute, subcoriaceous:

A Leaves golden sheen when dried, subteretely rolled to linear-

lanceolate, calyx glaucous, short deltoid

3. C. aurescens var. glauca

A A Calyx lobes lanceolate, falcate, long, acute, leaves becoming

reflexed 10. C. genistoides var. heterophylla

AAA Calyx lobes short, acute, leaves ascending:

□ Leaves elliptic-lanceolate, acute at base and apex, veined

on undersurface, margins inrolled, calyx acute, bracts

shorter than pedicels 18. C. dregeana

□ □ Leaves linear to spathulate, obtuse at apex, or subacute,

narrowing towards base, slightly veined, subcoriaceous,

calyx lobes short, obtuse, bracts small:

+ Pedicels short, stout, 2-3 mm. long, may elongate in fruit,

bracts clasping pedicels, 2-3 mm. long, leaves linear to

elliptic, veined 19. C. sessiliflora

+ + Pedicels elongated, bracts acute, much shorter than pedi-

cels, leaves attenuate at base, much smaller on axillary

branchlets, not veined, leaflets cuneate.

20. C. subternata

+ + + Leaflets narrow-linear spathulate

20. C. subternata var. laxiflora

1. C. bolusii Hofm. and Phill. Both. 1:109 (1922).

Small shrub; stem at first hairy, becoming glabrous; bark breaking off in squares;

leaf-bases prominent; leaflets at first hirsute, becoming glabrous, ericoid, subclavate,

obtuse, sulcate below, reflexed or ascending, curved or straight, laxly arranged, 5-7 mm.

long; flowers few, pedicel present; bracts lanceolate, hairy outside, keeled and sulcate

on either side of keel, as long as pedicel (5 mm.); calyx lobes lanceolate, subfalcate,

acute to obtuse, hirsute on outer surface, ciliate on margins, subglabrous within.

Prince Albert Division: Swartberg Pass, Bolus 11465 (type); Klein Swartberg,

Andreae 1250.

164

2. C. bowieana Harv. FI. Cap. 2: 9 (1862) “ Vleitee.” C. ashtonii Hofm. and Phill. l.c.

Small shrub, branches at first silky, becoming glabrous with tufts of hair persisting

in leaf axils; stem and leaves with a golden sheen when dried; leaflets from a thick

flattened node, up to 1-75 cm. long, 7 mm. broad, cordate at base, flattened or margins,

revolute and leaflets subterete, sulcate and pale glaucous grey on undersurface; bracts

4-8 mm. long, broadly ovate, ciliate, villous inside, glabrous externally; calyx 8 mm.

long, lobes ovate, obtuse, long ciliate on margins, hairy inside, glabrous externally;

flowers subsessile; calyx clasped at base by bracts.

No locality: Bowie (type in Kew Herb.). Swellendam Division: Burehell 7350;

Grootvadersbos, Ashton 2843; Lemoenshoek, Esterhuysen 10444, George Division:

Burehell 5893.

3. C. aurescens sp. nov.

Nearest C. bolusii, but calyx sub-glabrous and leaves somewhat flattened.

Shrub up to 75 cm. high; stem and leaves at first shortly hairy, becoming glabrous,

with a golden sheen when dried; stem ribbed and angled with prominent leaf bases;

leaflets thickly coriaceous, mucronate subterete and sulcate, becoming flattened, with

only margins slightly revolute; the upper surface shiny golden, undersurface dull,

glaucous; bracts sparsely pubescent outside, becoming glabrous, hairy inside and on

margins, 5-6 mm. long, ovate, acute, mucronate, coriaceous in fruit, shorter or sub-

equal to pedicel; calyx lobes 4 mm. long, lanceolate obtuse, subfalcate, convex, keeled,

pubescent inside and on margin, outside at first hairy, becoming glabrous. Var .glauca:

differs from type, being glabrous, and calyx lobes shortly deltoid, glaucous.

Fruticulus erectus, 75 cm. altus. Rami aurei, costati; juniores puberuli. Foliola

2-5 cm. longa, supra aurea, subtus glauca, coriacea, mucronata; juniora teretia, seniora

lanceolata, marginibus revolutis. Bracteae 5-6 mm. longae, ovatae, acutae, mucronatae,

extus glabrae, intus villosae, margine puberulae, pedicellis subaequales. Calycis /obi

4 mm. longi, lanceolati, subfalcati, obtusi, dorso rotundati, intus pubescentes, extus

glabrescentes. Fructus lanceolatum, obliquum.

Var. glauca a typo ramis glabris lobis calycis deltoideis glaucis differt.

Riversdale Division: Garcias Pass, Bolus 11249. Ladismith Division: Swartberg,

Marloth 3988. Laingsburg Division: Witteberg, Compton 2801. Prince Albert Divi-

sion: Swartberg, Andreae 1255; Pocock SI 44; Stokoe 9024; Sewe Weeks Poort, Primos

33, Stokoe 1919. Var. glauca Ceres Division: Bolus Herb. 15447; Koue Bokkeveld,

Schlechter 8922. No locality: Pappe, Herb. Hook.

4. C. galioides (Berg.) D.C. Prod. 2: 101 (1825).

Sophora galioides Berg. Cap. 141 (1767); Sophora genistoides Thb. var. 3, Prod.

(1794); Podalyria genistoides Willd. Sp. PI. (1799) partly; Galega genistoides

Thb. var. B, FI. Cap. (1806); Cyclopia genistoides Walp. in Linnaea 13 (1839)

partly; C. galioides E. and Z. Enum. (1826); E. Mey. Comment. 1 (1836);

Vogel in Linnaea 10 (1836); Benth. Ann. Mus. Vind. 2 (1839); Benth. in Hook.

Lond. Journ. Bot. (1843); Harv. l.c.; Hofm. and Phill. l.c.

Stout shrub; branches angular with prominent leafbases, at first densely villous

with greyish hairs, glabrous with age, bark purplish; leaflets up to 2-5 cm. long, at first

hairy, becoming glabrous, sessile, subterete, sulcate beneath, stout, straight, close

together on stem or somewhat lax, apex mucronate, ascending; bracts 4-8 mm. long,

acuminate, ovate to lanceolate, as long as the pedicel or exceeding it, glabrous outside,

puberulous inside; pedicels short, glabrous; calyx glabrous; lobes lanceolate falcate,

acuminate, apiculate, flowers in springtime.

165

No locality: Scholl in Herb. Mus. Vind. (1836); Grey, Herb. Kew., Cape Peninsula;

Vlaggeberg, Schlechter 459; Table Mountain, Flanagan 2414; Devils’ Peak, Alexander,

(1847); Muyzenberg, Burke and Zeyher\ Kalk Bay, Alexander (1846); Vis Hoek, Bolus

67; Simonstown, Jameson (1841); Wright ; Compton ; Pillans 3039.

5. C. capensis Salter in Journ. S.A. Bot. 5 (1939).

An erect shrub, 30-50 cm. high, rarely 1 -5 metres high; branches virgate, ascending,

at first leafy and sparsely grey villous, becoming nude and glabrous and light brown;

petioles 1 mm. long, pilose in the axils; leaflets slender, linear terete, curved, mucronate,

pilose at first, becoming glabrous, 1-1 -5 cm. long, canaliculate below; flowers congested

towards apex of branches pedicel 4-7 mm. long; bracts 5-8 mm. long, glabrous on both

surfaces, ovate-lanceolate, acute; calyx lobes, broadly lanceolate, falcate, apiculate,

6 mm. long; carina minutely cuspidate, emarginate at apex; flowers in late summer.

Cape Peninsula: Table Mountain, Wolley-Dodd 875; Wynberg, Esterhuysen 12643;

Paulsberg, Salter 8035, 8036; Compton ; Kommetjie, Galpin 12694; Smitswinkel,

Salter 8034; False Bay, Alexander (1846); Simon’s Bay, Alexander (1846); Bright

Water, Leighton 1616.

6. C. pubescens E. and Z. l.c.

Vog. in Linnaea 10 (1836); Walp. l.c. Benth l.c.; Harv. l.c.; Phill and Hofm. l.c.

Plant pubescent; leaflets linear, obtuse, mucronate, margins slightly revolute,

midrib prominent below, thinly pubescent to glabrous, up to 2-5 cm. long; bracts

wrinkled longtitudinally, apex acuminate recurved, pubescent in and outside or sub-

glabrous; flowers pedunculate, pedicel 1-5-1 -8 cm. long, calyx pubescent both in and

outside, distinctly nerved, lobes 0-6 cm. long, acuminate from a deltoid base.

Port Elizabeth Division: Ecklon and Zeyher 1146; Van Staaden’s River, Bolus 1604

(partly).

7. C. meyeriana Walp. l.c.

C. sessiliflora E. Mey, non E. and Z. l.c. ; Benth. l.c. ; Harv. l.c.

Stem at first densely villous, becoming glabrous; leaves crowded, persistently hairy

in the axils; leaflets pilose, becoming glabrous with age, slender, terete, sulcate below;

flowers sessile, crowded towards apex of branches; bracts broadly ovate, clasping

base of calyx, acuminate, 6-8 mm. long, shiny, glabrous, villous on margins; calyx

villous on margin, deciduously hairy on outside and inside, lobes lanceolate, straight;

standard usually clearly apiculate.

No locality: Drege (1836). Somerset West Division: Hottentots Holland, Stokoe

13/11/34; 7/11/38. Paarl Division: Drakenstein, Drege; Du Toits Kloof, Drege.

Worcester Division: Wilde Pardeberg, Stokoe 1055; Baviaans Kloof, Leighton 1357.

Caledon Division: Genadendal, Burchell 7770; Schlechter 9826; Landdroskop, Stokoe

7813, Esterhuysen 2636.

8. C. montana Hofm. and Phill. l.c.

Stem and leaves laxly silky haired; stem angled to subterete, bark reddish, with a

golden sheen when dried; leaves with a tuft of hairs persisting at the base, laxly arranged;

leaflets narrowly terete, cordate at base; flowers subsessile towards apex of branches;

bracts 4-6 mm. long, broadly ovate, acute, clasping calyx at base, glabrous in and

outside; lobes of calyx 5-7 mm. long, lanceolate, acute, straight, becoming glabrous,

at first densely long hairy inside, outside and on margins, standard not distinctly apicu-

late. Differs from C. meyeriana by glabrous bracts.

Var. glabra Hofm. and Phill. l.c., has calyx glabrous, ciliate on margins, petals

large and pale yellow.

166

Ceres Division: Ertjies Kloof, Compton 16087. Worcester Division: Witteberg,

Esterhuysen 9466, 8677. Tulbagh Division: Winterhoek, Phillips 11131. Var. glabra

Ceres Division: Conical Peak, Stokoe 7814, 7815. Worcester Division: Hex River,

Esterhuysen 9364, 9369, 8707; Waaihoek, Esterhuysen 9932, Pells 15/12/38; Fontein-

tjiesberg, Esterhuysen 8767; Brand wag, Esterhuysen 11012; Matroosberg , Mar/oth 2277,

2370, Esterhuysen 14221.

9. C. maculata ( Andr .) comb. nov. “ Vleitee.”

Gompholobium maculatum Andr. Bot. Rep. 6:427 (1803); Cyclopia tenuifolia Lehm.

in Linnaea 5 (1830) 373; Benth. l.c.; E. and Z. l.c. ; Harv. l.c.; Hofm. and Phill. l.c. ;

C. laricina E. Mey. l.c.

Shrub 60-90 cm. high, glabrous, with ribbed branches; leaflets linear filiform or

lanceolate, 2-3-5 cm. long, acute, margins inrolled, sulcate or open backed; pedicels

up to 1 cm. long, in fruit 1 -5 cm. and much thickened; bracts truncate, obtuse, coriace-

ous, thickened with age, keeled, sparsely hairy within, 3-5 mm. long, broader than long;

calyx lobes 3 mm. long, ovate-lanceolate, obtuse, glabrous with margins pubescent,

uninerved; vexillum of a deeper colour above the claw; legume oblong, coriaceous.

No locality: Zeyher in Nl. H. 25099; E. and Z. in Nl. H. 10019; “ with Wieb 47 ”

in Herb. Kew. Worcester Division : Chavonnesberg, Esterhuysen 8205. Paarl Division :

Bainskloof, van Breda 76. Stellenbosch Division: Eerste Rivier, Herb. Hook. Nov.

1840; French Hoek, Bolus 6980. Somerset West Division: Hottentots Holland,

Alexander (1846). Caledon Division: Elgin, F.D.H. 2840. Swellendam Division:

Storm Valley, Zeyher 2257 ; Rivier-Sonder-End, E. and Z. 1150; in N.H. 24359; Groot-

vadersbos, F. D. H. 3750; Burchell 7522. Riversdale, Corente River, Muir 59.

10. C. genistoides (L.) R. Br. Hort. Kew 3: 5 (1811) “ Heuningtee. ”

Sophora genistoides Linn. Sp. PI. Tom. I Ed. Ill (1764); Berg, l.c.; Burm. FI.

Cap. Prod. (1768); Thb. Prod. PI. Cap. (1794); Podalyria genistoides Willd. l.c.; Ga/ega

genistoides Thb. d, Prod. PI. Cap. 2: 192 (1800). Ibbetsonia genistoides Sims. Bot. Mag.

1259 (1810); Cyclopia genistoides E. Mey. l.c.; E. and Z. l.c.; Benth. l.c.; C. galioides

E. Mey. l.c.; C. teretifolia E. and Z. l.c.; C. heterophylla E. and Z. l.c.; var. linearifolia

E. and Z. l.c.

Shrub with branches ribbed, becoming terete, glabrous, or puberulous, erect, nude

at base, leafy above; leaves dense, shorter or longer than internodes, reflexed when

mature; leaflets ovate or linear, subterete or open backed with margins incompletely

rolled in, 0-5-3 -5 cm. long, obtuse, glabrous; pedicel 0-5-1 -5 cm. long, flowers crowded

at apex of branches ; bracts ovate, acute, glabrous outside, pubescent inside, 5-7 mm.

long, usually shorter than pedicel; calyx lobes lanceolate, falcate, acuminate, 4-6 mm.

long, glabrous or puberulous on margins; fruit lanceolate; wings sometimes auricled

on lower margin, just above claw.

Var. heterophylla Harv. l.c.: Leaflets terete to lanceolate, margins incompletely

inrolled. C. heterophylla E. and Z. l.c.

Var. teretifolia ( E . and Z.) comb. nov. : leaflets slender, curved, short. C. teretifolia

E. and Z. l.c.

Var. ovalifolia var. nov.: leaflets subterete to flattened, ovate; small shrub, flowers

long stalked.

Var. ovalifolia var. nov.: suffrutex humilis, foliolis ovatis vel subteretibus floribus

pedicellatis.

No locality: E. and Z. in N.H. 10018; Siebert 47; Herb. Harv. 745; Zeyher 353;

Drege (1876); Burchell 417; Pappe 9, Herb. Hook.

Malmesbury Division : Groenkloof, Drege, N.H. 24347, 24289; Bokbaai, Esterhuysen

3829. Worcester Division: Stettynsberg, Esterhuysen 11100; Du Toits’ Peak, Ester-

huysen 8568. Paarl Division: Wemmershoek, Mar/oth 819, Esterhuysen 9667, 11277;

167

Fransch Hoek, Stokoe; Herb. Harv. 770. Cape Peninsula: Wynberg, Bolus 7272B;

Kirstenbosch, Esterhuysen 11842, Hutchinson 29, MacOwan 76; Devils Peak, Wolley-

Dodd 1320; Hout Bay, Mar loth 280; Alexander (1846); Simon’s Bay, Wright 9/657;

Wolley-Dodd 1328. Stellenbosch Division: Burchell 8320; De Jongh, 4390; Guardian

Peak, Esterhuysen 11984. Somerset West Division: Hottentots Holland Mountains,

Stokoe\ Alexander (1846); Sir Lowry’s Pass, Hutchinson 336. Caledon Division:

Elgin, Hubbard 473, F.D.H. 2842; Bolus', Zeyher (1865); Kleinriviersberg, Zeyher in

N.H. 11833, Esterhuysen 2915; Table Mountain, Marloth 6046; Kogelberg, Stokoe 1015.

Bredasdorp Division: Potberg, Pillans 9473, 9328, 9500. Swellendam Division:

Elandspad, Pillans 9406. Var. heterophylla : Swellendam Division : Rivier-Sonder-End,

E. and Z. 1148, Stokoe 9004. Var. teretifolia: Caledon Division: E. and Z. 1145;

Hermanus, Bolus 9863. Var. ovalifolia: Caledon Division: Sneeukop, Esterhuysen

2635. Uniondale Division: Mannetjiesberg, Esterhuysen 6469.

11. C. burtonii Hofm. and Phill. l.c., “ Heuningtee,” “ Suikertee

A small much-branched shrub, 30-41 cm. high, with leaves, flowers and fruits

smaller than the average size for the genus ; leaflets lanceolate, flat or channelled below,

acute, 0-6-1 -2 cm. long, 2 mm. broad, coriaceous; bracts 2 or 3 on the same branch,

two bracts at base of pedicel, and a smaller one adnate to pedicel, acute, 1-2 mm. long,

lanceolate; pedicel 1-2 mm. long; calyx lobes acuminate, 1-2 mm. long, ovate, pubescent

on margins; vexillum 8-5 mm. long, 6 mm. broad, acute; alae narrowly oblong; carina

obtusely rostrate.

Oudtshoorn Division: Swartberg, Burton in F.D.H. 2914. Prince Albert Division:

Swartberg, Minnaar 5, Stokoe 9020.

12. C. filiformis sp. nov.

Small shrub, stem terete, bark glaucous, purplish when dried; leaflets up to 2 cm.

long, linear lanceolate, margins slightly reflexed, midrib prominent below, apex obtuse,

membranous, ascending or reflexed when mature; flowers lateral in leaf axils, smaller

than in the average size for the genus; bracts 2-3 mm. long, subtruncate, apex reflexed,

mucronate, keeled ; pedicels filiform up to 8 mm. long ; calyx lobes deltoid, acute, 1 • 5 mm.

long, exceeding the tube; standard ovate-lanceolate; wings narrowly oblong, claw

1 -5 mm. long, blade 9 mm. long; carina obtuse, as long as wings; legume lanceolate,

small; style persistent; flowers appear in November.

Fruticulus 20 cm. altus. Rami teretes, glabri, glauci. Foliola ascendentia deinde

reflexa, membranacea, 2 cm. longa, lineari-lanceolata, apice obtusa, marginibus revolutis

nerviis subtus prominentibus. Bracteae 2-3 mm. longae, subtruncatae carinatae,

glabrae, apice emarginatae. Flores laterales, pedicellis filiformibus 8 mm. longis.

Calycis lobi deltoidei, acuti. Vexillum ovato-lanceolatum. Alae oblongae, unguiculatae.

Carina obtusa, unguiculata, usque 9 mm. longam. Legumen parvum, lanceolatum,

stylo persistente.

Port Elizabeth Division: Van Staadensrivier, MacOwan 336/A.K.

13. C. longifolia Vog. in Linnaea 10: 595 (1836); Walp l.c.; Benth. l.c.; Harv. l.c.

Glabrous shrub, branchlets angled ; leaflets subcoriaceous, oblong, linear, attenu-

ate at base, obtuse at apex, 3-5 cm. long or more, flat with extreme edges revolute,

dark green above, paler below, midrib prominent below; pedicel exceeding bracts;

bracts ovate oblong, plicate-striate, apex recurved, 0-4 cm. long; calyx lobes acute,

short, 3 mm. long; petals 9 mm. long.

No locality; Mundt and Maire in Kew Herb.; ex Hort. Reg. Berol. Port Elizabeth

Division : Van Staadens, Bolus 1604 (partly).

168

14. C. plicata sp. nov.

Stem pubescent becoming glabrous, ribbed, with leafbases prominent, golden to

reddish in colour; leaves subpetioled, hairy in axils, stipulate; leaflets 3 or abnormally

4 or 5 where leaflets replace the stipules; stipules short, subulate, minute or wanting;

leaflets subterete, sulcate or open backed below, coriaceous, curved, ascending 0-8-

1 -8 cm. long; bracts obtuse, deeply plicate, apex recurved, coriaceous, glabrous outside,

hairy inside, 4-5 mm. long; pedicel 5-7 mm. long, pubescent; calyx pubescent in and

outside and on margin, becoming glabrous; petals 1-2 cm. long, legume 3 cm. long,

oblique and curved, several seeded.

Rami juniores pubescenti, seniores glabri aurei vel rubro-brunnei. Folia petiolata,

stipulata, 3-5 foliolata, axillis villosis. Foliola subteretia, 0-8-1 -8 cm. longa, glabra,

sulcata, marginibus revolutis. Stipulae subulatae, minutae vel nullae. Bracteae

4- 5 mm. longae, plicatae, apice recurvae, intus pubescentes, extus glabrae. Pedunculi

5- 7 mm. longi, pubescentes. Ca/ycis lobi 1-2 mm. longi, pubescentes, basi lati apice

obtusi. Legumen obliquum, leviter curvatum.

Uniondale Division: Hoopsberg, Esterhuysen 6549 (Type in Herb. Bolus); Mannet-

jiesberg, Esterhuysen 6466; Smutsberg, Esterhuysen 7035; Marloth Herbarium, under

C. longifiora, no locality or details.

15. C. falcata (Harv.) comb. nov.

C. vogelii var. falcata Harv. l.c. ; C. brachypoda Hofm. and Phill. l.c. pro parte,

non Benth.

Glabrous erect shrub, stems angled to terete, with a golden sheen or glaucous;

leaflets flat or margins revolute, oblong elliptical, linear, ovate or lanceolate tips mucron-

ate, base acute or obtuse, midrib prominent below, up to 2-8 cm. long, up to 1 cm.

broad, glaucous, coriaceous; bracts large, 5-9 mm. long, keeled, acuminate, ovate,

puberulous inside, glabrous outside, as long as, or exceeding petiole; calyx lobes acute

to obtuse, lanceolate, falcate, pubescent on margins, 4-8 mm. long.

Var. ovata var. nov.; leaflets short, ovate, margins crinkled slightly, bracts and

calyx shorter, leaves stipulate, leaflets 0-5-1 cm. long, 0-19-0-5 cm. wide, bracts 3-6 cm.

long, acuminate.

Var. ovata var. nov.: foliolis ovatis 0-5-1 cm. longis 0- 1-0-5 cm. latis, bracteis

0-3-0 -6 cm. longis acutis lobis calycis lanceolato-falcatis.

Tulbagh Division: Zeyher 354. Worcester Division: Onklaarberg, Stokoe 1158;

Paarl Division: Fransch Hoek, Bolus 5151. Caledon Division: Alexander, Oct. 1846.

Elgin, F.D.H. 2841, Swarteberg, Schlechter 5583. Stellenbosch Division: Helderberg,

Parker 3392. Robertson Division: Onklaarberg, Stokoe 7411. Var. ovata; Ceres

Division: Michells Pass, Esterhuysen 6167. Tulbagh Division: Witzenberg, Esterhuysen

1281. Worcester Division: Hexrivier, Esterhuysen 9362; Matroosberg, Gillett 3603;

Chavonnesberg, Esterhuysen 8175. Caledon Division: Rivier-Sonder-End, Stokoe

7391. Paarl Division: Wemmershoek, Esterhuysen 11272.

16. C. buxifolia (Burm. f.) comb. nov.

Genista buxifolia Burm. f. l.c.; Cyclopia latifolia D.C. l.c.; Harv. l.c.; Hofm. and

Phill. l.c.; C. cordifolia Benth. Ann. Mus. Vind. 2: 67 (1838). C. latifolia

Benth. in Hook. Lond. Journ. Bot. 2: 432 (1843).

Glabrous shrub; stem ribbed, downwards from leafbases, glaucous, purplish when

dried; leaflets ovate or ovate-lanceolate, truncate or cordate at base, 0-5-2 -5 cm. long

and 0-3-1 -4 cm. broad, coriaceous with impressed veins on undersurface of leaflet,

petiolate, apex obtuse; bracts lanceolate, keeled, acuminate, 4-7 mm. long, shorter or

equal to pedicel; calyx tube short, lobes acute, ovate, equal to or exceeding tube, rare.

169

No locality: Scholl in Herb. Mus. Vind. (1836). Cape Division: Table Mountain,

Bodkin in Bolus Herb. 8038; Esterhuysen 4014; Constantiaberg, Wolley-Dodd , 3450;

Waai Vlei, Wolley-Dodd 3399. Ladismith Division: Swartberg, Marloth 3988 b.

Prince Albert Division; Swartberg, Pocock 164.

17. C. intermedia E. Mey. l.c. exl. lit. C. “ Kouga Bush tea” Walp. l.c. ; Benth. in

Hook. Lond. Journ. Bot. 2 (1843); C. vogelii var. intermedia Harv. l.c.; C. brachy-

poda var. intermedia Hofm. & Phill. l.c. ; C. aurea Fourc. in Trans. Roy. Soc. S.A.

21 (1934); C. subternata Hofm. and Phill. l.c. partly.

Erect shrub 0-5-0 -7 m. high; stem and leaves with a golden sheen when dried;

twigs angular; leaflets on a decurrent shield, ciliate within, linear cuneate with revolute

margins, 16-30 mm. long, 2-5-6 mm. broad, coriaceous, obtuse, mucronate; bracts 2,

concave, ovate, obtuse, keeled, coriaceous, thickened with age, hirsute within, pubescent

on margins, 3-6 mm. long, shorter than pedicel; pedicels vary from 0-3-2 cm. in length;

calyx lobes short, deltoid-ovate, obtuse, roundbacked, lowest lobe keeled, longer pubes-

cent on margins, sparsely hairy within, 2-3 mm. long at first glaucous, becoming shiny,

coriaceous.

No locality: Drege (1836) Hort. Reg. Berl.

Riversdale Division: Marloth 5666; Garcia’s Pass, Galpin 3900; Langeberge,

Marloth 7186 (?). Oudtshoorn Division: Olifants River, Gill, Herb. Kew; Uniondale

Division: Langkloof, Burchell 4929; Kouga Mountains, Esterhuysen 10738; 7036.

Humansdorp Division: Zuur Anys, Fourcade 3049; Blouberg Forest, Holland 4063,

Dix 3.

18. C. dregeana nom. nov.

C. latifolia E. Mey. l.c. non. D.C. ; Walp, l.c. partly; non Benth.; C. vogelii var.

subternata Harv. l.c. partly; C. subternata Hofm. and Phill. l.c. partly.

Erect shrub, stem ribbed to terete, purplish; leaflets lanceolate, acute at base and

apex, slightly veined on undersurface, coriaceous, margins revolute, 2-5 cm. long,

0-8 cm. broad, shortly petiolate, and leaves minutely stipulate on some twigs, bracts

small, acute, keeled, about one-third length of pedicel, 3 mm. long; calyx lobes lanceo-

late, acute, overlapping at the base. Differs from C. intermedia by small acute bracts

and veined leaflets, from C. buxifolia by narrow lanceolate leaflets, from C. subternate

by acute leaflets, and slightly larger bracts and calyx lobes, and from C. falcata by

smaller size of bracts and calyx.

Ceres Division: Schurfteberg, Compton 16230. Tulbagh Division: Grant 2474.

Worcester Division: Du Toits Kloof, Drege (1836). Paarl Division: Fransch Hoek,

Bolus (1895).

19. C. sessiliflora E. and Z. non E. Mey, l.c. “ Heidelbergtee,” “ Gewone bossiestee,”

“ Hottentotstee ”.

Vogel l.c.; Walp. l.c.; C. brachypoda Benth. in Ann. Mus. Vind. 2, 1: 67(1837);

Hook, Lond. Journ. Bot. 2. (1843); C. vogelii var. brachypoda Harv. l.c. 6;

C. brachypoda Hofm. and Phill. l.c. partly; Rafnia retroflexa B. Thb. in Herb.

Thb.

Shrub with virgate, subfastigiate flexuous branches, stem cicatrised at base; leaflets

linear or oblong, obtuse or attenuate at base, margin flattened or revolute, shiny on

upper surface, veined below; membranous or subcoriaceous, dense to laxly leafy,

leaves ascending; bracts 1-2 -5 mm. long, ovate, clasping pedicel, pedicel scarcely exceed-

ing bracts, may elongate in fruit; calyx lobes 2-4 mm. long, short, subacute or obtuse;

vexillum emarginate; fruit obliquely lanceolate.

170

No locality: Burchell 6117, 7223; Herb. Hook., Mundt. Herb. Hook. Swellendam

Division; Puspasvlei, E. and Z. 1147; Lemoenshoek, Esterhuysen 10443. Riversdale

Division: Langberg, F.D.H. 2844, 2845, Marloth 9318; Muir 483, Burchell 6701.

20. C. subternata Vogel l.c. 595; “ Gewone bossiestee

Benth., Hook. Lond. Journ. Bot. II (1843) partly; Hofm. and Phill. l.c. partly;

C. grandiflora A.D.C. Not. 8 ml. PI. Rar. Jard. Genev. 29 (1840); C. grandi-

folia Benth. l.c. ; C. laxiflora Bth. Ann. Mus. Vind. 2 (1837); C. latifolia

E. and Z. non D.C. l.c.; C. vogelii var. subternata Harv. l.c. partly; C. vogelii

var. laxiflora Harv. l.c.; C. latifolia Walp l.c. partly; Benth. Ann. Mus. Vind.

II (1840) 67.

Shrub 1-3 m. high; twigs angled, becoming terete, at first glaucous; leaflets some-

times minute on young twigs, 2-3 mm. long, when expanded 2-5 cm. long, 0-5 cm.

broad; spathulate lanceolate or sublinear, obtuse at apex, tapering to base, flat or

margins slightly revolute, membranous to subcoriaceous ; bracts 2-5 mm. long, keeled,

about a third the length of pedicel or less, acute or acuminate, sulcate on either side of

keel; calyx lobes short acute, deltoid, 2-3 mm. long.

Differs from C. falcata by the small bracts and short calyx lobes, and from C. ses-

siliflora by the pedicels being much longer, and by the presence of axillary shoots with

small leaves.

Var. laxiflora (Bth.) comb. nov. has leaflets narrowly linear, spathulate, mem-

branous, otherwise as type.

No locality: Burchell 5519, 5549, 6016; Bowie : Herb. Berol.; Herb. Benth.; Mundt,

Herb. Hook. Riversdale Division: Garcias Pass, Phillips 341; Langeberg, Muir 798.

George Division: Alexander Aug. 1847; Outenique Mountains, Flops 54; Knysna,

Kapp 48; Rodin 1169; Salter 6938; Keet, Sept. 1918. Uniondale Division: Zitzikama,

Galpin 3899; Joubertina, Esterhuysen 10599. Humansdorp Division: Clarkson,

Thode A785; Coldstream, Marloth 13063; MacOwan A.K. ; Fourcade 1631. Var.

laxiflora: Knysna: Plettenberg, E. and Z. 1149.

OTHER SOURCES OF “BUSH TEA”.

Aspalathus tenuifolia D.C. Prod. 2, p. 143 (1825).

A slender shrub, with leaves needle-shaped, up to 6 cm. long, single or fascicled,

curved or straight ; flowers few, interruptedly racemose, flowers distant, pedicels 4-5 mm.

long. Calyx teeth short, distant on the tube; vexillum pubescent, longer than the

arched, glabrous or puberulous carina. Leaves, twigs and calyx laxly adpressed hairy.

The flowers are up to 1 cm. long, larger than in A. contaminatus. This shrub only

occurs in the Piquetberg District and is the source of “ Naald Tee ” according to

Marloth.

Piquetberg District: Drege, rocky gorge on mountain, Nov. 1830; Pillans 7316;

Bolus 7526; Edwards 221; Marloth 5361, 11063.

Aspalathus contaminatus (Thb.) Druce Rep. Bot. Ech. Cl. Br. Is. 1916; A. lebeckia

contaminata Thb. FI. Cap. (1823) 561; A. corymbosus E. Mey. Linnaea (1832)

159; Borbonia pinifolia Marl. Trans. Roy. Soc. S.A. II (1912) 238.

Shrub prostrate or erect and up to 2 m. high, with numerous intermediate forms,

leaves solitary or fascicled, long linear terete, 2 • 5-4 cm. long, stouter than in A. tenuifolia

and usually straight; flowers on short pedicels, few, in pairs or imperfect corymbose-

racemules; calyx teeth lanceolate, shorter than tube, 2 mm. long, corolla 0-5 cm. long,

vexillum pubescent, carina pubescent, equal in length to the vexillum. Marloth states

that this plant is used as the source of “ naald tee ”, “ koopmans tee ”, “ Black tea ” or

171

“ rooibostee ” on the Cedar Mountains near Clanwilliam. The form on the Cape Penin-

sula is prostrate, while the form on the Cedar Mountains attains the height of 2 metres,

but they cannot be separated on any other character.

No locality: Ecklon 56, Ecklon and Zeyher 1396; Herb. Harvey 804 (1839); Mund

Maire ; Lehman; Wallich, Pappe 431. Calvinia Division: Henrici 3413. Clanwilliam

Division: Cedar Mountains, Marloth 4728; 7428; Olifants River Mountains, Marloth

7369; Pakhuis Pass, Esterhuysen 7420, Salter 5045; Bath in Bol. Herb. 15421; Rogers

16817; Drege; Wipperthal, Marloth 8313. Ceres Division: Klein Vlei, Schlechter

10065. Worcester Division: Stokoe 7308, van Breda 5. Malmesbury Division:

Groenkloof, Zeyher 440. Paarl Division: Paarl and Pont. Drege (1830). Stellenbosch

Division: Simonsberg, Wolley-Dodd 2%6. Somerset West Division : Sir Lowry’s Pass,

Schlechter 7201. Cape Division: Table Mountain, Prior. MacOwan 59; Vlaggeberg,

Wolley-Dodd 720; Simonstown, Schlechter 308, Bolus 3815; Cape Town, Rogers 16139,

Burchell 917. Caledon Division: Burchell 8128.

SUMMARY OF COMMON NAMES OF SPECIES USED AS TEAS

AND THEIR LOCALITIES.

C. bowieana HarV.

(=C. ashtonii Hofm. and Phill.)

Only on Langeberge to Outeniquas. Swellendam to George.

C. maculata (Andr.) Comb. nov.

(=C. tenuifolia Lehm.)

Drakenstein Mountains, Rivier-Sonder-End Mountains and

Langeberge.

(Worcester, Paarl, Stellenbosch, Somerset West, Caledon,

Swellendam, Riversdale Districts.)

C. genistoides R. Br. (most common tea plant).

Table Mountain and Coast, Drakenstein Mountains, Rivier-

Sonder-End Mountains, and Potberg (with outlying speci-

men in Uniondale District).

(Malmesbury, Cape, Paarl, Worcester, Somerset West,

Caledon, Bredasdorp, Swellendam and Uniondale Districts.)

C. burtonii Hofm. and Phill.

Swartberg only. •

(Oudtshoorn and Prince Albert Districts.)

C. subteretifolia sp. nov.

Koue Bokkeveld Mts., Witteberg, Klein Swartberg, Groot

Swartberg, Langeberge.

(Ceres, Laingsburg, Prince Albert, Ladismith, Oudtshoorn,

Riversdale Districts.!

C. sessiliflora E. and Z.

(=C. brachypoda Hofm. and Phill.)

Langeberge only.

(Swellendam and Riversdale Districts.)

C. subternata Vogel

(=C. subternata Hofm. and Phill.)

Langeberge, Attaquas, Outeniquas, Zuuranys, Langkloof,

Zitzikama, Kouga Mts.

(Coastal ranges from Riversdale to Humansdorp Districts.)

Vlei tee.

Heuning tee.

Heuning tee f

Suiker tee ■{

Swartberg tee [_

Bossies tee

Swartberg tee

Gewone bossies tee. .

Heidelberg tee ■(

Hottentots tee

Gewone bossies tee. . . .

172

Kouga bush tea

Naald tee

Swart tee

Koopmans tee <(

Rooibos tee |

L

C. intermedia E. Mev.

(=C. aurea Fourc.)

Langeberge, Kamnassie, Kouga, Langkloof Zuuranys

Mts.

(Riversdal, Oudtshoorn, Uniondale, Humansdorp Districts.)

Aspalathus tenuifolia D.C.

Piquetberg only.

Aspalathus contaminatus (Thb.) Druce

( —Borbonia pinifolia Marloth.)

Cedar Mts., Olifants Rivier Mts., Drakenstein Mts., Table

Mountain.

(Calvinia, Clanwilliam, Ceres, Worcester, Paarl, Stellenbosch,

Somerset West, Caledon, Malmesbury and Cape Town

Districts.)

CYCLOPIA INDEX

C. ashtonii Hofm. & Phill

C. aurea Fourc

C. aurescens sp. nov

C. aurescens var. glauca var. nov

C. bolusii Hofm. & Phill

C. bowieana Harv

C. brachypoda Hofm. & Phill

C. brachypoda var. intermedia Hofm. & Phill.

C. brachypoda Benth

C. burtonii Hofm. & Phill

C. buxifolia (Burm. f.) comb, nov

C. capensis Salter

C. cordifolia Benth

C. dregeana nom. nov

C. elata Sweet

C. falcata (Harv.) comb, nov

C. falcata var. ovata var. nov

C. filiformis sp. nov

C. galioides (Berg.) D.C

C. galioides E. Mey

C. galioides E. & Z

C. genistoides (L.) R. Br

C. genistoides var. linearifolia E. & Z

C. genistoides var. heterophylla Harv

C. genistoides var. ovalifolia var. nov

C. genistoides var. teretifolia E. & Z. comb. nov.

C. genistoides E. Mey

C. grandifolia Benth

C. grandiflora A. D.C

C. heterophylla E. & Z

C. intermedia E. Mey

C. laricina E. Mey

C. latifolia Benth

C. latifolia D.C

C. latifolia E. & Z

C. latifolia E. Mey

C. laxiflora Benth

C. longifolia Vog

C. maculata (Andr.) comb, nov

C. meyeriana Walp

C. montana Hofm. & Phill

C. montana var. glabra Hofm. & Phill

C. plicata sp. nov

C. pubescens E. & Z

C sessiliflora E. & Z

C. sessiliflora E. Mey

= bowieana Harv.. . .

= intermedia E. Mey

= ashtonii Hofm & Phill

= falcata (Harv.) comb. nov. (partly)

= sessiliflora E. & Z. (partly)

= intermedia E. Mey. (partly)

= sessiliflora E. & Z. (partly)

= sessiflora E. & Z

= latifolia D.C.

= buxifolia (Burm. f.) comb. nov.

= latifolia E. Mey

= galioides D.C

= vogelii var. falcata Harv

= genistoides R. Br

= galioides (Berg.) D.C.

= genistoides R. Br.

= genistoides (L.) R. Br

= subternata Vogel

= genistoides var. heterophylla Harv.

= maculata comb, nov

= buxifolia (Burm. f.) comb. nov.

= subternata Vogel

= dregeana nom. nov

= subternata Vogel

= tenuifolia Lehm.. .

= sessiliflora E. Mey

= vogelii var. brachypoda Harv.

= meyeriana Walp

Page

164

169

164

163

164

168

169

167

168

165

168

169

164

168

167

164

166

170

170 \

166

169

166

168

170

169

170

167

166

165

168

165

169

165

173

C. subternata Vog

C. tenuifolia Lehm

C. teretifolia E. & Z

C. ternata Sweet

C. triphylla Sweet

C. vogelii Harv

C. vogelii var. subternata Harv. . .

C. vogelii var. brachypoda Harv. .

C. vogelii var. intermedia Harv. . .

C. vogelii var. laxiflora Harv

C. vogelii var. falcata Harv

Galega genistoides Thb

Galega var. B

Genista buxifolia Burm

Gompholobium maculatum Andr.

Ibbetsonia genistoides Sims

Podalyria genistoides Willd

Sophora galioides Berg

Sophora genistoides Berg

Sophora genistoides Burm

Sophora genistoides Linn

Sophora genistoides Thb

170

= maculata Andr. comb, nov 166

= genistoides R. Br. var. teretifolia E. & Z. comb.

nov 166

=Aspalathus ferruginea

= Argyrolobium sericeum

= subternata Vogel (partly) 170

= subternata Vogel 170

= sessiliflora E. & Z 169

= intermedia E. Mey 169

= subternata Vogel 170

= falcata Harv. comb, nov 168

= C. genistoides R. Br 166

= C. galioides D.C 164

=C. buxifolia (Burm. f.) comb, nov 168

= C. maculata (Andr.) comb, nov 166

=C. genistoides (L.) R. Br 166

= C. galioides (partly) 164

= C. genistoides (partly) 166

==C. galioides D.C 164

= C. genistoides (L.) R. Br 166

= C. genistoides 166

1. C. bolusii Hofm. & Phill.

Bolus 1 1465, Swartberg Pass.

2. G. bowiena Harv.

Bowie, Herb. Kew.

3. (a) C. aurescens Kies

Primos, 33, Seweweekspoort.

(b) C. aurescens var. glauca Kies

Pappe (Harv. t. XXIV), Cape of

Good Hope.

4. C. galioides DC.

Bolus 67, Vishoek.

5. C. capensis Salter

Galpin 12694, Cape Peninsula.

6. C. pubescens E. & Z.

E. & Z. 1 146, Uitenhage.

7. C. meyeriana Bth.

Drege (1836).

8. (a) C. montana Hofm. & Phill.

Phillips 11131, Tulbagh.

(6) var. glabra

Marloth 2277, Matroosberg.

9. C. maculata (Andr.) Kies

E. & Z. 1150, Swellendam.

10. (a) C. genistoides R. Br.

Drege (1836).

( b ) var. heterophylla E. & Z.

E. & Z. 1148, Swellendam.

(c) var. ovalifolia var. nov.

Esterhuysen 2635, Caledon,

Somerset, Sneeukop.

175

)5c ?

11. C. burtonii

t Burton, F. D. H. 2914 (Type).

12. C. filiformis sp. nov.

MacOwan 336/A. K. Nov.,

Van Staadensrivier.

13. C. longifolia Vogel

Mundt & Maire, Hort.

Reg. Berol.

14. (a) C. plicata sp. nov.

Scrap in Herb. Marloth.

(b) C. plicata sp. nov.

Esterhuysen 6549,

Uniondale, Hoopsberg.

15. (a) C. falcata (Harv.) comb. nov.

Zeyer 354.

(6) Alexander Oct. 1846, Caledon.

(c) C. falcata var. ovata var. nov.

Esterhuysen 6167, Ceres, Slab

Peak.

16. C. buxifolia (Burm.) Kies

Bolus 8038, Table Mt.

17. C. intermedia E. Mey.

Fourcade 3049, Humansdorp.

18. C. dregeana nom. nov.

Drege (1836) (Type), Du Toit’s

Kloof.

19. C. sessiliflora E. & Z.

E. & Z. 1 147, Puspasvlei.

20. (a) C. subternata Vogel

Burchell 5519.

(b) C. subternata var. laxiflora

(Bth.) comb. nov.

E. & Z. 1149, Plettenberg Bay.

/

C^ctpp\<x ^>pe-c-ie«c^

176

4

6 d ?

> - 'g

tS

4 cK

||ll!

5j- C .Si tj S ,u ,• ® i Ji Ji

5-- 0-4>i3 •£ -0 ^ ^ 2. Q,

<j cj j o U cjoddc^ec

K> i- »'

£ & d

S | s 1 1

riij<

| s j ! f 1 1 1

111! lull

j j 3iS i £ £ £

. \ ti ri d d d o ii d

fcredLaador^.

177

Notes on Asparagus Africanus and Related Species.

By

P. Kies.

On examining the material at Kew Herbarium, which Baker had placed under

Asparagus africanus and its varieties, it appeared that several unrelated plants had

been placed together. Several of the specimens belong to other species, and some of

his varieties are worthy of specific rank. The following changes in nomenclature are

proposed.

1. A. africanus Lam. Encyc. I, 295 (1783); Asparagus dependens Thb. Prod. PI. Cap.

( 1 794) ; Asparagopsis juniperina Kunth, Enum. V, 74 (1850); Asparagopsis niveniana

Kunth l.c., Asparagopsis lamarckii Kunth l.c.

Lamarck’s description of this species states that the stem is woody, angled, leaves

setaceous, curved, 10-20 per fascicle, and that at the origin of each branch and under

each fascicle one finds an acute reddish spine, bent slightly backwards. The specimens

agreeing with this description are those which according to notes C. E. Moss had pre-

viously to this, compared with the types of A. dependens and A. juniperina.

Stem twining, angled and furrowed (in dried state) green, minutely pubescent.

Branches short, twining, angled. Spines 6 mm. long on main stem, reduced to 1 mm.

on branches, reddish in colour, patent or reflexed at an acute angle, curved or straight,

Cladodes fascicled, 5-20 per cluster of various lengths, usually 1 cm. long, spreading

straight or curved, slender, subterete. Flowers 3 or more per fascicle, peduncle jointed

in the lower third.

Cape Province. — Cape Division: Table Mountain, Dec. 1846, Prior ; Rondebosch,

Prior, Burehell 389. Paarl Division: French Hoek, Schlechter 9218. Swellendam

Division: Burehell 7221. Uitenhage Division: Drege 8576, Zeyher 178. Alexandria

Division : Addo, Burehell 4269. Karoo : Gill. Reg. Orient : Prior.

la. A. africanus Lam. var. brevifolia Kies var. nov. Typo similis sed cladodiis adscen-

dentibus differt .Cladodia brevia, 5-6 cm. longa, adscendentia. Caules erecti; rami

adscendentes. Squamae ramorum et ramulorum basi in calcaria pungentia deflexa

rigida productae.

Differs from the type in that the cladodes are ascending, and short, not curved,

spreading and long. Cladodes short, 5-6 cm. long, ascending. Stem erect, branches,

ascending; squamae below branches and branchlets produced into a rigid, sharp, slightly

deflexed spine.

Cape Province. — Kaffraria: Cooper 191.

Natal. — Umvoti Division: Greytown, Wylie \ Bothas Hill, Hutchinson 4703.

Alexandra Division: Flanagan 253. Inanda Division: Wood 989. Zululand: Eshowe,

Gerstner 2405.

V

2. A. multiflorus Bak. in Journ. Linn. Soc. XIV, 610 (1874); A. dependens Bak. (non

Thnb.) FI. Cap VI, 266 (1896); A. africanus var. pubescens Bak. in Journ. Linn. Soc.

XIV, 610 (1874); A. africanus var. dependens Bak. FI. Cap VI; 266 (1896).

The specimens which Baker cites under A. africanus var. B, are conspecific with

A. multiflorus Bak. This species differs from A. africanus in the stems being pubescent,

not glabrous, and in the deflexed branches and deflexed cladodes.

178

Branches scrambling, lax, pubescent, green. Spur reduced, not spiny under the

fascicles of cladodes. Branchlets deflexed, cladodes ascending or deflexed, 6-12 per

fascicle, stout, terete, usually straight. Flowers 8 or many per cluster, peduncles

jointed in lower third.

Cape Province. — Somerset Division: Bowker (Type in Kew Herb.). Beaufort

Division: Cooper 523. Albany Division: Cooper 3250. Kaffraria: Cooper 404.

Uitenhage Division: Zeyher 4158, 236; Enon, Thode A 2769.

3. A. concinnus ( Bak .) Kies stat. nov.; A. africanus var. concinnus Bak. in Journ.

Linn. Soc. XIV, 619 (1874); A. laricino affinis sed humilior.

Radices incrassatae, fusiformes, 1 cm. diametro. Rhizoma squamis stramineis

1 cm. longis vestita. Caulis fruticosus, erectus vel subscandens, 60-90 cm. longus;

rami sulcato-costulati; costulis minute albo-scabridis demum laevigatis nitido-griseis vel

rubris; ramuli breves, fasciculati vel solitarii, erecto-patentes. Squamae ad basin

fasciculorum membranaceae, acuminatae, basi in calcaria, pungentia, recurva, brunnea

productae. Cladodia 0-4-1 cm. longa, 15-70 fasciculata, viridia, teretia, apice mucronu-

lata, leviter curvata vel structa. Flores 1-7 fasciculati. Pedunculi 5-6 mm. longi,

basin versus articulati. Perianthium 0-3 mm. longum, segmentis recurvatis cremeis

dorso laete viridi-carinatis. Filamenta patentia, basin versus perianthio adnata ; antherae

aurantiacae demum rubrae. Ovarium globosum, viride; stylus brevis, trifidus. Bacca

globosa, rubra. Semina globosa, nigra.

Roots thick, fusiform, 1 cm. diam.; bud-scales on rhizome 1 cm. long, horny,

yellowish. Shrublet 60-90 cm. high, unbranched for 10 cm., with adpressed squamae.

Branches scrambling, white, striate, with minute retrorse hairs on striae, the older stems

becoming smooth, shiny, blue-grey or reddish, branchlets short, single or fascicled,

erect-spreading. Squamae thin, membranous with acuminate apex, basal spur sharp,

recurved, brownish, spiny under each fascicle of cladodes. Cladodes- 4-1 cm. long,

15-70 per clustre, arising in axils of spines, pale green, terete (triquetrous in dried

specimens), mucronate, straight or slightly curved. Flowers 1-7, usually 3 per fascicle.

Peduncle jointed near the base, 5-6 mm. long. Perianth 3 mm. long, lobes reflexed,

creamy, with pale green midline to each segment. Stamens spreading, filaments adnate

to perianth at base; anthers orange, becoming dark red. Ovary green, globose, style

short, stigmas 3, papillate at tips. Berries globose, red, seed globose, black.

This is related to A. laricinus, both having white striate stems. It differs from

A. laricinus, in the low habit, short internodes and short cladodes. It differs from

A. microphyl/us in having straight cladodes and spines under each fascicle of cladodes.

Baker cites these specimens under A. africanus, but the white striate stems would indicate

a close relation with A. laricinus rather than with A. africanus.

Cape Province. — Somerset East Division: Bruintjies Hoogte, Burchell 3020 (Type in

Kew Herb.). Colesberg Division: Shaw, Kuruman Division: Burchell 2443.

O.F.S. — Fauresmith Division: Veld Reserve, Verdoorn 1186, 955; Mogg 17085;

Voelfontein, Kies 339. Wepener Division: Acocks 11172.

Transvaal. — Lydenburg Division: Wilms 1551-.

Fig. 1.

4. A. microphyllus Burch, ex. Bak. FI. Cap. VI, 266 (1896); A. africanus var. y Bak.

l.c. A. microphyllus Burch, mss.

Stem stout, scrambling, shrubby, curved between two nodes, white striate, minutely

hirsute on striae. Spines on main stem 2-7 mm. long, hirsute, deflexed, none under

fascicles of cladodes. Branches lax, similar to main stem. Cladodia 4-20 per fascicle,

8 mm. long, curved terete, glabrous. Flowers 1-3 per fascicle, peduncle jointed in

lower third, cladodes up to 12 mm. long.

179

Differs from A. africanus in the white ribbed stem, lack of spines under the cladodes,

and the much curved cladodes. Differs from A. retr of r actus by stouter habit and

shorter cladodes.

S.W.A. — Great Namaqualand: Karasberg, Naruda, Pearson 7964. Kubib,

Pearson 9478; Hoffnungsfelde 9536.

Bechuanaland. — Kosi Fontein, Burchell 2572 (Type in Kew Herb).

5. A. compactus Salter in Journ. S.A. Bot. VI, 165 (1940); A. africanus var. wrightii

Bak. Journ. Linn. Soc. 14; 610 (1874). Differs from A. africanus in the ribbed white

stems and trigonous cladodes.

Cape Province. — Cape Division: Prior, Cape Town. Caledon Division: Zwarte-

berg, Schlechter 10364.

6. A. pearsonii Kies sp. nov. ; A. multifloro Bak. affinis sed ramis glabris laxioribus

differt.

Suffrutex erectus vel subscandens, 1 m. altus. Caules laxe ramosi, subscandentes,

teretes, longitudinaliter striati, scaberuli, virides, recti. Rami subhorizontales vel deflexi,

sparsi. Squamae ramorum basi in calcaria pungentia productae, ramulorum minores.

Cladodia fasciculata, pauca, 0 • 3-2 • 5 cm. longa in quoque fasciculo 2-6, teretia, recta

vel curvata, deflexa vel adscendentia. Flores pauci, 2-4-nati, pedunculi breves, 2-3 mm.

longi, medio articulati. Perianthium albidum segmentis patentibus brunneo-carinatis.

Filamenta perigonio i breviora. Ovarium 1 • 5 mm. longum, ovoideum. Stylus 0 ■ 5 mm.

longus, trifidus.

Bush or scrambler 60-100 cm. high. Stems laxly branched, stout, longitudinally

striate, slightly zigzag or straight. Branches few, horizontal or deflexed. Spines on

main stem soft, on branchlets reduced to minute blunt spurs. Cladodes few, 2-6 per

cluster, varying from 0-3-2 -5 cm. long in same cluster, terete, straight or slightly curved,

deflexed or ascending. Peduncle short, 2-3 mm. long, jointed near the middle. Flowers

few, 2- or 4-nate, white, 5 mm. diam.; perianth lobes spreading, keeled with a brownish

stripe. Stamens i shorter than the perianth. Style 5 cm. long, with 3 short branches.

Ovary 1-5 mm. long, ovoid.

This species is near A. multiflorus, but the stem is not pubescent and it differs from

A. africanus by the reduced spurs on the branchlets, and it also differs in the cladodes

being of various lengths in each fascicle.

S.W.A. — Great Namaqualand: Great Karasberg, Pearson 7965; Klein Karasberg,

Pearson 9722; Ortendahl 83; between Sabiesis and Holoog, Pearson 4601; Ramans

Drift, Pearson 4548, 4003 (Type in Kew Herb.); Bysonderheid, Schinz 30.

7. A. rivalis {Burch.) Kies sp. nov. ; A. rivalis Burch, nomen nodum ; A. asiaticus Bak. pro

parte, FI. Cap. 6: 265 (1896); A. africanus Bak. pro parte, l.c.

The description of A. asiaticus given by Linnaeus is of an erect spiny stem, with

filiform branches and setaceous fasciculate leaves. There is no specimen of this species

in the Linnaean Herbarium, and the description was derived from the figure in Pluk.

Am. 54 t. 15 f. 4 (Mant. 366), which shows a stiff shrubby habit, with small spines under

each fascicle of cladodes. This does not agree with the South African species described

by Baker under this name, which is a laxly branched climber, with spines only on the

main stem, and none under the cladodes. The A. asiaticus of Hooker is closely related

to the South African A. asiaticus of Baker, but differs in having the pedicels bent at

right angles at the joint, while in Baker’s species this is never seen. I suggest naming

the species A. rivalis, taking the name which Burchell used (without description) for

his specimen from Kosifontein No. 2587. Baker cites this name as a synonym

under A. asiaticus, and quotes Burchell’s specimen. Baker also places under A. africanus

specimens which belong here.

180

Caules fruticosi, erecti vel scandentes, 60-180 cm. alti. Rami subscandentes,

sparsi, horizontales, virides, glabri vel puberuli, striati, angulati vel teretes. Squamae

ramorum albae, amplexicaules, fimbriatae, basi in calcaria pungentia non indurata

flavescentia productae; ramulorum minores basi in tubercula minuta productae.

Cladodia 2-20, recta, 2-20 min. longa, internodiis longiora. Pedunculi 1-3 vel plures

fasciculati, 6-10 mm. longi, prope basin articulati. Perianthium 6 mm. diam., lobis

stellato patentibus. Stamina perigonio J breviora. Bacca globosa, perianthio basi sub-

persistente.

Climber or erect shrub, 60-180 cm. high with branches 1-5 cm. apart, lax, at right

angles to stem, twining or curving. Stem green, glabrous or minutely puberluous,

striate, angled or terete, slender. Squamae on main branches white-membranous with

basal spurs soft, straight, yellowish; on the branchlets the squamae clasp the base of the

cladodes, and the spurs are reduced to minute tubercles. Cladodes straight, varying

in length from 2-10-20 mm. usually exceeding the internodes, 2-20 cladodes per fasicicle,

usually 10. Pedicels 6-10 mm. long, articulating in lower-third, 1-3 or more per

fascicle, straight. Perianth 6 mm. diam. lobes stellately spreading. Stamens 2/3

length of perianth. Berry sphaerical, with subpersistent perianth at base.

Cape Province. — Albany Division: Grahamstown, Galpin 284. Graaff Reinet

Division: Bolus 132. Colesberg Division: Shaw; Aliwal North, Drege 4481b; Gerstner

160. Kimberley Division: Warrenton, Adams 802. Herbert Division: Douglas

Orpen 237.

Bechuanaland.— Mochudi, Rogers 6252; Gaberones, Oct. 1913 Rogers; Lobatsi,

Rogers 6041; Kosifontein, Burehell 2587 (Type in Kew Herb.); Maadji Mountain,

Burchell 2376; Mafeking, Burtt Davy 11019, 11087.

S.W.A. — Hereroland: Kalkfeld, Engler 6443.

O.F.S. — Bethlehem Division: Phillips 3152." Bloemfontein Division: Gemmell

4955. Senekal Division: Goossens 880.

Transvaal. — Shilouvane, Junod 617. Waterberg Division: Burtt Davy 7051.

Lydenburg Division: Thode A 1648. Pretoria Division: Repton 632, Mogg 14839.

Brits Division: Mogg 14623. Barberton Division: Galpin 622.

Natal. — Colenso Division: Wood 4089. No locality, Gerrard 70, 126.

181

Asparagus concinnus (Bak.) Kies ( Kies 339).

Fig. 1. — Twig, x 1;2. — Dry twig showing fascicled short shoots, x 1; 3. —Flower, x 2;

4. — Perianth lobe and stamen, x 3; 5. — Gynoccium, x 3; 6. — Rhizome and tuberous

roots, x 1.

183

New and Interesting Records of South

African Fungi.

By

P. H. B. Talbot.

Sometimes the author has the good fortune to collect, or to be sent, fungi which

are apparently not recorded for South Africa. It seems desirable to place such species

on published record and consequently this paper, the first of what is hoped will be a

series, is presented.

It is a pleasure to thank all those who have assisted by sending collections. The

author is also deeply indebted to the Director of the Commonwealth Mycological

Institute and members of his staff, for their help and courtesy in checking some of the

determinations. Without such collaboration it is often impossible to be sure of the

identity of a fungus suspected of being a new record for the country.

No attempt has been made to list all the literature concerned with each species,

nor to present a complete synonymy, since many of the most important papers are not

available. Reference is made to papers which were useful in naming the collections,

and in some of these there are lists of synonyms.

Unless otherwise stated, the numbers quoted below refer to collections in the

Mycological Herbarium of the Department of Agriculture, Pretoria.

PATELLARIA ATRATA (Hedw.) Fries, Syst. Myc. 2 (1823) 160; Nannfeldt, Studien

fiber die Morphologie und Systematik der nicht-Lichenisierten inoperculaten

Discomyceten (1932) 328-330, f. 47; Phillips, Brit. Discomycetes, 2 Ed. (1893) 366;

Rehm in Rabenh. Krypt.-Fl. 1, Abt. 3 (1896) 334 and 295, f. 1-6.

Lichen atratus Hedw., Spec. Muse. Frond. 2: p. 61, t. 21, f. A.

Lecanidion atratum (Hedw.) Rabenh.; Boudier, leones Mycologicae 3 (1905-

1910), PI. 564; Butler in Mycologia 31 (1939) 612-623, Figs. 1 and 2; Ibid in

Mycologia 32 (1940) Fig. 1 b., f. 2, f. 4a-d., descr. p. 799-803.

For synonyms see Rehm, l.c.

Illustrations : Nannfeldt, l.c.; Rehm, l.c.; Butler, l.c.; Boudier, l.c.; Text Fig. 1.

Collections : No. 39186, P. H. B. Talbot on dead wood, Fountains, Pretoria,

Nov. 1948.

Apothecia superficial on wood, adhering closely right up to the margin, waxy to

subcoriaceous, black, sessile, plane or slightly convex, with raised margin, 0-5-1 mm.

in diam. Asci cylindric-clavate, hyaline, with thickened apex, with a peculiar mode of

dehiscence (see below), 11-13 x 60-78 / 1 . Ascospores hyaline, fusoid to subclavate

(5-6*5)-8-9-6 X (30)-43-50 /i, with (3)— 8—1 1 septa. Paraphyses exceeding the asci

and forming an agglutinated layer at the apex, filamentous, septate, branching near

the apex into short branches, the terminal cells of which are inflated and encrusted with

blackish material.

There is apparently no published record of this species in South Africa, though

there is a specimen of it, which the author has not seen, in the S.A. Museum [S.A.

Mus. No. 33426, MacOwan (1058), on living bark of Acacia karroo, Boschberg]. The

species is interesting on account of its peculiar mode of ascus dehiscence. This feature

has been carefully studied by Butler, loc. cit., who summarises his investigations thus:

“ Lecanidion has been placed in the inoperculate Discomycetes, but has a method of

spore discharge unique for that group. The outer ascus wall breaks at, or near, the

184

apex and rolls back ; an inner membrane, here termed the endoascus, pushes up above

the epithecium nearly one-third the length of the ascus; the spores are then shot out

successively and forcefully from the apex of the projecting endoascus.”

PERROTIA FLAMMEA (Alb. and Schw.) Boudier in Bull. Soc. Myc. de Fr. 17 (1901)

24, leones Mycologicae Vol. 2, PI. 321 : Seaver, North Am. Cup-Fungi (1928) 154.

Peziza flammea Alb. and Schwein., Consp. Fung. (1805) 319, PL 1, f. 6.

Lachnella flammea (Alb. and Schw.) Fries, Summa veg. Scand. (1849) 365; Phillips,

Brit. Discomycetes 2 Ed. (1893) 407; Rehm in Rabenh. Krypt. FI. 1, Abt. 3

(1896) 858.

For synonymy see Seaver, loc. cit.

Illustrations'. Boudier, leones Myc. PI. 321; Text Fig. 2.

Collections'. No. 39187, N-. J. G. Smith, on dead twig of Acacia sp., Hennops

River, Pretoria distr., Sept., 1948.

Apothecia at first round, commonly up to 0-5 mm. diameter, later elongated

1-1-3 x 0-6-0 -8 mm., occurring superficially on wood, characterised by bearing a

dense covering of brick-red hairs over the whole outer surface, especially well developed

towards the apex of the apothecium, its base eventually becoming glabrous and black

in colour. The margin is strongly inrolled. Asci are cylindric, tapering at the base,

90 X 8 p, operculate, octosporous. Ascospores biseriate, hyaline, fusoid-elliptical,

eventually two-celled, straight or typically slightly curved, 3-2 x 14-4-15-3 p. Para-

physes simple, hyaline, septate, filamentous, exceeding the asci in length. External

hairs simple, fuscous at the base, paling quickly above to hyaline, covered externally

with minute brick red granules, septate, somewhat thick-walled, 3 -2-4 -8 p wide.

The maximum dimensions of the various organs cited by Seaver far exceed the

measurements recorded here. These are nevertheless a very close match with those

given by Rehm. There appears no doubt as to the identity of this collection. The

species has not previously been recorded for South Africa.

ATICHIA GLOMERULOSA (Ach.) Stein in Cohn Crypt. Flora ii, Part 2, p. 356;

Rabenhorst, Krypt. -FI. Band 1, Abt. 3 (1896) 500; von Hohnel in Ann. Jard. Bot.

Buitenzorg suppl. iii (1910) 19; Saccardo, Syll. Fung. 33 (1913) 769; Cotton in

Kew Bull. Misc. Inf. (1914) 54-63; Fraser in Proc. Linn. Soc. New S. Wales 61

(1936) 279.

For synonymy see Cotton, loc. cit., p. 62.

Illustrations'. Text Fig. 3.

Collections : No. 36995, G. A. Hepburn, associated with sooty moulds and scale

insects on the upper surface of Valencia Orange leaves, Martindale, C.P.,

Dec. 1948.

Colonies globular, black, gelatinous when moist, 200-600 p diam., obscurely lobed

or sometimes expanded into a stellate body up to 2 mm. diam. and composed of (4)-

5-6-(9) distinct rays attached lightly to the substratum by a brief stalk and gelatinous

disk. There is no ordinary mycelium, the thallus being composed of branched chains

of yeast-like cells held in a gelatinous matrix. The individual cells are hyaline, or

dilutely coloured at the periphery, thick-walled, 8-9-6 p diam., subglobose or elongated

ovoid. The conidia are aggregated into balls known as “ propagula ” borne in a

longitudinal, somewhat elliptical cleft on the upper surface of each ray in the stellate

forms, or in groups near the surface in the globular forms. The propagula are roughly

globose, 20-30 p diam., composed of subglobose to elliptical cells, each 5-8-7 p diam.,

with thick integuments, and dilutely coloured in a mass. Asci are found below the

upper surface of the thallus, not in an apothecium; their exact location on the thallus

could not be determined. The asci are pyriform, hyaline, with wall thickened as much

185

as 7-8 n at the apex, 23-28 x 40-44 p. Ascospores colourless, bicellular, deeply con-

stricted at the septum, 5-8-7 x 13-6-5-6 /x. No free ascospores were seen, so these

may not be mature in size.

This collection appears to fit well the description of A. glomerulosa given by Fraser

(1936, l.c., p. 279) and differs notably from other species whose descriptions have been

traced. The species usually occurs on conifers, and in the globular form, but Fraser

reports the stellate form and also records that the species is associated with Capnodium

citricolum in some of McAlpine’s collections. McAlpine’s writings relevant to these

collections have not been seen [Proc. Linn. Soc. N.S.W. 21 (1896) 469-497], but the

present writer assumes that the host-plant was Citrus.

The systematic position of the Atichiaceae is discussed by Cotton (1914) and by

Fisher [in Ann. Bot. n. ser. iii (1939) 399-426]. The other genus of the Atichiaceae,

Phycopsis, is known in South Africa from the species P. africana Syd. [in Ann. Myc.

24 (1926) 265], but is distinguished from Atichia by an irregular distribution of the

propagula (i.e., they are not restricted to definite areas on the thallus) and less importantly

by the complete absence of a stellate thallus.

GLONIOPSIS LEVANTICA Rehm in Hedwigia 25 (1886) 142; Saccardo, Syll. Fung.

9 (1891) 1118; Bisby in Trans. Brit. Myc. Soc. 25 (1941) 130-135, Fig. 1 D; Ibid.

27 (1944) 26.

Illustrations : Text Fig. 4.

Collections : No. 36812, P. H. B. Talbot , on Acacia mollissima. Lions River, Natal,

17/5/48; No. 36994, P. H. B. Talbot, on Clerodendron glabrum, Meintjies Kop,

Pretoria, July, 1948.

Hysterothecia erumpent or more or less superficial, solitary or crowded, usually

straight, but occasionally curved at the centre when long, dull black, not striate, car-

bonous, fusoid, 0-5-0 -7-1 -4 x 0-2-0 -3 mm., about 0-3 mm. high. Asci cylindrical or

somewhat clavate, with short blunt base, fanning out in mounts, surrounded and exceeded

by paraphyses, the asci (14- 4)— 1 7—2 1 —(25 - 5) X 90-120 p. Ascospores hyaline, when

young divided into 2-4 cells, later dictyosporous with (6)-8-9 cells longitudinally and

2-3(-4) cells in width, the cells highly refractile, distichous, occupying almost all the

ascus. The spores are more or less elliptical with broad rounded ends, and constricted

at about the centre into two parts, the upper of which is slightly larger; surrounded by

a thick gelatinous envelope, scarcely visible, 1 • 6-3 • 5 jx thick, which appears to be

divided transversely at the central constriction of the spore. Spores (including envelope)

are (9 - 6)— 1 0 - 7—1 2 - 8— ( 1 6) X 20-27-(34) p. In some specimens the envelope is dis-

solved, leaving a thin spore wall visible. Paraphyses surrounding and exceeding the

asci in length, branched, intertwined, hyaline or pale coloured in mass near the apices,

indistinctly septate, 1 -4 p or less in width.

The species has apparently not been recorded before in the Union. The author is

indebted to Dr. G. R. Bisby for identifying No. 36812 as this species. Dr. Bisby noted

that the spores of this specimen were perhaps a trifle large for the species, but that it is

known to be rather variable. The spores of No. 36994 agree well in size with those

of European specimens.

This species was once found in association with Septonema spilomeum (in collection

No. 36972), but this is probably fortuitous, since G. levantica is not known to produce

S. spilomeum as an imperfect stage.

GLONIUM PARVULUM (Gerard) Saccardo, Syll. Fung. 2 (1883) 735; Ellis and

Everhart, N. Am. Pyrenomycetes (1892) 682; Lohman in Pap. Mich. Acad. Sci.

Arts and Lett. 13 (1931) 141-157, Ibid. 17 (1933) 248; Bisby in Mycologia 24

(1932) 326.

Hysterium parvulum Gerard in Bull. Torr. Bot. Club. 5 (1874) 40.

186

Illustrations : Lohman, l.c., PI. 29, 30, 31; Text Fig. 5.

Collections : No. 36814, W. G. Rump, on dead wood, Pietermaritzburg district.

Hysterothecia dull black, straight, not striate, superficial, carbonous, very closely

aggregated in lines almost touching each other, 0- 4-1*1 X 0-2-0 -3 mm., with narrow

straight cleft. Asci cylindrical, narrow, sometimes bent, 3: 3-4 -4 X (41 -5)-47-54-57 p,

including a long indistinct basal “ tail.” Ascospores hyaline, 2-celled with deep median

constriction, fusoid, 2 -7-3 -3 X 6 -5-8 -7 p, the upper cell usually slightly larger. Para-

physes, hyaline, branched, septate, intertwined, filamentous, sometimes with a very

small globose apex, about the same length as the asci, about 1-1 p wide.

This specimen was kindly identified by Dr. G. R. Bisby, who wrote: “ Rather

immature, but I found some spores 6-8 X 2-5 p that were mature enough to escape

from the asci. The fungus agrees closely with Glonium parvulum (Gerard) Sacc. des-

cribed from North America and carefully redescribed by Lohman. It has been reported

with some doubt from Europe.”

HYSTERIUM INSIDENS Schweinitz, Syn. N. Am. 2078; Ellis and Everhart, N. Am.

Pyrenomycetes (1892) 696; Bisby in Mycologia 24 (1932) 314 and 326; Lohman

in Pap. Mich. Acad. Sci. 17 (1933) 262-266, f. 38.

Hysterographium insidens (Schw.) Saccardo, Syll. Fung. 2 (1883) 778.

Illustrations : Text Fig. 6.

Collections'. Nos. 36957, 36943, 36993, P. H. B. Talbot, all on Acacia mollissima,

Atholl Expt. Stn., E. Tvl., Feb. 1949.. .Locally abundant.

Hysterothecia black, carbonous, closely gregarious, often confluent along the

sides, erumpent on wood and bark, linear, straight or curved, or short and navicular,

0-4-2 -5 X 0-4-0 -5 mm. Striations parallel to the cleft are evident. Asci clavate,

octosporous, interspersed with branched, threadlike, hyaline paraphyses; the asci

120 x 14-5 p (mature or nearly so). Spores uniformly brown with (5)-6-(7) cross

septa, rounded at both ends, somewhat fusiform with the third cell usually larger than

the others, and usually a constriction next to it, 6 -5-8-9 x 22-7-28-3-29-5 p. In

one collection the spores were mostly 7-6 X 24 p; in another 8 X 24-27 p. On the

same piece of wood were larger hysterothecia with spores 9-6-12-8 X 32-35-(40) p.

Dr. G. R. Bisby kindly confirmed that “ No. 36993 agrees very well on morphology

with Hysterium insidens, as suggested.” Lohman (1933, l.c.) found that American

collections of this species produced Septonema spilomeum Berk, as a conidial state.

Dr. Bisby later undertook ascospore cultures of No. 36993, and found that the same

Septonema spilomeum state as produced in North America developed slowly on the

culture plates of this South African material. He generously sent me part of the dried

cultures, from which it was possible to establish the identity of some collections of

Septonema spilomeum occurring naturally on Acacia mollissima at Atholl Experiment

Station. A description of these collections follows.

SEPTONEMA SPILOMEUM Berk, in Hook. Lond. Jour. Bot. 4 (1845) 310, T. 12,

Fig. 5; Cooke, Handbk. of Brit. Fungi 2 (1871) 481 ; Saccardo, Syll. Fung. 4 (1886)

399; Lohman in Pap. Mich. Acad. Sci. 17 (1933) 264-266, Fig. 38.

Illustrations'. Lohman loc. cit; Text Fig. 7.

Collections'. Nos. 36971, 36972, 36973, P. H. B. Talbot, on dead wood and bark

of Acacia mollissima twigs, Atholl Experiment Station, E. Tvl., Feb. 1949.

Locally abundant.

Sporodochia deepseated, usually gregarious but discrete and seldom coalescing,

more or less hemispherical, 1—1-5 mm. diam., pulvinate, dark reddy-brown to almost

black, pulverulent. Conidiophores not readily distinguished from the lowest, least

developed conidia, but dilutely coloured, minutely roughened, about 3-2 p diam.

Conidia cylindrical or ellipsoidal, opaque, dark brown, in simple chains which occasion-

187

ally fork, at first transversely septate, usually with 4-6 transverse septa, slightly con-

stricted at each septum and more so at the central septum, with rough, granular surface,

8-11-12-8 X 16-35-43 p. The conidia are occasionally separated in the chain by a

small colourless cell. Mycelium of the sporodochium is scanty, hyaline, infrequently

septate, branched, filamentous, about 2 p diam.

Lohman (loc. cit.) emphasises that the conidia are progressively larger from the

base to the apex of each filament, that the conidial filament is seldom oppositely branched,

that the conidia are much constricted at the central septum, and that the terminal

conidia are frequently muriform. She also notes that the isthmian chains are characteris-

tic of the form genus Polydesmus, and the chains of muriform conidia of Sirodesmium.

To judge by Linder’s description and figures of Bonordeniella aspera [Linder in Mycolo-

gia 26 (1934) 439 and illustrations] it is possible that this is synonym of Septonema

spilomeum, but the types have not yet been compared.

Dr. G. R. Bisby kindly confirmed the identification of No. 36972. He stated:

“ This specimen is like the type and other material of Septonema spilomeum from

America — even to the colour. Kew has some fifteen specimens from the eastern

United States and one from Australia; none from Europe (at least under the above

name) where Hysterium insidens may be absent. I did not touch the type, which is

now nearly bare, but your fungus is microscopically the same as Ellis 353, which Lohman

says is the same as the type. It is easy to find spores 40 p long and up to 12-5 p wide

in Ellis 353. The problem of the generic name to use cannot finally be settled until

the types of Septonema, etc., are restudied. Meanwhile the binomial 5. spilomeum

can safely be applied to No. 36972.”

HELICOMA CURTISII Berk, in Grevillea 3 (1875) 106; Linder in Ann. Mo. Bot.

Gard. 16 (1929) 312 and 241, PI. 21, f. 1-10; PI. 30, f. 1 ; PI. 31, f. 2.

Illustrations : Linder, loc. cit; Text Fig. 8.

Collections : No. 36998, P. H. B. Talbot, on Eucalyptus, Hennops River, Pretoria

district, September 1948 (mixed collection); Nos. 36917, 36999, 37001, 37000,

P. H. B. Talbot, on bark of Eucalyptus sp., Buffelspoort, Rustenburg district,

12/8/1948.

Fungus forming black discrete patches on bark, hirsute in appearance, brownish

under the dissecting microscope. Conidiophores arising from the short dark mycelial

cells which are little developed, erect, unbranched, septate, deep fuscous, lighter at the

apex, which is rounded and tapered to a single sporogenous tooth, 136-178 x 7-8-5 p,

slightly flexuous. Conidia hyaline to very dilutely coloured, coiled lj times in one

plane, 4-8 septate, 4 -5-5 -3 p thick, base tapering-rounded and slightly recurved,

diam. of coil 13-17-5 p.

In the collection No. 36917 the Helicoma was found in association with Lasio-

sphaeria pezizula (B. and C.) Sacc. which Linder [in Ann. Mo. Bot. Gard. 16 (1929)

239-240] has shown to be its perfect stage. In both stages it is a new record for South

Africa.

LASIOPHAERIA PEZIZULA (B. and C.) Sacc. Syll. Fung. 2 (1883) 195; Linder in

Ann. Mo. Bot. Gard. 16 (1929) 314.

Sphaeria pezizula Berk, and Curt, in Grevillea 4 (1876) 106.

Herpotrichia pezizula (B. and C.) Ellis and Everhart, North American Pyreno-

mycetes (1892) 160.

Illustrations: Text Fig. 9.

Collections: No. 36917, P. H. B. Talbot, on bark of Eucalyptus sp., Buffelspoort,

Rustenburg Dist. 12/8/1948.

Perithecia small, smooth, black, about 300 p in diam., the walls composed of dark,

polygonal cells. Perithecium globose with papillate ostiole, drying to a concave disc,

188

closely gregarious but not confluent. Asci 100 150 x 20-24 p, broad clavate, 8-spored,

with marked hyaline thickening at the apex. Ascospores hyaline (becoming dilutely

coloured), 5— 7— (8) septate, straight, or more often slightly curved, almost cylindrical,

tapering towards the ends to a rounded or somewhat pointed end, 41-56 X 5 -5-8 -5 p.

The identity of this, and its imperfect stage Helicoma curtisii, was kindly checked

at the Commonwealth Mycological Institute by Mr. E. W. Mason.

CINTRACTIA LEUCODERMA (Berk.) P. Henn. in Hedwigia 34 (1895) 335; Hennings

in Fungi in H. Baum, Botanische Ergebnisse der Kunene-Sambesi Expedition,

Berlin (1903) 156. Whetzel and Kern in Mycologia 18 (1926) 119-120.

Ustilago leucoderma Berk, in Ann. Mag. Nat. Hist, ii, 9 (1852) 200; Saccardo,

Syll. Fung. 7 (1888) 460.

Illustrations : Whetzel and Kern in Mycologia 18 (1926) PI. 16 A.

Collections : No. 37002, P.G.C. Brett, on Rhynchospora cyperoides, S.A.S A.

Experiment Stn., Mount Edgecombe, Natal, April 1949.

Sori 2 -5-3 -5 cm. in length, surrounding the stem or peduncle of the grass, or

forming elongated patches on these but not fully surrounding them. The stem or

peduncle forms a central columella to the sorus. Sori are black inside, composed of a

hard mass of agglutinated, not pulverulent spores, and white on the outside due to a

thick white membrane which soon flakes off. The membrane is composed of fungous

cells of a degenerated, gelatinous appearance under the microscope. The spores are

deep brown with a thick browny-black, minutely warted epispore, and are subglobose

to angular by compression, 14-4-16 X 16-19-2 p.

This fungus has been reported on Rhynchospora aurea and R. dolichostyla from

localities which include parts of tropical and East Africa, but not yet, it would seem,

from South Africa itself, though it has been compiled in certain works dealing with

South African Ustilaginales. [See Verwoerd in Ann. Univ. Stellenbosch 4 (1926)

22, and Zundel in Bothalia 3 (1938) 316.]

TRICHODERMA VIRIDE Pers. ex Fr., Syst. Myc. 3 (1829) 215; Persoon, Syn. Meth.

Fung. (1801) 231 ; Bisby in Trans. Brit. Myc. Soc. 23 (1939) 149-168, Text Figs. 1-5.

Illustrations : Bisby (1939) loc. cit. ; see also references in Wakefield and Bisby,

Trans. Brit. Soc. 25 (1941) 60.; Text Fig. 10.

Collections : No. 36808, P. H. B. Talbot, on prune agar cultures isolated from

rotten potato leaves from Buffelspoort, Transvaal, 10/8/1948; No. 37003,

C. van der Merwe, on soil and soil cultures from Kroondal Expt. Stn., Rusten-

burg, August, 1949.

Vegetative hyphae hyaline, 3-4-9 /x wide, thick-walled, branched, giving rise to

the finer hyphal system from which the branched conidiophores arise. Conidiophores

not easily distinguishable from the vegetative hyphae, 1-7-3 -4 g wide, hyaline, septate,

branched, usually with opposite branches arising at right angles from a characteristically

swollen node of the parent hypha, the branches producing terminal and lateral phialides

on which the conidia are borne. Phialides shaped like a ninepin, straight or curved,

arising singly or in opposite pairs, or in whorls of seldom more than three, 6-5-9- 1 X

2 -5-3 -4 [x. Conidia abstricted from the phialides, unicellular, hyaline to pale bright

greenish, globose, ovoid or broad elliptical, formed in slimy balls at the apex of each

phialide, or occasionally in short chains. Conidia 3-4 /z diam., or 3 • 4 X 4 • 6 p, smooth-

walled. [Bisby (1939) mentions occasional inconspicuous roughenings of the spore

wall.] Chlamydospores terminal or intercalary, subglobose to broad elliptical, very

pale yellow, scanty, thick-walled, 8 -8-9 -6 X 11-2 g. Colonies on prune agar fluffy

white to blue-green to deep green, with a marked smell of coconut.

This fungus is a common contaminant of culture plates at the Division of Botany

and Plant Pathology, but as far as the author is aware has not been recorded or described

from South Africa under its valid name. Records refer only to T. lignorum [see Doidge

189

and van der Plank in S.A. Dept. Agric. and Forestry Sci. Bull. No. 162 (1936) 23],

which Bisby points out is invalid. No specimen of the fungus had previously been

preserved in the National Herbarium at Pretoria. The Transvaal Chamber of Mines

Timber Research Laboratory isolated this fungus from underground mine timber

( T.R.L . 249) and from the coat of a native miner ( T.R.L . 427). I am grateful to Dr. G. R.

Bisby for confirming my identification of specimen No. 36808.

NIGROSPORA ORYZAE (B. and Br.) Petch in Journ. Ind. Bot. Soc. 4 (1924) 24.

Monotospora oryzae B. and Br. in Journ. Linn. Soc. Bot. 14 (1873) 99.

Illustrations'. Mason, Annot. Account of Fungi Received at the Imp. Myc. Inst.,

list ii, fasc. 2 (1933) 53, f. 17. Compare also Mason in Trans. Brit. Myc. Soc.

12 (1927) PI. 15, f. 1-9; Molliard in Bull. Soc. Myc. de Fr. 18 (1902) PI. 9.;

Text Fig. 11.

Collections'. No. 35552, G. J. M. A. Gorter, on Zea mays (Anveld var.), Prinshof,

Pretoria, April 1946; No. 36992, G. J. M. A. Gorter, on Zea mays (inbred

Hickory King), Potchefstroom Agric. Coll., June 1949.

Mould found in cobs and on the base of maize seeds, forming a thin hyphal web

punctuated by dark purple-black spots which are the conidia. Hyphae creeping, sep-

tate, dilutely coloured to fuscous, up to 9 y diam. (fide Mason, 1933, l.c. p. 55, they are

10-18 y wide). Conidiophores inflated globose to ampulliform, borne laterally and

terminally on a branched, septate mycelium which may become dilutely coloured.

Conidia jet black, opaque, shiny, smooth, globose to broad elliptical depending on the

angle of view, i.e., they are depressed spheres (1 1 - 4)— 1 3 - 7— ( 14-8) y, shot off at maturity.

Mason (1927) showed that the genus Basisporium Molliard was a synonym of

Nigrospora Zimm. and that three species could be provisionally accepted on the basis

of spore measurements, but not on the basis of hosts and geographical distribution.

For the small-spored form on maize (Mason, 1927, p. 157-160) the name N. oryzae was

suggested, and is accepted here. A form with larger spores was designated N. sphaerica

(Sacc.) Mason. Mason (1933), p. 61) and Reddy [Agric. Expt. Stn. Iowa State Coll.

Research Bull. No. 167 (1933) 5] both showed that in culture the large spored form

tended to produce smaller spores. For that reason and the fact that both forms have

the same host range and distribution, and also that Mason’s studies did not include

pathogenicity tests, Reddy included both forms under the name Basisporium gallarum

Molliard. Durrell [in Agric. Expt. Stn. Iowa State Coll. Research Bull. No. 84 (1925)]

preceded Reddy in attributing the dry rot of corn in the United States to B. gallarum.

B. gallarum was first described by Molliard [in Bull. Soc. Myc. de Fr. 18 (1902) 167-170],

and is not to be confused with the soil isolation named B. gallarum by Dale [in Ann.

Myc. 10 (1912) 466] which was renamed Monotospora daleae Mason (Mason, 1933,

loc. cit. p. 50, PI. 16).

Mason’s studies (Mason, 1927, l.c.) clearly indicated that the fungus in question

is congeneric with Nigrospora Zimm. It appears desirable in recording this fungus

as new to South Africa to place it in the genus Nigrospora, and, as the small-spored form,

it is recorded as N. oryzae. If it should subsequently be established that N. sphaerica

and N. oryzae are cospecific no harm is done, for the latter epithet has priority and will

stand. ( Monotospora oryzae B. and Br., 1873; Trichosporium sphaericum Sacc. 1882.)

TORULA HERBARUM (Pers.) Link ex Fries; Fries, Syst. Myc. 3 (1829) 501; Link

in Magaz. Ges. Naturf. Freunde Berlin 3 (1809) 21 ; Saccardo, Syll. Fung. 4 (1886)

256; Massee, Brit. Fung. Flora 3 (1895) 362; Ferraris, Flora Italica Cryptogama,

Pars. 1 Fasc. 8 (1912) 224; Mason, Annot. Account of fungi received at the Imp.

Myc. Inst. List ii Fasc. 3 (special part) (1941) 113-117, Figs. 33, 34.

Monilia herbarum Pers., Syn. Meth. Fung. (1801) 693.

190

Septonema toruloides Berlese in Bull. Soc. Myc. de Fr. 8 (1892) 103, PI. 10, Figs.

18-20.

Illustrations : Berlese, loc. cit. ; Mason, loc. cit.; Text Fig. 12.

Collections'. No. 36929, P. H. B. Talbot, Hennops River, Pretoria district, Oct.

1948.

Pulvinate to effused, black with an olivaceous tinge, pulverulent. Repent hyphae

superficial or just within the substratum, fuscous, septate, branched, 1-5-3 g diam.,

giving rise to laterally borne conidiophores. Conidiophores short, inflated, subglobose,

smooth, thin-walled, 4-4-8 g diam., single, sessile or on a very brief stalk. Conidia

in chains which are typically much branched, fuscous, minutely spinose or smooth

under the oil immersion lens, each conidium composed of 3-6-(20 or more) subglobose

cells, with constrictions at the septa; width 5-6 g\ the apical cell of each conidium

commonly the lightest in colour and smaller than the others.

Mr. E. W. Mason very kindly determined this species, and commented: “ Almost

exactly the same as Wallace No. 1507 on Zea mays from Tanganyika. . . . Many of the

chains are formed exclusively of short conidia, 3-6 celled ; but some preparations show

conidia of 20 or more cells. The breadth of the conidia is mainly between 5 and 6 g

diam. Except for these long conidia it is almost exactly the same as a soil isolation

from Great Britain kept in the culture collection, but in that the longest conidium so far

seen consists of only 8 cells.”

LACELLINA GRAMINICOLA (B. and Br.) Petch in Ann. Roy. Bot. Gard. Peradeniya

9 (1924) 171; Hopkins in Mem. Dept. Agric. S. Rhodesia No. 2 (1939) 6.

Chaetopsis graminico/a Berk, and Br. in Journ. Linn. Soc. Bot. 14 (1875) 90.

Mesobotrys graminicola (B. and Br.) Sacc., Syll. Fung. 4 (1886) 325.

Gymnosporium cinctum B. and Br. in Journ. Linn. Soc. Bot. 14 (1875) 90,

Coniosporium cinctum (B. and Br.) Sacc., Syll. Fung. 4 (1886) 243.

Illustrations: Text Fig. 13

Collections : No. 36996, P. H. B. Talbot, on dead leaves of Miscanthidium junceum,

Rustenburg Kloof, Transvaal, 24/5/1949.

Colonies dense black, velutinous, superficial, elliptical or oval, discrete, rarely

coalescing, up to 1 x 2 mm. Setae present, numerous, unbranched, smooth-walled,

opaque, black-brown at the base becoming paler and pellucid near the rounded apex

and there revealing septa, 9-6-15 X 500-700 g. Conidia fuscous, subglobose to broad

elliptical or ovate (flattened in one plane) with rather thick walls, smooth when mature,

minutely sculptured when young, 4 • 6-6 • 2-(6 ■ 6) X (5-4)-6-2-6-9~(7-3) g. A few

outsize conidia 4-6 X 8 -5-9 -6 g are present. Conidia borne terminally and laterally

in evanescent chains and groups, easily becoming detached as a fuliginous powder.

Conidiophores sub-hyaline or pale golden brown, 3 -5-4 -8 g wide, thick-walled, with

strongly verrucose walls, sparingly branched, suberect, the apex often swollen or

expanded and appearing shortly moniliform owing to the presence of young catenate

conidia, arising from a sparse repent mycelium of the same appearance and all forming

a loose network surrounding the base of the setae.

This interesting species is a new record for South Africa, though it is known from

Rhodesia. The author was forunate in being permitted to borrow Dr. Hopkin’s

Rhodesian material to study (Myc. Herb. Dept. Agric. S.R. No. 2308, on Aristida sp.,

leg. /. C. Hopkins, Salisbury, 1/3/1933). The colonies in Dr. Hopkin’s specimen were

somewhat smaller (0-5-1 mm. diam.) and the setae less pointed at the apex, but in all

other particulars there was very close correspondence of characters and measurements.

To make sure of their identity these specimens ( Hopkins , 2308 and Talbot, 36996)

were sent to the Commonwealth Mycological Institute. Mr. M. B. Ellis very kindly

oompared them with authentic material of Lacellina graminicola ( Thwaites 505), and

191

stated that there seemed to be no doubt that they matched this species. Mr. Ellis

reported: “ . . . the conidia are flattened in one plane, smooth to verruculose, pale

brown 5-7 (5-7) p X 3-4 p (measured in lactic acid) and are borne on subhyaline or

pale golden-brown rough-walled conidiophores which are 3-4 p in diameter. The

setae are dark brown, 11-15 p in diameter at the base and 10 p about the middle; in

Thwaites 505 they vary in length from 300-1,200 p, in your material (Dr. Hopkin’s)

from 270-700 p plus (most of the setae have had their tips broken or eaten off). . . .

Although we have not been able to examine material of L. libyca Sacc. and Trott. it is

quite likely that this again is the same thing.”

Lacellina libyca Sacc. and Trott. [in Ann. Myc. 11 (1913) 419; Saccardo, Syll.

Fung. 25 (1931) 781] is obviously closely related and in fact the author first disposed his

collection as this species. As the above measurements will show, there is a good deal

of variation in the size of the conidia and setae which may indicate that the two species

intergrade.

Petch states that the conidia form groups and chains only accidentally after abscis-

sion. My observations support the view of Saccardo and Trotter that the conidia are

in evanescent chains or groups before abscission. The young conidia, still attached,

are clearly connected by isthmi, while if a colony is inverted over a slide and tapped

gently the conidia fall onto the slide still retaining their positions in groups and short

chains.

GONATOBOTRYS RAMOSA (Karst.) Reiss in Fresenius Beitr. ii 44 (1852) Tab. 5,

Fig. 22-23; Saccardo, Syll. Fung. 4 (1886) 169; Lindau in Rabenh. Krypt. FI.

Band. 1, Abt. 8 (1905) 357.

Glomerularia ramosa Karst, in Bot. Zeit. 7 (1849) 368, Tab. 6, Fig. V., p. (fide

Lindau).

Illustrations'. Text Fig. 14.

Collections'. G. J. M. A. Gorter, on germinating seeds of Zea mays in the laboratory.

Pretoria, 17/3/49.

Colonies white, composed of arachnoid or densely interwoven hyphae and conidio-

phores. Repent hyphae hyaline, sparsely branched, septate, 2 -6-5 -3 p diam. Coni-

diophores erect, very long, slender, very rarely branched, hyaline, septate, up to 660 p

long and 4-8 p wide, bearing always at the apex, and sometimes at the penultimate cell

as well, or irregularly distributed, a slightly swollen area covered with numerous spicules

on which the conidia are borne. The spicules are cylindrical with rounded apex and

up to 5 • 3 p long by 2 p wide. Occasionally in the swollen conidium-bearing areas are

short lateral vescicular branches of the conidiophore which also bear a few sporogenous

teeth. The conidia are ovate to elliptical and basally apiculate, hyaline, thin-walled,

large, (10-4)— 11 -2—12-8 x 17-6-22-4 p, smooth.

This interesting species was noticed by Mr. Gorter and subsequently sent by him

to the Centraalbureau voor Schimmelcultures at Baarn, for determination. Unfortu-

nately no specimen was preserved in the Herbarium in Pretoria, and further attempts to

isolate the fungus from maize seeds failed.

ACREMONIELLA ATRA (Corda) Saccardo, Syll. Fung. 4 (1886) 302; Lindau in

Rabenh. Krypt. FI. Band 1, Abt. 8 (1906) 675; Mason, Annot. Account of Fungi

received at the Imp. Myc. Inst., List ii, Fasc. 2 (Special Part) (1933) 29-39,

Fig. 11; Bainier in Bull. Soc. Myc. de Fr. 23 (1907) 237, PI. 18.

For synonymy see Mason, loc. cit.

Illustrations'. Text Fig. 15.

Collections'. E. E. Schaefer, cultures from seeds of Zea mays from Onderstepoort,

Pretoria, August, 1949.

1599—7

192

The aleuriospore stage of this fungus was found by the writer in cultures made

from maize seeds, in which there was a preponderance of other mould species. The

stage showing aspergilliform phialospores (see Papulaspora aspergilliformis illustrated

by Bainier in Bull. Soc. Myc. de Fr. 23 (1907) pi. 18] was not seen, and attempts to get

the fungus into pure culture failed. The aleuriospore stage is briefly described below.

Hyphae branched, hyaline, septate, whitish, 2-5-5 p diam. Aleuriophores formed

as lateral branches from the hyphae, hyaline, tapering to a point at the apex, occasionally

branched, septate, 44-53 X 5 -3-6 -6 p, bearing a single spore at the apex. Aleurio-

spores obovate or elliptical, light cinnamon brown, smooth, 21-3-28 x 13-16-20 p.

TRICHURUS GORGONIFER Bainer in Bull. Soc. Myc. de Fr. 23 (1907) 229-234,

PI. 25, Figs. 1-6; Saccardo; Syll. Fung. 22 (1913) 1452.

Illustrations'. Bainier l.c. ; Text Fig. 16.

Collections'. No. 39188, C. O'Connor, contaminant on peanut husk cultures,

March, 1949; No. 39189, E. E. Schaefer, contaminant on potato slice culture,

March, 1949; No. 39190, L. Kresf elder, contaminant on peanut husk culture,

August 1949; No. 39197, P. H. B. Talbot, contaminant on prune agar culture.

Sept. 1949. All these collections were made at the Division of Botany and

Plant Pathology, Pretoria.

The fungus arises as a web of whitish then pale cinereous or pale violaceous-

cinerous mycelium interspersed later with synnemata. The synnemata are 1 -2-2 mm.

in length, the width of the fertile head being 80-140 p and that of the sterile part of the

stalk being 40-48 p. The fertile head is cylindrical to clavate, and in gross appearance

is pale cinereous, violaceous-cinereous or later becoming ashen or pale brownish in

colour. The stalk is blackish brown. The synnemata are usually unbranched, but

frequently in some cultures they are branched. They arise singly or in groups direct

from the mycelium or from a thick mycelial strand or even from peculiar balls of hyphae

resembling sclerotia but soft, subglobose or irregular in shape, brown, 160-290 p in

diam. The stalk of the synnema is composed of parallel hyphae, closely united, brown,

septate, 2-4-3 p in width and irregular in width along short pieces of a single hypha.

Mycelial hyphae are similar, 1-6-4 -4 p wide. In the fertile part, the synnema hyphae

have terminal or lateral branches which bear very compact penicillate structures ter-

minating in phialides and chains of conidia. Such penicillate structures may also

originate from the mycelial hyphae before the formation of synnemata. The phialides

are dilutely coloured, 2 -7-3 -3-4 -3 x (5 - 5)— 7 - 6—10 - 8/z. The conidia are catenate,

easily becoming separated, ovoid or broad elliptical, very dilutely coloured, (2-7)-

3-3 x 4-4-(4-9) p. In a mass they are pale violaceous-cinereous, later ashen or pale

brownish. Arising as branches among the penicillate phialides are numerous long,

unbranched, septate, dilutely coloured, straight or undulate or curved (but not coiled)

hairs, varying in width from (3)— 5 - 5 p, with rounded apex, and 55-150 p in length.

The sterile hairs arising from the fertile part of the synnemata are characteristic

of the genus Trichurus, and distinguish it from Stysanus. Bainier (l.c.) has compared

three species of Trichurus, and our collections agree very well with his description of

T. gorgonifer. The fungus grows readily on a medium composed of peanut husks and

wheat grains, used for the culture of Phytophthora, where it was first noticed and brought

to my attention by Miss C. O’Connor.

PASSALORA GRAMINIS (Fuckel) v. Hohnel in Centralblatt fur Bakt. 60 (1924) 6;

Sampson and Western, Diseases of Brit. Grasses and Herbage Legumes (1941) 22,

text Fig. 6, PI. 3, Fig. 6.

Scolecotrichum graminis Fuckel Symb. Mycol. (1869) 107; Johnson and Hunger-

ford in Phytopath. 7 (1917) 69; Wakefield and Bisby in Trans. Brit. Myc. Soc.

25 (1941) 87.

Illustrations : Sampson and Western, loc. cit.; Text Fig. 17.

193

Collections : No. 39198, comm. S. J. Truter, on Dactylis glomerata, Riversdale

Farm, Himeville, Natal, January, 1950.

Fungus forming numerous very small chocolate-brown spots on both sides of the

leaves, the infected parts of the leaves crinkling and withering. The spots are com-

posed of very densely fasciculate conidiophores which are brown coloured, straight,

curved or slightly geniculate, unbranched, obscurely septate near the apex, (4-6)-6-6

X 70-80 fi, with a single apical scar. Conidia lightly coloured, 2-celled, the upper cell

frequently somewhat narrower and more elongated than the lower, both cells tapering

towards the ends, the basal one marked with a very distinct scar, smooth, borne apically

on the conidiophores, 10-11-3 x 36 g.

Sampson and Western (l.c.) figure conidiophores with more than one conidial

scar, and the assumption is that the conidia may be produced acropleurogenously,

though in the present material only acrogenous conidia were found in a mature state.

In some of the young conidiophores two lobes were seen at the apex, one representing

a very young conidium and the other the growing tip of the conidiophore. Thus there

appears to be little doubt that in our material more than one conidium per conidiophore

may be produced. Passalora graminis is a new record for South Africa.

194

*

T3

<U

w C

a $

li

<D

2d >>

°n &

u

Cfl

o

. si

xi

o £

CQ 2

o"3

C <U

.5 si

<u ~

> o

J5 a

o «

&o <u

T3 •£

C

**3

X) (U

<3

O co

Q S

O

ii

•S o

p

C/5

^ .

CL 03

T ^

I. Lh

<N O

o §-

E o

sn

ci

showing the granular incrustation.

195

FIG. 3.

Fig. 3. — Atichia glomerulosa (Ach.) Stein.

A — Stellate colonies. B — Globular colony. C — Asci and ascospores. D — Yeast-

like cells composing the thallus. E — Groups of conidia, i.e., “ propagula.”

196

-Ascospores, showing the thick investment of each spore.

Fig. 5. — Glonium parvulum (Gerard) Sacc.

Asci, ascospores and paraphyses.

197

•Hyphae of the sporodochia. B — Chain of conidia separated at B by a small

FIG. 8

198

Fig. 8. — Helicoma curtisii Berk.

Conidiophores and conidia.

Fig. 9. — Lasiosphaeria pezizu/a (B. and C.) Sacc.

Asci and ascospores. The habit of the fungus on bark is shown at A.

100

199

.sy "a

Wh o

X «

<U -g

Oh <D

•3 2

^ a

.g £

13

tx ‘2

O <u

cd

• 43

O CL

ffi

Uh

43

O

"5

Oh

ca

co

,&o

<

d

£

Hyphae, conidiophores and conidia.

200

Fig. 12. — Torula herbarum (Pers.) Link ex Fries.

Conidiophores and branched chains of conidia.

Fig. 13. — Lacellina graminicola (B. and Br.) Petch.

Conidiophores, conidia and setae. A shows the apex, and B the base of a seta.

201

Conidiophores and conidia.

202

Fig. 15. — Acremoniella atra (Corda) Sacc.

Aleuriospores and aleuriophores.

203

Fig. 16. — Trichurus gorgonifer Bainier.

A, B and C show parts of the synnemata including penicillate structures, phialides,

sterile hairs and conidia ; D shows penicillate structures arising direct from the mycelium ;

E — Hyphae.

204

-100 j* |

FIG. 17.

Fig. 17. — Passalora graminis (Fuckel) v. Hohnel.

Conidiophores and conidia.

205

Encephalartos Natalensis.

By

R. A. Dyer and I. C. Verdoorn.

Encephalartos natalensis Dyer and Verdoorn, sp. nov., ab E. altensteinii Lehm. inter alia

foliolis majoribus infimis spinosis reductis ab E. woodii Sander foliis haud valde

arcuatis truncis minus robustis differt.

Planta e base ramosa truncis usque ad 6-5 m. cylindricis plus minusve 40 cm. diam.

Folia stricta vel apicem versus aliquot recurva, 1-3-3 -2 m. longa, usque 37 cm. lata,

foliolis infimis spinosis reductis; pulvinus bruneo-lanatus, 3 cm. crassus, 4 cm. latus et

5 cm. longus; pedunculus glaber usque ad 26 cm. longus; foliola rigida, late lanceolata

vel basin versus asymmetrica, 16-23 cm. longa, 2 -5-4 -5 cm. lata, marginibus integris

vel utrinque 1-5 spinosis. Strobilus masculinus circiter 45 cm. longus, 9-5 cm. diam.;

squamae circiter 4 cm. longae. Strobilus femineus 48 cm. longus, 24 cm. diam.: facies

squamarum 3 cm. prominentes, rugosi, apices versus constricti et umbilicos truncatos

producti, apicibus plus minusve rhomboideis 1-4 cm. latis concavis dense lanatis.

Semina scarlatina, 5 cm. longa, 2 cm. diam.

Natal. — Valley of Thousand Hills, in kloof, old female plant, Dyer 4475 (type);

Verdoorn and Christian 713, 713a; Dyer and Verdoorn 2367; 2368; 2370; 2371 (all

from same plant as type specimen); young plant; Dyer 4476; Verdoorn and Christian ,

713b; Dyer and Verdoorn 2369; plant on ledge of krantz. Dyer 4477; near Inchanga,

Ducasse (male cone); Railway Nursery on lawn, Yerdoorn and Christian 713c.

Plates I, II and III.

The following specimens are probably conspecific but further research is required

in the native habitats. Common in Karkloof River Valley below Howick Falls; Bayer

S. 7; Dyer and Codd 4785; Krantzkop, overlooking Tugela River Valley, Verdoorn

and Christian 714; Crundall.

Plant branched from the base to form a clump of several stems of unequal length;

stems eventually attaining a height of about 6-5 m. (20-22 ft.) and 40 cm. (15 in.)

diameter, more or less cylindrical, covered with compact leaf bases, the scars of which

are about 6-5 x 5 cm. (2| x 2 in.). Leaves in a moderately dense crown, erect-spread-

ing, sometimes slightly twisted and recurved towards the apex, 1 -3-3-2 m. long, up to

about 40 cm. broad; pulvinus woolly-tomentose about 3x4 cm. thick and 5 cm. long;

peduncle tomentose at first, glabrescent, 10-26 cm. long up to the lowest prickle; leaflets

spreading or forming a broad V at the base and an acute V towards the apex, dark

grass-green, shining on surface, paler and not glossy below, with veins not apparent

on face but visible on back, glabrous, hard and rigid, the medial broadly lanceolate,

16-23 cm. long, 2 -5-4 -5 cm. broad, tapering to the hard pungent or subpungent some-

times decurved apex, very obtuse symmetrical or asymmetrical above the base, narrowed

to 1-1-2 cm. and with line of attachment 1-4-1 -8 cm. long; the leaflets towards the

apex reduced in size to about 10 cm. long, 1 -5 cm. broad, with the basal ones reduced

in length, but little in breadth until they become reduced to prickles; margin a hard

rim entire or with 1-5 sharp teeth on one or both margins, more often on lower leaflets

where the teeth may be slightly to distinctly lobate. Male cone (from plant in veld)

about 45 cm. long, 9-5 cm. diam.; scales (dry) from middle section of cone about 4 cm.

long with the sporangial surface narrowly obovate-oblong about 2-7 cm. long, 1-8 cm.

broad at top and 1 cm. broad at base, terminating in a flat or slightly excavate rhomboid

beak; shoulders of scale acutely ridged, entire, rarely bluntly toothed; beak with grey

or foxy pubescence with the rhomboid top 8x7 mm., the lower edge of which protrudes

slightly (though more markedly so in scales towards apex of cone) the upper edge

rounded with a small hump about the middle; the basal scales shorter and broader

and much smaller near the top of the peduncle. Female cone (from type specimen)

206

up to about 50 cm. long, 25 cm. diam.; scale faces (as seen on fresh cone) prominent,

rough with an elevated brown woolly rhomboid apical beak; scales from middle section

of cone (removed from cone) with a stripe about 4 cm. long, yellowish, side arms from

shoulders about 2 cm. long, with the shoulders 3-cornered, toothed on margins; scale

face 5 cm. broad at base and 4 cm. high, rough with prominent blunt rugosities, lateral

angles obtuse, lower and upper angles rounded or humped, constricted near apex and

produced into a short truncate slightly concave umbilicus covered with brown wool.

Seed scarlet, 5 cm. long, 2 cm. diam.

The description under a new name of such a well-known plant as the giant cycad

in the Valley of the Thousand Hills, near Inchanga, Natal, requires explanation. Until

now it has been classified under the species Encephalartos altensteinii, which has its

headquarters in the coastal area of the eastern Cape Province from the banks of the

Bushmans River to Kei River. Hutchinson and Rattray [in FI. Cap. 5; 2:40 (1933)]

took a broad view of the limits of E. altensteinii to include the arborescent forms in

Natal, and some in the Transvaal separated earlier by Stapf and Burtt Davy under the

name E. transvenosus. Hutchinson’s protected himself, however, by pointing out that

they had not seen specimens from either Natal or the Transvaal. It follows from our

present action that we do not hold with Hutchinson’s and Rattray’s view, although

admitting that the differences relied on for specific separation may not be very convincing

to others. In a footnote to E. altensteinii in Flora Capensis attention is drawn specially

to the fact that the lower leaflets are not reduced to prickles in E. altensteinii. One

of the features of the plant in the Valley of a Thousand Hills, near Inchanga, is the

almost invariable reduction of the lower leaflets to prickles. The leaflets of E. natalensis

on the average are appreciably broader, especially the lower ones, than in E. altensteinii

and the texture is more rigid and the colour a darker green. In addition the plants

may become somewhat larger and branching from the base is more of a feature in the

Natal species than of those in the Eastern Cape Province.

One further consideration which favours the separation of E. natalensis is in the

apparent break in the distribution range between E. altensteinii and E. natalensis,

neither being recorded, so far, from the Transkei area.

E. natalensis is readily distinguished from E. transvenosus in that the leaflets of the

latter are lighter green, more curved-spreading and have an indumentum on the surface

which persists for some considerable time. E. woodii, only known from one or two

male clumps in the wild state, differs in its more robust trunk (which broadens somewhat

at the base) and the long, regularly curving and later drooping leaves with larger and

generally more leathery leaflets.

Henderson in the Journal of South African Botany, 1945, accepts the Inchanga

plant as part of his E. altensteinii, but like Hutchinson had not seen material from the

wild in Natal. He separated, however, E. transvenosus from synonymy in Flora Capensis

and also maintained E. woodii as a distinct species.

Only three specimens, all female, of E. natalensis have been examined by the

authors in the Valley of the Thousand Hills. Part of an old male cone, brought by a

Native to the gardener of the Railway Nursery, and a male plant, on the lawn of the

Railway Nursery cottage, said to have come from the neighbourhood, were used for the

description of the male plant. Although there were probably more specimens in the

wild formerly it could not have been common in the immediate vicinity within living

memory. The three female plants studied in the wild near Inchanga are: (a) a large old

specimen with several stems in a steep, densely wooded kloof ; ( b ) a young unbranched

plant about 3 ft. high nearby; and (c) a medium-sized plant, branched from the base

(tallest trunk 9 ft.) situated on the mountainside below a krantz about half a mile distant

from the other two. It will be seen from the specimens cited that plants at Howick

Falls are regarded as probably conspecific and others in the Krantzkop area have been

placed with them, but all require further close study for several seasons in the field,

which it has not been possible to do as yet.

207

To give further details of ( a ) there were in November, 1949, not less than 11 well-

developed trunks in the clump and several suckers which had not risen above ground.

The tallest trunk was about 20-21 ft. above ground level and the diameter was near a

mean of 15 in., the other trunks being all fairly uniform in diameter. Viewing the

clump from the side with the upper aspect of the kloof on the left: on the upper side

there are 3 erect trunks varying from 10-14 ft. in height and two broken trunks, one

being 3 ft. long, hollow at the top and the other 5 ft. long with a branch about 6 ft. long

from its top spreading horizontally with upturned crown. On the lower side of the

main trunk were 5 subsidiary trunks, the longest 15 ft., spreading nearly horizontally

from near the base and with upturned crown, one trunk about 8 ft. tall and three shorter.

The stem of the young plant ( b ) nearby was about 3 ft. tall, up to about 18 in.

diam., with larger leaves and leaf bases and showing obvious signs of far more rapid

growth than the old plant.

Speculation on the age of living cycads, often referred to as living fossils, has been

indulged in by many people interested in botany. Estimates have varied very widely

but almost invariably, we feel, on the generous, even excessive side. So far no scientific

method of calculation has been devised, since the trunks are fibrous and exhibit no such

character as annual rings.

Sgme idea of the age of cycads might be judged from the number of leaf scars

on the stem, but so far it has not been determined with what regularity whorls of leaves

are produced. The available information shows that new whorls of leaves are not

necessarily produced annually, in fact this is likely to prove exceptional in adult plants.

One cannot judge the average rate of growth from observations made during a short

period. There is a comparatively wide gap at the apex of the stem between two suc-

cessive whorls of leaves, but the gap between the leaf bases of the whorls decreases as

further growth extends them to the maximum circumference of the trunk. Thus leaf

bases of two adjacent whorls may be 6 in. apart at the apex of the trunk but later in the

life of the trunk the same whorls may almost merge into each other.

Meagre records allow of an early growth rate in arborescent species, such as E.

natalensis of 2-3 ft. in the first 30-50 years, but this rate would not be maintained even

in unbranched specimens.

In the case of the old plant (a) the estimation of age becomes a guess at a minimum

because of the fact that it is branched from the base and a succession of suckers down

to embryonic size is present to continue growth when the older branches die of senile

decay or accident. The present tallest stem of about 20 ft. in height may well have

been a sucker far removed in succession from the original parent but we have no means

of telling. In addition to the normal retardation of growth rate with age, the rate of

growth of each of a group of stems from one root is almost certainly slower than a

single stemmed plant, and so to hazard a guess the age of the largest branch of the

matriarch is something upwards of 250 years and of the rootstock itself possibly more

than 1,000 years (in keeping with the name of the valley in which it grows). If this old

plant and the young plant (now 30-50 years old) and our botanical records persist for

another 1,000 years, posterity will have some factual evidence to fall back on.

In this connection it is most gratifying to record that Dr. Seele, the owner of the

ground surrounding the old and young female plants, has very generously and wisely

set it aside as a park in his newly surveyed township of Monteseel. In addition the

matriach type specimen of E. natalensis is to be declared a national monument.

It stands in a boulder strewn clearing surrounded by low forest. The vegetation

has regained some of its former luxuriance since its comparatively recent protection

from grazing animals.

1599— 8

208

Appendix.

After the above account had gone to press it was possible to make a further visit

to the three plants (a), ( b ) and (c) mentioned above. With the assistance of Professor

A. W. Bayer and Dr. C. R. C. Heard of the Natal University, measurements were taken

on 19/2/51. The height of the tallest stem of the old plant (a) was 19 ft. 9 in., which

shows that previous estimates were too high. Even now one cannot say from how far

underground the tallest stem arises, but a spot was selected at the base of the plant for

the present measurements and from which to take all future measurements.

Photographs were taken to fix the position of all existing branches — eleven in all,

excluding the basal suckers which have not risen above the leaf-mould. The total

circumference round the base of the clump is 26 ft. 1 1 ins.

PLANT (a) (11 STEMS).

Height. Circumference.

(breast high).

(1) 9 ft. 10 ins. 3 ft. 10 1 ins.

(2) 9 ft. Ill ins 4 ft. 0 ins.

(3) 7 ft. 1 1 ins. 3 ft. 5- ins.

(4) 19 ft. 9 ins 3 ft. 9 ins.

(5) 15 ft. 4 ins 3 ft. 9 ins.

(6) 8 ft. 21 ins 3 ft. 9| ins.

(7) 10 ft. 0 ins 4 ft. 4 ins.

(8) 8 ft. 5 ins 3 ft. 7 ins.

(9) 7 ft. 8i ins. 4 ft. 01 ins.

( 10) 5 ft. 6 ins 4 ft. Oi ins.

(11) /4 ft. 4 ins. (base) 3 ft. 101 ins.

\6 ft. 0 ins. (branch) 3 ft. 6 ins.

PL ANT (b) ( YOUNG UNBRANCHED FEMALE).

Height. Circumference.

3 ft. 7 ins. (above firm humus) 4 ft. 6 ins.

3 ft. 10 ins. (to corner of adjacent rock)

This young plant is obviously making rapid growth by comparison with the branches

of the old female (a).

PLANT (c) ( SEVERAL STEMS).

Height (Tallest stem on north side). Circumference.

1 1 ft. 5 ins. total :

8 ft. 3 ins. (above rock ledge) 3 ft. Ill ins.

3 ft. 2 ins. (below rock ledge).

The circumference of the whole plant at base was 22 ft.

209

Plate I. — Encephalartos natalensis. Old female plant ( a ) in text, showing the

group of erect stems and the large one on the right spreading down hill (1945).

1599

210

E3

Plate II —Encephalartos natalensis. Top, old female (a) in text showing relative

size of whole clump of stems (1945); bottom, male cone taken from plant on

lawn near railway nursery cottage, Inchanga.

[Photo H. King.

211

Plate III. — Encephalartos natalensis. Top left, cone from medium sized trunk of

old female plant, (a) in text, showing wide space between consecutive whorls

of leaves while the portion of old trunk on its left shows how congested leaf

bases become with age (1945); top right, cones from young female plant,

(b) in text (1947); bottom, young female plant (, b ) in text (1949).

213

Newly Described Species

By

E. A. Bruce, A. Brueckner, R. A. Dyer, P. Kies and j. C. Yerdoorn.

ASCLEPIADACEAE.

Ceropegia decidua E. A. Bruce, sp. nov. (Asclepiadaceae); affinis C. caffroro Schltr., sed

foliis deciduis, tubo corollae intus basi tuberculato, corona exteriora cupuliforme

columnam staminum multo superantia valde distincta.

Herba volubilis, usque 40 cm. longa, basi ramosa; ramis minutissime scabridulis

basi 1 -5 mm. diam., internodiis 1-4 cm. longis inferioribus brevioribus; tuber depresso-

sphaericum, circiter 3 cm. diam. Folia decidua, oblongo-ovata, vel ovata. patentia, car-

nosiuscula, apice rotundata et apiculata vel abrupte et breviter acuminata, basi rotundata,

1 1 -8 cm. longa, 0-4 1 cm. lata, margine et subtus basin versus pilis paucis albidis

ornata; petiolus circiter 3 mm. Iongus. Cymae numerosae, 2-4 florae, breviter pedun-

culatae, e nodis lateraliter ortae, floribus succedaneis; pedunculus 4 10 mm. Iongus;

pedicelli 3-4 mm. longi; bracteae minutae, triangulares, vix 1 mm. longae. Calyx ad

basin lobatus; lobi anguste lineari-lanceolati, 2-2-5 mm. longi, glabri. Corolla erecta,

2- 2 2-5 cm. longa; tubus 1-6 1-9 cm. Iongus, basi 5 7 mm. ovoideo-infiatus, circiter

3- 5 mm. diam., deinde 10 11 mm. cylindricus, 1-1 mm. diam., fauce vix ampliatus,

extra glaber pallide viride-purpureus, intus basi tuberculatus, supra levis et parce pilosus;

lobi 5 6 mm. longi, subspathulati, apice connati, basin versus leviter replicati, apice

abrupte ampliati, margine pilis fuscis ornati. Corona exterior cupuliformis, circiter

1-5 mm. alta, lobis 5 latissime triangularibus 0-5 mm. longis basi 1 mm. latis apice

subacutis vel emarginatis glabris columna staminum longioribus; lobi coronae interioris

circiter 2 mm. longi, erecti, vel apicem versus leviter recurvo-patuli, spathulati, leviter

compressi, glabri, columnam staminum multo superantes. Folliculi non visa.

Transvaal. — Naboomspruit district, April, Montgomery s.n. (flower); Naboom-

spruit district, April, Phillips s.n. (flower); 3 miles south of Nylstroom on road to

Warmbaths, Jan., and flowered in Pretoria, Aug., Eric Codd 47 (type): 22 miles north-

east of Nylstroom on road to Naboomspruit, Jan., and flowered in Pretoria, Aug.,

Eric Codd 53 (flower).

Plate 1.

Ceropegia decidua was first discovered in April, 1947, by Mr. J. A. L. Montgomery,

a keen collector of Ceropegia, who thought the plant might prove to be an undescribed

species. The following year Mr. E. A. Phillips collected more material and a year later

still further material was added by Eric Codd. All these plants were gathered in the

Nylstroom-Naboomspruit area, though Mr. Phillips has since discovered the species

1 7 miles N.E. of Pretoria and three miles W. of Hammanskraal on the road to Zoutpan.

The plant grows in open bush country in sandy soil, under the shade of trees, very often

species of Acacia. An unusual characteristic of the species is that the slightly fleshy

leaves are shed in early winter (May-June) before the flowers emerge, hence the specific

epithet decidua. Mr. Phillips has made some observations on this point and states

that the flowers appear occasionally in winter (July) but August seems to be its flowering

period, going on till about October, the leaves appearing after August. Some of the

specimens he has found have been erect and branching from the base, though he has

also found twining specimens. It is possible that if no support is available the plant

adapts itself to an erect habit. The inconspicuous grey-green flowers, unlike most

other species of Ceropegia , are more prolific at the base of the stem, where they form

small erect clusters, opening successively from ground level along the whole length

of the stem.

In the shape of the corolla C. decidua is most closely allied to C. caffrorum Schltr.,

though the basal inflation in C. decidua is more ellipsoid and minutely tuberculate

within, and the flowers are grey-green in colour not dark red-brown, as described by

214

Medley Wood in C. caffrorum. Other points of difference are the deciduous leaves

and the deeper cup-like outer-corona, which far overtops the staminal column.

It is interesting to note that the enlarged apical portion of the corolla-lobe in

C. decidua varies in shape, in some cases being transversely elliptic, the five apices

forming a small canopy, whereas in other cases it is longer than broad, not, or scarcely

forming, a canopy.

BURSERACEAE.

Commiphora neglecta Verdoorn, sp. nov. (Burseraceae); inter species austro-africana

trifoliolatas et spinescentes distincta, a C. calcicola Engl, foliis glabris majoribus

a C. betschuanica drupis globosis haud acuminatis recedit.

Arbuscula vel arbor 8 m. alta nonnunquam suffrutescens. Caules virides vel

grisei, squamis membranaceis solubili. Ramuli divaricati, extimi spinescentes, novelli

cinereo-pubescentes. Folia viridia, trifoliolata; petiolus sparse pilosus, 0-5-4 -5 cm.

longus; foliola glabra, minute crenato-serrata (nonnunquam indistincte); medium

late obovatum, 1-6 cm. longum et 0-7-3 -5 cm. latum, apici breve acuminatum vel

rotundatum, basi cuneatum, lateralia orbiculare-elliptica, 0-9-4 cm. longa, 0-5-2 -5 cm.

lata, apice basique breve acuminata. Flores unisexuales vel hermaphroditi, pedicel-

lati, subfasciculati ramulis valde abbreviatis orti, hermaphroditi et feminei singulares,

masculi in cymos dichasia evoluti. Calycis tubus 1 mm. longus; lobi 4, 1 mm. longi.

Petala 4, libera, plus minus 3-5 mm. longa, 1-5 mm. lata. Stamina 8, 4 plus minus

3 mm. longa, 4 plus minus 1-5-2 mm. longa. Ovarium ovoideum, in stylum crassum

acuminatum; stigma globosa. Drupa laevis, rubrescens, subglobosa vel leviter com-

pressa, 1 cm. diam., minute apiculata, pericarpio bivalvi; pyrena laevis, nigra, minute

apiculata.

Transvaal. — Waterberg, 21 miles N. of Nylstroom, Codd 4821, Verdoorn 2362;

3 miles N. of Nylstroom, Smuts and Gillett 3340; 10 miles N. of Ellisras Codd 1013;

east of Oslo, Codd 4429, 3995, 4430; Kruger National Park, near Punda Maria, Lang

in Transvaal Museum Herbarium 31083; 15 miles E. of Skukuza Codd and Verdoorn

5498 (Type fruiting specimen); Codd 4396; Codd and de Winter 5052; 9 miles SE.

of Pretorius Kop, Codd and de Winter 5166.

Mozambique. — Hornby 2836, Polana Beach, Gerstner 6658.

Natal. — Gerstner 5025; False Bay, Hlabisa, Gerstner 6767; Mahlabatini, Gerstner

2222; Melmoth, Acocks 13004; Hluhluwe, Mangusi Forest, Forestry Herb. No. 5299;

near Umgeni Dam, Bayer 729 and 730.

Plate 2.

Small tree, sometimes up to 25 ft. tall, occasionally shrubby. Stem green (in some

localities green at base only, grey above) bark peeling in thin membranous flakes.

Branches divaricate the ultimate abbreviated and spine-tipped, at first ashy pubescent.

Leaves' spinach green, trifoliolate, varying in size; petiole sparsely pilose 0-5-4 -5 cm.

long; leaflets glabrous or with a few scattered hairs obscurely to distinctly, but finely,

crenate-serrate on the margins, lateral veins about 4; the terminal leaflets broadly

obovate, shortly apiculate or rounded at the apex, cuneate at the base, 1-6 cm. long

and 0-7-3 -5 cm. broad, the lateral sub-orbicular elliptic shortly acuminate at base

and apex, 0 • 9-4 cm. long and 0 • 5-2 • 5 cm. broad. Flowers unisexual or hermaphrodite

pedicelled and sub-fascicled on abbreviated shoots, the hermaphrodite and female

singly and the male in stalked dichasial cymes. Calyx tube 1 mm. long, lobes 4, 1 mm.

long. Petals 4, free about 3 • 5 mm. long, 1 • 5 mm. broad. Stamens 8, four about 3 mm.

long and four about 1 -5-2 mm. long (in female flowers aborted). Ovary ovoid narrow-

ing into a short thick style and globose stigma (in male flowers aborted). Drupe smooth,

green, turning scarlet when ripe, sub-globose or slightly compressed, about 1 cm. diam.,

minutely apiculate, 2-valved (and easily breaking into the two segments when ripe),

seed smooth black with a minute white apicule, aril scarlet (bleaching with age), with

4 subterete claws running from the base to two-thirds up the seed.

215

The species described here differs in the particular combination of characters

rather than in any outstanding characteristics. This may account for its not having

been described before, in spite of its wide distribution in well-collected areas. In all

the material seen so far the leaves are constantly trifoliolate and the branchlets spinescent.

In these respects, among the South African species, it agrees with C. calciico/a and

C. betschuanica and can be distinguished as follows: —

Leaflets pubescent; small shrubs C. calciicola.

Leaflets glabrous or rarely sparsely pubescent; trees sometimes shrubs:

Leaflets deeply crenate-serrate; drupes ovoid, acuminate C. betschuanica.

Leaflets not deeply crenate serrate usually obscurely so; drupes globose C. neglecta.

From the description of C. rehmanni Engl, it differs in that the leaves are always

trifoliolate and not “ trifoliolate and pinnately 5 foliolate on the same branch.” Also

the leaflets of our plant cannot be described as having the petiole and nerves of the

leaflets grey pilose. The description does not mention whether the branchlets are

spinescent or not.

The type of C. rehmanni which should be in the Berlin-Dahlem Herbarium was

collected at “ Klippan, Boschveld Transvaal.” To date the locality has not been

exactly established, and no specimen found that corresponds with the description.

CHENOPODI A CEAE.

Salsola humifusa Brueckner, sp. nov. (Chenopodiaceae); ab omnibus speciebus habitu

humifuso, ramis lateralibus longis gracilibus prostratis stoloniferis differt.

Suffrutex parvus, humifusus, 15-40 cm. diam., praerosus, ramis lateralibus longis

gracilibus prostratis stoloniferis. Rami seniores glabri, ramulis et pullulis erectis pilosis.

Folia bractiformia, griseo-viridia, opposita, carnosa, ovata, parva, c. 1-5 mm. longa,

glabra vel parce pubescentia, supra concava, apice carnosa, marginibus hyalinis ciliatis,

base auriculato amplexicaule breve calcarato. Flores axillares, sessiles, 2-bracteolati ;

bracteolae late ovatae, saepe inaequilateres, c. 1 - 8 mm. longae, 1 -5 mm. latae, carnosae,

leviter canaliculatae, marginibus late hyalinis ciliatis. Septula 5, lanceolata, c. 3 mm.

longa, 1 mm. lata, membranacea, ciliata, basi leviter carnosa, in fructu base indurantia

gibbosa, pilosa. Stamina 5, filamentis planis; antherae apice attenuatae. Ovarium

globosum, apicem versus in stylum bifidum attenuatum ; discus patelliformis, 5-lobulatus.

Cape Province. — Phillipstown Division: Henrici 3927, Brueckner 683, Kimberley

Division: Acocks and Hafstrom 1282.

O.F.S. — Fauresmith Division: near Koffiefontein, Rabie 15 a, C. A. Smith 4514

{type in National Herbarium); near Petrusburg, Rabie 14 a..

Fig. 1 and 2.

A procumbent shrublet forming mats from 15 to 40 cm. in diameter, often cropped

down, but with long, slender prostrate side branches, rooting at intervals. Older

branches glabrous with erect, hoarily-pubescent branchlets and shoots. Leaves bracteo-

late, greyish-green, opposite, fleshy, ovate, small, about 1 -5 mm. long, concave ventrally

glabrous or very sparsely pubescent, apex fleshy, margin hyaline and ciliate, base auricled,

amplexicaul, shortly spurred. Flowers, sessile, axillary, with two ensheathing bracteoles.

Bracteoles opposite, broadly ovate, often asymmetric, about 1 -8 mm. long and 1 -5 mm.

broad, fleshy, slightly keeled, with a wide hyaline, ciliate margin. Sepals 5, lanceolate,

about 3 mm. long and 1 mm. wide, membranous but slightly fleshy basally, becoming

indurated and lobed basaliy after fertilisation, with loose pubescence marginally and on

thickened portions. Stamens 5, filaments flat, anthers attenuated. Ovary globose,

narrowing into the style which is bifid at the apex; disc patelliform, 5-lobed.

The prostrate, matted habit of this Salsola is distinctive, and it was isolated as a

separate species in 1928 by Mr. C. A. Smith under the unpublished name Salsola humi-

fusa. Growing on the dry margins of pans and vleis, it is often well eaten down by

game and small stock, and may be inconspicuous. Characteristic long lateral branches

are produced from a long vertical taproot, and root at intervals along their length on

1599—9

216

Fig. 1. — Salsola humifusa Brueckner.

Photo of type sheet, Smith 4514.

217

Fig. 2. — Sa/so/a humifusa Brueckner.

1 — Tip of branchlet showing opposite flowers. 2 — Bract. 3 — Bracteole (dorsal

view). 4— -Bracteole (ventral view). 5 — Sepal. 6 — Sepal of fruiting perianth (dorsal

view), showing thickened indurated basal portion. 7 — Stamen. 8 — Gynoecium.

218

the soil surface or underground. The small greyish-green leaves are opposite and

decussate, or subopposite, on short erect branchlets. Small flowers are axillary, the

comparatively long sepals being conspicuous beyond the fleshy bracteoles which envelope

the base. The indurated basal lobes of the fruiting perianth would appear to be homo-

logous with the wings of the fruiting perianth of other Salsola species.

The vernacular names of this plant are Springbokganna and Kruipganna.

Salsola rabieana Verdoorn, sp. nov. (Chenopodiaceae); afflnis S. glabrescenti et S.

tuberculatae sed ab ilia plantibus humilioribus induratioribusque bracteis bracteo-

lisque appresse sericeis sepalis plus minus pubescentibus, ab hac plantis altis non

compactis bracteis bracteolisque haud in nodulis compressis differt.

Suffrutex calciphilus, circa 50 cm. altus; ramuli intricati, lignosi, rigidi, indurati.

Folia bractiformia, alterna, carnosa, 2 mm. longa, 1 -5 mm. lata, dorso appresse sericea,

apice carnosa, apiculata, marginibus hyalinis. Bracteolae 2 mm. longae, 3 mm. latae,

dorso appresse sericeae, apice carnosae, apiculatae, marginibus late hyalinis. Sepala

5, ovata, c. 3 mm. longa, 1-1 -5 mm. lata, centro triangulare pubescente viride carnoso,

in fructu ala late horizontali scariosa obcuneata circiter 3 mm. longa, 2-5-5 mm. lata

infra medium laborum instructa. Stamina 5. Ovarium depresse globosum 1-5 mm.

longum 2-5 mm. diam. in stylum bifidum attenuatum; discus patelliformis marginibus

hyalinis obscure lobatis.

Orange Free State/ — Fauresmith district: near Koffiefontein, W. A. Rabie 17a,

Driefontein, Henrici 3885, 3885a (type), 3887, 3579, 3884, Kies 289; Luckhoff, Farm

Rosemarie, Henrici 3826, 3889, 3889a, 3990b; Roodepoortjie, Henrici 3049; Kalabas-

drift, Henrici 2865; east of Jagersfontein, C. A. Smith 4388; Jacobsdal, H. G. Schweick-

erdt 1129; Bloemfontein district: Henrici 2738, 2740.

Cape Province. — Hopetown division: Potfontein, C. A. Smith 2815; Herbert

division, 10 miles west of Belmont straight along road to Douglas, H. G. Schweickerdt

1129, Clievedon, Brueckner 657. Kimberley division: Kamfersdam, Brueckner 93,

64. Hay division: Heuningkrantz, O. L. Cooke in McGregor Museum, no number,

Paauwfontein, O. L. Cooke, no number.

Transvaal. — Ventersdorp district: Raatsiehay, /. B. Pole Evans 3132.

Fig. 3.

Subshrub about 50 cm. high found in lime soils; branches intricate woody, rigid,

and hard. Leaves bractlike, alternate, fleshy, 2 mm. long and 1-5 mm. wide, dorsally

appressedly sericeus, apex with a fleshy apicule, margins with a hyaline border. Flowers

sessile axillary bi-bracteolate ; bracts very like the leaves but wider and with wider

winglike margins, about 2 mm. long and 3 mm. wide. Sepals 5, ovate, about 3 mm.

long, three about 1-5 mm. wide and two about 1 mm. wide, with central, triangular,

green, pubescent, fleshy areas which in fruit bear below the middle, horizontal, scarious

obcuneate wings about 3 mm. long, of which three are 5 mm. wide and two 2-5 mm.

wide. Stamens 5. Ovary depressed globose 1 -5 mm. long, 2-5 mm. diam. narrowing

above into a bifid style; disc patelliform with a hyaline border which is obscurely lobed.

This species, which is reported to grow on limestone formation, is close to S.

glabrescens and S. tuberculata and seems to fall between these two species. From

S. glabrescens it is distinguished by being a smaller bush with shorter and more intricate

branches and in its greyish appearance due to the bracts and bracteoles being appres-

sedly grey sericeus. S. tuberculata, while resembling our species in the plants being

greyish owing to a similar pubescence, differs in being a lower and more compact bush

with the bracts and bracteoles congested into nodules “ like small brussels sprout ”

as described by Mr. C. A. Smith.

Many years ago Mr. Smith recognised this as a distinct species and gave it the manu-

script name of S. rabieana in honour of the late Mr. Rabie of Fauresmith, who had a

good knowledge of the veld plants in that area and their vernacular names. The

common names by which this species is known are “ bloupanganna ” and “ hardeganna.”

219

Fig. 3. — Sa/sola rabieana Verdoorn.

1 — Branch of type specimen. 2 — Bracteal leaf, dorsal and front view x 5.

3 — Bracteole dorsal and front view x 5. 4 — Mature flower viewed from above, show-

ing the dorsal wings on the sepals x 3. 5 — Sepal and dorsal wing side view x 6.

6 — Stamen x 5. 7 — Disc x 5. 8 — Ovary, stvle and stigma x 12. 9 — Spiral embryo

X 15.

220

COMPOSITAE.

Pegolettia retrofracta (Thb.) P. Kies, comb. nov. (Compositae).

Eupatorium retrofractum Thb. FI. Cap. (1823).

Pegolettia po/yga/aefo/ia Less. Syn. Comp. 200 (1832).

Vernonia polygalaefolia Licht. ex. Less. Syn. Comp. 200 (1832), non Less. 1831.

Iphiona polygalifolia Benth. and Hook. Gen. II (1863)..

CYCADACEAE.

Encephalartos humilis Verdoorn sp. nov., affinis E. lanato Stapf & Burtt Davy et E.

laevifolio Stapf & Burtt Davy sed ab ambitu inter alia plantis minoribus diflfert.

Planta humilis. Caudex in toto circa 35 cm. longus, 13 cm. diam., c. 12 cm.

epigeus, lanuginosus, soboliferus. Folia recurvata, arete lanuginosa, glabrescentia,

30-55 cm. longa; pinnae anguste lineares, integrae, apice pungentes, mediae 9-13 cm.

longae, 4-6 mm. latae, plus minus 9-nervatae. Strobi/us masculinus usque 15 cm.

longus et 4 cm. diam.; squamae mediae 1 -5 cm. longae et 2 cm. latae, facies dense lanu-

ginosae. Strobi/us femineus ignotus.

Transvaal. — Schagen Hill, Dyer 4806 (with male cone) type, Verdoorn 2349

and 2350, Liebenberg 335; Schoeman’s Kloof J. C. Smuts 285; Krantz Nurseries,

Christian 568 (with male cone); Verdoorn 2348; Rosehaugh, Reynolds 3936 (with male

cone), Mogg 17363; Majuba in Berlin Forest Reserve, Verdoorn 2351.

Plate 3.

A low plant more or less hidden in the grass, suckering freely to form small clumps,

usually wedged among low rocks. Stem rising about 12 cm. above ground, in the wild

state (often higher in cultivation, one observed to be 45 cm. tall), in all about 35 cm.

long (30-50 cm. long), 13 cm. diam., usually narrowing to a cone-like apex, thinly lanate.

Leaves recurved with leaflets forming a V, glabrescent with small patches of wool per-

sisting, 30-35 cm. long; pinnae narrowly linear, entire, with a pungent apicule; the

median 9-13 cm. long and 4-6 mm. wide, more or less 9-nerved, nerves prominent

beneath. Male cone up to 15 cm. long and 4 cm. diam., median scales about 1 -5 cm.

long and 2 cm. wide; scale faces densely lanate and not very prominent. Female cone

not seen.

The specimens quoted under E. lanatus Stapf and Burtt Davy in the Flora Capensis

all have entire, narrowly linear leaflets and come from the Transvaal. Under this

species the monographers of the genus, Hutchinson and Rattray, sank E. laevifolius

Stapf and Burtt Davy. Recent field observations have shown that not only are these

two species distinct but that a third and undescribed species is involved. Although

closely related, they are nevertheless as distinct from each other, as are some of the

species in other groups of the genus.

Field observations on the three species are summarized as follows:

E. lanatus Stapf and Burtt-Davy. The type comes from Toevlugt (the Native

name of which is Botshabelo), Middelburg district, Transvaal, and a large number of

plants were examined there. They were found on the average to have stems about

1| to 4 ft. tall (rarely up to 5 ft. tall); to have the leaves retain the lanate covering on the

rhachis for quite a long time and the cones, male and female, to be thickly and persistently

lanate. The same species was seen on the Wilge River and on General Sir Pierre van

Ryneveld’s farm 50 miles east of Pretoria. All these places are in what is termed the

highveld of the Transvaal.

E. laevifolius Stapf and Burtt-Davy. This species occurs on some of the rocky

ridges of the high mountainous country around Barberton (which is in the lowveld).

The plants are usually taller than E. lanatus with stems 5 ft. to 1 1 ft. tall, and the leaves

longer ; they are soon glabrescent and have a slight glaucous look and the cones, especially

the female, become glaucous glabrous except for a small, short lanate spot in the depres-

sed centre of the scale face.

221

E. humilis n.sp. This newly described species occurs on some of the grass-

covered rocky hills below the mountain summits in the lowveld area. It is a small

plant, often practically hidden by grass, and the stem rises only a few inches above the

ground. In this respect it differs from both the preceding. Another difference is in

the male cone. Three male cones each from a plant in a different locality, were seen

and they were very much smaller than those of either of the other species, being only

up to 15 cm. long and 2 cm. diam. as against about 24-30 cm. long and 4-7 cm. diam.

The scale faces in our species are flatter than in either of the others and are a little less

thickly lanate than those of E. lanatus and much more so than in those of E. laevifolius.

In connection with the size of the cones it should be noted that since the description

of E. lanatus in the Flora Capensis is based on specimens of all three species, the measure-

ments given for the cones should not be taken as being those of typical E. lanatus.

They are in fact smaller than those of either E. lanatus or E. laevifolius as observed in

the field. This is accounted for in one instance by the presence on a herbarium sheet

of an immature female cone of true E. lanatus, and for the rest the measurements of

the male cones were probably taken from specimens of E. humilis. For instance the

quoted specimen Wilms 1355, from “ Crococile River near Piet Schoeman ” (Flora

Transvaal, Vol. 1, p. 99, and FI. Cap., Vol. 5, 2, p. 43) must be E. humilis according

to the locality and in the Flora of the Transvaal the specimens mentioned in the follow-

ing quoted note under E. laevifolius are most probably all E. humilis: “ Davy 32d, col-

lected on the lower slopes of Spitzkop, Lydenburg, 4,000 ft. alt., on the side near Sib-

thorpe’s (several in a clump, all young, among grass by the roadside), and a plant

collected by Sim at Rosehaugh, Lydenburg, may belong here.”

The late Mr. H. Basil Christian was the first to suspect that there was a third species

in this group. Among his very good collection of living cycads were some plants

which he had got from the Kranz Nurseries in Schoeman’s Kloof (plants said to have

come from the neighbourhood) and he noted that the stems were more cone-shaped

and the leaf scars smaller than any of his other species. The leaves, though resembling

those of E. lanatus remained smaller over the long period that these plants were observed

growing next to plants of E. lanatus and E. laevifolius. In connection with the leaves

he also observed that they did not arise “ like a bunch of asparagus ” in the centre

of the stem apex but appeared in close succession and the petiole was much twisted.

This may be a pathological condition, but it has been noticed since on several plants

of this species. Then a male cone appeared and it was like a miniature of the other two

species. These observations led to the investigations which resulted in the description

of this species.

EUPHORBIACEAE.

Euphorbia clivicola R. A. Dyer, sp. nov. (Euphorbiaceae); affinis E. schinzii Pax ramis

brevioribus congestis flavo-viridibus aculeis minoribus differt.

P/anta succulenta, perennis, armata. Radix tuberosa caudice continuata, plus

minusve 15 cm. longa et 2-3 cm. crassa, attenuata, radicibus tenuibus paucis instructa.

Rami plurimi, congesti, breves vel brevissimi, 2-6 cm. longi, 1 -5 cm. crassi, indistincte

4-angulati, tuberculati apicem versus angustiores, angulis tuberculis prominentibus,

podariis corneis haud confluentibus 2-aculeatis. Aculei circiter 5 cm. longi. Cyma

solitaria tribus cyatheis sessilis vel perbreviter pedunculata. Cvathium primum mas-

culinum; 2 bisexualia, lateralia, subsessilia. Involucrum 3-5 mm. diam., glabrum, lobis

5 parvis fimbriatis et 5 glandulis transverse oblongis integris 1-5 mm. latis contiguis

flavis munitum. Ovarium glabrum, sessile; styli 2 mm. longi, fere infra medium con-

nati, apice bifidi.

Transvaal. — Potgietersrust district: about 20 miles north of Potgietersrust on

Farm Lunsklip, quartzite ridge, Plowes in National Herbarium, Pretoria, No. 28386

(type), with photographs. Pietersburg district: near Pietersburg, Kirsten, photographs.

Plate 4.

222

Dwarf spiny perennial succulent with the main stem and root merging into each

other and forming an underground tuberous body about 15 cm. long and 2-3 cm.

thick, tapering to the base and with a few slender secondary roots. Stem repeatedly

branched and retracting further underground as plant enlarges, only the young ultimate

branches appearing above ground. Branches yellowish-green, congested into a dense

mass, about 2-3 cm. long but up to 6 cm. if the plant is shaded by grass, 1 • 5 cm. thick

towards base, narrower towards apex, being reduced at each successive tubercle, ob-

scurely 4-angled by the more or less decussate arrangement of the tubercles and their

paired spines; angles often with spine-pairs opposite but occasionally irregularly dis-

posed, usually armed with spines on the apical 2-4 tubercles; tubercles with the upper

margin truncate, sloping gradually below the spines; spine-shields extending to the leaf

scar above the spines and 1-2-5 mm. decurrent; spines up to above 5 mm. long, in

pairs at the apex of the tubercles, grey, and with minute prickles by the side of the leaf-

scar. Leaves rudimentary, soon deciduous. Inflorescence: cymes solitary from the

apical flowering eyes of the branchlets, sessile, usually consisting of 3 cyathia, the

central one male with two bisexual cyathia laterally disposed on very short peduncles

or sessile; involucre cup-shaped, 3-5 mm. in diam., glabrous, with 5 glands and 5 sub-

quadrate fringed lobes; glands bright yellow, transversely oblong, 1-5 mm. in their

greater diameter, entire. Ovary sessile, included in involucre, glabrous; styles about

2 mm. long, united into a column in the basal third, spreading above, with bifid tips.

There seems little doubt that E. clivicola is an evolutionary form from Euphorbia

schinzii- like stock. The present concept of E. schinzii is so broad that our plant could

be squeezed in also, but this would appear to be making an unsatisfactory position

worse. It is felt that some of the many forms now under E. schinzii could conveniently

be given definite taxonomic status, either with varietal or even specific rank.

The main characters distinguishing E. clivicola from E. schinzii are: the branches

are yellowish-green, broadest below the middle, furnished with few (2-4) spine-pairs,

congested into cushion-shaped masses rising only slightly above ground level and

forming a compact plant with a single tuberous main root.

It occurs between Potgietersrust and Pietersburg in the northern Transvaal, on

quartzite slopes, in company with Euphorbia clavarioides Boiss., Aloe pretoriensis Pole

Evans and Cara/luma lutea N.E. Br. The specific epithet refers to the hillslope habitat.

Mr. Darrel Plowes, the collector, added that the “ soil ” was sour, being derived from

the white quartzite rubble and consisted mostly of rock chips.

Euphorbia confinalis R. A. Dyer sp. nov., affinis E. triangulari Desf. et E. excelsae W. D.

& S., ab ilia habitu trunco uno erecto robusto ramis minus constrictis angulis

4 minus compressis, ab hac cymis 1-3, 3-cyatheis differt.

Arbor succulenta usque ad 8 m. alta, trunco cylindrico erecto, ramis numerosis

patenti-ascendentibus. Rami usque 1-1-5 m. longi, 2-5-6 cm. crassi, nonnunquam

ramosa, 4- vel 3-5-angulati, segmentis 5-20 cm. longis constricti, podariis corneis

2-aculeatis bruneis haud vel rariter confluentibus, aculeis abortivis vel ad 8 mm. longis.

Cymae 1-3, plus minusve 2-5 mm. supra aculeos dispositae, 3-cyatheis. Pedunculus

2 mm. longus, ramis 3 mm. longis. Cyatheum primum masculinum; cyathea lateralia

2, bisexualia; involucrum 6-8 mm. diametro, glabrum, lobis parvis obovatis fimbriatis

et glandulis 5 transverse elliptico-oblongis circiter 3-5 mm. latis contiguis integris

flavis munitum. Ovarium obtuse triangulare, stipitatum; styli 1-5-3 mm. longi infra

medium vel medio connati. Capsula plus minusve 8 mm. lata, pedicello circiter 5 mm.

rare usque 8 mm. longo exserta.

Transvaal. — Zoutpansberg District: Kruger National Park, 6-7 miles SW. of

Punda Maria, Cocld 3026; 5366; 10-11 miles SW. of Punda Maria, Dyer, photo;

Codd, photo; Pilgrims Rest district, Kruger National Park, 2 miles E. of The Gorge

Camp, Codd and de Winter 5580 (type); Barberton district, Komatipoort, Dyer^ 4799.

Portuguese E. Africa. — Near Ressano Garcia, van der Merwe 1698; E. 17;

ridges between Moamba and Ressano Garcia, Dyer 4799.

Plate 5.

223

Tree 15-25 ft. but occasionally up to 30 ft. tall, with a straight unbranched trunk

or occasionally with 1-2 trunk-like branches, with a crown of curved ascending branches

which wither with age and fall, leaving the naked trunk. Branches 3-4 ft. long, 4- or

occasionally 3-5-angled, constricted at intervals of 5-20 cm. with more or less parallel

sides, 2-5-5 cm. between adjacent angles, usually about 3 cm. on flowering segments;

angles moderately compressed, about 5 mm. thick near the margin and projecting

3-3-5 cm. from the centre, furnished with paired spines or their rudiments. Spine-

pairs 1-2 cm. apart, 0-5-8 mm. long, those on young plants conspicuous, becoming

obsolete on old trees, on a horny base which is discontinuous or sometimes continuous

along the angles. Cymes 1-3 together at the flowering eyes, 2-5 mm. above the spine

pairs; each cyme consisting of 3 cyathia, a central male and 2 lateral bisexual cyathia

arranged in a plane parallel to the main axis. Peduncle about 2 mm. long, branches

3 mm. long with bracts 2-5 mm. long and as broad, shortly ciliate on the margin,

keeled on back. Involucre 6-8 mm. diam. with 5 glands and 5 obovate deeply fimbriate

lobes; glands greenish-yellow, transversely somewhat oblong, 3-4 mm. in their greater

diam. Ovary 3-angled and raised on a pedicel; and subtended by a small triangular

calyx. Styles 1 -5-3 mm. long, divided up to or below half-way with bifid tips. Capsule

exserted from the involucre on a gynophore about 5 mm. or occasionally up to 8 mm.

long.

Information about this plant has accumulated slowly since it was first brought to

the notice of the writer in 1936 by Dr. F. Z. van der Merwe. It was referred to in a

note under E. triangularis Desf. in the Succulent Euphorbieae 2; 897 (1941). The

weight of evidence is now strongly in favour of according it separate specific rank.

In habit it resembles E. excelsa W. D. & S. more than E. triangularis Desf. but

differs in the cymes which in E. excelsa are single from the flowering eyes and each has

more cyathia. Whereas E. triangularis usually has a few to several trunk-like branches,

E. confinalis more often than not has a straight unbranched trunk crowned by the

curved ascending angled branches. The branches of mature trees are often practically

spineless and have more solid, or at least less winglike angles and larger involucres than

in E. triangularis.

E. confinalis occurs commonly on rocky hills on the western border of the Kruger

National Park near Punda Maria and extends at intervals southwards on to the Lebombo

Mts. and occurs on both sides of the border between the Transvaal and Portuguese

East Africa at Komatipoort and Ressano Garcia. This association with boundaries

suggested the specific epithet, the choice of which in such a large genus as Euphorbia

being decidedly limited as regards appropriate names.

It may be added here that the distribution of E. triangularis into Portuguese East

Africa seems to be proved by the record of a large colony of fairly typical plants not far

from the Swaziland border in the catchment area of the Umbeluzi River on the road to

Goba. A specimen collected by Mr. L. C. C. Liebenberg and recorded as rare at the

bridge near the Royal Sheba mine near Barberton, and another by Thorncroft, extends

the distribution into the Transvaal as well.

Euphorbia keithii R. A. Dyer , sp. nov. (Euphorbiaceae); affinis E. zoutpansbergensi

R. A. Dyer, ramis ramosis 3-6-angulatis stylis et ovulis differt.

Frutex vel arbor 2-6 m. alta, succulenta, spinosa, trunco teretiusculo ramis numero-

sis. Rami plus minusve patenti-ascendentes, ramosi, segmentis usque 25 cm. longis

3-4 cm. crassis constricti, 3-6- plerumque 5-angulati, inter angulos sulcati; anguli

compressi, podariis corneis griseis confluentibus aculeis 2, 5-8 mm. longis. Cymae

3, rariter 1-2, supra aculeos emittentae, subsessiles, 3 cyathia; pedunculus bibracteatus.

Cyathium primum masculinum deciduum, cyathia bisexualia breviter pedunculata.

Involucrum 4-5 mm. diam., glabrum, 5 lobis subquadratis fimbriatis et glandulis trans-

verse oblongis 2-2-5 mm. latis integris contiguis viridi-flavis munitum. Ovarium

stipitatum, glabrum; styli 1-5 mm. longi in columnam ad medium versus connati,

apice bifidi. Capsula stipite usque 6 mm. longo, 6-7 mm. lata, obtuse trilobata; semina

suborbiculata.

224

Swaziland. — On western edge of Ubombo Mts., cultivated, near Stegi, D. R.

Keith in Nat. Herb. Pretoria, 28421, 28422, 28423 (type), 28424; about 8 m. SE. of

Stegi on east-facing krantz, Dyer, photographs.

Plates 6, 7 and 8.

Shrub or small tree 6-20 ft. high with the older branches eventually withering and

falling from the stems. Branches dark green, spreading ascending, 1-2 m. long, 3-6-

usually 5 — angled, constricted at intervals into segments; segments varying in length

according to habitat conditions up to about 25 cm. long and 3-4 cm. broad with more

or less parallel sides or somewhat narrowed to the apex; angles winglike, about 0-7-

1 -5 cm. deep and 1-52 cm. between the ridges of adjacent angles, only slightly tubercu-

late; spine shields united into a continuous horny margin along the margin of the angles

and with the flowering eyes about 5 mm. above the spine pairs; spine pairs 1-1 -75 cm.

distant, 5-8 mm. long, moderately stout. Leaves rudimentary, about 5 mm. long,

ovate-cordate, spreading recurved, soon deciduous. Inflorescence with 3 cymes or

occasionally only 1-2 developing from each of the flowering eyes. Cymes subsessile,

each cyme consisting of 3 subsessile cyathia vertically disposed, i.e., one central male

cyathium and one bisexual cyathium on either side arranged in a plane parallel to the

main axis; bracts small, scale-like; involucre cup-shaped, 4-5 mm. in diam., glabrous

with 5 glands and 5 subquadrate fimbriate lobes; glands contiguous, yellowish-green,

transversely oblong, 2-2-5 mm. in their greatest diam. Ovary exserted on a curved

gynophore about 5 mm. long, moderately 3-angled, about 1 -5 mm. broad; styles about

1 -5 mm. long, united into a column for half their length, with bifid stigmas. Capsule

on a curved gynophore about 6 mm. long, reddish-green, obtusely 3-angled, 6-7 mm.

broad, flattened, i.e., broader than long, with ovules filling the cells, subglobose, smooth.

Capt. D. R. Keith, of Ravelston, near Stegi, first brought this species to my notice

about 1942. He had one young specimen in his garden, from which complete flowering

and fruiting material was eventually obtained for description in 1949-50. Efforts to

establish the distribution of the species indicate that it does not cover a very wide range.

It may well be abundant locally on krantzes on the western side of the Lebombo Moun-

tains, though not necessarily west-facing, since those plants photographed in their

natural habitat in 1947 were mainly on an east-facing aspect.

E. keithii has a general likeness to such arborescent species as E. sekukuniensis

R. A. Dyer, but the disposition of the cyathia of the cymes is distinctive. In this essen-

tial character it is seen to approach nearest to E. zoutpansbergensis among South African

species, but even to that species the relationship is not particularly close. Nearer

relatives are likely to be found on the mountains between Southern Rhodesia and

Mocambique.

Euphorbia restricta R. A. Dyer, sp. nov. (Euphorbiaceae); affinis E. enormi N.E. Br •

ramis luteo-viridibus immaculatis podariis corneis confluentibus plerumque 4-6-

angulatis differt.

Planta humilis, succulenta, perennis, armata. Radix tuberosa caudice continuata

apicem versus 4-8 cm. crassa. Caulis brevissimus, leviter ramosus, tuberculatus.

Rami pauci caudicis apice editi plerumque 4-6-angulati, usque 16 cm. longi, 2-3 cm.

crassi, maiores segmentis 1-2 cm. longis constricti, podariis corneis confluentibus

2-aculeatis; anguli compressi recti vel rariter leviter torti. Aculei usque 1 cm. longi

minimi ad ramorum constrictiones. Cyma singula, subsessilis, tribus cyathiis. Cyathium

primum masculinum, 2 bisexualia. Involucrum 5-6 • 5 mm. diam., glabrum, lobis 5 parvis

laceratis et glandulis transverse oblongis 3 mm. latis integris luteis munitum. Ovarium

glabrum, sessile; styli 3-5 mm. longi ad medium connati, apice bifidi. Capiu/a circiter

6 mm. diam., sessilis, obtuse trilobata; semina 2-5 mm. longa, subglobosa, brunea, plana.

Transvaal. — Letaba district: on farm The Downs, between P.O. The Downs and

Wandrags Asbestos Mine, 4,500 ft., Codd 3092 (type); Crundall in National Herbarium,

Pretoria, 28391.

Plate 9.

225

A dwarf succulent plant with a combined subtuberous main root and reduced

main stem, which usually branches into 2 or more stem-like branches, each of which

produces a tuft of aerial flowering branches. Main stem 4-8 cm. diam., irregularly

tuberculate with some tubercles over 0-5 mm. prominent. Branches up to about

16 cm. long, generally unbranched, but occasionally rebranched, withering with age and

finally falling away from stem, up to about 3 cm. in diam. and constricted into segments

1-2 cm. long, depending on seasonal conditions, 4-6-angled. Angles fairly acute and

winglike, armed with a continuous horny margin and paired spines. Spines up to

1 cm. long in pairs at intervals of about 7-5 mm. but closer on stem constrictions,

with a pair of prickles on the sides of the flowering eyes, which are slightly below midway

between the spine pairs. Cymes solitary from the flowering eyes of the terminal seg-

ments, with 3 cyathia, 1 central male and 2 bisexual produced in a plane parallel to the

main axis, occasionally 1 bisexual cyatheum suppressed; peduncle very short with 2

broadly ovate obtuse bracts subtending the cyathia; bisexual cyathia with a very short

bibracteate peduncle. Involucres cup-shaped, glabrous, 5-6 mm. diam. with 5 glands

and 5 subquadrate, somewhat lacerate lobes; glands yellowish, transversely oblong,

about 3 mm. in their greater diam., concave, with slightly upturned margins. Ovary

sessile, glabrous; styles about 3-5 mm. long united into a column for half their length

and bifid at the tips; connective tissue to ovary shortly hood-shaped. Capsule about

6 mm. diam. moderately 3-angled, 2-5 mm. long; subglobose, slightly longer than

broad, brown, smooth.

The first record in the National Herbarium of this species dates from April, 1945,

when Mr. A. H. Crundall forwarded material from the farm The Downs in the Letaba

district of the Northern Transvaal. The plant, however, was not in flower and it was

not until Dr. L. E. Codd and Mr. B. de Winter collected complete material on the same

farm, in October, 1947, that it was possible to draw up a fuller description. Dr. Codd

records that the plants were locally common, growing in shallow soil and in the crevices

of dolomite ridges, being associated with E. schinzii Pax. The new species is restricted

both in size and distribution and hence the specific epithet.

While the plant resembles some forms of E. enormis N.E. Br. in the Lydenburg

district, it is readily distinguished by the branches being unmarked, yellowish-green

with up to 6 angles; by the angles having a continuous horny margin; and by the sub-

sessile cymes with smaller cyathia. In general appearance it suggests a miniature

E. barnardii W. D. & S., which is also in the same geographical region and may well

be the closest relative in the evolutionary sense.

Euphorbia unicornis R. A. Dyer, sp. nov: (Euphorbiaceae); affinis E. corniculatae R. A.

Dyer podariis corneis 1-aculeatis glandulis connatis facile distinguitur.

Planta humilis, succulenta, perennis, armata, ramosa, usque 30 cm. alta. Rami

plus minusv€ erecti, ramosi, basin versus subcylindrici, 1 cm. crassi, superne indistincte

6-7-angulati, tuberculati; tuberculi pulvinati, 5-7 mm. longi, usque 5 mm. lati, podariis

corneis 1-aculeatis. Acu/eus singulus, 4-6 mm. longus, 2 superioribus parvissimis

1-1-5 mm. longis.- Cyma breviter pedunculata, tribus cyatheis; pedunculi 1-2 mm.

longi apice bibracteati. Cyathia apicem ramorum versus producta; primum mas-

culinum; bisexualia 2, lateraliter producta. Involucrum plus minusve 3 mm. diam.,

glabrum, lobis 5 parvis minute dentatis et glandulis rubris suberectis connatis munitum.

Ovarium sessile 3-angulatum; styli 2 mm. longi ad medium connati apice bifidi.

Mocambique. — Niassa Province: Quissanga district, Cuero Mt., 400 m. alt.,

Sept. Pedro and Pedrogao 5091.

Plate 10.

A branched succulent spinescent shrublet, up to about 30 cm. tall, greyish-green

with purplish tinge. Branches subcylindric, about 1 cm. thick, not perceptibly con-

stricted at intervals, possibly commonly with two lateral branches arising from the

same level, indistinctly 6-7-angled or subcylindric with 6-7 grooves about 1 mm. broad

at the base, tubercled along the angles and with rudimentary leaves on young growth.

226

Tubercles slightly prominent and cushion-like, almost completely covered by a horny

surface but with the horny surface interrupted narrowly between the tubercles at the

flowering eyes, bearing centrally a single spine 4-6 mm. long and above this 2 prickles

1-1 -5 mm. long, one on either side of the leaf-scar. Flowering-eyes 5-7 mm. apart,

slightly above the leaf-scars. Cymes solitary from apical flowering-eyes on peduncles

1-2 mm. long, consisting of a central male involucre and 2 lateral bisexual cyathia

arising in a plane at right-angles to the main axis. Bracts oblong, 1 mm. long. Invo-

lucre cup-shaped, about 3 mm. diam. with 5 subquadrate finely-toothed lobes and a

glandular rim. Glands 5, red, apparently forming a continuous rim but actually con-

nected by very narrow membranes. Ovary sessile, 3-angled. Styles 2 mm. long,

united for about half their length, the free portions becoming twisted, emarginate at the

slightly flattened tips.

The cushion-like tubercles almost completely covered by a horny shield seem to

distinguish this and the related species E. corniculata from previously described plants.

I have, however, seen a similar but incomplete and unnamed specimen with paired

spines collected near the Kenya -Somaliland border, and a close comparison of other

tropical African dwarf species is indicated. The pattern of the cyme is similar to that

displayed by E. schinzii and its relatives.

1 am very grateful to the two Portuguese collectors, Messrs. Pedro and Pedrogao,

who kindly forwarded the material with the following additional information : “ com-

mon name ‘ nya nyach ’, growing near the administrative officer’s residence on rocky

outcrops, greyish-green purpurascently shaded, involucre rosy to red.” The con-

nection of the glands by very narrow membranous strips into a continuous cup-shaped

structure is most unusual for the genus.

Jatropha messinica E. A. Bruce , sp. nov. (Euphorbiaceae); affinis J. pseudoglanduliferae

Pax, sed foliis fere basin lobatis, petalis florum masculinorum liberis, sepalis

femineorum lanceolatis, stylis liberis vel basi connatis differt.

Suffrutex parvus, glaber, 60-90 cm. altus; caules plures, erecti, basin versus ramosi,

substriati, pallido-cinerei vel stramenei. Folia chartacea, viridia, glabra, digitata,

usque basin 5-7 lobata; lobi lanceolati — anguste elliptico-lanceolati, 2-5-6 cm. longi,

5-7 mm. lati, apicem et basin versus angustati, margine glanduloso-serrati, nervis

lateralibus utrinque impressis usquel media circiter 16, nerva media supra impressa,

subtus prominenta; petiolus 2-3-5 cm. longus, glaber. Stipulae dissectae, setaceae,

4- 5 mm. longae, glanduliferae. Inflorescentia terminalia in cymis pedunculis disposita;

bracteae lanceolatae, acutae, 3-4 mm. longae, margine et apice glandulosa-dentatae.

Flores masculini subsessiles, apicem versus 5-6 dispositi, pedunculo circiter 6 mm.

longo. Sepala S 5, ovata, acuta, circiter 1-5 mm. longa, basi cohaerentia, glabra,

eglandulosa. Petala 5, libera, glabra, late ovata, circiter 2 mm. longa, 1-5 mm.

lata, apice rotundata, basi subunguiculata. Stamina 8, monadelpha, antheris vix 1 mm.

longis. Flores feminei , basi inflorescentia masculina solitarii, subsessiles vel breviter

pedicellati. Sepa/a $ 5, lanceolata, circiter 3 mm. longa, margine integra vel superne

glandulosa. Petala $ 5, libera vel leviter coherentia, glabra, obovata, circiter 6 mm.

longa, 3 mm. lata, apice rotundata, basin versus angustata, subunguiculata. Discus

5- lobatus. Ovarium ovoideum, glabrum, 3-5 mm. longum, 2-5 mm. diam., leviter

trilobatum; styli 3, erecti, liberi vel basi vix connati, persistentes, circiter 2 mm. longi,

stigmata crassa bifida tuberculata. Capsu/a ovoidea, leviter trilobata, 1 1 mm. longa,

9 mm. lata, rugosa, fusco-cinerea, petalis sepalisque basi persistentibus. Semina

suboblonga, c. 7 mm. longa, 4 mm. lata, pallido-brunnea vel straminea fusco-maculata,

dorso rotundata aliter compressa, apice carunculata.

Transvaal. — Zoutpansberg district, Messina, a few hundred yards from the rail-

way station west of Messina, May, 1945, Gerstner 5461, flower and fruit, type ;

Gerstner 5447; Vanetzi Valley, 2,500 ft., March, 1946, Gerstner 6052; Messina, Govern-

ment Experimental Farm Vergelegen, March, D. G. Steyn\ Messina, 1,900 ft., on margins

of river bed, June, Gatpin 9191; Dongola, Farm Greefswald No. 615, 1 mile north of

227

Mapungubwe, 1,700 ft., on a rocky sandstone ridge in light shade, April, Codd 4125.

Jatropha messinica is a small half-herbaceous shrub or soft bush with bright green

digitate leaves very similar to those of Cannabis sativa. There are six different collect-

ings of this species in the National Herbarium, Pretoria; four from Messina, one from

Dongola and one from the Vanetzi Valley about 20 miles south-east of Messina, so the

specific epithet messinica was thought to be appropriate. The Dongola specimen,

Codd 4125, was sent to Kew for identification and was returned as equal to specimens

of Moss and Rogers 76 and Rogers 19350 named J. pseudo-glandulosa. There is no

record of such a name being published and this is probably a clerical error for/, pseudo-

glandulifera, which the laison officer at Kew says is quite distinct from our specimen.

J. messinica belongs to the section Glandulifera Pax of the subgenus Adenoropium

(Pohl) Griseb., which is characterized by the flowers having both petals qnd sepals

and the stipules being fimbriate and persistent. It is most nearly allied to J. pseudo-

glandulifera Pax, but differs from this species in the deeply 5-7-lobed leaves, the segments

of which are almost free to the base. There are also slight differences in the floral

characters, the styles in /. messinica are free almost to the base, whereas in J. pseudo-

glandulifera they are connate for two-thirds of their length, the sepals of the female

flowers in the former species are lanceolate, whereas in the latter they are ovate. There

seems to be some variation in both male and female flowers among the specimens

quoted under /. messinica. A male flower of Codd 4125 was examined and was found

to be larger than those of the type ( Gerstner 5461), the sepals were oblong 2-5 mm.

long and the petals 3-5 mm. long and 2-5 mm. broad and just cohaerent at the base,

not free as in the type. In the female flowers the petals of the type were coherent at

the base, whereas in Codd 4125 they were free. These slight differences are interesting

to record, though they are not of specific importance.

LABIATAE.

Coleus vagatus E. A. Bruce, sp. nov. (Labiatae); affinis C. pentheri Gurke, sed habitu

prostrato vagato bracteis herbaceis viridibus subcrassis apice plus minusve trunca-

tis numquam acuminatis valde distinctus.

Herba perennis, aromatica, basi ramosa; caulis primus subquadrangularis, laterales

prostrati, vagati, subteretes, usque 60 cm. longi, circiter 2-5 mm. diam., pilis albidis

multicellulis patulis parce pilosis et pilis brevibus incurvis pubescentes et glandulis

aurantiacis parcissime induti, internodiis 2-6 cm. longis saepe e nodis radicati. Folia

opposita, petiolata; petiolus 0-5-1 cm. longus, planus, pubescens et pilosus; lamina

ambitu diversa, late ovata — obovata, 1 • 5-2-5 cm. longa, 1 • 5-2 cm. lata, apice rotundata,

superne crenata, inferne in basin cuneatim angustata et integra, pallido-viridis, crassa,

subtus plus minusve dense supra paricissime pubescens at glandulis aurantiacis punctata,

margine pilis albidis Incurvis ciliata; nervi vix conspicui, supra leviter impressi, subtus

prominentes. Inflorescentia compacta, sub anthesi circiter 10 cm. longa, verticillastris

sessilibus 4-6 floris; bracteae oppositae et decussatae, virides, oblongo-ovatae, 9 mm.

longae, 6 mm. latae, subcrassae, apice late rotundatae vel subtruncatae, margine inferne

pilosis multicellulis albidis ciliatae, subtus aurantiaco-glandulosae et appresso-pubes-

centes, supra parce pubescentes nec glandulosae; pedicelli c. 3 mm. longi, pubescentes,

suberecti. Calyx bilabiatus, glandulosus et pilosus; tubus brevissimus, c. 1-5 mm.

longus; lobus posticus suborbiculatus, 3 mm. longus, 4 mm. latus, apice plus minusve

apiculatus margine longe ciliatus; lobi laterales triangulares, c. 2 mm. longi, antici

lateralibus similes 1-5 mm. longi uterque ciliati. Corolla c. 1-6 cm. longa, patens,

atro-violacea ; tubi pars inferior cylindrica, 3-3-5 mm. longa, pars superior geniculatus,

ampliatus, compressus, 5 mm. longus, 4 mm. latus, breve pubescens et parce glandu-

losus; lamium posticum erectum vel leviter recurvatum, suborbiculatum 4 mm. altum,

4-5 mm. latum, parce glandulosum et pilis longis paucis instructum, superne leviter

4-lobatum, lobis posticis majoribus et latioribus 2-2-5 mm. latis, lobis lateralibus

minoribus 0-5 mm. latis subacutis; lamium anticum patens, cymbiforme, 1-1-1 cm.

longum, c. 3-5 mm. profundum, pubescens, glandulosum, et pilis longis paucis

228

Fig. 4. — Coleus vagatus E. A. Bruce.

-Plant, natural size. 2 — Bract, X 2. 3 — Calyx split open, X 2, 4 — Corolla,

> — Portion of stem with leaves, natural size.

229

instructum. Stamina 4, declinata, 2 labium anticum subequalia, vel paullum exserta,

2 breviora; filamentis parte superiore excepta in tubum antice apertum connatis;

antherae 1 -25 mm. longae. Ovarium glabrum, disco in glandulam late oblongam leviter

concavam vix 1 mm. longam producto; stylus gracilis, c. 2-2 cm. longus.

Transvaal. — Nelspruit district, Kruger National Park, 1| miles E. of Skukuza,

900 ft., in lowveld bush, on stream bank, May, Codd 5489 type. Zoutpansberg

district, Messina, Dec., Rogers 20708; Komatipoort, April, Dyke, 5516.

Swaziland. — On the road to Komatipoort, May, Pole Evans 3461.

Fig. 4.

Coleus vagatus is an inconspicuous, aromatic, perennial herb with a fairly short,

erect, quadrangular main stem, from which a number of long, prostrate, straggling

branches arise, which frequently root at the nodes. The species is most closely allied

to Coleus pentheri Giirke but differs from it in the prostrate habit, fleshy leaves and

green, ovate bracts, which are more or less truncate and not acuminate at the apex.

The bracts are arranged in densely packed opposite and decussate pairs, each bract

subtending a 3-flowered cyme, so that the whole forms a dense spike terminating a

lateral branch. The deep violet flowers only appear a few at a time and do not make a

conspicuous show, so that the green inbricate bracts are the main feature of the inflores-

cence. In common with some other species of the genus C. vagatus bears numerous

red-gold glands and long white, multicellular, bristle-like hairs, which are particularly

prominent on the bracts, though they are also present on the stems, leaves, calyces and

corolla and the glands are even present on the anthers. The specific epithet vagatus

meaning straggling or wandering refers to the habit of the plant.

LILIACEAE.

Asparagus glaucus Kies, sp. nov. (Liliaceae); A. suaveolenti Burch, affinis sed perianthiis

majoribus persistentibus differt.

Sujfrutex 30-60 cm. altus. Radices multi, crassi, fusiformes, 0-6 cm. diam.

Rhizoma squamis stramineis 2 cm. longis vestita Caules aggregati (juniores exteriores),

glauco-purpurei, squamis deltoideis coriaceis amplexicaulibus marginibus membraneceis

0-7 cm. longis ornati, caules seniores straminei, basi simplices, superne ramosi, ramis

patentibus; ramuli fasciculati, pungentes, squamis membranaceis parvis ornati. Clado-

dia 1-6 fasciculata, teretia, mucronata, leviter curvata, glauca, 3-7 mm. longa. Flores

solitarii vel binati. Peduneuli basin versus articulati, 3 mm. longi. Perianthium

cremeum, segmentis 3 mm. longis patulis dorso purpureo-carinatis. Filamenta

lanceolata; antherae luteae. Stylus brevis, brevissime 3-ramosus. Bacca glauco-

viridis, 3 mm. diam., globosa, perianthio persistente inclusa. Semina nigra, solitaria.

Cape Province. — Kimberley division: Mollers Pan, Acoeks and Hafstrom, H. 874;

Riet River, Acoeks 8464.

S.W.A. — Great Namaqualand: Great Karas Mountains, Ortendahl 425.

Orange Free State. — Fauresmith division: Petrusberg, Marais 124; Ventersvlei,

Verdoorn 1153; Veldreserve, Henrici 3987. Heuningberg, Marais 155; Groenvlei,

Mogg 13621, Kies 340 (Type in Nat. Herb. Pretoria); Rosemarie, Verdoorn 2148,

1623a.

Fig. 5.

Shrublet, 30-60 cm. high. Roots thick, fusiform, 0-6 cm. diam., many; rhizome

clothed by straw-coloured striate horny scales, 2 cm. long. Stems many, growing in

circular patches (young shoots round the edge of the patch), glaucous, purple or

greenish, bearing deltoid amplexicaul squamae of leathery texture and with mem-

branous margins, 0-7 cm. long; older stems becoming straw-coloured, simple at base,

branched above with branches stiffly at right angles. Branchlets fascicled, ending in

230

Fig. 5. — Asparagus glaucus Kies ( Kies 340).

1 — Twig, x 1. 2 — Twig with fruits, x 1. 3 — Flower, x 2. 4 — Perianth lobe

with stamens attached, X 3. 5 — Gynoecium, X 3. 6 — Rootstock, x 1.

231

sharp spines, bearing minute membranous squamae which subtend the cladodes.

Cladodes 1-6 per fascicle, terete, mucronate, narrowed towards the base, slightly curved,

glaucous, 3-7 mm. long. Flowers solitary or sometimes 2-nate. Peduncles jointed

near the base, 3 mm. long. Perianth lobes 3 mm. long, patently spreading, creamy

with purple line down centre on outside. Filaments broadly lanceolate, anthers yellow.

Ovary with style short and style branches very short. Fruit glaucous, green, 3 mm.

diam. covered by the dry, horny, keeled, persistent perianth. Seed single, black.

Found on surface limestone with Salso/a, and on alluvial soil. Plentiful and

characteristically glaucous in appearance. It differs from A. suaveolens principally

in the glaucous appearance and the larger perianth which persists and encloses the ripe

fruit.

Kniphofia ensifolia Bak. var. albiflora E. A. Bruce, var. nov. (Liliaceae); a typo floribus

albidis differ t.

Transvaal. — Pretoria district: 13 miles west of Witbank on the road to Pretoria,

November 1947, Codd and de Winter 3156 in National Herbarium Pretoria 28928

(type); September, 1948, Codd 4774; 14 miles south-east of Pretoria, December, 1947,

Bruce 27; 2 miles east of Middelburg, November, 1949, Codd 5767.

This striking Kniphofia with its narrow spikes of cream or white flowers, long

exserted stamens and broad glaucous leaves, was found in a marshy vlei in black clay

loam 13 miles west of Witbank. The 3J-5 ft. high plants were growing in small clumps

and were locally abundant in an area of about a quarter of an acre. At first this was

thought to be an undescribed species closely allied to K. ensifolia Bak., K. tuckii Bak.

(Bot. Mag. t. 7644) and K. rivularis Berg. (F.P.S.A. pi. 866). K. ensifolia was originally col-

lected in the Western Transvaal near the Matebe River about 12 miles north-west of

Zeerust. K. tuckii is a native of Colesberg in the northern Cape, and K. rivularis was

described from Modderfontein, north of Johannesburg. K. rivularis had previously

been found in the Rietvlei catchment area near Pretoria, so it was decided to visit this

area and make a comparison between these plants and the specimen from Witbank.

It was not until December 12th, over a month later, that K. rivularis was found in flower,

thinly scattered in a vlei about 14 miles south-east of Pretoria. The flowering spikes

of this species consisted of dull red flowers in the upper part and greenish-white below,

but one inflorescence was found with pure white flowers and no trace of red. This

plant was compared with our var. albiflora from near Witbank and no real distinction

could be found. Both were tall plants with broad glaucous-green leaves with serrulate

keels and margins, though this latter character was variable and some of the outer leaves

were smooth ; the peduncles in both cases were slightly longer than the leaves and the

inflorescences dense, white-flowered and sub-cylindrical. The only difference observed

was that the plant from Rietvlei was in flower a month later and that a basal sterile

bract 5-8 cm. long was conspicuous in the bud stage of this plant and was not observed

in the Witbank ones. It was now necessary to determine whether K. rivularis, the

most recently described of the three species, was distinct from the other two. The

type specimens of K. ensifolia Bak., K. rivularis Berg, and K. tuckii Bak. were kindly

loaned by the Royal Botanic Gardens, Kew, for comparison. The type of K. ensifolia

was unfortunately very fragmentary, but a careful examination of all three specimens

was made and though there were minor differences between them in the width and ser-

ration of the leaves and the size of the floral bracts, they agreed well in the shape and

density of the inflorescence, the shape of the floral bracts, the size and shape of the

perianth and the length of exsertion of the stamens. Since going to press both K.

ensifolia and K. tuckii have been collected from their type localities and from this

material it is clear that K. rivularis is not distinct from K. ensifolia, though K. tuckii

differs in its brighter flower colour and shorter leaves and should be upheld only as a

variety of K. ensifolia, while our white-flowered plant becomes K. ensifolia Bak., var.

albiflora E. A. Bruce.

1599-10

232

OCHNACEAE.

Ochna glauca Verdoom, sp. nov. (Ochnaceae); O. pretoriensi Phillips accedit sed a

qua et a ceteris speciebus africanis foliis ramulisque glaucis differt.

Frutex 1 • 5-2-5 m. altus; ramuli glauci, lenticellati lenticellis haud conspicuis. Folia

hysterantha, glauca, late oblongo-elliptica, 3 cm. longa, 1 • 8 cm. lata, marginibus minute

crenato-serratis ; petioli 2-5 mm. longi. Flores solitarii, apices ramulorum valde

abbreviatorum orti; pedunculi 10-12 mm. longi. Sepala subaequalia, plus minusve

oblonga, 7 • 5-8 mm. longa, 3 • 5-5 • 5 mm. lata, persistentia, deinde 1 cm. longa et reflexa.

Petala mox decidua, unguiculata, circa 10 mm. longa, 8 mm. lata. Stamina indefinita;

lilamenta circa 4 mm. longa; antherae 2-5 mm. longae, 0-75 mm. latae, basifixae,

rimis apicalibus dehiscentes. Carpella 5, 1 mm. longa, 0-5 mm. lata; stylus gynobasis

5-5 mm. longus apice breviter 5-lobatus; lobi 0-25 mm. longi apice stigmatosi. Drupae

nonnumquam 2 disco incrassato insidentes, circa 8 mm. longae, 6 mm. latae.

Transvaal. — Zoutpansberg district: de Klunderdt, Dongola Botanic Reserve, Pole

Evans 4494 (fruiting) type; 4590; 4588 (flowering); Verdoom 2292, 2333.

A glaucous straggling shrub 5-8 ft. tall; branches brittle, lenticels not very conspicu-

ous. Leaves crowded, glaucous, metallic green, hysteranthus, often folded about the

midrib and drooping, broadly oblong elliptic, about 3 cm. long and 1 • 8 cm. broad,

margin rather obscurely or finely crenate-serrate, teeth proclivent, midrib prominent

beneath, many slender lateral veins prominent above or on both surfaces and reticulate

veins fairly obvious; petiole 2-5 mm. long. Flowers precocious, solitary at the apices

of much abbreviated shoots which are numerous on the apical branchlets. Peduncle

slender 10-12 mm. long. Sepals 5, rather unequal, the outer 7-5 mm. long 5-5 mm.

broad, and the inner 8 mm. long and 3-5 mm. broad. Petals soon deciduous, obovate,

crinkled in bud, 10 mm. long 8 mm. broad, rounded or bluntly acuminate at the apex,

cuneate into a claw at the base. Filaments numerous, about 4 mm. long; anthers 2-5

mm. long, 0-75 mm. broad basifixed, dehiscing by 2 apical slits. Carpels 5, 1 mm.

long and 0-5 mm. broad; style gynobasic, 5-5 mm. long, very shortly 5-lobed at the

apex, lobes 0-25 mm. long, stigmatic at the apex. Fruit usually with 2 carpels maturing

on the fleshy disc, about 8 mm. long, 6 mm. broad, green, turning brownish (persistent

sepals reflexed from the disc turning reddish-brown).

All the specimens quoted above come from a granite hill on the farm de Klunderdt

in the north-western Transvaal, but they were collected at different times of the year.

The farm De Klunderdt was in the now abandoned Dongola Reserve. The species

occurred in large numbers on this particular hill but it has, to date, not been recorded

from anywhere else in South Africa. A duplicate sent to Kew was matched with a

specimen, Eyles Herb. 3765, from the Matopos, S. Rhodesia, so it is probable that the

species occurs in the granite hills of that country and our locality is the most southerly

one. (While in press it has been possible to confirm this through the courtesy of the

officer in charge of the Salisbury Herbarium.)

The glaucous green leaves are the most striking characteristic of this species. Like

other species of Ochna the shrubs are leafless in winter and the flowers appear profusely

before the leaves in spring. The leaves and fruits are borne during the very hot months

when few collectors are out in the granite hills of these sub-tropical regions.

The species, having the inflorescence solitary or 2-nate at the apex of abbreviated

shoots, falls in the same section in Phillip’s key (Bothalia 1, Part 2; page 87) as O. pre-

toriensis Phillips and it is nearest to that species. Besides the glaucous colour of the

leaves there are other characters in which our specimens differ from O. pretoriensis,

but further material is required before they can be said to be specific distinctions. There

seems to be some variation in leaf-shape in O. pretoriensis but the leaves of our plant

are shorter and broader than those of the type and less distinctly serrate. In the fruiting

material of O. glauca the enlarged persistent sepals are not erect and surrounding the

233

ripening fruits as seen on specimens of O. pretoriensis, but are strongly reflexed,

exposing 2 green drupes, turning brownish as they ripen, seated on the enlarged disc,

the sepals also turning a brownish-red colour. In O. pretoriensis the sepals eventually

become reflexed evidently only after the fruit has ripened. The flowers, especially the

petals of our specimens, are smaller than those of typical O. pretoriensis.

PEDALIACEAE.

Ceratotheca saxicola E. A. Bruce sp. nov. (Pedaliaceae) ; ab omnibus speceibus foliis

minoribus suborbiculatis vel late ovatis, floribus concoloribus hellebori-rubris

habitu minore multo ramosa saxicola valde distincta.

Herba perennis, parva, c. 45 cm. alta, basi et superne multo ramosa; rami graciles,

teretes, lignosi, dense pubescentes, basi c. 3 - 5 mm. diam., ramuli graciliores vix 1 - 5 mm.

diam., internodiis 1-2-4 cm. longis. Folia patuli, opposita, longe petiolata; petioli

foliorum inferiorum usque 1 • 5 cm. longi superiorum 3 mm. longi, dense pubescentes ;

laminae suborbiculares vel late ovatae, inferiores 1-3-1 -8 cm. longae, 1 -3-2-4 cm.

latae, basi truncatae, late cuneatae vel rare subcordatae, apice rotundatae, margine

crenatae, utrinque pubescentes, superiores minores, ovatae vel late ovatae, 9—1 1 mm.

longae, 8—13 mm. latae, basi cuneatae, aliter quam inferiores. Flores hellebori-rubri,

in axillis foliorum superiorum solitarii; pedicelli 4-5 mm. longi, pubescentes. Calyx

ad basin 5-lobatus; lobi leviter inaequales, oblongi, inconspicue falcati, 5-5-5 mm.

longi, 2-2-5 mm. lati, apice rotundati, utrinque pubescentes. Corollae-tubus c. 2 cm.

longus basi subglobosus, 3 mm. longus, 4 mm. diam., deinde inflatus oblique sub-

campanulatus, fauce 1-5 cm. diam., extra pubescens, intus e constrictione dense pilosus

aliter glaber; lobi 5, extra pubescentes, intus glabri, 1 inferiore majore subpendulo late

ovato 7 mm. longo 8-5 mm. lato, 4 superioribus suberectis transverse ellipticis 3 mm.

longis 6-5 mm. latis. Stamina 4, subdidynama, supra constrictionem inserta, fila-

mentis 1 • 1, et 1 - 2 mm. longis glabris, antheris anguste oblongis 3 mm. longis. Discum

annulum, c. 0-5 mm. profundum. Ovarium subcylindricum, c. 2 mm. longum, 1 -3 mm.

latum, apice rotundatum, pubescens; stylus gracilis, glaber, 1-8 cm. longus, stigmata

bilobata lobis planis oblongis apice acutis usque 1 - 5 mm. longis. Capsula oblongo-

ovoidea, compressa, pubescens, 8-9 mm. longa, c. 5 mm. diam., apice breviter bi-cornuta,

basi rotundata, cornubus patentibus vix 0-7 mm. longis. Semina subovoidea, com-

pressa, brunnea, c. 2 mm. longa, rugoso-tuberculata, margine pallidiore leve.

Transvaal. — Zoutpansberg district: Kruger National Park, 32 miles north-east

of Punda Maria, 1,000 ft., in crevices in sandstone krantzes overlooking the Levubu

River, May, Codd 5535 (type); April, Codd, 5980.

Ceratotheca saxicola is an attractive little plant with showy deep hellebore red

flowers (Ridgway, pi. 38), differing from other species of Ceratotheca in its branched,

bushy habit and small leaves. It appears to be very localized, as Dr. L. E. Codd, who

discovered the species, has found it growing in only the one place in the Kruger National

Park, though he has done extensive survey work on the flora of the whole of that

region. Only two or three plants were found growing in a more or less inaccessible

rock crevice on a steep rock face in Karoo sandstone formation, on the south bank

of the Levubu River some miles west of the Pafuri Rest Camp.

C. triloba E. Mey ex Bernh., the only other species recorded from South Africa,

is a common weed, which occurs throughout both Tropical and South Africa. Our

plant is easily distinguished from this, not only in the flower colour, habit and size of

leaves but also in the oblong obtuse calyx lobes, the broader anthers and stigma lobes,

the very short horns on the capsule and the lack of glandular pubescence.

The specific epithet saxicola, meaning “ rock-dweller ” refers to the plant habitat.

234

RUTACEAE.

Fagara humilis E. A. Bruce, sp. nov. (Rutaceae); affinis F. capensi Thunb. et F. magalis-

montanae Engl., a quibus foliolis minoribus latioribus numquam nitidis basi

rotundatis paribus subaequalibus nec apicem versus majoribus, rhachide piloso'

nec vel vix caniculato, floribus majoribus valde distincta.

Frutex parvus, 0-9-1 -5 m. altus, cortice pallido-cinereo ; rami e basi plures, sub-

erecti, ramosi, dense pubescenti, demum glabrescentes, 4-7 mm. diam., aculeis plerumque

curvatis 5-9 mm. longis basi valde dilatatis interdum e ramulis junioribus aculeis sub-

erectis armatis. Folia petiolata, 8-15 (plerumque 9-10) jugata paribus subaequalibus

vel apicem versus minoribus; petioli et rhachides 7-9 cm. longi (juniores breviores

3-6 cm. longi,) pilosi nec vel vix caniculati, aculeis parvis recurvatis armati. Foliola

aromatica, sessilia, coriacea, numquam nitida, ovata vel oblongo-ovata, leviter obliqua,

basi rotundata, apice rotundata vel rare subacuta, 1-1-3 cm. longa, 5-7 mm. lata,

margine superne crenata in quoque sinu glandula notata leviter incurvata pilis paucis

ciliata, nervis lateralibus utrinsecus circiter 4-6 patulis; foliola juniora tenuiora, minora,

supra fusca subtus pallidiora. <$ Inflorescentia racemosa, racemis e nodo caulis

apicem versus 4-7 fasiculatis; racemi 2-5-3 cm. longi, 4-6-flori, pedunculo pubescente

basi sublanato petiolo multo breviore; flores albi vel flavi, breviter (1-1-5 mm.) pedi-

cellati, inferiores mox decidui, pedicellis basi articulatis. Sepala 5, parva, late triangu-

laria, basi connata, c. 1 mm. longa, apice plus minusve lanata, acuta. Petala 5, oblongo-

elliptica, leviter cymbiformia, c. 4 • 5 mm. longa, 1 • 8 mm. lata, basi angustata, apice

obtusa vel subacuta. Stamina 5, filamentis 2 • 5 mm. longis, antheris anguste oblongis

2 mm. longis. Ovarium abortivum, parvum, subglobosum. $ Inflorescentia spicata,

spicis apice caule 3-5 fasciculatis ; spicae c. 2 cm. longae, pedunculo pubescente. Sepala

5, parva, ovato-triangularia, basi connata, c. 0 • 5 mm. longa, apice obtusa vel subacuta,

margine ciliata. Petala 5, oblongo-elliptica, 2-5-3 mm. longa. Stamina 0; staminodia

minuta. Ovarium late et oblique ovoideum, 1-5 mm. longum, stylo incurvato uni-

lateraliter juxto, stigmata subglobosa, disco parvo. Capsula oblique ovoidea, circiter

7 mm. longa, glandulosa, brunnea. Semina solitaria, nigra nitidaque.

Transvaal. — Letaba district: Kruger National Park, 4\ miles south of Shingwedzi

Camp, 1,000 ft., shrub 4 ft. high with yellow flowers and reddish twigs, in Mopane veld

on reddish granitic flats, Nov. 4th, 1948, Codd and Dyer 4671 fls.) type; Shing-

wedzi Camp, near Ranger’s house, 900 ft., shrub 3 ft., erect branches, aromatic leaves

and white flowers, in lowveld bush on rocky koppie, Nov. 3rd, 1948, Codd and Dyer

4651 (<£ fls.); Famont 25 (young fruit); 13 miles north-east of Punda Maria, 1,400 ft.,

shrub 5 ft. high with aromatic leaves in Mopane veld in leegte, Codd and Dyer 4579;

near Skukuza, in bush, leaves sweet-scented, Oct. 1934, Letty 68 ($ fls.). Vanetzi

River, 2,500 ft., shrub 4 ft. in dry bush veld, root-bark used against sores in the mouth,

oil cellules in corners of leaf serrations, March 2nd, 1946, Gerstner 6034 (leaf); near

junction of Limpopo and Pafuri Rivers, Smuts 2392 (leaf).

Portuguese East Africa. — On dry stony soil in forest of Copaifera and Acacia,

May 8th, 1944, Torre 6599 (fruit).

This interesting species of Fagara has been collected several times in the Kruger

National Park in addition to two outside localities. Material of both male and female

flowers, fruits and leaves has been gathered, so that it has been possible to draw up a

full description of the species. According to Engler’s revision of Rutaceae in Engl,

and Prantl. Natiirliche Pflanzenfamilien Bd. 19a, p. 221 (1931), this species should

belong to Section Maqueria Triana and Planch., subsection Paniculatae Engl. § Geronto-

gaeae Engl., and in the small leaflets is most closely allied to F. capensis Thunb. and

F. magalismontana Engl. It is, however, distinguished from these species in the dwarfer

habit, pubescent branches and leaf rhachis, dull surfaced leaves and fasciculate inflores-

cences.

235

F. humilis , as the name implies, is a small, comparatively insignificant shrub with

no outstanding characteristics. It grows singly or forms a low thicket of thorny stems,

which bear the small creamy-yellow flowers towards the ends of the branches in late

spring before the leaves are fully developed.

SCROPHULARIACEAE.

Lindernia dongolensis E. A. Bruce, sp. nov. (Scrophulariaceae) ; affinis L. stuhlmanii

Engl., sed caule simplice, foliis inferioribus plurimis, superioribus apicem caulis

versus longioribus, floribus axillaribus differt.

Herba erecta, simplex, gracilis, glabra, 5-7 cm. alta, caule subqudrangulare circiter

0-5 mm. diam. Folia sessilia, integra, glabra, inferiora plurima, linearia vel subulata,

5-7 mm. longa, basi 1 mm. lata, superiora linearia vel anguste lineari-lanceolata, usque

1 • 3 cm. longa, 1 • 5 mm. lata, apicem caulis versus in 2-4 paribus disposita. Flores

in axillis foliarum superiorum solitarii, pedicellis 4-8 mm. longis. Calyx campanulatus,

glaber, 2-2-5 mm. longus, 5-lobatus, lobis trianglari-ovatis circiter 1 mm. longis basi

vix 1 mm. latis apice rotundatis vel subacutis minute mucronatis. Corolla 9-1 1 mm.

longa, tubo infundibuliformi 4-5 mm. longo fauce circiter 4 mm. diam., labio superiore

anguste obovato glabro 4 mm. longo 2-5 mm. lato, inferiore longiore trilobo lobis

lateralibus anguste obovatis 4-4-5 mm. longis 3 mm. latis medio longiore obovato

4-5-6 mm. longo apicem versus 4-4-5 mm. lato basi 2 mm. lato inferne puberulo.

Stamina 4, filamentis 3-5-4 mm. longis, inferioribus basi 1 mm. incrassatis geniculatis,

superioribus c. 3 mm. longis fauce insertis; antherae c. 0-5 mm. longae divaricatae.

Ovarium ovoideum; stylus circiter 4 mm. longus, stigma bilamellata lobis suborbicu-

laribus circiter 1 mm. diam. margine fimbriatis. Capsula ovoidea, circiter, 4 mm. longa,

2 mm. diam. calyce longiora.

Transvaal. — Zoutpansberg district: Dongola area, De Klundert, growing in water

in a shallow rock pan on a rocky koppie, March, Bruce 66 (flower).

Fig. 6.

This small slender herb was found growing in about £-1 in. of water in a shallow

rock depression or pan about 4 ft. diam. on a large flat rock slab half-way up a koppie

in full sun. There were a number of these plants growing close toegether in little clumps,

their long slender roots intertwined and forming a dense mass, which held together

what little “ soil ” there was. The rather fleshy basal leaves of the plants form a

dense cluster or rosette and towards the top of the otherwise bare stem were two to four

opposite pairs of rather longer leaves. The plants were flowering freely and the white

corollas tinged with lilac with an orange spot at the throat, made a colourful display

against the bare grey rock. The upper lip of the corolla is entire and forms a little

hood under which the stamens are curved, the lower lip is large and 3-lobed, forming

a platform from which the lower pair of stamens arise. This pair of stamens has a

purple-tinged, knee-like, puberulous appendage at the base.

The genera Craterostigma, Ilysanthes, Torenia and Lindernia are very closely

related. Ilysanthes can be separated out in having only two stamens, whereas the other

three genera have four stamens. Both Torenia and Craterostigma have a winged or

plicate calyx, which is not the case in Lindernia. Our specimen, therefore, definitely

belongs to the genus Lindernia, as it has four stamens and the calyx-tube is neither

plicate nor winged. A duplicate specimen was sent to Kew and it was there confirmed

that it was a new species of Lindernia, though similar in habit to Ilysanthes welwitchii

Engl. It is interesting to note that Dr. Pole Evans, who knows the vegetation of the

Dongola area very well, has no recollection of having seen this little plant before.

236

Fig. 6. — Lindemia dongolensis E. A. Bruce.

1 — Corolla split down one side, x 4. 2 — Calyx, X 4. 3 — Stigma x 8.

237

VERBENACEAE

Vitex patula E. A. Bruce sp. nov. (Verbenaceae) ; affinis V. madiensi Oliv. et V. amboniensi

Giirke, a quibus foliolis sessilibus vel subsessilibus fructu minore, a V. madiensi

foliolis minoribus integris, a V. amboniensi foliis 3-foliolatis obovatis obtusis

facile distinguenda.

Frutex vel arbor parva, 2-4-5 m. alta; rami patuli, subteretes, cani vel pallido-

brunnei, parce pubescentes demum glabrescentes, longitudinaliter fissi; ramuli juniores

pallido-flavo-brunnei-tomentosi. Folia opposita, submembranacea nec coriacea,

plerumque 3-foliolata rare 4-5-foliolata, longe petiolata; petiolus inflorescentia plus

minusve subaequalis, 3 -5-6 -5 cm. longus, teres nec caniculatus, pubescens vel juniore

breviore tomentoso; foliola integra, sessilia vel subsessilia (terminale obovata 4-7-5

cm. longa 2-3-5 cm. lata lateralibus minoribus obovato-ellipticis 3-6-5 cm. longis

2-3 • 5 cm. latis) apice rotundata vel rare subacuta, basi cuneata, subtus parce glandulosa

et nervis crispo-pubescentibus, supra parce scabridula (juniora densiore pubescentia) ;

nervi laterales utrinsecus 8-11, subtus prominentes plus minusve patuli, nervis tertiariis

distinctis inter se paralleled. Cymae dichotomae, axillares, ramulos apicem versus

longe pedunculatae ; pedunculi 3-4 cm. longi, dense pubescentes vel subtomentosi;

bracteae lineares, 0-8-1 -3 cm. longae, pubescentes vel tomentosae. Flores parvi

pedicellati, mauvi. Calyx obscure bilabiatus, 5-dentatus, extus tomentosus, intra

glabrescens, tubo campanulato c. 2 • 5 mm. longo, lobis 3 anticis quam 2 posticis majori-

bus triangularibus acutis 1-1-5 mm. longis. Corollae-tubus anguste infundibuliformis

calyce longior, c. 4-5 mm. longus, basi 1 -5 mm. fauce 3-5 mm. diam.; lobi inaequales,

tomentosi apice rotundati, antico maximo transverse elliptico 3-5 mm. longo 4 mm.

lato lateralibus ovatis 2-5 mm. longis 2 mm. latis. Stamina 4, basi corollae inserta,

breviter exserta. Ovarium globosum, 1 mm. diam., glabrum, glandulis paucissimis

apicem versus obtectum. Fructus niger, nitidus, ovoideus, 1 • 5 cm. longus, 0 • 9 cm. diam.

Portuguese East Africa. — Maputo district: between Santaca and Mazimiama,

on sandy soil in open forest, Nov., Gomes e Sousa 3885; near the Transvaal border,

growing in sandy soil, small edible berry, Dec., Lamont 27.

Transvaal. — Zoutpansberg district: Kruger National Park, Dzundweni Hill,

11| miles S.E. of Punda Maria, in lowveld bush on rocky hillside, 1,500 feet, March,

Codd 5319, (type); miles north-east of Punda Maria, in dense bush on low rocky

ridge, 1,400 feet, June, Codd 4227; Punda Maria, in sandy veld on koppie, Jan., Lamont

45 ; 6 miles north-west of Punda Maria, in fairly dense bush, 1,500 feet, April, E. A. Bruce

172; Limpopo, Dec. 1928, Hutchinson 2101; 1£ miles north of P.O. Wylliespoort,

3,000 feet, on steep rocky hillside at foot of north slopes of Zoutpansberg Range, May,

Codd 4163; near Wylliespoort, Nov., Pole Evans 1951 ; Wylliespoort, April, Rodin 4232.

Vitex patula is a summer flowering shrub or small tree with rather straggling,

spreading branches, mauve flowers, 3-foliolate, rarely 4-5 foliolate, leaves and small

shiny, black, edible fruits. The species generally occurs in fairly dense bush on rocky

hillsides, but has also been found on sandy soil in open forest. Its range of distribution,

as represented by our material, is from Wylliespoort in the Zoutpansberg range east-

wards to Punda Maria and then a long jump south-east to the Maputo district in

Portuguese East Africa.

As in other species of Vitex there is variation in leaf size, texture and pubescence

and in the pubescence of the branchlets in the flowering and fruiting stages ; the leaves

and young branches particularly are much more densely pubescent in the flowering

than in the fruiting stages. According to Pieper’s Revision of Vitex in Engler’s Bot.

Jahrb. (1928), V. patula E. A. Bruce, on account of its glabrous, subglobose ovary and

more than 3-celled calyx hairs with an elongate unthickened tip, should belong to

subgenus Euvitex, sect. Axillares, subsect. Cymosae § G/andulosae: Eutriches. Material

238

sent to Kew for comparison was allied to V. amboniensis Giirke, a species belonging to

§Pilosae, which differs from our species in the 5-foliolate leaves with narrower, acute,

more pubescent, petiolulate leaflets, and hairy ovary. Our species is also related to

V. madiensis Oliv., a very variable species with a number of varieties. It can be dis-

tinguished from this in the smaller, sessile, thinner-textured leaflets in which the tertiary

veins are finely impressed, reticulate.

239

Plate 1.

Ceropegia decidua E. A. Bruce.

(Photo of type by H. King.)

1599-11

240

24 1

Plate 3.

Encepluilartos humilis Verdoorn. Top male cone on plant on Schagen

Hill ; below, the whole plant. [Photos by R. A. Dyer.

1 599—12

242

Plate 4.

Euphorbia clivico/a R. A. Dyer.

243

Plate 5.

Euphorbia confinalis R. A. Dyer; type specimen taken from the tree in

foreground in the Kruger National Park. [ Photo by L. E. Codd.

244

Plate 6.

Euphorbia keiihii R. A. Dyer, in natural habitat.

245

Plate 7.

Euphorbia keithii R. A. Dyer. Left, small tree in natural habitat on face of

krantz near Stegi; right, top of branch in young fruit.

[Photo by H. King.

246

Plate 8.

Euphorbia keithii R. A. Dyer, fruiting specimen.— (Photo H. King.)

247

Plate 9.

Euphorbia restricta R. A. Dyer. (Photo H. King.

248

Plate 10.

Euphorbia unicornis R. A. Dyer. (Photo H. King.)

Micromorphology of the lower Hymenomycetes.

By

P. H. B. Talbot.

To classify the lower Hymenomycetes, not often distinguished by their shape,

colour or hymenial configuration, mycologists eventually turned to the microscope

to seek more exact diagnostic characters. As far as the circumscription and

identification of species is concerned, microscopy has revealed many useful characters.

But the other ideal of taxonomy, namely a natural classification, is still far from

realisation as the significance of facts already observed has only partially been evaluated.

This paper is an attempt to review with illustrations the morphology of the

microscopic organs found in the lower Hymenomycetes, special attention being given

to the basidium. For amplification examples are cited from other groups of fungi

as well. Some standard of terminology has been aimed at, particularly with regard

to the basidium. The conclusions reached are that Linder’s terminology for the

basidium is unacceptable as it over-simplifies this organ, while Neuhoff’s terminology

is based on homologies which are unacceptable. Moreover, Neuhoff’s terminology is

flexible and in many instances would vary according to the subclass in which a particular

fungus is classified. We believe, with Donk, that the place or stage of karyogamy,

and the place or stage of meiosis, are fundamentals which should be used in formulating

a terminology for basidia.

The new term protosterigma coined by Dr. M. A. Donk, is published here for the

first time with its originator’s generous permission. The term is explained in a note

on sterigmata in general provided by Dr. Donk, which should be read in conjunction

with this paper.

Morphology, Cytology and Terminology of Basidia.

(Note. — In reading this section, reference should be made to Fig. 1 for terminology

and to Fig. 2 for cytology).

In the simpler Hymenomycetes, basidial characters are most important for the

recognition of genera and higher ranks, and studies of basidial types have led to a

better understanding of basic relationships on which to build a more natural

classification (Juel, 1898; Maire, 1902; Neuhoff, 1924; Gaumann, 1928; Donk, 1931;

Rogers, 1934; Martin, 1938; Linder, 1940; Heim, 1948, 1949). Rogers (1944)

summarised the position in stating that classification is implicitly an embodiment of

an hypothesis concerning phyletic relation and that arising from this there are two

views of reliable indications of relationship in the Hymenomycetes, viz. (1) that by menial

configuration is a reliable indication of kinship and (2) that the basidium and associated

characters are reliable indications of kinship, and that hymenial variations are

comparatively recent and trivial. There seems no doubt that the second view is the

more acceptable one. Increasing use of the microscope has demonstrated affinities

which were long unsuspected.

One of the most significant steps in the study of Basidiomycetes was the recognition

of the homo- and heterobasidial subclasses by Patouillard (1900), and the resultant

attention given to basidial morphology. Since then, many new facts have been observed

about the morphology and cytology of the various types of basidia, but there is still

much controversy in their interpretation, and particularly in the terminology applied

to the different parts of basidia.

250

Fig. 1 . — Three mature heterobasidial types, Tremella, Septobasidium and Tulasnella,

drawn diagrammatically to indicate the terminologies used by Linder, Neuhoff (as

elaborated by Rogers), and Donk.

251

In any science terminology is necessary to define objects unambiguously. For

objects such as basidia, which are composed of a number of different parts, it is necessary

that only those parts which are entirely homologous (and not merely similar, or

analagous) should receive the same name. By homologous is meant “ fundamentally

corresponding in value or relationship with one another ”. For such structures to

be homologous they must develop in the same way, serve the same function, have the

same relation to a fundamental type, or have the same relative position (Shorter Oxford

Fig. 2. — Diagrammatic representation of cytological changes in several types of

homo- and hetero-basidia. Details reconstructed from information or illustrations

by authors cited below: — -

1. Karyogamy followed by chiastic meiosis. 2. Karyogamy followed by

stichic meiosis. 3. Auricularia (Maire, 1902). 4. Sebacina (Whelden, 1935 b).

5. Agaricaceae, Hygrophorus (Maire, 1902). 6. Cantharellus (Maire, 1902).

7. Tulasnella (Rogers, 1932). 8. Vuilleminia (Maire, 1902). 9. Guepinia (Dacry-

opinax spathularia) (Bodman, 1938). 10. Brachybasidium (Gaumann, 1922 a).

1 1 . Jola (Gaumann, 1922 b). 12. Uredinales (Linder, 1940). 13. Hyaloria (Martin,

1937). 14. Phleogena (Shear & Dodge, 1925). 15. Mycogloea (Olive, 1950).

16. Helicogloea (Baker, 1936). 17. Clavaria (Maire, 1902). Note that there may

be minor variations in different species of the genera cited, but that the course

of events has been generalised. Note too that the spindle orientation is variable

in some species, and also the plane of septation.

252

English Dictionary, 1933). Terminologies have been proposed or clarified by three

recent exponents of various hypotheses, namely Neuhoff (1924) and his followers,

especially Rogers (1934) and Martin (1938), secondly Donk (1931) and thirdly Linder

(1940). These terminologies are illustrated diagrammatically in Fig. 1. Of these,

Rogers’s contribution has perhaps received the most support; but the time has come

to re-examine all three in an attempt at revaluation and clarification.

In the older descriptions of Basidiomycetes there was little differentiation in

terminology of the various parts of the basidia. The whole organ was known (as it

is now) as the basidium. Sterigmata and probasidia were recognised as parts of the

basidium. The term probasidium originated with van Tieghem (1893) to denote the

basal vesicles of the basidia of rusts and smuts, i.e. the teleutospores and chlamydospores

of these groups respectively; but it was later rightly extended by Donk to include all

primary tasidial cells in the same stage of development, and particularly the more or

less resistant or persistent vesicles in a similar position on the basidia of the

Auriculariaceae. The term sterigma was applied to such spore-bearing structures as

those found in the normal clavate homobasidium, as well as to tubular extensions of

the primary basidial cell found in Heterobasidiomycetes. One cannot accept Boedijn’s

(1937) contention that the primary basidial cell and its tubular extensions (e.g. in

Helicogloed) are separate organs; they are each part of a single organ, the basidium.

The first consideration is whether an extended terminology is essential or whether

we may be content, as Linder (1940) was, to return to the simple terminology of

probasidium, sterigma, and basidium. The bone of Linder’s thesis is phylogenetic,

namely that the septate basidium and germination of the basidiospores by repetition

are primitive characters. To quote, “ The basidium, once established, is essentially

the same throughout the class, but has undergone simplification as a result of loss of

septation. Also it has in higher forms taken over the functions of the probasidium

and become the locus of both caryogamy and meiosis. Because of the essential unity

in structure and function of this organ, the relevant terminology can be greatly reduced

and a return made to the simple descriptive terms which were in vogue previously

Although phylogeny has been called “ the playground of imaginative theorists ”,

there is a phyletic implication in all plant classification. It is likely that the simplification

or reduction in the basidium, which Linder suggests, has been the course of evolution

of these fungi, and not as Neuhoff (1924) suggested that heterobasidial types are

derivative from homobasidial types. Martin (1938) agrees that the clavate basidium

with four sterigmata at the apex of the gonotocont is a reduced, and not a primitive,

type. He also observes that reduction is carried still further in the Gasteromycetes,

whose sterigmata may be eliminated. Rogers (1934, 1947) comments at some length

on this reduction in the Gasteromycetes, in whose homobasidia there is degenerative

variability with loss of function. Their sterigmata, no longer serving as spore-projectors,

are either grotesquely long and tenuous, or progressively short and obsolescent.

Let us now examine Linder’s simple terminology and see how adequate it is in

homologising and defining the different parts of heterobasidia (See also Figs. 1 and 2).

(1) Probasidium. — The term used by van Tieghem (1893) in the connotation given

above, was extended by Donk (1931) to include all basidia in the stage between the

formation of the primary basidial cell until the time that it forms protuberant extensions

which bear the spores directly or indirectly. This is undoubtedly “ that part or stage

in which karyogamy occurs” (Donk, 1931). Such a definition homologises such

diverse structures as: the basal sac of Helicog/oea; the young, undivided cell of

Tremella, which later becomes four-celled; the persistent or resistant basal vesicles of

some Auriculariaceae, e.g. Septobasidium; the young basidium of the Dacryomycetaceae

before forking; and the young, clavate basidium of the higher Hymenomycetes before

production of sterigmata. Whether these cells are narrow or inflated, thickwalled and

resistant or thinwalled and evanescent, does not alter the basic fact of their homology.

The point is that these cells bear the same relation to one another in time of appearance

253

in cytological function for karyogamy, and in development to give rise to further

structures designed for spore formation and liberation. Variations such as sclerotic

walls are surely insignificant except as biological or ecological adaptations.

For Linder (1940) to limit the term probasidium to the precursor of the young

basidium in rusts, smuts and Auriculariaceae is to ignore the potentialities of essentially

the same type of cell in other Basidiomycetes.

In some basidia (see Fig. 2) both karyogamy and later meiosis take place in the

primary basidial cell which we have here termed the probasidium. In others, karyogamy

and meiosis are separated in place, as well as in time of occurrence. Rogers (1934)

expresses the view, concerning ancestral types, that if both these events took place in

the primary cell (hypobasidium), four nuclei were available to thrust out the four

extensions of the wall which he terms epibasidia (e.g. in basidial types such as Tremella

and Tulasnella). On the other hand, if meiosis were delayed and took place in an

extension of the hypobasidium, there was only one nucleus to thrust at the wall, and

thus only one extension (epibasidium) would be formed (e.g. in basidial types such as

Helicogloea and other Auriculariaceae). Although it has been shown by Whelden

(1935 a, p. 52) and also by Bodman (1938) that the epibasidia of some heterobasidio-

mycetes may sometimes be initiated before meiosis of the fusion nucleus, Rogers’

explanation given above would certainly appear to be a good one connecting

morphological and cytological development in heterobasidia, at least in ancestral types.

But can this be used, as by Rogers, to homologise the epibasidium of Tremella with

that of Auricularia ? If so, can it ignore the fact that the sterigmata of the clavate

homobasidium are produced in the same way as the epibasidia of Tremella, yet receive

a different name? We think it cannot. On the above explanation Rogers (1934) states

that there are two fundamental phyletic tendencies among Basidiomycetes (a) those

which have epibasidial meiosis and ( b ) those with hypobasidial meiosis; but he adds

that Donk’s distinction between the part or stage of karyogamy (probasidium) and

the part or stage in which meiosis occurs (metabasidium) introduces a false complexity.

If, however, the phyletic tendencies and the morphological differentiation depend upon

the place of meiosis, _ is this not a fundamental character which should be used in

formulating a terminology?

To put the argument another way: Baker (1936) has stated, “ If karyogamy and

meiosis both take place in the hypobasidium, several epibasidia result; if meiosis is

delayed and takes place in an extension of the hypobasidium there can be only one

epibasidium ”. Again we do not question the correctness of this statement, but only

its interpretation for the purposes of a terminology. All that is implicit in this statement

forms the crux of the differences between the terminology of Neuhoff and followers,

and Donk. From Rogers’ and Baker’s statements above, the conclusion must

inevitably be that the form of the mature basidium is conditioned by the time and

place of meiosis, and not that the time and place of memsis are conditioned by the

form of the mature basidium. Yet Rogers accepts the second alternative by basing

his terminology on the morphology of the mature basidium, and rejects Donk’s

terminology which is based on developmental morphology arising from cytological

change.

Now, is it possible to homologise the epibasidia of the Tremella and Auricularia

types ? Emphatically no ! Taking in order the criteria of homology as defined previously,

we shall see how the two views clash, and which view best meets the case.

(A) To be homologous the structures must develop in the same way.

The epibasidia of Tremella and Auricularia both develop supposedly in response

to the thrust of nuclei and protoplasm upon the wall of the young basidial cell in which

karyogamy has occurred; but in Tremella the nucleus first divides so that there are

four daughter nuclei available to thrust out four extensions. In Auricularia the whole

protoplast and its single nucleus thrusts out a single extension and the nucleus then

254

divides. In Tremella the epibasidia bear spores directly upon their narrowed apices.

In Auricularia the epibasidium divides into four cells, and in each cell a nucleus is

present to thrust out from the epibasidial segments, terminal or lateral extensions on

whose narrowed apices the spores are borne. The development of the epibasidia is

certainly not the same in both cases.

(B) To be homologous the structures must serve the same function.

The epibasidia of Tremella merely serve to carry the nuclei to the spores, and if

necessary they may extend so that the spores are produced in the open. The epibasidium

of Auricularia serves as the place of meiosis, then from its cells are produced the

extensions which are necessary for nuclear migration to the spores. These extensions

are capable of elongation to meet the ecological need; not so the epibasidium itself

which remains fairly constant in length. Surely it is the extensions which are homologous

with the epibasidia of Tremella. It is naturally realised that extension of part or the

whole of the basidium is only one aspect of the variability of heterobasidiomycetes.

For example, in Vuilleminia it is the metabasidia which extend to enable the spores to

be produced at the surface. Taken alone, the functional aspect of extending epibasidia

is perhaps not important in homologising these structures, but in conjunction with

other facts, particularly morphological, it seems to have a bearing on the subject.

(C) The structures must have the same relation to a fundamental type.

Here there is a phyletic implication, which may be largely speculative in fungi,

but nevertheless cannot be ignored as it is the basis of our ideas on classification. We

therefore agree with Rogers that Basidiomycetes show two distinct phyletic tendencies

(a) those with epibasidial meiosis and ( b ) those with hypobasidial meiosis. This only

serves to strengthen the argument that on account of the different places of meiosis in

Tremella and Auricularia their epibasidia cannot be homologised.

(D) The structures must have the same relative position.

In Tremella and Auricularia, the structures which are called epibasidia are both

borne as extensions of the cell in which karyogamy takes place. However, the sterigmata

of clavate homobasidia are formed as extensions of a cell in which karyogamy takes

place in exactly the same way as in Tremella by upthrust of four daughter nuclei. On

this view the sterigmata must be homologues of the Tremella epibasidia, which Rogers

does not accept. In Auricularia although the epibasidium is borne as an extension

of the cell in which karyogamy takes place, it is not another homologue because meiosis

has not yet occurred. It is the terminal or lateral extensions of the epibasidial segments

in Auricularia which have the same relative position as the epibasidia of Tremella and

the sterigmata of higher Basidiomycetes.

Neuhoff’s terminology does not stand the tests of homology, which have been

defined here in most inclusive terms.

(2) Basidium. — Linder (1940) defines a basidium as “ that organ which is in part

the homologue of the ascus, and which following karyogamy and meiosis bears the

basidiospores either directly or through the interpolation of the sterigmata

It must be accepted, as previously stated, that the whole organ including any basal

vesicle or cell, any septate or non-septate extensions, and any spicules on which the

basidiospores are borne, must be included in the term basidium. This inclusive sense

of the term may be what Linder intended, but actually his definitions limit the term to

the part of the basidium left after subtracting probasidia and sterigmata. If this

interpretation is correct, then obviously Linder’s simplification does not suffice. For

descriptive taxonomy alone it would cause unwarranted circumlocution and confusion.

The whole mature organ must be called the basidium, and other names sought for

its parts. It appears difficult to settle upon terms which convey these distinctions

satisfactorily.' Neuhoff’s epibasidium implies the same as Linder that the spore-bearing

255

tubules are something “ upon the basidium ” and not a part of the basidium itself —

though this interpretation is not qualified by its application. Donk’s term metabasidium

may be translated as something which is “ the conclusion of the basidium ”, although

the metabasidium is also only part of the whole basidium. The original meaning of

the term probasidium was “ that which comes before the basidium ”, but now as an

antithesis to metabasidium it may be interpreted as “ the first stage of the basidium ”.

(3) Sterigmata. — Linder (1940) applies this term to “ that structure which arises

from the basidium and bears the basidiospores.” He continues, “ The restriction of

the sterigmata to apply only to that portion of the spore-bearing filament which is very

slender and through which the nucleus squeezes in its passage to the spore, not only

leads to confusion but is illogical.”

It has been put forward that there are two criteria of a true sterigma: —

(a) Neuhoff (1924) gives the cytological criterion that the nucleus in passing through

a true sterigma to a spore, becomes elongated and assumes the staining properties

of chromosome-material. All Neuhoff’s followers have accepted this as the

criterion of a sterigma. Quite obviously the nucleus must elongate to squeeze

through a narrow passage, but it has been argued that the spatial relation is

inconsequential because in many species the nucleus has been observed to

elongate and change its staining properties before it actually reaches the narrow

passage. This is taken to indicate that the area approaching the passage is

distinguished in some way other than narrowness from the remainder of the

tubular appendage, and should be named differently. Early attenuation of

the nucleus was observed in He/icogloea (Baker, 1936), Exidia (Whelden,

1935 a) and Tremella (Whelden, 1934), and may be accepted as fact. If, as

in Helicog/oea, elongation of the nucleus is initiated in the epibasidial segments

well before the nucleus approaches the narrow passage, or if elongation occurs

at a variable distance from the passage, what are to be regarded as the

morphological limits of the “ true sterigma ”? Where is the point where one

may say, “ Here the epibasidium ends, and here the sterigma begins? ” There

are a few exceptions, e.g. Xenolachne (Rogers, 1947) where there is a fairly

sharp distinction on account of the grossly attenuated, long sterigma, and also

in Tulasnella tulasnei (Rogers, 1932, pp. 95-96) where the sterigmata elongate

greatly as an adaptation to the presence of a gelatinous matrix round the

basidia. Even in these it is possible that the nucleus might elongate at a variable

distance from the morphological “ sterigma ”. It is obviously impossible to

set the limits of the “ true sterigma ” without a cytological investigation of

each and every basidium in development, and even then it may be found that

this point is variable in a single species, or even in different basidia of the same

specimen. Further, may not the change in staining reaction be due simply

to the physical fact of elongation, so that the nuclear components are spread

out and of necessity become coloured differently from an opaque and compact

body? Such an effect is frequently seen in staining blood smears, where the

staining reaction of the nuclei of various types of cells depends to an extent

upon the thinness of the smear.

(i b ) Buller’s (1922, p. 31) criterion of a sterigma is functional and not cytological,

viz. that a sterigma is an organ for bearing and forcibly discharging the spore.

True, forcible discharge is not universal. There are gymnocarpous Basidio-

mycetes in which forcible discharge is either impossible or of no advantage to

the fungus, and in these the sterigmata have lost their function, becoming

reduced to very short attachments or alternatively becoming disproportionately

long and filamentous. In either of these instances the attachment of the spore

is delicate and enhances the possibility of its being set free by fracture. In

other words this reduction or distortion is an adaptation to loss of function,

256

and must strengthen the opinion that the functional criterion of a sterigma

is important. Rogers (1947) has given a clear account of such basidia, terming

them apobasidia with the definition of an apobasidium as “ a basidium whose

basidiospores are not apiculate, nor borne obliquely on the sterigmata, nor

forcibly discharged

Further to the functional aspect, it has been argued that the epibasidium

is an adaptation to ensure that the migrating nucleus shall be transported

through varying depths of jelly in gelatinous Heterobasidiomycetes to the

surface, where a true sterigma may be formed in the open to serve its function

of discharging the spore freely and with the greatest possible chance of subsequent

dispersal. We do not question that this process occurs, but only that it serves

further to distinguish the parts called epibasidia and sterigmata by Neuhoff.

Considering the whole organ as designed to bear and liberate the spores under

the best possible conditions, why, from the fact that part of it is capable of

elongation to meet the environmental conditions, should that part be held a

separate entity? And further, it is true that in some instances [e.g. in Tulasnella

tulasnei (Rogers, 1932)] the other part, namely the narrow terminal pedicel

or “ true sterigma ” of Neuhoff, is also capable of extension when embedded

in a gelatinous matrix. Baker’s work (1936) on Helicogloea shows that if the

epibasidia are immersed each cell may develop a tubular extension as an

adaptation to the immersed condition, but protruding epibasidia form

“ sterigmata ” directly without the need for tubular extensions. In this case,

are the lateral extensions and the sterigmata to be held apart, or are they merely

variations of a single organ? Surely the second alternative! Again, in

Vuilleminia the whole basidium may elongate according to the depth at which

it is embedded in a semi-gelatinous matrix. Variation in length to meet the

environmental conditions is certainly not solely an attribute of epibasidia.

It occurs in epibasidia, and in sterigmata (both sensu Neuhoff) and also in

metabasidia (sensu Donk). Variation, as Rogers (1934) justly observes, is

one of the hall-marks of Heterobasidiomycetes, and is largely a response to the

ecological conditions and especially to the water factor. One comes to the

conclusion that the whole tubular appendage is a single entity in which the

nucleus migrates to the apex, elongates on approaching, but at no fixed distance

from the narrow passage with change in staining properties, and enters the

spore which is finally discharged. Conditional upon the type of basidium

and the depth at which it is embedded, the basal part or sometimes the narrower

apical part of the spore-bearing organ, or even the whole basidium, may

elongate to ensure that the spore is formed and receives its nucleus in the air,

so that it may be freely liberated for dispersal.

From the foregoing we can agree with Donk (1931) and Linder (1949) to call the

whole tubular spore-bearing appendage a sterigma. But there is still the somewhat

exceptional case of Tulasnella to consider. In the development of the Tulasnella

basidium there are two divergences from the usual state in most hererobasidia, namely

that the “ epibasidia ” are cut off by basal septa, and that the first post-meiotic mitosis

takes place in the “ epibasidia ”.

The ovoid form of the Tulasnella epibasidium is suggestive of a spore; this fact

has lead to its being considered as a sessile basidiospore which germinates in place to

form a conidium. Heim (1949) has suggested that the basidia of Tulasnella are similar

to teratological basidia of certain Homobasidiomycetes, in which the basidium produces

a prolongation terminating in a sporoid body, the latter liable in turn to produce a

conidium. Heim terms this prolongation (epibasidium sensu Neuhoff) a hemibasidio-

spore, and its terminal “ spore ” a basidioconidium, which can produce secondary

conidia. He suggests that the whole structure composed of hemibasidiospore and

basidioconidium, corresponds to a sessile basidiospore produced incompletely by an

257

accelerated sporogenetic rhythm. This is essentially a more modern version of Juel’s

(1897) interpretation of the Tulasnella basidium as one bearing sessile basidiospores

germinating in place to give secondary spores or conidia. Rogers (1932) summarised

Juel’s points for this interpretation and provided convincing counter-arguments to

show that Juel’s points are, on the whole, not antagonistic to Neuhoff’s conception of

this basidium. Donk (1931, p. 115) has also dismissed Juel’s interpretation mainly

on the grounds of similarity in the basidia of Tulasnella and Botryobasidium ( Pellicularia ).

This would perhaps be more correct if the similarities offered were as between Tulasnella

and Ceratobasidium (some species of which were formerly placed in Botryobasidium).

In Tulasnella (Rogers, 1932) following karyogamy and two meiotic divisions of the

fusion nucleus, the epibasidia are formed and are then cut off by a basal septum from

the rest of the basidium. In the epibasidia there follows the first post-meiotic mitosis,

after which one of the daughter nuclei migrates to the apical sporoid body. Rogers

(1932, p. 100) states, “ Mitosis within the appendage is not the impossible behaviour

of an extension of the basidium that it would be for a sterigma.” This statement should

at least be qualified by the definition accepted for a sterigma, for both epibasidia (sensu

Rogers) and sterigmata (in our sense) are extensions of the basidial cell. Certainly the

“ sterigma ” of Rogers (i.e. Donk’s spiculum) would be too confined a space in which

to expect mitosis. But is it correct to say that this mitosis within the tubular organ

really distinguishes this organ from homology with similar organs which apparently

only serve for migration of the nucleus and spore-bearing? If it is held that they ara

not homologous solely on this account, then equally well these tubes cannot all be

called epibasidia, nor sterigmata. On the other hand, if they are homologous then they

may be called either sterigmata or epibasidia according to choice of terminology. Taken

alone, the first post-meiotic mitosis within the tubes of Tulasnella cannot serve as an

argument for not calling these organs sterigmata.

Martin (1938) observes that the place of meiosis should not be regarded as a

fundamental basidial character as it may take place within the limits of the original

cell (pro- or hypobasidium) or within extensions of the cell wall (epibasidium). He

considers it a secondary character, conditioned at least in some instances by the nature

of the zeugite. If we took up this premise then it would be illogical to regard either

Donk’s probasidium (the place of karyogamy) or his metabasidium (the place of meiosis)

as fundamental. But we have already observed that the two fundamental phyletic

tendencies in Basidiomycetes are indicated by epibasidial meiosis and hypobasidial

meiosis, and that the morphological differentiation of the basidium is connected with

the place or stage of meiosis. That being so, this is most surely a fundamental basidial

character, which governs the final form of the basidium. The possibility that meiosis

may take place in one of two set places and thus be bound up with the production of

either one or several extensions of the primary basidial wall, points to this feature

as fundamental. But on the contrary the place of the first post-meiotic mitosis cannot

be considered fundamental, for it does not alter the established pattern of morphological

change. In Hymenomycetes it may occur either in the metabasidium, or in the sterigmata

(our sense) or in the spores, and is not bound to a fixed site. In Tulasnella (Rogers,

1932, p. 103) all three possibilities seem to occur. In the basidia with more than four

sterigmata the first post-meiotic division is likely to occur in the metabasidium; in

some species it is known to occur in the sterigmata; and in a few other species it is

unknown either in the metabasidium or in the sterigmata, hence likely in the spores.

But note that in all these instances the general pattern of the Tulasnella basidium is

unchanged. In several other genera the first post-meiotic mitosis occurs in the spores

(cfr. Maire, 1902). As this mitosis may occur in the metabasidia or in the spores,

there is no reason why it should not be expected in an intermediate organ, the sterigma,

especially if the last passes through an accumulative phase in which it achieves a certain

degree of independence from the rest of the basidium, and if it is spacious enough to

permit nuclear division to occur.

258

In the Clavulinaceae (Gaumann, 1928, p. 532) the mature basidia contain eight

nuclei as a result of a triple division of the diploid nucleus, the number of sterigmata

is variable, the spores each contain one nucleus, and a variable number of nuclei

degenerate within the body of the basidium. The position regarding the third division

of the fusion nucleus within the basidium is much the same in the Cantharellaceae

(Gaumann, 1928, pp. 533-534). Also in a species of Kordyana (Gaumann, 1928, p. 530)

the basidia may be the seat of extra (abnormal for the species) divisions under certain

weather conditions, producing extra nuclei for which extra sterigmata are formed.

The foregoing arguments, it is hoped, serve to show that the occurrence of the

first post-meiotic mitosis within the spore-bearing organ is not a valid argument for

not calling this organ a sterigma in Tulasnella. Now, what about the basal septation

of these organs in Tulasnella ? There seems no a priori reason why a sterigma on

receiving its protoplast and nucleus, should not become inflated and separated from

its parent cell by a septum, particularly if this represents an accumulative phase in

which the nucleus is about to divide. The fact is, however, that apart from Tulasnella

and certain irregular cases recorded in Ceratobasidium and in some Dacryomycetaceae

(Rogers, 1934, p. 170; Rogers, 1935, p. 4), septa are not known at the base of the

sterigmata. It is part of the Neuhoff school of thought that epibasidia may be

characterised by production of septa; it is not commonly a way of characterising a

sterigma, although it could be argued that just as a spore may become septate, so also

could a sterigma without justifying a change of name. The standpoint we adhere to

is that this is merely another aspect of the variability of heterobasidia particularly in

a primitive type coming near to its Ascomycete ancestors.

There is a good deal of agreement between many workers (cfr. Rogers, 1932,

1934; Gaumann, 1928; Heim, 1949) that the spore-bearing organ of Tulasnella-typs

may have originated from a type of Ascomycete showing ascospores thrusting out the

wall of the ascus and germinating exogenously by a conidium. On this hypothesis

Rogers (1934, p. 168) suggests that “the Tulasnella basidium, as highly organised as

any existing type, is phylogenetically closest of existent types to the antecedent ascus ”,

and further that the epibasidia are essentially homologous with ascospores. The

same explanation may be given for the origin of the sterigma, and is accepted as that

here; or it may also be given as the origin of the sporoid hemibasidiospore of Heim’s

interpretation. Allowing now for the primitive nature of the Tulasnella type of basidium,

or of its immediate ancestors in which the process of exogenous spore formation may

be only lightly established, what is more likely than that there should be variations in

this process? And what variation is more likely than one involving the degree of

independence of the ascospore (or the derivative “ epibasidium ”) from its parent cell,

i.e. the formation of septa?

The close affinity of Ceratobasidium (Rogers, 1935, p. 4; Martin, 1948, p. 114)

with Tulasnella makes possible either the derivation of the former from the latter type

of basidium, with loss of septation, or suggests an independent derivation of

Ceratobasidium from an Ascomycete type, in this case no septa appearing at the base

of the sterigmata. The fact that there are occasionally adventitious septa in

Ceratobasidium (Rogers, 1935) is highly suggestive that this process was once a normal

occurrence in a related type which has now been lost. Alternatively it may argue the

derivation of the Tulasnella type of basidium from that of Ceratobasidium. Whichever

way about, it shows that the septation is variable as would be expected in a primitive

type. From types such as Ceratobasidium sterigmaticum, with two inflated sterigmata,

(Rogers, 1935, 7, fig. 4) could be derived the Dacryomycete basidium which is known

to produce occasional adventitious septa (Rogers, 1934, p. 170) in the sterigmata, while

Pellicularia flavescens with inflated cornute sterigmata and repetitive spores (Rogers,

1935; Rogers, 1943, p. 105) provides a link between Ceratobasidium and other

Pellicularia species and hence to the Corticiaceae. From the inflated and septate sterigma

to the filamentous and non-septate sterigma there are a series of known types which

259

indicate the possible phylogenetic changes, and simultaneously there has been a change

from variability in these features to stability. In essentials the sterigma has remained

the same throughout, an organ for the migration of the nucleus to the spore and one

which bears and discharges the spore.

If the sterigmata are called epibasidia in Tulasnella and Ceratobasidium, but

sterigmata in Pellicularia (which includes species such as P. flavescens with sterigmata

no different from the Ceratobasidium epibasidia) one is bound to establish a special

family Ceratobasidiaceae (Martin, 1948, p. 114). If all these structures are sterigmata,

and they undoubtedly are all homologous, such families are at least suspect.

From the foregoing we accept the tubular sporebearing structure(s) in toto as a

sterigma. Donk has differentiated the apex of the sterigma by the term spiculum

(coined by Tulasne, 1853) and the lower part of the sterigma by the term protosterigma

(coined by Donk, and used here for the first time in publication with his kind permission).

This differentiation in the terminology of the parts of the sterigma is best understood

by referring to Dr. Donk’s own “ Note on Sterigmata in General ”. (See p. 301).

This interpretation of the sterigma may not be accepted by followers of Neuhoff’s

terminology. But again consider the case of an auricularious basidium like that

figured for Septobasidium in Figs. 1 and 3. Here the p?rt which we call a sterigma is

entirely homologous with the same extensions in a Tremella basidium. To us they are

both sterigmata. To Neuhoff those structures in Tremella are four one-celled epibasidia

(or “ not separate cells, but extensions of the hypobasidial segments ’’—Martin, 1938).

But in the auricularious basidium, organs having the same morphology, function and

cytology are either known as sterigmata (if they are short consequent upon being

produced on emergent basidia) or as unnamed lateral extensions of each cell of the

four-celled epibasidium (if they are elongated consequent upon being formed on

deeply immersed basidia). In the latter instance Donk has ironically suggested (in Litt.)

that it would be logical to call them epiepibasidia ! It cannot be controverted that

the Neuhoff School homologises the four epibasidia (sterigmata in our view) of

Tremella, with the single four-celled epibasidium and its lateral extensions or sterigmata

of the auricularious basidium. The point is that the sterigmata, in our sense, of the

auricularious basidium are regarded as part or extensions of the epibasidium by Neuhoff,

but as the whole epibasidium in tremellaceous basidia. To equate the part (in Tremella)

with the whole (in Auricularia ) is preposterous, and on this contention alone it is

impossible to accept Neuhoff’s terms epi- and hypo-basidium.

On the other hand, the four-celled parts of the Auricularia and Tremella basidia

are entirely homologous and should receive the same name. Both develop from a

probasidium in which karyogamy occurs. In both meiosis takes place, followed by

septation (differing only in the plane of septation) and production of a sterigmatic

outgrowth from each cell. That the plane of septation is different is of little importance,

as is shown by the fact that in various Tremellaceae marked variation in the plane

of septation may occur (Whelden, 1935 c).

Now, having rejected Neuhoff’s and Linder’s terminologies can we accept Donk’s?

We have already accepted his probasidium, and sterigma composed of protosterigma

and spiculum; there remains to consider his metabasidium, for whose acceptance

some arguments have already been given. The definition of a metabasidium as that

part or stage of the basidium in which meiosis occurs, is a cytological one. It may be

held that a definition based on cytology is impracticable in routine work on morphology

and taxonomy, and thus an undesirable, even though fundamentally true, definition.

It may be held that Neuhoff’s terms are simpler to apply, that the swollen bas'd part

of the basidium is the hypo-basidium and that everything above that is the epibasidium,

except the spore-bearing spicules, which are true sterigmata. In practice Donk’s

terminology is no more difficult to apply: a probasidium, whether as a resistant or

merely as a primary basidial cell can be recognised at once; so can the sterigmata or

260

tubular spore-bearing filaments; what is left is the metabasidium, which in some cases

quite obviously replaces the primary cell.

Now consider the basidia of Phleogena faginea (Shear & Dodge, 1925) and

Mycogloea carnosa (Olive, 1950), both illustrated in Fig. 3. In Phleogena there is a

basidial stipe bearing a cell in which successively karyogamy, meiosis, transverse

septation and production of lateral spore-bearing filaments occur. In Neuhoff’s

terminology, the whole basidium here must be an epibasidium, but this epibasidium

is not a structure seated “ upon a basidium ” or even upon a hypobasidium. It is

instead simply seated upon a basidial stipe, i.e. a basidium-bearing hypha. Donk’s

terminology, however, accounts for a probasidium replaced by a (four-celled)

metabasidium each cell of which bears a sterigma and spore. It appears to the writer

that the developmental connotation in Donk’s terminology is important, and that it is

insufficient to define a basidium only on its mature morphology.

Or again, in Mycogloea, there is a basidial stipe bearing a cell which is similarly

changed from a probasidium to a metabasidium with sterigmata. Again Neuhoff’s

terminology is quite inadequate as it allows only for the epibasidium.

Further interesting basidial types, which put terminology to the test of constancy

to the same principles, are those in which the basidia are somewhat intermediate between

other well-known basidial types such as characterise different families or even sub-classes.

Their intermediate nature allows actual or potential disagreement on their taxonomic

position. Is the terminology applied to one and the same part of the basidium to

change according to the family or sub-class in which the fungus is classified ? A good

example is Patouillardina ( Atractobasidium ), illustrated in Fig. 3, 10. The author of

this genus did not describe it adequately or quite correctly (Bresadola, 1906 and 1920,

p. 52), but appeared to place it in the Auriculariaceae. Martin (1935, 1939, 1945)

has classified it in the Tremellaceae on account of the “ regular perpendicular alignment

of the second basidial septa with reference to the first septum ”, but also noted that

the basidia are somewhat intermediate between those of these two families. Thus

Martin would consider that the basidium is composed of one four-celled hypobasidium,

each cell bearing an epibasidium. Suppose now a good reason were found for allying

Patouillardina with the Auriculariaceae; the basidium would still be exactly the same

but the terminology would be reversed. The basidium would then be composed of a

four-celled epibasidium, each cell of which bears an extension. In Donk’s terminology,

no matter which classification is accepted, the names of the parts of the basidium remain

constant, i.e. in Patouillardina there is a probasidium which develops into a four-celled

metabasidium, each cell bearing a sterigma.

Another example is Peniophora heterobasidioides (Rogers, 1935, p. 30-31). This

has affinities with both homo- and heterobasidiomycetes, but is classified as homobasidial

with sterigmata and repetitive spores. If the sterigmata had been a little more swollen,

can we not suppose that they would have been called epibasidia and the fungus classed

as a heterobasidiomycete ?

The genus Ceratobasidium is placed as a heterobasidiomycete with a primitive

type of holobasidium (Rogers, 1935, p. 4; cfr. Martin, 1948). This assignation is

probably correct, as indicated by the germination of the basidiospores by repetition.

But suppose the point were controversial and it could be held that these were

homobasidia. If they are heterobasidia, Rogers’ terminology provides for a

hypobasidium surmounted by a number of stout epibasidia. If they were homobasidial,

the same basidia would be said to consist of basidia with four sterigmata. Donk’s

terminology for both homo- and heterobasidia of this type is inflexible. For him the

basidia of Ceratobasidium consist of a metabasidium and four sterigmata, no matter

which sub-class is to receive the genus.

261

Similar arguments apply also to the basidia of Pellicularia ( Botryobasidium ), which

has already been touched on. Rogers (1935, 1943) classes some species of Pellicularia

(e.g. P. flavescens ) as homobasidial, with sterigmate basidia; yet these species have

stout, cornute sterigmata no different from those of Ceratobasidium, which Rogers

classes as heterobasidial, with hypo- and epibasidia. Ceratobasidium also links with

Tulasnella through such species as C. anceps in which the epibasidia are not cut off

by a basal septum.

To what Extent are Morphology and Cytology linked?

From the examples given above in various connections it is obvious that the probable

sites of karyogamy and meiosis can be inferred from the mature morphology of the

basidium, given the basic knowledge we already have of the cytology of a number of

basidial types. The place of the first post-meiotic mitosis, variations in the plane of

formation of the septa, and the plane of the mitotic spindles cannot generally be inferred.

One might suppose that generally in a broad ovate or clavate basidium there would be

sufficient space for the mitotic spindle to lie horizontally (i.e. chiastic division). This

does not always happen, for in such basidia stichic division, with the spindle placed

longitudinally, sometimes takes place. Also in a narrow, cylindrical basidium the

spindle might be supposed to lie longitudinally, but again this is not constant. The

orientation of the spindle is not even to be correlated with the class of the basidium,

for stichic and chiastic types may occur in either homo- or heterobasidiomycetes.

Examples of such variations in spindle orientation may be seen by reference to Fig. 2.

The plane of the spindle may not even be constant in a single species (e.g. in

Helicogloea lagerheimi. Baker, 1936; Sebacina globospora, Whelden, 1935 c). It is

not impossible that this might be due to sectioning technique, but it might just as well

be due to natural variation. Despite this, there is probably not one species sufficiently

investigated in this regard that cannot be definitely classed as either chiastic or stichic,

especially if in deciding this character one restricts oneself to the first division, which

has the longest spindle, and neglects the following ones. In many cases cytology helps

to underline differences in basidial shape which would otherwise be underrated or even

overlooked, or are difficult to express in terms of basidial morphology (cfr. Clavulina,

stichic and Clavaria, chiastic). However, where the character of spindle orientation

has been used as a primary basidial character (Juel, 1898; Maire, 1902; Gaumann,

1928) it has resulted in a quite unnatural grouping of genera. Its use in such instances

has been properly censured by Rogers (1934) and Lohwag (1937). Its best use is as

a secondary character for the separation of small groups, and certainly not as a primary

one for the linking of large groups.

Size and Shape of Basidia.

In the higher Hymenomycetes the hymenium is compact and resists lateral

expansion of the basidia, with the result that they tend to be clavate or cylindrical.

In simpler forms, which are often gelatinous or composed of loosely arranged tissues

with an irregular or discontinuous hymenium, the basidia can expand laterally and

assume many different shapes. Several of these types are illustrated in Fig. 3, which

exemplifies the strange shapes encountered particularly among heterobasidia. A

more complete collection of illustrations is not possible here, owing to the difficulty of

obtaining suitable material for drawing, or in obtaining permission to reproduce

published illustrations.

Basidia may arise singly at the apices of subhymenial hyphae, or in botryose

clusters which are perhaps best seen in the genus Pellicularia. Basidia ranging in size

from 3x7// ( Corticium galzini) to 25 x 210 p {Aleurodiscus amorphus ) have been

noted by the writer. To see basidia clearl), crush preparations of very thin sections

mounted in 5 per cent potassium hydroxide solution to which a very small drop of

1 per cent aqueous phloxine has been added (Martin, 1934) have been found very

satisfactory.

262

Fig. 3. — A selection of basidial types: —

1. Aleurodiscus sp. 2. Craterellus sinuosus. 3. Trechispora raduloides (after

Rogers, 1944). 4. Galzinia cymosa (after Rogers, 1944). 5. Botryobasidium

coronatum ( = Pellicularia pruinata; after Rogers, 1934). 6. Ceratobasidium

fibrillosum (after Martin, 1948). 7. Vuilleminia coraedens. 8. Sebacina megaspora

(after Martin, 1936). 9. Tulasnella violacea (after Rogers, 1934). 10. Patouil-

lardina cinerea (after Martin, 1935, as Attractobasidium corticioides. 11. Femsjonia

luteoalba. 12. Neotyphula guianensis (after Martin, 1948). 13. Sirobasidium

brefeldianum (after Coker, 1920). 14. Septobasidium burtii (after Couch, 1938)

15. Cystobasidium sebaceum (after Martin, 1939). 16. Helicobasidium purpureum.

17. Auricularia sp. 18. Eichleriella (the two figs, to the right after Bourdot &

Galzin, 1928; the one to the left, E. incarnata, after Bresadola, 1903). 19. Dacry-

omyces deliquescens. 20. Helicogloea lagerheimi (=Saccoblastia sebacea; after

Wakefield & Pearson, 1923). 21. Mycogloea carnosa (after Olive, 1950). 22. Puc-

cinia caricis (after Grove, 1913).

263

Summary of Terminology for Basidia Accepted here.

Basidium. — That organ of the Basidiomycetes which is partly the homologue of

the ascus, and which following karyogamy and meiosis bears the basidiospores either

directly or on extensions of the gonotocont wall, the sterigmata. The term is taken

to include the probasidia, metabasidia and sterigmata as parts of the whole basidium.

Probasidium. — That part or stage of the basidium in which karyogamy occurs,

i.e. the primary basidial cell. Included also in this term are the teleutospores of Rusts,

the chlamydospores of Smuts, and the more or less persistent or resistant cells in the

same stage of development in the Auriculariaceae. The term is intended to denote

the “ first stage of the basidium ” rather than “ that which precedes the basidium ”.

Metabasidium. — That part or stage of the basidium in which meiosis of the diploid

nucleus occurs. In many basidia it obviously replaces the probasidium. The term is

intended to denote the final stage of the basidium as an antithesis to the first stage, or

probasidium.

Sterigma. — That part of the basidium which comes between the metabasidium

and the basidiospores, or the elongations of the metabasidium through which the

nuclei migrate to the spores which are borne terminally. The sterigma is composed

of a basal, filamentous or inflated part called the protosterigma, and an apical point

called the spiculum on which the spore is borne. (See also pp. 301-302).

Holobasidium. — An unseptate basidium. It is most common in homobasidio-

mycetes, but is sometimes encountered among heterobasidiomycetes, e.g. in

Ceratobasidium and Dacryomycetaceae.

Phragmobasidium. — A basidium which is divided by septa. The metabasidium

is divided into a number of cells (usually four) by cruciate or parallel septa after meiosis,

or in some instances the sterigmata are separated from the metabasidium by basal

septa. The term is the antithesis of a holobasidium.

Heterobasidium. — A basidium of the Heterobasidiomycetes.

Homobasidium. — A basidium of the Homobasidiomycetes.

Heterobasidiomycete. — A basidiomycete in which there is a phragmobasidium,

or if the basidia are holobasidia then the sterigmata are differentiated as stout, subulate

or cornute structures associated with basidiospores which germinate by repetition.

The definition of a Heterobasidiomycete based on the presence of an epibasidium is

inherently unsound because, as we hope to have shown, the epibasidium itself is not

defined in such a way as to represent a constant entity. Among holobasidia such as

found in Dacryomyces or Ceratobasidium the “ epibasidium ” is recognised in practice

as a swollen, stout sterigma associated with other characters such as spore repetition,

or elongation of tne sterigma to meet the environmental conditions, which give the

clue to the heterobasidiomyceteous nature of the fungus. The concept of an

epibasidium is quite unnecessary for the definition of a Dacryomycete basidium.

Homobasidiomycete. — A Basidiomycete other than a Heterobasidiomycete. The

group is characterised by basidia which are not septate, do not possess stout, swollen

sterigmata, and which produce basidiospores which germinate directly to form a

myceh’um.

Texture.

Although the character of texture is a complex of several properties, and eludes

precise definition, it has sometimes been used successfully in the taxonomy of the

resupinate Hymenomycetes, for example by Bourdot & Galzin (1928) in the genera

Corticium and Peniophora. Texture is a physical character, but may be modified by

chemical changes. This occurs in matrices which become gelatinous, mucilaginous

264

or ceraceous, or which have mineral matter deposited in the tissues. In the higher

Hymenomycetes, especially those which are coriaceous, suberose or ligneous, the

consistency of the pileus and cuticle may be correlated with the type of hyphae and their

direction in a particular plane (Ames, 1913, p. 220; Lohwag, 1940). In resupinate

Hymenomycetes the texture depends mostly on the compactness of the hymenium

in relation to that of the intermediate and basal layers, but it is also influenced by the

thickness, and particularly the moistness of the fungus.

The texture may vary with the age of the fungus, and also with the relative rates

of tangential and radial growth. Thus the margin of a fungus often differs in texture

from the older parts, owing to a difference in compactness of growth. Duportella

tristicula when young, is velutinate owing to the presence at the surface of numerous

upright hyphae; later the hymenium forms above these hyphae and the fungus becomes

membranous or coriaceous, with a rather waxy hymenium. Similarly, purely superficial

modifications of texture may occur in the genera Peniophora and Hymenochaete by

late production of cystidia and setae respectively. These often impart a glitter to the

surface, and feel scabrid if they are sufficiently robust.

The texture may be modified by chemical change. Ceraceous fungi appear waxy.

The tissues are usually compacted by a waxy substance which may obliterate the hyphae

or render them indistinct, giving the impression of degeneration. The wax often forms

an obstructive emulsion in alkaline mounts of sections. Gloeocystidia filled with oily

globules can also give the fungus a ceraceous appearance. This state, where the waxy

or oily substance is intracellular, contrasts with the commoner one where it invests

the outside of the cells. The latter condition is particularly common in the genera

Acia and Plilebia, but is by no means confined to those genera.

The hyphae of gelatinous species (Fig. 4, 1) often appear filamentous, only the

lumen showing distinctly with stains like phloxine. Gelatinised hyphae are best stained

with iodine or with ammoniacal aqueous methylene blue (Nannfeldt, 1947, p. 332),

which reveals that the hyphae are frequently loosely intertexed but held together by a

gelatinous substance. This has the power of imbibing water, in contrast with

non-gelatinous matrices which absorb water by capillarity (Buller, 1922, p. 157). The

gelatinous modification is characteristic of the majority of common Heterobasidio-

mycetes, where it would seem to be an ecological adaptation for securing quick

imbibition and retention of water. It may also be regarded as the author of the

heterobasidium, which requires a spacious hymenium for its development.

Mucilaginous species have much the same construction as gelatinous ones, but

differ in having a matrix of thinner, slimy consistency.

A corneous texture usually results from the loss of water from tissues which were

previously ceraceous, gelatinous or mucilaginous. This is exemplified in the genera

Tremella and Auricu/aria, and in such species as Peniophora gigantea.

For taxonomic work on the lower Hymenomycetes, the types of texture described

below are important: —

(1) Membranous (Fig. 4, 5). — The hymenium is a compact palisade but not

otherwise sharply differentiated from the subhymenial and deeper tissues, which

are usually also fairly compact. The fructification may be adnate, separable,

or held to the substratum by superficial intrusive hyphae. This group contains

most of the species with thick, moist hymenophores.

(2) Pellicular (Fig. 4, 4). — The hymenium is sharply differentiated as a thin crust

lying above loosely arranged inferior tissues. The hymenium is continuous and

compact at maturity, but at times may be discontinuous in patches thus appearing

minutely poroid under the lens. Pellicular species are generally rather thin and

265

Fig. 4. — Diagrams of sections of fructifications of resupinate Hymenomycetes

illustrating the arrangement of tissues in the following textural types: —

1. Gelatinous. 2. Arachnoid. 3. Byssoid. 4. Pellicular. 5. Membranous.

6. Crustose.

266

dry; the open nature of the tissues is doubtless responsible for quick drying

out. The hyphae are usually very distinct. Most of these species are separable

from the substratum by snapping of the basal tissues under tension.

(3) Byssoid or Floccose (Fig. 4, 3). — Here there are no compact tissues, even the

hymenium being irregular, discontinuous, and often staged at different levels.

A fairly thick, tufted or woolly structure results, in which the hyphae are loc sely

arranged throughout. The hyphae are easily detached from the substratum

by breaking under tension.

(4) Arachnoid (Fig. 4, 2). — This type cf texture is common in the young stages of

growth but may persist into maturity, and then the hymenial elements are quite

scattered and discontinuous. The hyphae are usually delicate (though not in

the genus Pellicularia), and more or less adpressed to the substratum in an

openwork structure reminiscent of cobwebs. When the elements of the fungus

are so dry and discontinuous as to appear scurfy or mealy, then the terms

furfuraceous or farinaceous are applicable; in a finer and more powdery state

the structure may be termed pruinose or pulverulent.

(5) Crustose or Arid (Fig. 4, 6). — Certain resupinate fungi are thin, dry, and

compact througout their tissues, and lack a distinct cortex. They are con-

veniently termed crustose. Usually they are adnate; frequently they are

charged with mineral matter (e.g. Aleurodiscus acerinus ). But it is more

common for fungi to be arescent, that is becoming crustose only on drying,

a character well shown in the section Coloratae (Bourdot & Galzin, 1928) of

the genus Peniophora.

The successful use of texture as a taxonomic character is much limited by the

difficulty in defining the types of texture unambiguously. With this in mind, Table 1

has been prepared so that the usage adopted by the writer may be checked by direct

reference to material of several species differing in texture.

Margin.

The type of margin depends typically on the compactness of the hyphae composing

it, and the direction in which they are intertwined. It may readily be observed that

as a resupinate fructification grows, new hyphae and young hymenial elements are

differentiated behind the radially expanding periphery. Radial growth generally precedes

the tangential growth which is responsible for knitting the hyphae together into a firm

fructification. It is thus probable that the presence or absence of marginal hyphal

strands depends largely upon the relative rates of radial and tangential growth. The

marginal character is fairly constant, and useful in taxonomy. The more important

types of margin (illustrated in Fig. 5) are fibrillose, fimbriate, arachnoid (byssoid),

villose or pubescent, farinaceous, and pruinose.

Hyphae may be aggregated into thick mycelial strands which sometimes meander

over the surface of the fructification or extend beyond the margin [e.g. the cords of

hyphae found in Coniophora arida, and especially in the section Radicatae (Bourdot

& Galzin, 1928) of the genus Peniophora ]. Certain of the lower Hymenomycetes also

develop rhizomorphs (e.g. Asterostroma rhizomorpha, Septobasidium bagliettoanum,

Helicobasidium compactum). Absence of suitable material has precluded a study of

these, but it would be interesting to know how far these thick mycelial cords approach

the degree of differentiation of tissues seen in the rhizomorphs of some Agaricaceae

[e.g. Armillaria mellea, where conducting vessels and hyphae are enclosed in a protective

sclerotic rind ; Marasmius species which form the rhizomorphs known as “ horse-hair

blights ” (Petch, 1915)].

267

Table 1 .

SOME SPECIES OF FUNGI WITH THEIR CHARACTERISTIC TEXTURE (MARKED O).

THE PRINCIPAL TEXTURE MAY BE MODIFIED BY CHEMICAL OR PHYSICAL CHANGE

TO OTHER TEXTURES (MARKED X).

Texture, or modifications of the principal texture.

268

Sclerotia.

Sclerotia are stromatic aggregations of hyphae, sometimes as big as a mustard

seed, with a compact rind and a mass of pseudoparenchyma within. They occur

especially among some of the parasitic species of lower Hymenomycetes, which are

pellicular or pruinose and liable to dry out, e.g. Pellicularia filamentosa (Rogers, 1943)

and Ceratobasidium anceps (sub Corticium, Gregor, 1932;. Probably owing to their

hard rinds these sclerotia are able to withstand dry conditions. When present they

are an aid to specific diagnosis, but they are not invariably found in any one species.

Relation of hyphae to the substratum.

In his studies of the Thelephoraceae, Burt (1925, 1926) stressed the character of

separability of the fungus from the substratum. The writer has examined the reasons

why some fungi adhere firmly to the substratum, while others are easily separable.

The property of adherence is not entirely an attribute of the fungus, but also derives

from the nature of the substratum, and is thus not constant for the same species of

fungus. The conclusion is that this character should not be stressed in taxonomy,

though it is very interesting biologically. The following are some of the reasons for

this conclusion:—

(1) Species which are usually separable, may be adnate or separable in small

pieces only, when growing on rough bark into the interstices of which the

hyphae may penetrate and there form a compact mass.

(2) Separation may be achieved either by snapping of the basal hyphae under

tension, or by carrying away small pieces of the substratum. The character

is thus dependent on the relative tensile strengths of the hyphae and the

substratum, and in turn upon the state of decomposition of the substratum.

(3) To some extent separability depends upon the composition of the basal tissues

of a fungus, which will distrupt more easily if they are loosely interwoven

even though such hyphae may be attached to the substratum by numerous

intrusive hyphae spread over their entire under surface. There are indeed

few resupinate fungi (e.g. Aleurodiscus amorphus and A. disciformis) which

radiate from their primordium without putting down an extensive system of

intrusive hyphae into the substratum.

Fig. 5. — Diagrams illustrating terms used in describing margins of lower Hymen-

omycetes : —

1. Fibrillose. 2. Fimbriate. 3. Arachnoid or Byssoid. 4. Villose or pubescent.

5. Farinaceous. 6. Pruinose.

269

The question of separability is closely linked with the way in which a resupinate

fungus develops from its primordium. One may imagine a fallen twig becoming infected

by a fungus. The sequence of events after infection must frequently be as follows:—

The hyphae first ramify through the subcortical layers of the twig then

emerge through cracks or lenticels to produce hymenophores. This is supported

by the observation that the wood is frequently permeated by hyphae, and that

the primordia of the fructifications develop usually above lenticels or cracks

in a hard substratum, such situations being paths of little resistance to emergent

hyphae. That these hyphae are emergent is deduced from the simultaneous

appearance of several primordia which later coalesce (e.g. in Peniophora cinerea,

P. quercina, P. aurantiaca) ; for it is improbable, knowing the low germination

rate of basidiospores in nature (Buller, 1909), that the tufts of mycelium at these

foci are each the product of separate spores which have lodged in lenticels or

cracks, achieved simultaneous germination, and sent down hyphae into the

substratum. In some other species, e.g. the Aleurodiscus spp. mentioned

above, there is often only one primordium and consequently only one deep-

seated mass of hyphae concerned in the nutrition of the hymenophore. Sec-

tioning of many fructifications which appear to have arisen from only one

primordium and then spread radially, will often reveal that the hyphae lying

behind the growing edge may penetrate into the body of the substratum. This

effect of secondary penetration by the hyphae is sometimes concerned in giving

rise to the adnate condition of resupinate fungi. From a biological point of

view, the adnate condition may be supposed to confer an advantage on the

fungus by assuring a supply of nutrients and moisture from a wider area of

the substratum, and by protecting the fungus from easy dislodgment.

New concepts in studying micromorphology.

Corner (1947, 1948a) has broken new ground in the study of micromorphology

of the Basidiomycetes by introducing mathematical expressions derived from careful

measurements of a great number of basidiospores, basidia and cystidia. These expres-

sions appear to be of general application. They relate the length and width of these

organs, and the volume and spore number of the basidia and basidiospores.

The ratio D fd (where D is the length, and d the width of the basidiospore) expresses

the shape of the spore. This ratio plotted against D gives a locus which is called a

“ sporograph ”. Corner shows that a locus, and not a point determined by the average

range of D and d, is necessary in studying the size and shape of basidiospores, i.e. that

a developmental analysis is necessary to understand the adult structure. For example,

in certain C/avaria species the sporograph clearly relates 5-5 p, subglobose spores

with 10-17 p cylindric-fusiform spores. It shows that “ giant spores ” may not be

freaks, but in certain species are spores of the same kind as normal ones only produced

on basidia of greater volume than usual. It shows that spores of different kinds may

have the same size and shape under certain circumstances. It also shows that in some

species at least, micro- and macrc-spores are inherently different kinds of spores.

Corner also derived the more general expression D = d(a + hD), where a and b

are constants for the species. The values for a and b define different loci for different

kinds of spores. A similar relationship holds for basidia, viz. / = w(a + bl), where /

is the length of the basidium, w is its width, and a and b are constants for the species.

This curve is termed a “ basidiograph ”. The same expression holds for certain types

of cystidia which were studied, and Corner proved that some types of cystidia were

sterile, precocious and overgrown basidia.

Further studies (Corner, 1948a) on the basidium show that it may be regarded

as a self-charged ampoule, containing dense protoplasm which is injected into the

basidiospores by enlargement of a basal vacuole acting as a plunger. He established

270

Fig. 6. — A selection of basidiospores

1. Corticium galzini. 2. Corticium microsporum. 3. Peniophora byssoidea.

4. Poria viridans. 5. Peniophora subalutacea. 6. Cytidia flocculenta. 7. Corticium

tulasnelloideum. 8. Corticium lacteum. 9. Corticium niveocremeum. 10. Cor-

271

the mathematical expression V — A = nv, where V is the volume of the basidium,

A is the volume of the initial vacuole, v is the volume of the spore, and n is the spore

number.

Mathematical expressions such as these are essential for an analysis of the develop-

ment of these organs, and may be used very successfully to determine whether certain

morphologically dissimilar organs are actually homologous. This new concept in the

study of micromorphology has been applied with striking success by Corner to a few

carefully studied species.

Basidiospores.

The basidiospores are borne singly at the apex of each sterigma, or they are rarely

sessile in some Gasteromycetes. There are usually two to four sterigmata per basidium,

but some species have a corona of 6-8 or more, which is particularly characteristic of

the Urnigera section (Bourdot & Galzin, 1928) of the genus Corticium, that is the

genus Trechispora Karst emend. Rogers (Rogers, 1944).

A selection of basidiospores is illustrated in Fig. 6. In shape they may be subglobose,

ovate, oblong, cylindrical, elliptical, navicular, reniform, allantoid, cordate or pyrami-

dical. They frequently have one side depressed and their shape may be modified by

an apiculus. They may be hyaline or coloured a light brown, yellowish-brown, sepia,

or less commonly brightly coloured in a mass, such colours usually being rosy,

violaceous or blue-green. Often the spores are guttulate. Chlamydospores may be

developed in the basidiospores of Jaapia argillacea (Fig. 7, 4). These are lightly

coloured, and form in such a way that the ends of the basidiospore are distinct and

colourless, thus this structure has sometimes been regarded as a guttulate basidiospore.

Rogers (1935, pp. 29-30) disposes of that interpretation effectively. The majority of

spores are smoothwalled, but in some species they are minutely sculptured, or in others

bear large spines, warts or reticulations. Sculpturing is frequently accompanied by a

slight thickening of the spore wall and also by coloration. Taken alone, sculpturing

and coloration are not good generic characters, though they have often been misused

as such. Miss Wakefield (1935) cites several examples of this misuse.

The range of spore size is large. To quote the two examples used in connection

with basidia, Corticium ga/zini has spores measuring 1 ■ 5-2 ■ 5 x 3-4 • 5 p, while those

of Aleurodiscus amorphus measure about 22-5 X 29 • 5 p excluding the spiny outgrowths.

The spores of the latter species give an amyloid reaction, turning blue in the presence

of Melzer’s chloral-iodine solution. This reaction is given by many species and is a

useful aid in diagnosis.

ticium centrifugum. 11. Corticium calceum. 12. Corticium laeve. 13. Corticium

salmonicolor. 14. Cyphella alboviolascens. 15. Cyphella punctiformis. 16. Corti-

cium karstenii. 17. Corticium confluens. 18. Corticium laetum. 19. Corticium

polygonium. 20. Vuilleminia comedens. 21. Corticium aurora. 22. Septobasidium

bogoriense. 23. Auricularia auricula-judae. 24. Epithele typhae. 25. Jaapia

argillacea (after Wakefield & Pearson, 1920). 26. Cyphella pelargonii. 27. Ecchyna

faginea. 28. Corticium vellereum. 29. Corticium albostramineum. 30. Tomentella

echinospora. 31. Corticium polyporoideum. 32 34. Tomentella spp. 35. Corticium

trigonospermum (after Wakefield & Pearson, 1913). 36. Pistillaria cardiospora.

37. Hypochnella violacea. 38. Pellicularia flavescens. 39. Cyphella sp. near

C. citrispora. 40. Merulius lacrymans. 41. Phaeocyphella galeata. 42. Dacry-

opinax spathularia. 43. Platygloea effusa. 44. Dacryomyces deliquescens. 45. Heli-

cobasidium compactum. 46. Eichleriella spinulosa. 47. Femsjonia luteoalba.

48. Coniophora arida. 49. Aleurodiscus aurantius. 50. Aleurodiscus oakesii.

51. Aleurodiscus amorphus. 52. Aleurodiscus mirabilis. With the exception of

Nos. 7, 24, 25, 35, the spores are all drawn to the scale shown.

272

The spores of homobasidiomycetes germinate directly to give a mycelium, and

are never septate. Those of heterobasidiomycetes may become septate (Fig. 6, 44, 47)

and germinate by putting out a small process termed a secondary sterigma, from which

arise in series one or more secondary spores (Fig. 6, 43, 46) which later give rise to the

mycelium. This sequence of events was called “ germination par renovation ” by

Patouillard. It is perhaps best translated as “ germination by repetition ” (Rogers,

1933, p. 183) but has also been known as “ renovation ” (Donk, 1931, in translation).

The secondary spores are the same shape as the original basidiospore, but smaller.

In Exobasidium and Dacryomyces, sessile secondary spores, or conidia, are formed

by gemmation from the sides of the basidiospore, either singly or in chains. This

process should not be confused with germination by repetition.

Donk’s term ballistospore is sometimes used to designate “ those spores of Basidio-

mycetes which are actively projected at maturity according to the specific mechanism

studied and amply described by Buller in his Researches on Fungi. Ballistospores are

found in Basidiomycetes with exposed hymenia, Uredinales, Tilletiaceae and Sporo-

bolomycetaceae (comprising the genus Tilletiopsis) ” (Derx, 1948, p. 468, translated;

see also Nyland, 1949).

Chlamydospores.

Though they are uncommon in the lower Hymenomycetes, chlamydospores are

present in some species and should not be overlooked. They may be defined as large,

asexual, resting spores, characterised by a thick, refractile wall and dense lipid contents.

They bccur in a terminal or intercalary position on the hyphae. The spore wall may be

smooth or variously sculptured.

Chlamydospores are found in cultures of many wood-rotting Basidiomycetes,

e.g. Trametes serialis group (Fig. 7, 3). Good examples may also be seen in the genus

Nyctalis, one of the parasitic Agaricaceae. In N. parasitica the chlamydospores are

smooth, and occur in the subhymenium (Fig. 7, 2). In N. asterophora they are stellate

and occur in the tissue of the upper part of the pileus (Fig. 7, 1). This interesting

genus has been well investigated by Buller (1924), Thompson (1936) and Ingold (1940).

In Sebacina epigaea (see McGuire, 1941, PI. 2, figs. 18-19) the basidiospores may

be transformed into resting spores resembling the chlamydospores of N. asterophora.

In Corticium laetum intercalary chlamydospores are sometimes found.

Fig. 7. — Chlamydospores: —

1. Nyctalis asterophora. 2. Nyctalis parasitica. 3. in culture of Trametes

serialis group. 4. basidiospores of Jaapia argillacea with internal chlamydospores,

the left-hand spore diagrammatic and not drawn to scale shown.

273

The basidiospores of Jaapia argillacea (Fig. 7, 4) may develop to contain a large,

thickwalled, yellowish chlamydospore separated from the poles of the basidiospore

by septa. Accounts of these structures have been given by Wakefield & Pearson (1920)

and by Rogers (1935, PI. iii, fig. 14).

Hyphae.

The hyphae of the lower Hymenomycetes may be hyaline or coloured, thick- or

thin-walled, with or without clamp connections, loosely or compactly arranged, with

frequent or infrequent septa, and a variable degree of branching. Hyphae tend to

branch most just beneath the hymenium. The hyphae range in width from a minimum

of about 1 -5 p in some species to a maximum of 12— 18— (29) p in species of the genus

Pellicularia, e.g. P. pruinata (Bres.) Rogers ex Linder. The walls of gelatinised hyphae

do not show distinctly except by special staining, and normally one sees only their

filamentous, refractile lumen. The gelatinised walls may be stained with strongly

ammoniacal methylene blue (Nannfeldt, 1947).

The colour of the hyphae appears to be significantly related to other structural

characters, particularly to the development of setae and asterosetae (see p. 291).

Recently great strides have been made in classifying the types of hyphae encountered

in the Polyporaceae (Corner, 1932 a; Corner, 1932 b; Cunningham, 1946). It has

been shown that a similar classification may be applied to the hyphae of the Thele-

phoraceae (Banerjee, 1942). Corner (1932 a) defined four types of hyphae, viz.

generative, skeletal, binding and mediate hyphae. In a second paper (1932 b) he defined

the systems into which these types could be combined, namely monomitic, dimitic and

trimitic systems, to form tissues. Cunningham (1946) extended this work and stan-

dardised the technique for examining hyphae. Largely on this basis, Cunningham

(1947-1950) has classified and described the polypores of New Zealand.

Hyphal types.

(1) Generative hyphae (Fig. 8, 1, 2): These are branched, thin-walled, septate*

hyaline or coloured, with or without clamp connections, usually staining readily, and

varying in width from about 1-5-10 p.

(2) Skeletal Hyphae (Fig. 8, 3, 4): These are of two general types, which

Cunningham distinguishes as “ bovista ” and “ long ” types. The bovista type has a

main axis 3-10 p in width, with several lateral branches which may in turn branch

and taper. They are usually aseptate, and may be hyaline or coloured. The long type

lacks the complexity of the bovista, and is slender, 3-7 p in diameter, septate or aseptate,

branched or typically unbranched, usually thick- but sometimes thin-walled, hyaline

or coloured and loosely interwoven.

(3) Binding hyphae (Fig. 8, 5): These also occur in the bovista and long forms.

They are aseptate, interweaving, branched or sparingly branched, and thickwalled.

As their name suggests they serve to consolidate the hyphae into firm tissues.

(4) Mediate hyphae: For practical purposes these need not be recognised. They

are intermediate between generative and skeletal, and generative and binding hyphae,

forming the transition from one to another of these types.

So far as the writer has been able to ascertain, “ bovista ” types of hyphae have

not yet been found in the Thelephoraceae, though well known in the Polyporaceae.

The “ long ” type of skeletal hypha is readily distinguished; not so the “ long ” binding

hypha.

274

Fig. 8. — Hyphal types: — A

1 . Generative hyphae from Polyporus angolensis. 2. Generative hyphae from

Trametes serpens. 3. Skeletal hyphae of the “ bovista ” type from Polyporus

sulphureus. 4. Skeletal hyphae of the “ long ” type from Polystictus pinsitus.

5. Binding hyphae of the “ bovista “ type from Polystictus pinsitus.

Hyphal systems.

(1) Monomitic: In this system there are only generative hyphae.

(2) Dimitic: The system is composed of generative and skeletal hyphae.

(3) Trimitic: Three hyphal types, generative, skeletal and binding hyphae, are

present.

Cunningham (1946) has made several useful generalisations which help in deciding

the kind of hyphal system present in a specimen. Thus : If all the hyphae are of one

kind, or all possess clamp connections, the system is monomitic; if clamps are present

in species with coloured hyphae of more than one type, the system is trimitic; clamps

are absent from species with coloured hyphae and a dimitic system; clamps are usually

present in the generative hyphae of hyaline dimitic systems. Cunningham has shown

that the presence or absence of clamps is significantly connected with the hyphal types

and systems.

Some clamp connections are illustrated in Fig. 8, 1, 2 and Fig. 9. The cytology

of clamp connection formation is reviewed by Rogers (1936), where it is shown that

clamps may be concerned in basidial proliferation as well as in a reproductive function

275

in the byphae where they serve to “ increase indefinitely the number of dikaryotic cells

from which may be formed gonotoconts Rogers also shows that the clamp is a

homologue of the crozier of Ascomycetes, and gives evidence of the derivation of

Basidiomycetes from Ascomycetes.

1. Stereum schomburgkii. 2. Pellicularia subcoronata. 3. Odontia arguta.

4. Corticium armeniacum. 5. Stereum bicolor. 6. Corticium gloeosporum.

Hyphal pegs.

These are erect, wartlike, papillate, tubercular or spiny structures projecting from

the hymenium and often big enough to see with the naked eye. They are composed

of sterile hyphae arising in coherent fascicles from the subhymenial or deeper tramal

tissue. The thelephoroid forms bearing hyphal pegs might mistakenly be classed in

the Hydnaceae without checking that the pegs are sterile.

The term “ hyphal peg ” has been commonly applied only to the Polyporaceae,

in which group the sterile tufts of hyphae project into the lumen of the tubes (Fig. 10, 1).

Overholts (1929) commented that hyphal pegs were never found in the genera Fomes,

Daedalea and Lenzites, nor in species with brown context, nor in the soft, white species

of Polyporus, nor in true species of Poria. He found that they were common in

Polystictus, were present in some species of Favolus, and in Trametes serpens, and

were very large in the genus Mycobonia. A common South African species, Trametes

meyenii, bears abundant hyphal pegs in most collections. Bose (1944) noted that

hyphal pegs, common in certain species at high altitudes, may be scarce or absent in

the same species at low altitudes.

As was done by Rogers (1935, p. 25) and Banerjee (1942), it seems desirable to

extend the term “ hyphal pegs ” to include the strictly comparable structures found

not only in the Polypores but also in the Thelephoraceae. In the latter group they have

276

been variously known as “ sterile emergences ” (Patouillard), “ emergent fascicles of

hyphae ” (Burt, 1919), or “ cystidiform synnemata ” (Langeron, 1945). As a synnema

implies a fructification it should be avoided as a term for these sterile aggregations

of hyphae.

Sometimes the presence of hyphal pegs has been the basis for splitting off new

genera (e.g. Dendrothele separated from Aleurodiscus : cfr. Rogers, 1935, p. 25); at

other times they are regarded only as useful confirmatory specific characters. Generic

distinctions involving hyphal pegs have been based on their hyphal composition and

upon their arrangement either as discrete spines or, extending the conception, as

confluent into sterile linear ridges or shallow poroid networks. To illustrate the use of

hyphal pegs in taxonomy some of the series which are apparent are set out in key form

below, but in doing so no close affinity of genera is implied: —

A. — Hyphal pegs discrete.

1. Pegs formed of coloured hyphae. Fructification dimidiate or sessile, but Veluticeps.

not resupinate

2. Pegs formed of colourless hyphae.

(A) Basidia tremelloid; texture sometimes subgelatinous Heterochaete.

(B) Basidia claviform.

(a) Pegs composed of delicate, much branched and inter- Dendrothele

woven dendrophytic hyphae, which are also present in the (Fig. 10, 3).

hymenium.

( b ) Pegs composed of delicate ordinary hyphae, little

branched or slightly forked, and arranged more or less

parallelly.

(x) Fructifications resupinate, lacking a well devel- Epithele

oped intermediate layer between substratum and (Fig. 10, 2).

hymenium

(y) Fructifications pileate, sessile or shortly stipitate, Mycobonia.

or if resupinate then having a broad intermediate

layer

B. — Hyphal pegs united linearly Grammothele

(see below).

C. — Hyphal pegs united to form shallow networks Hymenogramme and

Porogramme (see

below).

Fig. 10. — Hyphal pegs from:—

1. Trametes meyenii, inset showing a cross section of the pores. 2. Epithele

typhae. 3. Dendrothele (after Hohnel & Litschauer, 1907). All diagrammatic.

277

Veluticeps was regarded as a synonym of Hymenochaete by Killerman (fide

Ainsworth & Bisby, 1945), but is used by Burt (1919, p. 259) in the sense given above,

and is distinguished by him from Mycobonia by its coloured pegs. Malencon’s (1939)

description and illustrations of Veluticeps heimii show clearly that it is structurally

quite distinct from Hymenochaete, a fact which Patouillard (1894) correctly observed.

One may imagine hyphal pegs to be united laterally so that sterile ridges, or pores

surrounded by sterile walls, are formed over the surface of the hymenium. Two genera,

which when mature have a hymenium covered with small, sterile, poroid ridges, are

Hymenogramme and Porogramme. Patouillard (1899) states that Porogramme differs

from Hymenogramme in having pores which are isodiametric instead of elongated as

in the latter. They are genera which are insufficiently known to pass further comment

on their structure. Lloyd (1923, p. 1232) has suggested that Hymenogramme should

be united with Grammothe/e.

The genus Grammothele, though still imperfectly understood, requires comment.

The hymenium is covered with ridges which are sometimes quite regular and linearly

arranged, though not necessarily continuous (e.g. in G. liniata and G. polygramma),

or sometimes more tubercular and irregular (e.g. in G. mappa, G. pseudomappa and

G. cineracea). Through these ridges there project erect sheaves of hyphae surrounded

and capped by large aggregations of minerals which cause a distinct glitter. These

hyphae are very dilutely coloured and roughly parallelly arranged, not branched, and

together with their sheaf of minerals form a type of hyphal peg resembling a very large

cystidium at first sight (Fig. 16, 12). In G. cineracea these sheaves are branched and

the branchlets terminate in mineral caps. The hyphal pegs in G. mappa are more

often immersed than emergent.

The genus Gloiothele (Bresadola, 1920), founded on Poria lamellosa P. Henn.,

may be defined as a Grammothele possessing gloeocystidia. The conformation of the

hymenium resembles that of Heterochaete, but there are holobasidia and a crystal

incrustation of the hyphal pegs as found in Grammothele.

Ampoule hyphae.

In some lower Hymenomycetes, e.g. Grandinia farinacea and species placed by

Bourdot & Galzin (1928) in the section Humicola of tne genus Corticium, a number

of hyphae become swollen into shapes resembling ampoules (Fig. 11). These are not

always easy to find, and appear to be less obvious in material which is not fresh.

Ampoules are generally associated with clamp connections, which here and there

become much enlarged as the hyphae on one or both sides of the clamps become

inflated and bulbous. When ampoules develop on both sides of a clamp, they meet

obliquely in the middle. Sometimes the ampoule is viewed from a position where no

sign of a clamp is present (Fig. 11, 3). Such structures seen alone might be confused

with intercalary vesicles in the sense used elsewhere in this paper.

Fig. 11. — Ampoule hyphae drawn diagrammatically from: —

1-3. Corticium confine. 4. Corticium fastidiosum.

278

The species of Corticium grouped in the section Humicola are not related solely

by the possession of ampoule hyphae. They have other characters in common which

suggests that the grouping is a natural one, e.g. a rather pellicular or farinaceous

hymenium bearing veinlike mycelial strands or granules, and spores which are usually

sculptured.

Conducting hypbae.

In many Hymenomycetes the context contains, in addition to the usual hyphal

systems, a network of special hyphae with dense contents, which are known as

conducting or laticiferous hyphae. In their best-developed form they are seen to be

of indeterminate length, much branched, anastomosing, of wide diameter, relatively

thinwalled, at first non-septate and coenocytic but later becoming sparsely septate.

The dense sap may be hyaline, milky or brightly coloured, and in some species changes

colour on exposure to light and air. Such hyphae are reminiscent cf the latex tubes

of higher plants. Their contents stain easily with Melzer’s chloral-iodine, or with

phloxine. In some species the sap is coagulated by alcohol (especially in Agarics).

Such conductors may be up to 24 ^ in diameter. They are characteristic of the genus

Lactarius (Fig. 12, 2), and are present in some species of Mycena and Collybia, and in

some polypores, e.g. Polyporus sulphureus (Fig. 12, 1). Heim (1936) has studied the

conducting systems of several Agaric species and gives excellent illustrations of them.

Little appears to be known about the composition of the sap of conducting hyphae.

This would probably be a clue to their function, which is generally supposed to be the

conduction of nutrient emulsions. Conductors are present in the rhizomorph of

Armillaria mellea and if their function is nutritive it might explain in part the success

of rhizomorphs in penetrating far into unfavourable situations, this action being assisted

Fig. 12. — Conducting vessels of: —

1. Polyporus sulphureus. 2. Lactarius deliciosus. 3. Stereum spadiceum.

4. Corticium lactescens.

279

by their sclerotic rind. Confirmation of this supposition is given by Findlay’s (1951)

account of nutrients being transported distances of up to thirty feet along the rhizo-

morphs of Armillaria mdlea.

Conducting hyphae as described above are distinguished from gloeocystidia by

the fact that the latter are unbranched and limited in length. In Hydnum erinaceum,

long yellowish conductors in the tissues of the spines curve out into the hymenium

and their appear like gloeocystidia. The distinction between conductors and gloeo-

cystidia is not so obvious in another type of conducting hypha found in several species

of Stereum (Fig. 12, 3) which exude reddish juice on being bruised, and in Corticium

lactescens (Fig. 12, 4), which exudes a milky juice when fresh. Here the conductors are

comparatively narrow (4-9 p diam. in Stereum spadiceum), apparently unbranched or

very sparsely branched and relatively limited in length. Their connection with any

extensive network of conductors is obscure, if in evidence at all. In fact it is usual

to say that C. lactescens has gloeocystidia, though Miss Wakefield (1935) has pointed

out that these are the terminations of a system of conducting hyphae. Perhaps, then,

the best criterion of a conductor is that its contents exude when it is broken across

and are often brightly coloured, added to the fact that the contents are very dense and

that the conductors are long and indeterminate in length.

Vesicles.

Vesicles are pyriform or subglobose, hyaline, thinwalled swellings, usually borne

terminally but occasionally in an intercalary position on the context hyphae. In

section they may appear devoid of contents, or filled with a homogeneous, or globular

or granular, matter which stains deeply with phloxine. They occur typically in the

trama, and are conspicuous objects when numerous or large. They are figured in this

situation by Overholts (1929) and Burt (1920), pp. 125-134, figs. 13-16) for certain

species of Stereum. The vesicles seen in Stereum murraii (Fig. 13, 2) and S. purpureum

(Fig. 13, 1) illustrate typically what is meant by this term. Miss Wakefield (1911)

described the vesicles found in Grandinia mucida.

Structures similar to vesicles, but smaller, may be found with some difficulty in

the pore tissues of certain polypores, e.g. Polyporus subiculoides (Fig. 13, 3) and a

species of Poria near P. mucida (Fig. 13, 4). These are immersed in the subhymenium

and are probably better referred to as cystidia than vesicles. They are comparable

with the cystidia of Odontia bicolor (Fig. 16, 13), which, however, have a crystalline

crown.

To the writer there is little doubt that vesicles are a form of gloeocystidium, judging

not only from their position in the fructification but also from their general morphology :

there is also the observation that in Stereum murraii they are seen intergrading from

the typical pyriform shape of vesicles to the typical elongated shape of gloeocystidia,

with simultaneous change of their contents from homogeneous to granular or globular.

The gloeocystidium of Corticium polygonium (Fig. 17, 13) may equally well be considered

as a vesicle with dense, granular contents. In Stereum murraii at least, it would seem

that vesicles devoid of contents have simply lost their contents in the sectioning and

mounting treatment. If the organs are short, swollen, and without very dense contents

the tendency is to call them vesicles; if elongated they may be called gloeocystidia.

Three types of structure which might be confused with vesicles are worth mentioning.

Firstly the short, dumpy basidia of species of Amauroderma (Fig. 13, 5-6) may easily

be mistaken for vesicles or gloeocystidia in crush mounts if sterigmata are abseDt.

Secondly in some species (e.g. Stereum hirsutum) when surface scrapings of the hymenium

are made for mounting, one often cuts transversely across the underlying hyphae,

obtaining a structure which appears like a refractile, tbickwalled globose cell in optical

section (Fig. 13, 8). It is well to emphasise that vesicles are thinwalled and larger.

9102-2

280

I

Thirdly, some species (e.g. Daedalea biennis) have conidia embedded in the tissues

(Fig. 13, 7). These are small, subglobose or ovate, and appear thickwalled owing to

the presence of a large, dilutely coloured guttule. They arise from very delicate hyphae

which are seldom seen, but whose presence may be inferred from fragments still attached

to the conidia making them appear apiculate.

In a compact trama empty spaces are sometimes left when vesicles collapse.

Fig. 13. — Vesicles and structures resembling them:—

1 . Tramal vesicles in Stereum purpureum. 2. Tramal vesicles in Stereum murraii,

showing variation towards gloeocystidia. 3. Hymenial cystidia in Polyporus

subiculoides. 4. Hymenial cystidia in Poria sp. nr. P. mucida. 5-6. Immature

basidia of Amauroderma spp. 7. Conidia embedded in the subhymenial tissue of

Daedalea biennis. 8. Cross section of thickwalled hyphae from Stereum hirsutum

and part of a hypha to show their origin.

Coscinoids.

This term was given by Singer (1947, pp. 155-157) to the hyphae in a peculiar type

of conducting system in Paxillus lateritius, on the basis of which he erected a new genus,

Linder omyces, to accommodate this species.

Coscinoids are long, filamentous, brown, aseptate, non-clamped hyphae, with a

pitted sieve-like surface and a sponge-like interior. The surface perforations are the

orifices of a system of meandering tunnels of the same diameter as the orifices, and

separated by a wall substance coloured deep brown. The whole structure is sponge-like

and thus distinct from all other types of conducting system. When found in the

hymenium, these structures were referred to by Singer as “ coscinocystidia ”. Fragments

of coscinoids are illustrated in Fig. 14.

281

Fig. 14. — Fragments of coscinoids drawn semidiagrammatically from the type of

Paxillus lateritius in Kew Herbarium.

Paraphysoid structures.

A number of sterile accessory hymenial structures are often grouped for convenience

under the general name of paraphyses, though they are a heterogeneous collection and

probably not homologous with the paraphyses found in Ascomycetes, to which group

the term is sometimes restricted. For that reason the various types are sometimes

distinguished as pseudoparaphyses and subdivided according to their form, but the

terminology has been used in somewhat different senses by different authors.

In many of the Agaricaceae and in some of the Thelephoraceae there are hymenial

cells which separate and support the basidia in such a way that the efficiency of

basidiospore projection is increased. Langeron (1945, p. 323) terms these cells

pseudoparaphyses or basidioles, and considers that they are merely aborted basidia.

Miss Wakefield (1935) noted that in species showing differentiation of tissues resulting

from a compact growth form, it is possible that some of the terminal cells of the hyphae

remain permanently sterile and serve only to space out the basidia. Such sterile cells

may acquire a distinctive form and be known as paraphyses ; if they are indistinguishable

from immature basidia they may be known as basidioles. It is possible that by mathe-

matical analysis in the manner of Corner (1947; 1948 a) proof may be obtained

whether these cells termed basidioles might have assumed the function of basidia or

not. On the other hand there are sterile interbasidial cells which differ entirely from

those just mentioned, and which are conspicuous enough to be classified and to be

used as specific characters in taxonomy. Miss Wakefield (1935) observes that the

distinction between paraphyses and cystidia is not always very clear, and that in

practice the more or less elongated, filiform structures are liable to be called paraphyses

and the stouter bodies cystidia. The distinction at present is one of form rather than

of development or function. At present the emphasis must be on morphology alone,

for our knowledge of the function and development of all such structures is almost

negligable. Certain elements grouped here as paraphysoid structures are allied to

basidia, cystidia, gloeocystidia or hyphae in morphology, but may likely bear no close

relationship to any of these.

The various forms grouped as paraphysoid structures in the Hymenomycetes

(excluding Agaricaceae) will be considered now, following the usage proposed by

Pilat (1926) with slight modification. We shall endeavour to point out where this

differs from the usage of other authors, so that some uniformity may result.

(1) Simple paraphysoids (Fig. 15, A).— These are colourless, cylindrical or filiform,

elongated, unbranched, rarely septate, smooth, thinwalled, straight, flexuous or spirally

twisted elements which are prominent in the hymenium. In form there is often little

to distinguish them from hyphae or immature basidia or cystidioles. It may be that

their form can change in a manner analagous to that described by Bose (1940), who

observed that different moisture conditions could convert regular basidia of some

282

Fig. 15. — Paraphyses and paraphysoid structures: —

Group A. — simple paraphyses from 1. Aleurodiscus mirabilis. 2. Aleurodiscus

amorphus. 3. Aleurodiscus sp. 4. Aleurodiscus amorphus (after Burt, 1918).

5. Aleurodiscus oakesii (after Burt, 1918). 6. Corticium confluens. Group B. —

283

polypores into hyphal elongations, and vice versa. The distinction between simple

paraphysoids and dendrophyses in such species as Corticium roseum Pers. and Corticium

confiuens (Fig. 15, A, 6 and 15, E, 13) is merely one of degree of branching. In these

species there is no .fundamental difference, and their simple paraphysoids and dendro-

physes are only different stages in development of the same type of organ.

Simple paraphysoids were included as paraphyses by Pilat (1926), and by Hdhnel

& Litschauer (1907); they were distinguished as flexuous or spirally twisted paraphyses

by Burt (1918); in some senses they were termed pseudoparaphyses or basidioles by

by Langeron (1945).

(2) Pseudophyses (Fig. 15, B). — These structures are thinwalled, not septate,

hyaline, unbranched or occasionally forking slightly (Fig. 15, B, 2), somewhat clavate

and constricted especially near the apex into a number of bead-like parts. The cells

are homogeneous and stain deeply with phloxine. They would appear to be a form of

gloeocystidium, and the term used by Bourdot & Galzin (1928), torulose gloeocystidia,

is probably far more accurate than their description as a form of paraphysis.

Pilat (1926), Hohnel & Litschauer (1907), Singer (1945), Langeron (1945) and

Wakefield (1945) all termed these structures pseudophyses. Burt (1918) and Overholts

(1929) called them moniliform paraphyses. De Bary (1887) described them as

“ narrowly filiform hairs often constricted like a rosary ”.

(3) Acanthophyses (Fig. 15, C). — Pilat (1926) introduced this term for the peculiar

form of dendrophysis which is clavate or cylindrical and decorated with a number

of short, pin-like outgrowths chiefly near the apex. Fie specified also that this term

included Burt’s (1918) “bottle-brush paraphyses”, such as are common in many

species of Stereum and Aleurodiscus.

We have here extended the circumscription of this term to include an assortment

of paraphysoid structures all characterised by bearing short, pin-like outgrowths, but

probably differing widely in origin and function. For example, Singer (1945) compared

the “ dendrophyses ” of Favolaschia (cfr. Fig. 15, C, 9) with the “ cellules en brosse ”

of some Agaricaceae, and found that they belonged to different types. In a wide sense,

however, both might be included as acanthophyses following the conception given here.

In Favolaschia these cells occur both on the sterile surface of the pileus and in the

hymenium; in Agarics they are on the upper surface of the cap only. Cellules en

brosse have also been termed “ digitate cystidia ” (Corner, 1948 b, p. 243). While

observing the somewhat similar morphology of cellules en brosse , which might be

considered as acanthophyses with swollen bases, and other acanthophyses, it is felt

that the former term is useful in its application to the Agaricaceae, since these cells

do not occur in the hymenium.

Pseudophyses from : 1-3. Aleurodiscus amorphus. 4-5. Aleurodiscus aurantius.

6-7. Aleurodiscus oakesii; Group C. — Acanthophyses and related structures.

1-3. “ cellules en brosse ” from Marasmius (Androsaceus) epiphylloides (after

Conrad & Maublanc, 1924-1933). 4-6. “ cellules en brosse ” from Marasmius

sp. 7-8. Acanthophyses from Aleurodiscus oakesii. 9. from Favolaschia pezizoidea

(after Singer, 1945). 10-13. from Aleurodiscus mirabilis. 14. from Aleurodiscus

farlowii (after Burt, 1918). 15. A group of four “ cockroach cells ” from Aleuro-

discus weirii (after Burt, 1918): Group D. — Dichophyses from 1. Asterostromella

investiens. 2-5. Asterostromella sp.: Group E.— Dendrophyses from 1-3. Aleuro-

discus oakesii. 4-7. Aleurodiscus sp. 8-10. Aleurodiscus aurantius. 11-12. Aleuro-

discus amorphus. 13. Corticium confluens. All drawn to scale shown except Nos.

A, 4, 5; C. 1-3, 9, 14, 15.

284

In the sense used here acanthophyses are hyaline cells with hyaline or occasionally

dilutely coloured pin-like processes. They occur in the hymenium. They may have

thickish walls and be cylindrical or clavate, resembling some forms of cystidia (Fig.

15, C, 10-12), or thinwalled with a few basal processes (Fig. 15, C, 13) and in this

case may be modified basidia. In some species (Fig. 15, C, 14) they possibly represent

a modified hypha.

Pilat (1926) appears to have restricted the term acanthophysis to the bottle-brush

type of paraphysis, or to forms very similar to it. This usage is followed by Overholts

(1929) and Wakefield (1935). Hohnel & Litschauer (1907) and Langeron (1945)

grouped such forms as dendrophyses, and in part Singer (1945) did the same, e.g. in

Favolaschia. Included in our conception of acanthophyses are elements which Bresadola

named “ cystidia corniculato-pinnata ”, and those which Burt (1918) described as

paraphyses with aculeate prongs, bottle-brush paraphyses, and cockroach-shaped

paraphyses ” (Fig. 15, C, 15).

(4) Dichophyses (Fig. 15, D). — This term was introduced by Pilat (1926) for a

special form of dendrophysis distinguished by its more or less regular dichotomous

branching. They are clearly seen in species of Asterostromel/a, also in Stereum

albobadium, “ Stereum ayresii ” (“ S. midas ”), and in the genus Dichostereum Pilat

or Vararia Karst.

Dichophyses are undoubtedly a form of hypha having rather thick walls, narrow

lumina and subulate apices to the dichotomous branches. They are usually coloured

and may occur in the hymenium or in the tramal tissues, or both. Only in the hymenium

should they be considered for convenience as a type of paraphysis. Corner (1948 b, f. 8)

shows clearly the origin of dichophyses in Asterostromella sp. as lateral branches of

generative hyphae. Bourdot & Galzin (1928) emphasised their hyphal nature, referring

to them as “ corne de cerf ” (deer-antlers), or dichotomous-divaricate or dendroid-

branched hyphae. Hohnel & Litschauer (1907) described them as “ tree-like, branched

cystidia Burt (1918) included them under the category of “ granule-bearing para-

physes ”. Wakefield (1935) mentioned Pilat’s usage.

(5) Dendrophyses (Fig. 15, E). — These are thinwalled, usually colourless, vaguely

and irregularly but not dichotomously branched hyphae, which occur intermingled

with other hymenial elements. Pilat (1926) restricted the term to paraphyses of

tree-like form with irregular branching. Hohnel & Litschauer (1907) did not distinguish

between these and acanthophyses, naming both dendrophyses. Wakefield (1935)

followed Pilat’s usage. Overholts (1929) termed those with short lateral branches

“ antler paraphyses ”, though this description is usually reserved for dichophyses.

Burt (1918) shows both dendrophyses and dichophyses under forms classed as

“ granule-bearing paraphyses

Dendrophyses occur in many species of Aleurodiscv.s (Fig. 15, E, 1-12), in some of

Sebacina (e.g. S. sublilacina), and in various species of Corticium, especially the section

Aleurodiscoidea of Bourdot & Galzin (1928), e.g. Corticium roseum. In C. roseum

and Corticium confluens simple paraphysoids are found as well, and these differ from

the dendrophyses only in being unbranched and presumably in a younger state (cfr. Figs.

15, A, 6 and 15, E, 13).

Cystidioles.

A cystidiole is defined by Ainsworth & Bisby (1945) as “ a sterile structure in the

hymenium, at the same level as the basidia, wider than a paraphysis, thin walled and

little differentiated ”. They are usually inconspicuous, colourless, subulate or fusoid,

often with a sharply pointed apex, thinwalled, sometimes finely encrusted, or smooth,

sometimes projecting a short distance above the basidia. Usually they have no septa,

but in Corticium gloeosporum an occasional subapical septum may be found. In

Stereum purpureum cystidioles seem to appear under wet conditions, giving rise to the

285

form once known as S. rugosiusculum. There has been some doubt whether the

corticioid forms bearing cystidioles should be classified under Corticium or under

Peniophora. Thus on account of the cystidioles C. sambuci is sometimes classed as

a Peniophora, while in Peniophora cremea the cystidioles are large and could well be

called cystidia.

Among the species of resupinate Hymenomycetes bearing cystidioles are the

following: Corticium laeve, C. sambuci, C. ochraceofulvum, C. g/oeosporum (Fig. 16,

39), Stereum purpureum, Peniophora cremea, P. sanguinea, Acia subceracea, Odontia

crustosa (Fig. 16, 34), O. knysnana (Fig. 16, 35) and Hypochnella violacea.

Cystidia.

These are a rather heterogeneous assemblage of sterile bodies commonly occurring

in representatives of the Agaricaceae, Polyporaceae, the genus Peniophora, and in

certain species of the genera Stereum, Odontia, Mycoleptodon, Cladoderris and Dupor-

tella, to mention some of the better known. It is improbable that the so-called cystidia

of all these groups are homologous, for they differ considerably in form, origin and

probably in function. More information is urgently needed on these points, but for

the time they are all discussed under the general category of cystidia. Many more

studies such as those of Corner (1947) are needed, where it was shown that the cystidia

of some species of Clavaria and Oedemansiella conform to the same mathematical

equations as their basidia and are to be regarded as sterile, mostly precocious,

overgrown basidia.

Certain generic segregates have been based on the presence of cystidia. For example

Peniophora was segragated from other corticioid fungi originally on account of its

“ metuloids ”, but since then has included forms with very different cystidia ; Lloydella

was erected to include species of Stereum with metuloids; Kneiffia was ill-defined but

its type species, K. setigera ( = Peniophora aspera) has characteristic cystidia ; Hetero-

chaetella was segregated from Sebacina on account of its cystidia, and Coniophorella

from Coniophora on the same grounds.

In the Agaricaceae, where the term “ cystidium ” originated with Montagne and

was taken up by Leveille, the cystidia are sterile, hyaline, unicellular, ovoid cylindrial

or clavate, blunt or pointed, smooth or ornamented or encrusted, originating as the

terminal cells of branched hyphae and projecting from the hymenium. Buller (1924,

pp. 52-53) classified them according to their position on the fructification, thus:

Pilocystidiu on the surface of the pileus ; Cheilocystidia on the edge of the gills ; Pleuro-

cystidia on the sides of the gills; Caulocystidia on the stipe; Dematocystidia on the

cuticle. Langeron (1945, p. 325, fig. 278) classifies the principal types of pleurocystidia

giving examples of fungi in which they occur. There are five types, including those

with spinous, barbed, crowned, and hooked apices, and those which have an encrustation

resembling a muff round the middle of the cystidium. Langeron also noted a

dimorphism in the cheilocystidia found on the large and small gills of some Agarics.

The supposed functions of the cystidia in Agrics have been debated in various works,

but it is difficult to ascertain the original source of some of these ideas. Cystidia have

been thought to function as excretory hydathodes (Knoll) 1912. De Bary (1887)

mentions that cystidia are often covered with slime, but observes that this condition

is common in fungus cells rich in cell sap. Levine (1913, p. 161) considered that the

cystidia in Boleti were modified basidia with a glandular function, excreting mucilage

from their whole surface. Cystidia are also thought to have a mechanical value in

maintaining the rigidity of the gills, in spacing adjacent gills apart, or in loosening

appressed lamellae from the stipe. In the lower Hymenomycetes the pointed, encrusted

“ metuloids ” were suggested to have a protective function against small animals feeding

on the hymenium.

286

287

This study is more concerned with the cystidia of the lower Hymenomycetes

(See Fig. 16), which again are very heterogeneous even in a single genus such as

Peniophora. Thus it is impossible to give a general description of them. They may be

conical or cylindrical in shape, or sometimes narrow and flexuous. The walls are

sometimes characteristically thick and encrusted on the outside, but there are many

species which have smooth, thinwalled cystidia. Some species, e.g. Peniophora proximo

(Fig. 16, 15) have cystidia with finely sculptured walls. The lumen is usually narrow

towards the apex, but in Peniophora section Tubuliferae (Bourdot & Galzin, 1928)

the lumen is very narrow at the base and expands towards the apex of the cystidium .

Incrustation of the walls may take the form of very fine amorphous, to large crystalline,

deposits of mineral matter. In Odontia bicolor (Fig. 16, 13) the incrustation is in the form

of druses of crystals, while in Odontia arguta (Fig. 16, 36) and Peniophora pallidula

(Fig. 16, 16) the cystidia bear droplets of brownish resin. The small capitate cystidia

of Odontia bicolor are similar in shape to those of Polystictus subiculoides, and resemble

small vesicles except for their occurrence in the hymenium. Cystidia are usually

hyaline, but if dilutely coloured then the colour resides in their walls and not in their

contents. The contents are seldom well differentiated. In Peniophora section

Gloeocystidiales (Bourdot & Galzin, 1928) the cystidia are smooth and thinwalled and

have a fine, dense, granular content like that of gloeocystidia, but they originate in

the basal part of the trama and project far beyond the hymenium, which according

to Bourdot & Galzin distinguishes them from gloeocystidia. The distinction is not

particularly clear.

The type of cystidium termed a “ metuloid ” by Cooke, is thickwalled, heavily

encrusted and usually conical (e.g. Fig. 16, 1, 2, 4, 7, 17). They are sharply differentiated

from the hyphae from which they rise. The smooth cystidia of Peniophora section

Tubuliferae have a lumen which expands at the apex; they arise from the basal tissues

and often have a forked base, and possess the curious property of dissolving in

Potassium hydroxide solution so that only the shrunken cell contents are left behind.

Examples are Peniophora glebulosa, P. subalutacea, and similar cystidia found in

Stereum karstenii (Fig. 16, 31, 21, and 20 respectively). Septation is not usual in

cystidia, but is very well seen in Peniophora aspera (Fig. 16, 22), Coniophorella olivacea

and Peniophora tomentella (Fig. 16, 30). In the first species the cystidia also show

occasional clamp connections.

The place of origin is probably important in classifying the types of cystidia

encountered among the lower Hymenomycetes. Some cystidia are little modified

from the hyphae from which they arise, e.g. Peniophora tomentella (Fig. 16, 30)

Fig. 16. — Cystidia and Cystidioles: Cystidia from the following species: —

1 . Stereum cinerascens. 2. Lopharia mirabilis. 3. Poria eupora. 4. Peniophora

roumeguerii. 5. Peniophora pelliculosa. 6. Grammothele pseudomappa. 7. Dupor-

tella tristicula. 8. Poria obducens. 9. Cladoderris australica. 10. Cladoderris

infundibuliformis. 11. Peniophora pubera. 12. Hyphal peg from Grammothele

mappa (type) resembling a cystidium. 13. Odontia bicolor. 14. Daedalea quercina.

15. Peniophora proxima. 16. Peniophora pallidula. 17. Stereum involutum.

18. Peniophora laevis forma P. aftinis. 19. Peniophora velutina. 20. Stereum

karstenii in acid and alkaline mounts. 21. Peniophora subalutacea. 22. Peniophora

aspera ( = P. setigera). 23. Peniophora aegerita. 24. Peniophora longispora.

25. Irpex dregeanus. 26. Peniophora byssoidea. 27. Sebacina sp. 28. Odontia

papillosa. 29. Sebacina sublilacina. 30. Peniophora tomentella. 31. Peniophora

glebulosa. 32. Peniophora cinerea. 33. Peniophora aegerita. Cystidioles from the

following species: 34. Odontia crustosa. 35. Odontia knysnana. 36. Odontia

arguta. 37. Stereum purpureum. 38. Peniophora cremea. 39. Corticium gloeos-

porum. All are drawn to the scale shown except Nos. 16, 23, 24, 31, 33.

288

P. byssoidea (Fig. 16, 26) and Stereum umbrinum (Fig. 19, 11-13). Others, especially

the metuloid type, are sharply differentiated from the hyphae. Many cystidia originate

from the base of the trama, e.g. Peniophora glebulosa (Fig. 16, 31), P. subalutacea

(Fig. 16, 21) and Stereum karstenii (Fig. 16, 20). Especially in the Coloratae section

of Peniophora (Bourdot & Galzin, 1928) the cystidia are arranged in stages throughout

the trama, e.g. P. proximo (Fig. 16, 15) and P. roumeguerii (.Fig. 16, 4). Still other

cystidia arise in the hymenium or at least in the upper half of the trama, e.g. Peniophora

gigantea, P. longispora (Fig. 16, 24), P. byssoidea , P. pallidula, and P. pubera (Fig.

16, 26, 16 and 11 respectively.

Illustrated in Fig. 16, 12 is the form of hyphal peg found in the trama of

Grammothele mappa. At first it might be thought to be a large cystidium, but more

careful observation reveals the core of erect sheaves of brownish hyphae capped by

a heavy mineral incrustation (see also under Hyphal Pegs, p. 277).

Gloeocystidia.

In Fig. 17 are shown a number of bodies found in species of many genera and all

classed as gloeocystidia. These are sterile organs mainly distinguished by their thin

walls, dense and deeply staining contents, lack of sculpturing and incrustation, and

their position in the tissues of the fungus. The contents may be hyaline or lightly

coloured, yellow to brownish; they are highly refractile, homogeneous or granular

or globular, in the last instance usually of an oily nature. Commonly they arise from

hyphae situated in the subhymenium or deeper tissues and may be completely embedded,

or extend as far as the level of the basidia ; only in a few species are they found projecting

above the hymenium. Their walls are colourless. Any colour they may possess is

located in the cell contents.

Gloeocystidia stain deeply with phloxine or eosin in potassium hydroxide mounts,

and ones which are normally coloured take on a deeper, brownish colour with iodine

solutions. Singer (1945) recommends the use of brilliant cresyl blue for differential

staining of the contents and walls of gloeocystidia; with this stain the contents turn

blue, while the walls are stained lilac.

In shape, gloeocystidia may be narrow, flexuous and subcylindrial, or wider and

clavate, or uncommonly (e.g. Corticium polygonium, Fig. 17, 13) inflated and ovoid

or pyriform. They are not branched. Septation is very rare but is known to occur

in the gloeocystidia of Duportella tristicula (.Fig. 17, 15), where a single subapical

septum is sometimes seen. Many species show pseudoseptation of the gloeocystidia

when mounted in glycerine, which causes the contents to contract (Overholts, 1929).

The walls are normally smooth and not encrusted with mineral matter. However,

in Corticium pallidum (Fig. 17, 12) the gloeocystidia are capped with a brownish

resinous substance which is soluble in hot lactic acid but not in alkali (Hohnel &

Litschauer, 1907).

Despite the distinguishing features outlined above, gloeocystidia may be confused

with other organs in a few specific cases. Some examples are now given: —

(1) In the genus Aleurodiscus the basidia are very large and stain deeply, thus

immature basidia may be taken for gloeocystidia in crush mounts where the

hymenial elements are displaced. Basidia of Amauroderma may also be taken

for gloeocystidia or vesicles (see p. 279 and Fig. 13, 5-6).

(2) The conducting hyphae of the bleeding species of Stereum, and of Corticium

lactescens, do not form an obvious network, and their terminations may be

thought to be gloeocystidia. Conductors exude their contents readily, whereas

the contents of gloeocystidia usually cohere fairly well. In Hydnum erinaceum

there are “ gloeocystidia ” in the spines which connect with an extensive system

of conducting hyphae. They are part of one system and should not be

differentiated.

289

1 . Stereum diaphanum. 2. Stereum duriusculum. 3. Stereum affine. 4. Stereum

involutum. 5. Stereum thozetii. 6. Stereum bicolor. 7. Corticium praetermissum

( = Peniophora tenuis). 8. Corticium roseocremeum. 9. Corticium coroniferum.

10. Corticium albostramineum. 11. Corticium luteocystidiatum. 12. Corticium

pallidum. 13. Corticium polygonium. 14. Bourdotia cinerella var. trachyspora.

15. Duportella tristicula. 16. Asterostroma cervicolor. 17. Bourdotia eyrei.

18. Seismosarca sp. All drawn with camera lucida, but scale varying.

290

(3) Certain species of Peniophora (e.g. P. sphaerospora, P. tenuis ) with non-encrusted,

thinwalled cystidia which project above the hymenium, have frequently been

classed as Corticium as there is some doubt whether these organs are cystidia

or gloeocystidia. The tendency nowadays is to place these under Peniophora,

for example Rogers & Jackson (1943) refer Corticium praetermissum (Fig.

17, 7) to Peniophora under the species P. tenuis. Again in Stereum diaphanum

it is difficult to decide whether these organs are cystidia or gloeocystidia; they

may be embedded or project above the hymenium.

(4) Some cystidia, when mouhted in alkali, have their encrustation dissolved

away, and but for their position and thick walls may be confused with

gloeocystidia. Numerous illustrations of this condition are to be found in

Fig. 16, e.g. 6, 9, 10, 11, 17.

(5) The vesicles in some species resemble gloeocystidia. See note on p. 279.

(6) The pseudophyses (moniliform paraphyses) of species of Aleurodiscus are

probably best considered as a form of gloeocystidium. See note on p. 283.

(7) In Stereum bicolor, Peniophora incarnata, and some other species of Peniophora

in the section Coloratae (Bourdot & Galzin, 1928) there appears to be a distinct

intergrading of form between gloeocystidia and cystidia. This is well shown

in S. bicolor (Fig. 17, 6). In most species where cystidia and gloeocystidia

occur together, the two kinds of organ are distinct at all times. This bears

on the origin of these organs, which will be considered now.

i

As gloeocystidia are probably not all homologous organs no generalisation is

possible regarding their origin and funtion, and these aspects have only been studied

in a very few species. In the species just mentioned, there is a difinite intergrading

from gloeocystidia to cystidia. Massee (1887) concluded that cystidia were the terminal

growing points of laticiferous vessels, and that the young ones contained a hyaline,

dense protoplasm which was later replaced by a finely granular substance which

eventually escaped through an apical pore to nourish the spores. A very common

idea was that gloeocystidia exude at the apex and become changed into encrusted

cystidia in species such as P. incarnata. Present observations on Stereum bicolor

(Fig. 17, 6) suggest that the whole wall of the gloeocystidium dissolves and that the

contents, now very dense and coherent, remain as a highly refractile body which

resembles a fragmented and encrusted cystidium. Whelden (1936) carefully studied

the cystidia and gloeocystidia of Peniophora livida and concluded that the cystidia

originated in the same way as basidia from the apex of an hymenial hypha, while the

gloeocystidia arose laterally from hyphae near the substratum. In this fungus these

organs were always entirely distinct from one another. Mention has already been made

of Corner’s work on the origin of cystidia on some species of Clan aria and Oedemansiella

(see p. 285).

Nothing definite is known of the function of gloeocystidia, but it is likely that they

are concerned in nutrition. Overholts (1929) suggested that they might have a

protective function if their contents were unpalatable to animals.

Gloeocystidia are uncommon among higher Hymenomycetes, but are widespread

among the genera of the lower Hymenomycetes. For example, they are found in

many species of the genera Aleurodiscus, Stereum, Corticium, Sebacina, and Asteros-

tromella. They form the so-called “ colour cells ” found in the genus Favolaschia

(Singer, 1945). Their presence is the reason for splitting off new genera from old, e.g.

291

Gloeocystidiel/um from Corticium, Gloeopeniophora from Peniophora, Bourdotia from

Sebacina, Seismosarca from Exidia, and Gloeotulasnella from Tulasnella. This practice

is now ceasing, and at the most the presence of gloeocystidia is given subgeneric

significance.

Setae.

Setae may be defined as sterile, rigid, dark-coloured, thick-walled, spine-like

organs, usually having a more or less pointed apex, and possessing the property of

darkening in alkali. Burt (1918, p. 302) drew attention to this chemical reaction as

a fundamental way of distinguishing setae from similar structures such as elongated

coloured cystidia (e.g. in Stereilm umbrinum). Setae are yellowish- or reddish-brown

in colour, and the pigment is located mainly in the walls. The lumen is usually narrow.

Simple, unbranched setae vary in size and shape, and in the character of their

apices. Such variations are usually only of minor specific significance, for on the whole

their appearance is fairly uniform. Compound setae are not common; they may

assume bizarre shapes, but may be analysed as composed of a main axis united to

branches which individually do not differ much from forms encountered among simple

setae.

In different species, setae may be abundant or rare, emergent from the hymenium

or embedded deeply, or found upon the cuticular surface.

Cunningham (1946) has studied the origin of setae in some species of polypores.

They usually develop below the hymenium and project beyond the basidia. Cunningham

states that they arise from skeletal hyphae in species with a dimitic hyphal system, and

from generative hyphae in species with a monomitic system; they are absent from

species with a trimitic system. However, as Corner (1932 b, p. 59) showed that the

setae of Fomes laevigatus, a species with a dimitic system, arose from generative hyphae

perhaps no generalisation is yet possible. With most setae it is extremely difficult

to trace the type of hypha from which they arise.

Regarding the function of setae nothing definite is known, but it has been speculated

that they may provide rigidity to the fructification and protect the tissues from damage

by small animals. Overholts (1929) mentioned these speculations, but added that the

function is probably far more fundamental. Considering that small animals provide

a useful alternative to anemochory or water dispersal of spores of many resupinate

Basidiomycetes (Talbot, 1952), the writer is inclined to minimise any protective function

ascribed to setae, cystidia or gloeocystidia.

As with most other conspicuous organs, setae have been used in the erection of

new genera, e.g. Hymenochaete Lev. and Mucronoporus Ell. & Everh. They are common

in Phellinus Quel, and Xanthochrous Pat., are found in many species of Poria, and were

estimated by Overholts (1929) to occur in 5-10 per cent of species of Fomes and

Polyporus.

Of themselves, setae would not appear to have any generic significance. They

are, however, associated with other characters which occur in a series of fungi at present

classed in many different genera and families, and which seem to constitute a natural

series. Donk (1933) refers to this series as the Hymenochaetoideae, a sub-family of

the Aphyllophoraceae, while Corner (1948 b) calls it the Xanthochroic series. As

Corner (1948 b, p. 235) expresses it, “These Xanthochroic fungi are distinguishable

not so much by the Hymenochaete setae, which are absent from many of them, as by the

character of their hyphae. The absence of clamp connections, the lack of inflation

of the cells of the fruit body, and the ochraceous or brown colour of the hyphal walls,

which darkens to ferruginous or date brown with alkali, distinguish these fungi

IOO

292

Fig. 18. — Setae: —

1. Uncinate (hamate) seta from Poria uncinata (after Cunningham, 1946).

2. Yentricose seta from Polyporus tabacinus (after Cunningham, 1946). 3. Com-

pound seta of Polyporus grammocephalus (after Cunningham, 1946). 4. Subulate

seta (after Lowe, 1942). 5. Ventricose seta (after Lowe, 1942). 6. Setae from the

hymenium of Polyporus gilvus. 7. from Polyporus radiatus. 8. from Hymenochaete

fasciculata (type). 9. from Poria contigua. 10. from Hymenochaete nigricans.

293

This, and similar trends in classification are among the most significant to have

arisen recently. The Friesian classification was based on macroscopic characters.

Later with Patouillard (1900) came the first and greatest attempt to utilise microscopic

characters in taxonomy, but the trend was then to give too much prominence to them,

with the result that many genera became defined by single microscopic characters.

The present tendency is to bring into their right perspective the wealth of microscopic

characters which are already known, and through this to reveal a series of affinities

which cut right across the classical groups. A more natural classification will result

from this type of work, but only when more species have been scrutinised and more

is known of the function of their microscopic organs.

For the diagnosis of species, some of the differences seen in setae are the following : — -

(1) Size. Cunningham (1946) mentions a size range of 10-160 p. long and 6-12 p

wide.

(2) Shape. Subulate (Fig. 18, 4), ventricose (Fig. 18, 2, 5), conical and elongated

(Fig. 18, 8, 9) shapes are common.

(3) Colour.

(4) Apex. The apex is often sharply pointed, but sometimes blunt. It may be

hamate or uncinate (Fig. 18, 1).

(5) Position of the setae in the hymenium, trama or cuticle, and their relative

abundance.

(6) The type of hypha from which the setae arise.

(7) Branching. Simple, or compound setae (Fig. 18, 13).

Setoid structures.

In certain species of resupinate Hymenomycetes and polypores, structures are

found which resemble setae except that they are much more elongated and may differ

in their origin. They are usually more cylindrical, of comparatively great length, and

not limited in size and shape as is a well defined seta. This is because they are

apparently only modified thickwalled skeletal hyphae. The lumen is usually very

narrow, sometimes expanding a trifle towards the apex, which itself is frequently wider

than the rest of the hypha. The apex may be minutely sculptured as in Stereum

schomburgkii and Duportella tristicula (Fig. 19, 8-10). In other species the wall is

quite smooth, e.g. in Polyporus ochroporus and P. patouillardi (Fig. 19, 14-17). In

all these species the setoid structures darken slightly in alkali while the rest of the

trama shows an even darker reaction. Similar structures occur in P. tabacinas (fide

Cunningham, 1946). In some collections of Stereum hirsutum some of the hyphae

though only very faintly coloured, have thick walls and a lumen which expands near

the inflated apex in a way similar to that found in the species just mentioned. Pilat

(in Hedwigia, Vol. 70) records some collections of S. hirsutum of this type. These

structures are included here for comparison, but they do not darken in alkali and

are not setoid structures. In Stereum umbrinum there are mineral-encrusted organs

which are usually called cystidia (Fig. 19, 11-13) but which are frequently smooth-

walled and might then be considered as setoid structures. They too are only modified

skeletal hyphae and sometimes give a faint colour reaction with alkali.

11. from Polyporus cameofulvus (type). 12. from Polyporus ochroporus (cotype).

13. Compound cuticular setae from Polyporus cuticularis. 14. Favolus megaloporus.

15. Polyporus dryadeus. All except Nos. 1-5 drawn to the scale shown.

294

295

A form of setoid structure differing from those already mentioned is found in the

pore walls of Polyporus dictyopus. These terminate some of the hyphae. They take the

form of a blunt, irregular axis with irregular, narrower lateral outgrowths. The axis

is brown and darkens in alkali; the lateral outgrowths are paler and often almost

hyaline at their apices (Fig. 19, 1-7).

Asterosetae.

In the genera Asterostroma, Asterodon, and in some species of Lachnocladium

there occur brown, stellate organs composed of several rays, each like a seta, radiating

from a common centre which is sometimes expanded into a distinct boss. They

terminate some of the laterals of generative hyphae, which, in Asterodon, also give rise

to skeletal hyphae independently, and to extrahymenial setae and hymenial setae

(Corner, 1948 b). The largest asterosetae occur towards the base of the fructification;

they diminish progressively in size as the hymenium is approached. The rays of the

asterosetae may be simple or branched in the same fructification. When they are

branched, the organ is still regular in shape, and stellate. Asterosetae are related to

ordinary setae in composition and basic structure, and may be regarded as a special

form of compound seta. They darken in alkali and the species bearing them also show

other indications of belonging to the Xanthochroic series, or Hymenochaetoideae.

Corner’s work (1948 b) on the morphology and development of Asterodon

ferruginosus is especially interesting in relating the shape and direction of the asterosetae

to the forces controlling the growth and form of the fructification. Thus, in the

resupinate part of the fructification there are evidently no form factors in operation and

stellate setae are developed (Fig. 20, 6). In the hymenium the setae become simple,

or those parts of the setae lying in the subhymenium become shortly substellate (Fig.

20, 7). In the spines, where a positive geogropism is in action, the setae lying in the

context tissue are elongated downwards (Fig. 20, 8-10), while those near the hymenium

where the factor responsible for lateral production of basidia is in operation, become

drawn out into lateral branches simulating ordinary hymenial setae (Fig. 20, 10). As

Corner expresses it, “ The shape of the setae and the direction of the skeletal hypha

express the action of the form-factors of the fruit body This most significant work

in interrelating these forms of setae provides a reason for the bewildering shapes which

setae may assume.

Mineral inclusions.

Very little is known about the chemical composition of the minerals which so

commonly occur in the trama, or encrust the hyphae and other organs of fungi.

Calcium oxalate is a common mineral found in the form of octahedra, acicular crystals

or irregular nodules or granules; in fact most mineral substances found in fungi were

once assumed to be calcium oxalate without further inquiry. Recent investigations

into the metabolic products of fungi have shown that a vast number of organic

compounds may be isolated from growing fungi and then crystallised.

Added to the mineral forms mentioned above, it is also common to find small

mineral platelets, druses of crystals (e.g. in the trama and crowning the cystidia of

Odontia bicolor), and large irregular concretions (e.g. in the trama of Grammothele

Fig 19. — Setoid structures: —

1-7. Hyphal terminations in the pore walls of Polyporus dictyopus. 8-9. Hyphal

terminations in the hymenium of Stereum schomburgkii. 10. Hyphal terminations

in the hymenium of Duportella tristicula. 11-13. Modified skeletal hyphae

(cystidia) of Stereum umbrinum. 14-15. Setoid hyphae of Polyporus ochroporus

(cotype). 16-17. Setoid hyphae from the pore tissue of Polyporus patouillardii

(short examples).

OOI

296

297

spp.). Incrustation of the cystidia is particularly common in the genus Peniophora,

and in species of Stereum which were at one time segregated as the genus Lloydella

Incrusted hyphae also occur in many species and are then an aid to specific diagnoses

(e.g. in Peniophora filamentosa and Polyporus rutilans and Coniophora betulae). In

P. filamentosa and P. rutilans, which but for the difference in hymenial configuration

are identical, the crystals are rapidly soluble in potassium hydroxide solution and

react to give a vinous colour. Some mineral inclusions are soluble in alkali, others in

acid solutions, thus it is desirable to make mounts in more than one medium.

Occasionally the mineral incrustation has its own bright colour which colours the whole

fungus; more often the minerals are colourless or light yellowish and at the most

impart a glitter to the surface of the fungus.

Large airspaces which are frequently observed in sections of fungi may be due to

the former presence of mineral inclusions. These are sometimes torn out during

sectioning as the razor comes up against the hard obstruction, or they may sometimes

have been leached out prior to collection. It is possible that during its growth the

fungus produces minerals in solution, which, with the advent of dry conditions, become

deposited in the tissues as large concretions. On subsequent leaching of the trama

these may be redissoved and thus leave an airspace.

Acknowledgments.

The writer was dependent upon the generosity of many authors and publishers

for permission to reproduce illustrations of material which was not available, or

unsuitable, for first-hand study. The following authors, publishers and editors of

journals, as the case may be, co-operated most kindly in allowing the figures listed

below to be copied. The writer tenders to them his sincere thanks: —

The College of Forestry, State University of New York, Technical Publication

No. 60; J. L. Lowe, Polyporaceae of New York State, Illustr. ii, 26.

Bulletin of the Torrey Botanical Club: G. W. Martin in Vol. 62 (1935), 341,

Fig. 2, I. & J.

Lloydia: G. W. Martin in Vol. 11 (1948) 115, Fig. 1, 1 & 2 f.

R. Singer in Vol. 8 (1945) 209, PI. 3, Fig. 9.

University of Iowa Studies in Natural History: D. P. Rogers in No. 16 (1934)

PI. 7, Figs. 3, 4, 21.

Mycologia: D. P. Rogers in Vol. 36 (1944) 81, Fig. 3; 97, Fig. 12.

G. W. Martin in Vol. 28 (1936) 215, Fig. 1, 9.

G. W. Martin in Vol. 31 (1939) 509, Figs. 12, 13.

L. Olive in Vol. 42 (1950) 389, Fig. 2, 17, 20, 21.

Annals of the Missouri Botanical Garden: E. A. Burt in Vol. 5 (1918) 181,

Fig. 1, f; 184, Fig. 3, f; 183, Fig. 2; 203, Fig. 14, c & br.

British Mycological Society Transactions:

E. M. Wakefield in Vol. 4 (1913) PI. 3, Figs. 3-5.

E. M. Wakefield & A. A. Pearson in Vol. 6 (1920) text Fig. on p. 320; in

Vol. 8 (1923) 218, Fig. 4.

E. J. H. Corner in Vol. 31 (1948) 234, Fig. 1; 239, Fig. 5.

Journal of the Elisha Mitchell Society: W. C. Coker in Vol. 35 (1920) PI. 55, Fig. 1.

University of North Carolina Press : J. N. Couch, “ The genus Septobasidium ”

(1938) 351, PI. 28, Fig. 1.

Fig. 20. — Asterosetae: —

1-5. from Asterostroma cervicolor. 6-10. from Asterodon ferruginosus (after

Corner, 1948 b). 6. Stellate setae in the resupinate part of the fructification.

7. substellate setae near the subhymenium. 8-10. Setae in the tissues of the spines

and emerging laterally into the hymenium on the spines (simple, forked or substellate

setae). Nos. 1-5, only, drawn to scale shown.

298

Cambridge University Press: W. B. Grove, “The British Rust Fungi” (1913)

14, Fig. 16 a and b.

Department of Scientific and Industrial Research, New Zealand: G. H. Cun-

ningham in N.Z. Journ. of Science & Technology 28 (1946) Figs. 7, 8, 9.

The writer is much indebted to the following persons for their assistance in locating

the owners of copyright material: Prof. Roger Heim, Dr. D. P. Rogers, Dr. P. H.

Gregory, Dr. Theodor Just, Dr. C. A. Berger, Prof. Ralph G. Unger, Dr. G. W. Martin

and Dr. J. N. Couch.

The author’s obligation to Dr. M. A. Donk must be apparent in the part of this

paper dealing with basidial morphology, but he also wishes to record his thanks to

Dr. Donk for making available much literature, for illuminating criticism of parts of

the manuscript, and for every generous encouragement.

Literature Cited.

Atnsworth & Bisby. . (1945). A Dictionary of the Fungi. 2nd Ed., I.M.I., Kew.

Ames (1913). Ann. Mycol. 11, p. 220.

Baker (1936). Ann. Mo. Bot. Gard. 23, pp. 69-126.

Banerjee (1942). Journ. Indian Bot. Soc. 21, pp. 33-39.

Bodman (1938). Mycologia 30, pp. 635-652.

Boedijn (1937). Bull. Jard. Bot. Buitenzorg ser. iii, 14, pp. 187-196.

Bose (1940). Nature 145, p. 706.

(1944). Journ. Indian Bot. Soc. 23, pp. 153-157.

Bourdot & Galzin. . . (1928). Hymenomycetes de France. Soc. Myc. de Fr.

Bresadola (1903). Ann. Mycol. 1, Tab. 3, Fig. 1, c.

(1906). apud Rick in Broteria 5, p. 7.

(1920). Ann. Mycol. 18, p. 44, Fig. 3; p. 52.

Buller (1909). Researches on Fungi, Vol. 1. Longmans & Co.

(1922). Ibid. 2, p. 31 and p. 157. Longmans & Co.

(1924). Ibid. 3, pp. 52-53 and pp. 446^163. Longmans & Co.

Burt (1918). The Thelephoraceae of North America. Ann Mo. Bot. Gard. 5,

pp. 301-372.

(1919). Ibid, in Ann. Mo. Bot. Gard. 6, pp. 253-280.

(1920). Ibid, in Ann. Mo. Bot. Gard. 7, pp. 81-248.

(1925). Ibid, in Ann. Mo. Bot. Gard. 12, pp. 213-257.

(1926). Ibid, in Ann. Mo. Bot. Gard. 13, pp. 173-354.

Coker (1920). Journ. Elisha Mitchell. Soc. 35, pp. 113-182, PI. 55, Fig. 1.

Conrad & Maublanc (1924-1933). leones Selectae Fungorum 3, PI. 220, ii.

Corner (1932 a). Annals of Bot. 46, pp. 71-111.

(1932 b). Trans Brit. Myc. Soc. 17, pp. 51-81.

(1947). New Phytologist 46, p. 195.

(1948 a). New Phytologist 47, p. 22.

(1948) b). Trans. Brit. Myc. Soc. 31, pp. 234-245.

Couch (1938). The genus Septobasidium. Univ. of North Carolina Press.

Cunningham (1946). N.Z. Journ. Sci. & Technology 28, pp. 238-251.

(1947-1950). N.Z., D.S.I.R. Plant Diseases Div., Bulletins Nos. 72-83.

De Bary (1887). (English Translation). Comparative morphology and Biology of

the Fungi, Mycetozoa and Bacteria. Oxford.

Derx (1948). Bull. Bot. Gard. Buitenzorg. ser. iii, 17, p. 468.

Donk (1931). Revisie van de Nederlandse Heterobasidiomycetae en Homobasidio

-mycetae Aphyllophoraceae. Deel 1. Wageningen.

(1933). Revision der Niederlandischen Homobasidiomycetae-Aphyllophor-

aceae. Deel ii. Amsterdam-Haarlem.

Findlay (1951). Trans. Brit. Myc. Soc. 34, p. 146.

Gaumann (1922 a). Ann. Mycol. 20, pp. 257-271.

(1922 b). Ann. Mycol. 20, pp. 272-289.

(Transl. Dodge, 1928). Comparative morphology of Fungi. McGraw Hill.

Gregor (1932). Ann. Mycol. 30, p. 464.

Grove (1913). The British Rust Fungi, p. 14, Fig. 16. C.U.P.

Heim (1936). Rev. de Mycologie 1, pp. 223-256.

(1948). Rev. de Mycologie Suppl. Colonial 13, pp. 1-12.

(1949). Rev. de Mycologie Suppl. Colonial 14, pp. 1-14.

Hohnel & Litschauer (1907). K. Akad. Wiss. Wein. Sitzungsb. 116.

Ingold (1940). Trans. Brit. Myc. Soc. 24, pp. 30-31, Figs. 1-2.

299

i

JUEL

Knoll

Langeron .

Levine

Linder. . . .

Lloyd

Lohwag. . .

Lowe

Malencon.

Maire

Martin

Massee

McGuire

Nannfeldt . .

Neuhoff

Nyland

Olive

Overholts. . .

Patouillard,

Petch. .

PlLAT. . ,

Rogers.

Rogers & Jackson

Shear & Dodge

Singer

Talbot

Thompson

Tulasne

Van Tieghem

Wakefield

Wakefield & Pearson

Whelden

(1897). Bih. Svenska Vet.-Akad. Handl. 23, iii, 12, pp. 1-27.

(1898). Jahrb. Wiss. Botanik 32, pp. 361-388.

(1912). Jahrb. Wiss. Botanik 50, pp. 453-501.

(1945). Precis de Mycologie. Paris.

(1913). Bull. Torrey Bot. Club 40, pp. 137-181.

(1940). Mycologia 32, pp. 419^147.

(1923). Mycological Notes 7, p. 1232.

(1947). Ann. Mycol. 35, pp. 157-193.

(1940). Ann. Mycol. 38, pp. 401-450.

(1942). Polyporaceae of the New York State. N.Y. State College of

Forestry, Tech. Publ. No. 60.

(1939). Bull. Soc. Myc. de Fr. 55, pp. 42-52, Figs. 3, 4.

(1902). Recherches cytologiques & taxonomiques sur les Basidiomycetes.

Bull. Soc. Myc. de Fr. 18.

(1934). Mycologia 26, p. 264.

(1935). Bull. Torrey Bot. Club. 62, pp. 339-343, Fig. 2.

(1936). Mycologia 28, p. 215, Fig. 1, 9.

(1938). Anier. Journ. Bot. 25, pp. 682-685.

(1939). Mycologia 31, pp. 507-518.

(1945). Mycologia 37, p. 540.

(1948). Lloydia 11, pp. 111-122.

(1887). Joum. Roy. Microsc. Soc. pp. 205-208.

(1941). Lloydia 4, p. 16, PI. 2, Figs. 18, 19.

(1947). Svensk Botanisk Tidskrift 41, p. 332.

(1924). Botan. Archiv. 8, pp. 250-297.

(1949). Mycologia 41, p. 688.

(1950). Mycologia 42, p. 389, Fig. 2, 17, 20, 21.

(1929). Proc. Internat. Congress Plant Sci. 1926, Ithaca, 2, pp. 1688-1712.

(1894). Bull. Soc. Myc. de Fr. 10, pp. 77-78.

(1899). Bull. Soc. Myc. de Fr. 15, pp. 199-200.

(1900). Essai taxonomique sur les families et les genres des Hymenomycetes,

Lons-le-Saunier.

(1915). Ann. Roy. Bot. Gard. Perad. 6, pp. 43-68, Pis. 2-7.

(1926). Ann. Mycol. 24, pp. 201-230.

(1932). Bot Gazette 94, pp. 86-105.

(1933). Ann. Mycol. 31, pp. 181-203.

(1934). Univ. of Iowa Stud, in Nat. Hist. 16, pp. 160-183.

(1935). Univ. of Iowa Stud, in Nat. Hist. 17, pp. 3-43.

(1936). Mycologia 28, pp. 347-362.

(1943). Farlowia 1, p. 115.

(1944). Mycologia 36, pp. 70-103.

(1947). Mycologia 39, pp. 556-564.

(1943). Farlowia 1, pp. 263-336.

(1925). Journ. Agric. Res. 30, pp. 407-417.

(1945). Lloydia 8, pp. 170-230, PI. 3, Fig. 9.

(1947). Farlowia 3, pp. 155-157.

(1952). Trans. Brit. Myc. Soc. 35, pp. 123-128.

(1936). Mycologia 28, p. 223, Figs. 1-13.

(1853). Ann. Sci. Nat. ser. iii, 19, p. 196.

(1893). Jour, de Bot. (Morot) 7, p. 77 et. seq.

(1911). Trans. Brit. Myc. Soc. 3, p. 280.

(1913). Trans. Brit. Myc. Soc. 4, pp. 113-120, PI. 3, Fig. 3-5.

(1935). Differential characters in some resupinate Hymenomycetes (Manu-

script). Summary in Proc. 6th Internat. Bot. Congr. (Amsterdam).

Leiden, 1935.

(1920). Trans. Brit. Myc. Soc. 6, pp. 319-320.

(1923). Trans. Brit. Myc. Soc. 8, p. 218, Fig. 4.

(1934). Mycologia 26, pp. 415-435.

(1935 a). Mycologia 27, pp. 41-57.

(1935 b). Mycologia 27, pp. 503-520.

(1935 c). Rhodora 37, p. 121.

(1936). Amer. Journ. Bot. 23, p. 544.

301

A Note on Sterigmata in General.*

By

M. A. Donk. f

The sterigmata in Tulasnella and Gloeotulasnella develop in three phases. First,

they grow out into the swollen, spore-shaped bodies which have attracted so much

attention. Secondly, after these voluminous bodies have reached their maximal size

and gone through an outwardly stationary period (nuclear division in certain species),

they emit an outgrowth, rather conical and short in Tulasnella and rather finger-shaped

and longer in Gloeotulasnella. Finally these outgrowths produce apically pointed

tips on which are formed the basidiospores to be shot away at maturity. I propose

to call the successive portions protosterigma, secondary protosterigma, and spiculum (x)

respectively.

The formation of a secondary protosterigma is not of common occurrence among

hymenomycetes. Except in Tulasnella (where they may be indistinct) and Gloeotulas-

nella it is encountered in many gelatinous Fleterobasidiae, like Tremellaceae, for

instance in Tremella Fr. and Exidia Fr. (2). In these instances the basidia are deeply

embedded in a gelatinous matrix. They send out towards the surface (primary)

protosterigmata, starting as globular buds which elongate into hypha-like threads

and, after having stopped growth just below the surface of the fruit-body, become

more or less clavate in their upper portion. Much of the protoplasmatic contents

of the basidium-cells accumulates here. Next this clavate and rounded apical portion

emits a more slender outgrowth, the secondary protosterigma, which, as it reaches

beyond the surface, produces the spiculum and the basidiospore. It would appear

that the differentiation into a primary and secondary protosterigma is largely due to

a temporary interruption of the growth of the sterigma. This may be interpreted as

a rest-period, but it is more likely to be one of internal activity preparatory to the

quick formation of the spiculum and ballistospore at such a great distance from the

basidial body which usually plays the important role in the process of violent discharge

of the spore. More often the protosterigma develops without interruption until it

has formed its spore and no secondary protosterigma is discernible.

The definition of a protosterigma should perhaps be: that part of the sterigma

which bears the spiculum, or, in terms of development: a sterigma before it has formed

the spiculum. (The full-grown sterigma might be termed a metasterigma.) This makes

the protosterigma an outgrowth of the basidium, which, either very soon when it is

still a minute, more or less rounded bud, or after growing out characteristically and

considerably, forms the spiculum; it also makes the protosterigma a universal feature

of those basidiomycetes that discharge their spores forcibly. It may be assumed that

even the delicate sterigma starts as a nipple- to knob-like protuberance that rapidly

produces the spiculum.

* This note originally formed part of a manuscript dealing with what I would call the tulasnelloid

fungi. These are, for instance, characterised by their basidiospores exhibiting repetition.

t Keeper of Herbarium Bogoriense, Kebun Raya Indonesia. Published with permission of the

Director of Kebun Raya Indonesia.

C) The term “ spiculum ” is taken from L. R. Tulasne (in Ann. Sci. nat., Bot. iii 19: 196, 1853)

who interchanged it with “ sterigma Jt is here restricted, in accord with its etymology, to the pointed

tip of a sterigma upon which the violently discharged spore is produced.

(2) Compare G. W. Martin in Mycologia 37: 534, 1945. For figures, see for instance, Brefeld

(Unters. Gesamtgeb. Mykol. 7: pis. 5-7) and G. W. Martin fin Univ. Iowa Stud. nat. Hist. 18 (3):

f. 17, 1944.

302

That the protosterigma is often not noticed would be accounted for by the facts

(i) that its formation and the subsequent production of the spiculum take such a short

time that the intermediate stages are only exceptionally seen, and (ii) that the smaller

sterigmata when full-grown do not show any appreciable outer differentiation into

protosterigma and spiculum. This lack of differentiation at maturity is typical of all

horn-shaped (outward curved and slender-conical) sterigmata, but yet I have been able

to find the bud-like protosterigma in many cases and in modern literature it is even

now and then depicted. Among the heterobasidious fungi it is usually considerably

developed, often up to hypha-like or slender-clavate ; in Ceratobasidium and in certain

species of Botryobasidium it is finger-shaped and usually somewhat ventricosely swollen

at the time that the sterigma starts producing the spore; in Tulasnella it is a voluminous,

ovoid body before it grows out further.

Protosterigmata, when they are well developed, are now often called epibasidia,

and simultaneously the term sterigma is restricted to the spiculum. Since I can neither

accept a term that covers besides protosterigmata ( Tulasnella , Tremella ) also metabasidia

{Auricularia), nor a terminology that does not homologize the sterigmata of Tremella

with those of Auricularia, I have abandoned such a confusing term as epibasidia

altogether (3). For those appendages of the metabasidia called epibasidia by Neuhoff

and Rogers the new term protosterigmata might appear a convenient substitute, but

it should not be forgotten that less well developed protosterigmata occur throughout

the hymenomycetes. “ Epibasidia ” may be very small and practically totally reduced

even in Tremellales; compare, for instance Sebacina obscura G. W. Mart (4).

Not all hymenomycetes exhibit violent spore discharge. There are a few genera

that produce sticky spores which become freed by hydrolysis (slime spores) or by

breaking off by mechanical disturbance from outside (dry spores) (5). Examples of

the first are Hoehnelomyces Weese (Auriculariaceae-Phleogenoideae) and Sirobasidium

Lagerh. & Pat. and Hyaloria A. Moll (Tremellaceae); of the second Phleogena Link.

(Auriculariaceae-Phleogenoideae) and Xenolachne D. P. Rogers (Tremellaceae). Basidia

that do not forcibly discharge their spores have been called apobasidia by D. P.

Rogers (6); if such a term is needed at all, one is forced to provide also for a companion

term for basidia that do (ballistobasidia).

Spores that are not forcibly discharged may be either sessile {Phleogena, Hoehnelo-

myces, Sirobasidium) or borne on sterigmata {Hyaloria, Xenolachne). In the latter

case the sterigmata lack the spiculum (which evidently plays an important role in the

process of ballistospore discharge) and the whole of them would appear comparable

to the protosterigma, a term which thus becomes superfluous here.

(3) Donk in Meded. Nederl. Mycol. Ver. 18-20: 78-81. 3 fs. 1931.

(4) G. W. Martin in Lloydia 7: 70. f. 5. 1944.

(5) Mason, Annot. Account Fungi reed, List ii, Fasc. 3: 77, 1937.

(6) In Mycologia 39: 558, 1947. It should be recalled that the term apobasidium was originally

coined by Vuillemin (Champignons 349, 1912) for a much more inclusive notion.

303

The Genus Stereum in South Africa.

By

P. H. B. Talbot.

A preliminary examination of the many collections of Stereum in the National

Herbarium, Pretoria, suggested that these were greatly in need of revision. The result

of this work has been that of sixty-seven recorded names in Stereum only twenty-two

are accepted here as good species for South Africa.

The task of untangling the South African records has been considerably lightened

by Dr. E. M. Doidge’s check lists of species and pertinent literature (in Bothalia 5,

1950). This immensely valuable book, being largely a compilation, is inevitably a

record of wrong identifications as well as correct ones; but its value is that it is a

faithful and virtually complete record. The reader is referred to Doidge’s lists for

literature relating to the species of the genus Stereum in South Africa. In much of the

cited literature, the species are merely listed without description or illustration, but

there are papers by van der Byl, Lloyd, and Wakefield, which treat of taxonomy and

form the basis of our knowledge of this genus. Insofar as they affect South African

species, Lloyd’s comments are seldom very clear or disciplined.

In general the literature citations in this paper are limited to the original place

of publication of a species or record, and a reference to Doidge’s check lists (Bothalia 5,

1950). Any attempt to include full references to foreign descriptions of each species

would have been thwarted by the absence here of much of the important literature.

For the same reason the synonymy of each species is not given in full but is generally

limited to species of which type or authentic material (denoted by an exclamation

mark after the name of the species) has been seen by the writer.

Specimens are cited by number only, except for establishing synonyms. Unless

otherwise denoted, these numbers refer to collections in the National Herbarium,

Pretoria. A fuller citation may be traced by following up these numbers in Doidge’s

check lists.

In sorting out the herbarium material into taxonomic species before naming

them, it was considered that minor variations in colour, size or hairyness were relatively

unimportant. Habit and general appearance of the plants, linked with characteristic

microscopic features, were the chief guide in grouping specimens into taxonomic

species. Microscopic characters were stressed since in a group of species of the same

habit they provide reliable constant differences. This treatment has possibly resulted

in a certain amount of merging together of species, e.g. in the merging of S. kalch-

brenneri with S. hirsutum. It was felt that this was desirable if not carried to extremes.

The materials used in this study were all dried herbarium specimens, unless

otherwise stated. The mountant employed in microscopic work was 5 per cent

potassium hydroxide solution with the addition of 1 per cent aqueous phloxine as

stain. Measurements given for “ thickness in section ” exclude the hairy zone of the

abhymenial surface. Such measurements were made on sections mounted as above,

but not squashed out. For a clear view of the microscopic organs it is generally

necessary to squash the sections gently under the cover glass.

The records of Stereum for South Africa are arranged in alphabetical order of

the specific epithets. Synonyms and doubtful or excluded species are printed in italics;

accepted species are printed in bold face. A key to the accepted species is given at

the end of the paper.

304

For the loan of specimens, the author is much indebted to the following insti-

tutions: Herbarium Royal Botanic Gardens, Kew; Herbarium, Paris Natural History

Museum; Herbarium, British Museum (Natural History); Herbarium, South African

Museum; Myco logical Herbarium, Department of Agriculture, Southern Rhodesia.

Special thanks are due to Dr. R. W. G. Dennis for his kindness in comparing several

specimens with material in Kew Herbarium.

Deur die goedgunstigheid van die trustees van die Herbarium P. A. van der By l,

en van die Stellenbosse Universiteitsraad, is 'n geleentheid vir die bestudering van wyle

Prof, van der Byl se versamelings van Stereum soorte aangebied. Verwysing na hierdie

monsters sal in verband met die verskillende soorte gevind word.

Stereum Persoon ex S. F. Gray, A Natural Arrangement of British Plants 1 (1821)

652; Persoon in Roemer Neues Mag. Bot. 1 (1794) 110, Obs. Myc. 1 (1796) 35.

Fructifications coriaceous, membranous or subllgneous, stipitate, sessile, effuso-

reflexed or resupinate, or dimidiate, or infundibuliform, simple or branched. Stem

lateral or central or absent. Hymenium inferior, smooth, sometimes rugose or

exceptionally tubercular. Flesh pale, the context usually having an intermediate layer

of more or less horizontally arranged hyphae. Spores hyaline, smooth. Cystidia,

gloeocystidia or vesicles present or absent. Setae absent. Annual or perennial,

lignicolous or terrestrial.

The genus Stereum Pers. ex S. F. Gray is accepted as validly published without

conservation and is typified by the species S. hirsutum (Willd.) Pers. ex S. F. Gray.

Reasons for the selection of this species as the type are given by Rogers (in Farlowia

3, 1949, pp. 450 & 480) and by Donk (in Bull. Bot. Gard. Buitenzorg ser. iii, 18, 1949,

pp. 98-99).

(1) Stereum adnatum Lloyd (!) Myc. Notes 7 (1925) 1336, Fig. 3093; Doidge

loc. cit. p. 487.

= Stereum rimosum Berk. var. africanum Talbot (!), for reasons given in Bothalia

6 (1951) 39.

(2) Stereum affine Lev. in Ann. Sci. Nat. ser. iii, 2 (1844) 210; Saccardo Syll.

Fung. 6 (1888) 559; Doidge loc. cit. p. 487.

Fig. 6.

Pileus thin, coriaceous, lignicolous, solitary or gregarious, stipitate, arising from

a circular light-buff coloured mycelial pad 3-6 mm. in diam., the pad sometimes being

common to more than one pileus and the pilei in this case sometimes uniting above.

Pileus usually flabellace, rarely infundibuliform and then sometimes split down one

side, radius 1-3 cm. (rarely up to 7 cm.) from attachment, width 0-5-2 cm. Large

specimens may be deeply divided into a number of flabellate pileoli with a common

cuneate base, but this condition is rare. Surface smooth, yellow-brown to bay or

chestnut, with a “ shot ” lustre in a radial direction, not colour-zoned, sparsely

pruinose with scanty hairs, glabrescent. Hymenium light buff, pinky buff or light

reddy brown when old, smooth. Margin concolorous or somewhat paler on the

abhymenial surface, thin, undulate or with small incisions, often reflexed in infun-

dibuliform specimens. Stipe 1-2-5 mm. in diam., 0-5-2 cm. long, light buff colour,

minutely velutinate. Thickness in section 560-800 /x.

Basidia: 3-2-4 x 23-27 /x, subcylindrical, compact.

Spores: abundant, 3-2 x 4 /x, hyaline, subglobose, ovate or broad elliptical,

smooth, frequently uniguttulate.

Gloeocystidia: (8)— 9 * 6— 12 - 8 x 50-70-(100) fj., hyaline, with homogeneous,

contents, subulate or fusoid or irregularly cylindrical, rounded at apex, base attenuated,

found in the hymenium and arising from subhymenial hyphae, abundant.

305

Hyphae: Skeletal hyphae 3-2-4 ^ wide, hyaline, thick- walled, not septate,

unbranched, without clamps; Generative hyphae hyaline, 2 -4-3 -2 /x wide, thin-walled,

septate, scantily branched, with rare clamp connections.

Tissue differentiation: A faint, narrow, yellowish zone subtends the abhymenial

surface from which the scanty hairs arise.

Abhymenial hairs: 9-6-12-8 x 50-180 /x, hyaline, scanty, solitary, thick-walled

with wide lumen, septate or non-septate, simple or occasionally forking.

Specimens examined: 28310, 36716, 30880, 15555, 27774, 27335, 27336, 31653,

9206, 28911, 11628, 12048, 36799, 32474, 34227, 34952, 36869, 14909 (a): Universiteit

van Stellenbosch, Herbarium P. A. van der Byl Nos. 145, 694, 514 (as S. glabrescens).

Several South African specimens filed under Stereum glabrescens B. & C. prove

to be S. affine. Burt (in Ann. Mo. Bot. Gard. 7, 1920, 110) states that S. glabrescens

lacks gloeocystidia, which are present in all the South African specimens which the

writer has seen in this group. Wakefield (in Det. Kong. Norske. Vidensk. Selsk. Forh.

9, 1936, 52) suggested that the specimens referred to S. glabrescens by van der Byl

(in Trans. Roy. Soc. S. Afr. 10, 1922, 151, Fig. 1 and in Ann. Univ. Stellenbosch 7,

1929, 37) were probably S. affine. The writer has examined v.d. Byl’s material and

confirms that it is S. affine, showing quite obvious gloeocystidia when stained with

phloxine.

Apart from gloeocystidia, it is suggested that another difference between these

two species is that S. glabrescens is always flabehiform while S. affine may sometimes

be infundibuliform.

(3) Stereum albo-badium (Schw. ex Fr.) Fries: Recorded by Kalchbrenner in

Grev. 10 (1881) 58; Doidge loc. cit. p. 491.

The material referred to by Kalchbrenner as “ Stereum albo-badium Fr. Ep. 551

C.B. Spec, (sic.), in mont. Boschberg ”, was seen at Kew with the sheet annotated

thus: “ Stereum albo-badium Schwein. Afr. austral. Type of Kalchbrenner, Com.

MacOwan 9/83

This specimen lacks cystidia and branched paraphyses and cannot be S. albobadium.

The material is so scanty that it probably cannot be named with certainty.

S. albobadium is a North American species of characteristic appearance. It is

resupinate with a free margin. The hymenium is umber-bay in colour, and velvety.

The margin is narrow and whitish. Microscopically it has brownish branched

paraphyses and small encrusted cystidia. Ravenel material (not type) was seen at Kew.

(4) Stereum amoenum Kalchbr. & MacOwan (!) in Grev. 10 (1881) 58; Theumen

in Flora 59 (1876) 424; Doidge in loc. cit. p. 490 (Nec. S. amoenum Lev.).

This species was described from South Africa. As this name was preoccupied,

Saccardo redescribed the species under the name Stereum kalchbrenneri Sacc. (in.Sacc.

Syll. Fung. 6, 1888, 568). S. amoenum is thus an obligate synonym of S. kalchbrenneri.

Lloyd (Myc. Notes 4, 1915, L. 60, 10, Note 341) notes these name changes under

S. kalchbrenneri.

The writer has examined authentic material of S. amoenum in Herb. Macowanianum

Nos. 1084 & 1086 sub Herb. S.A. Museum No. 34269 as S. kalchbrenneri. On the

herbarium sheet, van der Byl has noted that S. kalchbrenneri is very close to S. hirsutum

and differs virtually only in having a more luxuriant fructification and darker hairs

on the upper surface than is usual in S. hirsutum. He added that they could scarcely

be considered as separate species.

The writer’s impression of S. kalchbrenneri, gained from examining MacOwan’s

specimens, is that this species is characterised by a combination of luxuriant growth,

rather dark brown hairs, a reddish hymenium and a tendency to develop cystidioid

hyphae. The hymenial colour varies between pinkish buff, cinnamon, mikado brown

and russet (Ridgway Pis. xxix & xv) while the colours seen on the abhymenial surface

are warm buff, clay colour, dark Sudan brown (Ridgway Pis. xv, xxix, iii) and whitish.

306

In the National Herbarium, Pretoria, the specimens showing the best agreement with

MacOwan’s material of S. kalchbrenneri are Nos. 31454, 11255 and 13793. However,

it is quite impossible to separate this species from S. hirsutum owing to the considerable

variation and intergrading seen in the specimens assigned to each. This is illustrated

by comments on the herbarium sheets such as “ pallid form of S. kalchbrenneri ” and

“ dark form of S. hirsutum ”. Since no constant difference can be demonstrated between

the two, it is thought best to refer S. amoenum and S. kalchbrenneri to Stereum hirsutum

(Willd.) Pers. ex S. F. Gray.

It may be noted here that Bresadola (in Ann. Myc. 14, 1916, 232) cites S. amoenum

Kalchbr. & MacOwan and S. vellereum Berk, as synonyms of S. friesii Lev. The

writer doubts whether this can be substantiated.

(5) Stereum atrocinereum (Massee) van der Byl in Ann. Univ. Stellenbosch

7 (1929) 44; Doidge loc. cit. p. 493.

Peniophora atrocinerea Massee (!) in Journ. Linn. Soc. Bot. 25 (1889) 141.

As noted by Doidge, this is a synonym of Stereum schomburgkii Berk. (!). Reasons

for this conclusion were given by the writer in Bothalia 6 (1951) 44.

The material of Herb. MacOwanianum No. 1197 sub Herb. S.A. Museum No.

34284, on which van der Byl based his description of S. atrocinereum, is undoubtedly

part of the type number of Peniophora atrocinerea, which the writer has also seen in

Kew Herbarium. The part in Herb. S.A. Museum shows spores which are hyaline,

cylindrical-depressed or bread elliptical, 3-4 x 6-8 p.

As previously recorded, MacOwan’s material is a pale form of S. schomburgkii,

which in turn is now found to be synonymous with Stereum fulvum (Lev.) Sacc .

Stereum atrocinereum is accordingly referred to Stereum fulvum (Lev.) Sacc. (see

p. 315).

(6) Stereum australe Lloyd (!) in Lloyd Myc. Notes 4 (1913) L. 48, 10, Note 115;

Ibid. 4(1915) L. 60, 15, Note 387; Ibid. 5(1917) L. 65, 2; Doidge loc. cit. p. 487.

Stereum tenebrosum Lloyd (!) nomen nudum in Lloyd Myc. Notes 5(1918) L. 67,

16, Note 692; Doidge loc. cit. p. 493.

As “ Stereum lobatum with cinereous hymenium ”, Lloyd in Lloyd Myc. Note

4(1913) L. 46, 3.

As “ Stereum lobatum (Kunze) Fr. var. cinereum Lloyd ”, Doidge loc. cit. p. 487 •

Stereum transvaalium v.d. Byl (!) in Ann. Univ. Stellenbosch 7(1929) 41; Doidge

loc. cit. p. 494.

Fig. 13

Pileus tough, coriaceous, lignicolous, solitary or more often gregarious, often

laterally connate, occasionally imbricate, flabellate or cuneate, attached by a reduced

base, or reflexed and attached by a long narrowly-effused base, or occasionally orbicular

and sessile attached by a central umbo; 2-5-4 cm. radius X 3— 5— ( 1 0) cm. wide.

Surface concentrically furrowed and zoned with velutinate hairs of reddy-brown to

light yellow-brown to greyish colour, becoming worn off and smooth in the ridges

when old and weathered. Hymenium reddish-brown when moist becoming cinereous

to cinereous-buff on drying, smooth, reflecting the abhymenial furrows, “ bleeding ”

red when fresh and bruised. Thickness in section 640-1000 p.

Basidia: compact, hyaline, clavate, 4^4-8 X 30 p.

Spores: uncertain. Possibly 2-5 x 5-6 p, hyaline, smooth, elliptical, with one

'side depressed. (3 x 4 p fide van der Byl; 4 x 6 p fide Lloyd).

307

Conductors: originating as modified skeletal hyphae in the trama and subhymenium,

curving into but not beyond the hymenium, thick-walled, non-septate, contents brownish,

walls hyaline, lumen often widening towards the apex, 4 -8-6 -4 /z diam.

Hyphae : Skeletal hyphae thick- walled, non-septate, unbranched, hyaline or some-

times dilutely coloured, 6-4 /z diam.; Generative hyphae rather thin-walled, hyaline,

septate, without clamps, 3 - 2 /z diam. The two hyphal types are intertwined throughout

the trama.

Tissue differentiation: Tissue hyaline above a horizontal yellow-brown zone

which subtends the abhymenial hairs.

Abhymenial hairs: hyaline, thick-walled (3 -2)-4-8-6-4 /z diam., fasciculate.

Specimens examined: 27519, 27520, 31033, 30890, 27522, 28490, 15557, 26390,

31816, 28966, 27611, 27721, 15559 (2 specimens), 30269, 30270, 28848, 31852, 40218

(T.R.L. 202), 40219 (T.R.L. 69), 40220 (T.R.L. 2); 8847, 1464 (as S', tenebrosum det.

Lloyd); Universiteit van Stellenbosch, Herbarium P. A. van der Byl No. 1472 (Type

of S. transvaalium).

In Lloyd’s writing cited above, there is a slight ambiguity in Letter 48, but the

writer takes this letter to mean that Lloyd recognised S. australe and S. tenebrosum

as the same species, i.e. that they are both the same as his “ S. lobatum with cinereous

hymenium ” mentioned in Letter 46. This opinion is confirmed by examination of

specimens in Herb. Pretoriae which were determined by Lloyd variously as S. australe

and S. tenebrosum and which the writer is confident represent only one species.

5. australe is well characterised by its smallish brown, velutinate pilei , the

cinereous hymenium (when dry) and the presence of conductors. Old specimens

may weather greyish with bare chestnut zones and then resemble S.fasciatum externally.

Regarding the inclusion of Stereum transvaalium as a synonym of S. australe,

see notes given under the former species on p. 328.

(7) Stereum bellum (Kunze) Saccardo, Syll. Fung. 6(1888) 563; Doidge loc. cit.

p. 488.

Thelephora bellum Kunze in Flora (1830) 370.

Fig. 8

Pilei small, dimidiate, sessile, laterally connate, imbricate, lignicolous, about

1 cm. radius from attachment and attenuated towards the base thus somewhat cuneate

in single specimens, rigid, tough and not flexible. Surface reddy-brown, often concen-

trically zoned and with a relatively wide light tan coloured margin, rather sparsely

velutinate. Hymenium smooth, yellow-orange colour. Thickness in section 720-

1200 |tz.

Basidia: (immature) cylindrical, 20 x 32 jz.

Spores: few seen, hyaline, ovate or subglobose, 2 X 2-5 /z.

Gloeocystidia : quite numerous in the hymenium, hyaline, thinwalled, deep-

staining, pyriform to fusoid or ventricose, 7-5-12 x ( 1 2)— 1 5— 25— (40) /x.

Cystidia: rare (see comment below), hyaline, thickwalled, apically encrusted, or

smooth (6 - 5)— 10— 12 x 30-37 /z; in the hymenium, fusoid to conical.

Hyphae: all hyaline and closely intertexed; skeletal hyphae non-staining, non-

septate, very seldom branching, thick- walled with a narrow or occasionally fairly wide

lumen, 3 -2-4 -8 fx wide; generative hyphae thin-walled, deep-staining, tortuous, with

clamps, branched, 2-4-3 -2 ^ wide.

Tissue differentiation: tissues uniformly composed of closely intertexed hyphae

without a notably differentiated abhymenial zone.

Abhymenial hairs: hyaline to pale straw colour, thickwalled with narrow lumen,

usually widening somewhat towards the apex; apex rounded or sometimes pointed;

hairs usually with 2-4 septa, 3-5-8 /z wide and projecting 80-104 /z from the surface.

308

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der B>1

No. 645.

In the absence of other material for comparison there is some doubt whether

this specimen is correctly named, but it is thought best to record this under van der Byl’s

determination as Stereum bellum. The specimen shows several features which suggest

an affinity with Stereum involutum, yet is sufficiently different in detail to be held apart.

It is noted by Lloyd (Syn. Stip. Stereum, 1913, 41) that S. bellum sensu Bresadola

non Kunze is Stereum bresadoleanum, a species synonymous with S. involutum. See

also under S. friesii (p. 314).

The cystidia described above were clearly seen in only one section made from

this specimen and may be merely a thickened and encrusted form of the gloeocystidia .

Gloeocystidia were seen in all sections. As far as habit is concerned, this specimen

might be taken casually for a small and immature S. hirsutum, but its reddish velutinate

surface is different, and the microscopic structure quite different. Saccardo (loc. cit.)

mentioned that S. bellum has an affinity with S. hirsutum.

(8) Stereum bicolor (Pers. ex Fr.) Fries, Epicrisis (1838) 549, Hym. Eur. (1874)

640; Doidge loc. cit. p. 488; Talbot in Bothalia 6 (1951) 39.

Thelephora bicolor Persoon, Syn. Meth. Fung. (1801) 568, Myc. Eur. 1 (1822)

122; Fries, Syst. Myc. 1 (1821) 438.

Stereum laxum Lloyd (!) in Lloyd Myc. Notes 4 (1915) L. 60, 10, Note 339;

Doidge loc. cit. p. 491.

Stereum fuscum (Schrad.) Quelet, Flor. Myc. de Fr. (1888) 14.

Stereum coffeatum Berk. & Curt. (!) in Grev. 1 (1873) 164.

Stereum pannosum Cooke (!) in Grev. 8 (1879) 56.

Fig. 21.

Occasionally entirely resupinate, usually resupinate-reflexed, sometimes pileate,

imbricate, soft, spongy texture. Abhymenial surface snuff-brown, concentrically

furrowed, floccose, becoming smooth; margin paler. Hymenium smooth, whitish to

creamy, not furrowed, sometimes rimose when dry, rather velvety. In section up

to 1000 p thick; hymenial layer hyaline, circa 100 p thick, the remaining tissues coloured

brownish. Not adnate.

Basidia: 3 -5-5 -5 X 22-27 p , cylindrical, with 2-4 sterigmata.

Spores: hyaline, smooth, elliptical, unilaterally depressed, or oblong, frequently

uniguttulate, 3-4*5 x 2-3 p.

Gloeocystidia: very abundant in the hymenium, not usually emergent, hyaline,

very refractile, thinwalled, cylindrical, fusoid or somewhat ventricose, about 4—1 1 X

90 p, borne on very narrow hyaline hyphae about 2 p wide, contents homogeneous

when young but appearing to solidify and become fragmented when old and then

resembling cystidia or mineral concretions.

Hyphae: all smooth, thin-walled, with occasional to numerous clamp connections,

much branched, septate; Subhymenial hyphae colourless or pale brownish; Tramal

hyphae brown, 3-6 p wide, loosely intertexed.

Tissue differentiation : There is no dark compact zone subtending the abhymenial

surface; that surface is floccose and composed of ordinary brown hyphae.

Specimens examined: 27596, 28964 (T.R.L. 200), 35422, 30891, 27784, 28555,

27766, 31728, 31851, 30698; van der Byl (2239) in Kew; MacOwan (1244) C.B.S.

in Kew; MacOwan (1244) as Thelephora biennis Fr. sub Stereum fuscum in Herb.

S.A. Museum No. 34292.

309

The type number of Stereum laxum Lloyd, 31321, A. V. Duthie (56), on decaying

leaves and twigs, Stellenbosch, agrees in every respect with S. bicolor and is accordingly

reduced to synonymy. Lloyd (loc. cit.) wrote that S. laxum has no cystidia, but its

gloeocystidia are quite obvious. His suggestion that S. laxum resembles an Hypochnus

in context, is also rather misleading.

S. bicolor has also appeared in South African literature (and elsewhere) under the

name S. fuscum (Schrad.) Quel., but the latter is nomenclaturally inacceptable (see

Talbot, loc. cit.). Doidge (loc. cit. p. 488) lists records of S. bicolor which were

improperly placed by other authors under Thelcphora biennis.

(9) Stereum bresadoleanum Lloyd (!), Syn. Stip. Stereum in Lloyd Myc. Notes

4(1913) 41; Doidge loc. cit. p. 488.

= Stereum involutum (Klotzsch) Fries; see p. 317.

(10) Stereum caperatum Lloyd (!) in Lloyd Myc. Notes 4 (1916) 549, Fig. 751-

[non S. caperatum (Berk. & Mont.) Massee]; Doidge loc. cit. p. 494.

As this name was a later homonym of S. caperatum (Berk. & Mont.) Massee,

Lloyd subsequently changed to it Stereum turgidum Lloyd (!), (Lloyd in Myc. Notes

5, 1916, L. 63, 15, Note 502). For reasons given in Bothalia 6 (1954), p. 339, the latter

species is regarded as synonymous with Stereum cinerascens (Schw.) Massee (!). See

also the following description.

(11) Stereum cinerascens (Schw.) Massee (!) in Journ. Linn. Soc. Bot. 27 (1890)

179; Doidge loc. cit. p. 488; Talbot in Bothalia 6 (1951) 40 and Ibid. 6 (1954), p. 339.

Thelephora cinerascens Schwein. in Amer. Phil. Soc. Trans, n.s. 4 (1832) 167.

Stereum turgidum Lloyd (!) in Lloyd Myc. Notes 5 (1916) L. 63, 15, Note 502;

Stevenson & Cash in Bull. Lloyd Library 35 (1936) 58; Doidge loc. cit. p. 494.

Stereum caperatum Lloyd (!) in Lloyd Myc. Notes 4 (1916) 549, Fig. 751 [non.

S. caperatum (Berk. & Mont.) Massee].

Lopharia mirabilis (B. & Br.) Patouillard in Bull. Soc. Myc. de Fr. 11 (1895) 14,

PI. 1 ; Doidge loc. cit. p. 501 ; Talbot in Bothalia 6 (1951) 56 and Ibid. 6 (1954), p. 339.

Radulum mirabile Berk. & Br. (!) in Journ. Linn. Soc. Bot. 14 (1873) 61.

Lopharia lirellosa Kalchbr. & MacOwan (!) in Grev. 10 (1881) 58.

Fig. 18.

Fructifications coriaceous, resupinate, effused with a narrow reflexed margin, or

effuso-reflexed ; not adnate. Abhymenial surface (when exposed) ochraceous, ashen

or warm buff, tomentose, obscurely zoned or concentrically furrowed. Hymenium

first smooth and whitish, later creamy-ochraceous or pinkish buff. Hymenial variations

include smooth or slightly scabrid states, papillate or tubercular developments, or

strongly developed warts or teeth or incised ridges arranged irregularly or in somewhat

concentric patterns and sometimes forming incomplete shallow pores. Thickness in

section, excluding ridges or teeth, 250-800 p.

Basidia: 40-65-(80) x (5)— 9—1 1 p, clavate, with 4 sterigmata.

Spores: hyaline, smooth, oblong-cylindric or oblong-elliptical, often with one

side depressed, 5-5-14-3 x 4 -4-8 -8 p (usually 10-6-11-8 x 6-1 p).

Cystidia: embedded or projecting, heavily encrusted with large crystals, thick-

walled, often faintly coloured at the base, conical or subfusiform, apex blunt or pointed,

50-150 x 12-24 p. They may be emergent up to 50 p or embedded in stages

throughout the trama.

310

Hyphae: 3-5 p diam., hyaline or faintly coloured, rather opaque and thick-walled,

those next to the substratum forming a denser, coloured layer.

Abhymenial hairs: adpressed or suberect, coloured, 3-5-5 p diam.

Specimens examined: As S’, cinerascens 28688, 28498, 28926, 34377, 35421, 35309,

33077, 33213, van der Byl (2732); As S', turgidum Lloyd & S. caperatum Lloyd, type

number 31332 (A. V. Duthie, 74); As Lopharia lirellosa, type, P. MacOwan, C. Bon.

Spei, in Herb. Kew; As Lopharia mirabilis, type, Thwaites 328, Peradeniya, Ceylon,

1868, in Herb. Kew; 27797, 28302, 27799, 31309, 31356, 31911, 27769, 28697, 31397,

28299, 34553, 36786, 33205, 36785, van der Byl (2261,2620, 551, 1429); as Raduium

lirellosa, 31309; as Raduium sp., 27556.

An extensive comparison of this species with Lopharia mirabilis and Stereum

turgidum, leading to the sinking of the genus Lopharia, is given by the writer in Bothalia

6 (1954), p. 339. There, the variability of this fungus is discussed and reasons are

given for associating under one name, S. cinerascens, what at first sight appear to

be totally disrelated forms, some with smooth hymenia and some with highly convoluted

hymenia apparently characteristic of the Hydnaceae. That these form the extremes

of a graded series with identical microscopic characters, is only realised when a large

number of collections have been seen and carefully studied.

(12) Stereum cinereum Lev.; recorded by van der Byl in Trans. Roy. Soc. S. Afr.

10 (1922) 153, f. 4; listed as doubtful by Doidge loc. cit. p. 488.

The collection cited by van der Byl as the basis of this record is apparently not

to be found in any herbarium. This species is omitted from van der Byl’s later summary

of the South African Thelephoraceae (in Ann. Univ. Stellenbosch 7, 1929) as though

he were doubtful of the determination.

(13) Stereum concolor (Jungh.) Sacc., Syll. Fung. 6 (1888) 561; Doidge loc. cit.

p. 491.

Lloyd recorded some of van der Byl’s collections under this name; for references

see Doidge loc. cit. Doidge lists this species doubtfully as a synonym of Stereum

lobatum (Kunze ex Fr.) Fr. Van der Byl omits reference to S. concolor in his writings,

and the specimen in his herbarium No. 807 which was determined by Lloyd as

S. concolor has been annotated by van der Byl as a “ young condition of 5. lobatum ”.

The writer has seen this specimen and taken alone it could well be referred to Stereum

concolor, for this species is distinctive enough in typical specimens to warrant a name

of its own. However, there is little doubt that 5. concolor is in fact only a young stage

of S. lobatum. This is confirmed by the finding of collections, part of which could

be confidently referred to S. concolor and part to S. lobatum, e.g. No. 11291 in the

National Herbarium. The principal differences between these growth forms are in

their colour and hairyness.

Forms classed as S. concolor have a uniform, even, velvety brown tomentum on

the surface, ranging in colour from light yellow-brown to a somewhat deeper brown

spaced wide apart in concentric zones. The impression is a continuous tomentum of

a rather light brown colour. In forms classed as Stereum lobatum the tomentum is

frequently interrupted by bare or glabrescent zones and the colour zonation includes

brownish, greyish or greeny-grey, and narrow hazel or chestnut stripes. There are no

apparent miscoscopic differences between these forms.

In view of the above, S. concolor is treated here as a synonym of Stereum lobatum

(Kunze ex Fr.) Fr.

311

(14) Stereum cyphelloides Berk. & Curt, in Journ. Linn. Soc. Bot. 10 (1868) 331;

Saccardo, Syll. Fung. 6 (1888) 558; Burt in Ann. Mo. Bot. Gard. 7 (1920) 112;

Martin Lloydia 7 (1944) 76.

Fig. 5.

Fructifications soft, terrestrial, flabellate, attached by a reduced base which is

vaguely continued into a faint mycelial pad, laterally substipitate ; radius 4-8 mm.

from the attachment, width 5-9 mm. Stem flattened, 1 • 5-3 mm. wide. Surface

radially striate with appressed fibrils which anastomose towards the base, lacking a

cuticle, the context hyphae merely running out into the surface fibrils; colour fight

creamy-yellow all over. Hymenium concolorous, smooth. Margin not noteworthy.

Thickness in section 580-750

Basidia: compact, hyaline, cylindrical, 3-2 x 29 p.

Spores: smooth, hyaline, typically pip-shaped with an attenuated apiculus,

2-4-3-2 X 5-6-7-2 /z.

Hyphae: all of one kind, hyaline, thin- walled, branched, septate, without clamps,

2 -4-3 -2 /z wide.

Tissue differentiation: No bordering zone beneath the abhymenial surface.

Texture soft and absorbent, the tissues fairly compact but easily teased apart into

individual hyphae.

Specimens examined: 31419; 11528 (sub Thelephora).

The species is distinctive for its small whitish pilei, the soft texture and the peculiar

pip-shaped spores. In No. 31419 all the spores seen were pip-shaped, but in No. 11528

many of them exhibited the “ curiously angled and distorted ” character described

and figured by Martin loc. cit. The writer is indebted to Dr. R. W. G. Dennis for

comparing No. 31419 with type material in Kew Herbarium.

The literature references given by Doidge loc. cit. p. 488, for this species, are

incorrect, but the writer has not been able to trace the correct reference.

(15) Stereum diaphanum (Schw.) Cooke ex Saccardo in Syll. Fung. 6 (1888) 558;

Doidge loc. cit. p. 488.

Thelephora diaphana Schweinitz apud Berk. & Curt, in Acad. Nat. Sci. Philad.

Journ. 2 (1853) 278 (fide Burt in Ann. Mo. Bot. Gard. 7, 1920, 98).

Fig. 4.

Fructifications coriaceous to fibrous and brittle when dry, terrestrial, growing on

humus and buried wood, centrally stipitate, deeply infundibuliform, usually single,

occasionally two pilei fusing above. Pileus (0-4)-2-5 cm. diam., (0 - 3)— 1 -8 cm. radius

from attachment; surface creamy to pale yellowish, silky-fibrillose, radially linear-

striate, sometimes showing one or two obscure concentric colour zones of a slightly

darker yellow-brown. Hymenium smooth, creamy, cracking radially when dry and

revealing a silky-fibrillose trama; margin sometimes entire, more often undulating

and shortly laciniate or splitting down the radius into lobes, very thin, sometimes

involute. Stipe 1-2 mm. wide x 7-10 mm. long, light creamy colour, clothed with

short, fine hairs especially near the base, where there is usually a small mycelial pad.

Thickness in section 480-800 /z.

Basidia: cylindric-clavate, about 35 x 3 -2-4 -8 p.

Spores: 3-2-4 X 6-4-8 p, elliptic-oblong with one side depressed and the end

attenuated into a lateral apiculus, hyaline, smooth, abundant.

Gloeocystidia : 8—9 • 6— ( 1 6) x 58-86 p, subcylindric or clavate, hyaline, smooth,

thinwalled, with homogeneous contents, immersed in the hymenial layer or often

projecting up to 32-48 p beyond it.

Hyphae: hyaline, branched, septate, with thin, firm walls, 3 -2-6 -4 /x wide,

acking clamps but with occasional H-anastomoses, all of one type.

9102-3

312

Tissue differentiation: there are no abhymenial hairs and no compact tissue

differentiated just below the abhymenial surface.

Specimens examined: 21099, 21208, 23160, 14516, 20403, 31858: Universiteit

van Stellenbosch, Herbarium P. A. van der Byl No. 662 (as S. pusillum Berk.).

The pilei of this species have a well-marked tendency to split radially down the

striae. They are brittle, and creamy in colour, when dry. Our description differs

from that of Burt (loc. cit.) in spore size and thickness of the pilei. Burt has referred

to the “ hairlike cystidia ”, which are here called gloeocystidia. Although these may

be emergent, their smooth walls and homogeneous, readily stained contents suggest

gloeocystidia. These organs are similar to those found in Corticium praetermissum

( = Peniophora tenuis ) and it is somewhat a matter of personal taste what they shall

be called.

This species is known so far from the Fountains-Groenkloof valleys near Pretoria,

where it is fairly common, and from Durban, where one collection has been made.

The writer is indebted to Dr. R. W. G. Dennis for comparing Nos. 21099, 21208,

with Schweinitz material in Kew Herbarium. Spores were lacking in the last-

mentioned, but otherwise there was very close agreement.

(16) Stereum durbanense van der Byl (!) in Trans. Roy. Soc. S. Afr. 10 (1922)

155, Fig. 8, in Ann. Univ. Stellenbosch 7 (1929) 45; Doidge loc. cit. p. 489.

Stereum tomentosum van der Byl (!) in Trans. Roy. Soc. S. Afr. 10 (1922) 156,

Fig. 9, in Ann. Univ. Stellenbosch 7 (1929) 45; Doidge loc. cit. p. 493.

Fig. 12.

Fructifications corky, drying tough, brittle and woody, lignicolous, somewhat

orbicular, attached by the centre and partly resupinate over wide areas then widely

reflexed, pilei sometimes connate and occasionally imbricate over the main pileus,

overall size up to 7 x 12 cm. Surface with a thick pad-like tomentum of closely

matted hairs, ochraceous to cinnamon, becoming greyish, if rubbed showing bright

cinnamon-ochre colour, concentrically furrowed with more or less glabrous areas in

the furrows and hence somewhat colour-zoned, colour lighter near the margin.

Hymenium light-brown to light-fawn colour, smooth or slightly rimose or convoluted

in places into small tubercles or larger humps, reflecting the abhymenial furrows near

the margin, cracking a little on drying. Margin acute, undulate or lobed, usually a

little darker than the rest of the hymenium. Thickness in section excluding tomentum

950-1500 p, the tomentum itself 1000-2000 p thick.

Basidia: hyaline, cylindric-clavate, forming a dense palisade, 3-2 x 17-26 p.

Spores: hyaline, smooth, oblong-cylindric with a small lateral apiculus, 3-2 x

4 -8-6 -4 p (3^1 p diam. fide van der Byl).

Hyphae: of two kinds. Skeletal hyphae unbranched, 5-6-8 p wide, dilutely to

darkly coloured, thick-walled, with narrow or wider lumen, showing septa in the wider

parts of the lumen and with contents darker brown than the walls. Generative hyphae

3 -4-4 p wide, branched, clearly septate, subhyaline to hyaline, thin-walled with wide

lumen.

Tissue differentiation: hyphae compact and very closely interwoven; a dark zone

subtends the abhymenial surface.

Abhymenial hairs: very thick and closely matted, dark red-brown, thick-walled,

4-8-8 p wide.

Specimens examined: 15613 (three specimens ex Natal Herb. 471, originally

determined by Lloyd as S. subpileatum Berk, and changed by van der Byl to S. durbanense

van der Byl); 31852 (N.H. 606); 31898 (N.H. 692); 35559 (W. G. Rump, 756); 32007

(N.H. 904); 32478 (N.H. 341) as Stereum tomentosum van der Byl; 15601 (N.H. 341)

313

as Stereum sp.; Universiteit van Stellenbosch, Herbarium P. A. van der Byl No. 293

(Type of S. durbanense ), 294, 517; Herbarium P. A. van der Byl No. 305, 132 (as

•S. tomentosum).

Careful microscopic examination of material of Stereum tomentosum has failed

to distinguish it from Stereum durbanense, while it is believed that the minor macroscopic

differences (in colour of the surface and the hymenium) are of no specific significance.

They may have been more apparent in the fresh material but are now hardly distinguish-

able in the dry specimens. Since S. durbanense has page priority in the original place

of publication of these two species, S. tomentosum is relegated to synonymy. Inciden-

tally the latter is a more suitable epithet, especially as the species is not limited to the

vicinity of Durban.

Authentic material of S. durbanense was found in the National Herbarium under

No. 15613 (Natal Herb. 471) split up into three separate packets. This material had

been determined by Lloyd as Stereum subpileatum Berk. (Lloyd Myc. Notes 5, 1917,

L. 66, 15, Note 634; listed by Doidge loc. cit. p. 489). In assigning some of van der

Byl’s collections to this species, Lloyd noted that they differed from S. subpileatum

in the absence of cystidia. Thus van der Byl (1922, loc. cit.) described the new species

S. durbanense to accommodate them, and later (van der Byl, 1929, loc. cit.) noted that

S. durbanense possessed a few hairlike cystidia, but that these were not encrusted as

in S. subpileatum. The writer has seen material of S. subpileatum in Kew Herbarium,

and there were bottle-brush paraphyses as well as cystidia. It is quite certain that

S. durbanense is different, and that S. subpileatum must be excluded from South African

lists.

The writer was unable to find in any of the material of S. durbanense the cystidia

which van der Byl described as “ 20-60 x 6-4 p, emergent up to 8 p, hairlike, colourless,

not encrusted, very few found, only in parts of the hymenium ”. There were, however,

scanty cystidioid hyphae which intruded into the hymenium but not beyond it. These

were smooth, almost colourless, thick-walled, cylindrical, about 4-8 X 48 p, and

were probably a form of skeletal hypha.

The absence of conductors differentiates S. durbanense from S. rimosum, S.

rimosum var. africanum, and S. rugosum, with which it might be confused. Small

specimens, though uncommon, might be confused with thick forms of Stereum australe,

but for the absence of conductors. The species is quite distinctive and the microscopic

check for conductors is seldom necessary. A further characteristic is that S. durbanense

has a rather narrow hyaline hymenium above a brownish context, when viewed with

the naked eye.

(17) Stereum duriusculum Berk. & Br. (!) in Journ. Linn. Soc. Bot. 14 (1873) 66;

Doidge loc. cit. p. 489.

Fig. 22.

In Bothalia 6 (1951) 51-53, the writer described and discussed this species, and

related ones, and proposed the new combination Asterostromella duriuscula (B. & Br.)

Talbot. It is clear that Stereum is an unsatisfactory genus for the reception of this

species, while many people will no doubt think the same of Asterostromella. What

is needed is a study of the genera Dichostereum Pilat and Vararia Karsten, where

possibly the true affinities of Stereum duriusculum will lie. For the present this species

is left as Asterostromella duriuscula, and the reader is referred to the above paper for

a description. For convenience the species is keyed out with other Stereum species

in this paper.

(18) Stereum elegans Mey.; recorded by Kalchbrenner in Grev. 10 (1881) 58;

Saccardo Syll. Fung. 6 (1888) 553; Doidge loc. cit. p. 493.

Kalchbrenner’s record of this species referred to MacOwan’s collection No. 1232.

Doidge loc. cit., notes this material under S. thozetii Berk., which was the determination

314

given it by van der Byl on the sheet in Herb. S.A. Museum No. 34266. Having studied

MacOwan’s material, and had cospecific material checked at Kew, the writer confirms

its determination as Stereum thozetii Berk.

Another collection, J. M. Wood No. 396 (National Herbarium Pretoria No. 10653)

is filed in Pretoria and at Kew under Stereum nitidulum. This collection was cited by

Bottomley (in S.A. Journ. Sci. 13, 1916, 440) as “ Stereum e/egans — Stereum nitidulum

B.”, and the sheet at Kew is annotated “ = Stereum e/egans ”, by Bresadola. This

material certainly does not correspond with Petch’s description of S. elegans (in Ann.

Roy. Bot. Gard. Perad. 9, 1924, 260). The writer enlisted the aid of Dr. R. W. G.

Dennis in comparing this specimen with material of S. elegans at Kew and was informed

that it was a good match with S. nitidulum but not with S. elegans. It is dealt with

in this paper under S. nitidulum.

It would appear that Stereum elegans must be excluded at present from South

African fists.

(19) Stereum fasciatum (Schw.) Fr., Epicr. Syst. Myc. (1838) 546; Doidge loc.

cit. p. 489.

The writer has seen no South African material which he can confidently refer to

S. fasciatum. Many specimens are so named in the herbaria, particularly in Herbarium

P. A. van der Byl. At first, van der Byl (in Trans. Roy. Soc. S. Afr. 10, 1922, 155,

Fig. 7) classified his specimens as Stereum lobatum. Later (in Ann. Univ. Stellenbosch

7, 1929) he referred them all to S. fasciatum , commenting that up till then he had

seen no typical specimens of S. lobatum in South Africa, although S. fasciatum was

common in parts.

The writer has examined van der Byl’s specimens in the Universiteit van Stellenbosch,

Herbarium P. A. van der Byl. Most of them are undoubtedly Stereum lobatum', a

few are rather small forms which may possibly be called S. fasciatum, but there are

many intermediate forms and nowhere to draw the fine between the two extremes.

Furthermore these small forms are not altogether like North American specimens

of S. fasciatum. Still other specimens labelled S. fasciatum have proved on examination

to be Stereum australe.

In Bothafia 6 (1951) 45 and 50 the writer cited a collection of J. M. Wood No. 163

(under Stereum luteobadium ) as Stereum fasciatum. This was an error, and re-exami-

nation of that collection has failed to distinguish it from S. lobatum.

It is felt that Stereum fasciatum must be fisted as a doubtful species for South

Africa. Notes differentiating 5 fasciatum and S. lobatum are given under the latter

species (p. 319).

(20) Stereum friesii Leveille. Zoll. Verz. p. 17; Saccardo, Syll. Fung. 6 (1888)

266; Doidge loc. cit. p. 489.

Fig. 9.

Pileus sessile, fignicolous, slightly effused, not flexible, semi-dimidiate or broadly

spathulate, 1-3 cm. x 1-1-5 cm., sometimes laterally connate. Surface velutinate or

with a somewhat thicker tcmentum, concentrically zoned, brown and fight yellow-

brown. Margin fight brown, broad in young specimens, narrower in old. Hymenium

creamy to cinnamon-cream, shading off near the attachment to a bay colour with a

smoky violet tint. Margin wide and creamy in colour on the hymenial side. Sections

700-1000 /a thick.

Basidia: small, cylindrical, 12-16 x 2 -5-3 -2 /a.

Spores: quite abundant, hyaline, smooth, ovate to subglobose, (1 -7)-2-4 x 3-2 /a

or about 2-5-3 /a diam.

Cystidia: in the hymenium, projecting 11-18 /a, or embedded, hyaline, fusoid

with fairly thin walls and wide lumen; walls minutely rugose and encrusted, easily

losing the encrustation and then smooth, 27-38 X 10-12 ju.

315

Gloeocystidia : smooth, thin-walled, deep-staining, embedded, fusoid, 7*5-8 x

12-30 fi.

Hyphae: skeletal hyphae 3 -3-4-4 -8 p wide, hyaline, smooth, unbranched, without

septa, thick-walled with a narrow lumen, the lumen occasionally wider; generative

hyphae thin-walled, deep-staining, much branched, with occasional clamps, somewhat

tortuous, 2 -4-3 -2 p diam.

Tissue differentiation: all tissues are rather compact and densely interwoven.

There is no denser, coloured zone differentiated below the surface.

Abhymenial hairs: rather scanty, hyaline, thick-walled at the base with a wide

lumen, the walls narrowing and the lumen widening towards the apex; 6-4 p diameter.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der

Byl No. 378.

This specimen is undoubtedly one of the same species as van der Byl No. 645

as Stereum bellum. It is not yet known whether S. bellum or 5. friesii is applicable

to the species, if either name is.

(21) Stereum fulvum (Lev.) Sacc., in Saccardo Syll. Fung. 6 (1888) 570; Doidge

loc. cit. p. 489.

Thelephora ( Stereum ) fulva Leveille (!) in Ann. Sci. Nat. ser. iii, 5 (1846) 149.

Stereum schomburgkii Berkeley (!) in Journ. Linn. Soc. Bot. 13 (1873) 168; Sac-

cardo, Syll. Fung. 6 (1888) 568; Doidge loc. cit. p. 493; Talbot in Bothalia 6 (1951) 43.

Stereum atrocinereum (Massee) van der Byl in Ann. Univ. Stellenbosch 7 (1929) 44.

Peniophora atrocinerea Massee (!) in Journ. Linn. Soc. Bot. 25 (1889) 141.

Stereum retirugum Cooke (!) in Proc. Roy. Soc. Edinb. (1882) 456; Saccardo

Syll. Fung. 23 (1925) 510; Doidge loc. cit. p. 492.

Hymenochaete olivaceum Cooke (!) in Grevillea 14 (1885) 11.

Fig. 20.

Resupinate or resupinate-reflexed, or conchiform and attached by a small umbo,

orbicular-confluent, thin, coriaceous cr papyraceous, loosely attached to the substratum.

Reflexed surface tomentose, glabrescent when old, buffy-brown, concentrically furrowed

in pileate specimens, the troughs being paler. Hymenium smooth, or more usually

furrowed concentrically and cracking radially; colour very variable, through umber,

brown, olive brown, light yellow-brown, brownish slate, depending on the state of

development. Margin narrow, yellowish, finely fibrillose, usually free. Context

concolorous, 200-(500) p thick in section.

Basidia: clavate, 4-5-6 x 20-25 p, in young stages in a regular palisade, later

interrupted and exceeded by the setoid hyphae.

Spores: 3-2 X 6 -4-6 -8 p, hyaline, broad elliptical or cylindrical-depressed,

smooth.

Setoid hyphae: 3-5-8 p wide, cylindrical, often wdth a fusoid apex, brown,

minutely rugulose near the apex, thick-walled with a very narrow lumen which sometimes

expands at the apex, immersed or emergent up to 16 p, originating as skeletal hyphae

which curve upwards into the hymenium, total length variable and indeterminable,

very densely arranged, darkening in alkali.

Hyphae: skeletal hyphae brown, thick-walled, smooth, without clamps, of the

same dimensions as the setoid hyphae; generative hyphae hyaline to light yellow-

brown intermeshing with the skeletal, 3 -5-4 -5 p diam., often indistinct, thin to thicker-

walled, with occasional clamp connections.

Tissue differentiation: The hyphae form a more or less horizontal weft without

a darker or more compact basal layer.

316

Abhymenial hairs: pallid to light yellow-brown, 5 p diam., fairly thick- walled,

septate, with occasional clamp connections.

Specimens examined: Type of Thelephora ( Stereum ) fulva Lev., Drege 9441, Cap-

de-Bonne-Esperance, in Herb. Mus. Paris; Type of Stereum schomburgkii, Schomburg,

Australia, in Herb. Kew.; Type of Peniophora atrocinerea Massee (sub “ Corticium

atrocinereum Kalchbr.”). P. MacOwan, Cape Province, in Herb. Kew.; as Stereum

atrocinereum (Massee) van der Byl, MacOwan 1197 (Herb. S. Afr. Mus. No. 34284);

as Stereum mebranaceum, A. Pegler (1234) in National Herbarium Nos. 8414 and

34454; 31863, 8756, 35237, 28942, 20944, 27644, 27544, 27552, 28496, 28499, 28304,

28504, 28687, 34951, 36800, 36801, 11968, 11770, 28565, 27758,27607,28493, 30881,

33074, 33565, 34454, 8414, 2301, 33990, 33179; Hoeg F. 67 in Herb. Kew.

This species is well known under the name Stereum schomburgkii. Recent

examination of the type of S.fulvum Lev., borrowed from Herb. Mus. Paris, has shown

that the two are synonymous, and consequently the earlier epithet must now be taken

into use.

In Bothalia 6 (1951) 44, the writer gave reasons for reducing S. atrocinereum

(Massee) van der Byl to synonymy, and also reasons for the variation in colour of the

specimens. The material which van der Byl described as 5. atrocinereum has now

been seen, in Herb. S. Afr. Mus. 34284. It corresponds with the material of Peniophora

atrocinerea in Herb. Kew, and is almost certainly part of the same collection.

Doidge’s record of Stereum retirugum refers to Mocambique only, but it is

confirmed that Cooke’s type of the species in Herb. Kew is synonymous with S.fulvum.

The only record of 5. membranaceum Fr. for South Africa (Pole Evans and

Bottomley in Ann. Bolus Herb. 2, 1918, 192) is based on a pale form of S. fulvum

and not on the species suggested. Lloyd (in Lloyd Myc. Notes 6, 1920, 960) suggests

that S. schomburgkii differs mainly from S. membranaceum in having an umber

hymenium instead of one which is violaceous but fades when old. Actually S. mem-

branaceum is a synonym of Stereum papyrinum, a species not unlike Stereum umbrinum.

The type number of Stereum fulvum, Drege No. 9441, is cited in mistake for Drege

No. 9442 under Stereum murrayi, by Doidge loc. cit. p. 491.

(22) Stereum fuscum (Schrad.) Quelet, Flor. Myc. de Fr. (1888) 14; Doidge loc.

cit. p. 488.

Doidge lists the South African records of this species and correctly indicates that

it is a synonym of Stereum bicolor (Pers. ex Fr.) Fr.

MacOwan’s specimen No. 1244 (as Thelephora biennis. Herb. S. Afr. Mus. 34292),

determined by van der Byl as S. fuscum, is S. bicolor. A note on the nomenclature

of this species was given in Bothalia 6 (1951) 40.

(23) Stereum glabrescens Berk. & Curt.; Recorded by van der Byl in Trans.

Roy. Soc. S. Afr. 10 (1922) 151, Fig. 1, and in Ann. Univ. Stellenbosch 7 (1929) 37;

Doidge loc. cit. p. 487.

A wrong record based on specimens of Stereum affine (see p. 305). Stereum

glabrescens is not known to occur in South Africa.

(24) Stereum hirsutum (Willd.) Pers. ex S. F. Gray, A Natural Arrangement of

British Plants 1 (1821) 652; Persoon in Roemer Neues Mag. Bot. 1 (1794) 110;

Saccardo Syll. Fung. 6 (1888) 563; Doidge loc. cit. p. 489.

Thelephora hirsuta Willdenow, Flor. Berol. Prod. (1787) 397; Fries, Syst. Myc.

1 (1821) 439; Persoon, Syn. Meth. Fung. (1801) 570, Myc. Eur. 1 (1822) 116.

Stereum amoenum Kalchbr. & MacOwan (!) in Grev. 10 (1881) 58; Doidge loc.

cit. p. 490 (nec S. amoenum Lev.).

317

Stereum kalchbrenneri Saccardo, Syll. Fung. 6 (1888) 568; Doidge loc. cit. p. 490.

Fig. 11.

Fructifications coriaceous, lignicolous, gregarious, usually laterally connate’

imbricate, effuso-reflexed or dimidiate, occasionally semi-resupinate and attached by

an umbo; radius 1-4 cm. from attachment, 0-7-5 cm. wide or several centimetres

wide by lateral confluence. Surface covered with strigose-fasciculate or matted

hairs, showing concentric zonation, concentrically furrowed, coloured greyish, light

yellow-brown, or a deeper reddish-brown. Hymenium smooth, creamy to buff",

cinnamon or light orange colour. Margin thin, entire to undulate or lobed, not notably

differentiated. Thickness in section 600-800 p.

Basidia: cylindric-clavate, 25-35 x 3 -5-4 -8 /x.

Spores: hyaline, cylindrical, sometimes slightly bent, smooth, 2-4 x 6-4-8 p.

Hyphae: generative hyphae hyaline, smooth, thin-walled with wide lumen, septate,

branching, lacking clamps, 2-4-3 -2-(4) p wide; skeletal hyphae hyaline, smooth,

thick-walled with narrow lumen, occasionally septate, unbranched, without clamps,

6-4-8 i a wide.

Cystidioid hyphae: present in many but not all specimens, arising as skeletal

hyphae and curving into the hymenium but not beyond it, hyaline, thick-walled with

a narrow lumen except at the apex where the lumen expands, contents of lumen hyaline

to yellow-brown, up to 9 - 6— ( 1 6) p wide at apex.

Tissue differentiation: A compact golden coloured dense zone subtends the

abhymenial surface; the trama is composed of closely intertwined hyphae arranged

more or less horizontally.

Abhymenial hairs: 6 -4-9 -6 p diam., hyaline or rarely dilutely coloured, smooth,

thick-walled with narrow lumen, occasionally septate, intertwined and matted or

arranged somewhat parallelly.

Specimens examined: 34953, 23374, 27517, 24822, 11292, 11255, 34956, 31525,

24873, 31298, 34955, 11290, 34479. 34957, 34954, 12173, 31709, 17299, 9060, 31234,

23670, 11290, 30516, 27722, 30720, 30261, 30803, 9148, 27539, 24822, 1952, 8776,

8773, 2344, 27518, 34072, 30735, 27538, 33067, 26697, 17803, 18151, 20586, 23671,

17101, 27518, 11255, 30948, 31454, 13793, 31737, 28967, 28880, 27647, 27340, 25879,

15484, 2344 (in part), 23474, 22085, 18146, 30893, 36707, 34200, 31031, 1017, 11292,

23374, 29719, 13073, 33067, 28859, 28835, 28756, 17788, 15496, 28951, 30719, 31298,

31807, 31892, 32504; Sub. S', spadiceum, 13793, 12993, 5651, 1707.

Reasons for sinking S. amoenum and S. kalchbrenneri in synonymy are given in

the notes on those species.

Compared with European specimens of S. hirsutum, the majority of South African

material is more luxuriant, not so generally greyish in surface colour, and very frequently

shows a strong development of cystidioid hyphae. These are features which contributed

to the erection of the species S. kalchbrenneri, but the species is so variable and merges

so closely into more typical S', hirsutum that it is felt that a separate name is not

warranted. See also notes on S', vellereum, p. 329.

(25) “ Stereum hirsutum forma kalchbrenneri ”. In S. Afr. Journ. Sci. 42 (1946)

133, Simpson & Talbot listed a collection under the name “ Stereum hirsutum (Willd.)

Fr. kalchbrenneri forma ”, under the impression that S. kalchbrenneri had already

been proposed as a form of S. hirsutum. It is possible that this citation constituted

the proposal of a nomen nudum for a new form. The material referred to may be

included in the species S. kalchbrenneri which we here regard as a synonym of Stereum

hirsutum (Willd.) Pers. ex S. F. Gray.

(26) Stereum involutum (Klotzsch) Fries, Epicrisis (1836-38) 546; Saccardo,

Syll. Fung. 6 (1888) 560; Lloyd, Syn. Stip. Stereum, in Lloyd Myc. Notes 4 (1913)

40; Doidge loc. cit. p. 490.

318

Thelephora involuta Klotzsch (!) in Linnaea 7, p. 499.

Lloydella involuta (Kl.) Bresadola in Ann. Myc. 18 (1920) 60.

Stereum bresadoleanum Lloyd (!), Syn. Stip. Stereum in Lloyd Myc. Notes 4 (1913)

41; Stevenson & Cash in Bull. Lloyd Library 35 (1936) 51; Doidge loc. cit. p. 488.

Stereum proximum Lloyd, Syn. Stip. Stereum in Lloyd Myc. Notes 4 (1913) 40;

Doidge in loc. cit. p. 488.

Fig. 7.

Fructifications coriaceous, drying tough and with a horny hymenium, lignicolous,

gregarious, attached by a reduced base, usually several pilei attached by reduced bases

and laterally connate above, semiflabellate or petaloid, radius 1-3 cm. from the

attachment, width 1 • 5-3 -5cm. Surface very finely velutinate, marked with very narrow

concentric zones coloured tawny, yellowy-orange, greyish or light brown, not distinctly

furrowed, general effect tawny when young becoming darker brown when old.

Hymenium waxy, reddish-bay colour, smooth, paler towards the margin, drying darker

and distinctly horny. Margin paler on both surfaces, involute or slightly crimped,

very thin. Thickness in section up to 1000 p.

Basidia: closely packed, rather indistinct, about 3 x 16-20 p.

Spores: uncertain, thought to be about 1-6-2 x 2-4-3 p, elliptical, hyaline.

Cystidia: fusoid, thick-walled, hyaline, staged in the hymenium, embedded,

scarcely ever emergent, encrusted at the apex or smooth, encrustation soluble in KOH,

always abundant, 10-15 x 27-66 p.

Gloeocystidia : usually subulate with a swollen base, sometimes more cylindrical

or fusoid, thin-walled, with homogeneous contents, hyaline, embedded in the hymenium

and subhymenium, abundant in thicker parts of the specimens seen, 7-3-10 x 40-66 p.

Hyphae: generative hyphae thin-walled, hyaline, with occasional septa, branches

and clamp connections, 3-2-4 p diam.; skeletal hyphae thick-walled, hyaline, without

septa, rarely branched, without clamps, 3 -2-4 -8 p diam. The two hyphal types are

densely intermingled.

Tissue differentiation: There is no well-marked colour zone subtending the

abhymenial surface. Unmounted sections show the dark, waxy hymenium and some-

times a similar dark abhymenial zone.

Abhymenial hairs: hyaline to very dilutely coloured, free or fasciculate, thick-

walled, 8-10-12 p diam.

Specimens examined: 14909 b, 31956, 31750, 15556; Type of Thelephora involuta

Kl., Mauritius, in Herb. Kew.; Universiteit van Stellenbosch, Herbarium P. A. van der

Byl Nos. 192, 193 (as S. proximum).

Specimens comprising this taxonomic species were found distributed in the

National Herbarium under the names S. proximum, S. bresadoleanum and S. involutum,

among which was authentic material of S. bresadoleanum.

Van der Byl (in Ann. Univ. Stellenbosch 7, 1929, 38) described S. proximum

and distinguished it from S. involutum by the more finely velutinate. surface of the former.

In the absence of authentic material of S. proximum, the writer must follow published

synonymy. By Lloyd’s own admission (in Lloyd Myc. Notes 7, 1922, 1115, Fig. 2095),

S. proximum Lloyd is a synonym of S. bresadoleanum Lloyd, which he previously

suggested (in Syn. Stip. Stereum in Lloyd Myc. Notes 4, 1913, 40) was a form

S. involutum. Bresadola (in Ann. Myc. 18, 1920, 60) united S. bresadoleanum and several

other species under the name Lloydella involuta, whose specific epithet is accepted here.

Having compared type material of S. involutum, the writer agrees that S. bresadoleanum

is synonymous.

The species is characterised by its finely velutinate surface and the waxy, reddish

bay hymenium, also by possession of cystidia, gloeocystidia and a dual hyphal system.

Gloeocystidia do not appear to be mentioned in available descriptions. They are

usually abundant, but might easily be missed without the use of a stain like phloxine.

319

(27) Stereum kalchbrenneri Saccardo (!), Syll. Fung. 6 (1888) 568; Doidge loc.

cit. p. 490.

= Stereum hirsutum (Willd.) Pers. ex S. F. Gray. See further notes under

Stereum amoenum (p. 305) and Stereum hirsutum (p. 316).

(28) Stereum laxum Lloyd (!) in Lloyd Myc. Notes 4 (1915) L. 60, 10, Note 339;

Doidge in loc. cit. p. 491.

= Stereum bicolor (Pers. ex Fr.) Fr. See p. 308.

(29) Stereum lobatum (Kunze ex Fr.) Fries, Epicr. (1838) 547; Saccardo, Syll.

Fung. 6 (1888) 568; Doidge loc. cit. p. 491.

Thelephora lobata Kunze in Weigelt Exsiccati, 1827; Fries in Linnaea 5 (1830)

527.

Fig. 10.

Fructifications coriaceous, sessile, typically wedge- or fan-shaped, tapering to a

reduced base which is attached by a small umbo (one centrally attached, infund ibuliform

specimen was seen), often produced singly, sometimes laterally connate, the connate

pileoli produced from one or more umbo. There is very infrequently any resupinate

part. Size varying from 4-10 cm. radius x 2-10-20 cm. in width. Surface concen-

trically furrowed and colour-zoned, with a thin velvety tomentum of closely matted

hairs inclined to rub off easily in old specimens leaving at least some bare zones.

Colour of surface uniform light yellow-brown to somewhat deeper brown, reddy brown

grey or greeny-gray, especially becoming greyish with hazel or chestnut coloured rubbed

zones. Hymenium smooth or reflecting slightly the surface furrows, coloured creamy

to fight buff, not dark coloured or cinereous. Margin acute, entire or undulate, or

vaguely lobate in connate specimens. Thickness in section 700-900 fx.

Basidia: 3-5-5 x 24-33 /x, cylindric-clavate, compact.

Spores : 2 • 7-3 -3x5- 3— (8) hyaline, cylindrical with one side a little depressed,

or oblong or broad-elliptical, smooth.

Hyphae: hyaline or some very dilutely coloured, unbranched, without clamps,

thin-walled, 2-5-4 /x diam., septate, sometimes with expanded parts up to 8 /x diam.

Other hyphae are thin- to thick-walled with a wide lumen, up to 8 p diam. Some

of the thick-walled hyphae about 6-4 /x diam., curve up and intrude into the hymenium.

Tissue differentiation: a yellow-brown zone is present beneath the layer of

abhymenial hairs.

Abhymenial hairs: 3 -2-6 -4 /j. diam., yellowish, thick-walled, septate, somewhat

f3.SClCUl3.tC

Specimens examined: 13794, 33379, 31559, 27332, 30741, 28971, 17098, 34950,

34126, 36615, 34210, 34196, 28503, 36613, 11544, 23347, 17811, 9150, 23232, 31560,

27537, 31667, 31296, 30837, 11291, 31559, 11523, 12049, 34516, 14907, 10654 (J. M.

Wood, 163), 20972 (MacOwan, 1269); Herb. S.A. Museum 34265 (MacOwan 1163,

1276, as S. versicolor then det. van der Byl as S. fasciatum).

There is difficulty in separating S. lobatum, S. fasciatum (Schw.) Fr. and S. concolor

(Jungh.) Sacc.

The above taxonomic species is composed of specimens which are large and

flabellate or slightly lobed, and not effuso-reflexed. They are all rather thin and with

a thin, velvety tomentum. Single, typical specimens of S. concolor may be held apart

from S. lobatum on account of their uniform fight brown tomentum, but there is little

doubt that such forms merely represent a young stage in the growth of S. lobatum

(see p. 310). The latter is typically greyish or greeny-gray with hazel or chestnut

coloured rubbed zones, but specimens with a brownish surface have also been included

here under S. lobatum.

320

Stereum fasciatum is without doubt very close to S. lobatum, but differs principally

in being thicker, with a thicker, shaggy tomentum, and in frequently being effuso-

reflexed when young, in which state it is strongly reminiscent of 5. hirsutum. Forms

referred to S. lobatum are larger, thinner and more flexible and have a finer, softer,

adpressed tomentum which rubs off with age leaving the pileus with smooth shining

chestnut-brown zones.

There is apparently no distinguishing microscopic character, but it may be

convenient to retain the two names for the extreme forms which look very different.

Intermediates are to be found, but on the whole S. fasciatum is smaller and most

frequent in the temperate regions, whereas S. lobatum is typically a tropical fungus

and sometimes becomes very large.

Some of the differences noted above are based on notes kindly supplied by Miss

E. M. Wakefield. Compare also the discussion under S. fasciatum (p. 314).

(30) Stereum lobatum (Kunze) Fr. var cinertum Lloyd ex Doidge in Bothalia

5 (1950) 487, nomen nudum.

Lloyd did not publish the varietal epithet which is attributed to him by Doidge.

As Doidge (loc. cit.) noted, this variety is a synonym of Stereum australe Lloyd.

(31) Stereum luteobadium Fries: recorded by Kalchbrenner in Grev. 10 (1881) 58;

Bottomley in S.A. Journ. Sci. 13 (1916) 440.

The collection (Wood 163, National Herbarium 10654, and in Kew Herbarium

under Hymenochaete luteobadia) cited in the above papers was referred to Stereum

fasciatum by van der Byl (in Ann. Univ. Stellenbosch 7, 1929, 50) and by Doidge (loc.

cit. p. 485), and also by Talbot (in Bothalia 6, 1951, 45 & 50). The writer would now

refer it to Stereum lobatum. In Herb. Kew there is also a specimen “ leg. P. MacOwan,

9/83, C.B.S., Herb. Kalchbrenner ” as S. luteobadium in the Hymenochaete luteobadia

folder. This too is S. lobatum. Hymenochaete luteobadia (Fries) Hohnel & Litsch. does

occur in South Africa.

(32) Stereum membranaceum Fr. Recorded by Pole Evans & Bottomley in Ann.

Bolus Herb. 2 (1918) 192.

The collection referred to (A. Pegler, 1234, in National Herbarium Nos. 8414

6 34454) is a pale form of Stereum fulvum Lev. (!) (see p. 315) with setoid hyphae

rather more encrusted than usual. This correction was noted by Doidge loc. cit. p. 493

under S. schomburgkii, a synonym of S. fulvum.

(33) Stereum murraii (Berk & Curt.) Burt in Ann. Mo. Bot. Gard 7 (1920) 131

(as S. murrayi ); Rogers & Jackson in Farlowia 1 (1943) 290; Doidge loc. cit. p. 491.

Thelephora murraii Berk. & Curt, in Journ. Linn. Soc. Bot. 10 (1868) 329.

Fig. 16.

Fructifications lignicolous, resupinate, effused, becoming rarely slightly reflexed

at the margin, rather tough, woody to corky. Surface hard, crustose, uneven, black,

showing as a narrow black seam when resupinate. Hymenium creamy to tan or buff

colour, smooth or uneven or somewhat tubercular, becoming shallowly cracked.

Margin entire. Thickness in section up to about 5 mm. Context with a veined, mottled

or marbled appearance.

Basidia: not seen.

Spores: not seen. (Hyaline, smooth, flattened on one side, 4-5-5 X 2-5 /a, fide

Burt).

Vesicles: embedded in strata, very numerous, hyaline, with thin firm walls,

homogeneous contents, pyriform, 11-15 X 12-25 /a. Sometimes the vesicles are

elongated, subcylindrical, or ventricose or fusoid thus appearing like gloeocystidia

and then 8-14 X 40-60 /a.

321

Hyphae: hyaline, suberect, densely interwoven; some are branched, septate,

tortuous, submoniliform, deep-staining, up to 5 /x diam. Others are straight, much

branched, filamentous, not staining, 1-2 /x diam.

Tissue differentiation: The context is divided into strata at the junctions of which

the vesicles are most numerous. Much mineral matter is present.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der

Byl No. 708; 39709 (J. Weese, Eumycetes selecti exsiccati No. 548).

Van der Byl’s collection No.. 708 of this species is old and in poor condition.

Apart from its greater thickness and the presence of only a few vesicles in good

condition, this collection is a good match with S. murraii. It is probable that the

numerous air spaces which are present in van der Byl’s material represent vesicles

which have degenerated with age, and the context is not as hard as described for

S. saxitas Burt (Burt loc. cit. p. 134).

The above description was drawn mainly from European material in the National

Herbarium. In it, the elongated form of vesicle was common. As noted by Rogers

and Jackson (loc. cit.) this kind of vesicle is encountered in the form of S. murraii

which was known as Corticium effusum Overholts.

(34) Stereum nitidulum Berk.; Saccardo Syll. Fung. 6 (1888) 552; Doidge loc.

cit. p. 491.

Fig. 2.

Fructifications terrestrial among grass roots, coriaceous, centrally stipitate, infundi-

buliform, radius 0-7-1 -2 cm., 0-6-1 -2 cm. wide. Stipe 5-7 mm. long, 1 mm. diam.,

light tan colour, smooth, glabrous. Surface glabrous, smooth, concentrically zoned

with bay and light yellow-brown bands. Margin thin, indented to shortly laciniate.

Hymenium light yellow-brown showing one or two darker bands corresponding to

the darkest of the bands on the abhymenial surface, i.e. the pileus is semi-translucent.

Thickness in section 500-800 /x.

Basidia: cylindrical to clavate, 20-30 X 3-4-5 g., sometimes showing a basal

clamp connection.

Spores: hyaline, smooth, ovate, subglobose, or uncommonly broad-elliptical,

3-2-4 x 4-4-8 /x, or about 4-5 g. diam.

Gloeocystidia : abundant, thin-walled, with homogeneous contents, flexuous,

subcylindrical to subfusoid or ventricose, embedded in and just beneath the hymenium,

40-80 x 7-10-5 ju..

Hyphae: generative hyphae hyaline, thin-walled, branched, staining deeply,

septate, with occasional clamp connections, up to 3 • 2—4 - 8 /x diam., but mostly

collapsed, intermingled throughout with the skeletal hyphae which are 3 -2-4 -8 /x

wide, hyaline, thick-walled, with lumen narrow or invisible, not staining, unbranched,

not septate, without clamps.

Tissue differentiation: There are no abhymenial hairs and no compact zone

subtending the abhymenial surface.

Specimens examined: 10653 (J. M. Wood No. 396); Wood A396, Inanda, Natal,

in Herb. Kew.

This species is not unlike S. thozetii but has spores only about half the size of

those of the latter. The sub-translucent character of the pileus is also characteristic,

so that the pileus appears brownish on both surfaces. Dr. R. W. G. Dennis kindly

compared our material with the type of S. nitidulum and found that they were a good

match. (See also under S. elegans, p. 313).

The material of S. nitidulum corresponds very closely with Welwitsch 427, British

Museum, as Stereum ravenelii, the latter differing only in having gloeocystidia about

twice as large and having skeletal hyphae in which the lumen is usually rather wide.

The size of the gloeocystidia in these stipitate Stereums is known to vary greatly (cfr.

322

Martin in Lloydia 7, 1944, 75) but with only a single specimen of each species available

for study the limits of variation remain unknown, and for that reason Welwitsch 427

is treated in this paper as 5. ravenelii.

(35) Stereum notatum Berk. & Br.: Recorded by Kalchbrenner in Grev. 10 (1881)

59; Saccardo Syll. Fung. 6 (1888) 581; Doidge loc. cit. p. 491.

The specimens backing this record have been examined. They are J. M. Wood

No. 109 (Herb. S.A. Museum No. 34285), and MacOwan No. 1091 in Kew Herbarium,

ex Herb. Kalchbrenner.

The description of this species given in Saccardo would apply to practically any

young, effuso -reflexed Stereum, but according to Petch (in Ann. Roy. Bot. Gard.

Perad. 9, 1925, 264) Stereum notatum, a Ceylon species, is a “ bleeder ”. There is

no indication of conducting vessels or discolouration of the hymenium in either of the

two South African specimens under this name. The writer considers that both these

specimens are merely young immature, examples of Stereum hirsutum, and that

S. notatum should not appear in South African lists.

(36) Stereum ochraceo-flavum Schw. ex Peck; Burt in Ann. Mo. Bot. Gard.

7 (1920) 183.

Doidge (in Bothalia 5, 1950, 491) cites two specimens as the basis of this record

for South Africa. One of these, No. 30822, is a pale form of Stereum hirsutum. Of

the other, No. 22001 (MacOwan, 1091 b), all material is missing except two slides with

a few sections on each. The sections do not show the cystidia which Burt (loc. cit.)

describes for the species. S. ochraceo-flavum is thus considered a very dubious record.

(37) Stereum ostrea (Blume & Nees) Fr.; Recorded by Lloyd in Lloyd Myc.

Notes 6 (1920) 952; Listed by Doidge loc. cit. p. 489 as a synonym of S. fasciatum.

The writer has not seen the specimen cited by Lloyd. Burt (in Ann. Mo. Bot.

Gard. 7, 1920, 155) gives S. ostrea as a synonym of S. fasciatum. Bresadola (in

Hedwigia 51, 1912, 321) treats the species S. lobatum, S. concolor and S. perlatum

as forms of Stereum ostrea. The present writer is not in a position to evaluate the

soundness of this treatment.

(38) Stereum percome Berk. & Br. (!). Recorded by Doidge loc. cit. p. 491;

Massee in Journ. Linn. Soc. Bot. 27 (1890) 185.

Massee’s record is based on a specimen in Kew Herbarium under S. percome,

namely “ MacOwan, C.B.S. 9/83 ”. The writer has compared this specimen with

the type of S. percome and found that MacOwan’s specimen is Hymenochaete nigricans

(Lev.) Bres.

(39) Stereum perlatum Berk, in Hook. Lond. Journ. 4 (1842) 153; Doidge loc-

cit. p. 492.

The specimen on which this record is based is Universiteit van Stellenbosch,

Herbarium P. A. van der Byl No. 144, which Lloyd determined in Lloyd Myc. Notes

6 (1920) 952. It is probably only an old and weathered example of Stereum lobatum.

It is old, somewhat broken up, and in very poor condition, and should certainly not

have been made the basis of a new record for the country, especially considering that

S. perlatum is only critically separable from S. lobatum.

(40) Stereum proximum Lloyd, Syn. Stip. Stereum in Lloyd Myc. Notes 4 (1913)

40; Doidge loc. cit. p. 488; van der Byl in Trans. Roy. Soc. S. Afr. 10 (1922) 152,

Fig. 2; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 38.

— Stereum involutum (Klotzsch) Fries, see p. 317.

(41) Stereum pruinatum Berk. & Curt. (!).

323

The MacOwan collection (1227), Somerset East, in Herb. Kew. has been compared

with the type of S. pruinatum and is not that species. It is entirely resupinate with a

more or less chocolate-coloured hymenium, pruinose under the lens and much cracked

on drying. The margin is lighter, yellow-brown, appressed and somewhat fibrillose ;

texture spongy; 520 p thick. The structure is corticioid. The microscopic characters

are not very distinct, the hyphae are dark brown, about 4 p wide, with roughish walls ;

the walls are thickened but the lumen is quite distinct. The hyphae are branched,

without clamps, and of one type only. The basal tissues are more compact and run

out into hairs of the same type as the hyphae. It appears to have basidia which are

short and dumpy, and gloeocystidia are present. No spores were seen, except some

traced to an Aspergillus sporophore.

(42) Thelephora { Stereum ) pulverulenta Leveille (!) in Ann. Sci. Nat. ser. iii, 5 (1846)

145; Doidge loc. cit. p. 491.

The type of this species, namely Drege No. 9442, was kindly lent to the author

by Herb. Museum Paris. It proves to be undoubtedly a species of Hymenochaete,

most probably Hymenochaete luteobadia (Fr.) Hohnel & Litsch.

The citation by Doidge (loc. cit. p. 491) of Drege No. 9441 in this connection is

an error. Her citation of S. puherulentum Lev. as a synonym of S. murraii is no doubt

also an error, firstly because Leveille’s species has priority and secondly because Burt

(in Ann. Mo. Bot. Gard. 7, 1920, 131) indicates that S. pulverulentum Peck is the

synonym of S. murraii.

(43) Stereum purpureum (Pers. ex Fr.) Fries; Persoon in Roemer Mag. Bot.

1 (1794) 110, Obs. Myc. 2 (1799) 92; Fries Epicrisis (1838) 548, Hym. Eur. (1874)

639; Doidge in Bothalia 5 (1950) 492.

Thelephora purpurea Persoon, Syn. Fung. (1801) 571, Myc. Eur. 1 (1822) 121;

Fries, Syst. Myc. 1 (1821) 440.

Stereum rugosiusculum Berk. & Curt, in Grevillea 1 (1873) 162.

Fig. 17.

Fructifications coriaceous then later tough, lignicolous, resupinate becoming

reflexed, or sessile, dimidiate, single or laterally confluent, often closely imbricate,

the pileate part 0-4-1 -3 cm. radius and 1-5 cm. wide by lateral fusion. Surface light

brown or fawn with matted villose hairs forming a soft thick covering, sometimes with

a few obscure colour zones and concentric furrows towards the margin. Margin

sometimes concolorous, usually paler or greyish, often involute, sometimes narrowly

lobed by fusion of the pilei. Hymenium ceraceous or subgelatinous when fresh,

becoming horny on drying, smooth, purplish or violaceous, becoming fawn or brown

with a livid tinge when dry.

Basidia: 25^-0 x 3-5 p, cylindric to subclavate, with 4 sterigmata, forming a

dense palisade.

Spores: usually abundant, hyaline, smooth, elliptical with one side depressed

and a small lateral apiculus, or oblong-cylindrical, 2-5-3 x 5 -5-6 -5 p.

Cystidioles: not always present, 42-60 x 5-7 p, hyaline, not encrusted, smooth,

subulate to subcylindrical with a pointed or rounded apex, formed in the hymenium

and usually projecting above it.

Vesicles: ( 1 1 )— 1 7 - 5 p diam. or the same width and up to 20-30 p long, globular,

ovate or pyriform, sometimes elongated especially when near the hymenium, smooth,

thin-walled, terminal, with homogeneous contents, formed in a rather loosely-woven

tissue below the hymenium, evidently a form of gloeocystidium.

Hyphae: all hyaline, 3-4 p diam., some thick-walled, unbranched, non-septate;

others with thin walls, branched, septate and with rare clamp connections.

324

Tissue differentiation: hyphae of the lower part of the trama densely intertexed,

upper tissue bearing vesicles more loosely intertexed; there is a narrow pale brown

seam of hyphae subtending the abhymenial hairs.

Abhymenial hairs: thin- or thick- walled, free or fasciculate, 3-4 p diam., like

the hyphae.

Specimens examined: 15523, 36729; A. E. Eaton, Cape, as S. vorticosum in Herb.

Kew.

Cystidioles are not always present, and may be difficult to see when they are

present. The form with cystidioles was once distinguished as S. rugosiusculum, but

it appears that production of cystidioles is probably dependent on weather conditions.

The presence of vesicles is a very helpful diagnostic character for the species. Fructi-

fications of S. purpureum are apparently rare in South Africa though the pathological

effect of this fungus in producing “ silverleaf ” disease of fruit trees is quite well

known.

Judging from specimens in the National Herbarium, there has been a tendency to

confuse S. purpureum with the conidial state of Punctularia affinis (B. & C.) Talbot.

This must have been due to the similarity in their colour alone.

(44) Stereum pusillum Berk.; Saccardo Syll. Fung. 6 (1888) 559; Doidge loc.

cit. p. 492.

The specimens cited by Doidge have all been examined. Two of them are referred

to S. thozetii. The other, 31858 (Universiteit van Stellenbosch, Herbarium P. A

van der Byl No. 662) is in rather poor condition, but appears to be Stereum diaphanum.

S. pusillum should be omitted from South African lists.

(45) Stereum radicans (Berk.) Burt in Ann. Mo. Bot. Gard. 7 (1920) 108; van der

Byl in Ann. Univ. Stellenbosch 7 (1929) 37, PI. 2, 13; Doidge loc. cit. p. 492.

The/ephora radicans Berk, in Hooker’s Lond. Journ. Bot. 3 (1844) 190.

Fig. 23.

The material on which this record is based was studied in Universiteit van

Stellenbosch, Herbarium P. A. van der Byl No. 2384, and another part of it was kindly

lent to the writer by Dr. J. C. F. Hopkins from Mycol. Herb. Dept. Agric. S. Rhodesia

No. 3878 (Eyles, 4153). A description of this material is appended.

Fructifications terrestrial, thick, distorted, somewhat spathulate or flabellate but

rather indefinite in shape. There is no proper stipe but the pileus narrows towards

the base and might be considered laterally stipitate. Both surfaces are rugose and

radially furrowed. The upper surface has a hint of fine radial striae near the margin.

Margin entire, thick. Upper surface coloured chocolate; hymenium almost concolorous

but with darker blackish patches suggesting a “bleeding” species. Spores not seen.

Hyphae all hyaline, some wide with thick walls; others narrow. Long, deep-staining

elements which are a form of conducting vessel are present; they are up to 560 p or

more in length, unbranched, non-septate, thin-walled (4 - 8)— 6 - 5—1 1 p wide, very

abundant. Other miscroscopic characters are rather indistinct. The pileus colour

suggests a species of Thelephora, but the texture is that of a Stereum. The context

is light-brown in colour.

Burt’s description of S. radicans states that no cystidia or gloeocystidia are present,

and this statement is repeated by Martin (in Lloydia 7, 1944, 77). The conducting

vessels in Eyles’ specimen are so conspicuous that it would seem that it must be

unrelated to S. radicans , for such a feature could not have been missed if it were present

in S. radicans. Burt also emphasises the longitudinal striae on the upper surface of the

pileus; there is a hint of these in the present specimen, but no more. Martin (loc.

cit.) infers that a radicating base is not a constant feature of this species; certainly

325

it is not shown in Eyles’ specimen. Martin further states that his collections spproach

Cladoderris, another feature which casts doubt on the determination of Eyles’ specimen

as S. radicans.

Although there may be similarities in external form with S. radicans, the absence

of spores and the presence of conductors in Eyles’ specimen make a confident identi-

fication impossible.

(46) Stereum ravenelii Berk. & Curt, in Grevillea 1 (1873) 162; Doidge loc. cit.

p. 492.

Fig. 1.

Fructifications coriaceous, terrestrial, centrally stipitate, infundibuliform, 7 mm.

radius from attachment, 1-5 cm. diam. Stipe 6 mm. long, 1 mm. wide. Surface

glabrous, concentrically furrowed, with red-brown, bay or yellow-brown zones.

Hymenium yellow-brown to brownish when dry. Margin thin, entire.

Basidia: compact, 4-8 x 26-36 /a.

Spores: abundant, hyaline, smooth, small, ovate or elliptical, 3-2 X 4-8 /x.

Gloeocystidia : thin-walled, with homogeneous contents, up to 12-8 x 144 /x,

with a ventricose base or sometimes widest near the apex, abundant, embedded.

Hyphae : generative hyphae about 2 /x wide, hyaline, thin-walled with wide lumen,

septate, with clamp connections; skeletal hyphae hyaline, about 5 p. wide, with thicker

walls but with a wide lumen.

Specimens examined: Welwitsch (427, 425) ex British Museum Nat. Hist.

Lorrain Smith’s determination of these specimens as S. ravenelii has been accepted

here, as the specimens accord fairly well with Burt’s description (in Ann. Mo. Bot.

Gard. 7, 1920, 90) and no other material of this species has been available for study.

Welwitsch 425 is old material in which the details are obscure. The infundibuliform,

bay-coloured pileus, gloeocystidia and small spores appear to be the chief characteristics.

In these features, however, the present specimens correspond very closely with

S. nitidulum (see p. 321) and are only separable by the size of their gloeocystidia and by

hyphal characters. It is possible that the specimens treated here as S. nitidulum and

S. ravenelii represent only a single species. If so, that species is more likely to be

S. nitidulum than S. ravenelii.

(47) Stereum retirugum Cooke (!) in Proc. Roy. Soc. Edinb. (1882) 456; Doidge

loc. cit. p. 492.

Doidge’s record of this species refers to Mocambique; there is no material under

this name in the National Herbarium, Pretoria. Cooke’s type at Kew is synonymous

with Stereum schomburgkii Berk (!), which is now referred to Stereum fulvum (Lev.)

Sacc. (!).

(48) Stereum rimosum Berk. (!); Recorded by Lloyd in Lloyd Myc. Notes 4 ^1913)

L. 46, 4; Doidge loc. cit. p. 492.

The writer is of the opinion that this species is not typically represented in South

Africa, and the specimens cited by Doidge are all referred to S. rimosum Berk, var

africanum Talbot (see following description).

(49) Stereum rimosum Berk, var africanum Talbot in Bothalia 4 (1948) 495, Fig. 5;

ibid. 6 (1951) 38; Doidge in loc. cit. p. 492.

Stereum adnatum Lloyd (!) in Lloyd Myc. Notes 7 (1925) 1336, Fig. 3093;

Stevenson & Cash in Bull. Lloyd Library 35 (1936) 49; Doidge loc. cit. p. 487.

Fig 14.

Fructifications effused, resupinate-reflexed, often sessile and attached by a broad

umbo, or sometimes only narrowly attached and then composed of several connate,

subdimidiate pilei, coriaceous, becoming thickened, lignicolous. Surface cinnamon

326

buff colour, concentrically furrowed, covered with a thick, felty, pad-like tomentum.

Margin even or lobate. Hymenium closely and conspicuously rimose and rugose,

sometimes appearing blistered, cracking to show a pallid silky context, often concen-

trically furrowed, warm buff or pinkish buff in herbarium, yellow when fresh, bleeding

when bruised and often drying adustus to cinereous colour especially where bruised.

Width in section excluding the tomentum, 700-1000 /x.

Basidia: closely aggregated, 4-4-5 /x wide at apex.

Spores: hyaline, smooth, elliptic-ovate, with one side frequently depressed, with

a small attenuated apiculus, 2-3 X 3 - 5—5 • 5 /x.

Conducting vessels: yellow, 5 -5-8 -5 /x wide, with rigid walls, in a layer about

200 /x thick, distributed in the subhymenium and curving upwards into the hymenium,

not emergent.

Hyphae: thin-walled, hyaline, frequently septate, 3-5 /u. wide.

Tissue differentiation: there is a narrow orange coloured, dense zone subtending

the abhymenial surface.

Abhymenial hairs: thick-walled, much intertwined, almost hyaline to pale

yellowish, 4-2 /x diam.

Specimens examined: Type, 30233; 30268, 30777, 28285, 27755, 28296, 28303,

40211, 40217, 27565, 34375, 27772, 34365, 28295, 36891; 1708 (Type of Stereum

adnatum Lloyd); Universiteit van Stellenbosch, Herbarium P. A. van der Byl No.

1646 (as Stereum transvaalium van der Byl); Uganda, Maitland (460, 19 A).

Reasons for reducing S. adnatum Lloyd to synonymy with S. rimosum var africanum

are given in Bothalia 6 (1951) 39.

The variety africanum was described as “ not differing from the type in microscopic

appearance, with a resupinate-reflexed habit (neither pileate nor sessile-umbonate),

with a thicker tomentum, with a rougher, paler and more zonate hymenium ” (Talbot

in Bothalia 4, 1948, 945). With further specimens to hand, some are now included

which are only narrowly resupinate, and in others the hymenial colour is predominantly

cinerous with only patches of yellowish colour remaining. The change to cinereous

colour is definitely associated with the ability of this fungus to “ bleed ” when bruised.

Thick specimens showing a perennating tendency have been collected.

(50) Stereum rubiginosum Fries. Recorded under this name by Montagne in

Ann. Sci. Nat. ser iii, 7 (1847) 174.

A specimen has not been seen by the writer, but “ Stereum rubiginosum Fr.” is

a synonym of Hymenoebaete rubiginosa Dicks, ex Lev. This is noted by Doidge loc.

cit. p. 485.

(51) Stereum rugosum (Pers.) Fr. Doidge loc. cit. p. 492 records two specimens

under this name.

W. Nelson, Hout-Bosch Berg, 1880, in Kew Herbarium has been seen by the

writer and is referred to S. rimosum var africanum. No. 34394 is a young species of

Stereum, but not S. rugosum. It lacks conductors and also differs in habit and colour.

There is a third specimen in Pretoria as S. rugosum. No. 35559, which proves to be

S. durbanense van der Byl.

(52) Stereum sanguinolentum (Alb. & Schwein. ex Fr.) Fr., Epicrisis (1838) 549,

Hym. Eur. (1874) 540; Saccardo Syll. Fung. 6 (1888) 564; Doidge in loc. cit. p. 493;

Talbot in Bothalia 6 (1951) 37.

Thelephora sanguinolenta Alb. & Schwein., Consp. Fung. (1805) 274; Fries Syst.

Myc. 1 (1821) 440, F.lenchus Fung. 1 (1828) 178.

Fig. 15.

Fructifications coriaceous, lignicolous, thin, resupinate or effused becoming

narrowly reflexed, orbicular-confluent. Margin acute, pallid. Surface villous-strigose,

with short, adpressed, silky hairs, zonate and striate, whitish to some tint of buff.

327

Hymenium more or less cinereous when fresh, becoming light brown, smooth or

cracking rimosely to show a silky subiculum, often zonate. Flesh exuding a reddish

juice when wounded in fresh state. In section 400-500 p, thick.

Basidia: clavate, 26-40 x 4 -5-6 -5 /x.

Spores: hyaline, smooth, cylindrical, unilaterally depressed (6 • 4)— 8— (9) x

3-3-5 fi. '

Conducting vessels: conspicuous, reddish-brown, numerous, in the intermediate

tissues and curving upwards into the hymenium, 3—4 — (9) /x wide, very occasionally

forked, walls hyaline, smooth, occasionally thickened.

Hyphae: hyaline, 2-5-5 /x diam., with thin, firm walls and a wide lumen, without

clamps, septate, sometimes branched, densely interwoven in a more or less horizontal

direction.

Tissue differentiation: There is a narrow, dense, yellowish-brown zone subtending

the abhymenial surface.

Abhymenial hairs: simple, thick-walled, agglutinated, short, adpressed, 4-5 /x

diam.

Specimens examined: 28933, 33248, 40462, 40494.

Stereum sanguinohntum is usually considered a North Temperate species, and

was possibly introduced to South Africa on imported conifers. Its thin, almost papery

texture and its occurrence on conifers is at once a difference from all other South

African species of Stereum possessing conducting vessels. S. sanguinolentum is suspected

of causing a serious disease of Pinus taeda in Northern Natal.

(53) Stereum schomburgkii Berk. (!) in Journ. Linn. Soc. Bot. 13 (1873) 168;

Saccardo Syll. Fung. 6 (1888) 568; Talbot in Bothalia 6 (1951) 43; Doidge loc. cit.

p. 493.

= Stereum fulvum (Lev.) Sacc. (!). See p. 315.

(54) Stereum spadiceum Fr. ; Recorded by Lloyd in Lloyd Myc. Notes 4 (1913)

L. 46, 8; Brown in S.A. Journ. Sci. 33 (1936) 388; Doidge in loc. cit. p. 493.

All the specimens cited by Doidge have been seen. Mrs. Brown’s material is

Stereum australe Lloyd. The other specimens all lack conductors and are characteristic

Stereum hirsutum.

(55) Stereum subpileatum Berk.; Recorded by Lloyd in Lloyd Myc. Notes 5 (1917)

L. 66, 15, Note 634; Doidge in loc. cit. p. 489.

The material cited by Lloyd was made the type of Stereum durbanense van der Byl

(see p. 312). Stereum subpileatum does not occur in South Africa.

(56) Stereum tabacinum Sow. ex Fr. var australis Mont.; Recorded by Kalch-

brenner in Grev. 10 (1881) 58; Doidge loc. cit. p. 485.

The writer has not seen MacOwan’s material on which this record is based.

Steieum tabacinum var australis is a synonym of Hymenochaete tabacina (Sow. ex Fr.)

Lev.

(57) Stereum tenebrosum Lloyd (!), nomen nudum, in Lloyd Myc. Notes 5 (1918)

L. 67, p. 16; Doidge loc. cit. p. 493.

This species is sunk under Stereum australe Lloyd, for reasons given on p. 307.

(58) Stereum thozetii Berk. Austral. Fung. No. 268; Saccardo Syll. Fung. 6 (1888)

557; Doidge loc. cit. p. 493.

Fig. 3.

Fructifications soft-coriaceous, terrestrial among grass, perhaps on grass roots,

solitary or sometimes gregarious and then 2-3 pilei from separate stems may fuse

above into a single connate pileus; centrally stipitate, infundibuliform becoming

328

somewhat flattened and discoid later. Pileus 0-4-2 cm. radius from the attachment,

the disc 0-7-3 cm. diam., larger sizes often resulting from fusion of up to 3 pilei. Stipe

4 mm. long, 1 mm. wide, smooth or pruinose, creamy to light-brown colour. Surface

of pileus glabrous to pruinose, not tomentose or velutinate, concentrically zoned in

shades of pale yellow-brown to red-brown or brown. Hymenium smooth, creamy

or whitish in colour. Margin thin, entire, undulate, concolorous. Thickness in section

400-1100 ft.

Basidia: cylindric-clavate, 26-35 X 4-6-6 [x, with 4 sterigmata.

Spores: 4 -8-6 -4 x (6-4)-7-2-9 -6 [x, hyaline, smooth, frequently uniguttulate,

broad ovate to broad elliptical, sometimes showing a small apiculus, thin-walled.

Gloeocystidia : usually abundant and long, 3 -2-9 -6 p. wide x 66-186 g (or more)

long, hyaline, arising from generative hyphae, sometimes showing a basal clamp,

subcylindrical, sometimes somewhat ventricose at the base, at other times not at all

swollen and completely hyphoid, thin-walled, smooth, with homogeneous deep-staining

contents.

Hyphae: skeletal hyphae hyaline, not staining, usually thick-walled, occasionally

branched, not septate, 2 -4-3 *2 /x diam.; generative hyphae hyaline, thin-walled,

staining deeply, 1 -5-3 -2 [x diam., smooth, with occasional clamp connections, branched,

septate.

Tissue differentiation: There are no abhymenial hairs, nor is a dense zone

differentiated below the abhymenial surface.

Specimens examined: 8933, 31456, 13009, 8807, 28895; Herb. S. Afr. Museum

No. 45947; Herb. S. Afr. Museum No. 34266 (MacOwan 1232).

The identification of this species was checked at Kew by Dr. Dennis. In the

National Herbarium this species was formerly confused with S. pusillum and S. nitidulum,

while MacOwan’s collection cited above was originally determined as S. elegans. As

far as can be ascertained neither S. elegans nor S. pusillum occur in South Africa, and

S. nitidulum may be distinguished from S. thozetii by the bay-zonate pileus of the

former and its smaller spores.

(59) Stereum tomentosum van der Byl (!) in Trans. Roy. Soc. S. Afr. 10 (1922)

156 Fig. 9, in Ann. Univ. Stellenbosch 7 (1929) 45; Doidge loc. cit. p. 493.

= Stereum durbanense van der Byl (!). See p. 312.

(60) Stereum transvaalium van der Byl (!) in Ann. Univ. Stellenbosch 7 (1929) 40;

Doidge loc. cit. p. 494.

Two specimens are preserved under this name in the Universiteit van Stellenbosch,

Herbarium P. A. van der Byl Nos. 1472 and 1646. No. 1472 is taken to be the Type,

since it is the only specimen cited in van der Byl’s latin description of the species.

Most of his description, however, appears to be taken from No. 1646, or at least it

is compounded from both specimens, which are both cited in the Afrikaans text.

In the writer’s opinion, No. 1472 cannot be differentiated from Stereum australe,

while No. 1646 is referred to Stereum rimosum var africanum. The fact that the

description was partly based on this specimen of S. rimosum var. africanum may be

the reason why van der Bjl stated that the general appearance of the fruit-bodies

differentiated them easily from S. australe.

Stereum transvaalium must be taken either as a synonym of Stereum australe, or

as a nomen confusum.

(61) Stereum turgidum Lloyd (!) in Lloyd Mj>c. Notes 5 (1916) L. 63, 15, Note 502;

Doidge loc. cit. p. 494.

Lloyd (in Lloyd Myc. Notes 4, 1916, 549, Fig. 751) first described this species

under the name Stereum caperatum Lloyd, but as this was a later homonym of

S. caperatum (Berk. & Mont.) Massee, he later changed its name to S. turgidum. For

329

reasons given in Bothalia 6 (1954), p. 339, this species is regarded as synonymous with ,

Stereum cinerascens (Schw.) Massee (!).

(62) Stereum umbrinum Berk. & Curt. (!) in Grevillea 1 (1873) 164; Doidge loc.

cit. p. 494; Talbot in Bothalia 6 (1951) 41.

Hymenochaete vinosa (Berk.) Cooke (!) in Grev. 8 (1880) 149; Saccardo Syll.

Fung. 6 (1888) 600.

Hymenochaete scabriseta Cooke (!) in Ravenel, Fung. Amer. (1882) 717.

Hymenochaete purpurea Cooke & Morgan (!) apud Cooke in Grev. 11 (1883)

106.

Hymenochaete kalchbrenneri Massee (!) in Journ. Linn. Soc. Bot. 27 (1890) 116;

Saccardo Syll. Fung. 9 (1891) 230.

Fig. 19.

Resupinate, effused, sometimes narrowly reflexed, never pileate. Margin shortly

villose. Context soft, spongy. Hymenium velutinous, cracking but little in drying,

sometimes pitted, umber, vinaceous purple, purple-brown, light sandy brown or snuff

brown in colour.

Basidia: hyaline or very faintly coloured, about 6 x 30 p.

Spores: cylindrical or ellipsoid, hyaline, smooth, 6-8 x 3-4 p.

Cystidia: originating in the basal or middle part of the trama, curving up into

the hymenium and frequently projecting 10-20 p beyond; dark yellow-brown, lighter

colour when young or when emergent, not very thick-walled, encrusted or rugose

especially near the apex, rarely quite smooth, 100-250 x 7-9 p, cylindric-clavate or

fusoid, arising as apical modifications of the skeletal hyphae.

Hyphae: skeletal hyphae 6 -4-9 -6 p diam., yellow-brown, unbranched, without

septa; generative hyphae lightly coloured, 3-4-5-(6) p diam., thin-walled, branched,

septate, without clamps.

Tissue differentiation: the tissues are formed of loosely interwoven hyphae without

any denser abhymenial zone.

Specimens examined: 20974, 22044, 30220, 27626, 28294, 28702, 28276, 28277,

27767, 33392, 33400, 34357, 34393, 35419, 34381, 36839, 36710; van der Byl (2737)

in Herb. Kew.; MacOwan (1055) sub. Hymenochaete pellicula in Herb. S. Afr. Museum

No. 34315; MacOwan (1054) sub Peniophora cinerea in National Herbarium.

The cystidia of this species are apically modified skeletal hyphae. They do not

darken in alkali and in some specimens lack conspicuous incrustation or roughness,

but they are quite different from setae. The lack of a distinctive horizontal layer of

densely interwoven hyphae as an intermediate or basal tissue is characteristic of

relatively few species of Stereum. Its closest affinity is with S. papyrinum Mont. (!)

( = S. membranaceum Fr.), which is frequently pileate and always possesses wider,

more encrusted, more peniophoroid cystidia with thicker walls. Most South African

specimens of S. umbrinum have a purplish tinge rather than the typical umber colour.

(63) Stereum vellereum Berk.; Doidge loc. cit. p. 494; van der Byl in Trans.

Roy. Soc. S. Afr. 10 (1922) 156, in Ann. Univ. Stellenbosch 7 (1929) 47; Lloyd in

Lloyd Myc. Notes 5 (1917) L. 66, Note 584.

Lloyd’s conception of this species, which has been followed in South Africa is

that it is a fungus very like Stereum hirsutum but differing in being somewhat thinner

and having very pale or colourless surface hairs instead of the more yellow-brown

hairs of S. hirsutum. Microscopically there is no difference. The writer is doubtful

whether these distinctions are sufficiently marked to merit a different specific name.

330

, The following specimens examined show the pale to colourless surface hairs and

might be considered as S. vellereum: Universiteit van Stellenbosch, Herbarium P. A.

van der Byl Nos. 1717 (Eyles 3892), 2229, 143, 191, 147; National Herbarium, Pretoria

Nos. 15610, 30879.

(64) Stereum versicolor (Swartz ex Fr.) Fr.; Doidge loc. cit. p. 494.

The writer has not seen all the specimens cited by Doidge, but those seen are

not referable to S. versicolor. MacOwan (1276, 1163, as Stereum versicolor, Herb.

S.A. Museum 34265) is referred to Stereum lobatum.

(65) Stereum ( Hymenochaete ) villosum Lev.; Recorded by Lloyd in Lloyd Myc.

Notes 5 (1916) L. 63, 2.

The specimen referred to (No. 15558) is Hymenochaete luteobadia (Fr.) Hohnel

& Litsch. Doidge loc. cit. p. 485 cites Stereum villosum Lev. as a synonym of

Hymenochaete nigricans (Lev.) Bres. This synonymy is also given by Bresadola (in

Ann. Myc. 14, 1916, 232). The present specimen is however not H. nigricans.

(66) Stereum vitile Fries, Fungi Natalenses (1848) 23; Doidge loc. cit. p. 494.

Doidge notes that “ fide Wakefield this fungus has not been recognised since the

original collection and it is doubtful whether a specimen exists ”.

A fungus not unlike Stereum umbrinum is called for from Fries’ description, and

affinity with this species is suggested by Saccardo in Sacc. Syll. Fung. 6 (1888) 569,

and repeated by Massee in Journ. Linn. Soc. Bot. 27 (1890) 193.

(67) Stereum vorticosum Fr.; Recorded by Berkeley in Journ. Bot. London 14,

n.s. V (1876) 175; Doidge loc. cit- p. 492.

Berkeley’s material was borrowed from Kew Herbarium. It was labelled “ Stereum

vorticosum Fr. Pale form. Cape. A. E. Eaton ”, and proved on examination to be in

no way different from Stereum purpureum.

Key to accepted species of Stereum: —

1. Pilei infundibuliform and centrally stipitate. 2.

Pilei not infundibuliform nor centrally stipitate. 7.

2. With gloeocystidia (some may be interpreted as smooth cystidia). 3.

Without gloeocystidia. Here may be located some rare forms of Stereum lobatum (29) which

are infundibuliform by fusion and thus often partially split down one side.

3. Without surface hairs. Sections do not show any well-marked denser coloured zone immediately

beneath the abhymenial surface. 4.

With (scanty) surface hairs. Sections show a well-marked denser coloured zone beneath the

abhymenial surface. Here are located unusual forms of S. affine (2) which are usually only

infundibuliform by fusion and thus are often partially split down one side.

4.

5.

6.

7.

9.

Spores, small, ovate, broad-elliptic or subglobose, not larger than 3^1 x 4—5 p.

Spores, larger, in the range of 3-6 X 6 ■ 5-9 p.

Gloeocystidia up to 12-8 X 144 p in size. • S. ravenelii (46).

Gloeocystidia smaller, up to 10-5 x 80 p. S. nitidulum (34).

Gloeocystidia usually rather narrow (3 -2-9 -6 p wide) and often hyphoid. Hyphae of two types,

some with clamp connections. S. thozetii (58).

Gloeocystidia usually wider (8-9-6-16 fi wide) and clavate. Hyphae of one type only and without

clamps. S. diaphanum (15).

Pilei laterally stipitate, or flabellate or spathulate, or subsessile and cuneate attached by a markedly

reduced base. 8.

Pilei dimidiate or effuso-reflexed or resupinate. 13.

Fresh pilei bleeding red when bruised. Conducting vessels present microscopically in fresh or

dried plants. S. australe (6).

Fresh plants not bleeding. Conducting vessels absent. 9.

Without gloeocystidia or cystidia. 10.

With gloeocystidia or cystidia, or both together. 1 1 .

331

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Pilei small (up to 1 cm. in any direction), soft, whitish, without colour zones on the surface. Spores

pip-shaped or later distorted and angled. Hyphae monomitic.

S. cyphelloides (14).

Pilei large, coriaceous or tough, surface coloured with zones of grey, brown, chestnut. Spores

cylindric-depressed. Hyphae dimitic. S. lobatum (29).

With gloeocystidia but no cystidia. Pilei with a definite stipe, flabellate, spathulate or infundi-

buliform. S. affine (2).

With both gloeocystidia and cystidia. 12.

Pilei usually merismatoid, i.e. a compound fructification consisting of a number of smaller pilei

growing together in a bush. Gloeocystidia 7-10 X 40-66 p. S. involutum [26]

Pilei not merismotoid, but single or sometimes dimidiate or fused laterally. Gloeocystidia

smaller 7-12 x (15)-25-(40) p. Here is located the species represented by S. bellum (7) and

S. friesii (20) in the sense used by van der Byl.

Without cystidia, cystidioles, gloeocystidia, vesicles or conducting vessels (distinguish carefully

between skeletal hyphae which intrude into the hymenium and conductors or cystidia). 14.

With any of the following organs : cystidia, cystidioles, gloeocystidia, vesicles, conducting vessels

(Avoid locating here species which have intrusive skeletal hyphae unless these are much swollen

like cystidia at the apex). 18.

Mature pilei small (1 cm. or less) soft, whitish, azonate. Spores pip-shaped becoming angularly

distorted. Hyphae monomitic. S. cyphelloides (14).

Mature pilei larger, or if immature then either not whitish or possessing more than one type of

hypha. 15.

Skeletal hyphae in context brown. The skeletal hyphae which curve up into the hymenium are

brown, rugose or encrusted. Hymenium usually dark-coloured, only rarely yellowish or light-

coloured. [Compare also S. umbrinum (62) where the skeletal hyphae are much expanded

and resemble cystidia in the hymenium.] S.fulvum (21).

Skeletal hyphae in the context not brown, but at the most only pale straw-coloured or hyaline.

Hymenium not dark, usually creamy, yellow, orange, fawn or sometimes changing to cinereous.

16.

Pileus with multi-coloured zones on the surface, usually flabellate, or if a uniform brown colour

then the pilei are relatively large and flabellate. S. lobatum (29).

Pileus without multi-coloured zones on the surface, or zoned in shades of brown ; usually smaller

than S. lobatum and effuso-reflexed or dimidiate, not flabellate. 17.

Pileus thin, coriaceous, effuso-reflexed or dimidiate with a shortly villose or matted hairy surface.

S. hirsutum (24).

Pileus thicker (usually more than 1 mm. thick) corky or subligneous, effuso-reflexed, with a thick

padlike tomentum of ochraceous to golden hairs. S. durbanense (16).

Fresh pilei bleeding red when bruised. Fresh or dried specimens possessing conductors in the

hymenial layer. 1 9.

Fresh pilei not bleeding, lacking conductors at all times. 21.

Pilei generally dimidiate or cuneate with a reduced base, rarely widely effuso-reflexed. Hyphae

dimitic. Hymenium cinereous, smooth. S. australe (6).

Pilei mostly resupinate-reflexed, rarely dimidiate. 20.

Hymenium rimose, i.e. blistered and cracking into small rough areas, yellow, tan or cinereous.

Not on conifers. Hyphae dimitic. Usually more than 700 p thick.

S. rimosum var africanum (49).

Hymenium smooth, not rimose, cinereous to light brown. Occurring on conifers. Hyphae

monomitic. Usually less than 600 p thick. S. sanguinolentum (52).

Species possessing pyriform or subglobose vesicles embedded deep in the trama (some of the

vesicles are sometimes elongated and must be distinguished from gloeocystidia). 22.

Species without vesicles. 23.

Fructifications more or less resupinate, sometimes narrowly reflexed, stratose with a veined or

marbled subligneous context and a glabrous black abhymenial surface showing as a black line

in wholly resupinate specimens. Hymenium yellowish.

S. murraii (33).

Fructifications effuso-reflexed or dimidiate, not stratose or veined, with a hairy brownish surface.

Hymenium purple to purple-brown. S. purpureum (43).

Species with cystidia but lacking gloeocystidia. 24.

Species with gloeocystidia and sometimes cystidia as well. 25.

Cystidia large (12-24 p wide) conical or fusoid, encrusted, hyaline or only dilutely coloured.

Spores averaging 6 x 11 p. Hymenium light-coloured.

S. cinerascens (11),

332

Cystidia dark yellow-brown, subhyaline where emergent, actually only apically swollen and

encrusted or rugose skeletal hyphae (rarely smooth at apex). Spores 3-4 x 6-8 ,u. Hymenium

usually umber brown or purplish, rarely a light sandy brown. [Compare S.fulvum (21) whose

skeletal hyphae in the hymenium are less like cystidia, being not much expanded and roughly

cylindrical.] S. umbrinum (62).

25. Species with gloeocystidia but no cystidia. 26.

Species with both gloeocystidia and cystidia. 27. .

26. Context pale creamy to pale yellow-brown, usually stratose. Spores subglobose, 6-7 n dianr

(Gloeocystidia sometimes seen With difficulty.) S. duriusculum (17).

Context brown, contrasting with a hyaline hymenial layer. Spores 3^1-5 x 2-3 /j. Gloeocystidia

abundant, sometimes fragmented and refractile like cystidia.

S. bicolor (8).

27. Cystidia and gloeocystidia clearly differentiated. Context pale-coloured throughout. Hyphae

hyaline. Here may be located the species represented by S. helium (7) and S. friesii (20) in the

sense of van der Byl.

Only gloeocystidia present, but older ones are fragmented and highly refractile thus resembling

cystidia or mineral aggregations. Context brown, contrasting with a hyaline hymenial layer.

Basal hyphae mostly brown. S. bicolor (8).

EXPLANATION OF THE ILLUSTRATIONS.

The following lettering has been used throughout the illustrations: —

333

Fig. 1. — S. ravenelii. Fig. 2. — 5. nitidulum. Fig. 3. — S. thozetii.

Fig. 4. — S. diaphanum.

334

335

Fig. 9. — van der Byl (378) as Stereum friesii. Fig. 10. — • S . lobatum.

Fig. 11. — S.hirsutum. Fig. 12. — S. durbanense.

336

Fig. 13. — S. australe. Fig. 14. — S. rimosum var africanum.

Fig. 15. — S. sanguinolentum. Fig. 16. — S. murraii.

337

338

Fig. 21. — S. bicolor. Fig. 22. — S. duriusculum. Fig. 23. — Eyles (4153) as

S. radicans.

339

On the Genus Lopharia Kalchbrenner &

MacOwan.

By

P. H. B. Talbot.

Summary.

Four species of Lopharia have previously been described. Of these, L. lirellosa

Kalchbr. & MacOwan is regarded as a synonym of L. mirabilis (B. & Br.) Pat. It is

shown that the external morphology of the hymenium, upon which the genus Lopharia

is based, is a variable and unreliable character. Over several collections, intergrading

states of the hymenium link the species L. mirabilis, Stereum turgidum Lloyd and

Stereum cinerascens (Schw.) Massee, and no constant differences can be demonstrated

in the internal structure of these three species. S. turgidum and L. mirabilis are

accordingly recognised as synonyms of Stereum cinerascens, and the genus Lopharia

is sunk under the genus Stereum Pers. ex S. F. Gray.

Lopharia dregeana (Berk.) Talbot is found to be cospecific with Irpex vellereus

B. & Br., and the new combination Irpex dregeanus is made. It is suggested that

Lopharia javanica P. Henn. & E. Nym. may be based on a collection of Lopharia

mirabilis with immature spores. A sporograph supports this supposition, but detailed

evidence is lacking as the type of L. javanica was not available for study.

The writer examined type or authentic material of most of the species discussed

here. Such material is indicated by an exclamation mark ( !) after the specific epithets

listed in this paper.

History of the genus Lopharia.

In 1873, Berkeley & Broome described Radulum mirabile from Ceylon, in the

following words (in Journ. Linn. Soc. Bot. 14, p. 61): —

558 RADULUM MIRABILE, B. & Br. Primum orbiculare tomentosum, demum

confluens; hymenio perfecto hispidulo (No. 328). On dead wood. 5 inches long,

2 broad.

From this description it is certain that the species could not be recognised again,

but the type specimen was preserved in Kew Herbarium, where Massee studied it and

in 1892 (in Grevillea 21, p. 2, PI. 182, fig. 8-9) erected a new genus, Thwaitesiella, with

the single species T. mirabilis (B. & Br.) Massee. The description and illustrations

were competently executed.

Meanwhile, in 1881, Kalchbrenner and MacOwan had erected the genus Lopharia

(in Grevillea 10, p. 58) on the single species L. lirellosa, with the following diagnosis: —

LOPHARIA, K. et M. On. Hymenium cartilagineo-membranaceum g/abrum,

contiguum, in rugas interruptas, cristato-incisas elevatum, Phlebiae maxime affine;

sed in hac rugae acie integerrima gaudent.

LOPHARIA LIRELLOSA, K. et M. On. Effusa, plana, pallide rufescenti

carnea, subpruinosa, ambitu determinato, villoso-ciliatulo , albidiore; plicis interrupts,

varie curvatis, subramosis, cristato-incisas. Somerset East (1. MacOwan ). Ligno

arete adnata, placas oblongas f ormans. Plicae ad formam lirellarum Graphidis

eurvatae.

340

In 1895, Patouillard (in Bull. Soc. Myc. de Fr. 11, pp. 13-15, PI. 1) recognised

two species of Lopharia, viz. L. lirellosa and L. mirabilis, the latter being transferred

from the genus Radulum. Massee’s genus Thwaitesiella lapsed into synonymy.

Patouillard distinguished the two species on the disposition of the hymenial crests

and teeth, which he stated were more or less concentric and radiating about a central

papilla i ni. mirabilis, but irregular in L. lirellosa. He observed that the specific difference

was but slight and that the two were generically inseparable. He commented on the

similarity in internal structure of Lopharia and Stereum but differentiated the two by

the external appearance of the hymenium, which is smooth in Stereum. He noted

that young specimens of L. mirabilis were smooth and resembled a resupinate, orbicular

Stereum. In effect, he said that Stereum was a simple form related to the more highly

developed Lopharia and Cladoderris forms, between which there were obvious

similarities.

With the accumulation in the herbaria of specimens determined as L. mirabilis

and L. lirellosa, it has become evident that their separation on the basis suggested by

Patouillard is untenable. Their type specimens may show minor differences, but no

specific difference is maintained when a large number of collections is examined. The

hymenial configuration is variable and intergrades between the two forms even in a

single collection. Petch (in Ann. Roy. Bot. Gard. Perad. 4, 1910, 410) noted von

Hohnel’s opinion that the two species might be identical. After examining the types

and several other collections of both, the writer is convinced that they are synonymous

and henceforth in this paper refers both to the species L. mirabilis. He is aware that

other mycologists have recognised this synonymy in practice, but was able to find

only one statement of it in the literature at his disposal (van der Byl, 1934, loc. cit. infra),

and accordingly published the following nomenclator (in Bothalia 6, 1951, 56). Stereum

turgidum is now added to this list of species: —

Lopharia mirabilis (B. & Br.) Patouillard in Bull. Soc. Myc. de Fr. 11 (1895) 14,

PI. 1, Essai Taxon, sur les Hym. (1900) 74; Petch in Ann. Roy. Bot. Gard. Perpa.

4 (1910) 410; van der Byl in Ann. Uhiv. Stellenbosch 12 (1934) 2, fig. 1.

Radulum mirabile B. & Br. (!) in Journ. Linn. Soc. Bot. 14 (1873) 61 ; Saccardo,

Syll. Fung. 6 (1888) 496.

Thwaitesiella mirabilis (B. & Br.) Massee in Grevillea 21 (1892) 3, PI. 182,

fig. 8-9.

Lopharia lirellosa Kalchbr. & MacOwan (!) in Grevillea 10 (1881) 58;

Patouillard (1895, loc. cit.).

Licentia yao-chanica Pilat in Ann. Mycol. 38 (1940) 66, text fig. 2, Tab. 1,

figs. 1-3.

Stereum turgidilm Lloyd (!) in Lloyd Myc. Writ. 5 (1916) L. 63, Note 502;

Stevenson & Cash in Bull. Lloyd Library 35 (1936) 58.

Stereum caperatum Lloyd (!) in Lloyd Myc. Writ. 4 (1916) 549, fig. 751.

[non S. caperatum (Berk. & Mont.) Massee].

The writer has not seen specimens of Licentia yao-chanica, but Pilat’s description,

photographs and figures leave no doubt that this is the same species as Lopharia mirabilis,

and that the genus. Licentia should lapse.

In the paper referred to above (Talbot, loc. cit.), the very marked similarity of

L. mirabilis and Stereum cinerascens (Schw.) Massee, was noted. Since then the type

specimen of Stereum turgidum has been seen and is obviously referable to L. mirabilis.

A detailed comparison of these three species has now been made and is presented

below.

341

Comparison of Lopharia mirabilis , Stereum turgidum and Stereum cinerascens.

(1) Gross morphology of the hymenium.

Stereum turgidum is interesting as a close link between extreme forms of L. mirabilis

and S. cinerascens. Its type specimen is effuso-reflexed with a light-coloured hymenium

covered with irregular warts, teeth or incised ridges, not very regularly arranged but

nevertheless more or less radiating. Part of the hymenium is perfectly smooth, and

on other parts there are only a few, minute, scattered tubercles. Most specimens of

S. cinerascens are quite smooth, and some tend to crack on drying to reveal a silky

subiculum. Some others show a feeble development of hymenial papillae. L. mirabilis

shows either a geometrical or an irregular pattern in the distribution of its papillae,

teeth, or incised ridges, and may even bear shallow pores bounded by incised ridges.

The hymenium is usually uncracked, but may crack as in S. cinerascens.

The possession of incised ridges and crests, or teeth, the chief character of the

genus Lopharia , is by no means constant.

(2) Tissue distribution.

It was first thought that the distribution of the cystidia in the tissues might provide

a constant difference between the three species. Sixteen sections of different collections

of the three species were made. Some are illustrated in Fig. 1 to show that intergrading

occurs.

Fig. 1 . See pg. 345.

Characteristic specimens of L. mirabilis tend to be thinner and possess a single

stratum of hymenial cystidia mostly projecting above the hymenium, while characteristic

S. cinerascens tends to be thicker and have deeply embedded cystidia as well as those

projecting from the hymenium. Air spaces are common in S. cinerascens. Every

possible variation between these two extremes is shown in the single piece of the type

specimen of S. turgidum preserved in Pretoria Herbarium.

As shown in Fig. 1, other characters of tissue distribution are also unreliable

as points of difference. Here it should be noted that the tissue indicated as a basal

seam subtending the surface hairs, varies in density. It is sometimes almost

undifferentiated, but more often forms a narrow or wide, fight-coloured zone, with

well-defined boundaries.

Measurements of the microscopic features, excluding spores, of L. mirabilis,

S. cinerascens and S. turgidum, are summarised in Table 1.

Table 1.

MEASUREMENTS OF MICROSCOPIC CHARACTERS.

342

The measurements given in Table 1 were made several months apart and without

reference to each other. They show some differences, as is to be expected from

measuring only a few representatives of each type of organ, but also show a close

enough correspondence to be accepted as additional evidence for the identity of the

three species.

(3) Spore size and shape.

Ninety spores of L. mirabilis distributed among eleven collections, ninety-one of

S. cinerascens distributed among eight collections, and fifty of S. turgidum from the

type collection only, were measured. After the manner of Corner (in New Phytologist

46, 1947, 195), a sporograph was plotted from the spore measurements of each species

and the result is shown in Fig. 2.

Fig. 2. See pg. 346.

The sporographs indicate that the spores of all three species correspond in shape

throughout the stages of their development. The sporograph links, for example,

globose spores measuring 5x5 micrometer units, with oblong or elliptical ones

measuring 6 x 12 micrometer units (Note: 1 micrometer unit = 1-1 p). This suggests

that the spores of Lopharia javanica (cited as 5-6 x 5-7 p) may have been immature,

and that this species, if apparently different from L. mirabilis only in spore characters,

may in fact be the same species. That is only supposition, and needs to be tested by

reference to the type specimen, which is not available.

Statistical calculations from the spore measurements were made, and are shown

in Table 2.

Table 2.

SPORE MEASUREMENTS.

Comparing S. cinerascens and L. mirabilis : —

Difference in lengths =0-10 micrometer divisions.

Standard Error of difference — ± 0-22. Difference insignificant.

Difference in widths =0-05 micrometer divisions.

Standard Error of difference = + 0-12. Difference insignificant.

Comparing Stereum cinerascens and Stereum turgidum : —

Difference in lengths =0-99 micrometer divisions.

Standard Error of difference = ± 0-17. Difference significant.

Difference in Widths =0-09 micrometer divisions.

Standard Error of difference = 0-21. Difference insignificant.

343

In these comparisons there is a statistically significant difference only between

the lengths of the spcres of S. cinerascens and S. turgidum. It should be remembered

that the comparison was made on a series of specimens growing on different substrata

and collected and preserved at different stages of development. In the case of

S. cinerascens, the fact that eleven collections were examined tends to smooth out

these differences, but there is no such effect in S. turgidum of which only one collection

was available. In general these statistics give a useful confirmation of the identity of

the three species. For the purposes of mycological taxonomy, and taking into account

the variety of conditions under which the fungi grew, a difference in mean length of

the spcres of only about 1 p is scarcely worth consideration.

The foregoing evidence of macro- and microscopic characters establishes the

identity of L. mirabilis, S. turgidum and S. cinerascens as a single variable species, for

which the specific epithet cinerascens has priority. To decide the genus in which this

species should be placed, it is necessary to consider the types of the genera Lopharia

and Stereum.

Typification of the genera Stereum and Lopharia.

The genus Stereum Pers. ex S. F. Gray is accepted as validly published without

conservation (Rogers in Farlowia 3, 1949, 450). An acceptable type species for the

genus has already been selected by Rogers (in Farlowia 3, 1949, pp. 450, 480) and

independently by Donk (in Bull. Bot. Gard. Buitenzorg, ser. iii, 18, 1949, 98-99), from

among the five species first listed by S. F. Gray (in A Natural Arrangement of British

Plants 1, 1821, 652). This species is Stereum hirsutum, an eminently reasonable selection

which preserves the name Stereum in its present sense and conforms to the Rules of

Nomenclature.

Lopharia was erected as a monotypic genus with the species L. lirellosa, which

is thus the type species. This species, as has been shown, is synonymous with

L. mirabilis, but even if it were possible to point out specific differences these would

not materially affect the subsequent discussion on the status of the genus.

The type species of the genus Lopharia possesses the internal structure and tissue

distribution characteristic of the type species of the genus Stereum, except for the

presence of cystidia in the former. As we do not recognise cystidia as of generic

significance in Stereum, and, taking into account the demonstrable variability of the

hymenium in Lopharia, there remains no bar to reducing Lopharia to synonymy with

Stereum. The correct name for the species represented by the morphological forms

known as Radulum mirabile (1873), Thelephora cinerascens (1832), and Stereum

turgidum (1916), is Stereum cinerascens (Schw.) Massee. This species may be considered

as the type species of the genus Lloy della Bres., being the first species listed under that

genus by Bresadola (in Lloyd Myc. Writ. 1, 1901, 51), and conforming to the author’s

description of that genus. However, we do not recognise the genus Lloy della, which

differs from Stereum only in the possession of cystidia.

By combining the nomenclators given for Lopharia mirabilis in this paper, and for

Stereum cinerascens in a previous paper (Talbot, loc. cit.) a rather full synonymy is

obtained for Stereum cinerascens, in which the writer has examined the type or authentic

material for all the following species: Stereum cinerascens (Schw.) Massee, Corticium

aschistum Berk. & Curt., Peniophora berkeleyi Cooke, Stereum moricola Berk., Stereum

dissitum Berk., Corticium ephebium Berk. & Curt., Peniophora occidentalis Ellis & Everh.,

Lopharia mirabilis (B. & Br.) Pat., Lopharia lirellosa Kalchbr. & MacOwan, Stereum

caperatum Lloyd, Stereum turgidum Lloyd.

Other species of Lopharia.

Two other species of Lopharia have been described. The first, L.javanica P. Henn. &

E. Nym. (in Monsunia 1, 1889, 144), is said by Petch (loc. cit) to appear to differ from

9102-4

344

L. mirabilis only in the size of its spores. It is suggested earlier in this paper that if

this is really the only difference, then the sporograph may be used to relate the two

species.

The last species of Lopharia to be described was L. dregeana (Berk.) Talbot, which

the writer transferred from Corticium (Talbot, loc. cit.), and at the time expressed

doubt in the choice of genus. As the genus Lopharia has now been shown to be

untenable, it is necessary to reconsider L. dregeana. That combination was made in

good faith, but too hastily, for the writer has now discovered that Irpex vellereus

Berk. & Br. (!) is cospecific with the irpicioid forms of L. dregeana which he noted

earlier. Here then is another group of species, all with identical microscopic characters

but varying in the convolution of the hymenium. The hymenial variations intergrade

and cannot be accepted as constituting specific differences. The group is composed of

Corticium dregeanum Berk. (!) having a smooth hymenium except for a few, small,

scattered tubercles, Lopharia dregeana (Berk.) Talbot in which the former species was

associated with further specimens bearing a lopharioid hymenium, and Irpex vellereus

Berk. & Br., with irpicioid, flattened teeth coalesced at the base or containing specimens

with longer, discrete, hydnoid teeth.

Irpex lacteus Fr. was chosen as the lectotype species of the genus Irpex Fries, by

Clements & Shear (The Genera of Fungi, 1931, p. 346), from among the species first

described by Fries (Elenchus Fung. 1, 1828, 142-148). Reasons for this choice were

not given, but as it is thought to preserve the name Irpex in its present usage, it is

adopted here. With the genus thus typified it is clear that Irpex vellereus has been

correctly classified as to genus, but as the specific epithet dregeanum has priority the

new combination Irpex dregeanus (Berk.) Talbot is necessary. The synonymy of this

species is set out below.

Irpex dregeanus (Berk.) Talbot comb. nov.

Corticium dregeanum Berkeley (!) in Hooker’s Lond. Journ. Bot. 5 (1846) 3;

Saccardo, Syll. Fung. 6 (1888) 636; Montagne in Ann. Sci. Nat. ser. iii, 7 (1847) 174.

Hymenochaete dregeana (Berk.) Massee in Journ. Linn. Soc. Bot. 27 (1890) 1 14.

Lopharia dregeana (Berk.) Talbot in Bothalia 6 (1951) 57.

Irpex vellereus Berk. & Br. (!) in Journ. Linn. Soc. Bot. 14 (1873) 61 ; Saccardo,

Syll. Fung. 6 (1888) 489; van der Byl in Ann. Univ. Stellenbosch 12 (1934) 4,

figs. 5-6.

The author is much indebted to Dr. R. W. G. Dennis of Kew Herbarium, to

Dr. D. P. Rogers, and to the Director and Staff of the Commonwealth Mycological

Herbarium, for illuminating discussions of some problems in nomenclature.

345

Fig. 1. — Sections of the fructifications of Stereum turgidum (A, B, C), Stereum

cinerascens (D, F) and Lopharia mirabilis (E, G), drawn diagrammatically with the

aid of a camera lucida to show variation and intergrading of internal characters, and

also to show the distribution of the tissues.

346

Fig. 2. — Sporographs of Stereum cinerascens, Lopharia mirabilis and Stereum

turgidum, relating the length of the spores to the ratio length/average width of spores.

Points determined by less than five readings are not shown on the graph.

347

Some Graminicolous Species of Helminthosporium

and Curvularia Occurring in South Africa.

By

K. M. Putterill.

The form-genus Helminthosporium is so broadly defined that it has accommodated

many types of dematiaceous fungi possessing phragmospores, despite fundamental

differences they may have shown. Since its foundation in 1809, the genus has become

an unwieldy and heterogeneous collection of species, out of which a number of smaller

genera need to be formed as natural groupings become apparent. Curvularia is another

form-genus, recently split from Helminthosporium, but even in this comparatively

small group there may be no true relationship between species, since their perfect

states are still unknown.

Cylindro-Helminthosporium, a subgenus discussed later, appears to show more

affinities among its species than merely a common characteristic conidial form. Several

of its species also produce a Pyrenophora stage, indicating close relationship throughout

their life-cycle, which is the only true criterion of affinity. Similarly the species of

Eu- Helminthosporium are conidial states of the ascomycete genus Cochliobolus.

In the present paper, twenty species distributed in Helminthosporium and Curvularia

are described. Most are new records for South Africa; some are previously undescribed

species. The perfect states of these conidial forms have not yet been seriously sought

in this country, but are recorded elsewhere for some of these species. The descriptions

given here are based primarily on South African material.

Not the least of the taxonomic problems concerned in this study was that of

determining the generic limits of Helminthosporium. For example in species such as

H. miyakei (see p. 367), the conidia are strongly tapering and unlike those found in

Eu- or Cylindro-Helminthosporium, but rather approaching the form of Cercospora

conidia. Chupp gives evidence that species such as Cercospora phaeocarpa Mitter

(56, p. 239) with thick-walled conidia might be better classified under Helminthosporium,

but such species differ in several important features from typical Helminthosporia.

C. phaeocarpa possesses a stromatic base, the upper cells of which form one-celled

conidiophores about 6-15 /x long. The development of conidiophores on a

pseudoparenchymatous foundation is not a feature ordinarily associated with

Helminthosporium, which typically produces separate or only basally fused conidiophores

each as a several-celled branch of a basal hypha. Obviously, dealing with form-genera,

there must be borderline cases and intergrading between genera, each setting its

individual problem for the taxonomist.

Collections in the National Herbarium, Pretoria, include a number of

Helminthosporium forms parasitic on species of Meliola, and not typical of either

Eu- or Cylindro-Helminthosporium. They have been described by Hansford (21).

The conidial forms present numerous other problems connected with their, great

variability in nature and in culture. First is the fact that many are collective species

consisting of several strains differing from one another in morphological or

physiological detail, but forming a continuous series within specific limits. The protean

nature of H. sativum, for example, has become so apparent with intensive study that

systematists have had to define the species in more and more elastic terms. This has

348

had the unfortunate result that other Helminthosporia associated with foot-rots

are often overlooked. Important parasites have been studied most, and are thus known

to have several strains; the tendency to regard newer or little-known species as more

rigid is quite unjustifiable. The policy adopted in this paper has been to use existing

names rather than tentative new ones, for experience has shown that the full range of

a species may not be covered in the first description. H. leersii (see p. 365) is cited

as an example of this conservative treatment.

In nature, as well as in artificial culture, new forms constantly arise by saltation.

In culture different strains may even be obtained from a single conidium. Some species

are more prone to saltation than others, and there are certain conditions not necessarily

identical for all species, which will predispose a fungus to behave thus. The literature

on H. gramineum, H. sativum and H. sacchari (q.v.) deals extensively with controlled

experiments to study conditions inducing saltation.

Related to the phenomenon of saltation are those of staling and deterioration in

culture. Some species are much more unstable than others under conditions generally

found suitable for the maintenance of Helminthosporia in the laboratory. Species

presenting the greatest difficulty are those which appear in insufficient quantity on the

natural substratum, yet deteriorate so rapidly in culture under standard conditions,

that it is impossible to continue obtaining characteristic samples. H. bicolor and

H. rostratum, in particular, soon resulted in cultural forms so different from the original

as to lead to conflicting identifications discussed under these titles.

The fact that various strains of a species may differ from one another in physiological

behaviour or pathogenicity is most important in phytopathology. As host varieties

may also differ in susceptibility to fungal attack, much caution is necessary in drawing

conclusions from varietal resistance tests.

The possibility of attack by more than one fungus at a time requires mention.

In association with other parasites, a weakly pathogenic strain may accomplish what

it could not do alone, or mixed strains may inhibit the growth of one another. Mixed

associations have been encountered repeatedly in the course of this study, and are

discussed under the heading of H. sativum and other foot-rotting species.

Though this paper is not primarily concerned with the disease aspect, yet any

account of species of Helminthosporium would be incomplete without reference to

their relationship to their hosts. This aspect is dealt with under each species, and

references to phytopathological literature are given. Disease symptoms are characteristic

only within certain limits; eye-spot, stripe, foot-rot, etc., are symptoms which may be

produced by several different species. Symptoms attributed to many species described

in this paper are uncertain, often because of the presence of other fungi. The need

for further investigation in view of the incompleteness of such descriptions is stressed.

Host differences are an aid in the distinction of species of Helminthosporium ,

but are not completely reliable. Although many Helminthosporium species are very

similar each must stand on its own morphological characteristics; too often spqcies

have been described as “ new ” because they were found on previously unrecorded

hosts. H. halodes is a case in point. Here there might have seemed ample reason

for prejudice against using, for a form from foot-rotted wheat, a name given to a species

from so different a substratum as Distichlis spicata (L.) Greene, growing in a salt-marsh

subject to occasional inundation by sea water.

In this study, prune agar, Brown’s standard agar, and autoclaved host tissues

were the media used. Cultures were incubated at 25° C. Every species received in

living condition was studied in pure culture as well as on the host. Measurements of

conidia and other structures were taken from frequency charts compiled from at least

one sample from each medium, and more where obtainable from the host. Up to

100 spores were measured in a sample, depending on the variability of the spore

population.

349

My thanks are due to Mr. E. W. Mason of the Commonwealth Mycological

Institute at Kew, Dr. Johanna Westerdijk of the Centraalbureau voor Schimmelcultures

at Baarn, and Dr. Mitra of the Agricultural Research Institute at Pusa, India, for

confirming my identifications of three of the species and for naming two others dealt

with here.

Natural Subdivisions of Helminthosporium, and Associated Ascigerous Forms.

Mention has been made in the Introduction of natural groups within the large

form-genus Helminthosporium. It will be convenient to examine in this section all

available reports of ascigerous stages connected with Helminthosporium species, none

of which have been known to develop the “ perfect ” state in South Africa. The reason

for this may be that under our milder winter conditions there is no need for the fungus

to pass into the more resistant sexual state; the conidia, especially thick-walled ones,

are sufficiently resistant themselves. Alternatively Helminthosporia from cereal crops

may be maintained between seasons on secondary wild grass hosts, or even on self-sown

seedlings of the crop itself. This is known to happen in coastal and other districts

where the winter is mildest.

Sometimes the mere beginnings of a perithecial body are laid down; there are

many records of “ immature perithecia ”, “ immature sclerotia ” and “ sclerotial

bodies ”. In culture, many species develop thick “ knots ” or mycelial complexes

containing abundant anastomoses. Similar sclerotial growths occur naturally. But

even if perithecia should be formed, and asci within them, it is well known that if

suitable conditions intervene before the ascospores are mature, then further development

of these spores may be suspended in favour of the immediate formation of conidiophores

and conidia, even from the surface of the perithecium. Thus the influence of climate

or environment on the life-cycle is profound, and it is partly this fact that has caused

our understanding of the complete life-cycles of these fungi to develop so slowly.

Species of Helminthosporium are the conidial stages of some Pyrenomycetes.

In the fact that they do not all pass into the same ascigerous form we have the most

striking corroboration of the belief that Helminthosporium is a large composite genus,

presently to be split up into more natural groups. A start has now been made. In

1929, Nisikado (43) proposed the division of Helminthosporium into two subgenera.

He based his work on Drechsler’s (15) observation that the conidial shapes and modes

of germination fall into two main groups, and that this division corresponds with

the possession of Ophiobolus or Pyrenophora perithecial stages so far as these are known.

Thus Nisikado’s Eu-Helminthosporium possesses fusiform conidia germinating from

the poles only, together with an Ophiobolus stage; Cylindro-Helminthosporium is

characterised by cylindrical conidia germinating from intermediate as well as polar

cells, and by a Pyrenophora stage. Not all the species of Helminthosporium studied

by Nisikado have yet been finked with a perithecial stage; but a glance at the summary

below will show that if any one can be satisfactorily classified on conidial characters,

its ascomycetous stage may be predicted with fair certainty.

With the heterogeneous nature of Helminthosporium thus exposed, the need for

a special generic name under which to describe the full life-cycle of a fungus such as

H. teres — Pyrenophora teres becomes insistent. That need was not met by Ito (28)

when, in 1930, he raised Nisikado’s Cylindro-Helminthosporium to generic rank under

the name Drechslera, for apparently the term was to refer only to the conidial stage.

In 1930, Pyrenophora was no more precise in meaning than Helminthosporium, thus

Ito could have chosen to break away from both old generic names and describe the

full fife-cycle under a new genus of Ascomycetes. The genus Pyrenophora Fuckel

became confused when Saccardo incorporated into it various non-sclerotioid forms

previously known as species of Chaetoplea. However, in 1934, Drechsler (19) ably

surveyed the field and pointed out that “ through the elevation of Chaetoplea to generic

350

rank by Clements & Shear (7), Pyrenophora is automatically rehabilitated as a natural

genus in the sense defined and applied by Fuckel, being reserved for the hard, sclerotioid

perithecial forms having their asexual stages in Helminthosporium- forms such as

H. teres, H. bromi, H. tritici-repentis with indiscriminate germination, and corresponding

broadly with Nisikado’s subgenus Cylindro-Helminthosporium and Ito’s genus

Drechslera ”. And further, “ to this sense it would seem to be highly desirable that

further usage and application should rigorously conform”.

Species of Pyrenophora included by the authors discussed above comprise the

following : — -

P. phaeocomes (Reb.) Sacc. (19; 49, vol. 11). Type species.

P. relicina (Fuckel) Sacc.

P. teres Drechsl. (14, 19, 60). Connected with H. teres.

P. bromi (Died.) Drechsl. (14, 19). Connected with H. bromi.

P. tritici-repentis (Died.) Drechsl. (14, 19). Connected with H. tritici-repentis.

P. gr amine a Ito (28). Connected with H. gramineum.

P. avenae. Connected with H. avenae.

P. japonica Ito (28).

P. polytricha da Camara (9). Connected with H. olisiponense.

Forms possibly connected with the same series: P. chaetomioides Speg. (49,

Vol. 16). Said to be connected with H. penicillosum Speg.

Forms not yet connected with Pyrenophora, but named by Ito as species of

the genus Drechslera (28) : D. arundinis (H. arundinis ) ; D. catenaria (H. catenarium ) ;

D. gigantea (H. giganteum ); D. tritici-vulgaris (H. tritici-vulgaris).

Forms connected with the same series by Drechsler: H. erythrospilum Drechsl.

(19 a); H. dictioides Drechsl. (19 a); H. siccans Drechsl. (19 a); H. vagans Drechsl.

(18).

The present use of the name Pyrenophora in preference to Pleospora of so many

authors requires comment. Pyrenophora was erected in 1849 and Pleospora eight years

later. Fuckel restricted the genus Pyrenophora to forms with hard sclerotioid perithecia.

Saccardo, however (49, v. 11) distinguished Pyrenophora as “ Pleospora , but with

setose perithecia ”. The latter simple distinction has been generally adopted, though

some authors, notably Winter, considered the possession of setae insufficient reason

for the maintenance of a separate genus. Drechsler (19) discusses the point and

concludes that “ excessive emphasis on the presence or absence of setose outgrowths

as a criterion for distinguishing Pyrenophora and Pleospora has obscured much more

important differences in life-histories and structure ”. The type species of Pleospora,

P. herbarum (Pers.) Rab., is associated with a Macrosporium conidial stage, while the

restricted group of species referred definitely to Pyrenophora by Fuckel have all been

associated with Helminthosporium. Stripped of non-sclerotioid forms, and distinguished

also from Pleospora, Fuckel’s genus Pyrenophora now stands out as a natural one.

Under this name, therefore, we leave Cylindro-Helminthosporium Nisikado.

Turning now to Eu-He/minthosporium Nisikado, it is clear that a similar need

exists here for a name which will indicate the occurrence of both Helminthosporium

and Ophiobolus stages in this combination only. Drechsler has shown that the majority

of species in the collective genus Ophiobolus are not connected with a Helminthosporium

stage at all, but instead with a Phoma type of asexual stage. Further, those that do

have a Helminthosporium stage agree among themselves (and differ from other

Ophiobolus species) in having a pronounced helicoid arrangement of the ascospores

within the ascus, and consequently unusually long ascospores and wide asci. The

mutual relationship of forms with Eu-Helminthosporium — Ophiobolus life-cycles being

351

without doubt natural, the group may be removed from both genera and established

independently. For these species, Drechsler (19) erected the new genus Cochliobolus.

So far the genus Cochliobolus includes the following species: —

C. heterostrophus Drechsl. (19) type species. Synonyms Ophiobolus

heterostrophus Drechsl. (15, 16) and Helminthosporium maydis Nisikado & Miyake

(44).

C. sativus (Ito & Kuribayasbi) Drechsl. (31, 19). Connected with H. sativum.

C. kusanoi (Nisikado) Drechsl. (42). Connected with H. kusanoi.

C. miyabeanus (Ito & Kuribayashi) Drechsl. (29). Connected with H. oryzae.

C. setariae (Ito) Drechsl. (28). Connected with H. setariae.

C. stenospilus (Carpenter) Drechsl. Connected with H. stenospilum.

It is clear that the conidial stages of Pyrenophora and Cochliobolus account for a

large number of the known graminicolous species of Helminthosporium, but there still

remain other species from grasses that cannot be assigned to either genus, and these

will now be considered.

H. dematioideum, which germinates by production of germ-tubes from the basal

cell, is one of these. Another may be H. miyakei with conidia of unique form. There

are also the species parasitic on Meliola in which the conidial shape is also different.

H. giganteum Heald & Wolf (17), produces a whorl of germ-tubes at each end of

the conidium. In addition to producing normal conidia, it sporulates abundantly in a

manner like Hormodendrum. A sexual stage has not yet been found, but these two

features alone are enough to show that the affinities of this species lie in another direction.

Then there is the relationship between H. sigmoideum Cav. and Leptosphaeria

salvinii Catt., demonstrated by Tullis (58). H. sigmoideum is unusual in having conidia

with rather pointed, differently coloured end cells, and the sharp pointed “ sterigmata ”

of the conidiophores are unique. Finally it produces the well-known sclerotial stage,

Sclerotium oryzae, which is commoner than the conidial stage itself. Examination

of the genus Leptosphaeria appears to justify the suspicion that L. salvinii is misplaced

in this genus, and that its independence should be recognised by a new generic name.

In the literature at her disposal, the author has found no other Leptosphaeria having

a Helminthosporium stage. Several Leptosphaeria species have phomaceous conidial

stages, while what appears to be the type species of the genus, L. doliolum (Pers.) de Not.,

is stated by Saccardo (49) to have a Periconia conidial stage. Leptosphaeria as it appears

in the Sylloge Fungorum is a large assemblage in which smaller groups will certainly

prove to be independent natural genera, and the fungus under discussion is probably

one. We cannot, however, go further than this interesting speculation until we have

seen authentic specimens and examined the work of da Camara (9) on the genus, neither

of which is available to us at present.

Finally there is the related series of small-spored species for which Boedijn (2)

erected the genus Curvularia, under which name they are discussed in this paper. Most

of these develop more or less cylindrical stromata, but apparently perithecia have not

yet been reported for any of them.

Species of Helminthosporium found in South Africa.

(1) Helminthosporium teres Saccardo, in Fung. Italici (1881) t. 833, Sacc. Syll.

Fung. 4 (1886) 412.

Helminthosporium hordei Eid. in Der Landwirt, Bd. 27 (1891) 509.

Illustrations'. Fig. 1.

352

Conidia cylindrical, straight or nearly so, usually with a slightly uneven contour,

both ends rounded; basal septum often very slightly constricted so that the basal

segment is characteristically rounded in outline. Mature conidia, especially when

beginning to germinate, are frequently slightly constricted at the septa; walls thin;

colour sub-hyaline or tinted faint smoky yellow-green, to deeper smoky yellow; hilum

large, included in basal contour. Conidia 30-195 X 10-23 p [over 200 p long and

with 1-11 septa fide Smith & Rattray (65)], borne on geniculations or swellings of

the conidiophores. Conidiophores stout, septate, usually simple, yellowish-brown or

olivaceous, usually 80-230 x (6)— 7— 9— ( 11) p, wider across the swollen basal cell,

emergent from stomata or between epidermal cells in groups of two or occasionally

three, possessing large distinct scars left by fallen conidia.

Ascigerous stage: Pyrenophora teres Drechsl. Cultures of H. teres are characterised

by abundant anastomoses and the consequent formation of small, round black sclerotia

in large numbers. Sclerotial bodies form naturally on dead plants during autumn

and if conditions are suitable, give rise to peritbecia in the spring. Diedicke (11)

originally suggested the name Pleospora teres for the perithecial stage, though perithecia

had not then been found. The complete ascigerous stage was first described by

Drechsler (14) under Pyrenophora and in view of the fact that Diedicke’s fungus was

purely hypothetical, Wakefield & Moore (60) have observed that the correct citation

of the species is Pyrenophora teres Drecnsl.

Drechsler (14) states that conidia of H. teres always have a slight constriction of

the basal septum so that the basal segment is characteristically rounded. Smith &

Rattray (53) have not found this feature constant in South African specimens; the

present writer finds the feature well represented, if not quite constant in all specimens

examined. In South Africa, the conidia seem to reach somewhat larger dimensions

than those recorded elsewhere.

H. teres Saccardo (1881) was later described by its author in terms which make

it impossible to be distinguished from either of the two other species of Helminthosporium

common on barley. Ravn (47) taking into account differences in destructiveness,

distinguished the then known leaf-stripe disease (which he attributed to H. gramineum)

from what he called Helminthosporiosis (i.e. “ net-blotch ”) of barley. To the fungus

causing the latter, he applied Saccardo’s binomial, H. teres, an opinion later confirmed

by Saccardo himself. Further comment differentiating these species will be found

under H. gramineum.

Specimens examined: F. 693, on Hordeum, Pretoria, 6/10/09; 8411, W. Fletcher,

on Hordeum vu/gare, Elim, 20/10/14; 21787, Krieger Schadlicbe Pilze Exsicc., on

Hordeum distichum, Europe; 25884, N. J. G. Smith, on Hordeum, Belmont Valley,

Albany, 1929; 25885, N. J. G. Smith, on Hordeum, Grahamstown, June 1930; 25886,

N. J. G. Smith, on Hordeum, Swellendam, July 1930; 25887, N. J. G. Smith, on Hordeum,

Swellendam, 7/7/30; 30377, J. Sel/schop, on Hordeum vulgare, Louis Trichardt, Sept.,

1938; 30855 and 30856, J. Sellschop, on Hordeum, Louis Trichardt, Aug., 1939.

Pathological notes : H. teres causes “ net-blot cb ” of barley, which, though common,

is not so serious as leaf-stripe caused by H. gramineum. In primary infections the

seed-borne fungus may infect the first leaf, as found with H. gramineum, resulting

similarly in a pale stripe, which may turn brcwn. Severely infected young plants may

closely resemble those suffering from leaf-stripe disease. The later leaves are not

ordinarily infected by contact, so that the stripe form of the disease is rarely seen after

the seedling stage. Instead, conidia blown from primary lesions cause secondary

infections of the foliage, and the irregular blotches thus produced are by far the

commonest symptoms to be seen. Typically these blotches or streaks are smallish,

reticulate patches of narrow dark brown fines over a fighter background, and usually

surrounded by a yellowish border. Very elongated lesions may resemble the stripes

caused by H. gramineum, though their placing is more haphazard. The grain is similarly

353

contaminated by windblown conidia, but though secondary infections may reach

epidemic proportions they rarely reduce the yield seriously. Ascospores maturing

in spring cause infections of the secondary type.

“ Net-blotch ” is common wherever barley is grown. Because of the conditions

necessary for secondary infections, it is most prevalent in damp places or seasons.

Perithecia have not yet been recorded in South Africa. In this country it is not even

necessary for the fungus to be seed-borne, for wherever the climate is suitable it may

live on self-sown plants, or plant remains, all the year round.

(2) Helminthosporium gramineum Rabenhorst, Herb. Myc. No. 332 (1856);

Saccardo, Syll. Fung. 10 (1892) 615.

Brachysporium gracile (Wallr.) Sacc. var. gramineum Rabenh., Sacc. Syll. Fung.

4 (1886) 430.

Napicladium hordei Rostrup, Sygdomme landbr. for. Snyl. (1893) 130-132.

Illustrations'. Fig. 2.

Conidia nearly always straight, typically cylindrical, often somewhat narrower

in the upper half, broadly rounded at each end, the hilum included in the basal contour,

subhyaline to yellowish-brown; walls fairly thin; rarely slightly constricted at the

septa. Conidia measure 20-120 x 11-22 g, with 1—7— (8) septa. Germination from

polar and intermediate segments. Smaller secondary spores are frequently formed

on short processes from germinating primary conidia. In isolates containing secondary

spores, these and the smaller primary spores are indistinguishable, so that the range

of spore size is extended at the lower end. The larger spores in an isolate are thus

to be regarded as more characteristic of the species and more important in its

identification.

Conidiophores: light yellowish-brown, lighter in the upper part, simple, septate,

bearing successive conidia on moderate-sized geniculations leaving distinct scars after

separation. Conidiophores 30-200 g long, and 5-9 g wide above the swollen basal

cell; arising singly or in groups of 2-6 from stomata or between epidermal cells.

Ascigerous stage: Pyrenophora graminea Ito. Diedicke (10, 11) reported an

ascigerous stage of H. gramineum, referring it to Pleospora trichostoma (Fr.) Wint.

He regarded this as a collective species and later split it up, naming the form from

barley P. graminea. Drechsler (14) considered that the fungus thus described by

Diedicke was actually the perfect stage of H. teres. Ito & Kuribayshi (30) agreed with

Drechsler’s conclusion. Although Drechsler mentioned that Paxton had found the

mature perithecial stage of H. gramineum neither of these workers described the fungus.

Ito (28) first published a description of Pyrenophora graminea as a new species.

H. gramineum was one of the first parasitic graminicolous Helminthosporia to be

described but it was many years before it was clearly distinguished frcm other species

of Helminthosporium on the same host. Ravn (47) first differentiated H. gramineum

and H. teres and correctly observed the disease symptoms attributable to each. In

1910 a third parasite of barley, H. sativum, was described by Pammel, King & Bakke.

Early descriptions of H. gramineum and barley leaf-stripe are therefore confused.

Ravn pointed out that microscopically the conidia and conidiophores of H. teres

and H. gramineum are hardly distinguishable from one another. The writer has found

that Drechsler s distinctions (14), though useful, do not invariably hold good. It

was found that “ typical ” conidia of H. gramineum attained a greater maximum length

and breadth than recorded by Drechsler and others, and may have up to 8 septa, while

the conidiophores were net markedly narrower than those of H. teres. As Smith (53)

states: “ the two are best distinguished from one another by the commonness of spores

in H. teres samples which are superior in length to any conidia developed by the other

species ”. H. sativum is easily distinguished microscopically from the above two species,

by its conidia.

354

Specimens examined: 21785, 21786, Krieger: Schddliche Pilzen, on Hordeum

vulgare; 25888, N. J. G. Smith , on Hordeum, Grahamstown, 27/6/30; 25889,

N. J. G. Smith, on Hordeum, Aylesbury, Albany, C.P., Aug. 1929; 30861, 30862,

J. Sellschop, on Hordeum, Branddraai, Lydenburg, Tvl., Sept. 1939.

Pathological notes: H. gramineum causes leaf-stripe disease of barley, which is

one of the most widespread and destructive maladies of this crop, especially in colder

climates. Many workers in colder countries have claimed that infected plants cannot

recover. In South Africa the disease does not ordinarily cause great losses, for conditions

are such that the host plant is able to outstrip the parasite. Smith has described

leaf-stripe very fully, both in its typical form (50) and as it occurs in South Africa (53).

Leaf-stripe is most easily recognised if the manner of formation of the strip is

clearly understood. The disease is seed-borne, and Smith (50, 53) has shown that the

fungus does not grow up with the plant by maintaining itself in a smut-like fashion

in the growing point, as Ravn (47, 48) claimed. Instead the coleoptile becomes infected

by the seed-borne fungus and transfers its infection to the first small enclosed leaf,

which, as it grows itself enwraps and so infects the second young leaf. Each successive

leaf, and finally the young ear, become infected in turn from its wrapping. As the

elongating leaves and sheaths brush past the source of infection continuous vertical

strips become infected and show the typical stripe. At first the stripes are yellow but

later turn brown. The leaves may split along the stripes. The development of a

typically striped plant depends on the relative growth rates of the host and parasite.

A vigorous plant which outstrips the fungus may show only a partially developed

form of the disease. If the fungus reaches the ear it may prevent normal seed formation,

while if it invades growing points the plant may be killed before maturity, as happens

in severe outbreaks of the disease.

It is evident that symptoms of this disease may be confused with elongated

secondary lesions caused by H. teres or H. sativum. In these latter species, however,

the lesions are caused by wind-blown conidia and are usually entirely haphazard.

Wind-blown conidia of H. gramineum do not seriously affect plants during the growing

season.

Under certain conditions H. gramineum has been known to cause foot-rot of

barley. It has also been isolated from wheat, oats and maize.

(3) Helminthosporium mediocre sp. nov.

Illustration : Fig. 3 a.

Conidia brownish-yellow, 40-108 X 13-18 /x, with 2-10 septa, sub-cylindrical or

long elliptical, usually straight, sometimes slightly curved, widest in the middle or

slightly below it, narrowing somewhat towards the ends both of which are broadly and

similarly rounded; walls of medium thickness, hilum included in the basal contour.

Germination bipolar.

Conidiophores: light olive-brown, 80-150 g. long and 5-5-8 ^ wide, usually with

a swollen basal cell; emerging singly from the epidermis.

Ascigerous stage: unknown.

Helminthosporium mediocre sp. nov.

Conidia: flavo-fusca, 40-108 X 13-18 p, 2-10 septatis, subcylindrica vel longe

ellipsoidea, crassitudo maxima ad vel paullo infra medio, fere recta, interdum leviter

curvata, plerumque attenuata, ad basin et apicem late et similiter rotundata; membranae

mediocres, hilum in extrema basalis inclusum. Germinatio bipolaris. Conidiophora :

pallide olivacea, 80-150 p longis et 5-5-8 p latis; cellulis basalis fere incrassatis, ex

epidermi singulariter emergens. Typus No. 30756.

355

H. mediocre, both in culture and on the host plant, is devoid of any extraordinary

distinguishing feature. No record of any similar species of Helminthosporium on

Pennisetum could be traced. A portion of the type was sent to the Commonwealth

Mycological Institute but no name for it could be suggested.

H. leucostylum Drechsl. and H. nodulosum B. & C. have been found capable of

attacking Pennisetum typhoideum under experimental conditions (39), but both of these

differ from H. mediocre in the size and shape of their conidia. Other records of species

of Helminthosporium on Pennisetum occur in phytopathological literature, but none

were found to include a description of the fungus.

In culture on prune agar, the conidia (Fig. 3 b) were found to be smaller, darker,

with thicker walls and with the end cells often somewhat lighter. The end cells were

sometimes emphasised. The conidia measured only 33-82 x 12-15 p. The

conidiophores appeared longer in culture as they merged gradually into long, narrow

hyphae below.

Specimens examined: 30756, L. C. C. Liebenberg, on Pennisetum clandestinum,

Schagen, Tvl., June, 1939.

Pathological notes: The fungus occurred on discoloured basal sheaths and leaf

blades of Pennisetum clandestinum in numerous brown spots about 0-2-1 -5 mm. x

1-2 mm., or larger by coalescence. The spots were slightly elongated lengthwise.

As ether fungi were also growing in these lesions, and no inoculation tests were made,

it is not known whether this species was parasitic.

(4) Helminthosporium sativum Pammel, King & Bakke in Iowa Agr. Expt. Sta.

Bull. 116 (1910) 178-190; Saccardo Syll. Fung. 25 (1931) 822.

Helminthosporium acrothecioides Lindfors. Svensk. Bot. Tidskr. 12 (1918) 227.

Illustrations : Figs. 4a & 4 b.

Conidia'. Long-elliptical, widest near the middle, tapering towards the rounded

ends, typically slightly or distinctly curved, greenish or brownish-olivacecus when

mature, often opaque, lighter at the ends, 26-120 x 12-26 p, with 3-10 septa; wall

thick, thinning towards the ends; hilum well marked. Germination bipolar.

Conidiophores: Erect, simple, septate, 6-8 p wide, 60-300 p long, olivaceous,

paler at the apex; arising from dead host tissue from between epidermal cells or

stomata, either singly or in groups of 2-4, bearing about 6 conidia. Variable in culture,

sometimes branching irregularly and bearing many conidia.

Ascigerous stage: Cochliobolus sativus (Ito & Kuribayashi) Drechsl. (19).

The perfect stage of this fungus was unknown until 1929, when it was described

by Kuribayashi (31) as Ophiobolus sativus.

Helminthosporium sativum was first described (Pammel et al. loc. cit.) from barley.

Its presence undoubtedly caused a great deal of confusion in descriptions of other

congeneric cereal parasites before that date.

H. sativum is most difficult to define satisfactorily. Its astonishing variability

in morphology and parasitic behaviour has been extensively studied by many workers

(6, 14, 37). The species has been found to comprise many races, which, though

morphologically inseparable, differ greatly in pathogenicity and cultural behaviour.

Morphological characters may vary widely under different conditions. New races

arise with ease as saltants in culture, and apparently also in nature. The large-spored

species of Helminthosporium, as Drechsler (20) observes, offer the greatest possibilities

for variation in their length and septation with the consequent tendency for such species

to be divided into new races or strains.

The writer found that conidia produced in cultures of H. sativum differed greatly

from those occurring naturally (Fig. 4 b). In culture the conidia were often irregularly

shaped, bent rather than merely curved, and sometimes showing bifurcate ends. Many

were scarcely longer than wide, and had none or few septa, which were often irregularly

356

spaced or set at odd angles. Conidia formed on greenhouse plants, and occasionally

under natural conditions, tended to show the same peculiarities. It is thus hardly

possible to cefine a typical conidium, but we may assume that the most typical are

those from cultures.

Specimens examined: 25890, N. J. G. Smith, on Hordeum, Grahamstown, 27/6/30;

30421, J. Pons, on Triticum vulgare, Lydenburg, Tvl., 17/10/38; 30493, J. Sellschop,

on Triticum vulgare, Bethlehem, Nov. 1938; 30857, 30858, 30859, J. Sellschop, on

Hordeum, Louis Trichardt, 10/8/39 ; 30860, J. Sellschop, on Hordeum, Lydenburg,

Tvl., Sept. 1939; 30990, 30991, J. Pons, on Triticum vulgare, Lydenburg, Tvl., Oct.

1939.

Pathological notes: H. sativum is parasitic on barley, wheat, rye and other cereals

and wild grasses, on which it causes more than one type of damage. It appears to

be world-wide in distribution (6, 13, 14, 22, 24-27, 33, 36, 37, 38, 40, 53, 54). It is

especially severe on wheat and barley in the United States. In South Africa it is common

on both wheat and barley, and has also been isolated from oats and wild grasses.

The nature and severity of the disease symptoms vary with the mode of infection,

environment, strain of the fungus and the host concerned. This variability is noticed

even on a single host. Strains from different hosts have been found to be

interchangeable. The types of damage described below occurred on barley and wheat,

but though a set of experiments with a single strain cannot be typical of all cases, it

applies roughly throughout the host range.

The fungus may be carried in the grain, where, for example, it causes “ black

point ” of wheat. In such cases, if germination is not prevented, seedling blight may

result. Plants which survive or escape this may be attacked by root- or foot-rot caused

by the same fungus. Such plants display marked discolouration round the base and

sometimes a bronzing of the foliage. Similar rots may be caused in plants from clean

seed by conidia that have lodged in the soil. Severely affected plants die or are dwarfed,

and may be prevented from tillering. If infection is less severe, or the host plant

withstands the infection, the parasite may actually stimulate the plant to excessive

tillering. The heads, if formed, rarely fill properly, and the grain may be shrivelled.

H. sativum may also cause a disease known as “ spot-blotch Here, conidia

formed on the diseased parts cause secondary infections of the later foliage or stems

in the form of small spots or blotches. These are up to 10 mm. long, or longer on the

stems, at first yellow but later turning brown with or without a yellowish margin. The

ears and grain may also be infected by the conidia. Where the foliage is much reduced

by disease, the grain usually ripens prematurely if it is formed at all.

Though it has been found (e.g. Mitra & Bose, 38) that the optimum conditions

for damage are plentiful moisture and warmth this result may be misleading. Plants

weakened by drought may fall prey to parasites which they might otherwise have

resisted. Foot-rot occurs in dry as well as in moist areas. Mourashkinsky (40),

working on the control of cereal diseases in arid districts in Russia, states that seedling

wilts of cereals, caused by H. sativum and other species of Helminthosporium and

Fusarium, are of some importance in dry regions. He finds it erroneous to believe that

drought is so unfavourable for disease development that control measures may be

relaxed. The longevity of the thick-walled conidia of H. sativum, together with its

ability to grow saprophytically on dead or dying tissue of the Gramineae, makes this

fungus difficult to control.

South African conditions are seldom such as to encourage the development of

serious disease from secondary infections. The writer’s experience has been that

although the “ spot-blotch ” form of the disease is common in the summer-rainfall

areas, the foot-rot form is much more destructive.

Examination of numerous plants affected by foot-rot in different parts of the

country has shown that H. sativum occurs commonly in company with a fairly constan t

set of other fungi. These fungi have been isolated from the bases of many living plant

357

showing foot-rot, but not from dead material. Species of Fusarium (especially F.

avenaceum ) and Curvularia (e.g. C. spicifera) were the most constant and important

associates. Other fungi that were isolated more than once were: H. bicolor Mitra,

H. halodes var. triiici Mitra, H. rostratum Drechsl. and H. dematioideum. As many

as seven such associates were isolated on one occasion from a single plant which was

not particularly reduced. Sometimes H. bicolor or H. halodes was associated with a

Fusarium and/or a Curvularia on plants attacked by foot-rot, though H. sativum was

not present. No doubt there are other fungi which may participate.

In the writer’s opinion, foot-rot of wheat in South Africa is very often attributable

to the combined action of a complex of fungi, rather than to H. sativum alone. If so

it would explain why H. sativum isolated from foot-rot specimens often proves to be

feebly parasitic on re-inoculation. But the tendency of this species to produce saltants

in culture should not be overlooked. The point can only be established by a series of

experimental inoculations of these fungi singly and in combinations. Hynes (24, 26),

in Australia, has undertaken just such a series of experiments. He found damage to

be greatest when plants were inoculated simultaneously with H. sativum, Fusarium

culmorum and Curvularia ramosa — a remarkably close parallel with what the writer

has found in South Africa. Hynes did not, however, find in Australia all the large-spored

species of Helminthosporium named in the present paper, though most of these have been

associated with foot-rot elsewhere. Hynes concluded that “ the large-spored species

of Helminthosporium associated with foot-rot throughout Australia is predominantly,

if not exclusively, H. sativum ”. This conclusion does not find support in South Africa.

(5) Helminthosporium N Henry in Minn. Agr. Expt. Sta. Tech. Bull. 22 (1924).

Illustrations : Fig. 5.

Conidia: Long-elliptical, widest near the middle, tapering somewhat towards

the rounded ends, slightly or distinctly curved, sometimes straight; wall fairly thick,

thinner towards the ends; hilum well-marked, not .protruding, but flat rather than

included in the rounded basal contour; conidia fuliginous or greyish-olivaceous,

lighter at the ends, 16-82 x 8-16-5 p, with 1-10 septa. Germination bipolar.

Conidiophores: variable, often branching below the conidium-bearing part,

geniculate, up to 7 p wide. Only seen in culture.

Ascigerous stage : unknown.

This species has conidia very similar to those of H. sativum, but rather more

uniform in shape. The most pronounced difference is in the width of the conidia,

although differences in length and colour are also apparent. Henry (22, p. 35) gives

a table comparing variations and constants for the conidia of the two species. Both

he and Drechsler considered Helminthosporium N sufficiently distinct from H. sativum

to be regarded as a new species but neither named it specifically. Both noted a

resemblance in the spores of Podosporiella verticillata O’Gara (45).

Specimens examined: 30497, J. Sellschop, on Triticum vulgare, Bethlehem, Nov.

1938.

Pathological notes: Helminthosporium N was originally reported on wheat, where

it produced symptoms similar to those caused by H. sativum. It was, however, less

virulent. It is probably more common than it appears from the records, being very

liable to be identified as one of the many variants of H. sativum.

The South African isolation was made from the culm bases of wheat attacked

by foot-rot. No inoculation experiments were done, so its relationship to the host

and to other foot-rotting fungi present remains in doubt.

358

(6) Helminthosporiura bicolor Mitra in Trans. Brit. Myc. Soc. 15 (1930) 286.

Illustrations : Fig. 6.

Conidia: roughly cylindrical with bluntly rounded ends, or long-elliptical and

widest near the middle then narrowing towards the rounded ends, straight or slightly

curved; wall thick, thinning towards the end segments; hilum included in the basal

contour or occasionally showing slightly outside; olivaceous-brown or yellowish- to

smoky-brown to dark brown, sometimes opaque, paler in the end segments;

20-90-(108) X 8—18 /x; septa 1-12, the terminal septa thicker and darker. Germination

bipolar.

Conidiophores: brownish, variable in culture, from 160 p long; the short fertile

region markedly geniculate, sometimes forked, 5-7 p wide, narrower below; conidia

borne at short intervals on the geniculations.

Ascigerous stage: unknown.

All the South African isolates of this taxonomic species agree well with the

description of H. bicolor Mitra, except that the majority of their conidia are longer

and have marked end-septa. In view of this discrepancy, two isolates (Nos. 30496,

30432) were sent to Dr. Mitra, who confirmed their identity as H. bicolor, but noted

that accentuated septa are not shown in the type. In these two strains isolated from

Wheat and sorghum respectively, and from different localities, conidia up to 108 p

in length occurred. The maximum length in other South African isolates was 75-90 p,

and 79 p in the type specimen. These unusually long conidia were slender, sometimes

distinctly curved, and much lighter in colour (see three conidia figured at top of Fig. 6)

than the usual conidia for this species. They occurred only in the fresh specimens and

in the first culture isolations from fresh material, but by means of monospore cultures

it was proved that subcultures from the long conidia later produced only average

conidia. The two extremes thus belonged to the same species, which showed many

intermediate forms in culture. In addition, examination of standing conidiophores

in culture showed that the first-formed conidia were of the long, light-coloured type,

and that the shorter, darker conidia were formed later and more abundantly.

Cultures of this species were among the most rapidly staling of any encountered

in this study (cfr. introduction), and staling was apparently accompanied by a decrease

in the size of the conidia. Measurements of conidia in staling cultures of No. 30496

agreed well with those for H. euchlaenae Zimm. In the description of the latter there

is no mention of emphasied terminal septa or lighter end cells. While it is possible

that H. bicolor and H. euchlaenae are synonymous this point requires further

investigation, and at present the South African material is better disposed under

H. bicolor. A summary of measurements relevant to this discussion is given in Table 1,

below.

Table 1.

359

Specimens examined: 30432, L. Codd, on Sorghum halipense, Pretoria; 30458,

J. Sellsthop, on Triticum, Clarens, O.F.S., Nov. 1938; 30496, J. Sellschop, on Triticum,

Bethlehem, Nov. 1938.

Pathological notes : H. bicolor was described as the cause of foot-rot and root-rot

of barley and wheat seedlings in India. In South Africa it has been isolated from

wheat showing foot-rot, grown in various parts of the country. This fungus has also

been found associated with lesions surrounded by dull-red borders, on leaves of

Sorghum halipense and S. verticilliflorum. In some cases Helminthosporium turcicum

was also present.

(7) Helminthosporium halodes Drechsler in Journ. Agric. Res. 24 (1923) 707.

Illustration: Fig. 7.

Conidia long-elliptical, straight or slightly curved, widest at or above the middle,

narrowing somewhat towards the broadly-rounded apex and more markedly towards

the narrowly-rounded or roughly triangular basal cell; walls fairly thick, becoming

thinner towards the ends; hilum protruding, large, conspicuous; septa 1-11, the

terminal septa thick and dark; colour greenish- or yellowish-brown to dark or

smoky-brown, lighter in the end segments; 20-85 X 10-21 p. Germination bipolar.

Conidiophores (in culture) simple, usually long, 5-8 p wide in the enlarged fertile region.

Ascigerons stage: unknown.

The above description is based on material in culture. It is evidently identical

with H. halodes Drechsl. var tritici Mitra (36, p. 287), which Dr. Mitra himself kindly

confirmed. The variety differs from Drechsler’s species in spore size and septation.

Mitra gives the extreme spore measurements as 23-73 x 13-20 p, with 2-9 septa, while

the extreme measurements for Drechsler’s type are 20-105 x 10-14 p with 1-12 septa.

In length and septation the conidia of the South African strain thus fall between the

two, and the conidia may attain a slightly greater width than Mitra claimed.

The present description, like Mitra’s, was based on cultures obtained from

foot-rotted wheat, and no naturally occurring conidia were seen. Drechsler found that

H. halodes showed a tendency towards the production of shorter, thicker and darker

conidia in culture. This is a variation towards the form shown in H. halodes var.

tritici, and it is thus not improbable that the latter is an ill-founded variety. In view

of the known cultural irregularities and abnormalities of H. halodes, as striking as

those found in H. sativum, the author would minimise the seemingly important difference

in the width of the conidia and dispose her material under H. halodes rather than under

the variety tritici.

Specimens examined: 30434, K. M. Putterill, on Triticum, Brits, Oct., 1938.

Pathological notes: H. halodes was originally described on Distichlis spicata in

the United States of America. The symptoms it produced resembled those caused by

H. sativum. It has been reported as a parasite of sugar-cane seedlings, maize, sorghum,

wheat and barley in India (55). The variety tritici associated with foot-rot of wheat

in India (Mitra, 36) was found to infect barley seedlings as well. In South Africa,

H. halodes was isolated from foot-rotted wheat.

(8) Helminthosporium rostratum Drechsler in Journ. Agric. Res. 24 (1923) 722.

Illustrations : Fig. 8.

Conidia long-elliptical, straight cr slightly curved, widest near the middle, with

a broadly rounded to subconical base; apex broadly rounded in the shorter conidia

but narrowing markedly in the longer conidia, often elongated into a rostrate extension;

walls thick, thinning towards the ends; hilum large and strongly protruding; septa

3-21 (usually more than 10), the terminal septa usually accentuated except the apical

septum in rostrate conidia; greenish-yellow when young, olivaceous- or yellowish-brown,

dark olivaceous or blackish-brown at maturity, more dilutely coloured in the end

360

segments and with a very light area round the hilum and the extreme apex;

20-180 X 10-26 /x, usually 13-20 /x wide. Germination bipolar. Conidiophores (in

culture) variable, simple, 160 /x long, or longer, the fertile region expanded, 6-9 p wide,

geniculate, olivaceous.

Ascigerous stage: unknown.

Both isolations made of this fungus showed extreme conidial variation in culture.

Drechsler (loc. cit.) states, “ on artificial media, the fungus produces conidiophores

bearing conidia altogether similar to those found in nature ”. In the author’s material,

rostrate conidia agreeing well with H. rostratum were produced in the first set of cultures,

but appeared less readily in the second set and were absent from subsequent cultures.

In being half their original length and tending to become obovate in the shortest conidia,

the later conidia strongly resembled those of H. halodes. In a private communication

Dr. Westerdijk confirmed the author’s identification of No. 30456 as “ H. rostratum

Drechsl. with typical rostrate conidia ”, but within two months on subcultures the

same fungus, still in pure culture, was indistinguishable from H. halodes Mitra, a fact

which Dr. Mitra kindly confirmed. In view of these facts and the variability of this

fungus in culture, which is paralleled by the cultural behaviour of H. sativum and H.

bicolor, the South African isolations are best assigned to H. rostratum.

Specimens examined: 30456, J. Sellschop, on Triticum, Bethlehem, Nov. 1938;

30460, J. Sellschop, on Triticum, Ferraira, Nov. 1938.

Pathological notes: Helminthosporium rostratum was described (loc. cit.) from

the dry leaves ofr Eragrostis major in the U.S.A. Its author stated, “ the mature

condition of the grass precluded any inquiry into the biological relation of the fungus

to its substratum ”.

The South African isolations were made from the base of wheat plants affected

by foot-rot. One isolate (No. 30456) was found in association with H. sativum,

H. bicolor, Helminthosporium N, Curvularia spicifera, Curvularia sp., and Fusarium

avenaceum. Five of these fungi have been found on grasses affected by foot-rot. The

second isolate was associated with H. sativum, H. dematioideum, Fusarium sp. (probably

F. avenaceum ) and Curvularia sp. No inoculations were made to test the ability of

H. rostratum to cause foot-rot alone.

(9) Helminthosporium turcicum Passerini in Bol. Comiz. Agr. Parmense 10 (1876);

Saccardo Syll. Fung. 4 (1886) 420.

Helminthosporium inconspicuum Cooke & Ellis in Grevillea 6 (1878) 88.

Illustrations'. Fig. 9.

Conidia long elliptical, straight, less often slightly curved; broadest near the

middle and narrowing decidedly towards the ends; apex rounded; basal cell conical,

often more rounded; thin-walled; hilum prominent, characteristically protruding.

Septa l-8-(ll). Subhyaline with a grey-green tint, to light yellow-brown.

50— 1 40— ( 154) x 15-26 /x. Germination bipolar. Conidiophores simple, few-septate,

fertile region moderately geniculate, light olivaceous; 100-280 X 7-10 p; emerging

in small groups from the stomata.

Ascigerous stage: unknown.

Specimens examined: F. 890, H. G. Mundy, on Zea mays, Rhodesia, Feb. 1910;

1134 May Varney, on Sorghum vulgare, Kansas, U.S.A., Oct. 1889; 7913, Briosi &

Cavara on Zea mays, Pavia, Italy, 1889; 2187 J. Fisher, on Zea mays, Cedara, Natal,

March 1912; 9729, Angus & Co., on Sorghum verticilliflorum, Dalton, Natal, March

1916; 10099, J. Pope-Ell is, on Sorghum halipense, Ashburton, Natal, April 1919;

15448, Th. Eruleben, on Sorghum sudanense, Pietermaritzburg, March 1922; 15462

361

R. R. Staples, on S. sudanense, Cedara, Apr. 1922; 20354, Entomologist, on Zea mays,

Belgian Congo, March 1935; 20386, McKay, on Zea mays, Naboomspruit, Apr. 1935;

28600, L. C. C. Liebtnberg, on S. halipense, Prinshof, Pretoria, March 1936; 30431,

30432, K. M. Putterill, on S. halipense, Prinshof, Pretoria, Dec. 1938; 30708, K. M.

Putterill, on Sorghum verticilliflorum, Prinshof, Pretoria, Jan. 1939; 30709, K. M.

Putterill, on S. sudanense, Prinshof, Pretoria; 30710, K. M. Putterill, on S. halipense,

Prinshof, Pretoria, Jan. 1939.

Pathological notes: Helminthosporium turcicum is the cause of leaf blight or

“ white-blast ” of maize, a conspicuous and often serious disease. It occurs in all

maize-growing countries. The same disease has also been described on various species

of Sorghum.

In South Africa, Doidge & Bottomley (12) found that leaf blight “ causes

considerable damage (to maize) in Natal in wet seasons, but appears to have slight ill

effects elsewhere ”. In the summers of 1937-39 entire plantings of Sorghum sudanense

at Prinshof, Pretoria, were rendered useless by the disease. In Rhodesia (23), and

Madagascar it completely destroyed maize crops, especially of newly introduced

varieties, while it is regarded as important enough in Kenya to justify extensive work

on the breeding of resistant varieties.

The symptoms of this disease are evident as small straw-coloured lesions, often

surrounded by a darker outline, which may later merge into one large blotch covering

almost the entire surface. Fully grown plants are not usually much affected, but where

the infection occurs early in the wet season on young plants, the destruction of large

leaf-areas with considerable subsequent damage to the crop, may result. On species

of Sorghum the symptoms resemble those on maize, except that the smaller lesions have

reddish-brown borders, or may consist of yellowish blotches turning reddish or blackish

later.

Elsewhere H. turcicum has been found capable of infecting barley, oats, rice,

sugar-cane and wheat, under experimental conditions. Although various workers

have found the strains on these hosts interchangeable, it is suspected that the strains

from maize and Sorghum may be biologically different.

Drechsler (15) believes that leaf blight caused by H. turcicum may often have masked

the simultaneous appearance of leafspot caused by Cochliobolus heterostrophus (H.

maydis Nisi. & Miyake) on maize plants. Thislatter fungus would otherwise have

been more widely known in maize areas. It has not yet been recorded in South Africa.

(10) Helminthosporium leucostylum Drechsler in Journ. Agric. Res. 24 (1923) 710.

Illustrations'. Fig. 10.

Conidia usually straight, widest about one-third from the base, tapering evenly

and characteristically towards the narrow apex; base paraboloid or sometimes broadly

rounded; wall thick, thinning in the terminal cells to conspicuous thin areas at the

extreme apices and around the hilum which is conspicuous but not protruding; medium

to dark olivaceous, or olive-brown, lighter in the terminal cells; 1-8 septate, not

constricted at the septa ; 15-80 x 8-17 /x.. Germination bipolar. Conidiophores simple,

very short, the fertile region markedly geniculate, the tip somewhat flattened; typically

subhyaline to light fuliginous, darkening lower down; 25-105 x 5—6 -5 /u.; emerging

in groups from the stomata. Well-spaced conidia leave prominent scars on separation

from the conidiophores.

Ascigeious stage : unknown.

Drechsler named this species leucostylum because of the characteristic colourless

or light-fuliginous conidiophores. According to his description and figures, the

conidiophores are as light at the base as at the apex. Drechsler’s figures show the

conidiophores to be of even thickness throughout, and rounded at the top. The

conidiophores in South African material differed somewhat in being subhyaline at

362

the apex but deeper in colour towards the base. The most typical conidiophores were

fairly light fuliginous, but others (e.g. in No. 23689) were not much lighter in colour

than the olive-brown conidia.

Mitra & Mehta (39) described an isolate from Eleusine coracana which differed

from the type in having slightly larger, olive-brown conidia and conidiophores which

were “ conspicuously slender at the basal part, gradually becoming broader at the

top which is flat or anvil-shaped These features are not mentioned by Drechsler,

but they agree well with what was seen in South African material.

There is a noticeable variation in the conidiophores in recorded isolations of

H. leucostylum, but conidial features have remained characteristic. Taking this into

consideration, and despite the fact that H. leucostylum was named for the lack of

coloration in the conidiophores, the South African material is too typical to be assigned

to any other species.

Specimens examined: 23689, P. Cazelet, on Eleusine indica, White River, Tvl.,

Jan. 1929; 25908, L. Doidge, on Eleusine indica, Ladysmith, Natal, March 1931;

10036, L. C. Turner, on E. indica, Vereeniging, Feb. 1917.

Pathological notes. Drechsler (loc. cit.) in describing H. leucostylum from Eleusine

indica, also found it common on Eragrostis major in the U.S.A. It has been reported,

too, on E. coracana in India and Uganda, and on E. aegyptiaca in India. It was found

in India (39) to be capable of infecting maize, sorghum, Panicum frumentaceum and

Pennisetum typhoideum. On E. coracana and E. aegyptiaca it is stated to cause leaf-blight

or leaf-spot, seedling-blight and head-blight.

In South Africa this species was recorded on Eleusine indica by Smith & Putterill (52),

who stated that diseased plants were “ characterised by a black-dusted withering of

the extremities Tips of inflorescence branches showed empty, discoloured spikelets;

tips of leaves withered and tended to split lengthwise, while the lower parts were green

and apparently healthy. Black, dusty stripes parallel to the veins on the brown,

withered parts resulted from abundant spore formation. Wind-blown conidia may

also cause a brcwn-spotting of the lower parts of the older green leaves. Seedling

plants may be killed by this fungus.

Drechsler described similar symptoms on Eleusine indica, though their significance

was obscured by a considerable overgrowth of other fungi.

(11) Helminthosporium cynodontis Marignoni in Micromiceti di Schio (1909);

Saccardo Syll. Fung. 22 (1913) 394.

Illustrations: Fig. 11.

Conidia usually long-elliptical, slightly curved, often straight, widest near the middle

and tapering gradually to the rounded ends; wall fairly thin; hilum included in the

basal contour; subhyaline to light fuliginous; 20-83 x 8-15 p, with 3-11 septa.

Germination bipolar. Conidiophores simple, dark brown, 60-200 x 4-6-5 p, arising

from stomata or between epidermal cells singly or in pairs.

Ascigerous stage; unknown.

The South African material agrees well with that from America dealt with by

Drechsler (14).

Specimens examined: 27359, R. Gardner, on Cynodon dactylon, Pilgrims Rest,

Feb. 1934; 29976, B. Wilson, on C. bradleyi. Redan, March 1938; 30242, A. M.

Bottomley, on C. bradleyi, Pretoria, March 1939.

Pathological notes: The type of H. cynodontis occurred on dry leaves of Cynodon

dactylon in Italy. The species has also been recorded by Drechsler (14) in the U.S.A.

on Cynodon dactylon and Eleusine indica, on C. dactylon in Kenya (41) and on

C. dactylon in South Africa (51). The author has found it to be fairly common on

C. dactylon and C. bradleyi in the Transvaal, and on C. dactylon in the Eastern Cape

363

Province. Drechsler also regarded this fungus as common, so that lack of records

probably indicates a comparative unimportance of the hosts rather than rarity of the

fungus.

The author has found this fungus on dark brown lesions on living leaves and stems,

but it is most plentiful on dying or withered parts. In wet seasons, during February

and March it was abundant on lawns turning brown and in brown patches on lawns.

Species of Curvularia , Rhizoctonia and also Helminthosporium dematioideum have

sometimes been found on the same plants with H. cynodontis. Although the relationship

of the last-named to the host has not been determined it appears to be often secondary

to some other agent (e.g. nematodes or Rhizoctonia) only becoming conspicuous in

unusually wet weather.

(12) Helminthosporium brizae Nisikado in Berichte Ohara Inst, fur Landw. Forsch.

4 (1929) 122.

Illustrations : Fig. 12.

Conidia subcylindrical or long-elliptical, straight, only exceptionally curved,

widest near the middle, tapering gently towards the abruptly rounded ends; hilum

large, conspicuous, included in the basal contour; walls fairly thin, thinner over the

ends; subhyaline or light fuliginous at first, later becoming yellowish or yellow-brown,

lighter in the basal or both end cells; 26-75 x 9 ■ 5-15 p, with 2-7 septa. Conidiophores

simple, light to dark brown or olivaceous, lighter above; 40-150 x 4 -5-7 -5 p above

the bulbous basal cell; emerging from between epidermal cells singly or in groups of

2-3.

Ascigerous stage: unknown.

In Nisikado’s description of H. brizae, some discrepancies are evident. Although

he described the conidia as “ usually sub-cylindrical, but widest at a point near the

middle and tapering towards both ends ”, his illustrations are of conidia mostly widest

below the middle while tapering decidedly more towards the apex than the base.

Because of the lack of material for comparison, the writer decided to accept Nisikado’s

description rather than his figures.

The South African material differed from the description in only two respects.

Firstly, Nisikado made no mention cf a rather noticeable hilum. Then, the conidia

in the author’s material had a shorter average length (40 p instead of 55 p) than the

Japanese material, although the range in conidial length was the same in both instances.

Conidia were not plentiful in the South African material though abundant

conidiophores were present. It seems possible, therefore, that most of the riper, heavier

spores had been lost, leaving behind an unrepresentative spore population. Despite

these differences there is a good general agreement between the South African material

and Nisikado’s description of H. brizae.

Specimens examined: 28530, K. M. Putterill, on Briza maxima. Port Elizabeth,

Oct. 1935.

Pathological notes: H. brizae was described from the leaves of Briza minor in

Japan. On this host it was found to cause rectangular or elliptical lesions at first

2-3 mm. x 1-2 mm. with a scaled appearance, later bleached and up to

10-20 mm. x 3-5 mm. in size, the leaf at last shrivelling up. The host in South Africa,

Briza maxima, showed small purplish-black spots on the leaf blades and sheaths. The

larger oblong or elliptical lesions, 1-5 mm. x 0-5-2 mm., each had a minute bleached

spot in the centre. There is thus a difference in the symptoms cn Brize maxima in

South Africa and B. minor in Japan. Whether this difference is significant is unknowm

as the host and conditions may determine the appearance of the disease as much as

the parasite aoes.

364

(13) Helminthosporium sacchari (Breda de Haan) Butler in Mem. Dept. Agric.

India Bot. ser. 6 (1913) 204-208, PI. 6; Saccardo Syll. Fung. 25 (1931) 823.

Cercospora sacchari Breda de Haan in Meded. Proefsta. Suikerriet West-Java

3 (1892) 15-21.

Helminthosporium ocellum Faris in Phytopathology 18 (1928) 753-773.

Illustrations : Fig. 13.

Conidia long-elliptical or elliptic-fusiform, with a slight or marked crescentic

curve, or sometimes straight, widest at the middle tapering slightly to each rounded

end; walls thin; hilum included in the basal contour; yellowish or sub-hyaline with

a faint smoky greenish to smoky olivaceous-brown tint; (33)— 50— 105— (125) X

1 1 - 5— 1 7— (1 8) p, with 3-11 septa. Germination bipolar. Conidiophores, medium smoky

olive-brown colour, lighter at the tips, frequently swollen at the base, with geniculate

fertile regions, showing distinct conidial scars; 100-260 X 4-10 p; arising from brown

mycelium on the surface of the leaf or below it.

Ascigerous stage: unknown.

The causal organism of eye-spot disease of sugarcane was first described (3) as

Cercospora sacchari under which name it was later recorded from many other

sugarcane-growing areas. Butler (4) noted that the fungus was better classified in

Helminthosporium, and made the transfer. Van der Byl (59) recorded Cercospora

sacchari from Natal, but also noted that it was more properly a Helminthosporium.

Faris (20) claimed that in Cuba the eye-spot fungus of de Haan and others differed

from Butler’s H. sacchari, and distinguished the former as Helminthosporium ocellum

n. sp. Mitra (37) showed that “ H. sacchari Butler and its saltants have a range of

variation in spore dimensions sufficiently wide to include all the forms of

Helminthosporium causing eye-spot of sugarcane ”. The same may be said of the

range of symptoms on different varieties of cane under different conditions. McRae

(34, 35) in reporting Mitra’s work, stated, “ The conclusion drawn so far is that

H. sacchari Butler and H. ocellum Faris are identical, though they are different strains,

and the name H. sacchari Butler with its amplified description (Mitra 36, p. 290) should

be maintained for the eye-spot disease in Cuba.” This is now generally accepted,

and H. sacchari, like H. sativum, is recognised as a species composed of many different

strains.

The conidia of the South African strains are large and resemble those of H. ocellum

as described by Faris.

Specimens examined: 30245, 30988, A. McMartin, on Saccharum officinarum,

Mt. Edgecombe, July 1939; 32708, on Pennisetum sp., Honolulu, Hawaii.

Pathological notes: Eye-spot disease of sugarcane, caused by H. sacchari, is

recognised by lesions of a dull red colour, usually with a light-brownish centre, formed

on the leaves. Young lesions are often surrounded by a pale halo. These spots often

spread or coalesce into long streaks with or without a definite eye.

The disease occurs in sugar-belts all over the world. In South Africa it has not

been serious. It is most prevalent in cool, damp weather.

The frequency of leaf-spotting and the acuteness of the attack vary with the variety

of sugarcane. An Indian cane, Co. 419, was badly attacked, the fungus causing a

rot of the unfolded leaves and even attacking the stems which dried and became light

in weight.

H. sacchari has not been found on hosts other than sugarcane in South Africa,

but elsewhere has been recorded on Napier grass, Pennisetum sp. It has been found

capable of infecting wheat and barley under experimental conditions.

365

(14) Helminthosporium leersii Atkinson in Bull. Cornell Univ. (Sci.) 3 (1897) 47;

Saccardo Syll. Fung. 14 (1899) 1087.

Illustrations: Fig. 14.

Conidia curved, occasionally straight, widest at or just below the middle, tapering

evenly towards the rounded ends; walls very thin and not constricted at the septa;

hilum inconspicuous and included in the basal contour; light fuliginous; 52-124 x

^13)— 1 6 -5-120) p, with 2-12 septa. Germination bipolar. Conidiophores simple,

thick-walled, rigid; basal cell swollen; fertile region straight or moderately geniculate;

light to medium brown becoming dark brown, lighter near the apex; 130-375 x

6-5-8-5-(10) p; emerging singly from the host epidermis.

Ascigerous stage: unknown.

In Atkinson’s description the conidiophores were characterised as “ amphigenous,

brown, irregularly nodulose or flexuous, 200-350 x 4-6 p Drechsler, who noted the

peculiar growth of H. leersii in culture, illustrated simple, straight conidiophores taken

direct from the host. In South Africa the conidiophores are stout and stiff rather than

“ irregularly nodulose or flexuous ”. Though Drechsler did not mention the dimensions

of the conidiophores, he indicated the size of the conidia as 50-95 x 11-14 p, thus

narrower than the average of 16 • 5 p in the South African isolations. The South African

form is apparently not exactly the same as the American. There seems to be a significant

difference in conidial width, if not in length, for although conidial length may fluctuate

considerably in many species of Helminthosporium, the width usually remains rather

characteristic.

Nisikado’s figures for the conidial measurements of the Japanese form of

H. leersii (45-125 X 12-20 p), agree well with the South African form, but no mention

was made of conidiophores.

Although the South African fungus agrees very well with the type in general

conidial features (shape, colour, unusual thinness of wall) and appears to match

H. leersii, there is at least a possibility that it is a distinct strain. However, it must be

admitted that the comparatively meagre information available on this species might

make it appear more limited than it really is. Until proper comparison of material

should demonstrate constant differences it is not advisable to add confusion with a

new species, and the South African fungus is provisionally referred to H. leersii.

In conidial size and general appearance the author’s material bears a resemblance

to H. sacchari, but the former is recognised by the stiff, dark conidiophores.

Specimens examined: 33115, A. P. D. McClean, on Leersia hexandra, Nottingham

Road, March 1939; 33116, A. P. D. McClean, on Setaria sp., Nottingham Road,

March 1939.

Pathological notes: H. leersii was described from leaves of Leersia virginica in

the United States of America. Drechsler (14) later redescribed it from the same host.

Nisikado (43) recorded it from Japan on Leersia hexandra. In South Africa this fungus

was collected on L. hexandra and Setaria sp. at the same time and place.

The symptoms on the two South African grasses are small indefinite brown spots

1-3 mm. long, elliptical or elongated, which later develop into irregular dark brown

eye-spots 8-10 x 2-3 mm. in size, with a straw coloured or bleached ashen centre.

Lesions occur chiefly on the leaf-blades of both grasses and are less definite in form

on Setaria. Fructifications are found on these spots and also on the withered lower

leaves. The brown and gray eye-spots are exactly like Drechsler’s description of those

on L. virginica.

(15) Helminthosporium urochloae sp. nov.

Illustrations: Fig. 15 A & B.

Conidia usually straight but sometimes with a slight or pronounced crescentic

curve, occasionally slightly sigmoid; widest one third to one half the distance from the

366

base, tapering towards the rounded ends, often more tapered towards the apex; about

half the conidia slightly narrowed near the first or second basal septa, these conidia

having a characteristically narrowed proximal part; wall thick, thinner at the ends;

hilum not projecting; yellow-brown to olive-brown, later opaque; (62)— 90— 1 60— ( 189) X

(13)— 18— (21) /x, with 6-16 septa. Germination bipolar. Conidiophores simple, with

swollen basal cell; fertile region slightly geniculate; dark olive-brown or olivaceous

at the base, becoming lighter towards the tips which are light coloured or subhyaline;

160-300 x 5-10 /x; mostly emerging singly.

Ascigerous stage: unknown.

Helminthosporium urochloae sp. nov.

Conidia: flavo-fusca vel olivacea, deinde opaca, (62)-90-160-(189) x (13)-

18- (21) /x, 6-16 septatis, fere recta sed interdum leviter vel manifeste curvata,

nonnunquam leviter sigmoidea (crassitudo maxima \ supra basin), plerumque

attenuata (praceipue ad apicem) et rotundata; conidia dimidia pars leviter ccntracta

ad septis basalis primae et secundae, istae septae pars proxima proprie attenuata;

membrana crassa, ad extremitatis tenuioris; hilum non eminens. Germinatio bipolaris.

Conidiophora: simplices, cellulis basalis incrassatis; regio fertilis subgeniculata, basin

versus obscure olivacea, apicem versus pallidiora, apices pallidis vel subhyalinis, fere

singulariter emergens. Typus No. 26148.

Some features of H. urochloae suggest comparison with H. oryzae B. de H. as

described by Drechsler (14), which it resembles in the general shape, colour and size

of the conidia. H. urochloae differs from the latter species in having a large proportion

of conidia with characteristically narrowed basal ends ana in the absence of a mycelial

crust anywhere on the host. Its conidiophores are somewhat wider than those of

H. oryzae and in the latter branched conidiophores also occur. H. urochloae is easily

distinguished by its conidial features from other Helminthosporia on grasses.

Specimens examined: 26148, V. A. Wager, on Urochloa helopus, Barberton,

March 1932; 26182, L. C. C. Liebenberg, on Urochloa helopus, Nelspruit, March 1932.

Pathological nctes: H. urochloae was found on Urochloa helopus in the Transvaal.

No inoculations were made but it is suspected that the fungus was parasitic and

responsible for an eye-spot disease. The symptoms were very numerous brown to

black lesions varying in size from pinpoints to irregular blotches about 2x6 mm. or

larger, on the leaf blades. Some remain as ill-defined brown stains, but most develop

into irregular eye-spots each with a small straw-coloured or dirty white centre,

surrounded by a very dark zone. This peripheral zone is lighter towards the outside

and may fade off into the surrounding tissue of the leaf. The fungus was plentiful on

ail affected parts but especially on the lower withered leaves. No other possible parasites

were found..

(16) Helminthosporium ravenelii Curtis in Amer. Journ. Sci. ser. ii, 6 1 848) 349-353 ;

Saccardo Syll. Fung. 4 (1886) 412.

Helminthosporium hojfmanni Berk. & Curt., Intro, to Crypt. Bot. (1857) 298.

Helminthosporium tonkinense Karst. & Roum. in Rev. Myc. 12 (1890) 78.

Helminthosporium crustaceum P. Henn. in Hedwigia 41 (1902) 147.

Illustrations'. Fig. 16.

Conidia straight or with a slight crescentic or sigmoid curve, widest near the middle,

narrowing slightly to either end, often decidedly narrower towards the base; wall

thin; hilum inconspicuous and included in the basal contour; light smoky-yellow

to smoky brownish-yellow; 25-82 x 12-18 /x with fl)-3-4-(7) septa. Germination

bipolar. Conidiophores very distinctive, long, much branched, torulose; light fuscous

yellow; 550 x 5-1 1-5 /x; arising close together from a velvety basal layer of interwoven

j yphae encrusting the affected parts of the host.

367

Ascigerous stage: unknown.

H. ravenelii differs from other species in the unusual conidiophores rather than in

conidial characters. These torulose, much branched conidiophores, crowded into a

compact crust, are very distinct from the usual simpler, separate conidiophores of

many other species. Only one other species, H. miyakei, studied by the author displayed

similar features.

Specimens examined:

On Sporobolus capensis : 1453, I. B. Pole Evans, Harden Heights, 11/4/11;

1826, E. M. Doidge, Zoutpansberg, 14/8/11; 2201, T. Parkins, Eshowe, 5/4/12;

6692, E. Baker, Cedara, 15/5/13; 6921, A. Pegler, Kentani, 26/5/13; 9764, I. B.

Pole Evans, Tweedie, 4/7/16; 9769, P. v. d. By/, Illovo, Natal; 10065, A. O. D.

Mogg, Tweedie, 13/4/17; 10097, P. J. Pienaar, Nelspruit, 17/4/17; 11643,

A. O. D. Mogg, Tweedie; 20371, C. N. Campbell, Mooi River, Apr. 1925; 31783,

H. Power, Illovo, 20/5/16.

On Sporobolus fimbriatus : 6667, A. Pegler, Kentani, 14/5/13; 23390, A. O. D.

Mogg, Balgowan, 3/4/19.

On Sporobolus pyramidalis: 13995, F. Eyles, Salisbury, 20/2/19; 26093,

L. C. C. Eiebenberg, Lydenburg, Apr. 1931; 29835, E. M. Doidge, Donnybrook,

5/9/37.

Pathological notes : Helminthosporium ravenelii is the cause of “ false smut ”

of Sporobolus spp. in most warm countries where it is widely distributed. In South

Africa the commonest host is S. capensis. As the popular name of the disease indicates,

the inflorescences of the grass are attacked. It appears at first as a velvety, olive or

brown coating on the floral parts, and later turns into a dense black crust which often

cakes the separate parts together.

This fungus may not be as economically unimportant as is usually stated.

Investigators have proved that ravenelin, a metabolic product of the parasite (46)

causes cattle poisoning, of which false smut of Sporobolus has been suspected in South

Africa and elsewhere (32).

(17) Helminthosporium miyakei Nisikado in Berichte Ohara Inst, fiir Landw.

Forsch. 4 (1929) 122.

Illustrations : Fig. 17.

Conidia bent, or with a crescentic or irregular sigmoid curve, sometimes straight,

widest just below the middle, tapering suddenly and often irregularly to a narrow

apex, narrowing gradually towards the abruptly rounded base; basal cell often out

of alignment; immature conidia mostly without a tapering apex, obovate long-elliptic

or irregular; wall thin; hilum included in the basal contour; light yellow-fuliginous;

25-113 x 11-5-18 p, with (2)— 3— 5— (8) septa. Condiophores torulose, profusely

branched, arising close together from a mat of interwoven hyphae ; up to 260 x 4-1 1 • 5 /x.

Ascigerous stage : unknown.

H. miyakei approaches the genus Cercospora in the shape of its conidia, but in

every feature except conidial shape it is closely allied to Helminthosporium ravenelii.

The genus Helminthosporium seems the most suitable for its disposal at present.

Although the conidial shape in H. miyakei differs from that of the other species discussed

in this paper, it does find a parallel in some of the conidia of H. rostratum. Many

conidia of the latter however, do not taper, while the tapering of others is never as

marked as in H. miyakei.

With only the conidial stage known, it seems unlikely that H. miyakei will fall

into either of the two main groups Cylindro- and Eu-Helminthosporium. Further work

may reveal H. miyakei and H. ravenelii as members of another group possessing conidia

of the Helminthosporium type.

368

Specimens examined: 7761, T. R. Sim, on Eragrostis curvula, Pietermaritzburg,

26/4/14.

Pathological notes: H. miyakei forms an olivaceous, later black, crustose growth

on the inflorescences of its graminaceous hosts, much like H. ravenelii. It has been

found only once in South Africa but has been recorded as common on Eragrostis

tef in Western Ethiopia (5). Nisikado described his type from Eragrostis pilosa in

Japan. No other records of this interesting and distinctive species have been traced.

(18) Helminthosporium dematioideum Bubak & Wroblewski in Hedwigia 57 (1916)

337; Saccardo Syll. Fung. 25 (1931) 821.

Illustrations: Fig. 18.

Conidia straight, oblong-cylindrical, widest near the middle, hardly tapering

towards the rounded apex but narrowing decidedly towards the rounded or sub-conical

base; walls thin, sometimes somewhat constricted near the basal septum; hilum

included in the basal contour; light yellowish-brown; (20)-30-36-(43) x (6)— 8—10—

(13) p, with 3-5-(7) septa. Germination basal. Conidiophores: Simple, septate, usually

swollen at the base; medium brown, paler at the tips; emerging singly or in small

groups from the host epidermis; 50-100 x (3)— 4 — 6— (7) p.

Ascigerous stage: unknown.

In his discussion of the American isolates, Drechsler (14) states, “ the principal

circumstance suggesting a possible lack of identity of the European and American

forms, is the recorded occurrence of the former only on the inflorescence of

Anthoxanthum odoratum while the latter is much more abundant on the foliage “. He

suggests that this might be due to a poor original specimen. From the writer’s

experience it would seem that Drechsler attached too much importance to this slight

possible difference in substratum. The South African form fits the description of

H. dematioideum so well that it must be assigned to this species despite its occurrence

on hosts different from the type. The strain from wheat was morphologically

indistinguishable from the rest. The identity of Nos. 30241 and 30498 was confirmed

by Dr. Westerdijk.

In culture it was found that there was a tendency towards greater variation in

conidial size (8-50 p), conidial colour (darker and more olivaceous) and longer

conidiophores. Such variations were also found in other species of Helminthosporium.

It is notable that the mode of germination of conidia in H. dematioideum is not

typical of either of the two main classes in which most graminicolous Helminthosporia

fall, namely those connected with Pyrenophora or Cochliobolus perfect stages.

Specimens examined: 27358, Naude, on Cynodon dactylon, Pretoria; 27359,

R. Gardner, on Cynodon dactylon. Pilgrims Rest, 8/2/43 ; 28779, v. d. Merwe, on Cynodon

bradleyi, Pretoria, 18/12/36; 30241, K. M. Putterill, on Cynodon bradleyi, Pretoria,

20/2/39; 30243, K. M. Putterill, on Cynodon transvaalensis, Pretoria, 30/4/39; 30250,

J. Roderick, on Cynodon transvaalensis, Johannesburg, 15/4/39; 30492, J. Sellschop,

on Triticum, Ferreira, O.F.S., Nov. 1938.

Pathological notes: H. dematioideum was described from Galicia on Anthoxanthum

odoratum-, it was later found by Drechsler (who amplified the description, 14, p. 683)

on the same host in the United States of America. Drechsler also found it on

Agrostis spp. often accompanied by fructifications of larger-spored species of

Helminthosporium.

In South Africa this fungus was found to be common on withering or dead leaves

of Cynodon transvaalensis, C. bradleyi and C. dactylon. On Cynodon lawns it was

frequently found mixed with Helminthosporium cynodontis, species of Curvularia, and

occasionally with Rhizoctonia. These fungi are known to be responsible for causing

browning of lawns in summer, thus it is not known what part was played by

369

H. demcitioideum in producing this condition. Drechsler was similarly unable to decide,

from the specimens found, whether H. dematioideum was present as a parasite or

saprophyte, but he seemed to believe that it was more saprophytic on Agrostis and

possibly parasitic on Anthoxanthum. The relationship of H. dematioideum to its

Cynodon hosts can be established only by inoculation experiments. At present it may

be said that the circumstances under which it appears suggest that it may be saprophytic,

or a weak parasite capable of attacking plants already weakened by other fungi. This

view is supported by the fact that this fungus was unexpectedly isolated from

foot-rotted wheat, in company with a number of foot-rotting fungi.

Although there are not many records of this fungus, the writer’s experience suggests

that it is common.

Species of Curvularia found in South Africa.

The genus Curvularia was erected by Boedijn (2) to accommodate a number of

small-spored species that had unnaturally been placed in Helminthosporium,

Brachysporium, Brachycladium and Acrothedum.

Species of Curvularia are characterised by unequally-sided or strongly bent conidia,

with 3^1 septa. At least one of the central cells is larger and darker than the others.

Most species form branched cylindrical stromata in pure culture. These have a black

rind, often with white growing points, and are white inside. No perfect stage has

yet been found for any Curvularia species.

Boedijn divides the genus into three groups:—

(A) Maculans group: characterised by straight or merely asymmetrical,

3-septate conidia, with the two middle cells larger and darker than the end cells.

Stromata are never formed in culture.

(B) Lunata group: characterised by 3-septate conidia, more or less strongly

bent, with only one of the middle cells enlarged and darker than the end cells.

Most species form stromata.

(C) Geniculata group: characterised by 4-septate conidia and forming

stromata in culture.

Boedijn noted that the species of Curvularia often consisted of several strains

showing minor morphological differences. This is well illustrated in the South African

isolations of C. spidfera in which a strain from Cynodon has very much larger conidia

than one from wheat. Host differences cannot be regarded as important among these

fungi, which may possibly live in the soil.

The species of Curvularia described here belong to the Maculans and Lunata

groups, but there are doubtless many other species in South Africa.

(19) Curvularia spicifera (Bainier) Boedijn in Bull. Jard. Bet. Buitenzorg, ser. iii,

13 (1933) 81.

Brachycladium spiciferum Bainier in Bull. Soc. Myc. de Fr. 24 (1908) 81.

Helminthosporium tetramera McKinney in U.S.A. Dept. Agr. Bull. 1347 (1925).

Illustrations: Fig. 19.

Conidia oblong or ellipsoid, straight, with broadly rounded ends; walls fairly thick,

thinner towards the ends ; hilum protruding slightly outside the basal contour ; greenish-

fuliginous or brownish, lighter over the ends; (1 1)— 25— 35— (43) x (7)-9-ll-(13) p,

with 3 septa, occasionally 2 or 4 septa. Conidiophores erect, of variable length; fertile

region long and geniculate, olivaceous or brownish, 5-7 p diam. Conidia borne freely

in long clusters.

Ascigerous stage: unknown.

370

Boedijn placed C. spicifera in the Maculans group. Hynes (25) found that

McKinney’s (33) Helminthosporium tetramera was synonymous with C. spicifera.

However, McKinney reported that his fungus formed stromata in culture, a phenomenon

described by Bqedijn as specifically non-existent in the Maculans group.

In the writer’s material of C. spicifera the central cells of the conidia were not as

noticeably darker and larger than the end cells as in other material she saw, or as figured

by Boedijn. The terminal cells were slightly paler where the wall thinned out over

the end cells. Bainier and McKinney did not mention this feature, while Hynes described

his conidia as “ of. a uniform straw brown colour

Apart from this difference in colour, the strain from Cynodon yielded much larger

and darker conidia in culture than in nature. In culture the length of the conidia

might be up to 54 p, and the diameter up to 14-5 p. This strain was indistinguishable

from that on wheat except for conidial size. A culture sent to the Imperial Mycological

Institute was identified as “ C. spicifera except in spore measurements ”.

Although the South African material differs from the description of the type in

some respects, there is a good general agreement between the two. Differences observed

and recorded by various workers indicate that this species is composed of a number

of strains of varied character.

Specimens examined : 30246, J. Roderick, on Cynodon transvaalensis, Johannesburg,

Apt. 1939; 30494, J. Sellschop, on Triticum sp., Bethlehem, Nov. 1938.

Pathological notes: Curvu/aria spicifera has been isolated from foot-rotted parts

of wheat, oats, barley, rye and rice. Its recorded distribution is The United States of

America (8), India (57), Australia (25), Angola (1), and South Africa. The writer has

also found it on withering leaves and stems of Cynodon transvaalensis in South Africa,

occurring among brown patches on lawns of this grass.

C. spicifera has so far only been found by the writer in association with other fungi

known to be capable of causing foot-rot. Its relationship to its host and its associates

is unknown. Neither McKinney nor Hynes regarded it as an important pathogen.

(20) Curvularia species.

Illustration : Fig. 20.

Conidia straight or unequal-sided to slightly curved; broadly rounded at the apex

and tapering towards the base; base somewhat triangular, squared off in the

well-marked hilum; yellowish-brown to brown; 3-septate, with the third cell larger

and darker than the others, and with thicker walls; most strongly curved in the third

cell; (15)-24-30-(36) X (8)— 10— 14 — (16) p. Conidiophores brownish, very variable

in length in culture, 3-6/x diam. in the fertile region. Conidia borne spirally on the

geniculate protuberances.

Ascigerous stage: unknown.

This species apparently belongs to the Lunata group, with its curved, 3-septate

conidia having the third cell from the base larger than the others. But it differs from

the Lunata group in that the third cell is not markedly darker than the end cells, and the

conidia are seldom strongly bent, often being straight.

A specimen of this fungus was sent to Mr. E. W. Mason, who commented as

follows: “It strongly resembles Brachycladium ramosum as figured by Bainier [Bull.

Soc. Myc. de Fr. 24 (1908) PI. 4], the only differences being that Bainier said his fungus

formed elongated black stromata, covered with conidiophores, and that the conidia

were 3 or 4 septate. However, Dr. Boedijn has applied the name Curvularia ramosa

(Bain.) Boedijn to a distinctly more variable form isolated from wheat in Australia,

with longer conidia, 2-5 septate, and with one or more of the central cells often distinctly

darker than the two end cells. I would prefer to refer this to Curvularia lunata group,

at any rate until more is known about the South African species ”.

371

The writer has observed white, branched “ roots ” of fungous tissue growing down

into the agar in month-old tube cultures of this species. These structures were probably

an arrested stromatic development. Such positively geotropic stromata were seen by

Boedijn in cultures of the Geniculata group. Especially in old cultures, the conidia

sometimes became 4-septate. These septa, which were thinner than the rest and divided

the large third cell, were obviously late additions. The addition of further septa in the

conidia was one of the variations observed by Boedijn, who found it particularly in the

Lunata group.

From the above evidence it seems best to follow Mr. Mason’s advice and place this

species provisionally in the Lunata group, until more is known about it.

Specimens examined: 30495, J. Sellschop, on Triticum, Bethlehem, Nov. 1938.

Pathological notes: This species was isolated from foot-rotted wheat plants in

association with other fungi known to be capable of causing foot-rot. No inoculation

tests were made. It is possible that the fungus was a secondary invader.

LITERATURE CITED.

(1) BENSAUDE, M :

(2) Boedijn, K. B

(3) Breda de Haan, J. van

(4) Butler, E. J. & A. Hafiz Khan . .

(5) Castellani, E. & Ciccarone, A.. .

(6) Christensen, J. J

(7) Clements, F. E. & Shear, C. L. . .

(8) Crosier, W. & Weimer, D

(9) da Camara, E. de S

(10) Diedicke, H

(11) Diedicke, H

(12) Doidge, E. M. & Bottomley, A. M.

(13) Dosdall, L

(14) Drechsler, C

(15) Drechsler, C

(16) Drechsler, C

(17) Drechsler, C

(18) Drechsler, C

(19) Drechsler, C

(19a) Drechsler, C

(20) Faris, J. A

(21) Hansford, C. G

(22) Henry, A. W

(23) Hopkins, J. C

(24) Hynes, H. J

Comptes Rendus Soc. de Biol. 103 (1930) 1265-1266.

Bull. Jard. Bot. Buitenzorg, ser. iii, 13 (1933) 120-134.

Meded. Proefsta. Suikerriet, West-Java 3 (1892) 15-21.

Mem. Dept. Agr. India, Bot. ser. IV (1913) 204-208.

Reg. 1st. agron. per l’Africa italiana (1939) 32 pp. (Seen

Abstract in Rev. Appl. Myc. XIX: 85.)

Univ. Minn. Agric. Expt. Sta. Tech. Bull. 37 (1926).

The Genera of Fungi (1931).

Proc. Ass. Off. Seed Anal. N. Amer., 1939. (1940) 120-124.

(Seen Abstract in Rev. Appl. Myc. XIX: 601.)

Rev. Agron 20 (1932) 5-63. (Seen Abstract in Rev. Appl.

Myc. XII: 725.)

Centralbl. Bakt. Abt. 11, Bd. 9 (1902) 317-329.

Centralbl. Bakt. Abt. 11, Bd. 11 (1903) 52-59.

Bot. Survey of S. Africa, Mem. No. 11 (1931).

Univ. Minn. Agr. Expt. Sta. Tech. Bull. 17 (1923).

Journ. Agric. Res. 24 (1923) 641-740.

Journ. Agric. Res. 31 (1925) 701-726.

Phytopathology 17 (1927) 414.

Journ. Agric. Res. 37 (1928) 473-492.

Journ. Agric. Res. 40 (1930) 447.

Phytopathology, 24 (1934) 953-983.

Phytopathology 25 (1935) 334-361.

Phytopathology 18 (1928) 753-754.

Imp. Myc. Inst. Mycol. Pap. No. 15 (1946) 213-214.

Univ. Minn. Agr. Expt. Sta. Tech. Bull. 22 (1924).

Rept. Plant Path, for 1929. Rept. of Secy., Dept. Agric.

S. Rhodesia for 1929, (1930) 84-86.

Sci. Bull. Dept. Agric. N.S.W. 47 (1935).

372

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

(50)

(51)

(52)

(53)

(54)

(55)

(56)

(57)

(58)

(59)

(60)

Hynes, H. J

Hynes, H. H

Ito, S

Ito, S. & Kuribayashi, K. .

Ito, S. & Kuribayashi, K.

Kuribayashi, K

Marchionatto, J. B

McKinney, H. H

McRae, W

Journ. Roy. Soc. N.S.W. 70 (1937) 378-391.

Jour. Aust. Inst. Agric. Sci. iii, 4 (1937) 212-218.

Sci. Bull. Dept. Agric. N.S.W. 61 (1938).

Proc. Imper. Acad. Tokyo 6 (1930) 352-355. (Seen Abstract

in Rev. Appl. Myc. X: 232.)

Ann. Phytopath. Soc. Japan 11 (1927) 1-8. (Seen Abstract

in Rev. Appl. Myc. VII : 54.)

Journ. Fac. Agric. Hokkaido Imper. Univ. Sapporo, Japan

29 (1931) 85-125. (Seen Abstract in Rev. Appl. Myc.

X: 514.)

Trans. Sapporo Nat. Hist. Soc. 10 (1929) 138-145, 162.

(Seen Abstract in Rev. Appl. Myc. VIII: 438.)

Bol. Min. Agric. Nac., Buenos Aires 29 (1930) 457^162.

U.S. Dept. Agric. Bull. 1347 (1925).

Scient. Repts. Imper. Inst. Agric. Res. Pusa (1931-2)

122-140.

Scient. Repts. Imper. Inst. Agric. Res. Pusa (1932-3)

134-160.

Mitra, M

Mitra, M. & Bose, R. D

Mitra, M. & Mehta, P. R

Mourashkinsky, K. E

Nattrass, R. M

Nisikado, Y

Nisikado, Y. & Miyake, C

O’Gara, P. J

Raistrick, H j,

Ravn, F

Saccardo, P. A '

Smith, N. J. G

Smith, N. J. G. & Putterill, K. M.

Smith, N. J. G. & Rattray, J. M.

Stevens, F. L

Subramaniam, L. S

Sydow, H., Mitter, J. H. &

Tandon, R. N.

Thomas, K. M

Tullis, E. C

van der Byl, P. A

Wakefield, E. M. & Moore, W. C.

Trans. Brit. Myc. Soc. 15 (1930) 254-293.

Trans. Brit. Myc. Soc. 16 (1931) 115-127.

Indian Journ. Agric. Sci. 5 (1935) 449-484.

Indian Journ. Agric. Sci. 4 (1934) 943-975.

(Abstract seen in Rev. Appl. Myc. XIX: 391.)

Rept. Dept. Agric. Kenya for 1938 (1939) 42-47. (Seen

Abstract in Rev. Appl. Myc. XIX: 71.)

Japan. Jour. Bot. 4 (1928) 99-112. (Seen Abstract in Rev.

Appl. Myc. VII: 519.)

Ber. Ohara Inst, fur Landw. Forsch. 4 (1929) 111-126.

Ber. Ohara Inst, fur Landw. Forsch. 3 (1926) 221-266.

Phytopathology, 5 (1915) 323-326.

Rept. London. Sch. Hyg. 1934-5, (1935) 27-30. (Seen

Abstract in Rev. Appl. Myc. XV: 170.)

Bot. Tidsskrif Bd. 23 (1900) 101-322.

Zeits. Pflanzenkr. 11 (1904) 1.

Sylloge Fungorum, 25 volumes (1882-1931).

Ann. Appl. Biol. 16 (1929) 236-260; Ibid. 17 (1930).

S. Afr. Journ. Sci. 29 (1932) 286-295.

S. Afr. Journ. Sci. 30 (1933) 198-205.

S. Afr. Journ. Sci. 27 (1930) 341-351.

U.S.A. Dept, of Reg. & Educ. Div. Nat. Hist. Survey Bull.

14 (1922) 77-185.

Indian Journ. Agric. Sci. 6 (1936) 11-16.

Ann. Myc. 35 (1937) 239.

Rept. Govt. Mycologist, Madras, 1938-39 (1939). (Seen

Abstract in Rev. Appl. Myc. XIX: 258.)

Journ. Agric. Res. 47 (1933) 675.

Union of S. Afr. Dept. Agric. Sci. Bull. 10 (1918).

Trans. Brit. Myc. Soc. 20 (1936) 97-109.

373

FIG.3.A

FIG.3.B

Fig. 1. — H. teres. Fig. 2. — H. gramineum. Fig. 3a, 3b. — H. mediocre.

Fig. 4a, 4b. — H. sativum. Fig. 5. — Helminthosporium N. Fig. 6. — H. bicolor.

375

Fig. 7. — H. halodes. Fig. 8. — H. rostratum. Fig. 9. — H. turcicum.

Fig. 10. — H. leucostylum.

9102-5

•OSmm.

376

Fig. 11. — H. cynodoniis. Fig. 12. — H. brizae. Fig. 13. — H. sacchari.

Fig. 14. — H. leersii.

377

378

Fig. 18. — H. dematioideum. Fig. 19. — Curvularia spicifera.

Fig. 20. — Curvularia species.

379

Concerning some South African Pestalotiopsis

Steyaert.

( Pestalotia Auct. non de Not.)

By

R. L. Steyaert.

(Institut National pour VEtude Agronomique du Congo Beige.)

In an attempt to study type specimens of Pestalotiopsis Steyaert ( Pestalotia Auct*

non de Not.), I have been privileged to borrow type material of some South African

species. Unfortunately some of this type material, of which only fragments were

available to my study, was insufficient; it was either too poor in conidia or bore none

or only broken ones.

However, I have been able to gather sufficient data to substantiate some funda-

mental modifications in the list of South African species.

As the study of this batch of material progressed it soon became apparent to me

that one and the same species had been considered different according to the host on

which it occurred.

It is common knowledge that mycologists, specially those dealing with phytocolous

species, have always been impressed by the “ Host-species ” relationship. With this

host specificity in mind many species unrecorded on a host have been considered as

new. Considering the enormous difficulties involved in the determination of species

for such extensive genera as Pestalotiopsis it is a point of view that has been profitable

for practical purposes, mainly because specimens have been described and published

and are now on record and available. It is my opinion that authors should not be

extensively criticised on that account. In the absence of a thorough and well-made

revision it is a situation that is inevitable, but as the systematics of Mycology progresses

it is to be remedied. The species concept with saprophytic or semi-saprophytic fungi

such as Pestalotiopsis should be based on a much broader view. One is lead to this

conclusion when a great number of specimens can be examined. Dimensions may be

misleading and I do not uphold that I have never been misled myself. Morphological

characters are of course the main discriminating factors and on this basis it is apparent

that within the South African species many synonyms have been published. The

ones that were offered to my study are a typical example of this contention.

The following species were published or redescribed in Bothalia 4: 821-831:

1948 by E. M. Laughton: Pestalotia burchel/iae Laughton, P. caffra Sydow, P. laugh-

tonae Doidge, P. encephalartos Laughton, P. milletiae Laughton, P. pelargonii Laughton,

P. podocarpi Laughton non Dennis, P. rapaneae Laughton non Viegas.

The conidia of all these species are claviform, with the two upper, coloured cells

a deep, opaque, fuliginous brown and with an opaqueness more pronounced round

the septum separating the cells. The setulas are relatively thick, usually with a clearly

visible if not wide lumen. The dimensions of the conidia have a range that is given in

Table 1, in which my measurements are compared with those of E. M. Laughton.

Mine are the result of the standard technique I have used throughout my studies on this

genus. About twenty conidia are drawn for each specimen with a camera lucida and

with the same optical equipment. All drawings are thus immediately comparable

and measurable with the same scale in all their details. In Figure 1 are given sample

drawings of approximately the biggest and the smallest conidia of each specimen of

the above-mentioned species.

380

SOf*

381

If for some specimens there are some plus or minus variations the distinction

cannot be, in my opinion, above the varietal level. Even so, I think the variations

are due more to the host difference than to varietal distinctions.

If these measurements are compared with those I have given for Pestalotiopsis

glandicola (Cast.) Steyaert in Bull. Jard. Bot. Etat, Brux. 19: 352-353: 1949, it will

be seen that there is a very definite identity. Also, the same morphological characters

as described above can be observed with this species.

The following synonymy is therefore justified: —

Pestalotiopsis glandicola (Cast.) Steyaert in Bull. Jard. Bot. Etat Brux. 19:

330-334, 352-353: 1949.

Pestalotia caffra Sydow in Ann. Mycol. 12: 266: 1914, fide isotype in Nat.

Herb. S. Afr. Pretoria, E. M. Doidge 6630, Isipingo, Natal, Nov. 1913.

P. burchelliae Laughton in Bothalia 4: 821: 1948, fide F. S. Laughton 34912

holotype in Herb. loc. cit.

P. laughtonae Doidge in loc. cit., p. 824, fide E. M. Laughton 35145 holotype,

35146, 34916 in Herb. loc. cit.

P. encephalartos Laughton in loc. cit., p. 823, fide Malvern 861, in Herb,

loc. cit.

P. milletiat Laughton in loc. cit., p. 825, fide E. M. Laughton 33437 holotype

in Herb. loc. cit.

P. pelargonii Laughton in loc. cit., p. 827, fide E. M. Laughton 34918 holotype

in Herb. loc. cit.

P. podocarpi Laughton non Dennis in loc. cit., p. 827, fide E. M. Laughton

34917 holotype in Herb. loc. cit.

P. rapaneae Laughton non Viegas in loc. cit., p. 829, fide E. M. Doidge 17171

holotype in Herb. loc. cit.

Unfortunately I have been able to retrieve only fragments of conidia, usually

the coloured cells only, from the fragments of the types of P. laurophylli and P. ocoteae,

but what I have observed on these suggests strongly that they also might be considered

as synonyms.

On the specimen Doidge 17169, I have been able to observe the beautiful conidia

of Pestalotia trichocladi Laughton for which I make the new combination: —

Pestalotiopsis trichocladi (Laughton) Steyaert comb. nov.

Pestalotia trichocladi Laughton in Bothalia 4: 829: 1948, fide Doidge 17169

holotype in Nat. Herb. S. Afr., Pretoria.

This species has conidia of remarkable size (see Table 1) specially in regard to

the setulas. Laughton did not record that they are spathulate. My observations show

that they are definitely so, with a wide lumen nearly the whole length. The coloured

cells are concolorous. This feature places the species very close to P. planimi (Vize)

Fig. 1. — Conidia of Pestalotiopsis glandicola (Cast.) Steyaert, from type specimens

of: a-b, Pestalotia butchelliae Laughton {F. S. Laughton, 34912); c-d, P. caffra Sydow

{Doidge, 6630); e-f, P. encephalartos Laughton {Malvern, 861); g-h, P. laughtonae

Doidge {E. M. Laughton, 35145), i-j {Laughton, 34916), k-1, {E. M. Laughton, 35416);

m-n, P. milletiae Laughton {E. M. Laughton, 33437); o-p, P. pelargonii Laughton

{E. M. Laughton, 34918); q-r, P. podocarpi Laughton non Dennis {E. M. Laughton,

34917); s-u, P. rapaneae Laughton non Viegas {Doidge, 17171).

382

Steyaert, which has much smaller setulas and bigger conidia. The drawings of three

conidia of P. trichocladi are given in Fig. 2.

Several other specimens ( P . pterocelastri Laughton in loc. cit., p. 828; P. cassinis

Laughton in loc. cit., p. 822; P. watsoniae Verw\ et Dipp. in S. Afr. Journ. Sci. 27:

327: 1930) gave such scanty or incomplete material that no opinion can be offered

as to the validity of these species. Regarding the type specimen of P. cassinis, there

were individual acervuli in which the conidia bad coloured cells ranging from con-

colorous, and with sizes as given by E. M. Laughton for P. cassinis, to the versicolored

type of conidium of P. glandicola. It may be that immature stages of P. glandicola

have been mistaken for a new species.

With regard to P. pterocelastri, unfortunately no conidia were retrievable from the

part of the specimen sent, but judging by Laughton’s measurements this species is

very close to, if not identical with, P. ilicis (West.) Steyaert.

The specimen of P. watsoniae gave abundant conidia but these were incomplete

and lacked setular ornaments. There is no doubt that this species belongs to the genus

Truncatella Steyaert, but in the absence of setulas I cannot decide on the validity of

the species. Authors have so often described as plurisetulate, species which in reality

possess a single but ramified setula, that I cannot decide without having personally

seen the arrangement of the setae. As far as the coloured cells are concerned, P. wat-

soniae is no different from Truncatella ramulosa (V. Beyma) Steyaert or from

T. conorum-piceae (Tubeuf) Steyaert.

Fig. 2. — Conidia of Pesta/otiopsis trichocladi (Laughton) Steyaert, X 1000. From

type specimen of Pestalotia trichocladi Laughton ( Doidge , 17169).

Table 1.

LONS OF THE CO]

383

385

The Morphology and Anatomy of Utricularia

Transrugosa Stapf.

By

J. Slinger.

INTRODUCTION.

Kamienski, in 1897, recognised five genera of the family LENTIBULARIACEAE.

These are, Pinguicula, Genlisea, Polypompholyx, Utricularia and Biovularia. It is

now known that all these plants are carnivorous and that prey is captured and digested

by means of specialised vegetative organs which constitute the “ traps Among the

LENTIBULARIACEAE occur examples of the simplest traps (Pinguicula), the most

complex of the pitfall type (Genlisea), and the trap or bladder of Utricularia which has

attained a degree of structural complexity and perfection of mechanism for which

there is no analogue among other plants.

The Utricularias show a wide range of variation in form and habit. The plants

may be freely-floating or anchored aquatics, or epiphytic, or they may be terrestrial

in wet to moist sandy soils. Among the aquatic forms are found the larger (e.g.

U. stellaris) and the smaller species of Utricularia (e.g. U. cymbantha ), while the

terrestrial species, with few exceptions are small. The epiphytic forms are often

remarkable for the size and showiness of their flowers.

The genus is of world-wide distribution, the most widely distributed species being

the submerged or semi-submerged aquatics. They are found throughout North

America including Greenland, in Europe and in Asia. Related species extend

throughout the tropics into South America, South Africa, Australia and New Zealand.

Terrestrial species are widely distributed in the tropics of the Old and New worlds.

Fig. 1 shows the distribution of Utricularia in Africa. It will be seen that aquatic

and terrestrial species are well represented, while only one epiphytic species has been

recorded ( U . bryophila). Most common are the terrestrial species, of which U. trans-

rugosa Stapf is an example. It will also be seen from Fig. 1 that six species, all

terrestrial, are limited to Africa south of the central lake area. These species are

U. transrugosa, U. kirkii, U. capensis, U. livida, U. ecklonii and U. sandersonii.

Twenty-three species are restricted to west, central and northern regions, while the

nine remaining species are distributed throughout Africa. Several of the African

species are also found in tropical America ( U. obtusa, U. foliosa, U. subulata ), while

others occur in Algiers and Portugal ( U . exo/eta) and through India to China, Malaya,

and tropical Australia (U. stellaris, U. striatula).

As previously stated, U. transrugosa is one of the small terrestrial species of

Utricularia and is found growing in vleis or boggy ground along stream banks. It will

be seen from Fig. 1 A that the species has been recorded from the Transvaal and also

from Southern Rhodesia near Salisbury, but is not recorded from any locality north

of the Zambesi.

386

Fig. 1. — The Distribution of the Genus Utricularia in Africa.

(Legend for Fig. 1 at foot of page 387).

mX

387

U. transrugosa was first described by O. Stapf in Flora Capensis IV 2, 428 (1904).

His description was based on specimens from Barberton, Johannesburg and Rustenburg.

Rand 727, which was collected in boggy ground near Johannesburg, was one of these

and had previously been misidentified by Moore in Journal of Botany (1903), 405,

as U. sanguined Welw. It is the first published record of U. transrugosa from the

Witwatersrand. Later, in 1951 and 1952, the plant was found at Bryanston, near

Johannesburg, in a semi-dry vlei associatd with sedges, Drosera cf. burkeana and

Lobelia cf. decipiens. It was also found in very wet mud along the bank of a nearby

stream.

Aq. Aquatic including submerged, anchored or surface-floating forms.

Ter. Terrestrial.

Sub-ter. Sub-terrestrial.

Epi. Epiphytic.

388

In Hooker’s leones Plantarum, Tab. 2796 (1903) this plant was described as

U. livida var. transrugosa Stapf. The two species U. transrugosa and U. livida,

(Fig. 1, h) are distinguished by the markings on the palate of the lower corolla lip;

in the former the palate is distinctly rugose, while in the latter it is narrower and

tubercled.

Bladders, stolons, leaves, capsules and seeds were not available to Stapf for his

description of U. transrugosa, and F. E. Lloyd (1942) in “ The Carnivorous Plants ”

does not even refer to U. transrugosa. The object of this report is to give an adequate

description of the anatomy and morphology of U. transrugosa, together with facts

concerning the biology of this species which have become known during these

investigations.

MORPHOLOGY.

(a) VEGETATIVE CHARACTERS.

This is a small perennial plant reaching a height of 6-10 cms. when in flower.

Unless in flower, the plant is likely to be overlooked, as apart from the inflorescence,

only the tips of the leaves appear above ground.

The vegetative morphology of Utricularia cannot be likened to that of any other

flowering plant. In order to avoid confusion, the terms leaf, stolon and rhizoid will

be used in this report, though it must be remembered that in Utricularia the distinctions

between these three are ill-defined. The vegetative portions of the plant are almost

entirely below soil-level, with just the tips of the recurved leaves appearing above

ground. These leaves are simple, entire and spathulate in shape tapering towards

the base into a cylindircal portion which is continuous with a stolon. In its expanded

portion, the leaf is dark green above, paler below, often becoming reddish-brown

with age. Because of the soft, succulent texture of the leaves, the dichotomously-

branched midrib is often indistinct and cannot be seen at all with the naked eye in

the lower cylindrical portion.

The stolons are very delicate, cylindrical, filiform strands of tissue which have a

characteristic white, glistening transluscence. They are much branched, and as will

be seen later, this branching follows a pattern.

U. transrugosa is rootless. The rhizoids described by Stapf are problematical.

When present, they occur in tufts at the base of an inflorescence, but cannot really be

distinguished morphologically from the stolons. Indeed, the vegetative structures of

U. transrugosa are so plastic, that a rhizoid can continue growth, branch and bear

leaves and thus become a stolon. The leaves too, as will be seen from Fig. 2 are no

more than unbranched, expanded stolons.

Minute stalked bladders (traps) are borne in great numbers on leaves, stolons and

rhizoids. There appears to be no definite arrangement of bladders; they may be

opposite or alternate and are not more frequent in one region than in another. As

it is difficult and inpracticable to separate the anatomy from the morphology of these

bladders, this will be discussed collectively in a later section.

Numerous small glands are present over the outer surface of the vegetative parts

of the plant and the epidermis is thickly cuticularised. (See anatomical section for

structure of glands.)

U. transrugosa was frequently found associated with filamentous green algae

(Zygnema, Oedogonium) which had become densely twisted and matted around stolons

and leaves. These, at first glance, give the plant a hairy appearance, and can lead

to confusion.

389

( b ) BRANCHING , AND ORIGIN OF THE INFLORESCENCE.

In spite of the plasticity of stolons, leaves and rhizoids these organs are arranged

in positions bearing a clear relationship to one another.

The main stolon (St. 1, Figs. 2 and 3) is of indefinite growth. At various points

along its length it gives rise to aerial shoots which will bear flowers. These aerial

shoots or aerial stems are not to be confused with the “ air-shoots ” of Goebel (Lloyd,

p. 223). Secondary stolons (St. 2) arise at the base of an aerial shoot at right angles

to the main stolon and a short distance above the junction of the aerial shoot with the

main stolon. A single primary leaf (Lloyd, Plate 23, Fig. 7) is always associated with

an aerial shoot and originates in an angle between St. 1 and St. 2. These leaves are

circinate away from the aerial shoot.

Rhizoids are formed in consecutive pairs alternately on opposite sides of the

aerial stem, the first pair being at right angles to the secondary stolons. As the

rhizoids may later become stolons, in Fig. 3, at a later stage of development, R1 may

be St. 3, R2 would then be Rl, and so on.

Secondary stolons and secondary leaves may also arise alternately along the

main stolon between aerial shoots, while tertiary stolons and leaves may arise alternately

along a secondary stolon. It will be seen that in this way a much branched plant

bearing inflorescences may be formed, and that the branching conforms to a pattern

with unexpected regularity.

Fig. 2. — Vegetative structures of U. transrugosa Stapf.

The plant was freed of organic debris and soil particles with the aid of a fine jet

of water.

A.S.: aerial shoot. St. 1: main stolon. St. 2: secondary stolon. St. 3: tertiary

stolon. LI: primary leaf. L2: secondary leaf. Rl, R2: rhizoids.

(c) FLORAL CHARACTERS.

The erect aerial stem of U. transrugosa bears from 1-3 personate flowers which

have a delicate honey-like scent. Pedicels are about 2 mm. long and are subtended

by a bract which is often recurved in the fruit. 1-2 barren bracts are found near the

base of the aerial stem. Two bracteoles arise at the base of the pedicel above the

fertile bract. They are narrower than this bract but of equal length. The calyx consists

of two reddish-purple sepals, 3 mm. long, which are persistant and become papery

in the mature fruit.

Fig. 3. — Young inflorescence showing origin, and branching of stolons.

This is an enlargement of a portion of the plant represented in Fig. 2. Lettering

as for Fig. 2.

The corolla varies greatly in colour from mauve or purple with yellow markings

on the palate, to pale yellow or white. Its upper lip is 6 mm. long and constricted

in the centre, the lower half being broader than the upper half. The apical margin

of this lip shows slight variation in that it is in some cases smoothly rounded and in

others, emarginate. The lower lip rises from a reflexed spur, 7 mm. long, to arch

forward and meet the upper lip; thence it is reflexed backwards and expands to form

a platform 11 mm. broad, the margins of which are upturned. In the mouth formed

with the upper lip it bears two lateral rugose swellings; each swelling usually has

20-30 transverse rugosities which are dark purple and velvety in appearance. These

rugosities often converge or bifurcate forming in places a zig-zag pattern. The

depression formed between the two lateral rugose swellings is variously marked with

yellow patches.

Two stamens are attached to the lower lip at the mouth of the spur. When mature,

they face away from the ovary and towards the mouth of the spur. Fig. 4 shows the

stamen-ovary relationship in three stages of development of the flower. In (a), a very

young bud, filaments are short and anthers very large and closely adpressed to the

ovary. In ( b ) the filaments have elongated and anthers have separated and rotated

so that they face each other, while in (t) the mature anthers have completed a rotation

through 180° so that their previously outer margins become the inner margins, and

the anthers face away from the ovary instead of towards it as in the bud. During

this roration the filaments become twisted. When fully mature the anthers adhere

at the centre of their inner margins. Dehiscence is by means of longitudinal slits.

391

Fig. 4. — Stamen-Ovary Relationship.

a, b and c are drawn from flowers in progressive stages of development. Note

broad point of attachment of anther to filament.

The ovary is globose and constricted at the apex into a short style which broadens

out into a 2-lobed stigma, the upper lobe of which is narrow and acute and is adjacent

to the upper corolla lip. The lower lobe is wide and fan-shaped and bears the stigmatic

surface.

The fruit is a capsule with numerous seeds tightly packed on a free central placenta.

Seeds are angular, predominantly wedge-shaped (Fig. 5) and the placenta has become

convoluted and folded between them, so that when all the seeds are removed the

placenta has the appearance of a fine honeycomb. The calyx and the ovary wall become

papery in the fruit and dehiscence is by means of irregular longitudinal tears.

Fig. 5. — Seeds of XJ. transrugosa Stapf in various positions.

The broad, flattened surface of the seed faces externally (430 x nat.).

ANATOMY.

The microscopic and macroscopic structures of this plant proved to be different

from other angiosperms. The anatomy of aerial shoot, leaf, stolon and bladder was

studied. Observations were made primarily from microtomed sections, 15 g and 20 g

thick. Material which had been preserved in 4 per cent formalin was embedded in

55 paraffin wax and the sections were later stained with safranin and fast green. As

392

the tissues of Utricularia are so delicate, hand sections and dissections were necessary

to substantiate evidence from prepared slides. Especially in the case of bladders,

microtomed sections should not be relied upon, as tearing and displacement of tissues

might have lead to a wrong interpretation of bladder structure.

Fig. 6. — Portion of transverse section through aerial stem of Utricularia transrugosa Stapf

E: epidermis. O.P: outer parenchyma (chloroplasts are not shown).

S: sclerenchyma band. Ph.: phloem. X: xylem vessel.

I.P: inner parenchyma.

393

(a) ANATOMY OF THE AERIAL SHOOT.

Epidermal cells of the aerial shoot are large and heavily cuticularised. Their

radial walls appear to be thickened towards the outside. Many stomata are present,

particularly towards the base of an aerial shoot. (See leaf anatomy for structure of

stomata.)

Immediately within the epidermis is a zone of parenchymatous tissue about 4 cells

wide. These cells are rounded and arranged in longitudinal rows with numerous

intercellular air-spaces. Chloroplasts are abundant in this region. No distinct

endodermis coula be seen. The outer parenchyma zone was bounded internally by

a complete sclerenchyma band, 1-3 cells wide.

The vascular tissues of U. transrugosa are embedded in a groundwork of angular

parenchyma cells, which are smaller towards the sclerenchyma zone becoming larger

towards the centre of the stem. No air-spaces are present between these cells.

Xylem vessels are arranged in a circular zone a short distance from the sclerenchyma

band. Vessels are scattered, occurring singly or in pairs. Thickening of their walls

is either spiral or annular and both types of thickening may occur together in a single

Fig. 7. — Transverssection the rough stolon of U. transrugosa Stapf.

' E: epidermis. P: parenchyma. G: gland. X: xylem vessel. ?Ph.: phloem.

394

vessel element. Small, thin-walled patches of tissue are found scattered in the angular

parenchymatous region near the vessels, and similar patches occur abutting on the

sclerenchyma zone. These patches were interpreted as phloem.

o b ) ANATOMY OF THE STOLON.

The epidermis and parenchyma of the stolons are similar to those found in the

aerial shoot. In Fig. 7, one of the glands, which occur abundantly on the outer surface

of the stolon, is sbov/n. The inner basal cell is embedded in the epidermis; the middle

cell is narrow and collar-like and bears the single rounded capital cell, which is not

cuticularised.

Towards the centre of a stolon, parenchyma cells become smaller and closely

packed. Here they surround a strand of thin- walled, tightly packed, angular cells

in which a single xylem vessel is embedded. The xylem vessel is similar to those found

in the aerial shoot. It is suggested that the angular cells may be phloem tissue.

(c) ANATOMY OF THE LEAF.

The leaf, as stated previously, is no more than an expanded stolon.

Fig. 8 shows that the central strand of a stolon continues into a leaf and there

it divides dichctomously once, or more rarely, twice. The single xylem vessel, present

in the stolon, passes into a leaf but terminates before the vascular strand branches

into two, the veins consisting thereafter of angular cells only. Chloroplasts are

numerous in the parenchyma cells of the leaf. Mesophyll is undifferentiated.

Fig. 8. — Schematic representation of leaves of

U. transrugosa Stapf

B: bladder. VS: vascular strand.

Epidermis and glands are similar to those described for stolons. Stomata are

found towards the apex of the upper surface on the leaf, and very occasionally on the

lower surface. They are not present in the lower cylindrical portion. This limited

occurrence of stomata is to be expected as most parts of the leaf are usually embedded

in mud or soil. The stomata are slightly sunken (Fig. 9) and, as will be seen from

Fig. 10, they conform to the anomocyclic or “ Ranunculaceous ” type of Metcalfe

and Chalk (1950) in that they are surrounded by cells which are indistinguishable in

size, shape or form from those of the remainder of the epidermis.

The leaf epidermis does not strip readily, thus in order to study stomata, leaves

were soaked overnight in Eau de Javelle and then the internal parenchymatous tissue

395

was scraped away, leaving the epidermis intact. For material which has been preserved

in formalin, previous soaking in Eau de Javelle can be omitted.

Fig. 9. — Transverse section through upper portion of leaf.

This section is cut through the line 1-2 of Fig. 8 b.

E: epidermis. G: gland. P: parenchyma with chloroplasts. VS: vascular

strand.

Note absence of xylem vessels in this region.

Fig. 10. — Epidermis, showing stoma and gland in surface view.

G: gland, which consists of C: capital cell, M: middle cell and B: basal cell.

G.C: guard cell. S: stoma pore. E: epidermal cell.

The middle cell of the gland is represented by a dark line due to refraction caused

by the spherical capital cell.

396

THE MORPHOLOGY AND ANATOMY OF THE BLADDER.

It has already been noted that the bladders or traps are present in large numbers

on stolons and leaves of U. transrugosa. They are small ovoid-shaped structures,

2 mm. long and 1 mm. broaa, and are slightly flattened on the sides. They are

attached to a stolon or leaf by a narrow stalk which varies considerably in length.

The cells at this point of attachment are slightly thickened and possibly prevent the

bladder from being torn away. The stalk is attached near the base of one end of the

bladder, while at the opposite end the walls of the bladder are produced into an upper

and a lower lip. The two bps are joined laterally by folds of tissue which form the

cheeks. These structures together form a protective funnel-like entrance to the mouth

of the bladder. Lloyd has stated that the stalk end of the bladder is ventral, the mouth

end dorsal (p. 233). It is more convenient however to describe dorsal and ventral

surfaces of the bladder to correspond with upper and lower lips. From Fig. 11 it

will be seen that the lips carry numerous glandular hairs which interlock with one

another across the mouth entrance. Each hair consists of 4 cells. The lowermost

or “ wall cell ” is a prolongation of an epidermal cell. This is followed by a basal

cell, a narrow cylindrical middle cell, and a capital glandular cell (see Fig. 12 b). The

glandular hairs are arranged in six longitudinal rows on both upper and lower lips.

On the cheeks, glandular hairs are shorter and uniseriate. The large number, and

Fig. 11. — Cellular representation of bladder of U. transrugosa Stapf

Only the outer wall is shown with the quadrifids being visible because of trans-

parency of the bladder wall. The stippled area in the upper left of the bladder is the

mouth region.

d: door, q: quadrifid. c: cheek, s:' outer spherical gland. 1: the shaded

circular line indicates extent of lumen, vs: vascular strand, gh: glandular hair.

(Note. — Soaking bladders overnight in 20 per cent citric acid solution assists

considerably in cleaning them from soil, detritus, algae, etc.)

397

Fig. 12 (a). — Upper lip of bladder seen from below.

(b). — A single glandular hair from lip of bladder.

w: “ wall cell b: basal cell, m: middle cell, c: capital cell.

interlocking positions of these glandular hairs probably serve two functions : (a) they

form a protective covering to the mouth and prevent soil particles and detritus from

clogging the entrance; ( b ) they assist in maintaining a continuous film of water around

the trap mechanism.

the mouth. Water is essential to the proper functioning of the trap mechanism.

There are no antennae or bristles (Lloyd, p. 233) on the lips of the bladder of

U. transrugosa.

The bladder wall is composed of two layers of cells except in the lip region where

it becomes thicker. The outer wall layer is continuous with the epidermis of stolon

or leaf. Scattered over this outer layer are the small, 3-celled, spherical glands which

are present over the rest of the underground plant (see Fig. 7). The cells of this layer

are elongated along the profile of the bladder (Fig. 11) and are smaller and compressed

in the cheek region, but become more equidimensional and wavy-walled on the

surface view

Fig. 13. — Quadrifids in surface and side view.

c: projection of capital cell, m: middle cell, b: basal cell, n: nucleus.

398

flattened sides of the bladder. These irregular walls may assist in retaining bladder

shape during action and resetting of the trap when the sides are alternately convex

and concave.

The cells of the inner wall layer are in general larger in surface area than the outer

layer cells. In the lip and cheek region the two layers are separated by one to many,

large, thin-walled parenchymatous cells. The two walls enclose a central cavity or

lumen which is reached from the outside through the mouth.

A single vascular strand passes into the bladder through the stalk, travels dorsally

in the wall layers without branching, and finally terminates in the tissue of the

upper lip.

Projecting into the lumen are glandular hairs bearing two or four armed capital

cells. Darwin called them bifids and quadrifids (1888). The binds are limited to the

mouth region. Each gland arises from a cylindrical basal cell and has a disc-shaped

middle cell, and a 2- or 4-armed capital cell which is devoid of cuticle. The arms of

the quadrifids are equal in length and are not spreading (Fig. 13).

Fig. 14. — Longitudinal section through mouth region of bladder.

owl: outer wall layer, iwl: inner wall layer, q: quadrifid. b: bifid, d: door,

v: theoretical position of velum, pe: pavement epithelium, c: balloon-like cuticles,

ul: upper lip. 11: lower lip. gh: glandular hair, tb: tripping bristle, be: but-

tress cell.

399

Fig. 15. — Sections through bladder along the lines 1-2, 3-4, and 5-6 of Fig. 14.

vs: vascular strand, q: quadrifid. s: outer spherical gland, pe: pavement

epithelium, iwl: inner wall layer, d: door, g: gland on threshold.

400

The basic structure of the mouth region can be studied from Figs. 14, 15 and 16.

In longitudinal section it will be seen that the upper and lower lips project beyond

the rest of the bladder and that they are several cells in thickness. The lower lip also

extends back in towards the lumen of the bladder as a wedge-shaped shelf — the collar

or threshold. The bladder entrance is guarded by two valves; a larger one, the dcor,

hanging obliquely inwards at an angle of 45°, and a smaller one, the velum. The

door is an oval flap of tissue which hangs downwards from the inner margin cf the

upper lip. Its base rests against the threshold (Fig. 14). In its upper half, the sides

of the door are attached to the cheeks, which are the flaps of tissue connecting upper

and lower lips. In its lower half, the sides of the door are attached to the “ pavement

epithelium ” (of Goebel), which is a glandular layer arising from the threshold. In

this way the cheeks and pavement epithelium form a tunnel-like entrance with the

door hanging obliquely across the tunnel and being attached on all sides except at

the base.

The angle at which the door hangs downwards is about 45°, and this together with

the general structure of the mouth region, classifies the bladder of V. transrugosa as

one of the “ long tubular entrance ” type of Lloyd (p. 258).

From Fig. 15 it will be seen that the door in transverse section is not flat but it

hangs like a hammock, the concave side being towards the upper lip, the convex side

towards the lower lip. Fig. 14 shows the door in longitudinal section. It is evidently

curved in this plane also.

In lower surface view, Fig. 16, it will be seen that the door is roughly oval in

shape with the lower free edge truncated and appearing slightly concave due to

flattening of the door in making the preparation for Fig. 16. Two distinct regions

are visible; an upper and a lower. The lower half of the door is composed of small,

thin-walled cells and is non-glandular. Cells in the upper half of the door are somewhat

large and bear numerous glands. These glands radiate outwards from a central point

where they are very short, and are also found on the invaginated cheeks and on the

inside of upper and lower lips where they are much longer. The glands are slightly

different in structure from other glands of U. transrugosa in that they are “ sessile ”

i.e. the normal basal cell is lacking, and they arise directly from prolongations of the

cells which produce them. The capital cells of these glands in the central region of

the door are small and spherical, but in the upper and outer regions of the door they

are greatly elongated. From the central point at the junction of upper and lower

halves of the door and from which the glands radiate outwards, arise two long, stiff,

tapering bristles. Very rarely, three of these bristles were seen. They form the tripping

mechanism. Each bristle is of two cells, the upper being much smaller than the lower,

and the lower cell arising directly from the door without any bulbous swelling at its

base. These bristles in all cases, projected outwards and upwards towards the

upper lip.

In longitudinal section (Fig. 14) the door is seen to consist of two layers of cells;

an outer one from which the glands and tripping bristles arise, and an inner layer of

cells. Two distinct regions of the door are again visible; an upper glandular, and

a lower non-glandular region. In the upper portion, cells of the inner layer are slightly

broader than outer layer cells (in a ratio of about 2:1) and are similarly deeper, except

at the point of attachment of the door to the upper lip where cells of the inner layer

are about three times as deep as the outer layer cells. In the central region of the door

the cells of both layers are approximately equidimensional.

The lower non-glandular portion of the door has a characteristic and peculiar

structure, similar to that found in U. kirkii Stapf. (Lloyd, p. 232, 260, Plate 33.) Cells

of the inner door layer are greatly enlarged and form two lateral bulges separated by

a narrow groove (Fig. 15). In Fig. 14 one of these bulges is shown in longitudinal

section.

401

The door is attached to the rest of the bladder along a semicircular line, and

considerable stress is placed on cells along these lines of attachment during action

of the trap.

This is compensated for by the presence of thickened “ buttress ” cells in the

inner layer of the door. These buttress cells are shown in surface view in Fig. 16.

They are visible because of transparency of the outer cell layer of the door. It was

not possible to confirm Lloyd’s interpretation of these “ buttress ” cells. He states

that “ the cells are constricted at regular intervals ” and that the spaces between the

Fig. 16. — Door of bladder in lower surface view.

This preparation was obtained by slitting the threshold along a median line and

spreading it out on either side of the door. This was then viewed flattened back against

the upper wall and lip of the bladder.

11: lower lip. c: cheek, sg: sessile gland, oz, mz, iz: outer, middle and inner

zones of pavement epithelium, ic: balloon-like cuticles from inner zone, fe: free

edge of door, uw: upper wall of bladder, q: quadrifid. tb: tripping bristle,

be: buttress cell, ul: upper lip.

402

constrictions are mistaken for single cells. (See Lloyd, p. 250 and Plate 33, No. 2.)

Microtomed sections failed to confirm this and the view is expressed that these are

indeed single cells. They are shown in transverse and longitudinal sections in Figs. 15

and 14. Buttress cells are not found where the door is attached to the pavement

epithelium.

The pavement epithelium is a glandular layer arising from the threshold. The

glands of the pavement epithelium are also sessile in that their lowermost cells are the

outer layer of the threshold, and a true basal cell is lacking. In Fig. 14 it can be seen

that these glands are tightly packed together in the central region, becoming wider

spaced and separated from each other at either end. Three zones are thus distin-

guishable; an outer, middle and inner zone.

The velum described by Lloyd (p. 247) was not seen intact during these investi-

gations. Its theoretical position has been indicated by dotted lines in Fig. 14. The

velum is formed by the cuticles which are shed from the capital cells of the pavement

epithelium glands. These cuticles remain attached to one another and to the capitals

of the outer zone glands, and form a thin membranous flap which covers the free edge

of the door. It is only the cuticles of the outer and middle region glands which form

this velum, while those from the inner zone of glands behave individually. These

enlarge and burst forming a cushion of balloon-like tissue which blocks the small

chink between the edge of the door and the pavement epithelium glands. These were

observed and have been represented in Figs. 14 and 16.

Thus it may be seen that the bladders of Utricularia transrugosa are structurally

and anatomically complex. This allows for the intricate mechanism of trap action,

which has been fully described by Lloyd.

POINTS OF BIOLOGICAL INTEREST.

(a) PERENNATION.

It has been established during these investigations that U. transrugosa is perennial

under field and laboratory conditions.

A block of earth, about 10 inches x 10 inches X 10 inches was dug from the

vlei where U. transrugosa has previously been collected. At the time cf removing

the earth (May, 1952) there were no visible signs of the presence of the plant on the

surface of the vlei. The earth was kept in the laboratory at room temperature

t24-28° C.) and was watered daily with tap water. No artificial nutrients were supplied

to the soil. After two months (July, 1952) the plant had not reappeared in the vlei,

but small green leaves appeared on the surface of the soil which had been kept in

the laboratory. The plant from which these leaves were produced was then dissected

free of soil. A considerably thickened stoloniferous structure was noticed, and from

this arose “ new ” stolons bearing young green leaves. Clinging to the thickened

portion were the remains of old, decayed stolons and leaves. The regular branching

previously described for U. transrugosa Stapf was not evident in this perennating

portion of the plant. As thickened structures were not found to be present in dissections

of mature flowering plants, one may assume that as the young plants establish them-

selves they either break away from the thickened portion or this shrivels due to

depletion of food reserves. Plants grown in this way in the laboratory did not develop

rhizoids and all the bladders present were empty. Thus all nutrients for the developing

plant must have come from the thickened stoloniferous portion or have been absorbed

from the soil through the non-cuticularised capitals of the spherical glands.

Under field conditions, too, similar thickened perennating parts of the plant were

found. A very young plant of which only the tips of the leaves were visible, was taken

from mud along the banks of a stream and dissected free of soil and detritus. Fig. 17

shows two of the thickened areas found on this plant. It was noted that young stolons

arising from these thickened parts were green, especially at the rounded growing

403

apices. The preparation from which Fig. 17 was drawn was left soaking in tap water

in the laboratory for two weeks, after which time the small green outgrowth (Fig. 17, z)

had developed from one end of the thickened portion.

The ability to develop perennating structures is not an uncommon phenomenon

among the Utricularias. U. globulariaefolia, a fairly large terrestrial species, becomes

perennial by its stout, tough stolons, similarly to U. transrugosa. In U. volubilis the

plant body is an upright corm which grows at the top and dies behind. From the

corm are produced numerous leaves. Wager (1928) described the ability of U. stellaris

to form resting buds towards the end of a season. These buds were then able to develop

the following spring. Many species of Utricularia have now been found to produce

resting buds and the structural differences of these buds have been used as a basis

for classification of the species (Rossbach, 1939).

(b) GERMINATION.

All efforts to germinate seeds of U. transrugosa were unsuccessful. Seeds were

selected, and sterilised by soaking for 2 mins, in mercuric chloride (0-1 per cent in

50 per cent alcohol). They were then washed in sterile distilled water and left to soak

overnight under sterile conditions.

Seeds were then placed on moist sterile filter paper, or on agar in petri plates,

10 seeds per plate, and were kept in the laboratory at room temperature which fluctuated

between 24-28 C. Two kinds of agar were used; Difco bean pod agar which had

Fig. 17. — Portion of perennating plant of U. transrugosa Stapf

(Taken from natural habitat.)

The density of stippling is proportional to the greenness of the plant areas,

pp: perennial portion, yl: young leaf, yz: young stolon, os: old stolon,

db: developing bladder, z: young green shoot (see text).

404

a final pH of 5 • 3, and a second medium prepared as follows : 150 g soil, taken from the

vlei in which the plants had been found growing, was thoroughly shaken up with

150 ml distilled water, and left to soak for 3 hrs. The suspension was then filtered

and 2 per cent agar was added to the fiquid obtained. The final pH of the medium

was between 4 -6-4 -8.

It was expected that as the plants grew in acid soils, media having low pH values

would be conducive to the germination of seeds. Unfortunately no definite reason

is known for the failure of germination.

(c) POLLINATION.

The manner in which pollination is effected in U. transrugosa can only be conjec-

tured. The floral characteristics of the plants are highly suggestive of adaptions to

cross-pollination by insects. The colour of the flowers together with their honey-like

scent would act as a lure to insect visitors, which would then be directed toward the

mouth of the flower by the conspicuous transverse rugosities. In entering the mouth

of the flower the insect would brush against the broad upwardly facing stigmatic lobe,

and on continuing downwards into the spur of the flower it would collect pollen from

the anthers. It is of interest, that if the upper corolla lobe is slightly depressed (as

by the entry of an insect) the anthers are drawn away from the mouth of the spur and

face obliquely upwards thus being in a position which makes it impossible for an insect

to enter or lave the spur without contacting the anthers.

In retreating from a flower the insect would brush the non-receptive underside of

the stigmatic lobe and self pollination would be avoided.

(d) VARIATION IN NATURE.

Several instances of variation within the species U. transrugosa have already been

given and need only be summarised here. The most conspicuous variation occurs in

flower size and colour, the purple flowers being generally the larger. The apical margin

of the upper corolla lip varies in shape.

Vegetative structures are more constant however, the only variations that have

been noted being the length of bladder stalk, and the presence of 2 or 3 tripping bristles

on the door of the bladder.

Kamienski (1895) and Oliver (1867) have both noted a wide range of variation

in floral characteristics within the species of Utricularia. This variability is of importance

when one studies the characteristics which are used to distinguish species of Utricularia

especially in the case of U. lividalU. transrugosa and U. kirkii or in the case of U.

capensisjU. brachyceras and U. ecklonii. This will be treated more fully in the following

discussion.

(e) FOOD OF U. TRANSRUGOSA STAPF.

If one accepts that Utricularias obtain or supplement their nitrogenous food

requirements by the capture and digestion of animal prey, for which processes water

is essential, the survival of U. transrugosa in an acid semi-dry vlei is intersting. Exami-

nation of bladders from this plant, however, showed that although soil water content

was so low, the bladders were functioning efficiently. Decayed animal remains were

found in most bladders and appeared to consist generally of small arthropod larvae,

crustaceans and nematodes. In some cases prey was extracted alive from the bladders.

An attempt to grow Utricularias in the laboratory was only partially successful.

The plants were kept in a trough containing peat and large quartzite pebbles, and

were watered regularly. The water was taken from an outdoor alga-tank which

contained abundant small animal and plant life. The Utricularias remained alive,

but only one plant produced an aerial stem, which however failed to bear flowers even

though watered with beef extract solution.

405

(/) PHENOLOGY.

U. transrugosa has a relatively short flowering period (about 3 weeks) and its

time of flowering is largely dependent on adequate early summer rains. The plant

was collected at Bryanston in September, 1951, after good rain, but in 1952 two

flowering periods were noticed. In August and September 1952, a very dry period,

isolated dwarf specimens were collected from the vlei and stream at Bryanston. Later

in the year (November), after rains had fallen, flowers were again collected from the

stream locally. Unfortunately, the vegetation covering the original vlei was burnt

between September and November and the vlei has now completely dried out. U.

transrugosa did not reappear in this site, even after heavy rains, and whether its

disappearance is temporary or not cannot be told until the following summer.

It was evident that the long dry period in 1952 had a deleterious effect on the

flowering of U. transrugosa, which had flowered far more abundantly in 1951.

DISCUSSION.

The most striking point arising out of these investigations is the anatomical and

morphological similarity between JJ. transrugosa Stapf and U. kirkii, Stapf. In his

analysis of the genus in Flora Capensis, Stapf separates the two species by flower

size, U. kirkii being considerably smaller, and by the presence of rugosities on the

palate of U. transrugosa contrasted with the minute tubercles in U. kirkii. However,

as many species of Utricularia show variation in flower size, the only tangible difference

between U. transrugosa and U. kirkii is the presence of rugosities or tubercles on the

palate. That this is the principal difference is further substantiated by Lloyd, whose

figure illustrating the bladder of U. kirkii shows this structure to be almost indistin-

guishable from the bladders of U. transrugosa, described in this report. Both species

are terrestrial, and Fig. 1 shows that they have been recorded from similar localities

i.e. in the Transvaal, and a short distance from the Zambesi in Central Africa. Lloyd,

(p. 260) has further stated, with regard to the bladder, that “ U. kirkii, occurring in

Central Africa, is apparently unique ”. Also (p. 232) “ U. kirkii is an African species

with apparently few associates if any, and has a distinct form of trap ”. The two

lateral bulges formed by the inner cells of the door (called tubercles by Lloyd), have

not yet been described for any other species. Bladders of U. transrugosa and U. kirkii

are also similar in that they lack a “ doorstep ” at the point where the lower lip turns

inwards to produce the pavement epithelium. In other species of Utricularia, the lower

lip forms a non-glandular, flat, steplike ridge of tissue leading to the pavement.

Two alternatives present themselves either: ( a ) U. transrugosa is a large flowered

form of U. kirkii, or ( b ) the two species are a vicarious pair.

The earlier indentification of U. transrugosa as U. livida var. transrugosa leads

one to consider the differences between the two species U. livida and U. 1 transrugosa .

Again it is found that the distinguishing features are corolla size, and the presence

of tubercles on the palate of U. livida and of rugosities in U. transrugosa. Unfortunately

no details of bladder structure are available for U. livida. If bladders were examined

and found to conform to the U. kirkii/ U. transrugosa type, an interesting position,

worthy of further investigation, would arise. U. exilis Oliver from Flora of Tropical

Africa (Fig. 1, F) which is very similar to U. kirkii and has a smooth palate, is of the

same affinity also.

Miss L. Stevens (1938) attempted to unravel the relationship between the species

U. capensis I U. brachyceras/U . ecklonii. Stapf (1904) maintained that U. brachyceras

was a short spurred form of U. ecklonii and that U. capensis was a distinct separate

species. Miss Stevens implies, however, that U. ecklonii is a growth form of U. capensis

and should be sunk in that species* while U. brachyceras is quite distinct from the

combined U. capensis — U. ecklonii. Her opinion is based on relative sizes of the

stigmatic lobes.

406

The work of Lloyd, however, has made it obvious that in cases of close floral

similarity, differentiation of species cannot be based on floral characteristics alone.

The pattern of branching and bladder structure are sufficiently constant and different

between species, to be used as features for taxonomic classification.

Although Lloyd does not refer to U. transrugosa, his descriptions of bladder types

occurring among the genus Utricularia have proved useful guides.

SUMMARY.

A morphological and anatomical description of the floral and vegetative characters

of Utricularia transrugosa Stapf is given.

Anatomical details of the bladder place it in the “ long tubular entrance ” group

of Lloyd, and show a striking similarity between U. transrugosa and U. kirkii Stapf.

Biological findings indicate that U. transrugosa is perennial, and like many

Utricularias shows considerable variation in flower size and colour.

ACKNOWLEDGEMENTS.

The writer is grateful to Professor N. P. Badenhuizen for permission to carry out

these investigations in the Botany Department, University of the Witwatersrand, and

to Dr. H. B. Gilliland for his constant guidance and assistance in collecting material.

The Chief, Division of Botany, and the Director, Transvaal Museum are thanked

for the use of the facilities of the National Herbarium and of the Transvaal Museum

respectively.

BIBLIOGRAPHY.

Compton, R. H 1909. The Morphology and Anatomy of Utricularia brachiata, Oliv. New

Phytologist 8, 117.

Darwin, C 1888. Insectivorous Plants (London).

Kamienski, F 1902. Utricularia in Bot. Jahrb. 33, 92.

1897. Utricularia in Engler-Prantl Nutur. Pfianzen. IV Teil Illb abt., 118

Lloyd, F. E 1942. The Carnivorous Plants (Waltham, Mass.).

Metcalfe, C. R., and

Chalk, L 1950. Anatomy of Dicotyledons, 2, 993.

Oliver, D 1867. African Lentibulariae. J. Linn. Soc. 9, 153.

Rossbach 1939. Winter buds among Utricularia. Rhodora 41, 114.

Solreder, H 1908. Systematic Anatomy of Dicotyledons, 1, 694.

Stapf, 0 1904. Lentibulareaceae in FI. Cap. (London) IV. 2, 429.

Stapf, 0 1906. Lentibulareaceae in FI. Trop. Afr. (London), IV. 2, 468.

Stapf, 0 1931. Lentibulareaceae in FI. W. Trop. Afr. (Lndon) 2, 232.

Stapf, 0 1916. Utricularia papillosa Stapf. Kew Bull. 41.

Stevens, E. L 1938. Notes on three South African Utriculariae. Journ. S.A. Bot.

Wager, V. A 1928. Notes and descriptions of a Few Uncommon Water Plants. Trans.

Roy. Soc. S.A. 16, 191.

Weintroub, S 1933. A Preliminary Account of Aquatic and Sub-aquatic Vegetation and

Flora of the Witwatersrand. J. Ecol. 21, 44.

407

The Nomenclature of Species of Allophylus in

South Africa.

By

B. de Winter.

No complete revision of the African species of the genus Allophylus has been

attempted since Radlkofer published his monograph of the Sapindaceae in Engler’s

Pflanzenreich in 1932-1934. The nomenclature adopted in the latter work is not in

agreement with the latest edition of the Int. Code of Bot. Nomenclature (1952). This

note is an attempt to clear up the confusion existing re the use of the correct names

and also the correct citation of authors. Fortunately, name changes were found to be

unnecessary except in two species.

1. Allophylus decipiens ( Sond .) Radik, in Engl. & Prantl. Pflanzenfam. 3 (5) 313 (1895).

Rhus spicatum Thunb. FI. Cap. 11, 217 (1818).

R. decipiens E. Mey. in Drege Cat. PI. Exsicc. Afr. Austr. Febr. 1839; Arnett*

in Hook. London Journ. Bot. 3, 153 (1841). nomen nudum.

Schmide/ia decipiens (E. Mey.) Presl. Bot. Bemerk. 41 (1844) nomen nudum.

S. decipiens Sond. in Harv. et Sond. FI. Cap. 1, 239 (1860).

Allophylus spicatus (Thunb.) Fourcade in Trans. Roy. Soc. S. Afr. 21, 100 (1932),

non A. spicata (Pair.) Radik, in Engl. & Prantl, Naturl. Pflanzenfam. 3 (5), 312 (1895).

This species was misidentified by E. Meyer as Rhus undulata Jacq.

2. Allophylus dregeanus ( Sond .) de Winter, comb. nov.

Rhus monophylla E. Mey. in Drege, Zwei Pflanzengeogr. Doc. 216 (1843), nomen

nudum.

Schmiclelia monophylla (E. Mey.) Presl. Bot. Bemerk. 470 (1845) nomen nudum,

non Hook. f. in Hook. Icon. t. 775 (1845).

S. dregeana Sond. in Harv. et Sond. FI. Cap. 1, 239 (1860).

S. monophylla var. natalitia Szyszyl. Enum. Polypet. Rehmann 2, 47 (1888).

Allophylus monophyllus Radik, in Engl, et Prantl, Naturl. Pflanzenfam. 3 (5),

312 (1895). nomen nudum, et in Engl. Pflanzenr. 4, 165. 518 (1932).

3. Allophylus melanocarpus {Sond.) Radik, in Engl, et Prantl. Naturl. Pflanzenfam. 3 (5),

313 (1895).

Rhus melanocarpa E. Mey. et Rhus leucocarpa E. Mey. in Drege Cat. 3 (1840);

Arnott in Hook. Journ. Bot. 3, 153 (1841); Meyer in Drege, Zwei Pflanzengeogr.

Doc. 216 (1843), nomina nuda.

' Schmidelia melanocarpa Sond. in Harv. et Sond. FI. Cap. 1, 238 (1860).

'' S. leucocarpa Sond. l.c.

S. rehmanniana Szyszyl. Polypet. Rehmann. 2, 46 (1888).

* Arnott has often been cited as the author of this species. A short description is given by him

when contrasting the species with Rhus melanocarpa and Rhus erosa E. Mey. non Thunb. The short

insufficient diagnosis is not valid, however, since Arnott still uses Rhus as the generic name although

he points out that the plants belong to the genus Schmidelia. (See Art. 43, Int. Code of Bot.

Nomenclature, 1952). The same applies to Rhus melanocarpa and Rhus leucocarpa. Sonder (in FI. Cap.)

gives good descriptions and the names must be attributed to him.

9102-6

408

4. Allophylus natalensis ( Sond .) de Winter comb. nov.

Schmidelia natalensis Sond. in Harv. et Sond. FI. Cap 1. 239 (1860).

A. erosus Radik, in Engl. Pflanzenr. 4, 165, 544 (1932).

This species was validly described for the first time by Sonder who gave it the

name S. natalensis. Earlier, the species had masqueraded under the illegitimate name

“ Rhus erosus, E. Mey.” The origin of the latter name is plainly a misidentification

which can be traced to E. Meyer who in the various works on Drege’s plants named

this species as Rhus erosus Thunb. When it was realized that the latter was a distinct

species subsequent authors used the name Rhus erosus E. Mey. non Thunb.

Radlkofer recognized this fact and to rectify the error described Allophylus erosus

Radik, as an independent species based on the plant niisidentified by E. Meyer as

Rhus erosus Thunb. This would have been legitimate had there been no other name

available.

The following species complete the list of members of the genus Allophylus recorded

for South Africa.

5. Allophylus transvaalensis Burtt Davy in Kew Bull. 280 (1921).

This is probably only a hairy variety of A. melanocarpus (Sond.) Radik.

6. Allophylus gazensis Bak. f. in Journ. Bot. 57. 182 (1919).

Galpin 909, Lakwongwa Forest, Barberton, Transvaal; matches the type of

A. gazensis in Kew.

409

The Nomenclature of the Cape Acacia.

By

I. C. Verdoorn.

By the “ Cape Acacia ” is meant the species which occurs naturally and abundantly

to within about 75 miles of Cape Town. It is the only Acacia species in that region

but is not restricted to the Cape and further to the north and east other species of

Acacia occur as well. The early travellers such as Barrow, Lichtenstein, Sparrman,

Thunberg and later Burchell make mention of this Acacia for it is a feature of the

landscape in the Hex River Pass and in quite a wide radius from that point to the

interior.

Besides the common name “ Cape Acacia ” it is also known as the “ Karoo Thorn ”,

“ Cape Thorn Tree ”, “ Doring Boom ”, “ Witdoring ” or “ Mimosa ”. In the

Transvaal it is commonly called “ Sweet Thorn ” or “ Soetdoring ”. According to

Simon van der Stel’s Journal the Namaquas name for it was “ Choe ” while in Natal

the Zulus call it “ UmuNga ”.

The species is characterised by the following features : the habit of growth varies

from shrubby thickets to tall, well-shaped trees, all armed with long, straight, white,

paired spines; the sweetly scented flowers are yellow (the colour of egg yolk), massed

in small globose heads which are axillary in the upper leaves and run into leafless

panicles at the apex of the branchlets; the leaves are only 2 to 5 jugate with

comparatively large leaflets; and the pods are linear and sickle shaped.

The following is an account of the botanical name and synonyms of this plant.

Acacia karroo Hayne Arzneyk. Gebr. Gewachse 10, descr. et t. 33 (1827); Glover in

Ann. Bol. Herb. Vol. 1, p. 150 (1915); Burtt Davy in Kew Bull. 1922 p. 328 and

in Tvl. FI. 2, p. 346; Marloth in FI. of S.A. 2, p. 51 (1925); and Baker in Leg. of

Trop. Afr. p. 843 (1926).

Mimosa nilotica Burm. f. FI. Cap. p. 27 (1768); Thb. in FI. Cap. Ed. Schult. p. 432

(1823) non Linn.

Mimosa capensis Burm. f. 1. c. p. 27 (1768) partly.

Mimosa leucacantha Jacq., Hort. Schoenbr. 3, 75 (1798-1800) non Vatke.

Acacia horrida Willd. Sp. PI. (1805) as to Jacq. fig.; Harv. and Sond. in FI. Cap. 2,

281 (1865).

Acacia capensis (Burm.) Burchell. Trav. 1, p. 114 and 189 (1822); Sw. in Hort. Britt.

1, 103 (1826) nomen; Colla in Mem. Acad. Torin 35 p. 175 (1831); Eckl. and

Zeyher Enum. PI. Afr. Aust. 260 (1835) nomen.

Mimosa eburnea Bojer Hort. Maurit, 115 (1837) non Linn.

Mimosa nilotica Burm. f. For this species Burman quotes Plukenet t. 123 fig. 1

and mentions that there are dried specimens and other figures of it. Plukenet t. 123

fig. 1, although showing only a leafy twig, could very well represent the “ Cape Acacia.”

The leaflets are too big for Mimosa nilotica L. and besides it is known that that species

does not occur near the Cape. Through the kindness of Prof. Baehni, Director of

the Conservatoire et Jardin Botaniques, Geneva, photographs of the only two Acacia

specimens in Burman’s herbarium were received. The one specimen has written on it

“ Mimosa nilotica Linn ” and “ Plukn. Tab. 123. Fig. 1.” (the name Mimosa capensis

410

Fig. 1. — Photo of Acacia specimen in Burman’s herbarium, Geneva. Sent by

Prof. Baehni in July, 1952.

411

was later written over Mimosa nilotica, see our Fig. 1 . As far as one can tell from a

leafy twig this also represents the “ Cape Acacia So M. nilotica was a wrong

identification by Burman of the “ Cape Acacia

The same is true for Mimosa nilotica Thunb. Barrow made this clear when he

wrote in his “ Travels ” p. 89 (1801) “ . . . skirted by a thicket of the doornboom

or thorn tree, a species of Mimosa, called erroneously by the Swedish travellers

(Sparrman and Thunberg) the nilotica, or that which produces the gum Arabic. The

pods of this are very long and moniliform . . . whereas the Karroo Mimosa has

short sickle-shaped pods The sickle-shaped pod of our Acacia is the characteristic

by which it can be distinguished from other related species in S.A., Egypt, Asia, and

India.

Mimosa capensis Burm. f. In this case Burman quotes only Plukenet t 123 fig. 2

(see right hand figure reproduced here under our Fig. 2), and he does not mention dried

material. This figure is not recognisable as the “ Cape Acacia ” nor indeed as any

S. African species of Acacia. It is true that the same figure, somehow got into the

“ Van der Stel’s Journal of the Expedition to Namaqualand 1685-6 ” published by

Waterhouse in 1932 and is supposed to represent the Acacia met with on that expedition

which from description and locality probably was the “ Cape Acacia ”. On the other

hand the same figure is quoted by Linnaeus for his Mimosa reticulata. Bentham in

his Mimosae p. 507 comes to the decision that both the names based on Pluk. 123 fig. 2,

that is M. capensis and M. reticulata, must be rejected since the figure is not recognisable

and there are no dried specimens of either species. In following Bentham’s decision

the present author is supported by the Advisory Committee on Problems of Botanical

Nomenclature and Associated Matters in South Africa, which rules that since Burman

obviously did not understand the species when he published the name it must be

rejected It appears that later Burman learned of his mistake, for on the herbarium

sheet where “ Mimosa nilotica Linn.” had been written the name “ Mimosa capensis ”

was later written over it, as stated above (see again our Fig. 1).

Mimosa leucacantha Jacq. Jacquin’s plate is the first easily identified figure

published with a definite name but the combination of the epithet with Acacia is

invalidated by the earlier name Acacia leucacantha Vatke. for an entirely different

species.

Acacia horrida Willd. Under the name Acacia liorrida in Sp. PI. (1805), Willde-

now cites three distinct species: —

(a) Mimosa horrida L. Sp. PI. 1505 (,1764).

(b) Mimosa leucacantha Jacq. Hort. Schoenb. 3 p. 75 t. 393 (1798).

(c) Mimosa orfora Forsk. descr. 177 (1775).

(a) This is the type of Willdenow’s new combination. Linnaeus first described

Mimosa horrida in Sp. PI. Ed. 1, p. 521 (1753). Willdenow cites the description as

being in Sp. PI. Ed. 3, 1505 (1764) but since the wording is practically the same in both

editions it is not significant. In both editions the species is said to be an Indian species

and Pluk. t. 121, Fig. 1, is cited. So Willdenow’s name must be restricted to whichever

Indian species agrees with Pluk. t. 121, fig. 1.

(b) As pointed out above Jacquin’s figure is of the “ Cape Acacia ” and different

from Mimosa horrida L.

* As a matter of interest the other specimen of Acacia in Burman’s herbarium is that of a seedling

grown from seed and cannot be identified.

412

Fig. 2. — Photo of Plukenet Tab. 123: Fig. 1 Quoted by Burman for Mimosa

nilotica\ also quoted by Linnaeus for Mimosa scorpioides; Fig. 2 Quoted by Burman

for Mimosa capensis', also quoted by Linnaeus for Mimosa reticulata.

(c) Mimosa orfota (not orfora as written by Willdenow) is combined with Acacia

by Schweinfurth in Bull. Herb. Boiss. 4, appendix 2, p. 213, 1896 and Acacia nubica Bth.

is sunk under it. This is a species quite distinct from the “ Cape Acacia

Harvey and Sonder’s use of the name Acacia horrida Willd. in the Flora Capensis

is therefore also incorrect since Willdenow’s name must be restricted to Linnaeus’

specimen and cannot be applied to the Cape species.

Acacia capensis (Burm.) Burchell. In his “Travels” Vol. 1, p. 114, Burchell

mentions the Karroo-thorn-tree at Genadendal and gives it the name Acacia capensis B.

but he does not give a Latin diagnosis as is his custom with new species. This suggests

that it may have been a new combination but nowhere does Burchell confirm this.

On page 189 he describes the Acacia unmistakably and there is a drawing of it at the

head of the chapter commencing on that page. In spite of the absence of a Latin

description, this might be considered legitimately published under the provisions of

articles 37 and 43 of International Rules of Botanical Nomenclature, where it is stated

that in certain circumstances a plate with analysis is accepted as equivalent to a

description (if the plates and names were published before Jan. 1, 1908). But in the

face of Bentham’s statement in Mimosae p. 507 this cannot be looked upon as a new

413

species but as a new combination. Bentham, who worked with Burchell on his

Leguminosae, writes “ Burchell adopted the name A. capensis for this species, supposing

it to be the Mimosa capensis Burm. Cap. p. 27, which it probably is.” Burchell’s name

is therefore the combination with Acacia of Mimosa capensis Burman and must be

rejected.

Acacia capensis Colla. This is a doubtful species because (a) the type is a seedling

grown from seed which the donor averred was that of Acacia capensis of the Cape of

Good Hope, and ( b ) the description is based on Burchell’s description of his Acacia

capensis. In any case the name is redundant because as shown above the combination

with “ capensis ” had already been published.

Mimosa eburnea Bojer. The Cape species was introduced into Mauritius and

became naturalised there. It was erroneously identified as M. eburnea L.f. by Bojer

and listed in his Hort. Maurit. 115 (1837).

Acacia karroo Hayne is, therefore, the correct name for the “ Cape Acacia ”.

415

Newly Described Species.

By

L. Britten, A. Brueckner, H. B. Christian, R. A. Dyer,

M. D. Henderson, D. J. B. Killick, R. Story,

I. C. Yerdoorn and H. Wild.

ASCLEPIADACEAE.

Tavaresia meintjesii R. A Dyer, sp. nov., affinis T. angolensi Welw. corollae tubo

breviore lobis longiorib usque, coronae exterioris lobis facile distinguitur. i

Planta parva, succulenta cactiformis e basi ramosa. Caules 8-10 cm. alti,

1-2-1 -5 cm. crassi, 6-8-angulati, angulis tuberculatis spinosis, spinis pubescentibus 4-5

mm. longis, plus minusve hastatis setis lateralibus 1-5-2 mm. longis. Flores pauci

cymis sessilibus ramorum basin versus emittentes; pedicelli 1-2 cm. longi, puberuli.

Sepala lanceolata vel lineari-lanceolata, 1-1 -2 cm. longa, puberula. Corolla 7-7-5 cm.

longa, infra medium tubulosa tubo obconico 3-3 • 5 cm. longo 2 • 5 cm. lato pilis longis

induto, lobis triangulari-lanceolatis circiter 4 cm. longis, minutissime pubescentibus,

pilis usque 3 mm. longis ciliatis. Coronae exterioris lobi basi breviter connati, 9-10 mm.

longi profunde 2 vel rariter 3 segmentis linearibus supra medium recurvis divisi.

Coronae interioris loborum cornu exteriore 2-3 mm. longum, interiore filiforme, 7-8 mm.

longum.

Transvaal. — Zoutpansberg Distr.: 40 m. east of Messina near n’Wanedze River,

Meintjies (ex Steenkamp ) 523 ; about 20 east of Messina near n’Jelele River, Meintjes

(ex Myburgh) 663, cult. Johannesburg (type); Pietersburg Distr.: Mariepskop among

rocks on koppie, Lubbe in Nat. Herb. Pretoria 28520.

Plate 1.

A tufted dwarf perennial succulent, leafless and spinose. Stems branching from

near the base, erect, 8-10 cm. tall, 1-25-1-5 cm. thick, glabrous or very minutely

pubescent, angled; angles 6-8, tuberculate; young tubercles terminated by an

ascending, pubescent, hastate-shaped bristle; bristle with a central lanceolate lobe

4-5 mm. long and with the two sharply pointed lateral lobes 1-5-2 mm. long, with

margin and keel thickened. Flowers few in sessile cymes from near the base of young

branches, developed successively; pedicel 1-2 cm. long, puberulous. Sepals lanceolate

or linear-lanceolate, 1-1-2 cm. long, puberulous down the central portion of the back.

Corolla 7-7-5 cm. long, tubular at the base, cream-coloured within with spots and

transverse bands of maroon, and maroon on the margins; tube obconical, 3-3-5 cm.

long, 2-5 cm. wide at the mouth, long-pubescent, especially tov/ards base; lobes

triangular-lanceolate, about 4 cm. long and about 1 • 5 cm. broad at base, spreading,

minutely pubescent on surface with spreading white to red cilia 3 mm. long on margin.

Outer corona maroon, 9-10 mm. long, united into a tube 2 mm. long at the base, each

lobe deeply divided to give 10 linear segments or occasionally one or more of the lobes

may produce a third central filiform segment; segments erect from the base and

spreading-recurved above and slightly thickened towards apex. Inner corona lobes

416

maroon, united at the base to the tube of the outer corona, 2-horned; outer horn

erect-spreading, 2-3 mm. long; inner horn filiform, erect, 7-8 mm. long.

The first record of this species in the National Herbarium dates back to 1943

when Mr. B. Meintjes of Johannesburg submitted a specimen for identification. The

flower was so different in its proportions from other species of Tavaresia that doubt

arose whether the flower was normal or not. Eventually a second specimen was located

in the same area and when this too flowered in Mr. Meintjes collection in Johannesburg

it proved identical to the first. In addition a specimen, which was received from the

Petersburg district, agrees sufficiently closely to the other two to be included in the

same species. The species is remarkable in the genus for the shortness of the corolla

tube by comparison with the length of the lobes, and the outer corona lobes, while

slightly thickened at the ends, do not have the knob-like tips of the three other known

species. In vegetative character it shows a close similarity to T. angolensis Welw.

In the type locality in the Limpopo River Valley, east of Messina, Mopane veld

[Colophospermum mopane (Kirk ex Benth.) Leonard] is dominent. Other succulent

plants in the same area include Stapelia gigantea N.E.Br., S. getleffii Pott, Caralluma

rogersii Bruce & Dyer and Monadenium lugardae N.E.Br.

CHENOPODIACEAE.

Atriplex erosa Brueckner and Verdoorn sp. nov., ab A. semibaccatae annuls foliis valde

incisis marginibus involutis bracteis fructiferis latioribus plerumque tuberculatis

dififert. — A rosea Wright in FI. Cap. 51, p. 446 quoad Bolus 656, non Linn;

A tartarica Paul Aellen Bot. Jahrb. 70, 1940, non Linn.

Herba sub-lignosa, annua, erecta vel decumbens usque ad 50 cm. alta, ramosa;

rami angulati furfuracei. Folia alterna, subsessilia vel breve petiolata, ovata vel ovato-

lanceolata, 0-6-3 cm. longa, 0-4-1 -5 cm. lata, supra leviter subtus dense furfuracea,

marginibus involutis valde incisis apicibus acutis basibus cuneatis. Flores monoici,

glomerati; glomeruli axillares et in spicas terminales dispositi (folia subtendentia

apicem versus valde reducta); superiores floribus maculinis atque paucibus ad basin

femineis, inferiores floribus femineis.

Flores masculi: lobi perianthii membranacei, dorso carnosi, obovati, apice obtusi

basi connati c. 0. 1 cm. longi; stamina 5, lobis opposita, filamentis brevibus planis basi

connatis, antheris dorsifixis. Flores feminei: 2-bracteati bracteis sub-carnosis

tuberculatis; perianthium nullum; ovarium compresso-globosum; styli 2, filiformes,

basi connati; ovulum 1. Bracteae fructiferae 2, late rhomboideae vel sub-rhomboideae

circiter 4 mm. longae, 3 mm. latae, fere medio connatae, basi cuneatae, apice

subrotundae vel late triangulares, margine irregulariter dentatae, dorso leviter pilosae

lepidotae distincte tuberculatae. Semen compresso-orbiculatum, circiter 2 mm. longum,

circiter 1-7 mm. latum, nitidum, fuscum.

Cape Province. — Middelburg Division; Grootfontein, Gill 45; Kimberley

Division; Wesselton, Wilman no num. in McGregor Museum Herb ; Alexandersfontein,

Burtt Davy 9561; Brueckner 805 (type) and 806; Aliwal North Division: Aliwal North,

Derrington 2560.

Orange Free State. — Bloemfontein Division; Bloemfontein, Potts in Herb.

Grey University College, 2891; Fauresmith Division; Ventersvlei, Verdoorn 1159;

Wheeldon, Henrici, 2440; Jagersfontein, Smith 5171; Knoffelsfontein, Smith 5392;

Acocks 13506.

5.

Fig. 1. — Atriplex erosa Brueckner and Verdoorn, 1, apical portion of flowering

branch ; 2, male flower, x 5; 3, rudimentary ovary in base of corolla tube x 5; 4,

anther x 10; 5, rudimentary ovary x 5; 6, fruiting bracts separated, the lower showing

the seed within, x 10; 7, larger mature fruiting bracts intact, X 10.

Annual herb somewhat lignified, erect or decumbent (erect plants with a single,

erect, branched stem up to 50 cm. high: decumbent usually with several decumbent

stems radiating from the tap root). Branches angular lepidote, greenish-grey to reddish-

grey. Leaves alternate, subsessile or shortly petioled, ovate to ovate-lanceolate,

0-6-3 cm. long, 0-4-1 -5 cm. broad, ashy or mealy lepidote on both surfaces, sparsely

so on upper, densely on lower surface, midrib prominent beneath, margins involute

and deeply incised, apex acute and base cuneate. Inflorescence with flowers clustered

into close terminal and axillary glomerules, the subtending leaves much reduced towards

the apex; clusters of male flowers, with a few female flowers at the base, in the axils

of the upper leaves and female flowers in the lower. Male flowers with 5-merous

perianth, the lobes membranous, thickened dorsally, obovate, obtuse, more or less

0-1 cm. long, joined at the base into a short tube; stamens 5, opposite the perianth

lobes, filaments short, flattened, connate at the base; anthers dorsifixed, shedding

418

pollen through- vertical slits; ovary rudimentary, conical. Female flowers with two

compressed fleshy, usually tuberculate bracts ; perianth o; ovary globose, compressed;

styles 2, filiform, connate at the base; ovule 1, surrounded by a delicate pericarp.

Fruiting bracts 2; broadly rhomboidal or sub-rhomboid more or less 4 mm. long and

3 mm. broad cuneate at the base, connate for about their length, the free apices

semi-orbicular to broadly triangular and irregularly toothed, the outer surfaces slightly

pilose-lepidote, with conspicuous wart-like tubercles. Seed round, compressed, more

or less 2 mm. long and 1-7 mm. broad, smooth, shiny, dark brown.

The distribution of this plant is, according to existing records, limited to portions

of the southern Orange Free State and the northern Cape Province where it flourishes

in somewhat moist and brackish denuded localities. Its occurrence in restricted patches

is evidently on account of it not being relished by stock, so it grows fairly vigorously

and is propagated by seeds which are not carried abroad. The nature of its distribution,

and the habitat leads us to assume that it is indigenous.

In South Africa the identification of Atriplex species has been rendered difficult

by the large scale introduction of Australian species as fodder plants in the more arid

and saline areas. Some of these have become naturalised and hybridization takes place.

Dr. Paul Aellen in his “ Atriplex und Blackiella in Sudafrika ” Bot. Jahrb. 1940,

placed specimens which belong to the species described here under Atriplex tartarica L.

but recently, on seeing the specimens quoted here, he agreed that they are distinct

from the European species and should be described as intended by the present authors.

We greatly appreciate Dr. Aellen’s help and advice as he is an authority on the world’s

Chenopodiaceae whereas we know only our local representatives of the family.

Atriplex suberecta Verdoorn sp. nov., A. muelleri affinis sed bracteis in floribus femineis

apicibus deltoideis basi pseudo-stipitatis incrassatis foliis non truncatis differt.

A. muelleri Paul Aellen in Bot. Jahrb. 70, 1940 p. 390 non Benth.

Herba annua vel perennis, basi sub-lignosa, usque 75 cm. alta, plus minusve pilis

diaphanis squamuliformibus obtecta; rami decumbente-erecti, usque 1 m. longi. Folia

oblonga, ovata vel rhomboidea, basi cuneata, breve petiolata, crasse et irregulariter

sinuato-dentata vel lobata,l-5 cm. longa, 0-5-3 -5 cm. lata, plus minusve pilis diaphanis,

squamiformibus obtecta. Flores monoici glomerati; glomeruli axillares, floribus

masculinis atque ad basin paucibus femineis in axillis superioribus, femineis in

axillis inferioribus dispositis. Flores masculini: perianthium breve tubulatum,

plerumque 5-lobatum, laete viridum nonnumquam roseum; lobi c. 1-5 mm. longi;

stamina plerumque 5 lobis perianthii opposita filamentis basi connatis; antherae

apud apicem dorsifixae. Pistillum rudimentum. Flores feminei: bracteae compressae c.

4 mm. longae, 3-5 mm. latae, basi connatae rhomboideae pilis farnaceis obtectae

apicibus deltoideis dentatis basi pseudo-stipitatis incrassatis solidis. Ovarium

compresso-globosum pericarpis membranaceis ; styli 2; ovulum 1; testa Crustacea.

Fructus sub-globosus, compressus c. 3-5 mm. longus, 3 mm. latus, basi incrassatus

induratus-que aliquantum osseus apicibus deltoideis bractearum persistentibus. Semen

compresso-orbiculatum nitido-bruneum.

Cape Province. — Oudtshoorn Div Oudtshoorn, Gill 37; du Plessis no num.

Albany Div.: “Lakeside”, Lindstedt 24; Grahamstown, Story 3741. Fort Beaufort

Div.: Fort Beaufort, Taylor 1. Graaff Reinet Div.: Graaff Reinet, Sister Tarcisia 10.

Middelburg Div. ; Grootfontein, Verdoorn 1739; Theron 47,48,79. Stutterheim Div. :

Bolo, Acocks 9699. Tarka Div.: Tarka Conservation Area, Whitlock 3. Calvinia

Div.: Calvinia, Smith 2463; Schmidt 398. Hanover Div.: Hanover, Herb. Univ.

Witwatersrand 24925. Aliwal North Div.: Aliwal North, Gerstner 157B; Henrici

419

3053. Hopetown Div. : Olifantskop, Blcinkenburg 2. Kimberley Div. : Kimberley,

Acocks 143 ; Brueckner 772. Barkly West Div.: Boetsap, Pagan no num. Hay Div. :

Postmasburg, Esterhuysen 1120. Prieska Div.: Prieska, Bryant 206.

Orange Free State. — Fauresmith Div.: Fauresmith, Veld Reserve, waste lands,

Henrici 3854 (type), 3860, 3860, 3855, 3855a; Verdoorn 1194; Breuckner 812; Schansen,

Verdoorn 2206; Jagersfontcin, Smith 5229. Bloemfontein Div. : Glen College, Mostert

1, 2, 3, 4, 700.

South West Africa. — Luderitz Div. : Luderitz, Kinges 2020. Swakopmund Div. :

Swakopmund, Moss 6967 ; Rogers 29708; Bradfield 576; Nonedas, Bradfield 472;

Palmerhorst, Wiss 953. Bulspoort, Liebenberg 5122.

Transvaal. — Pretoria Dist. : Pretoria, Skanskop, Repton 1098. Benoni Dist.:

Benoni, Bradfield 293, Bradfield 125.

Fig. l.—Atriplex suberecta Verdoorn, 1, apical portion of flowering branch;

2, bracts enclosing female flower, x 5; 3, fruiting bracts beginning to swell and

indurate at stipe-like base, X 9 ; 4, maturing fruit with bracts completely indurated except

for toothed apical portion, x 9; 5, one bract removed showing the gynoecium, x 9.

420

i. 2. 3 4

Fig. 3. — 1, Fruiting bracts of A. semibaccata R. Br. ; 2, fruiting bracts of A. suberecta

Verdoorn; 3, fruiting bracts of A. muelleri Bth.; 4, fruiting bracts of A. erosa Brueckner

& Verdoorn.

An annual or perennial herb, sub-woody at the base, forming a bushy growth

up to 75 cm. tall, more or less scaly with diaphanous swollen-based hairs (which

become scale like in dried specimens); branches decumbent erect, up to 1 m. long.

Leaves very variable in size, shape and dentation, oblong, ovate or rhomboidal, cuneate

into a short petiole, coarsely and irregularly sinuate-dentate or lobed, 1-5 cm. long,

0-5-3 -5 cm. broad, more or less mealy pubescent on upper surface and -scaly on

lower surface. Flowers monoecious, in glomerules, the males, with a few females at

the base, aggregated in the axils of the upper leaves and the females in the axils of the

lower leaves. Male flowers pale green with a dark green keel (not seen on dried specimens)

or occasionally suffused with red (red may be present in some flowers on the same plant) ;

perianth tube short, lobes usually 5, about 1 • 5 mm. long, cucullate. Stamens usually 5,

opposite the perianth segments, filaments linear in dried specimens, terete and turgid

in fresh, united at the base about 1 mm. long; anthers attached near the apex, the cells

free below. Rudimentary ovary present in the base of the perianth tube. Female

flowers bi-bracteate; bracts flat, enclosing the ovule to above the middle, cuneate below

and with a stipe-like base which becomes solid and swollen as the fruit matures; apical

lobes with a large central deltoid tooth and with 1 or more smaller teeth on each side,

pubescent with diaphanous swollen hairs, 3-nerved from the base, nerves disappearing

as the fruit matures and the swelling which starts at the stipe-like base spreads upwards;

ovary compressed globose with a membranous pericarp; styles 2; ovule 1, testa

crustaceous. Mature fruit enclosed in the persistent bracts, sub-globose, compressed,

becoming swollen and indurated from the stipe-like base until all but the deltoid apex

of the bracts becomes smooth and somewhat bone-like, usually 3-5 mm. long and

3 mm. broad. Seed compressed globose, brown, shiny.

This species, which is very common in a wide area in South Africa, has for long

been identified with A. rosea of the Flora Capensis (Vol. 5, p. 44) probably because

it runs to that species in the key. In that work, however, the only specimen from

South Africa cited by Wright, Bolus 656, as represented in the Bolus Herbarium, is

the species described above as A. erosa Brueckner and Verdoorn (it is certainly not

A. rosea L.).

In the article “ Atriplex und Blackiella in Siidafrika ”, Bot. Jahrb. 1940, Paul

Aellen cites several specimens, which are here described as A. suberecta, under A.

muelleri Bth. After seeing a large collection sent to him from the National Herbarium

in 1952 Dr. Aellen wrote that he is now convinced that the South African specimens

he had placed under A. muelleri are distinct from that species and should be described

as a new species. He added that the opportunity of seeing the S. African material

and notes had made him realize that similar plants occur in Europe and in Australia

and that he had considered them to be the juvenile form of A. muelleri Bth. (see “ Die

in Europa wolladventiv beobachteten Australischen Atripliceen ” by Paul Aellen in

Verh. Naturf. Gesellschaft in Basel Vol. 50 p. 159 (with figure) and “ Revision der

Australischen und Neuseelandischen Chenopodiaceen ” in Bot. Jahrb. 68, p. 369).

Since I have not been able to examine the specimens either from Europe or

Australia, only the South African plants are dealt with here. I am grateful to Dr.

Aellen for his help and advice and for encouraging us to publish the two new species.

421

It is hoped that in his future work on the family, as the authority on Chenopdiaceae,

he will take the matter further.

The following key should help to distinguish the two new species A. erosa and

A. sube recta from their nearest relatives in South Africa, A. muelleri Bth. and A. semi-

baccata R. Br., the Creeping Salt Bush. The two last-mentioned have been introduced

into S. Africa in recent times. The group is characterised by the bracts of the female

(lowers being compressed, more or less rhomboid in outline or at least cuneate at the

base and united to about the middel, persisting and thickening with the fruit which is

usually about, or under, 5 mm. long. The flowers are in glomerules, those in the axils

of the upper leaves consisting of male flowers with a few female flowers at the base and

those of the lower leaves purely female. The plants are procumbent or decumbent

to sub-erect, usually with many stems radiating from the tap root but in some specimens

of A. erosa with a single erect stem.

It must be borne in mind that these species hydridise easily and since all but

A. erosa are relished by stock they become widely spread and the introduced species,

or their hybrid progeny, may be met with away from the areas where they were

introduced. In this way it is difficult to tell whether a species is indigenous or not.

KEY.

Subtending leaves much reduced towards the apices of the branchlets giving the inflorescence a

spike-like appearance; fruiting bracts very broadly sub-rhomboid in outline, usually

tuberculate; leaves deeply incised with involute margins A. erosa.

Subtending leaves not reduced towards the apex of the branchlets, inflorescence distinctly axillary;

fruiting bracts never tuberculate:

Plants prostrate; leaves usually under 2 cm. long and 6 mm. broad; fruiting bracts

distinctly rhomboid in outline with deltoid apical lobes, usually entire or with 2 small

teeth near the base; fruit becoming semi-baccate when ripe and often turning red

or orange; female flowers 1 to few in the axils of the lower leaves A. semibaccata.

Plants decumbent-erect or erect; fruit never becoming baccate when ripe but more or

less boney; female flowers many in the axils of the lower leaves:

Leaves usually truncate at the apex; fruiting bracts with short, more or less

rounded and toothed apical lobes; fruit not stipitate, indurating as a

whole A. muelleri.

Leaves rounded at the apex; fruiting bracts with broadly deltoid apical lobes,

toothed, the middle tooth the largest and deltoid; fruit with a stipe-like

solid base which is the first to swell and indurate A. suberecta.

COMPOSITAE.

Helichrysum coactum M. D. Henderson sp. nov., affinis H. umbraculigero Less, sed

caulibus non ramosis tenuioribus, foliis angustis linearibus basi non attenuatis

utrinque dense lanatis concoloribus et non fasciculatis differt.

Herba erecta, radicibus fibrosis. Caules 40-45 cm. alti non ramosi, graciles, dense

albo-lanati, foliati. Folia linearia, 3-3-5 cm. longa, 3 mm. lata, basi non attenuata,

utrinque dense albo-lanata, concolora. Capitula 3-4 mm. longa, homogama, 5-flora,

cymis densissimis terminalibus planis usque ad 2 cm. diam. disposita. Involucri

bracteae nitido-flavae. Receptaculum planum. Flores hermaphroditi. Achaenia

glandulosa; setae 1-2.

Natal. — Bergville Div. : Cathedral Peak Forest Research Station Catchment 1,

at 6,050 ft. Jan., Killick 1354.

Note. — Mr. B. de Winter reports that Hutchinson 4551 and 4593 in Kew

Herbarium belong to this species.

422

Herbaceous plants. Roots numerous, fine, fibrous. Stems 40-45 cms. high,

unbranched, slender, erect, several growing together, densely white-woolly especially

at the apex, becoming glabrous at the base, leafy. Leaves linear, 3-3-5 cm. long,

3 mm. broad, very slightly or not narrowed to the base, amplexicaul, sometimes shortly

decurrent; densely whitish-woolly on both surfaces concolorous; midrib conspicuous

on undersurface. Heads 3-4 mm. long, homogamous, 5-flowered, terminal in very

dense flat-topped cymes (up to 2 cms. across); peduncles felted together as in

H. umbraculigerum Less. Involucral bracts bright yellow, few, in 3-4 rows, erect,

oblong or more or less spathulate, truncate, longer than the florets. Receptacle smooth.

Florets hermaphrodite, 1-5-2 mm. long; apex of corolla lobes with thickened margins,

Achenes glandular; pappus of 1-2 caducous bristles.

These plants are to be found growing in vleis on the Drakensberg in Natal and

are reported to match two specimens in Kew collected by Hutchinson. The heads

are very similar to those of H. umbraculigerum Less but the leaves differ considerably,

those of H. umbraculigerum being more or less ovate and much attenuate to the base,

discolorous and sometimes fasciculate, while those of H. coactum are linear, not or

but very slightly narrowed to the base, concolorous and never fasciculate.

Helichrysum grandibracteatum M. D. Henderson, sp. nov., affinis H. randio S. Moore,

sed planta tota caniore, capitulis dense corymbosis, pedunculis brevioribus,

involucri bracteis acutioribus nitido-albis et floribus multo longioribus differt.

Caules erecti, 15-20 cm. alti, e rhizomate perenni lignes orti, basi dense foliati.

Folia radicalia et inferiora caulina linearia, acuta, 6-7 cm. longa, 0-3 cm. lata, basi

non attenuata, infra prominentia multi-nervosa ; folia caulina superiora 2-4 cm. longa,

lineari-lanceolata, acuta, sessilia; folia omnia infra albo-lanata, supra viridia (interdum

purpurata) et glabra. Inflorescentia compacta terminalis corymbosa. Capitula late

obconica, 1 cm. longa, 1-5-2 cm. lata, homogama, flores flavi. Involucri bracteae

nitido-albae, acutae; bracteae interiores flores duplo longiores. Achaenia breviter

hirsuta. Pappi setae numerosae, tenues, liberae, basi pilis patentibus intermixtae.

Natal. — Bergville Div. : Cathedral Peak Forest Research Station 6,600 ft., Nov.,

Killick 1149 (Type); Mont-aux-Sources, Nov., Schweickerdt s.n., 8,000 ft., Oct., Sidey

2004.

Cape. — Mt. Currie, Nov., Goossens 311; Victoria East, 5,500 ft.. Nov.-Dec.„

Dyer 780.

Note. — Mr. B. de Winter reports that Baur 541 and McOwan 2091 in Kew

Herbarium also belong to this species.

Stems 15-20 cm. long, sometimes shorter, many arising from a perennial, woody,

sometimes divided, rootstock, white-woolly, striate, densely leafy at the base. Leaves

white-woolly on undersurface, green and more or less glabrous on upper surface,

many-nerved ; nerves prominent on under surface ; radical and lower cauline linear,

acute, 6-7 cm. long, about 3 mm. broad not narrowed to base; upper cauline similar,

2^1 cm. long, linear-lanceolate, acute, sessile, not narrowed to the base, sometimes

purplish on upper surface when dry. Inflorescence a compact few-headed terminal

corymb. Heads shortly pedunculate, homogamous, many-flowered, broadly obconic,

1 cm. long, 1-5-2 cm. wide at the top when pressed. Involucral bracts shiny-white,

glabrous, in many rows; outer shorter than inner, lanceolate, acute imbricate; inner

almost twice as long as florets. Receptacle honeycombed with margins of pits shortly

produced. Florets hermaphrodite and fertile. Corolla yellow, 5 mm. long; tube

sub-cylindric, very slightly widened upwards, shortly 4-5-lobed. Achenes shortly

hirsute, angled when young, becoming more or less cylindric when mature. Pappus

of numerous fine bristles about as long as corolla; bristles free, but with intermingling

patent hairs at the base, apex with looped hairs.

423

The earliest collection in the National Herbarium is that made by R. A. Dyer in

1926 on the main ridge towards The Hogsback in the Victoria East Division. The

specimen was labelled “ near H. argentissimum ” from which species, however, it is

quite distinct. Subsequent collections were matched with this specimen. The species

occurs in grassveld at high altitudes in the mountains of Natal and eastern Cape.

Although resembling H. randii S. Moore, this species is readily distinguished by the

corymbose heads with long acute shiny-white involucre bracts and the more grey colour

of the leaves and stems. H. argentissimum J. M. Wood has large solitary white heads

and densely woolly-tomentose leaves.

Helichrysum scopulosum M. D. Henderson, sp. nov., affinis H. fulgido Willd., sed foliis

basalibus dense rosulatis molliter et dense lanatis late obovatis differt.

Herba 8-30 cm. alta. Folia basalia numerosa, rosulata, sessilia, obovata, obtusa,

dense albo-lanata, 1-10 cm. longa, 1-4 cm. lata. Peduncu/i erecti, graciles, e rosula

orti, dense albo-lanati, foliis parvis 2-4 cm. longis et 0-5 cm. latis lanceolatis acutis

sessilibus et lanatis interdum foliis inferioribus oblongis. Capitula heterogama,

multiflora, floribus marginalibus femineis paucis. Receptaculum plusminusve planum,

alveolatum. Achaenia glandulosa. Pappi setae paucae liberae.

Natal. — Bergville Div. : Cathedral Peak Forest Research Stn., Catchment 2,

7,400 ft., Nov., Killick 1554 (Type); Cathedral Peak 8,000-9,000 ft., July, Esterhuysen

10228 ; 12864; Mont-aux-Sources, Hutchinson, Forbes & Verdoorn 66; Weenen

Div.: MnWeni Drakensbergen 8,000-9,000 ft., July, Esterhuysen 14533; National

Park Area, Drakensbergen 8,000 ft., July, Esterhuysen 15552, Saddle Rockeries

Drakensbergen 8,000 ft., July, Esterhuysen s.n.; Ndedena area, Drakensbergen 8,000 ft.,

July, Esterhuysen 17346.

Herbaceous plants 10-40 cm. high. Stem erect, sparingly branched or simple,

densely clothed with leaves at the base, slender above, densely white woolly with wide-

spaced leaves. Leaves densely and softly, whitish-woolly on both surfaces at the base

becoming thinly woolly or setose on the slender part of the stem; basal crowded on

the stem to form a cushion-like rosette on the ground, sessile, obovate, obtuse, 1-10

cm. long, 1-4 cm. broad, venation inconspicuous; upper lanceolate, acute, 2-4 cm.

long, 0-5 cm. broad, sessile not narrowed to the base. Heads large, 3-4 cm. broad,

solitary, terminal, heterogamous with a few marginal female florets and numerous

hermaphrodite florets. Involucral bracts bright yellow, radiating, glossy, lanceolate,

acute, inner shorter than median, outermost foliaceous, woolly. Receptacle flat or

slightly convex, honeycombed. Style sometimes with a globose swelling at the base

within the corolla tube. Achenes glandular. Pappus bristles few, free.

This species has been collected only on the Drakensberg of Natal where it is

common on steep rock faces with its roots in crevices. Specimens have remained

unnamed in the National Herbarium for some time with the exception of the one

collected by Hutchinson, Forbes & Verdoorn in 1930 which was at first called

H. fulgidum Willd, which species the new one most closely resembles. It is, however,

readily distinguishable by the . oft wolliness and cushion-like habit of the lower rosette

of leaves. The heads produced by H. fulgidum are usually smaller than those produced

by H. scopulosum, but are otherwise very similar.

Helichrysum tenax M. D. Henderson sp. nov., affinis specibus Polylepideae, sed foliis

inferioribus (30 cm. sub inflorescentia) multo majoribus et foliis superioribus

oblongis non lanceolatis differt.

Planta herbacea, basi semi-lignea, 90 cm. alta. Folia glanduloso-setosa, utrinque

viscida, concolora; inferiora magna, 12-20 cm. longa et 4-8 cm. lata, obovata, sessilia,

obtusa, superiora parva, 7 cm. longa et 2 cm. lata infra inflorescentiam ad usque

3 cm. longa, oblonga, sessilia, amplexicaulia. Inflorescentia terminalis corymbosa

424

vel paniculata, pendunculis albo-lanatis, capitula heterogama, discoidea, 1*5-2 cm.

diam. Invo/ucri bracteae nitido-flavae, splendidissimae. Achaenia glandulosa. Pappi

setae paucae, caducae.

Natal. — Bergville Div. : Cathedral Peak Forest Research Station, 5,800 ft.,

Dec., Killick 1632 (Type); Dec., Codd & Dyer 6249; Cathedral Peak, Nov., Bayer-,

Top of Mont-aux-Sources 11,000 ft., March, Evans 542. Estcourt Div.: Cathkin

Peak Hostel 6-7,000 ft., West 26; Pasture Research Station 4,300 ft., Oct; Acocks

9847; Weenen Div. : Draycott Hill, 4,300 ft., June, Acocks\ Currie Div. : Kokstad,

Dec., Phillips 3493.

Basutoland. — Maluti Mts.: 7-8,000 ft., Staples 86.

A herbaceous much branched plant with a semi-woody base, 90 cm. high. Stem

more or less white woolly pubescent especially in the upper part, ribbed, with a pith

in the centre. Leaves glandular-setose, viscid on both surfaces, concolorous; lower

large, 112-20 cm. long, 4-8 cm. broad on the average, obovate, sessile, apex more or

less obtuse, midrib and veins hairy beneath, internodes very short; at approximately

20-30 cm. below the apex leaves abruptly reduced in size from 7 cm. long, 2 cm.

broad to less than 3 cm. long under inflorescence, oblong, broadly acute or obtuse,

sessile, auriculate, sometimes very shortly decurrent, internodes longer, midrib sometimes

slightly woolly at the base beneath. Inflorescence a terminal corymb or panicle with

very woolly peduncles. Heads heterogamous, discoid, 1 • 5-2 cm. in diameter with

one row filiform female florets surrounding very numerous hermaphrodite florets.

Involucral bracts bright yellow, glossy, radiating, inner shorter than median, outer

small, woolly on outside. Receptacle honeycombed, flat or slightly convex. Corolla

of female florets inflated in lower half, upper half narrow cylindrical with spreading

triangular lobes ; of hermaphrodite florets broader than female corolla, slightly inflated

in lower half, upper half gradually widened to the triangular lobes. Anthers with

Tong acuminate basal tails, apex with a short obtuse appendage. Style branches truncate,

glandular at the apex. Achenes obscurely angled, glandular. Pappus bristles few,

caducous.

The first record of this species in the National Herbarium is a specimen collected

by M. S. Evans in 1898 at the top of Mont-aux-Sources. It was originally identified

as H. setosum Harv., which species, in the Polylepidea group, it most closely resembles.

The new species is readily distinguishable from all members of the above group by

the very large sticky leaves on the lower part of the stem and the oblong, not lanceolate,

upper leaves. The plants are bushy, reported to be up to seven or more feet high and

are common on hillsides in the mountainous regions of Natal, Basutoland and the

eastern Cape. Glandular hairs on the leaves trap and hold fast small insects which

alight on the surface, hence the specific name. The plants were thought to be insecti-

vorous when the remains of the bodies of these insects were first noticed in 1945 but

no further evidence has yet been produced.

Helichrysum tenuifolium Killick, sp. nov., affinis H. trilineato DC., sed habitu robustiore,

foliis longioribus tenuioribus et acutioribus differt.

Frutex ligneus, 0*9-2 m. altus. Rami albo-lanati maxime apicibus. Folia patentia,

sessilia, linearia, 0*7-1 *8 cm. longa, 1-1*5 mm. lata, revoluta, 3-venata; apex acutus,

recurvatus, mucronulatus. Capitula campanulata, 3*5-5 mm. longa, 3-4*5 mm.

lata, apicibus ramorum sub-corymbosa. Pedunculi 1-5 mm. longi, albo-lanati. Bracteae

involucri 5-seriatae, exteriores albae apicibus saepe rubicundae, intimae flavae. Recep-

taculum leviter convexum, fimbrillatum. Flores 20-30, flavi; marginales 9 filiformes;

interiores ^ tubulosi, limbo ampliato. Achaenia breviter pubescentia.

Natal. — Bergville District: on the banks of the Mlambonjwa River, Cathedral

Peak Area, Schelpe 960; on the boulder-bed of the Inhlwasine River, Cathedral Peak

Forest Influences Research Station, Killick 1438, 1591 (type).

425

A robust, woody shrub, 0-9-2 m. high. Branches white-woolly especially at the

ends, older portions rough with persistent leaf bases. Leaves patent, sessile, linear,

0-7-1 -8 cm. long, 1-1-5 mm. broad, revolute; apex acute, recurved, mucronulate;

upper surface with 3 parallel veins, glabrous to sparsely hairy, lower white-woolly.

Capitula campanulate, 3-5-5 mm. long, 3-4-5 mm. broad, arranged in dense sub-

corymbose terminal cymes. Peduncles 1-5 mm. long, white-woolly. Involucral bracts

5-seriate; the innermost linear-spathulate, reflexed at the tips, yellow, glabrous; the

outer successively shorter, lanceolate, white-woolly, often salmonrpink at the tips.

Receptacle slightly convex, fimbrillate. Florets 20-30, yellow; marginal $ filiform

inner tubular, widening above. Achenes minutely pubescent.

FI. tenuifolium is a robust shrub, 3-6} ft. high with darkish-green, linear leaves

and showy sub-corymbose cymes of yellow heads. It flowers from November to

February. Closely related to H. trilineatum DC., it differs in its taller habit and in the

longer, narrower and acuter leaves.

Senecio brevidentatus M. D. Henderson, sp. nov., affinis; S. tugelensi Wood & Evans,

sed inflorescentia paniculata, caulibus basi pilosis, marginibus foliorum breviter

dentatis non tenuiter serratis differt.

Herba erecta, 25 cm. alta; rhizoma fibrosum. Caules basi parce pilosi. Folia

radicalia oblanceolata, obtusa, 6-7 cm. longa, 1-1-5 cm. lata, attenuata, marginibus

breviter dentatis; folia inferiora caulina oblonga, obtusa vel plusminusve acuta, 2-5

cm. longa, 0-5-1 cm. lata, sessilia, auriculata, marginibus breviter dentatis et ciliatis;

folia suprema parva, lanceolata, ad 2 cm. longa, acuta auriculata, marginibus integris

ciliatis. Inflorescentia paniculata; capitula radiata, flava ample calyculata; involucri

bracteae spice nigrae. Achaenia juvenes leviter pilosa.

Natal. — Bergville Div. : Cathedral Peak Forest Research Station in vlei in

Catchment 1, 6,050 ft., Nov., Killick 1152 (Type). Note: Mr. De Winter at Kew

reports that Wood 1010, 391, 162 in Kew Herbarium belong to this species.

Herb about 25 cms. high, erect, simple or branching above the base; rootstock

fibrous; roots robust; Stems striate, sparsely covered with white hairs at the base,

becoming glabrous. Leaves glabrous or with ciliate margins; radical oblanceolate,

obtuse 6-7 cm. long, 1-1-5 cm. broad, narrowed to the base; margins shallowly

dentate; lower cauline oblong, obtuse or more or less acute, sessile, 2-5 cm. long

0-5-1 cm. broad auricled at the base but not stem clasping, margins shallowly dentate,

ciliate; upper cauline reduced in size, lanceolate, acute, auricled at the base 2 cm.

long, margins entire, ciliate. Inflorescence a lax panicle; penduncles elongate bearing

minute lanceolate bracts. Heads 1 cm. long, radiate. Involucre of about 20 linear,

glabrous involucral bracts with minute black tips, calycled, with 8-9 linear basal bracts.

Ray florets 10-12 mm. long, limb longer than tube. Disc florets numerous, apices of

lobes more or less obtuse with a minute glandular cushion on the outside. Anthers

with a short obtuse apical appendage. Achenes slightly hairy when young. Pappus

of numerous fine barbellate bristles.

Although there is only one collection of this species in the National Herbarium

it is reported to match three unnamed specimens in Kew collected by Wood, probably

also from Natal. It was reported to be fairly common in a vlei on the Drakensberg in Natal.

The species most closely resembling S. brevidentatus is S. tugelensis Wood & Evans

which also occurs at high altitudes in Natal, but the two are easily distinguishable

by the several headed paniculate inflorescence, the hairiness of the basal part of the

stem and the shallowly dentate, sometimes ciliate, margins of the leaves of the new

species, opposed to the one or two headed inflorescence, glabrous stem and glabrous

serrate margins of the leaves of S. tugelensis.

426

Senecio macroalatus M. D. Henderson, sp. nov., affinis S. decurrenti DC. et

S. digitatifolio DC., sed planta omnino glabra, foliis multo profundioribus et

acutioribus serratis, capitulis inconspicue calyculatis, involucri bracteis apicibus

nigris differt.

Herba erecta, usque ad 67 cm. alta,. glabra. Folia irregulariter et profunde serrata,

apicibus longis acutis; folia radicalia 25 cm. longa, petiolis longis; folia inferiora

caulina 17 cm. longa et 3-5 cm. lata, valde decurrentia, alis profunde serratis usque

ad 1 cm. latis; folia suprema parva, lanceolata, basi auriculis incisis. Inflorescentia

paniculata. Capitula radiata, flava, inconspicue calyculata. Involucri bracteae, 7-8 mm.

longae, apice nigrae. Achaenia glabra, 10-costata, 2 mm. longa.

Natal. — Bergville: Cathedral Peak Forest Research Station, 9,800 ft., in Organ

Pipes’ Pass, March, Killick 1486.

Herb up to 67 cm. high, glabrous. Stem erect, striate leafy, especially at the

base. Leaves deeply and irregularly serrate on margin, apex long acute, midrib and

reticulate venation prominent on undersurface; radical oblanceolate, 25 cm. long,

3 cm. broad, narrowed to a winged, clasping petiole; cauline oblong-lanceolate,

17 cm. long, 3-5 cm. broad, strongly decurrent in deeply serrate stem-wings, up to

1 cm. wide; uppermost small lanceolate with incised basal auricles clasping the stem.

Inflorescence a dense terminal panicle, with small subulate bracts on the peduncles.

Heads radiate, ray- and disc-florets yellow. Involucre cylindrical to campanulate,

7-8 mm. long; bracts 12, linear-lanceolate, more or less obtuse, with minute black

bristly tips and narrow membranous margins; basal bracts few, a third the length

of involucral bracts, with black or reddish tips. Receptacle honeycombed, margins

of pits usually produced on one side only. Corolla of ray-florets 16 mm. long, limb

2-3 times as long as tube; of disc-florets 8 mm. long, lobes 1-5 mm. long, thickened

and glandular. Filaments thickened below anthers. Achenes glabrous, 2 mm. long,

10-ribbed. Pappus of all florets of fine copious barbellate bristles.

Although common in the Koeleria-Festuca-Danthonia grassveld of Organ Pipes’

Pass in the Drakensberg this species has not previously been collected for the National

Herbarium and is not matched in the Kew Herbarium. The glabrous nature of the

whole plant and the inconspicuously calycled heads with black-tipped involucral bracts

very readily distinguish it from both S. decurrens DC. and S. digit alifolius DC., which

like S. macroalatus have broadly decurrent leaves.

CYCADACEAE.

Encephalartos eximius Vercloorn sp. nov., affinis E. ghellinckii Lem. sed plantis

minoribus valde soboliferis 1- (rariter 2-) strobiliferis foliolis leviter latioribus

differt.

Planta humilis valde sobolifera; truncus 25-60 cm. longus, circiter 25 cm. diam.

Folia c. 65 cm. longa, sub-erecta; petiolus c. 10-20 cm. longus, 6 mm. diam., priori

lanuginosus; pulvinus 3-4 cm. longus 2-5-3 cm. latus dense lanuginosus; rachis

lanuginosa et nonnunquam villosa, glabrescens; foliola angusto-linearia, usque 9-13 cm.

longa, 4-5 mm. lata, ad basin et apicem rachis minora, patentia, apicibus pungentibus,

marginibus revolutis, infra prominente 5-nervata; nervi pilosi. Strobilus masculinus

1-natus (vel rariter strobili 2-nati), plus minusve cylindricus apicem et basin versus

leviter attenuatus, 13-22 cm. longus, 4-5-7 cm. diam., dense lanuginoso-tomentosus;

pedunculus 3-11 cm. longus; bracteae longae angustae dorso dense tomentosae;

squamae c. 2 cm. longae, 1-8 cm. latae, subtus numerosis micro-sporangiis obtectae;

facies dense tomentosa, sub-orbicularis vel sub-rhomboidea, c. 1-2 cm. X 5-8 mm. X

1-2 mm. ; sub tomentum leviter concava. Strobilus femineus 1-natus, dense tomentosus,

20-30 cm. longus, 16-18 cm. diam.; pedunculus c. 3-5 cm. longus, basin 2-3 cm.

diam., versus apicem 4 cm. diam., bracteae multae, angustae, 2-5 cm. longae, 3-8 mm.

427

latae, dorso dense tomentosae; squamae c. 3 cm. longae 4 cm. latae, facies plana

(sub tomentum leviter concava), plus minusve rhomboidea 4 cm. x 2-5 cm. x 1-3 cm.

Semina citrina pallide salmonea suffusa deinde succineo-brunea.

Cape Province. — Bedford Div. : Eildon, Dyer 5331. Story 4017; 4019 (male

cone); 4021 (type, female); Huntly Glen, Dyer 5333; Story 4014; 4022; 4023;

Daggaboersnek, Story 4018. Cradock Div.: Cradock Mts. towards Tarkastad,

Marloth 2151. (Cult in Kirstenbosch Gardens, origin unknown, Henderson 1566).

Plates 2, 3 and 4.

Plants dwarf, suckering freely at the base, stems above ground level usually 25-60

cm. long and about 25 cm. diam., leaf-scars variable, the average diagonal measurements

3x1-5 cm.; remains of bracts grey, smooth, indurated, about 3 cm. long and 2-5 cm.

broad at the base. Leaves about 65 cm. long, more or less erect, sometimes rather

spirally twisted; petiole 10-20 cm. long, bluntly three-angled, woolly pubescent in

parts when young, sometimes with long straight hairs as well, about 6 mm. diam.,

widening slightly towards the base where it is about 9 mm. broad just above the pulvinus ;

pulvinus about 3-4 cmv long, and 2 • 5-3 cm. broad in the middle, densely lanuginose ;

rhachis woolly pubescent in parts with curly and sometimes straight grey hairs,

becoming glabrous, rounded below and flatly rounded above, sometimes with a rather

faint ridge down both surfaces; leaflets narrowly linear, the median about 9-13 cm.

long, 4-5 cm. broad, reducing gradually in size towards base and apex of rhachis,

entire with revolute margins and 5 prominent, sparsely pilose nerves on the lower

surface, pungent at the tip, narrowing very slightly at base before widening into the

foot-like connective to the rhachis, spreading or forming a wide V, not overlapping.

Male cones 1- or rarely 2-nate, more or less cylindrical, tapering slightly to base and

apex or rounded at the apex, 13-22 cm. long, 4-5-7 cm. diam. with scales in about

17 oblique rows; peduncle 3-11 cm. long with long, narrow, dorsally tomentose bracts;

median scales removed from the cone about 2 cm. long, 1-8 cm. broad, the lower

surface completely covered by numerous microsporangia; scale faces densely felted

with tawny tomentum, appearing sub-circular to sub-rhomboid in outline, varying

in size, about 1-2 cm. by 5-8 mm. and only 1-2 mm. thick (when tomentum is removed),

flattish, with a central depression, hidden by the tomentum. Female cones solitary

(in all specimens seen), densely lanuginose with tawny or off-white tomentum,

20-30 cm. long, 16-18 cm. diam., scales in about 9 oblique rows, peduncle about 3-5

cm. long, 2-5 cm. diam. at base, up to 4 cm. diam. at top; bracts many, narrow,

2-5 cm. long, 3-8 mm. broad, dorsally densely tawny tomentose; median scale removed

from the cone, 3 cm. long (from base of stipe to exposed surface) and 4 cm. broad at

greatest width; stipe and sinus arms yellowish; scale faces more or less flat (not

prominent), with a shallow concave rhomboid umbilicus hidden by tomentum,

sub-rhomboid in shape, diagonal measurements up to 4 cm. by 2-5 cm. and about

1 • 3 cm. thick. Seed pale orange-yellow through salmon suffused orange to amberbrown.

In the article “ Materials for a Revision of the South African Species of

Encephalartos ”, Journ. of S.A. Bot. Jan. 1945, page 62, M. R. Henderson described

a plant growing in Kirstenbosch Gardens of which the place of origin was unknown.

All efforts by interested botanists to find the species in the veld failed until 1951. In

April of that year Dr. R. Story, Botanical Survey Officer, stationed at Grahamstown

located a group of cycads in the mountains N.E. of Bedford. They were growing in

very little soil on dolerite and seemed in poor condition. Being unable to match it

with any species in the Albany Museum and suspecting it to be an undescribed species

Dr. Story sent material to the National Herbarium, Pretoria, where it was recognised

as being the same species as the Kirstenbosch plant mentioned above.

It was found to be closely related to E. ghellinckii Lem. of the Drakensberg and

in some respects like E. cycadifolius Lehm. which occurs on koppies in the grassveld

of Queenstown District. It is similar to these species in that the leaflets are narrow

428

and entire, the cones densely and thickly felted tomentose and the seeds yellow-orange

to brown, never scarlet. It differs in being smaller than both these species, suckering

more freely at the base, and in the cones being borne singly and only rarely with 2

male cones on a plant. Besides these 3 differences it may be distinguished from the

nearer neighbour, E. cycadifolius, in that the leaflets are narrower and more spreeding,

that is the pairs are not arranged V-wise along the axis like those of E. cycadifolius.

In comparison with E. ghellinckii, which occurs on the eastern slopes of the

Drakensberg range and as far south as near Flagstaff in the Cape Province, the leaflets

of E. eximius are slightly broader. They do not have the margins as strongly inrolled

as E. ghellinckii and the prominent nerves of the undersurface are therefore obvious.

In our species too, these nerves are sparsely pubescent with long curly hairs. The

pubescence in general on these two species is rather similar but on the whole E. ghel-

linckii is more generally woolly-villous, having curly and straight hairs. Straight hairs

are, however, sometimes found on the new species.

From existing revisions on the genus one might be led to assume that the early

travellers and botanists overlooked this species completely. But since this seemed

doubtful old records and likely descriptions and specimens were combed for possible

clues. One definite find was a specimen in the Marloth Herbarium, Marloth 2151,

collected in the “ Cradock Mts. towards Tarkastad altitude 1200 m., Oct. 1895 ” which

was obviously the same as our species. It had been confused with E. Friderici-Gui/ie/mi,

that is E. cycadifolius Lehm. Another case of possible confusion of our species is the

record of E. lehmannii, the Karoo cycad, from Bedford (see FI. Cap. Vol. 5, sect. 2,

page 36). In all probability this is our species but since the record is not supported

by a specimen among the citations proof of this is lacking.

ERICACEAE.

Erica gazensis El. Wild sp. nov., E. lanceoliferae S. Moore affinis, pilis glandulosis

distinguenda; ab E. milanjiana Bolus antheris appendiculatis differt.

Suffrutex ad 60 cm. altus, adscendens vel erectus, pilis eglandulosis brevibus et

glandulis stipitatis longioribus paucioribusque indutus. Ramuli numerosi, recti vel

flexuosi, rubro-brunnei, diametro circa 1 mm., teretes; internodiis 2-7 mm. longis.

Folia quaterna; petiolus 0-2 mm. longus, complanatus, breviter pubescens; lamina

oblongo-lanceolata, subacuta, revoluta, griseo-viridis, minima 2-6 mm. longa, 0-8 mm.

lata, maxima 4 mm. longa, 1-5 mm. lata, supra minute muricata, pilis brevibus eglandu-

losis et principue ad marginem pilis longis glandulosis sparsim vestita, subtus pallidior,

dense eglanduloso-puberula, pilis glandulosis breviter stipitatis admixtis, nervo mediano

prominente. Inflorescentia ad ramulorum ultimorum apices in umbellas subglobosas

disposita. Flores quaque umbellae 8-12; pedicelli tenues, 2-4 mm. longi; bractea 1,

aliquando deessens, pedicelli dimidio superior vel inferior, oblongo-linearis, breviter

eglanduloso-pilosa et margine stipitato-glandulosa, 0*2 mm.-0-8 mm. longa, 0-15

mm.-0-8 mm. lata; bracteolae 2, saepe deessentes, prope calycem, oppositae vel

sub-oppositae, bracteae similes sed minores, ca. 0-2 mm. longae; calyx quadripartitus,

segmentis lanceolatis vel ovato-lanceolatis, 1 mm. longis, basi 0-6 mm. latis, uninervatis,

membranaceis, pilis brevibus eglandulosis et glandulis marginalibus longistipitatis

vestitis; corolla roseo-punicea, campanulata vel paulo urceolato-campanulata, glabra

vel sparsim pilis minutis eglandulosis marginata, 1- 7-2-1 mm. longa, 1- 7-2-1 mm.

lata, lobis 4 apice rotundatis, erectis vel leviter recurvis, 0-75 mm. longis; stamina 7-8,

inter disci lobos emergentes, filamentis complanatis, 1 • 2 mm. longis, antheris breviter

exsertis, oblongis, 0-75-0-9 mm. longis, purpureo-fuscis, scabridis, bi-appendiculatis,

appendiculis supra basin dorsifixis, deflexis, lamelliformibus, basi latis, 0-35-0-5 mm.

longis, secundem marginem paulo scabrido-pubescentibus ; ovarium depresso-globosum,

villosum, eglandulosum, altitudine 0-38-0-8 mm. diametro 0-42-1-2 mm.; discus

7-8-lobatus; stylus tetragonus, glaber, 2-4 mm. longus, 1 mm. exsertus, apice

subcapitatus.

429

S. Rhodesia: Melsetter: Chimanimani, 2,200 m., along open streams, July,

Thompson 16 (type in S.R.G.H., Salisbury); Greenmount, 2,000 m., Bracken grassland,

Sept., Crook 127 (S.R.G.H. 31413); Albany, 1,800 m., in Brachystegia spiciformis

woodland, July, Crook 11 (S.R.G.H. 29010); Rocklands, 1,500 m., Oct., Chase 2976

(S.R.G.H. 31197); Chimanimani Mts., 2,000 m., June, Munch 203 (S.R.G.H. 24452).

Fig. 4. — Erica gazensis H. Wild. A — Flower, X 8; B — Anther, x 17; C — Ovary

and disc, x 26; D — Leaf, abaxial surface, x 17; E— Leaf, adaxial surface, x 17.

The relationships of E. gazensis can be readily seen if it is inserted in the key given

by Aim & Fries in their monograph “ Die tropischen Arten der Gattung Erica L.”

Ark. Bot. Band 2A, No. 7 (1927). The revised portion of this key would then be as

follows: —

Branches and leaves glandular:

Anthers muticous E. milanjiana.

Anthers appendaged E. gazensis.

Branches and leaves eglandular E. lanceolifera.

E. gazensis is rather a variable species and the variations seem haphazardly

distributed among the available material ; however, the structure of the floral parts is

comparatively uniform. Chase 2976 has its umbels noticeably crowded together;

Crook 127 a continuous minute pubescence along the margins of the corolla lobes whilst

the type has merely a very few minute hairs on a very small proportion of the corolla

lobes and the remainder have completely glabrous corollas. Crook 127 and 11 both

differ from the rest in having less revolute leaf margins. This last difference could

be due to differing conditions of humidity and shade. The specific name is derived

from the name of that part of S. Rhodesia and Portuguese East Africa which extends

on both sides of the border to include the Melsetter, Chipinga, Chimanimani areas

and Southern Manica province. It is defined by Swynnerton in J. Linn. Soc. Bot.

XL, p. 2 (1911).

Erica eylesii Bolus var. blaeriodes H. Wild, var. nov., staminium numero vulgo 4,

rarius 5, nunquam usque 8, a typo distinguitur.

Portuguese East Africa. — Manica; Chimanimani Mts., 2,000 m., June, Munch

206 (type of var. in S.R.G.H., Salisbury); Chimanimani Mts., 1,700 m., on rocky

outcrops at streamside July, Pedro & Pedrogao 7331; Chimanimani Mts., 1,700 m.,

430

431

margin of R. Mevumozi, in ericoid scrub, July, Pedro & Pedrogao 7318; Chimanimani

Mts., 1,700 m., on rocky outcrops, by streamside, July, Pedro & Pedrogao 7332;

Chimanimani Mts., 1,500 m., by stream, June, Wild 2903.

S. Rhodesia — Melsetter; Chimanimani Mts., 2,000 m., near stream, June, Wild

2937; Melsetter town lands, 1500 m., July, Crook 4; Melsetter town lands, 1,500 m.,

very locally common, July, Crook 6; Chimanimani Mts., Bonde R., 1,500 m., June,

Wild 2854; Chimanimani Mts., Musapa Gap, 1,000 m., Oct., Munch 338; Melsetter,

1,500 m., riverine, Aug., Crook 105.

In the last few years a good deal of material has been collected in the Chimanimani

area which bears the closest resemblance to Erica eylesii Bolus but has flowers with only

four stamens and sometimes much broader ovate leaves. The possibility was examined

that this represented an undescribed Blaeria sp. with a superficial resemblance to

E. eylesii but, apart from the stamen number and variation in leaf shape, no other

character could be found to distinguish it from E. eylesii. Although the undetermined

specimens and E. eylesii fell fairly readily into two groups with either 4 or 8 stamens,

odd flowers were found with 5, 6 or 7. Flowers with 5 stamens were fairly common

on predominantly 4-anthered gatherings while 7-anthered flowers were fairly common

in 8-anthered gatherings and only a very few flowers with 6 stamens were discovered,

mainly on predominantly 8-anthered specimens.

It was concluded therefore, that the 4-anthered specimens constituted merely a

variety of E. eylesii. This is of some significance since the only reliable character

distinguishing the genus Blaeria L. from Erica L. is the possession of 4 stamens as

against 8 (or rarely 6). It is fairly evident that this is a rather artificial distinction and

the existence of an Erica sp. which can have from 4-8 anthers emphasises this more

strongly. It would seem that the genus Blaeria can only be retained on the grounds

of convenience. This otherwise rather poor reason does have some force, however,

in practice, when the large size of the genus Erica is taken into consideration.

In the course of these investigations it was found that the leaves are often ovate

and revolute at the margins only. On the same plant, however, linear-oblong or acicular

leaves are often present and on examination prove to be derived by the folding of the

cordate leaf along two adaxial ridges more or less parallel with the midrib. In many

specimens one or other kind of leaf predominates and this at first sight suggests the

presence of two species.

It is in the flower structure, however, that polymorphism is most apparent. Bolus

has described the corollas of the type as being cyathiform and slightly constricted at

the throat, Aim & Flies [who redescribed this species by mistake some three years

later and, by a strange coincidence, gave it the same specific name — Arkiv Bot., Stock-

holm, 21A, No. 7, 6 (1927)], in describing corollas from the same gathering, define

them as inflated tubular to subovoid. An urceolate-globose form is perhaps the com-

monest on a wide range of gatherings but quite commonly a gathering will only shew

tubular flowers with parallel sides. The buds are narrowly tubular or have a con-

striction in the middle of the tube. In addition, a small proportion of the flowers are

very broadly pyriform with a narrow cylindric neck. In this case the anthers, although

fertile, never become exserted and this may render self-pollination more likely. In

all cases fertile capsules are produced. That these differences are of no taxonomic

significance is proved by the fact that they are all represented on one branch on the

type sheet of the variety. The accompanying illustration shews three flower forms,

a bud, two leaf forms and the ovary, all drawn from the type sheet of the variety.

It should be noted further that the anther appendages of this species are sometimes

much reduced and difficult to see. In this form it approaches closely E. thryptomenoides

S. Moore, which appears to differ from E. eylesii only in its muticous anthers. The

relationship of the two species requires further investigation with a wider range of

material.

432

The author wishes to acknowledge with thanks the assistance of Miss I. C. Yerdoorn

of the National Herbarium, Pretoria and of Mr. N. S. Pillans and Miss Esterhuysen

of the Bolus Herbarium who have seen the majority of the material used in drawing

up this account.

GERANIACEAE.

Pelargonium jacobii R. A. Dyer sp. nov., affinis P. carnoso L. foliis obovatis vel

oblanceolatis dentatis facile distinguitur.

Planta suffrutescens succulenta, sparse ramosa, basi 2-4 cm. crassa, usque 40 cm.

longa. Folia oblanceolata vel obovata, 4-9 cm. longa basin versus attenuata, pilosa,

margine dentato vel leviter lobato. Pedunculus 5-20 cm. longus plus minusve ramosus,

breviter pubescens ramis 3-7-floribus; bracteae lanceolatae vel oblongo-lanceolatae,

5-7 mm. longae, 2 -5-3 -5 mm. latae. Flores pedunculati, pedicellis gracilibus 2-3 cm.

longis glabris calycis calcari usque 2 mm. longo eis adnato. Sepala oblonga, 8 mm.

longa, 2 -5-3 -5 mm. lata, obtusa. Petala subaequalia, oblanceolata, spathulata, 9 mm.

longa, apicem versus 3 mm. lata. Ovarium dense villosum; fructus 4-5 cm. longus,

rostro pilis patulis albidis hirsuto.

Cape Province. — Richtersveld ; Kubus, August, Marloth 12345 (type); Arris

Drift, Marloth 12345 b.

South West Africa. — Sandy flats east of Buchuberg, July, Dinter 6437.

Plate 5.

Plant up to about 20 cm. tall and up to 40 cm. with inflorescence. Stem succulent,

2-4 cm. thick near base, sparsely branched, pubescent near tips of branches. Leaves

obovate to oblanceolate, 4-9 cm. in total length, tapering into a petiole-like lower half

and a swollen base, pilose; margin dentate or moderately lobed. Peduncles slender

or stout, 5-20 cm. long, simple or up to 3-branched, mjnutely pubescent. Bracts

subtending the pedicels more or less lanceolate or oblong-lanceolate, 5-7 mm. long,

2 -5-3 -5 mm. broad. Flowers white, pedicellate; pedicels slender, 2-3 cm. long, with

the fused nectariferous calyx spur extending 2 mm. down, glabrous. Sepals oblong,

8 mm. long, 2 -5-3 -5 mm. broad, obtuse, glabrous or with a few hairs and becoming

glabrous. Petals oblanceolate-spathulate, more or less equal, 9 mm. long and 3 mm.

broad towards the apex. Ovary densely hairy; the beak elongating in the fruiting

stage up to about 4 cm. long; seeds hirsute.

This species is closely related to P. carnosum L. from which it is readily distinguished

by the entire leaves. It may be even more closely allied to P. mirabile Dinter, of which

no authentically named specimen has been seen. Miss Esterhuysen of the Bolus

Herbarium states that according to specimens in that Herbarium, the calyx of

P. mirabile is very hairy and quite distinct from that of the present species.

The naming of this plant as P. jacobii, in honour of Colonel Robert Jacob Gordon

of the Old Dutch East India Company, is resorted to because the surname has already

been used. Gordon did much to assist botanical exploration prior to 1800, and

must surely have been the first person to paint this species, so that it is desired to

commemorate him in this way. A measure of appreciation of Gordon’s work was

expressed by the writer in an address to the South African Biological Society, which

was published in the society’s Pamphlet No. 14, 1949. The writings of V. S. Forbes

in Afrikana Notes and News (Afrikana Aantekeninge en Nuus), June, 1952, and

elsewhere, should also be consulted if one wished to get a broad view of Gordon’s

remarkable personality.

The accompanying reproduction was made from a photograph of the original

coloured illustration by Gordon which is in the Rijks Museum, Amsterdam. There

is a copy of this painting in an album of “ Paterson’s Drawings ” in the library of Sir

Ernest Oppenheimer, Johannesburg.

433

GESNERIACEAE.

Streptocarpus kentaniensis Britten and Story, sp. nov., S. meyeri Burtt et S. luteae

C.B. Cl. affinis, a S. meyeri tubulo breviori different, et quod tubulus fundo

faucium maculatus est, a S. lutea quod flores violacei sunt, nec candidi nec

lutei, ab ambabus limbo floris obliquo et foliis angustis fere linearibus.

Herba acaulis, rosulata, multifoliata; foliis crassulis, angustis, rigidis, ascendentibus,

distincte petiolatis, inflorescentias multas ferentibus. Folium plerumque 10 cm. longum,

1-8 cm. latum, rugosum, acuminatum, villis brevibus non glanduliferis utrinque

contectum, margine leviter crenato et leviter revoluto, nervis infra plurimis

prominentibus, villis appressis; petiolus usque ad 3 cm. longus, rubens, pedunculis

serie 3-4 in extremo emergentibus. Pedunculus 9-12 cm. longus, villis plerumque

non glanduliferis, infra fuscus, deinde paullatim viridis, floribus 2, rariter usque ad 5.

Bractea circiter 1 mm. longa, villis plerumque non glanduliferis. Pedicellus 1-2 cm.

longus, villis plerumque glanduliferis. Flos plerumque 2 -5-2 -9 cm. longus. Calyx

circiter 3 mm. longus, ad basin partitus segmentis angustis villis glanduliferis et

eglanduliferis contectis. Corolla similiter pubescens; tubulus sub-violaceus, intus

pulla viola diverse maculosus, cylindratus, plus minusve curvatus, faucibus ad circiter

6 mm. dilatatus, intus villosus, villis fundo longis, alibi brevioribus, ruga in longitudinem

per medium fundum; limbus 1-7-2 cm. latus; lobus anticus et lobi ex utroque latere

6 mm., lobi aversi 5 mm., omnes apice orbiculati. Stamina filamentis tortis, purpureis,

violaceis vel candidis, 4 mm., antheris circiter ad dimidiam partem tubuli allatis;

staminodia duo, 1-5 mm. longa. Pistillum 1-3 cm. longum; stigma ultra antheras

eminens, ovarium stylusque dense villis longis glanduliferis et eglanduliferis contecti;

ovarium viride, purpura reticulatum; stylus ovario aequilongus; stigma capitatum,

ovatum; orbis annularis, colore croceus. Capsula gracilis, circiter 5 cm. longa, scabra;

semina circiter 0-5 mm. longa et dimidio angustiora, ferme ovata, utroque fine rostrata,

fusca, inaequaliter angulata et rugosa.

Cape Province. — Kentani or Komgha district: on rocks in woods near Kei

Mouth, July, Flanagan 2337; Kentani district: Britten, unnumbered; 2 miles S.E. of

Kentani, mossy rock in rather dry forest, rare, 1,000 ft., August, Story 4038 in Albany

Museum Herbarium (type).

A stemless herb with thick, stiff, narrow leaves, ascending, distinctly petiolate,

bearing many inflorescences. Leaf usually 10 cm. long, but varying from 3 to 20 cm.,

usually about 1 • 8 cm. wide, rugose, covered with non-glandular short hairs on both

sides, apex acute, margin slightly crenate and slightly revolute, upper side dark green,

lower side pale, sometimes reddish ; midrib very prominent below, with hairs appressed,

other smaller veins numerous and also prominent, projecting to give the lower side a

honeycombed appearance with the green tissue scarcely visible; petiole up to 3 cm.

long, reddish, with peduncles arising in series of three or four at its distal end.

Peduncle 9-12 cm. long, with hairs mostly non-glandular, brownish at base merging

into green at top, usually two-flowered, but sometimes bearing as many as five flowers.

Bract about 1 mm. long, with hairs mostly non-glandular. Pedicel 1-2 cm. long, with

hairs mostly glandular. Flower 2 -5-2 -9 cm. long, smaller if growing under dry

conditions. Calyx about 3 mm., cut to base, segments narrow, covered with hairs

on outside, many glandular. Corolla similarly hairy, tube light lobelia violet, spotted

variously with aconite violet inside, cylindrical, slightly curved, widening towards

the throat where it is about 6 mm. wide, with hairs inside long on the floor, shorter

elsewhere, and with a fold running down the centre of the floor; limb 1 -7-2 cm. across;

front lobe and side lobes 6 mm., back lobes 5 mm., all rounded at the apex. Stamens

with characteristic twisted filaments, purple, lilac or white, 4 mm., bringing the anthers

about half-way up the tube, anthers connivent, explosive, cells confluent, with powdery

pollen oval-shaped when dry, round in water; staminodes two, 1-5 mm. long.

434

Gynoecium 1-3 cm. long; stigma protruding beyond the anthers; ovary and style densely

covered with long hairs, many of them glandular; ovary green veined with purple; style

white, nearly as long as ovary; stigma capitate, oval; disc annular, xanthine orange.

Capsule slender, about 5 cm. long, scabrous; seeds a little over 0-5 mm. long and half

as broad, roughly oval, beaked at each end, dark brown, irregularly angled and rugose.

435

The following notes are by the second author, R. Story.

Plants of this species, grown by Mrs. Turnbull near Kentani, were sent by Mr.

S. Stent in July and November, 1940 to Miss L. L. Britten in Grahamstown. Mrs.

Turnbull sent more in 1943. They had been found growing in a “ rather dry place ”,

and were kept moderately watered in cultivation.

Fig. 7. — Streptocarpus kentaniensis, X 0-6.

436

In a search by the writer for wild specimens of this species, six localities in the

Kentani district were visited without success at various times between January and

June, 1952. In August, 1952 directions were obtained from Mr. Mills, of Nyutura,

near Kentani, and plants were located growing on a pile of rock in a forest patch

estimated to receive an annual rainfall of about 25 inches. Although there are many

similar piles of rock in this and other forest patches, and although another species of

Streptocarpus is common in most of them, no S. kentaniensis was found except in this

one place. The plants were of all sizes, and several were in flower in spite of the dry

conditions then prevailing. Even when they are not flowering they may be readily

recognised from the shape of their leaves.

The plants have many leaves developing in orderly succession and forming a

rosette. There has so far been no sign of root runners, but occasionally a new rosette

may develop adventitiously on the petiole. The flowers appear to need a specific

pollinating agent, for few capsules are set naturally in cultivation. The species is

unusual in flowering in winter or early summer.

The plant was recognised as an undescribed species by Miss L. L. Britten, who

studied it in cultivation for many years, and who died before her work upon it was

ready for publication. Her comprehensive notes were used in conjunction with the

type specimen in drawing up this account.

IRIDACEAE.

Moraea culmea Killick, sp. nov., affinis M. tenui Ker., sed ungue lamina perianthii

segmentorum exteriorum aequilongo differt.

. Cormus 1-2 cm. diam. Caulis 30-55 cm. longus. Folia 2, caulina, inferiora linearia

9-20 cm. longa, superiora spathis similia 5 -5-7 -5 cm. longa. Spathae lanceolatae,

acuminatae, cuspidatae, apicibus scarioso-fuscae, exteriores 3-2-5 cm. longae, interiores

4-2-6 cm. longae. Perianthii segmenta exteriora oblongo-unguiculata ; lamina 7-11

mm. longa, 3-5 mm. lata, reflexa, albo-lutea; unguis cuneatus, 8—10 mm. longus,

interior pubescens, nectario operculato ut basin. Segmenta interiora trifida, ungue

1 • 2 cm. longo cuspide media lineari 4-7 mm. longa et lobis duobus falcatis lateralibus

1-2 mm. longis. Filamenta connata, 6-8 mm. longa; antherae 4 mm. longae. Styli

rami 6-9 mm. longi, cristis lanceolato-acuminatis incurvatis 3-4 mm. longis; stigma

integra et truncata vel 2-lobata. Ovarium clavatum, 7-9 mm. longum.

Natal. — Bergville District: Cathedral Peak Forest Influences Research Station,

Killick 1209; 1588 (type); Estcourt District: near Tabamhlope Police Station, Acocks

10781.

Corm globose, 1-2 cm. diameter; tunics with brown parallel fibres and short

transverse strands. Stem 30-55 cm. long, slender, 1-2 mm. thick. Leaves 2, cauline;

lower linear, sheathing at base, 9-20 cm. long, 2-4 mm. broad; upper spathe-like,

completely sheathing, 5 -5-7 -5 cm. long, 2-5-4 mm. broad. Spathes cylindrical, 1-3-

flowered; valves herbaceous, lanceolate, acuminate, cuspidate, scarious-brown at the

tips, outer 3-2-5 cm. long, inner 4-2-6 cm. long. Outer perianth segments oblong-

unguiculate; lamina 7-11 mm. long, 3-5 mm. broad, reflexed, outside white tinged

with yellow and green-dotted, inside white with yellow puberulous blotch at base

surrounded by few large green dots; claw cuneate, 8-10 mm. long, 4 mm. broad,

tapering to 1 mm. at base, outside greenish-white, inside light green, pubescent, with

an operculate nectary at the base. Inner perianth segments trifid, with a claw 1 • 2 cm.

long and a central linear cusp 4-7 mm. long and two diverging falcate lateral lobes

1-2 mm. long. Filaments connate, except for upper fifth, 6-8 mm. long; anthers

4 mm. long. Style branches 6-9 mm. long, deeply bifid; crests lanceolate-acuminate,

ncurved, 3-4 mm. long. Stigma entire and truncate or 2-lobed with minute deltoid

obes. Ovary clavate, 7-9 mm. long.

437

Fig. 8. — Moraea culmea: a, flower, x 3; b, longitudinal section of outer perianth

segment, K 4; c, staminal column and style, x 4.

Our only records of M. culmea are from the lower slopes of the Drakensberg in

the Cathedral Peak Area and Tabamhlope in the Estcourt District. In the former

locality it grows socially in moist situations in Themeda triandra grassveld, while at

Tabamhlope it is frequent — generally in Highland Sourveld. It flowers during the

months of November and December. The details of flower colour given in the

description above were observed in the field. On drying the flowers become yellow.

The plant has a straw-like appearance, hence the specific epithet culmea.

M. culmea is closely related to M. tenuis Ker., but differs principally in that the

claw of the outer perianth segments is equal in length to the lamina instead of half

as long. In addition, it differs in flower colour and vegetatively, in not having a

subradical produced leaf.

Moraea modesta Killick, sp. nov., affinis M. tripetalae Ker., sed perianthii segmenta

interiora trifida non linearia differt.

Cormus globosus, 1-1*5 cm., diam. Caulis 10-25 cm. altus. Folia basalia 2,

filiformia, 30-55 cm. et 7 cm. longa; folia caulina 2, spathis similia, 3-5 cm. longa.

Spathae lanceolatae, apicibus scarioso-fuscae, exteriores 2-3*5 cm. longa e, interiores

3-5*5 cm. longae. Segmenta exteriora perianthii ovato-unguiculata; lamina 8-11 mm.

longa, 7 mm. lata, alba, purpurato-venata; unguis sub-erectus, 1-1*2 cm. longus,

1*5 mm. latus. Segmenta interiora perianthii trifida, 1*5-2 *5 mm. longa, 0*5-0 *7

mm. lata, purpurata, lobo medio 0*8 mm. longo et lobis duobus lateralibus minutis.

Stamina filamentis liberis; antherae 4-7 mm. longae, apiculatae. Styli rami 1*4 cm.

longi, cristis lanceolato-acuminatis 4 mm. longis; stigma integra, minuta. Ovarium

cylindricum, 6-10 mm. longum.

438

Natal. — Bergville District: Cathedral Peak Forest Influences Research Station,

Killick 1028 (type), 1551A; below the Sentinel, Mont-aux-Sources, Galpin 10372;

Little Berg, Royal National Park, West 1269.

Conn globose, 1-1 *5 cm. diameter; outer tunics with brown parallel fibres and

short transverse strands; inner tunics with thicker anastomosing fibres. Stem 10-25 cm.

high, slender, 1 mm. thick. Produced leaves 2, basal, filiform, revolute, 1 mm. broad;

the longer 30-55 cm. long, somewhat falcate, the other 7 cm. long, erect; cauline

leaves 2, spathe-like, 3-5 cm. long. Spathes cylindrical, 1-3-flowered; valves herbaceous,

lanceolate, scarious-brown at the tips, occasionally shortly cuspidate, the outer 2-3 • 5

cm. long, the inner 3-5-5 cm. long. Outer perianth segments ovate-unguiculate ; lamina

spreading, 8-11 mm. long, 7 mm. broad, white, purple-veined, with a yellow blotch

at the base decurrent on the claw; claw sub-erect, 1-1-2 cm. long, 1-5 mm. broad,

widening slightly just below half-way. Inner perianth segments trifid, minute, 1*5-2 -5

mm. long, 0-5-0 -7 mm. broad, with a central cusp 0-8 mm. long and two very short

lateral lobes, purple. Filaments free, 4 mm. long; anthers 4-6 mm. long, apiculate,

Style branches 1-4 cm. long, bifid, with lanceolate-acuminate crests 4 mm. long;

stigma entire, minute. Ovary cylindridal, 6-10 mm. long.

This new species of Moraea grows on grassy slopes of the Drakensberg between

6 and 9,000 feet. The plant is rare in the Cathedral Peak Area, but common further

north in the Royal National Park at Mont-aux-Sources. It flowers from October to

December. As its name implies, it has a modest and rather inconspicuous appearance.

M. modesta is closely allied to M. tripetala Ker., but can be readily distinguished

by having trifid instead of linear inner perianth segments. Miss G. J. Lewis of the

S.A. Museum, Cape Town, who kindly confirmed that this is an undescribed species,

points out that it also differs from M. tripetala in that the claw of the outer perianth

segments is glabrous instead of minutely bearded.

439

LABIATAE.

Micromeria grandiflora Killick, sp. nov., affinis M. pilosae Benth., sed caulibus decum.

bentibus, internodiis brevioribus, floribus bracteisque multo majoribus differt^

Herba perennis, aromatica, 15-30 cm. alta, basi parce ramosa, pilosa, glanduloso-

punctata. Caules decumbentes. Folia subsessilia, ovato-cordata, 1 . 2-2 • 5 cm. longa,

1- 2-2 cm. lata, serrata. Flores in axillis foliorum solitarii. Pedicelli 1-2-2 cm. longi,

bibracteati. Bracteae foliis similes, 7-11 mm. longae, 6-10 mm. latae. Calyx 7-8

mm. longus; tubus subcampanulatus, 4-5 mm. longus, 15-nervus; calycis dentes 5,

subaequales, triangulo-lanceolati, 4-5 mm. longi. Corolla pallide ad saturate cobalto-

violaceum; tubus 1-3-1 -7 cm. longus, parte inferiore subcylindrica 4 mm. longa

2 mm. lata, parte superiore campanulata compressa 7-9 mm. lata ostio, latere antico

2- sulcato, intus pilis longis crassis maxime duarum lirarum; labium posticum emargi-

natum, 2-5 mm. longum, 7-10 mm. latum; labium anticum trilobatum, lobis lateralibus

emarginatis vel integris, 4 mm. longis, 5-6 mm. latis, lobo medio emarginato, 4-5 mm.

longo, 6 mm. lato. Stamina 4, inclusa; stamina postica 3 mm. longa; antica 7 mm.

longa; antherae biloculares, parallelae, cum 6-14 glandulis globosis flavis ventralibus

inter cellas. Discus leviter lobatus. Stylus inclusus, 1-0-1 -3 cm. longus; stigma

bilobata. Nucellae oblongae, 3-5 mm. longae, 2-5 mm. latae.

Natal. — Bergville District: Cathedral Peak Forest Influences Research Station,

Codd and Dyer 6241. Killick 1102, 1600. 1684 (type).

Fig. 10. — Micromeria grandiflora : a, flower, x 2; b, corolla slit down upper

side, x 3; c, ventral side of anther with the cells slightly parted to show glands, x 10;

d, dorsal side of anther, x 10; e, pistil, x 5.

9102-7

440

A perennial, aromatic herb, 15-30 cm. high, sparingly branched from the base,

pilose with unicellular glandular hairs and multicellular eglandular hairs, golden

gland-dotted. Stems decumbent, often with small branchlets in the axils of the lower

leaves, pilose. Leaves subsessile, occasionally with petioles 3 mm. long, ovate-cordate,

1-2-2 -5 cm. long, 1-2-2 cm. broad, serrate, teeth few, golden gland-dotted, pilose.

Flowers solitary, axillary. Pedicels 1-2-2 cm. long, bibracteate about the middle,

pilose. Bracts resembling the leaves, 7-11 mm. long, 6-10 mm. broad. Calyx in

flower 7-8 mm. long; tube subcampanulate, 4-5 mm. long, 15-nerved, golden gland-

dotted, pilose outside, glabrous inside; calyx teeth 5, sub-equal, triangular-lanceolate,

4-5 mm. long, pilose outside, glabrous inside except near the apex; calyx in fruit

9-10 mm. long. Corolla pale to deep cobalt-violet, 2-2-4 cm. long; tube 1-3-1 -7

cm. long, subcylindric for basal 4 mm., flattened-campanulate above, 7-9 mm. wide

at mouth, 2 mm. wide at base, 2-furrowed on lower side (showing as ridges inside

tube), pubescent and gland-dotted outside, with long thick brown hairs inside chiefly

on the two ridges; upper lip emarginate, slightly reflexed, 2-5 mm. long, 7-10 mm.

broad, pilose and gland-dotted outside, glabrous inside; lower lip 3-lobed, lateral

lobes emarginate or entire, 4 mm. long, 5-6 mm. broad, sparingly pilose and golden

gland-dotted outside, glabrous inside, median lobe emarginate, 4-5 mm. long, 6 mm.

broad, glabrous outside and inside. Stamens 4, didynamous, included, arcuate;

posticous pair 3 mm. long; anticous 7 mm. long, shortly decurrent; filaments glabrous;

anthers 2-celled, parallel, 1 mm. long, with 6-14 round yellow glands situated ventrally

on the connective between the cells. Disc slightly lobed. Style included, 1-1 • 3 cm. long ;

stigma bilobed, lobes acute, the posticous the shorter. Nutlets oblong, 3 • 5 mm. long,

2*5 mm. broad, brown, shortly hairy.

The first gathering of this new species of Micromeria was made by the author

in November, 1950. It was subsequently collected by Drs. L. E. Codd and R. A. Dyer

in December of the same year and again by the author in November, 1951 and February,

1952. The plant forms small societies in undisturbed Themeda triandra grassveld at

about 6,400 feet on the Little Berg in the Cathedral Peak Area. All the specimens

cited were collected from the same spot. This species appears to be very localized.

The author has spent over a year in the Drakensberg doing botanical survey work,

but has seen it growing in only two small catchment areas.

M. grandiflora is a summer flowering perennial producing flowers from the

beginning of November until the middle of March. The plant is strongly aromatic

producing a mentha-like smell when bruised. Responsible for this are the numerous

gland-tipped hairs and sunken golden glands covering most parts of the plant.

Besides M. grandiflora, there are two other species of Micromeria in South Africa,

M. biflora Benth. and M. pilose Benth., the former occurring in the Eastern Cape,

Transvaal and Basutoland, while the latter is found in Natal and Pondoland. M.

grandiflora has affinities with M. pilosa, but differs in that the stems are decumbent

instead of prostrate, the internodes are shorter and the flowers and bracts are very

much larger. M. pilosa also grows in the Cathedral Peak Area, but as a rule occupies

moister situations than M. grandiflora.

LILIACEAE.

Aloe babatiensis Christian and Verdoorn sp. nov. in Procrassae Berger ab A. percrassae

inter aha plantis caulescentibus foliis brevioribus, ab A. rubroluteae plantis

minoribus perianthiis formis faciliter distinguitur.

Truncus simplex, circa 50 cm. altus, apicem versus dense foliatus infra foliis

exsiccatis deflexis obtectus. Folia rosulata, patentia, apicem versus recurvata,

ovato-lanceolata vel lanceolata, acuta, plus minusve 27 cm. longa, 6 cm. lata, supra

salignea nitidula, obscure lineata, immaculata vel interdum basi maculis albidis paucis

441

nstructa, subtus pallido-viridula, covexa, obscure lineata, immaculata; margines

acuti, sinuato-dentati ; dentes deltoidei, 4 mm. longi, 5-12 mm. distantes apicibus

rubro-bruneis pungentibus. Inflorescentia erecta 40-80 cm. longa, infra medium

ramosa; pedunculus viridis, apicem versus bruneolus, compressus, 25 mm. latus,

basin versus anguste hyalino-marginatus, marginibus minute dentatis; bracteae

subamplexicaulis, 8-nervatae, 35 mm. longae, basi 35-40 mm. latae, sensim acuminatae,

apicem versus abrupte apice obliquo-acuminatae; rami 3, erecti bracteis vacuis paucis

vestiti. Racemi cylindraceo-acuminati, subdensi terminales usque 30 cm. longi;

bracteae pallido-virides 7-8-nervatae, 27 mm. longae, basi 2-15 mm. latae,

oblongo-ovatae; pedicelli pallido-virides vel flavo-virides, erecti deinde cernui, 18-25

mm. longi. Perigonium armeniacum vel salmonium, ca 3-7 cm. longum, cylindraceo-

subtrigonum leviter decurvum, basi rotundatum vel truncatum circa ovarium 5-6 mm.

diam. supra ovarium leviter constrictum 4-5-5 mm. diam. deinde apicem versus

ampliatum, segmentis rectis vel sub-patentibus; segmenta exteriora per 12 mm. libera

obtusa ; segmenta interiora ad margines libera, dorsifixa. Antherae croceae sub-exsertae.

Ovarium viridulum, 6-8 mm. longum, 2-3 mm. latum apicem versum leviter acuminatum;

stylus citrinus, deinde exsertus.

Tanganyika Territory. — About 50 m. N.W. of Babati on road to Ngorongoro

Crater, Pole Evans & Evens 872 in Herb. Christian 985 and in N.H. No. 28489, flowered

at Ewanrigg 16/9/46 (type), leaves from same plant collected 13/7/50.

Plate 6.

Stem simple about 50 cm. high foliate with a crown or rosette of leaves on the

apical portion and old dry leaves below. Leaves crowded above with only some of the

lower leaves of the rosette sometimes showing a short clasping base, spreading, decurved

towards apex, ovate-lanceolate to lanceolate, acute, about 27 cm. long and 6 cm. wide,

upper surface dark shining green, shading to copper colour in the dry season where

exposed to sun, shallowly and broadly concave, sometimes obscurely lineate, immaculate

or sometimes with one to few white spots near the base; lower surface pale green,

lineate, immaculate, convex, margins acute, sinuate dentate; teeth deltoid spreading

with reddish-brown pungent tips 4 mm. long, 8-12 mm. apart, closer lower down.

Inflorescence erect, 40-80 cm. long, branched from below the middle, branches 3 (in

specimens seen) erect or arcuate, crowding together; peduncle green shading to fawn

above, laterally compressed low down; about 25 mm. broad, flat on one surface,

rounded on the other, edges acute and, in lower half, with a very narrow hyaline margin

which is very minutely toothed; bracts at base of branches sub-amplexicaul, about

8-nerved, 35 mm. long and 35-40 mm. broad at the base gradually narrowing towards

apex where they narrow abruptly to an oblique acute apex, sterile bracts below the

racemes somewhat smaller. Racemes up to 30 cm. long, cylindric-acuminate, sub-dense;

floral bracts pale green, 7-8-nerved, 27 mm. long, base 12-15 mm. broad oblong ovate;

pedicels pale green or yellowish-green, erect, those of mature flowers cernuous, 18-25 mm.

long. Perianth apricot-orange to salmon-pink (buds green-tipped; lobes of open flowers

yellowish), about 3-7 cm. long cylindric-subtrigonous, slightly decurved, rounded or

sub-truncate at the base, about 5-6 mm. diam. over the ovary, slightly constricted

above the ovary to 4 • 5-5 mm. diam. and widened above to 9 mm. and then contracted

at the mouth with the lobes eventually sub-spreading; outer segments free for 12 mm.,

pale pink on margins with a broad deeper pink median portion, apices obtuse straight

to sub-spreading; inner segments free on the margins, dorsifixed to tube, white with

a broad reddish median line, obscurely nerved, apices obtuse straight to slightly

spreading. Stamens with white filaments and terra-cotta anthers which are sub-exserted.

Ovary green 6-8 mm. long 2-3 mm. broad, tapering slightly to the apex; style yellow

to pinkish eventually exserted.

442

The following notes are by the second author, I. C. Verdoorn.

Plants of this Aloe were originally collected on the Pole Evans’ Central and East

African Expedition, Sept. 1938. They were not flowering at the time and for years

they did not flower either in the Transvaal or in Rhodesia. Plants were distributed to

different gardens in both these regions in the attempt to strike the conditions which

would induce flowering. Eventually (Sept. 1946) one plant flowered at Ewanrigg,

Mr. Christian’s farm near Salisbury where Mr. Christian then described and

photographed it. The herbarium specimen which he prepared at the time is the type.

In October, 1951, one of the plants of the same collecting, which Mr. Christian had

given to Mr. Munch of Rusapi to cultivate, flowered and the inflorescence was sent

by air to Mr. Reynolds of Johannesburg. This material was kindly donated to the

National Herbarium and afforded the opportunity of checking the late Mr. Christian’s

description for publication.

With regard to the relationship of this new species, unfortunately Mr. Christian

did not leave any notes expressing his views. The broad bracts, which are longer

than and envelop or conceal the pedicels, together with the rounded base (not stipitate)

of the perianth and the general habit place it in the series Procrassa of Berger, Das

Pflanzenreich 1908. It is readily distinguished from the described species in this series.

In general appearance it is most like Aloe percrassa but differs, in the first place, by

having a stem, and also in the smaller leaves and slightly longer perianths. In the

large bracts it resembles the well known species, Aloe cryptopoda, in Berger’s series

Latebracteatae but otherwise has very little in common with this species and the others

in the Latebracteatae.

The most striking characters about the inflorescence of Aloe babatiensis are the

large pale green, rather thick bracts, the erect branches crowding the racemes together

and the conspicuous sterile bracts below the racemes. The colour of the flowers is

something between apricot-orange and salmon-pink and the buds are green-tipped.

The habit is characterised by the spreading leaves which narrow perceptibly from a

broad base which may be described as deltoid acuminate, and are crowded in a rosette

at the top of the stem. Some of the lower leaves of the rosette show a very short

amplexicaul base and below them the stem is covered with the remains of old dried

leaves. The upper surface of the leaf is rather dark green and shining.

According to one of the collectors, Mr. Erens, the plants were found about 50 miles

N.W. of Babati on the road to Mbulu and the Ngorongoro Crater “ growing on granite

outcrops in a valley with plenty of leafmould, young plants forming on the stems ”.

Asparagus multituberosus R. A. Dyer, sp. nov., affinis A. asparagoidei (L) Wight caulo

subterraneo elongato multituberoso stylo tripartito differt.

Herba perennis. Caulis subterraneus attenuatus usque 35 cm. longus vel longior

tuberis multis densis circiter 1 cm. longis, 2 mm. latis indutus. Rami lateraliter

producti, plus minusve ramosi, graciles, glabri, espinosi, 20-45 cm. longi, angulares.

Folia parva, ovata, acuminata, membranacea. Phyllocladia sessilia, ovata vel basi

cordata, 1-5-2 -5 cm. longa, 1-2 cm. lata, nervis 8-10. Flores axilares, penduli, 1-3,

pedunculis plus minusve 5 mm. longis apicem versus articulatis. Perianthemum 7 mm.

longum segmentis oblongis albis carina viride. Filamenta lineari-lanceolata, basi lobata.

Ovarii loculi 12-ovulati. Stylus tripartitus.

Cape Province.— Ceres Division: among shrubs on hills near Karoopoort,

Aug., Marloth 9006 (type); Calvinia Division: Calvinia commonage among shrubs,

Mar loth 12778.

Perennial herb with rhizomatous rootstock. Rhizome elongating up to 30 cm. or

more, unbranched or sparsely branched, producing along its whole length densely

packed sessile tuberous roots about 1 cm. long and 2-3-5 mm. thick, and with a few

long non-tuberous roots. Stems 1 — few arising laterally from the rhizome,

443

subscandent, glabrous, spineless, 20-45 cm. tall, ribbed or angled. Leaves small,

scale-like, ovate, acuminate, membranous. Phyllocladia sessile, ovate to broadly

ovate or cordate, 1-5-2 -5 cm. long, 1-2 cm. broad, slightly oblique, with 8-11 main

veins. Flowers axillary, 1-3; pedicels slender curved, about 5 mm. long, articulated

slightly below the perianth. Perianth 7 mm. long with the segments united at the base;

segments linear-oblong, obtuse, 2-5 mm. broad. Stamens with filaments attached

1- 1-5 mm. above the base of the perianth segments; filaments lanceolate, eared at the

base; anthers oblong, 1-5 mm. long. Ovary oblong, 2-2-5 mm. long, 12 ovules in

each cell. Styles 3, free, about as long as the ovary; stigmas truncate, minutely

papillate.

The branches and leaves of this plant are very similar to those of A. asparagoides (L)

Wight (= A. medeoloides Thunb.), and A. ovatus Salter, but the root system is very

different. The tripartite style or 3 free styles of A. multitub erosus is apparently an

unusual feature in the genus and is an important distinguishing character of this species.

A duplicate specimen in the National Herbarium of Schlechter 8069, from Windhoek

in the Clanwilliam district, Cape Province, possibly belongs to this species. The

specimen does not possess roots but the flowers have a 3-partite style.

Eriospermum bifidum R. A. Dyer, sp. nov., folio erecto racemo denso segmentis

aequalibus filamentis bifidis distinguitur.

Tuber 4-9 cm. longum, 4-5 cm. crassum, simplex, rariter ramosum. Folium

solitarium; lamina erecta, oblongo-lanceolata, 7-12 cm. longa, 2-3 cm. lata, concava,

subglauca; petiolus erectus, 5-10 cm. longus, gracilis. Pedunculus gracilis, 20-25 cm.

longus, erectus, basi bractea parva subtentus. Racemus 3-5 cm. longus, 20-30-florus,

densus, pedicellis erectopatentibus 5-15 mm. longis. Perianthemum regulare, 4-5 mm.

longum; segmenta basi breviter connata, obovata, 3-4 mm. longa, supra medium

2- 2-5 mm. lata, obtusa. Filamenta oblongo-elliptica, 2-2-5 mm. longa, leviter crassa,

apice bifida, antheris oblongis, 1 mm. longis. Ovarium 1-1-5 mm. longum, obtuse

3- angulatum. Stylus cylindricus; stigma subcapitata.

Cape Province. — Bathurst Division; near Clumber, on dry hillside near scrub

bush, Dyer 383a (GRA); Albany Division: Committees in arid scrub, Dyer 513

(GRA); 8-10 miles from Grahamstown at Bothas Hill, in arid scrub, Dyer 670 (PRE);

1299 (GRA) type; 35 miles from Grahamstown on old Port Elizabeth Road in scrub,

Dyer 1084; 22-24 miles from Grahamstown towards Carlisle Bridge, arid flats, Dyer

2113 (GRA).

Tuber sausage-shaped or somewhat pyriform, simple or occasionally branched,

4- 9 cm. long, 4-5 cm. thick, shrivelling in very dry weather, about 10 cm. deep in soil

and with old persistent fibrous petiole-like leaf bases. Leaves tough; lamina erect,

oblong-lanceolate, glabrous, 7-12 cm. long, 2-3 cm. broad, concave on upper surface,

somewhat glaucous; petiole wiry, 5-10 cm. long. Peduncle subtended by small

sheathing bract, 20-25 cm. tall, erect. Raceme 3-5 cm. long, dense, 20-30-flowered ;

pedicels 5-15 mm. long, spreading-erect; bracteoles minute. Perianth 4-5 mm. long,

regular; segments shortly united at the base, greenish-yellow with darker green stripes

down back, all equal, obovate, 3-4 mm. long, 2-2-5 mm. broad. Filaments greenish-

yellow, elliptic-oblong, 2 mm. long, bifid at the apex; anthers versatile 1 mm. long,

attached at the sinus of the filament lobes. Ovary 1—1 - 5 mm. long, slightly 3-angled.

Style about equal to ovary in length; stigma somewhat capitate.

When flowers of this species were first examined in 1927, doubt arose whether it

could be included in the genus Eriospermum at all on account of the bifid filaments.

444

The late Dr. R. Marloth, a specialist on the genus at the time, reported as follows:

“ It is different from all others that I have seen, but some I have possess spathulate

stamens with the anthers attached to a small mucro at the apex facing introrsely.

There seem to be all kinds of transitional stages from the linear to the ovate, oblong,

spathulate forms on to yours with an emarginate obovate filament.” A. V. Duthie

also makes reference to the specimens in her “ Contribution to our Knowledge of

the Genus Eriospermum ” in Ann. Univ. Stellenbosch, 1940, p. 11. The filaments

are more than emarginate at the apex and cannot strictly be described as obovate.

Most workers interested in the genus Eriospermum have pointed out the many

difficulties which face one in trying to classify herbarium material. The main difficulties

arise because so many species produce leaves and flowers at different times of the year,

and incomplete specimens in herbaria are the rule rather than the exception.

Eriospermum zeyheri R. A. Dyer sp. nov., folio prostrato racemo leviter denso perianthi?

segmentis aequalibus distinguitur.

Tuber subpyriforme usque 3 cm. crassum. Folium solitarium; lamina leviter

carnosa, glabra, cordata vel lobis plus minusve imbricata, 5-9 cm. longa et lata,

rotundata vel apiculata. Pedunculus erectus, 20-30 cm. longus basi bractea parva

subtentus. Racemus densus, 8-12 cm. longus, 20-40-florus, pedicellis erecto-patentibus

5 mm. vel non-numquam usque 1 cm. longis. Perianthemum subcampanulatum,

4- 5 mm. longum; segmenta basi breviter connata, oblanceolato-oblonga, supra medium

2-2-5 mm. lata, obtusa. Filament a oblongo-elliptica, 2-2-5 mm. longa, 1-25-1-5 mm.

lata, antheris oblongis 1 mm. longis. Ovarium 1-2 mm. longum, loculis 3-ovulatis;

stylus cylindricus 1 mm. longus.

Cape Province. — Albany Division: 6 miles from Grahamstown along road to

Cradock, Sept, (leaves) and flowers added later, Dyer 557 (type in Herb. Alb. Mus.);

Port Elizabeth Division: on hills near Port Elizabeth, March (flowers) Zeyher 4281.

Tuber simple, inverted, subpyriform, with a groove on one side along which the

persistent petiole-like leaf bases pass, 3 cm. broad towards the growing end and

narrowing to the old base which is uppermost in the ground, with a fairly tough skin,

5- 7 cm. below ground. Leaves with soft lamina which is somewhat fleshy, glabrous,

adpressed to ground, usually fight green and glossy, deeply cordate or with basal lobes

overlapping, about as broad as long, 5-9 cm. wide, obtuse or apiculate. Peduncle

with short sheathing bract, 20-30 cm. tall, usually stiffly erect. Raceme 8-12 cm. long,

dense 20-40-flowered; pedicels about 5 mm. long, spreading-erect; bracteoles minute.

Perianth usually yellowish-green with a darker central stripe on the segments, regularly

campanulate in outline; segments almost free to base, spreading with age,

oblanceolate-oblong, obtuse, the outer very slightly shorter and narrower than the

inner, 4-5 mm. long, 2-2-5 mm. broad. Filaments elliptic-oblong, 2-2-5 mm. long,

1-25-1-5 mm. broad; anthers oblong, 1 mm. long. Ovary 1-2 mm. long, 3-6 ovules

in each ovary cell. Style cylindric, 1 mm. long.

The leaves of this species appear in spring and die back in summer while the

inflorescences develope about March and soon mature. It is not known whether

Baker had a complete specimen for study when he prepared the account of the genus

for Flora Capensis Yol. 6 (1896). A specimen of E. zeyheri Zeyher 4281 in the National

Herbarium consists of an inflorescence only and the number is cited by Baker under

the species E. bellendini. It is certainly not conspecific with the typical plant figured

in Curtis’s Botanical Magazine, t. 1382 and it appears evident that Baker used the

latter name as a dumping ground.

445

The present species is characterised by the glabrous glossy, light-green leaves pressed

to the ground and the rather short dense and stiffly erect racemes of more or less

regular yellowish flowers. So far it appears that the distribution is restricted to the

eastern Cape, from Uitenhage and Port Elizabeth to Albany Division.

PHYTOLACCACEAE.

Psammotropha alternifolia Killick, sp. nov., affinis P. androsaceae Fenzl., sed habitu

minore compactiore, pedunculis pauco-nodosis, foliis alternatis differt.

Herba procumbens, 2 -5-6 -2 cm. alta, multo-ramosa. Folia alterna, ramorum

apicibus conferta, elliptico-oblonga, 2-3-5 mm. longa, 1-1-2 mm. lata, albomarginata,

breviter mucronulata. Pedunculi solitarii, 0-4-1 -9 cm. longi, verticillo bractearum

terminale et subterminale. Flores umbellulati, minuti, cremeoflavi. Pedicelli 1 mm.

longi, basi bracteolati. Sepala oblonga, cucullata, 1-5-2 mm. longa, 1-3 mm. lata.

Stamina 1 • 5 mm. longa. Ovarium 3-angulatum, disco hypogyno ; stylus 1 mm. longus,

3-fidus.

Natal. — Bergville District: Mont-aux-Sources, Mann 2865 b, Potts 2971, Hut-

chinson, Forbes and Verdoorn 108; Beacon Buttress, Galpin 10369; near Cleft Peak,

Schelpe 475; near Castle Buttess, Killick 1482 (type); Organ Pipes Pass, Killick 1842.

Procumbent herb, 2 -5-6 -2 cm. high, much branched. Leaves scattered alternately

on the branches and tufted at the apices, elliptic-oblong, 2-3 • 5 mm. long, 1-1 • 2 mm.

broad, white-margined, shortly mucronulate. Peduncles solitary, 0-4-1 -9 cm. long,

with a terminal and subterminal whorl of bracts each whorl supporting an umbellule.

Flowers minute, cream-yellow. Pedicels 1 mm. long, bracteolate at base. Sepals

Fig. 11. — Psammotropha alternifolia : a, portion of flowering branch, x 3; b, leaf,

x 10; c, bract, x 10; d, flower from above, x 10; e, sepal, dorsal view, x 10;

f, pistil and disc, X 20.

446

oblong, cucullate, 1-5-2 mm. long, 1-3 mm. wide. Stamens 1-5 mm. long. Ovary

3-angled, situated on a saucer-shaped disc; style 1 mm. long, 3-fid.

The known range of this species on the Drakensberg is between the Organ Pipes

Pass in the Cleft Peak Area and Mont-aux-Sources further north.

P. alternifolia is essentially a rock -crevice plant, but occasionally it grows in mud

patches present in alpine grassveld. Like many other plants of high altitudes it has

a tufted, dwarfed habit. It flowers from November to late April. In spite of its small

size, it is quite conspicuous when in flower; the profusion of cream-yellow flowers

makes the plant stand out strikingly against the brown of the adjacent basalt.

This species is nearest P. androsacea Fenzl., but can be easily distinguished by

its dwarfed, compact habit, by the 2-noded peduncles and by the presence of leaves

scattered alternately on the branches. It derives its name from this latter character.

It also differs in that the leaves are less conspicuously mucronate.

447

Plate 1 . — Tavaresia meintjesii R. A. Dyer. [Photos B. Meintjes.

9102-8

448

Plate 2. — Encephalartos eximius Verdoorn, habitat on hills between Bedford and

Tarkastad [Photo R. A. Dyer.

Plate 3. — Encephalartos eximius Verdoorn, on hill between Bedford and Tarkastad.

[Photo R. A. Dyer.

449

Plate 4 .—Encephalartos eximius Verdoorn, female cone on left, Story 4021 ; male

cone on right, Story 4019; from plants on hills between Bedford and Tarkastad.

[Photo H. King.

9102-9

450

Plate 5. — Pelargonium jacobii R. A. Dyer.

451

Plate 6. — Aloe babatiensis Christian & Verdoorn.

[Photo G. W. Reynolds.

Entorrhiza calospora sp. nov., and Some Other

Parasitic Fungi in Limeum Roots.

By

P.H. B. Talbot.

Summary.

A fungus, described here as Entorrhiza calospora sp. nov.* and assumed to belong

to the Tilletiaceae in the absence of proof by germination of the chlamydospores, is

associated with large galls on the roots of Limeum glomeratum, L. viscosum and

Trianthema pentandra in the Transvaal and on L. viscosum from Ficksburg, O.F.S.

The galls range in diameter from less than 1 mm. to 3-5 cm. They are composed of

parenchymatous cells in which the fine mycelium, often coiled, bears globose, strongly

verrucose, binucleate chlamydospores 17-4-22 p in diameter, produced singly but

forming groups of up to twenty per host cell. The dry chlamydospores are lemon

yellow to golden in a mass. A dusty sorus of chlamydospores is not formed; instead

the spores are liberated by the gradual disintegration of the galls when the host dies

in autumn. Germination of the spores and re-infection of the hosts were not observed.

The parasite does not appear to harm the plant as a whole, since its aerial parts show

no signs of disease nor of stimulation. Olpidium brassicae (Woronin) Dangeard,

Ligniera junci (Schwartz) Maire & Tison, and an unidentified Phycomycete, were also

found in the roots of a single plant of Limeum viscosum which bore Entorrhiza galls.

Occurrence of Entorrhiza calospora.

In 1943, Mr. J. J. O. Pazzi collected some specimens of Limeum glomeratum at

Brummeria, near Pretoria, and brought them to the Division of Botany and Plant

Pathology. On their finer roots were many large galls whose cells contained abundant

warted resting-spores which were thought at the time to belong to a Phycomycete.

The following year Mr. Pazzi again sent specimens from the same area. These were

referred to a student of Phycomycetes, but without a full knowledge of the life-cycle

of the fungus this person rightly would not commit himself to an identification,

replying that although the fungus appeared to be a Phycomycete even the youngest

galls lacked stages of the fungus prior to the formation of resting-spores.

A morphologically identical fungus in similar galls on the roots of Trianthema

pentandra was collected by Mr. B. N. Wolff near Malati in 1945. The subject was then

dropped until 1953, when the writer became interested and endeavoured to locate

fresh material to study, beginning his search at the Botanical Reserve, Pretoria

University Farm, which is close to Brummeria. This report deals only with the galls

on Limeum viscosum from the Botanical Reserve, and on L. glomeratum from Brummeria.

Miss M. Henderson kindly identified the host plants.

* Entorrhiza calospora sp. nov.

Sori radicicoli, tumores globosi vel elongati usque ad 3-5 cm. diam. inducentes. Hyphae

intracellulae, hyalinae, 10-2-6 p diam., tenuissime tunicatae, apice ramulorum lateralium

chlamydosporas magnas generantes. Chlamydosporae globosae, hyalinae vel luteae, binucleatae,

17-4-22 /x diam.; episporio verrucis prominentibus, 1 -3-3 p crassis, 2-4 p altis, exomato. Hab. in

radicibus Limei glomerati, L. viscosi, Trianthemae pentandrae. Typus: Herb. Pretoriae No. 33770,

leg. J. J. O. Pazzi.

129828-2

454

Limeum viscosum is common on the ridge included in the Botanical Reserve, and

was abundant for a season on a patch of ploughed sandy soil west of the ridge. Later,

veld grasses invaded the ploughed soil and ousted the Limeum plants. The host plant

also grew along a little-used road on the south-east side of the ridge. From all these

sites Limeum plants were dug up and a rather small proportion of them bore the galls

that were being sought. The presence of the galls could not be inferred from the

aboveground appearance of the plants. On the other hand most Limeum plants at the

Reserve had insect-galls or thickenings at the collar or on the stems. Sections of these

thickened tissues were free of fungus but occupied by insect larvae.

Limeum viscosum is an annual herb of prostrate, procumbent or decumbent habit,

and favours cultivated or disturbed sandy soils. According to herbarium records, it

is commonest on sandy or gravelly soils. Fungus-galls are more easily found on those

plants growing in deep sand where the roots are better developed than in shallow

gravelly soils.

Phillips (1951) places Limeum in the Phytolaccaceae and Trianthema in the

Aizoaceae, two adjacent families in the classification he adopts. In Limeum he notes

eight species widely distributed in South Africa, while in Trianthema there are four

species, some of which occur in each Province.

When the writer began this work the following herbarium specimens of the

fungus-galls were available in the National Herbarium: 33770, Pazzi, on Limeum

glomeratum, Brummeria, 1943; 35291 and 35391, Pazzi, on L. glomeratum, Brummeria,

Spring 1944; 39031, Wolff, on Trianthema pentandra, Mahale, P.O. Malati, 1945.

In the phanerogamic section of the National Herbarium, the writer found a specimen

of Limeum glomeratum collected by H. B. Terry, 30/1/1943, sine loc., with typical galls

on its roots. Finally specimens have been received from Ficksburg, O.F.S., on

L. viscosum collected by Dr. Meredith, March, 1955.

With the preconceived idea that the fungus was a Phycomycete, the writer spent

some time collecting galls of all sizes and dissecting, sectioning and examining them

in an attempt to trace stages of a Phycomycete life-cycle, but only mycelium and

resting-spores were found. He accumulated a good deal of information about these

without coming any nearer to classifying the fungus, and all attempts at germinating

the chlamydospores failed. Then a paper by Schwartz (1910) came to hand, and at

once the imposing similarity between the Limeum fungus and Entorrhiza was evident.

Technique.

The freshly collected galls were fixed in the field and sectioned after embedding

in paraffin. The following fixatives were tried: Formol-alcohol, Carnoy, Gilson,

Bouin, and Chicago Chrome-acetic-osmic acid, according to the formulae given by

Chamberlain (1915). Fixation for two hours in Carnoy or twenty-four hours in Bouin

gave satisfaction, the latter possibly being the better.

Paraffin sections were cut in thicknesses between 3 p and 15 p. Sections about

7 p thick were generally the most useful, but those at 3 p were necessary for seeing the

contents of the chlamydospores. The mycelium was of course only seen in short

lengths in the sections, and was best studied by teasing out coarse hand sections mounted

in warmed lactic acid with cotton blue as stain, or in Cartwright’s picro-aniline blue

(Cartwright, 1929). The stained, teased out portions were washed and mounted in

lactic acid. The cover glass was ringed with nail lacquer. In such preparations the

mycelium was easily seen but its cytological details were not clear.

Fifteen common stain combinations were tested on the paraffin sections. Many

gave pleasing colour combinations but insufficient cytological detail. The cytology

was well shown, however, by using Heidenhain’s haematoxylin with a light counterstain

of Orange G, or with Giemsa stain, or with Safranin O counterstained with Delafield’s

haematoxylin. In the last combination it was found that after differentiating the

haematoxylin in acid alcohol it was desirable to intensify the remaining coloration

455

with ammonia alcohol to bring out the colour in the hyphae and host nuclei

satisfactorily. Used in this way, Safranin and Delafield’s haematoxylin was perhaps

the best of the combinations tested.

Description of the Galls and of E. calospora. (Figs. 1-38).

The galls measure less than 0-5 mm. to 3-5 cm. in diameter. The smaller galls

are rounded and more or less smooth. They may be formed laterally on a rootlet,

or may surround the root thus later interrupting the continuity of the conducting

system. A laterally formed gall may often have only the slightest connection with

the root, whose structure then remains quite unimpaired. Here one is reminded of

the galls of Sorodiscus radicicolus described by Ivimey Cook (1931) except that in the

Entorrhiza galls the youngest stages of the fungus are not necessarily found adjacent

to the root. Mycelium and young chlamydospores were present even in the smallest

galls. A root may bear a linear series of galls which coalesce on further growth into

an irregularly elongated mass. Some of the larger galls are obviously formed like

this, but others with an irregularly globose form and a tubercular surface are apparently

formed by prolonged growth of a single gall.

The galls consist of parenchyma cells and interrupted conducting tissues. The

host nuclei are somewhat enlarged and clearly seen in young galls, but later disappear.

The interior of an old gall is pale yellowish, marbled or granular, owing to

innumerable chlamydospores occupying practically all the parenchyma cells. A dusty

sorus is not produced. Scattered and displaced Xylem elements may be present

depending upon the degree of root involvement, but the Xylem never contains

chlamydospores or mycelium. A very thin peripheral layer of suberised tissue may

surround old galls.

The galls of E. calospora are far larger than those of any other species of Entorrhiza .

Descriptions of other species mention galls tending to be ovoid, cylindrical, or digitate,

and none measuring more than 3 mm. wide and 10 mm. long.

The hyphae of E. calospora are colourless, 1-0-2 -6 p in diameter, with very thin

walls. They are intracellular, but may perhaps be intercellular too, though this

condition was not seen. The hyphae are often continuous with the cell walls of the

host. Within the cells they are sharply and irregularly bent at intervals, or spirally

coiled. Coiled hyphae are typical of those bearing chlamydospores. The coiled and

contorted hyphae form unentangleable masses which commonly enmesh the host

nucleus, and so close is their connection that the hyphae may appear to arise from the

host nuclei. The hyphae vary very much in diameter over short distances. Their

contents are difficult to observe as they resist differential staining and their alternately

strongly and weakly refractile zones appear respectively under- and overstained. Minute

dark-staining granules, possibly nuclear material, are scattered in the hyphal cytoplasm

which is itself either dense or vacuolate. The hyphae are sparingly branched, with

occasional septa, and the chlamydospores are formed singly and terminally on lateral

branches, which are somewhat narrower than the purely vegetative mycelium.

The chlamydospores are pale lemon to golden in a dried mass when mature, but

appear almost hyaline and strongly refractile in the microscope. Early in their

development, as soon as they are recognisable as spores, they are more or less globose

with a lobed outline of strongly refractile warts. The warts are scanty at first but

later cover the whole surface and become very conspicuous. The mature chlamydospores

are globose, 17-4-22 p in diameter (including the warts). The warts are 2-4 p high

and 1 • 3-3 p broad at the base. They are cylindrical or conical with a rounded apex

which is narrower than the base. Some rather dark chlamydospores have warts which

differ in being abruptly narrowed towards the apex. The wall of the chlamydospore

456

totals 3 -3-4 -7 p in thickness and consists of three layers, the outermost warted layer

and two smooth inner layers. Refractile (? oily) masses or vacuolate cytoplasm fills

the chlamydospore. Two relatively large nuclei, 1-3 yu. diameter, are present in the

mature spores, and the vacuolate cytoplasm often carries a few small dark-staining

granules. As many as twenty chlamydospores have been seen in a single host cell;

the average number is 4-5 per cell. The greater numbers occur in greatly enlarged

cells. It is possible that the latter are formed by the breaking down of the intervening

walls of several smaller cells, but only in one instance was this clearly seen. The

immature chlamydospores stain deeply with all the common stains; the mature ones

are more resistant to stains but show up well on account of their refractility.

Galls appear in early spring and continue to grow, new galls also being formed,

throughout the summer. In the autumn, about March, the host plants begin to die

and the galls dry out and quite rapidly disintegrate into a powder composed of

aggregations of chlamydospores and pieces of the softer host tissues, often stranded

together by the more resistant Xylem elements of the host.

Germination Experiments.

In the foreword to Ainsworth & Sampson’s “British Smut Fungi’’, 1950, the

Director of the Commonwealth Mycological Institute wrote, “ The compilation of

this monograph focuses attention on the gaps in our knowledge of the germination

of many of the species and it is hoped that its publication will stimulate interest in

this group of fungi, which is of such great importance to agriculture.” In the text of

the book, the authors write of Entorrhiza (p. 88), “ Spore germination is not well

known ...”

The writer has tried several times to germinate the chlamydospores of E. calospora

using spores taken from fresh material, from newly disintegrated galls and from

disintegrated galls which have been placed in a refrigerator for one month at 8° C.

in an attempt to simulate overwintering. After refrigeration some of the material

was also left at room temperature for the months of June-August inclusive. The

refrigerated galls were divided into two batches kept respectively in dry and moist

sand. Hanging drop cultures in water, acidified water, and Czapek solution were

made, and chlamydospores deposited in a thin film on slides were exposed to alternately

wet and dry atmospheres, but none of these methods resulted in germination after

four weeks. The spores that were used appeared to be mature but this point certainly

merits further investigation. Work on germination is continuing.

Synopsis of the genus Entorrhiza C. Weber.

This genus was erected by Weber (1884) to accommodate the smut described

by Magnus (1878) as Schinzia cypericola, which causes root swellings on Cyperus

flavescens. The genus Schinzia Naegeli (Naegeli, 1 842) had been based on two uncertain

species found in Iris roots, and moreover was a later homonym of Schinzia Dennstatt

(1818). Entorrhiza cypericola (Magn.) Weber is thus the type species of the genus

Entorrhiza. But in his work, Weber wrongly attributed the root swellings on Cyperus

flavescens and Juncus bufonius to the single species E. cypericola, and his biological

observations were made on the material of Juncus bufonius. Magnus (1888) corrected

this error by showing that the smut on C. flavescens had finely reticulate or punctate

spores, whereas that on J. bufonius differed in having coarsely warted spores. Still

accepting the genus Schinzia, Magnus named the smut on Juncus as Schinzia

aschersoniana Magn. In Aug. 1888 Lagerheim (1888) made the combination Entorrhiza

aschersoniana (Magn.) Lagerh. and was followed independently in this by De Toni

in Oct. 1888 (De Toni, 1888).

457

Nine species described or combined under Entorrhiza have been traced in the

literature, namely: —

E. cypericola (Magnus) Weber in Bot. Zeit. 42 (1884) 370.

= Schinzia cypericola Magnus in Verh. bot. Vereins Brandenburg 20 (1878) 53;

Magnus in Ber. deutsch bot. Gesellsch. 6 (1888) 102.

E. aschersoniana (Magnus) Lagerheim in Hedwigia 27 (1888) 262.

= Schinzia aschersoniana Magnus in Ber. deutsch bot. Gesellsch. 6 (1888) 103.

= E. cypericola (Magnus) Weber, pro parte, in Bot. Zeit. 42 (1884) 370.

E. casparyana (Magnus) Lagerheim in Hedwigia 27 (1888) 262.

= Schinzia casparyana Magnus in Ber. deutsch bot. Gesellsch. 6 (1888) 103.

E. digitata Lagerheim in Hedwigia 27 (1888) 264.

E. cellulicola (Naegeli) De Toni in Saccardo Syll. Fung. 7 (1888) 498.

= Schinzia cellulicola Naegeli in Linnaea 16 (1842) 281.

E. solani Fautrey in Rev. de Myc. (1896) 11.

E. scirpicola (Correns) Sacc. & Syd. in Saccardo Syll. Fung. 14 (1899) 425.

= Schinzia scirpicola Correns in Hedwigia 36 (1897) 40.

E. caricicola Ferdinandsen & Winge in Dansk Bot. Arkiv. 11 (1914) 10.

E. raunkiaeriana Ferdinandsen & Winge in Dansk Bot. Arkiv. 11 (1914) 8.

Of the above, E. cellulicola in Iris roots and E. solani in Solarium tuberosum are

doubtful members of the genus. Until now, all species referred with certainty to

Entorrhiza have been found exclusively on members of the Juncaceae and Cyperaceae.

E. casparyana with globose, strongly verrucose chlamydospores 17-22 g diameter, is

the only one sufficiently like the fungus in Limeum root galls to require close comparison.

It is doubtful if authentic material of E. casparyana exists. In these circumstances,

and also in view of the difference in host families, the writer was advised by Dr. M. B.

Ellis of the Commonwealth Mycological Institute to describe the fungus on Limeum

as a new species. Such treatment is in accordance with the species concept in Smut

fungi recently proposed by Fischer & Shaw (1953), whose views are summarised as

follows (in Rev. Appl. Myc. 33, 1954, 116): “Any two or more smuts of similar

morphology and symptomatology, parasitising different species and genera of the

same host family, would be regarded as belonging to one morphological species. On

the other hand, smuts of comparable morphology attacking members of different host

families would be deemed distinct species.” As a practical convenience this treatment

has much to commend it. Nevertheless, the writer is disinclined to propose yet another

new species for the Entorrhiza present in Trianthema when it is so patently

indistinguishable from that in Limeum.

Other parasitic fungi of Limeum viscosum roots.

Some L. viscosum plants potted with their own soil at the Botanical Reserve were

kept in the laboratory to watch the development of galls known to be on their roots.

During subsequent transfer of one plant to another pot, some of its rootlets were

sampled to see whether the root hairs might contain sporidia of Entorrhiza. These

were not seen, but instead the following fungi were found: (1) a single rootlet infected

with Ligniera junci (Schwartz) Maire & Tison; (2) abundant rootlets heavily infected

with Olpidium brassicae (Woronin) Dang. ; (3) several hypertrophied root hairs infected

with an unidentified Phycomycete in the resting-sporangium stage; (4) a rootlet with

a perithecium of Chaetomium adhering to it; (5) several roots bearing a mantle of

sterile mycelia, both brown and hyaline. The sterile mycelia and the Chaetomium

were evidently only fortuitously associated with the rootlets and were not investigated

further. The first two parasitic fungi mentioned above are now described.

458

Ligniera junci (Schwartz) Maire & Tison in Comptes Rendus Acad. Sci. Paris

152 (1911) 206, Ann. Mycol. 9 (1911) 235; Ivimey Cook in Trans. Brit. Myc. Soc.

1 1 (1926) 196-213, PI. 8, Archiv fur Protistenkunde 80 (1933) 222-227, PI. 1 1 ; Karling,

The Plasmodiophorales (1942) 60, PI. 11.

Sorosphaera junci Schwartz in Annals of Bot. 24 (1910) 512, PI. 40.

See Figs. 39-41.

A small lateral rootlet was seen with twelve cystosori in the cortical tissues. The

root was not hypertrophied. The cystosori appeared solid and were globose or

ellipsoidal. They were rather small, the largest measuring 17 x 22 p. They consisted

of closely aggregated resting spores lacking a common membrane. In the smallest

cystosorus ten resting spores were visible, in the largest twenty-five, but other spores

were hidden behind those that could be seen. The resting spores were 3 -8-6-1 p in

diameter, and varied somewhat in shape. Basically they are globose but the central

ones become angled and the peripheral ones wedge-shaped by compression. Some of

the peripheral spores have a slightly concave outer wall giving them a bluntly bicuspid

appearance. Such spores are illustrated in figs. 14, 15, 17, 20 of Schwartz’s paper

(Schwartz, 1910, loc. cit.). The spore walls are smooth, very thin and hyaline. No

zoosporangia or other stages in the life-cycle were seen in the very limited material

available.

The form of the cystosori was not well shown in the single preparation in which

this fungus was found, but there can be little doubt of its identity with Ligniera junci.

Our knowledge of the life-cycles, host relationships and taxonomy of the species of

Ligniera is quite fully covered in the literature cited above. Probably only two good

species of Ligniera are known (Cook, 1933), namely L. junci and L. verrucosa

Maire & Tison, the latter having verrucose resting spores. Karling (1942) however,

includes a further two species, L. pilorum Fron & Gaillat and L. isoetes Palm, which

he regards as possibly distinct from L. junci.

L. junci attacks many different hosts, and consequently has been described under

several different names. Cook (1933, p. 227) lists the recorded hosts of L. junci and

concludes, rather sweepingly, “ From such a variety of hosts it seems probable that

this fungus is able to infect any vascular plant under suitable environmental conditions.”

Olpidium brassicae (Woronin) Dangeard in Ann. Sci. Nat. Bot. ser. vii,

4 (1886) 242-343; Sampson in Trans. Brit. Myc. Soc. 23 (1939) 199-205, figs. 1-22,

PI. 5.

Chytridium brassicae Woronin in Jahrb. Wiss. Bot. 11 (1878) 548-574.

See figs. 42-51.

Abundant material was found consisting of solitary or aggregated zoosporangia

occupying the cortical tissues of small lateral roots without causing hypertrophy.

Zoosporangia globose, 12-5-70 p diam., or elongated 10-36 X 32—194 p, with

homogeneous or large globular contents, thinwalled or sometimes somewhat thickened.

Exit tubes of the zoosporangia are 2 -9-8-0 p diameter and 4-6-23 p long depending

on the distance to the exterior of the host. Zoospores more or less globose, about

3 p diam. (with a single flagellum up to 17 p long fide Sampson, loc. cit.). Resting

sporangia (cysts) are angular-globose, 13-6-28-5 p diam., or sometimes elongated and

then angular-ellipsoid up to 22 x 29 p, with an exospore which is coarsely and deeply

ridged giving a stellate appearance in optical section, with 7-9 points to each star;

endospore smooth, thinwalled.

Our material did not show the full life-cycle of this fungus, nor any sporangia

bearing more than one exit tube. Preparations stained with Heidenhain’s haematoxylin

showed the ridges of the cysts very well and we were able to confirm Miss Sampson’s

observations on their morphology. Miss Sampson has made an excellent case for

459

reducing Asterocystis radicis De Wild., Olpidium borzii De Wild., Olpidium radicicolum

De Wild., and Olpidiaster radicis (De Wild.) Pascher to synonymy with Olpidium

brassicae. The reader is referred to her paper for an authoritative account of the

life-cycle, morphology and taxonomy of Olpidium brassicae, and to another paper

(Sampson, 1932) for differentiation of the species of Olpidium which have been recorded

on higher plants.

As neither Ligniera junci nor Olpidium brassicae had previously been recorded

in South Africa, attempts were made to secure further material by planting cauliflower

seeds in the soil from which these fungi had been isolated. The seedlings were kept

rather wet, which favours the development of O. brassicae according to Miss Sampson,

but they remained uninfected.

Perhaps the greatest interest concerning this discovery of L. junci and O. brassicae

lies in their association in the roots of a single plant, which also bore Entorrhiza galls.

We believe that L. junci and Entorrhiza have not been recorded in association with

one another since L. junci was originally described (Schwartz, 1910, as Sorosphaera

junci), when it occurred with species of Entorrhiza in the roots of Juncus bufonius,

J. articulatus and J. lamprocarpus. On the other hand, associations of Ligniera spp.

with Olpidium brassicae, and of other Plasmodiophoraceae and chytrids, are well

documented (Maire & Tison, 1911, b, pp. 241-242; Guyot, 1927; Ivimey Cook, 1927,

1933; Bartlett, 1928; Fitzpatrick, 1930, pp. 62-63; Karling, 1942, pp. 64-68). Often

such fungi were found occupying the same tissues in a single root and not unnaturally

were thought to be a single organism in various stages of development. Thus certain

species and genera were proposed (e.g. Rhizomyxa hypogaea Borzi, Sorolpidium betae

Nemec and Anisomyxa plantaginis Nemec) which now appear to have been based on

the accidental association of two different species. The identities of some of the

component fungi of such “ mixed species ” are still debated, but it is now generally

accepted that the infection is indeed a mixture of species and not a single organism in

various developmental stages. Cook (1933, p. 241), giving several examples of

associated chytrids and Plasmodiophoraceae, states, “ The use of Sorolpidium betae,

Anisomyxa plantaginis and Rhizomyxa hypogaea as evidence of a relationship between

Plasmodiophoraceae is unsatisfactory as these species are mixtures.” No one who

has seen this phenomenon of mixed infection is likely to contest Cook’s statement.

Literature Cited.

Bartlett, A. W

Cartwright, K. St. G.

Chamberlain, C. J

Cook, W. R. I

De Toni, G. B

Fischer, G. W. & Shaw, C. G.

Fitzpatrick, H. M

Guyot, A. L

Karling, J. S

Lagerheim, G

Magnus, P

Maire, R. & Tison, A

Naegeli, C. W. von

Phillips, E. P

Sampson, K

Schwartz, E. J

Weber, C

(1928). Trans. Brit. Myc. Soc. 13, p. 221.

(1929). Annals of Botany 43, p. 412.

(1915). Methods in Plant Histology, 3rd Ed.

(1927). Trans. Brit. Myc. Soc. 12, p. 282.

(1931). Ann. Mycol. 29, p. 313.

(1933). Archiv fur Protistenkunde 80, p. 179.

(1888). Saccardo Syll. Fung. 7, p. 498.

(1953). Phytopathology 43, p. 181.

(1930). The Lower Fungi, Phycomycetes.

(1927). Rev. Path & Entomol. Agric. 14, p. 176.

(1942). The Plasmodiophorales, 1st Ed.

(1888). Hedwigia 27, p. 261.

(1878). Verh. hot. Vereins Brandenburg 20, p. 53.

(1888). Ber. deutsch bot. Gesellsch. 6, pp. 100-104.

(1911). Comptes Rendus Acad. Sci. Paris 152, p. 206.

(1911, b). Ann. Mycol. 9, pp. 226-246.

(1842). Linnaea 16, p. 281.

(1951). The Genera of South African Flowering Plants, 2nd Ed.

(1932). Trans Brit. Myc. Soc. 17, pp. 190-191.

(1910). Annals of Botany 24, pp. 511-522, PI. XL.

(1884). Botanische Zeitung 42, pp. 370-379, PI. IV.

460

Explanation of the Figures.

Figs. 1-38: Stages in the development of Entonhiza calospora.

Figs. 1, 2, 6, 11, 13, 14, 15, 17, 34, show the relationship of the host cell nucleus and

the hyphae. Note that the host nucleus is often enmeshed by a tangled knot

of hyphae.

Figs. 1-6, 10-12, 14-18, 25-38, show the hyphae of abruptly bent and contorted form.

The characteristic spiral coiling of the hyphae is seen in figs. 16, 28, 32, 33, 38.

Hyphae knotted into unentangleable masses are depicted in figs. 1, 2, 6, 11, 15,

17, 18.

Fig. 24 represents a host cell containing seventeen chlamydospores and presents the

only evidence seen of fusion between adjacent cells to form larger ones containing

the greatest numbers of chlamydospores.

Figs. 7-10, 13-18, 25-38, trace the development of the chlamydospores. The youngest

chlamydospores, globose with an indented, sparsely verrucose outline are seen

in figs. 7-10, 25-30, 33, 37, 38. Older stages with strongly verrucose exospores

are also shown in some of these and in the remaining illustrations of the

chlamydospores. Figs. 19-22 depict mature chlamydospores in section, showing

the wall structure and contents. The binucleate condition of the chlamydospores

is shown in figs. 19-20.

Figs. 39-41 : Cystosori of Ligniera junci composed of closely aggregated resting spores.

Figs. 42-51: Stages in the life-cycle of O/pidium brassieae. Zoosporangia with exit

tubes are shown in figs. 42-45, 47-50. Young zoosporangia without exit tubes

are seen in figs. 45-46. Stellate resting sporangia (cysts) are depicted in fig. 51.

461

462

463

465

The Cyphelloid Fungi of South Africa.

By

P.H.B. Talbot.

The recorded species of South African cyphelloid fungi are examined in this paper.

These are included in the genera Aleurocystus, Aleurodiscus, Cyphella, Cytidia,

Dendrothele, Gloeosoma, Porotheleum, Solenia and Stigmatolemma, some of which

are invalid names requiring comment. These genera, with several others, have been

classed among the families Thelephoraceae, Cyphellaceae and Leptotaceae, but

similarities in their structure permit of their treatment together as “ cyphelloid fungi ”,

until such time as their true relationships have been clarified.

Extensive use has again been made of Doidge’s check-lists of South African fungi

(in Bothalia 5, 1950). Herbarium numbers, unless otherwise stated, refer to specimens

in the National Herbarium, Pretoria. Genera and species are considered in alphabetical

order. Synonyms or doubtful records are printed in italics, and accepted species in

bold face. Keys to the accepted species are given at the end of the paper.

Grateful thanks are due to Dr. Wm. Bridge Cooke for reading the manuscript,

suggesting some alterations that have now been made, and generously allowing

publication here of two new species that he has described from South African material.

Some peculiarities of nomenclature that remain in the paper are the author’^

responsibility and are explained in the appropriate places.

The writer is indebted to the following for the loan of specimens: The Curator

of Fungi, Smithsonian Institution; The Director, Royal Botanic Gardens, Kew;

The Director, Museum National D’Histoire Naturelle, Paris. The Trustees of the

P. A. van der Byl Herbarium, University of Stellenbosch, kindly authorised a visit

to the Herbarium to examine specimens in their keeping.*

ALEUROCYSTUS Lloyd.

Aleurocystus Lloyd (as McGinty) in Lloyd Myc. Notes 6 (1921) 1088, nomen

provisiorum. Further referred to in: Stevenson & Cash in Bull. Lloyd Library

No. 35 (1936) 42; Martin in Lloydia 5 (1942) 162; Donk in Reinwardtia 1 (1951)

204-206.

Donk (loc. cit.) reasons that the names mischievously proposed by Lloyd under

the pseudonym “ N. J. McGinty ” are to be regarded as nomina provisioria, and that

under Art. 37 of the International Rules of Botanical Nomenclature their publication

was invalid. This is certainly so regarding Aleurocystus, which Lloyd only suggested

could be a segregate based on Aleurodiscus capensis, whose texture and cystidia set

it apart from other species of Aleurodiscus. Donk also holds that Stevenson & Cash’s

(loc. cit.) reference to Aleurocystus amounts to “ incidental mention ”, and again

does not constitute valid publication under the same Article of the Rules. Stevenson &

Cash listed the combination Aleurocystus capensis, attributing it to Lloyd who, however,

made it only by implication and not literally.

* Deur die goedgunstigheid van die Universiteit van Stellenbosch is ’n geleentheid aangebied om

eksemplare in die Herbarium P. A. van der Byl te ondersoek. Vir sy hoflikheid en hulp gedurende my

besoek wil ek graag dr. J. G. Smith bedank. Verdere verwysing na verskillende soorte in die van der

Byl-Herbarium word hieronder gemaak.

466

It seems clear that the names Aleurocystus and Aleurocystus capensis are nomina

provisioria and therefore invalid. For Aleurocystus capensis see under Cytidia habgallae

(B. & Br.) Martin.

ALEURODISCUS Rabenh. ex J. Schroet.

Aleurodiscus acerinus (Pers. ex Fr.) Hohnel & Litsch. in K. Akad. Wiss. Wien Sitzungsb.

116 (1907) 804, Tab. 2, f. 6; Doidge loc. cit. p. 479.

Corticium acerinum Persoon, Obs. Myc. 1 (1796) 37.

Thelephora acerina (Pers.) Pers. ex Fr., Syst. Myc. 1 (1821) 453, Hym. Eur. (1874) 648;

Persoon, Syn. Fung. (1801) 581, Myc. Eur. 1 (1822) 152; Leveille in Ann. Sci.

Nat. ser. iii, 5 (1846) 150.

Stereum acerinum (Pers. ex Fr.) Fr., Epicrisis (1838) 554; Saccardo Syll. Fung. 6 (1888)

587.

Saccardo’s reference to South African material does not specify a collection or

locality, and cannot be traced further. Leveille published Drege 9452 from the Cape

as Thelephora ( leijostroma ) acerina Pers. This collection is not deposited in the Paris

Natural History Museum, and consequently it has not been possible to decide whether

the species is represented in South Africa. The species differs from the variety

A. acerinus var. alliaceus in the size and particularly the shape of its spores. A. acerinus

has spores which are ovoid, with a ratio of length/width of about 1-6 (always less than

2-0), and a size of 5 -6-6 -4 x 8-0-12 y. (6-7 x 10-13 /x fide Hohnel & Litschauer).

Aleurodiscus acerinus (Pers. ex Fr.) Hohnel & Litsch. var alliaceus (Quel.) Bourd. & Galz.

in Bull. Soc. Myc. de Fr. 28 (1912) 352, Hym. de Fr. (1928) 334; Pilat in Ann.

Myc. 24 (1926) 217, Tab. 15, f. 11.

Corticium alliaceum Quelet, F. Myc. de Fr. (1888) 5; Saccardo Syll. Fung. 6 (1888)

629.

Aleurodiscus acerinus (Pers. ex Fr.) Hohnel & Litsch. var longisporus Hohnel & Litsch.

(!) in K. Akad. Wiss. Wien Sitzungsb. 116 (1907) 805, PI. 2, f. 4; van der Byl

in Ann. Univ. Stellenbosch 7 (1929) 25; Doidge loc. cit. p. 479; Talbot in Bothalia

6 (1951) 26, PI. 18.

Fig. 1.

Since seeing more material of the Type of A. acerinus var longisporus (viz. MacOwan

1074, distributed under Nos. 20840 and 21947 in the National Herbarium, Pretoria)

the writer has found it necessary to emend the description and comments given

previously on this variety (Talbot, 1951, loc. cit.). Further, accepting the synonymy

of A. acerinus var. longisporus and A. acerinus var. alliaceus, the latter name should

be used as it has priority.

After much search, eight spores were found in the Type of var. longisporus. These

measured 4-8 x 9—17 - 6 /x (smallest 4 X 9/x; two largest 8 x 16 /x and 6-4 x 17-6 /x).

They were long-elliptical, or almost cylindrical, with one side usually somewhat

depressed. Their length was 2 -0-2 -8 times as great as their width. In European

material of A. acerinus spores were found to have a length 1-4-1 -9 times as great as

their width. Literature records of the spore measurements of the species and its

varieties were analysed, with the following results: —

A. acerinus : Ratio length/width = 1-4-1 -8: average 1-6.

A. acerinus var. longisporus : Ratio length/width = 1 -9-2-9: average 2-4.

A. acerinus var. alliaceus : Ratio length/width = 2 -0-2- 5: average 2-25.

467

It appears from this that the spores of A. acerinus are more ovoid (besides being

usually shorter) than those of the varieties longisporus and alliaceus, thus substantiating

a varietal difference from the main species. Also it is evident that the spores of the two

varieties have essentially the same shape and order of size, and that only one variety,

alliaceus being its prior name, is merited.

If only the size ranges of these various spores are taken into account, and their

shape (i.e. ratio length/width) is disregarded, then the spores of the main species may

fall within the size range of the spores of the varieties, and the erection of varieties

would be questionable. For example, the spores of A. acerinus have a range of

6-7 x 10—13 p (fide Hohnel & Litschauer) which lies within the range of 4-8 x 9-17-6 p

measured by the author for spores of A. acerinus var. longisporus. Yet the spore shapes

are distinct and the varietal difference is legitimate.

An emended description of A. acerinus var. alliaceus follows: —

Fructifications resupinate, irregularly expanded, crustose, adnate, chalk-white

becoming pallid ivory-white, finely and closely cracked, having a determinate margin;

in section up to 80-100 p thick.

Basidia: cylindrical to subclavate with a wavy outline, 35-60 x 5-8-5 p, with 2 or 4

sterigmata.

Spores: long elliptical, often with one side flattened, or subcylindrical, more than

twice as long as broad, 4-8 x 9-17-6 p, smooth, hyaline.

Paraphyses: (1) basidia modified into pseudophyses which are bulbous with a short

to long, narrow, apical outgrowth; 10-15 x 25—35 p, with the outgrowth up to

16 x 3 p, (2) racemosely branched paraphysoids (dendrophyses) between and beyond

the basidia, very heavily encrusted with minerals, extensions of the hyphal system,

2 p or less in width.

Hyphae: 2 p or less in width, hyaline, erect, much branched and closely arranged

in the basal part, very heavily encrusted, emerging as dendrophyses.

Specimens examined: 20840, 21947 (both MacOwan 1074, type of A. acerinus var.

longisporus ). Two other collections, 27566 and 39058 may belong here, but no spores

were seen with certainty.

Aleurodiscus acerinus (Pers. ex Fr.) Hohnel & Litsch. var. longisporus Hohnel & Litsch.

(!) in K. Akad. Wiss. Wien Sitzungsb. 116 (1907) 805, PI. 2, f. 4; van der Byl

in Ann. Univ. Stellenbosch 7 (1929) 25 ; Doidge loc. cit. p. 479 ; Talbot in Bothalia

6 (1951) 26, PI. 18.

= Aleurodiscus acerinus (Pers. ex Fr.) Hohnel & Litsch. var. alliaceus (Quel.)

Bourd. & Galz. See previous entry.

Aleurodiscus botryosus Burt in Ann. Mo. Bot. Gard. 5 (1918) 198, f. 10; Bourdot &

Galzin, Hym. de Fr. (1928) 332, f. 102.

Fig. 3.

Fructifications effused, resupinate, thin firmly adherent, white then later straw-

coloured, smooth, becoming cracked when old into small areas; margin similar,

thinning out to pruinose. Up to 360 p thick in section. Fructifications sometimes

confluent.

Basidia: none seen mature, clavate, 60-70 x 8-10 p when immature.

Spores: hyaline, smooth, broad elliptical to ovoid, amyloid, 13-5-15 x 7-5-8 p.

Gloeocystidia : embedded in stages from the basal tissues to the hymenium, thinwalled,

with yellowish granular or homogeneous contents, subcylindrical, flexuous, or clavate

or pyriform, 50-110 x 9-17-(20) p.

468

Paraphyses: dendrophyses of a peculiar type found in the hymenium and in lower

tissues composed of a short main axis 1 • 5-4 p diam., with short lateral branches bearing

minute spherical granules which show a strong amyloid reaction.

Minerals: abundant throughout.

Hyphae: thinwalled, hyaline, 2-2-5 p diam., suberect, no clamps seen.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der Byl

nr. 2261, Knysna (as Aleurodiscus cerussatus).

The presence of yellowish gloeocystidia and dendrophyses together, distinguishes

this species from other Aleurodiscus species so far seen in South Africa. This fungus

differs from A. cerussatus, where v. d. Byl placed it, mainly in the nature of its

dendrophyses occurring throughout the tissues and bearing amyloid granules.

The accessory organs of A. cerussatus are to be classed rather as acanthophyses

(bottle-brush paraphyses) than dendrophyses, and are confined to the hymenium.

Aleurodiscus roseus is distinguished by the rose-coloured hymenium, lack of

gloeocystidia, presence of clamp connections and presence of normal type of

dendrophysis with large, irregular branches.

Aleurodiscus capensis Lloyd (!) in Lloyd Myc. Notes 6 (1920) 930, fig. 1687, ibid.

6 (1921) 1088 ; Stevenson & Cash in Bull. Lloyd Library No. 35 (1936) 43 ; Doidge

loc. cit. p. 484.

On examination of the type specimen (v. d. Byl No. 833 in the Lloyd Mycological

Collection No. 34029), the writer agrees with Martin (in Lloydia 5, 1942, 161-162)

and W. B. Cooke (in Mycologia 43, 1951, 208-209) that this species must be referred

to Cytidia habgallae (Berk. & Br.) Martin.

Aleurodiscus cerussatus (Bres.) Hohnel & Litsch. in K. Akad. Wiss. Wien Sitzungsb-

116 (1907) 807, PI. 4, f . 1 ; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 25,

f. 8; Doidge loc. cit. p. 479.

The specimen on which this record is based is Universiteit van Stellenbosch,

Herbarium P. A. van der Byl nr. 2261, Knysna. This specimen is considered to be

Aleurodiscus botryosus Burt, and consequently A. cerussatus should be omitted from

South African lists.

Aleurodiscus corneus (Lloyd) Lloyd (!) in Lloyd Myc. Notes 6 (1920) 930, f. 1688;

Stevenson & Cash in Bull. Lloyd Library No. 35 (1936) 43; Doidge loc. cit. p. 484.

Cytidia cornea Lloyd (!) in Lloyd Myc. Notes 5 (1917) 656, figs. 935-937; Doidge

loc. cit. p. 484.

On examination of the type of this species, Duthie 154 in the Lloyd Mycological

Collections No. 34063, the author agrees with Martin (in Lloydia 5, 1942, 161) and

W. B. Cooke (in Mycologia 43, 1951, 208) that the species should be referred to Cytidia

habgallae (Berk. & Br.) Martin.

Aleurodiscus disciformis (DC. ex Fr.) Patouillard in Bull. Soc. Myc. de Fr. 10 (1894)

80; Burt in Ann. Mo. Bot. Gard. 13 (1926) 305; v. d. Byl in Ann. Univ.

Stellenbosch 7 (1929) 24, f. 7; Doidge loc. cit. p. 479.

Thelephora disciformis DC. ex Fr., Syst. Myc. 1 (1821) 443.

Fig. 4.

Fructifications effused, resupinate, white, eventually greyish or yellowish and

cracked. Section about 500 p thick.

Basidia: 12-5-21 x 65-100 p, long-clavate, with 2-4 sterigmata up to 2-3 p wide

and 10-20 p long.

Spores: hyaline, smooth, 18-22 x 12-5-16 p, amyloid, ovate to subglobose, sometimes

showing a distinct apiculus.

469

Paraphyses: (1) gloeocystidioid, with a tendency to become irregularly swollen or

moniliform, but usually subcylindrical, hyaline, thinwalled. (2) simple cylindrical

paraphyses, thinwalled, 2 -5-3 -5 p diam.

Hyphae: 3-5 p diam., with rather rigid walls, colourless, encrusted with mineral

matter.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der Byl

nr. 658, on Podocarpus, Stellenbosch.

This specimen was evidently determined by Burt, who cited it, loc. cit. It appears

to fit A. disciformis. The author however, saw no submoniliform paraphyses 5-9 p

wide, which are seen in European material. This may have been due to the hurried

circumstances under which the examination was made in the v. d. Byl Herbarium,

for van der Byl mentions paraphyses of that sort in his description.

Aleurodiscus mirabilis (Berk. & Curt.) Hohnel in K. Akad. Wiss. Wien Sitzungsb.

118 (1909) 818.

Psilopeziza mirabilis Berk. & Curt, in Journ. Linn. Soc. Bot. 10 (1868) 364.

Aleurodiscus apiculatus Burt in Ann. Mo. Gard. 5 (1918) 186, f. 4. An extensive

synonymy is given by Rogers & Jackson in Farlowia 1 (1943) 267.

Fig. 3.

Fructifications resupinate, effused with a narrow reflexed margin, membranous

to coriaceous, pinkish when fresh, drying light yellow-brown or buff; hymenium even,

not cracking, minutely pruinose under a lens; margin reflexed and rolling slightly

inwards, villose or byssoid, pale coloured, almost white; thickness in section 350-600 p.

Basidia: clavate, up to 120-170 x 26-33 p, with four stout conical sterigmata up to

25 x 6 /u.; some basidia have a bulbous base decorated at the constriction with hyaline,

unbranched, blunt spines up to 4 p long.

Spores: hyaline, biapiculate, semilunate, up to 22-8 x 13-7 p, with one side depressed,

smooth (becoming minutely echinate, and 20-25 x 12-15 p, fide Burt, loc. cit.).

Paraphysoids: (1) very numerous acanthophyses, some representing aborted basidia,

others modified hyphae; clavate, hyaline, stoutwalled, with a long attenuated base,

of the same dimensions as the basidia, decorated towards the base of the clavate

part with hyaline spines. (2) hyphoid, hyaline, thickwalled, 53-66 p long, decorated

with hyaline aculeate spines, closely or sparsely arranged, the terminal branches of

some subhymenial hyphae. (3) simple, unbranches paraphysoids, hyaline, somewhat

flexuous, subcylindrical, about 130 x 8-9 p, some quite smooth, others tending towards

acanthophyses with a few basal spines.

Hyphae: hyaline, thickwalled, 4-5-3 p wide, with occasional clamp connections,

septate, branched, not encrusted.

Tissues: basal tissues composed of loosely intertexed hyphae which give the byssoid,

white appearance to the margin; other tissues compactly arranged.

Specimens examined: 36768, Talbot (54), on Hypericum, Qudeni.

As remarked by Rogers & Jackson (loc. cit.), the sublunate, biapiculate spores

and the abundant thickwalled long-appendiculate acanthophyses are the chief diagnostic

features of this remarkable species. In the present specimen there are minor variations

from Burt’s description, e.g. the presence of clamp connections and the fact that no

echinulate spores were seen. I have no doubts, however, in referring it to A. mirabilis.

When fresh, the rosy pink colour of the fungus, effused widely over the substratum,

was most conspicuous.

470

Aleurodiscus polygonioides (Karst.) Pilat in Ann. Mycol. 24 (1926) 221.

Corticium polygonioides Karsten, Symb. ad Myc. Fenn. 8 (1881) 12; Bourdot & Galzin

in Bull. Soc. Myc. de Fr. 27 (1911) 232, Hym. de. Fr. (1928) 227, f. 69.

Fig. 5.

Fructifications at first orbicular, resupinate, later confluent and widely effused,

remaining resupinate, membranous. Hymenium smooth, pruinose, eventually somewhat

cracked, purplish, violet or lilac coloured when fresh, becoming duller later or changing

to brownish or clay coloured. Margin whitish, pruinose when young, later concolorous

with the hymenium. Thickness in section 195-325 p.

Basidia: about 50-70 X 7-11 p but immature; ovoid to pyriform and clustered when

young, developing to irregularly clavate or subcylindrical.

Spores: (4-8)-6-4-8-(9-6) x (8-8)~10-4-14-4 /x, ovate or broad elliptical, with a

distinct apiculus, hyaline, smooth, not amyloid.

Paraphyses: Dendrophyses present in profusion, hyaline to faintly coloured, bearing

faintly coloured granules, arising from the tramal hyphae, their main axis 2-4 p wide.

Hyphae: hyaline, branched, septate, 2 -5-4 -8 p wide, with clamp connections, the

walls thin or a little thickened, lying more or less horizontally in the basal tissues,

giving rise to erect dendrophyses.

Specimens examined: 40226, H. King, on Lonicera, Pretoria; 36903, 39072, Talbot,

Hennops River; 36904, Talbot, on Buddelja salvifolia, Hennops River; 17817, F. Eyles

(2102), on Combretum, Salisbury, S.R. (as Stereum purpureum).

This species has much the same structure as Aleurodiscus roseus but is readily

distinguished by its colour. It is clear from Bourdot & Galzin’s (1928, p. 336) notes

that Aleurodiscus ionides is remarkably similar to A. polygonioides. I have disposed

the present specimens under the latter name solely on the strength of Bourdot & Galzin’s

comment that A. ionides differs from A. polygonioides in having a ceraceous texture,

not membranous, later becoming hardened and subcrustaceous and adhering strongly

to the substratum.

Aleurodiscus roseus (Pers. ex Fr.) Hohnel & Litsch. in K. Akad. Wiss. Wien Sitzungsb.

115 (1906) 1568; Pilat in Ann. Mycol. 24 (1926) 219; Rogers & Jackson in

Farlowia 1 (1943) 299.

Corticium roseum (Pers. ex Fr.) Fr., Epicrisis Syst. Myc. (1838) 560; Burt in Ann.

Mo. Bot. Gard. 13 (1926) 224; Bourdot & Galzin, Hym. de Fr. (1928) 227.

Thelephora rosea Pers. ex Fr., Syst. Myc. 1 (1821) 451.

A fuller synonymy is given by Pilat, Rogers & Jackson, and Burt.

Fig. 6.

Fructifications at first orbicular with a slightly raised margin, later confluent,

effused, resupinate with an appressed or raised margin. Hymenium light rose-coloured,

paling later, smooth, pruinose. Margin white, shortly fibrillose. Texture membranous

to pellicular. Thickness in section 200-220 p.

Basidia : discontinuous, remaining immature for a long time, 30-55 X 5-7 p (immature),

exceeded by the paraphyses, ovate to pyriform when young, becoming cylindrical to

clavate.

Spores: smooth, hyaline, ovate to broad elliptical, apiculate, 6 -4-8 -8 x 8-10-4 p.

Paraphyses : Dendrophyses abundant, their main axis about 2 • 4 p wide, richly branched,

arising from the lower hyphae, encrusted with minerals.

Hyphae: about 2-4 p wide, hyaline, branched, septate, not encrusted, with clamp

connections, the walls thin or a little thickened.

471

Specimens examined: All on Acacia mollissima : 39055, 39056, Talbot, Atholl Expt.

Stn.; 39054, Simpson & Talbot (31), Qudeni.

A. roseus falls in the same group as A. polygonioides but the two are readily

distinguished by their colour. The spore measurements given above, while falling

in the range for this species in Europe (8-16-5 X 6-10 u, fide Bourdot & Galzin),

are rather small.

CYPHELLA Fries.

Donk (in Bull. Bot. Gard. Buitenzorg ser. iii, 17, 1941, pp. 159-160; ibid. ser. iii,

18, 1949, pp. 88-89 and 159) proposed the conservation of Aleurodiscus Rabenh. ex

J. Schroet. against Cyphella Fr. The proposal was supported by Rogers (in Farlowia

3, 1949, p. 433) and was adopted by the Stockholm Congress. Thus the use in this

paper of the name Cyphella is nomenclaturally improper, and an explanation is required.

The series of studies of which this paper forms a part, are intended to check the

validity of the recorded species of South African Thelephoraceae, and to indicate

synonyms and incorrect specific determinations where possible. Such work is essential

before the many species that await collection may be identified and described. With

few exceptions the type of revisionary work making a fundamental contribution to

the broader aspects of classification and nomenclature can only be expected from

the older countries with their better herbarium and library facilities. The dilemma

in which the writer is now placed is whether to withhold publication of the facts

uncovered about the South African species of Cyphella, or whether to act against the

decision of Congress by retaining this invalid name in publication. It appears that

as a temporary measure, until someone more favourably situated has monographed

the genus and placed the species in generic segregates likely to have some stability,

the interests of progress in mycology are better served by adopting the latter course.

Cyphella alboviolascens (Alb. & Schw. ex Fr.) Karsten in Finska Vet.-Soc. Bidrag

Natur och Folk 37 (1882) 133; Saccardo Syll. Fung. 6 (1888) 669; van der Byl in

Ann. Univ. Stellenbosch 7 (1929) 12; Doidge loc. cit. p. 483.

Peziza alboviolascens Alb. & Schw. ex Fr., Syst. Myc. 2 (1823) 96; Alb. & Schw.,

Consp. Fung. (1805) 322, PI. 8. f. 4.

Cyphella curreyi Berk. & Br. (!) in Ann. Mag. Nat. Hist, iii, 7 (1861) 379; Kalchbrenner

in Grevillea 10 (1882) 104.

Fig. 7.

Fructifications sessile, lacking a subiculum, globose, later cupulate, 1-3 mm.

diam., white, covered with abundant long, villose or strigose hairs arising from the

base and sides of the cupule. Margin involute, closing the fructification when dry.

Hymenium smooth, pale coloured, very light tawny (in European specimens also pale

violet, grey-lilac, glaucous or greenish).

Basidia: clavate, 40-60 x 8-10 /x, occasionally with a basal clamp.

Spores: broad ovate to pip-shaped, often with a prominent apiculus, inequilateral,

11-2-14-4 x 9-9-6-(ll -6) /x.

Hyphae: more or less horizontally arranged at the base of the cupule, hyaline, smooth,

with occasional clamps, 1 • 5-3 /x diam., compact and entangled beneath the hymenium.

Surface hairs: up to 200 x 5-7 /x, very thick-walled, with a narrow lumen, rounded

at the apex, with finely roughened walls becoming smooth towards the base, and often

encrusted with mineral granules.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der Byl

nr. 2503, E. D. Loseby, on Vitis, Stellenbosch.

472

This collection corresponds very well with European material of Cyphella

alboviolascens. Another specimen filed as this species in the van der Byl Herbarium

No. 2524, proved on examination to be Cytidia flocculenta. The material on which

Kalchbrenner based his record of Cyphella curreyi has not been traced; C. curreyi

B. & Br. (!) is a synonym of Cyphella alboviolascens.

Cyphella applanata sp. nov.

Fig. 12.

Fructifications 0-5-2 mm. diam., sessile, resupinate with upturned margin forming

a flat, saucer-like body of membranous texture. Margin lobed, white,- villose, the

hairs covered with a white mineral incrustation. Hymenium smooth, dirty white to

lightly greyed fawn.

Basidia; clavate, with 4 sterigmata, 17-25 >< 4 -5-5 -5^.

Spores: hyaline, smooth, cylindrical to cyhndric-curved with a depressed side, often

cohering in groups of four, 3 -2-4 -8 x (6 - 4)— 8—1 1 -2 p.

Hyphae: very fine, hyaline, with occasional clamps.

Hairs: thinwalled, pale yellowish, without damps, heavily encrusted with acicular

crystals, 2-4 p wide without the crystal layer, forming a thick white, villose covering

to the upturned margin.

Specimens examined: Type, 33350, E. M. Laughton, Eshowe, Zululand, Nov. 1941.

In its flat, discoid fructification with upturned, lobate margin, this species is not

unlike C. bloxami and C. disciformis as described by Pilat (in Ann. Mycol. 22, 1924,

212) but differs in the shape and size of its spores and in having incrusted hairs.

Cyphella applanata sp. nov.

Carposomata disciforma, applanata, 0-5-2 mm. diam., tenuissime membranaceo;

margo lobata, extus pilis densis, simplicibus, 2-4 p crassis, crystallis densissime

incrustatis. Hymenium leve, sordide albidum vel griseo-fulvum. Basidia clavata,

hyalina, 17-25 x 4- 5-5 -5 p, sterigmatibus 4. Sporae hyalinae, leves, cylindraceae

vel saepe subcurvulae et uno latere compressae. Hyphae hyalinae, tenuissimae, saepe

nodosae. Pili tenue tunicati, haud nodosi, 2-4 p crassi, crystallis granulosissimis

dense villosi, substraminei. Typus, 33350, leg. E. M. Laughton, Eshowe.

Cyphella cheesmani Massee apud Cheesman in Journ. Linn. Soc. Bot. 38 (1909) 411;

Saccardo Syll. Fung. 21 (1912) 422; Doidge, loc. cit. p. 483.

The writer has not traced any specimen of this species, but as it appears well

characterised by its spores it is accepted and keyed out with the other species of

Cyphella. A translation of Saccardo’s description follows.

Membranous, sessile, cupular then flattened, with villose exterior, white, 1 mm.

diam.; hymenium lightly plicate; basidia clavate with four sterigmata: spores

ovoid, finely asperulate, hyaline, 12-15 x 7-8 p.

Cyphella curreyi Berk & Br. (!) in Ann. Mag. Nat. Hist, iii, 7 (1861) 379; Kalchbrenner

in Grevillea 10 (1882) 104.

Recorded by Kalchbrenner for an unnumbered collection of MacOwan from

Somerset East. This collection has not been traced. C. curreyi is a synonym of Cyphella

alboviolascens (Alb. & Schw. ex Fr.) Karst.

Cyphella farinacea Kalchbr. & Cooke (!) in Grevillea 9 (1880) 18; Kalchbrenner

in Grevillea 10 (1882) 104; Saccardo Syll. Fung. 6 (1888) 671; Doidge loc. cit.

p. 483.

Fig. 8.

473

Cupules gregarious, often touching one another, very shortly stalked or sessile,

0-3-1 -5 mm. diam. when dry, discoid, becoming flattened, with inrolled margin when

dry. Exterior covered with greyish-white meal composed not of hyphae but of mineral

matter. Hymenium olivaceous or usually cinereous to blackish. General appearance

is greyish with a dark interior.

Basidia: clavate, with 4 sterigmata, 6-5 x 19-20 n, with basal clamps.

Spores: hyaline to faint yellow, smooth or appearing very minutely punctate, ovate

or almost globose, 4 -8-6 -4 n diam.

Hyphae: hyaline, septate, branched, with frequent large clamps which are often looped,

subgelatinous, much entangled, 1 • 5-1 • 8 ^ diam.

Specimens examined: Type, MacOwan (1221) in Herb. Kew and in Pretoria sub

No. 20807.

Cyphella friesii Quelet, Champ, du Jura et des Vosges iii (1875) 15: Kalchbrenner

in Grev. 10 (1882) 104.

This species was recorded by Kalchbrenner for an unnumbered collection of

MacOwan from Somerset East. Doidge (loc. cit. p. 483) takes up this record under

Cyphella friesii Crouan (described by Crouan in Florule du Finistere, 1867, 62).

The author is indebted to Dr. R. W. G. Dennis for drawing attention to yet another

Cyphella friesii, described by Weinmann (in Hymeno-et GasteroMycetes hucusque

in imperio Rossico observatos, 1836, 523) and antedating both Crouan’s and Quelet’s

species, neither of which thus have any nomenclatorial standing. The different spelling

of the epithet “ Frisei ” in Weinmann’s description is clearly a misprint since it appears

in Weinmann’s own index as “ Friesii Weinmann ” and is obviously proposed in honour

of Fries.

The material of Kalchbrenner’s record has been examined (see Fig. 20), but is in

too poor a condition to describe under a new name or to identify with a species already

named. A brief description of this material follows: —

Cupules membranous, shortly stipitate or cuneate tapering to a sessile, reduced

base; about 3-3-5 X 1-5-2 mm. in size; light dirty yellowish-brown, with a surface

composed of cohering and flattened matted fibrils. Hymenium light yellowish-brown.

Margin indistinctly incised-lobate. No basidia were seen. Abundant spores were

present but it is doubtful whether they belong; they were globose, lightly coloured,

minutely echinate, 3-5-4 ^ diam. The context hyphae were indistinct, but hyaline

and about 1-5 y. diam. The surface hyphae were hyaline, thickwalled with a narrow

lumen, branched, 4-6-5 p diam.

Cyphella fulvodisca Cooke & Massee (!) in Grev. 18 (1890) 50; Scott Elliot in Hedwigia

29 (1890) 66, Tab. 1; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 11;

Saccardo Syll. Fung. 9 (1891) 247; Doidge loc. cit. p. 483.

Fig. 11.

— Cyphella variolosa Kalchbr. (!).

Cyphella pelargonii Kalchbr. (!) apud von Thuemen in Flora 68 (1875) 378; Saccardo

Syll. Fung. 6 (1888) 677; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 11;

Doidge loc. cit. p. 483.

Fig. 13.

Cupules closely gregarious in groups of 6-20, somewhat coherent with one another,

rather horny when dry, light beige colour but appearing whitish owing to the presence

of white pruinose hairs on the outside, about 1 mm. long and 0-5-0 -8 mm. wide, more

or less cylindrical with a tapered base. Hymenium yellow-brown.

Basidia: clavate, 25-35 x 6-8 /i.

474

Spores: light yellow-brown, smooth, with a small apiculus, elliptical to broad ovate,

5 • 5-6 • 8 X 8-3-10 p.

Hyphae: colourless, thickwalled, with only a narrow lumen staining, somewhat

gelatinised, with clamp connections, 3-4-8-(6-4) p wide.

Surface hairs: yellow-brown with a lighter, rounded apex, appearing white macro -

scopically owing to the presence of mineral incrustation, (25)-40-60 x (3-5)-4-6 p,

occasionally with a basal clamp connection.

Specimens examined: Type, MacOwan, Cap. b. sp.. Herb. Kalchbr., in Herb. Kew.

See also the discussion under Cyphella variolosa.

Cyphella punctiformis (Fr.) Karst., Fung. Fenn. Exsicc. N. 714; Saccardo Syll. Fung.

6 (1888) 678; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 12; Doidge loc.

cit. p. 483.

Peziza punctiformis Fr., Syst. Myc. 2 (1823) 105.

The above-cited records of this species for South Africa would appear to rest on

the MacOwan collection from Somerset East (see next entry) which Kalchbrenner &

Cooke determined as C. punctiformis var. strigosa, except for Doidge’s citation of

No. 28673 ( Rump 311). This latter collection was wrongly determined, being in fact

a very minute agaric, with most distinct lamellae and cheilo- and pleurocystidia. On

present knowledge this species is not considered to be represented in South Africa.

Cyphella punctiformis (Fr.) Karst, var. strigosa Kalchbr. & Cooke in Grevillea 9 (1880)

18; Kalchbrenner in Grevillea 10 (1882) 104.

Peziza punctiformis Fr. pr. parte, fide Kalchbrenner loc. cit.

The habitat on rotten leaves is suggestive of Cyphella punctiformis, but apparently

no specimen is in existence. The “ description ” given by Kalchbrenner & Cooke

consists of three words, “ Pilis elongatis, granulatis ”, which hardly may be taken

to validate a new variety, especially as it applies equally well to the main species. While

not accepting the species or variety on present knowledge as being represented in South

Africa, I have thought it worth while to give a description of C. punctiformis, translated

from Bourdot & Galzin (Hym. de Fr., 1928, p. 160): —

Sessile, very small, 0-15-0-5 mm., thin, cupuliform with an inrolled margin and

globose in dry times, pure white, villose or bristling; hymenium smooth, white.

Hyphae generally indistinct, 1^4 p diam.; hairs of the cuple 40-100 x 3-4 p,

thickwalled, pointed, rarely septate, asperulate in the upper parts; basidia

15-26 X 5-8 p, with 2-4 sterigmata up to 2-3 p long. Spores oblong or amygdaliform,

rather variable, 5-10 x3-4 p.

Cyphella tabacina Cooke & Phil. (!) apud Cooke in Grev. 10 (1882) 123; Saccardo

Syll. Fung. 6 (1888) 672; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 12;

Doidge loc. cit. p. 484.

Fig. 10.

= Cyphella variolosa Kalchbr. (!). See next entry.

Cyphella variolosa Kalchbr. (!) in Grevillea 10 (1882) 104; Saccardo Syll. Fung. 6

(1888) 671; Doidge loc. cit. p. 484.

Cyphella tabacina Cooke & Phil. (!) apud Cooke in Grev. 10 (1882) 123; Saccardo

Syll. Fung. 6 (1888) 472; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 12;

Doidge loc. cit. p. 484.

475

Cyphella fulvodisca Cooke & Massee (!) in Grev. 18 (1890) 50; Scott Elliot in Hedwigia

29 (1890) 66, Tab. 1 ; van der Byl in Ann. Univ. Stellenbosch 7 (1929) 1 1 ; Saccardo

Syll. Fung. 9 (1891) 247; Doidge loc. cit. p. 483.

Fig. 9.

Fructifications cupular, urceolate to hemispherical, subsessile or shortly stipitate,

up to 2 mm. diam. when moist, 0-3-0 -7 mm. diam. when dry; the margin entire, rolled

inwards irregularly when dry. Cupules gregarious, seated upon a red-brown subiculum

which may be thick, scanty or absent, exterior white or greyish, mealy, composed of

mineral matter and very short hairs. Hymenium drying tawny to olivaceous.

Basidia: cylindric-clavate, with 4 sterigmata, 4-2-8 x 23-40 p, with basal clamps

Spores: abundant, coloured, pale yellow-brown, broad elliptic, ovate or broad fusoid,

with one to several guttules and a minute apiculus, 4 -8-6 -4 x 7 -3-9 -6 /a.

Hyphae: hyaline, 1 -6-2-4 p diam., septate, with clamps, branched, thinwalled.

Hairs of cupule: pale yellow-brown, thin- to thickwalled, the wall sometimes minutely

roughened, more or less cylindrical to fusoid, sometimes with a basal clamp,

3 • 5— 5— (7) x 35-58 p.

Subiculum: absent, scanty or well developed, reddy-brown to chocolate coloured,

shortly velutinate, composed of branched, interwoven, repent hyphae, pale coloured

to yellow-brown, thickwalled with occasional clamps, 2 -4-6 -4 p wide, terminating

in more or less erect hairlike hyphae 3-5-6 -4 /a wide and 64-120 p long.

Specimens examined: As C. variolosa : Type, MacOwan (1381) ex Herb. Kalchbr.

in Herb. Kew and sub No. 20939 in Pretoria; 40924, Talbot, on Populus deltoides,

Piet Retief, Tvl. ; As C. tabacina: Type, 11146, J. M. Wood (524) and in Herb. Kew.;

As C. fulvodisca : Type, Scott Elliot, Fort Dauphin, Madagascar, in Herb. Kew;

Farquharson (53), S. Nigeria, in Herb. Kew; Dummer (4158) and (1157), Uganda,

in Herb. Kew; 35563, Rump (753), Newsel Beach, Natal; 12017, Dummer (1157),

Uganda.

A close comparison of the types of C. variolosa, C. tabacina and C. fulvodisca

leads me to believe that they are all the same species. The absence of a subiculum

in the type of C. variolosa is not regarded as an important distinction; for in

C. tabacina, the species in which the subiculum is best developed, there are parts almost

devoid of it, while in C. fulvodisca the subiculum is thin and scanty and sometimes

hardly distinguishable. The size of the cupules in all three species is variable, as also

are their stipes. C. fulvodisca tends to be sessile, but cupules showing a short stipe

are present. In C. tabacina the stipes are short while in C. variolosa they are

comparatively long. The only specimens of C. tabacina and C. variolosa are their

types. The author believes that they are merely developmental stages in a species

which is most widely represented by the stage which received the name C. fulvodisca.

In microscopic characters all three species are a close match, as is shown in Table 1 .

Table 1.

476

In microscopic characters C. pelargonii and C. variolosa are very close, and it is

possible that they are synonymous, in which case the former name would have priority.

With only one specimen for examination it is difficult to assess how far the greater

cupular size of C. pelargonii may be reconciled with the known variation in this feature

in C. variolosa. I have differentiated the two here on account of the greater size of the

cupules in C. pelargonii and because its hyphae are thickwalled and commonly wider

than those of C. variolosa.

Cyphella Species Insufficiently Known.

(1) The species recorded by Kalchbrenner as Cyphella friesii Quel. See p. 473,

and Fig. 20.

(2) The species represented by the type of Femsjonia natalensis Cooke (in Grevillea

10, 1882, 123). This is cited by Saccardo Syll. Fung. 6 (1888) 780, and by Doidge

loc. cit. p. 477. See Fig. 19.

The type. No. 11047 (J. M. Wood, 476) has been examined and is certainly not

one of the Dacryomycetaceae. It seems referable to Cyphella, but owing to lack of

spores (also mentioned by Cooke) and difficulty in interpreting some other structures

it is not considered sufficiently well characterised to be included in the known species

of Cyphella.

The cupules are yellowish drying light orange-yellow, 0-3-0 -6 mm. diam. when

dry (0-5-1 mm. fide Cooke), somewhat fleshy-gelatinous in texture, sessile, discrete

but occurring on a ffoccose to pruinose greyish-white subiculum. The cupules are

wide open, their exterior being glabrous or covered with a fine scanty meal. In the

hymenium are clavate organs 25-35 X 6 -4-8 -5 p of the appearance of basidia, also

clavate organs 13-18 x 3 -5-5 -5 p again like basidia. No sterigmata or spores were

seen. Also in the hymenium were ovate vesicular bodies, 10-15 x 15—23 p, whose

nature and origin could not be made out. The hyphae were hyaline, subgelatinous,

with clamps, thinwalled, 1 • 8-3 p wide. The subiculum was composed of scanty hyaline

hyphae, thinwalled, 2 -8-3 -2 p diam., surrounded by heavy deposits of minerals.

CYTIDIA Quelet.

Cytidia cornea Lloyd (!) in Lloyd Myc. Notes 5 (1917) 656, figs. 935-937; Doidge

loc. cit. p. 484.

The author has examined the type of this species (Duthie 154, in Lloyd Mycological

Collections No. 34063) and concludes with Martin (in Lloydia 5, 1942, 161) and

W. B. Cooke (in Mycologia 43, 1951, 208) that it must be referred to Cytidia habgallae

(Berk. & Br.) Martin.

Cytidia flocculenta (Fr.) Hohnel & Litsch. in K. Akad. Wiss. Wien Sitzungsb. 116

(1907) 758; Burt in Ann. Mo. Bot. Gard. 11 (1924) 9, PI. 1, f. 7; W. B. Cooke

in Mycologia 43 (1951) 204; Doidge loc. cit. p. 484.

Thelephora flocculenta Fries, Elenchus Fung. 1 (1828) 184.

Corticium flocculentum Fries, Epicrisis (1838) 559, Hym. Eur. (1874) 647.

Cyphella amp/a Lev. in Ann. Sci. Nat. ser. iii, 9 (1848) 126.

Further synonyms are given by W. B. Cooke, loc. cit.

Fig. 16.

Fructifications discoid to shallow cupular, membranous-gelatinous in texture,

up to 1-5 cm. diam., sessile, attached more or less centrally; surface greyish-white,

hairy, with a ffoccose tomentum composed of hyaline hyphae united into fascicles,

without much mineral matter. Hymenium smooth, but usually with a few prominent

veins, coloured fawn to reddish brown, darkening on drying to a dull reddy-black.

Thickness in section excluding tomentum about 700 p.

Basidia: 30— (40) x 4-5 p, cylindrical to clavate.

Spores: hyaline, smooth, cylindrical to allantoid, 1-5-3 x 7-10 p.

477

Hyphae: gelatinised, thickwalled, usually only the lumen showing clearly, with

numerous clamp connections, much branched and anastomosed, 1-3 p diam.

Tissue differentiation: a dark compact zone subtends the surface hairs.

Hairs: thickwalled, unbranched, without septa or clamps, hyaline, sometimes a faint

straw colour, 4-5 p diam.

Specimens examined: Universiteit van Stellenbosch, Herbarium P. A. van der Byl

nr. 2524 (as Cyphella alboviolascens) ; 39103, van Hoepen, on Populus ; 28804,

Doidge & Bottomley, on Populus ; 27569, Louwrens; 32505, Jacot-Guillarmod; 24852,

on Quercus', 23719, Reinecke; 40474, Talbot.

Cytidia habgallae (Berk. & Br.) Martin in Lloydia 5 ( 1942) 161, figs. 4-12; W. B. Cooke

in Mycologia 43 (1951) 208; Doidge loc. cit. p. 484.

Corticium habgallae Berk. & Br. in Journ. Linn. Soc. Bot. 14 (1873) 72.

Cytidia cornea Lloyd (!) in Lloyd Myc. Notes 4 (1917) 656, figs. 935-937.

Aleurodiscus corneas (Lloyd) Lloyd (!) in Lloyd Myc. Notes 6 (1920) 930, f. 1688.

Aleurodiscus capensis Lloyd (!) in Lloyd Myc. Notes 6 (1920) 930, f. 1687, ibid. 6 (1921)

1088; Stevenson & Cash in Bull. Lloyd Library No. 35 (1936) 43.

Gloeosoma capensis Lloyd (!) (as McGinty) nomen provisiorum in Lloyd Myc. No tes

6 (1921) 1088; Stevenson & Cash in Bull. Lloyd Library No. 35 (1936) 42.

Aleurocystus capensis Lloyd (!) (as McGinty) ex Stevenson & Cash in Bull. Lloyd

Library No. 35 (1936) 42; Lloyd (as McGinty) in Lloyd Myc. Notes 6 (1921)

1088, nomen provisiorum.

An extended synonymy is given by Martin (loc. cit.) and by W. B. Cooke (loc. cit.).

Fig. 17.

Fructifications sessile, attached by the central part to the substratum, discoid,

irregularly circular with lobed, slightly upturned margin, single or confluent, coloured

whitish to pale creamy or with a pinkish-buff tint when old (may become darker still,

fide Martin). Texture fleshy-gelatinous when moist, drying ceraceous-horny. Diameter

when dry 2-6 mm. Thickness in section up to 1000 p.

Basidia: large, clavate, 10-17 x 70-95 p (immature) with 2-4 sterigmata up to 10 p

long and 4 p wide at the base.

Spores: at first hyaline, thinwalled, smooth, with a prominent apiculus, ovate or

broad elliptic, 15-19 x 13-14 p (up to 18-25 x 10-8-18 p, fide W. B. Cooke); finally

the walls become thickened to as much as 2-5 p thick and are penetrated by small

peglike projections of the lumen in two (rarely more) places, one opposite the apiculus

and one at the opposite pole of the spore; the thickened spores are commonly found

embedded in the tissues of the fructification.

Paraphyses: accompanying the basidia, hyaline simple paraphysoids or dendrophyses

with short branches, thinwalled, not encrusted, 1 • 5-3 p wide and 40-60 p long.

Cystidia: embedded or projecting above the hymenium, usually fusoid, with thin to

thick walls covered with detersile pale yellowish-brown mineral granules, (8)— 1 6— 25 x

60-100 p.

Hyphae: smooth, hyaline, thinwalled, gelatinised, with clamp connections, 1-2 p

wide, arranged more densely towards the outer parts of the fructification.

Specimens examined: A. V. Duthie 154 (Lloyd Mycol. Collections No. 34063), type

of Cytidia cornea', P. A. van der Byl 833 (Lloyd Myc. Coll. No. 34029), type of

Aleurodiscus capensis', 23490, F. Eyles (675), Mazoe, S.R.

478

The cystidia and the presence of peculiarly thickened basidiospores characterise

this species. According to Martin (l.c.) the imperfect state of this fungus is represented

by Matula poroniaeformis (B. & Br.) Massee and M. rompelii (Rick) Lloyd.

Cytidia simulans Lloyd (!) in Lloyd Myc. Notes 6 (1920) 991, f. 1772; Doidge loc. cit.

p. 484; W. B. Cooke in Mycologia 43 (1951) 207.

Fig. 18.

Fructifications discoid, centrally attached, sessile, 1-1-5 cm. diam., with a fleshy-

gelatinous texture; hymenium smooth but bearing small wrinkles, pale brown;

surface white, bearing a few hyaline hairs encrusted heavily with mineral granules and

thus appearing pruinose.

Basidia: cylindric-clavate, 20-35 x 3-4 -5 /x, with (fide W. B. Cooke) 2 or 4 sterigmata.

Spores: of two kinds (1) abundant, small, hyaline, smooth, cylindrical or allantoid,

2- 2 • 5-(3 • 2) X 6 • 4—9 • 6— ( 1 1 -2) /x, borne on the basidia with four sterigmata, (2) scanty,

larger, hyaline, smooth, ovate to elliptic-depressed, 4 -8-6 -4 x 9-6-12-8 /x, borne on

basidia with two sterigmata.

Hyphae: hyaline, more or less thickwalled with only the lumen showing clearly,

gelatinised, with numerous clamp connections, much contorted and sometimes

undulating, 1-5-3 /x diam., branched and anastomosed.

Specimens examined: Type, A. V. Duthie 243, in Lloyd Myc. Coll. No. 33589.

The larger spores are not easy to find in the type, and only, material of this species,

while the smaller spores are much like those of Cytidia flocculenta. The latter may

be distinguished by its long floccose hairs on the surface, where C. simulans has only

a few short hairs and derives its white, pruinose appearance from the mineral granules

encrusting its surface.

DENDROTHELE Hohnel & Litsch.

Dendrothele duthieae sp. nov.

Fig. 22.

Fructifications resupinate with a narrowly upturned margin, not firmly adnate,

discoid, up to 5 mm. diam., or elongated-confluent, bearing on the hymenium white

papillae (hyphal pegs) visible to the naked eye, emerging from a whitish to pale tan

coloured hymenium; texture pellicular, the hymenium being membranous and the

basal tissues very soft and byssoid.

Basidia : long, clavate, with a wavy outline, often with a basal clamp, 50-1 10 X 10-14 /x,

with four sterigmata.

Spores: hyaline, smooth, pip-shaped or ovate-elliptical to long-elliptical with the

attached end narrowed to a prominent apiculus and commonly curved, not amyloid,

abundant, 8-9 - 6-( 11-2) X (1 5)-l 7-19-(22) /lx.

Hyphal pegs: abundant, visible to the naked eye, white, superficial, i.e. easily detached

and leaving a crater when removed, composed of dendrophytic hyphae with short,

irregular branches, heavily encrusted with minerals, often showing a clamp near the

base of the main axis, 1 -5-3-5 y. diam., with firm walls. Hyphal pegs sterile.

Paraphyses: in the hymenium are dendrophyses like those of the hyphal pegs, and

aborted basidia, some with a tendency to fork and others with vague submoniliform

constrictions.

Hyphae: basal hyphae loosely, horizontally interwoven, byssoid, hyaline, septate,

branches, with numerous clamps and with thin, firm walls and a wide lumen, about

3- 2 /x wide. The hymenium and subhymenial layers contain much mineral matter.

Specimens examined: Type, No. 31393, A. V. Duthie (187), without locality but

probably Knysna area.

479

The spores of this species have a distinctive apiculus which remains uncoloured

by phloxine stain and is easily picked out. It is possible that hyphal pegs are not always

produced in this species, but on the basis of the present specimen it must be disposed

in Dendrothele rather than Aleurodiscus.

Dendrothele duthieae sp. nov.

Fructificationes resupinati, discoidales, usque ad 5 mm. diam., elongato-confluentes.

Margines minute revolutae. Hymenium albicans vel dilute alutaceum, minute

albopapillosum. Textura pellicularis. Basidia clavata, 50-110 x 10-14 p. Sporae

hyalinae, laeves, amygdaliformes vel ovato-ellipticae, haud amyloidae, (15)— 17— 19— (22) x

8-9-6-(ll-2) p. Papillae steriles, oculo nudo conspicuae, superficiales, ex hyphis

dendrophyticis et corporibus mineralibus compositae. Paraphysae (1) dendrophyticae

(2) basidia abortiva, nonnula submoniliforme constricta. Hyphae basales laxe

intertextae, hyalinae, septatae, nodoso-septatae, ramosae, tenue tunicatae. Typus:

No. 31393, leg. A. V. Duthie (187).

GLOEOSOMA Bresadola.

Gloeosoma capensis Lloyd (as McGinty) (!) in Lloyd Myc. Notes 6 (1921) 1088, nomen

provisiorum. Further referred to in: Stevenson & Cash in Bull. Lloyd Library

No. 25 (1936) 42; Martin in Lloydia 5 (1942) 162.

At the same time suggesting a new genus, Aleurocystus, to accommodate

Aleurodiscus capensis, Lloyd (loc. cit.) proposed that the latter might also be transferred

to the genus Gloeosoma. The arguments given under Aleurocystus, apply here as well,

and one concludes that Gloeosoma capensis is also a nomen provisiorum and was not

validly published. See under Cytidia habgallae (B. & Br.) Martin.

POROTHELEUM (Fr. ex Fr.) Fr.

Porotheleum incanum (Kalchbr.) Saccardo, Syll. Fung. 6 (1888) 423.

Stigmatolemma incanum Kalchbr. in Grevillea 10 (1882) 104; Doidge loc. cit. p. 544

(as Stigmatolemina incanum).

Fig. 21.

Fructifications composed of sessile cup-shaped fruit-bodies very closely aggregated

and confluent, at first closed, later opening and showing the hymenium lining the

interior of each cupule. The cupules are all seated on a white subiculum which projects

as a white, very narrow, byssoid margin beyond the aggregate fructification. Walls

of cupules are grey, i.e. light cinereous in colour giving the whole a light grey appearance,

and are shortly byssoid on the exterior. Hymenium very dark, almost black when dry.

Occupying areas several cm. long and 1-2 cm. wide on bark and wood. Diameter

of the cupules ± 325 p.

Basidia : clavate, 1 6 • 6-25 -6x7- 2-9 • 6 p, often with a large apical guttule. Sterigmata

2-4, conical, stout, 1-6 p long. Basidia are formed in clusters and show occasional

basal clamps.

Spores: hyaline, smooth, cylindrical with one side depressed, or occasionally seen

as broad elliptical, 7-2-8 x 3 -2-4 -4 p, abundant.

Hyphae: of the cupules and the subiculum are hyaline, very narrow, up to 1-6 p

diam., branched, with clamp connections, smooth in the cupules but often coated with

very fine mineral matter in the subiculum.

Specimens examined: No. 31058 ( b ), A. M. Bottomley, dead wood, Knysna, 1939.

480

Tn the original description of Stigmatolemma incanum the spores are said to be

15 p long. The author had been unable to trace authentic material of this species

and was about to propose a new species for No. 31058 ( b ) based on the difference in

spore size, when Dr. Wm. Bridge Cooke kindly informed him that there was apparently

type material of S. incanum, ex Herb. Karsten, preserved in the Lloyd collections sub

No. 17601 in the U.S. National Fungus Collections. Dr. Cooke states that this specimen,

collected by MacOwan at Somerset East, has spores measuring 7x4 p. It therefore

corresponds very well with Miss Bottomley’s new collection.

In using the generic name Porotheleum, I have not overlooked Donk’s argument

(in Reinwardtia 1, 1951, 217-218) that Porotheleum Fr. (later spelled Porothelium by

Fries himself) is a later homonym of Porothelium Eschweiler, and so should be replaced

by Stromatoscypha Donk (1951, loc. cit.). An alternative interpretation of the facts

and Code of Nomenclature has been suggested to me by Dr. Cooke in a private

communication. I feel that a full statement must await Dr. Cooke’s own publication

on this group and that meanwhile it would be inadvisable to make a new combination

of P. incanum under Stromatoscypha.

SOLENIA Pers. ex Fr.

Solenia Candida Persoon in Roemer Neues Mag. Bot. 1 (1794) 116, Syn. Fung. (1801)

676, Myc. Eur. 1 (1822) 334; Fries, Syst. Myc. 2 (1823) 200; Saccardo Syll.

Fung. 6 (1888) 424; Doidge loc. cit. p. 487.

This record rests on the specimen cited by Doidge, 28297, W. G. Rump (223),

in the National Herbarium, which is the type specimen of Solenia natalensis W. B. Cooke

sp. nov.

Solenia minima Cooke & Phil. (!) apud Cooke in Grevillea 10 (1882) 123; Saccardo

Syll. Fung. 9 (1891) 207; Doidge loc. cit. p. 487.

Fig. 15.

Cupules delicate, membranous, sessile, whitish to pale lemon yellow when dry,

short-cylindrical to barrel-shaped, discrete but gregarious, small, 250 p high, 130 p

diam.

Basidia: clavate, 16-19 X 4-8-6^.

Spores: hyaline, smooth, subglobose to broad ovate, 2-5-3 x 3-4 p.

Hyphae: hyaline, rather indistinct, about 1 -6 p diam., no clamps seen.

Specimens examined: Type, Wood (482), on Strelitzia, Inanda, under Nos. 10706

and 11145 in National Herb., Pretoria, and specimen in Kew.

The species is distinguished by its very small, subcylindrical, sessile cupules, which

are yellowish when dry, and lack surface hairs.

Solenia natalensis W. B. Cooke sp. nov.

Receptacles cylindric, 0-5 mm. tall at margins of colony, separate, in centre of

fruiting area up to 1-0-1 -5 mm. tall, sometimes up to 2 mm., becoming densely

gregarious, apparently adnate in sheaths; subiculum evanescent or absent, not evident;

surface hairs when present strongly appressed, up to 3 p in diameter, with thick walls,

the outer half in the shape of a medium to tight spiral, mostly toward the upper part

of the receptacle; toward the lower part scattered, hyaline, thick-walled, branched

hyphae occur; context hyphae 2 -0-2 -5 p, up to 3 • 5 p in diameter, simple septate

and with clamp connections; hymenium formed of a tight palisade of basidia; basidia

with basal clamps, mostly 12-5 x 5-5 p, 4-sterigmate; spores hyaline, ovate, apiculate,

smooth, somewhat flattened on one side, 3-4 x 2 * 5—3 - 5 p.

Hab.: on indigenous wood.

481

Type: Town Bush, Pietermaritzburg, Natal, Union of South Africa, Coll. W. G. Rump

(223), Oct. 1934 (Pretoria 28297).

This specimen was determined originally as Solenia Candida. It can be distinguished

from that species by its smaller, ovate spores and its surface hairs which are spiral

in the outer half rather than dichophysoid. Its marginal cupules resemble superficially

those of S. Candida while those in the centre of a colony appear like those of the variety

polyporoidea. On examination the adjacent cupules are not grown together, which

would be the case if the cupules were truly adnate, and are not damaged when separated

in mounting.

Solenia natalensis W. B. Cooke sp. nov.

Receptacula cylindracea, 0-5-2 mm. alta, separata deinde dense gregaria, adnata.

Subiculum fugiens vel absens, non manifestum. Pili superficiei plerumque versu,

partem superam receptaculi, forte appressi, usque ad 3 p diam., crasse tunicatis

dimidiis extremis in helicem glomeratis; hyphae hyalinae, crasse tunicatae, ramosae

plerumque ex parte inferiore receptaculi orientes. Hyphae contextae 2-0-2- 5 p, usque

ad 3-5 ju. diam., vel septatae vel nodoso-septatae. Basidia 12-5 x 5-5 p, infra nodulis,

sterigmatibus 4. Sporae hyalinae, ovatae, apiculatae, leves, uno latere compressae,

3-4 x 2 -5-3 -5 p. Typus No. 28297, leg. W. G. Rump (223).

Solenia rhoina W. B. Cooke sp. nov.

Receptacles densely crowded in definite areas, with a thin layer of arachnoid

subicular hyphae on substratum where cups do not occur; receptacles up to 0-5 mm.

in diameter, sessile, broad cupulate, cream-colour when dry; margin entire, silky,

appearing like a minute Stereum; context very thin, the several layers of hyphae giving

rise on the one hand to surface hairs, on the other to basidia; surface hairs hyaline,

solid, parallel, appressed, progressively shorter to the margin, less appressed near

the base, 25-75 x 2 -5-4-0 p; hymenium of tightly packed basidia each with a basal

clamp, 4-sterigmate, 14-5-18 x 5-7 p; spores smooth, hyaline, ovate, apiculate,

flattened on one side, 6-7-5 x 4 -5-5 -5 p.

Hab.: on Rhus lancea (Karreeboom).

Type: Hennops River, Pretoria District, Union of South Africa, collected by

P. H. B. Talbot, Aug. 15, 1948. (Pretoria 39049).

This species is distinct from other species in the genus. The subiculum is Solenia-Uke

rather than Porotheleum- like, but the colonies are as discrete as in some species in

that genus.

Solenia rhoina W. B. Cooke sp. nov.

Receptacula usque ad 0-5 mm. diam., in locis definitis dense conferta, sessilia,

cupulata, cremicolora. Subiculum tenue, arachnoideum. Margo integer, bombycinus.

Contextus tenuissimus. Pili superficiei hyalini, solidi, appressi, paralleli, versus

marginem breviores, 25-75 x 2 -5-4-0 p. Basidia infra nodulis, sterigmatibus 4,

14-5-18 x 5-7 p. Sporae hyalinae, leves, ovatae, apiculatae, uno latere compressae,

6-7-5 x 4 -5-5 -5 p. Typus No. 39049, in ramo arboris quae Rhus lancea appellatur,

leg. P. H. B. Talbot.

STIGMATOLEMMA Kalchbrenner.

Stigmatolemma incanum Kalchbr. in Grevillea 10 (1882) 104; Doidge loc. cit. p. 544

(as Stigmatolemma incanum).

— Porotheleum incanum (Kalchbr.) Sacc.

482

Keys to Accepted Species.

Aleurodiscus.

1. Both yellowish gloeocystidia and encrusted dendrophyses present together. Amyloid granules

encrusting the dendrophyses. Spores broad elliptic to ovate, 13-5-15 x 7-5-8 y.

A. botryosus.

Either gloeocystidia or dendrophyses present, or both absent. Spores otherwise. 2.

2. Spores semilunate, biapiculate, smooth or minutely echinate, about 14 x 23 y. Acanthophyses

and simple unbranched paraphysoids present. A. mirabilis.

Spores otherwise. 3.

3. Fructifications whitish or light-coloured but not rosy, purplish, violet, lilac or brownish to

clay-coloured. 4.

Fructifications rosy, purplish violet, lilac, brownish to clay-coloured. 6. ,

4. Spores ovate-subglobose, 1 8-22 x 12-5-16 y. Simple cylindrical paraphysoids and gloeocystidioid

submoniliform paraphysoids present. A. disciformis.

Spores not more than 8 y broad. Encrusted dendrophyses and bulbous pseudophyses with a

short to long apical outgrowth present. 5.

5. (Spores ovoid, 5-6-6-4 x 8-12 y, ratio length/width less than 2-0. A. acerinus.)

Spores broad elliptic to ovate, 4-8 x 9-17-6 y, ratio length/width more than 2-0.

A. acerinus var. alliaceus.

6. Fructifications rose coloured, paling to a fleshy tint on storage. A. roseus.

Fructifications purplish, violet, lilac or partly changing to brown or clay colour.

A. polygonioides.

Cyphella.

1. Fructifications cupulate, sessile or shortly stalked. 2.

Fructifications discoid with reflexed margin (i.e. saucer-shaped). Encrusted, villose hairs on

reflexed part. Spores cylindrical, straight or curved, 3 -2-4 -8 X (6 - 4)— 8—1 1 y.

C. applanata.

2. Fructifications villose with long hairs on the exterior. 3.

Fructifications pruinose with short, sparse hairs or mineral encrustations, or nude. 4.

3. Spores smooth, broad ovate to pip-shaped, 11-2-14-4 x 9—9 - 6— ( 11-6) y. Cupules 1-3 mm. diam.

C. alboviolascens.

Spores finely asperulate, ovoid, 12-15 x 7-8 y. Cupules 1 mm. diam., then becoming flattened.

C. cheesmani.

4. Spores hyaline, ovate to subglobose, 4 - 8-6-4 y diam., smooth or very minutely punctate. Cupules

0-3-1 -5 mm. diam., not hairy, pruinose with mineral deposits. C. farinacea.

Spores lightly coloured, cupules with sparse, short hairs on exterior. 5.

5. Cupules up to 1000 y high, lacking a subiculum. Hyphae thickwalled, 3-4-8-(6-4) y wide.

C. pelargonii.

Cupules up to 600 y high, commonly seated on a reddy-brown subiculum which may be scanty

or absent. Hyphae thinwalled, 1-6-2 -4 y wide. C. variolosa.

Cytidia.

1 . Spores large, ovate or broad elliptical, at first hyaline, later pale coloured with the wall thickening

to 2-5 y and penetrated by peglike projections of the lumen. Fusoid, encrusted cystidia present.

C. habgallae.

Spores otherwise. Cystidia absent. 2.

2. Spores of one kind only, cylindrical to allantoid, 1-5-3 X l-\0 y. Cupular hairs long and floccose.

C. flocculenta.

Spores of two kinds (a) cylindrical or allantoid, 2-2 - 5— (3 - 2) x 6 - 4-9 - 6— ( 11-2) y, ( b ) ovate to

elliptic-depressed, 4 -8-6 -4 x 9-6-12-8 y. Cupular hairs short and scanty.

C. simulans.

Solenia.

1. Cupules 170-250 y diam., 1 mm. long; creamy when dry, with spiral surface hairs in the outer

half. S. natalensis.

Cupules 130 y diam., 250 y long; yellowish when dry, lacking surface hairs.

S. minima.

Cupules up to 500 y diam., creamy yellow, with solid parallel surface hairs.

S. rhoina.

Explanation of the Illustrations.

Throughout the illustrations the following symbols apply: —

A = Acanthophyses.

B = Basidia.

C = Cystidia.

D = Dendrophyses.

G = Gloeocystidia.

H = Hyphae.

HA = Habit.

P = Pseudophyses.

S = Spores.

SE = Section.

SH = Surface hairs.

483

Fig. 1. — Aleurodiscus acerinus var alliaceus drawn from the type of A. acerinus var

longisporus ; Fig. 2. — Aleurodiscus disciformis; Fig. 3. — Aleurodiscus botryosus;

Fig. 4. — Aleurodiscus mirabilis; Fig. 5. — Aleurodiscus polygonioides; Fig. 6. —

Aleurodiscus roseus.

484

Fig. 7. — Cyphella alboviolascens; Fig. 8. — Cyphella farinacea, from the type; Fig. 9. —

Cyphe/la variolosa, from the type; Fig. 10. — Cyphella tabacina, from the type;

Fig. 1 1 . — Cyphella fulvodisca.

485

Fig. 12. — Cyphella, applanata from the type; Fig. 13. — Cyphella pelargonii, from the

type; Fig. 14. — Solenia natalensis, from the type; Fig. 15. — Solenia minima,

from the type.

129828-3

486

Fig. 2\.—PorotheIeum incanum.

487

489

New and Interesting Records of South

African Fungi. Part II.*

By

P.H. B. Talbot.

Sixteen species of fungi newly recorded for South Africa are described here, three

of which are new species of helicosporous Fungi Imperfecti.

Arthrobotrys Oligospora Fres. : Drechsler in Mycologia 29 (1937) 464, f. 3.

Illustrations-. Drechsler, loc. cit. ; Fig. 1.

Collections : 41012, C. O'Connor, on Potato leaves and filter paper in a culture, Div.

of Plant Pathology, Pretoria, Aug. 1954.

A branched, septate, hyaline, repent mycelium, 5-8 p wide, gives rise to erect

conidiophores. These are hyaline and either unbranched or rarely forked near the

base, 4-7 p wide, 0-5-1 mm. long, septate in the sterile basal part. The fertile part

bears swollen nodes at intervals, each having a whorl of short denticles to which the

conidia are attached. The apex of the conidiophore is similarly swollen and denticulate.

The conidia are hyaline, thinwalled, obovate to obpyriform, with a single cross-septum

situated well below the middle of the conidium. A short truncate apiculus is present

and the conidia are sometimes slightly constricted at the septum. The conidia measure

13-16 X 23-27 /*.

This attractive species has not previously been recorded in South Africa. In

determining the species, Drechsler’s (loc. cit.) account of the species of Arthrobotrys

has been followed. The size and shape of the conidia, and the situation of the conidial

septum, are reliable characters on which to differentiate the species.

Dendryphiopsis Atra (Corda) Hughes in Canad. Journ. Bot. 31 (1953) 655, 638, figs.

89-90.

Dendryphion atrurn Corda, leones Fung. 4 (1840) 33.

Illustrations-. Hughes, loc. cit.; Fig. 2.

Collections : 40993, P. H. B. Talbot, on fallen Populus deltoides, “ Goedehoop ”, Piet

Retief distr. Apr. 1954.

The conidiophores are blackish-brown, stout, thickwalled, septate, about 9 p

wide, and arise from the woody substratum singly, forming conspicuous black patches.

The apex of the main axis of the conidiophore is terminated by a conidium. Near

the apex, “ primary, secondary, tertiary, and even quaternary distinct lateral branches

arise each in turn being terminated by a solitary conidium ” (Hughes, loc. cit. p. 638).

The conidia are very dark, thickwalled, with 1—4— (5) septa and a distinct basal hilum.

The five-celled conidia measure 58-65 X 13-15 p.

No chains of conidia were seen, a fact stressed by Hughes, but a few of the conidia

possessed deep constrictions suggesting that their growth had been halted and then

resumed. This species is a new record for South Africa.

Ceratocystis adiposa (Butler) Moreau in Rev. de Mycol. 17 (1952), Suppl. Colonial

No. 1, p. 22.

Part i of this paper appeared in Bothalia 6 (1951) 183-204.

490

Sphaeronema adiposum Butler in India Dept. Agric. Mem. Bot. ser. i, 3 (1906) 40;

Saccardo Syll. Fung. 22 (1913) 926.

Endoconidiophora adiposa (Butler) Davidson in Journ. Agric. Res. 50 (1935) 802.

Ceratostomella adiposum (Butler) Sartoris in Journ. Agric. Res. 35 (1927) 577.

Illustrations'. Sartoris, l.c., figs. 1-4; Davidson, l.c., figs. 3, A-F; Fig. 3 (conidial

stage), Fig. 4 (perithecial stage).

Collections'. 39199, van Niekerk, on pine shooks, Pretoria, Feb. 1951.

The conidial stage of this fungus was found on pine shooks sent in for examination;

subsequently the fungus was grown on malt agar, producing conidia readily and

perithecia after 2\ months in culture.

Conidial Stage: The repent mycelium is first hyaline then pale coloured, branched,

septate, 1-5-3 -4-6 -8 p wide. Erect phialophores arise at right angles to the hyphae.

They are brown, 1-4 celled, commonly two-celled, somewhat constricted at the septa,

thinwalled, surmounted by a one-celled phialide. The phialides are cylindrical or slightly

urn-shaped, brown, ( 1 8)— 23— 26 p long, expanding into a minute collarette at the apex.

Together the phialide and phialophore measure 25-50 x 5 -7-6 -8 p.

The endogenous conidia are produced in basipetal succession from the phialides,

and form chains of up to 20 or more extruded conidia. The conidia vary in form

according to age. When young they are almost hyaline and smooth, subglobose,

broadly elliptical or roughly cylindrical with rounded ends. Later they become deeply

coloured, minutely warted or strongly striately verrucose, broad elliptical to cylindric-

oblong, j(9)— 14— 18 x 8-10 p. Sometimes the last-formed conidium (nearest the

phialide) may be greatly swollen, 18-22 p in diameter, very darkly coloured, with a

strongly warted or semi-spinose wall.

Perithecial Stage : The perithecia are dark, globose, about 200 p in diameter, embedded

in the substratum, with a beak 1-2 mm. long, tapering from 90 p wide at the base to

24 p wide at the apex. The apex is rostrate, divided into a number of brown, septate

hairs, 2 -4-3 -2 p wide, becoming paler towards the ends. Asci were not seen but

copious ascospores extruded from the perithecium and formed an opalescent, slimy

droplet at the apex of the beak. The ascospores are hemispherical, or semilunate

in section, 5-6-4 x 3-3-4 8 p.

Professor W. J. Lutjeharms very kindly suggested the genus Endoconidiophora

thus enabling me to identify this most interesting fungus, which has not before been

recorded in South Africa.

Botryosporium longibrachiatum (Oud.) Maire in Ann. Mycol. 1 (1903) 341; Saccardo

Syll. Fung. 18 (1906) 510.

Botrytis longibrachiata Oudemans; Saccardo Syll. Fung. 10 (1892) 537.

Radaisiella elegans Bainier in Bull. Soc. Myc. de Fr. 26 (1910) 382, PI. 20.

Cephalosporium dendroides Ellis & Kellerman in Journ. of Mycol. 9 (1903) 5, figs. 1-4.

Illustrations'. Bainier, loc. cit. (as Radaisiella elegans)', Kellerman, loc. cit. (as

Cephalosporium dendroides)'. Fig. 5.

Collections'. 40518, J. E. van der Plank, on dead potato stems and leaves in greenhouse,

Div. of Plant Pathology, Pretoria, Apr. 1951.

The conidiophores are 9-12-5 p wide, hyaline, septate, bearing numerous lateral

fertile branches, or pegs where these have been detached. The fertile branches are

(48)— 80— 1 10 p long, expanding in width from the base to the apex, which is rounded

or diamond-shaped in section and is 10-14 p wide. The swollen fertile apex bears in

turn a number of ampullae with innumerable, minute spicules which carry the conidia.

The conidia are hyaline, smooth, one-celled, elliptical, 2 -3-2 -8 X 4-5-5 p.

491

This fungus occurs characteristically in greenhouses on dead or decaying substrata,

and is a strikingly beautiful object in the microscope. The above collection is a new

record for South Africa.

Xenosporella larvalis (Morgan) Linder in Ann. Mo. Bot. Gard. 16 (1929) 322, PI. 25,

figs. 12-15.

Helicoma larvale Morgan in Cincinnati Soc. Nat. Hist. Journ. 15 (1892) 45, f. 9.

Illustrations: Linder, loc. cit. ; Fig. 6.

Collections: 36840, P. H. B. Talbot, Garstfontein, Pretoria distr., Aug. 1948; 36997,

W. G. Rump (595), on mangrove, Isipingo, 1944.

Colonies thin, effuse, powdery, gray-black or brownish. Conidiophores

4-5-3 x 33-40 p, branched, septate, hyaline or dilute fuscous above, the basal cells

fuscous, 3 - 5—5 * 3 p wide, arising from repent, septate, branched fuscous hyphae.

Conidia 18 • 8-22 p diam., the filament 8 p wide, hyaline to dilute fuscous and frequently

coiled round a central cell, the cells divided off by dark septa, acrogenous, coiled once,

divided longitudinally and transversely into 15-17 cells, basal cell oblong and truncate.

The subhyaline conidia divided by dark septa and the subhyaline conidiophores,

also the measurements, place this species as X. larvalis. According to Linder the

conidiophores may be up to 75 p long, and the conidia rarely show a central cell. The

collection cited on mangrove was mixed with much other fungous material, some

of which was another, undetermined helicosporous mould.

Xenosporella rosea sp. nov.

Illustrations : Fig. 7.

Collections: Type, No. 36976, P. H. B. Talbot, on Eucalyptus stump, Hendrina road,

near Ermelo, Feb. 1949.

Colonies effused over bark, coarsely granular, rosy-coloured due to the massed

colour of the conidia. Conidiophores short, hyaline or very dilutely coloured, up

to 10 p long and 4-5-3 p wide, arising from a repent mycelium composed of branched,

septate, hyaline to deep fuscous hyphae, 4-7-5 p wide, sometimes with the shorter

cells constricted to appear moniliform. Conidia acrogenous, coiled |-1£ times round

1-5 deep-fuscous central cells, which are 4-6-7 p in diameter and subglobose. The

conidia are hyaline to fuscous, usually only very dilutely coloured or slightly darker

at the septa, divided first by transverse walls then by longitudinal walls into 26-34 cells

when mature, the coil 33-44 p diam.; filament 14-6-18-5 p diam.; conidia widest

at the middle, tapering to the ends, the distal end rounded, the base rounded or

occasionally somewhat truncate. In optical cross section the filament appears to be

divided by radial walls into 4-6 cells.

This species corresponds with none in Linder’s monograph of the Helicosporous

Fungi Imperfecti (in Ann. Mo. Bot. Gard. 16, 1929, 318). In the size of its conidia

it comes nearest to X. thaxteri, but the latter has black colonies, longer conidiophores

and wider conidial filaments. The presence of several central cells in the conidium is

characteristic of the new species.

Xenosporella rosea sp. nov.

Fungus effusus, pulverulentus, rosaceus. Conidiophora brevia, usque ad 10 p

longa, hyalina vel pallidissime colorata, 4-5-3 p diam., ex hyphis repentis orientia.

Hyphae repentae, ramosae, septatae, hyalinae vel atro-fuscae, interdum moniliformae,

4-7-5 p diam. Conidia 33-44 p diam., filo 14-6-18-5 p diam., hyalina vel fusca,

murjforma, cellulis 26-34, acrogena, in spiram circa 1-5 cellulas centrales, fuscas,

subglobosas, f-l^-ter glomeratis.

492

Helicoma interveniens sp. nov.

Illustrations : Fig. 8.

Collections'. Type, No. 40519, P. H. B. Talbot, on dead wood, Transvaal sine loc., 1950.

Effused, forming a blackish, hirsute colony. Repent mycelium sparse, fuscous,

septate, branched, 5-6 /z wide. Conidiophores erect, single or in small groups, not

branched, straight or often irregularly bent, deep brown but pellucid, the terminal

cell almost hyaline and rounded when immature, but tapering to a minute sporogenous

tooth when fertile. The conidiophores are multiseptate, 125-230 x 5-6 /u,, with up

to 7 lateral sporogenous teeth in old conidiophores. Conidia acropleurogenous, borne

on minute sporogenous teeth, coiled tightly 1^ times, hyaline to very pale yellowish,

with (7)— 1 1—13 hyaline septa, the filament 6-5-8 ^ wide, with both ends rounded and

without constrictions at the septa; diameter of the coiled conidia 18-22 \x.

This fungus has characters in common with several other species, but differs from

each in certain well-marked features. It is rather similar to H. curtisii Berk., but the

latter differs in having fewer septa in the conidia, tapering-rounded instead of rounded

basal cells to the conidia, very rarely pleurogenous conidia, and coiled spores of a

smaller diameter. H. curtisii was figured and described in Bothalia 6 (1951) 187, f. 9.

Other species which required careful comparison were H. velutinum Ellis, H. mulleri

Corda and H. ambiens Morgan. The new species differs from H. velutinum in having

pleurogenous conidia without constrictions at the septa. It differs from H. mulleri

in having minute, unbranched sporogenous teeth and larger conidia and from

H. ambiens in having unbranched conidiophores and more septa in the conidia.

Helicoma interveniens sp. nov.

Fungus effusus, subniger, hirsutus. Hyphae repentae parcae, fuscae, septatae,

ramosae, 5-6 /z diam. Conidiophora erecta, non ramosa, recta vel curvata, atro-fusca,

pellucida, multiseptata, 125-230 X 5-6 /z. Conidia in dentibus minutis orientia,

acrogena, 18-22 n diam., filo 6-5-8 /z diam., hyalina vel substraminea, septis hyalinis

(7)— 1 1—13, non constricta, in spiram 1^-ter glomeratis.

Helicosporium gracile (Morgan) Linder in Ann. Mo. Bot. Gard. 16 (1929) 281, PI. 13,

figs. 4, 7; PI. 14, figs. 9-11.

Helicomyces gracilis Morgan in Cincinnati Soc. Nat. Hist. Journ. 15 (1892) 40, f. 2.

Illustrations'. Linder, loc. cit. ; Fig. 9.

Collections'. 36949, P. H. B. Talbot, on dead Acacia mollissima twigs, Atholl Expt.

Station, E. Tvl., Feb. 1949.

The colonies form inconspicuous, greenish-yellow patches on wood and bark,

composed of a repent mycelium of fuscous, septate hyphae, 3-4-3 /z wide, bearing

occasional sporogenous teeth. The conidiophores arise erect from the repent hyphae.

They are fuscous, septate, unbranched, with acute apex, 58-140 x 3-5-4-5-(7) /z,

bearing minute sporogenous teeth acropleurogenously, from which the conidia arise.

The conidia are hyaline, glaucous-yellowish in a mass, coiled up to 3^ turns, filamentous,

the filaments up to 1*5 /z wide, seldom showing the septa, the coils 9-15 /z diam., partly

uncoiling when wetted.

Linder emphasises that H. gracile has dilutely coloured conidiophores and hyphae.

The present specimen is pellucid throughout, though not very dilutely coloured. It

differs from H. vegetum Nees (fide Linder, loc. cit. p. 277) in its creeping habit and the

repent mycelium which bears conidia either directly or upon conidiophores.

Helicosporium lumbricoides Sacc. emend Matruchot; Linder in Ann. Mo. Bot. Gard.

16 (1929) 282, PI. 15, f. 5.

Illustrations : Linder loc. cit.; Fig. 10.

493

Collections : 40521, P. H. B. Talbot, under bark of Eucalyptus, Fountains, Pretoria,

Aug. 1952.

The colonies are effused, light- to sooty-grey, readily separable from the substratum.

Conidiophores arising from a dark repent mycelium, ascending or more or less erect,

subhyaline to pellucid deep brown, septate, branched and much anastomosed,

3 -4-5 -7 p wide. Conidia pleurogenous, borne singly on small sporogenous teeth

situated on the lower parts of the conidiophores, hyaline, the filament hygroscopic,

1-5-2 p wide, coiled 2£-3 times, indistinctly many-septate (10-14 septa seen); Coiled

conidia 20-25 p diam.

This specimen was found together with Lasiosphaeria pezizula and its conidial

stage, Helicoma curtisii, on the same piece of bark.

It is evident from Linder’s descriptions (loc. cit.) that Helicosporium lumbricoides

and H. lumbricopsis Linder are rather similar. I have determined the present collection

as H. lumbricoides because the mycelial mat is easily detached, the conidiophores are

of the width given for this species and form an anastomosed network, and the conidial

filaments are 2 p or less in width and have fewer septa than in If. lumbricopsis.

Helicosporium ramosum sp. nov.

Illustrations : Fig. 11.

Collections: Type, No. 40546, P. H. B. Talbot, on dead wood, Fountains, Pretoria»

Nov. 1951.

The colonies are sparse, white and pulverulent. The conidiophores are erect,

30-240 p long, 3-2-4 p wide, branched below and often anastomosing above, fuscous,

subhyaline near the apex, arising from a repent mycelium. The repent hyphae are

subhyaline to fuscous, 2 -5-4 -5 p diam., branched and anastomosed. The conidia

are hyaline, white in a mass, acropleurogenous, borne on lateral denticles on the main

axis of the conidiophores, or on terminal or lateral denticles on branches of the

conidiophores or repent mycelium. These branches are up to 40 p long. The conidial

coil is 19—24 — (27) p diam., the filament 1-6 p wide, with about 15 septa, coiled 1^—24-

times.

This species is distinguished by its white colonies and by the branched anastomosed

conidiophores and conidial size, a combination of characters not found in any description

I have traced. It somewhat resembles H. lumbricoides, but differs in the way in which

the conidiophores branch.

Helicosporium ramosum sp. nov.

Fungus parcus, albidus, pulverulentus. Conidiophora ex hyphis repentis orientia,

erecta, 30-240 x 3-2-4 p, fusca, subhyalina prope apicem, infra ramosa, supra

anastomosa. Hyphae repentae subhyalinae vel fuscae, ramosae, anastomosae, 2 • 5-4 • 5 p

diam. Conidia hyalina, acropleurogena, in dentibus lateralis in axe principali

conidiophorum vel in ramis conidiophorum et hypharum repentium orientia. Rami

usque ad 40 p longi. Conidia 19—24 — (27) p diam., filo 1-6 p diam., septis circa 15,

in spiram 1^-2-Mer glomeratis.

Coremiella ulmariae (McWeeney) Mason apud Hughes in Canad. Journ. Bot. 31 (1953)

640, figs. 92, 93.

Stysanus ulmariae McWeeney in Irish Naturalist 4 (1895) 273-277, figs. 1-6.

Coremiella cystopoides Bubak & Krieger apud Bubak in Ann. Mycol. 10 (1912) 52-53,

f. ii.

Illustrations: McWeeney, loc. cit.; Hughes, loc. cit.; Bubak, loc. cit.; Fig. 12.

Collections: 41016, H. J. Potgieter, College of Agriculture, Potchefstroom, Dec. 1954.

494

The fungus forms small synnemata up to 1 mm. high, which are composed of a

stalk up to about 0-2 mm. wide, made up of loosely interwoven parallel hyphae, which

expand near the apex by repeated dichotomous branching into a subglobose head

up to about 0-5 mm. diam. The whole structure is whitish to grey, with the stalk

usually a darker grey. The hyphae of the synnema are sparingly branched, smooth,

hyaline to pale brown, 4 -8-6 -4 p diam. The terminal branches composing the head

are closely septate and become transformed into rows of endoconidia, the rows being

20-140 /x long and 5 p wide. The endoconidia are cubical to rounded, with a thickened

wall bearing a minute pore in each end wall. They are dilutely coloured and about

5 /x wide. The conidia are separated by intermediate cells whose lateral walls remain

thin and eventually collapse inwards to free the conidia.

This species has not been recorded in South Africa before, though Prof. Liitjeharms

tells me he has isolated it many times in Bloemfontein on decaying plant stems.

Hughes (loc. cit.) gives an interesting account of the way in which the conidia

are formed. I am indebted to the Director of the Commonwealth Mycological Institute

for a copy of McWeeney’s diagnosis of C. ulmariae and for the information that the

Institute regards C. cystopoides as a later synonym.

Piricularia oryzae Cavara, Fungi Longob. exsicc. No. 49, 1891; Briosi & Cavara,

I Funghi Parass. No. 188, 1892; Cavara, Atti 1st. bot. Univ. Pavia, ser. ii, 2 (1892)

280.

Illustrations: Watts Padwick, Manual of Rice Diseases (1950) f. 3.; Fig. 13.

Collections : 41033, Gerrit Moerdyk, on Oryza sativa, Nylstroom distr., Jan. 1953.

The fungus forms discrete to confluent whitish, greyish or light brown spots on

leaves of rice, surrounded by a dark reddish brown discoloration. The whitish area

is about 0-5-3 mm. in its greatest diameter, becoming tinged greyish when the fungus

is well developed. The conidiophores arise singly or more often in groups. They

are unbranched, hyaline to pale brown at the base and hyaline at the apices, with two

or three septa, 3-2-5 /x wide and 90-140 p long, cylindrical, attenuated to a sharp

apex. The conidium is borne terminally on the conidiophore and is pushed aside by

a lateral offshoot of the conidiophore which in turn bears a terminal conidium. This

process may be repeated 3-4-(7) times as lateral offshoots arise in succession. Each

conidium is solitary, acrogenous, hyaline to pale olivaceous, usually 3-celled (1-4 celled),

subglobose when young, becoming ovate or typically pyriform, with a small spicular

point of attachment, 16-25 x 7-11 p.

Doidge (in Bothalia 5, 1950, 685) lists this species from Mocambique, but it has not

hitherto been recorded in South Africa. Since the above collection was made, it has

been seen fairly frequently. Watts Padwick (Manual of Rice Diseases, 1950, pp. 1-21)

gives a comprehensive account of this and allied species of Piricularia. According

to Watts Padwick, Nishikado was able to separate four species of Piricularia “ each

having distinct morphological characters and a rather narrow host range ”. The

following were the species dealt with by Nishikado: —

Piricularia oryzae Cav., on Oryza sativa ;

Piricularia grisea (Cooke) Sacc., on Digitaria sanguinalis ;

Piricularia setariae Nishikado, on Setaria italica and S. viridis;

Piricularia zingiberi Nishikado, on Zingiber mioga and Z. officinale.

In the following two entries, the determination of the fungus as to species has been

made not on the morphology of the fungus, but purely by following the host range

given by Nishikado, since these fungi are only critically separable from Piricularia

oryzae. Each constitutes a new record for South Africa.

Piricularia setariae Nishikado: Watts Padwick, Manual of Rice Diseases (1950) 8.

Collections: 39097, S. J. Truter, on Setaria italica, “ Ukulinga ” farm, Pietermaritzburg,

Aug. 1950.

495

Piricularia grisea (Cooke) Saccardo, Mich. 2, p. 148; Sacc. Syll. Fung. 4 (1886) 217;

Watts Padwick, Manual of Rice Diseases (1950) 8.

Trichothecium griseum Cooke in Ravenel Amer. Fungi n. 580.

Collections : 41015, S. J. Truter, on Digitaria sanguinalis, Pietermaritzburg distr.,

Sept. 1954.

Cordyceps tuberculata (Lebert) Maire in Bull. Soc. d’Hist. Nat. de l’Afrique du Nord

8 (1917) 165; Petch in Trans. Brit. Myc. Soc. 16 (1931) 213-217; Dingley in

Trans. Roy. Soc. N.Z. 81 (1953) 338.

Akrophyton tuberculatum Lebert in Zeitschr. f. wiss. Zool. 9 (1868) 375.

Cordyceps sphingum (Schw.) Berk. & Curt, in Journ. Linn. Soc. Bot. 10 (1868) 375.

Illustrations : Petch in Trans. Brit. Myc. Soc. 9 (1924) PI. 2, f. 12 (as Torrubiella

ochracea ); Lloyd Myc. Notes 5 (1916) 591-592, figs. 830-832 (as Cordyceps sphingum) ;

Herre in Mycologia 15 (1923) 280, f. 1. (as C. sphingum ); Fig. 14.

Collections : 39185, R. Ripley, on dead sphingid moth, College of Agriculture, Cedara,

Apr. 1949; 41014, S. J. Truter, on dead moth, Pietermaritzburg distr., Sept. 1954.

Mycelium whitish, covering the whole under surface of a dead moth, arising from

which are numerous closely aggregated cylindrical to clavate or irregular clavae, bearing

perithecia in groups towards or at the apex, or frequently only on one side of the clava.

The clavae are sometimes subulate, pointed and sterile. The clavae are whitish to

very pale yellow, 2-3 mm. long and 0-3-0 -5 mm. wide, composed of very closely

interwoven hyphae, which are hyaline, septate, much branched, narrow. The perithecia

are found in groups, very slightly immersed in the hyphae of the clavae, bright pale

lemon yellow all over, becoming ochraceous, ovoid to conical, glabrous, somewhat

pellucid, 0-3-0-4 mm. high and 0-2-0-3 mm. diam. Paraphyses absent. Asci

cylindrical, tapering to the base, rounded at the apex, 3—4 - 8 x 225-310 g. Ascospores

8 per ascus, filamentous, hyaline, almost as long as the asci, septate, becoming divided

into part spores.

A full account of the history, synonymy and morphology of this species is given

by Petch (1931, loc. cit.). This species is a new record for South Africa.

Phymatotrichum silvicolum Taubenhaus & Watkins in Amer. Journ. Bot. 24 (1937)

387-390, figs. 1-22.

Illustrations'. Taubenhaus & Watkins, loc. cit.; Fig. 15.

Collections'. No. 41033, P. H. B. Talbot, on moist soil, Nel Street, Pretoria, Jan. 1955.

The fungus forms a cottony to pulverulent, white mycelial mat on moist soil,

later changing to a light tan or greyish colour. The hyphae are hyaline, branched,

septate, commonly 7-10 g wide, forming synnemata or remaining separate. Lateral

swellings from the hyphae develop into branches which at first are simple, but later

form somewhat swollen, terminal clusters of branches. These are the conidiophores

and together with a part of the main stalk become minutely papillate. The conidia

are attached by minute stipes and cover the whole surface of the conidiophores so

that the latter are practically obscured. The conidia are hyaline, globose or subglobose

to ovoid, with thin walls and without septa, 2 -5-4 -5 g in diameter, produced in vast

numbers.

In their excellent description of this species Taubenhaus and Watkins differentiate

it from other species of Phymatotrichum by “ the length and slenderness of the

conidiophores, the characteristic apical cluster of dendroid or digitate branches which

are clavate or moderately inflated, and the distribution of the actual spore-bearing

area from well down on the stalk of the conidiophore to the apices of all its branches.”

This fungus was first brought to my notice by Dr. S. J. Truter who found it in the

Union Buildings gardens after a spell of wet weather in February. Subsequently it

was found under pine trees near the Botha Statue. I am much indebted to Prof.

W. J. Liitjeharms for identifying this fungus and informing me that at one time it was

quite common under pine trees at the University of the Orange Free State, Bloemfontein.

496

Fig. 1. — Arthrobotrys oligospora\ Fig. 2.— Dendryphiopsis atra; Fig. 3. — Ceratocystis

adiposa (conidial); Fig. 4. — Ceratocystis adiposa (perithecial).

497

Fig. 5. — Botryosporium longibrachiatum; Fig. 6. — Xenosporella larva /is;

Fig. 7. — Xenosporella rosea.

498

Fig. 8. — Helicoma interveniens; Fig. 9. — Helicosporium gracile;

Fig. 10. — Helicosporium lumbricoides'. Fig. 11. — Helicosporium ramosum.

499

Fig. 12. — Coremiella ulmariae ; Fig. 13. — Piricularia oryzae ;

Fig. 14. — Cordyceps tuberculata.

500

Fig. 15. — -Phymatotrichum sihicolum.

501

Three Species of Curvularia from Pinus.

By

G. C. A. van der Westhuizen.

Since Boedijn (1) created the genus Curvularia in 1933, species have been reported

from the soil (8), on Gladiolus bulbs (5), on vegetable seeds (3), and on grasses (2, 4, 6).

This paper records two species found on pine trees and one species causing blue stain

of pine timber. No previous records of Curvularia on such substrata have been traced.

Boedijn (1) divided the genus into three groups on the basis of the number and

location of the darker, enlarged cells in the conidia. Two of the fungi described here

have three-septate conidia with the two central cells larger and darker in colour than

the others; these features are characteristic of the Maculans group. The third specimen

has the three-septate conidia of the Lunata group, in which only the penultimate cell

is larger and darker.

A. Maculans Group.

Curvularia maculans (Bancroft) Boedijn in Bull. Jard. Bot. Buitenzorg, ser. iii, Vol. 13,

1933, pp. 120-134.

Illustrations'. Fig. I; Plate 1 A.

On Czapek agar (7), the colonies are at first subhyaline but soon turn black. They

are appressed, spreading, velvety, sometimes with superficial dull brown cottony hyphae,

raised and tufted in the centre, with the reverse black.

The submerged mycelium is subhyaline to light brown, septate, branched, smooth

or sometimes constricted at the septa and 3-6 p. wide. The aerial mycelium is darker

brown, straight-walled, branched, septate and 3-6 p wide.

The conidiophores arise as lateral branches of the aerial and submerged mycelium

and are light brown in colour, and narrow at the base, mostly 3-4 p wide. Towards

the fertile region the conidiophores widen gradually. The fertile region is dark brown

and geniculate with many protuberances on which the conidia are borne in whorls or

irregular spirals. Upon separation of the conidia, circular scars are left on these

protuberances. In this region the hyphae are 4-6 p wide. The conidiophores vary in

length from 76 p to over 200 p.

The conidia are elliptical to obovate and broadly rounded at the ends. They

are widest at or above the middle, smooth and often slightly constricted at the proximal

septum. The terminal cells are small and light brown with a conspicuous scar in the

basal cell. The central cells are larger, thicker-walled, and of a dark, almost opaque

brown colour. In size they vary from 18-24 p long and 12-16 p wide, being mostly

22 x 16 p.

Habitat : On shoots of Pinus patula infected with Diplodia pinea , Louis Trichardt,

Transvaal. May 1953.

This fungus, with its more or less elliptical conidia of which both central cells

are enlarged and dark, agrees very closely with Boedijn’s description except for a small

but insignificant discrepancy in spore size (Boedijn’s measurements were 19-26 x

11-17 p).

Curvularia brachyspora Boedijn loc. cit.

Illustrations: Fig. 2; Plate 1 B.

502

On Czapek agar, the colony is at first subhyaline, soon turning dark greenish-brown

to black, spreading, appressed, cottony, with the reverse black.

The submerged mycelium is hyaline or subhyaline, light greenish-brown or light

brown, branched, septate, straight-walled, or constricted at the septa and 2-6 p wide.

The aerial mycelium is darker brown, straight-walled, smooth, septate and 2-6 p wide.

The conidiophores are brown, erect, more or less straight, single, unbranched,

arising as lateral or terminal branches of trailing or aerial hyphae. They are narrow

at the base but widen gradually towards the darker, geniculate, fertile region where

conidia are borne in spirals on fairly prominent protuberances. Their length is variable,

30->200 x 3-6 /a.

The conidia are elliptical, widest at the middle, markedly unequal sided, rounded

at the ends, three-septate with the wall slightly constricted at the outer septa. The

middle septum is thick and dark. The two central cells are larger than the light brown

terminal cells, and of a dark translucent brown colour. The conidia measure

16-24 X 10-14 /a, being mostly 22 x 14 /a.

Habitat : On shoots of Pinus pa tula infected with Diplodia pinea, Rhenosterhoek,

Oct. 1952.

Like the previous one, this fungus has both the middle cells of the conidia larger

and darker than the terminal ones. It differs from C. maculans , however, in that these

cells appear clear and translucent although dark in colour, while the conidia are markedly

unequal sided.

In all respects, this fungus agrees closely with Boedijn’s (1) description of

C. braehyspora.

B. Lunata Group.

Curvularia pallescens Boedijn, loc. cit.

Illustrations'. Fig. 3; Plate 1, C.

On Czapek agar, the colony is at first hyaline but soon turns dark brown to black.

It is slightly raised in the centre, appressed, spreading, velvety, with the reverse black.

The submerged mycelium is subhyaline to light greenish-brown, later turning

olivaceous, septate, branched, more or less straight-walled or constricted by the septa,

2-8 /a wide. The aerial mycelium is more or less straight-walled, septate, light brown,

and 2-8 p wide.

Conidiophores arise as terminal or lateral branches of the submerged and aerial

hyphae. They are brown, simple, erect, more or less straight, septate, narrow at the

base but widening gradually towards the darker, geniculate, fertile region where spores

are borne in spirals on well developed protuberances which show circular scars upon

separation. Their length is very variable, 60 /a to over 200 /a x 3-6 p.

The conidia are elongate elliptical, unequal sided or sometimes slightly bent,

rounded at the ends, with a marked hilum at the proximal end. They are pale brown,

smooth, three-septate with the third cell slightly larger and thicker-walled than the

others, often slightly darker coloured or both central cells slightly larger and darker

than the terminal cells which are sub-hyaline or pale brown, 18-26 x 8-10 p.

Habitat : In blue-stained sapwood of Pinus patula, Sabie. March 1952.

This fungus differs from the previous two in having narrower conidia without

marked contrast in the colour of the cells. Nor is the third cell markedly larger than

the others, as is characteristic of the Lunata group.

Boedijn (1) described the conidia of C. pallescens as elongate elliptical, inequilateral,

or slightly bent, very faintly brownish, three-septate, the third cell or both central

cells larger and slightly darker than the almost- colourless terminal cells,

21-5-30 x 7-11 p. Groves & Skolko (3) described and illustrated a species of

Curvularia whose pale brown conidia were closely similar to Boedijn’s and had the

third cell slightly larger and darker than the others, which is a feature of the species

503

described above. Marchionatto (6) reported from the Argentine that a species of

Curvularia isolated from rice had acropleurogenous, ellipsoid, straight, geniculate

conidia, 24-30 x 7-10 p, consisting of four cells of irregular size and shape, the top

and bottom ones hyaline and the middle ones greyish. Although he identified this

fungus as C. pallescens, there is some doubt whether this is correct as it does not match

Boedijn’s description. Nor does it resemble the species described by Groves & Skolko,

or the present fungus from Pinus. The last, however, agrees closely with both Boedijn’s

and Groves & Skolko’s (3) descriptions and illustrations of C. pallescens, except that

it appears to be slightly darker in colour than Boedjin’s description would imply.

Groves & Skolko’s (3) illustrations, however, indicate that the conidia of their fungus

were slightly darker in colour than is apparent from their description, so that there is

very close similarity between their fungus and this one. The conidia of the fungus from

Pinus are definitely brown but of a light, translucent tint, which is much lighter than

in the other two species described. Other than this, there is a close similarity between

this fungus and the descriptions of C. pallescens. For the present it appears, therefore,

that this fungus should be identified as C. pallescens rather than be given independent

specific rank.

Of great interest too is the sap-staining property of this fungus. It was isolated

from a piece of log showing bluish grey streaks. Inoculation of sterilised blocks of

sapwood with this fungus caused pale greenish grey stains to develop in the blocks

after 14 days incubation. Continued incubation caused intensification of the stain

to a bluish grey colour. This intensification of colour was associated with a colour

change of the hyphae from pale greenish brown to dark brown, in the tracheids and

rays, together with an increase in the number of hyphae present. Chips removed

from these blocks and incubated on malt agar, yielded pure cultures of the fungus

again. (Plate 1, D).

Literature Cited.

(1) Boedijn, K. B liber einige phragmosporen Dematiacien. Bull. Jard. B.

Buitenzorg, ser. iii, 13 (1933) 120-134.

(2) Bugnicourt, F Des especes du genre Curvularia isolees des semences de Riz-

Rev. gen. Bot., 57 (1950) 65-77.

(3) Groves, J. W. & Skolko, A. J... Notes on seedborne fungi. Ill Curvularia. Can. Journ.

Res. Sect. C. 23 (1945) 94-104.

(4) Hynes, H. J Species of Helminthosporium and Curvularia associated with

Root-rot of wheat and other graminicolous plants. Journ .

Proc. Roy. Soc. N.S.W. 70 (1937) 378-391.

(5) Limber, D. P Notes on Curvularia species on Gladiolus with five new

distribution records. Plant Dis. Reptr. 33 (1949) 66-69.

(6) Marchionatto, J. B El 'manchado' do los granos de Arroz y los hongos que lo

acompanon. Rev. Argent. Agron. 10 (1943) 114-116.

(Seen Rev. Appl. Myc. 23, 1944, 356).

(7) Smith, G An introduction to Industrial Mycology, 2nd Ed., London,

1942, p. 172.

(8) Timonin, M. T Soil microflora in relation to manganese deficiency. Sci.

Agric. 30 (1950) 324-325.

505

Plate 1.

A. — C. maculans.

B. — C. brachyspora.

C. — C. pallescens.

D. — Hyphae of C. pallescens in medullary ray of Pinus patula (radial longitudinal

section).

507

Notes on Certain South African Erythrina Species.

By

L. E. Codd.

Read at the meeting of the South African Association for the Advancement of Science

at Bloemfontein, July 1954.

The notes recorded here are supplementary to the review of South African

Erythrina species by Collett in Bothalia, 4: 219 (1941), which presents a sound

reflection of the position, except in regard to the line of demarcation between

E. lysistemon Hutch, and E. caffra Thunb. Further information also seemed desirable

on the distinction between E. latissima E. Mey. and E. abyssinica Lam. (= E. tomentosa

R. Br. ex A. Rich.). Collett omits to supplement his conclusions with supporting

data and therefore it is not always clear what evidence influenced him in making his

decisions. It was in order to remedy this lack, especially with regard to the correct

naming of species occurring in the Kruger National Park, that the present studies

were undertaken.

1. E. lysistemon Hutch, and E. caffra Thunb.

E. lysistemon Hutch., Kew Bull. 422 (1933) is a name that has not generally been

taken up. It is based on certain Transvaal specimens, namely: Belfast District,

Rietvlei, Smuts 66 (type); Pretoria District, Nelson 84; Magaliesberg, Burke ;

Rustenburg District, Nation', Soutpansberg District, Hutchinson and Gillett 4155.

In distinguishing his species from E. caffra Thunb., he states: “ Superficially the

two species are very much alike, but I was induced to investigate further on account

of the small and narrow vexillum and the peculiar shape of the leaflets of the Transvaal

plant. That these were of importance was revealed on dissection, for in the new species

the vexillary stamen was found to be quite free from the split staminal tube, whilst in

E. caffra it is adnate to it below the middle ”.

Collett upholds E. lysistemon, but to a more limited extent citing, in addition

to the type, Leemann, N.H. 22845 from Pretoria and Taylor 732 from the Soutpansberg.

The rest of the Transvaal specimens and all the Natal specimens he places under

E. caffra. An examination of the material indicated that this division was not in

agreement with natural affinity and that the line of distinction between the two species

would more likely be found in the area between the Kei River and southern Natal,

an area from which specimens were almost completely lacking. An opportunity to

study this critical zone more closely occurred in August 1953, when a tour was made

from Grahamstown to Durban. A visit to the type area of E. lysistemon was made

shortly afterwards in September 1953.

The examination of many dried and fresh flowers indicates that the stamen

character stressed by Hutchinson is not a reliable one, but that general appearance

of the inflorescence, together with combined measurements of floral parts, enables

one to classify the material into two natural groups. The distinguishing characters

are summarised in Table I and, although there is overlapping in some of the dimensions,

there was practically no impression in the field of the two species grading into one

another. Diagnosis in the Herbarium is most reliably based on length of the wings

and keel, and the ratio of vexillum length to calyx length. In only two cases did the

impression gained in the field disagree with subsequent measurement, and these were

508

both cultivated trees, one near Mqanduli and one at the Marburg Mission near Port

Shepstone. It is felt that these two isolated cases, although to some extent intermediate

in character, do not break down the distinction between the two species. It may be

mentioned that the difference in leaflet shape put forward by Hutchinson does not hold

good. The leaflets of E. caffra are constantly ovate-deltoid, while those of E. lysistemon

vary from ovate-deltoid to hastate, so that it is not always possible to distinguish

between the two species when flowers are absent. A small difference, which may have

significance, became apparent when seeds of the two species were germinated in seed

tins. In E. caffra, the hypocotyl always lengthened, raising the cotyledons about 2 cm.

above soil level while, in E. lysistemon , the cotyledons invariably remained at soil level.

Table I.

Comparison of E. caffra and E. lysistemon.

E. caffra Thunb., Prodr. PI. Cap. 121 (1800); FI. Cap. ed. Schult. 559 (1823); E. Mey.,

Comm. PI. Afr. Austr. 149 (1835); E. & Z., Enum. PI. Afr. Austr. 259 (1836);

Harv. in FI. Cap. 2: 236 (1862), pro parte; Sim, For. FI. Cape Col. 201, t.53

(1907); Collett, l.c., pro parte.

E. viarum Tod., Nuovi Gen. e Spec. 62 (1861).

E. constantiana Micheli in Rev. Hortic. 68: 524 (1896).

Plates 1 and 2.

Type: A Thunberg specimen at Uppsala “e Cap. b. Spei ”.

Natal. — Port Shepstone, Codd 8000.

Cape. — Libode, Codd 7990. Mqanduli, Codd 7988. Elliotdale, Codd 7985.

Willowvale, Kriel FD 8865. Kentani, Pegler 89; 1134; Marais 476. Komgha,

Flanagan 319. East London, Smith s.n.; Codd 1919. Albany, Galpin 67 ; Codd 7976.

Alexandria, Sim 2135; Archibald 5906. Uitenhage, Zeyher 1691. Humansdorp,

Four cade 1406; 4466.

Cultivated. — Durban, Collett s.n.; Forbes s.n. Stegi, Swaziland, Keith s.n.

E. caffra occurs near the coast from Port Shepstone to Humansdorp Districts,

the furthest it was seen from the coast, in the wild state, being about 35 miles in the

Albany District. It is associated with coastal forest and wooded stream banks,

averaging 30 to 40 feet high, but often attaining to 60 feet or more. The flowers are

normally orange-scarlet in colour, though near East London an occasional tree with

cream flowers is found. They are characterised by having a relatively short, broad

vexillum which arches strongly away from the flowering axis, exposing the stamens.

The wings and keel are comparatively long, exceeding 2 cm. in length, while the peduncle

is short, rarely exceeding 5 cm. in length.

509

In 1861, Todaro described two species which appear to show a close relationship

to E. caffra, namely, E. viarum and E. insignis. A reply to an enquiry sent to the

Director, Instituto ed Orto Botanico dell’ Universitata di Palermo, reveals that

corresponding type specimens apparently do not exist, but that trees referred to under

the name of E. viarum are growing in Palermo. A specimen at Kew collected by Wilman

in Palermo under the name E. viarum is, according to our liaison officer at Kew,

identifiable as E. caffra. E. insignis is described at some length by Todaro but, in the

absence of supporting measurements, the species cannot be identified with certainty.

It may be the same as the species described later as E. lysistemon Hutch. As there is

no type specimen and hence little chance of its ever being satisfactorily placed, it is

felt that the name may justifiably be disregarded.

There is little doubt that E. constantiana Mich, can be relegated as a synonym

of E. caffra Thunb. The plate and description are based on a tree of unknown origin

in the garden of M. Constant at Golfe Juan in the South of France. The dimensions

of the flowers and leaves are somewhat larger than is usual in E. caffra , but this is no

doubt due to cultivation. It is fairly evident that E. lysistemon came into cultivation

in Europe and America as E. caffra before the true E. caffra, with the result that the

latter received a new name. This was confirmed when cultivated specimens were sent

to us for naming by Miss Elizabeth McClintock of the California Academy of Sciences,

San Francisco. The trees grown in America as E. caffra were, in fact, E. lysistemon,

while true E. caffra was cultivated under the name E. constantiana.

E. lysistemon Hutch, in Kew Bull. 422 (1933); Collett in Bothalia, 4: 223 (1941).

E. caffra auct. plur. non Thunb. e.g., Wood, Natal PI. 6: t.542 (1912); Phillips in

FI. PI. S. Afr. 2: t.59 (1922); Baker, Trop. Legum. 369 (1925); Burtt Davy,

FI. PI. Ferns Tvl. Swaz. 2: 415 (1932); Collett, l.c., pro parte; Codd, Trees

and Shrubs K.N.P. 70, t.66 (1951).

Plate 3.

Type: Smuts 66, Rietvlei, Belfast District, Transvaal.

S. Rhodesia. — Gilliland 706. Matobo, Miller 1643, 1879; Plowes 1461. Bikita,

Wild 4417. Belingwe, Wild 4343.

Mocambique. — Manica e Sofala, Pedro and Pedrogao 148, 7549, 7560, 8053.

Bechuanaland. — Kanye, Hillary and Robertson 570.

Transvaal. — Soutpansberg, Hutchinson and Gillett 4155; Taylor 732; Bailey in

Hb. Burtt Davy H3015. Letaba, Botha s.n.; Krige 25. Waterberg, Story 1528;

Meeuse 9335. Zeerust, Gerstner 3356. Rustenburg, Rose Innes 47; Turner 16.

Krugersdorp, Bunting 31. Brits, Prosser 1356. Pretoria, Burt Davy 1777, 1850;

Phillips 1648; Mogg 15157; 15742; Leeman in Nat. Herb. 22845; C. A. Smith 309;

Codd 6149, 6547; de Winter 395; Marais 1; Prosser 1038. Belfast, Smuts 66 (PRE,

isotype); Codd 8003. Lydenburg, Barnard 534; Codd 8005, 8006. Pilgrims Rest,

Smuts and Gillett 2344. Nelspruit, Pole Evans 3927 ; Codd 8008 ; Codd and de Winter

5127; van der Schyff 13, 402. Barberton, Galpin 494; Legal H2462.

Swaziland. — Codd 7807.

Natal. — Hlabisa, Pole Evans 3648. Mtunzini, Thode A1528. Entonjaneni, Codd

1916. Durban, Medley Wood 10016; Thode A1514; Lansdell s.n. Umzinto, Smuts

s.n.; Codd 8002. Port Shepstone, Codd 7998, 7999; Dyer and Dohse 5417.

Cape. — Bizana, Codd 7993, 7994, 7997. Libode, Codd 7991. Mqanduli, Codd

7986, 7987. Elliotdale, Codd 7983. Engcobo, Codd 7981, 7982.

510

According to Baker, l.c., the species occurs also in Nyasaland and Angola. Its

distribution overlaps that of E. cajfra from Port Shepstone to Bashee Mouth and

extends further inland. It inhabits drier situations than E. cajfra, such as scrub forest,

coastal sand dunes, dry savannah and rocky, wooded slopes, rarely exceeding 40 feet

in height and usually considerably less. The flowers are normally brilliant scarlet, the

main variants being a pink-flowered form near Umkomaas and a tree with orange

scarlet flowers at Marburg Mission near Port Shepstone. The vexillum is relatively

long and narrow, arched to a smaller extent than in E. cajfra and folded to enclose the

stamens. The wings and keel are comparatively short, rarely exceeding 1 • 5 cm. in

length, while the peduncle is seldom shorter than 6 cm.

Specimens from Southern Rhodesia and the Northern Transvaal (Soutpansberg

District) are characterised by having a shorter vexillum, varying from 3 -7-5-1 cm.

as against 4 -6-6 -2 cm. for those with a more southerly distribution. In all other

respects they appear to be identical with the more southern (typical) specimens and

therefore there are insufficient grounds for separating this northern form as a distinct

variety.

2. E. latissima E. Mey. and E. abyssinica Lam.

There was at one time a tendency to place E. latissima E. Mey. (type: Drege,

between Umgazana and Umzimvubu Rivers) as a synonym of the more tropical species,

E. abyssinica Lam. (= E. tomentosa R. Br. ex A. Rich.). Examples are to be found

in Wood, Natal PI. 4: t. 384, 385 (1906); Sim, For. FI. Cape Col. 201 (1907); and

Marloth, FI. S. Afr. II. 1 : 81 (1925). More recent authors, such as Burtt Davy (1932)

and Collett (1941), uphold the name E. latissima for the South African plant, but they

omit to give reasons for their decision.

An examination of available herbarium material indicated that corolla length

could serve as a basis for sorting the material into two groups, which differed to some

extent in leaf size and texture and also in geographical distribution. The two groups

meet in Southern Rhodesia but do not appear to grade into one another, thus providing

support for the view that they represent two distinct species. It has not been possible

to study this problem personally in the field, but useful confirmatory evidence has

been supplied by Dr. H. Wild of Salisbury Herbarium and Mr. D. C. H. Plowes of

Nyamandhlovu Pasture Research Station. The main distinguishing characteristics are

summarised in Table II. In addition, observers state that in E. abyssinica the calyx

is bright red, especially in the young stage. The colour is not evident in herbarium

specimens, so it is not known how constant a feature it is. In E. latissima, the calyx

has a rusty brown colour.

Table II.

Comparison of E. latissima and E. abyssinica.

511

On the basis of the above distinctions, it appears necessary to add E. gibbsae Bak.

(type: Gibbs 73, Matoppos) to the list of synonyms of E. latissima given by Collett,

l.c. Although the type has not been seen, no constant characters could be found for

separating specimens from the Matoppos from E. latissima .* A point arising out

of this investigation but falling outside its scope, is the validity of the name

E. abyssinica Lam. The original description is hardly sufficient to sustain it and

enquiries have failed to reveal the existence of a type specimen. The species is fully

described by A. Richard in his Tent. FI. Abyss. 1 : 213, 214 (1847), where it is identified

with a Quartin Dillon specimen, and it may be noted that E. tomentosa R. Br. ex A. Rich,

is described for the first time in this work also.

The distribution of E. abyssinica is from Abyssinia, through tropical Africa to

Southern Rhodesia and Mocambique, without entering the Union of South Africa,

while E. latissima extends from Southern Rhodesia and, probably, the highlands of

Mocambique, through eastern Transvaal and Natal to the Transkei.

* In a recent communication, Mr. Plowes informs me that, having compared flowering specimens

of E. gibbsae and E. latissima, he considers that the two may be distinct.

1

513

Plate 1. — Erythrina cajfra Thunb., photograph of type specimen at

Uppsala Herbarium.

514

Plate 2. — Erythrina caffra Thunb., East London (Codd 7979).

Plate 3. — Erythrina lysistemon Hutch., Rietvlei, Nelspruit District (Codd 8008).

515

The Schotia Species of Southern Africa.

By

L. E. Codd.

Schotia is essentially a genus of southern Africa and the species occurring in

Southern Rhodesia, the southern part of Mocambique, the Union of South Africa and

South West Africa form a closely related group. An examination of available material

led to the conclusion that only four species, namely S. afra (L.) Bodin, S. latifolia

Jacq., S. brachypetala Sond. and S. capitata Bolle, can be upheld in this region out of

some 15 specific names published.

The specimens seen include those of the National Herbarium, Pretoria, and of

the following Herbaria, whose assistance in sending their material on loan is gratefully

acknowledged: Bolus Herbarium; South African Museum Herbarium; National

Botanic Gardens, Kirstenbosch ; Albany Museum Herbarium, Grahamstown; and

Natal Herbarium, Durban. When citing specimens seen, the abbreviation indicating

the Herbarium is omitted where the gathering is represented in the National Herbarium,

Pretoria. Otherwise, only one Herbarium is cited, though this is not necessarily the

only Herbarium with representative material.

Among the specimens examined were eight which presented difficulty because

they showed floral characters intermediate between those of S. afra and S. latifolia.

Three resembled S', latifolia in general appearance but, unlike that species, possessed

stamens free to the base; the remaining five showed a close similarity to S. afra, but

the stamens were united to a variable extent at the base. These two groups were not

themselves very uniform and did not convey the impression of being worthy of specific

or varietal rank. They are treated here as hybrids and it may be of significance that

they occur in the area where both S. afra and S. latifolia are found. However, it is

desirable that they should be subjected to further study in the field and from a

cyto-genetical view-point in order that a reliable evaluation of their status can be made.

One of the published species, S. alata Thunb., proved, on examination of a fragment

of the type kindly sent to us by the Director of Uppsala Herbarium, not to be a Schotia,

but is identifiable as Erythrophysa alata (E. & Z.) Hutch.

The Schotia species of southern Africa have little economic value. The wood is

described as being hard and close grained with a pleasing finish when worked, but

stems suitable for timber purposes are seldom found. The leaves of S. afra are browsed

by livestock, while the young seeds, roasted or boiled, formed an article of diet of

primitive tribes and early settlers. It is said to be in reference to this property that

the popular name Boerboon (literally, “ farmer’s bean ”) came to be applied to the

species. S. latifolia has since been called the Bosboerboon (Forest Boerboon) and

5. brachypetala the Huilboerboon (Weeping Boerboon), the latter containing copious

quantities of nectar in the flowers. Schotias have attractive flowers in shades of pink

and red, but their slow growth reduces their value as horticultural subjects.

The generic description which follows is based on the species included in this

treatment. It has not been possible to form an opinion on the species described from

the more tropical parts of Africa owing to the scanty material of them in South African

herbaria. The few specimens seen give the impression of having a somewhat distant

relationship with the typical species. Baker (1926) suggested that the genus should be

limited to plants with an obconical or cylindrical receptacle and broad, imbricate

sepals which are not reflexed at flowering time. It is probable that closer study will

indicate that some tropical species at present in Schotia are, in fact, generically distinct.

516

SCHOTIA.

Jacq., Coll. 1: 93 (1787), nomen conservandum ; Schreber, Gen. Plantar. 279 (1789);

Lam., 111. 2: t. 331 (1797); Bodin in Thunb. Gen. Nov. PI. (as “ Scotia ”), 9: 130

(1798); Thunb., FI. Cap. ed. Schult. 388 (1823); DC., Prod. 2: 507 (1825);

Harv., Gen. S.A. PI. ed. 1: 92 (1838); Endl., Gen. 6785 (1839); Harv. in FL

Cap. 2: 273 (1861); Benth. & Hook., Gen. PI. 1: 581 (1865); Oliv. in FI. Trop.

Afr. 2: 309 (1871); Marloth, FI. S. Afr. II. 1 : 56 (1925); Baker, Leg. Trop.

Afr. 708 (1926); Phillips, Gen. S.A. FI. PI. ed. 2: 394 (1951).

Guajacum Linn., Sp. PI. ed. 1: 382 (1753), pro parte.

Theodora Medik., Theod. (1786); Harv., Gen. S.A. PI. ed. 1: 92 (1838); Taub. in

Engl. & Prantl, Pflanzenfam. III. 3 : 138 (1894); Pflanzenw. O.-Afr. C 198 (1895);

Harms in Yeg. der Erde, 9: 454 (1915).

Oniphalobium Jacq. f. ex DC., Prod. 2: 508 (1825).

Shrubs or trees, unarmed. Leaves paripinnate, alternate; stipules small, caducous;

leaflets coriaceous, often oblique, entire, rarely less than 3 pairs and not exceeding

18 pairs, the lower usually smaller than the upper. Inflorescence many-flowered in

short lateral or terminal panicles, in some species produced from the old wood; branches

of inflorescence stout, often woody. Bracts small, membranous, caducous. Pedicels

usually short. Calyx leathery, often red in colour; tube obconical, subcampanulate

or cylindrical, persisting in the fruiting stage; lobes 4 (rarely 5), often unequal,

imbricate, longer than the calyx tube, broad and not reflexed at flowering, deciduous.

Corolla pink or red in colour; petals 5 or, in one species, some or all petals may be

reduced to linear filaments, imbricate, inserted in the mouth of the calyx tube, early

deciduous. Stamens 10, inserted with the petals, free or connate below; filaments

linear, usually of two lengths, alternating, often coloured; anthers elliptic, attached

medially, versatile, 2-celled, dehiscing longitudinally. Ovary oblong, oblique, stipitate,

the stipe attached to one side of the calyx tube; ovules several to many; style terete;

stigma terminal, capitate. Pod woody, 1 to several seeded, oblong or broadly linear,

compressed, sometimes curved, beaked, sub-indehiscent, with a hard, peripheral margin

which persists, often with the seeds attached, after the eventual dehiscence of the valves;

seeds ovoid to globose, slightly compressed, light brown, with or without a yellow aril.

Type Species: S. afra (L.) Bodin.

The genus is named in honour of Richard van der Schot, Chief Gardener of the

Imperial Garden of Schonbrunn, and friend and travelling companion of Jacquin’s,

during the latter’s travels in America.

Key to the Species of Southern Africa.

Stamens free to the base

Leaflets usually exceeding 5 pairs and less than 1 cm. in width.

Leaflets 6-12 pairs, 3-10 mm. wide 1 (a). S. afra var.

afra.

Leaflets 11-18 pairs, 1 -5-3 mm. wide 1 (6). S. afra var.

angustifolia.

Leaflets 4 or 5 pairs, usually exceeding 1 cm. in width 5. Hybrid Form A.

Stamens connate at the base

Petals 5, normally developed

Inflorescence subglomerate on abbreviated lateral branchlets;

leaflets rarely exceeding 2-5 cm. in length

Staminal tube not forming a distinct sheath 6. Hybrid Form B.

Staminal tube forming a distinct sheath which projects

as a narrow rim 2. 5. capitata.

Inflorescence relatively lax, terminal; leaflets usually exceeding

2-5 cm. in length 3. S. latifolia.

Petals all, or some, reduced to linear filaments 4. S. brachypetala.

517

1. S. afra (L.) Bodin, aggregate species.

(a) var. afra.

(b) var. angustifolia.

Shrub or small tree up to 7 m. high, often with a twisted trunk; branchlets

glabrous to tomentulose. Leaves paripinnate, 2-5-8 cm. long and 2-3 cm. broad,

tending to fold along the rhachis, glabrous to tomentulose; stipules ovate, acuminate,

2 mm. long, caducous; petioles 2-4 mm. long; rhachis channelled above, rarely

narrowly winged; leaflets 6 (rarely less) to 18 pairs, usually opposite, sessile, linear or

oblong to elliptic, 5-25 mm. long and 2-10 mm. broad, rounded and often oblique

or gibbous at the base, apex obtuse, usually mucronate. Inflorescence a congested

panicle, subglobose, 4-8 cm. in diameter, borne on short lateral branchlets or, rarely,

terminal, composed of many flowers; inflorescence branches abbreviated, usually

hidden by the flowers, semi-woody, glabrous to tomentulose. Bracts small, scale-like,

caducous. Pedicels 3-9 mm. long. Calyx red, leathery; tube obconical to cylindrical,

3-8 mm. long and 3-5 mm. broad; lobes usually 4, sometimes one lobe emarginate,

rarely divided to the base producing 5 lobes, obovate to oblong, concave, 6-11 mm.

long and 5-10 mm. broad. Corolla red to pink; petals 5, oblanceolate, slightly clawed,

veined, 10-18 mm. long and 3-7 mm. broad. Stamens 10, free to the base; filaments

15-20 mm. long, exceeding the corolla by 3-8 mm.; anthers elliptic,' 2 mm. long.

Ovary oblong, compressed, 5-7 mm. long and 2 mm. broad, oblique, stipitate; stipe

5-6 mm. long; style 8-14 mm. long; stigma capitate. Pod as in generic description,

5-5-12 cm. long and 3-3-5 cm. broad. Seeds ovoid to roundish, slightly compressed,

12-15 mm. long and 8-12 mm. broad, pale brown, aril very small or absent.

Thunberg published the binomial Schotia afra without supporting description or

citations in his Prodromus (1794), but this nomen nudum does not invalidate the later

effective publication of the combination by his student, Bodin, in Thunberg’s Gen.

Nov. PI. 9: 130 (1798). Some authorities credit Thunberg with the combination,

e.g. S. afra (L.) Thunb. in Juel’s Plantae Thunbergiana (1918). However, Bodin is

clearly cited as the author of the work on Schotia in Thunberg’s Gen. Nov. PI. and

so it seems reasonable to accept him as the author of the combination. It may be

noted that Thunberg, in his Flora Capensis (1823), reverts to the epithet S. speciosa.

According to Acocks (Veld Types of South Africa, 1953), S. afra is a common

species of the Karroid Valley Bushveld and the allied Fish River and Sundays River

Scrub of the eastern Cape Province. It also occurs in the Noorsveld, the Succulent

Mountain Scrub or Spekboomveld, the Little Karoo and, further north, in the Orange

River and Namaqualand Broken Veld. In a personal communication he expresses

the opinion that the variation seen in plants occurring in these various veld types could

well be included within the limits of one species.

Harvey in Flora Capensis upheld three varieties in addition to the typical form,

namely: var. angustifolia, var. tamarindifolia and var. ovalifolia. In the present

treatment, only one variety is separated as distinct from typical S. afra, namely

S. angustifolia, with leaflets more numerous and narrower than the typical form. Var.

tamarindifolia is recognisable in flowering specimens by the calyx tube being cylindrical

instead of obconical but otherwise it appears to be identical with typical S. afra and

is not considered to be varietally distinct. Var. ovalifolia shows characters intermediate

between S. afra and S. latifolia and evidence points to its being a hybrid between these

two species. It is discussed more fully under the heading Hybrid Form A (p. 529).

Var. (a). S. afra ( L .) Bodin var. afra. S. afra (L.) Bodin in Thunb. Gen. Nov. PI. 9 : 1 30

(1798) (as “Scotia”); Thunb., Prodr. PI. Cap. 79 (1794), nomen nudum.

Guajacum afrum L., Sp. PI. ed. 1: 382 (1753); Mill., Gard. Diet. ed. 8: (1768).

518

5. speciosa Jacq., Coll. 1: 93 (1787); Icon. PI. Rar. 1: 8, t. 75 (date ?); Schreber,

Gen. Plantar. 279 (1789); Willd., Sp. PI. 2; 537 (1799); Andr., Bot. Rep. 5:

t.348 (1804); Thunb., FI. Cap. ed. Schult. 388 (1823); DC., Prod. 2: 508 (1825);

E. & Z., Enum. PI. Afr. 261 (1836); Harv. in FI. Cap. 2: 274 (1861); Sim, For.

FI. Cape Col. 207, t.57 (1906); Marloth, FI. S.Afr. II. 1: 57 (1925). Var.

tamarindifolia (Afz. ex Sims) Harv., l.c.

Theodora speciosa Medik., Theod. 16 (1786); Harv., Gen. S.A. PI. ed. 1: 92 (1838);

Taub. in Engl. & Prantl, Pflanzenfam. III. 3: 138 (1894).

5. tamarindifolia Afz. ex Sims in Bot. Mag. 1. 1 153 (1809); DC., Prod. 2: 508 (1825);

E. Mey., Comm. PI. 161 (1836).

Citations prior to 1753. — Boerh., Lugd. 2: 57 (1727); Walth., Hort. Leipzig,

2: t.2 (1735); Linn., Hort. Cliff. 489 (1737); Van Roy., Hort. Lugd. 536 (1740).

Figs. 1 and 2.

Type. — No. 532-4 in the Linnaean Herbarium. Of the two specimens in the

Linnaean Herbarium, both sterile, numbered 532-3 and 532-4, respectively, the latter,

being more characteristic of the species, is chosen as the lectotype. Types of synonyms:

S. speciosa Jacq., the plate in his Collecteana; S. tamarindifolia Afz. ex Sims, Masson,

“ Africa australis prope bon. spei ”, in the British Museum; Theodora speciosa Medik.,

no information is available as to whether a corresponding specimen exists.

Cape. — Albany: 16 miles E. of Grahamstown, Codd 9237; Fraser’s Camp,

Maguire 633 (NBG). Bathurst: Port Alfred, Rogers 16637; Tyson s.n.; White

57 (GRA); Schonland s.n. (GRA); Kowie, Salisbury 121 (GRA). Alexandria:

Debega Valley, Galpin 10661; Bushmans River Poort, Johnson 863. Uitenhage:

Ecklon and Zeyher 47 (BOL); Theron 232; Swartkops River Valley, Zeyher 2447;

Aloes, I.L. Drege 3126; Addo, Brynard 202; Gill 10 (BOL); Lang in Hb. Marloth

6840; Despatch, Holland, 337 (GRA). Port Elizabeth; Redhouse, Marloth 13040;

Patterson 524 (BOL); Enon, Thode A2643; Perserverance, Rodin 1282 (BOL);

Coegakamma to Addo, I. L. Drege 584 (GRA). Humansdorp: Phillips 3350; Slang

River, Fourcade 1795 (BOL). Knysna: Keurbooms River, Taylor 29. Oudtshoorn:

Between Calitzdorp and Oudtshoorn, Barker 628 (NBG). Mossel Bay : McGaffin s.n.

(GRA). Laingsburg: Seven Weeks Poort, de Jager s.n. (BOL). Ladismith: Pole

Evans H19101; Swartberg, Marloth 2945. Riversdale: Still Bay, Muir 92; Taylor

92; Skilpadgat, 5 miles N.W. of Albertina, Taylor 209 (NBG). Swellendam: Eierpoort,

Compton 11920 (NBG); Anysberg, Stokoe 8390. Montagu: Marloth 8129; between

Dobbelaarskloof and Karreevlakte, Hall 881 (NBG). Robertson: Along Komas

River, Acocks 8618.

Jacquin’s plates of S. speciosa are among those whose dates of publication are

not known with certainty. The subject is discussed by Schubert in Contrib. Gray Herb.

154: 3-23 (1945), who tentatively places the date of plate 93 of the Collecteana

(accepted as the earlier) as 1787. The date of publication of Jacquin’s leones Plantarum

Rariorum, Vol. I, is frequently accepted as 1781, the date of the title page to that work

but, on page 8, in the text relating to plate 75, he refers to the plate in his Collecteana,

and to the publication of Theodora speciosa by Medikus. Therefore plate 75 of the

Tcones is undoubtedly the later of the two.

Fig. 1. — Schotia afra var. afra, a, leaf; b, flower; c, longitudinal section; d, petal;

e, ovary. Fig. 2. — S. afra var. afra, form with cylindrical calyx tube, a, leaf;

b, flower; c, petal; d, longitudinal section. Fig. 3.— S. capitata, a, flower;

b, petal; c, longitudinal section. Fig. 4. — S. afra var. angustifolia, a, leaf;

b, flower; c, petal; d, longitudinal section. Fig. 5. — S. latifolia, a, flower;

b, petal; c, longitudinal section; d, ovary. Fig. 6. — S. brachypetala, a, flower •

b, flower with calyx lobes removed, showing reduction of all petals; c, form

with petals partially reduced. All life size.

(bo-.

520

Included in var. afra are the plants with, normally, 6-11 pairs of leaflets, oblong

to elliptical, 10-17 mm. long and 4-10 mm. wide. These plants may again be subdivided

into two groups on the basis of calyx shape, one group with the calyx tube obconical

in shape while, in the other group, the calyx tube is cylindrical, though the distinction

is not always clear cut. The form with the obconical calyx tube appears to have been

in cultivation in Europe from early in the 18th Century and is the form figured by

Jacquin, so it can be fairly safely assumed that the specimens in the Linnaean Herbarium

belong here. This form is centred mainly in the Little Karoo, occurring in the Montagu,

Swellendam, Ladismith, Laingsburg and Oudtshoorn Districts, with outliers as far

east as the Knysna District.

The second group with the calyx tube cylindrical in shape is indistinguishable

vegetatively from the preceding, having much the same number, size and shape of

leaflets. Thus no advantage can be seen in separating it as a distinct variety. It was

described as S. tamarindifolia Afz. ex Sims and was placed as S. speciosa var.

tamar indifolia by Harvey in Flora Capensis. Plants of this nature are distributed

mainly to the east of the preceding group in the districts of Humansdorp, Uitenhage,

Alexandria, Bathurst and Albany, but they are also recorded further west in the

Riversdale District.

S. tamarindifolia is described as having five calyx lobes. This is unusual but not

unknown. The usual number is four but, occasionally, the largest lobe is notched or

split at the apex. In extreme cases, which are very rare and apparently not constant

even in one inflorescence, the division continues to the base of the lobe, with the result

that five lobes instead of four are present. Of the 41 flowering specimens examined,

10 possessed flowers with one calyx lobe notched for a quarter or more of its length,

while in two of these the division continued to the base of the lobe, resulting in five

instead of four calyx lobes.

Var. ( b ). S. afra (L.) Bodin var. angustifolia ( E . Mey.) Harv., in FI. Cap. 2: 274 (1861).

S. parvifolia Jacq., Fragm. 85, t. 136 (1809).

S. angustifolia E. Mey., Comm. PI. Austr. Afr. 161 (1836).

S. venusta Mason in Journ. Roy. Hort. Soc. Lond. 39: t. 14 (1913), nomen nudum.

Figs. 4 and 7.

Fig. 7. — S. afra var. angustifolia, pod, life size.

521

Type. — Drege, between Kunap and Kat Rivers. Of the two specimens cited by

E. Meyer, this appears to be the more representative of the variety and is therefore

chosen as the lectotype. An isotype is at the British Museum Herbarium.

Cape. — Butter worth: Kei River, Bayer in N.H. 31017 (NH); Kei Bridge, Codd

9242. Komgha: Flanagan 1322. Queenstown: Saunders in N.H. 10602 (NH);

Gwatyn, Galpin 8108; near Balotwa, Thorns s.n. (NBG). King William’s Town:

Sim 2139; north side of Mitchell’s Pass, Esterhuysen 13235 (BOL). Keiskammahoek :

Boma Pass, Acocks 91 1 1. Middledrift: Salisbury 290 (GRA). Victoria East: Lewis

in SAM 66637; 4 miles above Fort Cox, Gillett 4592; Curries Kloof, Bennie 441.

Graaff Reinet: Bolus 621 (BOL). Prieska: 20 miles N.W. of Prieska, Bryant in

Hb. Bol. 25613 (BOL). Kenhardt: 7 miles W. of Pofadder, Thorne s.n. (SAM).

Namaqualand: Krapohl in Hb. Bol. 20589 (BOL); Groot Rozynbosch, Pearson

3824 (SAM). Gordonia: Bokputs, Marsh in Hb. Marl. 6527; Aughrabies Falls,

Esterhuysen (BOL).

South West Africa. — Warmbad: Skunsberg Quelle, south of Warmbad, Dinter

5133; Eendoorn, Galpin s.n.

Var. angustifolia is distinguished from the typical by the more numerous, smaller

and narrower leaflets, possessing 12-18 pairs of leaflets which are 5-12 mm. long and

1-5-3 mm. wide. The calyx tube is obconical, resembling the typical form, but the

pedicels are slightly shorter. S. parvifolia Jacq. appears to belong here, though Jacquin’s

figure shows too few leaflets. The distribution of this variety is further east and more

inland than the typical, occurring in the Districts of Butterworth, Komgha, King

William’s Town, Victoria East, Queenstown, Graaff Reinet and, further north, in

Prieska, Gordonia, Kenhardt, Namaqualand and South West Africa, to mention the

most important.

2. S. capitata Bolle in Peters Reise Mossamb. Bot. 18 (1862); Oliv. in FI. Trop. Afr.

2: 310 (1871); Sim, For. FI. Port. E. Afr. 51 (1909); Baker, Leg. Trop. Afr.

710 (1926).

5. tamarindifolia Afz. ex Sims var . forbesiana Baill. in Adansonia, 6: 197 (1865-66).

Theodora capitata (Bolle) Taub. in Engl. & Prantl Pfianzenfam. III. 3: 108 (1894);

Pflanzenw. O.-Afr. C 198 (1895).

S. transvaalensis Rolfe in Kew Bull. 248 (1906); Burtt Davy, FI. PI. Ferns Tvl. Swaz.

2: 326 (1932); Phillips in FI. PI. S. Afr. 15: t. 574 (1935) excl. descr. “ up to

30 ft. high”; Codd, Trees and Shrubs, Krug. Nat. Park, 68, t.63 (1951).

Fig. 3.

Shrub or slender tree up to 7 m. high, sometimes sub-scandent; branchlets

glabrous to sparingly pubescent; bark grayish. Leaves paripinnate, 3-5-8 cm. long

and 2 -2-4 -5 cm. broad, tending to fold along the rhachis, glabrous to sparingly

pubescent; stipules 2-3 mm. long, ovate, acuminate, caducous; petioles 1-6 mm.

long; rhachis narrowly winged, especially towards the apex; leaflets 3-5, rarely

6 pairs, opposite, sessile, elliptic or sub-rotund or obovate, 1 • 5-2 • 5 (rarely up

to 3-5) cm. long and 1-1-8 (rarely 2) cm. broad, obtuse to acute at the base, often

oblique, obtuse or acute at the apex, usually mucronate. Inflorescence a congested

panicle, subglobose, 3-8 cm. in diameter, borne terminally or, more usually, on short

lateral branchlets; inflorescence branches much abbreviated, semi-woody, glabrous

to tomentulose. Bracts less than 1 mm. long, scale-like, caducous. Flowers scarlet,

sessile or subsessile with pedicels rarely exceeding 1-5 mm. long. Calyx deep red,

leathery; tube obconical, 3-4 mm. long, persistent; lobes 4, subequal, obovate to

elliptic, concave, 6-9 mm. long and 3 mm. broad. Corolla scarlet; petals 5,

oblanceolate, clawed, veined, 10-14 mm. long and 3-4 mm. broad. Stamens 10 united

at the base for 3-4-5 mm. forming a sheath round the ovary stipe, the sheath being

129828-4

522

split open on the side where the stipe is attached to the calyx tube, and projecting as

a narrow, irregular rim; one or two stamens often free to the base on the split side;

filaments linear, 12-16 mm. long, exceeding the corolla by 3-5 mm.; anthers elliptic,

1 ■ 5-2 mm. long. Ovary oblong, compressed, 4-5 mm. long and 2 mm. broad, stipitate;

stipe 4-5 mm. long and 1 mm. thick, adnate to one side of the calyx tube; style terete,

13-15 mm. long; stigma capitate. Pod as in the generic description, 4—12 cm. long

and 2 -5-3 -5 cm. broad. Seeds ovoid, slightly flattened, 8-12 mm. long, pale brown,

with a large, compressed, yellow aril.

Type. — Peters, Inhambane. The holotype in Berlin-Dahlem Herbarium appears

to have been destroyed and enquiries have thus far failed to reveal the existence of

an isotype. Types of synonyms: S. transvaalensis Rolfe, P. Oranje s.n., Barberton

District, in Herb. Kew; S. tamarindifolia var. forbesiana Baill., Forbes 32, Delagoa

Bay.

Mocambique. — Sul do Save: Mutamba, near Inhambane, Gomes e Sousa 2171;

Moambo, Pedrogao 25, 26, 218; between Inharrime and Panda, Pedro and Pedrogao

1 887 ; between Matola and Umbeluzi, Pedro and Pedrogao 897, 901 ; between Unianga

and Mazivila, Pedro and Pedrogao 1441 ; between Chamusca and Nalasi, Pedro and

Pedrogao 1271, 1294; Canissado, Pedro and Pedrogao 1297; between Ginja and

Chibuto, Pedro and Pedrogao 1536; Ginja, Pedro and Pedrogao 2133; Lourenco

Marques, Borle 18, 88, 597; Torre 6370; Bremekamp in T.M. 30198; Bela Vista,

Swierstra in T.M. 16390; near Goba, Torre 6470; Maziminhama, Myre 1247;

Maputo, Hornby 2537.

Transvaal. — Pilgrims Rest: 1\ miles N.E. of Skukuza, Codd and de Winter 5063.

Nelspruit: Skukuza, Stevenson-Hamilton s.n.; 21 miles N.E. of Pretorius Kop, Story

3934; Klokwene, van der Schyff 196; Sigaas, van der Schyff 774. Barberton:

Komatipoort, Rogers 12609; Kaapmuiden, Thorncroft 611.

Swaziland.- — Flats below Stegi, Rodin 4559; Sitilo River, near Gollel, Pole

Evans 3395 (4).

Natal.- — Ingwavuma: Ndumu, Gerstner 4004 (NH); Pongola River, Boocock 47.

Ubombo: Gerstner in NH 22865; between Otobotini and Ubombo, McClean in

NH 18472. Ngotshe: Candover, Galpin 13608. Hlabisa: False Bay, Gerstner 4761;

Ward 697, 690. Mahlabatini: 6\ miles N. of Mahlabatini, Codd 1932; near

Mahlabatini, Acocks 11666; Black Umfolosi, Swynnerton 47. Lower Umfolosi:

Umfolosi River, Wager in T.M. 22386; Umfolosi Game Reserve, Ward 1468 (NH);

Nagana Research Station, Kluge 18 (NH); Umhlatuzi Valley, Lawn 795, 1744 (NH).

The holotype of S. capitata was apparently lost with the partial destruction of

Berlin-Dahlem Herbarium and our liaison officer at Kew has been unable to locate

an isotype in any of the British Herbaria. Thus, in the absence of a type specimen,

circumstantial evidence had, to some extent, to be relied upon in deciding to place

S. transvaalensis as a synonym of S. capitata. All the Schotia specimens seen from

Mocambique can be placed in two species, namely, S. brachypetala and the species

here outlined. Gatherings made in the neighbourhood of Inhambane, the type locality

of S. capitata, fail to reveal a third species. The description of S. capitata by Bolle

is based on a flowering specimen without leaves; it could apply to our present

conception of the species and, fortunately, the description of the five petals excludes

S. brachypetala. The name S. capitata was taken up by Oliver in the Flora of Tropical

Africa and subsequent authors, for example, Taubert (1894), Sim (1909) and Baker

(1926), usually with the opinion expressed that 5. tamarindifolia var. forbesiana is

probably a synonym. There is a specimen at Kew of Forbes 32, type of the latter

variety, and it has been reported as being conspecific with Oranje s.n., type of

S. transvaalensis.

523

S. capitata shows a superficial resemblance to some of the specimens cited under

the heading “ S. latifolia X S. afra, Form B ” but is distinguished from this group and,

in fact, any other species, by the characteristic and relatively specialised staminal sheath

which encloses the stipe of the ovary. The sheath is composed of, usually, 8 stamens

fused at the base for 3-4-5 mm., being open to the base on the side to which the stipe

is attached to the calyx tube. The sheath is smooth on the inside and is projected as

an uneven rim about 1 mm. above the junction of the stamens. One or two stamens

are usually free to the base on the open side of the sheath. The leaf rhachis is usually

more distinctly winged in 5. capitata than in any of the other species.

In the absence of flowers, it is often difficult in the herbarium to separate

S. brachypetala from some forms of S. capitata because there is overlapping in the

size of leaflets. In the field, they can be distinguished by the difference in habit,

S. brachypetala forming a round-topped tree from 25 to 40 feet or more in height,

with a trunk up to 18 inches in diameter. S. capitata is normally a shrub with several

slender, ascending branches, frequently intertwined in thorn scrub, with which it is

frequently associated. More rarely it forms one to several slender, erect stems up

to 20 feet in height and 4 to 6 inches in diameter.

3. S. latifolia Jacq., Fragm. 23, 1. 1 5, f.4 (1809); DC., Prod. 2; 508 (1825); Jacq.

f., Eel. PI. 2: 1. 1 26 (1844), in textu; E. Mey., Comm. PI. 162 (1836); E. & Z.,

Enum. PI. Afr. 262 (1836); Harv., Gen. S.A. PI. ed. 1: 92 (1838); Harv. in FI.

Cap. 2: 274 (1861); Sim, For. FI. Cape Col. 206, t.57 (1906); Marloth, FI.

S. Afr. II. 1: 57 (1925).

Omphalobium Schotia Jacq. f. ex DC., Prod. 2; 508 (1825), in synonomy.

S. diversifolia Walp. in Linnaea, 13; 541 (1839).

S. cuneifolia Gandoger in Bull. Soc. Bot. Fr. 60: 462 (1913).

Figs. 5 and 8.

Tree up to 10 m. high; trunk up to 25 cm. in diameter; branchlets glabrous to

tomentulose; bark smooth, dull grey to reddish brown. Leaves paripinnate, 6-10 cm.

long and 5-8 cm. broad, glabrous or pubescent; stipules ovate, acuminate, 4 mm.

long, caducous; petioles 0-5-2 cm. long; rhachis slightly channelled above, winged

in the juvenile state; leaflets 3-5 pairs, usually opposite, sessile, elliptic-oblong to

obovate, 2-5 (rarely less)-6 cm. long and 1-4-3 cm. broad, cuneate to rounded at the

base, oblique, rounded of acute at the apex, rarely mucronate, the upper leaflets the

largest. Inflorescence a relatively open panicle, terminal to main or lateral branches,

8-15 cm. long and 6-10 cm. wide; inflorescence branches patent, semi-woody, 3-8 cm.

long, glabrous to tomentulose. Bracts scale-like, caducous. Pedicels less than 2 mm.

long. Calyx reddish brown, leathery; tube very short, turbinate, 1-5-2 mm. long,

persistent; lobes 4, subequal, obovate, concave, 5-7 mm. long and 4-5 mm. broad.

Corolla pink to flesh colour; petals 5, equal or, occasionally, 1 slightly reduced,

oblanceolate, veined, 9-11 mm. long and 2-5-3 mm. broad. Stamens 10, connate at

the base for 0-5-3 mm., the staminal sheath split open on the side to which the ovary

stipe is attached to the calyx tube; filaments linear, 10-14 mm. long, exceeding the

corolla by 2-3 mm. ; anthers elliptic, 1 • 5-2 mm. long. Ovary oblong, compressed,

4-5 mm. long and 2 mm. broad, stipitate; stipe 2 mm. long, adnate to one side of

the calyx tube; style terete, 10-11 mm. long; stigma scarcely differentiated. Pod

as in the generic description, 5-14 cm. long and 3-4-5 cm. broad. Seeds ovoid,

flattened, 11-14 mm. long, pale brown with a large, compressed, yellow aril.

Type. — The plate in Jacquin’s Fragmenta. Types of synonyms: S. diversifolia

Walp., syntypes, Ecklon and Zeyher from “ Uitenhage and Albany ” and three Drege

specimens collected at Glenfilling, Keiskamma and Zandplaat, respectively. S.

cuneifolia Gand., Penther 2516 (or 2561, see later discussion), Cape of Good Hope.

524

Fig. 8. — 5. latifolia, leaf and pod, life size.

525

Cape. — Engcobo: 2 miles N. of Engcobo, Codd 7981a. Kentani: Pegler 593.

Komgha: Flanagan 712; Queenstown: Junction Farm, Galpin 8109. Stockenstroom:

Brambledene, between Alice and Seymour, Barker 2898 (NBG). Stutterheim: Dohne,

Acocks 8955, 9501. East London: Penther 2561 (W); Galpin 3141; Nanni 152;

Blind River Valley, C. A. Smith 3841 ; mouth of Keiskamma River, Galpin 7695.

Peddie: Ebb and Flow, Maguire 619 (NBG). Albany: Mario th 10877; Howieson’s

Poort, MacOwan 77 (BOL); Gowieskloof, Bennie 669 (GRA); 1 £ miles S.W. of

Grahamstown, Comins 983; 15 miles N. of Grahamstown, Maguire 661 (NBG).

Bathurst: Fish River Valley, Schlechter 6104; Port Alfred, Rogers 959; Potts 169

(GRA); Kowie West, Burn Davy 7887. Uitenhage: Gill 16 (BOL); Patterson

2209 (GRA); “Uitenhage and Albany”, Ecklon and Zeyher 1701; Paarde Poort,

Bolus 1547 (BOL); Kamachs, Long 1277; Enon, Thode 22644; Zuurberg, Holland

300 (GRA). Port Elizabeth: Kensit 7 (BOL); between Coega and Sundays River,

Zeyher 672 (BOL) in part (?); Van Staadensberg Gorge, MacOwan 2458 (GRA),

Baakens River, I. L. Drege 659; Van Staadens Kloof, Taylor 832; Bethelsdorp;

Paterson in T.M. 12299, 25844; Swartkops River, Springfield, Forest Dept. 150 (GRA).

Humansdorp: Loerie, Dix 208. Uniondale: 9 miles from Joubertina. Fourcade

5079 (BOL). Knysna: Keurbooms River, Fourcade 169 (BOL); Compton 4457

(NBG); Keurbooms River Pass, Fourcade 1992 (BOL). George: Pappe s.n. (GRA).

Cultivated. — Kirstenbosch, Henderson 1891 (NBG); Botanic Garden, Durban,

Medley Wood 1891, 5017, 8689 (NH); near Durban, Medley Wood 9260 (NH);

Barberton (?), Thorncroft ex hort NBG; Cape Town, Marloth 9315.

Jacquin’s plate lacks flowers and was made from an immature plant stated to be

five years old, grown from seed sent by Georg Scholl, who collected in the Cape Province

for the Imperial Gardens at Schonbrunn from 1785 to 1797. As far as is known, his

travels did not take him further eastward than East London, which would not bring

him into the area of distribution of S. brachypetala, a species indistinguishable

vegetatively from S. latifolia. Fortunately this is not the only evidence we have relating

to the identity of this plate. De Candolle, Prod. 2: 508 (1825), gives a description

of the flowers of 5. latifolia, based on an unpublished plate (published later in Eel.

PI. 2: 1. 1 26, 1844) by Jacquin fil., who had given it a manuscript name of Omphalobium

Schotia. It may be presumed that this is the same plant as that figured by the elder

Jacquin, and subsequent authors have accepted its identity. It is not known if herbarium

specimens corresponding to the plates are in existence.

S. diver sifolia Walp. is based on Ecklon and Zeyher 1701 (represented in

Grahamstown, South African Museum and the National Herbarium, Pretoria) and

three Drege specimens collected respectively at (a) Glenfilling, Albany District; ( b )

Keiskamma; and (c) between Zandplaat and Komga. The Drege specimens have

not been seen, but all the localities fall within the distribution range of S. latifolia,

and the description by E. Meyer in his Comm. PI. 162 (1836) agrees with this species.

The following is a copy of the description of S. cuneifolia Gandoger: “ Foliola

ampla oblonga, apice subrotundata, mucronata nec retusa, basi cuneata, marginibus

undulatis, rami floriferi glaberrimi, patentes, late angulosi, cymae laxiflorae. HAB.:

Cap. (Penther n. 2516). A S. latifolia Jacq. recedit foliolorum coriaceorum forma

staminibusque longioribus. Legumen maximum, 9-10 cm. longum, 3 cm. latum,

glabrum.” In response to an enquiry made to the Naturhistorisches Museum, Vienna

Dr. Rechinger informs me that he was unable to find a Schotia specimen with the

Penther number 2516. There is, however, a sheet of S. latifolia labelled Penther 2561

from “ Distr. East London ”. A portion of this specimen was kindly sent to us on

loan and was confirmed as being S. latifolia Jacq. Whether Gandoger’s No. 2516

is a typographical error is not certain, but it seems possible that the type may in fact

be No. 2561. In any case, the description gives no adequate grounds, in the light of

modem information on Schotia species, for separating this species from S. latifolia.

526

Sim in his Forest Flora of the Cape Colony, page 206, states that “ it grows

to a tree 30 to 50 feet and is common throughout the Eastern and Transkei Conservancies

and Natal.” Acocks (Veld Types of South Africa, 1953), in addition to recording the

presence of 5. latifolia in the Fish River Scrub, Southern Valley Bushveld and

Alexandria Forest, cites it as occurring in the Zululand Thornveld (pp. 39, 40). In

the absence of flowers, it is practically impossible to distinguish S. latifolia from

S. brachypetala and thus the two might be confused in the field. S. brachypetala is

known to occur as far south as Umtata, but no herbarium specimens identifiable as

S. latifolia are known from the wild to the east of Engcobo and Kentani Districts in

the Transkei. The Medley Wood specimens from Durban and the Thorncroft specimen

from Barberton are considered to be from cultivated trees. Thus there is need for

more collecting to be done in the Transkei and Pondoland areas in order to demonstrate

clearly whether the two species overlap or not.

S. latifolia grows in forest margins and dry scrub and forms a small, round-topped .

tree 15 to 30 feet in height (Marloth’s reference to a 50 foot tree in the Pirie Forest

must be regarded as exceptional). It is probably evergreen or, like some other species

of Schotia, the leaves may be shed for a very short period before the flowers appear.

The flowers are relatively small, pinkish in colour and are borne in profusion in terminal

panicles.

4. S. brachypetala Sond. in Linnaea, 23: 39 (1850); Harv., Thes. 21. t.32 (1859);

FI. Cap. 2: 274 (1861); Wood, Natal PI. 4: t.390 (1906); Marloth, FI. S.Afr. II.

1 : 57 (1925); Baker, Leg. Trop. Afr. 709 (1926); Burtt Davy, FI. PI. Ferns Tvl.

Swaz. 2: 326 (1932); Dyer in FI. PI. S.Afr. 20: t. 777 (1940); Codd, Trees

and Shrubs Krug. Nat. Park, 66, t.63 (1951). Var. pubescens Burtt Davy, l.c.,

xxx, 326 (1932).

5. rogersii Burtt Davy, l.c., xxx, 326 (1932).

S. semireducta Merxm. in Mitt. Bot. Munchen, 6: 199 (1953).

Fig. 6.

Tree up to 16 m. high; trunk up to 45 cm. in diameter; branchlets glabrous to

tomentulose; bark rough, reddish brown. Leaves paripinnate, 6-16 cm. long and

5-8 cm. broad, glabrous or pubescent; stipules ovate, acuminate, 4-5 mm. long,

caducous; petiole 5-25 mm. long; rhachis slightly channelled, often narrowly winged,

especially towards the apex; leaflets 4-7 pairs, opposite or subopposite, sessile or

with petiolules up to 2 mm. long, elliptic, oblong, ovate-oblong, ovate or obovate,

2 -5-8 -5 cm. long and 1-2-4 cm. broad, obtuse or rounded at the base, oblique, often

with a gibbosity near the base, rounded at the apex, rarely mucronate, the upper leaflets

the largest. Inflorescence a congested panicle, subglobose, 6-13 cm. in diameter,

usually cauliflorous on older branches, occasionally borne terminally; inflorescence

branches woody, abbreviated, glabrescent or tomentulose. Bracts ovate, 6 mm. long

and 4 mm. wide, early caducous. Flowers deep red, pedicellate; pedicels 5-11 mm.

long, glabrous or pubescent. Calyx deep red, leathery; tube obconical to

subcampanulate, glabrous or sparingly pubescent, 3-5 mm. long; lobes 4, subequal,

ovate to elliptic, 8-12 mm. long and 4-6 mm. broad, concave, the largest often

emarginate. Corolla reduced to 5 linear filaments 2-6 mm. long, or 1-4 petals

developing per flower, in which case petals red, spatulate to oblanceolate, clawed,

1 • 3-1 • 8 cm. long and 3-6 mm. broad. Stamens 10, connate at the base for 1 • 5-3 ■ 5 mm.,

forming a staminal tube which is entire, or split at one side; filaments linear, red,

1-8-2 cm. long, exceeding the calyx by 8—10 mm.; anthers elliptic, 2-2-5 mm., long.

Ovary oblong, compressed, 4—6 mm. long and 2 mm. broad, tuberculate along the

margins, stipitate; stipe 4 mm. long, straight, adnate to one side of the calyx tube;

style terete, 9-11 mm. long; stigma capitate. Pod as in generic description, 5-11 cm.

long and 3-5-4 cm. broad. Seeds ovoid or oblique, flattened, 10-14 mm. long, with a

large, compressed, yellow aril.

527

Type. — Gueinzius 33, Port Natal, at the Stockholm Natural History Museum

Herbarium. Types of synonyms: var. pubescens Burtt Davy, Junod 635, Shilouvane,

Transvaal; S. rogersii Burtt Davy, Rogers 21246, Waterpoort, Soutpansberg District;

S. semireducta Merxm., syntypes, Dehn 42, Rusape, and Wild in Salisb. Hb. 40239,

Rusape, S. Rhodesia.

Southern Rhodesia. — Salisbury: Brain 9503; Evles 844; Twentydales, Wild

4122. Umtali: Chase 1602.

Mocambique. — Maputo: Gomes e Sousa 3854; near Salamanga, Hornby and

Pedro 19.

Transvaal. — Soutpansberg: Waterpoort, Rogers 21246 (isotype of S. rogersii,

PRE); Smuts and Gillett 3149; Pole Evans HI 8282; Smitsdrift, Burtt Davy 1697;

Wylliespoort, von Wolff in T.M. 34861; Codd 4443, 4444; Erens s.n. ; Repton 689;

28 miles west of Louis Trichardt, Codd 4442; Soutpan, Galpin 15143; Dongola area,

Codd 4332a, 4453; Punda Maria area, Rowland Jones 23. Pietersburg: Blauwberg,

Leipzig Mission, Leipoldt 2; van Dam in T.M. 23998; Tscheuschner in T.M. 29521;

Munnik, van Rensburg s.n.; near Boyne, Gerstner 5376; Chuniespoort, Obermeyer

in T.M. 34676, 34677; Pole Evans H19450; Repton 451. Letaba: Shilouvane, Junod

635 (isotype of var. pubescens, PRE); Tzaneen, Pole Evans HI 58 19; Altenroxel in

Hb. Burtt Davy H3065; Charter in Hb. Burtt Davy H3182; between Munnik and

Duivelskloof, Mogg in Hb. Burtt Davy H 10684; near Leydsdorp, Dist. Forest Off.

LD5; Codd 1672. Waterberg: 25 miles N.W. of Vaalwater, Smuts 351; near

Hermanusdoorns, Codd 1005, 4426. Potgietersrus : Leendertz 1249; Pruizen, Burtt

Davy 2218; Pyramid Estate, Galpin 9135. Lydenburg: Excelsior, Keet 1110;

Sekukuniland, Schoonoord, Barnard 115; van Warmelo 12; Lulu Mts., Mogg 16975;

10 miles S.W. of Penge Mine, Codd 6695. Pilgrim’s Rest: Bushbuckridge, Pritchard 43.

Nelspruit: Kruger National Park, near Skukuza, Letty 96; Codd 4386; 18 miles N.

of Malelane, Codd 4375; Numbi, van der Schyff 63; Crocodile Bridge, van der Schyff

3893. Barberton: Thorncroft 3010; Hyslops Creek, Thorncroft 727 (NH);

Lampogwane River, Galpin 653; Malelane, Pole Evans H15767; Louws Creek,

Wager in T.M. 22442; Kaapmuiden, Prosser-, Greenheart, Oranje in Hb. Burtt Davy

1124.

Swaziland. — Near Mbabane, Rodin 4534; between Bremersdorp and Stegi,

Miller S/69; Lebombos, Hornby 2832.

Natal. — Zululand: Watt and Brandwyk 1004; Curson s.n. Ubombo: near

Mkuzi, Galpin 13342; s.n. (BOL); Utrecht: Thode A1287. Estcourt: West 324.

Msinga: Fabene Halt, C. A. Smith s.n. Hlabisa: Hluhluwe Game Reserve, Ward

1499 (NH). Lower Umfolosi: Umhlatuzi, Lawn 508 (NH). Mapumulo: Darnall,

Schmidt 55, NH37707 (NH). Durban: Gueinzius in SAM 15632; Medley Wood

9943; MacOwan in Hb. Bolus 1106; Rogers 1457; Inanda, Medley Wood 1367.

Umzinto: Dumisa, Rudatis 686 ; Umkomaas River, Allan in Hb. Norm. 70; Lansdell

in NH 37519. Port Shepstone: Lower Umzimkulu, Medley Wood 9575 (NH). No

locality: Flanagan s.n.; Sanderson s.n.

Cape. — Tabankulu: near Welsh Bridge, Story 4211; Acocks 13844. Umtata:

Commonage, Miller B/95. Ngqeleni: Buntingville, Conservator of Forests 2058.

Of the four species recognised in southern Africa, S. brachypetala has the widest

distribution, occurring in dry savannah and scrub forest in Southern Rhodesia,

Mocambique, Transvaal, Swaziland, Natal and extending as far south as Umtata and

Ngqeleni in the eastern Cape Province. As here envisaged, the species exhibits a certain

amount of variation in characters such as pubescence, number and size of leaflets,

and degree of suppression of the petals. Plants from the extremes of the distribution

range may give the impression of being distinct, but an examination of ample material

shows a series of intermediates, making the separation of even varieties of doubtful

value.

528

Of the above variable characters, pubescence has the least significance, showing

complete gradation from a persistent tomentum to a practically glabrous condition.

Burtt Davy’s var. pubescens, which is equal to typical 5. brachypetala in every other

respect, is therefore not upheld.

S. rogersii Burtt Davy, based on Rogers 21246 from Waterpoort in the

Soutpansberg, was separated mainly on the grounds that two or three normal petals

develop in each flower while, in typical S. brachypetala, the petals are all reduced to

linear filaments. In all other respects, the two species are practically identical. The

significance of the petal character is a problem that has been given particularly careful

attention both in the field and in the herbarium. In the first place, the suppression of

all petals is not invariably complete even in typical S. brachypetala, as will be seen

from the description given by Wood (Natal Plants, 4: 390) and from reference to

Table I. Further evidence bearing on the question was gathered in the Soutpansberg,

the type area of S. rogersii. Here most of the trees examined showed the development

of one to three normal petals per flower but, growing among these trees, and

indistinguishable from them, are trees whose flowers have all the petals suppressed as

in typical S. brachypetala. The reasons for the suppression or non-suppression of

petals on adjacent and otherwise identical trees are not known, but the character appears

to be constant for a particular tree. Another fact emerging from this study is that the

tendency to produce normal petals becomes more marked toward the northern limits

of the species. This is indicated by Table I which summarises the examination of 54

herbarium specimens.

Table I.

Relation between Geographical Distribution of Schotia brachypetala and number of Normal

Petals per Flower.

It is of interest to note that no case has yet been found of a flower with 5 normal

petals. On the basis of the petal character alone, it is evident that a separate species or

even variety cannot be upheld, and the conclusion is unavoidable that S. rogersii must

be relegated as a synonym.

The status of 5. semireducta Merxm. is not as easy to decide. Most of the specimens

seen from Southern Rhodesia show a combination of deviations from typical

S. brachypetala. There is the development of 2 to 4 normal petals per flower and,

in addition, most specimens differ in having up to six or, rarely, seven pairs of leaflets,

which are often more markedly and persistently tomentose.

In describing his species, Merxmuller contrasts it with S. brachypetala, S.

brachypetala var. pubescens and S. transvaalensis, but he fails to consider S. rogersii.

Thus the character which he stresses most, namely, the development of two or three

petals per flower, has already been used as a distinguishing feature for the separation

of a species and, on the basis of the findings here reported, the character must be

considered to have practically no taxonomic value in this particular group. The

remaining characters to consider in evaluating the status of S. semireducta are number

529

of leaflets and pubescence. Of these, pubescence has already been shown to have the

least meaning and, in fact, one specimen from Southern Rhodesia, Brain 9503, has

leaves and twigs that are glabrous, though the inflorescence branches are pubescent.

Regarding number of leaflets, typical S. brachypetala has almost consistently 4 pairs

or, rarely, 5 pairs. In the Transvaal, some specimens have up to 6 pairs, though this

is exceptional and the usual number is 4 pairs. In Southern Rhodesia the specimens

appear normally to have 6 pairs, rarely 5 or 7 pairs. Thus there is a progressive change

from south to north and the three main characters deviating from typical S. brachypetala

are found in S', semireducta, though the combination is not yet constant. This is, no

doubt, an evolutionary trend and, on available evidence, it is considered that the

differences are not sufficient to sustain S. semireducta as distinct.

S. brachypetala is an attractive, round-topped tree usually about 25 to 35 feet

high, though known to grow up to 45 or 50 feet. As pointed out by Wild, Rhod. Agr.

Joum. 49 : 3-9 (1952), it is often associated with termite mounds while, in the Lowveld,

it is one of the species found on river banks. It is practically evergreen, the leaves

usually being shed for a short while immediately before the clusters of red flowers

appear in October.

5. S. latifolia Jacq. X S. afra ( L .) Bodin, Form A.

S. stipulata Ait., Hort. Kew. ed. II. 3: 33(1811); DC., Prod. 2: 508 (1825); E. & Z.,

Enum. PI. Afr. 261 (1836).

S. speciosa var. ovalifolia Harv. in FI. Cap. 2: 274 (1861).

Figs. 9 and 11.

Shrub or small tree. Leaflets 3-5 pairs, 1 -7-4-2 cm. long and 1 -2-2-3 cm. broad,

sessile, elliptic-oblong to obovate, oblique, base rounded to cuneate, apex obtuse or

acute, often mucronate. Inflorescence terminal, more or less lax; inflorescence branches

semi-woody, patent. Pedicels 1-5 mm. long. Calyx leathery; tube 3-4 mm. long;

lobes 8-9 mm. long and 5-6 mm. broad, concave. Corolla reddish; petals 5, 12-13 mm.

long and 3-4 mm. broad. Stamens 10, free to the base. Ovary oblong, compressed,

4-5 mm. long, stipitate. Pod as in generic description. Seeds ovoid, flattened, pale

brown, with a yellow aril.

Cape. — Uitenhage (?): “Between Winterhoeksberge and Zwartberge ”, Ecklon

and Zeyher 1700 (SAM); “ Zwartkopsriver near Addo ”, Zeyher 2446 (SAM). Port

Elizabeth: Redhouse (on Swartkops River), Paterson 662 (GRA).

The description given above is based mainly on Paterson 662, which is complete

with leaves, flowers and fruits. Superficially, this specimen looks like S. latifolia,

with its 3-5 pairs of leaflets which vary considerably in size, its terminal, relatively lax

inflorescence and arillate seeds. However, it also possesses characters which normally

belong to S. afra, namely, pedicels 4-5 mm. long, calyx tube 4 mm. long and stamens

free to the base. There is thus evidence to suggest that it is of hybrid origin.

Two further specimens in the South African Museum Herbarium show a

resemblance to Paterson 662. These are Ecklon and Zeyher 1700 and Zeyher 2446

which are remarkably similar to one another and, in spite of the different localities

given, lead to the suspicion that they may be from the same gathering. They consist

of two twigs and one twig, respectively, having immature inflorescences which are

terminally placed and relatively lax. The leaflets are in 3-5 pairs, intermediate in size

between S. afra and S. latifolia while, as far as can be determined from the flower

buds, the stamens appear to be free.

530

It seems that the name S. stipulata Ait. is applicable to this small group of specimens

intermediate between S. afra and S. latifolia. This supposition is based on information

supplied by Mr. D. J. B. Killick, our Liaison Officer at Kew, who was asked for a

description of the type of Alton’s species, if it could be found. Alton based his species,

published in 1811, on a plant cultivated at Kew and said to have been grown from

seed supplied by Francis Masson in 1794. The only specimen found that comes close

to Aiton’s description is one in the British Museum, labelled “ Hb. Kew 1808 ” with

no further details. Two names are on the sheet: “ S. macrophylla ” and “ S. stipulata

It has 4 or 5 pairs of large leaflets, 2 -5-3-0 cm. long and 1-2-1 -6 cm. broad,

conspicuously mucronate, while the rhachis is distinctly winged (both these conditions

are found in the juvenile state of all Schotia species). The pedicels are short, about

1 mm. in length, but the flowers are larger than those of S', latifolia and the stamens

are free to the base. The plant has conspicuous stipules 7 mm. long and 4 mm. wide

but, as very few herbarium specimens have stipules, the significance of this character

cannot be judged. It may be inferred that this is probably Aiton’s type.

Masson travelled with Thunberg as far east as the Uitenhage District and was

at the Swartkops River, more or less the locality of Zeyher 2446 and Paterson 662,

on 15th December, 1773 (Phil. Trans. Roy. Soc. Lond., 66: 297, 1776). It would

have been useful corroborative evidence if Masson had made a herbarium specimen

corresponding to the seeds supplied to Kew, but no such specimen has been found.

Three Schotia specimens collected by Masson are at the British Museum; two of these

are clearly S', afra while the third, endorsed in pencil “ S. macrophylla ” is equal in

every respect to S. latifolia.

In Flora Capensis, Flarvey reduced S. stipulata Ait. to S. speciosa var. ovalifolia

citing, in addition, “ E. & Z. 261 ”. No Schotia specimen with this number has been

found and it is thought that Harvey cited, by mistake, the page number of Ecklon and

Zeyher’s Enumeratio because, on page 261 of that work, is listed “ 1700. S. stipulata

Ait.”.

To give this small group of intermediate specimens specific rank is clearly not

justified. To place them as a variety of either S. afra or S. latifolia would be equally

illogical as it would mean an undesirable widening of the conception of one of these

species. It is therefore preferred to keep them apart as an “ intermediate ” or

“ hybrid ” form, as has also been done with the specimens listed under the following

heading.

6. S. latifolia Jacq. X S. afra (L.) Bodin, Form B.

Fig. 10.

Shrub or small tree. Leaflets 4-10 pairs, 0-9-2- 5 cm. long and 4-10 mm. broad,

oblong to elliptic-oblong, rounded at the base, often oblique, usually obtuse at the

apex, sometimes mucronate. Inflorescence usually on abbreviated side branchlets,

more or less glomerate; inflorescence branches abbreviated, semi- woody. Pedicels

3-3-5 mm. long. Calyx leathery; tube obconical, 3-4 mm. long; lobes 7-9 mm.

long and 5-7 mm. broad, concave. Corolla pink to reddish; petals 5, 13-17 mm.

long and 3-4-5 mm. broad. Stamens 10, united at the base for 1-3 mm. with the sheath

split open on one side. Ovary oblong, compressed, 4-5 mm. long, stipitate. Pod as

in generic description. Seeds ovoid, flattened, pale brown, with a small or large aril.

Cape. — East London: Amabele Siding, Flanagan s.n.; Bonza Bay, East London

Museum in NBG. Bathurst: Lower Kariega Valley, Acocks 13280; Port Alfred,

R. Verdoorn 10. Alexandria: Between Alexandria and Grahamstown, Burtt Davy

12130.

531

Fig. 9. — Hybrid Form A, a, flower; b, petal; c, longitudinal section. Fig. 10. —

Hybrid Form B, a and b, leaves; c, petal; d, flower; e, longitudinal section.

All life size.

Fig. 11. — Hybrid Form A, leaf, life size.

532

Of the above specimens, Flanagan s.n. is fruiting, while the remainder are in

flower. The floral characters show a fairly uniform pattern, resembling S. afra in

their usually glomerate inflorescences, pedicel length and length of the calyx tube but,

like S. latifolia, the stamens are united at the base. The leaflets show more variation

though, in general, they are fewer and larger than in S. afra, being intermediate in

character between this species and S. latifolia. The Flanagan specimen has tomentulose

twigs and leaflets, and seeds with large arils as in S. latifolia, while the Acocks specimen

has seeds with much smaller arils. One gains the impression that these characters

could be derived through hybridisation between S. latifolia and S. afra, though there

is no direct evidence to support this view. As in the case of the specimens considered

under the previous heading, no advantage can be seen, at this stage, in giving these

few specimens varietal or specific rank. It is, perhaps, significant that they occur

in the area where both S. latifolia and S. afra are found.

Further field work is desirable in order to form a better idea of the distribution

and frequency of these aberrant forms, but the fact that they are not well represented

in herbaria suggests that they are relatively rare. A cyto-genetical study of seeds from

such plants may also help to decide whether they are really of hybrid origin and thus

throw light on their taxonomic position.

Fig. 12. — Map showing distribution of Schotia species in southern Africa.

533

Index.

Page.

Guajacum afrum L. = S. afra 517

Omphalobium Schotia Jacq. f. ex DC. = S. latifolia 523

Schotia afra (L.) Bodin 517

S. afra var. angustifolia (E. Mey.) Harv 520

S. alata Thunb. = Erythrophysa alata (E. & Z.) Hutch 515

S. angustifolia E. Mey. = S. afra. var. angustifolia 520

S. brachypetala Sond 526

S. brachypetala var. pubescens Burtt Davy = S. brachypetala 526

S. capitata Bolle 521

S. cuneifolia Gandoger = S. latifolia 523

S. diversifolia Walp. = S. latifolia 523

S. latifolia Jacq 523

S. latifolia X S. afra, Form A 529

S. latifolia X S. afra, Form B 530

S. parvifolia Jacq. = S. afra var. angustifolia 520

S. rogersii Burtt Davy = S. brachypetala 526

S. semireducta Merxm. = S. brachypetala 526

S. speciosa Jacq. = S. afra var. afra 517

S. speciosa var. angustifolia (E. Mey.) Harv. = S. afra var. angustifolia 520

S. speciosa var. ovalifolia Harv. = S. latifolia X S. afra, Form A 529

S. speciosa var. tamarindifolia (Afz. ex Sims) Harv. = S. afra var. afra 517

S. stipulata Ait. = S. latifolia X S. afra, Form A 529

S. tamarindifolia Afz. ex Sims = S. afra var. afra 517

S. tamarindifolia var. forbesiana Baill. = S. capitata 521

S. transvaalensis Rolfe = S. capitata 521

Theodora speciosa Medik. = S. afra 517

T. capitata (Bolle) Taub. = S. capitata 521

535

New and Interesting Records of South African

Flowering Plants.

By

E.E.A. Archibald, L.E. Codd, R.A. Dyer,

A. D. J. Meeuse and D. van Druten.

ACANTHACEAE.

Sclerochiton ilicifolius A. Meeuse, sp. nov.

Frutex multiramosus, 1-2 m., interdum usque ad 3 m., altus. Rami ramulisque

teretes, cortice griseo longitudinaliter pauceque rugoso obtecti; ramuli ulteriores

subtereti vel subquadrangulares, laeves, puberuli. Folia parva, rigida, tenuiter coriacea,

lucida, subtus pallidiora, oblonga, elliptica vel ovato-elliptica, acuta vel acuminata,

pungentia, plerumque pauce spinoso-dentata, interdum parce integra, supra tota

subtus ad costam puberula, 1-2 cm. longa et 0-6-1 cm. lata; petioli puberuli, 0-3 mm.

longi. Inflorescentiae terminales, spicatae, pauciflorae, densae; bracteae bracteolaeque

ellipticae, pungentes, integrae vel interdum 1-2 spinas laterales gerentes, dense

glanduloso-pubescentes. Calyx extus dense glanduloso-pubescens; sepala 5, inaequalia,

pungentia, posterius majus, lineari-oblongum, c. 3 cm. longum, plerumque bicuspidatum;

laterales et anteriora lanceolata, longe attenuata, acutissima. Corolla 1-labiata, alba;

lobi supra dilutissime violacei, c. 34 mm. longi; tubus brevis, ad apicem contractus,

in fauce dense barbatus; limbus cuneato-oblongus, sparse pubescens et intus in lineis

longitudinalibus 2 densissime sericeo-villosus, ad apicem 5-lobatis lobis 5 subaequalibus

oblongis apice rotundatis leviter deflexis et concavis c. 6 mm. longis. Stamina in fauce

inserta, exserta, subaequalia; filamenta arcuata, purpureo-brunnescentia, ad basin

pilosa; antherae 1-loculatae, depresse semiorbiculares, lanatae; connectivum ad

apicem non productum. Pollen ut in typo, subglobosum. Ovarium cylindrico-fusiforme,

glabrum, 4-ovulatum; discus flavus, annuliformis; stylus arcuatus, subulato-filiformis;

stigma subnullum minute 2-lobatum. Capsula oblonga vel ellipsoidea, glabra, lignosa,

15-20 mm. longa, 2-4-sperma. Semina subquadrata vel trapezoidea, magis compressa,

brunnea, glabra, rugulosa, 5-6 mm. longa.

Transvaal. — Warm Baths district, hills west of Warm Baths: Mogg s.n.; Waterberg

district, between Hermanusdoorns and Elmeston, Meeuse 9655 (type) in National

Herbarium, Pretoria; Meeuse 9657, Codd 401 1 ; Soutpansberg district, Schlesinger’s

Sawmill W. of Louis Trichardt, Gerstner 5929, 1946 (all in PRE; isotypes and

duplicates of Meeuse 9657 distributed to BM, BOL, BR, COI, EA, K, L, P, SRGH,

and elsewhere).

Description. — Shrub, erect or sometimes somewhat straggly, much branched from

the base and throughout, 1-2 m., occasionally up to 3 m. high. Main stems and branches

woody, up to about 2 cm. in diam., covered with a greyish longitudinally wrinkled

bark; the branches usually bearing numerous short leafy lateral shoots; ultimate

twigs terete or faintly quadrangular, firm, yellowish or pale buff to pale yellowish green,

smooth, densely puberulous, rather densely leafy mainly towards the apex and many

-'’•niinating in an inflorescence. Innovations glandular-pubescent. Leaves resembling

536

those of the European holly {Ilex aquifolium L.) in shape and texture but smaller,

usually oblong or elliptic varying to somewhat ovate or obovate, rather dark green

and somewhat shiny above, pale green and duller below, rigid, thinly coriaceous,

somewhat tapering to cuneate at the base, acute to acuminate and spine-tipped at the

apex, 1-2-5 cm. long and 0-6-1 cm. wide; the margin slightly thickened to subrevolute,

usually with 1-3 spine-tipped triangular teeth on either side, rarely some leaves entire;

upper surface glabrous or somewhat puberulous at the base in mature leaves; lower

surface glabrous, marked with minute cystoliths, puberulous on the midrib; midrib

faintly impressed above, slightly prominent below; lateral nerves (3-)4-7(-8) on

either side, hardly impressed above, in fresh leaves impressed but in dried specimens

prominent on lower surface, emerging at an angle of about 45°, rather straight, mostly

reaching the margin and some of them ending in the thin spine of a lateral tooth of the

leaf-blade; tertiary nervation inconspicuous; petioles 0-3 mm. long, somewhat

depressed above, convex below, puberulous. Inflorescences terminal on the flowering

branches, spicate, ovoid or shortly cylindric, 4-6 flowered; upper leaves gradually

changing into a few lower sterile bracts; sterile and fertile bracts entire or occasionally

with a lateral spine on one or both sides, ovate or ovate-elliptic to somewhat obovate-

elliptic, slightly concave, acuminate and spine-tipped, in colour and texture resembling

the leaves but slightly paler and densely covered with fine gland-tipped patent hairs

intermixed with shorter non-glandular hairs on outer (lower) side, on their margins

and on their tips inside, and on lower half of inner (upper) surface with some short

adpressed hairs, 20-25 mm. long and about 10 mm. wide; main nerve ending in the

terminal spine and on either side with 1-3 nerves arising from the base and running

more or less parallel with the main nerve, all reaching the margin in upper half of the

bract; bracteoles 2, very similar to the bracts but slightly narrower, 7-8 mm. wide.

Calyx green, unequally 5-lobed, the lobes ending in soft spines, densely glandular,

pubescent outside and near the tips inside, upper (posticous) lobe the largest-

linear-oblong, acute to acuminate and undivided, or more or less deeply bicuspidate

with acute, spine-tipped cusps at the apex, about 3 cm. long and 6-5 mm. wide; 2 lateral

lobes linear-lanceolate, long-tapering into a very acute apex, about 2-5 cm. long and

2-2-5 mm. wide near the base; basal (anticous) lobes linear-lanceolate, tapering

into an acute apex, 26 mm. long and 3-3-5 mm. wide near the base. Corolla white

tinged with very pale mauve on the inside of the lobes and more creamy-white on

lower side of lip, 34 mm. long, tube short, smooth, shin^, glabrous outside, cylindric-

campanulate, thence widening into an oblique throat which is partly covered with soft

white adpressed hairs outside, mainly on upper side, and densely bearded within; limb

large, 1 -lipped, cuneate-oblong, sub-equally 5-lobed, about 25 mm. long and 21 mm.

wide near the apex, nearly glabrous below, with a few short hairs and two very densely

white-silky palatal ridges extending from the throat to the corolla-lobes on upper (inner)

side; the lobes oblong, rounded at the apex. Stamens 4, subequal, inserted in the

throat in two pairs, filaments brown, densely bearded at the base and 2 upper

(posticous) ones thinly hairy throughout, about 15 mm. long; anthers 1-thecous,

depressed semi-orbicular, completely attached to the filament, yellowish brown, those

of each pair more or less connate by woolly hairs; connective greenish. Pollen

subglobose, smooth, with three longitudinal very shallow chinks not reaching the

poles. Pistil about 19 mm. long; disc annular, bright yellow; ovary ovoid-oblong,

rather dark green, glabrous, 4-ovuled; style subulate-filiform, rather stout, slightly

upcurved in basal part, thence faintly arcuate-recurved, very pale mauvish green with

a line of white hairs near its base on lower (abaxial) side otherwise glabrous; stigma

truncate, faintly bilobed. Capsule oblong or ovoid-oblong, somewhat laterally

compressed, tapering to apiculate at the apex, turning a shiny brown when ripe, woody,

15-20 mm. long. Seeds 2-4, much compressed, subquadrate or trapezoid often more

or less distinctly oblique or produced on one side at the base, brown when ripe, finely

rugose, without hygroscopic hairs, 5-6 mm. long.

537

Fig. 1. — Sclerochiton ilicifolius flowering branch; 2, leaf; 3, pubescence of floral leaves

and calyx, showing gland-tipped hairs and short non-glandular ones; 4, corolla,

seen from above; 5, stamen, seen from the side; 6, pistil; 7, capsule and 8,

seeds. 1, 2 and 4: natural size. 5, 7 and 8: X 1-5, 3, X 100 and 6, x 2.

538

a .

b.

c.

c(. Qj .

Fig. 2. — a bract, b bracteole, c posticous calyx-segment, d lateral calyx-segment, e one

of the anticous calyx-segments, natural size.

Fig. 3. — Pollen grain of Sclerochiton ilicifolius seen from the aequatorial side (left)

and from the polar side (right), appr. x 400.

Note. — Some flowers of the specimens collected under the number Meeuse 9655

are abnormal. They have more than 5-calyx-segments, 6-9 corolla-lobes and 5 or 6

stamens or 4 stamens and 1 or 2 sterile staminode-like structures. This is obviously

a monstrosity or a teratologicum because the flowers of the Acanthaceae are generally

4- or 5-merous with 4 or 2 stamens.

Prior to this investigation the two specimens Gerstner 5929 and Codd 4011

were placed tentatively in the genus Crossandra and the specimen collected by Mogg

in 1924 was unearthed recently in a parcel of unnamed specimens in the National

Herbarium preparation room. In April, 1955, ample fresh flowering material was

obtained in the Waterberg district near the site of Codd 4011 and in the beginning of

June of the same year excellent fruiting material was collected in exactly the same

locality.

The new species evidently belongs to the tribe Acanthoideae-Acantheae on account

of the one-lipped corolla, the 4 stamens with one-loculed anthers and the pollen

morphology and was compared with the genera belonging to this tribe.

It differs from Acanthus in its habit, in the 5-partite calyx (4-partite in Acanthus),

in the subequally 5-lobed corolla (unequally 3- or 5-lobed in Acanthus) and in the

globose (not ellipsoid) pollen. It differs from Blepharis and Acanthopsis in the 5-partite

calyx, the sub-equally 5-lobed corolla, the pollen and the glabrous seeds (the seeds

have hygroscopic hairs in the other two genera) ; in addition, Blepharis has the filaments

of the anticous segments produced behind the anthers and the ovary has 2 hollows

filled with glands.

539

It differs from Crossandra, among other things, in the short corolla-tube (long

in Crossandra ), in the colour of the corolla (red, orange or yellow in Crossandra), in

the exserted stamens (included in Crossandra ), in the shape of the pollen and in the

glabrous seeds (those of Crossandra are covered with hygroscopic hairs or with scales

fringed with hygroscopic hairs). It was evident that the plant agreed with the genus

Sclerochiton (including Pseudoblepharis ) in all essential characters such as the shape

of the inflorescence, the 5-partite calyx, the subequally 5-lobed corolla, the morphology

of the stamens and pollen, the stigma and the glabrous seeds. It differs from all

previously described species of Sclerochiton in a few less important, mainly vegetative

characters such as the spine-toothed and pungent leaves, the pungent floral leaves

and calyx-segments and the glandular pubescence on the floral leaves and calyx. The

leaves in all other species are entire, although those of the type species, Sclerochiton

harveyanus Nees, are sometimes faintly crenate or serrate. Pungent floral leaves and

calyx segments, and a glandular pubescence were also not previously recorded for

this genus. These differences do not, however, warrant separate generic status. All

descriptions of species of Sclerochiton, a genus restricted to Africa and Madagascar,

were compared and not one is even remotely suggestive of the species discussed here.

No other genus proposed in the tribe Acantheae from Africa or Madagascar was

encountered that could possibly fit our plant, so that it is most unlikely that it was

ever described before. A specimen was kindly compared by Dr. H. Wild of the

Government Herbarium, Salisbury, Southern Rhodesia, who reported that he did not

know the plant and that it was not represented in the Salisbury herbarium. This would

indicate that the plant does not occur in tropical Africa and is restricted to the Transvaal.

Its known distribution in the Transvaal is very scattered, and although the plant, even

when not in flower, is rather remarkable with its bright green foliage consisting of

small spine-toothed leaves and would attract the attention of botanists on the solid

quartzite outcrops forming its natural habitat, it must be a rare or very local species.

On the other hand it can be quite frequent locally. The leaves, though smaller, are

strongly reminiscent of those of the European holly, Ilex aquifolium L., and the specific

epithet “ ilicifolius ” is considered to be appropriate as it not only expresses this

resemblance but also emphasizes one of the main differences between the new species

proposed here and all other species of Sclerochiton.

ANACARDIACEAE.

Rhus batophylla L. E. Codd, sp. nov., affinis R. tomentosae L. foliolae marginibus in

toto serrato-dentatis, ramulis infra foliisque dense lanatis, inflorescentiis longioribus

et angustioribus differt.

Frutex 1—1 - 5 m. altus e basi profunde ramosus. Rami ascendentes, sparse ramosi,

superne dense lanati. Folia petiolata, petiolis 1-2 cm. longis, supra leviter canaliculatis,

dense lanatis; foliolae sessiles, ellipticae, ovatae vel obovatae, coriaceae, serrato-

dentatae, dentibus triangularibus mucronatis 2-3 mm. longis basi 3-4 mm. latis, infra

dense tomentosae, supra glabrescentes ; foliola terminalis 4- 5-6 • 5 cm. longa, 2 • 3-3 cm.

lata, foliolae laterales 2 -5-3 -8 cm. longae, 1-5-2 -6 cm. latae. Paniculae term in ales

vel non-nunquam axillares, angustae, 8-18 cm. longae, 2-4 cm. latae, dense lanatae.

Flores aggregati, pedicellis 0-5-1 mm. longis; sepala triangularia, plusminusve 0-7 mm.

longa; petala glabra, 1-1 -5 mm. longa, in floribus masculis staminibus 5 brevioribus.

Drupae compressae, 5-6 mm. latae, glabrae, rubrae.

Plate No. I.

Transvaal. — Lydenburg District, Sekukuniland : Mogg s.n.; Mooihoek Chrome

Mine, \% miles west of Driehoek, Codd and Dyer 7699, female {type), 7700, male;

Story 4051; 5 miles south-east of Steelpoort Post Office, Codd 6700.

540

Shrub 1-1-5 m. high, branching freely from the base. Branches stiff, ascending,

sparingly branched, unarmed, the younger branches densely lanate, with reddish-brown

bark under the tomentum. Leaves trifoliolate, petiolate; petioles 1-2 cm. long, slightly

channelled on the upper surface, not winged, densely lanate; leaflets sessile, elliptic

to ovate or obovate, obtuse to cuneate at the base, acute, coriaceous, tending to fold

along the midrib, regularly and conspicuously serrate-dentate from base to apex, the

teeth broadly triangular, mucronate, 2-3 mm., long and 3-4 mm. broad at the base;

under surface of leaf covered with dense, white to cream coloured, woolly, persistent

tomentum, the upper surface covered thinly in the young stage, becoming glabrescent

with age; secondary nerves prominent on under surface; tertiary nerves reticulate

above, obscured by the tomentum below in the young stage but becoming conspicuous

with age; terminal leaflet 4 -5-6 -5 cm. long and 2-3-3 cm. broad with 8-12 pairs of

secondary nerves ; lateral leaflets 2 • 5-3 • 8 cm. long and 1 • 5-2 • 6 cm. broad with

6-9 pairs of secondary nerves. Inflorescence terminal and also axillary in the upper

leaf axils, forming a narrow, elongate panicle 8-18 cm. long and 2-4 cm. wide, the

main axis and branchlets densely lanate tomentose. Flowers borne in clusters ; pedicels

0-5-1 mm. long. Calyx pubescent, 5-partite, segments triangular, less than 1 mm.

long. Corolla glabrous, yellow-green; petals 5, 1-1-5 mm. long. Male flowers with

5 stamens, shorter than the petals ; disc present ; ovary absent. Female flowers usually

with 5 staminodes; ovary flattened. Drupe glabrous, flattened, 5-6 mm. long, red

and wrinkled when mature.

Rhus batophylla is a very distinctive shrub showing no close affinity to any known

species. The dense, woolly tomentum persisting on the branchlets and on the underside

of the leaves suggests that its relationship is to Schonland’s section Tomentosa, but

it is readily distinguished from any species in this group and, in fact, from any other

South African species, by the margins of the leaflets being regularly serrate-dentate

from base to apex, and in the long, sparingly branched, narrow inflorescence.

Specimens were sent to our liaison officer at the Kew Herbarium, where they were

unmatched.

With its discolorous, markedly serrate leaflets and its habit of branching near

ground level, producing numerous stiff, ascending branches 3 to 4 feet high, the plants

bear a striking resemblance to a species of Rubus, and the specific epithet is derived

from the Greek batos, bramble. The plants tend to grow communally in the dry

bushveld to the north and west of Burgersfort, in eastern Sekukuniland, where the

altitude is about 3,000 feet. The species has not been recorded outside this area.

ASCLEPIADACEAE.

Brachystelma montanum, R. A. Dyer, sp. nov., affinis B. crispo Grah., sed habitu minore

corollis multo minoribus interne immaculatis externfc glabris differt.

Herba perennis; tuber 2-5 cm. latus, plus minusve 2 cm. altus, subglobosus vel

depresso-globosus. Caules perennes, subterranei, solitarii vel bini, ad 5 cm. longi,

caules annui 1-4, usque ad 6 cm. alti, superne ramosi. Folia oblonga ad lineari-

lanceolata, basalia 1-5-2 -3 cm. longa, 2-4 mm. lata, superiora breviora et angustiora

obtusa ad acuta, infra et marginibus scabrida, supra tenuiter scabrida. Flores ad

nodos plerumque 2, pedicelli graciles, 1-3-5 mm. longi, dorso pilis paucis instructi.

Corolla usque ad 1 • 5 cm. longa, basi tubulosa; tubus campanulatus, usque ad 4 - 5 mm.

longus, 4-5 mm. diametro; lobi e basi ovato-lanceolato lineares, circiter 1 cm. longi,

minus 1 mm. lati, infra medium ciliati. Coronae lobi inferne in seriei uno confluentes,

exteriores poculos V-formatos formantes, interiores supra incumbentes, partibus liberis

linearibus usque ad 1 mm. longis.

541

Cape Province. — Wodehouse Division, northern slopes of Andriesberg, 6,000 ft.

Dec. Galpin 6278 in National Herbarium Pretoria, type (duplicate at Kew).

Perennial herb; tuber 2-5 cm. broad and about 2 cm. thick, subglobose or

depressed, with 1-2 perennial underground stems up to 5 cm. long; annual branches

1-4, up to 6 cm. tall, branched above. Leaves oblong to linear-oblong; basal ones

1 • 5-2-3 cm. long, 2-4 mm. broad and upper ones considerably reduced in size. Flowers

usually 2 from each node; pedicels slender 1-3-5 cm. long, thinly pubescent. Sepals

lanceolate, 3-3-5 mm. long with a few hairs on their backs. Corolla up to 1-5 cm.

long, tubular at base; tube campanulate, up to 4-5 mm. long and 4-5 mm. diam.

at the mouth; lobes linear from an ovate-lanceolate base, about 1 cm. long, less than

1 mm. broad above, ciliate below the middle. Corona lobes united more or less into

one series about 2 mm. high; the outer lobes are like V-shaped pockets with their

margins appearing as shoulders at the base of the inner lobes; inner lobes incumbent

on the backs of the anthers with the free portion linear, up to 1 mm. long.

Galpin’s specimen has remained unnamed since its collection in 1901. It is well

known that the flowers of Stapelieae and of allied genera such as Brachystelma shrivel

during drying unless particular care is taken to prevent it. In the case of Galpin 6278

parts of the collection have obviously contracted considerably in drying while others

appear to have retained their normal size and the measurements given above were taken

from these.

Brachystelma pilosum R. A. Dyer, sp. nov., affinis B. galpinii N.E. Br., sed robustior,

pilis patulis hirsuta, coronae lobis exterioribus glabris, coronae lobis interioribus

oblongis obtusis differt.

Herba perennis, erecta usque ad 25 cm. alta, prope basin ramosa, pilis patulis

1- 1 -5 mm. longis hirsuta. Caulis basi 4-5 mm. diametro; internodiis 2-4 cm. longis.

Folia sessilia vel subsessilia, oblonga, 1-1-5 cm. longa, 3-5 mm. lata, marginibus satis

involutis et leviter undulatis. Flores patentes vel suberecti, plerumque 2 ad nodos.

Sepala lanceolata, 2-2-5 mm. longa. Corolla 1-6-1 -8 cm. longa, parte inferiore

patelliformi 1-1-5 mm. connata, lobi basi lanceolati, medio ad 0-5 mm. contracti,

superne ad 1 mm. lati, apicibus connati, externe dimidio inferiore, superne tenuiter

hirsuti, interiore glabri. Corona columnae staminalis supra basin circiter 0-75 mm.

inserta; coronae lobi exteriores 1-5 mm. longi, in segmentis linearibus profunde

bilobati; coronae lobi interiores oblongi, supra antheras incumbentes et leviter

longiores, basin versus dorsale leviter umbonati.

Transvaal.— Potgietersrust Division; near Naboomspruit on farm Nooitgedacht

in lightly wooded grassland, rare, July, Galpin 11573 in National Herbarium, Pretoria,

type; Crawley, Transvaal Mus. No. 10402, at Kew and N.H.P.

Perennial herb, erect, up to about 25 cm. tall, branched from near the base. Stem

4-5 mm. diam. near base; branches hirsute with spreading hairs 1-1-5 mm. long;

internodes 2-4 cm. long. Leaves sessile or subsessile, more or less oblong, 1-1 • 5 cm.

long, 3-5 mm. broad, with the margins somewhat folded upwards and slightly undulate.

Flowers spreading or suberect, usually 2 together at the nodes. Sepals lanceolate,

2- 2 • 5 mm. long. Corolla 1 • 6-1 • 8 cm. long, with the basal united portion saucer-shaped,

1-1 • 5 mm. long; lobes united at tips with a lanceolate base and contracted to 0 - 5 mm.

then expanded towards the tips to 1 mm. broad, hirsute on the backs in the lower half

and thinly so above, glabrous on the inner surface. Corona arising from the staminal

column about 0-75 mm. above its base; outer corona lobes about 1-5 mm. long,

deeply bilobed into linear segments; inner corona lobes oblong, incumbent on the

backs of the anthers and slightly exceeding them but not meeting over the centre of the

staminal column, slightly thickened or humped on the backs towards the base.

542

In Kew Bulletin 1953 p. 358, Mr. Bullock proposes the resuscitation of the genus

Dichaelia Harv. for the species of Brachystelma in which the tips of the corolla-lobes

are connate. In describing B. pilosum, which has this character, I have followed

N. E. Brown and used the generic name Brachystelma because I consider the observations

by Brown in FI. Cap. 4, 1 : 834 (1908) are well founded and his treatment consistent

with the known facts.

LILIACEAE.

Albuca crudenii, Archibald sp. nov., affinis A. glandulosae, Bak., folio uno, perianthio

glabro, stigmate convexo differt.

Bulbus subglobosus, circiter 30 mm. diam., membranaceo-tunicatus. Folium

unum, viridi-flavum, lineario-lanceolatum, planum vel paulum plicatum, usque ad

45 cm. longum, 16 mm. latum, novellum apice subulato, marginibus venisque puberulis.

Scapus 45 cm. longus, puberulus ; pedunculus 28 cm. longus, 3 mm. diam. ; racemus

laxus secundus; bracteae membranaceae, rubro-nervatae, deltoideae, apiculatae, 9 mm.

longae, 8 mm. latae; pedicelli cemui, puberuli. Flores 3—10, penduli, odorati; petala

exteriora glabra, patentia, aurea, medio viridi-striato, oblonga, 15 mm. longa, 6 mm.

lata, apicibus haud cucullatis; petala interiora glabra, erecta, colore pallidiora quam

exteriora, ovata, 12 mm. longa, 5 mm. lata, apicibus cucullatis 1 mm. longis. Stamina

exteriora infertila, interiora fertila. Gynoecium glabrum; ovarium 5 mm. longum,

3 mm. diam., viride, basin versus sine scuta; stilus oblongus, 5 mm. longus, 1 mm.

diam., luteus; stigma convexum, viride.

Cape Province.— Albany District: Dassieklip, October, Archibald 5289 in Albany

Museum Herbarium (type); Alicedale, September, Cruden 14a in Albany Museum

Herbarium.

Bulb nearly spherical, up to 30 mm. diam.; scales tunicate, outer light brown,

membranous, inner white, fleshy; roots fibrous. Leaves usually one, only very rarely

two, light yellowish-green, lax, mesophytic, linear-lanceolate, flat or slightly folded,

up to 45 cm. long, 16 mm. broad, tip subulate in young leaf but withering off early,

veins marked on upper and lower surfaces by fine channels, leaf margins and

vein-channels each with a single row of minute, glandular, transparent, erect hairs.

Inflorescence a lax, secund raceme up to 45 cm. long; peduncle up to 28 cm. long,

3 mm. diam., with minute, sparsely distributed, glandular hairs in inconspicuous

channels, raceme a quarter to half as long as peduncle, similarly puberulus, bracts

deltoid, apiculate, up to 9 mm. long, 8 mm. broad at base, membranous, transparent,

with several light reddish-brown veins; pedicels cernuous, becoming erect as fruit

ripens, lowest up to 20 mm. long becoming shorter towards apex of raceme, puberulus.

Flowers 3-10, pendulous, with faint vanilla-like scent. Perianth glabrous; outer

segments up to 15 mm. long, 6 mm. broad, patent, oblong, with very slightly cucullate

apex, bright yellow with narrow green median stripe; inner segments up to 12 mm.

long, 5 mm. broad, erect, connivent, ovate, pale yellow with narrow green median

stripe, apex cucullate, 1 mm. long. Androecium three sterile and three fertile stamens ;

outer filaments ovate-lanceolate, up to 9 mm. long, 1 • 5 mm. broad, with mucronate,

sterile tip; inner filaments spade-shaped, lower part 1-5 mm. long, 1-5 mm. broad,

upper part 5 • 5 mm. long, less than 1 mm. broad; anthers 4 mm. long, apex rectangular,

base spreading. Gynoecium glabrous; ovary ellipsoidal in longitudinal section,

carpels rounded dorsally, without basal shields, 5 mm. long, 3 mm. diam., green; style

oblong with slightly narrower section in upper third, triquetrous, 5 mm. long, 1 mm.

diam, yellow; stigma convex, 3-lobed, lobes simple each with a row of minute papillae,

green. Fruit a trilocular capsule, apex acute when green; seeds numerous.

543

ing fruit.

544

In describing this new species I wish to pay tribute to the late Dr. S. Schonland,

former Curator of the Albany Museum and founder of the Department of Botany at

Rhodes University, Grahamstown. In the Albany Museum Herbarium there is a

single specimen of this species labelled “ Albuca Crudeni Schonl. n.sp.”. It was

collected by Mr. F. Cruden at Alicedale in September, 1917 (Cruden, No. 14a). Dr.

Schonland made copious notes on many other species of Albuca in the herbarium

but this sheet is devoid of any other comment than the manuscript name.

In October 1953 I collected plants at Dassieklip and subsequently found that

they matched Cruden’s specimen. As the description was drawn up from the fresh

material collected at Dassieklip this has been designated as the type, but I retain the

name suggested by Dr. Schonland to commemorate Mr. Cruden who discovered

several interesting plants in the eastern Cape Province.

Dassieklip is an interesting little poort in the Albany District where the old road

between Port Elizabeth and Grahamstown passes through a fault of Witteberg Quartzite

on the east bank of the Bushman’s River. The krantz faces south and consequently

is in the shade for a large part of the day. On its rock ledges are to be found small

ripens, lowest up to 20 mm. long becoming shorter towards apex of raceme, puberulus.

colonies of A. crudenii and many other small geophytes as well as that rather uncommon

fern Ophioglossum nudicaule , L.f.

A. crudenii falls into the section Eu-albuca because it has three fertile and three

sterile stamens. It is quite distinct from any other species so far attributed to this

section in that it has only one leaf, or, very rarely, two. The only other species in the

section that has glandular leaves is A. glandulosa, Bak. which, however, has white

flowers with glandular hairs on the perianth segments, a turbinate, puberulus style

and flat stigma as well as many other minor dissimilarities.

Aloe alooides (Bolus) van Druten comb. nov.

Urginea alooides Bolus in Journ. Linn. Soc. 18: 395 (1881).

Notosceptrum alooides Benth. in Gen. PI. 3: 775 (1883); Bkr. in FI. Cap. 6: 285

(1897).

Aloe recurvifolia Groenewald in Tydskrif vir Wetenskap en Kuns 39 (1935); Reynolds

in FI. PI. S. Afr. 16: Plate 801 (1935), in Journ. of S.A. Bot. 6: 111, Plate 10

(1940), The Aloes of S. Afr. 436 (1950).

During research work at the National Herbarium the iso-type of Urginea alooides

was obtained on loan from the South African Museum, Cape Town. The specimen

consists of an inflorescence which was found to be an excellent match of the type of

Aloe recurvifolia Groenewald, which is the name in current use for the plant.

Bolus had originally described this plant as Urginea alooides and stated ” A

remarkable species with the aspect of an Aloe. It is readily distinguished by its long

densely flowered spike. I have only seen the flowering scapes; but the leaves were

described to me by the late Mr. MacLea, an energetic collector in the region indicated ”.

In a note attached to his collection MacLea states “ I considered it an Aloe of some

kind, its habit of growth and long fleshy strap-shaped leaves gave me that impression.

It affects same habit as most Aloes, grows on slopes of mountains facing North-East.

The present was growing amongst quartzite rock and shingle near Mac-Mac. I

remember distinctly the reason I gathered no plants or leaves was on account of their

size and thick fleshy nature — not so sure about the leaves being strap-shaped or

linear-lanceolate. I think they were the former and pendulous at the extremities and

not less than 2 ft. long. More than one flower spike to each plant.” It seems strange

545

that, with this information, Bolus should have placed the plant in the genus Urginea

but he apparently ruled out the genus Aloe because of the very small-flowered dense

spike.

In his Genera Plantarum Bentham transferred Bolus’s species to his new genus

Notosceptrum and it was retained there, tentatively, by Baker in Flora Capensis.

Nothing further was done until the present investigation, and although it is unfortunate

that so apt a specific epithet as recurvifolia must be changed, there is no doubt that the

plant must bear the name Aloe alooides.

VITACEAE.

Cissus natalitia (Szysz.) L. E. Codd, comb. nov.

Vitis natalitia Szysz., Polypet. Rehmann., 2: 45 (1889).

Cissus connivens Lam. var. meyeriana Planch, in DC. Monog. Phan. 5: 584 (1887),

partly; Gilg and Brandt in Engl. Bot. Jahrb. 46: 516 (1912), partly; Suessenguth

in Pflanzenfam. 20d: 250 (1953), partly.

C. glabra E. Mey. in Drege, Zwei Pfl. Docum. 173 (1843), nomen nudum; Planch,

l.c. 584, in synonomy.

C. orientalis Harv. in FI. Cap. 1: 253 (1859/60), partly, non Lam.

The specimens cited by Szyszylowycz are as follows: —

Natal— Ladysmith, Rehmann 7136; Biggarsberg, Rehmann 7098; Vildshill,

Pinetown, Rehmann 7991.

In Flora Capensis, Harvey mis-identified two specimens as Cissus orientalis Lam.,

namely, Drege, from between the Umtata and Umzimvubu Rivers, and Burke and

Zeyher, from the Magaliesberg. Planchon placed the two specimens under C. connivens

Lam. var. meyeriana Planch., a treatment followed by Gilg and Brandt, while the

species C. orientalis Lam. he removed to the genus Ampelocissus. A study of available

herbarium material, together with observations in the field, led to the conclusion that

these specimens are representative of two distinct species.

The Burke and Zeyher specimen was given specific rank as C. sulcatus C. A. Smith

in Burtt Davy’s Flowering Plants and Ferns of the Transvaal and Swaziland, 2: xx,

477 (1932). Thus the question arose as to whether a specific name was available for the

Drege specimen. Judging by the localities and descriptions, it seemed probable that

it would prove to belong to the same species as the Rehmann specimens described as

Vitis natalitia Szysz., and this view was confirmed by our liaison officer at the Kew

Herbarium.

A comparison between the species C. sulcatus, C. natalitia and C. connivens was

made possible through the kindness of the Director of the Museum National d’Histoire

Naturelle, Paris, who sent us a portion of Lamarck’s type on loan. The conclusion

reached is that C. sulcatus is more closely related to C. connivens than is C. natalitia

and that, while there are sufficient grounds for regarding C. natalitia as a good species,

more material is required from Madagascar, the type area for C. connivens, for a

correct evaluation to be made of the status of C. sulcatus. In separating C. sulcatus

from C. connivens, C. A. Smith states very briefly that his species differs in the

multipinnate not bipinnate leaves, and in the fruits, though he fails to indicate how

the fruits differ. In the present state of our knowledge, the three species may be

distinguished as in the following key: —

Stipitate glands usually present; leaflets regularly dentate from base to apex; stems

thick, sparingly branched C. natalitia.

Stipitate glands absent; leaflets crenate-dentate mainly towards the apex; stems thin,

much branched

Leaves bipinnate; leaflets cuneate or acute at the base C. connivens.

Leaves often multipinnate; leaflets rounded or obtuse at the base C. sulcatus.

546

C. natalitia is distributed from northern Natal to the Transkei and, in addition

to the Drege and Rehmann specimens mentioned previously, is represented by the

following specimens: —

Natal. — Ngotshe, Codd 7024. Nongoma, Gerstner 718. Hlabisa, Codd 2010.

Nqutu, Codd 2401. Dundee, van der Merwe 2649. Klip River, Acocks 13855. Weenen,

Medley Wood 10132; Acocks 10709. Eshowe, Gerstner 2406. Mtunzini, Mogg 5909,

6013. Pietermaritzburg, Killick 262; Codd 6795. Camperdown, McClean 143. Port

Shepstone, Galpin 11012.

Cape. — Port St. Johns, Howlett 43; Bruce 451. Kentani Pegler 890.

C. sulcatus is so far recorded only from the central Transvaal, as indicated by

the following collections: —

Transvaal. — Brits, Hutchinson 2326; C. A. Smith 3443; Mogg s.n.; Hofmeyr

s.n.; Leendertz 713. Rustenburg, Codd 6364; Pole Evans s.n; Waterberg, van Son

in TM. 30435.

547

Plate I. — Rhus batophylla L. E. Codd, sp. nov. (Codd and Dyer 7699).

549

The Oleaceae of Southern Africa

By

I. C. Verdoorn.

The area covered in this revision is the Union of South Africa, Swaziland, South

West Africa, and the region of the Bechuanaland Protectorate which lies between

South West Africa and the Transvaal. To date, collecting in South West Africa and

the Protectorate has not been extensive and so the citations from these regions may

give a wrong impression — the family may be better represented there than it appears

to be.

All the material of Oleaceae from the following herbaria was sent on loan for

examination and most of the specimens seen have been cited (the symbols used in the

citations follow the names): Bolus Herbarium (BOL); South African Museum

(SAM); Kirstenbosch (NBG); Albany Museum (GRA); Natal Herbarium (NH);

Witwatersrand University (J); Forest Department (FD); Ryksherbarium, Leiden (L);

and the Natal University Herbarium, Pietermaritzburg (PMB). A number of specimens

specially asked for were sent from the Royal Botanic Gardens, Kew (K).

In the citations of the specimens, if a specimen is in the National Herbarium,

Pretoria (PRE), no symbol follows it; if not at Pretoria but in one or more of the

other South African Herbaria, the symbol of ONE of them is added — e.g. Pillans 5271

(BOL), even if it is also in the Kirstenbosch herbarium; if the specimen is in the Kew

or Leiden herbarium or both these overseas herbaria their symbols are added in every

case together with one of the South African herbaria (including Pretoria in this instance)

if it is also in that herbarium — e.g. Burchell 8236 (K, L, GRA) or Grenfell 869

(K, PRE). In view of the large number of specimens examined this economy in the

use of symbols, instead of cutting out some citations, seems to be justified.

Among the abbreviations used in the citations of literature the only one that may

not be easily recognised is F.T.E.A. for the “ Flora of Tropical East Africa ”.

The decision as to the correct authority for certain species needs some explanation.

Harvey’s species Menodora juncea was published in Harvey’s Genera Plantarum,

2nd edition, four years after his death, edited by J. D. Hooker. It has been suggested

that the name should be attributed to “ Hook, f.” rather than to Harvey. Article 58

of the International Code of Botanical Nomenclature provides that if the publishing

author supplied the description his name should be retained rather than that of the

one who proposed the name but did not publish it. In this case, however, all the

description there is of Menodora juncea is Harvey’s, based on the material in his

herbarium and therefore, in my opinion, Harvey is the correct author of the species.

In like manner the descriptions of Jasminum breviflorum, J. gerrardi and Olea

enervis, published several years after Harvey’s death by Wright in the Flora Capensis,

were essentially Harvey’s and therefore if abbreviating the authority “ Harvey in

Wright ” Harvey’s name should be retained in preference to the publishing author

who did not supply the original description.

In the case of Jasminum streptopus, the authority given in the Index Kewensis

and the Flora Capensis, is “ E. Mey. ex DC.”, but Meyer described this species in his

“ Commentariorum ” in 1837 and is therefore the author without qualification. This

applies also to J. angulare var. glabratum E. Mey. and a few more of his varieties in

this genus.

550

I am much indebted to the Directors of the herbaria mentioned above, not only

for the loan of the specimens but, in many cases, for assistance and advice from them

and members of their staffs. I am also indebted to the Directors of the herbaria at

Berlin-Dahlem, Geneva and Paris for photographs of type specimens.

FAMILY CHARACTERS.

Trees, shrubs, climbers or occasionally low herbaceous woody-based plants.

Leaves opposite (rarely alternate in the low sub-herbaceous plants), exstipulate, simple

or compound, acarodomatia sometimes present on underside of leaves in axils of veins.

Inflorescence cymous usually paniculate or occasionally only one flower developing,

or flowers appearing fascicled. Flowers regular, hermaphrodite or occasionally

unisexual, sometimes with heterostylous arrangement. Calyx shortly or deeply

2-15-lobed. Corolla gamopetalous, 4 or more lobed, sometimes lobed almost to the

base. Stamens normally 2, epipetalous. Ovary superior, bilocular with 2 to 4 ovules

in each loculus. Fruit dry or fleshy, dehiscent or indehiscent.

Key to Genera.

Fruit a woody capsule; corolla salver-shaped, tube well developed, lobes spreading,

with a group of swollen brown to purplish hairs at the base of the lobes ....

Fruit a drupe, berry or membranous capsule, never a woody capsule; corolla-lobes

without a group of swollen brown to purplish hairs at their base:

Fruit a drupe or berry; trees, shrubs or climbers:

Corolla-tube long and slender; lobes narrow; fruit a berry, 2-lobed,

one usually aborting; climbers or bushy shrubs

Corolla-tube short to almost none:

Leaves without acarodomatia; calyx toothed or shallowly lobed;

corolla-tube short but distinct, lobes only slightly

longer than broad; inflorescence terminal or

axillary, very many flowered; ovules pendulous,

seed albuminous

Leaves on flowering branches usually with acarodomatia; calyx

deeply lobed; corolla-tube very short, usually slit to

base between alternate lobes, lobes usually appearing

distinctly longer than broad* because of deeply

infolded margins; apex cucullate; inflorescence

axillary often on old wood, few-flowered; ovules

attached ventrally; seeds exalbuminous, cotyledons

thick

Fruit a membranous capsule; low virgate shrublet or herbaceous plant with

woody base; corolla yellow

* Except in L. latipetala in trop. Africa.

1. SCHREBERA.

Roxb., PI. Coromandel 2: 1, t. 101 (1798) nomen conservandum ; Harv. ex Wright

in FI. Cap. 4, 1: 482 (1907); Phill. Gen. S.A. FI. Plants Ed. 2: 571 (1951);

Nathusia Hochst. in Flora (1841).

Trees or shrubs (sometimes sub-scandent?). Leaves opposite, compound (simple

in some species outside S. Africa), rhachis usually winged. Inflorescence a paniculate

cyme. Flowers heterostylous. Calyx campanulate, loosely enveloping the corolla,

truncate or irregularly and obscurely lobed. Corolla salver-shaped, white, sometimes

tinged with pink or puce; tube well developed, cylindrical; segments 6 or more,

spreading to reflexed, each with a group of swollen brown to purplish hairs at the base.

Stamens 2, inserted on the corolla either in the throat with the anthers exserted or a

little below the throat with the anthers included; filaments short, anthers large, introrse.

1. Schrebera.

2. Jasminum.

3. Olea.

4. Linociera.

5. Menodova.

551

Ovary bilocular, small, truncate or obscurely bi-lobed at apex; ovules 4 in each loculus;

style filiform; stigma included or exserted, subcapitate or oblong in outline. Capsule

bi-valved, woody with loculicidal dehiscence; seeds produced into a long solitary

subapical wing.

Type species: S. Swietenioides Roxb., Burma.

The genus Schrebera is found in Africa and India, mostly in the tropics. The

two species that extend into S. Africa have compound leaves. There is not a single

record of a simple-leaved species from inside the Union, the nearest being S. trichoclada

Welw., found quite frequently in the Zambesi basin with one record as far south as

Lundi, S. Rhodesia. In P.E.A. the southernmost record is on the north bank of the

Save River.

Key to Species.

Branchlets glabrous or occasionally puberulous; leaves glabrous; branches of

inflorescence and calyx persistently puberulous, never tomentulose

nor quite smooth; cymes not very compact 1. S. alata.

Branchlets grey or silvery tomentulose; leaves pubescent or glabrescent; branches

of inflorescence and calyx tomentulose, the calyx sometimes becoming

quite smooth (never puberulous); cymes compact 2. S. argyrotricha.

1. Schrebera alata ( Hochst .) Welw. in Trans. Linn. Soc. Bot. 27: 41 (1869); Turrili

in F.T.E.A. (1952); Nathusia alata Hochst. in Flora 24. 1 : 25 and 2: 672(1841);

Schrebera saundersiae Harv. in Thes. Cap. 2: 40, tab. 163 (1859); Wright in FI.

Cap. 4, 1: 483 (1907); Schrebera latialata Gilg in Bot. Jb. 30: 73 (1901).

Tree (said to occur occasionally as a sub-scandent shrub in Natal), found in the

marginal flora of forests and the coastal bush in Natal, sometimes small, more or less

12 feet tall with slender stem and branches, or about 25 feet tall with stem 1 foot in

diameter, bark greyish or light brown in colour. Branchlets glabrous or puberulous,

not tomentulose. Leaves glabrous, 5-13 cm. long, 5-foliolate, sometimes 3-foliolate ;

leaflets varying considerably in size, texture and shape, the lateral elliptic to broadly

elliptic or oblong, unequally cuneate at the base, about 3-7 cm. long and 1-4 cm. broad,

the terminal usually a little larger than the other leaflets, more obovate oblong, sometimes

elliptic, subsessile or cuneate into a petiolule up to 1 cm. long; petiole and rhachis

winged, wing variable, narrow or broad, auricled at base or not. Inflorescence

puberulous, about 6 to 11 cm. long, the lateral branches slender, usually over 2 cm.

long; bracts deciduous variable in size, some 5 mm. long and 2 mm. broad, often

narrower, rarely broader. Calyx more or less tubular-turbinate, persistently puberulous

outside sometimes densely so, almost velvety, pubescent in upper portion within, more

or less truncate and slit or variously lobed, persisting in fruit. Flowers sweetly scented,

white suffused with pink or puce at different stages, tube cylindrical about 1-4 cm.

long, glabrous or variously puberulous in parts outside, often pubescent with rather

long hairs near the base within; lobes suborbicular, about 5 mm. long, crenate on

margins, spreading, with a cluster of glandular reddish-brown hairs near the base on

inner face. Stamens 2, heterostylus, inserted near the apex of the corolla and then

exserted or in other specimens below the apex and then included. Ovary 2-celled,

about 1 • 5 mm. long, somewhat flattened and puberulous on top, obscurely 2-lobed ;

ovules 4 in each cell, pendulous; style about 1-5 cm. long, glabrous or sometimes

sparsely pubescent with short patent gland-tipped hairs, long exserted in flowers with

stamens included and vice versa; stigma about 1-5 mm. long, obscurely 2-lobed.

Capsule green when young (turning black on drying) becoming light brown at maturity,

puberulous except at base within calyx, obovate-oblong, slightly laterally flattened at

right angles to the septum, parallel to the wall of division with a ridge along the septum.

Seeds 8 mm. long with an oblong 1 • 1 cm. x 6 mm. subterminal wing (degree to which

wing runs down one side varies, also width).

Fig. 1.

552

Type: Schimper 245, Abyssinia. Types of synonyms: S. sounder siae Harv.,

Gerrard 1153, Durban; S. latialata Gilg, Medley Wood 5201, Durban.

Transvaal. — Soutpansberg: Entabini, Maviewa Kop, Poynton 78 (F.D.); Sibasa,

Junod in Tvl. Mus. 21199; nr. Sibasa, Tshaulau, Codd 6913 (leafy branch, 3-foliate);

van Warmelo s.n. (3 sheets). Pilgrims Rest: Kowyns Pass, Codd & Verdoorn 7604;

Story 4000; Mariepskop, van Son in Tvl. Mus. 30712; Fitzsimons & Van Dam in

Tvl. Mus. 26,537; Keet in F.D. Flerb. 5916; Forester in F.D. Herb. 8119. Fydenburg:

Magalieskop, Kotze in F.D. Herb. 2816. Barberton: Summit Rimers Creek Gorge,

Galpin 1115.

Fig. 1. — Schrebera alata (Hochst.) Welw. from Graskop, Transvaal. Branchlets

glabrous but branches of the inflorescence and calyces puberulous; wings on petiole

and rhachis not obvious on drawing; a, dehiscing capsule; b, seeds.

553

Natal. — Ingwavuma: Codd 7034. Ubombo: Gerstner 4551 (leafy specimen,

wings broad). Eshowe: Lawn 863 (NH). Mapumulo: Nonoti, Medley Wood 10366.

Inanda: Tongaat, McKen 17 (NH); Inanda, Medley Wood 819 (NH). Durban: Gerrard

1153 (type no. of S. saundersiae Harv.); Medley Wood 11839; Medley Wood 12016

(NH); Berea, Medley Wood 5201 (type no. of 5'. latialata Gilg); Medley Wood 11042

(L). Umlazi: Umlaas, Maurice & Evans 329 (mixed sheet, centre S. argyrotricha)

(NH). Camperdown: between Drummond & Inchanga, Butcher in NH. 17694.

Also found northwards to Angola and Abyssinia.

Morphological variation in this genus makes the definition of its species very

difficult and as it becomes better known in Africa the tendency is to return to a broad

view of the species. Turrill in “ Flora of Tropical East Africa ”, 1952, takes this broad

view and sinks under S. alata (Hochst.) Welw. the following East African species:

S. obliquifoliolata Gilg, 5. merkeri Lingelsh. and S. nyassae Lingelsh. Under this

treatment S. saundersiae Harv. from Natal must be sunk under it too and in fact, as far

back as 1869 when Welwitsch made the combination “ Schrebera alata ” for Nathusia

alata Hochst., he was of the opinion that S. saundersiae Harv. was conspecific with

Hochstetter’s Abyssinian plant and his own specimens from Angola, and cited all

these specimens under the name Schrebera alata (Hochst.) Welw. Gilg, however,

resuscitated it and described several new species, most of which are now again sunk.

With its capacity to vary in size, shape, and texture of leaves and the degree to which

the rhachis is winged one may find specimens from neighbouring trees appearing very

different, sometimes more so than a specimen from Abyssinia and one from Natal.

Among Gilg’s species is Schrebera latialata, the type being a Medley Wood specimen

from Berea, Durban, that is from the same area as S. saundersiae Harv. The characters

which distinguish Gilg’s species, for instance the broad wings on the petiole, have been

proved to vary from season to season, probably according to the rains and the amount

of shade the plant gets, and it can therefore not be upheld and goes, with S. saundersiae,

into the synonymy of S. alata.

The species is distinguished from S. argyrotricha Gilg mainly by the glabrous or

glabrescent branchlets (not densely tomentulose) ; glabrous leaves; branches of the

inflorescence and calyx being evenly puberulous; and the rather open, not compact,

cymes. In some cases it may be difficult to distinguish between these species and the

suggestion that the differences are not specific may present itself. But the difference

in the pubescence is not merely one of degree but a different type of pubescence. In

S. alata the new growth is glabrous or puberulous, never silvery tomentulose, and the

branches of the inflorescence are puberulous and the calyx densely and persistently so,

whereas in S. argyrotricha the branches of the inflorescence are silvery tomentulose

while the calyx sometimes becomes quite smooth, especially in the upper half. In

this latter species one often finds a smooth calyx and silvery pubescent pedicel while

in 5. alata this is never the case, the pedicel and calyx being evenly puberulous.

S. alata is usually a bigger tree with rather larger leaves and more open inflorescence.

It is found only in forest or on the fringes of forests in mountainous country and in

the south in the coastal bush of Natal.

2. Schrebera argyrotricha Gilg in Bot. Jb. 30: 74 (1901); Wright in FI. Cap. 4,

1: 483 (1907); S. gilgiana Lingelsh. in Pflanzenr. 72: 108 (1920); S.mazoensis

Sp. Moore in J. Bot. Lond. 45: 48 (1907).

Small trees (or shrubs) in dry bushveld, often with a short main stem about 3 feet

long, and several long erect smooth branches, but sometimes trees about 18 feet tall;

young branchlets grey or silvery tomentulose. Leaves 3-5-foliolate, 4-10 cm. long,

grey green in effect owing to the pale ashy pubescence which is persistent in western

area plants, while those from eastern regions are glabrescent; leaflets variable in

shape and size, the lateral from 1 • 5-7 • 5 cm. long and • 8-3 • 5 cm. broad, usually very

broadly elliptic-oblong, the terminal subsessile, slightly larger, more or less obovate

554

and shortly cuneate at the base, upper surfaces slightly darker than the lower, both

surfaces pubescent or (in eastern regions) glabrous; petiole and rhachis narrowly

to broadly winged, variously auricled at or near the base or without auricles.

Inflorescence tomentulose, 2-7 cm. long, cymes compact; bracts deciduous, variable

in size, lower up to 1 cm. long, 9 mm. broad, upper usually 5 mm. long and 3 mm.

broad, innermost narrower, more or less spathulate. Calyx silvery pubescent to

glabrescent outside, smooth at least in top half, minutely puberulous on upper portion

within, tubular-turbinate, truncate, slit or shallowly lobed, persistent in fruit. Corolla

sweetly scented, white suffused with pink or puce at different stages; tube cylindrical,

smooth outside and with a few long hairs about the centre within, about 1 cm. long;

lobes often 7, spreading abruptly to slightly reflexed, about 4-5 mm. long and

3 -5-4 -5 mm. broad, irregularly crenate on the margins with a cluster of glandular

reddish brown hairs near the base on inner face. Stamens 2, heterostylus, inserted

near the apex of the tube and then exserted or, in other specimens, below the apex

and then included. Ovary 2-celled, about 1-5 mm. long, subtruncate and obscurely

2-lobed at the apex, very shortly pubescent on top; style about 1 cm. long, usually

glabrous; stigma long exserted in flowers with stamens included and vice versa, bifid,

lobes about 1 • 5 mm. long, adhering to each other. Capsule green and pubescent when

young (turning black on drying), light brown at maturity, oblong obovoid slightly

laterally flattened, with a ridge along the septum. Seeds about 7 mm. long, with a

subterminal, oblong wing about 9 mm. long and 6 mm. wide, sometimes narrower

(wing varies in width and degree in which it runs down one side).

Fjg. 2.

Type: Wilms 201, Lydenburg. Types of synonyms: S. mazoensis Sp. Moore,

Eyles 202, S. Rhodesia; S. gilgiana Lingelsh., syntypes, Marloth 3403, S. Rhodesia,

Rehmann 5950, Transvaal, Landauer 150, Natal.

Transvaal. — Soutpansberg : Machava, Acocks 8861; Elim, Botha 2 (two

gatherings) and in F.D. Herb. 10735. Pietersburg: Blaauwberg Smuts and Pole

Evans 919 (small leaves, 3-foliolate); Bremekamp and Schweickerdt 113, (leaves small);

Codd 8676 (leaves large, 3-5-foliolate) ; Codd and Dyer 9178 and 9120 (small leaves).

Letaba: Houtbos, Rehmann 5950, syntype of S. gilgiana (Bol); Modjadjies Reserve,

Krige 206. Waterberg: Palala Heights, Meeuse 9638; near Naboomspruit, Yischgat,

Galpin 8855; Nylstroom hills, Mogg s.n. and in Moss Herb. Wits. 26475; near

Warmbaths, Smuts s.n. ; Codd 2961; Marais 268; 269; Verdoorn 2400; 2401; 2402

(coppice shoots); 2403 (after flowering). Pilgrims Rest: Graskop, Lothian Reserve,

Joubert in F.D. Herb. 8720; near Leydsdorp, Stapelton in F.D. Herb. 8462.

Lydenburg: Waterfall (type locality), du Plessis 1; 2; 3; Marais 63; 64; 66;

Verdooen and Codd 7614; 7615; 7618; Penge Mine, Codd and Dyer 7755; Buffelsvlei,

Story 4076. Nelspruit: near Pretorius Kop, Codd 6053; Codd and de Winter 4883;

Acocks 16620; v. d. Schyjf 169; Shabin, v. d. Schyff 1469 (glabrous); Klokwene,

v. d. Schyff 181; Skukuza, Cholmondeley s.n.; near Nelspruit, Mogg s.n.; Plaston,

Holt 303 (NH); Junction Nels and Corodile Riv., Liebenberg 2621. Barberton:

Thorncroft 606 (NH); Williams in Tvl. Mus. 10447; Thorncroft 606 (NH); Roses

Creek, Mathews 144.

Swaziland. — Mohlangotsha Mtn., Mankiana, Miller S/261.

Natal. — Nkandhla, Insuzi Gorge, Acocks 12730; Mapumula: Monoti, Medley

Wood 8789 (NH).

Also occurs in S. Rhodesia and Portuguese East Africa (possibly also in

Tanganyika for Turrill in F.T.E.A. cites a specimen which he names S. mazoensis

from Kondoa, T.T.).

555

Fig. 2. — Schrebera argyrotricha Gilg., from the waterfall, Lydenburg. Branchlets

and branches of the inflorescence tomentulose, calyces glabrescent in parts ; a , dehiscing

capsule; b, seeds.

Two ecotypes are found in this species, the specimens from the western and central

areas being densely and persistently silvery tomentulose while those in the eastern

regions have a tendency to glabrescence in parts. In these eastern specimens the

calyces are often quite smooth while the pedicels and the rest of the inflorescence are

silvery tomentulose; the leaves may be glabrous but new growth usually has the

characteristic tomentulose pubescence. By a strange coincidence the two ecotypes,

the western and eastern forms both occur at the type locality of S. argyrotricha, the

Waterfall near Lydenburg. Through a personal visit to that locality the error of

considering these forms as distinct species was avoided. The Waterberg and Rhodesian

specimens with the conspicuous tomentulose twigs had been taken to be S. mazoensis

while specimens with smooth calyces and less hairy leaves were named S. argyrotricha.

On the visit to the Waterfall where Wilms collected the type specimen described as

having a glabrescent calyx, specimens of Schrebera were found to be quite plentiful.

Those growing in exposed positions were as densely and persistently tomentose as the

specimens in the Waterberg and S. Rhodesia, while others, sometimes only a few yards

away, but growing in the shade of the cliff, had smooth calyces and more or less glabrous

129828-5

556

leaves. Further examination of specimens from Rhodesia, Transvaal, Swaziland

and Natal supported the view that they belong to one species but occur in the two

forms described.

As in S. alata so in this species there is a great deal of variation in the size, texture

and shape of the leaflets, in their number, 3 or 5, and the degree to which the rhachis

is wniged and auricled. For this reason S. gilgiana Lingelsh. can hardly be maintained

on the small size of the leaflets since the measurements fall within the range of size of

our species. The 3 specimens cited come from 3 distant localities, Rhodesia, Transvaal

and Natal but also well within the range of distribution of S. argyrotricha. To illustrate

the unreliability of leaf size, two specimens in the National Herbarium from the

Blaauwberg, collected at different times, both had small 3-foliolate leaves. In a recent

expedition to that remote spot the botanist was asked to look out for this Schrebera

and material was brought back with the normal sized leaflets and most of the leaves

5-foliolate.

Notes that may assist in distinguishing this species from S. alata are given under

the latter species.

Among the tropical species 5. argyrotricha is most like the description of

S. tomentella (Welw.) Gilg which however has larger leaves, outside the range of

variation in size of our species.

Another tropical species, S. greenwayi Turrill, is very like the eastern form of

5. argyrotricha, having almost glabrous leaves and the calyx smooth in part in contrast

to the pubescent pedicel but in this case the tomentum on the new growth may be

considered to differ in some respect from that of our species.

2. JASMINUM.

Linn., Gen. PI., ed. 5: 7 (1754); Harv. ex Wright in FI. Cap. 4, 1 : 479 (1906); Phill.

Gen. S.A. FI. Plants ed. 2: 571 (1951).

Climbing, scrambling or erect shrubs or shrublets. Leaves compound or reduced

to one leaflet, digitately or pinnately 3 or more foliolate, usually opposite. Inflorescence

consisting of a solitary flower or a few flowered simple or compound cyme, terminal

on main and lateral branchlets, sometimes axillary as well. Flowers white or yellow,

sometimes flushed with rose, sweetly scented. Calyx-tube campanulate, lobes 4-13,

often long and subulate. Corolla-tube long and slender, lobes 4-11, imbricate, spreading

abruptly. Stamens 2, inserted near the top of the tube, not exserted; filaments short.

Ovary 2-locular with 2, rarely more, ovules in each loculus, attached near the base;

usually only one ovule developing; style slender; stigma clavate or bi-lobed. Fruit

a berry, deeply bi-lobed, usually one lobe aborting.

Type species: J. officinale L., India.

Under the common name “ Jasmine ” this genus is largely grown in gardens as

an ornamental plant. At least two of our species are frequently found in cultivation,

J. multipartitum and J. angulare.

Among a number of Indian species found in South African gardens there is one,

identified here as J. multiflorum (Burm.) Andr.,* which seems to have become

naturalised along the Umgeni River near Durban and possibly also in Barberton where

it is frequently grown as a hedge. J. multiflorum is easily distinguished from the

indigenous species by the ovate-cordate leaves which are mostly over 3-5 cm. long

and 2-5 cm. broad and by the densely hirsute-tomentose twigs, pedicels and calyces

including the long subulate calyx-lobes.

* J. multiflorum (Burm.) Andr. = J. pubescens Willd. and J. hirsutum Willd. pro. parte non

N. hirsuta L. (see Andr. Bot. Rep. t. 496 and B.M. 1991).

557

Another tropical Asian species which is quite commonly Cultivated and which,

when dried and pressed, may be confused with a S. African species is J. humile L. It

has angled stems like our J. angulare and the calyx lobes are similar; although the

leaves in the S. African species, J. angulare, may be pinnate occasionally they are

constantly, or more commonly so, in J. humile. The growing plant, when in flower,

could not be confused with our plant because the flowers are yellow instead of white.

A specimen in the Albany Museum Schonland 945 named J. angulare, is rather

J. humile, and must have been found in cultivation or it may have been an escape.

Key to Species.

Leaves compound, 3-5-7-foliolate :

Leaves usually digitately 5-7-foliolate, sometimes 3-foliolate but then

specimens from the T ransvaal ; leaflets more than twice as

long as broad; shrublets about 11 feet tall

Leaves usually 3-foliolate, sometimes some pinnately 5-foliolate ; shrubs,

twiners or scramblers :

Leaves narrowly linear-landeolate to linear-oblong, more than

twice as long as broad, glabrous and without

acarodomatia, only petioles pubescent

Leaves ovate to oblong, seldom up to twice as long as broad :

Inflorescence fairly compact; petioles sub-erect; leaves

(on pressed specimens) covering the base of

the flowers, occasionally pinnately 5-foliolate

Inflorescence of broad lax cymes, usually broader than

long; petioles spreading; leaves (on

pressed specimens) not covering the base of

the flowers:

Inflorescence terminal and axillary only in upper

leaves; acarodomatia, if present, in

axils of lower veins only; leaflets

usually under 5 cm. long; calyx

small, under 2 mm. long

Inflorescence terminal and axillary; acarodomatia

in axils of central and upper veins

also ; leaflets usually over 5 cm . long ;

calyx usually over 2 mm. long. . . .

Leaves unifoliolate :

Calyx-lobes much shorter than the tube, very shallow or up to T5 mm.

long, then thickened and conduplicate, apiculate, but never

produced into a subulate upper portion ; acarodomatia if

present mostly in axils of basal or lower veins; petioles

articulate in upper portion

Calyx-lobes almost as long as to much longer than the tube, subulate at

least in upper half, if rather short then leaves with

acarodomatia also in axils of upper veins:

Whole plant glabrous or inconspicuously puberulous in parts;

no acarodomatia:

Glabrous plants with lanceolate-elliptic, glaucous leaves

usually about 3 times as long as broad;

petiole articulate in upper portion ; inflores-

cence normally 3-flowered; restricted to

south-western Cape

Twigs, at least, shortly densely and persistently puberulous ;

leaves rarely 3 times as long as broad;

petiole articulate in lower half; flowers

usually single on short pedicels

Plants, at least in parts, conspicuously pubescent, either tomen-

tulose or patently pubescent with straight or

curled hairs:

Pedicels rather short and thick, usually about 5-7 mm.

long, densely pubescent; calyx lobes up to

13, subulate, usually definitely longer than

the tube; leaves without acarodomatia;

twigs more or less tomentulose; petioles

articulate, usually in lower half

1 . J. quinatum.

2. J. tortuosum.

3. J. angulare.

4. J. fluminense.

5. J. abyssinicum.

6. J. breviflontm.

7. J. glaucum.

8. J. multipartitum.

9. J. stenolobum.

558

Pedicels rather long and slender, sparsely pilose or

glabrous ; leaves usually with acarodomatia ;

petioles short, articulate near apex 10./. streptopus

(aggregate sp.).

Leaves oblong-elliptic to ovate-oblong mostly

4-7 cm. long; inflorescence usually

3-5-flowered 10a. var. streptopus.

Leaves on the whole smaller and ovate-acuminate,

mostly 1-5-3 -5 cm. long; inflores-

cence usually 1 -flowered 10b. var. transvaalense.

1. J. quinatum Schinz in Vjschr. Naturf. Ges. Zurich. 55: 245 (1910). J. tortuosum

Harv. ex Wright in FI. Cap. 4: 482 (1907) partly, as to specimens from Kalahari,

non Willd. J. lupinifolium Gilg & Schellb. in Bot. Jb. 51: 83 (1914).

Sub-shrubs with erect or straggling stems about 40 cm. long from a creeping

rhizome. Twigs ridged or angled, glabrous or thinly to densely pubescent with straight

or crisped hairs. Leaves digitately 5-foliolate, sometimes 3-, 4- or 7-foliolate; petiole

flattened, 1-5-5 mm. long, rarely up to 12 mm. long; leaflets with the central the

largest, 12-40 mm. long, and 2-9 mm. broad, rarely larger, the lateral smaller and the

outermost still smaller, glabrous or sparsely to densely pubescent with straight or

crisped hairs, linear-lanceolate, lanceolate to ovate-lanceolate, apiculate at the apex,

cuneate at the base into a petiolule-like claw, claw varying in length, that of the central

leaf usually the longest, up to 5 mm. long, rarely longer. Inflorescence terminal,

sometimes also axillary in the upper pair of leaves, cymes 1- to 3-flowered, the terminal

and axillary forming what appears to be a sub-corymbose inflorescence; peduncle

0 to 20 mm. long, glabrous or pubescent; pedicels 3-12 mm. long, of some lateral

1 -flowered cymes from axils of upper leaves up to 20 mm. long, glabrous or pubescent.

Calyx glabrous, tube about 2-5 mm. long, usually 6-toothed, teeth somewhat unequal,

1-2-5 mm. long, rarely up to 3 mm. long, triangular with margins folded inwards,

sometimes conduplicate, apiculate, sinuses U-shaped. Corolla white, tube 18-22 mm.

long, usually 6-lobed; lobes 10-12 mm. long and about 4 mm. broad. Stamens

apiculate, included. Ovary quadrate, about 1-5 mm. long; style 2-lobed, lobes 3 mm.

long, eventually exserted. Fruit twin berries, often only 1 developing, globose, about

5 mm. diam.

Plate 1.

Type: Sehlechter 3914, Corocodile Riv., Lydenburg. Type of synonym:

J. lupinifolium Gilg & Schellenb., Wilms 924, Lydenburg.

Transvaal. — Pretoria: Hennops River, Leendertz in Tvl. Mus. 8174; 8212;

Koedoespoort, Mogg 11788. Carolina: Rademacher in Tvl. Mus. 7473; Waterval

Boven, Rogers 22639 (NH); 18386 (Tvl. Mus. 2511); Pole Evans 2619; Waterval

Onder, Prosser 1228 (K, PRE); Bergendal, Galpin 12365; Steynsdorp, Dieperink 30.

Belfast: near Draaikraal, Codd 8053; Schoenmanskloof, Young A355. Lydenburg:

Wilms 924, syntype of J. lupinifolium Gilg and Schellenb. (L); near Corocodile River,

Sehlechter 3914 (type number); Spekboom River, Young A455; A456; Secocoeni

Land, Gray 4186; along Sabi Road, Smuts and Gillett 2494; near Lydenburg, Codd

5646; 6663; 6664 (hairy form); Codd and Verdoorn 7599; Rooidraai, Liebenberg

3489; Origstad Valley, Walters 10762. Pilgrims Rest: Rogers 24260 (GRA). Piet

Retief : Galpin 9667.

Not known to occur outside the Transvaal.

This species has a superficial resemblance to J. tortuosum Willd. of the eastern

Cape and Wright, in the Flora Capensis, cites Wilms 924 and Rogers 2511 from the

Transvaal, under J. tortuosum whereas both these specimens are J. quinatum. The

resemblance is in the long narrow leaflets without acarodomatia, but the species are

easily distinguished in that the leaves of J. quinatum are mostly digitately 5-foliolate

559

and only occasionally with some leaves 3- or 7-foliolate. Then the habit of the latter

is a low shrublet forming a fairly dense growth from a creeping rootstock whereas

J. tortuosum is a high climber, the slender voluble branchlets probably accounting

for the name. Also the area of distribution for these two species is distinct.

In the Botanische Jahrbiicher, 1914, Gilg and Schellenberg described Jasminum

lupinifolium on 3 gatherings of Wilms, among them Wilms 924 which, as mentioned

above, was cited by Wright under J. tortuosum Willd. From the description, and

from the examination of Wilms 924 as represented in the Ryksherbarium, Leiden,

it is obvious that this species is the same as J. quinatum and being a later homonym

must be sunk under J. quinatum Schinz. The epithet “ lupinifolium ” is very apt, for

the leaves are reminiscent of those of lupins. Although more often five, the leaflets

vary in number from three to seven, the central being the largest with the longest

petiolule. They all stand more or less erect and are so crowded on the pressed specimens

that it is difficult to distinguish them.

Like in other species in the genus, J. quinatum may occur in a completely glabrous,

a conspicuously hairy or an intermediate form. It is found on mountain slopes, among

rocks on hillsides, in the open, along sandy stream banks and on shaly slopes in open

woodland.

2. J. tortuosum Willd. Enum. 1: 10 (1809); DC. Prodr. 8: 311 (1844) excl. vars.;

Harv. ex. Wright in FI. Cap. 4, 1 : 482 (1907) pro parte; J. flexile Jacq. Schoenbr.

4: 46, t. 490 (1804) non Vahl.

Twiner or scrambler with branches angled in parts, the ultimate usually hirsute in

parts, hairs white, crisped. Leaves 3-foliolate; petiole hirsute especially along the

upper surface, about 5—10 mm., rarely 2 cm. long; blade glabrous and without

acarodomatia, 1-7-4 cm. long and -5-1-4 cm. broad (rarely larger 4-5 x 2 cm.);

acute or broadly rounded at apex, mucronate, cuneate at base into a petiolule, central

petiolule the longest; midrib prominent below and running into the petiolule at the

base. Inflorescence glabrous, terminal on the branches and on the lateral branchlets,

and axillary in axils of the upper pair of leaves only, 3-5-flowered, peduncle 1-2-5 cm.

long; pedicels 1-1-5 cm. long (rarely 2 cm. long). Calyx usually 5-lobed, sometimes

6, 2 -5-5 -5 mm. long; lobes from much shorter than the tube to about as long as the

tube, 1-2-5 mm. long, triangular, acuminate, conduplicate, apiculate, sinuses U-shaped.

Corolla white, usually 6-lobed, tube 1 • 5-2-7 cm. long; lobes up to 1-2 cm. long, 5 mm.

broad. Fruit not seen.

Plates 2 & 3.

Type: Willdenow in Berlin-Darlem Herbarium, Cape Province.

Cape. — Mossel Bay: Grootplaats, Muir 2400. Oudtshoorn: Huis Riv. Pass,

Compton 20327, rather broad leaves (NBG). Caledon: Elbrecht in Tvl. Mus. 19010.

Somerset West: Gordons Bay, Parker 4305. Tulbagh?: Drege s.n. (L).

Not known to occur outside the Cape Province.

This species is not well represented and it is possible that the specimens cited are

merely variants of J. angulare. They agree quite well with the type (see plate 2) and

with Jacquin’s figure, cited by DC. under J. tortuosum Willd., with the exception of

Compton 20327 which has larger leaves, but agrees in the other characters, such as

leaves glabrous, without acarodomatia, and usually more than twice as long as broad.

The 4 specimens from the Cape Province, cited by C. H. Wright in the Flora

Capensis under J. tortuosum, have been examined and they are all J. angulare. The

specimens cited from the Transvaal are J. quinatum, as mentioned in the notes under

that species.

560

The Drege specimen in the Ryksherbarium, Leiden, cited here may be of the

same Drege gathering as seen by De Candolle and identified with J. tortuosum Willd.

(DC. Prodr. 8: 311). A photograph of the Leiden specimen is reproduced here,

on the right-hand side of plate 3. As may be seen from a comparison of plates 2 and 3,

it compares well with the type of the species.

3. J. angulare Vahl Symb. 3: 1 (1794); Willd. Sp. PI. 38 (1797); Hooker in Bot.

Mag. t. 6865 (1886); Harv. ex Wright in FI. Cap. 4, 1: 481 (1907) excl. Burtt

Davy 360. J. capense Thb. Prod. 2 (1794-1800); Thb. FI. Cap. Ed. 1 : 4(1807-13).

J. angulare var. glabratum E. Mey. Comm. 1 fasc. 2: 14 (1837); DC. Prodr.

8: 311 (1844). J. natalense Gilg. and Schellenb. in Bot. Jb. 51 : 86 (1914) in part,

excluding Schlechter 11749 from Komati Poort.

Shrub, usually scrambling or climbing, sometimes climbing up to 20 feet high in

trees. Ultimate branchlets 4-30 cm. long, angled (at least in parts), angles ridged,

glabrous to variously pubescent from thinly so to tomentulous and densely pubescent

with crisped hairs. Leaves 3-foliolate, or occasionally a few pinnately 5-foliolate;

petioles usually ascending, rarely patently spreading, 0-4-2 cm. long, glabrous, thinly

pubescent or tomentulose; leaflets glabrous, thinly pubescent or tomentulose on

both surfaces, acarodomatia often present on under surface in axils of lower veins,

usually broadly ovate, sometimes oblong, variable in shape, with apex acute, mucronate

or rounded, the terminal 1-3-4 -5 cm. long and 0-6-2 -5 cm. broad, with a petiolule

•3-2 cm. long, rarely lobed at the base, but when deeply lobed forming a pinnately

5-foliolate leaf; lateral leaflets usually distinctly smaller than the terminal with a

petiolule 2-6 mm. long, sometimes longer. Inflorescence of 1 or 2 terminal, rather

compact cymes (the leaves in pressed specimens usually concealing the base of the

inflorescence), pedicels of lateral flowers about 1-2 cm. long, usually glabrous, sometimes

thinly pubescent or rarely densely so. Calyx campanulate, usually glabrous, tube

2-5-3 mm. long, 7-toothed, teeth about 1-5 mm. long, varying in size. Corolla white,

usually greenish outside (rarely pink, fide Galpin); tube 1-7-3 -5 cm. long; lobes

5, about 1-1-5 cm. long, 6-7 mm. broad; stamens inserted in upper portion of tube,

included filaments 2 mm. long, anthers 2-5 mm. long, distinctly apiculate, apicule

1 mm. long. Ovary brown, 1 • 5 mm. long, style ultimately exserted, 2-lobed, lobes

about 5 mm. long. Fruit a globose berry, often in pairs, about 7 mm. diam.

Plate 4.

Type: Drege s.n., Cape Province. Types of synonyms: J. capense Thb., Thunberg

s.n., Zeeko Riv., Uitenhage; J. natalense Gilg & Schellenb., syntypes, Wood 940,

Weenen, Kuntze s.n., Ladysmith, Bachmann 1029, and Beyrich 77, Pondoland ( Schlechter

11749 is J. fluminense).

Transvaal. — Volksrust: Jenkins in Tvl. Mus. 9300.

Natal.— Utrecht : Wahl in Tvl. Mus. 15388, some leaves 5-foliolate. Newcastle:

Charlestown, Medley Wood 5166, 1 leaf 5-foliolate. Dundee: Blesboklaagte, Codd

2412. Nqutu: Codd 7666. Bergville: Oliviers Hoek Pass, Medley Wood 3515

(K, NH); Upper Tugela, Gillett 1146. Klip River: Ladysmith, Acocks 9987, some

leaves 5-foliolate. Estcourt: West 1844 (specimen with fruits, leaves very small);

Acocks 9939, some leaves 5-foliolate; Colenso, Hutchinson 1861; Mooi River, Mogg

7218, 5-foliolate leaf. Nkandhla: Qudeni Gerstner 2629 (NH). Pietermaritzburg;

Scottsville, Fisher 701 (superficially looks like J. fluminense); Umsindusi River, F.G.C.

669. Harding: Oliver 97 (NH). Umzinto: Dumisa, Campbelltown, Rudatis 1860 (L).

Without precise locality in Natal: Gerrard & McKen 626 (NH); Tintern, Maurice &

Evans 501 (NH); Cooper 1166 (K, BOL); Gerrard 280 (K).

O.F.S. — Without precise locality. Cooper 2704 (K).

561

Cape. — Port St. Johns: Hutchinson 1778. Umzimkulu: Clydesdale, Tyson 2017.

also at (K). Kentani: Pegler 1810, 1200, 605 (BOL); 2103 (BOL). Butterworth:

Hector 993 (L). Queenstown: Gwatyn, Galpin 8267. Stockenstrom : Barker 2901

(NBG); Katberg, Sole 400. Komgha: Flanagan, 858; Medley Wood 5166 (SAM),

East London: Wood 3390; Smith 3645; 3667; Breyer in Tvl. Mus. 16542; Barker

1415 (NBG); Rattray 184, rather small calyx (GRA); Hilner 139 (GRA); Vincent

Barker 3512 (NBG). Stutterheim: Rogers 12703. Keiskama Hoek: Story 2500.

Mkubiso Forest, Stayner 30 (GRA). King William’s Town: Tyson 2150; 2887

(SAM); Compton 19345 (NBG); Kei Road, Rogers 3209 (GRA); Tamache Heights,

Leighton 2645, 2830. Peddie: Leighton 2645 (BOL), near Keiskama River, Galpin

7691; Breakfast Vlei, Taylor 1715. Bedford: Acocks 17633; Hutton s.n. cited in

FI. Cap. under J. tortuosum (K); Theron 579. Victoria East: near Fort Hare, Alice,

Grand and Blenkiron 2705; Hogsback, Rattray 356 (GRA). Albany: Cooper 2712 (K);

Rogers 2182; Blaauwkrantz Bridge, Galpin 266 ; Fish River, Marloth 10869; Oaklands

Park, Galpin 281; Botha’s Hill, Rogers 3408 (SAM); Grahamstown, Britten 1560

(GRA). Port Alfred: Tyson in Govt. Herb. PRE. 12990; Hutton s.n. (GRA);

Salisbury s.n. (GRA). Somerset East: Zuurberg, Holland 303; Compton 20276

(NBG); Boschberg, MacOwan 1946, 3 sheets cited in FI. Cap. under J. tortusum

(SAM); Alexandria: Copeman 84. Uitenhage: Sunday Riv., Gill s.n. (K); Aloes,

Drege 3146; Prior s.n. Zwartkops River, E. & Z. s.n.; 496 (BOL); 231, cited in

FI. Cap. (K, SAM); 3372 (SAM); Addo, Drege b (K, L); Drege B. b (L); Wilman

Bol. Herb. 25407 (BOL); Enon, Thode A 2699; A 2698; Drege B. a., type gathering

of var. glabratum E. Mey (K, L); between Enon and Zuurberg, Drege a, (K, L). Port

Elizabeth: West 18 (K, GRA); St. George’s Strand, Long 852; Redhouse, Paterson

240 (GRA); /. L. Drege 1 13 (GRA); Long 853 (GRA); Algoa Bay, Cooper 2702 (K);

Kabeljouws River, Bolus 1667, cited under/, tortuosum in FI. Cap. (K, BOL). Ladismith:

Suar Mts. Liebenberg 703. Without precise locality in Cape Province: Castelnau 462;

E. & Z. s.n. (GRA); Barber 23 (GRA); E. & Z. 286 (L); Ecklon s.n. (L); E. & Z.

s.n. (L); British Kaffraria, Hutton s.n., cited in FI. Cap. under /. tortuosum (K).

Not known to occur outside South Africa.

J. angulare Vahl and J. capense Thb. are synonymous and since both were published

during 1794, and more exact dates cannot be found for the publications, the author

who first cites one of these as a synonym of the other must be followed (International

Rules, Art. 67). Willdenow in 1797 cited J. capense as a synonym of J. angulare :

therefore, of these two the latter, /. angulare Vahl, is the correct name.

With regard to the pubescence, in this species some of the plants are completely

glabrous or appear glabrous with sparse pubescence in parts while others are

conspicuously tomentulose. The examination of a range of material shows that the

tomentulose specimens are not restricted to any particular region and that in some

places both pubescent and glabrous forms occur near each other, such as two collections

of Thode at Enon., A 2699 (glabrous) and A 2698 (tomentulose), and Drege B. b and b

from Addo. It was found too that in some cases on the same specimen both glabrous

and tomentulose twigs occurred. On this account a species or variety separated from

J. angulare on the presence or absence of pubescence cannot be maintained. This

applies to J. natalense Gilg & Schellb. and /. angulare var. glabratum E. Mey. Gilg and

Schellenberg write of J. angulare var. glabratum E. Mey. (vide Bot. Jb. 51: 84, 1914)

that since the type in Vahl’s herbarium in Copenhagen is glabrous this variety must

be dropped. Making no reference to Vahl’s original description which gives the

branchlets as villose, they then proceed to describe J. natalense (l.c. page 86) to take the

distinctly pubescent specimens. Although the cited specimens from Natal and

Pondoland have not been seen, judging from the description and several specimens

from the same locality there can be no doubt that they are J. angulare in the broad sense

as here defined. (The specimen cited from the Transvaal, Schlechter 11749, is

J. fluminense).

562

J. angulare is characterised by the angled branchlets seen best in the glabrous

specimens, the terminal rather compact cymes with fairly long pedicels to the lateral

flowers, and the calyx usually 3 mm. or more long. The leaves are 3-foliolate with an

occasional pinnately 5-foliolate leaf, and the leaflets are usually ovate with acarodomatia

on the under surface in the axils of the lower veins. In the pinnately 5-foliolate leaves

this species resembles the tropical african species, J. goetzianum Gilg, but in this latter

species the leaves are normally pinnate and the inflorescence lateral as well as terminal.

J. angulare has been collected among boulders on hill sides, near rivers, in scrub and

scrub forest, and in coastal bush. The flowering time is from October to January.

4. J. fluminense Veil. FI. Flumin. 10 (1825); op. cit. Atl. 1. t. 23 (1827); Dandy

in Kew Bull. 1950, p. 368 (1951); Turrill in F.T.E.A. p. 19 (1952). J. mauritianum

Boj. ex DC. Prod. 8: 310 (1844); Harv. ex Wright in FI. Cap. 4, 1 : 482 (1907).

J. schroeterianum Schinz Verh. Bot. Brand. 30: 256 (1888).

Woody climber or scrambler, sometimes shrubby. Ultimate branchlets usually

long, 10-20 cm. or more, densely shaggily pubescent to rather thinly tomentose

sometimes partly glabrescent. Leaves 3-foliolate; petioles patently spreading with

apex only upturned, 0-5-2 -2 cm. long, usually densely pubescent, rarely partly

glabrescent; leaflets densely to sparsely pubescent on both surfaces sometimes

sub-glabrous, ovate, gradually narrowing to an acute or rounded apex, acarodomatia

when present usually in axil of lower veins only, the terminal leaflets about 2 • 5 to 4 • 8 cm .

long and 1-4-3 cm. broad on a petiolule 1 to 1-9 cm. long, lateral markedly smaller

with petiolule 3-10 mm. long. Inflorescence terminal and axillary in the upper leaflets

only, together forming a broad cymose panicle, branches of the inflorescence densely

pubescent with crisped hairs; pedicels 1-5-6 mm. long, similarly pubescent. Calyx

shortly campanulate, pubescent like the pedicels or sub-glabrescent, 1 • 5-2 mm. long

with 5-6 teeth or lobes, up to 1 mm. long, oftern smaller or almost obsolete. Corolla

white, fragrant, tube 1-5-2 -5 cm. long, lobes 6-8, about 1-2 cm. long, 5 mm. broad.

Stamens inserted in upper half; filaments about 1-5 mm. long; anthers 4-5 mm. long,

acute at apex not reaching the mouth of the corolla tube. Style 2, lobed at apex, lobes

subexserted. Fruit brown to shiny black when mature, globose, about 7 mm. in

diameter, often twin berries developing.

Plate 5.

Type: A specimen from Santa Cruz, Brazil, where it had been introduced and

was already established in 1825. Types of synonyms: J. mauritianum, (specimen?),

Mauritius; J. schroeterianum Schinz, Schinz s.n., South West Africa.

South West Africa. — Kaokoveld: Kunene River, Hall 460 (NBG) .

Ovamboland: Omupande, Rautenen 790 (K); Okavango, van Dam s.n. (NBG);

Lugard 230 (GRA). Oshikango: Loeb 291. Caprivi Strip: Curson 1024. Ngamiland:

Curson 806, 144, 343.

Bechuanaland. — Maun: Van Son in Tvl. Mus. 28959: Kahako, Schoenfelder

SI 38; near Kachikau, Erens 382.

Transvaal. — Soutpansberg: Chipese, Verdoorn 2013; Lam and Meeuse 4894;

Dongola, Dyer 4311; Waterpoort, Moss 5311 (J); Fogwells, Smuts and Gillett 3134;

Mpefu, Smuts 2043; Maslangani, Breyer in Tvl. Mus. 16031 ; Elim, Obermeyer 29254;

Messina, Rogers 18737; Griffin Mine, Breyer 15630. Pietersburg: Mokeetsi,

Reynolds 113; Blaauwberg, Smuts and Pole Evans 940; Codd 8706; Chuniespoort,

Maguire 22049. Piet Potgietersrust : Magalakwin River Drift, Hutchinson 2665.

Letaba: Gravelotte, Galpin 13522; Selati Railway, Rogers in Tvl. Mus. 12985.

Carolina: Waterval Onder, Rogers 422 (GRA). Pilgrims Rest: Newington, Pole

Evans H. 18889. Nelspruit; Crocodilepoort, Dyke 5391, near Skukuza, Letty 46;

Johnson 449 (NBG); v. d. schijff 847 ; Pretorius Kop, v. d. Schijff 4 ; 492; Codd

563

5085; 5655. Barberton: Kaapmuiden, Mogg s.n.; Rogers in Tvl. Mus. 24290;

Boulder Siding, Mogg s.n. ; Komatipoort, Schlechter 11749 (cited in original description

under J. natalense Gilg and Schellenb.); Leendertz in Tvl. Mus. 2601; Rogers in

Tvl. Mus. 2601; 19352; Rogers 22202; 2677 (GRA); Komati Falls, Burtt Davy

360, cited in FI. Cap. under J. angulare, (K); Lomati River, Jenkins in Tvl. Mus. 9919;

Crocodile River Drift, Bolus 7848, cited in FI. Cap. (BOL); Highlands Creek, Moss

10702.

Swaziland.— Mbabane : near Bremersdorp, Compton 19726 (NBG). Asoko:

Dohse 210.

Natal. — Ingwavuma: Gerstner 4002 (NH); Codd 2082; Ward 2003.

Also occurs northwards to Nigeria, Abyssinia and Eritrea, in Mauritius, the

Seychelles and Arabia; naturalised in the West Indies and S. America.

Until quite recently this species was generally known in Africa, Asia and the

Mascarenes as J. mauritianum Boj. In the Kew Bulletin 1950 (published 1951) Dandy

pointed out that this name is antedated by J. fluminense which Velloso had described

in 1825. His specimen was from Brazil where the tropical African species had

established itself.

This species, which is so common and widely distributed in Africa, is recognised

by the following combination of characters: petioles patently spreading with the

apex only upturned; petiolules of lateral leaflets 3-10 mm. long, of terminal longer;

acarodomatia, if present, in axils of basal or lower veins only; inflorescence terminal

and in axils of upper leaves, often forming a broad, more or less flat-topped cymose

panicle; pedicels comparatively short and the calyx tube small, under 2 mm. long,

with short or almost obsolete teeth. On the whole the pubescence is fairly characteristic

being rather dense and of crisped hairs covering even, or especially, the branches of

the inflorescence, but this character can be misleading as some specimens, especially

in the eastern regions, are glabrescent and occasionally a specimen of J. angulare is

found with a similar pubescence.

These two species, J. fluminense and J. angulare, are closely related and are best

distinguished from each other by the very small calyx of J. fluminense, usually not over

2 mm. long, its broad inflorescence borne clear of the leaves, and short pedicels. The

distribution is distinct, J. fluminense being a tropical species reaching only as far south

as the north-eastern Transvaal and northern Zululand (Ingwavuma) whereas J. angulare

occurs commonly in the Cape and as far North as Volksrust in the Transvaal and

Utrecht in Natal. Evidently Gilg and Schellenberg confused these species because,

when describing J. natalense (the pubescent form of J. angulare, see note under that

species) in the Jahrbucher 1914 they cited also Schlechter 11749 from Komati Poort

which is definitely J. fluminense. In the same way Wright in the Flora Capensis cited

Burtt Davy 360 from Komatipoort under J. angulare and it too is J. fluminense.

Turrill in the Flora of Tropical East Africa, page 21 (1952), divides J. fluminense

into subspecies and varieties. It has not been found possible in practise to uphold

these. From the description, J. schroeterianum Schinz is also a synonym of

J. fluminense. This is supported by a specimen named schroeterianum — J.

mauritianum ”, at Kew, Rautanen 790, which is obviously J. fluminense as here defined.

With regard to the affinity of J. fluminense with another tropical species, /. abyssinicum,

see the notes under this latter species.

In South Africa, J. fluminense is found along rivers or in watercourses in dry

country, scrambling over small bushes or climbing in trees on wooded slopes in the

lowveld.

5. J. abyssinicum ( Hochst . ex) DC., Prodr. 8: 311 (1844); Bak. in FI. Trop. Afr.

4: 11 (1902); Gilg & Schellenb. in Bot. Jb. 51 : 84(1913); Turrill in F. T. E. A.

18 (1952). J. wyliei N.E. Br. in Kew Bull. 1909 p. 419.

564

Shrub, usually with profuse and high climbing branches, ultimate branchlets

long, terete, ascending. Leaves 3-foliolate; petioles patently spreading with apex only

ascending, 1 • 5-3 • 3 cm. long; petiolules of lateral leaflets 1-6 mm. long, of the terminal

10-20 mm. long; leaflets broadly ovate to orbicular-ovate, often shortly acuminate

at the apex, the terminal usually 2 -5-7 -5 cm. long and 2-4-5 cm. broad, glabrous

(in South African specimens), acarodomatia, when present, in axils of central and

upper as well as basal veins on the lower surface. Inflorescence terminal and lateral

in the axils of the leaves, cymose-paniculate, lax (in South African specimens), the

branches minutely puberulous; pedicels varying in length on the same inflorescence,

from 4 mm. (in terminal flowers) to 15 mm. (on lateral flowers), puberulous. Calyx

campanulate, puberulous, about 3 mm. long, shortly 5-6-toothed; teeth -5-1 mm.

long with wide sinuses between. Corolla white, suffused reddish purple without;

tube about 2-2 cm. long; lobes 5-6, about 1 cm. long and 6 mm. broad. Stamens

inserted in upper half of the corolla tube; filaments 1-5 mm. long; anthers included

5 mm. long, apiculate, apicule about 1 mm. long. Ovary sub-quadrate, about 1-5 mm.

long and broad; style long, terete; stigma 2-lobed, lobes oblong subcylindrical in

outline, about 4 mm. long. Fruit a twin berry, one often aborting, subglobose, about

3 cm. diam.

Plate 6.

Type. — Schimper 169, Aduwa, Abyssinia. Type of synonym: J. wyliei N.E. Br.,

Wylie in Herb. Medley Wood 8860, Nkandhla, Natal.

Transvaal.— Zoutpansberg: Entabini Forest near Louis Trichardt, Galpin 9679

(fruiting specimen). Pilgrimsrest: Marieps Kop, Marsh s.n.

Natal. — Nkandhla, Medley Wood 8860 (type number of J. wyliei N.E. Br.).

Also occurs northwards to Abyssinia and Uganda.

Among the specimens in South African herbaria there are no less than 6 sheets of

Medley Wood Herb. No. 8860, the type number of J. wyliei N.E. Br. These all agree

with the description of J. abyssinicum as given by Turrill in his revision of the Oleaceae

(FI. Trop. East Africa, 1952) and compare with several specimens from tropical Africa

so named in the National Herbarium. Following Turrill’s identification of these

tropical African specimens, J. wyliei N.E. Br. goes into synonymy. The gap in the

distribution can probably be explained by the lack of any extensive collecting in the

intervening high forests. This was to some extent proved to be so when, after being

requested, an Officer of the Department of Forestry, searching for this species, found

it in the forest at Marieps Kop in the north-eastern Transvaal. The material was sent

to the National Herbarium in April, 1954 and the specimen is cited above. This material

assisted in confirming the tentative identification of a fruiting specimen from a forest

near Louis Trichardt, Galpin 9679, as belonging to the same species and so another

link in the distribution is provided.

J. abyssinicum is usually found in high forest. It is closely related to J. fluminense,

from which it can be distinguished principally by the slightly larger calyx and the

axillary as well as terminal inflorescences. In the South African specimens it can also

be distinguished by the usually rather larger leaves which are thinner in texture and

glabrous, and the acarodomatia which, when present, occur in the axils of upper as

well as basal veins. One or other of these features may occasionally be found in eastern

forms of J. fluminense, but then the size of the calyx will be decisive, or the shape of

the inflorescence, which is more or less flat-topped in J. fluminense and rather oval in

outline in J. abyssinicum (compare plates 5 and 6).

6. Jasminum breviflorum Harv. in Wright in FI. Cap. 4, 1: 480 (1906). J. gerrardi

Harv. in Wright l.c. (glabrous variant).

565

Climber, usually rampant, occasionally shrubby. Twigs densely or sparsely

pubescent with patent crisped hairs or short stiff hairs, glabrescent. Leaves simple,

variable in shape and size, lanceolate-oblong, lanceolate-ovate, ovate to ovate-sub-

orbicular, up to 4-5 cm. long and 2-5 cm. broad, usually smaller, pubescent on both

surfaces, especially along the veins beneath, glabrescent in parts, sometimes glabrous;

petiole short, up to 3 mm. long, pubescent, articulated at or above the middle, often

at the apex, articulation not obvious. Inflorescence terminal on the twigs, usually

3-flowered and often also with solitary flowers in the axils of the upper pair of leaves ;

pedicels long, up to 1-5 cm. long, densely or thinly pubescent to glabrous. Calyx

thinly pubescent, rarely densely, sometimes glabrescent; tube about 3 cm. long;

lobes usually 5, very shallow (calyx appearing truncate), or up to 2 cm. long and

thickened and conduplicate, usually with recurved apicule but not produced into a

subulate erect apical portion. Corolla white, fragrant, tube 1 • 5-2 cm. long (in eastern

Cape up to 3 cm. long); lobes about 7, usually 1-2 cm. long and 2-5-5 mm. broad.

Stamens 2, inserted on the tube in upper half, included, filaments very short; anthers

about 3-5 mm. long. Ovary about 1 mm. long, truncate, 2-celled with 1 ovule in each;

style filiform 1-5 cm. long; stigma lobes 3-5 mm. long, exserted. Fruit a twin berry

but usually only one developing, sub-globose, about the size of a large pea turning

black when ripe; seed 1, exendospermous.

Plate 7.

Type: Burke s.n., Magaliesberg, near Pretoria. Type of synonym: J. gerrardi

Harv., Gerrard 1477, Nonoti Riv., Mapumula, Natal.

Transvaal. — Waterberg: near Naboomspruit, Galpin 494 M.; near Nylstroom,

Repton 3481; Bremekamp & Schweickerdt 2; Middelfontein Station, Galpin 13989;

Mogg 17557; Warmbaths, Leendertz 6542; near Warmbaths, Bolus 12113; Burtt

Davy 2182; Thode A1729; N.E. of Thabazimbi, Codd 4793. Rustenburg: Nation

346 (BOL); Zwartruggens, Sutton 809; near Rustenburg, Hutchinson 2933;

Buffelspoort, Turner 39; Assen, van Nouhuys in Tvl. Mus. 31046. Brits: Welgevonden,

Mogg s.n., Beestkraal, Jenkins in Tvl. Mus. 6943. Pretoria: Magaliesberg, Burke s.n.,

holotype (K); Hartebeespoort Nek, Prosser 1578; Hamanskraal, Repton 3544.

Natal. — Without precise locality, Buchanan 17 (SAM). Dundee: Vants Drift,

Codd 1517? poor. Msinga: Tugela Ferry, Galpin 14789; Dyer 4381 poor.

Babanango: between Babanango and Entonjaneni, Umhlatuzi Bridge, Lawn 2040

(NH). Greytown: Muden Valley, Galpin 14737. Eshowe: Nkwaleni Valley, Lawn

362 (NH). Mtunzini, near Mtunzini, Lawn 1631 (NH). Mapumula: Nonoti, Gerrard

1477, holotype (K); Gerrard & McKen 1477 (NH). Camperdown: Rehmann 7706

(K); Acocks 10855; Umlaas Drift, Medley Wood 1827, cited in FI. Cap. under

J. streptopus (NH). Maritzburg: Rajah 4 (NH); nr. Maritzburg on Durban Road,

Compton 4417 (NBG). Umzinto: Dumisa, St. Mcihael, Rudatis 2123 (NH).

Cape. — Tembuland, on the Kei River, Bolus 25405 (BOL). Butterworth: Kei

Bridge, Flanagan 1204, 1205. Queenstown: Junction Farm, Galpin 8145; Spence in

Galpin Herb. 8038. King William’s Town: Galpin 5912; 5913 (GRA); Sim 2724 ;

Tyson 2232. Peddie: Keiskamma River, Sim 6281; Kaffir Drift, Compton 17824

(NBG).

Also found in Portuguese East Africa and may occur in other territories north

of South Africa.

Harvey’s species, J. gerrardii and /. breviflorum, were published posthumously

by C. H. Wright in the Flora Capensis at the same time, the former from Natal and the

latter from the Transvaal. They were classed together on the simple leaves and the

short to almost no calyx lobes, and distinguished from one another by the glabrous

twigs and leaves of J. gerrardii, as opposed to the very pubescent ones of J. breviflorum.

According to the description these features seemed to be supported by the shape of

566

the leaf which was described as “ ovate-lanceolate, acute at each end ”, in the Natal

species and “ ovate, obtuse or acute ” in the Transvaal species. An examination of a

number of specimens from the Transvaal, Natal and the eastern Cape showed (1) that

both leaf-shapes could be found among pubescent specimens as well as among glabrous

specimens ; (2) although the glabrous specimens mostly came from Natal, their habitat

was not different from the pubescent form in that Province; and (3) that no constant

supporting feature to that of pubescence could be found. It was also found that some

specimens from the eastern Cape were pubescent in part only while on a Flanagan

specimen one unattached branch is glabrous and the other pubescent. These species

can therefore not both be upheld, not even with one as a sub-species or variety of the

other. Since they were published at the same time and, as far as can be ascertained,

neither has been sunk before, J. gerrardii is here sunk under J. breviflorum. The choice

of the name to be retained, although it is second in order on the page in Flora Capensis,

was made because the pubescent varient, J. breviflorum, is by far the more common and

widely spread.

As the name implies, the flowers are shorter in this species than in most others,

but it is not a reliable character since several specimens from the eastern Cape have

long flowers. Whether this is a regional variation or whether there is hybridization

with the long-flowered eastern Cape species, J. angulare, is not known. In spite of

the variations, these specimens with the almost truncate to shortly lobed calyces form

a recognisable group. The calyx-lobes when present are rather thickened and

conduplicate and usually have a recurved apicule, but are never produced into an

erect subulate apical portion. The petiole is short and articulate in the upper half.

From description the nearest ralative in tropical Africa would be J. stolzeanum

Knobl. and our species should be compared with it to ascertain the differences. If

they are not distinct J. breviflorum Harv. in C.H. Wr. takes priority over J. stolzeanum

Knobl. (1936).

7. J. glaucum (L.f.) Ait. Hort. Kew. ed. 1, 1: 9 (1789); Harvey ex Wright in FI.

Cap. 4, 1: 480 (1906), in part, excl. var. parvifolium E. Mey, and citations from

eastern Cape, Natal and Transvaal. Nyctanthes glauca Linn. f. Suppl. 82 (1781);

J. ligustrifolium Lam. Encycl. 3: 218 (1789); J. glaucum var. lanceolatum et var.

latifolium E. Mey. Comm. 1, 2: 13 (1837); DC. in Prod. 8: 305 (1844); Harv.

ex Wright in FI. Cap. 4, 1: 481 (1906).

Shrubs rarely scandant, glabrous in all parts. Leaves simple, glaucous, usually

about 3 times as long as broad, lanceolate, lanceolate-elliptic, lanceolate-ovate, slightly

narrowed to a rounded mucronate apex, or acute, sometimes long acuminate, 3-6-5 cm.

long, -7-2-5 cm. broad, narrowing gradually to the 3- (or more) nerved base and

then abruptly narrowed into a petiolule like basal portion; petiole short, 1-5 mm.

long but with the cuneate leaf-base appearing about 1 cm. long, articulate at the apex

but articulation often appearing to be about midway depending on the length of the

leaf-base. Inflorescence terminal on ultimate branchlets, occasionally also in axils

of upper pair of leaves, normally 3-flowered; bracts setaceous, usually about 1 cm.

long; pedicels 3-12 mm. long. Calyx with usually 5-7 subulate lobes; lobes as long

as or longer than the tube, sinuses rounded. Corolla white, fragrant; tube usually

2-2-5 cm. long; lobes 6-8 in number, 1 -7-2-5 cm. long, up to 5 mm. broad. Fruit

black, a twin berry, one often aborted.

Plate 8.

Type: Thunberg, Langekloof, Caledon*. Types of synonyms: J. ligustrifolium

Lam. Cape of Good Hope, without precise locality; J. glaucum var. lanceolatum

E. Mey. Drege, Olifants Riv., Van Rhynsdorp; var. latifolium E. Mey. Drege,

Clanwilliam.

* Photograph of type in Uppsala herbarium seen since going to press.

567

Cape. — Van Rhynsdorp: Ebenezer, Drege s.n., cited in FI. Cap. under “ var.

lanceolatum E. Mey.”. Clanwilliam: Drege s.n. “var. latifolium ” (L); 6 miles

N.W. of Clanwilliam, Leipoldt s.n.; 4451 (BOL); 1 mile N. of Keerom, Pillans 8675;

near Hex River; Compton 18837 (NBG); Pakhuis Pass Middlemost s.n. (NBG);

beyond Pakhuis Village, Compton 4752 (NBG); near Olifantsriver, Zevher s.n. (BOL);

Krakadouw Poort, Esterhuizen 12269 (BOL). Piquetberg: Piqueniers Kloof, Dickson

in herb. Bolus 5699 (BOL). Tulbagh (or Worcester?): Witsenberg, E. & Z. s.n.

(77. 10); Tulbaghkloof, Zeyher 1149 (K, PRE). Worcester: Hex River, Leighton

2379 (BOL); Pillans 5271 (BOL). Paarl: Gt. Drakenstein Mts., Esterhuizen 9526;

Wasserfall 721 (NBG). Caledon: Witwater, Baur 8742.

The distribution is limited to the south-western Cape.

It is obvious that this species as treated in the Flora Capensis is a mixture of

J. glaucum and J. multipartitum. The specimens cited there from the south-western

Cape, that is from the Tulbagh, Van Rhynsdorp and Clanwilliam districts, answer

best to the original description of J. glaucum (L. f.) Ait. and form a group easily

distinguished from the rest. The group is characterised by the wholly glabrous plants,

the glaucous leaves which are usually about 3 times as long as broad, mostly

lanceolate-oblong to lanceolate-ovate in shape, and the normally 3-flowered

inflorescences.

Among the other specimens cited in the Flora Capensis, that is those from outside

the winter rainfall area, Krauss 458 is the type number of J. multipartitum Hochst.,

the very next species described in that work, and Cooper 378 and Galpin 266 are also

J. multipartitum Hochst. The other two specimens cited have not been seen, namely

Burchell 3657 and Wilms 925, but, judging from the localities where they were collected

they too are in all probability J. multipartitum.

The collector’s notes, on the herbarium specimens examined, are very meagre;

not one mentions the habit of the plant or the glaucous leaves. Thunberg, the collector

of the type specimen, is quoted by Lindley in the Botanical Register t. 2013 (1837)

as saying that it, meaning J. glaucum, is “ as high as a man in Lange Kloof (Caledon)

by the great stream called Zonder End and in the neighbourhood of the Brederiver ”.

The plate, however, does not convincingly depict J. glaucum. It seems rather to be one

of the rare 3-flowered specimens of J. multipartitum.

8. J. multipartitum Hochst. in Flora 27, 2: 825 (1844); De Wildeman, Ic. Sel. Hort.

Hen. 4 t. 134 (1903); Wood and Evans in Natal Plants 4, PI. 328 (1906); Harv.

ex Wright in FI. Cap. 4, 1 : 480 (1906); J. glaucum Harv. ex Wright l.c. pro parte,

non Ait.; J. glaucum var. parvifolium E. Mey. Comm. 1, 2: 173 (1837); DC.

Prodr. 8: 305 (1844).

Shrub 18 inches to 9 feet tall, sometimes scandent, twiggy. Twigs puberulous,

usually short about 2 cm. long. Leaves simple, usually glabrous, suberect, shining,

green, often drying blackish and brittle, very variable in size and shape, sometimes

aborted or very small on the flowering twigs, oblong, ovate-oblong or ovate-lanceolate,

1 -5-3-5 cm. long (usually about 2-5 cm. long) 4-28 mm. broad (usually about 9 mm.

broad), rounded or sometimes acuminate and acute at apex, rounded at base and

abruptly narrowed into a petiolule-like base; petiole glabrous or minutely puberulous,

appearing articulated in the centre or above or below, 1-5 mm. long. Inflorescence

terminal on short lateral twigs, usually 1 -flowered., rarely 3-flowered; pedicels shorter

than the calyx, 1-5 mm. long, glabrous or rarely minutely puberulous. Calyx glabrous

or rarely minutely puberulous in parts; tube 2-3 mm. long; lobes 5-9, margins

inrolled, appearing subulate with wide sinuses between them, 3-5 mm. long sometimes

minutely puberulous at least on margins. Corolla white, pink flushed or lobes dorsally

brick red, fragrant; tube 2-3 cm. long; lobes about 11, more or less 1-7 cm. long

568

and 4 mm. broad. Stamens inserted on tube in upper half, included ; filaments 1 • 5 mm.

long; anthers 3-5 mm. long, apiculate. Ovary 2 mm. long quadrate, truncate; style

2-lobed at apex, eventually exserted, lobes oblong about 2 mm. long. Fruit a twin

berry joined at base only, but often only one-half developing, oblong to oblong-globose,

slightly compressed, about 1-1-5 cm. long, 8-10 mm. broad and 4 mm. thick, green,

shiny, turning black when ripe.

Plate 9.

Type: Krauss 458, near the bay, Durban, Natal. Type of synonym: J. glaucum

var. parvifolium E. Mey., Drege , Bothasberg, Albany.

Transvaal. — Pietersburg: near Boyne, Gerstner 5345 (poor specimen). Potgieters-

rust: Leendertz 6595; 7406; Makapan Valley, B.M. in Moss Herb 22664 (J).

Waterberg: Naauwpoort, Galpin 133133; Naboomspruit, Galpin 13988; near

Nylstroom, Story 1540; Krantzberg, Prosser 1713; near Warmbaths, Sidey 1343;

Marais 551, 552, 553; near Rooiberg, Sandspruit, Forssman 262. Lydenburg:

Sekukuniland, Barnard 147; Parys, Barnard and Mogg 756. Pilgrimsrest; Bushbuck-

ridge, Forest Officer 28; Branddraai, Young A641. Nelspruit: Breyer 18001;

Pretorius Kop, Codd 6173; v. d. Schijff, 961; Shabin, v. d. Schijff 670; Clobe,

v. d. Schijff 1171 ; Near Nelspruit, Liebenberg 2632 ; Plaston, Holt 92; White River,

Rogers in Tvl. Mus. 20528; Rogers 2348 (GRA).

Swaziland. — Stewart 8851; Lebombos, 25 m. S. of Stegi, Hornby 2819.

Natal.— Without precise locality, Kate Saunders in Bol. Herb. 25402 (BOL);

Medley Wood 356 (BOL). Nongoma: Mkusi River Valley, 5 m. N.E. Mkuzi, Codd

2058. Hlabisa: St. Lucia, Lansdell in N.H. 37644 (NH); Lower Umfolosi: 11m.

S.W. of Empageni, Codd 1879. Nqutu: Amangi Valley, native for F. Bayer in NH.

21766 (NH). Babanango: Umhlatuzi Valley, Melmoth Road, Lawn 939; 1280,1890

(NH). Mtunzini: Lawn 1777 (NH); Ginginhlova, Galpin 133129; Lawn 1199;

1200 (NH); Myrlyn, Inyoni, Johnson 478. Umvoti: Rietvlei, Craigie Burn, Fry in

herb Galpin 2751. Weenen: Pentz 539; West 1157; Sandspruit, Acocks 10560.

Estcourt: near Muden, Repton 1192; Research Station, West 418; Mooi River

Valley, Sutherland s.n. Pietermaritzburg: Killick and Marais 1999; Table Mountain,

Bond 1250 (NBG); Durban Road, Barker 442 (NBG); Albert Falls, Comins 498;

Tugela River, Gerrard 264 (NH). Camperdown: Cato Ridge, McClean and Ogilvie

in NH. 27921 (NH). Durban: Medley Wood 94 (SAM); 12411; s.n.; Marloth

4176; woods near Durban Bay, Krauss 458, type number (K); Berea, Medley Wood

150 (GRA); Clairmont, Medley Wood 11195 (NH); “ Claremont”, Schlechter 2951;

Umhlanga, Medley Wood 10520 (NH); Springfield, Indian in NH. 17818 (NH);

Bellair, Lansdell in NH. 34288 (NH). Umzinto: Umkomaas, Tyson 2666 (SAM);

Drift Halt, Lansdell in NH. 34710 (NH).

Cape.— Pondoland, without more precise locality, Bachmann 1030 (K). Bizana:

Bruce 441 ; Mount Ayliff, Story 567. Port St. Johns: Edwards in Moss Herb. 8427 (J);

Umtata: Umtata River near falls, Schonland 3775 (GRA); Kentani: near Kentani,

Pegler 1253 (NBG). Komgha: near Komgha, Flanagan 59; 479; Kei Mouth,

Compton 17685 (NBG). Queenstown: Gwatyn, Galpin 8266. East London: John

Wood 3372 (poor); King William's Town: Pirie Forest, Muden Dam, Acocks 9290;

Buffalo River Valley, Galpin 5912; Keiskama Hoek, Cooper 378 (BOL). Stockenstrom :

Katberg, Sole 401; Fort Armstrong, Martin 158 (NBG). Elands River, Scully 84

(SAM). Victoria East: near Kweza Siding, Acocks 1115; near Alice, Gillett 4588;

Hutch and Dyer 1688; Dyer 1796 (GRA); Barker 1414 (NBG). Peddie: Kafir

Drift, Compton 17819 (NBG). Albany: between Blaauwkrantz and Kowie Poort,

Burche/I 3657, cited under J. glaucum in FI. Cap. (K). Koonap Heights, Dyer 724;

Alicedale, Cruden 143 (GRA); Plutos Vale Barker 2810 (NBG); Blaauwkrantz

Bridge, Galpin 266 (GRA); Blaauwkrantz, Hutchinson 1569; Fish River Heights,

569

Hutton s.n. (K) and in Bolus Herb. 2540 (BOL); Fish River Valley, Lotsy and Goddyn

6 (L); Fort Brown, Lotsy and Goddyn 727 (L). Bathurst: Trapps Valley, Daly 626

(GRA). Somerset East: Zuurberg, Compton 20256 (NBG). Uitenhage: near

Uitenhage, Prior s.n.; Thode A678 (NH); West 19 (GRA); Enon, Thode A1108;

A2697; A678; Sand Fontein, Burke s.n. (K).

Also occurs in Portuguese East Africa and possibly in Southern Rhodesia.

In South Africa this species has quite a wide distribution and it is often found

cultivated in gardens. It occurs frequently in the Waterberg district of the Transvaal,

and from there eastwards to Swaziland, Natal and eastern Cape as far as the Uitenhage

district. It is characterised by the usually single flower on a rather short pedicel (under

1 cm. long) at the end of short slender lateral twigs which are characteristically shortly

and densely puberulous. The leaves on these twigs are usually young or stunted,

smaller than those on the sterile branches and they frequently dry black and brittle

giving the herbarium specimens a characteristic appearance. The sweetly scented

flowers have very white radiating corolla lobes which are dorsally, wine-coloured or

maroon, the buds and corolla tube being the same maroon colour.

In the Flora Capensis this species is partly confused with J. glaucum Aiton (see

notes under that species), and the type number of J. multipartitum, Krauss 458, is

cited under the latter species; on the other hand Burchell 3657 (K) is wrongly cited

under J. glaucum. When correctly delimited, these species are clearly distinct. Not

so clear is the difference between some of the northern specimens of J. multipartitum

and some of J. stenolobum Rolfe. The main difference between these species is the

obvious pubescence of J. stenolobum and the less conspicuous but characteristic short

dense pubescence of J. multipartitum. In some of the Transvaal specimens where the

distribution of the two species overlap there may, when material is not adequate, be

some difficulty in distinguishing these species (see notes under J. stenolobum ).

Specimens from Rhodesia and Portuguese East Africa which, except for the

inflorescences being more often 3-flowered, fit into J. multipartitum as treated here,

fall into the broad concept of J. meyeri-johannes Engl, as given in the Flora of Tropical

East Africa (1952). In my opinion they should be classified as J. multipartitum Hochst.,

which seems to be distinct from the tropical species, judging from some of the specimens

seen. If not distinct, J. multipartitum is the older name and would have to stand.

9. J. stenolobum Rolfe in Oates, Matabele Land, ed. 2: 403 (1889) Baker in FI.

Trop. Afr. 4, 1 : 4 (1902); Harv. ex Wright in FI. Cap. 4, 1 : 481 (1907); Gilg &

Schellenb. Bot. Jb. 51: 91(1913); Turrill in F.T.E.A.: 24 (1952).

A bushy or straggling shrub, sometimes scrambling or climbing. Branches terete,

pubescent, glabrescent; ultimate twigs 1-5 cm. long, rarely longer, densely or

persistently pubescent to tomentulose with spreading or reflexed hairs. Leaves very

variable in size and shape, from ovate-lanceolate or elliptic to ovate or ovate-oblong,

usually 1-4 cm. long and 0-4-1 -8 cm. broad, pubescent on both surfaces, without

acarodomatia ; petioles pubescent, 3-8 mm. long articulated near the base (rarely

higher up). Inflorescence terminal on the short ultimate twigs, usually 1-flowered

(occasionally 2-3-flowered) ; pedicels pubescent, usually short, 1-7 mm. long,

occasionally up to 1 cm. long. Calyx pubescent, pubescence variable, hairs short or

long and curly; tube 2-3 mm. long; lobes 6—13, subulate with U-shaped sinuses,

2-5-6 mm. long. Corolla white (buds red, fide Comins), fragrant; tube 2-2-5 cm.

long; lobes 7-12, 1-3-1 -7 cm. long, 3-5 mm. broad. Stamens inserted in upper half

of the tube, included, anthers up to 5 mm. long, apiculate. Ovary sub-quadrate,

compressed, about 2 X 1-5 mm. ; style 2-lobed at apex. Fruit a twin berry (one

sometimes aborting), narrowly oblong, black, those seen about 1 cm. long and 6 mm.

broad.

Plate 10.

57Q

Type: Oates, Matabeleland, without more precise locality.

Bechuanaland Protectorate. — Kazungula, Miller B155; B383. Chobe,

Miller B1121. Ngamiland, Curson 209; 210; 219.

Transvaal. — Soutpansberg: N.W. of Soutpan, Obermeyer, Schweickerdt and

Verdoorn 92; Wyllies Poort, Pole Evans 1783; Punda Maria, Lang in Tvl. Mus.

32108; 32139; Baiandbai, Lang in Tvl. Mus. 32264; Dzundweni Hill, Codd and Dyer

4601. Petersburg: Leipsig, Blaauwberg, Bremekamp and Schweickerdt 60; Chunies

Poort, Pole Evans H. 19446. P.P. Rust; Crawley s.n. (or in Tvl. Mus. 7159);

Swerwerskraal, Rowland 119; 126. Waterberg: Nylstroom, van Dam in Tvl. Mus.

20750. Middelburg: Middelburg, Rogers 24864. Lydenburg: Ohrigstad-Branddraai

Rd., Young A 608; A 609; Waterfall, Codd and Verdoorn 7611; Nelspruit: Sigaas,

K.N.P., van der Schijff 1302. Barberton: Rogers 24879; Thorncroft 4964; Berea

Ridge, Galpin 621.

Natal. — Entonjaneni: N. of Nkwaleni, Codd 1840. New Hanover: Albert

Falls, Comins 497. Mzinto: Umpanpanyoni, Rudatis 2010.

Also occurs in Portuguese East Africa, the Rhodesias, Nyasaland and Tanganyika.

In South Africa J. stenolobum is most like J. multipartitum from which it is mainly

distinguished by the conspicuous pubescence on the twigs, leaves and calyx. The

pubescence is mostly rather dense, tomentulous, with spreading or recurved, often

crisped, hairs which vary somewhat in density and length. J. multipartitum on the

other hand appears to be glabrous but is persistently, minutely and densely, puberulous,

at least on the twigs. Between these species it is not a matter of one being glabrous

and the other not, but a matter of a different type of pubescence and that, in my opinion,

justifies upholding the two species, which in many other respects are so very similar.

Both bear solitary flowers (or occasionally more) at the apex of short lateral twigs, on

comparatively short and stout pedicels. The calyx lobes are usually definitely longer

than the tube, the petioles are conspicuously articulate, and the leaves have no

acarodomatia in the axils of the veins below. The leaves of J. multipartitum are

extremely soft, the young ones shiny in the sun whereas in J. stenolobum the leaves

are usually so pubescent that they give a different impression. From collector’s notes

it is impossible to discover whether the flowers of J. stenolobum have the dorsal face

of the lobes and the tube a maroon colour as in J. multipartitum.

The areas of distribution of the two species overlap in the Transvaal and Natal

and from these regions, at certain stages of growth and when the specimen is inadequate,

it may be difficult to decide to which of the two a specimen belongs, but on the whole

they are very easily distinguished. It has not been found, to date, that J. multipartitum

occurs in the Kalahari sand veld such as in Bechuanaland, Ngamiland, Matabeleland,

the Soutpansberg region and so forth, where J. stenolobum is frequent, and in the same

way the latter species has never been found in the eastern Cape where J. multipartitum

abounds.

Outside South Africa, in the more tropical regions, J. stenolobum has a tendency

to scramble or climb more freely and to bear more than one flower in an inflorescence.

The legend on a specimen from Cloma, Northern Rhodesia, reads: “ Berries

brown, produce state of coma when swallowed and inability to swallow

10. J. streptopus E. Mey. aggregate species.

(a) var. streptopus.

(, b ) var. transvaalense (Sp. Moore ) Verdoorn stat. nov.

Climbing or scrambling shrub; branchlets often slender, pubescent, hairs of

different lengths, appressed or patent, straight or crisped. Leaves simple, rather thin,

oblong-elliptic, ovate-oblong or ovate acuminate, 1 • 5-7 cm. long and • 8-3 cm. broad,

rounded or acute at the apex, mucronate, pubescent on both surfaces, especially along

the midrib, forming a fringe on each side of it, usually with acarodomatia on the lower

571

surface in the axils of the lateral veins; petioles pubescent, short, 1-4 mm. long, rarely

5 mm. long, obscurely articulate near the apex and somewhat twisted. Inflorescence

terminal on ultimate branchlets, 1-5-flowered, often 4-flowered. Pedicels thinly

pubescent with crisped or patent hairs, glabrescent, usually 5-20 mm. long. Calyx

pilose, glabrescent; tube 1-2-5 mm. long; lobes 4-7, subulate from a triangular

base, subulate portion varying in length (with age of flower?), sometimes very short

1-5-7 mm. long. Corolla white, fragrant; tube slender 2-3 cm. long; lobes 6-8,

about 1-6 cm. long and 3 mm. broad. Fruit a twin berry often one only developing,

globose, brownish to black, about 7 mm. diam.

Plates 11 and 12.

The species is characterised by a combination of the following features: Crisped

pubescence on the twigs ; rather short petioles articulated near the apex ; acarodomatia

on lower surface of leaves usually in axils of upper as well as basal veins; comparatively

long slender pedicels; calyx lobes subulate from a triangular base, the subulate portion

varying in length (very short or up to 7 mm. long), sometimes on the same specimen.

The character which evidently gives the species its name, the slightly twisted petiole,

is not restricted to J. streptopus but may be found to some extent in almost all the

species.

The calyx lobes in this species call for some special mention. In the South African

species already dealt with these lobes are either short and thick or clearly subulate. In

J. streptopus, and some tropical species like it, there seems to be a combination of the

two types of calyx lobes, the basal portion being fairly short and thick and produced

at the apex into an erect subulate portion. This subulate portion varies considerably

in length sometimes on the same species. It has been suggested that on young flowers

the subulate portion is absent or short and elongates with the age of the flower. While

this is borne out in some specimens it is not always the case and the variation seems

to occur haphazardly.

The South African specimens of this species fall into two groups which are here

given varietal rank and are described below. It is probable that some of the closely

related species in tropical Africa are also no more than varieties of the species. Mention

is made of two of these tropical species, one under each of the varieties below.

This is an example, as pointed out in the introduction, of the authority for an

E. Meyer species being given as “ E. Mey. ex DC.” whereas E. Meyer described it

in his “ Commentariorum ” in 1837.

( a ). var. streptopus; J. streptopus E. Mey. Comm. 1, fasc. 2: 173 (1837); DC. Prod.

8: 307(1844); Wood & Evans in Natal Plants 1, tab. 50 (1899); Harv. ex Wright

in FI. Cap. 4, 1 : 481 (1907).

Type: Drege, “margin of woods near Port Natal”.

Natal. — Durban: Medley Wood 900 (NH); 10113 (NH); Schlechter 2967

(GRA); Berea, Medley Wood 3786 (NH); Stella Wood, Lavoipierre 95; 96; N.E.

of Waterfall, Dohse 81. Inanda: Medley Wood 1191, cited in FI. Cap.

Probably also occurs in Southern Rhodesia and Portuguese East Africa.

The typical variety (see plate 11) is characterised by the pubescent, oblong-elliptic

to ovate-oblong leaves, mostly 4-7 cm. long and 1-6-3 cm. broad, and the 3-5-flowered,

often 4-flowered, inflorescences, terminal on the ultimate branchlets. From the material

examined it would seem that it is very localised for all the S. African specimens seen

come from the Durban area. Outside South Africa a couple of specimens from Umtali

and one from Portuguese East Africa may prove to be this variety. On the other hand

they may belong to the closely related J. pauciflorum Bth. which differs in having the

inflorescences in the axils of the lateral leaves as well as terminal. The above specimens

may be merely portions of the plant where the lateral inflorescences are not present.

On the other hand it is possible that the lateral inflorescence is not a specific character

572

and that J. pauciflorum is no more than another variety of J. streptopus, with the

tendency for the inflorescence to be axillary as well as terminal. Further study of

tropical specimens should settle this point and may show that a few more closely related

species in those regions are no more than varieties of J. streptopus, which is the oldest

described species in this group.

( b ). var. transvaalense (Sp. Moore) Verdoorn stat. nov. J. transvaalense Sp. Moore

in J. Bot., Lond. 56: 10 (1918).

Type: Rogers 18108, Modjadjes, Letaba, Transvaal.

Transvaal. — Soutpansberg: Louis Trichardt, Breyer 22718; 22719; Malta

Gorge, Junod 4437 ; Elim, Obermeyer 29253 ; Makonda, Westphal in Tvl. Mus. 29114;

Pepeti Falls, Smuts and Gillett 3233; 3180. Pietersburg: Blauwberg, Codd & Dyer

9128; 9168; Woodbush, Bolus 11117 (BOL); Rehmann 5952 (BOL); Eliovson in

Wits Herb. 26953; The Downs, Junod 4158. Belfast; Dullstroom, Galpin 13297.

Ermelo: Mavieriestad, Pott 4909. Letaba: Selati Mission, Gerstner 5510. Pilgrims

Rest: Sabie Hoek, Burtt-Davy 1519. Lothian, Forest Officer 22; Bushbuck Ridge,

Smuts and Gillett 2351. Piet Retief: Sidey 2054; Pole Evans 16.

Natal. — Eshowe: on margin forest, Gerstner 3886 (NH); Hlinza Forest edge,

Lawn 1335 (NH); Emkazeni Forests, Fernando 10399 (NH). Camperdown: Fairfield,

Rudatis 2042 (NH). Pietermaritzburg: Townhill, F.G.C. 339; Howick, Shafton,

Hutton 1180 (GRA).

Cape. — Bizana: Acocks 12230; Emagushen, Tyson 2815; 3150 (SAM).

This varietal form has not been collected outside South Africa, but in the tropical

regions there are very closely related species.

J. streptopus var. transvaalense (see plate 12) is more widely spread than the typical

variety. It is distinguished by the smaller, ovate-acuminate leaves, 1 • 5-3 • 5 cm. long

and -8-1-4 cm. broad, and the inflorescences frequently being 1 -flowered. Although

these characters give the specimens a distinct appearance, they are not fixed and therefore

cannot be considered to be specific. A few of the leaves often become large and take

on the shape of the typical form while the inflorescence occasionally is more than

1 -flowered. In the shape of the leaf and the 1 -flowered inflorescence it resembles

J. swynnertonii Sp. Moore from Portuguese East Africa, the Rhodesias and Tanganyika,

but this latter species has slightly larger leaves with more prominent veins. It is

doubtful whether J. swynnertonii (described in 1911) can be anything more than another

variety of J. streptopus.

3. OLEA.

Linn., Gen. PI., ed. 5, 8 (1754); Benth. & Hook, in Gen. PI. 2, 2: 679 (1876); Harvey

ex Wright in FI. Cap. 4, 1: 485 (1907); Phillips Gen. S.A. FI. PI. ed. 2: 572

(1951) pro parte.

Shrubs or trees, sometimes forest trees up to 100 feet tall. Leaves opposite, entire,

lepidote scaly, sometimes minutely and sparsely so. Inflorescence trichotomous panicles,

many flowered, axillary or terminal. Calyx persistent, cucullate, 4-toothed or shortly

4-lobed. Corolla united into a short tube, 4-lobed, lobes rather broad forming a

sub-globose bud, ultimately reflexed with tips ascending, deciduous. Stamens 2,

inserted on the corolla-tube, filaments short, anthers dorsifixed, relatively large with

extrorse dehiscence. Ovary subglobose, narrowed into a short style, stigma terminal,

bi-lobed ; ovules 2, pendulous. Fruit, a drupe with a thin fleshy layer, endocarp rather

hard with large seed cavity; seed usually solitary, endosperm present, cotyledons

thin.

573

Type species: Olea europaea Linn., the cultivated Olive.

The description by Phillips in Genera of S.A. Flowering Plants was drawn up

from both Olea and Linociera species in South Africa and therefore only partly applies

to Olea. The reasons for keeping Linociera separate are given in the notes on that

genus.

The leaves of Olea species never have acarodomatia in the axils of the veins below ;

the inflorescences are usually many to very densely many-flowered and the flowers

are small, remaining in the bud stage rather long, with the buds subglobose, and, in

all Oleas examined, the ovules were attached at the apex and endosperm was present

in the seed.

Key to Species.

Inflorescence axillary as well as terminal :

Leaves usually densely covered with small scales below, linear-lanceolate

to narrowly oblong-elliptic 1. O. africana.

Leaves fairly sparsely and very minutely scaly below appearing minutely

pitted, lanceolate-elliptic to elliptic, usually broadest about

the middle and from there tapering to base and apex;

ultimate branchlets slender usually whitish, in parts

obviously quadrate 2. O. woodiana.

Inflorescence terminal and axillary only in the two upper pairs of leaves:

Leaves linear-oblong, up to about 1 cm. broad, broadest in upper

two-thirds, long cuneate to the base 3. O. exasperata.

Leaves of various shapes and sizes but never linear-oblong, usually

more than 1 cm. broad 4. O. capensis

(aggregate sp.).

Fruit subglobose to oblong-elliptic, up to 1 cm. long; shrubs or

trees up to 50 feet tall ; leaves crowded on

herbarium specimens; inflorescence densely

flowered :

Leaves very variable often rounded at apex; midrib

prominent beneath from base to apex;

branchlet dark grey to greyish brown ; Cape

coastal regions 4a. subsp. capensis.

Leaves usually broadly elliptic, shortly acuminate to base

and apex; midrib prominent only in lower

half beneath, disappearing in upper half;

branchlets grey to whitish, leaves deciduous

with prominent leaf-scars. Transvaal and

Natal 4b. subsp. enervis.

Fruit large, oblong-elliptic about 1-5 cm. long; forest trees

sometimes 90 feet tall; leaves elliptic, tapering to

base and apex; petiole usually long and spreading

and leaves not very crowded on herbarium speci-

mens; inflorescence many flowered but not

densely so 4c. subsp. macrocarpa.

1. O. africana Mill. Gard. Diet. Ed. 8, n. 4 (1768); Adamson in Flora of the Cape

Peninsula p. 669 (1950). O. chrysophylla Lam. Tabl. Encycl. 1 : 29 (1791) et

Diet. 4: 544 (1794); Baker in FI. Trop. Afr. 4, 1 : 18 (1902); Chevalier in

Rev. de Bot. Appl. no. 303-304 (1948); Turrill in F.T.E.A. “ Oleaceae ” p. 9

(1952); O. europaea Thb. Prod. PI. Cap. 2 (1794) non Linn.; O. similis Burch.

Trav. 1: 177 (1822); O. europaea var. nubica Bkr. in FI. Trop. Afr. 4, 1 : 18

(1902); O. verrucosa Link. Enumer. PI. Hort. Berol. 1: 33 (1821); DC. Prodr.

8: 285 (1844); var. brachybotrys DC. l.c. ; Harv. ex Wright in FI. Cap. 4: 486

(1907), excl. syn. O. woodiana Knobl.

[According to Chevalier l.c., O. somaliensis Bkr., O. ferruginea Royle, O. cuspidata

Wall, ex G. Don., O. schimperi Gandoger and O. monticola Gandoger (all unknown

in South Africa) are also synonyms of O. chrysophylla and therefore of O. africana .]

574

Trees often 10 to 40 feet tall, sometimes stunted bushy growths; branchlets

verrucose. Leaves with a tendency for the sides to curl downwards as well as marginal

rim being reflexed, dark green above, paler beneath where it is densely covered, rarely

fairly sparsely so, with small silvery, golden or pale green scales, linear-lanceolate or

narrowly oblong elliptic, narrowed at base and apex (apex sometimes bluntly rounded),

1-9-8 -5 cm. long and 0-7-1 -5 cm. broad, rarely broader (see Clanwilliam and

Barberton specimens), mucronate; midrib impressed above, prominent beneath;

lateral veins obscure or faintly obvious, loops forming a more or less continuous line

within the margin, petiole usually 3-10 mm. long. Panicles axillary, sometimes a

short terminal panicle too, varying in size, usually shorter than the subtending leaf

the branches verrucose and scaly; bracts deciduous. Flowers small, white. Calyx

cupular, up to 1 mm. long, very shortly or obscurely 4-toothed. Corolla with a short

tube, about 1 mm. long; lobes more or less connivent, eventually spreading to reflexed,

about 2 mm. long, 1 • 5 mm. broad, margins narrowly infolded. Stamens inserted on

the corolla, filaments under 1 mm. long, more or less terete, anther attached near the

base, 1-5 mm. long, 1 mm. broad. Ovary sub-globose, narrowing into a short style;

stigma 2-lobed, forming a conico-globosa head; ovules pendulous. Drupe green

with whitish spots turning black or prune-coloured, sub-globose to oblong in outline,

up to 1 cm. by 9 mm. on dry specimens.

Fig. 3 and Plate 13.

Type: no specimen preserved, but plant described came originally from the

“Cape of Good Hope”. Types of synonyms: O. chrysophylla Lam., Sonnerat,

Reunion (fide Turrill); O. europaea Thb., Thunberg, Drakenstein near Cape Town;

O. similis Burch., Burchell, Olyvenhout Bosch, near Cape Town; O. verrucosa Link.,

plant described came originally from the “ Cape of Good Hope ”; O. europaea var.

nubica Bkr., Schweinfurth 249, near Suakin, Nubia.

Transvaal. — Soutpansberg: Wyllies Poort, Hutchinson 2064 (BOL); Gerstner

5865. Pietersburg: Blaauwberg, Smuts and Pole Evans 899; Codd and Dyer 9135;

Houseman in Col. Herb. PRE. 5328; Leipsig, Tsheuschaner in Tvl. Mus. Herb. 29509

(stunted); Letaba: Woodbush, Hoffman 18; The Downs, Rogers 20165.

Potgietersrust : Makapansberg, Rehman 5395 (SAM). Waterberg: Naboomspruit,

Galpin M. 208 (SAM). Marico: near Zeerust, Mar/oth 9535; Thode A1443;

Wonderfontein, Burtt Davy 7222; banks of Klein Marico, Burn Davy 7243.

Lichtenburg: near Lichtenburg, Kinges 1513 and 1738. Ventersdorp: Goedgedacht,

Sutton 620. Rustenburg: Swartruggens Sutton 851; 850; near Rustenburg,

Hutchinson 2932; Boshoek, Rose-Innes 57; Buffelspoort, Turner 5. Pretoria: Lotsy

and Goddyn s.n. (L.); Aapies Poort, Rehmann 4052 (SAM); Magaliesberg, Zeyher

1133 (SAM); Fountains Valley, Repton 309a and 125; Hutchinson 2317; Verdoorn

421 and 605; Curtis Hill, Pole Evans 257. Brits: Gun 2; 3 (stunted). Krugersdorp;

Hekpoort, Cohen 495: 1083; and 1240. Witbank: near Loskop Dam, Mogg 17285.

Lydenburg: Sekukuniland, Barnard 68; 269; Barnard and Mogg 905. Letaba:

Modjadjies Reserve, Krige 134; Kruger National Park, van der Schyff 56, 183; banks

of Letaba, Burtt Davy 2548. Nelspruit: Malelane, Codd 5223. Barberton: Rogers

Thorncroft in Roger’s Herb. 30059; Pott 5695 (leaves broad, marginal veins obvious)

Swaziland. — Ubombo Mountains, Miller S/54; Hornby 2812.

Natal. — Eshowe: Empangeni Rd., Lawn 1667 (NH). Weenen: West 1211;

Umhlumba, West 1453. Hlabisa: False Bay, Gerstner 5242; White Umfolosi,

Gerstner 3528 (NH). Misinga: Killick & Marais 2116. Umvati: Killick & Marais

2118 (stunted). Durban: Medley Wood 7750; Berea Bush, Medley Wood 3156 (NH)

(this specimen is O. africana and not O. woodiana, mixed gathering under nos. 548 &

3156 explains citation under O. woodiana).

575

Basutoland. — Mafeteng: Likhuele Mtn., Dieterlen 1242. Leribe: Dieterlen

313, and 715 (stunted). Mamathes: Guillarmod 77 (NH).

Orange Free State. — Senekal: Goossens 977. Ladybrand: Patterson 5090

(GRA). Bloemfontein: Pot 626 (BOL); Naval Hill, Compton 15662 (NBG).

Fauresmith: Vaalberg, Smith 5469; reserve hill, Smith 5588; Henrici 1883; Verdoorn

2375; south townlands, Verdoorn 940; Bergplaas, Verdoorn 1660, 1655; and 2759

(stunted).

Bechuanaland. — Kanye: Miller B/248; Pharing, Hillary and Robertson 523;

Ootsi, Miller B/231.

South West Africa. — Aus: Pillans 5964, Griqualand West variant (BOL)r

Mar/oth 4654. Grootfontein: Liebenberg 4880; Bristoes farm, Maguire 1740 (NBG);

farm Kumkauas, Kinges 2890; Venterspost, Schoenf elder S464. Otavi: Dinter 5294.

Rehoboth: Bullspoort, Rodin 2965.

Cape.— Komgha; Flanagan 17. Queenstown: Bowker’s Park, Galpin 2566;

Long Hill Peak, Galpin 7730. King Williams Town: Sim 2155. Keiskama Hoek:

Ngumeya Forest, Stayner 70 (GRA). Somerset East: MacOwan s.n. (GRA);

Boschberg, MacOwan 1364 (SAM). Graaff-Reinett : Bolus 610 (BOL). Bathurst:

Smuts 1329; Britten 2385 (GRA). Alexandria: Sim in F.D. Herb. 3166. Uitenhage:

Drege s.n.; Ecklon & Zeyher s.n.; Prior s.n.; Krauss 1795 (BOL); Paterson 2329

(BOL); Springfields, Paterson 1913 and 2194 (GRA). Humansdorp: Bitouw River,

Fourcade 619 (GRA); Hankey Reserve, Sim 3958; and 3995 (Forestry). Knysna:

Belvedere, Duthie 33 (GRA). Oudtshoorn: Gango, Britten 1743 and 1649 (stunted)

(GRA). Riversdale: Corenti River, Muir 342; 5147 (GRA); near Riversdale,

Schlechter 2005 (GRA). Gordonia: Upington, Kotze 839, Griqualand West variant.

Taungs: Rodin 3629 (BOL); Buxton, Bruechner 1214. Kimberley: near Taungs,

Rodin 3629; Nooitgedacht, Mogg 151 1 1 . Barkly West: Droogegrond, Whitlock s.n. ;

between Delpoort Hope and Kneukel, Acocks in Hafstroom Herb. H1215; Boetsap,

Brueckner 132. Hay: near Campbell, Pole Evans 6; 8; and 30; Griquatown,

Marloth s.n. (stunted); Blaauw Poort, Wilman 1316; Papkuil, Mar/oth 9946. Victoria

West: Brakfontein, Thode A2167. Namaqualand: Rattel Poort, Pearson 2966 (BOL).

Calvinia: Lokenburg, Story 4284. Clanwilliam: Greys Pass, Barker 6190; Maguire

2029; Pillans 9839, large leafed specimens, leaves about 1 -9 cm. broad (BOL); Uitkyk

Pass, Bond 1412 (NBG); Pakhuis Pass, Compton 4741 (NBG); Esterhuizen 14983,

short broad leaves, (BOL). Piquetberg: Pillans 8629; 7178 (BOL); Pickeniers Pass,

Pillans 5153 (BOL); Modderfontein, Howes 224. Ceres: Mitchells Pass, Esterhuysen

20719; 14736 (BOL). Ladysmith: Sewe Weeks Poort, Bond 258 (NBG). Worcester:

Hexriver Mountains, Rehmann 2704. Paarl: Paarl Mountain, Drege s.n.; Klein

Drakenstein, Galpin 11041. Stellenbosch: Jonkershoek, Parker 4730 (SAM);

Helderberg: Parker 4093 (NBG). Peninsula: Table Mountain, Drege s.n.; Gerstner

6142; Lotsy and Goddyn (L.); Devil’s Mountain, Drege s.n.; Kirstenbosch,

Esterhuizen 72 and 662 (NBG); Llandudno, Compton 21029 (NBG); west coast,

Humbert 9505; Blinkwater Waterfall, Marloth s.n.; Karbonkelberg, Adamson 1209.

Also occurs in North and East Tropical Africa and the Mascarenes and, according

to Chevalier, in India and Arabia as well.

It is to be expected that a species with such a wide distribution will vary to some

extent and that regional forms will occur. An example of such a form may be seen

in specimens from the Griqualand West region. Here the leaves are on the whole

smaller with very silvery scales on the lower surface and with the margins incurled, but

this form grades into the more usual and it is not strictly confined to the region

mentioned. A large leafed form has been collected in the Clanwilliam area and at

Barberton. Another variation which, however, is not a regional one, is the dwarfed

or stunted form. Such specimens may be found in any region and they are usually

growing in the neighbourhood of normal trees.

576

That this stunted form is very different from the normal growth is remarked on

by Chevalier,* but it certainly cannot be regarded as a variety for it has been observed

that from such a stunted bush a normal branch may develop.

This is borne out by the finding of Mohammed Drar, published in Bulletin 149,

Technical and Scientific Services, Egypt, pages 85-88 (1936), where he shows that

O. europaea var. nubica Bkr. is a “ sucker specimen ” (that is the stunted form, from

his description and photograph) of “ O. chrysophyllum ”. Among the specimens cited

these stunted ones and the regional forms mentioned are indicated.

The habitat of O. africana ranges from forest and riverside bush to open grassveld,

flats, stony ground, mountain kloofs or rocky ridges. It may be found in flower from

October to March.

The species is distinguished by the densely scaly under surface of the leaf, which

is linear-lanceolate or narrowly oblong-lanceolate, rarely some leaves obovate-oblong,

margins usually recurving as the leaf dries, lateral veins anastomosing near the margin,

the loops forming a more or less continuous line along, but a short distance from,

the margin; by the axillary inflorescence and the fruit which is oblong-globose and

up to 1 cm. long. It is closely related to the cultivated olive, Olea europaea Linn.,

and may be the origin of it.

Chevalier in the above-mentioned paper keeps the wild species separate specifically

from the cultivated and this seems a wise and orderly treatment. The only tangible

differences between the two species are the larger flowers and the larger and more

fleshy fruits of O. europaea which may have come about through selection and

cultivation but, if so, these characters have now been established and reproduce true

to type.

For the wild species, Chevalier gives O. chrysophylla Lam. (1791), as the oldest

name with O. verrucosa Link., etc., as synonyms, but in doing so he seems to have

overlooked the fact that O. africana Mill, was described in 1768 and therefore has

priority. It is possible that Chevalier ignored Miller’s name as that of an insufficiently

known species. At first it did seem doubtful to some of us whether, in the absence of

a type specimen, Miller’s species could be identified with certainty. The question was

asked, is there sufficient evidence to establish that O. africana Mill, is the same as the

species which had come to be known as “ O. verrucosa Link.” at the Cape and is

described under that name in Flora Capensis. After months of investigation which

included writing to Prof. Adamson of the University of Cape Town and the Director

of the Royal Botanic Gardens, Kew, it was finally settled that this is so. The consensus

of opinion is that Miller’s notes leave one in no doubt, for he writes (1) that the species

“ grows naturally at the Cape ” and (2) that it grows to the height of and bears some

resemblance to O. gallica Mill., that is O. europaea L., the cultivated olive. This cannot

be said of any of the other species at the Cape, and also Miller’s description of the

leaf does not fit that of any of the others. Furthermore Miller cites, as a synonym

of his O. africanum, Boerhaaves phrase-name for a species from the “ Cape ” and

it too describes unmistakably the species under discussion.

Specimens of O. chrysophylla Lam. from Mauritius and Reunion, the type locality

of the species, seen among sheets sent on loan from the Ryksherbarium, Leiden, are

obviously conspecific with certain specimens from South Africa and also some from

tropical east Africa and Abyssinia. It does not seem feasible to separate the specimens

from these three remote countries even into distinct varieties for in some cases they

compare better with each other than for instance, the Griqualand West form with

the majority of specimens from the rest of South Africa. No attempt is therefore being

made to follow Chevalier’s division of the species into 8 varieties, evidently based

principally on the country of origin, or to give any sub-specific rank to the ecotype in

S. Africa.

“ L’Origine de 1’Olivier cultive et ses variations ” in Rev. Bot. Appl. n. 303-304 (1948).

577

The timber of O. africana is considered to be valuable and durable. The leaves

are eaten by stock and the natives use an extract from the leaves boiled in water as

coffee.

The common name “ Wild Olive ” is generally used for this species in South

Africa. There is no modern record of the name “ Slagenhout ” mentioned by Boerhaave.

Fig. 3. — Olea africana Mill.; a, example of leaves from a specimen from the

Peninsula; b, from Griqualand West; c, stunted form, from Fauresmith district.

2. O. woodiana Knobl. in Bot. Jb. 17: 532 (1893); Medley Wood in Natal Plants,

3, plate 237 (1902); Sim, Forest FI. C.C. p. 266, pi. 108, (1907). O. mackenii

Harv. ex Wright in FI. Cap. 4, 1 : 488 (1907). O. listeriana Sim ex Lister in Rep.

Conserv. For. Cape for 1897, 98 (name only); Wright in FI. Cap. 4, 1: 1129

(1909); Sim Forest FI. C.C. p. 266 in obs.

Tree 12-50 feet, rarely 100 feet tall, with whitish bark, smooth or fluted; ultimate

twigs pale grey or whitish, more or less lenticular, at least some of the upper internodes

4-angled. Leaves lanceolate-elliptic to elliptic, usually broadest about the middle and

from there narrowing to an acute base and apex (apex sometimes rounded), 4-8 cm.

long, -8-3-3 cm. broad, flat with just the marginal rim reflexed and often loosely

undulate, minutely scaly, giving the appearance of being minutely pitted, especially

on lower surface, midrib more or less impresed above, prominent below, lateral veins

sometimes faintly obvious, then prominent above, anastomosing in large loops near

the margins (loops not forming an almost straight line along the margin); petiole

4-10 mm. long. Panicles axillary and quite frequently terminal too, many flowered

but not dense and compact; branches slender, sub-terete, 4-angled or variously

flattened and fluted, internodes and peduncles relatively long, pedicels short; bracts

up to 4 mm. long, sub-deciduous. Calyx small cupular and shortly 4-lobed, up to

578

1 mm. long. Corolla white; tube -75 mm. long; lobes about 2-25 mm. long and

1 • 5 mm. broad, forming a sub-globose bud, eventually reflexed. Stamens with filaments

inserted on the tube, free for about 1 mm., anther 1-5 mm. long, 1-25 mm. broad,

attached about at the middle. Ovary sub-globose; style very short; stigma 2-lobed

conico-globose; ovules pendulous. Drupe drying blackish, semi-ovoid, narrowing

to apex and oblique at base, usually about 1 cm. long and 5 mm. broad in dried

specimens, occasionally slightly larger.

Fig. 4 and Plate 14.

Type: Medley Wood 548, near Durban. Types of synonyms: O. mackeniiHarv.,

Gerrard 380, Tugela, Natal; O. listeriana Sim ex Lister, Sim 2143, East London.

Transvaal. — Pilgrims Rest: van der Merwe Bush, Burtt Davy 1428.

Swaziland. — Stegi: Ubombo mountains, Miller S55 and S20 (leafy specimens

only).

Natal. — Mtunzini: near Inyoni mouth, Gerstner 1957; at Inyoni mouth,

Gerstner 2444. Durban: near Durban, Medley Wood 548, isotype (SAM); Medley

Wood 7975; shore near Durban, Medley Wood 7879, cited Bot. Jb. 51: 76 (NH);

Bluff", Medley Wood 12634 (NH). Pinetown: Isipingo Beach, Ward 649.

Cape. — Port St. Johns: Galpin 11464; Doran in F.D. Herb. 2265, 2136, 2137

(galls on last two specimens); Robertson in F.D. Herb. 1862; Miller 6130 (galled

specimen). Komgha: Flanagan 618. East London: Sim 2143, small leaves,

distributed as O. listeriana; Hunter in F.D. Herb. 1725. Bathurst: Port Alfred,

Rogers 905; Hopewell, Acocks 11047. Port Elizabeth: Alexandria, Strauch in F.D.

Herb. 3163 & 3246; Sim in F.D. Herb. 3164.

This species has so far not been recorded outside South Africa.

In connection with the type of this species, it must be made clear that a mixture

was distributed under Medley Wood’s Garden Distribution No. 548, which is the same

as his own No. 3156. The specimens with either or both these numbers at Berlin and

in the South African Museum, Cape Town, are O. woodiana whereas specimens with

the same numbers in the Kew Herbarium and the National Herbarium, Pretoria are

O. africana (= O. verrucosa). This error probably accounts for the fact that

O. woodiana Knobl. is cited as a synonym under O. verrucosa Link, in the Flora Capensis

while O. mackenii Harv., which is synonymous with O. woodiana Knobl., is there

described as a new species.

From O. africana, the only other species with axillary inflorescences, O. woodiana

is distinguished principally by the shape of the leaves which are more or less elliptic,

broadest about the middle and from there narrowing to the base and apex (that is the

margins are not more or less parallel for a certain distance as in the linear- or narrowly

oblong-lanceolate leaves of O. africana)', the under surface being minutely and fairly

sparsely scaly, appearing pitted, instead of obviously and densely scaly; the flatter

leaves (not inclined to curl up longitudinally with the under surface often concave);

the anastomosing loops of the veins not forming a more or less straight line inside the

margin; having rather large terminal panicles as well as axillary; and the fruit being

somewhat longer, usually oblique and narrowing towards the apex (semi-ovate in

outline).

Among the species with only terminal inflorescences, O. woodiana resembles

O. capensis subsp. macrocarpa to a certain extent, especially in the leaves. It differs

in the ultimate twigs being whitish and at least in part 4-sided; the branches of the

inflorescences being very slender and having longer internodes; peduncles with the

flowers more or less clustered at the ends on relatively short pedicels; and fruits not

as large, up to 1 cm. long.

579

O. woodiana has a comparatively restricted distribution being found only in the

eastern regions of S. Africa. At Durban, according to records, it grows in the vicinity

of O. africana.

The timber is described as “ steel-like ” by Gerstner.

Fig. 4. — Olea woodiana Knobl.; a, example of leaf-shape from type locality;

b, from East London; c, from Port St. John’s, with fruit.

3. O. exasperata Jacq. Hort. Schoenbr. 3, 1, t 251 (1798); DC. Prodr. 8: 287 (1844);

Knobl. in Bot. Jb. 17: 533 (1893); Harv. ex Wright in FI. Cap. 4, 1: 486 (1907);

O. humilis Eckl. South Africa quart, journ. 1: 370 (1830); DC. Prod. 8: 287

(1844); Sim in Forest FI. of C.C. p. 266, pi. 120 (1907).

Bushy or straggling shrubs or small umbrella shaped trees, from about 2 feet to

20 feet tall, rarely taller, branchlets rough with numerous raised lenticles. Leaves

with a tendency for sides to curl downwards as in O. african, linear-oblong, 4-8-5 cm.

long, 6-10 mm. broad, rarely 1-5 cm. broad; broadest in the upper two thirds,

narrowing gradually to the base, minutely pitted on both surfaces, especially on lower,

midrib raised on lower surface, sometimes also on upper, a few lateral veins sometimes

obvious and raised above but disappearing about half way to the margin, anastomosing

seldom seen, but if so loops curved ; margin rim reflexed ; petiole 4 mm. long rarely

5 mm. Panicles terminal, short, broad, many flowered; buds sub-globose; bracts

very small, pointed, sub-persistent. Calyx short, cupular, glandular, shortly 4-toothed.

Corolla with a short tube under 1 mm. long; lobes up to 2-75 mm. long, margins

narrowly infolded. Stamens with filament almost 1 mm. long; anthers large brownish

about 2 mm. long and 2 mm. broad, dorsifixed. Ovary narrowing into a short style;

stigma 2-lobed forming a conico-globose head; ovules pendulous. Fruit “black

purple ” when ripe, up to 1 cm. long and 8 or 9 mm. broad, sub-globose to oblong-

globose.

Fig. 5 and Plate 15.

Type: figure in Jacq. Hort. Schoenbr. 3, 1, t 251, plant originally from “Cape

of Good Hope”. Type of synonym: O. humilis Eckl., Ecklon, “in dunes at the

Cape ”.

580

Cape. — East London: Sim 2310. Bathurst: Kariega Mouth, Story 3254; Kowie,

Burchell 3829; Tyson s.n.; Britten 5009 (GRA); Salisbury s.n. (GRA); Kasouga,

Britten 2110 and 2339; Port Alfred, Rogers s.n. (GRA). Uitenhage: Addo, E. & Z.

s.n.; Drege s.n.; Koega and Swartkop Rivers, Zeyher 3373 (SAM). Port Elizabeth:

Strauch 4233; Paterson 852 (BOL); St. Georges Strand, Long 856; Baakens River,

Drege 343 (GRA); near Schoenmakers Kop, F.D. Herb. 19 (GRA); Swartkop Riv.

Mouth, Drege s.n. (L); Patterson 1974; Humewood, Lanham 130. Humansdorp:

Phillips 3349; Tsitsikama, Fourcade 232a (GRA); between Sland and Kromme Rivier,

Sim 3 (GRA). Knysna : Buffalo Bay, Keet 427 ; Woodbourne, Keet 751 in F.D. Herb.

George: Compton 14347 (NBG). Riversdale: S.S.E. of Riversdale, Acocks 14592;

Albertinia, Muir 838. Bredasdorp: Brandfontein, Smith 4974; Taylor 324 (NBG);

Cape Agulhas, Ecklon s.n. (only specimen on right hand of sheet), (SAM). Malmesbury :

Bok Point, Compton 9405 (NBG); Bokbaai, Esterhuysen 3840 (BOL). Caledon:

Pole Evans 4329; Middelvlei Reserve, Hubbard 236 (BOL); between Eerste Rivier

and Swart Klip, Pillans 9211 (BOL); Buffels River Mouth, Pillans 8301. Stellenbosch:

Strand, Parker 3591 and 3662 (BOL). Peninsula: Constantia, Marloth 8407; Camps

Bay, Adamson 571; Cape Town, Marloth 7542; Hout Bay, Acocks 638; Bond 117

(NBG); Bolus 13749 (BOL); near Muizenberg, Schlechter 1264 (GRA); Uitvlugt,

Muller in F.D. Herb. 5120 (GRA); Little Lions Head, Compton 18583 (NBG);

Karbonkelberg, Esterhuysen 21172 (BOL); Witsand, Lotsy and Goddyn 1587 (L).

Endemic to the western and southern coasts of the Cape Province.

The most common habitat of this species is on the sand dunes in coastal bush or

scrub forest. It also occurs on limestone hills, such as in Bredasdorp, or in open grass

veld or valleys along the coastal belt. It is an easily recognised species with its narrow

leaves which are linear oblong, broadest in the upper two thirds and gradually narrowing

in the lower half to a short petiole.

The leaves resemble those of O. africana in the tendency of the margins to roll

backwards but the lower surface in O. exasperata appears minutely and densely pitted,

not scaly as in O. africana. The terminal inflorescence of O. exasperata is another

distinguishing character between these two species.

In Sim’s Forest Flora of the Cape Colony, this species is under the name Olea

humilis Eckl., a latter homonym.

Fig. 5. — Olea exasperata Jacq. ; a, example of leaf-shape from Constantia;

b, from Caledon; c, from Camps Bay.

581

4. O. capensis Linn., aggregate species:

(a) subsp. capensis.

(b) subsp. enervis ( Harv .) Verdoorn, stat. nov.

(c) subsp. macrocarpa (C.H. Wr.) Verdoorn, stat. nov.

Trees or shrubs, from shrubby growth about 2 feet high or taller, to trees from

5-35 feet tall, or forest trees up to 90 ft. tall. Leaves, light to dark green, undersurface

somewhat paler than the upper or sometimes concolorous, occasionally suffused

purplish, very variable in texture, size and shape, but not linear or oblong-linear,

usually much over 1 cm. broad; petiole green or purplish; margins sometimes faintly

to very decidedly undulate, the rim only recurved; scales minute appearing like minute

pits especially on under surface; midrib prominent below, at least at base, lateral

veins, when visible, forking just beyond halfway to the margin the anastomosing branches

looped. Panicles terminal and in axils of the 2 upper pairs of leaves, many-flowered,

branches glabrous, scaly, variously angled and sulcate; bracts small spreading ovate

to subulate about 1 mm. long. Calyx cupular 4-toothed or 4-lobed almost to the

middle, minutely ciliate. Corolla up to 3 mm. long, lobed almost to the base, lobes 4.

rarely 3 to 5, ovate-oblong to oblong, rounded at the apex but with a minute incurved

mucro and the rather thick margins slightly incurved. Fruit sub-globose, ovoid,

oblong-globose to oblong-elliptic, rarely at some stages pointed (see Bolus 23227 from

Caledon and Pillans 7887 Piquetberg) from 5 mm. long and 4 mm. diam. up to 2 cm.

long and 1 cm. diam.

The distribution of this aggregate species, is restricted to southern Africa, that

is if the four closely related species in tropical Africa are kept distinct by future workers

on the genus. The tropical species are: O. welwitschii (Knobl.) Gilg and Schellenb.

O. hochstetteri Bkr. ; O. urophylla (Gilg) Gilg and Schellenb.; and O. guineensis

Hutch, and C.A. Sm. (= O. hochstetteri fide Turrill F.T.E.A.), all described later

than O. capensis L. In appearance, judging from the few tropical specimens seen,

these differ in general from the S.A. species in the inflorescences being coarser with their

branches thicker and the flowers fewer, which when in bud are broader than long.

The differences might also come to be considered as merely subspecific.

After years of study and the examination of material from all the principal herbaria

in South Africa and some overseas, including the Royal Botanic Gardens at Kew, the

following conclusions were reached: (1) that O. capensis L., O. laurifolia Lam.,

O. enervis Harv. and O. macrocarpa C.H. Wr. of the Flora Capensis comprise one

variable and complex species; (2) that the material can be divided into 3 fairly distinct

groups, but these groups cannot be given higher rank than that of subspecies because

they grade into one another and because in some cases where the habitat of the specimens

is not known or there are no mature fruits present it will not be possible to determine

to which group the specimen belongs; and (3) that although there is considerable

variation within these groups especially in one of them, no further subdivisions can be

made at this stage because in no further instances was it found that certain

combinations of features were repeated giving a similar appearance to a fair number

of specimens and so isolating them as a group ; in most cases only single variants were

found. The reasons for these conclusions should become more evident from the notes

under the sub-species.

582

(a) subsp. capensis. O. capensis Linn. Sp. PI. Ed. 1: 8 (1753); Dill. Hort. Elthm.

t. 160, pi. 194 (1732); Bot. Reg. t. 613 (1822); Harv. ex Wright in FI. Cap. 4,

1: 487 (1907); Adamson in FI. of the Cape Peninsula p. 669 (1950); O.buxifolia

Mill. Gard. Diet. ed. 8 (1768); O. laurifolia Lam. III. I; 29 (1791); J. Burm.

Rar. Afr. PI. 233 t. 81 fig. 1 (1739); Harv. ex Wright in FI. Cap. 4, 1: 487 (1907)

pro majore parte, excluding Burchell 5225 (= subsp. macrocarpa) and Wood 500

(= subsp. enervis); Adamson in FI. of the Cape Peninsula p. 669 (1950);

O. undulata Jacq. in Hort. Schoenbr. 1, 1, t2 (1797); var. planifolia E. Mey. in

Comm. PI. Afr. Austr. 176 (1837); O. concolor E. Mey. in Comm. PI. Afr. 176

(1837) and in DC. Prod. 8: 286 (1844); O. laurifolia Lam. var. concolor Harv.

ex Wright in FI. Cap. 4, 1: 487 (1907).

This subspecies is distinguished by the following features: Shrubs or trees, up to

35 ft. tall, not taller forest trees. Leaves crowded on herbarium specimens, very variable,

obovate-oblong, oblong, broadly oblong, sub-orbicular, elliptic, ovate- to lanceolate-

oblong, or obovate- to oblanceolate-oblong; apex broadly rounded, obtuse, sub-acute

or acute, sometimes acuminate, usually mucronate; petiole 3 mm. to 1-7 cm. long;

midrib prominent on lower surface. Panicles compact and densely many flowered.

Fruits variable in size and shape but not over 1 cm. long and 6 mm. diam.

Figs. 6, 7 and 8; Plates 16, 17 and 18.

Type: Linnaen Herbarium No. 204, plant from Cliffort’s garden, originally

from “ Cape of Good Hope ” (see plate 16). Types of synonyms: O. laurifolia Lam.,

specimen in Paris Herbarium (see plate 17); O. undulata Jacq., tab. 2, in Jacq. Hort.

Schoenbr. 1, 1 ; O. undulata var. planifolia E. Mey., Drege, Zuurbergen; O. concolor

E. Mey., Drege , between Nieuwekloof and Elandskloof, Tulbagh district. (See plate 18).

Natal. — Port Shepstone: Uvongo, coastal bush, Letty s.n.

Cape. — Lusikisiki: coast near Umkwani River, Tyson 2657 (SAM); Fraser

Falls, Acocks 13431. Kentani: coast, Pegler 994; Pegler 826a (BOL). Komgha:

Flanagan 655; Kei Mouth, Schlechter 6194 (GRA). East London; Breyer 23242;

Rattray 509; coast, Sim 2102 and 2577 (NH); Bonza Bay, Story 4484; 11 m. W. of

East London, Hilner 158 and 268 (GRA); Fort Pato Forest, in F.D. Herb. 1724.

King William’s Town: Pirie, Sim 1336 (BOL). Somerset East: Zuurberg Pass,

Story 2300. Albany: near Hamilton Dam, Dyer 234; Hope Fontain, Acocks 12119;

nr. Grahamstown, MacOwen 1210 (BOL); Liebenberg G. 300; Amos Kloof, Galpin

361; Howison’s Poort, Zeyher 3377 (BOL). Bathurst: Southwell, Story 3135;

4494; 4495; Salt Vlei near Port Alfred, Story 4491; Kowie, Britten 2850 (GRA).

Port Elizabeth: Van Stadens River, Bolus 1210 (BOL); Longmore Forest Reserve,

Long 1009; Witteklip, Rodin 1047. Port Elizabeth: Kemsley 257 (GRA);

Krakakamma, Zeyher 3378 (SAM); Swartkops River Valley in F.D. Herb. 4445;

Hankey Reserve in F.D. Herb. 3952. Uitenhage: Van Stadens Gorge, Long 398;

Hoffmankloof to Driefontein, Drege in Ryksherbarium 908161-541 (L); Dornnek

and Bontjiesrivier, Zuurberg, Drege (K, L); Addo, Zeyher 547 (BOL). Humansdorp:

Klipdrift, Thode A 2492; Clarkson, Thode A959; sanddunes at Slang River, Phillips

3390; Lottering Bush, Zitzikama, Galpin 4322; Storms Rivier Forest, in F.D. Herb.

3928; 4022; 4027; Fourcade 704 (GRA); Groot River Pass, Fourcade 675 (GRA);

Ratels Bosch, Fourcade 392 (GRA, BOL). Uniondale: Prince Alfreds Pass, Fourcade

5870 (BOL). Knysna: in forest Theron 981; The Heads, Laughton in F.D. Herb.

8949; Schonland 3575 (GRA); Kaffir Kop Forest, Keet in F.D. Herb. 3566; nr.

banks of lagoon, Williamson 30 (GRA); Noetzie Taylor 1224; 1225; Keurboomstrand,

Taylor 596; Plettenberg Bay, Smart and Rogers 26812; Belvedere Brenton, Duthie

8 and 620 (GRA); Forests, Duthie 676 (GRA); Portland, Duthie 924 (GRA); near

Knysna, Burchell (K, GRA); Kaatjies Kraal, Burchell 5227 (L); Deepwalls, Rodin

1167 (BOL); Taylor 700; Phillips in F.D. Herb. 5487 (3 sheets); Keet in F.D. Herb.

2661 and 3598; Farleigh Forests, Keet in F.D. Herb. 2337. George: Kaaimans

583

River, Wilman s.n.; Jonkersberg, Burton in F.D. Herb. 3775 (suggests a touch of

“ subsp. macrocarpa ”) ; Mt. Pleasant, Martin 120 (NBG); Groenkop, Robertson

in F.D. Herb. 7593. Swellendam: Grootvadersbosch, Marloth 3524. Bredasdorp:

Cape Agulhas, Galpin 11254; Road to Stanford, Maguire 80 (NBG). Caledon:

between Houwhoek Mountains and Palmiet River, Burchell 8161, small leafed variant,

(K) ; Grabouw, Britten 3111 (GRA); Kogel Bay, Parker 4150; Mossel River, Pots

in S.A. Museum 5051, accuminate; Potts 1652, rounded apex, (SAM); near Hermanus,

Mossel River, Guthrie s.n.; Hermanus, Burtt Davy 18539 (BOL); Hangklip, Rodin

3107; Bolus 23227, pointed fruits, (BOL). Somerset West: Hottentots Holland,

Zeyher 3374 (BOL). Stellenbosch: Sir Lowrys Pass, Burchell 8236 (K, L, GRA);

Schlechter 7267 (GRA). Wynberg: between Wynberg and Constantia, Burchell

781 (K). Peninsula: without precise locality Wallich s.n. (K); Sieber 219; 220 (L);

Kirstenbosch, Zeyher 182, named O. undulata Jacq., (BOL); Pearson 16627 and

25303 (BOL); Kies 11; Compton 10849; 10040 (from various trees pointed leaves

and rounded; 10041; 10019 (NBG); Esterhuysen 11744 (elliptic-oblong fruits 1 cm.

long, leaves broad and rounded at apex); 11834 (fruits up to 1 cm. long, leaves broad);

12853 (small leafed variant); 17516 (some leaves obtuse some acuminate); 264

(NBG); 650 (NBG); 11683, fruits up to 1 cm. long, broad leaves, (NBG); Henderson

1408 (NBG); Compton 8077 (NBG); Bond 82 (NBG); Pi/Ians in BOL Herb. 1772,

large purple leaves, twigs purple, (BOL); Groote Schuur, Smuts 1100 (pointed leaves);

1165 (broad apex); Bishops Court, Galpin 4814; Claremont, Hutchinson 3; Table

Mountain, Drege s.n. ; Ecklon and Zeyher 84-7 (some leaves rounded, some acuminate),

labelled O. undulata Jacq. var. planifolia; Pappe s.n. (SAM); Marloth 5662, 7406,

11926; Smuts 1088; Esterhuysen 11401 (BOL); Andreae 263; Pole Evans 4327;

4328; Camps Bay, Ecklon and Zeyher 69 • 1 ; Prior s.n. ; Marloth 7501; Zeyher s.n.

(K, BOL); Muizenberg, Bolus 3904 (BOL); Pillans 3706; Adamson 944; Galpin

10344; Hout Bay, Smuts 1076; Pillans 3689; Lotsy and Goddyn 1929; 1840 and 1867

(L) ; Orange Kloof, Wolley Dod 866; Marloth 16628; Noordhoek Forest, Lotsy &

Goddyn 1599 and 1662 (L); Die Kommetjie, Lam. & Meeuse 4200B (L). Without

locality, Sieber 219 and 220 (cited in FI. Cap., former under O. laurifolia and later

under O. capensis, two sheets of each) (L); Simons Bay, Macgillivray 665 (K). Paarl:

Berg River, source on mountain slope, Pillans 8129 (BOL); Happy Valley, Bains

Kloof, Esterhuysen 12820 (BOL); French Hoek, Hubbard 307 (BOL). Ceres: Mitchell’s

Pass, Andreae 214; Bond 9 (BOL); exact locality? Thode A2268 (leaves small in

3 last named specimens; same tree in Mitchell’s Pass?). Tulbagh: between

Nieuwekloof and Elandskloof, Twenty-four Rivers, Drege s.n.* (type gathering of

O. concolor E. Mey., 3 sheets in Ryksherbarium, Leiden (L); E. & Z. 77-9 (that is

along same route as Drege, passing Nieuwekloof). Piquetberg: Kapiteins Kloof,

at base of mountain, Pillans 8024, 2 sheets, one fruiting, one flowering (BOL); [Pillans

7887 (BOL) from “ upper slopes of Kapiteins Kloof ” may belong here, distinct

appearance suggests hybrid, needs investigation]; Pikeniers Kloof, Zeyher s.n. (looks

like same collection as the following specimen Zeyher 1150) (BOL); Zeyher 1150,

3 sheets, cited in FI. Cap. under O. laurifolia var. concolor (BOL). Clanwilliam:

Grasruggens Nek, Pillans 8716.

Not known to occur outside South Africa.

This subspecies includes a large number of specimens from the coastal regions

of the Cape, stretching from the borders of Natal southwards and westwards to the

Peninsula and northwards to Clanwilliam. The leaves of the specimens vary very

much in texture, shape and size, but all attempts at grouping specimens with more

or less similar leaves failed completely. This was not only because intermediates are

found between all extremes, but also because the variations in leaves do not combine

with any other feature to form a recognisable group.

* The specimen in the National Herbarium, Pretoria, supposed to be the same Drege gathering,

is not an Olea.

584

Of the name changes necessary in this revision the one that will cause some upset

and needs explanation is the sinking of O. laurifolia Lam. under O. capensis. This is

because (1) the features given as the distinguishing characters seem to work well, since

some specimens have leaves broadly obtuse while others narrow to a sub-acute or

acute apex, and (2) the name O. laurifolia has wrongly come to be applied to the forest

tree here classified as O. capensis subspecies macrocarpa. The leaves of this forest

tree are somewhat similar in shape to those of several specimens from the Peninsula

which would fall under O. laurifolia Lam. (cfr. figs. 8a and b with leaf shapes on fig. 10),

but the fact that the forest tree produces large elliptic fruits is, in my opinion, more

diagnostic than similarity in leaf-shape. The figure of O. laurifolia cited by Lamarck

(tracing of a leaf and fruit of this figure is reproduced here, see fig. 6b) is described

by Burman (Rar. Afr. pi. 233) as having “ thick ” leaves which are “ almost sessile ”

and “ rotund fruits ” of the size of the drawing, all of which does not describe the

forest tree (see for instance its long petioles, pi. 20). Throuhgout the 200 odd years of

collecting no specimens with fruits larger than the one figured has been collected

anywhere in the Cape except in the forests at Knysna or similar forests to the east.

Therefore the name O. laurifolia Lam. cannot be applied to the forest tree with large

fruits.

Having settled this point, it still remains to be explained why O. laurifolia cannot

be separated from O. capensis at least as a variety. The examination of the types of

the two species (PI. 16 and 17) and of the drawings of leaves from a number of specimens

collected in a fairly circumscribed area at or near Kirstenbosch (Fig. 7 and 8), will

help to elucidate this. From each specimen two leaves were drawn, a small and large

one. On Fig. 7, a and b, the small leaves are very similar but in the one case (Fig. 7a)

they develop into a leaf with a very broad apex (like typical O. capensis, see Fig. 6a),

and in the other (Fig. 7b) into a leaf which narrows slightly to a blunt point, very like

typical O. laurifolia (see Fig. 6b). Similar pairs of leaves from different plants in the

Kirstenbosch area show an even greater diversity in the shape of the full grown leaf

than those of the types (see Figs. 8a-g and compare with photos of the types on plates

16 and 17). All these specimens whether small- or large-leaved, rounded or acute at the

apex, wavy or not on the margins, are considered by the local botanists, who know

the growing plants, to belong to a single species. In the same way taxonomists working

with the pressed specimens have found it impossible to group them and distinguish the

groups from each other. Among these heterogeneous specimens or within their

range of variation would fall the type of O. undulata Jacq. (see Fig. 6c) which was sunk

under O. laurifolia in the Flora Capensis and now is a synonym of O. capensis.

It was not quite so easy to decide whether O. concolor, from Tulbagh, is

synonymous with O. capensis. The isotypes, Drege s.n., which were seen among the

specimens sent on loan from Kew and the Ryksherbarium, Leiden, look rather distinct

(see PI. 18 and Fig. 8c). Most of the leaves are rather small, broadest in the upper

half, long cuneate to the base and abruptly narrowed near the apex into a cusp-like

acumen. After examining all other available specimens from the regions north of the

Peninsula, it was obvious that here too were odd variants that could not be grouped

together (see Fig. 8c-g). To illustrate, Wright in the Flora Capensis cites under

O. laurifolia var. concolor Harv., with the type (Drege s.n.) a specimen Zeyher 1150

from Piquetberg. This latter specimen has a long leaf, broadest about the middle and

long tapering to base and apex (see Fig. 8d), quite unlike the type of O. concolor. From

Piquetberg came also two specimens, Pi/Ians 7887 (see Fig. 8e) “ from upper slopes ”,

with leaves very like the type of O. concolor, and Pillans 8024, from “ base of mountain ”,

with very different leaves (see Fig. 8g). From Clanwilliam, the most northerly locality

known, comes yet another form of leaf (see Fig. 8f). Most of these specimens from

the northern districts, with the possible exception of Pillans 7887 which should be in-

vestigated (especially from the point of view of possible hybridization with O. exasperata),

are best included as variants in the complex subspecies O. capensis L. subsp. capensis.

585

Fig. 6. — Olea capensis L. subsp. capensis; a, leaves from the figure in Dill. Elth.

t. 160, cited with the original description of O. capensis L.; b, leaf and fruit from

the figure in J. Burm. Rar. Afr. PI. t. 81, fig. I, cited with the original description of

O. laurifolia Lam.; c, leaf and fruiting inflorescence from the figure in Jacq. Hort.

Schoenbr. I, I, t2, the type figure of O. undulata Jacq.

586

Fig. 7. — Olea capensis L. subsp. capensis. Samples of leaf-shapes from specimens

growing on the Peninsula, 2 from each specimen, to show the range of variation;

a, Drege s.n. Table Mountain ; b, E. & Z., mountains round Cape Town ; c, Esterhuysen

12853, Kirstenbosch; d, Esterhuysen 11744, Kirstenbosch.

587

Fig. 8. — Olea capensis L. subsp. capensis. Examples of various leaf-shapes;

a, two leaves, Smuts 1100, Groote Schuur; b, Marloth 5662, Table Mountain, like

“ O. undulata”; c, isotype of O. concolor, Drege, Tulbagh district; d, Zeyher 1150

cited in FI. Cap. as “ var. concolor”; e, Pillans 7887, Kapiteins Kloof, Piquetberg;

/, Pillans 8716, Clanwilliam; g, Pillans 8024, also from Kapiteins Kloof.

129828-6

588

Besides the variants from Kirstenbosch and the northern districts, there are others

that could be mentioned, such as the specimens with pointed fruits, that seem to occur

haphazardly, and certain growth forms.

Among the latter is a low, shrubby, sea-side form described in one instance by

the collector, John Phillips, as an “ impenetrable hedgelike consocies 12-36 in. high ”.

The leaves on such plants are usually large and thick and broadly rounded at the apex.

Investigation with the help of Forest Officers shows that the appearance is probably

caused by environment and in the early stages of growth. The particular patch

described by John Phillips as 12-36 in. high is today, 20 odd years after, 6 to 8 feet

high with smaller leaves, while a little deeper inland are small trees that link this form

with the normal tree.

With regard to the common name, the most generally used is “ Black Ironwood ”.

A record exists that in the Tzitzikamma it is known as “ Lemoen Ysterhout ” owing

to the yellowish colour of the underside of a fresh leaf. This yellow colour of the

leaf-undersurface is probably another example of the variations possible in this

subspecies; the notes usually describe the leaves as lighter on the undersurface, but

there are quite a number which are reported to have concolorous leaves while in a

few cases they are described as purplish beneath.

(b) subsp. enervis ( Harv .) Verdoorn stat. nov., O. enervis Harv. in Wright in FI. Cap.

4, 1: 488 (1907); O. laurifolia Harv. ex Wright in FI. Cap. 4, 1: 487, in part

( Wood 500) non Lam.

The essential features by which to distinguish this subspecies are the flatter leaves

of more uniform size and shape (compare drawings in Fig. 9 with those in Figs. 7,

8 and 10), on the average broadly elliptic, very rarely oblong elliptic, 4-5 cm. long and

1-5-2 -5 cm. broad; the short petiole; pale branchlets; lower leaves on herbarium

specimens deciduous, leaving prominent scars; and the midrib usually prominent

only in the basal half on the undersurface.

Plate 19 & Fig. 9.

Type: Gerrard 1151, from between the “ Buffalo River and Mooi River ”, that

is probably in the Msinga district, Natal.

Transvaal. — Soutpansberg: near Lake Funduzi, Gillett 3085 (BOL); Story

4857. Pietersburg: Blaauwberg, Codd 8694; 8695; Codd and Dyer 9001; Bremekamp

and Schweickerdt 112. Rustenburg: Breedtsnek, Repton 3905; Pisangkloof,

Rose-Innes 23254 (J). Brits: Kloof in Magaliesberg opposite Wolhuterskop turn-off,

Story and Rose-Innes 1392; Marais 279 and 285. Waterberg: Groothoek, Codd

3955; Hangklip, Maguire 1441 (BOL). Potgietersrust: near Potgietersrust, Galpin

8827; between Potgietersrust and Palala, Pole Evans s.n. ; S.E. of Palala, Story 1667;

Bokpoort, Codd 2364; Galpin 11685. Lydenburg: Sekukuniland, Barnard 502;

270; Barnard and Mogg 80, 768, 1003; Mogg 16908; 16888; Moss 22464; Van

Warmelo 93; 105; Steelpoort, Keet in F.D. Herb. 6058; Mooihoek Chrome Mine,

Codd & Dyer 7701. Barberton: between Louws Creek and Maid of the Mist

Mountain, Hutchinson 2434; Rimers Creek, Thorncroft 2005; Berlin Plantation,

Keet 6728; Thode A1622, leaves rather too long acuminate?

Swaziland. — Lebombos, east side, Hornby 2830.

Natal. — Ingwavuma: Gerstner 3114 (NH); Melmoth: Entonjaneni, Gerstner

2685 (NH). Nkandhla: Qudeni, Gerstner 3572, one specimen on sheet with all small

leaves, as on type specimen, (NH), Davis 113 (NH). Weenen: Umhlumba Mountain,

west slopes, Acosks 10615; Blaauwkrans River, Pentz 218. Msinga: Buffalo River

and Mooi River, Gerrard 1151, holotype (K); isotype, the majority of leaves small,

(PRE). Eshowe: Ngoya, Boocock in F.D. Herb. 5992. Ndwedwe: Inyoni Hill,

Oliver 502 ; Inanda, Medley Wood 500 (cited in FI. Cap. under O. laurifolia). Pinetown :

Umzinyati, Medley Wood 11443. Port Shepstone: Lower Umzimkulu, Medley Wood

11596, 9982 (NH).

589

Not known to occur outside South Africa.

This subspecies occurs in the bushveld of the Transvaal and Swaziland and the

dry inland regions of Natal. Although falling within the range of characteristics of

O. capensis L., specimens from these regions show a certain similarity in general

appearance that makes them distinguishable from the rest. Because the differences

are merely differences of degree and because occasionally specimens from the eastern

Cape are almost indistinguishable this group has been given the rank of subspecies.

The type specimen of O. enervis Harv., that is now the type of this subspecies,

Gerrard 1151 (see Fig. 9a), has leaves on the whole smaller than the average (see Fig.

9c-f), but they are within the range of size and may be matched with one or two of the

leaves on any average specimen of the subspecies.

Gerrard 1151 was collected between the Buffalo and Mooi Rivers in Natal, that

is probably in the Msinga district which lies between the Weenen and Nkandhla districts,

from which specimens obviously belonging to this group have been collected, for

instance Gerstner 3572, collected at Qudeni, Nkandhla.

The majority of the leaves on the flowering branches of Gerstner 3572 are small

like those on the type specimen (see left hand leaf of Fig. 9b), while on a branch on

the same sheet most of the leaves are much bigger (see right hand leaf of Fig. 9b). Fig.

9c depicts an average leaf from a specimen from Weenen and Fig. 9d is of a specimen

from Rustenburg in the Transvaal. Lastly Figs. 9e and 9f are drawings of representative

leaves from two different specimens collected on the Blaauwberg.

Fig. 9. — Olea capensis L. subsp. enervis (Harv.) Verdoorn. Examples of leaves

showing range in size and shape; a, leaf from type specimen, Gerrard 1151, between

Buffalo and Mooi Rivers, Natal; b, two leaves from Gerstner 3722, Qudeni, different

size and shape from same gathering; c, Acocks 10615, Weenen; d, Codd 2364,

Bokpoort; e, Codd 8695, Blaauwberg; /, Codd 8694, Blaauwberg.

590

A single specimen Gillett 3085 from near Lake Funduzi, has rather unusual leaves,

being mostly narrowly oblong, rounded at apex and base and with the midrib prominent.

Investigation has proved that this specimen, which is in the Bolus Herbarium, is not

representative of the trees there, the leaves of which are on the whole more typical.

Another specimen that differs somewhat from the general pattern is a specimen from

Barberton, Thode A1622, in which the leaves are rather long acuminate to an acute

apex.

(c) subsp. macrocarpa (C.H. Wr.) Verdoorn, stat. nov. O. macrocarpa C.H. Wr. in

FI. Cap. 4, 1 : 1129, addenda, (1909) and in Kew Bull. p. 186 (1909). O. laurifolia

Sim, Forest Flora of C.C. 264, t. 106 (1907) pro parte, fig. inch; Harvey ex Wright

in FI. Cap. 4, 1: 487 (1907) pro minore parte e.g. Burchell 5225; J. Phillips in

Trans, of the Royal Soc. of S. Africa, 16: 170 and 180 (1928); Acocks in Veld

Types of S. Africa p. 36 and 122 (1953), non Lam.

Forest trees, sometimes up to 90 feet tall, ultimate branchlets slender (in herbarium

specimens branchlets somewhat longer and more slender, with leaves less crowded

than specimens of “ subsp. capensis ”). Leaves usually narrowly elliptic, tapering to

the base and apex, sometimes fairly broad a shortly acuminate to base and apex,

5-10 cm. long and 1-3-5 cm. broad; apex often acute with subulate point, sometimes

obtuse; petiole 1-2 cm. long, often dark in lower half (purple when fresh?) usually

patently spreading. Panicles many flowered but not very compact. Calyx shortly

4-toothed, lepidote-pitted, minutely ciliate. Fruit purple when mature, 1 • 5-2 cm.

long and -6-1-1 cm. diam.

Plate 20 and Fig. 10.

Lectotype: Grenfell 869 (K), Pilgrims Rest, Transvaal. This specimen has been

selected as type from the two cited by C. H. Wright in the original description of

O. macrocarpa because it bears fruits and the size of the fruit is the principal

distinguishing feature.

Transvaal. — Soutpansberg: Hanglip, Gerstner 6015. Letaba: Woodbush,

D.F.O. in F.D. Herb. 4329 (syntype of O. macrocarpa C.H. Wr.); 541; 1129; 3272;

Botha in F.D. Herb. 3560; O'Connor 1920. Pilgrims Rest: forest near Pilgrimsrest,

Grenfell 869 (lectotype); Marieps Kop, Keet 1393 in F.D. Herb. 5971; forester in

F.D. Herb. 8101; 9386.

Natal. — Hlabisa: Hluhluwe Reserve, Ward 2247. Nkandhla: Qudeni Forest,

Bayer 811; 804; 805.

Cape. — Tabankulu: Tabankulu Forest, Kriel in F.D. Herb. 6129. Lusikisiki:

Ntsubane Forest, Fraser in F.D. Herb. 6042. Stutterheim: Fort Cunymghame,

Sim 2104. Humansdorp: Storms River Forest, forester in F.D. Herb. 4083; Zahn

in F.D. Herb. 4026. Knysna: Deepwalls Forest, Keet 917, in F.D. Herb. 3939; Keet

in F.D. Herb. 2590; 4085; 4086; Schonland 3595 (GRA); Laughton in F.D. Herb.

8400; 8399; Phillips in F.D. Herb. 7356; forest near Kaatjies kraal, Burchell 5225

cited in FI. Cap. under O. laurifolia (GRA, K); forest, Lam and Meeuse 4723 (L).

[Within bounds of possibility, Burchell 1221, Grootvadersbosch, Swellendam, belongs

here but leaves small and inflorescence very immature.]

Not known to occur outside S. Africa.

This subspecies has been found in the Transvaal, Natal and the eastern Cape

as far west as Knysna, and it is confined to the forest patches in these areas. According

to plant geographers, forest was much more extensive in South Africa 300 years

ago than it is now. Today only patches remain, some in the Transvaal, for example

at Woodbush and Marieskop, in Natal, as at Qudeni, the Transkei, at Tabankulu,

and the well known patches at Knysna and George. Some less well known patches

may be found further west, such as the one at Grootvadersbosch in the Swellendam

district. In the north eastern areas these forests are on the mountain ranges but in the

south they are along the coastal belt.

591

Fig. 10. — Olea capensis L. subsp. macrocarpa (C.H. Wr.) Verdoorn; a, leaf

and fruits from isotype, Pilgrimsrest; b, specimen from Marieps Kop; c, specimen

from Hluhluwe, Natal; d, specimen from Knysna.

592

The identification of the forest trees at Knysna, which bear large fruits and which

had erroneously come to be called O. laurifolia, with O. macrocarpa C.H. Wr. of the

Transvaal, is fully supported by Dr. John Phillips who, as an officer of the Department

of Forestry stationed at Knysna, wrote the following in 1928: “The foliage of the

so-called O. macrocarpa C.H. Wr. from the Zoutpansberg, anatomically is

indistinguishable from that of O. laurifolia [meaning the forest tree with large fruits].

This fact, together with the similarity in the anatomy of the ‘ bark ’ and the similarity

in the flowers and fruits, leads the writer to think that the plants are co-specific ”

(Trans. Roy. Soc. of S.A. 14: 179).

In the Knysna forest, subsp. capensis overlaps in its distribution with subsp.

macrocarpa. There do not seem to be any notes by foresters or forest officers about

hybridization but there are remarks about the difficulty of distinguishing between

the two, which they referred to as “ O. capensis ” and “ O. laurifolia ” respectively.

In the same paper mentioned above Phillips writes: “ The foliage of O. laurifolia

and O. capensis [meaning subsp. macrocarpa and subsp. capensis ], in the absence of

flowers and fruits, is often confused by botanists and foresters. The leaves of these

species are readily distinguished in that the walls of the dorsal epidermal cells of

O. laurifolia (sic) present a definitely wavy appearance in superficial section whereas

O. capensis shows stouter non-wavy walls

As pointed out in this work there is a great deal of variation in subsp. capensis

and in many cases, when it was thought that features had been found by which to

separate it from subsp. macrocarpa, when still more material had been examined it

was found that the differences did not hold. In like manner the feature of the wavy

cell walls found by Dr. Phillips might break down if a large number of specimens are

examined.

The similarity of the leaf shape in subsp. macrocarpa with a few specimens of

subsp. capensis can be seen if those on Fig. 10 are compared with some on Figs. 7

and 8. (Compare Fig. 10a, a drawing of a leaf and fruit from the isotype of subsp.

macrocarpa in the Bolus Herbarium with the leaf-shape of Fig. 8a, Smuts 1100, from

Groote Schuur).

The leaf, Fig. 10b, from a Mariepskop specimen, shows the attenuate point that

is quite frequently found in this subspecies, but never in subspecies capensis. In this

character it approaches the leaf-shape of O. woodiana (see Fig. 4). It will be seen

from the 4 figures on Fig. 10 that in subsp. macrocarpa there is more uniformity of

leaf-shape than in subsp. capensis.

The common name “ Ironwood ”, “Ysterhout”, is usually applied to subsp.

macrocarpa but sometimes it is also called “ Black Ironwood ”, the common name for

subsp. capensis.

4. LINOCIERA.

Sw. in Schreb. Gen. PI. 2: 784 (1791) nomen conservandum. Benth and Hook, in

Gen. PI. 2, 2: 678 (section Ceranthus); Bkr. in FI. Trop. Afr. 4, 1: 19 (1904);

Turrill in F.T.E.A. (1952). Olea Phill. Gen. S.A. FI. PI. ed. 2: 572 (1951) in part,

not Linn.; Harvey ex Wright in FI. Cap. 4, 1: 485 (1907) in part. Dekindtia

Gilg in Bot. Jb. 32: 139 (1902); Turrill in F.T.E.A. (1952). Campanolea Gilg &

Schellenb. in Bot. Jb. 51: 73(1914).

Trees , small to large forest trees. Leaves variable in size and texture, often large,

usually oblong to elliptic or obovate-oblong, opposite, entire, minutely and sparsely

lepidote, especially on undersurface, usually with acarodomatia in the axils of the

veins beneath. Inflorescence axillary, often on previous as well as on this years growth,

cymes sessile or more often paniculate and laxly flowered, peduncles sometimes reduced

or absent giving the impression of fascicled inflorescences; pedicels also sometimes

reduced and flowers then appear glomerate. Calyx deeply lobed sometimes to the

base, lobes broad to rounded at the apex, densely setulose pubescent to glabrous

593

without. Corolla with a very short tube or lobed to the base, sometimes lobes cohering

in pairs (with stamen inserted between them) but tube slit to the base between the

pairs ; lobes with margins deeply infolded and so usually appearing longer than broad,

cucullate at the apex (in African species). Stamens 2 (rarely 4 found in some flowers),

anthers shorter than the corolla-lobes (in African species; as long as, and enveloped

by the corolla-lobes, in the typical species); anther basifixed. Ovary sometimes hispid,

subglobose, narrowed into a short style, sometimes up to as long as, or slightly longer

than the ovary; stigma sub-capitate, obscurely bi-lobed; ovules collateral, attached

ventrally (in typical Linociera attached at apex). Fruit a drupe with a thin fleshy layer,

endocarp rather hard and with a large seed cavity; seed usually solitary, no endosperm,

cotyledons thick.

Fig. 11.

Type species: L. ligustrina Sw., from the West Indies.

The above generic description was drawn up from all available tropical African

as well as South African material because the species throughout Africa form a definite

group and obviously belong to the same genus. This group and the features by which

it differs from Olea were obviously not recognised by past workers. For instance

(1) several members of the group were originally in the genus Olea , such as O.foveolata

and O. peglerae in South Africa; (2) one typical Olea in tropical Africa was at one

time in the genus Linociera, O. welwitschii (Knobl.) Gilg & Schellenb. ; and (3) recently

in the Flora of Tropical East Africa, where Olea and Linociera are both maintained, a

species which fits best in the latter genus is included in the former, O. mildbraedii (Gilg

and Schellenb.) Knobl.

Fig. 11. — Transverse sections of fruits; a, Linociera battiscombei, showing

thick cotyledons and no endosperm; b, Olea capensis L. subsp. macrocarpa, showing

thin cotyledons in endosperm.

The question as to which of the two genera this group should belong is a moot

one. The species comprising it are no more typical of Linociera than of Olea, but

could form a section in either or constitute a separate genus. The reasons for here

putting them under Linociera rather than Olea are because (1) both in Africa, including

the Mascarenes, and in the East (India, Burma, China and New Guinea, etc.), similar

specimens are placed in this genus (see especially Flora of Tropical East Africa, 1952),

and (2) in Bentham and Hooker’s General Plantarum this group cannot be put under

Olea on account of the ex-albuminous seeds, whereas it can be fitted into the section

Ceranthus of Linociera.

a.

b

594

The decision to leave the group under Linociera instead of giving it generic rank

was made because, to take the latter course, would require an investigation of all the

synonyms and related genera and an examination of specimens of Linociera in both

the Old and the New World. Since this is not possible at this stage, the group is

characterised and defined as clearly as possible and is treated in such a way that any

future worker will be able to move it as a whole to whatever position in the family

he considers the correct one. The combination of characters used in the key to separate

this group in Linociera from the genus Olea seem to be sound and divide the specimens

into two natural classes. In one species from tropical Africa, Linociera latipatala

M. R. F. Taylor, one of the most useful distinguishing characters breaks down to a

certain extent, the corolla segments are not much longer than broad but the other

characteristics are present, such as acarodomatia on the leaves, the infolded margins

of the corolla lobes and the ventrally attached ovules. There can be no doubt that in

spite of this one feature it belongs to the group under discussion. This species was the

basis for Gilg and Schellenberg’s monotypic genus Campanolea and was described

by them as Campanolea mildbraedii in 1913. In 1932 it was transferred to Olea by

Knoblach. In 1940 Taylor described Linociera latipetala, which is obviously the same

species and which Turrill cites together with Campanolea mildbraedii as a synonym

of Olea mildbraedii (Gilg and Schellenb.) Knobl. Since there is already a Linociera

mildbraedii Gilg and Schellenb., under this genus it becomes Linociera latipetala M. R. F.

Taylor. Flaving access to the specimen Dummer 5473 which is cited by Taylor, this

study of the species was made possible and provided the authority for sinking the

genus Campanolea.

If future workers decide on giving this group generic rank it is possible that

Dekindtia will be found to be the first generic name applied to a member of the group

[see discussion under Linociera battiscombei (= Dekindtia africana)].

Key to Species.

Inflorescence a few- to several-flowered axillary cymose panicle:

Leaves on flowering branches rarely more than 7 cm. long, if longer then

under 3 cm. broad 1. L. foveolata,

(aggregate sp.).

Ultimate twigs glabrous, if somewhat pubescent leaves more

than twice as long as broad:

Small trees up to 30 feet tall, fruit up to 2 cm. long; leaves

usually more than twice as long as broad. . la. subsp. foveolata.

Forest trees up to 90 feet tall; fruits up to 2-7 by 2 cm.;

leaves usually not more than twice as long as

broad lc. subsp. major.

Ultimate twigs tomentulose; leaves not usually twice as long as

broad mostly ending in a broad acumen with

recurved or retuse apex lb. subsp. tomentella.

Leaves on flowering branches usually 9-13 cm. long, often broadest in

upper half, narrowing gradually to the base, shortly

acuminate to the apex and abruptly narrowed into a broad

obtuse 5-10 mm. long acumen; petioles 1 cm. or more

long 2. L. peglerae.

Inflorescence a reduced cyme with flowers sub-sessile and appearing glomerate

in the leaf axils; young inflorescences densely pubescent sessile

globules; calyx segments up to 4-5 mm. long; trop. African

species 3 . L. battiscombei.

1. L. foveolata (E. Mey.) Knobl., aggregate species.

(a) subsp. foveolata.

(b) subsp. tomentella Verdoorn subsp. nov.

(c) subsp. major Verdoorn subsp. nov.

595

Small trees, sometimes up to 30 feet tall or forest trees up to 90 feet tall. Ultimate

twigs short, usually with 2 pairs of leaves, glabrous, or tomentulose. Leaves varying

in size and texture, glabrous or sparsely pubescent at base and on midrib, ovate-oblong,

oblong-elliptic, or oblong, rounded or shortly cuneate at base (in coppice shoots rounded

to cordate at base), shortly or gradually narrowed to an obtuse apex, or with a broad,

sometimes retuse, acumen at apex, 2-7 cm. long (if longer then under 3 cm. broad)

and 1 • 3-3 • 7 cm. broad, very minutely and obscurely scaly, especially on lower surface,

appearing as if minutely pitted, acarodomatia usually present in the axils of the leaves

below; petiole 2-7 mm. long. Inflorescence short, laxly flowered, cymose panicles,

axillary and on the old wood, glabrous, or bracts and calyces setulose, glabrescent,

0- 5-3 cm. long, peduncles sometimes reduced, giving the appearance of fascicled

inflorescences. Calyx 4-lobed to beyond the middle; lobes rounded at the apex,

usually ciliate. Corolla white sometimes tinted pink, sweetly scented, 4-lobed almost

to the base and between alternate lobes right to the base; lobes about 4 mm. long with

margins deeply infolded and apex cucullate (like a mocassin toe), appearing longer

than broad. Stamens normally 2 (4 found in some flowers), inserted on the short tube

between alternate lobes, filaments fused on corolla-tube and only slightly longer than

it; anthers about 1-75 mm. long and 1-5 mm. broad. Ovary subglobose, stigma

subcapitate, obscurely 2-lobed; ovules 2, attached ventrally, usually near the base.

Fruit green turning blackish or purplish black when mature, sub-globose to oblong,

1- 5-2 -8 cm. long and 0-8-2 cm. broad.

The aggregate species, Linociera foveolata, is characterised by the comparatively

small leaves, seldom over 7 cm. long but if so then under 3 cm. broad and the laxly

flowered, glabrescent, axillary panicles, usually 1 to 3 cm. long.

This species divides naturally into 3 groups and these groups have been given

the rank of subspecies because although distinguishable seem to be more closely related

to each other than to the other species of Linociera , also the distinguishing characters

occasionally grade into each other and while in general it will be easy to decide to

which subspecies a specimen belongs there will be some cases when it will not be

possible, although the specimen could doubtlessly be classified as L. foveolata in the

broad sense.

The distribution of the three subspecies can be seen on the map reproduced here as

fig. 12.

{a) subsp. foveolata. L. foveolata (E. Mey.) Knobl. in Repert. Nov. Spec. 41: 151

(1937); Olea foveolata E. Mey. Comm. PI. Afr. Austr. 176 (1837); DC. Prod.

8: 285 (1844); Harv. ex Wright in FI. Cap. 4, 1: 485 (1909) pro parte; L.

marlothii Knobl. in Repert. Nov. Spec. 41: 151 (1937).

The typical subspecies is distinguished by its glabrous or sparsely pubescent

branchlets, never tomentulose; its leaves usually more than twice as long as broad,

shortly acuminate to an obtuse apex; and the fruit usually under 2 cm. long and 1 • 1 cm.

broad but not larger.

Plates 21 and 22.

Type: Drege s.n., from “between Hoffman Kloof and Driefontein ” in the

Zuurbergen, Uitenhage. Type of synonym: L. marlothii, Knobl., Rudatis 1416,

Dumisa, Natal.

Transvaal. — Lydenburg: Waterfall, Codd & Verdoorn 7613; Groblersdal:

Loskop Dam, Codd 8434. Rustenburg: near Ananda Guest Farm, Rose Innes 207,

211,286; Marais 280, 281, 286. Brits: Magaliesberg, opposite Wolhuterskop turn

off, Marais.

129828-7

596

Natal. — Ingwavuma: Gwoloweni Forest, Bayer 765. Entonjaneni: Umhlatuzi

Valley, Lawn 1925 and 1927, fruit rather large, 2 cm. long and 1-2 cm. broad (NH).

Estcourt: Dalton Bridge, Wright, West and Acocks 12; Umhlumba Mountain,

West 1463; Tabamhlope, West 1135. Ndwedwe: Groenberg near Inanda, Wood

1290. Pietermaritzburg: Impendhle, Acocks 13772; Emkazeni, Houshold in F.D.

Herb. 3180; Polela: Bulwer, Good in F.D. Herb. 6121; Boocock in F.D. Herb.

4101 (narrow leaf variant); Emkayeni forest, Fernando 91. Umzinto (Alexandra):

Umgaye, Dumisa, Rudatis 1416, isotype of L. marlothii Knobl. (3 sheets).

Cape. — Umzimkulu: Miller in F.D. Herb. 6123; Adam forest, Leigh in F.D.

Herb. 6123 (long, narrow rather thin leaf, like L. marlothii); Insikeni: Dlokolwana

forest, Miller in F.D. Herb. 6048. Libode: Hobokazi forest, Miller in F.D. Herb.

6048. Komgha: in woods, Flanagan 270 (3 sheets). Stutterheim: Fort Cunynghame,

Sim 2103; Kabaku Hills, Acocks 8964. East London: in Queens Park but wild,

Galpin 8219; King 1. Keiskamahoek: Mpamba, Stayner 52. King Williams Town:

Pirie, Sim 1334 (BOL). Albany: Fern Kloof, Story 4497, 4498; Britten s.n. (GRA).

Bathurst: Kowie, Britten 1847. Alexandria: south of town, Story 270. Port

Elizabeth: Longmore Kloof, Taylor 840. Uitenhage: Zuurbergen between

Hoffmanskloof and Driefontein, Drege s.n., isotype (L).

Not known to occur outside South Africa.

An isotype of the species, Drege from the Zuurberg, labelled in Meyer’s

handwriting “ O. foveolata E. Mey. ” and seen by Knoblauch was sent on loan to the

National Herbarium, Pretoria, from the Ryksherbarium, Leiden (see pi. 21). An

examination of this specimen showed that it compared well with specimens from regions

eastwards and northwards of this locality, but not to the west of it. These specimens,

although varying within limits, form a definite group, which is distinguishable from

certain plants found on the Peninsula, at Knysna and elsewhere which, both Wright

in the Flora Capensis and Adamson in the Flora of the Cape Peninsula, confused with

O. foveolata E. Mey. The main distinguishing feature of these two groups is the

glabrous ultimate branchlets in the group containing the Drege specimen, and the

tomentulose branchlets of the group in which specimens from the Peninsula are found.

Added to this the leaves in the former group are usually more than twice as long as

broad, narrowing evenly to an obtuse apex and with the lateral veins not spreading

very widely, as opposed to the rather broad (not twice as long as broad) leaves, shortly

narrowed towards the apex, and then abruptly narrowed into a broad retuse acumen,

with the lateral nerves widely spreading, in the tomentulose group, described below

as a new subspecies.

Of the three subspecies, L. foveolata subsp. foveolata is the commonest and the

most widely spread. It is also the most variable, but, to date, the variants have not

been found to occur in sufficient numbers to form a group that could be constituted

a variety. L. marlothii Knobl. has, for this reason, been put into synonymy without

even varietal rank. In the type gatherings, which were collected at Dumisa, Natal,

the leaves are long, up to 8 cm. long and of a thin texture with a light yellowish green

colour on the lower surface. No specimen seen compares exactly with this material,

of which there are a number of duplicates.

Another variant in Natal has small narrow leaves, about 3-5 cm. long and 1 cm.

broad, but it too has not been found to occur repeatedly in the same form. In both

these variants some of the leaves grade into the typical shape and size and so, for the

present, they are classes as L. foveolata subsp. foveolata.

The distribution of this typical subspecies extends from the Transvaal through

Natal and the Transkei to Port Elizabeth and Uitenhage, but not further westward.

Fig. 12.

597

(b) subsp. tomentella Verdoorn, subsp. nov. a typica ramulis ultimis tomentulosis

differt. O.foveolata Harvey ex Wright in FI. Cap. 4, 1 : 485 (1907) in part; Adam-

son in FI. of the Cape Peninsula p. 670 (1950), non E. Mey.

Besides the tomentulose branchlets this subspecies can be recognised by the

following features: leaves rather thin, rarely subcoriaceous, broadly elliptic to oblong,

mostly less than twice as long as broad, 2-6 cm. long and 1-3 cm. broad shortly

narrowed to base and apex, and usually ending in a broad acumen which is retuse at

the apex, sometimes sparsely pubescent at the base and on the midrib, at least on leaves

on the new growth, lateral veins spreading at a wide angle, petiole very short, 2-5 mm.

long; inflorescence -5-1-5 cm. long with the bracts and calyx often densely setulose,

glabrescent; fruit oblong-elliptic, green turning black, up to 1-5 cm. long and -8 cm.

diam.

Plate 23.

Type: Bur che.il 5539 (PRE), Knysna forest, near the Knysna River ford.

Cape. — Lusikisiki: near Fraser Falls, Acocks 13433 (rather thin small-leaved

form). Engcobo: Egossa forest, Sim 2472 (rather thin and small leaves). Kentani:

Pegler 826 (BOL). East London: Ndhni 137. Albany: Fernkloof, Story 4496;

4483. Bathurst: Port Alfred, Story 4493 and 4529 ; Kowie, Britten 1879, 2104 (GRA)

and 2998 (GRA). Alexandria: Coast Reserve, Strauch in F.D. Herb. 3247; Main

Forest, Sim in F.D. Herb. 3165. Port Elizabeth : Sister Antony 29 (GRA) ; Krakakama,

Burchell 4515. Knysna, Burchell 5539 (type, PRE); Gouna Forest, Keet 913; Salt

River Forest, Keet 526; Belvedere, Duthie 584 (GRA); Sanjulie, Phillips 139 (GRA);

598

Buffalo Bay, Keet in F.D. Herb. 2591 ; Deepwalls, Laughton 461 ; Keet in F.D. Herb.

2393; Kaffirkop Rd., Phillips in F.D. Herb. 139a. Bredasdorp: Bosch Kloof, Pillans

9448, leaves only, large, no new twigs? Peninsula: without precise locality, Mund &

Maire s.n.; Table Mountain, Orange Kloof, Adamson 877; Slangoolie Gorge,

Marloth 1 1925 a & b; 4405 (BOL); Hout Bay, Compton 17069 (NBG); Skoorsteenkop,

Acocks 639, coppice branch; Kirstenbosch, Zeyher 243, leaves only, (BOL); Skeleton

Gorge, Compton 10020, coppice leaves, (NBG); Wynberg Hill, Compton 15452 (NBG);

Oudekraal, Kolbe 2519 (BOL); Millers Poort, Pillans 9890, leaves only, (BOL);

Llandudno, Isaac in Bolus Herb. 25301 (L, BOL); Witsand, Lotsy and Goddyn 1867.

[Gerstner 4591, from the Nkandhla forest, Natal, may belong to this subspecies. It

is a variant with thin leaves, some twice as long as broad, twigs slender, the inflorescence

very short and densely setulose with yellowish hairs.]

Not known to occur outside South Africa.

As pointed out under the typical subspecies, specimens which belong here were

cited under Olea foveolata E. Mey. in the Flora Capensis. One of the cited specimens,

Burchell 5539 has been selected as the type of the new subspecies, the holotype being

the specimen in the National Herbarium, Pretoria. Isotypes were seen in the Bolus

Herbarium, the Albany Museum and Ryksherbarium, Leiden.

L. foveolata subsp. tomentel/a occurs on the Peninsula, at intervals eastwards along

the coast and in the forests in the eastern Cape, the Transkei and probably in Zululand.

In the coastal localities the specimens are fairly uniform and typical, but the forest

specimens show some variation. In these the leaves are in a degree thinner and the

inflorescences inclined to be shorter and more densely setulose pubescent. In other

respects, such as leaf size and shape, these forest specimens vary among themselves

and do not form a recognisable group. They are, therefore looked upon merely as

variants within the subspecies. (For distribution see Fig. 12).

(c) subsp. major Verdoorn subsp. nov. a L. foveolata typica fructis majoribus usque

2-8 cm. longis 2 cm. latis, arbore majore, usque 30 m. alto, differt.

This subspecies is distinguished principally on the large fruits which are about

the size of a small walnut. Before maturity these drupes have a rather characteristic

shape being more or less ovate in outline, sub-quadrate and with a ridge over the obtuse

apex, rather like a duck’s bill (see Fig. 13). When mature they are more or less oblong,

almost 3 cm. long and 2 cm. in diameter. The flesh is rather thin and the mesocarp

hard with quite a large seed cavity. Other distinguishing characters are the glabrous

branchlets, the larger leaves, up to 7 cm. long and 3-7 cm. broad, and the habit being

a tall forest tree.

Plate 24.

Type: Urry in National Herbarium, Pretoria, 28568, Marieps Kop, Pilgrims

Rest, Transvaal.

Transvaal. — Letaba: Woodbush, Botha in F.D. Herb. 3986. Lydenburg:

Magalieskop, Kotze in F.D. Herb. 2823 and 2829 (leaves only). Pilgrimsrest :

Mariepskop Forest, Loock s.n. (3 sheets); Urry in National Herbarium 28568, type,

2 sheets. Keet in F.D. Herb. 5938; Loock in F.D. Herb. 9387 and 9538; Scheffler

n F.D. Herb. 9929; 10048; 10049; [O’Connor in F.D. Herb. 2014, Woodbush

(no fruit, leaf rather long and narrow) and Renny DE3 (no mature fruit, leaves broad)

cannot be determined subspecifically but may belong here.]

Not known to occur outside South Africa.

To date this subspecies has been found only in high altitude forests in the eastern

Transvaal (Fig. 12). Without seeing the fruit some leafy specimens could be mistaken

for the typical subspecies for although, in general, the leaves are broader yet some of

the broad leaved variants in one may grade into the narrow leaved form of the other.

599

In the large fruits L. foveolata subsp. major resembles L. peglerae and

L. battiscombei, differing from the former in the smaller leaves, up to 7 cm. long as

against 13 cm. long in L. peglerae, and from the latter in the inflorescence being branched,

not glomerate.

The following are some of the notes from forest officers on this tree: “ Large

angular drupes, not mature; common at higher altitudes and one of the more important

species (commercially), but trees short and crooked in bole and contour, little

merchantable timber.” The following note is in part rather contradictory, “ a large

tree 80-90 feet high with a single large erect cylindrical bole 40-50 feet to first branch,

DBH 24 in., bark grey, fissured, \ in. to 1 in. thick, heavy crown about 30-40 feet

across. The fruits are eaten by birds, monkeys and bushpigs.”

Fig. 13. — Fruit of Linociera foveolata (E. Mey.) Knob, subsp. major Verdoorn

X 2.

2. L. peglerae (C.H. fVr.) Gilg & Schellenb. in Bot. Jb. 51 : 71 (1914). Olea peglerae

(in error O. pegleri) C.H. Wr. in FI. Cap. 4, 1: 485 (1907).

Tree in tall forest, about 60 feet high, new growth glabrous. Leaves large, the

majority 8-13 cm. long and 3-6-5 cm. broad (coppice leaves 19 by 8 cm., petiole short,

base cuneate, not subcordate as in coppice shoots of L. foveolata), oblong, oblong-

elliptic or obovate-oblong, often widest in upper half, usually long cuneate at base,

shortly narrowing towards apex and abruptly narrowed int* a broad obtuse acumen

at the apex, acumen up to about 1 cm. long, midrib prominent beneath, lateral veins

obvious, spreading at about 45°, usually with acarodomatia in the axils; petiole 1 cm.

or more long, glabrous. Inflorescence of cymose panicles up to about 5 cm long

glabrous, branches long and pedicels short; bracts sparsely pubescent glabrescent,

sometimes ciliate. Calyx lobed to middle or beyond, lobes rounded at the apex,

minutely ciliate. Corolla cream, lobes more or less 4 mm. long, united in pairs, slit

to the base between pairs, margins infolded and apex cucullate (mocassin-like).

Stamens inserted on joined portion between united pairs, anthers basifixed, 1 -75 mm.

long. Ovary subglobose, somewhat 4-lobed; style short, stigma terminal, sub-capitate,

more or less 2-lobed; ovules attached along ventral side. Fruit 1-7-2 -5 cm. long,

1-2-1 -4 cm. broad, before mature somewhat ovate and quadrate in upper half with

ridges across the top (duck’s bill), sometimes faint. (Quite mature fruit not seen).

Plate 25

Type: Peg/er 819, Kentani. *

Cape. — Mount Curry: Fort Donald, Forester in F.D. Herb. 9437. Engcobo:

Gora Forest, Mannina, Laughton in F.D. Herb. 9162 (leaves only); Zahn in F.D. Herb.

2045. Kentani: near Kentani, Peg/er 819 (type number); Manubie, Forest Officer

in F.D. Herb. 7965? Cult. Kirstenbosch, new area s.n. (NBG).

600

[N.B. — A specimen, Jackson in Natal Herb. 37523, from St. Lucia Bay, looks

as if it might be this species; the leaves are poor (galled), up to 10 cm. long and

2 ’8-3 -5 cm. broad, acumen pronounced on one, 3 fruits, 2 by 1-5 cm., oblong, only

very slightly narrowed at top, ridge not obvious; also a leafy specimen collected by

Bayer in Qudeni Forest may be this species.]

Not known to occur outside South Africa.

L. peglerae has the more typical leaf characters of the common tropical African

species of Linociera being rather large, long cuneate at the base with a fairly long petiole

and the lateral veins clearly raised on the under surface. In fruit characters it is like

L. foveolata subsp. majora and L. battiscombei (— Dekindtia africana ) with drupes

developing to a large size even before they become mature and fleshy and, in the early

stages, having a blunt ridge across the apex (like a duck’s bill).

To date this species has not been collected very often and so is not at all well known.

The following notes from sheets in the Forest Department Herbarium are therefore

useful: “It is known at Manubie Forest as Bastard Black Iron Wood and by the

Natives as Umdlebe. [On another sheet the Native name ‘ Umqumaswele ’ is given].

This tree flowers in August and seed ripens in November or early in December. It

grows fairly straight, to about 50 feet in height, with a diameter from 9-18 in. It is

rarely purchased by Sawyers although I have frequently known them to work the

wood up and attempt to sell it as O. laurifolia [meaning O. capensis subsp. macrocarpa ]

to wagon builders. It is reported to be much softer than [subsp. macrocarpa ] and is

subject to heart crack. It is fairly plentiful at Manubi Forest and generally throughout

mountain forests of Transkei ”. Miss Pegler writes of it “ Large forest tree; glossy

leaf, insignificant cream flowers.” She does not mention a scent whereas several

collectors of L. foveolata subsp. foveolata mention that the flowers of that subspecies

are sweetly scented.

When first published in the Flora Capensis the epithet was incorrectly given as

“ pegleri ” whereas it should be “ peglerae ” since the collector is cited as “ Miss

Pegler ”.

3. L. battiscombei Hutch , in Kew Bull. 1914: 17. Dekindtia africana Gilg in Bot.

Jb. 32: 139 (1902) non Linociera africana Knobl. (1834) nec (Welw.) Gilg &

Schellenb. (1913).

Tree, mostly 12-15 feet tall with spreading canopy (in tropical african specimens

up to 30 feet tall); ultimate branchlets (new growth) appressedly pubescent or

puberulous, glabrescent. Leaves narrowly to broadly elliptic or oblanceolate to

obovate-elliptic, 5-7-5 cm. long and 2-3-5 cm. broad (in tropical african material

seen from 5-12 cm. long and 2-4-5 cm. broad), long or shortly cuneate at the base,

acuminate, sometimes shortly so, often suddenly narrowed near the apex into a blunt

acumen, lateral veins obvious on both surfaces, spreading at a fairly wide angle, usually

over 45°, often with acarodomatia in the axils of the veins; petiole 3-9 mm. long.

Inflorescence a reduced cyme, flowers appearing glomerate in the axil of the leaves,

few to many in clusters; bracts and calyx densely adpressedly setulose pubescent

without. Calyx deeply 4-lobed, the lobes up to 5 mm. long and opposite pairs slightly

unequal in size. Corolla united at the base into a very short tube, lobes with margins

infolded, apex cucullate (slipper toe). Stamens inserted on tube, filaments very short;

anthers basifixed. Ovary slightly narrowed into a very short style; stigma terminal,

2-lobed; ovules attached along ventral face. Fruit blackish purple when ripe, about

1-8 cm. long and 1 cm. diam.; cotyledons thick, no albumen.

Plate 26.

Type: Battiscombe 517, Nairobi forests, Kenya. Type of synonym: Dekindtia

africana Gilg, Dekindt 73, Chella Mountains, Angola.

601

Transvaal. — Zoutpansberg: Tshakoma, Obermeyer 1080? (leafy specimen only).

Petersburg: Blaauwberg, Codd 8737; 8737a; Codd & Dyer 9118.

Also occurs in S. Rhodesia and northwards to Kenya and Angola.

When describing Dekindtia the author, Gilg, contrasted his new genus with Olea

and did not mention its relationship to Linociera. The principal distinguishing

characters mentioned are “ the compact axillary inflorescences in the form of sessile

nodules . . ., the shape of the corolla with its short tube and strongly elongated lobes

with their turned-in margins and apices.” In these corolla characters it fits in with the

group of specimens here put under Linociera and therefore in this genus the apparently

glomerate flowers would be the only distinguishing character. But there is a species,

L. congesta Baker, from the Cameroons which is described as having flowers “ in

sessile or nearly sessile clusters in the axils of the leaves,” also among the specimens

under L. foveolata subsp. tomentella one from Zululand has a very much reduced cyme.

Therefore even this character breaks down and the genus cannot be maintained as

distinct from the section of Linociera here defined. The epithet “ africana ” has

already been combined with Linociera [see L. africana Knobl. Bot. Centralbl. 61 : 129,

1894, and L. africana (Welw.) Gilg & Schellenb. Bot. Jahrb. 51 : 61, 1913], for species

distinct from this one. The next specific epithet for a specimen which is obviously

the same species, is Linociera battiscombei Hutch, and this is cited by Turrill as a

synonym under Dekindtia africana in the Oleaceae, Flora of Tropical East Africa.

It is unfortunate that there is no way of indicating when writing the name and the

author that the species was previously described under another name.

From collectors notes it appears that this species grows at high altitudes along

mountain streams. The Transvaal specimens were collected on the Soutpansberg

and the Blaauwberg, the latter an isolated mountain up to 6,700 feet high, rising from

the flats to the west of the Soutpansberg (see Fig. 12). The collector’s notes state that

the trees on the Blaauwberg were growing on a wooded stream bank at about 4,500 feet.

The other record of this species in South Africa, is only a leafy specimen from Tshakoma

in the Soutpansberg.

Among the South African species, L. battiscombei is in some respects near

L. foveolata subsp. tomentella. Where the last mentioned has the new growth

tomentulous, it is puberulous in the former. Both have rather broad leaves and, in

some specimens in the subspecies the inflorescence is much reduced and the bracts

and calyx lobes quite densely setulose; in these last mentioned instances the much

smaller flowers, especially the calyx lobes, which are usually under 2 cm. long

distinguish subsp. tomentella from L. battiscombei, which has calyx lobes up to 5 mm.

long.

5. MENODORA.

Humb. & Bonpl. PI. Aequin. 2: 98, t. 110 (1809); Harv. ex Wright in FI. Cap. 4,

1 : 483 (1907); Steyermark in Ann. Missouri Bot. Gard. Vol. 19: 87 (1932);

Phill. Gen. S.A. FI. Plants ed. 2: 572 (1951). Bolivaria Cham. & Schltr. in

Linnaea 1 : 207 (1826). Calyptrospermum A. Dietr. in Linn. Sp. PI. ed. 6, 1 : 226

(1831).

Perennial, suffruticose or sub-herbaceous from a woody base, erect or diffuse

with a strong tap-root. Leaves simple or divided opposite, sub-opposite or alternate.

Inflorescence cymose, paniculate or sometimes reduced to a single flower and then with

monochasial development. Calyx persistent, united at the base with 5 to 10 lobes,

lobes sometimes cleft. Corolla united, tube usually half the length of the lobes, usually

pilose within at the insertion of stamens, lobes usually 5, imbricate. Stamens 2, filaments

inserted on the tube and decurrent to its base, free above and exserted ; anthers attached

602

near the base, erect, oblong, dehiscing by longitudinal slits. Ovary 2-lobed, 2-celled,

ovules 2-4 in each cell, collateral, attached to the wall of partition about midway.

Fruit a bi-locular capsule, usually both cocci developing, circumscissilely dehiscent.

Seeds ex-endospermous, 1 to 4 in a coccus, reticulated.

Type species: M. helianthemoides Humb. & Bonpl. from Mexico.

The distribution of the genus Menodora is very interesting and can be used as

evidence of the surmised land-bridge, Gondwana, that is thought at one time to have

connected South America with South Africa. The three areas of distribution can be

seen on Steyermark’s map reproduced here (see Fig. 14) from his revision of the genus

in the Annals of the Missouri Bot. Garden, Yol. 19, 1932. They are (1) S.W. United

States and Mexico, (2) Central and southern S. America and (3) South Africa. The

area marked in South Africa is not quite correct and must be extended to Namaqualand

in the western Cape, and the Little and Great Karoo in the South (see notes under the

species M. juncea and fig. 15).

At first sight it may seem hard to believe that this genus belongs to the family

Oleaceae and indeed the history of its classification makes interesting reading (see

Steyermark’s Revision l.c.) but a critical examination will show that there can be no

doubt at all about its close relationship to Jasminum, and that its correct place is

therefore in the same family. It is interesting to know that the authors of the genus

originally, but only tentatively, referred the genus to “ Jasmineae ” which is now

Oleaceae. In the next three decades it was placed in turn in Acanthaceae, Gentianaceae

and a separate family Bolivariaceae described to take Menodora and its synonym

Bolivaria (Bolivaria was considered distinct at the time).

In the family Oleaceae the genus Menodora is distinguished by its fruit which is

a bi-locular capsule, both cocci usually developing. Except for one species in the

United States the dehiscence is by a horizontal suture, the top of the capsule coming

off like a cap and exposing the seeds.

Key to Species.

Rigid suffrutex; leaves simple, usually much reduced; pedicel erect in fruit. ... 1. M. juncea.

Diffuse sub-herbaceous plants from a woody base; leaves mostly lobed or

pinnatisect; pedicel cernuous in fruit;

Leaves mostly bi-pinnatisect, segments subulate: calyx-lobes multipartite . . 2. M. africana.

Leaves mostly 3 to 5-lobed, segments 1-5-2 -5 mm. broad; calyx-lobes

usually undivided 3. M. heterophylla

var. australis.

1. M. juncea Harv. Gen. PI. ed. 2: 220 (1869);* Harvey ex Wright in FI. Cap. 4,

1: 484 (1907); Steyermark in Ann. Missouri Bot. Gard. 19: 150.

Suffrutex, virgate, 1-5 feet tall, branches rigid, suberect, terete, striate, minutely

silvery puberulous. Leaves sessile, remote, linear, cuneate at the base, often much

reduced, close pressed, 0-3-3 cm. long and up to 3 mm. broad. Inflorescence of one

or several terminal, 3-flowered cymes. Calyx 5-6-lobed, tube usually 2-5-3 mm.

long, lobes linear to linear-acuminate from a triangular base, 2-5-8 mm. long. Corolla

yellow; tube about 6 mm. long; lobes oblong to oblong-obovate, 1-2-2 cm. long

and -7-1-4 cm. broad, broadly rounded at apex, mucronate. Stamens exserted;

anthers 4-6 mm. long. Ovary deeply 2-lobed; style filiform up to 1-4 cm. long;

stigma small, capitate. Fruit a 2-lobed capsule, sometimes only one lobe developing,

sub-globose up to 9 mm. long and 8 mm. broad, the coat becoming chartaceous and

dehiscing circumscissilely. Seeds black (seen through capsule wall on one of the

borrowed specimens).

Plate 27.

* Since this edition was published posthumously and edited by J. D. Hooker it has been suggested

that “ Hook, f.” is the correct authority but see introduction.

603

Type: Whitehead s.n., Modderfontein, near Springbok, Namaqualand.

Cape. — Namaqualand: Richtersveld, Kubus, Marloth 12282b; west of Anenous

Mt., Taylor 1132 (BOL); Kamieskroon, Acocks 14996; Pearson 5639 (BOL);

Soebatsfontein Road, Thorns s.n. (NBG); Numees Mine, Pillans 5108 (BOL);

Stinkfontein, Mathews in Herb. Bol. 25412. Laingsburg: Near Grootfontein, Marloth

8347. Prince Albert: near Prince Albert Road, Marloth 4519 (BOL); Beaufort

West: near Rosesberg Pass, Acocks 15885. Without locality, Scully in S.A. Mus.

Herb. 41539 (SAM).

Endemic in the western and south western Cape.

The distribution of this species is in the karroid areas of Namaqualand, in the

western Cape, and the Great and Little Karoo in south central Cape. It does not

occur very generally in these regions, but is found in restricted localities within them.

At such localities the individuals occur in fair numbers.

As pointed out in the notes under the genus the distribution in South Africa shown

by Steyermark in his map (see Fig. 14) has to be extended. Steyermark had only one

gathering, a Zeyher specimen, of Menodora juncea. The locality for Zeyher’s specimen

(see Linnaea 19: 590 No. 94) is rather vague for it reads “ Georg, Karoo in the clearing

of the Gouritz River, 1,000 ft. (IV, B, b, 14) ”. Today the district of George is not

near the Gouritz, but in the maps of Zeyher’s day it is shown to extend westwards

to that river. Today the Mossel Bay and Riversdale districts lie one on each side of

the Gouritz river. From Drege’s map “ IV, B, b ” is a strip inland from the coastal

strip and this places the locality somewhere about the northern boundries of the

districts mentioned that is near Herbertsdale which is similar veld to that in which the

species occurs elsewhere. So Steyermark’s locality which is shown by a circular spot

near the coast in the south (see Fig. 14) should be moved slightly to the west and

extended to the Great Karoo and the Namaqualand Karroid veld to complete the

distribution of the genus as it is known today (see Fig. 15).

Fig. 14. — General distribution of the genus Menodora taken from Steyermark’s

revision in the annals of the Missouri Botanic Gardens Vol. 19 (1932).

604

Menodora juncea is very distinct from the other two South African species and

has its allies in the Americas. It is a stiff erect suffrutex with remote leaves, most of

them much reduced. The flowers are in compound cymes with the pedicels erect in

fruit instead of, as in the other two species, 1 -flowered inflorescences with monochasial

development which results in the pedicels being cernuous in fruit (compare plates 27

and 28).

2. M. africana Hook. Ic. PI. t. 586 (1843); Wood & Evans in Natal Plants 3: 17,

PI. 240 (1902); Harv. ex Wright in FI. Cap. 4, 1: 484 (1907); Steyermark in

Ann. Missouri Bot. Gard. 19, 1: 123 (1932); FI. PI. of Africa 30, pi. 1187 (1955).

Undershrub with many slender branches from a woody base; branches

sub-herbaceous, slender, about 5-25 cm. tall, ridged, the ridges formed by decurrent

bases of the petioles, sparsely scabrid. Leaves alternate, sub-opposite or opposite

towards the base, sub-sessile to petioled, up to 1-5 cm. long, bi-pinnatisect with

occasional simple or simply pinnate leaves, segments narrow, more or less subulate

with inrolled scabrid margins. Inflorescence a reduced cyme with monochasial

development; flowers solitary, terminal, appearing lateral when the main branch

turns aside and a secondary one develops as the main axis and overtops the flower.

Calyx sparsely or densely scabrid without; tube 2-3 mm. long; lobes longer than the

tube, multipartite, segments more or less subulate. Buds red. Corolla yellow, tinged

reddish; tube about 4 mm. long, narrow below widening at the mouth, sparsely pilose

in the throat; lobes oblong about 1 cm. long and 4-5 mm. broad, rounded at the

apex, mucronate, slightly narrowing towards the base. Stamens 2, filaments inserted

in the corolla-tube, free for 3-4 mm. pilose where affixed to tube; anthers 3-4 mm.

long, 1-5 mm. broad, sub-basifixed, erect, sometimes minutely mucronate. Ovary

2-celled, bi-lobed, ovules 4 in each cell, axillary; style about 9 mm. long; stigma small,

capitate, terminal. Fruit a bi-locular capsule sometimes only one locule or coccus

developing, borne on a cernuous pedicel, coat parchment like, dehiscing circumscissilely

when ripe; seeds usually 2 or 3 in each coccus, about 1 cm. long, more or less oblong,

outer coat reticulated.

Plate 28.

Type: Burke 134 (“ 1341 ” in error in Hooker’s leones). Yet River, Orange Free

State.

Bechuanaland Protectorate. — Kanye: Pharing, Hillary & Robertson 480.

Transvaal. — Petersburg: near Pietersburg, v. d. Merwe 2272; Meeuse 9153.

Waterberg: near Palala, Smuts & Gillett 3371 ; Warmbaths, Sidey 1308; Leeuwpoort,

Rogers 22393 (J). Pretoria: Moss 10142; 6 m. S. of Pretoria, Verdoorn in National

Herbarium 28544; Comins 860; 3 m. S. of Pretoria, Codd 1740; near Irene, Smith

1099; Swingbridge area, Repton 3332; Arcadia, Stent in herb. 9563; Wonderboom

Reserve, Repton 2765; Brooklyn, Mogg 15217; Derdepoort, Leendertz 369 (L, GRA);

Hatherly, Rogers 109 (GRA); Magaliesberg, Zeyher 1132 (BOL); near Pretoria,

Bolus 25415 (BOL); McLea in Herb. Bol. 25414 (not “ 3104 ” as cited in FI. Cap.?)

(BOL); Rooi Kop, Smuts & Gillett 2543; Pole-Evans 1249; Rust-der-Winter,

Gerstner 5527. Lydenburg: Wilms 1068 (L). Barberton: Queen’s River, Galpin

1071b; near Corocodile Poort, Galpin 1071 (GRA). Standerton: near Val Station,

Smuts 397. Heidelberg: Lagerspoort, Prosser 1650. Germiston: Palmietfontein,

Gilliland 26812. Modderfontein : Haagner s.n. (GRA). Johannesburg: Thorntree

Kloof, Moss 6605 (Wits.). Vereeniging: Gilfillan 148. Potchefstroom : West of

Potchefstroom, Story 760; Panfontein Reserve, Louw 1977; Boskop, Louw 355;

Welverdiend, Louw 72; School of Agriculture, Liebenberg 934; 953; 985.

Wolmaransstad: Liebenberg 3022. Christiana: Commonage, Burt Davy 12480;

Kameelpan, Theron 439.

605

Fig. 15. — Distribution of 3 species of Menodora in Southern Africa.

Cape. — Mafeking: Moshesh, Brueckner 420; Appleyard in SAM. Herb. 18090

(SAM); Pitsani Road, Bolus 6433. Vryburg: Armoedsvlakte, Mogg 8024. Between

Kuruman and Vaal, Cruickshank in Herb. Bol. 2537 ; Between Kuruman and Vryburg,

Thorne in SAM. Herb. 54476 (SAM). Griquatown: Postmasburg, Wilman 9116.

BarklyWest: Mar loth 958 ; Hebron, Flanagan 1472. Kimberley: Elliott s.n.; Moran

s.n. ; Moran 86 (GRA); Hafstroom H 902; Oliver s.n. (SAM); du Toit’s Pan,

Tuck in SAM. Herb. 18089 (SAM); Bolus 25413 (BOL); Mostert’s Hoek, Acocks

and Hafstroom H992.

Orange Free State. — Without precise locality: Vet River, Burke 134, Isotype

(BOL); Draaifontein, Rehmann 3620 (BOL). Parys: Moss 13499 (J). Heilbron:

Goossens 410. Kroonstad, Pont 228; Sehweickerdt 1094. Senekal: Doornkop,

Goossens 782. Bloemfontein: Glen, Pole-Evans 19627. Fauresmith: Heuningberg,

Marais 172; Klipnek, Marais 140; Petrusburg, Henrici 4292.

Natal. — Without precise locality: Tugela River, Medley Wood 3550; Muden

Valley, Repton 1187; Dundee: Indumeni Mtn., Truscott 157; near Dundee, Pegler

s.n. Klip River: Ladysmith, Medley Wood 7948 (L). Estcourt: Research Station,

West 528; 420; Pentz 483; Colenso, Schlechter 3369 (GRA); Colenso, Tugela

River, Medley Wood s.n. (GRA); bank of Tugela, Medley Wood 758 (SAM).

Not known to occur outside Southern Africa.

Burke on his expedition to the Transvaal with Zeyher, collected this species along

the Vet River in the Orange Free State and on the Magaliesberg near Pretoria. When

working on Burke’s plants Hooker recognised this as a “ congener with Menodora . . .

hitherto supposed to be exclusively an inhabitant of the New World ”. For the first

606

time then in 1843, botanists learned of this example of a connection between the flora

of S. Africa and the Americas. Hooker described the species as M. africana (see Hook

Ic. t. 586) and the plate accompanying his description is such a good one, giving the

details of the distinguishing features, that it is reproduced here (see plate 28). The

bi-pinnatisect leaf can be seen on this plate and the circumscissilely dehiscent fruit.

There is one example of the cernuous pedicel to the fruit and this illustrates the

monochasial development, for it can be seen that in flower the pedicel is erect and

terminal and that it then turns aside and a secondary shoot develops as the main axis,

the fruit therefore appearing lateral. The species is also illustrated and described in

detail in Flowering Plants of Africa (Vol. 30, PI. 1187).

M. africana has been found to occur in all four provinces of the Union and in the

Bechuanaland Protectorate. For the distribution as it is known today see the map,

Fig. 15. It is the most widely spread of the three South African species and is plentiful

in the regions of its distribution as may be judged by the long list of citations.

This species is closely related to the following species M. heterophyUa var. australis,

see the notes under that species.

3. M. heterophyUa Moric ex DC. var australis Steyermark in Ann. Missouri Bot.

Gard. 19: 127 (1932). Menodora heterophyUa Oliver in Hook. lc. PI. t. 1459

(1884); Wright in FI. Cap. 4, 1: 484 (1907).

Undershrub with several to many slender, more or less erect branches from a

woody base; branches about 7 to 25 cm. tall, slender, more or less angled and ridged,

sparsely scabrid. Leaves very variable in size and lobing, alternate, approximate or

sometimes opposite, sessile or petioled, irregularly and pinnately 3 to 5-lobed or simple,

4- 18 mm. long, 3-6 mm. broad, the segments acute, 1 -5-2-5 mm. broad, margins

scabrid, inrolled, surface fairly sparsely glandpitted. Inflorescence a reduced cyme,

monochasial, flowers terminal appearing lateral when overtopped by a secondary

branch which takes the place of the main axis. Calyx persistent but not accrescent,

5- 10 mm. long; tube about 2 mm. long; lobes 10-15, narrowly linear, entire or

occasionally lobed, unequal in length and up to 1 mm. broad, acute, scabrid on the

margins. Corolla yellow (sometimes with red infusion in parts?); tube about 3 mm.

long, infundibuliform; lobes about 1 cm. long and 4-5 mm. broad, oblong, slightly

narrowed at apex, or rounded with a mucro. Filaments inserted on the tubular portion

of the corolla tube and pilose there, free for about 3 mm. Anthers about 3 mm. long,

1-5 mm. broad, sub-basifixed, erect, minutely apiculate. Ovary 2-lobed, style about

6- 8 mm. long (appears to be red sometimes), stigma terminal, broadly capitate,

obscurely bi-lobed. Fruit on a cernuous pedicel, a bi-locular capsule, occasionally

only 1 coccus develops, dehiscing circumscissilely when ripe, coat parchment-like;

seeds 1-4 in each coccus, 6-9 mm. long, 4-6 mm. broad, outer skin reticulated.

Type: Peg/er 950 (BER), from near Rustenburg. [Type of the species, Berlandier

1499, between Laredo & Bejar, Mexico],

Bechuanaland Protectorate.— Lobatsi: Rogers 6225; 6 m. N.E. of Gaberones,

Codd 8937. Mochudi: Harbor 6529 (BOL); Rogers 6372 (BOL). Mahalapye:

Mansergh in Bol. Herb. 25416 (BOL).

Transvaal. — Marico: Zeerust, Thode A. 1442; 4 m. S. of Zeerust, Acocks

12415; Matebe Valley, Holub s.n. Rustenburg: near Rustenburg, Pegler 950 (on

several sheets of this number there is a portion of M. africana, see notes); near race

course, Galpin 9668; Nation 229 (BOL); Wonderfontein, Gray in Col. Herb. 4131;

Zwartruggens, Sutton 852; 1113. Klerksdorp: Phillips 53; Convent 28 (GRA);

Sister Lucy 11 (GRA).

Not known to occur outside above regions.

607

This variety was first collected by Dr. Em. Holub in the “ Matebe Valley ”, Marico,

Transvaal. From information obtained recently at the Mission Station at Linokana,

the “ Matabe ” is the stream passing through that village which is a few miles to the

west of Zeerust. When the well known Kew Botanist N. E. Brown came upon the

specimens in the collection presented to Kew by Dr. Holub in 1883 he identified them

with the New World species Menodora heterophyl/a. A note in N! E. Brown’s own

handwriting on a Holub duplicate in the Bolus Herbarium reads in part “ A most

interesting discovery as the plant has hitherto only been found in Texas. I have

minutely examined these Transvaal and the Texan specimens and can find no difference

at all between them.” Steyermark in his Revision, 1932, agrees that the S. African

specimens are not specifically distinct, but owing to certain “ morphological differences ”

he describes them as a variety of the New World species. The main difference is in

the size of the leaves, which in the Texan plants are up to 4 cm. long and 3 cm. broad

while in our plant they are seldom even half that size. Working with the African plants

only it is not possible to form an independent opinion on this point, therefore the

recent revisor of the genus is followed. In South Africa this variety is very close to

M. africana and differs principally in the leaves being lobed rather than bi-pinnatisect,

and the segments not being subulate. The calyx segments are usually simple and

only occasionally lobed, instead of being usually multipartite; the habit also is rather

more erect and rigid. If the plate from “ Hooker’s leones ”, which depicts Holub’s

specimen and is reproduced here (see PI. 29), is compared with that of M. africana

(see pi. 28), these differences will be very evident but, owing to the variation in size

and shape of the leaves of M. heterophyl/a var. australis, it is not always quite so easy

to distinguish between it and M. africana. Miss Pegler, for instance, whose specimen

No. 950, is the type of the South African variety, evidently did not see the difference

for her notes on the duplicate of the specimen in the National Herbarium, Pretoria,

reads: “observed from Pretoria to Woodstock,” Rustenburg district. It has also

been found that on several herbarium sheets of her gathering 950 (including one in

Pretoria and others in the Bolus and Albany herbaria), there is at least one piece of

M. africana among the, to us, distinct specimens of M. heterophyl/a var. australis. To

date M. africana has not been collected anywhere near Rustenburg but it is plentiful

in the neighbourhood of Pretoria. It is therefore not possible at this stage to account

for the mixture on Miss Pegler’s specimen, but one is inclined to assume that she

collected some of the material in the early stages of her trip when she observed the

plant “ from Pretoria to Woodstock ” and the bulk of the material at Woodstock.

It could only have been at intervals on the way that she could have seen plants, for

neither species has an extended or general distribution but is found in patches under

certain veld conditions.

When Oliver described Holub’s find in the “ leones ” he writes of the second South

African species of Menodora, but it was really the third for in 1869 the second, Harvey’s

species M. juncea, was published.

x

f

■

609

Index.

PAGE

Acanthaceae 602

Bolivariaceae 602

Bolivaria Cham. & Schltr 601

Calyptrospermum A. Dietr 601

Campanolea Gilg & Schellenb 592, 594

mildbraedii Gilg & Schellenb 594

Ceranthus, section Li node fa 592, 593

Dekindtia Gilg 592, 601

Dekindtia africana Gilg 594, 600, 601, 636

Gentianceae 602

Jasmineae 602

Jasminum Linn 550, 556

abyssinicum (Hochst, ex) DC. [PI. 6] 557, 563, 564, 616

angulare Vahl. [PI. 4] 556, 557, 559, 560, 561, 563, 566, 614

angulare var. glabratum E. Mey 549, 560, 561, 562

breviflorum Harv. [PI. 7] 549, 557, 564, 565, 566, 617

capense Thb 560, 561

flexile Jacq 559

fluminense Veil. [PI. 5] 557, 560, 561, 562, 563, 564, 615

gerrardi Harv 549, 564, 565, 566, 617

glaucum (L. f.) Ait. [PI. 8] 557, 566, 567, 569, 618

glaucum var. lanceolatum E. Mey 566

glaucum var. lati folium E. Mey 566

glaucum var. parvifolium E. Mey 567, 568

goetzianum Gilg 562

hirsutum Linn 556

humile Linn. Cult 557

ligustrifolium Lam 566

lupinifolium Gilg. & Schellenb 558, 559

mauritianum Boj 562, 563

meyeri-johannis Engl 569

multiflorum (Burm.) Andr. Cult 556

multipartitum Hochst, [PI. 9] 556, 557, 567, 569, 570, 619

natalense Gilg. & Schellenb 560, 561, 563

offlcenale Linn 556

pauciflorum Bth 572

pubescens Willd 556

quintum Schinz. [PI. 1] 557, 558, 559, 611

schroetarianum Schinz 562, 563

stenolobum Rolfe [PI. 10] 557, 569, 570, 620

stolzeanum Knobl 566

streptopus E. Mey 549, 558, 570, 572

streptopus E. Mey. subsp. streptopus [PI. 11] 558, 570, 571, 621

streptopus E. Mey. subsp. transvaalsense (Sp. Moore) Verdoorn [PI. 12]. .558, 570, 572, 622

swynnertonii Sp. Moore 572

tortuosum Willd. [PI. 2, 3] 557, 558, 559, 560, 612, 613

transvalense Sp. Moore 572

wyliei N.E. Br 563, 564

Linociera Sw 550, 573, 592, 593, 594, 601

africana Knobl 600, 601

africana (Welw.) Gilg & Schellenb 600

battiscomei Hutch. [Fig. 11, PI. 26] 593, 594, 597, 599, 600, 601, 636

congesta Bkr 601

foveolata (E. Mey.) Knobl 594, 595, 598

foveolata (E. Mey.) Knobl. subsp. foveolata [PI. 21, 22], .594, 595, 596, 597, 600, 631, 632

foveolata (E. Mey.) Knobl. subsp. majora Verdoorn [Fig. 13, PI. 24]. .594, 596, 597, 598,

599, 600, 634

foveolata (E. Mey.) Knobl. subsp. tomentella Verdoorn [PI. 23].. 590, 597, 598, 601, 633

latipetala M.R.F. Taylor 594

ligustrina Sw 593

marlothii Knobl 595, 596, 632

mildbraedii Gilg & Schellenb 594

peglerae (C.H. Wr.) Gilg & Schellenb [PI. 25] 594, 597, 599, 635

610

Menodora Humb. & Bompl 550, 601, 603

africana Hook. [PI. 28] 602, 604, 605, 606, 607, 638

helianthemoides Humb. & Bonpl 602

heterophylla Mor. ex DC 607

heterophylla Mor. var. australis Steyermark [PI. 29] 602, 605, 606, 607, 639

heterophylla Oliver non Moric 606

juncea Harv. [PI. 27] 549, 602, 603, 604, 605, 607, 637

Nathusia alata Hochst 550, 551, 553

Nyctanthus glauca Linn, f 566

Olea Linn 550, 572, 573

africana Mill. [Fig. 3, PI. 13] 573, 574, 576, 577, 578, 579, 580, 623

buxifolia Mill 582

capensis Linn 573, 581, 582, 584, 626

capensis Linn, subsp. capensis [Fig. 6, 7, 8. PI. 16, 17, 18]. .573, 581, 582, 584, 585, 586, 587, 592,

626, 627, 628

capensis Linn, subsp. enervis (Harv.) Verdoorn [Fig. 9, PI. 19] 573, 581, 588, 589, 629

capensis Linn, subsp. macrocarpa (C.H. Wr.) Verdoorn [Fig. 10, PI. 20], .573, 578, 581, 584, 590,

591, 592, 593, 600, 630

chrysophylla Lam 573, 574, 576

concolor E. Mey 582, 584

cuspidata Wall, ex G. Don 573

enervis Harv 549, 581, 588, 589

europaea Linn 573, 574, 576

europaea Thb. non Linn 573

europaea var. nubica Bkr 573, 574, 576

exasperata Jacq. [Fig. 5, PI. 15] 573, 579, 580, 584, 625

ferruginea Royle 573

foveolata E. Mey 593, 595, 597, 598

gallica Mill 576

guineensis Hutch. & C.A. Sm 581

hochstetteri Bkr 581

humilis Eckl 579, 580, 625

laurifolia Harv. ex Wright non Lam 588, 590

lanrifolia J. Phillips non Lam 590, 592

laurifolia Lam 581, 582, 584, 588, 590, 592, 600, 627

laurifolia Lam. var. concolor Harv 582, 584

laurifolia Sim. non Lam 590

listeriana Sim ex Lister 577, 578

mackenii Harv 577, 578

macrocarpa C.H. Wr 581, 590, 592

mildbraedii (Gilg & Schellenb.) Knobl 593, 594

monticola Gandoger 573

peglerae C.H. Wr 593, 599, 600

schimperi Gandoger 573

similis Burch 573, 574

somaliensis Bkr 573

undulata Jacq 582

undulata Jacq. var. planifolia E. Mey 582, 584

urophylla (Gilg) Gilg & Schellenb 581

verrucosa Link 573, 574, 576, 578, 623

verrucosa Link. var. brachybotrys DC 573

welwitschii (Knobl.) Gilg & Schellnb 581, 593

woodiana Knob. [Fig. 4, PI. 14] 573, 574, 577, 578, 579, 592, 624

Schrebera Roxb 550

alata (Hochst.) Welv. [Fig. 1] 551, 552, 553, 556

argyrotricha Gilg. [Fig. 2] 553, 555, 556

gilgiana Lingelsh 554, 556

greenwayi Turril 556

latialata Gilg 551, 552, 553

mazoensis Sp. Moore 553, 554

merkeri Lingelsh 553

nyassae Lingelsh 553

obliquifoliolata Gilg 553

saundersiae Harv 551, 552, 553

swietenioides Roxb 551

tomenteila (Welw.) Gilg 556

trichocalda Welw 551

611

Plate 1. — Jasminum quinatum Schinz; specimen from near the type locality in

Lydenburg; whole plant, showing rhizome.

129828-8

612

Plate 2. — Jasminum tortuosum Willd.; holotype in the Berlin-Darlem Herbarium.

613

Plate 3. — Jasminum tortuosum Willd.; on right, Drege s.n. in Ryksherbarium,

Leiden (may be same gathering as cited by DC. under this species); on left, Muir 2400

from Mossel Bay district (PRE).

614

E

615

Plate 5. — Jasminum fluminense Veil.; Burtt-Davy 360, Komatipoort (K) cited

n FI. Cap. under J. angulare but is not that species; note the very small calyx and

broad inflorescence borne clear of the leaves.

616

E* MfROAAIO NATAICN3I

Plate 6. — Jasminum abyssinicum Hochst. ex DC.; Wylie in Medley Wood

Herb. 8860, isotype, from Nkandhla, Natal.

617

Plate 7. — Jasminum breviflorum Harv.; on right, Burke s.n. Magaliesberg,

type (K) with leaves pubescent, not tapering to base and apex; on left top, two portions

of Gerrard 1477, type of J. gerrardi (K), Nonoti Riv., Natal, with leaves glabrous

tapering to base and apex; lower left hand, Rehman 7706, Natal, cited in FI. Cap.

under J. gerrardi, with glabrous leaves tapering to apex only.

618

Plate 8. — Jasminum glaucum (L.f.) Ait.; specimen from Clanwilliam characteristic

of species.

619

Plate 9. — Jasminum multipartitum Hoclist.; Krause 458, isotype (K), from

near Durban; on same sheet Burke s.n. from Uitenhage district.

620

Plate 10. — Jasminum stenolobum Rolfe; specimen from the Soutpansberg

district, Transvaal.

621

Plate 11. — Jasminum streptopus E. Mey. var streptopus; Drege s.n., holotype,

Geneva, collected at “ Port Natal

622

Plate 12. — Jasminum streptopus E. Mey. var. transvaalense (Sp. Moore) Verdoorn

a characteristic specimen from the Pilgrimsrest district, Transvaal.

623

Plate 13. — Olea africana Mill.; after Sim, Forest Flora of C.C., plate 105, under

the name O. verrucosa.

624

Plate 14. — Olea woodiana Knobl. ; after Sim, Forest Flora of C.C., plate 108.

625

Plate 15. — Olea exasperata Jacq.; after Sim, Forest Flora of C.C. plate

under the name O. humilis.

626

627

Plate 17. — Olea capensis L. subsp. capensis, photo of type specimen of O. laurifo/ia

Lam. in the Paris Museum.

129828-9

628

Nieuwekloof and Elandskloof, Tulbagh district, isotype of 40. concolor E. Mey. in

Ryksherbariun Leiden.

629

Plate 19. — Olea capensis L. subsp. enervis (Harv.) Verdoorn; two specimens of

Gerrard 1151, type (K).

630

Plate 20. — Olea capensis L. subsp. macrocarpa (C.H. Wr.) Verdoorn; Grenfell

869, isotype (BOL).

631

JypjtslJ

; io

Plate 21. — Lionciera foveolata (E. Mey.) Knobl. subsp. foveolata; Drege s.n.,

Uitenhage district, isotype of L. foveolata E. Mey.; branchlets glabrous and leaves

comparatively narrow, (L).

129828-10

632

Plate 22. — Linociera foveolata (E. Mey.) Knobl. subsp. toveolata; Rudatis

1416, Dumisa, Natal, isotype of L. marlothii Knobl.; leaves somewhat thinner in

texture and rather longer than in more typical specimens.

Flora Africa* «y*t.ra»8

M

Alox«ndrft

.... ^

633

L&ieUro fovaox^va

toot;, sum*. ttMUUi Varaosrn.

Determinavit 1<c. v.rteon.

Plate 23. — Linociera foveolata (E. Mey.) Knobl. subsp. tomentella Verdoorn;

a characteristic specimen from Knysna; branch lets tomentulous, leaves broadly

elliptic ending in a retuse acumen.

129828-11

634

Plate 24. — Linociera foveolata (E. Mey.) Knobl. subsp. major Verdoorn, a

specimen from the type locality, Marieps Kop; note fruit large, though immature,

ridged at apex and leaves rather broad.

635

Plate 25. — Linociera peglerae (C.H. Wr.) Gilg & Schellenb.; Pegler 819, from

Kentani, isotype (PRE); leaves large with comparatively long petiole and blunt acumen

at apex.

636

Plate 26. — Linociera battiscombei Hutch (= Dekindtia africana Gilg.) ; specimen

from Blaauwberg, Northern Transvaal; the inflorescence glomerate in the axils of

the leaves.

637

Plate 27. — Menodora juncea Harv. ; specimen from Kamieskroon, Namaqualand.

638

Plate 28. — Menodora africana Hook.; from Hookers leones Plantarum, tab,

586, figure of type.

639

Plate 29. — Menodora heterophylla Mor. ex DC. var. australis Steyermark;

from Hooker’s leones Plantarum, tab. 1459.

The South African Convolvulaceae.

By

A. D. J. Meeuse.

Recent work on the Convolvulaceae of the Transvaal made it evident that the

account in the Flora Capensis requires revision. Several name changes have to be

made and many new records incorporated.

The first important contribution to our knowledge of the South African Convol-

vulaceae was made by Thunberg who described or mentioned about 12 species. Choisy

in DC., Prodr. 9 (1845) added many species collected by Burchell, Ecklon, Zeyher

and Drege. H. Hallier (fil.) revised the family, beginning with his paper on the “ natural

system” of the Convolvulaceae in Engl. Bot. Jb. 16: 453-591 (1893), and treated

the African Convolvulaceae in a series of papers, the most important of which are

in Engl. Bot. Jb. 18: 81-160 (1893), 28: 28-54 (1899) and in Bull. Herb. Boiss. 6: 529-

548 (1898), 7: 41-55 (1899). Hallier’s system has been generally followed with only

minor changes. His works, together with publications by Engler in Engl. Bot. Jb.

10:53-56 (1888), by Schinz in Verh. bot. Ver. Brandenb. 30: 270-276 (1888), by Baker

in Kew Bull. 1894: 67-74 (1894) and by Rendle in Jl. Bot. 39: 12-22, 55-64 (1901),

40: 189-191 (1902), were the main sources of reference for the treatments of the

Convolvulaceae in the Flora Capensis 4, 2 (1904) by Baker and Wright and in the

Flora of Tropical Africa 4, 2 (1905-1906) by Baker and Rendle. Since these publications

only few papers treating species from Southern Africa have been added. Pilger

continued Hallier’s papers on Convolvulaceae Africanae in Engl. Bot. Jb. 41 : 293-

297 (1908), 45: 218-222 (1910) and 48: 348-352 (1912); N. E. Brown published on

the collections made by the Lugards in Ngamiland (Kew Bull. 1909, Convolvulaceae

p. 122-125). These publications are important for the area under discussion because

they deal with species some of which were described or have since been recorded from

the Union. A monograph of the genus Cuscuta by Yuncker appeared in 1936.

The genera were treated by E. P. Phillips in his Genera of South African Flowering

Plants (Ed. 2, 1951). The deliminations of a few genera differ slightly from those

given by Phillips and several genera not mentioned by Phillips prove to occur in the

Union. The delimitations given here are in agreement with those in the Flora of West

Tropical Africa and in Flora Malesiana, Ser. 1, Vol. 4, part 4 (1953). Recently G.

Roberty, in Candollea 14: 11-60 (1952-1953), proposed a new system for the Convol-

vulaceae. He described eight new genera, reinstated some eight others and at the same

time reduced to synonymy some genera of long standing. The few new genera proposed

since Hallier’s system were mostly omitted, and among other things several violations

of the International Code were committed; for instance, the type species of the genus

Evolvulus ( E . nummularius F.) is the basis of a new genus, Volvulopsis Roberty. There

seems no improvement on Hallier’s system, which has stood essentially unchanged and

unchallenged for sixty years, and Roberty’s treatment is not supported here.

Material of the following herbaria was studied, apart from the collection of the

National Herbarium (PRE): BOL, COI, GRA, J, JE, KMG, L, LM ( = Herbarium

of the Repartigao Tecnica de Agricultura, Lourengo Marques), NBG, NH, NU, SAM,

US (US only Cuscuta). The abbreviations are those of the Index Herbariorum. Thanks

are due to the Directors of these herbaria for kindly sending their material on loan

to the National Herbarium.

642

I am much indebted to Mr. B. de Winter and Mr. D. J. B. Killick, South African

botanists at Kew, who assisted greatly in comparing types, looking up literature not

available lochlly and supplying other information. I also wish to express my gratitude

to Dr. S. J. van Ooststroom (Leiden), Dr. T. G. Yuncker (Greencastle, Ind.) and

Dr. B. Verdcourt (Nairobi) for helpful advice.

\c-

Specimens are not quoted unless there are only a few records of a species or if

they are interesting for some other, reason ; specimens from outside the area are cited

only in exceptional cases.

FAMILY CHARACTERS.

Flowers hermaphrodite (very rarely unisexual), minute to large, often showy,

regular, rarely slightly zygomorphic, generally axillary, solitary or cymose, often

aggregated into heads, rarely paniculate or approximated in a terminal spike, pedicelled

or sessile, but the cymes usually peduncled ; as a rule solitary flowers or cymes with

2 bracteoles (or bracts), these bracteoles either small or occasionally large, often

deciduous but sometimes persistent and accrescent. Calyx inferior, sepals 5, rarely 4

or 3, free or rarely united at the base, much imbricate, equal or more or less unequal,

persistent, often, especially the outer, accrescent. Corolla gamopetalous, usually

funnel-shaped, but sometimes hypocrateriform, campanulate, nearly tubular, urceolate

or rotate; the limb subentire or 5-lobed, more rarely 5- (or 4- or 3-) partite, in bud

generally induplicate-valvate, often contorted, glabrous inside, frequently more or less

distinctly marked on the outside by more or less hairy and veined midpetaline zones

(which are outside in the bud), separated by glabrous episepaline areas. Stamens 5,

rarely 4 or 3, inserted low down in the corolla-tube opposite the sepals or sometimes in

the throat (Cuscuta)', filaments equal or unequal, filiform, generally dilated and/or

glandular-hairy at the base, rarely stipulate, rarely attached to scales; anthers introrse,

ovate or oblong to nearly linear, often more or less sagittate at the base, dorsifixed,

dehiscing longitudinally; pollen either spherical and spinose all over ( Echinoconiae

Hall.f.) or more or less spherical and smooth or ellipsoid and marked with folds

( Psiloconiae Hall.f.). Hypogynous disc annular or cup-shaped, often shallowly 5-lobed,

. sometimes obsolete. Ovary superior, built up by 2 (rarely 3) united carpels, 2 (or 3)-

loculated, sometimes only 1 -chambered or by development of spurious septa 4-celled,

usually entire but occasionally 2- or 4-partite; ovules 2 in each carpel, rarely solitary,

erect, anatropous, sessile; style terminal or rarely gynobasic, filiform, simple or

more or less deeply 2-fid, or styles 2, very rarely 3, equal or unequal, rarely reduced;

stigma terminal and entire or bilobed, rarely 3-lobed, or stigmas 2 (rarely 3 or 4) globose,

ellipsoid, filiform, or flattened, elliptic or linear, rarely peltate. Fruit 1-3-, or spuriously

4-celled, mostly a capsule dehiscing by valves or more rarely transversely or irregularly

dehiscent, sometimes indehiscent with membranous, leathery, woody or fleshy pericarp.

Seeds as many as there are ovules or less by abortion, becoming triquetrous on the

inner face by mutual pressure, convex on the back, glabrous or hairy (especially on

the angles), sometimes verrucose; embryo straight; radicle directed towards the

hilum; cotyledons foliaceous, generally folded or spirally coiled, often with emarginate

or bilobed apex, sometimes small or 0 (in Cuscuta)', endosperm scanty to fairly copious,

between the folds of the cotyledons. Annual or perennial herbs or woody plants

without tendrils, unarmed or rarely spinescent, often climbing or prostrate, more

rarely erect and shrubby, very rarely trees. Roots in herbaceous perennials often

developed as thick taproots or as large fusiform to globose tubers. Leaves alternate,

simple, often cordate, hastate or sagittate at the base, entire, lobed or deeply pinnately,

palmately or pedately dissected, exstipulate but occasionally pseudostipulate by the

leaves of developing or suppressed axillary shoots, usually petiolate, rarely absent or

reduced to minute scales (Cuscuta).

Note.—' The structure of the fruit is an important distinguishing character between

genera and to overcome this difficulty the species belonging to the genera Stictocardia

and Turbina are included in the generic key of Ipomoea.

Another character of importance is the structure of the surface of the pollen, i.e.,

whether, it is smooth or spinulose, but this is only used in the key where it cannot be

avoided. .

Key to the Genera.

Leaves reduced to minute scales or absent. Flowers small, usually in clusters,

often numerous; herbaceous twining parasites 1. Cuscuta

Leaves well-developed, green:

Ovary 2- or 4-lobed and fruit split into 2 or 4 lobes; styles 2 (sometimes

connate below), inserted between the lobes of the ovary; small

prostrate herbs with oblong or cordate to reniform leaves:

Ovary 2-cleft with 2 ovules in each chamber; fruit 2-lobed 2. Dichondra.

Ovary 4-cleft with 1 ovule in each chamber; fruit 4-lobed 3. Falkia.

Ovary not deeply lobed; fruit not split into 2 or 4 lobes; style simple or,

if styles 2, terminal; plants of various habit:

Styles 2, free or more or less united at the base:

Styles forked; stigmas 4, linear or subclavate 4. Evolvulus.

Styles not forked; stigmas 2, peltate or capitate:

Sepals (in S. African species) broadly ovate to lanceolate,

more or less acute, hairy 5. Seddera.

Sepals (in only S. African species) broadly ovate-spathulate to

suborbicular, glabrous with membranous and some-

times (partly) ciliolate margins 6. Bonamia.

Style 1; stigmas globose to linear:

Pollen smooth or with folds or thickened ridges, but not spinose:

Stigmas filiform, terete or subclavate:

Bracteoles small or narrow, usually more or less remote

from the calyx

Bracteoles large, more or less enclosing the calyx as an

involucre

Stigmas globose, ovate, oblong or elliptic:

Ovary 1 -celled; stigmas ovate, complanate above; sepals

unequal, outer ones broadly ovate, 2 inner ones

much narrower, lanceolate; corolla (in S. African

plants) white or cream with purple “eye”

7. Convolvulus.

8. Calystegia.

9. Hewittia.

Ovary 2(-4)-celled :

Flowers (in S. African species) in pedunculate,

subglobose heads, usually blue; leaves

cordate at the base, entire or slightly lobed. . 10. Jacquemontia.

Flowers (at least in S. African species) not in heads,

but in cymes or dichasia, or solitary, white or

yellow, often with dark centre; leaves usually

palmately or pinnately lobed or dissected,

sometimes auricled at the base, rarely entire:

Fruit a 4-valved capsule or shedding the 4-valved

pericarp as a whole by basal circumscission 1 1 . Merremia.

Fruit, when ripe, shedding the outer layer of the

pericarp (by basal circumscission) as an

“ operculum ”, leaving the membranous

inner layer of the pericarp which later

more or less irregularly splits into strips

(Doubtful record, included in key to the

Species of Merremia) 12. Operculina.

644

Pollen spinose (Echinoconiae) :

Plants erect, suffruticose or shrubby, sometimes trailing, with

rather large leaves, covered in all vegetative parts and

on peduncles, pedicels, bracts and calyces with stellate

hairs (often tomentose); stigmas oblong, thick, fruit

dehiscent, 4-valved

Plants of various habit, but, if hairy, not with stellate hairs,

stigma biglobose or 3-Iobed:

Fruit* a dehiscent capsule; pericarp almost invariably

completely separating into 4 valves; leaves without

minute black glands on lower surface

Fruit* completely enclosed by the much enlarged calyx,

globose, with transversely winged and somewhat

thickened persistent dissipiments and a thin wall

between these winged portions which ultimately

becomes more or less irregularly detached from the

wings, exposing the seeds and thus the fruit

becoming lantern-shaped with 4 openings through

which the seeds are visible

Fruit* indehiscent, often 1 -seeded, with erect or more or

or less spreading sepals which usually do not

enclose the capsule; pericarp woody or leathery

13. Astripomoea.

14. Ipomoea.

15. Stictocardia.

16. Turbina.

1. CUSCUTA

L. [Sp. PI. Ed. 1 (1753), p. 124]; Gen. PI. Ed. 5 (1754), p. 60; T. G. Yuncker, “The

Genus Cuscuta'", Mem. Torrey Bot. Cl. 18, no. 2 (1932); Phillips, Gen. S. Afr.

Flow. PI. Ed. 2 (1951) p. 620; Verdcourt in E. Afr. Agric. Jl. 18, No. 2 (1952),

p. 85-86.

Parasitic, usually glabrous herbs, without chlorophyl, annual or rarely perennial

in the tissues of the host. Stems usually terete and slender to filiform, often whitish,

yellowish or reddish, twining or rambling, attached to the host by means of numerous

haustoria. Leaves reduced to minute scales or none. Flowers small, in cymose clusters,

sometimes paniculate, generally 5-merous but sometimes 4- or 3-merous. Calyx

5-lobed, 5-parted or sepals free; calyx-lobes or sepals broad or narrow, rounded or

obtuse to very acute. Corolla with a tubular, urceolate, campanulate or semi-globose

tube; the lobes shorter or longer than the tube, often patent to reflexed; the tube

inside usually with crenulate or fimbriate episepalous membranous scales, rarely scales

reduced to crenulate rims or absent. Stamens inserted on the corolla above the scales;

filaments often short; anthers often broadly elliptic; pollen smooth. Ovary 2-celled,

4-ovuled; styles 2, distinct, or connate and style 1; stigmas capitate or elongated.

Fruit an ovoid or subglobose capsule, opening irregularly, or circumscissile near the

base, or indehiscent. Seeds 4 or less, almost invariably glabrous ; embryo acotyledonous,

straight, filiform, enlarged at one end.

Type Species: Cuscuta europaea L. Cosmopolitan, the majority of the species

American.

Yuncker recognises 158 species in his monograph, but in the opinion of the present

writer and also of Dr. B. Verdcourt, Yuncker’s conception of specific differences is

rather narrow, and we would recognise about 140.

Several species attack economically important crops such as clover, lucerne

(= alfalfa), flax, etc., and have become or threaten to become cosmopolitan weeds,

e.g., C. campestris, C. suaveolens, C. epithymum, all three now occurring in S. Africa.

Although the flowers in Convolvulaceae are generally 5-merous in calyx, corolla

and androeceum (except in Hildebrandtia), those of Cuscuta are not infrequently 4- or

occasionally 3-merous, some species normally having 4-merous flowers.

* If in doubt, or if fruits are lacking, consult the key to the species of Ipomoea.

645

Some authors recognise a separate family Cuscutaceae for this genus, but apart

from the parasitic habit there is not one character to be found in the structure of the

flowers, fruits or seeds which would justify this separation.

Styles united almost to the apex

8. C. cassytoides.

Styles 2, free to the base or nearly so:

Stigmas capitate-globose or more or less peltate:

Stigmas (in dried specimens) more or less peltate with convolute edges;

styles shorter than the ovary; intrastylar aperture of capsule

large; flowers 4-5 mm. long and broad; calyx- and corolla-lobes

broad, obtuse or rounded; N. Transvaal, Port. E. Africa, in

forests on shrubs 6. C. kilimanjari ,

Stigmas usually globose, more rarely somewhat flattened:

Fringed scales in corolla-tube below the stamens present (scaies

very rarely subentire):

Scales in corolla-tube bifid at the apex:

Flowers normally 5-merous, subsessile, in globose clusters;

corolla-lobes ovate to suborbicular, usually shorter

than the tube; scales ± bilobed at the apex,

usually not reaching the base of the free portion of

the filament; Transvaal and throughout tropical

Africa 1. C. australis.

Flowers usually 4-merous, pedicellate, in ± umbellate

cymes; corolla-lobes narrow, usually distinctly

longer than the tube; scales bifurcate, the terminal

lobes exerted beyond the sinuses of the corolla-

tube, deeply fringed; Cape Province (endemic)... 2. C. bifurcata.

Scales in corolla-tube free, not bifid at the apex:

Scales mostly reaching the base of the free part of the

filaments, or, if not reaching the filaments, lobes

of calyx and corolla very obtuse:

Flowers in usually dense clusters, usually sessile or

nearly so, wider than long or about as wide as

long, ± rounded at the base; calyx nearly

enclosing the corolla-tube, not very loosely

about the base and without wartlike or saccate

basal processes; styles shorter than or nearly

equalling the ovary:

Scales not reaching the base of the filaments,

oblong, variously fimbriated to almost

entire; corolla-lobes broad, obtuse or

rounded; ovate to suborbicular, erect;

Transvaal and wide-spread in tropical

and N.E. subtropical Africa, mostly on

marsh plants

Scales usually reaching the base of the filaments,

ovate, abundantly fringed with fairly long

processes, corolla-lobes broadly triangular,

acute with often indexed tips, rarely

obtuse, usually spreading; very common

and wide-spread, often as a pest on

lucerne and other cultivated plants

Flowers in loose paniculate cymose clusters, pedicel-

late, slightly longer than wide; pedicels not

infrequently longer than the flowers, often

papillose and rather rough; calyx shorter

than the corolla-tube; (in older flowers at

least), usually very loose about the base of the

corolla to almost spreading, with wart-like or

saccate processes at the base; styles usually

longer than the ovary

1. C. australis.

3. C. campestris.

5. C. appendiculata.

646

Scales usually not reaching the base of the free part of

< . the filaments ; ■ lobes of calyx and corolla (sub-)

acute; calyx narrowed at the base, without warts

or processes at the base, corolla-tube campanulate,

longer than wide ,l; . . M . . .

Scales in corolla-tube adnate as far as the apex and consisting

of twe fringed ridges attached on either side of the

adnate portion of the filament, or rarely in some of

the flowers free at the apex (this variety as yet not

recorded frem the Union, but may be looked for)..

No fringed scales in corolla-tube below the stamens

Stigmas conical to clavate or filiform, usually much longer than wide:

Calyx as long as the corolla-tube, protruding at the sinuses between the

calyx-lobes to form prominent wings, 5-angled

Calyx not 5-angled by protruding wings:

Flowers 5-merous, or if 4-merous, corolla-lobes obtuse or style not

very short:

Stigmas and often also the styles shorter than the ovary (if

stigmas about as long as the ovary, styles longer than

the stigmas); stigma usually considerably shorter than

the style:

Corolla-lobes with cucullate apices; flowers ± 2 mm.

long; stigmas conical, sausage-shaped to sub-

clavate, much shorter than the styles..

Corolla-lobes often with indexed or reflexed tips, but

never cucullate; flowers more than 2 mm. long;

stigmas various:

Lobes of calyx and corolla acute; flowers pedicellate;

styles as long as or longer than the ovary:

Calyx usually distinctly shorter than the corolla-

tube ; if calyx about as long as the corolla-

tube, stigmas considerably shorter than-

the styles:

Flowers usually 3-4 mm. long; corolla-

lobes acute or obtuse; scales about

reaching the base of the filaments;

styles distinctly longer than the stigmas

Flowers 4-7 mm. long; corolla-lobes very

acute, often with reflexed tips; scales

not reaching the base of the filaments;

styles about as long as the stigmas. . .

Calyx about as long as the corolla-tube. Stigma

about as long as the style see

Lobes of calyx and corolla obtuse; flowers sessile,

in few-flowered dense clusters; styles shorter

than the ovary

Both stigmas and styles longer than the ovary. Stigma

filiform, as long as or longer than the style:

Calyx considerably shorter than the corolla-tube; stems

medium; flowers pale, whitish, pedicellate, 4-7

mm. long

Calyx not considerably shorter than the corolla-tube (in

some forms of C. epithymum calyx distinctly shorter

than the corolla-tube, but if so, stems very slender,

capillary, and flowers ± 3 mm. long, sessile in dense

globose clusters, often reddish):

Flowers somewhat fleshy, 3-5 mm. long, in loose

cymose clusters; scales usually reaching the

bases of the filaments; stems medium; S.W.

Cape districts, on wild plants

4. C. suaveolens.

‘ . : -■ i-1 :. ' i.

7. C. hyalina var.

nubiana.

7. C. hyalina.

13. C. angulata.

9. C. gerrardii.

10. C. africana.

11. C. natalensis.

12. C. nitida.

14. C. planiflora var.

madagascarensis.

11. C. natalensis.

12. C. nitida.

647

Flowers more herbaceous, ± 3 mm. long, sessile, in

dense, compact, globose clusters; scales

usually not reaching the base of the filaments;

stems very slender, capillary; wide-spread, on

wild and cultivated plants 15. C. epithymum.

Flowers 4-merous in small, dense, sessile globose clusters. Corolla-

lobes erect, acute; stigmas longer than the very short styles;

scales in corolla-tube entire of bifid, the fringes distinctly

thickened at their apices 14. C. planiflora var.

mossamedensis.

1. C. australis R. Br., Prodr. FI. Nov. Holl. Ed. 1 (1810) p. 491; Yuncker, op. cit.,

p. 124, fig. 1: Ooststr. in Blumea 3 (1938), p. 66; Verdcourt in E. Afr. Agric.

Jl. 18, no. 2 (1952), p. 85-86; Ooststr. in Steenis, FI. Males Ser. 1. 4.4 (1953),

p. 392. C. cordofana (Engelm.) Yuncker, op. cit., p. 127, fig. 2. ?C. obtusiflora

H.B. et K., Nov. Gen. Spec. PI. 3 (1818), p. 122 (= p. 96 of folio edition).

Type; Caley in herb. R. Brown from “ New Holland ” (K). Isotype in the

herbarium of the Missouri Botanical Garden (fide Yuncker, who erroneously called

this specimen the type).

Flowers often somewhat glandular, 2-3 mm. long, subsessile in dense subglobose

clusters. Calyx about as long as the corolla-tube, lobes broad, ovate to suborbicular,

rounded at the apex. Corolla campanulate; its lobes shorter than or as long as the

tube, erect to somewhat spreading, broad, rounded at the apex. Stamens shorter than

the corolla-lobes; filaments usually stout, more or less subulate and often shorter

than (sometimes about as long as) the broadly elliptic to suborbicular anthers. Scales

oblong, shorter than the tube, bifid or bilobed to entire, variously fimbriated. Ovary

globose; styles shorter than the ovary, intrastylar aperture large. Capsule globose

or somewhat obovoid, not dehiscent by circumscission. Seeds ellipsoid to ovoid,

1-5-2 mm. long.

Yuncker mentions under C. australis: “ Turkestan to Japan and to Australia ”

and under C. cordofana: “ throughout central Africa to Madagascar ”. If, what is

not unlikely, C. obtusiflora H.B. & K. is also conspecific or at most worthy only of

varietal rank, the range of C. australis would include South America. It extends into

the Transvaal and Portuguese East Africa.

Transvaal. — Potgietersrust, Naboomspruit: Galpin 11622 (PRE, BOL).

Hosts. — Yuncker, Van Ooststroom and Verdcourt mention Hydrocotyle, Poly-

gonum and several other plants. Verdcourt reports: “ often found on marsh plants

The specimens cited above were collected “ in swampy places ” on Polygonum tomen-

tosum and Jussiaea fluitans.

The reduction of C. cordofana to C. australis is done here in agreement with

Verdcourt’s above-cited preliminary treatment of the genus Cuscuta which is a summary

of his unpublished account for the Flora of Tropical East Africa.

The flowers are normally 5-merous, but 4-merous ones are not rare.

A specimen leg. Bradfield from Benoni, Transvaal (PRE) which I tentatively refer

to this species has only 3- and 4-merous flowers. Abnormality is indicated by the

fact that a trimerous flower had 3 anthers on one filament making 5 anthers in all

■suggesting the normal 5-merous flowers. This specimen was studied by Dr. Yuncker

and by Dr. Verdcourt and both agree upon its close relation to C. australis. Dr.

Yuncker thought it might represent an undescribed species. Dr. Verdcourt suggested

that the specimen is an abnormal form of C. australis. As it is sometimes difficult,

to distinguish between C. australis and the cosmopolitan weed C. campestris, the

648

abnormal specimen might even belong to the latter. The plant was collected on a

Pelargonium, probably in a garden, and this would point to C. campestris rather than

C. australis.

2. C. bifurcata Yuncker, op. cit., p. 131, fig. 7.

Type: Paterson 578, from Port Elizabeth (K).

Flowers 2-2-5 mm. long, glandular, usually 4-merous, in few-flowered umbellate-

cymes on short pedicels. Calyx about equalling the corolla-tube, its lobes oblong to

lanceolate, obtuse or sub-obtuse with rounded sinuses between them. Corolla divided

halfway down or slightly more; the tube broadly campanulate; the lobes acute or

subacute to obtuse, erect to spreading, lanceolate-oblong. Stamens shorter than the

corolla-lobes or subequalling them; anther shorter than the filaments. Scales bifurcate

at the apex, adnate to the sinus between the apical lobes, the latter exserted between

the corolla-lobes. Ovary globose, styles slender, shorter than or nearly equalling the

ovary. Fruit depressed-globose, not dehiscent by circumscission. Seeds about 2 mm.

long, subglobose, flattened on two sides, slightly attenuate on one end; hilum short,

oblong, oblique.

Endemic.

Cape Province. — Ceres, Kouebokkeveld: Schlechter 10110 (BOL, GRA, PRE).

Port Elizabeth: Paterson 578 (GRA, PRE, isotypes).

Hosts. — The type collected on Falkia repens (Convolvulaceae) ; Schlechter’s

specimens on Ursinia (Compositae).

3. C. campestris Yuncker, op. cit., p. 138, fig. 14; Ooststr. in Blumea 3 (1938), p. 68;

Yerdcourt in E. Afr. Agric. Jl. 18, no. 2 (1952), p. 85-86.

Type: From Texas (fide Yuncker).

Flowers 2-3 mm. long, occasionally longer, often glandular, subsessile or on

short pedicels in compact, globose clusters. Calyx enclosing the corolla-tube or nearly

so, broadly campanulate; its lobes ovate to orbicular or broadly triangular, usually

obtuse to rounded. Corolla-tube campanulate; the lobes broadly triangular to ovate,

spreading, with often inflexed tips, usually acute. Stamens shorter than the corolla-

lobes; filaments longer than or equalling the anthers. Scales ovate, elliptic or obovate,

often exserted between the corolla-lobes, free at the apex, usually abundantly fringed.

Ovary globose; styles often somewhat unequal, slender, becoming thicker and

conspicuous in fruit. Capsule depressed-globose with a depression around the style-

bases, 2-2-5 mm. high and 3-4 mm. in diam., not circumscissile. Seeds about 1 -5 mm.

long, usually flattened on one side; hilum terminal, oblong, transverse.

A native of America and introduced into Europe, S. and E. Africa, E. Asia,

Australia and Polynesia. Recorded widely in the Union of South Africa but as yet

not from South West Africa, on a wide range of hosts.

The oldest records in the Union are from 1907 so that it is likely that this plant

was introduced in 1907 or shortly before that date.

4. C. suaveolens Ser. in Ann. Sci. Phys. Nat. Agric. Indust. 3 (1840), p. 519; Yuncker,

op. cit., p. 148, fig. 22. C. medicaginis C. H. Wright in Dyer, FI. Cap. 4,2: 86-

(1904).

Type: A specimen from Lyon, France ( fide Yuncker).

649

Flowers in racemose clusters on short pedicels, more or less glandular, membranous

when dry, 3-4 mm. long. Calyx shorter than the corolla-tube, lobed about halfway

down; its lobes ovate-triangular, more or less acute often with revolute edges, separated

by usually rounded sinuses. Corolla long-campanulate to funnel-shaped; the lobes

ovate-triangular, with acute, indexed tips, about \ to f as long as the tube. Stamens

shorter than the corolla-tube; filaments about as long as the anthers. Scales usually

not reaching the stamens, oblong to ovate or triangular-ovate, free at the apex, fringed

with rather short processes. Ovary globose; styles slender, often distinctly unequal,

about as long as the ovary. Capsule globose, not circumscissile. Seeds 4 to 2, 1 • 5-2

mm. long, subglobose; hilum oblong, longitudinal.

Originally a native of South America but, as a contaminant of Medicago sativa,

wide-spread and now almost cosmopolitan.

Cape Province. — Somerset East, Pearston: Div. Council 16982 (PRE). Grahams-

town: Blomfield s.n. (GRA). Queenstown: Galpin 1760 (PRE, GRA, isotypes of

C. medicaginis C. H. Wright), Galpin 7781 (PRE, GRA). “ Fish River”: White 485

<GRA).

Transvaal. — Johannesburg: Rattray 830 (PRE). Belfast, Machadodorp: N.N.

in Govt. Herb. no. 6302 (PRE).

Hosts. — On a variety of plants but in S. Africa almost exclusively on lucerne.

5. C. appendiculata Engelm. in Trans. Acad. Sci. St. Louis 1 (1859), p. 503; Baker &

Wright in Dyer, FI. Cap. 4, 2 (1904), p. 86; Yuncker, op. cit., p. 152, fig. 26,

inch var. macroflora Yuncker; Salter in Adams. & Salt., FI. Cape Penins. (1950),

p. 687.

Type: Krauss 1816 from Swellendam, C.P. (Not seen.)

Flowers in loose, paniculate, cymose clusters, 1 • 5-4 mm. long, more or less

glandular particularly the calyx. Pedicels shorter to longer than the flowers, papillose

or verrucose. Bracteoles often papillose-verrucose and saccate at the base by wart-like

projections. Calyx shorter than the corolla-tube; its tube verrucose towards the base

and with wart-like basal processes; the lobes triangular, acute. Corolla campanulate;

the erect to spreading lobes nearly as long as the tube, oblong- or ovate-lanceolate,

acute to acuminate, with inflexed tips. Stamens shorter than the corolla-lobes ; filaments

about as long as the anthers. Scales oblong-ovate or obovate, much fimbriate, about

equalling the corolla-tube. Ovary globose; styles slender, equalling the ovary or

slightly longer, often somewhat unequal. Capsule ovoid or globose, somewhat

contracted and roughened around the style-bases, not circumscissile. Seeds subglobose,

somewhat flattened, about 1-5 mm. long; hilum oblong, longitudinal.

Endemic.

Cape Province. — Cape Flats: Andreae 397 (PRE). Near Cape Town: Moss 9305

(J). Fraserburg: Pons s.n. (PRE). Oudtshoorn: Schoeman s.n. (PRE). Riversdale:

Burgers s.n. (PRE); Muir in BOL H. no. 25024. Mossel Bay, Bankfontein: Muir

1281 (PRE). Uniondale: Fourcade 2087 (BOL). Humansdorp: Fourcade 5171

(BOL). Zitzikamma: Schlechter 5973 (BOL, GRA, PRE). Somerset East : MacOwan

in BOL H. no. 25023, 1958 (SAM). Bedford: Bennie 250 (GRA, type of var. macroflora

Yunck). Victoria East: Rattray 17 (GRA). King William’s Town, Green River:

Flanagan 1707 (BOL, GRA, PRE). Transkei: Barber 32 (GRA, PRE). Kimberley:

Hutton in herb. McOwan & Bolus 923 (BOL, GRA, SAM); McOwan 2915 (NH, PRE);

Flanagan 1421 (BOL, PRE); Moran 496a (KMG).

Transvaal. — Middelburg, Klein Olifants River: Schlechter 3811 (BOL, GRA,

NH, PRE).

650

Hosts. — Lycium , Nicotiana glauca, Hermannia, Combretum, Pelargonium, Falkia,

Compositae, Campanulaceae, Sporobolus pungens, Thesium sp.

Yuncker distinguishes a var. macroflora Yuncker with larger flowers than the

typical form, but admits that “ some of the specimens are intermediate in their characters

connecting this variety with typical C. appendiculata

6. C. kilimanjari Oliv. in Johnston, Kilimanjaro Exped., Append., p. 343 (1886), nomen

tantum, and in Trans. Linn. Soc., 2nd Ser., Bot. (1887), p. 343, descr.; Baker,

& Rendle in Dyer, FI. Trop. Afr. 4, 2 (1906), p. 205; Yuncker, op. cit., p. 187, fig.

58; Verdcourt in E. Afr. Agri. J. 18, no. 2 (1952), p. 85-86; Brenan in Mem.

New York Bot. Garden 9 (1954), p. 9.

Type specimen: Johnston 86 (K).

Stems of medium thickness to rather stout. Flowers in few-flowered cymes on

pedicels shorter than the flowers, 4-5 mm. long and in diam., somewhat coriaceous

when dried. Calyx cupulate, its lobes ovate-orbicular, obtuse, overlapping at the

base, rather thick and often more or less carinate. Corolla campanulate-cylindric,

the tube longer than the ovate-orbicular, obtuse to rounded and often somewhat

revolute lobes. Stamens shorter than the corolla-lobes, filaments as long as or Slightly

longer than the anthers. Scales triangular or oblong, the apex often truncate, irregularly

and unevenly fringed to almost entire along the edge, usually reaching the bases of

the stamens, but sometimes smaller or reduced. Ovary globose; stigmas shorter than

the ovary; stigmas often flattened with convolute edges. Capsule globose, intrastylar

aperture large ; irregularly circumscissile near the base. Seeds 2-4, ovoid, 1 • 5-3 mm.

long, somewhat attenuated at the base; hilum oblong.

From Abyssinia to Portuguese E. Africa and N. Transvaal, as a rule at altitudes

above 1,000 metres (3,000 ft.), in forests.

Transvaal. — Petersburg, Magoeba’s Kloof: Wager s.n. (PRE); Doidge s.n.

(PRE); Taylor 758 (PRE).

Hosts: Various, mainly shrubby. ( Plectranthus , Podranea and, according to

Verdcourt, often on Acanthaceae).

Verdcourt has pointed out that Oliver’s original description is inaccurate is that he

reported the absence of epistamineal scales so that this species keys out to “ C. obtusi-

flora ” in the Flora of Tropical Africa, and indeed many specimens were found in

the herbaria under the latter name.

7. C. hyalina Roth, Nov. PI. Spec. (1821), p. 100, Baker & Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1906), p. 205; Yuncker, op. cit., p. 235, fig. 107; Verdcourt in E. Afr.

Agr. Jl. 18 (1952), No. 2, p. 85—86. C. epitribulum Schinz in Bull. Herb. Boiss.

2me ser., 1 (1901), p. 880; Baker & Rendle, op. cit., p. 236.

Type: Heyne (India).

Flowers 2 -5-3 -5 mm. long, in umbellate cymes on short pedicels, thin in texture,

5- or sometimes 4-merous, shining and yellowish when dry. Calyx campanulate-

turbinate, the triangular or ovate-lanceolate lobes very acute to acuminate, longer

than or about as long as the tube, erect to reflexed. Stamens shorter than the corolla-

lobes, anthers as long as or shorter than the filaments. Scales none. Ovary globose;

styles slender, as long as or longer than the ovary, somewhat unequal. Capsule globose,

irregularly circumscissile near the base. Seeds ovoid, about 1-5 mm. long, hilum

short.

India to Abyssinia, Sudan, and drier areas of Southern Africa.

651

S.W. Africa. — Okahandja: Dinter 1338 (SAM); 4500 (PRE); Bradfield 344

(PRE). Windhoek: Bosch H. No. 25022 (BOL).

Cape Province.— Prieska : Bryant 357, H. no. 3038 (PRE).

8. C. cassytoides Nees ab E. in Linnaea 20 (1847), p. 196, nomen tantum; ex Engelm.

in Trans. Acad. Sci. St. Louis I (1859), p. 513; Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 86; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1906), p. 206;

Wood, Natal PI. 6 (1912), t. 534; Yuncker, op. cit., p. 250, fig. 123; Verdcourt

in E. Afr. Agric. Jl. 18, No. 2 (1952), p. 85-86. C. timorensis Decne. ex Engelm.,

Trans. Acad. Sci. St. Louis 1 (1859), p. 514; Yuncker, op. cit., p. 250, fig. 124;

Ooststr. in Blumea 3 (1938), p. 69 and in Steenis, FI. Males. Ser. I, 4-4 (1953), p. 393

Type: Drege 8307 (the specimen in the herb. Missouri Botanical Garden).

Stems coarse, up to 2 mm. in diam. and over. Flowers nearly sessile, 2-5-4 mm.

long, arranged in few-flowered clusters in paniculate spikes, or inflorescence reduced

to a single short raceme. Calyx cupulate, the lobes broadly ovate to orbicular, broadly

rounded, overlapping, with more or less unequal edges. Corolla campanulate; its

lobes as long as the tube, ovate, obtuse to rounded, erect to reflexed. Stamens subsessile,

filaments much shorter than the anthers. Scales triangular or shorter, truncate or

represented by pairs of narrow wings, shallowly and irregularly fimbriate to almost

entire, free at or adnate to the apex. O vary ovoid-conical or globose-conical; style

longer than the depressed, small stigmas. Capsule globose-ovoid to ovoid-oblong,

circumscissile near the base, 5-8 mm. long. Seeds often less than 4, about 3 mm.

long, hilum long, narrow, terminal.

5. Africa (from Swellendam into Natal and the Transvaal) to E. Africa, Java

and Lesser Sunda Islands.

Cape Province. — Swellendam, Grootvadersbosch: Ecklon & Zevher 22 iPRE).

Knysna: Doidge s.n. (PRE). Ditenhage: Drege 8037 (L, is type!). Humansdorp,

Zitzikamma: Fourcuse 677 (BOL, GRA). Somerset East, Boschberg: Burchell 3178

(PRE); MacOwan 1959 (OS). Albany, Grahamstown: Schlechter 2755 (GRA);

Ga/pin 16 (GRA), PRE); Howieson’s Poort: Zeyher 363 (PRE); Goldspring: Glass

350 vPRE). Keiskamma Hoek: Britten 2887 (GRA). Stutterheim, Dohne Hill:

Sim 2480 (NU). King William’s Town: Ranger 134 (PRE). Peddie: Sim 2479 (NU).

Kentani: Pegler 494 (BOL, PRE). Komgha: Flanagan 1285 (BOL, PRE, SAM).

Willowvale, Qoha Mouth: Meeuse (PRE).

Natal. — Durban, near Durban: Wood s.n. (GRA, SAM); Wood 11075 (NH);

Franks in herb. Wood No. 11704 (NH); Fansdell 1197 (PRE). Pietermaritzburg:

Killick 336 (PRE). Weenen: Thomasset s.n. (PRE). Paul Pietersburg: Acocks 11601

(NH, PRE).

Transvaal. — Barberton: Thorncroft 794 (NH); Galpin 949 (BOL, GRA, NH,

PRE).

Hosts. — Parasitic on many woody plants such as Grewia, Trema, Ficus, in forests,

and thickets.

Yuncker and Van Ooststroom remarked on the close relationship between C.

cassytoides and C. timorensis and Verdcourt, who has studied both types, came to the

conclusion that they are conspecific with which view I agree.

C. cassytoides has page priority (apart from the fact that the name was published

by Nees in 1847, as a nomen nudum it is true, whereas C. timorensis was only published

in 1859, when Engelmann took up Decaisne’s manuscript name) and the name C.

cassytoides N. ab E. ex Engelm. must be adopted.

652

9. C. gerrardii Baker in Dyer, FI. Cap. 4, 2 (1904), p. 84; Yuncker, op. cit., p. 264,

fig. 133. C. cucullata Yuncker, op. cit. (1932), p. 263, fig. 132.

Type: Gerrard 1337, Zululand (K).

Stems slender to medium. Flowers ± 2 mm. long, in loose clusters, more or less

glandular, on short pedicels. Calyx shorter than the corolla-tube or about as long,

more or less thickened at the basal part, its lobes ovate, obtuse or almost acute, often

loose about the corolla. Corolla-tube campanulate ; the lobes erect to slightly spreading,

ovate to more or less triangular, acute, indexed with cucullate tips, about as long as,

or longer, than the tube. Stamens slightly shorter than the corolla-lobes; filaments

somewhat subulate, longer than or about as long as the oval anthers. Scales oblong

or ovate, usually truncate or rounded at the apex, not deeply fimbriate; adnate below

the middle. Ovary globose, more or less depressed and slightly but distinctly attenuated

into a short thickened base; styles shorter to slightly longer than the ovary, often

divergent; stigmas conical or sausage-shaped to oblong, much shorter than the styles.

Capsule globose or depressed-globose, not circumscissile; styles divergent, intrastylar

opening large. Seeds (in the mature capsules studied) 1-2, black, finely tuberculate or

rugose; hilum small, inconspicuous, nearly basal.

Natal. — Eshowe: Lawn 1535 (NH). Entumeni: Wylie in herb. Wood No. 8761

(NH, PRE); Haygarth H. No. 12988 (NH, PRE). “Zululand”: Gerrard & McKen

1337 (NH, isotype!). Umzinto, Dumisa: Rudatis 827 (PRE, isotype of C. cucullata

Yunck.).

Hosts: several unidentified herbs and shrublets.

The original descriptions of C. gerrardii and of C. cucullata were based on single

specimens and an amended description is given above.

The non-circumscissile capsule places C. gerrardii in Yuncker’s subsection

Cucullatae of the section Pachystigma and not in subsection Africanae.

10. C. africana Willd., Sp. PI. (1797), p. 703; Thunb. in Hoffm. Phyt. Blatt. (1803),

p. 17, and FI. Cap. Ed. Schult. (1823), p. 568; Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 85, inch var. capensis Baker; Marloth, FI. S. Afr. 3 (1932),

p. 108; Yuncker, op. cit., p. 264, fig. 134. C. capensis Choisy in DC. Prodr. 9

(1845), p. 454. C. alpestris Fourcade in Trans. Roy. Soc. S. Afr. 21 (1932), p. 89.

Type: Yuncker, in his monograph, does not mention Willdenow as the author

of the name C. africana, although he is correctly credited as such by Thunberg and

in the Index Kewensis. Willdenow’s description is poor but he quotes Cuscuta

americana Thunb. Prodr. 32 (non L.) as a synonym. Cuscuta americana Thunb. is

undoubtedly C. africana and represented by a specimen in Herb. Thunberg [fide Juel,

Plantae Thunbergianae (1918), p. 385]. Moreover, the only specimen in the Willdenow

herb. (no. 3161) that is quoted by Yuncker in the subsection Africanae of the section

Pachystigma belongs to C. africana. This specimen (Willdenow H. No. 3161), identified

by Yuncker, is in my opnion the type of Cuscuta africana, and accordingly, the name

should be quoted as “ C. africana Willd.” (1797) and not as “ C. africana Thunb.”

(1803 or 1813). As it is highly probable that Willdenow received his specimen from

Thunberg, the specimen in the Thunberg herbarium at Uppsala can be taken to

represent an isotype. Willdenow, and others, have pointed out that this plant was

probably part of Linnaeus’s species Schrebera schinoides [Sp. PI. Ed. 2 (1763), p. 1662],

which includes parts of the host plant, Myrica aethiopica L. (= M. conifer a Burm. f.).

However, as Yuncker (op. cit., p. 265) pointed out, Linnaeus’s figure and description

are not sufficient to recognise the species and the plant is not represented in the Linnaean

herbarium. This and the fact that the description of Schrebera schinoides L. contains

653

part of the host, made Yuncker hesitate to substitute Linnaeus’s specific epithet for

the universally applied name africana. I am also of the opinion that Schrebera schinoides,

being a nomen confusum (see Green in Kew Bull. 1935, p. 482) is to be rejected according

to Art. 64 of the Rules.

Stems medium. Flowers on pedicels shorter to longer than the flowers in loose

to compact clusters. Calyx usually distinctly shorter than the corolla-tube; lobes

triangular-ovate, obtuse to subacute. Corolla campanulate-funnelshaped, the lobes

triangular, spreading or reflexed, sometimes suberect, acute or obtuse, usually about

as long as the tube. Stamens shorter than the lobes, filaments longer than the anthers.

Scales about reaching the stamens or longer, oblong with rather small fringes, free at

the apex. Ovary globose; styles slender, longer than the ovary and also longer than

the thickened oblong-cylindrical stigmas. Capsule globose, irregularly circumscissile

near the base. Seeds subglobose; hilum terminal, narrow.

CapeProvince. — Rivier Sonder End (District not quite certain): Thorne H. No. 45784

(SAM). Swellendam, Grootvadersbosch: Zeyher 3447 (PRE, SAM); Tradouw Mts. :

Marloth 8644 (PRE). Riversdale, Langebergen: Schlechter 1842 (US), 5780 (GRA);

Muir 156 in herb. Galpin No. 5155 (BOL, GRA, PRE). George, Kaymans Gat: Drege

7833 (L, PRE, isotvpes of C. capensis ); “ Georgetown”: Hops H. No. 25021 (BOL);

Outeniquas: Esterhuysen 19396 (BOL); Montagu Pass: Fourcade 6495 (BOL); nr.

Touws River: Burche/I 5730 (BOL, L); George: Schlechter 5780 (GRA); Esterhuysen

10858 (BOL); Fourcade 3469, 5307 (BOL). Knysna: Phillips 154 (GRA); Keet 1009

(BOL, GRA, PRE); McN aught on H. No. 18609 (SAM); Fourcade 1508 (BOL, GRA),

5307 (BOL); Barker 6032 (BOL, NBG). Uniondale, Joubertina: Esterhuysen 7075

(BOL), 10699 (BOL); Compton 4490 (BOL); Kammanassie Berg: Esterhuysen 4740a,

Zinn H. No. 5411 (SAM); Headwaters of Wagenboom River: Fourcade 2389 (type

of C. alpestris Fourcade, BOL); Zitzikamma, E. of Klein Bosch River: Fourcade 339

(BOL, GRA). Humansdorp, Rietvlei: Esterhuysen 6651 (BOL, PRE); Kromme Rivier:

Bolus 2406 (BOL). Uitenhage, Van Staadens Mts.: Paterson 892 (GRA); Hutchinson

1497 (BOL); Long 396 (PRE); nr. sources of Bulk River: McOwan 1933 (SAM).

The numbers Wolley Dod 859 from the Cape Peninsula and Schlechter 9043 (not

9093, as erroneously cited by Yuncker, op. cit., p. 267) from Tulbagh, both quoted

under C. africana in Flora Capensis, are C. nitida.

Hosts: Mainly on woody plants such as Laurophyllus capensis Thunb., Myrica,

Berzelia, Phylica, Protea, Geissoloma, Virgi/ia, Ursinia, Clutia, etc. According to

Marloth, C. africana is a perennial which hibernates inside the bark of the hosts [FI.

S. Afr. 3 (1932), p. 108],

11. C. natalensis Baker in Dyer, FI. Cap. 4, 2 (1904), p. 85; Yuncker, op. cit., p. 266,

fig. 135.

Type: Wood 596 from Inanda nr. Durban (Kew), designated by Yuncker.

Stems medium. Flowers 4-7 mm. long, in cymose clusters on pedicels shorter

than the flowers. Calyx much shorter than the corolla-tube, its lobes triangular-ovate,

acute to subacute, the tips often spreading. Corolla cylindrical-campanulate; the lobes

triangular-lanceolate, acute or acuminate, erect to spreading or reflexed, shorter than

or nearly as long as the tube. Stamens shorter than the corolla-lobes; the filaments

about as long as the anthers. Scales oblong, nearly reaching the stamens or shorter,

fringed with rather short processes, free at the apex. Ovary globose; styles slender,

about as long as the thickened cylindrical stigmas, both together considerably longer,

than the ovary. Capsule globose, circumscissile near the base; intrastylar opening

large. Seeds 1-1-25 mm. long, subglobose; hilum a short oblique line.

1332796-2

654

Cape Province. — Komgha: Pegler 911 (PRE, BOL). Kentani: Pegler 1508

(SAM, GRA).

Natal. — Durban, near Durban: Wood 596 (BOL, NH, PRE, SAM, isotypes!).

Botha’s Hill: Wood s.n. (GRA, SAM). Pietermaritzburg: Doidge s.n. (PRE); Natal

Univ. Students 8 (NH). Ixopo: Acoeks 13789 (PRE). Umzinto, Dumisa: Rudatis

1666 (PRE, L).

12. C. nitida E. Mey. ex. Choisy in Mem. Soc. Phys. Hist. Nat. Geneve 9 (1841), p. 272,

pi. 2, fig. 1, and in DC. Prodr. 9 (1945), p. 454; Baker & Wright in Dyer, FI.

Cap. 4, 2 (1904), p. 85; Yuncker, op. cit., p. 266, fig. 136; Salter in Adams. & Salt.,

FI. Cape Penins. (1950), p. 687. C. nitida E. Mey. ex Drege, Zw. Pflanzeng. Doc.

(1843), p. 87, 176, nomen tantum.

Type: “ Ad Paarlberg alt. 2000 ped. rep. Drege! ” (Choisy), in herb. De Candolle

(Geneva).

Stems medium. Flowers in loose cymose clusters on short pedicels, often somewhat

fleshy, granulate and reddish, 3-5 mm. long. Calyx campanulate-turbinate, deep;

its lobes usually short, triangular, acute, but sometimes longer, lanceolate and if so,

longer than the corolla-tube. Corolla-lobes spreading or reflexed, lanceolate or

triangular-lanceolate, acute, as long as or longer than the tube. Stamens shorter than

the corolla-lobes; filaments longer than the anthers. Scales large, oblong, reaching

the stamens, fringed, free at the apex. Ovary globose; styles about as long as the

oblong-cylindrical thickened stigmas, both usually longer than the ovary. Capsule

globose, depressed near the style-bases, circumscissile near the base. Seeds about

1-5 mm. long, ovoid-oblong or ovoid; hilum small, circular to oblong.

Cape Province. — Clanwilliam, Olifants Riv., near Rondegat: Schlechter 10789

(BOL, GRA, PRE); Cedarbergen: Thode A2170 (NH, PRE), s.n. (NH); Compton

4990 (NBG); between Hex River and Kriedouw Krantz: Pearson 5248 (BOL).

Piquetberg, Elandsberg: Pillans 7879 (BOL); Het Kruis: Stephens & Glover 8760

(BOL); Piquetberg: Bolus 7561 (BOL). Tulbagh, Nieuwe Kloof: Schlechter 9043

(BOL, GRA, L, PRE, US; erroneously cited as 9093 by Yuncker and under C. africana

in FI. Cap.); Tulbagh Road: Rogers 17342 (BOL, J, PRE); St. Helena Bay: Marloth

8014 (PRE). Malmesbury, Riebeeks Kasteel: Drege (BOL). Paarl, Wemmershoek:

Esterhuysen 4031 (BOL); Bond 728 (NBG); Paarlberg: Drege (L, isotype!); Klein

Drakenstein: Drege (L); Salter 5006 (BOL). Worcester, Worcester: Ecklon & Zeyher

1-11 (L, BOL, US, GRA); Zeyher 20 (SAM); Diep River: Marloth 7277; Marloth

7449, p.p. Wynberg and Cape Peninsula: Ecklon & Zeyher 62-11 (US, L, GRA,

PRE); Zeyher 1235 (PRE, SAM); Wolley Dod 859 (erroneously cited under C. africana

in FI. Cap.) (BOL); Bolus 4427, 4427a (BOL); Marloth 4252, 5616, 12742 (PRE),

Alexander s.n. (PRE); Moss 7522, 9094 (J); Phillips 146 (SAM); Smith 2908 (PRE);

Guthrie s.n. = H. No. 16961 (BOL); Leighton 376 (BOL); Young 207 (PRE); Salter

2866 (BOL); Pillans 10049 (BOL); Brain 6022 (SRGH); Compton 15508 (NBG).

Stellenbosch: Meyer s.n. (PRE); Smith 6036 (PRE). Somerset-West, Sir Lowry’s

Pass: Schlechter' 727 1 (BOL, GRA, L, PRE, SAM, US). Caledon: McNae 1097

(SAM).

Hosts : Mostly woody plants such as Rhus, Proteaceae, Montinia, Aspalathus,

Passerina, Ericaceae, Phy/ica, suffruticose Compositae, Pelargonium, Oftia and Selago.

The specimen “ leg. Drege, Port Natal ”, cited by Yuncker, must be an error

since C. nitida does not occur in Natal and most probably a mistake was made during

labelling or relabelling.

655

Marloth’s observations on “ C. africana ” being a perennial plant may well apply

to C. nitida, because in Marloth’s herbarium there are several sterile young plants

of a Cuscuta, growing on woody plants (e.g. Montinia ) and collected at Camps Bay

and Somerset West, where C. africana does not occur.

13. C. angulata Enge/m. in 7’rans. Acad. Sci. St. Louis 1 (1859), p. 474; Baker &

Wright in Dyer, FI. Cap. 4, 2 (1904), p. 84; Marloth, FI. S. Afr. 3 (1932), PI. 27;

Yuncker, op. cit. p. 267, fig. 137; Salter in Adams. & Salter, FI. Cape Penins.

(1950), p. 687.

Type: Drege s.n. from Dutoitskloof, Worcester Div. (the specimen in the herb.

Missouri Bot. Garden).

Stems slender. Flowers 3-4 mm. long, often glandular, on pedicels longer to

shorter than the flowers, in loose fasciculate cymes; bracts often numerous on the

longer pedicels. Calyx as long as and enclosing the corolla-tube, protruding at the

sinuses between the lobes to form prominent wings and widest about the middle; the

lobes triangular, acute. Corolla lobes narrowly triangular to lanceolate, erect to

spreading, longer than or equalling the campanulate, more or less angular tube, which

is usually angled opposite the protruding wings of the calyx. Stamens shorter than

the lobes; filaments longer than the anthers. Scales oblong-spathulate, fringed with

processes of medium length, adnate near the base and free for the greater part of their

length. Ovary globose or somewhat oblong, styles slender, longer than the ovary,

and usually longer than the oblong-cylindrical thickened styles. Capsule globose to

oblong or pear-shaped, circumscissile near the base. Seeds usually solitary, ovoid;

hilum terminal.

Cape Province. — Cape Flats: Marloth 4249 (PRE). Cape Peninsula: Salter 8728

(BOL). Hottentots Holland Mts. : Esterhuysen 9805 (BOL). Worcester, Dutoitskloof:

Drege (-‘ Cuscuta africana Th. c.”, isotypes, in L, PRE). Caledon, Houw Hoek:

Schlechter 7381 (BOL, GRA, L, PRE, SAM, US); Onrust Riv. : Schlechter 9506

(GRA, PRE). Hermanus: Taylor 1529 (PRE); Guthrie s.n., H. No. 25025 (BOL,

PRE); Paardeberg Foothills: Stokoe 9214 (BOL, PRE); Cloete H. No. 60916 (SAM).

Bredasdorp: Du Toit s.n. (BOL H. No. 25026); Wolwekloof: Smith 5021, 5045 (PRE);

Elim: Bolus 8580 (BOL, PRE, NH, NBG), Barker 7781 (NGB); Potberg (Potteberg):

Pillans 9490 (BOL, PRE, NBG).

Hosts: Mostly ericoid shrubs: Staavia, Berzelia, Ericaceae, Proteaceae, Phylica,

Passerina, Penaeaceae, Cliffortia , etc.

14. C. planiflora Ten., FI. Napoli t. 3 (1824-1829), p. 250, pi. 220, f. 3; Yuncker. op.

cit., p. 292.

This species is adopted here in a wider sense, as is done by Verdcourt in his

unpublished account of the genus Cuscuta for the Flora of Trop. E. Africa. The following

forms are regarded as synonyms or varieties: C. abyssinica A. Rich., C. balansae Boiss.

& Reutt., C. madagascarensis Yunck. and probably also C. brevistyla A. Braun.

The typical C. planiflora var. planiflora is a plant which occurs in the Mediterranean

region. The various varieties are found in the Mediterranean area (extending into

Persia and India), throughout tropical Africa and in Madagascar. The following

two varieties occur in Southern Africa.

14a. C. planiflora Ten. var. mossamedensis Welw. ex Hiern., Catal. Welw. Afr. PL

1-3 (1898), p. 743; Baker & Rendle in FI. Trop. Afr. 4,2 (1906), p. 203. C.

balansae Boiss. & Reutt. var. mossamedensis (Welw. ex Hiern) Yuncker, op. cit.,

p. 291, fig. 154, F, G.

656

Type: Welwitsch 6141 (the specimen in K, according to Yuncker; more correctly,

in my opinion, the specimen in BM).

Stems slender. Flowers 4-merous, in dense, few-flowered clusters, 2-3 mm. long.

Calyx somewhat fleshy, the lobes triangular or triangulat-ovate. Corolla campanulate

with spreading, triangular-ovate lobes which are shorter than the tube. Stamens shorter

than the corolla-lobes, the subglobose anthers on short filaments. Scales usually

shorter than the corolla-tube, rather deeply fringed, the fringes clavate, distinctly

thickened at their tips. Ovary globose, styles very short, shorter than or equalling the

cylindric stigmas. Capsule globose, circumscissile near the base. Seeds small, ovoid,

0-5-0-75 mm. long.

S.W. Africa. — Okavango, Andara: de Winter 4459 (PRE). Okahandja, Okakuja:

Grosserth in Herb. Dinter s.n. (SAM, H. No. 61987); Okahandja: Dinter 37 (SA.M,

PRE).

Hosts: Indigofera sp., Barleria sp., Monocotyledons, and Welwitsch’s type on

Merremia multisecta Hall. f. Also recorded from Angola.

The specimens cited above agree very well with Hiern’s and Yuncker’s description,

except that the scales are different in that the fringes are somewhat clavate, thickened

towards their apices, a character not mentioned by Hiern or Yuncker. However, an

isotype of this variety ( Welwitsch 6141 in COI) proved to possess the same thickened

fringes and is identical with the S.W. African material.

14b. C. planiftora Ten. var. madagascarensis ( Yuncker ) A. Meeuse, comb. nov.

C. madagascarensis Yuncker, op. cit., p. 276, fig. 277, inch var. schlechteri Yunck.,

op cit., p. 277.

Type of variety: Baron 3466 from Madagascar (K).

Stems medium. Flowers about 3 mm. long, in compact few-flowered clusters,

4- or 5-merous. Calyx rather loose about the corolla, the lobes ovate, often somewhat

fleshy at the tips. Corolla membranous, campanulate; the lobes erect, ovate, obtuse,

shorter than to about as long as the tube. Stamens shorter than the lobes, filaments

slightly longer than the anthers. Scales ovate, oblong or spathulate, entire or somewhat

bilobed, fringed mostly at the top, free at the apex. Ovary subglobose; styles and

stigmas erect, rather thin, subequal, together usually about as long as the ovary.

Capsule depressed-globose, circumscissile near the base. Seeds ovoid, hilum short

oblong.

Madagascar, S. Africa, E. Africa.

Natal. — Hilton Road: Schlechter 6761 (BOL, GRA, isotypes of C. madagas-

carensis Yunck. var. schlechteri Yunck.).

Transvaal. — Letaba, Duiwelskloof : Gerber H. No. 5678 (PRE).

Hosts: Apparently low, herbaceous or suflruticose Dicotyledons. The var.

madagascarensis seems to be a very rare form.

Dr. B. Yerdcourt, in a private communication, pointed out that there is one

specimen collected in East Africa (Kenya, Aberdares: Fries 2216, in K).

15. C. epithymum Murr. in Linn., Syst. Veg. Ed. 13 (1774), p. 140; Yuncker op. cit.,

p. 283, fig. 151. Cuscuta trifolii Babingt. in Phytologist 1 (1843), p. 467.

Type: No type specimen, but Pinax (219) of Bauhin used as basis of the species

(t. Yuncker).

657

Stems very slender, often reddish. Flowers about 3 mm. long, in dense many-

flowered globose clusters, often reddish. Calyx usually about as long as the corolla-tube;

the lobes triangular, acute. Corolla-lobes triangular, acute, spreading, shorter than the

tube. Stamens shorter than the corolla-lobes; filaments longer than the anthers.

Scales more or less spathulate, shorter than the corolla-tube, fringed mainly in the

upper part, free at the apex. Ovary globose: stigmas filiform, slightly longer than the

styles, together about twice as long as the ovary. Capsule globose, circumscissile near

the base. Seeds usually 4, ovoid, about 1 mm. long, rather rugose, angular, compressed,

hilum short, oblong, transverse.

A native of Europe, but spread with commerce throughout the world, mainly as

an infestant of leguminous hosts.

Cape Province.— E. London: Dodd s.n. (GRA). Port Alfred: Hutton 934

(GRA); Storms Riv. : Laughton s.n. (PRE).

Orange Free State. — Kretzmar s.n. = H. No. 10188 (PRE).

Transvaal.— Pretoria : Van Gass H. no. 947 (PRE); Smith 6215, 6218, 6221,

6235 (PRE).

Hosts: A great variety of plants, but outside its natural area of distribution chiefly

on leguminous crops. In S. Africa this species has been recorded occasionally; the

first records date from about 1894. It does not appear to be such a serious pest as

C. campestris.

2. DICHONDRA

J. G. et R. Forst., Char. Gen. (1776), p. 39, t. 20; Choisy in DC., Prodr. 9 (1845),

p. 451; Benth. & Hook., Gen. PI. 2 (1876), p. 879; Peter in Engl. u. Prantl, Nat. Pfl.

Fam., Ed. 1, 4-3a (1891), p. 13; Hall. f. in Engl. Bot. Jb. 16 (1893), p. 569 and 18

(1893), p. 82; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 83; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 65; Ooststr. in Blumea 3 (1938). p. 72 and in

Steenis, FI. Males. Ser. 1, 4-4 (1953) p. 394; Phillips, Gen. S. Afr. Flow. PL, Ed. 2

(1951), p. 621.

Small creeping herbs. Leaves simple, petioled, cordate-orbicular or reniform,

entire. Flowers small to minute, solitary, axillary, pedicellate; bracteoles 2, minute,

subulate. Sepals 5, free or rarely so, subequal, often spathulate, somewhat accrescent.

Corolla widely campanulate, deeply 5-lobed, not much longer than the calyx, hirsute

outside, the lobes induplicate- valvate. Genitalia included. Stamens 5; filaments

filiform, short, inserted in the corolla-tube between the lobes; anthers oblong to sub-

globose; pollen smooth. Ovary deeply 2-lobed ; each lobe 1 -celled, 2-ovuled; styles 2,

gynobasic and inserted between the lobes, short, filiform; stigmas capitate. Capsule

2-lobed; the lobes erect, membranous, 1- or rarely 2-seeded, indehiscent or irregularly

2-valved. Seeds subglobose, smooth, with thin crustaceous testa; cotyledons linear-

oblong or elliptic, more or less spirally coiled or plicate (twice folded); radicle curved,

terete ; endosperm scanty.

Type Species: Dichondra repens J. G. et R. Forst.

A small genus of 4-5 species, principally American, with one species in the tropical

and subtropical regions of both hemispheres.

D. repens J. G. et R. Forst., Char. Gen. (1776), p. 40, t. 20; Choisy op. cit. p. 451;

Hall, f., op. cit., 18 (1893), p. 82; Baker & Wright, op. cit., p. 83; Baker & Rendle,

op. cit., p. 65; Van Ooststr., op. cit. (1938), p. 72 (1953), p. 395.

Type: The exact holotype of most of the Forsters’ species cannot be indicated,

their specimens being widely distributed and represented in many herbaria [See E. D.

Merrill in Chron. Bot. 14: 208-211 (1954)].

658

Creeping perennial herb. Stems slender, rooting at the nodes, shortly hairy. Leaves

long-petioled, reniform to cordate-orbicular, 4-25 mm. in diam., broadly cordate at

the base, broadly rounded or emarginate at the apex, adpressed-hairy to strigose on

lower surface usually glabrescent on upper surface ; petioles usually densely adpressed-

hairy. Flowers solitary, pedicels usually shorter than the petiole, terete, hairy like the

petioles. Sepals obovate-oblong to spathulate, obtuse, 2-3 mm. long, hairy on back

and margins. Corolla shorter to slightly longer than the calyx, deeply 5-lobed, greenish-

yellowish. Stamens shorter than the corolla ; filaments filiform; anthers small; pollen

smooth. Capsule 2-lobed, lobes pilose, 1 - or rarely 2-seeded ; indehiscent or irregularly

2-valved. Seeds subglobose, smooth, glabrous.

Widely spread in the warmer regions of both hemispheres. Wide-spread in the

Union, but probably often overlooked on account of its very small, inconspicuous

flowers and its resemblance to species of Centella and Hydrocotyle , which occur in

similar habitats (moist places).

Cape Province. — Cape Peninsula: Ecklon 15 (= ? No. 406 mentioned in FI.

Cap.) (SAM); Schlechter 723 (BOL); Compton 3493 (BOL, NBG); Salter 8694 (BOL) ;

Momis 101 (NBG). Rondebosch and Newlands: Schlechter s.n. (BOL, H. no. 25045).

Knysna: Duthie 695 (BOL). Port Elizabeth: I. L. Drege s.n. (GRA). Grahamstown:

Archibald 1781 (PRE). Griqualand East: Baur 490 (SAM, probably from Baziya,

Umtata).

Natal. — Isipingo: Franks in herb. Wood no. 11062 (NH). Durban, Bluff:

Meebold s.n. (NH, H. no. 23127). Umzinyati Valley: Wood 1379 (BOL, NH, PRE,

SAM). Estcourt: West 1845 (NH).

Transvaal.- — Marico, Zeerust: Leendertz 4170 (PRE). Ventersdorp: Sutton 699

(PRE). Potchefstroom : Louw 1055 (PRE); Leeuwpoort: Mogg 23007 (PRE).

Krugersdorp : Moss 8125 (J); Mogg 23202 (PRE). Johannesburg: Wallace s.n.

(PRE); Gilliland vJ, H. no. 25415). Brits, Wolhuterskop: Nunns s.n. (PRE, H. no.

18772). Pretoria, Pretoria: Chippindall 30 (PRE); Hennops Riv. : Prosser 1627

(J, PRE); Meeuse 9625 (PRE). Middelburg: Mogg 16921 (PRE). Pietersburg,

Spelonken: Junod 38 (PRE). Zoutpansberg, Louis Trichardt: Young s.n. (PRE,

h. no. 26963); between Louis Trichardt and Entabeni : Meeuse 9215 (PRE). Transvaal,

without precise locality: Bur tt- Davy 15144 (J).

3. FALKIA

Linn, f, Suppl. (1781), p. 30; Thunb., Nov. Gen. (1781), p. 17 (“ Falekia ”);

Choisy in DC., Prodr. 9 (1845), p. 451; Benth. & Hook., Gen. PI. 2 (1876), p. 878;

Peter in Engl. & Prantl, Nat. Pfl. fam. Ed. 1, 4- 3a (1891), p. 14; Hall. f. in Engl. Bot.

Jb. 16 (1893), p. 569 and 18 (1893), p. 84; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904),

p. 81 ; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 65; Phillips, Gen. S. Afr.

Flow. PI. Ed. 2 (1951), p. 621.

Dwarf, prostrate, often matted, perennial herbs with petioled, ovate or cordate-

orbicular to oblong-lanceolate, entire leaves and small solitary, axillary, peduncled

flowers. Calyx shorter than the corolla-tube, sepals broad, divided nearly to the base

or forming a short tube, subequal, accrescent. Corolla-tube funnel-shaped or cam-

panulate; limb plicate, 5-angled or shortly and broadly 5-lobed. Stamens 5, inserted

on the corolla-tube, included; filaments linear; anthers oblong. Ovary deeply 4-lobed,

with a single ovule in each lobe, hairy; styles 2, gynobasic, almost equalling the corolla-

tube, linear or filiform; stigmas subglobose. Fruit divided into four membranous

utricles (sometimes fewer by abortion). Seeds obovoid, or subglobose; testa crus-

taceous; embryo curved; cotyledons flat, plicate; endosperm scanty.

Type Species: Falkia repens Linn, f., Suppl. (1781), p. 211.

659

Two species, natives of Africa, one confined to the Cape Province, the second

wide-spread from Natal and the Orange Free State to Abyssinia.

Leaves broadly ovate to reniform, about as long as broad or slightly longer than

broad, distinctly cordate at the base 1 . F. repens.

Leaves oblong to lanceolate-oblong or lanceolate, usually about twice as long

as broad, but varying from less than twice as long to several times longer than

broad, never distinctly cordate at the base 2. F. oblonga.

1. F. repens Linn, f., Suppl. (1781), p. 211; Thunb., Nov. Gen. (1781), p. 17; Choisy,

op. cit., p. 451; Hall, f., op. cit., 18 (1893), p. 84 (inch var. sericea ); Baker &

Wright, op. cit., p. 81 inch vars.; Adamson & Salter, FI. Cape Penins. (1950),

p. 686. F. villosa Hall, f., op. cit. 18 (1893), p. 85. F. diffusa, (Choisy) Hall,

f., l.c. (pro parte?) F. dichondroides Baker in Dyer, FI. Cap. 4, 2 (1904), p. 82.

Type: It is evident that Thunberg must have shown his specimen of F. repens

to the younger Linnaeus (or sent him a duplicate), but as there is no specimen in the

Linnaean herbarium or in the Stockholm herbarium, Thunberg’s specimen in the

herbarium Uppsala must be considered to be the lecto-type.

The publication by Linn. f. prior to Thunberg’s publication of the same species

in the same year is probably one of the cases of “ snatching ” emphasized by Otto

Kuntze in his “ Revisio According to Juel [in Plantae Thunbergianae, (1918),

p. 8 ff.] the publication of many of Thunberg’s plants by the younger Linnaeus was

done by mutual co-operation. At any rate, the publication of the younger Linne’s

Supplementum Plantarum is generally assumed (by inference) to antedate Thunberg’s

Nova Genera Plantarum and, therefore, “ Linn, f.” has to be cited as the author of

both the genus Falkia and the species Falkia repens. I am of the opinion that Juel

was wrong when he cited Falkia (or Falckia) repens “ Thunb.” As regards the spelling,

Thunberg named it after Falck and spelled it “ Falckia ”, but Linnaeus the younger

adopted the spelling Falkia and this spelling has to be retained.

Creeping perennial herb, hairy to glabrous, often covering large patches of ground.

Stems slender but firm, terete, up to 50 cm. long and over, rooting at the nodes. Leaves

cordate-ovate to orbicular, 6-25 mm. long and wide; base distinctly cordate, apex

rounded to emarginate, petiole shorter or longer than the blade. Peduncle 1-flowered

about as long as leaf and petiole, recurved in fruit. Calyx 4-7 mm. long, lobes at

first oblong, becoming ovate and deltoid to subcordate, crisped along the edges and

enlarged in fruit. Corolla about as long to twice as long as the calyx, white to

pale pink drying yellow, shallowly to deeply lobed. Stamens inserted low down in

the corolla-tube; anthers roundish; pollen smooth. Ovary deeply 4-lobed, each

lobe 1-ovuled; styles 2, subulate, gynobasic; stigmas capitate. Capsule membranous,

completely enclosed by the accrescent, marcescent calyx, about 2 mm. long.

F. repens has been recorded from the following districts: Tulbagh, Cape Peninsula,

Wynberg, Somerset, Caledon, Robertson, Bredasdorp, Riversdale, Mossel Bay, George,

Port Elizabeth, Uitenhage, Somerset East, Bedford, Albany, Victoria East, Bathurst,

Keiskammahoek, Stutterheim, King William’s Town, East London, Komgha, Kentani,

Umtata, Umzinkulu, Port Shepstone.

Falkia dichondroides is distinguished in Flora Capensis as having the corolla

scarcely longer than the calyx and distinctly lobed, whereas the corolla in F. repens

is mentioned as “ twice as long as the calyx ” and scarcely lobed. It is impossible to

distinguish the two forms in the herbarium, because lobed corolla-limbs occur in

specimens with long corollas, and slightly lobed limbs in specimens with short corollas.

Dr. R. Story, who studied the vegetation of the Eastern Cape Province, kindly supplied

ased upon observations in the field apart

660

from herbarium specimens) most plants in the Eastern Cape Province could be referred

to either “ F. repens ” or to “ F. dichondroides ” of the Flora Capensis.

Although the name F. dichondroides is reduced to a synonym, it is necessary from

a nomenclatural point of view to point out that this name is illegitimate. Hallier (in

Engl. Bot. Jb. 18, p. 85) described a species F. diffusa and quotes “ F. repens a diffusa

Choisy in DC. Pr. 9, p. 451 ” as a synonym. Choisy did not cite any specimens which

he referred to his variety diffusa, but Professor Baehni of the Geneva herbarium kindly

sent me those annotated specimens which Choisy himself had named F. repens a diffusa

and those which Hallier referred to F. diffusa. There is only one sheet (leg. Drege,

from Enon, Zuurberg, with an original label in E. Meyer’s handwriting “ Falkia repens

Th. d.”) on which Choisy (in 1841) has annotated: “ F. repens var. diffusa ”. A label

attached by Hallier (1898) says: “ Falkia repens L.”. Another sheet of Drege’s “ Falkia

repens Th. d.” 'in G bears only a label by Hallier, dated 1892, “ Falkia diffusa m.”.

Hallier, in 1893, quoted “ F. repens Th. d, Drege ” among the specimens he referred

to F. diffusa , and although he may have referred two sheets of the same Drege gathering

to two different species, he had perhaps changed his mind in 1898. However, another

sheet of the same Drege gathering “ Falkia repens Th. d.” in the Leiden herbarium,

annotated by Hallier in 1909 or 1910, was referred by him to F. diffusa Hall. f.

At any rate, the name Falkia diffusa can only be retained for forms of which the

sheet annotated by Choisy is the type. In my opinion, the two sheets leg. Drege in

the Geneva herbarium represent the same thing, i.e., the form named F. dichondroides

in Flora Capensis. The gathering “ F. repens Th. d ” of Drege is cited in Flora Capensis

under F. dichondroides , so that Choisy’s type is mentioned among the specimens quoted

under this newly described species. For this reason F. dichondroides becomes a synonym

of F. diffusa.

F. repens buries its fruits actively into the soil by a downward bending of the

penduncles after flowering. The persistent, hard calyx which encloses the ripe fruit

serves as a protection during this process.

2. F. oblonga Bernh. apud Krauss in Flora 27 (1844), p. 830; Hall. f. in Engl. Bot*

Jb. 18 (1893), p. 84, and in Bull. Herb. Boiss 7 (1899), p. 41; Baker & Wright

in Dyer, FI. Cap. 4, 2 (1904), p. 82, inch var. minor C. H. Wright; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 65. F. abyssinica Engl., Hochgebirgsfl.

Trop. Afr. (1892), p. 344; Hall, f., op. cit. (1893), p. 84.

Type: Krauss 359 from Natal, nr. Umlaas Riv. (isotype K).

Leaves about twice as long to several times as long as broad, up to about 7 cm-

long and 1 • 5 cm. wide, never distinctly cordate but usually rounded at the base, otherwise

as F. repens.

South Africa, Portuguese East Africa, also in Abyssinia, Eritrea and Somaliland.

Bechuanaland. — (Omaramba u Omatako).

Griqualand-West. — Vryburg, Mafeking, Barkly West, Kimberley.

Transvaal. — Lichtenburg, Marico, Potchefstroom, Vereeniging, Johannesburg,

Krugersdorp, Springs, Boksburg, Benoni, Pretoria, Pietersburg, Lydenburg, Bethal,

Carolina, Heidelberg, Standerton.

Orange Free State. — Vredefort, Kroonstad, Senekal, Bloemfontein, Fauresmith.

Eastern Cape.- — Albert, Aliwal North, Richmond, Fort Beaufort, Middelburg,

Queenstown, Pt. St. Johns.

Natal. — Durban .

Portuguese E. Africa. — Lourengo Marques.

661

An isotype ( Krauss 359, K) was studied by Mr. de Winter, who reported that the

leaves are up to about 2 cm. long. Actually the leaves vary a great deal in shape,

length and pubescence; they can be from lanceolate to oblong-orbicular, from ±

1 cm. to 7 cm. in length, from acute to obtuse, from villous or tomentose to almost

glabrous. In my opinion, these forms are only ecotypes and not worth varietal rank.

F. oblonga is very close to F. repens and it is even possible that they are not more

than different subspecies or varieties. They are, however, almost completely geographi-

cally separated and that is, apart from the difference in the shape of the leaf-base,

why I hesitate to regard them as one species.

F. oblonga is wide-spread in S. Africa, but very scattered, most probably because

it is found in wet places (as is F. repens ) and its ecological requirements make it a

species of local occurrence.

4. EVOLVULUS

L., Sp. PI. Ed. 2 (1762), p. 391 ; S. J. van Ooststroom, “A Monograph of the Genus

Evolvulus ”, Meded. Bot. Mus. Elerb. Utrecht No. 14 (1934); Philips, Gen. S. Afr.

FI. PI. Ed. 2 (1951), p. 62.

Type Species: E. nummularius L., Sp. PI. Ed. 2 (1762), p. 391.

Annual or perennial herbs or suffrutices, not twining. Leaves usually small,

entire, often sessile. Inflorescences usually axillary, 1-3-nate; bracteoles small. Sepals 5.

Corolla funnel-shaped or campanulate to subrotate, the limb 5-angled or shortly

5-lobed. Stamens 5, inserted above the middle of the corolla-tube, rarely near the base;

filaments linear; anthers linear. O vary 2-celled, or 1 -chambered by abortion, 4-ovuled;

styles 2, free from the base, each divided above or almost to the base, stigmas 4, linear-

terete or sub-clavate. Capsule subglobose, 2-4-valved. Seeds 4 or fewer by abortion,

glabrous.

Distribution: 97 species mentioned by Van Ooststroom, all American, with the

exception of two species, viz. the circum-tropical E. alsinoides, and E. nummularius

(America, Africa, Madagascar, introduced in India). One species in S. Africa.

E. alsinoides (L.) L., Sp. PI. Ed. 2 (1762), p. 392; Van Ooststr. op cit., p. 26 (for full

synonymy, cf. Van Ooststr., l.c.).

Type: In the Linnean herbarium ( teste Van Ooststroom).

Perennial herb, thinly or sometimes rather densely covered with patent pilose

hairs. Stems few to several, erect or decumbent, slender, up to 30 cm. sometimes to

60 cm. long. Leaves nearly sessile, ovate-oblong to lanceolate, entire, obtuse, distinctly

mucronate, 10-20 (-30) mm. long. Peduncles filiform, shorter to much longer than the

leaves, one- to few-flowered; bracts minute, linear; pedicels as long as or longer than

the calyx, spreading. Calyx densely silky or villous; sepals lanceolate, acute to

acuminate, 3^4 mm. long. Corolla rotate, 6-8 mm. in diam., bright light blue or

occasionally white. Ovary 2-celled, each cell 2-ovuled, glabrous; styles 2, stigmas 4,

long, terete or subclavate. Capsule depressed-globose, glabrous, 4-valved, fragile,

about as long as the calyx. Seeds 4 or less, black, smooth, glabrous.

Northern districts of S.W. Africa; Griqualand-West: Kuruman, Vryburg, Barkly-

West, Hay, Herbert, Kimberley; Bechuanaland; Transvaal: Marico, Rustenburg,

Potchefstroom, Pretoria, Lydenburg, Barberton and districts to the N. of these;

Swaziland; Natal and 7Mluland as far S. as Umzinto. Common and wide-spread.

Note. — This species is very variable and Van Ooststroom distinguished a number

of varieties. All the South African specimens I have seen are referable to the var.

662

linifolius (L.) Baker. I have not seen a single specimen referable to the var. glaber;

Flora Capensis and Van Ooststroom mention only one specimen from South Africa,

viz. Gerrard 1907 from Natal, which I have not seen.

5. SEDDERA

Hochst. in Flora 27 (1844), Bes. Beil. 7, t. 5; Choisy in DC. Prodr. 9 (1845),

p. 440; Hall. f. in Engl. Bot. Jb. 16 (1893), p. 572, and 18 (1893), p. 88; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 73. Breweria R. Br., Prodr. FI. Nov. Holl. (1810),

p. 487, pro parte; Benth. & Hook., Gen. PI. 2 (1876), p. 876; Peter in Engl. u. Prantl,

Natiirl Pflanzenfam., ed. 1., 4- 3a (1891), p. 16; Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 79; Phillips, Gen. S. Africa FI. PI. Ed. 2 (1951), p. 622, pro parte.

Type Species: Seddera virgata Hochst. et Steud. in Flora 27 (1844), Bes. Beil. 8.

t.5, figs. 1-20.

Small prostrate to suberect shrubs or suff'rutiees, never climbing. Leaves small’

entire. Flowers axillary, sessile or pedunculate, solitary or aggregated into few-flowered

clusters or pedunculate cymes which sometimes form leafy terminal spikes or panicles.

Bracteoles usually small. Sepals 5, acute or obtuse, subequal or the outer ones slightly

larger. Corolla small, (always?) white, funnel-shaped, the limb usually shallowly lobed;

the midpetaline areas hairy at least at the tips. Stamens inserted low down in the

corolla-tube; filaments filiform, dilated at the base and often appendaged; anthers

oblong; pollen smooth. Disc none or small. Ovary 2-celled, 4-ovuled, hairy at the

apex, style bifid almost or quite to the base; stigmas more or less peltate and orbicular,

sometimes bilobed. Capsule 4-valved. Seeds dark brown or black, glabrous.

Distribution: About 15 species, mainly restricted to Africa, Madagascar and

Arabia, one in India.

As regards the synonymy, Hallier (Engl. Jb. 16, p. 563 in the key and p. 572)

pointed out the differences between Seddera and Bonamia (= Breweria R.Br. s.s.).

Seddera is always erect or prostrate, suffruticose or shrubby, with small flowers, glabrous

stamens and peltate stigmas; Bonamia is herbaceous, suffruticose or woody and climbing

(all African species are shrubby and usually climbing) with usually large flowers, with

filaments which are usually glandular-villose at the base, very rarely glabrous (glabrous

in the only S. African species) and (in the African species) with 2 globose stigmas.

Phillips, l.c., pointed out that the styles in Seddera are free or more or less united’

so that “ the only difference ” between Seddera and Breweria does not hold true, but

Hallier does not mention the style character at all, on the contrary, in his generic

diagnosis of Seddera he mentions “ stylus 2-fidus vel styli 2 ”. Seddera can always

be distinguished from Bonamia by a combination of characters and by its habit, so

that it is better to retain Seddera Hochst. as a separate genus, distinct from Bonamia

Thouars (which includes Breweria).

Flowers always solitary, sessile or very rarely pedicellate; calyx 7-10 mm. long,

the sepals ciliate with bulbous-based hairs; corolla ± 10 mm. long or

longer; mid-petaline ares with a few long strigose hairs; leaves usually

more or less ciliate with bulbous-ba«ed hairs and, therefore, in most

cases appearing minutely dentate or serrate; pubescence of stems, leaves

and calyx usually distinctly brown or ferrugineous 1. S. capensis.

Flowers usually in few-flowered axillary cymes, rarely all solitary, sessile or

pedunculate (peduncles up to 35 mm. long); calyx usually only 4-7 mm.

long (if longer, densely tomentose, the sepals not ciliate); corolla usually

6-7 mm. long (if longer, midpetaline areas rather densely pubescent);

leaves more or less silky or shortly pubescent, not with bulbous-based

hairs and appearing quite entire; pubescence of all parts almost invariably

white or grey : 2. S. suffruticosa.

663

1. S. capensis ( E . Mey. ex Choisy) Hall.f. in Bull. Herb. Boiss. 6 (1898), p. 529; Baker

& Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 77. Evolvulus capensis E. Mey.

ex Drege, Zw. Pfl. geog. Doc. (1843), p. 46, nomen tantum; ex Choisy in DC.,

Prodr. 9 (1845), p. 444; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 86. Breweria

capensis (E. Mey. ex Choisy) Baker in Dyer, FI. Cap. 4, 2 (1904), p. 80. Bonamia

capensis (E. Mey. ex Choisy) Burtt Davy in Ann. Transv. Mus. 3 (1912), p. 121.

Type: A specimen leg. Drege from Shiloh (E. Cape) in herb. Geneva (isotypes

seen in L).

Suffruticose perennial. Stems several from a firm woody taproot, up to 30 cm.

long but often much shorter, prostrate to suberect, firm, terete, covered with rusty-

brown adpressed to patent stiff hairs as are petioles, leaves, pedicels, calyces and mid-

petaline areas of the corolla. Leaves ovate, ovate-lanceolate or oblong, sessile or shortly

petioled, strigose on both sides, more laxly so when older, 8-25 mm. long and 4-12

mm. wide, obtuse or subacute, minutely mucronate, rounded to somewhat narrowed

or truncate at the base, ciliate with bulbous-based hairs along the margin. Flowers

axillary, solitary, usually subsessile, pedicels rarely up to 10 mm. long. Bracteoles 2,

lanceolate, shorter than the sepals. Sepals broadly lanceolate, acute, 6-8 mm. long.

Corolla broadly funnelshaped, white, 8-12 mm. long. Ovary hairy at the apex. Capsule

subglobose, usually crowned with a tuft of hairs, about 5 mm. in diam. Seeds black,

glabrous smooth.

Recorded from the following areas: Beclmana/and (Mochudi, one record only);

Gri'tualand W. (Kuruman, Barkly West); Transvaal (Marico, Bloemhof, Christiana,

Potchefstroom, Vereeniging, Pretoria, Leta’oa); Orange Free State (Hoopstad, Kroon-

stad, Ventersburg, Bloemfontein); Natal tEstcourt, “Upper Tugela’’); E. Cape

(Queenstown, nr. Kei River). Extends into Southern Rhodesia.

2. S. suffruticosa ( Schinz ) Hall.f. in Engl. Bot. Jb. 18 (1893), p. 88; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 77. Breweria suffruticosa Schinz in Verhandl.

Bot. Ver. Brandenb. 30 (Sept. 1888), p. 275; Baker & Wright in Dyer, FI. Cap.

4, 2 (1905), p. 80. Convolvulus mucronatus Engl, in Engl. Bot. Jb. 10 (Oct. 1888),

p. 246. Seddera mucronata (Engl.) Hall, f., op. cit., p. 88. Breweria baccharoides

Baker in Kew Bull. 1894, p. 68. B. sessiflora Baker, l.c. Seddera welwitschii

Hall, f., op cit., p. 88; Baker & Rendle, op cit., p. 77. Bonamia suffruticosa

(Schinz) Burtt Davy in Ann. Transv. Mus. 3 (1912), p. 121.

Type: Schinz 750 from Oshando, S.W. Africa, in herb. Zurich (isotypes seen in

BOL, GRA, and L).

Suffruticose or herbaceous perennial, extremely variable. Stems several to many

from the base, often woody, firm, but young ones often slender, herbaceous; suberect

or erect, rarely prostrate, terete or subterete, at first more or less densely covered with

stiff, almost invariably whitish hairs to villous, ultimately glabrous and making a thin

corky bark, 15-50 cm. long. Leaves varying from lanceolate to broadly elliptic-oblong

(5-) 10-25 (-40) mm. long and (2-) 4-10 (-18) mm. wide, acuminate, acute or rounded

at the apex, mucronate, strigose with whitish hairs on both sides, margin entire;

petioles up to 4 mm. long. Flowers in axillary clusters or sometimes, or partly, solitary,

sessile, rarely pedunculate; penduncles, if present, up to 3 cm. long, terete, slender,

bearing one to several flowers in a capituliform cyme. Bracteoles shorter than the calyx,

lanceolate. Calyx 5-9 mm. long, thinly strigose to densely villous; three outer sepals

lanceolate, acuminate, inner ones slightly shorter. Corolla somewhat campanulate,

white, 6-12 mm. long; limb rather spreading; midpetaline areas more or less densely

silky-strigose. Ovary hairy at the apex. Capsule ovoid-subglobose with some white

hairs near the apex, about 5 mm. in diam. and about 6 mm: long. Seeds black, glabrous,

smooth.

664

South Africa, extending into Angola, Bechuanaland, Southern Rhodesia, Portuguese

East Africa and possibly into tropical East Africa.

Recorded from: S.W. Africa (N. districts; one of the sheets, viz., Dinter 55,

bears a manuscript name, Evolvulus Juttae Dinter Ms., in SAM); Bechuanaland;

Griqualand West (Yryburg, Taungs, Kuruman, Barkly West, Hay, Herbert, Kimberley);

Transvaal (Brits, Pretoria, Bronkhorstspruit, Waterberg, Groblersdal, Potgietersrust,

Pietersburg, Zoutpansberg, Letaba, Barberton); Portuguese East Africa (Louren^o

Marques).

This species is rather variable in its pubescence, the size of the leaves and the

length of the peduncles. Hairy forms have been placed in a var. hirsutissima Hall. f.

[in Bull. Herb. Boiss. 6 (1898), p. 531], but as far as I can see all these forms intergrade.

The peduncles, usually undeveloped, may attain a length of 35 mm. as, for instance,

in Hutchinson 2478 (from Barberton, Louw’s Creek), Turner 10 (Waterberg, Tvl.),

Gomes e Sousa 144 (Lourengo Marques) and these specimens approach, according

to Mr. de Winter who compared the type, S. welwitschii Hall, f., but are not quite

identical. Several authentic specimens of S. welwitschii I saw in COI ( Welwitsch 6152,

6154 and 6159) are very similar to forms of S. suffruticosa. The specimen Welwitsch

6159, which number was referred by Hallier [in Bull. Herb. Boiss. 5 (1897), p. 1009],

to S. welwitschii var., and by Hiern [in Cat. Welw. Afr. PI. (1898), p. 735] and Rendle

(in FI. Trop. Afr. 4, 2, p. 77) to S. welwitschii var. bakeri Hiern, is indistinguishable

from the type of S. suffruticosa ( Schinz 750). Therefore, in my opinion, 5. welwitschii

is only a synonym of S. suffruticosa. This does not alfect the name of the South African

plants, because the oldest epithet is Breweria suffruticosa Schinz (Sept. 1888), which

antedates Engler’s Convolvulus mucronatus (Oct. 1888) by a few weeks and Seddera

welwitschii (1893) by several years. The types or isotypes of Convolvulus mucronatus,

Breweria baccharoides and B. sessiliflora were either compared by myself or by Mr.

de Winter and they all do not appear to be specifically distinct.

6. BONAMIA

Dupetit-Thouars in Hist. Veg. Isl. Austr. Afr. (1804), p. 33, t. 8, et in Diet. Sci.

Nat. 5 (1806), p. 145, nom. cons.; Poir. in Lam., Encycl. Meth., Bot., Suppl. (1810),

p. 677; Choisy in DC., Prodr. 9 (1845), p. 439; Benth. and Hook., Gen. PI. 2 (1876), p. 877;

Peter in Engl. u. Prantl, Natiirl. Pfl. fam. 4- 3a (1891), p. 17, 376; Hall. f. in Engl. Bot.

Jb. 16 (1893), p. 527, 573, and in Bull. Herb. Boissier 5 (1897), p. 804, 996; Baker

& Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 78; Oostr. in Blumea 3 (1938), p. 75,

and in Steen., FI. Males., Ser. 1, 4-4 (1953), p. 398. Breweria R.Br., Prodr. FI. Nov.

Holl. ed. 1 (1810), p. 487. Metaporana N.E.Br. in Kew Bull. 1914, p. 168.

Type species: Bonamia madagascariensis Poir., l.c. [see Kew Bull. 1935, p. 381,

where also the conservation of Bonamia Dupetit-Thouars against Bonamya Neck.

(= Stachys L.) is discussed].

Herbaceous or woody twiners, rarely erect undershrubs. Leaves herbaceous or

occasionally subcoriaceous, entire, lanceolate, ovate or elliptic. Flowers axillary,

solitary or cymose, cymes sometimes forming terminal leafy panicles, bracteoles usually

small. Sepals 5, equal or subequal, rarely very unequal, orbicular to lanceolate, her-

baceous or coriaceous, never membranous. Corolla funnel-shaped, small or medium-

sized, blue or white, 5-lobed, with the midpetaline bands hairy outside. Stamens 5,

included or slightly exerted, filaments glandular at the base or glabrous, anthers oblong,

cordate or sagittate at the base, pollen smooth. Ovary 2-celled, 4-ovuled; style bifid

or 2 free styles, often unequal in length; stigmas 2, globose or peltate, rarely 2-partite

disc small or none. Capsule 2-, 4— or 8-valved, 2-loculated, 4- or, by abortion, less-

seeded. Seeds glabrous or pilose.

Number of species about 40, widely spread in the tropics.

665

This genus is very similar in most floral characters to Seddera Hochst., but it can

usually be easily distinguished by its habit and by the stamens (often with 2 basal small

teeth in Seddera, often glandular in Bonamia). The S.W. African species, here referred

to Bonamia, was originally described by Hallier as a Seddera but the resemblance of

the species under discussion to Bonamia poranoides Hall. f. (= Porana densiflora

Hall, f.), the pedunculate inflorescence, the spathulate-suborbicular, imbricate, more

or less rounded calyx-lobes, the stamens without teeth at the base and its general habit

(straggling or climbing shrub) place it in Bonamia rather than in Seddera.

N. E. Brown described a genus Metaporana, based on this S.W. African species

of Bonamia and on Porana densiflora Hall, f., but he overlooked the fact that Hallier

had already transferred Porana densiflora to Bonamia as B. poranoides Hall. f. [in Bull.

Herb. Boiss. 5 (1897), p. 1007]; the combination B. densiflora could not be applied

to this species because of B. densiflora (Baker) Hall, f., op. cit., p. 999, based on

Breweria densiflora Baker from Madagascar. This transfer was also overlooked in

FI. Trop. Afr., 4, 2 (1905), in which the species was still mentioned under Porana

densiflora Hall. f. N. E. Brown only redescribed the genus Bonamia and Metaporana

N.E.Br. is clearly a synonym of Bonamia Thouars. It is, in my opinion, doubtful

if the genus Porana occurs in Africa at all.

B. schizantha {Hall. /.) A. Meeuse , comb. nov. — Seddera schizantha Hall. f. in Bull.

Herb. Boiss 6 (1898), p. 532. Metaporana angolensis N.E.Br. in Kew Bull. 1914,

p. 169.

Type: Newton (1883) no. 18 in Herb. Zurich (fide Hallier).

Much branched, erect to climbing shrub or undershrub. Branches up to at least

75 cm. long, terete, adpressed-pubescent; ultimate branchlets slender, terete, divaricate

and subflexuose. Leaves very shortly petioled; petiole 1-5-3 mm. long; blade ovate-

elliptic to oblong, obtuse, mucronate, rounded, subtruncate or subcordate at the base,

1-4 cm. long and 4-15 mm. wide, entire, penninerved, with sparse adpressed hairs

or almost glabrous; lateral nerves 3 or 4 on either side, ascending, prominent below;

reticulate coarse nervation distinct below, less conspicuous above. Peduncles in the

leaf-axils, solitary or occasionally binate, slender, erecto-patent, subglabrous or

adpressed-puberulous, 3-18 mm. long, bearing a dense dichasium or a single unilateral

cyme, often collected into terminal panicles at the ends of the branches. Bracts and

bracteo/es minute, lanceolate, adpressed-puberulous, 1-1 -5 mm. long. Pedicels slender,

filiform, 2-3 mm. long. Sepals broadly ovate-spathulate to suborbicular, obtusely

rotundate, subcoriaceous, 2 mm. long and 1-5-2 mm. wide, glabrous or with sparse

adpressed pubescence, the margins membranaceous and partly ciliate. Corolla white;

the tube cylindrical, 2 mm. long, the limb deeply 5-lobed, cut down to the calyx; the

lobes 3 mm. long, 2-2-5 mm. wide, elliptic, obtuse, patent, outside with brownish

strigose hairs in the midpetaline zones. Filaments glabrous, dilated but not toothed

at the base, 2-3 mm. long, anthers nearly 1 mm. long. Ovary ovoid, glabrous except

a few hairs at the apex; disc very low; styles 2, free to the base, equal, filiform, 4-4-5

mm. long. Capsule subglobose with conical apex, 3-5-4 mm. long and wide, exerted

from the calyx, glabrous. Seeds small, glabrous, minutely punctate, trigonous, convex

at the back, flattened at the sides, brown or black.

Angola (Southern part) and S.W. Africa.

S.W. Africa. —Fransfontein: Liebenberg 4948 (PRE); Klein Ameib: Dinter

7070 (PRE, BOL); between Okahandja and Swakopmund, Dorstrivier: Dinter 196

(SAM, PRE). Also collected by Mr. de Winter in the Omaruru district (PRE);

apparently fairly common in the Kaokoveld {Story, de Winter in PRE).

The plants agree in every respect with Hallier’s excellent description, and Liebenberg

4948 was compared by Miss Kies at Kew with Pearson 2873 and 2391, two of the original

666

numbers cited by N. E. Brown, of which it proved to be a very good match. Dinter

7070 was compared with a specimen named Seddera schizantha (from Damaraland,

collector unknown, prob. Dinter ) in herb. Kew by Dr. Schweikerdt in 1938 and was

reported to match it.

7. CONVOLVULUS

L. [Sp. PI. Ed. 1 (1753), p. 153]; Gen. PI. Ed. 5 (1754), p. 76; Choisy in DC.,

Prodr. 9 (1845), p. 399; Benth. et Hook, f., Gen. Plant. 2 (1876), p. 874; Peter in

Engl. u. Prantl, Natiirl. Pflanzenfam. Ed. 1, 4 -3a (1891), p. 33; Hall. f. in Engl. Bot.

Jb. 16 (1893), p. 579; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 70; Baker

& Rendle in Dyer, FI. Trop. Afr. 4,2 (1905), p. 88; Phillips, Gen. S. Afr. Flow PI.

Ed. 2 (1951), p. 622.

Type species: Convolvulus arvensis L., Sp. PI., Ed. 1 (1753), p. 153 (priority of

place, as Linnaeus mentioned a number of species).

Perennial, sometimes annual, herbaceous to suffruticose, prostrate or climbing,

rarely erect, shrubby. Leaves entire or lobed, often cordate, hastate or sagittate at

the base, petiolate or sometimes sessile. Flowers regular, axillary, solitary or in few-

flowered, sometimes subumbellate, cymes, small to medium-sized. Sepals 5, equal

to unequal, persistent. Corolla funnel-shaped, in Southern Africa white or pale mauve

to pink; the limb shallowly 5-lobed; midpetaline areas often hairy towards the apices.

Stamens 5, usually unequal, filaments filiform or somewhat linear, often dilated at the

base; pollen smooth, ellipsoid. Disc annular or cup-shaped. Ovary 2-celled, 4-ovuled;

style filiform; stigmas 2, linear or filiform, rarely short, oblong. Capsule 2-celled,

dehiscent, 4-valved. Seeds 4, sometimes less by abortion, black or brown, usually

glabrous.

In the temperate and subtropical regions of both hemispheres, but much rarer

in the tropics. Species ca. 250, of which 15 occur in S. Africa.

The mainly subtropical genus Convolvulus seems to be more or less vicarious

with Ipomoea , the latter being mostly confined to the tropical regions.

It is difficult to divide Convolvulus into sections. An attempt was made by Peter

(l.c.), but his system seems to be rather artificial; the leaf-shape in particular varies

in many species.

There is considerable variation in leaf-shape in several South African represen-

tatives: from cordate or sagittate-hastate to narrowly hastate or sagittate with sometimes

bifid basal lobes, and also from undivided, cordate or sagittate, to palmately or

pinnately lobed; the palmately lobed leaves show a tendency towards reduction of

the lateral lobes so that the leaves become linear with hastate or sagittate base, or

auricled. The various shapes are sometimes found on one specimen [a typical example

is C. dregeanus, this is also occasionally found in specimens of C. boedeckerianus,

C. “ ornatus ”, (= C. ocellatus), C. capensis, and others], sometimes in a series of

specimens covering the whole range of variation. The extremes are often strikingly

different and many have been described as separate species, so that several pairs or

trios of conspecific, extreme forms appear under different names; for example: C.

capensis Burm. f. (dissected leaves) — C. falkia Jacq. non Thunb. (leaves not dissected) — -

C. inconspicuus Hall. f. (not dissected); C. capensis Burm. f. (leaves ovate to pal-

matifid) — C. filiformis Thunb. non Desr. (= Merremia bowieana Rendle; leaves linear

with basal auricles); C. natalensis Bernh. (leaves more or less cordate) — C. transvaalensis

Schltr. and C. bullerianus Rendle (leaves becoming narrowly hastate to linear with

auricled base); C. dregeanus Choisy [leaves varying from suborbicular-ovate (lower

ones) to palmately 5-fid (upper ones) on one specimen] — C. liniformis Rendle (leaves

linear, minutely auricled at the base, occasionally upper ones becoming palmately

5-fid); C. ornatus Engl, (leaves elongate-hastate to lanceolate and auricled, upper

667

ones often palmately 5-fid) — C. dinteri Pilger (leaves linear or lanceolate to subhastate,

basal auricles sometimes bifid) — C. ocellatus Hook, (leaves linear with small auricles

or no auricles).

Once this tendency was recognised, it was fairly easy to group the South African

species of Convolvulus into a dinstinct number of taxa by using distinguishing characters

based on the flowers, peduncles, pedicels, etc. Almost needless to say, the number

of species is less than that in the Flora Capensis (15 against 21). One species is not

infrequently cultivated as an ornamental (C. mauritanicus Boiss. from N. Africa with

blue flowers), but it is not included in this treatment. Convolvulus arvensis is an

introduced weed which has become very common and wide-spread. As regards the

key to the species of Convolvulus , the great variation in characters such as leaf-shape,

lengths of peduncle and degree of pubescence, makes it very difficult to distinguish

every not very typical specimen of a given species easily. For instance, Convolvulus

ulosepalus in very depauperate specimens with 1 -flowered penduncles resembles

C. boedeckerianus ; the polymorphous C. sagittatus sometimes resembles small-flowered

specimens of C. bidentatus; C. bidentatus is sometimes difficult to separate from forms

of C. capensis; C. thumbergii seems to grade into narrow-leaved forms of C. natalensis.

Without the aid of sufficient authenticated herbarium specimens the key (and every

other key for that matter) is not quite adequate to name every more or less atypical

or depauperate specimen of any given species.

Corolla 4-5 times as long as the calyx, usually quite glabrous; sepals obtuse or

rounded at the apex, usually glabrous but often minutely ciliate; peduncles

mostly longer than the leaves, 1- or sometimes 2-3-flowered; leaves

hastate-sagittate, entire (introduced weed)

Corolla less than 4-5 times as long as the calyx:

Flowers almost sessile (peduncles plus pedicels not exceeding 5 mm. in

fruit), in few-flowered clusters or occasionally solitary; calyx 8-13

mm. long, accrescent, the outer sepals ultimately subcordate, crisped

along the edges; corolla scarcely longer than the calyx (S.W. Africa)

Flowers distinctly pedunculate or at least pedicellate; peduncles plus pedicels

usually over 5 mm. long, often much longer; corolla 1-5-3 X the

length of the calyx:

All vegetative parts and flowers quite glabrous; peduncles 1 -flowered;

leaves varying in shape from, linear, entire or with basal auricles,

to palmatisect, often on one specimen, but always small, under

20 mm. long, rarely attaining 40 mm. in length

Plants hairy on either stems, leaves or inflorescences, very rarely plants

quite glabrous, but, if so, leaves different in shape or well over

20 mm. long and/or peduncles few-flowered:

Leaves (at least the upper ones) linear, hastate at the base with

patent, often deeply bifid basal lobes; peduncles usually

2-flowered, sometimes 1-, rarely up to 6-flowered, usually

rather long to long (3-14 cm.); sepals quite glabrous or

occasionally obscurely pubescent, coriaceous with mem-

branous edges, broad, rounded or obtuse, sometimes muc-

ronate; corolla usually about 20 mm. long (coastal districts

from Bredasdorp to East London, Uitenhage, Albany)...

Leaves various, but, if linear and hastate or auricled at the base,

either whole plant densely tomentose or basal lobes entire

to faintly bilobed or calyx distinctly hairy ; sepals often acute

or without membranous edges:

Peduncles 0 or rarely up to ± 2 mm. long (bracteoles placed

in the leaf axils or nearly so); flowers solitary; corolla

up to ± 14 mm. long, its lobes usually distinctly acute:

Sepals densely tomentose, obtuse or subacute, often

somewhat abruptly narrowed above the middle, 6-8

mm. long; corolla 12-14 mm. long; leaves often

thick with revolute edges, densely tomentose,

usually palmately 5-fid with linear lobes

15. C. arvensis.

1. C. argillicola.

2. C. dregeamis.

10. C. bidentatus.

3. C. ocellatus.

668

Sepals more or less laxly covered with silky, shiny,

golden-brown hairs, more or less ovate; corolla

± 9 mm. long; leaves flat and rather thin, pinnately

lobed with the central lobe the largest, usually

thinly hairy (also depauperated specimens of

C. ulosepalus and C. sagittatus)

Peduncles usually distinct although occasionally very short

(if so, flowers 14-16 mm. long or longer and/or lobes

of corolla rounded or obtuse):

Leaves up to ± 25 mm. long, with 5-9 narrow, linear or

filiform, obtuse, palmately arranged lobes; plant

usually finely, densely and very shortly white- or

silvery-, more rarely fawnish-tomentose ; calyx

6-8 mm. long, corolla 14-16 mm. long with more

or less rounded lobes ; peduncles always 1 -flowered,

short, up to 20 mm. long

Leaves different in shape or well over 25 mm. long or

plant not very shortly tomentose; peduncles often

more than 20 mm. long:

Peduncles usually 2-6-flowered, rarely 1-flowered:

Vegetative parts and peduncles densely and

shortly, more or less adpressed-hairy to

sericeo -tomentose, often canescent, some-

times fulvous; flowers up to ± 14 mm.

long; leaves oblong to linear-oblong,

pinnately and more or less irregularly

toothed to pinnatilobed, often with

toothed or dissected lateral lobes at the

base

Vegetative parts not so densely hairy or plants

farinose; corolla 8-15 mm. long:

Corolla ± 8 mm. long; sepals very

unequal, the inner ones rotun-

date, abruptly apiculate; stamens

not papillose at the base, leaves

usually obtuse, mucronate linear

to linear-oblong with hastate

base and usually bifid basal

lobes, margin usually entire,

more rarely leaves wider, sub-

sagittate and/or lobed or dis-

sected at the base or margin

irregularly dentate or crenate. . 7

Corolla ± 10 mm. long; sepals sub-

equal; leaves linear to linear-

oblong with hastate or sagittate

base, entire or basal lobes dis-

sected, usually obtuse, mucro-

nate, margin entire or subentire

Corolla 12-14 mm. long; sepals not so

unequal as in C. ulosepalus ,

stamens papillose at the base;

leaves in typical specimens cor-

date-deltoid or more or less

sagittate, not dissected (basal

lobes often more or less bilobed),

more rarely basal lobes dissected,

apex usually acute to acuminate,

edge usually crenate

4. C. boedeckeri-

anus.

5. C. multifidus.

6. C. aschersonii.

. C. ulosepalus.

8. C. sagittatus.

9. C. farinosus.

669

Corolla 15-20 mm. long or longer:

Leaves usually not dissected, often

cordate or cordate-oblong,

crenate, occasionally oblong to

linear-sagittate; stems prostrate;

corolla white or cream to

greenish-white; sepals often

crisped along the edges; bracts

sometimes more than two

together on one peduncle (not in

S.W. Cape districts) 12.

Leaves either dissected or, if entire,

found in S.W. Cape districts;

corolla often pale pink or rose-

coloured (especially after drying) :

Leaves pinnately nerved, linear to

linear-oblong with hastate

and toothed to dissected

base; the margin of the

central lobe often sinuous,

or more or less irregularly

pinnatilobed to deeply pin-

natisect; not in S.W. Cape

districts 13.

Leaves palmately nerved, palmati-

lobed or palmatifid, occas-

ionally linear with palmately

arranged smaller basal lobes

or auricled; if entire, more

or less oblong-cordate,

ovate-cordate to subreni-

form; S.W. Cape districts 14.

Peduncles usually 1 (—2) flowered:

Corolla usually under 15 mm. long, often smaller,

rarely up to ± 20 mm. long, but if more

than 15 mm. long, either leaves linear-

sagittate or linear, or plants from S.W.

Africa and Rhodesia:

Erect or prostrate, suffruticose, rarely

climbing ; whole plant densely tomen-

tose or sericeous up to the sepals,

leaves often with crenate and/or

reflexed margins; peduncles up to

± 10 mm. long, often shorter;

pedicels up to ± 15 mm. long (much

shorter if peduncles are rather long) 3.

Prostrate to climbing, herbaceous, never so

densely tomentose but usually thinly

hairy; leaves not with reflexed

margins; peduncles often more than

10 mm. long; pedicels usually short 8.

Corolla usually over 15 mm., often over 20 mm.

long : leaves usually not linear, often more

or less cordate or dissected, not found in

S.W. Africa or Rhodesia:

Whole plant usually densely hairy to tomen-

tose or shortly villous ; leaves ovate-

cordate-deltoid or oblong-cordate to

deltoid, entire or finely crenate;

calyx villous; sepals sharply and

abruptly acuminate-apiculate ; ± 20

mm. long (only known from E. Cape

Province) 11.

C. nmtalensis.

C. thunbergii.

C. cape mis.

C. oceliatus.

C. sagittatus.

C. gaipinii.

1332796-3

670

Plant either not densely villous-hairy, or, if

so, sepals usually not abruptly

acuminate-apiculate and/or leaves

different in shape:

Leaves usually not dissected, often

cordate or cordate-oblong,

crenate, occasionally oblong to

linear-sagittate ; stems prostrate ;

corolla white or cream to

greenish-white ; sepals often

crisped along the edges; bracts

sometimes more than two

together on one peduncle (not in

S.W. Cape districts) 12. C. natalensis.

Leaves either dissected or, if entire,

found in S.W. Cape districts;

corolla often pale pink or rose-

coloured, especially after drying :

Leaves pinnately nerved, linear to

linear-oblong with hastate

and toothed to dissected

basal lobes; the margin of

the central lobe often sin-

uous; or leaves irregularly

pinnatilobed to pinnatisect ;

not in S.W. Cape districts 13. C. thunbergii.

Leaves palmately nerved, palmati-

lobed or palmatifid, occa-

sionally linear with or with-

out palmately arranged small

basal lobes, or auricled ; if en-

tire, oblong-cordate, ovate-

cordate to subreniform. ... 14. C. capensis.

1. C. argillicola Pilger in Engl. Bot. Jb. 48 (1912), p. 348; Dinter in Fedde Repert.

16 (1919), p. 240.

Type: Pilger mentioned two specimens, a flowering one ( Dinter 1892) and a

fruiting one ( Dinter 2153), which represent the type for the flowers and for the fruits,

respectively (herb. Berlin, now destroyed); the isosyntypes in SAM I now consider

to be the neosyntypes.

Stems several from a perennial taproot, suffruticose at the base, prostrate, like

the young parts, leaves, petioles and calyces densely covered with stiff rusty-brown

or whitish hairs, ultimately glabrescent, rather firm and stout, terete and somewhat

longitudinally striate, up to 70 cm. long. Leaves ovate or oblong in outline, obtuse

or acute, usually with minute apical mucro, truncate to broadly and shallowly cordate

at the base with the blade somewhat cuneately decurrent on the petiole, irregularly

crenate or pinnatilobed to pinnatisect; the lobes usually obtuse often again crenate

to somewhat lobed; length of blade 2-4 (-5) cm., width 1-2-5 (-3) cm., petioles

slender, 0-5-2 (-3) cm. long. Flowers solitary or few together in the leaf-axils, nearly

sessile; bracts subulate or lanceolate-subulate, 7-10 mm. long, hairy. Sepals unequal,

the outer ones herbaceous, long-acuminate from an ovate base, 12-13 mm. long, acute,

hirsute, the 2 inner ones shorter and narrower, indurate, caudate-acuminate, about

10 mm. long, glabrous except at the hirsute tips. Corolla white, about as long as the

calyx, widely funnel-shaped to rotate, with a few stiff hairs towards the tips of the

lobes on the midpetaline areas. Ovary glabrous. Capsule subglobose, about 9 mm.

in diam. Seeds glabrous, verruculose or rugose, black, 5-6 mm. long.

S.W. Africa. — Okahandja; Dinter 1006 (SAM, PRE). Windhoek, nr. Seeis:

Codd 5811 (PRE); Liechtenstein: Dinter 4284 (SAM); Northern part of Auas Mts. :

Dinter 1892 (SAM, neotype for the fruits). Haris, Kuisib River bed: Pearson 9562

671

(SAM, BOL). Ukos: Pearson 9427 (BOL). Quartel nr. Rehoboth: Dinter 1892

(SAM, neotype for the flowers).

2. C. dregeanus Choisy in DC., Prodr. 9 (1845), p. 411; Hall f. in Engl. Bot. Jb. 18

(1893), p. 105; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 74. C. liniformis

Rendle in Jl. Bot. 39 (1901), p. 61; Baker & Wright, op. cit., p. 71.

Type: Drege 7828 from Richmond. Winterveld (Cape Province), in Herb.

Geneva.

Perennial, glabrous in all its parts, forming several annual stems from a common

thick taproot. Stems trailing or ascending to suberect; sometimes branched from

the base, very slender, up to about 30 cm. long; but often much shorter. Leaves

small, variable in shape, sometimes lanceolate to linear, often with 2 minute basal

auricles (and often close-set), acute, nearly sessile, 4-15 (-40) mm. long, but usually

the lower ones oblong to ovate-suborbicular, entire, subcordate or more or less lobed,

but gradually changing upwards into more deeply dissected leaves and upper ones deeply

palmately 3-5-fid with linear lobes of which the middle one is the longest and some-

times somewhat pinnatilobed, up to 20 (-25) mm. long; petioles of upper leaves up

to 10 mni. long, but usually petioles much shorter. Peduncles 5-30 (-45) mm. long,

1-flowered, bracteoles lanceolate-linear to somewhat spathulate, about 2 mm. long,

usually green and somewhat foliaceous; pedicels 1-5 mm. long or occasionally peduncles

reduced and pedicels up to about 10 mm. long. Calyx 4-7 mm. long; sepals much

imbricate, obovate or the outer oblong, obtuse. Corolla 12-20 mm. long, pinkish-

white or white, glabrous. Capsule globose, glabrous, about 6 mm. in diam. Seeds

scabrid to muriculate, about 4 mm. long.

Endemic.

Cape Province. — Barkly West: Lewis h. no. 54140 (SAM); Esterhuysen 948

(BOL). Kimberley: Wilman h. no. 2962 (KMG) = prob. H. no. 25030 (BOL); Oliver

h. no. 42282 (SAM); “ Griqualand West, near the Vaal River ”: Nelson 212 (PRE).

Colesberg, Klein Tafelberg: Burke 284 (SAM). Hanover: Sim in herb. Galpin 6255

(PRE). Richmond: Bolus 13819 (BOL), id. Winterveld: Drege 7828 (L, isotype!).

Middelburg: Hutchinson 3100 (BOL); Gill 85 (BOL). Graaff-Reinet: Bolus 1825

(BOL, GRA). Tarka: Acocks 16284 (PRE).

Orange Free State. — Kroonstad: Pont 161, 579 (PRE). Bloemfontein: Gemme/f

4952, 4976, 5640 (PRE). Fauresmith: Verdoorn 899 (PRF).

Transvaal. — Wolmaransstad: Sutton 418 (PRE); Liebenberg 2983 (PRE).

“ Mooi Rivier ” (Klerksdorp/Potchefstroom): Burke 283 (PRE, SAM, isotypes of

C. liniformis Rendle). Potchefstroom: Louw 1690 (PRE). Johannesburg: Galpin

6255 (PRE).

The specimens described as C. liniformis Rendle represent prostrate shoots and

they are linked up with “ typical ” C. dregeanus by intermediate forms. The specimen

Nelson 212 in PRE, for instance agrees very well with C. liniformis Rendle, but the

duplicate specimen of Nelson 212 at Kew was referred to C. dregeanus in Flora Capensis.

Several specimens cited above show stems with entire, linear leaves in addition to

stems showing the characteristic transition from entire, more or less ovate or oblong

leaves to the dissected upper leaves, and some herbarium sheets contain several specimens

showing plants of either type, evidently collected together in one locality. There is

no difference in floral characters, capsules or seeds.

C. dregeanus is quite distinct and only C. boedeckerianus shows a superficial

resemblance. However, C. dregeanus is the only South African species that is always

quite glabrous in all its vegetative parts, and has a glabrous calyx and corolla. i:

672

3. C. ocellatus Hook, f in Bot. Mag. 70 (1844), t. 4065; Choisy in DC., Prodr. 9

(1845), p. 404; Hall. f. in Engl. Bot. Jb. 18 (1894), p. 102; Baker & Wright in

Dyer, FI. Cap. 4, 2 (1904), p. 71.

This plant is rather polymorphous and two varieties can be distinguished: —

Leaves usually undissected, without basal auricles, not or rarely bullate, usually

acute; peduncles developed though sometimes very short; sepals

usually acute var. ocellatus.

Leaves often palmately 5-fid or with hastate basal auricles, more rarely entire; often

bullate and obtuse; peduncles 0 or occasionally developed, up to

2(— 5) mm. long; sepals often suddenly attenuate from the middle

into an oblong, obtuse apex, rarely acute var. ornatus.

3a. C. ocellatus Hook. f. var. ocellatus. This form includes Convolvulus randii Rendle

in J. Bot. 40 (1902), p. 189; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 95.

Type of var.: The type is evidently the plate in Bot. Mag. (t. 4065). Hooker

stated: “ . . . raised from seeds in the greenhouse of the Earl of Derby, at Knows-

ley . . . ”, most probably from seeds collected by Burke, because Hooker mentioned:

“ discovered by Mr. Burke at Macalisberg ”, and it is very unlikely that the seed could

have been obtained from any other source. Consequently, Burke 1 19 from the Magalies-

berg, Transvaal, is equivalent to a type specimen, though technically only a topotype.

Erect to decumbent perennial forming annual suffruticose stems from a woody

taproot. Stems several from the base, densely brownish or greyish tomentose, up to

about 60 cm. high. Leaves linear to linear-oblong, entire usually acute, nearly sessile,

9-20 (-30) mm. long and 1-2-5 (-5) mm. wide; thick, but not usually bullate as in

the var. ornatus with the stout midrib and the lateral nerves impressed above and

prominent below, covered (as are peduncles, bracts, pedicels, calyx and midpetaline

zones) with the same brownish or greyish tomentum as the stems. Peduncles ascending,

1-flowered, very rarely some 2-flowered, shorter than the leaves and up to about 10 mm.

long; bracteoles small, linear or subulate, about 3 mm. long; pedicels usually longer

than the peduncles and up to about 15 mm. long. Calyx about 6 (-10) mm. long;

sepals oblong or ovate, acute or somewhat acuminate, rarely obtuse, the outer ones

completely covered with the tomentum on the outside, inner ones with a median hairy

zone. Corolla white with a dark reddish-purple “ eye ”, widely funnel-shaped, about

1 5 (-20) mm. long and the limb 20 (-25) mm. in diam. Capsule ovoid-conical, shortly

apiculate, hairy at the apex, 5-6 mm. long and about 5 mm. in diam. Seeds glabrous.

Transvaal. — Prob. Potchefstroom (“Hills near Mooi River”): Zeyher 1322

(BOL). “ Magaliesberg ” : Burke 119 (PRE). Pretoria, Irene: Leendertz 669 (PRE),

Pole-Evans s.n. (PRE). Quagga’s Poort: Verdoorn s.n. (PRE).

Southera Rhodesia. — Gwelo: Rand 274 (Photograph of type of C. randii ex

BM in PRE); Walters h. no. 2433, in herb. Eyles sub no. 3469 (SRGH). Lalapansi:

Ingle 1 (SRGH).

The specimen Burke 119 is somewhat different in that the leaves are bullate and

the sepals are obtuse. In Verdoorn s.n. the sepals are obtuse or acute. These two

specimens form a transition to the var. ornatus, which cannot be very sharply distin-

guished from the var. ocellatus.

Convolvulus randii is somewhat more robust, with longer and wider leaves than the

Transvaal specimens, and has occasionally 2-flowered peduncles, but it cannot be more

than a somewhat luxuriant form of C. ocellatus.

673

The types of C. ocellatus (i.e. the plate in Bot. Mag. t. 4065), C. ornatus ( Marloth

716) and C. dinteri ( Dinter 812) are very different in appearance. C. ocellatus (and

C. randii) are erect, suffruticose with linear leaves, usually without a trace of basal

auricles, usually distinct peduncles, acute sepals and rather large flowers which (always ?)

have a purplish-red centre. C. ornatus is usually prostrate to ascending, with linear-

oblong or hastate to palmately 5-fid leaves with revolute edges. C. dinteri is inter-

mediate in several respects but sometimes distinctly climbing. The best solution is

to treat these very closely related forms as two varieties of C. ocellatus , including

C. dinteri in the var. ornatus.

The extreme forms are so different that they resemble several other species. Typical

“ C. ornatus ” sometimes resembles C. multifidus and C. boedeckerianus; typical

“ C. dinteri ” resembles certain forms of C. sagittatus (but is much more hairy). For

the difference between “ C. ornatus ”, C. multifidus and C. boedeckerianus , see under

C. multifidus.

3b. C. ocellatus Hook. f. var. ornatus (Engl.) A. Meeuse, stat. nov. C. ornatus Engl.

in Engl. Bot. Jb. 10 (1888), p. 247; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904),

p. 76. C. multifidus Hall. f. in Engl. Bot. Jb. 18 (1893), p. 102 non Thunb. (1794).

C. dinteri Pilger in Engl. Bot. Jb. 45 (1910), p. 219; Dinter in Feddes Repert.

16 (1919), p. 240.

Type: Marloth 716 (originally in B, now destroyed; duplicate-neotype in PRE).

Perennial, covered with a sericeo-tomentose, usual velvety-brown, fawn or drab,

more rarely white pubescence, producing several to many annual stems from a woody

rootstock. Stems prostrate, ascending to suberect, often sinuous and occasionally

climbing at the ends; the suberect ones short, prostrate and climbing ones up to about

40 cm. long, terete, often suffruticose. Leaves varying from linear-oblong with hastate

or sagittate base to palmately 5-fid or linear with bifid basal auricles, very rarely linear

or oblong with narrow base, usually rather thick with distinctly revolute edges and the

nerves much impressed above, prominent below, hence leaves often appearing sub-

bullate or plicate; linear-oblong ones 10-25 mm. long and 2-4 (-8) mm. wide, shortly

and abruptly acuminate or obtuse, basal lobes often bifid, often gradually changing

on the same plant to the palmately 5-fid leaves of which the central lobe is always the

longest; the lobes entire to crenate or more or less wavy at the margins because the

nerves are strongly impressed; lobes 1-4 mm. wide, obtuse or acute, the basal ones

often narrower than the central one, with more strongly revolute edges; occasionally

leaves flat, up to 15 mm. wide at the base and with somewhat pinnately lobed or toothed

lobes; petioles 0-3- (-5) mm. long, those of digitately compound leaves generally

longer but rarely attaining 10 mm. Peduncles 0 or rarely up to 2 (-5) mm. long;

bracteoles subulate or linear, up to about 5 mm. long but often shorter and minute;

pedicels 3-10 mm. long. Calyx 6-8 mm. long, densely tomentose to sublanate; sepals

ovate, ovate-lanceolate or elliptic to oblong, usually from an elliptic or ovate basal

part abruptly narrowing into a linear-oblong, obtuse apical portion, sometimes gradually

narrowed into an obtuse, acute or somewhat acuminate or minutely awned apex.

Corolla 12-14 mm. long, white, pink or pale mauve, the limb 12-14 mm. in diam.,

5-angled with usually acute points, densely hairy on the midpetaline areas. Capsule

globose-ovoid, 5-7 mm. in diam., 6-8 mm. long, apiculate, glabrous when mature.

Seeds dark brown, glabrous, smooth.

Transvaal. — Lichtenburg: Sutton 416 (PRE). Bloemhof/Christiana : Burn- Daw

11380 (PRE, NBG), 11858, 14347 (PRE).

Cape Province. — Vryburg: Rogers H. no. 26975 (SAM), Burtt-Davy 1 1 127

(PRE), 14674 (PRE, SAM); Lear (or Phear) H. no. 25032 (BOL); Henrici 28, 53

(PRE). Barkly West: Wilman h. no. 2995 (KMG); Lawson h. no. 25033 (BOL),

Acocks 1461, 17853 (PRE). Kimberley: Marloth 716 (PRE, isotype of C. ornatus

674

Engl.); Moran 77 (GRA)=prob. H. no. 476 (KMG)=H. no. 52259 (SAM); Acocks

■and Hafstrom H780 (PRE) ; Esterhuvsen 1 288 (BOL) ; Henderson 86 (KMG), Warrenton :

Adams 204 (GRA). Kuruman: Burchell 2412 (PRE); Silk 215 (BOL, SAM). Hay:

Aucamp H. n. 477 (KMG); Wilman H. no. 1321, 3091 (KMG); Louw s.n. (PRE).

Prieska: Acocks and Hafstrom H. 1120 (PRE).

Orange Frfe State. — Bloemfontein: Grist H. no. 7286 (PRE); Marais 195

(PRE).

S.W. Africa. — Aukas, Kraaifontein: Dinter 812 (SAM, PRE, isotypes of C.

dinteri Pilger) — Grootfontein : Schoenfelder S 403, S 404 (PRE).

4. C. boedeckerianus Peter in Engl, and Prantl, Natiirl. Pfl. fam. Ed. 1, 4-3A (1891),

p. 36, nomen subnudum; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 102, descr. latina;

Baker & Wright in Dyer, FI. Cap. 4,2 (1904), p. 76.

Type: Peter’s original description is very short and he did not mention any

specimens at all. A good description was only given by Hallier, who quoted several

specimens. On account of its specific name, the specimen leg. Boedecker from the

Orange Free State, cited by Hallier, must be considered to be the type (in herb.

Gottingen, according to Hallier). I have not seen the type specimen, but have examined

specimens which had been compared with the specimens in the Kew herbarium

collected b> Shaw and one of which was quoted by Hallier.

Perennial, forming several woody stems from the crown of a long woody taproot,

covered on stems, young parts, leaves, petioles, pedicels, bracteoles and calyx with

adpressed, usually golden-brown or fulvous, more rarely silvery-white, hairs. Stems

slender, prostrate or occasionally twining at the ends, rarely suberect, terete, usually

finely silky, glabrescent or sometimes nearly glabrous, up to about 60 cm. long (suberect

ones up to 20 cm.). Leaves usually pinnately to subpalmately 5-lobed, the lower ones

often ovate or oblong, coarsely dentate to pinnatilobed, gradually changing upwards

into the divided, more typical leaves, 10-35 mm. long, only in exceptional cases up

to 60 mm. long; middle lobe of the blade usually distinctly the largest, often irregularly

toothed or pinnatilobed to pinnatisect, rarely quite entire; basal lobes (auricles)

usually bifid; lateral lobes usually linear, entire or slightly toothed, sometimes pinnati-

lobed; pubescence usually rather sparse to glabrescent, rarely almost completely

disappearing in older ones; petioles 1-5 (-8) mm. long. Flowers solitary; peduncles

usually wanting, rarely developed, if present, very short; bracteoles subulate, minute;

pedicels 2-6 (-10) mm. long, somewhat elongating and glabrescent in fruit. Calyx

4-6 mm. long, tomentose to laxly hairy but never quite glabrous; sepals subequal,

ovate, acute, 2-3 mm. wide. Corolla white, pale pink or pale pinkish mauve, 7-10 mm.

long, the limb 9-12 mm. in diam., 5-angledwith acute triangular lobes; midpetaline areas

silky-pubescent outside. Capsule glabrous, globose or ovoid-globose, apiculate,

straw-coloured, 5-7 mm. long and in diam. Seeds dark brown, when ripe distinctly

verrucose-rugose or somewhat muriculate, about 4 mm. long.

Endemic but may extend into Bechuanaland Protectorate.

Cape Province. — Hay: Wilman s.n. (GRA)=prob. H. no. 2241 (KMG); Acocks

and Hafstrom H. 1044 (PRE); Hafstrom H 1042 (PRE). Barkly West: Bowker 593

(GRA); Pagan H. no. 473 p.p. (KMG, mixed with C. multifidus)', betw. Barkly West

and Kimberley: Bolus 6837 (BOL). Kimberley: Marloth 880 (PRE); Tvson s.n.

(SAM); Moran 47, 60 (GRA) H. no. 480 (KMG), H. no. 15895 (BOL); Elliot H.

no. 472 p.p. (KMG, mixed with C. multifidus)', Wilman H. no. 15843 (BOL); Lewis

H. no. 54142 (SAM); Brueckner 600 (PRE). Herbert: Wilman h. no. 475 (KMG).

“ Kaap Plateau ”, Griqualand West: Hafstrom H 1239 (PRE). Middelburg: Verdoorn

1530 (PRE). Queenstown: Galpin 2343 (PRE).

Transvaal. — Bloemhof: Tardrew s.n. (PRE, SAM); Louw 1976 (PRE).

675

Orange Free State— Kroonstad : Pont 240 (PRE). Winburg: Prosser 1529 (J).

Bloemfontein: Flanagan 2112 (PRE, GRA); Bolus 8217 (BOL); Marais 31 (PRE);

Potgieter 43 B (PRE). Fauresmith: Henrici s.n., 1875, 1877, 2558, 2662 (PRE); Smith

3926, 4606 (PRE); Marais 493, 506 (PRE); Verdoorn 978 (PRE). “nr. Caledon

River”: Burke 189 (PRE, SAM); “between Bethulie, O.F.S. and Aliwal North,

C.P.”: Thorne h. no. 51870 (SAM).

C. boedeckerianus is usually rendered quite distinct by its hairy calyx, small corolla,

obsolete peduncles and solitary flowers. The hairy calyx distinguishes it from C.

dregeanus which is very similar in habit but quite glabrous; the absence of peduncles

and the usually distinctly dissected or pinnatilobed leaves distinguish it from depauper-

ate 1 -flowered specimens of C. ulosepalus (which has distinct penduncles and usually

linear, auricled leaves). It is closely related to C. multifidus and for the differences

between C. boedeckerianus, C. ocellatus (C. ornatus) and C. multifidus, see under the

latter.

The specimens of Burchell 1839 I have seen are referable to C. multifidus, in my

opinion, and not to C. boedeckerianus as was done in FI. Cap. This gathering has

been quoted by Hallier under “ C. thunbergii R. et S. ” (evidently on account of Choisy’s

treatment in Prodr. 9; however, Choisy quotes “Burchell 1836 ” under this name).

For the various interpretations of “ C. thunbergii ”, see under the latter.

5. C. multifidus Thunb., Prodr. PI. Cap. (1794), p. 35, and in FI. Cap Ed. Schult. (1820),

p. 170; Choisy in DC., Prodr. 9 (1845), p. 410 excl. spec. Burchell 2412; Baker

and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 76. C. thunbergii Flail, f. in Engl.

Bot. Jb. 18 (1893), p. 102, non R. et. S.

Type: A specimen collected by Thunberg nr. Loeri River, Uitenhage, in herb-

Thunberg in Uppsala (photograph of type in PRE).

Perennial forming annual stems from a woody rootstock. Stems few to many,

suberect to prostrate, herbaceous but firm, suffruticose at the base, slender, terete,

covered, like all young vegetative parts, pedicels, bracteoles and alyces, with a dense

villous, tomentose to somewhat silky pubescence, 15-75 cm. long. Leaves palmately

5-fid with at least the central lobe irregularly pinnatifid or pinnatisect and the basal

lobes hastate and bifid, ovate or oblong in outline, 5-25 (-35) mm. long; ultimate

segments always narrow, usually under 2 mm. wide, often with revoluts edges or

somewhat concave; upper surface somewhat glabrescent, lower surface persistently

densely hairy, tomentose or villous ; petioles short, up to about 1 cm. long in tne largest

leaves. Peduncles 1 -flowered, up to 1 cm. long, sometimes 0; bracteoles small, linear-

subulate; pedicels longer than the peduncles, up to 15 mm., gradually becoming a

little thicker upwards. Sepals broadly ovate, outer ones 5-5-7 mm. long, often

somewhat thinner and crisped along the edges, obtuse or acute; inner ones usually

shorter, relatively broader, membranous-coriaceous, glabrous or nearly so except at

the apices, innermost mucronate to apiculate. Corolla funnelshaped, very pale pink

or white, 10-12 mm. long and 12-15 mm. in diam.; the lobes rounded to obtuse,

midpetaline areas densely silky. Capsule subglobose or ovoid-globose, apiculate,

glabrous, about 5 mm. in diam. Seeds very dark brown or black, glabrous, somewhat

longitudinally rugose on the back, verrucose-muriculate on the angles.

Endemic.

Cape Province. — Uitenhage, nr. Loeri River: Thunberg (photograph of type in

PRE), id. Grasrug: Baur 1020 (SAM). Herbert, Mazelsfontein: Anderson 602 (GRA,

PRE); Wilman s.n. (PRE); Victor H. no. 481 (KMG). Hay: Wilman H. nos. 2242,

7102 (KMG); Esterhuysen 4077 (BOL, KMG); prob. Hay (“ between upper Campbell

and Griquatown ”) : Burchell 1839 (BOL, GRA, L; in FI. Cap. referred to C.

boedeckerianus!). Kimberley: Levey in herb. Galpin 6332 (PRE); Elliott H. no. 472,

676

p.p. (mixed C. boedeckerianus, KMG); Wilman s.n. (BOL, h. no. 15843; SAM, h. no.

26794). Barkly West: Brueckner 150 (PRE); Pagan H. no. 473, p.p. (mixed with

C. boedeckerianus, KMG). Vryburg: Henrici 168 (PRE); Burtt-Davy 14675 (PRE,

NBG); without precise locality, but most probably Kimberley: Marloth 8417 (PRE).

Orange Free State.— Bloemfontein: Marais 96 (PRE).

This species has frequently been misinterpreted. Choisy, l.c., referred Burchell

2412 to this species, but Burchell 2412 belongs to C. ocellatus. Hallier at first confused

C. multifidus with C. ornatus Engl. (=C. ocellatus) and with “ C. thunbergii”, but

later he was able to study the type specimen and annotated: “ Convolvulus multifidus

Thunb. ! 1818 — C. capensi Burm. arete affinis et forsan mera varietas, a Choisy in

DC., Prodr. IX p. 410 false cum Burchell 2412, a me in Engl. Jahrb. XVIII p. 102

false cum Conv. ornato Engl, conjuncta. 19— III— 1909

The type, of which a photograph was kindly sent by Mr. Aim of the Uppsala

herbarium, enabled me to identify specimens referable to this species. I cannor agree

with Hallier, however, that it is very closely related to C. capensis. It is much more

closely related to C. boedeckerianus and forms of C. ocellatus ( C . ornatus ) and in some

specimens the differences are only very slight, but it is always possible to separate them

as follows: —

677

6. C. aschersonii Engl., Hochsgeb. fl. Trop. Afr. p. 349 (1892). C. penicellatus. Hall.

f. in Engl. Bot. Jb. 18 (1893), p. 103, non A. Rich. C. sagittatus var. australis

subvar. abyssinicus (“ abyssinica ”) Hall. f. in Bull. Herb. Boiss. 6 (1898), p. 533.

C. sagittatus var. abyssinicus (“ abyssinica ”) (Hall, f.) Baker & Rendle in Dyer,

Fl. Trop. Afr. 4, 2 (1905), p. 96.

Type: Schimper 660 from Abyssinia in B (now destroyed).

Perennial forming annual prostrate or sometimes twining stems from a woody

taproot. Stems terete, herbaceous, but firm, densely covered (as are all vegetative

parts, peduncles, bracteoles, pedicels and outer sepals) with a fulvous or silvery- white

pubescence, up to about 2 m. long. Leaves variable in shape, often on the same plant,

but generally oblong to lanceolate-oblong, varying to linear-hastate or ovate, usually

hastate at the base with the basal auricles lobed or dissected and the central lobe

irregularly crenate-repand-pinnatilobed, not infrequently with two small oblong lateral

lobes at the base and leaf subpalmately 5-7-lobed, sometimes pinnatisect, 3-6 cm.

long and 0.5-2 -5 (-3-5) cm. wide; pubescence usually strigose on upper surface,

less adpressed on lower one; the apex usually obtuse to rounded, emarginate and

minutely reflexed-mucronate, rarely acute; main nerves and midrib impressed above,

prominent below, petioles usually rather stout, terete, 5-20 mm. long. Peduncles usually

longer than the petioles shorter or longer than the subtending leaf, slender, terete,

subumbellately to cymosely 1-6-flowered, gracteoles linear or oblanceolate, 3-4 mm.

long; pedicels short. Sepals unequal, outer ones herbaceous, oblong, elliptic or ovate-

oblong, hairy and ciliate, obtuse or acute, sometimes crisped along the margins, 6-7

mm. long; inner ones blabrous or gradually less hairy to glabrous except at the tips,

membranous to thinly coriaceous, marcescent, shorter, much broader (innermost

broadly ovate to suborbibular, about 4 mm. long), much imbricate. Corolla funnel-

shaped, white or very pale mauve-pink, sometimes with mauvish markings in the

throat, 7-10 mm. long and about as much in diam., midpetaline areas pale greenish

outside, silky towards the apex and terminating in a beard of hairs. Stamens with

a few short, rounded papillae near the base. Capsule globose, apiculate, glabrous,

6-7 mm. in diam. Seeds very dark brown, very finely verrucose-rugose, not distinctly

so along the angles, 3-4 mm. long.

From Eritrea and Abyssinia to East Africa, extends into Angola, Bechuanaland,

Southern Rhodesia and Transvaal.

Angola. — Ambaca: Welwitsch 6204 (COI, K).

Transvaal. — Brits, near Wolhuterskop: Meeuse 9274 (PRE); Pretoria, near

Pretoria: Leendertz 552 (PRE), Moss 4720 (J), Thode A440 (PRE); Hammanskraal :

Hutchinson 2895 (BOL, GRA, PRE); Potgietersrust: Galpin s.n. 9038 (PRE); Peters-

burg, nr. Pietersburg: Meeuse 9191, 9237 (PRE), Woodbush: Wager s.n. (PRE),

Blaauwberg: Codd 8732 (PRE).

S. Rhodesia. — Bulawayo: Eyles & Johnson 54 (GRA); Rogers 5754 (BOL); Kolbe

4018 (BOL). Matopos: Kolbe 4345 (BOL). Plumtree: McCosh 13 (SRGH). Gwelo:

Kolbe 4283 (BOL), Eyles 1820 (PRE, SAM, SRGH). M’Sonedi: Hopkins H. no.

6762 (SRGH). Enkeldoorn: Eyles 8960 (SRGH). Gwebi: Baines 18 (SRGH).

Marandellas : Corby 496 (SRGH, PRE), Dehn 189 (SRGH). Salisbury: Kolbe 4270

(BOL), Hopkins H. no. 7828 (SRGH, PRE), Blenkinson in herb. Moss no. 14811 (J),

Young 109 (PRE), s.n. in herb. Moss 18497 (J), Morris 288, 303 (NBG). “Premier

Mine” Moss 18497 (J).

Eritrea. — Acrour: Schweinfurth & Riva 1061 (K). Saganeiti: Schweinf'urth

& Riva 1739 (K).

678

This species was reduced to a variety of C. sagittatus by Hallier and in FI. Trop.

Afr., but I am of the opinion that it is a distinct species, much more closely related

to C. ulosepalus Hall. f. than to C. sagittatus. The type was not seen, but other numbers

cited in FI. Trop. Afr. agree with material from Southern Africa. C. aschersonii can

be distinguished from C. sagittatus by its usually few-flowered inflorescences (the

flowers are usually solitary in the latter), small flowers (8-10 mm. long) and different

leaf-shape (middle-lobe up to 6 cm. long and 12 mm. wide, often crenate or sinuate,

occasionally dissected; basal lobes often toothed or multifid). These characters do

not occur together in C. sagittatus. From C. ulosepalus it is distinct by its depressed

short silky pubescence on both surfaces of the leaf (C. ulosepalus is never so densely

hairy) and by the sepals which are not so unequal as in C. ulosepalus.

The specimens from tropical Africa and Eritrea quoted by Hallier and in FI. Trop.

Afr. vary in the shape and size of the leaves, length of peduncles, etc. Some of the

South African and Rhodesian specimens match some of these quoted specimens:

Hopkins h. no. 7828, Galpin s.n. (Potgietersrust), Wager s.n. (Woodbush) are excellent

matches of Schweinfurth & Riva 1061, and Eyles 1820, Meeuse 9151, 9237 of Schweinfurth

& Riva 1739.

7. Convolvulus ulosepalus Hall. f. in Engl. Bot Jb. 18 (1893), p. 103; Baker & Wright

in Dyer, FI. Cap. 4, 2 (1904), p. 73; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), p. 95; Salter in Adams, and Salter, FI. Cape Penins. (1950), p. 685. C.

rhynchophyllus Baker & Engl, in Engl. Bot. Jb. 10 (1888), p. 247, nomen tantum;

Hall f. in Engl. Bot. Jb. 18 (1898), p. 104 (name only) and in Bull. Herb. Boiss.

6 (1898), p. 534, descr. latina.

Type: Not designated, because Hallier mentioned several specimens (some of

which I was able to examine).

Perennial, forming several to many annual stems, from a woody taproot. Stems

prostrate or climbing, herbaceous, slender, attaining a length of 1-2 m. or sometimes

more, when young hairy or nearly glabrous, when old glabrescent or completely

glabrous. Leaves 2-10 cm. long and 3-10 (-35) mm. wide, hastate or somewhat

sagittate with a linear to linear-lanceolate or oblong central lobe and entire or bifid

basal lobes, rarely leaves palmately 5-lobed with a large median lobe, much smaller

patent lateral lobes and hastate, often angular, toothed or bifid basal lobes; central

lobe usually entire to somewhat repand or crenate occasionally irregularly serrate or

lacerate, usually obtuse and mucronate, both surfaces of older leaves thinly hairy or

glabrous; petioles hairy like the stems, 5-20 (-30) mm. long. Peduncles 2-6 (-8)-

flowered, on young stems and depauperate specimens sometimes 1 -flowered, usually

much longer than the petioles, terete, slender, hairy like the stems and petioles,

bracteoles, linear, linear-lanceolate or -oblanceolate, hairy, 3-6 mm. long; pedicels

short. Sepals unequal, outer ones herbaceous, often purplish towards the tips, ovate,

elliptic or ovate-lanceolate, acute, pubescent outside, 5-6 mm. long; inner ones shorter,

glabrous or nearly so, marcescent to coriaceous, suborbicular, abruptly mucronate.

Corolla white or pale pink, funnel-shaped, 7-9 mm. long; and 9-12 in diam. ; lobes

acute, midpetaline areas hairy at the tips. Stamens without hairs or papillae at the base.

Capsule globose, apiculate, glabrous, 5-7 mm. in diam. Seeds dark brown to almost

black, glabrous, nearly smooth, very minutely and indistinctly verrucose. rugose or

punctate, 2 -5-3 -5 mm. long.

South Africa, wide-spread, extends into Southern Angola and Bechuanaland.

Recorded from.— South West Africa (wide-spread, except in the driest regions);

Cape Province (from the Cape Peninsula eastwards, also Wittebergen and Laingsburg,

throughout central Cape and Griqualand West as far east as Queenstown and Albany.

Orange Free State (wide-spread); Basutoland; Natal (only recorded from Newcastle,

679

Weenen and “ Upper Tugela ”); Transvaal (Marico, Zeerust, Christiana, Wolmarans-

stad, Klerksdorp, Potchefstroom, Vereeniging, Johannesburg, Brits, Pretoria, Lyden-

burg, Carolina, Ermelo, Standerton, Barberton).

Some interesting specimens are the following: Drege 741a (Mooiplaats, prob.

Albert Div., in L); Tyson 124 (Murraysburg; in SAM, NH); Bolus 252 (Graaff-

Reinet, in BOL, GRA, PRE, SAM); Baur 901 (PRE, SAM; according to the label

in PRE, Hallier (1893) and FI. Cap. collected at Shiloh, Queenstown, but according

to the label in SAM from Baziya, Transkei; the first locality is most probably correct);

Marloth 979 (from Boetsap, Barkly W., in PRE); Cooper 547 (PRE, from Beaufort

East). All these were referred to C. rhynchophyllus by Hallier (1893, 1898). Drege

741e (L, from Beaufort-West), Drege 7829a (L, from Richmond), these two were among

the original numbers cited by Hallier under C. ulosepalus.

The oldest name is C. rhynchophyllus Baker ex Engl., but this name remained

a nomen nudum till 1898 when it was validated by Hallier by a latin description. The

name C. ulosepalus Hall. f. (1893), therefore must be retained, because C. ulosepalus

and C. rhynchophyllus cannot be separated. The oldest specimen of this species is

found in herb. Thunberg. Thunberg named it “ C. sagittatus 2 ”. Hallier annotated

this specimen: “Convolvulus ulosepalus Hall. f. 1893. Non Convolvulus sagittatus

Thunb. FI. Cap.” The specimen “ C. sagittatus 1 ” in the Thunberg herbarium agrees

with Thunberg’s description and with the conceptions of other authors of C. sagittatus,

and is to be taken as the lecto-type of the latter.

Sometimes the leaves of C. ulosepalus are considerably wider than in “ typical ”

specimens and they resemble those of some forms of C. farinosus L. with dissected

leaves. C. ulosepalus can be distinguished by the rotundate inner calyx lobes, the

non-papillose anthers and the smaller flowers (8-10 mm. against 12-15 mm. in C.

farinosus). C. ulosepalus differs from C. sagittatus in the usually distinctly bifid basal

lobes of the leaf (mostly not or sightly lobed in C. sagittatus), the 2-6-flowered inflores-

cences, the smaller flowers and the rotundate inner calyx-lobes. For the distinction

between C. ulosepalus and C. aschersonii, see under the latter. Depauperate specimens

may resemble C. boedeckerianus, but differ from the latter in the leaf-shape, peduncled

flowers and the shape of the sepals. C. ulosepalus not infrequently becomes a noxious

weed which is difficult to eradicate on account of its perennial rootstock.

8. C. sagittatus Thunb., Prodr. PI. Cap. (1794), p. 35; Choisy in DC., Prodr. 9(1845),

p. 407; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 103 and in Bull. Herb. Boiss. 6

(1898), p. 533; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 72; Baker &

Rendle in Dyer FI. Trop. Afr. 4, 2 (1905), p. 96, pro parte (exclus. var. abyssinicus).

C. hastatus Thunb. var. natalensis Baker in Dyer, FI. Cap. 4, 2 (1904), p. 72.

Type: A specimen in the Thunberg Herbarium, Uppsala (photographs in PRE).

There are two sheets on which Thunberg has written “ Convolvulus sagittatus ”, the

one is marked “ 1 ” and agrees with Thunberg’s description of C. sagittatus, the second

sheet, marked “ 2 ” is Convolvulus ulosepalus Hall. f. Hallier annotated the second

specimen in 1909: “ Convolvulus ulosepalus Hall. f. 1893. Non Convolvulus sagittatus

Thunb. FI. Cap.” As Thunberg’s description agrees with the specimen “ 1 ” and not

with the specimen “ 2 ”, “ 1 ” must be taken as the lecto-type.

Perennial, forming several annual stems from a long thin perennial taproot.

Stems usually only branched from the base, slender, prostrate or occasionally in some

forms climbing, usually hairy, usually under 60 cm. long but in some forms attaining

1-2 m. Leaves varying from linear or linear-sagittate to hastate-sagittate or oblong

with truncate base, usually entire except the basal lobes, usually somewhat hairy,

sometimes densely so or nearly to quite glabrous; petioles usually very short and

often hairy. Peduncles 1 -flowered or in some forms few-flowered, usually terete, slender.

680

up to 4 cm. long, often hairy; bracteoles small, pedicels short or occasionally up to

3 cm. long. Calyx usually hairy, sepals varying from lanceolate to orbicular, acute

to obtuse, sometimes mucronate or ciliate. Corolla funnel-shaped, white or sometimes

pale pink or mauve-pink, midpetaline areas hairy near the tips. Capsule subglobose,

glabrous. Seeds usually 4, dark brown or black, glabrous, when ripe, somewhat

tuberculate-rugose.

Wide-spread in South Africa, in addition Eritrea, Abyssinia and East Africa;

extends into Northern and Southern Rhodesia and Angola, also in Arabia and probably

also in Madagascar.

As regards the taxonomic subdivision of C. sagittatus the following division,

mainly based on Hallier’s publication in Bull. Herb. Boiss. 6 (1898), p. 533-534, is

presented : —

Corolla 8-12 mm. long, sometimes up to 15-18 mm. long subsp. sagittatus.

Corolla 15-20 mm. long subsp. grandiflorus.

The subspecies sagittatus can be divided as follows: —

Leaves linear-sagittate with entire, rounded or rarely bi-lobed basal auricles;

peduncles usually 1 -flowered and usually under 3 cm. long; sepals

ovate or broadly ovate, acute, hairy to nearly glabrous; plants thinly

hairy to nearly glabrous with adpressed hairs; stems prostrate, rarely

climbing var. sagittatus.

Leaves sagittate or oblong-sagittate or somewhat hastate, up to 3 cm. long

and 10-15 mm. wide; basal auricles entire, peducles 1 -flowered, under

2 cm. long; sepals subspathulate, elliptic or obovate, obtuse, mucronate,

with crisped margin, usually quite glabrous; plants usually densely and

shortly pubescent on stems and petioles; stems prostrate var. phyllosepalus.

Leaves lanceolate- or linear-hastate, up to 4 cm. long, narrow but measured

across the entire, bifid or 2-3 toothed basal auricles up to about 20 mm.

wide; peduncles few-flowered or 1 -flowered, short or sometimes up to

4 cm. long; sepals broadly elliptic to suborbicular, abruptly apiculate,

acute, subhirsute; plants thinly to rather densely covered with rather

short stiff (subhirsute) hairs; stems prostrate var. hirtellus.

Leaves linear-sagittate with cordate or hastate-sagittate base and bifid basal

lobes which are entire or dentate; peduncles few-flowered, 5-5-5 mm.

long; sepals ovate or elliptic, acuminate; whole plant farinose and thinly

sericeous; stems usually climbing var. namaquensis.

The subspecies grandiflorus can be divided as follows: —

Leaves hastate-subcordate to oblong with truncate base, up to about 3 cm.

long and 2 cm. wide, rarely longer and narrower; peduncles 1-flowered,

usually under 3 cm. long; sepals lanceolate to elliptic or ovate, acute

or cuspidate, hairy; plants usually prostrate to suberect var. grandiflorus

(var. subcordatus ).

Leaves linear, up to 4 cm. long and 2-3 mm. wide; peduncles usually 1-flowered,

up to 4 cm. long; pedicels 2-3 mm. long; sepals ovate-lanceolate,

acuminate, subglabrous, plants usually prostrate var. graminifolius.

Leaves linear with 2 (often minute) auricles at the base, up to 7 cm. long and

4 mm. wide; peduncles usually 1-flowered, up to about 2 cm. long;

pedicels about as long as the peduncles ; sepals ovate-lanceolate or ovate,

thinly silky outside, glabrescent or occasionally tomentose, plants

often climbing var. linear ifoli us.

Not included are C. sagittatus var. parviflorus Hall. f. subvar. villosus Hall, f.,

op cit., p. 533 from E. Africa = Convolvulus thompsoni Baker in Kew Bull. 1894,

p. 67, which I have not seen (it may be a depauperated form of the following), and

the subvar. abyssinicus Hall. f. 1. c., which is Convolvulus aschersonii and must be

excluded.

681

8 a. C. sagittatus ssp. sagittatus. C. sagittatus var. parviflorus Hall. f. in Bull. Herb.

Boiss. 6 (1898), p. 533.

8 a. 1. C. sagittatus ssp. sagittatus var. sagittatus. C. sagittatus var. parviflorus Hall.

f. subvar. australis Hall, f., 1. c.

The following specimens I refer to the var. sagittatus: —

Cape Province. — Karroo ” : Thunberg (photograph of type in PRE). “ Blaauw-

pan, Karroo” (= ? Blouput, Prince Albert): Moss 17985 (J). Uitenhage: Ecklon

& Zeyher 2-9 (L). Somerset E., Cookhouse: Kensit in herb. Bolus no. 9316 (BOL).

Graaff Reinett, Ripplemead: Hutton 464 (GRA), 514 (BOL). Albert, Burghersdorp :

Pocock 99(GRA). Queenstown: Cooper 266 (PRE) ; Galpin 2009 (PRE). Sterkstroom:

Sim 4108 (GRA), this specimen forms a transition to the forma graminifolia. Albany,

near Grahamstown: McOwan 950 (GRA, SAM); Sole s.n. (GRA); Lotsy & Goddijn

(L); Martin 698 (NBG). Bathurst, Trappe’s Valley: Daly 671 p.p. (GRA); 671

(BOL). Kentani or Willowvale: Drege s.n. (L). Willowvale: Drege s.n. (L).

Orange Free State. — Fauresmith: Henrici 1963 (PRE). Bloemfontein: Mostert

651 (PRE), Gemniell 4973 (PRE). Heilbron, Coalbrook: Gilmore 2129 (PRE).

Basutoland.— Leribe : Dieterlen 97b (PRE, NH, SAM).

Cape Province. — Griqualand-W., Kimberley: Moran s.n. (BOL).

Natal.- — Estcourt: Schlechter 3362 (BOL, GRA, PRE); West 341 (PRE); Acocks

10552 (PRE). Winterton, Grantleigh: King 10 (PRE). Colenso: Wood s.n. (SAM).

Weenen: Acocks 10696 (PRE, NH). “Upper Tugela ”: Wood 3430 (HN). Pieter-

maritzburg: Killick & Marais 1998 (PRE). Vryheid: v.d. Merwe 2452 (PRE).

Transvaal. — Potchefstroom : Louw 1032 (PRE). Vereeniging: Leendertz H.

no. 10818 (PRE). Witbank, Zondagsfontein: Thode A2842 (NH, PRE). Without

precise locality: Stainbank in Herb. Wood no. 3650 (NH).

Bechuanaland. — Schoenfelder S 178 (PRE); Munro ML 8 (PRE).

The specimens Gilmore 2129, Dieterlen 97b, Schlechter 3362, West 341, Acocks

10552, 10696, Wood 3430, van der Merwe 2452, Killick & Marais 1998, Leendertz H.

no. 10818, Thode A 2842 and Stainbank in h. Wood 3650 form a transition to the forma

phyllosephalus; the specimens Henrici 1963, Smith 5169 and Pagan s.n. to the forma

hirtellus ; the specimens v.d. Merwe 2452 and Galpin 2029 to the ssp. grandifolius

var. linearifolius.

The following specimens from tropical Africa are, in my opinion, indistinguishable

from typical var. sagittatus , although there seems to be a gap in the distribution:

Nyasaland, Lake Nyasa: Galpin 15019; N. Rhodesia, Mazabuka: CRS 395,

470; Kenya, Nairobi: Verdcourt 368 (all PRE).

8. a. 2. C. sagittatus ssp. sagittatus var. phyllosepalus {Hall. /.) A. Meeuse, nov. stat.

C. phyllosepalus hall, f., op. cit., p. 535; Baker & Wright, op. cit., p. 75. C.

sagittatus, var. latifolius C. H. Wright in Dyer, FI. Cap. 4, 2 (1905), p. 72.

Type of var.: Hallier mentioned three specimens in herb. Zurich viz. Rehniann

3796, 4131 and 4674. These specimens were kindly sent on loan by the Zurich her-

barium and proved to be identical with Wright’s “ var. latifolius ” of which original

specimens had been compared with material of the National Herbarium, Pretoria.

One of Rehmann’s numbers (no. 3796) bears a label “ Typus ”. It is not known if

Hallier was responsible for this typification; at any rate. I propose Rehmann 3796

(in Z) as the lecto-type of this variety.

682

Occurs in the Orange Free State and the Transvaal, with intermediate forms in

Natal (see under var. sagittatus).

Orange Free State— Bethlehem, Clarens: Van Hoepen s.n. (PRE). Harrismith,

Witzieshoek: Junod s.n. (PRE). Kroonstad: Pont 36 (PRE). Ventersburg: Acocks

12501 (PRE). Bloemfontein: Rehmann 3796 (herb. Zurich). Glen: Glen School of

Agr. s.n. (PRE).

Transvaal. — Wolmaransstad: Sutton 86 (PRE). Marico, Zeerust: Pott s.n.

(PRE). Potchefstroom : Theron 1053 (PRE, NH); Liebenberg & Phillips B.L. 938,

949, 971A (PRE). Johannesburg: Bryant D31, 207 = D38 (PRE); Moss 10572 (J).

Springs, Geduld: Moss 15610, 15611 (J). Heidelberg, Suikerbosrand: Schlechter

3484 (BOL, GRA. PRE). Pretoria, Pretoria: Rehmann 4131, 4674 (Z); Leendertz

s.n. (PRE); Goosens 15 (PRE); Moss 9633 (J); Liebenberg 3399 (PRE); Mogg 15245

(PRE); Meeuse 9030 (PRE); Makkink s.n. (PRE); Coinins 859 (PRE). Irene: Burtt-

T>ayy2316(BOL). Bronkhorstspruit: Rogers s.n. (PRE). Bethal : Leendertz s.n. (PRE).

Standerton: Jenkins s.n. (PRE; this specimen is somewhat intermediate between the

var. phyllosepalus and the var. hirtellus ), Moss 17651 (J).

8. a. 3. C. sagittatus ssp. sagittatus var. hirtellus (Hall. /.) A. Meeuse, stat. nov.

C. hirtellus Hall, f., op. cit., p. 536.

Type of var. : Hallier mentioned two specimens, viz. Burke s.n. (K) and Rehmann

3848 (Z). The Burke specimen was kindly sent on loan by the Kew herbarium, and

enabled me to identify Hallier’s species. I propose Burke’s specimen as the lecto-type

of the variety.

Occurs in the Northern Cape Province, Orange Free State, Basutoland and

Transvaal High veld.

Cape Province. — Aliwal North: Gerstner 253 (PRE); this specimen is inter-

mediate towards the var. sagittatus.

Orange Free State. — Bloemfontein: Moraile 500 (PRE). Kroonstad: Pont 383

(PRE); without precise locality, near Vaal River; Burke s.n. (herb. Kew).

Basutoland. — Mokhotlong: Liebenberg 5820 (PRE).

Transvaal. — Potchefstroom: v.d. Westhuizen 269 (PRE). Benoni: Moss 14131

(J). Pretoria: Burtt-Davy 44 (PRE). Nigel, between Devon and Leslie: Dyer &

Verdoorn s.n. (PRE). Heidelberg: Codd 8508 (PRE). Bethal: Leendertz s.n. (PRE).

Ermelo: Leendertz s.n. (PRE); Henrici 1682 (PRE; this specimen is somewhat

approaching the var. sagittatus ).

8. a. 4. C. sagittatus ssp. sagittatus var. namaquensis A. Meeuse, var. nov. Convol-

vulus namaquensis Schltr. ms. on Schlechter 11124. Tota planta farinosa et sparse

sericea. Folia lineari-sagittata, basi cordata vel hastata-sagittata, lobi basales

bifida, dentata vel integra. Pedunculi pauciflori, 5-55 mm. longi. Sepala ovata

vel elliptica, acuminata. Corolla 14—18 mm. longa.

Type of var.: Schlechter 11124 in PRE (iso types in BOL, GRA, L). This form

is mainly found in Namaqualand, but also in Laingsburg and a very similar plant,

which is however, more densely silky-pubescent, was collected in the Ceres district.

Cape Province.— Little Namaqualand, Brakdam: Schlechter 11124 (BOL, GRA,

PRE); Pillans 5605 (BOL); Spektakel: Bolus 9423 (BOL); W. Morris in herb. Bolus

n. 25040 (BOL); Khamiesberg: Esterhuysen 1301 (BOL); Laingsburg, Kl. Roggeveld,

Schietfontein: Compton 8121 (NBG). A very hairy form which in other respects

agrees with this form was collected in the Cape Province, Worcester, Brandvlei ( Barker

7512 in NBG).

683

8. b. C. sagittatus subsp. grandiflorus {Hall, /.) A. Meeuse, stat. nov. C. sagitatus

var. grandiflorus Hall. f. in Bull. Herb. Boiss. 6 (1898), p. 533.

Type of subspecies: Not designated.

8. b. 1. C. sagittatus ssp. grandiflorus var. grandiflorus. C. sagittatus ssp. grandiflorus

var. subcordata (Hall, f.) Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 97. C. steudneri Engl., Hochgeb.-fl. Trop. Afr. (.1892), p. 350; Hall. f. in Engl.

Bot. Jb. 18 (1893), o. 104. C. angolensis Baker in Kew Bull. 1894, p. 67; Baker

& Rendle op. cit., p. 95. Ipomoea huil/ensis Baker op. cit., p. 70. Convolvulus

sagittatus var. grandiflorus subvar. subcordata Hall. f. in Bull. Herb. Boiss. 6

(1898), p. 534. C. huil/ensis (Baker) Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 97.

Type: Not designated.

Recorded from Abyssinia and Arabia, also in Angola and South West Africa.

South West Africa. — Dinter 696 (SAM). Auros, Otavi: Dinter 5614 (BOL,

SAM, PRE).

Angola. — Welwitsch 6131 (isotype of C. huillensis, COI).

Abyssinia.— Addis Abeba road: McLoughlin s.n. (PRE).

8. b. 2. C. sagittatus ssp. grandiflorus var. graminifolius {Hall. /.) Baker & Wright ,

op. cit., p. 72. C. sagittatus var. grandiflorus subvar. graminifolia Hall, f., op.

cit. (1898), p. 534.

Type of var.: Rehmann 7823 (herb. Zurich).

Endemic.

Transvaal. — Johannesburg, Modderfontein : Haagner s.n. (GRA).

Natal. — Camperdown: Rehmann 7823 (Z).

This variety may have to be united with the var. linearifolius.

8. b. 3. C. sagittatus ssp. grandiflorus var. linearifolius {Hall, f.) Baker & Wright in

Dyer, FI. Cap. 4, 2 (1904), p. 72; Baker & Rendle op. cit., 97 C. sagittatus var.

grandiflorus subvar. linearifolia Hall, f., op. cit. (1898), p. 534.

Type of var.: Galpin 1037 (herb. Zurich).

Recorded from South Africa, Angola, Rhodesia.

Cape Province. — Umtata: Schonland 3787 (GRA). Baziya: Baur 350 (GRA,

SAM). Umzimkulu, Clydesdale: Tyson 2135 (BOL, SAM).

Transvaal. — Belfast, Machadodorp: Galpin 13223 (PRE, BOL). Barberton,

Barberton: Galpin 1037 (PRE, BOL, GRA, NH, SAM, Z), Kaapsche Hoop:

Prosser 1474 (PRE, J). Pietersburg, Frischgewacht : Leendertz 829 (PRE, BOL).

Angola. — Baum 180 (COI).

S. Rhodesia. — Inyanga: Eyles 8473 (SRGH; a very hairy form, calyx densely

tomentose).

A variable species, or a species complex. Several more or less distinct forms can

be distinguished, but there is not a single good criterion to separate them with certainty.

I prefer to follow Hallier’s broad concept of the species, but include C. phyllosepalus

and C. hirtellus and exclude C. penicillatus A. Rich., which is, in my opinion, a distinct

species much more closely related to C. ulosepalus.

684

C. phvllosepalus was described as having broad (hastate-sagittate) leaves and

foliaceous, obtuse, mucronate, crisped, glabrous or subglabrous sepals, but intermediate

forms occur with narrower leaves and pubescent sepals (as in typical C. sagittatus),

with broad leaves and pubescent sepals and with narrow leaves and glabrous sepals.

C. hirtellus Hall. f. is a form in which the pubescence is more hirsute and the basal

auricles of the leaves are 2-3-lobed; it was described as being always 1 -flowered,

but intermediate forms occur which link this form with typical C. sagittatus, and also

specimens are found with C. hirtellus characters, but with 2-3-flowered inflorescences.

The forms with larger flowers (var. grandiflorus) are sometimes quite different in

appearance from typical C. sagittatus, but if the variation in leaf-shape (from linear-

sagittate to cordate-ovate-sagittate) and in pubescence is not considered to be essential

they only differ from typical C. sagittatus in the size of the corolla and this character

is not quite constant either.

Some specimens collected in S.W. Africa are similar to those described as C.

angolensis and C. huillensis from Angola. In Namaqualand, farinose specimens with

few-flowered inflorescences are found. Although farinose vegetative parts and several-

flowered inflorescences are also occasionally found in specimens collected elsewhere,

the Namaqualand form appears to be worthy of varietal rank.

C. sagittatus can usually be distinguished from the other African species of

Convolvulus, but sometimes one-flowered specimens of C. bidentatus and depauperate

specimens of C. ulosepalus and C. farinosus may closely resemble forms of C. sagittatus.

C. bidentatus has a different type of calyx, usually a long peduncle, large flowers, and

the basal lobes of the leaves bi-fid, and this combination of characters is sufficient for

a clear distinction (apart from its different distribution).

C. ulosepalus and C. farinosus can usually be distinguished by their different

leaf-shape, the long peduncle and, in the case of C. ulosepalus, the sepals, which are

much more unequal than in C. sagittatus.

C. aschersonii can be distinguished from C. sagittatus by a combination of

characters: the strigose, dense pubescence, few-flowered inflorescences, small flowers

and the often toothed or multifid basal lobes of the leaves never occur together in

C. sagittatus.

A form of C. ocellatus, described as C. dinteri Pilger, also resembles certain forms

of C. sagittatus, but the tomentum of C. ocellatus on the vegetative parts and the sepals

distinguishes it from C. sagittatus.

9. C. farinosus L., Mant. (1771), p. 203; Choisy in DC., Prodr. 9 (1845), p. 412; Hall.

f. in Engl. Bot. Jb. 18 (1893), p. 104; Baker and Wright in Dyer, FI. Cap. 4, 2

(1904), p. 74; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 98; Salter

in Adams, and Salter, FI. Cape Penins. (1950), p. 685. C. cordifolius Thunb.,

Prodr. FI. Cap. (1794), p. 35; Choisy, op. cit., p. 413. C. penicellatus A. Rich.,

Tent. FI. Abyss. 2 (1851), p. 74.

Type: Most probably the specimen in the Linnean herbarium which is proposed

here as the lecto-type.

Herbaceous perennial. Stems many, long, weak, slender, climbing and usually

branched, pubescent or farinose-puberulous, the younger shoots often silvery. Leaves

usually cordate-deltoid or sagittate, acute with terminal mucro, rarely obtuse, subentire

to irregularly and shallowly crenate, herbaceous drying membranous, glabrous above

except when young, glabrous or more or less pubescent beneath, distinctly netted-

veined; basal sinus broad, basal auricles rounded or pointed, sometimes with a few

685

teeth; very rarely leaves with additional triangular lobes above the basal auricles;

length of blade 4-9 (-12) cm., width 3-7 (-9) cm.; petiole about half as long as the

blade, finely and densely pubescent. Peduncles usually about as long as, or longer

than the subtending petioles, finely pubescent, subumbellately 1-4 (-6)-flowered ;

bracts minute, linear or lanceolate, pubescent; pedicels short. Sepals unequal, oblong

or elliptic (outer ones) to suborbicular (inner ones), much imbricate, acute or obtuse

with terminal mucro, 6-8 mm. long; two outer ones more or less pubescent, three

inner ones gradually less pubescent to nearly glabrous and wider. Corolla funnel-

shaped, white, very pale pink or very pale pinkish mauve, 12-15 mm. long; the lobes

short, the tube rather narrow; midpetaline areas hairy towards the obtuse, mucronate

tips of the lobes. Stamens bearing short thick lateral papillae in their dilated basal

portions; anthers a dirty purple. Ovary glabrous. Capsule subglobose, shortly

apiculate, glabrous, 5-8 mm. in diam. Seeds usually 4, blackish when ripe, scabridulous-

rugose.

Western Mediterranean area, tropical and subtropical E. Africa, S. Africa,

Mascarene Islands.

In S. Africa it is not found in the more arid regions (S.W. Africa, Griqualand-W.,

■Orange Free State and W. Transvaal).

Recorded from the Cape Province: From the Cape Peninsula and Ceres along

the coast, in the Eastern Cape Province also more inland (Albany, Fort Beaufort,

Keiskamma Hoek); Natal and Zululand (wide-spread); Swaziland; Transvaal:

Pretoria, Bethal, Petersburg, Zoutpansberg, Letaba, Nelspruit, Barberton; Por-

tuguese E. Africa (Lourengo Marques). I have seen only one specimen from S.

Rhodesia (nr. E. border, Chirinda: Wild 2151 in SRGH) and one from Ngamiland,

Bechuanaland ( Curson 6 in PRE).

Some interesting specimens are the following: Drege 7830 (between Zuurberg

and Klein Bruintjeshoogte, Somerset E., in L), and Drege “ C. cordifolius ” Th. a ”,

prob. from King William’s Town (L) quoted by Choisy and by Hallier.

The usually cordate-deltoid or sagittate, sharply acute and mucronate, often

more or less crenate leaves, the rather long, 1-6-flowered inflorescences and rather

small (12-15 mm. long) flowers characterize this species. However, forms with dissected

leaves are sometimes very similar to some forms of C. ulosepalus. For their distinction,

see under the latter.

Dr. B. Yerdcourt of the East African Herbarium (Nairobi) kindly pointed out

that the type of C. penicellatus A. Rich, which he studied (in P) is undoubtedly C.

farinosus L. Hallier treated Richard’s name as a synonym of C. aschersonii (see no. 6)

and he was followed in Flora of Tropical Africa 4, 2 (1905), p. 96.

10. C. bidentatus Bernh. apud Krauss in Flora 27 (1844), p. 829; Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 105, in syn. C. hastatus Thunb., Prodr. FI. Cap. 1 (1794), p. 35,

and in FI. Cap. 2 (1818), p. 17; Choisy in DC., Prodr. 9 (1845), p. 407, ex parte;

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 105; Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 72, exclus. var. natalensis Baker, non C. hastatus Forsk. (1775),

nec Desr. (1789).

Type: Originally the specimen Krauss s.n. from George in B, now destroyed;

if no duplicates of the Krauss gathering are extant (not one could be traced), I propose

Thunberg’s specimen in the Thunberg herbarium provisionally as the neotype.

Probably a perennial. Stems several from the base of the taproot, slender (basal

parts up to about 3 mm. thick), prostrate or climbing at the ends, up to about 3 m.

1332796-4

686

long, usually somewhat quadrangular, glabrous or sometimes pubescent. Leaves

narrowly hastate to linear with hastate base, up to 45 (occasionally 70) mm. long and

the middle lobe 1-6 (-8) mm. wide; glabrous or pubescent, the basal lobes up to

25 mm. long, usually deeply bi-fid; the apex usually obtuse, mucronate; the margin

entire, lowermost leaves if present sometimes broader, oblong or sagittate-oblong or

occasionally palmately 5-7 lobed with the middle lobe the longest, up to 15 mm. wide.

Peduncles almost invariably 2-flowered, more rarely 1 -flowered or cymosely 3-5-flowered,

3-8 (-14) cm. long, rarely shorter, pubescent or glabrous; bracts lanceolate, 3-5 mm.

long; pedicels short, pubescent, usually somewhat 4-angled and subclavate, 5-10 (-15)

mm. long. Calyx 6-8 mm. long, glabrous or very rarely obscurely pubescent; sepals

broadly ovate or obovate, rather chartaceous with membranous edges, much imbricate,

obtuse to rounded and often more or less mucronate at the apex. Corolla (15-) 20-22

mm. long, white or pale pink; midpetaline areas slightly hairy near the apex outside.

Capsule globose, glabrous, about 6 mm. in diam., almost completely enclosed in the

calyx. Seeds black, minutely rugose and subpuberulous with tufts of very minute

brown hairs.

Cape Province. — Bredasdorp, between Bredasdorp and Malagas: Esterhuysen

4339a (BOL). Riversdale, Still Bay: Muir 3537 (PRE). Knysna: Breyer s.n. (PRE),

Duthie 913 (BOL), Four cade 1571, 6304 (BOL). Uitenhage, Zwartkopsrivier : Zeyher

239 (BOL. PRE), Ecklon & Zeyher 13 (SAM), Brehm 126 (SAM), locality illegible:

Brehm 339 (SAM), nr. Uitenhage: Marais 155 (PRE). Pt. Elizabeth: Holland 3757

(BOL), Laidley 115 (L), Long 822 (GRA, PRE), Williamson 64 (GRA), Redhouse:

Paterson 1088 (GRA), s.n. (PRE). Alexandria: Archibald 4517/17 (GRA, PRE),

4820b (GRA), Johnson 1105 (GRA, PRE). Albany, Alicedale: Cruden 137 (GRA),

Trappe’s Valley: Daly 671 p.p. (GRA), “Lower Albany”: Bowker s.n. (PRE).

Bathurst, Kowie: Britten 712 (GRA). E. London, N. mouth of the Buffalo River:

Murray 100 (SAM); Cape, without precise locality : Thunberg (photograph of proposed

neotype in PRE); “ Regio orientallis ” : Alexander s.n. (PRE).

The name Convolvulus hastatus cannot be used for this species on account of

C. hastatus Forsk. and C. hastatus Desr. [ = C. arvensis L. and Merremia tridentata (L.)

Hall. f. ssp. hastatus Ooststr., respectively]. The only other specific epithet available

is bidentatus. Unfortunately the Krauss specimen on which this name is based, and

which was evidently the holotype, must have been destroyed. However, it is evident

from Hallier’s treatment of this species (1893), from the other specimens he quoted

and from some specimens which were annotated by Hallier himself, and were available

for study, that his conception of the species C. bidentatus Bernh. is limited to the

specimens agreeing with the type of C. hastatus Thunb., so that I have no doubt that

C. bidentatus Bernh. and C. hastatus Thunb. are synonyms and retain the first name

for this species.

Stranger is Hallier’s distinction of a var. major (“ Flores dupla magnitude, 2 cm.

longi ”), because the flowers in Thunberg’s type and in most specimens I have seen are

about 2 cm. long. The inclusion of the var. major in the species in Flora Capensis was,

therefore, quite correct.

As Hallier pointed out, Thunberg’s short diagnosis is quite sufficient to charac-

terise the species: “ Folia hastata, lobi laterales bifidi; pedunculi raro uniflori, saepius

biflori; calyx glaber ”. Hallier added that the broad, rounded to emarginate brown,

pale-edged, much imbricate sepals form a very good distinctive character. In spite

of all this, the species was not properly understood in the treatment in FI. Cap., because

several specimens referred to C. hastatus in the FI. Cap. do not belong here, such as

those included in the var. natalensis (= C. sagittatus! ), the Transvaal specimens, and

the gathering Leipoldt 321 (= C. capensis!) from Clanwilliam.

687

Some specimens of C. sagittatus resemble C. bidentatus, but these two species

differ in several characters and there is always at least one character (either pubescence

of calyx, shape of sepals and/or leaves) to distinguish them. Also some forms of C.

capensis may resemble C. bidentatus, but they are usually hairy on the calyx, have

mostly larger flowers with more densely hairy midpetaline areas, and but rarely a

similar leaf-shape.

11. C. galpinii C. H. Wright in Dyer, FI. Cap. 4, 2 (1904), p. 75.

Type: Galpin 2110 from Queenstown (K).

Probably a perennial; whole plant densely villous or tomentose with fulvous or

grey short hairs. Stems prostrate or twining, slender, terete, 60 cm. long. Leaves

sagittate-deltoid to deltoid-cordate or ovate-cordate, 2-4 cm. long, usually somewhat

irregularly crenate-serrate, usually acute, basal sinus wide; basal lobes often somewhat

toothed or with angular small lobes; petioles up to about 12 mm. long. Peduncles

1-2-flowered, slender, terete, up to 6 cm. long, bracteoles linear, 5-8 mm. long;

pedicels up to about 1 cm. long. Calyx 6-10 mm. long; outer sepals ovate, abruptly

acuminate, villous, inner ones shorter and relatively broader, ovate-orbicular, less

hairy. Corolla funnel-shaped, white, 16-20 mm. long and as much or a little more in

diam. ; midpetaline areas densely silky. Capsule subglobose, glabrous, apiculate,

6-8 mm. in diam. Seeds dark brown, 4-5 mm. long, subpuberulous with fine, yellowish

brown rugosities of the testa.

Endemic.

Cape Province. — Queenstown, Queenstown: Galpin 2110 (PRE, GRA, BOL,

isotypes!). Stutterheim: Leighton H. no. 26651 (BOL); Evelyn Valley: Compton

19246 (NBG). King William’s Town, Mt. Coke: Compton 17001 (NBG). E. London:

Hilner 277 (GRA). Albany, Grahamstown, Collingham: Britten 6511 (GRA), Ather-

stone: Rogers s.n. (GRA). Tarka, Fairfield, Great Winterberg: Acocks 17637 (PRE).

Mt. Ayliff, Insizwa Mts. : Schlechter 6469 (GRA).

This species is closely related to C. natalensis Bernh. apud Krauss and may be

only a form or variety of the latter, but the specimens I have seen can be separated

from C. natalensis by their abruptly acuminate and shorter sepals, smaller corollas,

more slender, often twining stems and usually smaller, often more or less sagittate or

deltoid-cordate leaves [C. natalensis has usually prostrate, robust stems, lanceolate to

ovate-lanceolate, not abruptly acuminate sepals which are (10-) 12-20 mm. long

against 6-10 mm. in C. galpinii, a corolla which is (18-) 20-32 mm. long against 16-20

mm. in C. galpinii ]. In addition, the seeds of C. galpinii are puberulous, whereas those

of C. natalensis are finely tuberculate-rugose, glabrous, but this character cannot be

used to separate specimens without seeds.

12. C. natalensis Bernh. apud Krauss in Flora 27 (1844), p. 829: Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 105; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 77. C.

calycinus E. Mey. ex Drege, Zw. pfl.geogr. Doc. (1838), p. 154, nomen tantum.

C. calycinus E. Mey. ex Choisy in DC., Prodr. 9 (1845), p. 408, non Roxb., nec.

H.B.K.; Hall, f., op. cit. p. 105; Baker & Wright, op. cit., p. 77. C. transvaalensis

Schltr. in Jl. Bot. 34 (1896), p. 402. C. bullerianus Rendle in Jl. Bot. 39 (1901),

p. 62; Baker & Wright, op. cit., p. 77.

Type : Krauss 465 from Natal, “ Table Mountain ” near Pietermaritzburg (originally

in B?; isotype in BOL).

Endemic.

Two varieties can be distinguished.

688

12a. C. natalensis var. natalensis. C. calycinus E. Mey. ex Choisy, non Roxb. C.

natalensis var. integrifolia C. H. Wright in FI. Cap. 4, 2 (1904), p. 77.

Perennial forming prostrate or occasionally somewhat twining annual stems from

a woody taproot. Stems usually rather stout and firm, terete, up to 1 m. long, covered,

as are all vegetative parts, peduncles, pedicels, bracteoles and calyces with rather short

hairs which are sometimes stiff and hirsutulous, sometimes silky and sometimes

somewhat woolly, fawn, rusty-brown to silvery-white, more densely so in all the younger

parts, at length glabrescent. Leaves cordate-ovate, cordate-oblong or sometimes

cordate or narrowly reniform, 1-5 (-8) cm. long and 0-75-3 (-4) cm. wide, usually

undissected with an undulate, crenate or somewhat dentate or serrate to subentire

margin but sometimes faintly, rarely distinctly and irregularly pinnatilobed; the apex

usually broadly rounded and shortly cuspidate, sometimes obtuse or acute, the basal

sinus broad and shallow, the basal auricles often somewhat angular; petioles 5-10

(-20) mm. long. Peduncles 1- to^ few-flowered, terete, short (up to 7 cm. long), occasion-

ally 0; bracteoles linear or lanceolate, 6—10 mm. long, often more than 2 when

inflorescence few-flowered; pedicels very short or up to 15 mm. long. Calyx 10-15

(-18) mm. long; sepals ovate-lanceolate, oblong, elliptic or lanceolate, more or less

unequal; the outer ones herbaceous, sometimes subcordate at the base, broader and

crisped along the margins; inner ones a little shorter and less pubescent to glabrous,

somewhat marcescent, sometimes broadly ovate; all sepals obtuse to acute, the inner

usually more acute than the outer ones. Corolla white, cream-coloured or white tinged

with green, funnel-shaped, 20-35 mm. long and 22-40 mm. in diam.; midpetaline

areas greenish, densely silky with rusty-brown, fawn or silvery-white hairs. Capsule

sub-globose, or somewhat ovoid, apiculate, glabrous, 8-10 mm. long and in diam.

Seeds dark brown, glabrous, distinctly verrucose-rugose, 5-6 mm. long.

Cape Province. — Bathurst, near Pt. Alfred: Burchell 4040 (L). East London:

Galpin 7346 (PRE); Rattray 588 (GRA). Kei Mouth (Komgha-Kentani): Flanagan

1812 (BOL, PRE, SAM). Kentani: Pegler 213 (BOL, PRE, SAM). Umzimkulu,

Clydesdale: Tyson 2170 (BOL, SAM); Drege s.n. (isotype of C. calycinus E. Mey. ex

Choisy, L).

Basutoland. — Cooper 929 (BOL).

Natal. — Port Shepstone, Pt. Edward: Huntley 888 (NU, PRE). Ixopo, Maxwell:

Evans 272 (NH). Durban: Mogg 11018 (PRE). Pine Town, Umbogintwini: Wylie

s.n. (NH, PRE), Gillits: Wood s.n. (SAM). Impendhle: Levett 97 (NH). Inanda:

Wood 288 (BOL, NH). Pietermaritzburg: Compton 23736 (NBG); Huntley 275 (NU,

PRE); Mogg 2181, 2224 (PRE); Table Mountain: Krauss 465 (isotype, BOL), near

Allerton: Mogg 2181 (PRE). Lion’s River, Tweedie: Pegler in h. Wood 11026 (NH,

PRE); Torwood, Rosetta: Young s.n. (NH); Balgowan: Mogg 3548, 3844 (PRE).

Estcourt: Wood 3462 (NH, BOL). Dalton Bridge: Acocks 10568 (NH), 10624 (NH).

Winterton: Reyburn s.n. (NH). Greytown, Rietvlei, Greenwich Farm: Frey in herb.

Galpin no. 2730 (PRE). Eshowe: Lawn 1090, 1217 (NH). Entumeni: Forbes 795

(NH). “Zululand”: Gerrard 1331 (PRE). “ W. Zululand”: Baker in herb. Evans

uo. 561 (NH). Nqutu: Codd 7655 (PRE). Newcastle: Wood 5979 (PRE). Charles-

town: Wood 4702 (PRE).

Transvaal. — Wakkerstroom: Galpin 9818 (PRE); Beeton 240 (SAM). Ermelo,

GoedeHoop: Pott s.n. (PRE). S.E. Transvaal: “ Inter Delagoa Bay et Drakensbergen,

Transvaal”: Bolus 9710 (BOL). Lydenburg: Obermeyer 320 (PRE), Pietersburg,

Haenertsburg: Pott 4713 (PRE); Woodbush: Jenkins s.n. (PRE). Of these, the following

numbers are more or less intermediate between the forma natalensis and forma trans-

vaalensis : Huntley 275, Mogg 2181, 2224, 3548, Jenkins s.n.

689

12b. C. natalensis var. transvaalensis (Schltr.) A. Meeuse, stat. nov. C. transvaalensis

Schltr. in Jl. Bot. 34 (1896), p. 402. C. bullerianus Rendle in Jl. Bot. 39 (1901),

p. 62. C. natalensis var. angustifolia C. H. Wright, l.c.

Leaves linear with hastate base to oblong, or long-triangular with cordate to

sagittate or hastate, sometimes truncate or rounded base, entire or pinnatilobed, rarely

pinnately-palmately dissected, 3-6 (-9) cm. long and 0-2-2 (in some forms up to 5)

cm. wide. Sepals more often acute and lanceolate, otherwise as the var. natalensis.

Cape Province. — Cathcart: Thomas River, Barker 3494 (NBG). McLear: Britten

4527 (GRA). Mt. Currie, Vaalbank near Kokstad: Haygarth in herb. Wood no.

4179 (BOL, NH).

Natal. — Pietermaritzburg: Fairall 89 (NBG); Scottsville: Allsopp 404a, 900

(NU); Alexander Park: Allsopp 906 (NH, NU). Estcourt, Mooi River: Wood 4071

(BOL, NH, PRE), 6206 (isotype of C. bullerianus Rendle, PRE) ; Meteor Ridge : Mogg

3165 (PRE). Weenen, South Downs: Wood 4382 (NH). Utrecht, Kafir Drift: Thode

A241 (NH, PRE); Tweekloof, Altemooi: Thode A1176 (NH, PRE).

Swaziland. — Hlatikulu: Stewart s.n. (PRE).

Transvaal. — Barberton: Galpin 430 (isotypes of C. transvaalensis Schltr., BOL,

GRA, PRE); Williamson 77 (PRE); Edwards in herb. Moss no. 10182 (J); Codd

8131 (PRE). Nelspruit, White River: Rogers s.n. (PRE). Belfast, Draaikraal: Codd

8056 (PRE). Carolina: Acocks 13937 (PRE); Burtt-Davy 7356 (NBG).

The oldest available name is C. natalensis Bernh., validly published by Krauss

in 1844. The name C. calycinus was published with a description by Choisy in 1845,

but apart from the evident priority of the name C. natalensis, the specific epithet

“ calycinus ” had already been used twice before in Convolvulus, viz. by Roxburgh and

by H.B. et K. (see under Ipomoea sinensis on p. 729). The differences between C.

natalensis and C. calycinus E. Mey., as indicated in FI. Cap., break down altogether,

so that it is not necessary to find another name for the latter. C. transvaalensis Schltr.

was altogether overlooked in FI. Cap. (the type of this species — Galpin 430— is quoted

under C. natalensis var. angustifolia C. H. Wright). C. bullerianus Rendle is inseparable

from Wright’s var. angustifolia of C. natalensis. I do not think the varieties mentioned

in FI. Cap. can be upheld, because they are based only on the shape of the leaves and

more or less intermediate forms are found which link up the typical form (with ovate-

cordate leaves) with the narrow-leaved var. angustifolia (= C. transvaalensis and

C. bullerianus ). They are treated here as varieties, but these are not sharply defined^

There is also considerable variation in the shape of the sepals which can be very

broad, leafy and crisped on the edges to lanceolate; in the number of bracts (2-5);

in the inflorescence (1- to few-flowered) and in the stamens (glandular at the base or

eglandular).

The forms with narrow leaves sometimes resemble C. thunbergii very much and;

in my opinion, C. thunbergii is indeed much more related to C. natalensis than to C.

capensis (to which it was reduced as a variety in FI, Cap.). They can be distinguished

as follows: —

(a) leaves often dissected in C. thunbergii, usually entire in C. natalensis (or only

basal auricles dissected);

(b) peduncles usually 1-flowered in C. thunbergii, 1- or more-flowered in C.

natalensis ;

(c) bracts: 2 in C. thunbergii, 2-5 in C. natalensis;

(d) corolla about 20 mm. long in C. thunbergii, usually 25-35 mm. in C. natalensis;

(i e ) seeds black, smooth in C. thunbergii, brown and tuberculate-rugose in C.

natalensis.

690

13. C. thunbergii R. et S., Syst. Veg. 4 (1818), p. 268, ex descr. ; Drege, Zw. Pfl.geog.

Doc. (1838), p. 46. C. altheoides Thunb., Prodr. FI. Cap. (1794), p. 35, ex parte,

non L. C. capensis Burm. f. var. plicatus Baker in Dyer, FI. Cap. 4, 2 (1904),

p. 78, non Convolvulus plicatus Desr.

Type: ? (see notes).

Perennial forming several prostrate annual stems from a thin woody rootstock.

Stems slender or occasionally somewhat stout, herbaceous but firm, terete, covered

with rusty-brown or fawn, rarely silvery-white short stiff hairs, up to about 70 cm.

long. Leaves oblong in outline varying to lanceolate or ovate, cordate to hastate or

truncate at the base, undivided and crenate to somewhat pinnatilobed or usually

palmately 5-lobed with the middle lobe the longest and all lobes crenate to pinnatifid

or even bipinnatifid or whole leaf somewhat palmately bipinnatifid, ultimate lobes

sometimes very fine; length of blade 2-5 cm., width 0-5-3 cm., the apex acute or

occasionally obtuse; both surfaces thinly to densely covered with the same hairs as

on the stems, especially on the nerves; petioles hairy like the stems, slender, up to

10 mm., rarely 18 mm. long. Peduncles 1-flowered, sometimes 2-flowered, slender,

te,rete, hairy like the stems as are bractioles and pedicels, 1-4 cm. long; bracteoles

subulate or linear-subulate, 6-9 mm. long or sometimes minute, pedicels up to 12 mm.,

rarely to 15 mm. long. Calyx with short stiff hairs outside like the vegetative parts;

sepals unequal, outer ones herbaceous to chartaceous, ovate, ovate-lanceolate or

elliptic, acuminate, usually very acute, 8-12 (-15) mm. long; inner ones less hairy,

distinctly shorter, suborbicular, somewhat marcescent. Corolla white or pale pink,

funnel-shaped, 18-25 mm. long and 20-30 mm. in diam.; midpetaline areas greenish,

hairy towards the apex outside and usually very densely so. Capsule subglobose,

apiculate, glabrous, 7-9 mm. in diam. Seeds black, glabrous, very minutely punctate,

about 5 mm. long.

Endemic.

Cape Province. — Albert, Burghersdorp: Cooper 790, 1366 (BOL), Guthrie 4202,

4209 (NBG); Pocock 132 (GRA). Aliwal-N. • Flanagan 1507 (BOL, PRE, SAM);

Gerstner 107 (PRE); Jamestown 2133 (NBG); Ruigtefontein: Thode A 1828 (NH,

PRE); Elandshoek: F. Bolus 144 — L. Bolus no. 6847 (BOL). Barkly East: Gerstner

675 (PRE). Woodhouse, Dordrecht: Acocks 12532 (PRE). Graaff-Reinet: Bolus 230

(BOL). Queenstown: Galpin 2010 (PRE); Shiloh: Drege s.n. (“ C. thunbergii a”, L),

Baur 921 (SAM). Albany, Botha’s Hill: McOwan 586 (SAM). Mount Currie:

Goossens 219 (PRE). Locality not known to me (Sterkstroom?): Drege s.n. “ C.

thunbergii b ”, in L).

Basutoland. — Leribe: Dieterlen 387 (GRA, NH, PRE, SAM); Phillips 966

(SAM). Berea Hills: Guillarmod 419 (PRE). Quacha’s Nek: Houston s.n. (NH).

Drakensberg: Stokoe 1551 (PRE), Thaba Tsuen: Page s.n. (BOL).

Orange Free State. — Zastron: Maree 89 (PRE). Ficksburg: Galpin 13955

(BOL, PRE); Fawkes 228 (NBG). Senekal: Goossens 882 (PRE). Bethlehem, Clarens :

Van Hoepen s.n. (PRE). Harrismith, Witzieshoek: Junod s.n. (PRE).

Transvaal. — Rustenburg: Leendertz s.n. (PRE); Pegler 949 (BOL, PRE); Roe

in herb. Bolus no. 25035 (BOL). Marico, Zeerust, Rietfontein: Riekert s.n. (PRE).

Krugersdorp: Gilfillan in herb. Galpin 6053 (PRE); Scheerpoort: Obermeyer s.n.

(PRE). Lichtenburg: Sutton 316, 317 (PRE). Potchefstroom: Louw 1017 (PRE).

Witwatersrand : Moss 7122, 16911, 19704 (J). Brits: Pole Evans s.n. (PRE). Johan-

nesburg: Tucker s.n. (BOL); Gilfillan s.n. (PRE); Mass 13520 (J); Frankenwald: Cohen

in herb. Moss no. 21212 (J); Florida: Hutton 627 (GRA, BOL). Pretoria: Leendertz

691

s.n.; Verdoorn 150; Mogg 14118, 14122, s.n.; Smith 6244\ Repton 685; Codd 3064; Burtt-

Davy s.n.; Meeuse97>16\ Comins 861 (all PRE); Moss 4719 (J); Onderstepoort : Theiler

s.n. (PRE). Waterberg, Geelhoutkop: Breyer s.n. (PRE). Middelburg: Schlechter

3793 (BOL, GRA, PRE, NH). Belfast: Schlechter 3479 (BOL, GRA, PRE); Galpin

12456 (PRE). Heidelberg: Leendertz s.n. (PRE). Carolina: Galpin 3494 (PRE, BOL).

“Transvaal”: Burtt-Davy 711 (NH).

N.B.—A specimen Hutton 438 labelled “ Shafton, Howick, Natal ” (GRA) is

this species, but it is likely that it was wrongly labelled and was actually from the

Johannesburg area — no other specimen was ever recorded from Natal.

This species is treated in the Flora Capensis as a variety of C. capensis, but it is

not identical with C. plicatus Desr. This assumed identity was probably based on

Choisy’s identification (in DC., Prodr. 9, p. 410) of a Drege specimen (“ C. Thunbergii

a ” from Shiloh) with Sonnerat’s type specimen of C. plicatus. However, the Sonnerat

specimen, which is in my opinion a form of C. capensis (see no. 14), does not agree

at all with the Drege specimen “ C. Thunbergii a ” (in L) which I have studied.

The name “ Convolvulus Thunbergii ” was applied by Choisy to a specimen

collected by Burchell (he cites “ Burch.! cat. n. 1836 ”, but this is most probably a

mistake for no. 1839) and to a Drege specimen which is the type of C. inconspicuus

Hall. f. (see under C. capensis). Burchell 1839 is referable to C. multifidus Thunb.

(see no. 5).

The only plants to which the original description of C. thunbergii applies are those

referred to “ C. capensis var. plicatus ” in FI. Cap., so that the first name is adopted.

A type specimen could not be traced and Juel [in Plantae Thunbergianae (1918), p. 385]

mentions that he could not find a plant in the Thunberg herbarium to which Roemer

and Schultes referred when they cited one of his specimens.

The following interpretations of the name Convolvulus thunbergii have been

given : —

(a) E. Meyer applied it correctly to some Drege specimens, as is apparent from

the names on the original labels.

(b) Choisy used the name for specimens referable to C. multifidus and to “ C.

inconspicuus ” (= C. capensis).

(c) Hallier in Engl. Bot. Jb. 18 (1893), p. 102 followed Choisy and applied the

name to Burchell 1839 (which is C. multifidus). Later he corrected this on

a label on the type of C. multifidus Thunb. in the Thunberg herbarium (see

under C. multifidus).

(d) Baker and Wright in FI. Cap. 4, 2, p. 78 placed it as a synonym under C.

capensis var. plicatus which was due to a misinterpretation of Convolvulus

plicatus Desr.

C. thunbergii is, in my opinion, much more closely related to C. natalensis than

to C. capensis. For the differences between the first two, see under C. natalensis.

From C. capensis it can be distinguished by the shape of the leaves (which are ovate

to oblong in outline and pinnatisect or 5-9-lobed with the central lobe the longest,

in C. capensis they are undivided or palmately lobed or dissected, or the middle lobe is

linear or filiform), the smaller flowers (calyx 8-12 mm. long, corolla ± 20 mm. long,

in C. capensis usually calyx 7-12 mm. long and the corolla 20-40 mm.), the longer

stigmas (very short in C. capensis) and the seeds (smooth in C. thunbergii, muriculate-

tuberculate-rugose in C. capensis). In addition, C. capensis and C. thunbergii are

almost completely geographically separated.

692

14. C. capensis Burnt, /., Prodr. FI. Cap. (1768), p. 5; Choisy in DC., Prodr. 9 (1845),

p. 410; Hall.f. in Engl. Bot. Jb. 18 (1893), p. 105; Baker & Wright in Dyer, FI.

Cap. 4, 2 (1904), p. 78, exclus. vars. /S and y. C. plicatus Desr. in Lamk., Encycl.

3 (1789), p. 558; Choisy, op. cit. p. 410; Hall.f., op. cit., p. 106. C. alceifolius

Lamk., Encycl. Method., Bot. 1 (1791), p. 461; Choisy, op. cit., p. 410; Hall.-

f., op. cit., p. 105. C. altheoides Thunb., Prodr. FI. Cap. (1794), p. 35 (pro parte?),

non L. C. falkia Jacq., Hort. Schoenbr. 2 (1797), p. 38, t. 198; Choisy, 1. c.;

Hall.f., op. cit., p. 106, non Thunb. (1794). C. filiformis Thunb., FI. Cap. Ed.

2 (1818), p. 16 and Ed. Schult. (1824), p. 168; Baker & Wright, op. cit., p. 71,

non Desr. (1789). C. inconspicuus Hall.f., op. cit., p. 106, Baker & Wright, op.

p. 71. Merremia bowieana Rendle in Jl. Bot. 39 (1901), p. 63. Ipomoea bowieana

(Rendle) Baker in Dyer, FI. Cap. 4, 2 (1904), p. 52.

Type: In herb. Burman (G-Del.) without collector or precise locality (photo

in PRE).

Endemic.

This variable species can be divided into three more or less distinct varieties, which

are also more or less distinctly geographically separated:

Leaves palmately dissected with the lobes oblong to linear, obtuse or sometimes

acute, more or less equal or middle lobe longer, or the lower or all

leaves undivided, cordate-oblong to cordate-reniform, obtuse or rounded

at the apex with rounded basal lobes ; sepals obtuse or acute; peduncles

often long var. capensis.

Leaves undissected to shallowly lobed or rarely dissected, triangular-cordate

to almost sagittate, subacute to acuminate at the apex and with usually

acute basal lobes; sepals acute to acuminate; peduncles usually short var. plicatus.

Leaves, at least the upper ones, linear to filiform, hastate with much smaller

lobes at the base or auricled, rarely without lobes or auricles; sepals

obtuse, apiculate var. bowieanus.

14a. C. capensis var. capensis. C. capensis Burm. f., C. alceifolius Lamk., C. capensis

a dissectus Hall, f., op. cit., p. 105 and C. capensis j8 malvcefolius Hall, f., op. cit.,

p. 106, C. inconspicuus Hall. f.

Perennial. Stems climbing or sometimes prostrate, herbaceous, occasionally

suffruticose, ascending, slender, terete, up to at least 1 • 50 m. long, usually clothed (like

leaves petiole and calyx) with brown pubescence, glabrescent, more rarely glabrous

or nearly so (if so, leaves and petioles also less hairy). Leaves variable in shape on a

single specimen, the lower (= older ones) narrowly reniform to cordate-oblong or

almost hastate-oblong, rounded to subacute or emarginate at the apex, up to 35 mm.

long, with subentire to repand, crenate or dentate margin; petioles 5 mm. long;

upwards leaves becoming more and more pinnately incised or palmately 5-fid, the

uppermost usually consisting of 5 linear lobes of which the central one is the longest;

the lobes usually variously lobed or incised, the linear lobes of uppermost leaves less

so to entire, length of blade (1-) 3-5 (-7) cm., width (0-5-) 2-3 (-4-5) cm.; petiples

0-5-2 -5 (-4) cm. Peduncles 1- to cymosely few-flowered, usually slender, terete, less

hairy than the stem or sometimes glabrous, often 4-10 cm. long, sometimes longer,

rarely shorter (mainly in young or stunted specimens like those described as C. incon-

spicuous Hall, f.); bracteoles lanceolate, linear or subulate, minute or up to 8 mm.

long, hairy; pedicels slender, usually more densely pubescent than the stems (25-)

15-5 mm. long. Calyx silky-villous outside, rarely nearly glabrous (6-) 8 (-10) mm.

long; sepals broadly ovate to oblong, much imbricate, usually obtuse but often

apiculate or mucronate, chartaceous (outer ones) to somewhat membranous (inner

693

ones). Corolla funnel-shaped, white inside and pale pink outside with greenish mid-

petaline areas or pink to pale rose-colour, (15—) 20-30 (-35) mm. long and as much

in diam., brownish-strigose on the midpetaline areas. Ovary glabrous. Capsule

globose, glabrous, about 8 mm. in diam. Seeds dark brown, glabrous, verrucose,

4-5 mm. long.

Cape Province. — Namaqualand, near Leliefontein: Drege (“ C. thunbergii var”.;

isotypes of C. inconspicuus Hall, f., in L), near same locality: Esterhuysen 1370 (BOL,

PRE); Khamiesbergen : Esterhuysen 736a (BOL). Van RJiynsdorp: H. Andreae 484

in herb. Marloth (PRE), Acocks 14813 (PRE). Calvinia: Schmidt 363 (PRE), Story

4302 (PRE), Taylor 2814, Compton 9802, Maguire 187 (all NBG). Clanwilliam: Pappe

s.n. (SAM), Leipoldt 321 (BOL, SAM), Galpin 10544 (PRE), Pillans 9128 (BOL), Thode

A 2045 (PRE, NH), Esterhuysen 7138. (BOL), Gillett 4013 (BOL), Thorne s.n. (SAM),

Schlechter 10774 (BOL, GRA, L, PRE), Weintroub in herb. Moss 19490 (J), Steyn

389 (NBG), Barker 3615 (NBG, BOL); near Warmbad: Pearson 7252 (BOL). Piquet-

berg: Bolus 25027 (BOL), Pillans 8014 (BOL), Compton 9501, 15001, 15075, 15093

(NBG), 10904, 15025 (NBG, BOL), Esterhuysen 5520 (BOL), Guthrie 2663 (NBG),

Howes 175 (PRE). Piquetberg/Clanwilliam, Grey Pass: Steyn 372 (NBG). Malmes-

bury: Drege s.n. (“ C. alceifolius a ” and “ b ”, L), Pappe s.n. (BOL, SAM); Bachmann

88 (JE), 91 (BOL), Bolus 9971 (BOL, PRE), 25038 (BOL), Barker 4062 (NBG), Parker

4618 (BOL, PRE, NBG), Letty 60 (PRE). Tulbagh: Bolus 5211 (BOL). Wellington:

Thompson s.n. (PRE). Ceres: Guthrie 3374 (NBG); Paarl: Pappe s.n. (BOL), Ester-

huysen 9053 (BOL). Caledon: Bolus 25037 (BOL). Victoria- West: Thode A2169

(PRE); without precise locality: Specimen in herb. Burman (G-Del., photo in PRE,

type!); specimen in herb. Lamarck (P, photo in PRE, type of C. alceifolius Lamk.),

Zeyher 1231 (BOL, SAM).

Of these, Pappe s.n. from Clanwilliam (SAM) and Thorne s.n. from Clanwilliam,

among other ones, approach the var. plicatus and var. bowieanus, respectively.

14 b. C. capensis var. plicatus ( Desr .) Baker in Dyer, FI. Cap. 4, 2 (1904), p. 78, as

to new combination only, exclus. descr., synonyms and specimens cited. C. plicatus

Desr. ; C. falkia Jacq. non Thunb.

Type of variety: A specimen leg. Sonnerat in herb. Lamarck (P, photo PRE).

Very similar to C. capensis var. capensis but differs in the shapes of the leaves

which are ovate-cordate, ovate or triangular-cordate, toothed or crenate-serrate, more

rarely dissected, usually very acute, 1-4 cm. long and 0-75-2 cm. wide; petioles 0-5-1

cm., in the peduncles which are usually short, under 4 cm. long and in the sepals which

are usually narrower, acute or acuminate.

Cape Province. — Swellendam: Pappe s.n. (SAM, GRA), “ Ecklon" s.n., but

probably leg. V. Ludwig (BOL), Ecklon & Zeyher 70. 10 (GRA, L), Zeyher 3440 (SAM),

Bolus 25036 (BOL), Galpin 4352 (GRA, PRE), Thode A 2379 (PRE, BOL); Wurts 327

(NBG). Uitenhage: Ecklon 9 (SAM); without precise locality: Sonnerat s.n. in

herb. Lamarck (P, photo PRE), Pappe 33 (SAM).

The specimen Pappe s.n. from Swellendam is more or less intermediate between

this variety and the var. capensis.

14c. C. capensis var. bowieanus ( Rendle ) A. Meeuse, stat. nov. C. filiformis Thunb.

(1818), non Desr. (1789). Merremia bowieana Rendle in Jl. Bot. 31 (1901), p. 63.

Ipomoea bowieana (Rendle) Baker.

Type of variety: Bowie s.n. (BM, photo of type specimen in PRE).

694

Whole plant (except the calyx) usually much more glabrous than in the var.

capensis. Leaves linear, frequently with small basal auricles or hastate at the base

with bifid basal lobes sometimes almost filiform with revolute edges, 2-5 (-7) cm.

long, the lowermost often palmately 5-fid and dissected changing upwards into oblong

ones with hastate base, petioles generally under 5 mm. long. Peduncles as a rule under

4 cm. long; bracteoles often somewhat broader, linear-oblong or somewhat spathulate,

pedicels often densely hairy. Calyx usually densely hairy, rarely nearly or almost

completely glabrous, sepals usually broader and rounded or obtuse to emarginate,

minutely apiculate. Midpetaline areas densely hairy; otherwise as the var. capensis.

Cape Province.— Swellendam to George : “ On roadsides in the Districts of

Swellendam and George ”, Bowie s.n. (BM, type of variety, photo of type in PRE).

Bredasdorp: Esterhuysen 4449a (BOL). Riversdale: Schlechter 1834 (BOL, GRA),

Muir 2014 (BOL, PRE); George. Four cade 3425 (BOL). Uniondale: Four cade

1720 (GRA), 2105, s.n. (BOL), Esterhuysen 6811 (BOL), Compton 10535 (NBG).

Humansdorp: Four cade 2626, 5916 (BOL), Esterhuysen 6672 (BOL). Oudtshoorn:

Compton 21767, 23155 (NBG). “Uitenhage”: Ecklon & Zeyher 93.10 (GRA, JE,

L). Port Elizabeth Drege 7831a (L), Tyson 2273 (SAM), Paterson 146 (BOL, GRA,

KMG), 1065 (GRA), West 354 (BOL), I. L. Drege 103A (GRA, PRE) = ? 105 (SAM),

Long 883 (GRA, PRE). “ Klipdrift in Great Karroo ”, Schlechter 2276 (GRA, J).

Albany: Britten 5649, 5816 (GRA). Without precise locality: specimens leg. Thunberg

in Thunberg Herb., Uppsala (photos in PRE), type material of C. filiformis Thunb.

non Desr.

The specimens Muir 2014, Esterhuysen 6811 and some others form a transition

to the var. capensis.

The var. plicatus (Desr.) Baker of FI. Cap. is Convolvulus thunbergii R. et S. (see no.

13). The combination based on C. plicatus Desr., made by Baker, applies to the type

of the latter, but the synonyms, the description and quoted specimens must be excluded

C. capensis is geographically separated from C. thunbergii (except the var. bowieanus )

and although C. thunbergii was treated as a variety of C. capensis in Flora Capensis,

I am of the opinion that C. thunbergii is a very distinct species, much more close related

to C. natalensis than to C. capensis. C. capensis differs from C. thunbergii in the usually

longer and thicker peduncles, usually larger flowers, the less acute or obtuse sepals

(except in the var. plicatus ), the often palmately nerved to palmatisect leaves (pinnatisect

or penninerved in C. thunbergii ) and especially in the very short stigmas. The very

short stigmas account for the redescription of the var. bowieanus in the genus Merremia.

However, the generic features of this plant agree with Convolvulus (e.g. often pink

flowers; those of Merremia are orange or yellow to white, often with a dark centre)

and the stigmas are never quite globose, but even in the extreme case of the var.

bowieanus always flattened on the inner side. For the third variety, the epithet

bowieanus was taken up, because Thunberg’s older name Convolvulus filiformis (1818)

is invalidated by C. filiformis Desr. (1789).

Especially in the var. capensis, several types of leaves are often found on one

specimen. The variation in leaf-shape is mainly responsible for the various synonyms

under which it was redescribed. In the var. capensis the first leaves formed are undis-

sected as I could observe on seedlings and young cultivated plants grown in Pretoria.

If the plants are depauperated or cannot climb they produce flowers before any

dissected leaves are formed. Such specimens are for instance the type of C. inconspicuus

Hall. f. (leg. Drege) and Esterhuysen 1370. Later, especially on vigorously growing

stems, dissected leaves are formed. Specimens showing only undissected leaves often

have short peduncles, whereas specimens which show only palmatifid leaves often

have long, not infrequently few-flowered peduncles, so that these extremes look very

695

different. However, there are many herbarium specimens which show the transition

in leaf-shape on a single stem and the observation of plants grown from seed confirms

the identity. This tendency to develop dissected leaves towards the tips of the stems

is also present, though less pronounced in the var. plicatus and in some other South

African species of Convolvulus ( C . dregeanus, C. aschersonii).

Another tendency is the reduction of the lateral lobes and the predominance of

the central lobe of the leaves. This is frequently seen in the var. capensis and in some

cases the leaves resemble those of the var. bowieanus very much. The reduction of

the lateral lobes is extreme in some specimens of the var. bowieanus in which the

reduction can be so complete that the leaves are linear to filiform, such as in the type

specimens of C. filiformis Thunb. and Merremia bowieana Rendle.

The specimens here referred to the var. bowieanus are usually densely hairy on

the calyx and the stems and peduncles are rather stout, but in some specimens such

as the type of C. filiformis Thunb. the stems and peduncles are slender and the calyx

can be almost completely glabrous as in some specimens from Oudtshoorn ( Compton ,

21767, 23155). These specimens resemble C. bidentatus in the glabrous calyx and narrow

leaves, but they can be distinguished by the absence of the hastately spreading basal

lobes of the leaf as found in C. bidentatus, by the usually 1 -flowered peduncles (often

2-flowered in C. bidentatus) and also by the shape of the sepals which are in addition more

herbaceous and not distinctly membranous along the edges as those of C. bidentatus.

15. Convolvulus arvensis L., Sp. PI. ed. 1 (1753), p. 153; Choisy in DC., Prodr. 9

(1845), p. 406; Baker and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 75; Baker

and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 97; Phillips, Weeds of S. Africa

(1938), p. 44, fig. 84.

Type : Linne’s original description was completely or mainly based on pre-Linnaean

works, but the species is also represented in the Linnaean herbarium and the preserved

specimen can be regarded as the type.

Perennial herb forming several to many annual stems from a long taproot. Stems

prostrate or twining, 0-75-1-75 m. long, angular, sparsely pubescent to glabrous.

Leaves often secund, entire, ovate-oblong, oblong or lanceolate with hastate or

sagittate base, usually obtuse and mucronate at the apex, 2-5 cm. long, glabrous or

thinly hairy; petioles shorter than the blades. Flowers axillary, solitary or sometimes

in 2-3- (occasionally more) -flowered cymes; peduncles angular, shorter or longer

than the leaves; bracteoles linear, about 3 mm. long, pedicels always much longer

than the cal>x. Sepals slightly unequal, 3-5-5 mm. long; the out ones a little shorter,

eliptic-oblong, obtuse, often shortly ciliate, glabrous or hairy ; inner ones broader,

to almost orbicular, obtuse to slightly retuse, more or less distinctly mucronulate,

usually glabrous. Corolla white or pink, white ones sometimes with pink or red

mid-petaline areas, broadly funnel-shaped, glabrous, except at the very tips of the

mid-petaline areas, 1 -5-2-5 cm. long and 2-3 cm. in diam., the limb shallowly lobed.

Stamens slightly unequal; filaments with broadened base which is papillose at the

margins. Ovary glabrous. Capsule ovoid-globose, glabrous, 5-8 mm. long. Seeds

dark brown or black, glabrous, minutely verrucose-rugose.

Originally a native of Europe and parts of Asia, but now a common weed in

temperate and subtropical areas throughout the world, rarely in the tropics. As a

weed it frequently occurs in grain lands.

In S. Africa it is recorded from all Provinces and very common in several areas,

where it is often a serious pest.

696

Doubtful Species of Convolvulus.

C. burmannii Choisy in DC., Prodr. 9 (1845), p. 405; Baker 3c Rendle in Dyer,

FI. Cap. 4, 2 (1904), p. 71.

This species was described by Choisy from a specimen in Burman’s herbarium

(Geneva-Delessert). However, when upon my request Professor Baehni tried to

locate the specimen, he could not trace it. He reported that the specimen must already

have been lost about 1890 when Hallier revised the Convolvulaceae in the Geneva

herbarium, because Hallier did not treat this species in his paper on Convolvulaceae

Africanae (1893). Hallier evidently only knew Choisy’s description, which is very

short and incomplete, and he suggested in Engl. Bot. Jb. 18 (1893), p. 147 that it might

be referable to “ Ipomoea plantaginea ” (= I. simplex ).

I cannot recognise Choisy’s plant from the description but, if the specimen was

indeed collected in S. Africa in Burman’s time, there can hardly be any other species

but I. simplex to which it could be referred. The pubescence of the stems and leaves

indicated by Choisy in conjunction with the fruticose stem also suggest Turbina oeno-

theroides, but the sepals are stated to be glabrous.

Even if it is to be regretted that this species remains doubtful, it can hardly be

of any importance from a nomenclatural point of view, because practically all other

species of Convolvulus and Ipomoea which occur in the area explored in Burman’s

time have older names or contemporary names (also given by Choisy in 1845), whereas

the species described later (such as C. galpinii, I. pellita) are altogether different from

the description of C. burmannii.

To BE EXCLUDED FROM THE SOUTH AFRICAN FLORA.

C. petraeus Lee ex Choisy in DC., Prodr. 9 (1845), p. 413. This species was

entirely overlooked by Hallier in his treatment of the African Convolvulaceae and is

not mentioned in the Flora Capensis either.

Professor Baehni kindly sent a photograph of the type specimen preserved in

Geneva. He added the information that the specimen had been filed in a wrong place

many years ago and Hallier, who studied the Convolvulaceae at Geneva about 1890,

did not see the specimen.

The photograph of the type showed that this plant is not identical with any South

African plant. Mr. de Winter, who visited the Geneva herbarium in 1953, was able

to solve this problem. The type of C. petraeus is a specimen of C. massonii Dietr.

a species occurring in Madeira and Teneriffe. Masson probably sent his plant, or seeds

of it, to Lee, who erroneously regarded it as South African and sent it as such to Choisy.

Choisy did not recognise the identity because the type or isotype of C. massonii (also

a Masson specimen !) studied by Choisy is a mature twig, whereas the type of C. petraeus

is from quite a young plant. Other specimens from Madeira show that the difference

in general appearance between the two types is due only to differences in maturity.

C. petraeus must, therefore, be exluded from the South Africa flora.

The synonymy is as follows: C. massonii Dietr., Lex. Nachtr. 2 (1816), p. 377

(“ Massoni ”); Choisy in DC., Prodr. 9 (1845), p. 413; Hall f. in Engl. Bot. Jb. 18

(1893), p. 109. C. petraeus Lee ex Choisy, op cit., p. 413.

8. CALYSTEGIA

R. Br., Prodr. FI. Nov. Holl. ed. 1 (1810), p. 483, nomen gener. conserv.; Choisy

in DC., Prodr. 9 (1845), p. 433; Benth. & Hook., Gen. Plant 2 (1876), p. 874; Peter

in Engl. u. Prantl., Natiirl Pfl.fam., ed. 1, 4- 3a (1891), p. 36; Hall. f. in Engl. Bot.

Jb. 16 (1893), p. 580, and 18 (1893), p. 110; Baker & Rendle in Dyer, FI. Trop. Afr.

697

4-2 (1905), p. 99; Phillips, Gen. S. Afr. Flow. PL, Ed. 1 (1926), p. 511; Van Ooststr.

in Blumea 3 (1939), p. 284, and in Van Steen., FI. Males. Ser. 1, 4-4 (1953), p. 437.

Volvulus Medic., Phil. Bot. 2 (1791), p. 42 and in Staatsw. Vorl. Churf. Phys. Oek.

Ges. 1 (1791), p. 202. Convolvulus L., Auct. pro. parte; Salter in Adamson & Salter,

FI. Cape Penins. (1950), p. 683.

Type Species: Calystegia sepium (L.) R. Br.

Characters generally as in Convolvulus, but bracteoles large and clasping the calyx.

Corolla medium-sized to large, white or pink. Pollen globose, smooth. Ovary 1 -celled

or imperfectly 2-celled, 4-ovuled; stigmas 2, oblong or elliptic, flattened. Seeds black,

smooth or verrucose.

About 25 species in the temperate and tropical regions of both hemispheres. Two

species in South Africa:

Leaves hastate-sagittate, more or less acute; climbing plant with subacute

bracts, introduced (Cape Peninsula) 1. C. sepium.

Leaves reniform, obtuse ; prostrate plant with obtuse bracts, growing on loose

calcareous sand near beaches (only records from Riversdale and

Knysna) 2. C. soldanella.

1. C. sepium (L.) R.Br., Prodr. FI. Nov. Holl., ed. 1 (1810), p. 483; Choisy in DC.,

Prodr. 9 (1845), p. 433. Convolvulus sepium L., Sp. PL, Ed. 1 (1753), p. 153,

Salter in Adams, and Salter, FI. Cape Penins. (1950), p. 685.

lype: The original description was based on Bauh. Pinax 294, but the species

is also represented in the Linnean herbarium and the specimen may be taken to represent

the lecto-type.

A perennial climbing herb. Stems terete, 1-3 m. long. Leaves herbaceous,

triangular in outline, sagittate or hastate-sagittate, 4-8 (-12) cm. long and 2-5 (-6)

cm. wide at the base; apex acute or acuminate; basal lobes slightly divergent, obliquely

truncate or acute, often more or less angulate; petioles 1-4 (-5) cm. long. Peduncles

1 -flowered, longer than the leaves ; bracteoles ovate or ovate-cordate, acute or subobtuse,

about 18 mm. long, longer than the calyx and more or less concealing it. Sepals

broadly lanceolate, very pale green, up to 10 mm. long. Corolla funnel-shaped, white

or pink (with white stripes), 5-5-5 cm. long. Capsule subglobose.

A native of Europe and probably also N. America, introduced as a weed elsewhere.

Cape Province. — Cape Peninsula, railway side S. of Steenberg Station: Salter

889 (BOL), Muizenberg, Sand Vlei: Moss 7497 (J, BOL).

2. C. soldanella ( L .) R.Br. ex R. et S., Syst. Veg. 4 (1819), p. 184; Choisy in DC.,

Prodr. 9 (1845), p. 433; Hall. f. in Engl. Bot. Jb. 18 (1893), p. Ill; Muir in Kew

Bull. 1934, p. 44—45. Convolvulus soldanella L., Sp. PL Ed. I (1753), p. 159.

Type: Linnaeus based this species on Bauhin’s Pinax, but the species is also

represented in the Linnean herbarium and this specimen may be taken to represent

the lecto-type.

Straggling perennial herb, growing in calcareous sand near the sea shore. Rhizome

10-60 cm. long, producing short upright slender stems. Leaves somewhat fleshy,

renifirm, with deep basal sinus and rounded basal lobes, 1-4 cm. long and about

as wide, but usually about 2 cm. diam., obtuse or emarginate, more or less repand;

basal lobes rounded, petioles usually longer than the blades. Peduncles mostly longer

than the leaves, sharply quandrangular; bracteoles broadly oblong, rounded at the

apex, slightly shorter than the calyx, 10-15 mm. long. Corolla 2-4-4 cm. long, pink,

mauve or pale purple. Capsule ovoid, acute.

698

W. Europe to the coast of the North Sea and the Baltic; N. Africa; Asia; N. and

S. America; Australia; S. Africa, always in calcareous loose sand near the sea shore,

its seeds distributed by sea currents.

This species was recorded by Muir from Riversdale near Morris Point (Muir,

l.c.); the only other record is the following:

Cape Province. — Buffalo Bay (Knysna): Keet 861 (PRE, GRA).

The combination Calystegia soldanella is generally attributed to Robert Brown,

but Brown only mentioned (in Prodr. FI. Nov. Holl. (1810), p. 483) that in his opinion

Convolvulus soldanella L. belongs in Calystegia and did not actually make the com-

bination. The first time the combination occurs is in Roemer et Schultes, Syst. Veg.

4 (1819), attributed to Robert Brown, and the correct citation is, therefore, Calystegia

soldanella (L.) R.Br. ex R. et S.

9. HEWITTIA

Wight et Arnott in Madr. Jl. Sci., ser. 1, 5 (1837), p. 22; Benth. et Hook., Gen.

PI. 2 (1876), p. 873; Peter in Engl. u. Prantl, Nat. Pfl. fam., ed. 1, 4-3a (1891), p. 32;

Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 68: Baker & Rendle in Dyer, FI.

Trop. Afr. 4-2 (1905), p. 100, Phillips, Gen. S. Afr. Flow. PI. Ed. 2 (1951), p. 623;

Ooststr. in Steen., FI. Males. Ser. I, 4, 4 (1953), p. 438. Shutereia Choisy in Mem.

Soc. Phys. Geneve 6 (1833), p. 485, t. 2, fig. 11, et in DC., Prodr. 9 (1845), p. 435;

Van Ooststr. in Blumea 3 (1939), p. 286, non Shuteria Wight et Arnott (1834), q.e.

nomen conservandum.

Herbaceous, pubescent, twining or prostrate. Leaves entire, angular or lobed,

cordate at the base. Flowers axillary, solitary or in few-flowered subcapitate cymes;

bracteoles oblong or linear-lanceolate, acuminate, at some distance from the calyx.

Sepals 5, acute, herbaceous. Corolla medium-sized, campanulate to funnel-shaped,

5-angled. Stamens 5, included; filaments linear with dilated base; pollen smooth.

Disc annular. Ovary hairy, 1-celled or imperfectly 2-celled at the apex, 4-ovuled,

style simple, included; stigmas 2, ovate-oblong. Capsule 1-celled, 4-valved, 4- or by

abortion less-seeded. Seeds black, glabrous, opaque.

The only species, H. sublobata (Linn, f.) O.Ktze., is found in tropical Africa,

southwards to Natal, N. Transvaal and S.W. Africa; also in tropical Asia, Malaysia

and Polynesia.

Although Shutereia Choisy (1833) need not necessarily be considered to be an

orthographical varient of Shuteria W. et A. (1834), it is better to consider it to be

illegitimate on account of the nomen conservandum Shuteria W. et A. (Leguminosae).

Moreover, Hewittia has been used by most authors and is retained here.

H. sublobata {Linn, f.) O.Ktze ., Rev. Gen. (1891), p. 441; Hall. f. in Engl. Bot. Jb. 18

(1893), p. Ill; Hutch. & Dalz., FI. W. Trop. Afr. 2 (1931), p. 212; Van Ooststr.

in Van Steen., FI. Males. Ser. 1, 4, 4 (1953), p. 438; Brenan in Mem. New York

Bot. Garden 9 (1954), p. 9. Convolvulus sublobatus Linn, f., Suppl. (1781),

p. 135. Convolvulus bicolor Vahl, Symb. 3 (1794), p. 25, non Desr. (1789).

Shutereia bicolor (Vahl) Choisy, op. cit. (1833), p. 48, t. 2, fig. 11; (1845),

p. 435. Hewittia bicolor (Vahl) Wight et Arn., op. cit. (1837), p. 22; Baker

& Wright, op. cit., p. 68; Baker & Rendle, op. cit., p. 100, nomen illeg.

Shutereia sublobata (Linn, f.) House in Bull. Torrey Bot. Cl. 33 (1906), p., 318

Van Ooststr. in Blumea 3 (1939), p. 287.

Type: No recognised type specimen could be located.

699

Stems slender, 1-2 m. long, angular, usually more or less pubescent, occasionally

rooting. Leaves oblong or ovate to broadly ovate in outline, adpressed-pubescent

to nearly glabrous, cordate or sometimes truncate at the base; the auricles entire or

angular, occasionally spreading and blade more or less hastate; the apex acuminate

to obtuse, mucronulate: the^edge entire or grossly dentate; blade 3-12 cm. long and

4-10 cm.- wide: petiole pubescent, 1-6 cm. long. Peduncles 1-10 cm. long, pubescent,

1-flowered or occasionally bearing 2-3 flowers in a dense head; bracteoles oblong-

lanceolate or narrower, much exceeding the very short, up to 3 mm. (in fruit 5 mm.)

long pedicels. Sepals more or less hairy and ciliate, unequal, outer ones much larger,

more or less ovate, 9-15 mm. long, the third more or less oblique, inner two smaller,

ovate with broadened and scariously margined base, 7-7-5 mm. long. Corolla 2-2-5

cm. long, cream or yellow, usually with a maroon or purple “eye”, the limb with,

5 very short, rounded, emarginate, mucronulate lobes; midpetaline areas pilose outside.

Ovary densely hairy with long white hairs, also a few long hairs on the basal part of the

style. Capsule depressed-globose to more or less quadrangular, crowned by the

persistent style, pilose, about 8 mm. long and 10 mm. in diam. Seeds 4-2, black,

opaque, glabrous except the pubescent hilum, 5-6 mm. long.

This well-known species was recorded from the following districts:

S.W. Africa. — The northernmost part near the Cunene River (Ovamboland)

and near the Okavango.

Transvaal.— Barberton, Nelspruit, Letaba, Zoutpansberg.

Natal. — From Zululand to Durban and along the coast southwards, extending

into the E. Cape (Port St. Johns). Also rather frequent in Angola and Portuguese

East Africa, but the only Rhodesian specimens I have seen came from the eastern border

near Melsetter.

10. JACQUEMONTIA

Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 476 and in DC., Prodr. 9 (1845),

p. 396; Benth & Hook., Gen. PI., 2 (1876), p. 874; Peter in Engl. u. Prantl, Natiirl.

Pfll. farm, Ed. 1, 4-3a (1891), p. 33; Hall. f. in Engl. Bot. Jb. 16 (1893), p. 578; Baker

& Wright in Dyer, FI. Cap. 4, 2 (1904), p. 69; Baker & Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1905), p. 85; Phillips, Gen. S. Afr. Flow. PI., Ed. 2 (1951), p. 622; Van

Ooststr. in Van Steen., FI. Males. Ser. 1, 4-4 (1953), p. 431.

Type species: Probably J. ferruginea (Steud.) Choisy (Brazil).

Herbaceous to woody, twining or prostrate, usually hairy, with stellate hairs.

Leaves variable, often cordate at the base, entire, rarely lobed. Flowers in axillary,

pedunculate, umbellate or capitate cymes, with or without an involucre; bracteoles

small, linear to lanceolate or the outer ones larger, foliaceous; pedicels very short

or 0. Sepals 5, often unequal. Corolla small to medium-sized, funnel-shaped or

campanulate, blue, mauve or pink, rarely white: the limb 5-toothed or nearly entire,

rarely lobed; midpetaline areas distinct. Stamens included, filaments linear, anthers

oblong; pollen smooth. Disc small or none. Ovary 2-celled, 4-ovuled; style simple,

included; stigmas 2, ovate or oblong, more or less flattened. Capsule globose, 2-celled,

4- or 8-valved, 4- or by abortion less-seeded. Seeds usually glabrous.

Species about 120, mainly American, a few in tropical Africa, Asia and Australia;

one species in South Africa.

700

J. tamnifolia ( L .) Griseb., FI. Br. W. Ind. (1861), p. 474; E. A. Bruce in Kew Bull.

1940, p. 63. /. capitata (Desr.) G. Don, Gen. Syst. 4 (1837), p. 283; Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 95, Wood & Evans, Natal PI. 1 (1899), t. 13; Baker

& Wright, op. cit., p. 69; Baker & Rendle, op. cit., p. 85; Hutch & Dalz., FI.

W. Trop. Afr. 2 (1931), p. 211.

Type: Linnaeus based this species on the figure in Dill. Hort. Elth. p. 428, t. 318,

f. 410 (1732).

Annual. Stems several from the base, twining or trailing, occasionally suberect,

up to about 75 cm. long, finely pilose with usually brownish, rarely whitish pubescence.

Leaves ovate, oblong to broadly cordate, 4-9 cm. long, 2-8 cm. wide, entire : the base

shallowly cordaie to truncate or occasionally abruptly cuneate; the apex acuminate or

acute, the blade glabrescent or more or less pilose with brownish or white hairs, ciliate;

petioles slender, shorter than the corresponding blades, usually much more densely

pilose than the blade or stem. Peduncles usually longer than the leaves, terete, often

pilose, bearing at the apex a dichotomously forked condensed cyme which forms a

dense globose head up to 3 cm. in diam., bracteated by reduced leaves with narrowing

base, the inner ones becoming smaller and more hairy ultimately resembling the sepals.

Sepals about 5 mm. long, subequal, lanceolate, acute, shaggy with soft ferrugineous

or rarely white hairs. Corolla blue, very rarely mauve to white, fugacious, funnel-

shaped, 5-angled, obscurely 5-lobed, about 10 mm. long, glabrous; midpetaline areas

conspicuous. Capsule globose, 4-5 mm. in diam., glabrous. Seeds usually 4, bright

brown, 2-2-5 mm. long, glabrous, finely scabrid-rugose.

Its known distribution is America (Southern United States, Central America,

West-Indies, Guyana, etc.), tropical and South Africa, Mascarenes. In South Africa

recorded from: South West Africa (Northernmost part); Bechuanaland (Caprivi-

strip and Ngamiland); Transvaal (Zoutpansberg, Letaba, Nelspruit, Barberton);

Natal and Zululand, along the coast South to the Umlaas River.

The African specimens had always been referred to as J. capitata , but Hallier in

Med. Rijksherbarium Leiden 35 (1918), p. 15, already reduced/, tamnifolia to/, capitata

(he maintained the name /. capitata, probably on account of the “ Kew-Rule ’")• Miss

E. A. Bruce independently came to the conclusion that they are identical (cf. Kew

Bull. 1940, p. 63, for additional details and full synonymy).

11. MERREMIA

Dennst., Schliiss. Hort. Malab. (1818), p. 12, 23 nomen nudum; ex Hall. f. in

Engl. Bot. Jb. 16 (1893), p. 581, descr.; Peter in Engl. u. Prantl, Nat. Pfl. fam. 4- 3a,

Nachtrage (1895), p. 377; Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 101 ;

Ooststr. in Blumea 3 (1939), p. 292, and in Steen., FI. Males. Ser. I, 4-4 (1953), p. 439;

Phillips, Gen. S. Afr. Flow. PI., Ed. 2 (1951), p. 623. Ipomoea, pro parte, Auct.; Baker

and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 46, exclus. type.

Type species: Merremia convolvulacea Dennst., l.c = M. hederacea (Burm. f.)

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 118 = Evolvulus hederaceus Burm. f., FI. Ind.

(1768), p. 77, t. 30, f. 2.

Herbaceous or woody twiners, or prostrate, rarely suberect. Stems terete, more

rarely winged. Leaves entire, dentate, auricled, lobed or palmately or pedately partite

to compound. Flowers axillary, solitary or in few- to many-flowered variously ramified

inflorescences, but usually cymose, bracteoles usually small. Sepals 5, usually subequal,

acute to rounded, narrow to orbicular, herbaceous to coriaceous, in several species

accrescent in fruit. Corolla funnel-shaped or campanulate, white or yellow to orange,

often with darker brownish or purplish centre, usually with distinct midpetaline areas.

701

the limb faintly 5-lobed, occasionally distinctly so. Stamens included; filaments filiform,

often unequal; anthers often spirally contorted, pollen espinose, ellipsoid. Disc

saucer-shaped. Ovary 2-celled or 4-celled, 4-ovuled; style filiform; stigma biglobose,

Capsule usually dehiscing by 4 valves, sometimes also circumscissile at the base, or

irregularly dehiscing, 4- to 1 -celled. Seeds 4, or less by abortion, glabrous or pubescent,

especially on the angles.

Species about 80, widely spread in the tropics of both hemispheres.

Van Ooststroom (1939), p. 293, has pointed out that Merremia Dennst. ex Hall. f.

can be maintained, because the older synonyms Skinneria Choisy (1833) and Spiranthera

Boj. (1837) are illegitimate, being later generic homonyms of Skinneria Forsk. (1776)

and Spiranthera St. Hil. (1823), respectively, so that Merremia stands.

Generally speaking, the species of Merremia have white, cream or yellow flowers,

often with a dark (reddish, purplish or brownish) centre. This is in striking contrast

to the genus Ipomoea, which often has pink, mauve to purple or magenta flowers and

in which yellow flowers are of rare occurrence.

The South African species belong to the sections Xanthips (Griseb.) Hall. f. (species

1-3) and Streptandra Hall. f. (species 4-9).*

Leaves pinnatisect, usually with patent hairs; flowers small: calyx 6-8 mm.

long, corolla 7-8 mm. long

Leaves palmately compound, entire, auricled at the base or, if pinnatisect,

leaves glabrous and flowers much larger:

Main stems distinctly winged; ultimate branches 4-ribbed-4-angled;

calyx glabrous, coriaceous, concolorous:

Leaves palmately 3-5 (-9)-lobed, perennial woody climber

Leaves cordate-orbicular to cordate-lanceolate, entire

Main stems not distinctly winged, or, if more or less angular, calyx

inflated, plicate-ribbed, with purple longitudinal stripes and purple

dots:

Calyx inflated, plicate-ribbed, the ribs purplish, sepals herbaceous,

pubescent, unequal; annual

Calyx not plicate-ribbed, concolorous; sepals smooth, glabrous or

sometimes pubescent, often coriaceous:

Leaves deeply palmatisect or pinnatisect to bipinnatisect, or lobes

of palmate leaf more or less dissected:

Calyx 7-10 mm. long; sepals elliptic, rounded at the apex,

pale green with membranous edges; corolla 2-2-5 cm.

long, 3-4 cm. in diam

Calyx 14-20 mm. long, sepals obtuse or acute but not

rounded at the apex, drying brownish; corolla 4-6

cm. long the limb 5-7 cm. in diam.:

Leaves herbaceous, palmately 5-7 lobed, 6-15 cm. in

diam., the lobes entire to pinnatisect; tall woody

climber

Leaves somewhat fleshy, irregularly palmately-pinnately

dissected to bipinnatisect. 3-7 cm. in diam.;

stems herbaceous, annual from perennial root-

stock (S.W. Africa)

* When the manuscript was ready for the press, Dr. R. Story collected another species in the

Kaokoveld, South West Africa. The material was in a very poor condition when collected and it

could not be named with absolute certainty. It is most probably M. quercifolia Hall, f., thus far only

known from Angola.

8 M. pinnata.

1 . M. pterygocaulos.

see Opercul'ma

turpethum.

2. M. verecunda.

3. M. palmata.

4. M. kentrocaulos.

5. M. bipinnipartita.

1332796-5

702

Leaves not deeply dissected (sometimes palmately lobed), often

auricled at the base:

Leaves deltoid-ovate to subreniform, more or less lobed or

crenate with more or less cordate base, hairy below

(only recorded from Eastern Cape) 6. M. malvaefolia.

Leaves linear or lanceolate to oblong, usually hastate or

auricled at the base:

Flowers up to about 16 mm. long; calyx glabrous;

basal auricles of leaves usually with several acute

teeth 7. M. tridentata ssp.

angustifolia.

Flowers 20-35 mm. long; calyx usually densely hairy

outside; basal auricles of leaves entire or bifid See Convolvulus capen-

sis var. bowieanus.

1. M. pterygocaulos ( Choisy ) Hall.f. in Engl. Bot. Jb. 16 (1893), p. 552 and 18 (1893),

p. 113; Baker and Wright in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 105; Hutch, and

Dalz. , FI. W. Trop. Afr. 2 ( 1 93 1 ), p. 2 1 2. Convolvulus pterygocaulos Steud. , cheironym

(Schimper, PI. Abyss., 2, No. 630, printed herbarium label). Ipomoea pterygocaulos

(Steud. ex) Choisy in DC., Prodr. 9 (1845), p. 381. Ipomoea tetraptera Baker

in Dyer, FI. Cap. 4, 2 (1904), p. 65. Merremia tetraptera (Baker) Hall. f. in Meded.

’s Rijksherb. Leiden 1 (1910), p. 21.

Type: Schimper 630 from Abyssinia in herb. Geneva; isotypes at K and probably

elsewhere.

Perennial, more or less shrubby climber, glabrous in all its parts except the corolla.

Main stems with 4 membranous wings, ultimate branches slender, 4-angled. Leaves

cordate-ovate to suborbicular in outline, 3-15 cm. long and wide, palmately 3-7 (-9)-

lobed to about the middle, the lobes usually ovate, acute or cuspidate and mucronate,

entire or subrepand; petioles more or less 4-angled, shorter than the blade. Peduncles

up to 15 cm. long, but often much shorter; more or less 4-angled, cymosely few-

flowered or (by reduction) rarely 1 -flowered; bracteoles minute, linear; pedicels up

to about 2 cm. long, ultimately distinctly angled, thickened, subclavate and remaining

erect in fruit. Calyx 9-11 mm. long; sepals obovate-oblong or oblong, obtuse, much

imbricate, chartaceous, pale yellowish green, accrescent and becoming broadly ovate

to orbicular and ultimately spreading in fruit. Corolla broadly funnel-shaped, pale

yellow, cream or white, densely long-silky with glistening silvery hairs on the mid-

petaline areas outside, 2 • 5-3 cm. long and 3 • 5-4 • 5 cm. in diam. Capsule ovoid-conical,

more or less truncate or flattened-depressed at the apex and crowned with the persistent

style-base, 12-15 mm. long and about as wide at the base, glabrous. Seeds black,

smooth, glabrous, about 7 mm. long when quite ripe.

Distribution. — Practically the whole of Africa south of the Sahara, as far north

as Abyssinia, as far south as Angola and Southern Rhodesia, extends into the Eastern

Cape Province.

Cape Province. — Kentani, Quora Bridge: Acocks 12295 (PRE).

Natal. — Palmiet nr. Durban: Wood 7542 (NH, one of the original numbers

cited by Baker under I. tetraptera) — Camperdown: Franks in Herb. Wood no. 11077,

(BOL, PRE, SAM) — Kranskop: McKen 1 (NH, also cited by Baker) — “ Bank of Little

Tugela”; Wood 3500 (NH, also cited by Baker).

Upon examination of original specimens of Ipomoea tetraptera Baker and a

considerable number of specimens referred to M. pterygocaulos (among them several

authenticated or frequently cited ones), I came to the conclusion that I. tetraptera is

identical with the wide-spread African species.

M. pterygocaulos seems to be rare in Southern Africa and has been recorded only

a few times.

703

2. M. verecunda Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 110; Pilger in Engl-

Bot. Jb. 48 (1912), p. 349; A. Meeuse in R. A. Dyer, Flow. PI. Afr. 30 (1955),

pi. 1193. Ipomoea quinquefolia Hochst. ex Hall. f. var. pubescens Baker in Dyer,

FI. Cap. 4, 2 (1904), p. 66. Ipomoea verecunda (Rendle) N.E.Br. in Kew Bull.

1909, p. 123.

Type: Not designated. Baker (in fl. Cap.) quoted several specimens under

Ipomoea quinquefolia var. pubescens (such as Burke 413 and Muskett in Herb. Bolus

9285). Rendle (l.c.) mentioned only two specimens {Mrs. Lugard 134 and Lugard

227) from Bechuanaland. Considering that Rendle did not mention a single specimen

from S. Africa and that the description of I. quinquefolia var. pubescens is very vague,

he described a new species altogether and the original specimens, therefore, are the

two gatherings from Bechuanaland. I have seen a sheet of Lugard 134 and consider

this number to be the lecto-type (BM, photograph in PRE, isotype in GRA).

Annual. Stems usually several from the base, herbaceous, green, prostrate or

twining, more or less angular, sulcate or minutely winged, glabrous or thinly hairy.

Leaves deeply palmately to pedately 7-9 (-ll)-sect, 2-8 (-10) cm. in diam, glabrescent;

the lobes linear-oblong-oblanceolate or in the lowermost leaves broader, more or less

apiculate or mucronate, contracted at the base, the lateral ones gradually smaller;

petioles 2-6 cm. long. Peduncles 1-flowered or occasionally cymosely 2-3-flowered,

0-5 cm. long; bracteoles 2 at the base of each pedicel, linear or linear-lanceolate,

5-8 mm. long; pedicels varying in length from about 0-5 cm. (if peduncles long) to

about 3 cm. (if peduncles short), at first erect or patent, ultimately reflexed. Calyx

turbinate, in flower 10-15 mm. long; sepals unequal, herbaceous, subobtuse to

acuminate; outer three very concave, pale green with 6-7 well marked longitudinal

purplish red veins, between the veins deeply sulcate-plicate and with small purplish

spots, hairy on the nerves, 5-7 mm. wide; inner ones shorter and narrower, 2-4 mm.

wide, less concave, without or with only a few purplish stripes, not sulcate or plicate,

acute or acuminate even if the outer ones are obtuse, in fruit hardly accrescent in

contradistinction to the much accrescent (up to 22 mm. by 14 mm.) outer sepals.

Corolla widely funnel-shaped, 15-20 mm. long and 20-30 cm. in diam., light yellow

with a purplish-red “eye”; the limb shallowly 5-lobed-5-angled; midpetaline areas

not distinct, glabrous or thinly pubescent towards the base. Capsule completely

enclosed by the inflated turbinate calyx, 4-lobed and somewhat depressed if 4-seeded

(3-lobed, 2-lobed or ovoid with only 3, 2 or 1 seed, respectively), 8—12 mm. in diam.

when 4-seeded; pericarp very thin, membranous, scarious white, subhyaline, irregularly

dehiscent to 4-valved. Seeds black, shining through the thin pericarp, 5-6 mm. long,

long, 4-5 mm. broad, glabrous except for lines of very short velvety pubescence on

the angles and the shortly pubescent hilum; the areas between the lines of hairs

microscopically pitted.

Endemic but may extend into southern Angola and western Southern Rhodesia^

Recorded from: S.IV. Africa, S. to Gibeon, N. at least to Namutoni; Griqualand-

W.: Hay, Barkly-W., Kimberley, Vryburg, Hopetown; Orange Free State: Winburg,

Bloemfontein, Fauresmith; Transvaal: Christiana, Bloemhof, Wolmaransstad, Rus-

tenburg, Ventersdorp, Potchefstroom, Johannesburg, Vereeniging, Brits, Pretoria,

Bronkhorstspruit, Groblersdal, Witbank, Lydenburg, Pietersburg; Bechuanaland:

Mochudi and Ngamiland; Basutoland: Leribe.

The calyx is very characteristic with its purplish raised nerves (it resembles that

of Hibiscus trionum). The seeds become very rugose-papillose after boiling because

the outermost layer of the testa swells to form yellowish-grey, sinuous folds, which

persist after drying and can then easily be rubbed off.

This species is fairly wide-spread, but not very frequent or perhaps often overlooked.

Only about 60 specimens were found in all the S. African herbaria.

704

3. M. palmata Hall, f in Engl. Bot. Jb. 18 (1893), p. 112; Baker & Rendle in Dyer,

FI. Trop. Afr. 4, 2 (1905), p. 108.

Type: Not designated. Hallier quoted Be/ck 52 and Liideritz 175, both from

S. W. Africa. As the original Liideritz specimen was probably destroyed I designate

Belck 52 in the Haussknecht herbarium, Jena, as the lecto-type.

Glabrous, prostrate or occasionally twining perennial herb. Stems herbaceous,

up to 2 m. long or even longer, sulcate and/or ribbed to almost winged (at least in

dried specimens). Leaves deeply palmately 5-7 (-9)-lobed (lowest lobes not infrequently

unequally forked); lobes narrowly linear to oblanceolate, acute or obtuse, mucronate,

3-6 cm. long and 0-5-8 mm. wide, of lowermost leaves often broader; petiole 2-25

mm. long. Peduncles 1-6 cm. long, 1 -flowered or occasionally cymosely 2-3-flowered;

bracteoles small, linear; pedicels 0-5-3 cm. long, somewhat thicker upwards. Sepals

pale yellowish green, subequal, elliptic with rounded apex, glabrous, coriaceous with

a membranous edge, 7-10 mm. long, 2 -5-5 -5 mm. wide. Corolla pale yellow or

sulphur-yellow with a deep-red, maroon or deep magenta “ eye ”, 2-3 cm. long, broadly

funnel-shaped with spreading, 3-4 cm. wide, faintly lobed limb; midpetaline areas

not distinct, sparsely pubescent towards the apex. Style linear with membranous

“ winged ” edges. Capsule usually distinctly exserted from the calyx, globose- or

ovoid-conical, 8-12 mm. long and 8-10 mm. in diam., glabrous, pale yellowish green

turning straw-colour, valves papyraceous. Seeds dark greyish brown to black, rather

dull, glabrous, 6-7 mm. long, about 5 mm. wide and about 3 mm. thick; testa nearly

smooth to distinctly rugose.

S. W. Africa, Bechuanaland, Transvaal, S. Rhodesia, Angola, possibly extending

into Trop. E. Africa.

Recorded from: S.W. Africa as far S. as Windhoek; Mafeking in the Cape

Province; Bechuanaland; Transvaal: Marico, Rustenburg, Brits, Pretoria,

Bronkhorstspruit, Waterberg, Petersburg, Zoutpansberg, Sibasa and the eastern

districts South to Barberton.

4. M. kentrocaulos (C. B. Clarke) Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 103;

Hall. f. in Meded. Rijksherb. Leiden I (1910), p. 21; Hutch, et Dalz., FI. W.

Trop. Afr. 2 (1931), p. 212; A. Meeuse in R. A. Dyer, Flow. PI. Afr. 30 (1955)

pi. 1194. Convolvulus kentrocaulos Steud., cheironym (PI. Schimp. It. Abyss. II,

no. 800, printed label). Ipomoea kentrocaulos (Steud. ex) C. B. Clarke in Hook.

f., FI. Br. Ind. 4 (1883), p. 213; N. E. Br. in Kew Bull. 1909, p. 124, incl. var.

pinnatifida N. E. Br. Operculina kentrocaulos (Steud. ex) Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 119.

Type: Not designated by Hall, f., but Steudel’s name, taken up by C. B. Clarke

and by Hallier, designates Schimper 800 from Abyssinia as the type gathering (K, type).

Large, glabrous perennial twiner. Stems becoming woody and up to at least

15 m. long, the younger ones slender, herbaceous but firm, terete, usually distinctly

muriculate with reddish papillae (as are petioles, peduncles and pedicels). Leaves

pentagonal in outline, 6-15 cm. long and as wide, palmately dissected nearly to the

base; base cordate with a narrow sinus; the lobes 5, oblong to lanceolate in outline,

obtuse to subacute, entire to irregularly and jaggedly pinnatilobed or pinnatifid; petioles

2-6 cm. long. Inflorescences cymose, few-flowered or reduced to a single flower.

Peduncles patent to suberect, 3-9 cm. long; bracteoles ovate, acute, concave, 3-5 mm.

long, early deciduous, occasionally larger and dissected like the leaves; pedicels up

to 3 cm. long, at first defiexed, patent to suberect when the flowers open and ultimately

cernuous in fruit. Sepals ovate-oblong or elliptic, coriaceous with thinner submem-

branous edges, glabrous, concave, somewhat unequal, obtuse or rounded and minutely

705

mucronate, 23-26 mm. long (inner ones longer than the outer ones) and about 12 mm.,

wide. Corolla funnelshaped, dull pale-yellow to pale buff with dark purple “ eye ”

glabrous, 4-6 cm. long and 6-8 cm. in diam., the limb faintly 5-angled, finely and

faintly plicate, midpetaline areas not sharply defined but smoother than the remainder

of the corolla. Capsule at first enclosed in the accrescent, brown and coriaceous calyx

but ultimately exposed just before dehiscence when sepals spread out, 12-15 mm. in

diam., pale brown, dehiscing by 4 valves and circumscissile at the base. Seeds brown,

minutely hairy and often with a few irregularly distributed white scales (insect-eggs?),

8-9 mm. long and about 6 mm. broad.

Africa south of the Sahara to Angola, Bechuanaland, the Northern Transvaal

and Portuguese East Africa; also in India.

Transvaal. — Lydenburg, Steelpoort: Story 4063 (PRE); v.d. Merwe 2356

(PRE); Waterval River Valley: Galpin 12256 (PRE). Letaba, Kruger National Park,

nr. Gorge: v.d. Schijff 2281, 2324A (PRE); Shingwedzi: v.d. Schijff 2985 (PRE).

Zoutpansberg, Messina: Rogers 20846, 22554 (PRE); Mastrangani: Breyer s.n.

(PRE, h. no. 16036). Sibasa, Kruger National Park, near Punda Maria: Obenneyer

684 (PRE); Codd 5342 (PRE); Klopperfontein: v.d. Schijff 3324 (PRE).

The seeds I have seen are not glabrous as stated by Rendle (l.c.).

Hallier referred this plant to Opercu/ina, but the capsule, although dehiscent at

the base, is also 4-valved and its wall is thin, papery, not consisting of two layers as

in Operculina (see the remarks by Van Ooststroom in Blumea 3 (1939), p. 326 regarding,

Merremia tuberosa (L.) Rendle, which has exactly the same type of capsule as M.

kentrocaulos). Hallier later (1910) agreed that his Operculina kentrocaulos had to be

referred to Merremia.

There is a considerable degree of variation in the dissection of the lobes of the

leaves: They vary from entire to pinnatilobed or almost pinnitifid, so that there is

no reason to maintain the var. pinnatifida.

Although I have not seen the type, I was able to study several specimens (e.g.,

Codd 5342) that had been compared with authentic material.

As regards the correct citation of the authors, Hallier (1893), p. 119 cited Ipomoea

kentrocaulos C. B. Clarke under Operculina tuberosa and used the name “ Operculina,

kentrocaulos (Steud.) Hall, f.” for the species under discussion. I cannot see why

Ipomoea kentrocaulos C. B. Clarke is not the first legitimate and correctly applied

name, because Clarke cites: “ Ipomoea tuberosa A. Rich, the old-world plant, not of

Linn.,” and “ Convolvulus kentrocaulos Steud. in PI. Schimp. n. 800 ”, so that he

expressly indicated that this species is distinct from Ipomoea tuberosa L. and took up

Steudel’s manuscript name. In addition, Clarke’s description applies to the species

with the muriculate stems and the adopted specific epithet implies the same. He also

defined Schimper 800 as the type according to the present rules of nomenclature.

The correct citation, therefore, is Merremia kentrocaulos [(Steud. ex) C. B. Clarke]

Rendle.

5. M. bipinnipartita {Engl.) Hall.f. in Engl. Bot. Jb. 18 (1893), p. 115; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 104. Ipomoea bipinnipartita Engl, in Engl.

Bot. Jb. 10 (1888), p. 246.

Type: Marloth 1289 from Usakos, S.W.A. As the original type specimen in the

herbarium Berlin-Dahlem was destroyed, I consider the specimen in PRE to be the

neo-type (isotype in SAM).

706

Perennial , glabrous. Rootstock tuberous. Stems procumbent, with rather long

internodes, sparsely and minutely tuberculate as are the petioles and the rhachis of

the leaves. Leaves somewhat fleshy, orbicular to broadly ovate in outline, 3-7 cm.

long and wide, bipinnatisect with 2 pairs of lateral leaflets and a terminal one ; secondary

lobes oblong or linear, usually obtuse but the terminal one very acute, almost entire

or with a few distant teeth to nearly pinnatisect; petioles up to 1 cm. long. Peduncles

short, about as long as the petioles, bearing slightly above the middle the small, ovate,

3-4 mm. long bracteoles. Sepals obovate, subpuberulous, about 16 mm. long and about

12 mm. wide, subcoriaceous. Corolla funnel-shaped with obtuse shallow lobes,

described as white with a black centre by Marloth, but by other collectors as “ cream

with deep red eye ” or “ white with purple eye ”, about 5 cm. long and 6-7 cm. in

diam., glabrous; midpetaline zones with thicker nerves but not well defined. Capsule

(only one seen) ellipsoid or obovoid, completely enclosed by the calyx (may become

ultimately exposed just before dehiscence as in M. kentrocaulos), about 9 mm. long

and 6 mm. in diam., glabrous, 1-seeded (if 2^1-seeded capsules are formed, they may

be considerably larger and of a different shape). Seed ellipsoid, truncate at one end,

7 mm. long and about 5 mm. in diam., glabrous (if more seeds develop, the shape

and size may be somewhat different).

S. W. Africa. — Brandberg (W. of Omaruru): Liebenberg 4982 (PRE). Usakos:

Marloth 1289 (PRE, type, SAM, isotype); Bradfield 662 (PRE). nr. Karibib: Kinges

3168,3322; de Winter 2691, 3220 (PRE). Rehoboth, Buellsport: Strey 2089 (PRE).

Gibeon, Orab: Dinter 2009 (PRE, SAM). Nuis: Range 1344 (BOL, SAM).

Although this plant has apparently a very limited distribution, some of the labels

mention: “ locally abundant ”.

6. M. malvaefolia Rendle in Jl. Bot. 39 (1901), p. 63. Ipomoea malvaefolia (Rendle)

Baker in Dyer, FI. Cap. 4, 2 (1904), p. 65.

Type: McOwan 403 (K) from Kowie (Bathurst).

Annual? Stems very slender, trailing, shortly hairy. Leaves more or less triangular

in outline, up to 3 cm. wide, palmately 5-lobed beyond the middle, brownish tomentose

beneath; lobes more or less obovate, contiguous, obtuse, apiculate; petiole much

shorter than the blade. Peduncles slender, ascending, 1-flowered, 4-10 cm. long;

bracteoles small, remote from the calyx. Calyx 8-13 mm. long, shortly pubescent;

sepals chartaceous, oblong, obtuse, much imbricate, the outer hispid. Corolla widely

funnel-shaped, probably pale yellow, 3-5 cm. long, silky on the midpetaline areas.

Capsule unknown.

Apparently a very rare species, recorded only from Bathurst, Albany and

Somerset East and as far as 1 can ascertain, not collected in the last 60 or 70 years.

I know this species only from the isotypes (McOwan 403 in GRA, BOL), but in

FI. Cap. specimens from Albany (leg. Bowker) and Somerset (leg. Bowker), are quoted.

1. M. tridentata (L.) Hall. f. ssp. angustifolia ( Jacq .) Ooststr. in Blumea 3 (1939),

p. 323, and in Steen., FI. Males. Ser. I, 4-4 (1953), p. 446; Brenan in Mem. New

York Bot. Garden 9 (1954), p. 8. Ipomoea angustifolia Jacq., Collect. 2 (1788),

p. 367, and Icon. Rar. 2 (1786-1793), p. 10, t. 317; Baker and Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 55; N. E. Br. in Kew Bull. 1909, p. 122. Ipomoea convol-

vuloides Schinz in Verhandl. bot. Ver. Brandenb. 30 (1888), p. 273; Dinter in

Fedde Repert. 18 (1922), p. 430. Merremia angustifolia (Jacq.) Hall. f. in Engl.

Bot. Jb. 16 (1893), p. 552 and 18 (1893), p. 117; Baker and Rendle in Dyer, FI.

Trop. Afr. 4, 2 (1905), p. Ill; Hutch, and Dalz., FI. W. Trop. Afr. 2 (1931), p.

211. Convolvulus longipedunculatus Dinter ms. on Dinter 7538.

707

Type of the subspecies: Jacquin’s plate was taken to be representative.

Annual , glabrous or rarely pubescent, prostrate and sometimes also twining herb.

Stems slender, subterete to angular, striate-ribbed, up to at least 1 m. long. Leaves

linear to narrowly oblong, more or less obtuse, mucronate, 2-8 cm. long, usually 2-6

mm. wide, with small toothed hastate basal auricles; petiole 0-5-3 (-5) mm. long

or leaves almost completely sessile. Peduncles very slender, 1-6 cm. long, 1-, more

rarely 2- or 3-flowered; bracteoles minute, lanceolate or cuspidate, persistent; pedicels

thickened upwards, 0-5-2 cm. long. Sepals oblong to ovate-oblong or elliptic, obtuse

to subacute, cuspidate or mucronate, glabrous, often brownish at least when dried,

5-10 mm. (usually 6-7 mm.) long; two outer ones slightly shorter than the inner ones.

Corolla pale yellow or occasionally darker yellow or bright lemon-yellow, with or

without a darker, reddish or brownish “ eye ”, funnel-shaped, 12-20 mm. long; the

limb shallowly 5-lobed with 5 more or less broadly triangular, acute lobes; midpetaline

areas well defined, glabrous. Capsule globose to ovoid, about 6 mm. in diam. ; the

valves papery, straw-coloured. Seeds 4 or less, yellowish brown to dark greyish brown,

3-4 mm. long, glabrous.

Africa south of the Sahara.

Recorded from: S. W. Africa (south to Okahandja); Griqualand-W. (Kuruman,

Vryburg, Mafeking, Hay, Kimberley); Orange Free State (Hoopstad, Kroonstad);

Transvaal (practically all districts); Swaziland; Zululand and Natal as far south as

Pt. Edward.

• This species is very common in many of the regions indicated and represented in

South African Herbaria by hundreds of specimens.

8. M. pinnata ( Hochst . ex Choisy) Hall, f in Engl. Bot. Jb. 16 (1893), p. 552 and 18

(1893), p. 116; Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 113;

Hutch, and Dalz., FI. W. Trop. Afr. 2 (1931), p. 211. Ipomoea pinnata Hochst.,

cheironym (Kotschy, herb. it. nub. no. 262, printed herbarium label); ex Choisy

in DC., Prodr. 9 (1845), p. 353; N.E.Br. in Kew Bull. 1909, p. 124.

Type: Kotschy 262 from Kordofan (the only specimen cited by Choisy), location

of holotype specimen probably Geneva.

Annual. Stems herbaceous, slender, trailing or twining, up to about 75 cm. long,

pubescent with soft, more or less distinctly bulbous-based spreading hairs (at least

when young), as are the leaves, peduncles, bracts, calyces and capsules. Leaves sessile,

pinnate, 1-4 cm. long and 0-5-1 -5 cm. wide, with 8-10 pairs of entire, linear segments.

Peduncles about as long as, or longer than the leaf, 1 -3-flowered; bracteoles subulate

or linear-subulate, 2-8 mm. long; pedicels usually under 6 mm. long (and few-flowered

inflorescences appearing subcapitellate) or occasionally up to 15 mm. long. Sepals

6-8 mm. long including the about 3 mm. long, aristate apex, 1 • 5-2 mm. wide,

herbaceous; the inner ones slightly narrower and less hairy on the wider basal portion.

Corolla glabrous, yellow, narrowly funnel-shaped, 7-8 mm. long; its limb distinctly

5-lobed, lobes broadly ovate, obtuse, midpetaline areas not differentiated. Ovary

with long stiff hairs longer than the ovary. Capsule globose or ovoid, straw-coloured,

about 6 mm. long and 5-6 mm. in diam. Seeds dark brown to black, glabrous, with

grey, minutely raised markings, about 2 mm. long, 1-1-5 mm. wide.

From Senegal to the Sudan and southwards to South West Africa, Bechuanaland,

Northern Transvaal and Portuguese East Africa.

708

S.W. Africa.— Tsumeb: Dinter 7572 (PRE, BOL), 7572a (PRE). Gaub (Groot-

fontein): Dinter 2431 (PRE, SAM).

Transvaal. — Rustenburg: Codd 3749 (PRF). Zoutpansberg, Messina: Rogers

19482 (BOL, J); Dongola: Bruce 51 (PRE); Codd 4120 (PRE), nr. Salt Pan; Schwei-

kerdt and Verdoorn 568 (PRE). Sebasa, Kruger National Park nr. Punda Maria:

Codd 5322 (PRE). Nelspruit: Breyer h. no. TRY 17704 (PRE).

Bechuanaland. — N’gamiland, Kwebe Hills: Lugard 176 (GRA).

12. OPERCULINA

5. Manso, Subst. Bras. (1836), p. 16; Peter in Engl. u. Prantl, Nat. Pfl. fam.,

Ed. 1, 4-3a (1891), p. 32; Hall. f. in Engl. Bot. Jb. 16 (1893), p. 582; Ooststr. in

Blumea 3 (1939), p. 361, and in Steen., FI. Males, Ser. I, 4-4 (1953), p. 454. Ipomoea,

p.p. and Merremia, p.p. Auct. plur.

Type species: Operculina turpethum (L.) S. Manso.

Characters generally as in Merremia, but fruit different. Large herbaceous twiners;

stems, peduncles and petioles often winged. Leaves entire, angular or digitate, often

cordate at the base. Corolla large, white or yellow. Pollen ellipsoid, espinose. Ovary

2-celled, 4-ovuled. Capsule large, with circumscissile epicarp, the upper part (which

comes off as an operculum or “lid”) more or less fleshy; the endocarp scarious, at

first entire, ultimately irregularly splitting. Seeds 1 or more, large, trigonous or globose,

black, glabrous.

Tropics generally (about 15 species).

Rendle in FI. Trop. Afr. 4, 2, p. 101, says “ 1 cannot distinguish generally the

larger-flowered and -fruited specimens which have been regarded as forming a distinct

genus Operculina. The transverse dehiscence of the fruit is not general in this small

group ”. Hallier [in Meded. Rijksherbarium Leiden 1 (1910), p. 21] agreed that some

species he had previously referred to Operculina should be put back into Merremia

such as M. kentrocaulos. Van Ooststroom (op. cit., p. 327) has pointed out that in

some species of Merremia, such as M. tuberosa (L.) Rendle, the wall of the capsules,

becomes detached as a whole at its base from the receptacle. In Operculina the “ oper-

culum ” is formed by the epicarp only, leaving the membranous endocarp (see figure 32

in FI. Males. Ser. 1, 4-4, p. 455) so that, according to him (the latest monographer),

there is no reason to unite Merremia and Operculina.

The only record in South Africa of a species of Operculina is Ipomoea saundersiana

Baker in FI. Cap. 4, 2 (1904), p. 60, which was based on Wood 1635, a specimen taken

from a cultivated plant in the Durban botanical garden, which had been raised from

seeds received from a Mrs. Saunders. Wood later gave a detailed description and

a plate of the cultivated specimen in question in Natal PI. 4 (1906), t. 353 and 354 and

remarked that the seeds had come “ from the interior of Africa ”. The plate in question,

and a duplicate of Wood 1635 (in NH) clearly show that Ipomoea saundersiana Baker

is a species of Operculina, and all the available evidence indicates that it is identical

with Operculina turpethum (L.) S. Manso, the only species of Operculina occurring

in Central Africa as far as I know.

Operculina turpethum has never been collected in the wild state in South Africa,

and I do not think it has ever been cultivated there outside the Durban botanical garden.

The species is only put on record here because of its inclusion (as Ipomoea saundersiana)

in FI. Capensis.

709

13. ASTRIPOMOEA

A. Meeuse, nom. no w— Astrochlaena Hall. f. in Engl. Bot. Jahrb. 18 (1893), p. 120;

Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 68; Baker & Rendle in Dyer, FI.

Trop. Afr. 4, 2 (1905), p. 118; Phillips, Gen. S. Afr. Flow. PI., Ed. 2 (1951), p. 623;

Meeuse in Taxon 4-8 (1955), p. 198, non Asteroch/aena Corda (1845), nee Asterochlaena

Garcke (1850).

Type species: Ipomoea lachnosperma Choisy = Astrochlaena lachnosperma (Choisy)

Hall. f. = Astripomoea lachnosperma (Choisy) A. Meeuse (proposed in Taxon 4-8

(1955), p. 199 sub. Astrochlaena). Hallier mentioned several species when he described

his genus Astrochlaena, the first being Astrochlaena floccosa (Vatke) Hall. f. As,

however, the latter is not a very well known species and may well be conspecific with

other previously described species, the earliest species described, viz. A. lachnosperma

(Ipomoea lachnosperma Choisy), a wide-spread and well-known species, was selected

as the type species.

The name Astrochlaena Hall. f. is invalidated by Asterochlaena Corda, a name

givent to a fossil “ fern ” stem (Astrochlaena and Asterochlaena are clearly only ortho-

graphic variants).

It is also invalidated by Asterochlaena Garcke, a synonym of Pavonia L. (Malvaceae).

There was certainly a case for the conservation of Astrochlaena Hall f. against Astero-

chlaena Garcke (see Taxon 4, 8), but the name of the fossil genus is well-known in

palaeobotany, is mentioned in recent text books on fossil botany, has apparently no

recognised older synonyms and was properly described, so that a new name for the

Convolvulaceous genus is inevitable.

Erect to decumbent, sometimes prostrate but never climbing, annual or perennial

herbs or undershrubs, covered with soft stellate hairs on all vegetative parts, peduncles,

bracteoles, pedicels and sepals. Stems usually simple or branched upwards, often

firm to stout. Leaves petiolate, usually ovate, oblong to subcordate, entire to coarsely

dentate-sinuate. Inflorescences cymose, few to many-flowered or by reduction

occasionally 1 -flowered; cymes axillary but often forming terminal leafy panicles at

the apices of the stems; long or short; bracteoles often small; pedicels usually short.

Sepals 5, usually more or less unequal, often ovate or oblong to lanceolate; outer

ones often subcarinate on the back. Corolla funnel-shaped, but the limb spreading,

almost entire: midpetaline areas well defined. Stamens included, unequal in length;

filaments filiform or linear; anthers oblong, somewhat lobed or sagittate at the base,

sometimes covered with swollen hairs; pollen spherical, spinose. Disc annular at the

base of the ovary. Ovary 2-celled, with 2 ovules in each cell; style simple, included;

stigmas 2, thick, oblong. Fruit a dehiscent usually glabrous capsule. Seeds usually

4, usually covered with a velvety or cobwebby to villous tomentum; embryo straight;

cotyledons flat, broad, deeply bilobed; endosperm scanty.

Three species in South Africa:

Corolla up to 2 cm. long; its tube purple, the limb white 1. A. lachnosperma.

Corolla 2-5-5 cm. long, concolorous (purple, mauve or magenta):

Corolla 2-5-4 cm. long; leaves usually ovate, cuneate to rounded at the

base; seeds very minutely velvety-pulverent ; East Africa to Natal 2. A. malvacea.

Corolla 4-5-5 cm. long; leaves broadly ovate to rotundate, shallowly and

broadly cordate to rounded at the base ; seeds distinctly villous; only

recorded from S.W. Africa 3. A. rotimdata.

710

1. A. lachnosperma ( Choisy ) A. Meeuse, comb. nov. — Ipomoea laehnospenna Choisy

in DC., Prodr. 9 (1845), p. 356. Astrochlaena laehnospenna (Choisy) Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 121; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), p. 119; N. E. Brown in Kew Bull. 1909, p. 124; Hutch, and Dalz., FI. W.

Trop. Afr. 2 (1931), p. 213.

Type: Kotschy 260 (from Kordofan) in herb. Geneva (isotypes in JE, L, PRE).

Perennial. Stems erect, up to at least 1-25 m. high, covered with pale fawn to

whitish stellate tomentum, as are petioles, peduncles, pedicels and calyces, firm, becoming,

woody and glabrescent at the base. Leaves ovate, varying to broadly elliptic, ovate-

lanceolate or subrhomboid, entire to slightly or occasionally distinctly repand, 3—10—

(-15) cm. long and 2-8 (-12) cm. wide; rounded or truncate to broadly cuneate at

the base, usually gradually narrowing into an acute or acuminate to subaristate apex,

thinly covered with white stellate hairs above, more densely so and paler beneath;

petiole 1-5 cm. long. Flowers solitary or umbellate; peduncles short, usually under

1 cm. long; pedicels usually shorter than the peduncle; sepals ovate-lanceolate,

acuminate-aristate, 6-8 mm. long. Corolla cream to white with purple in the tube

inside, funnel-shaped, about 18 mm. long. Capsule globose, glabrous, 6-8 mm. in

diam. Seeds villous-silky.

From Eritrea to Nigeria and southward, extending into South West Africa and

Bechuanaland.

South West Africa. — Grootfontein, Abenab: Dinter 7399 (PRE, BOL); Schoen-

felder S639 (PRE).

Bechuanaland. — Kwebe Hills: Lugard 91 (K, t. N. E. Brown). N’gamiland:

Curson 408 (PRE).

2. A. malvacea ( Klotzsch ) A. Meeuse, comb. nov. Breweria malvacea Klotzsch in

Peters, Reise Mossamb., Bot. (1861), p. 245, t. 367. Astrochlaena malvacea

{Klotzsch) Hall. f. in Engl. Bot. Jb. 18 (1893), p. 121; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 69; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 121; Hutch. & Dalz., FI. W. Trop. Afr. 2 (1931), p. 213; Brenan in Mem.

New York Bot. Garden 9 (1954), p. 8.

Type: Peters s.n. in herb. Berlin from Inhambane, Portuguese E. Africa (now

destroyed). If no isotype is extant in another herbarium, the plate (t. 367) in Peters,

Reise Mossamb., must be taken to represent the type.

Perennial. Stems 60-200 cm. long, erect, ascending or decumbent, firm, more

or less densely clothed with short whitish to fawn pubescence, glabrescent and becoming

woody at the base. Leaves usually ovate, sometimes varying to broadly ovate, ovate-

lanceolate or subrhomboid, 3-8 (-12) cm. long and 2-5 (-8) cm. wide, entire or repand,

rounded, truncate or sometimes somewhat cunate at the base, acuminate to rounded

at the apex, dull green drying brownish and sparsely stellate-hairy to glabrescent above,

matted with white stellate tomentum beneath; petioles varying considerably in length,

but usually much shorter than the correponding blades, densely stellate-hairy as are

peduncles, pedicels and calyx. Inflorescences axillary and terminal, often forming a

leafy panicle at the tops of the stems, peduncles rather slender, lower ones up to 12 cm.

long, upper ones usually much smaller, erect to patent, subumbellately 1- to few-

flowered; bracteoles minute, early deciduous, pedicels 5-15 (-20) mm. long, distinctly

subclavate. Sepals elliptic to ovate, acute to obtuse, 6-8 (-10) mm. long. Corolla

funnel-shaped, described as mauve or purple (probably bright magenta), 2-5-4 cm.

long and the limb up to 5 cm. in diam., glabrous or nearly so. Capsule globose or

somewhat ovoid, glabrous, 6-12 mm. long, 6-9 mm. in diam. Seeds minutely velvety-

pulverulent with a tuft of fawn hairs round the hilum.

711

From West Tropical Africa eastwards and extending into Natal.

Recorded from the coastal zones of Zululand and Natal as far south as Durban,

in the northernmost part fairly common (St. Lucia Bay, Richard’s Bay, etc.).

3. A. rotundata (Pilger) A. Meeuse, comb. nov. Astrochlaena rotundata Pilger in Engl.

Bot. Jb. 45 (1910), p. 222; Dinter in Fedde, Repert. 15 (1918), p. 345.

Type: Not designated, because Pilger mentioned two specimens ( Dinter 817 and

Dinter 817a). The original type specimens in the Berlin Herbarium having been lost,

I consider Dinter 817 in SAM to be the neo-type.

Probably perennial. Stems prostrate, up to at least 1-20 m. long, firm and stout,

densely covered with a short stellate brown tomentum, as are petioles, main veins of

lower surface of leaves, peduncles, pedicels and calyx. Leaves broadly ovate-subcordate

to almost orbicular-subcordate, 4-8 cm. long and 3-7 cm. wide, entire or somewhat

repand, rounded, truncate, subcordate to subcuneate at the base, acute to subacute

at the apex and often with a short mucro, rather thinly stellate-hairy above, more

densely so between the prominent veins and distinctly paler below; petiole 1-5 cm.

long. Peduncles mainly axillary, 4-7 cm. long, 1-3-flowered; bracteoles small, early

deciduous; pedicels subumbellate, 1-2-5 cm. long. Sepals ovate-lanceolate or elliptic,

obtuse, sometimes mucronate, 10-12 mm. long, the inner ones narrower and more

acute. Corolla funnel-shaped, “bright purple”, 4-5-5 cm. long; the limb about as

much in diam. Capsule subglobose to ovoid, glabrous, 10-12 mm. long and about

10 mm. in diam. Seeds densely and shortly villous.

South West Africa. — Grootfontein, Otavi. Auros: Dinter 5617 (NH, PRE,

SAM, L), Otavi: Dinter s.n. (herb. no. SAM 61986). Aukas-Kreyfontein : Dinter

817 (SAM, type!).

Notes. — The specimens cited above were distributed by Dinter under the name

Astrochlaena rotundata and agree very well with the description and with Dinter 817

in SAM. The hitherto undescribed capsule resembles that of A. malvacea very much,

but the seeds are covered with a villous, buff-brown or drab pubescence.

A. rotundata resembles A. malvacea and possibly some other species, but it is

distinct from A. malvacea in the leaf-shape, its larger flowers and the distinctly pubescent

seeds. I cannot identify A. rotundata with any other described species and it seems to

be endemic in South West Africa.

14. IPOMOEA

L. [Sp. PI. Ed. 1 (1753), p. 159]; Gen. PL, ed. 5 (1754), p. 76; Choisy in DC.,

Prodr. 9 (1845), p. 348; Benth & Hook., Gen. PI. 2 (1876), p. 870; Peter in Engl.

& Prantl., Natiirl. Pfl. fam. ed. 1, 4 -3a (1891), p. 28; Hall. f. in Engl. Bot. Jb. 16

(1893), p. 583 and 18 (1893), p. 123; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904),

p. 46; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 128; Ooststr. in Blumea

3 (1940), p. 481, and in Steenis, FI. Males, ser. 1, 4-4 (1953), p. 461; Phillips, Gen.

S. Afr. Flow. PL ed. 2 (1951), p. 624. Pharbitis Choisy in Mem. Soc. Phys. Geneve

6 (1833), p. 441 and in DC., Prodr. 9 (1845), p. 345; Peter, op. cit., p. 31. Calonyction

Choisy, op. cit. (1838), p. 441 and (1845), p. 345; Peter op. cit., p. 26; Hall. f. in Engl.

Bot. Jb. 16 (1893), p. 583 and 18 (1893), p. 153. Rivea, Choisy in DC., Prodr. 9 (1845),

p. 325; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 155; Phillips, Gen. S. Afr. Flow. PL

ed. 2 (1951), p. 624, pro parte, exclus. type; non Rivea Choisy in Mem. Soc. Phys.

Geneve 6 (1833), p. 407, nec. Hall. f. in Engl. Bot. Jb. 16 (1893), p. 584.

Type species: I. pes-tigridis L., Sp. PL Ed. 1 (1753), p. 1 62 {fide Phillips 1951).

712

Annuals or perennials, herbaceous, suffruticose or sometimes woody, usually

twining or prostrate, rarely floating or erect. Leaves varying in shape and size, entire,

lobed, or deeply divided, often cordate at the base, petiolate or rarely sessile.

Inflorescences axillary, cymose, but sometimes dense and pseudo-capitate, few- to

many-flowered but not infrequently 1-flowered by reduction; the cymes rarely forming

a terminal leafy panicle: peduncles short or long, rarely almost 0; bracteoles various,

usually small, linear or lanceolate, sometimes leafy and/or forming an involucrum;

pedicels usually distinct, sometimes long, occasionally very short. Sepals 5, herbaceous

to subcoriaceous, often ovate or lanceolate, linear or elliptic, obtuse to aristate,

unequal or equal, glabrous or hairy, persistent, often more or less enlarged in fruit.

Corolla small to large, usually quite regular, rarely faintly zygomorphic, usually funnel-

shaped or the tube somewhat campanulate, sometimes hypocrateriform; the limb

shallowly, occasionally deeply 5-lobed; midpetaline areas well defined by 2 distinct

nerves, often hairy towards their tips. Stamens 5, usually unequal, inserted near the

base of the corolla-tube, subincluded or sometimes exserted; filaments filiform or

somewhat linear, often dilated and hairy or papillate at the base; anthers ovate to

linear; pollen globose, spinulose. Disc annular. Ovary usually 2- or sometimes

4-celled, 4-ovuled, rarely 3-celled and 6-ovuled; style filiform (sub)-included to exerted;

stigma capitate entire or often 2-3-globular. Capsule globose to ovoid, dehiscent

by 4, rarely 6 valves. Seeds as many as the number of ovules, or less by abortion,

glabrous or hairy.

Circumtropical.

The generic limits are in agreement with Van Ooststroom’s latest treatment of

the genus (1953). The genera Colony ction Choisy and Pharbitis Choisy (and Quamoclit

Moench, which is only represented in Southern Africa by cultivated species) have been

reduced to its synonymy. Hallier in Meded. Rijksherb. Leiden 46 (1922), p. 19-20

had already reduced Calonyction (and Quamoclit) to subsections of Ipomoea section

Leiocalyx. Van Ooststroom (1940) followed him, but later (1953) placed them as

separate sections under Ipomoea next to Leiocalyx.

On the other hand, as in the system of the Convolvulaceae much importance is

attributed to the structure of the fruit and of the pollen, many species described under

Ipomoea have to be referred to other genera such as Merremia and Operculina (with

smooth pollen), and Stictocardia, Turbina and Argyreia (with indehiscent or only

partly dehiscent fruits). The genus Rivea as defined by Choisy in 1833 was later extended

by himself (1845) and by Hallier in Engl. Bot. Jb. 18 to include many more species

which do not agree with the original diagnosis, because they have bi-globose instead

of the oblong stigmas of the type species of Rivea, or dehiscent capsules instead of the

indehiscent fruit of Rivea. Van Ooststroom in Blumea 5 (1953), p. 353-355 has pointed

out that Hallier’s later conception of Rivea contained a mixture of heterogeneous

elements, which have to be referred partly to Rivea proper (two species in south-east

Asia and Ceylon), partly to Turbina or Argyreia (those forms which have a bi-globose

stigma and indehiscent fruits) and partly to Ipomoea sect. Eriospermum. The single

species of “ Rivea ” mentioned by Phillips in the second edition of his “ Genera ”

[i.e., Rivea adenoides (Schinz) Hall. f. ], is an example of the last group: its dehiscent

capsule excludes it from Rivea and its seeds with very long hairs, together with its

other characters, indicate that it belongs to Ipomoea sect. Eriospermum and its original

name Ipomoea adenoides Schinz has to be restored. For a more detailed discusion see

Van Ooststroom (1943), l.c., and under Turbina on p. 774.

The sections of Ipomoea , as indicated by Hallier in Engl. Bot. Jb. 18 (1893), have

been adopted by several authors, and are, with only few changes, maintained by Van

Ooststroom in Flora Males. Ser. 1, 4-4 (1953).

713

The species are arranged below in practically the same sequence as indicated by

Hallier, although it seems doubtful if they are all correctly placed in the proper section.

I made an exception for I. pellita which Hallier placed in his section Dasychaetia, but

to my mind, is perhaps much more closely related to species of section Calycanthemum,

such as I. crassipes.

Species not mentioned by Hallier I have tentatively referred to sections, although

it is difficult to place some of them satisfactorily. I. ommaneyi, for instance, shows

characters of section Calycanthemum (habit) and of Pharbitis (inflorescence).

The South African species are referable to the following sections: —

Calycanthemum (Klotzsch) Hall. f. (spec. 1-9).

Pharbitis (Choisy) Griseb.

subsection Chorisanthae Hall. f. (spec. 11-15).

subsection Cephalanthae (Choisy) Hall. f. (spec. 16-22).

Batatas (Choisy) Griseb. (spec. 23).

Leiocalyx Hall. f. (spec. 24-40).

Calonyction (Choisy) Griseb. (spec. 41).

Eriospermum Hall. f. (spec. 42-47).

The species of Turbina being very similar to Ipomoea in floral characters, are

included in the key to the species of Ipomoea because fruiting specimens are not always

available.

Ai Leaves simple, entire, lobed, emarginate or toothed, but not pinnately,

palmately or pedately lobed nearly to the base nor compound:

Flowers in pedunculate heads, surrounded by a single large bicuspidate

bract:

Sepals lanceolate, acute, glabrous or sparsely hairy, mainly at the tips

and along the margins; corolla distinctly funnel-shaped 3-5 cm.

long, the limb 3-5 cm. in diam

Sepals oblong-spathulate to oblong, obtuse, glabrous below the tips

inside, in fruit recurved and showing the glabrous surface outside;

corolla-tube slender, subcylindric, 2-3 cm. long; the limb

spreading, 1-3 cm. in diam

Flowers not in heads, or, if so, heads not subtended by one large and bicus-

pidate boat-shaped bract:

Bj Corolla narrowly to widely funnel-shaped:

Leaves with minute black dots beneath; sepals about 8 mm. long,

much imbricate, subequal, glabrous, obovate, obtuse (Zululand)

Leaves without black dots (sometimes minutely pitted or pellucidly

glandular):

Cx Corolla small, under 18 mm. long and or its limb less than about

15 mm. in diam., in fascicles, clusters or solitary, sometimes

in pedunculate 1- to few-flowered cymes or heads:

Leaves when dry pellucidly glandular when seen in transmitted light,

crisped along the margin; peduncles 0-2 mm. long; pedicels

short, only up to 2 cm. long in fruit; sepals very unequal,

outer ones ovate from cordate base, obtuse, inner ones

lanceolate, outer ones markedly accrescent, crisped along

the margin; corolla about 16 mm. long, mauvish-pink ;

capsule and seeds hairy

21. /. involucrata.

22. I. pileata.

see Stictocardia.

5 I. hackeliana.

714

Leaves not pellucidly glandular:

Flowers in distinctly pedunculate, bracteate, dense heads;

peduncle over 2 cm. long; leaves entire, elliptic to linear-

oblong, obtuse, tomentose beneath; coroHa slightly longer

than the about 6 mm. long calyx, pale yellow; erect to

trailing annual 16. I. chloroneura.

Flowers not in pedunculate bracteate heads; if fascicled or

clustered, peduncles never longer than 2 cm.:

Leaves, at least the majority, hastate at the base with entire or

toothed basal auricles, very rarely all lanceolate; flowers

solitary, rarely fasciculate; peduncles 0; pedicels 6-10

(-30) mm.; sepals linear, acuminate, equal, erect, hairy,

6—1 2 mm. long; corolla pinkish mauve with darker

centre, occasionally white; capsule hirsute 4. I. gracilisepala.

Leaves cordate, cuneate or truncate at the base, never hastate :

Leaves usually 3-lobed, albotomentose beneath, flowers

mauve-purple, in pedunculate dichasial 3-1 1 -flowered

cymes 13. I. arachnosperma.

Leaves entire, not albotomentose:

Leaves cuneate to truncate-rounded, rarely subcordate

at the base; flowers very small, sessile; corolla

hardly longer than the usually bristly calyx, white 1. I. coscinosperma.

Leaves distinctly cordate at the base:

Peduncles distinct, slender, over 1 cm. long, cymosely

1-3-flowered; pedicels distinct; outer sepals

subcordate at the base, attenuate into an acute

apical portion, inner sepals narrower; corolla

about twice as long as the calyx, its tube sub-

cylindric, 12-16 mm. long 1.1. sinensis.

Peduncles 0 or short and thick ; flowers solitary, binate

in axial clusters or fasciculate; corolla a little

longer than the calyx*:

Flowers usually in clusters, practically sessile; sepals

very hairy, ovate-lanceolate with linear-

acuminate, spreading tips; capsule hairy;

seeds glabrous 2. 1. eriocarpa.

Flowers usually binate, pedicellate; sepals obtuse or

subacute, ciliate or nearly glabrous, sometimes

thinly bristly outside; capsule glabrous or

with a few bristly hairs; seeds densely pubes-

cent 3. I. plebeia.

C2 Corolla larger to very large, over 18 mm. long; flowers usually

pedunculate :

Leaves bilobed at the apex, coriaceous or subcoriaceous, the

whole plant glabrous; stems stout, trailing; growing only

on sandy beaches 31. 1. pes-caprae.

Leaves not bilobed (sometimes emarginate) at the apex, or not found

on sandy beaches:

Leaves cordate-ovate in outline, shallowly 3-lobed, or entire,

cordate, cordate-ovate or cordate-deltoid, covered with

a thin white cobwebby tomentum beneath (this tomentum

occasionally reduced to a few scattered floccose patches);

flowers mauve or magenta; bracteoles and sepals lanceo-

late or linear-lanceolate, acute to acuminate-aristate:

Flowers in dense globose pedunculate heads; corolla 2-3 (-4)

cm. long 14. I. wightii.

Compare the footnote on p. 722 regarding I. leucanthemum.

715

Flowers in few- to many-flowered dichasia or cymes, sometimes

solitary; if peduncles and pedicels very short, flowers

few and corolla 3-5 cm. long:

Corolla 18-25 mm. long

Corolla 30-50 mm. long

Leaves either not with a white cobwebby tomentum or bracts and

sepals different (and corolla white or pale mauve):

D[ Leaves linear or lanceolate with narrow or sometimes

somewhat rounded base, entire or sometimes toothed,

repand, lobed or more or less emarginate; plants erect

to prostrate, often suffruticose:

Vegetative parts pubescent; calyx tomentose or pubescent:

Vegetative parts (at least on the younger shoots) and calyx

usually covered with an adpressed, short and

silvery pubescence; if only densely hairy sepals

subequal

Vegetative parts usually not with an adpressed short and

silvery tomentum, or sepals very unequal:

Sepals unequal; plant hairy . . . compare extreme

forms of

Sepals equal or subequal:

Sepals acuminate to aristate, leaves usually long-

attenuate at the base and well over 5 cm.

long, usually also some toothed or pinnatisect

leaves present; leaves hairy or glabrous above;

pubescence soft, white

Sepals acute or acuminate; leaves under 5 cm. long,

on a short petiole or nearly sessile, glabrous

above, pubescence soft, white

Sepals acute or acuminate; leaves usually over 5 cm.

long, hairy above and below; pubescence

usually stiff, adpressed on the leaves, yellowish

or brownish

Vegetative parts (at least the leaves) glabrous, stems

occasionally minutely scabrid-hirsutulous:

Corolla white, 20-25 mm. long; plant stemless or with

a very short stem; leaves entire or with a few

teeth

Corolla magenta or pale mauve with darker centre, over

25 mm. long:

Leaves usually over 5 mm. wide; corolla pale mauve

with darker centre; stems usually stout, some-

times minutely scabrous or hirsutulous

Leaves usually less than 5 mm. wide; corolla magenta;

stems often slender, glabrous, terete

D2 Leaves broader, not linear or lanceolate or, if narrow,

broadly truncate, hastate, sagittate or cordate at the base:

Ej Annual or perennial plants, herbaceous, prostrate or

climbing, sometimes suffruticose and suberect but, if so,

under 50 cm. high, never tall and woody:

Stems trailing, firm, often rooting at the nodes; leaves

broadly ovate to suborbicular, entire, angular, or

3-7-lobed; peduncles 3-18 cm. long, stout, angular;

pedicels 0-3-1 -2 cm. long; sepals equal and 7-8

mm. long or inner ones longer, 9-12 mm.; corolla

glabrous, pale mauve, 30-45 mm. long; ovary

usually hairy; cultivated for its edible tubers and

occasionally found as a culture relic

13. I. arachnosperma.

15. I. ficifolia.

see Turbina.

8. I. eras sipes.

6. I. oenotherae.

see Turbina robertsiana.

see Turbina oblongata.

32. I. simplex.

33. I. welwitschii.

34. I. bolusiana.

23. I. batatas.

716

Plant not cultivated for its edible tubers (sometimes as an

ornamental) :

Stems thick, trailing, hollow or spongy, rooting at the

nodes ; leaves linear to ovate with cordate, hastate,

sagittate or truncate base, entire or with dentate

basal auricles, glabrous, 3-15 cm. by 1-9 cm.;

inflorescence cymosely 1- to few-flowered;

peduncles 1-12 cm. long; bracts minute;

pedicels 2-6-5 cm. long; calyx 7-8 mm. long,

glabrous; corolla pink or pale mauve, often

with darker centre, glabrous, 3-5 cm. long; in

moist, marshy inundated localities or even com-

pletely aquatic, floating 30. I. aquatica.

Stems not thick, hollow or spongy or rooting at the

nodes and plant not usually found in very moist

localities :

Fj Leaves distinctly cordate or sagittate at the base:

Corolla either pale yellow, concolorous, or white

or yellow with a purple zone at the base of

the corolla-tube inside, 15-30 mm. long;

sepals subequal, 6-8 mm. long 24. I. obscura.

Corolla pink, mauve, magenta, purple to blue or

pure white, if white or yellowish with purple

or magenta centre, sepals very unequal and

over 8 mm. long:

Gx Sepals entirely glabrous:

Petiole short, under 15 mm. long; leaves

usually under 4 (-5) cm. long; plant

usually hairy in some part other than

the calyx; corolla with small tufts of

hairs protruding from the midpetaline

areas (especially conspicuous in older

buds) 25. I. transvaalensis.

Petioles 15-200 mm. long:

Herbaceous, quite glabrous twiner; leaves

cordate, entire, long-acuminate;

stems and peduncles fistulose; in-

florescence cymosely few- to many-

flowered; calyx 4-5-6 mm. long;

corolla 4-6 cm. long, violet, purple

or (in S. Africa usually) bright blue

with a white tube; cultivated and

occasionally found as an escape 35. I. tricolor.

Plants not cultivated, corolla never blue:

Leaves cordate-elliptic or rounded to

subreniform, obtuse with minute

cusps; basal sinus narrow and

deep; sepals obtuse, 12-18 mm.

long; corolla pale mauve with

magenta centre, 3-6 cm. long 26. 7. bathycolpos.

Leaves cordate-deltoid or cordate-ovate,

usually suddenly acuminate from

a broad, usually irregularly and

coarsely few-toothed base into an

entire apex; basal sinus broad;

sepals 6-8 mm. long; corolla

magenta, 2-3 cm. long 21. I. papi/io.

717

G2 Sepals thinly to densely hairy or tomentose:

Sepals long-attenuate or linear-acuminate at

the apex, 14-25 mm. long; twining

herbaceous plants with retrorse hairs

on stems and petioles; leaves often

3-lobed; often cultivated:

Outer sepals lanceolate at the base with long

and linear acumen, patently hirsute

in the basal portion, 17-25 mm. long;

cultivated 10.

Outer sepals lanceolate to ovate-lanceolate,

gradually attenuate towards the apex,

with adpressed pubescence, 14-22

mm. long; wild and cultivated.... 12.

Sepals acute, obtuse or mucronate but not

with a long and narrow acumen, if

sepals long-acuminate, plant not twining

and without retrorse hairs on stems

and petioles:

Calyx with bristly patent hairs in basal

portion, glabrous towards the apex;

twining annual with retrorse hairs on

stems and petioles 11.

Calyx not with bristly patent hairs in basal

portion and glabrous towards the

apex:

Twining plant, growing in moist places

(river banks, marshes, pools), often

among reeds, sedges. Papyrus, etc.;

petioles 3-12 cm. long; peduncles

2-12 cm. long; carrying one to

few flowers in a subumbellate

cyme; calyx 7-10 mm. long,

corolla 4-6 cm. long, mauve

(Okavango marshes, tropical

Africa) 42.

Prostrate to suberect plants, or an erect

shrub:

Plants prostrate to suffruticose, if

suberect never more than 50 cm.

high, corolla usually magenta:

Bracteoles usually close to the calyx,

linear or lanceolate, rarely

broader; sepals subequal.. . . see

Bractioles usually distant from the

calyx, often foliaceous; sepals

very unequal: outer ones

from a broad base ovate-

acuminate; inner ones much

narrower, linear from an ovate

or elliptic basal portion 8. j

Plant shrubby, erect; stems woody;

sepals very broad, unequal;

corolla pale mauve with magenta

centre see

F2 Leaves not distinctly cordate or sagittate at the base:

Sepals entirely glabrous, sometimes rugose or

muriculate :

Sepals 6-8 mm. long, corolla magenta, leaves

acuminate from a broad base, coarsely

toothed in lower half 27.

I. nil.

I. congesta.

I. purpurea.

I. riparia.

Turbina oblongata.

'. crassipes.

Turbina holubii.

I. papilio.

133279 6-6

718

Sepals 10-16 mm. long; corolla pale mauve with

darker centre; leaves linear-oblong to

elliptic, entire emarginate, or 3-lobed,

not broad or toothed at the base

Sepals (12-) 16-22 (-25) mm. long; corolla

magenta; leaves oblong, elliptic, ovate to

linear-oblong, entire . . .

glabrous specimens of

Sepals more or less hairy to tomentose:

Plant shrubby, erect, over 50 cm. high; sepals

very broad, suborbicular, rounded, un-

equal, inner ones larger

Plants prostrate; sometimes suffruticose,

suberect but if so, under 50 cm. high and

sepals not very broad and rounded:

Flowers in pedunculate, few- to many-flowered

heads (heads occasionally binate):

Bracteoles and sepals linear, narrow, bristly;

bracteoles (15)-20-30 mm. long;

sepals 20-35 mm. long; petioles

usually over 2 cm. long; leaves ovate

or ovate-oblong, mostly or all under

9 cm. long

Bracteoles and sepals lanceolate to ovate:

Leaves lanceolate or ovate-lanceolate from

a broad base to long-triangular,

gradually narrowed towards the

apex, up to 30 cm. long and 10 cm.

wide; petiloes usually very short;

corolla bright magenta, 3-5 cm.

long, densely silky on the mid-

petaline areas

Leaves usually oblong to ovate and not

often gradually tapering to the

apex from a broad base, petioles

usually over 1 cm. long; corolla

magenta, 5 cm. long or longer,

usually not densely silky but thinly

strigose on the mid-petaline areas

Flowers solitary or 2 (-3) together, or not in

heads but in cymes:

Sepals very unequal; outer ones ovate or

lanceolate from a broad, sometimes

subcordate base, inner ones linear

from a short elliptic or ovate basal

portion; bracteoles usually distant

from the calyx and often foliaceous

Sepals equal or not so distinctly unequal:

Leaves up to 4 (-5) cm. long, broad at

the base, often subcordate; brac-

teoles very minute, calyx 6-8 mm.

long; corolla bright magenta-rose-

colour, 2-3 cm. long with small

tufts of hairs protruding from the

midpetaline areas (especially con-

spicuous in older buds)

33. I. welwitschii.

Turbina oblongata.

see Turbina holubii.

9. I. pellita.

17. I. ommaneyi.

18. /. atherstonei.

8. /. crassipes.

25. /. transvaalensis.

719

Leaves, bracts, sepals or corolla longer

or corolla, if hairy on the midpeta-

line areas without protruding hair-

tufts:

Leaves oblong to narrowly ovate, often

crisped along the margin, up to

about 3-5 cm. long; the apex

often reflexed, mucronate;

pubescence soft, spreading,

white, the hairs never silky, not

yellowish or brown, lower sur-

face of leaves and calyx neither

densely hairy nor almost tomen-

tose; stems and peduncles usually

slender:

Bracteoles distant from, more rarely

contiguous to the calyx ; leaves

quite glabrous above; corolla

of open flowers thinly hairy to

glabrous on the midpetaline

areas (Transvaal)

Bracteoles contiguous to the calyx;

leaves usually hairy on both

sides; corolla of open flowers

silky on midpetaline areas

(Eastern Cape)

Leaves usually well over 3-5 cm. long,

rarely crisped along the margins;

pubescence often hirsute (or

stiffly strigose on the leaves);

the hairs not infrequently yellow

or brown, or occasionally silky-

tomentose on lower surface of

leaves, sepals and sometimes the

midpetaline areas of the corolla

E2 Tall perennial woody climbers or shrubs over 50 cm. high:

Calyx glabrous:

Sepals orbicular to oblong, obtuse; leaves usually

lobed, rarely entire

Sepals ovate, cuspidate to acute; leaves entire, usually

appearing after the flowers

Calyx (at least partly) hairy or tomentose:

Sepals broadly elliptic or obovate to orbicular, obtuse,

6-10 mm. long, the two inner ones conspicuously

larger and usually longer than the outer ones;

leaves cordate-ovate, usually under 5 cm. long,

petioles slender; erect shrub with the ultimate

twigs sometimes flexuous or twining

Sepals not as above; plants climbing, leaves usually

much larger, young parts densely tomentose, as

are the peduncles:

Young leaves with very conspicuous reticulate,

somewhat floccose tomentum on the main

veins beneath, older ones glabrescent; calyx

at first tomentose at least near the base;

corolla white, 6-8 cm. long (an eastern species)

Young leaves with a complete tomentum beneath,

tomentum usually persistent ; calyx tomentose;

corolla pale mauve, 6-10 cm. long (a western

species)

see Turbina robertsiana.

28. /. crispa.

see Turbina.

43. I. digit at a.

46. /. shirambensis.

see Turbina holubii.

44. /. albivenia.

45. I. verbascoidea.

720

Young leaves with a short silvery tomentum beneath;

tomentum usually persistent; flowers in a

terminal panicle; calyx silvery-tomentose;

corolla white, about 2 cm. long

Bt Corolla hypocrateriform, completely white or white with the inside of

the tube purple or magenta:

Plant woody and climbing; calyx glabrous, sepals obtuse

Plant shrubby, erect; calyx densely silky outside

Plant herbaceous, twining:

Corolla-tube up to about 35 mm. long; leaves ovate to lanceolate,

entire, cuneate to rounded at the base; sepals not awned..

Corolla-tube at least 6 cm. long; leaves cordate at the base, entire

or 3-lobed; sepals awned

A* Leaves deeply lobed (nearly to the base), dissected or compound:

Leaves albo-tomentose beneath with the veins marked out, palmately 3-5

or almost 7-lobed

Leaves glabrous or sometimes hairy but not albo-tomentose:

Flowers in dense pedunculate and bracteate heads; leaves hairy

Flowers solitary, fascicled or in cymes:

Corolla small, white tinged with mauve, 10-15 mm. long; leaves

palmately 5-fid with dentate or serrate to pinnatisect segments

Corolla usually larger, rarely white; leaf-segments not dentate or

serrate, entire or the basal ones with a lateral lobe:

Large glabrous perennial climber; leaves palmately (3-) 5-7 (-9)-

lobed, 6-14 cm. by 6-16 cm., not pseudostipulate; sepals

orbicular or the outer ones narrower, obtuse, 6-12 mm. long;

corolla 5-6 cm. long, pale mauve, the tube darker inside . .

Annuals or prostrate perennials, leaves much smaller, usually under

8 cm. in diam. and often pseudostipulate by the small leaves

of developing or suppressed axillary shoots:

Leaves usually pseudostipulate, distinctly palmately or pedately

5-9-fid, quite glabrous; climbing or occasionally prostrate

annuals:

Calyx 8-10 mm. long; outer sepals saccate at the base; corolla

yellow with mauve tube, 3-6 cm. long and in diam.;

leaves usually biternately pedate with often 9 segments

Calyx up to about 7 mm. long, outer sepals not saccate; corolla

mauve, purple or white; leaves palmately partite with

usually 5 segments:

Corolla 3-6 cm. long and 4-6 in diam., mauve

Corolla up to about 25 mm. long:

Peduncles very slender, filiform, corolla often under 16 mm.

long, purplish-mauve

Peduncles rather stout, not filiform, corolla 15-25 mm.

long, purplish-mauve or white

Leaves not pseudostipulate, pinnately dissected, tridentate or

palmately compound; stems prostrate or rarely suberect

from a perennial tuberous rootstock; corolla usually over

3 cm. long:

Young parts and sepals with soft white hairs; sepals very acute;

basal leaves, if present, linear; cauline leaves tridentate

to pinnatisect or all leaves pinnatisect; flowers solitary

on short peduncles, magenta (not cultivated)

see Turbina shirensis.

see Turbina stenosiphon.

47. I. adenioides.

29. I. lapathifolia.

41. I. alba.

19. I. magnusiana.

20. J. pes-tigridis.

36. I. coptica.

43 I. digitata.

37. I. dasysperma.

38. I. cairica.

40. I. heptaphylla.

39. I. hochstetteri.

6. I. oenotherae.

721

Sepals glabrous or, if hairy, flowers in 1 - to few-flowered cymes

on peduncles 3-18 cm. long and corolla pale mauve:

Leaves palmately (3-) 5-9-fid with linear segments under

5 mm. wide (often very narrow); corolla magenta;

stems often rather slender, glabrous, terete 34. I. bolusiana.

Leaves trisect to palmately 5 (-7) sect with linear, lanceolate or

oblong segments usually over 5 mm. wide; stems

usually stout; flowers pale mauve:

Peduncles and pedicels very short, flowers solitary; corolla

pale mauve with magenta centre (3-) 5-8 cm. long;

plant not cultivated 33. I. welwitschii.

Peduncles 3-18 cm. long, flowers 1 or few, cymose; corolla

pale mauve, 3-4-5 cm. long; plant cultivated for its

edible tubers and occasionally found as a culture-relic 23. I. batatas.

1. I. coscinosperma Hochst. ex Choisy in DC., Prodr. 9 (1845), p. 354; Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 124; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), p. 138; Hutch & Dalz., FI. W. Trop. Afr. 2 (1931), p. 216.

Type; Not designated. Choisy mentioned Kotschy 17 and 376 (from Kordofan).

There is a specimen of Kotschy 376 at Kew and this gathering is proposed here as the

lectotype (actual holotype specimen in herb. Geneva).

Annual. Stems several from the apex of the main root, usually stout, often angular,

thinly hairy, glabrescent, at first suberect but soon prostrate, up to several metres

long (3 m. in fertile soil on the Springbok Flats). Leaves often secund, from linear-

lanceolate to oblong, 3-12 cm. long, 0-5-5 cm. wide, entire or somewhat repand,

acute when young, but later subacute to rounded at the apex, usually distinctly

mucronate, usually gradually narrowed into a narrowly cuneate base, occasionally

broadly cuneate to almost rounded at the base, sparsely strigose mainly on the nerves;

petioles usually much thinner than the adjoining internodes, more or less thinly

subhirsute 0-5-3 (-6) cm. long. Inflorescences consisting of few-flowered clusters on

very short peduncles, on young stems solitary, or the short pedicels fasciculate; bracts

linear-subulate, hairy, about 4 mm. long. Sepals covered with long, white, rather

stiff hairs, sometimes only ciliate, ovate-lanceolate to lanceolate, attenuate into a

linear-lanceolate narrow apex, more abruptly acuminate from an ovate base in fruit,

subequal, the inner ones slightly narrower, 6-8 mm. long, in fruit up to 10 (-12) mm.

long. Corolla slightly longer than the calyx, in all annotated specimens from southern

Africa white, narrowly funnel-shaped. Capsule globose, glabrous, 6-8 mm. in diam.,

crowned by the persistent stylebase. Seeds very shortly pubescent, about 3 mm. long.

Abyssinia, Sudan, West Tropical Africa, Southern Rhodesia, Transvaal, Griqua-

land-West, South West Africa.

South West Africa. — Grootfontein, Tsumeb: Dinter 7607 (BOL, PRE); Groot-

fontein, Schoenfelder S560 (PRE); de Winter 2952 (PRE); farm Kumkauas: Kinges

2854 (in herb. Kinges; sepals quite glabrous), 2788 (PRE, sepals hairy); district

unknown, Otjisambira (?) Volk 2929 (PRE).

Cape Province. — Barkly West, Newlands: Wilman h. no. 3141 (KMG also BOL).

Transvaal. — Rustenburg: Sinclair s.n. (PRE). Brits: Dyer & Verdoorn 3424,

v.d. Linde s.n., Mogg s.n. (all PRE). Pretoria, Roodeplaat: de Wet s.n.; nr. Rust de

Winter: Meeuse 9526 (PRE). Waterberg, Warmbaths: Munro s.n. (PRE), Meeuse

9109 (PRE, L); Marais 262 (PRE, SRGH). nr. Naboomspruit: Galpin M. 530, M. 593

(PRE). Potgietersrust, Springbok Flats: Pole Evans 3937 (PRE). Letaba, Shingwedzi:

v.d. Schijff 3867 (PRE). Nelspruit, Kaapmuiden: Mogg s.n. (PRE).

Southern Rhodesia. — Wankie: Eyles 1298 (SAM, SRGH).

722

Some of the plants cited above were compared by Mr. B. de Winter with Kotschy

376 at Kew. He reported that they agree with the Abyssinian I. coscinosperma, although

the flowers in the South African specimens are always white and those of I. coscino-

sperma were reported to be red. However, there are specimens at Kew from Abyssinia

with pink flowers and the colour of the flowers does not seem to be very important;

in all essential characters the South Africa plants agree with the specimen Kotschy 376

and similar plants from Abyssinia and the Sudan.

I. coscinosperma is a very variable species. The leaves may vary from narrowly

lanceolate to linear-oblong or oblong, but they are usually narrowed towards the base,

which is cuneate or occasionally rounded to truncate, but (in the specimens seen)

never distinctly hastate-sagittate as in I. leucanthemum (see below). The sepals vary

from glabrous or glabrous with ciliate margin to densely hirsute. The form with

glabrous sepals has been distinguished as var. glabra (e.g., in FI. Trop. Afr.), but I

do not think this variety can be upheld.

I. coscinosperma is closely related to I. leucanthemum (Klotzsch) Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 124; Baker & Rendle, op. cit., p. 137 [= Calycanthemum leucan-

themum Klotzsch in Peters, Reise Mossamb., Bot. 1 (1861), p. 244, t. 40], and I even

thought that they might be identical. However, Dr. H. Wild (Salisbury) kindly pointed

out to me that in his opinion they are distinct. He had collected specimens (Wild

4077 in SRGH and PRE) in Elrungwe, Zambezi Valley, Menswa Pan (S. Rhodesia),

not very far from the type locality of I. leucanthemum, which are an excellent match

of the plate in Peters, Reise Mossamb. (t. 40). After having seen his specimen I fully

agree with him. The differences are: (a) the leaves of the specimens under discussion

are truncate to hastate-sagittate at the base and only up to about 5 cm. long; they are

obtuse and minutely mucronate (in I. coscinosperma usually over 5 cm. long and often

acute to acuminate and with a fairly long mucro); (b) the flowers appear to be solitary

(they are almost invariably clustered in I. coscinosperma)-, (c) I. leucanthemum is a

perennial, whereas I. coscinosperma is an annual. There may also be differences in

the fruits and seeds but unfortunately the fruits of I. leucanthemum are not known.

1. leucanthemum seems to be a rare and/or very local species (Lower Zambesi Valley).*

I. coscinosperma has been reported to be noxious in cultivated land in the Transvaal

(Springbok Flats) by rapidly covering patches of ground with its long creeping stems.

2. I. eriocarpa R. Br., Prodr. FI. Nov. Holl., ed. 1 (1810), p. 484; Choisy in DC.,

Prodr. 9 (1845), p. 369; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 136; Exell, Cat. Vase. PI. S. Thome (1943), p. 250; Ooststr. in Steenis, FI.

Mak, Ser. I, 4-4 (1953), p. 462. I. hispida (Vahl) R. et S., Syst. 4 (1819), p. 238,

non Zuccagni (1806); Hall. f. in Engl. Bot. Jb. 18 (1893), p. 123; Hutch. & Dalz.,

Fh W. Trop. Afr. 2 (1931), p. 216; Ooststr. in Blumea 3 (1940), p. 490. Convol-

vulus hispidus Vahl, Symb. Bot. 3 (1794), p. 29. Jacquemontia thomensis Henriq.

in Boh Soc. Broter. 10 (1893), p. 143 (fide Exell, l.c.); Baker & Rendle, op. cit.,

p. 86.

Type: Not seen.

Annual. Stems twining or prostrate, herbaceous, slender to rather stout, slightly

striate, retrorsely to patently pilose, 1-2 m. long. Leaves herbaceous, thin, cordate-

lanceolate to cordate-oblong, occasionally ovate-cordate, thinly strigose to subglabrous

but the nerves usually rather densely pilose, 2-9 cm. long and 0-5-5 -5 cm. wide; the

apex long-attenuate to acuminate, obtuse or acute, mucronate, the base cordate with

rounded shallow sinus and rounded basal lobes, the margin entire, ciliate or subciliate;

* Recently it was collected by Mr. B. de Winter in the Okavango, South West Africa. It would

key out in the key on p. 714 near /. eriocarpa and I. plebeia.

723

petioles thin, terete, pilose like the stems, 1-8 cm. long. Inflorescences axillary, sessile

or with a very short pilose peduncle which only in exceptional cases attains a length

of 1 cm., solitary or occasionally 2 in one axil, 1-3- or occasionally densely many-

flowered; flowers sessile or on very short up to 3 mm. long pilose pedicels; bracteoles

linear or lanceolate, pilose; lower ones 3-8 mm. long, upper ones shorter. Sepals

linear-acuminate from an ovate base, subequal, the inner ones slightly narrower, 7-8

mm. long, pilose. Corolla tubular to funnelshaped, pink or mauve, darker inside,

7-9 mm. long; midpetaline areas pilose. Ovary with long hairs. Capsule ovoid-

globose to globose, hairy, apiculate by the style-base, 5-8 mm. in diam. Seeds glabrous,

minutely reticulate, 3-3-5 mm. long.

Tropical Africa (extending into the Transvaal), Madagascar, tropical Asia to

N. Australia.

Cape Province. — Ceres Div., a single record, probably locality incorrect: Joubert

s.n. in BOL, H. no. 23269.

Transvaal. — Petersburg, Zoutpansberg, Letaba, Pilgrim’s Rest, Nelspruit, Bar-

berton.

3. I. plebeia R. Br., Prodr. FI. Nov. Holl., ed. 1 (1810), p. 484; Ooststr. in Blumea 3

(1940), p. 492 and in Steen., FI. Males. Ser. 1, 4-4 (1953), p. 463. ? Convolvulus

biflorus L., Sp. PL, ed. 2 (1763), App., p. 1668. ? Ipomoea biflora (L.) Pers.,

Synops. PL 1 (1805), p. 183; Hemsley in J. Linn. Soc., Bot., 26 (1890), p. 160.

Aniseia biflora (L.) Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 483, and in

DC., Prodr. 9 (1845), p. 431. Ipomoea cynanchifolia C. B. Clarke in Hook., FI.

Br. Ind. 4 (1883), p. 208, ex parte; emend. Hall. f. in Bull. Herb. Boiss. 6 (1898),

p. 538, non Meissn. (1869).

Type: There is no type of Convolvulus biflorus L. in the Linnean Herbarium or

in the Stockholm herbarium. The identity with the species I. plebeia R. Br. is highly

probable on account of the following considerations. Linne’s original description

agrees very well with the plants referred to I. plebeia or I. cynanchifolia, except for the

phrase: “ Corollae . . . lobis apice trifidis medio minore’\ However, this might refer

to the midpetaline areas of the corolla-lobes. The Convolvulaceae of China were

enumerated by Hemsley in J. Linn. Soc., Bot., 26 (1890) and there are only a few species

of Ipomoea with small flowers known from that area. That the Linnean species is an

Ipomoea is practically certain, because he described the stigma as follows: “ Stigmata

duo, capitata ”. The only other small-flowered Ipomoea from China, I. sinensis, has

a different inflorescence and does not fit the description of Convolvulus biflorus. Among

the unpublished notes left by Hallier in the Rijksherbarium, Leiden, is a sheet of

Ipomoea biflora with the synonymy “ Aniseia biflora ” and “ Ipomoea cynanchifolia ”

(Van Ooststroom, in litt.). Dr. Hallier had apparently arrived at the same conclusion,

probably also by a process of elimination. In the absence of a recognised type specimen,

the name I. plebeia is retained here. The type of this species is Banks & Solander s.n.

(in BM) from Australia.

This species occurs in two different forms of which the typical one, I. plebeia

subsp. plebeia is restricted to Asia, Indonesia and Australia and the other one, for

which the name I. plebeia subspec. africana is proposed here, is widespread in Africa:

I. plebeia R. Br. subsp. africana A. Meeuse, nom. nov. — I. gemimflora Welw., Apontam

(1859), p. 79, pro parte?; Baker & Wright in This. -Dyer, FI. Cap. 4, 2 (1904),

p. 62. I. cynanchifolia Baker & Rendle in This. -Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 137.

724

The African material differs slightly but constantly in the pubescence (which is

always laxer than in Australasian material) and in the sepals (which are narrower at

the base than in the typical form). The African subspecies was amply described in

Flora Capensis 4, 2, p. 62 as “ I. geminiflora Welw.” but this name cannot be applied

(see “ Notes ”). It was also described in Flora of Tropical Africa 4, 2, p. 137, as

“ I. cynanchifolia ” but this name is preoccupied in Ipomoea. In order to avoid any

difficulties connected with these names, not only a new name for the African subspecies

but also a neotype is proposed (Neotype: Codd 7785 from Barberton, Transvaal,,

in PRE).

Annual. Stems herbaceous, prostrate or climbing, slender, laxly pilose to hispi-

dulous, up to at least l m. long. Leaves oblong-cordate to cordate or triangular-cordate,

glabrous or nearly so except on the veins mainly on the lower surface, 2-10 cm. long,

and 0-5-6 -5 cm. wide; the basal lobes rounded; the basal sinus broad or narrow,

rounded; the margins entire, ciliate; the apex attenuate to shortly acuminate with an

acute or subacute point ending in a terminal mucro; the petioles slender, striate, pilose

to hispidulous like the stems, 0-5-7 cm. long. Inflorescences axillary, sessile or with

a peduncle up to 1 cm. long; flowers usually geminate or solitary, rarely 3 together,

on slender, terete, pilose or glabrous, 5-8 mm. long pedicels: bracteoles minute,

lanceolate, pilose 2-2-5 mm. long. Sepals herbaceous to subcoriaceous, unequal,

glabrous except the ciliate margins, or sometimes also hairy on the back or occasionally

almost completely glabrous, obtuse to subacute, 5-7 (-8) mm. long; two outer ones

ovate-acuminate or linear-acuminate from a broadly triangular, slightly cordate to

truncate base, third sepal somewhat oblique, fourth and fifth sepals much narrower,

oblong-lanceolate to linear. Corolla white, funnel-shaped, slightly longer than the

calyx; the limb 5-lobed-5-angled, about 9 mm. in diam. ; midpetaline areas distinct,

pilose outside mainly towards the tips. Ovary glabrous. Capsule broadly ovate to

globose, cuspidate by the persistent style-base, glabrous, 6-8 mm. in diam. Seeds

shortly brownish- or greyish-pubescent, 3-4-5 mm. long.

Africa (extending into South West Africa, the Transvaal and Natal).

South West Africa. — Oshikango: Loeb 15 (PRE). Grootfontein, Tsumeb: Dinter

2926 (SAM), 7498 (BOL, PRE); Grootfontein: Schoenfelder S610 (PRE), prob.

Grootfontein: Schoenfelder S852 (PRE); Namutoni: Breyer s.n. (PRE).

Transvaal. — Petersburg, Chuniespoort: Codd & Dyer 7757 (PRE); Blaauwberg:

Codd 7967, 8728 (PRE). Zoutpansberg, Mara: Eastwood s.n. (PRE); Louis Trichardt:

Gerstner 5938 (PRE); Wylies Poort: Meeuse 9196 (PRE). Letaba: Letaba, Krige 154

(PRE). Nelspruit, Alkmaar: Rogers 74 (GRA); Kruger National Park: v.d. Schijff

2417, 2502, 2728 (PRE); Kaapmuiden: Mogg s.n. (PRE). Barberton, Crocodile

Poort: Rogers 23986 (NH, J); Louw’s Creek: Van Dam s.n. (PRE); Barberton;

Liebenberg 2364 (PRE); Taylor 6946 (PRE); Codd 7785 (PRE).

Natal. — Amanzimtoti: Forbes 646 (NH). Verulam: Schlechter 2914 (BOL).

Inanda: Wood 258 pro parte (NH, with I. sinensis subspec. blepharosepala)\ Pt.

Dunford: Gerstner s.n. (NH).

Portuguese East Africa. — 40 km. from Lourengo Marques on Goba Road:

Esteves de Sousa 66 (PRE); near Goba: Myre & Balsinhas 670 (LM, PRE).

Southern Rhodesia. — Bulawayo, Linyati River: Moss 18548 (J). Ndanga:

Wild 2131 (SRGH). Hartley, Poole: Wild 1012 (SRGH). Mazoe: Wild 3770 (SRGH).

Nuanyadzi Riv., Sabi River: Wild 2497 (SRGH). Umtali: Chase 656 (SRGH).

Lomagundi, Doma: Whellan 283 (SRGH).

The name Ipomoea geminiflora Welw. was based on Welwitsch 6235. The sheet

of this gathering in BM to be taken as the holotype appears to be a mixture of I. plebeia

(only a scrap) and /. verticillata Forsk. (the two larger pieces). The same number

725

(6235) in other herbaria contains either the one species or the other. According to

Hallier, this number in the herb. De Candolle and Berlin is, or was, I. plebeia; in Kew

it is I. verticillata. I saw a duplicate in COI and this is only I. verticillata. Welwitsch’s

own description is non-committal and cannot be used to decide which form is meant.

However, both species have older names [I. plebeia (1810), and I. verticillata Forsk.

(1775), respectively] and I. geminiflora Welw. must either be regarded as a nomen

confusum (because the holotype is a mixture of two species) or be placed in the synonymy

of I. verticillata Forsk.

The name I. cynanchifolia C. B. Clarke is based on “ Wallich, Cat. no. 1399 ”,

but as Hallier pointed out, this number 1399 (in herb. Geneva) contains a mixture of

three species: No. 1399/1 contains the species under discussion, but not the nos. 1399

B-D, which represent I. eriocarpa R. Br. No. 1399/1 again is a mixture of two species,

viz. Burmese plants (= I. cynanchifolia C. B. Clarke sensu Hall, f.) and plants from

Sikkim which Hallier referred to I. tenuirostris Choisy. There can be no difficulty

caused by discrepancies between the herbaria having specimens of Wallich “ 1399 ”

because the name I. cynanchifolia C. B. Clarke (1883) non Meissn. (1869) is illegitimate

in any case and the other plants found under the number Wallich 1399 also have older

names (/. eriocarpa R. Br. 1810, I. tenuirostris Choisy 1845).

4. I. gracilisepala Rend/e in Jl. Bot. 39 (1901), p. 12; Baker & Wright in Dyer, FI.

Cap. 4, 2 (1904), p. 58; A. Meeuse in R. A. Dyer, Flow. PI. Afr. 31 (1956),

pi. 1217b. I. xiphosepa/a Baker in Dyer, FI. Cap. 4, 2 (.1904), p. 58, non I. xiphose-

pala Baker in Kew Bull. 1894, p. 69.

Type: Zeyher 1224 in BM (photo in PRE).

Annual. Stems several from the base, prostrate, or occasionally ascending (mainly

when still young), up to about 75 cm. long, subterete, more or less angular or sulcate,

at first densely, later finely and shortly pubescent. Leaves often secund, erect on creeping

stems, oblong or lanceolate-oblong to lanceolate, 2-6 cm. long and 0-5-2 cm. wide,

usually hastate-truncate or auricled at the base, often narrowed above the basal lobes,

sometimes some lanceolate with narrow base, very rarely (mainly in young plants)

all lanceolate; the apex obtuse, minutely mucronate, the basal lobes entire, bifid or

with 2-3 teeth; the edge entire to subrepand or sometimes with a few minute teeth,

upper surface dark green, glabrescent, lower surface paler and more densely hairy;

petioles 5-10 (-14) mm. long, glabrescent. Peduncles almost invariably 1-flowered,

short or 0, in fruit up to 7 mm. long, thinly hairy with rather long hairs as are bracteoles,

pedicels and sepals; bracteoles linear-lanceolate, 4—7 mm. long, obtuse or subacute;

pedicels usually under 8 mm. long, in fruit up to 12 (-15) mm. long. Sepals erect,

subequal, lanceolate to linear, acuminate, obtuse, 7-11 mm. long, 1-2 mm. wide, in

fruit spreading especially at the apex and attaining 12 mm. by 4 mm. Corolla narrowly

funnel-shaped, pinkish mauve with darker centre or rarely white, 12-16 mm. long,

the limb spreading, somewhat 5-angled, 10-12 mm. in diam. ; midpetaline areas distinct,

hairy towards the tips. Capsule globose, hirsute, 6-8 mm. in diam. Seeds brown

with a villous tomentum of adpressed stiff grey hairs, 4-5 mm. long.

South Africa and one record from South West Rhodesia.

Cape Province. — Griqualand W., Yryburg: Brueckner 398 (KMG); Rodin 3506

(BOL); Henrici 52 (PRE); Foley 2770 (PRE); Mogg 8671 (PRE).

Transvaal. — Lichtenburg: Sutton 402 (PRE). Potchefstroom: Louw 1727 (PRE).

Rustenburg or Brits, Castle Gorge: Meeuse 9368 (PRE). Brits, nr. Wolhuterskop:

Meeuse 9276 (PRE). Pretoria: Smuts & Gillett 2046, 2075, 2081, 2520 (PRE); Meeuse

9023, 9047, 9070 (PRE). Bronkhorstspruit : Meeuse 9252 (PRE). Groblersdal: Codd

726

8442. Waterberg, Warmbaths: Bolus 12162 (BOL); Leendertz H. no. 6671 (PRE);

Moss 16585 (J); nr. Ons Hoop: Codd 8457; nr. Hermanusdoorns: Meeuse 9653.

Pietersburg: Bolus 10905 (BOL, NH, GRA, PRE); Codd 7933 (PRE). Lydenburg:

Codd & Dyer 7726, 7739 (PRE). Without exact locality, “ Magaliesberg ” : Burke,

isotype of “ I xiphosepala ” Baker of FI. Cap. (SAM).

Orange Free State. — Vredefort, Parys: N.N. 557 (BOL).

The gathering of “ I. xiphosepala ” quoted by Baker, viz., a specimen leg. Burke

from the Transvaal (in SAM) is a very good match of the type of I. gracilisepala. The

specimen of Burke s.n. (SAM) I have seen has hastate leaves and the peduncles are not

ebracteate as indicated by Baker. Of all the specimens I have seen, not one had only

lanceolate leaves except very young plants; at least some of the leaves are hastate.*

Mr. de Winter compared the actual types and he agrees that I. xiphosepala of FI. Cap.

is a synonym of I. gracilisepala. I. xiphosepala Baker (1894) is a different plant from

Angola, which is identical with I. linosepala Hall. f. (1893).

A specimen collected by Codd & Dyer (no. 7739) in the Lydenburg district after

a very rainy season is very luxuriant in that the flowers are present in fascicles of up to

7 flowers per axil.

This species is probably often overlooked on account of its small flowers, but it

is locally very common, at any rate in the Pretoria district, and probably more generally

distributed than the specimens cited above indicate.

5. I. hackeliana ( Schinz ) Hall. f. in Engl. Bot. Jb. 18 (1893), p. 126; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 146. Aniseia hackeliana Schinz in Verhandl.

Bot. Ver. Brandenb. 30 (1888), p. 274.

Type: Schinz 749 from Olukonda (Ovamboland), South West Africa, in herb.

Zurich.

Annual. Stems several from the base, up to about 75 cm. long, prostrate, with

soft patent white hairs when young, glabrescent. Leaves ovate-cordate, usually secund,

more or less distinctly crenate or crenate-dentate, acute to rounded, sparsely pilose

on both surfaces, pellucidly glandular when dry, 1-4-5 cm. long and 0-75-3-5 cm.

wide; basal sinus usually wide and shallow with the blade cuneately decurrent into

the petiole, petioles rather slender, pilose like the stems, 4-28 mm. long. Flowers

solitary or in a few-flowered fascicles; peduncles 0; bracteoles practically sessile,

subcordate at the base, acute, ciliate on midrib and margin, 2-4 mm. long; pedicels

slender, patently hairy, up to 20 mm. long. Sepals green, often suffused with purplish

red, about 10 mm. long, hairy, very unequal; the outer ones ovate from a cordate

base, obtuse or subacute, the two inner ones lanceolate, acute ; all accrescent, becoming

papyraceous with distinct finely raised veins, attaining 15 mm. in length and the outer

ones 1 1 mm. in width. Corolla funnel-shaped, pale pinkish mauve with darker centre

or occasionally white, 12-14 mm. long; midpetaline areas slightly hairy towards the

tips. Ovary hairy. Capsule subglobose, densely pilose with long white hairs, 6-8

mm. in diam. Seeds black, hairy, 3-5 mm. long.

South West Africa. — Ovamboland, Olukonda: Schinz 749 (BOL, GRA,

isotypes!). Grootfontein, Tsumeb: Dinter 7522 (BOL, L, PRE). Okahandja: Dinter

951 (SAM, PRE); Bradfield 385 (PRE). Gibeon: Range 1451 (BOL, SAM).

* The leaves are narrowed at the base in the closely related lpomoea lyciifolia Merxm. (in Trans.

Rhod. Sci. Assoc. 43 (1957), p. 40] of which I have seen several gatherings from Southern Rhodesia.

Its ether characters are so close to /. gracilisepala that it cannot be more than a variety of the latter

and thus becomes /. gracilisepala Rendle var. lyciifolia (Merxm.) A. Meeuse, nov. stat.

727

Bechuanaland Protectorate.— Mochudi: Rogers 6518 (BOL, GRA, J); Harbor

in herb. Rogers no. 6528 (BOL); Guive Pits: v.d. Merwe 22 (BOL).

Transvaal. — Zoutpansberg, Wylies Poort: Rodin 4228 (PRE), Meeuse 9194

(PRE). Waterberg, nr. Ons Hoop: Codd 8468 (PRE).

6. I. oenotherae {Vatke) Hall, f in Engl. Bot. Jb. 18 (1893), p. 125; Baker & Wright

in Dyer, FI. Cap. 4, 2 (1904), p. 49; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), p. 145. Convolvulus oenotherae Vatke in Linnaea 43 (1882), p. 520.

Ipomoea petunioides Baker in Dyer, FI. Cap. 4, 2 (1904), p. 63. I. cecilae N.E.Br.

in Kew Bull. 1906, p. 166. I. pachypus Pilger in Engl. Bot. Jb. 41 (1908), p. 296;

Dinter in Fedde, Repert. 18 (1922), p. 431. I. cecilae N.E.Br. var. quinquesecta

Merxm. in Mitteil. botan. Staatssamml. Mtinchen H.6 (1953), p. 203, and var.

anomophylla Merxm., op. cit., p. 204. I. linear iloba Chiov. in Reale Acad. d. Ital.

(1939) 4, p. 174, f. 51, e descr. et icon.

Type: Hildebrandt 2767 from East Africa in herb. Berlin (now destroyed). Isotypes

may be extant elsewhere.

Perennial , with a fusiform rootstock up to 3 cm. thick and up to about 30 cm.

long, producing annual rosulate leaves followed by several unbranched prostrate

stems up to 40 cm. long in the specimens seen; the latter rather thick, angular, the

younger parts with silvery- white pubescence, the older parts glabrescent. Basal (rosulate)

leaves, if present, long-petioled (petiole up to 7 cm. long), the blade 4-10 cm. long

either lanceolate or linear, entire, up to 1 cm. wide, or with 1-2 lateral teeth or lobes,

or repando-pinnatisect, with 3-7 lobes, the terminal lobe usually distinctly longer

to much longer than the lateral ones. Cauline leaves shorter (2-6 cm. long) and on

shorter petioles (0-5-4 cm. long), entire or more often 3-7-lobed, cuneate at the base,

the lobes entire or slightly repand, 2-6 mm. wide; originally, as is the petiole, laxly

covered with soft silvery- white hairs, blade soon glabrescent; main nerves prominent

beneath. Flowers solitary on very short (0-5 mm. long) peduncles; bracteoles linear-

subulate or linear-filiform, 10-14 mm. long and ^—1 mm. wide; originally covered

with the characteristic silvery- white pubescence as are pedicels and sepals; pedicels

rather stout, subclavate, 5—1 5 mm. long. Calyx often rose-red in bud and in flower,

12-15 mm. long; sepals ovate or ovate-lanceolate, aristate, ciliate and with median

keel of hairs, sometimes with a few teeth on the margin, outer ones herbaceous, inner

ones paler and more membranous, shorter than the outer ones, all becoming broader

and brown in fruit. Corolla narrowly funnel-shaped with spreading limb, 3-5 cm. long,

described as mauve, pink or purple, those of living specimens I have seen a deep mauvish

pink to light magenta; midpetaline areas glabrous. Capsule globose, glabrous, straw-

coloured, 8-10 mm. in diam. Seeds 3-4 mm. long, rather thick, densely velvety

pubescent with (often pinkish-grey) adpressed hairs.

Widely distributed, but apparently not very common or of local occurrence, in

a more or less continuous range from South West Africa, Transvaal, Natal and Southern

Rhodesia to East Africa, Abyssinia and Somaliland.

S. W. Africa. — Grootfontein, Otavi: Dinter 5259 (BOL, PRE, SAM). Aukas:

Dinter 619 (SA.M) — District unknown: Volk 2824 (PRE).

Cape Province. — Mafeking: Brueckner 578 (KMG, PRE).

Transvaal. — Lichtenburg: Sutton 403 (PRE: this number on two different

sheets with different dates, apparently two different collections, both representing

I. oenotherae)', Liebenberg 21 (PRE); Acocks 12475 (PRE). Pretoria, near Pienaars

River: Meeuse 9540 (PRE); Rust de Winter: Meeuse 9022 (L, PRE); Bapsfontein:

728

Meeuse 9659 (PRE). Benoni: Bradfield 322 (PRE). Middelburg, Wonderfontein::

Bolus 12169 (BOL). Potgietersrust, near Grass Valley: Meeuse 9601 (PRE).

Natal.— Mooi River, The Thorns (nr. Greytown): Wood 4490 (NH).

Southern Rhodesia. — Matobo: West 2485 (SRGH); Hunyani: Eyles 4590-

(SRGH); Marandellas: Dehn 194 (SRGH); South Marandellas: Myres in herb.

Eyles no. 7400 (SRGH); Mazoe, Umvukwes: Wild 3923 (SRGH).

Northern Rhodesia. — Mazabuka: Veterinary Off. C.R.S. no. 54, 99, 407, 451,.

517 (PRE).

Tanganyika Territory. — W. slopes of Kilimanjaro: Greenway 6901 (PRE).

Kenya. — Nairobi: Verdcourt 510 (PRE).

No type or isotype was available, but Mr. de Winter has seen three S. African1

specimens, named I. oenotherae by Hallier. The variability of the species is considerable,

and the original description is well within the limits of variation. Easr- African specimens

seen by Mr. de Winter and myself are conspecific with the S. African and Rhodesian

specimens, so that the indentification of the S. African and Rhodesian plants with the

E. African ones (as already done in FI. Cap.) seems to be well founded. The types of

I. petunioides and I cecilae were compared by Mr. de Winter at Kew and they are not

essentially different, but fall within the range of variation. The type of I. pachypus

Pilger ( Ellenbeck 1209 from Somaliland) I have not seen, but Dinter (l.c.) has pointed

out that in Engler’s Pflanzenwelt Afrikas 1, 2 (1910), p. 589 this species was recorded

from the Otavi-valley in S. W. Africa by Dinter, so that the S. W. African specimen

had evidently been identified with the Somaliland plant in the Berlin Herbarium. At

any rate, Dinter 619 (probably the gathering referred to by Engler) in SAM and other

specimens from S. W. Africa agree with specimens from other areas, and also with

Pilger’s description of I. pachypus.

Merxmiiller recently described two varieties of I. cecilae, but in view of the con-

siderable variability, these varieties cannot be upheld.

I. oenotherae is closely related to I. commatophylla A. Rich, (which, however, is

an annual without a thick swollen root and white flowers with or without a purple

tube) and especially so to I. commatophylla var. angustifolia Oliv., which also lacks

a swollen root and has branched, rather shrubby shoots and, according to Mr. de

Winter, is worthy of specific rank. These two forms are also closely related to I. poly-

morpha R. et S. (= I. heterophylla R.Br. non Ortega) from the E. Indies and N.E.

Australia, or even identical, according to Hallier [Engl. Bot. Jb. 18 (1893), p. 125]

and Van Ooststroom in Blumea 3 (1940), p. 493.

7. I. sinensis ( Desr .) Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 469; and in DC.r

Prodr. 2 (1845), p. 370; Hemsley in Jl. Linn. Soc. Bot. 26 (1890), p. 162; Gagn.

et Courch. in Lecomte, FI. Gen. Indo-Chine 4 (1915), p. 236. Convolvulus sinensis-

Desr. in Lamk., Encycl. Meth., Bot. 3 (1789), p. 557. C. calycinus Roxb., Hort.

Beng. (1814); p. 13, nomen nudum; id., FI. lnd. ed. Carey 2 (1824), p. 51, descr.,

non alior. C. hardwickii. Spreng., Syst. 4, cures post. (1827), p. 60, nomen illeg.

Aniseia calycina (Roxb.) Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 482 and

in DC., Prodr. 9 (1845), p. 429. Ipoemoea calycina (Roxb.) Benth. ex C. B. Clarke

in Hook f., FI. Brit. India 4, 1 (1883), p. 201, non Meissn. (1869); Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 129. I. hardwickii (Spreng.) Hemsley in Jl. Linn. Soc., Bot.

26 (1890), p. 160; Chiov. in Acad. Reale d’ltal. 4 (1939), p. 174, nomen illeg.

I. auxocalyx Pilger in Notizbl. Bot. Gart. Berlin 11 (1933), p. 819.

Type: A specimen leg. Moreau in herb. Jussieu (P), photograph of type in PRE.

729

There are two forms, only one of which is represented in Southern Africa. In

agreement with Dr. B. Verdcourt who is engaged in the treatment of the Convolvulaceae

for the Flora of East Tropical Africa, these two forms are referred to as follows:

I. sinensis ( Desr .) Choisy subsp. sinensis.

This form includes the specimens from Asia and is also represented in East and

Central Africa, but does not occur in Southern Africa.

I. sinensis (Desr.) Choisy subsp. blepharosepala ( Hochst . ex A. Rich.) Verdcourt,

stat. nov., in litt. I. blepharosepala Hochst. ex A. Rich., Tent. FI. Abyss. 2 (1857),

p. 72. I. cardiosepala Hochst. ex Choisy in DC., Prodr. 9 (1845), p. 393, 429,

in syn., nomen nudum. I. cardiosepala Hochst. ex Baker & Wright in This.-Dyer,

FI. Cap. 4, 2 (1904), p. 61; Baker & Rendle in This.-Dyer, FI. Trop. Afr. 4, 2

(1905), p. 147, non I. cardiosepala (H.B. et K.) Meisner (1869).

Type of subspecies: Schimper 1780 in P.

Annual herb. Stems several from the apex of the taproot, prostrate or twining,

or young shoots occasionally suberect; terete, slender, more or less densely covered

with white spreading hairs, at least when young, up to about 1 m. long (rarely longer).

Leaves thinly herbaceous, cordate, cordate-ovate or cordate-oblong, entire, obtuse or

acute, mucronate, 3-8 cm. long, 2-5 cm. wide; the basal sinus usually deep and rather

narrow but sometimes broad and shallow; basal lobes rounded: upper surface green,

thinly pilose to quite glabrous except for a few hairs on the main nerves, lower surface

usually a little more densely hairy and paler, the margin ciliate; petioles slender,

1-9 cm. long, usually much more densel) pilose than the blade with patent sometimes

subhirsute hairs. Inflorescences cymosely 1-3-flowered; peduncles slender, pilose to

subhirsute, 0 - 5—4 t-6) cm. long; bracteoles minute, lanceolate; pedicels 0-5-2 cm.

long, pilose, somewhat thickend towards the apex, at first erect, in fruit reflexed. Sepals

erect, very unequal, herbaceous, glabrous to pilose or subhirsute but almost invariably

ciliate with stiff white long hairs, 5-8 mm. long, accrescent in fruit; outer ones cordate

or auricled at the base, long-acuminate, very acute; inner ones much narrower,

lanceolate or linear, very acute. Corolla funnel-shaped, pale mauve with darker mauve

to magenta inside the corolla-tube and the well-defined midpetaline areas somewhat

darker mauve, occasionally completely white; the tube 8-12 mm. long, subcylindric;

the limb spreading, faintly 5-lobed-5-angled, 15-20 mm. in diam. ; midpetaline areas

strigose outside at least towards the tips. Stigma pale mauve (also in white-flowered

specimens). Caysule globose, glabrous, 6-9 mm. in diam., straw-coloured, apiculate

by the persistent style-base. Seeds densely and shortly velvety hairy to almost villous

with grey or fawn hairs, 4-5 mm. long. (Seeds not glabrous as erroneously stated in

FI. Cap.).

Throughout tropical Africa, Sudan, Abyssinia, extending into southern Africa,

also in Arabia. Recorded in the Union of South Africa from the following regions

and districts: South West Africa (northern part), Griqualand West, Bechuanaland,

Transvaal (Marico, Rustenburg, Brits, Pretoria, Groblersdal, Lydenburg, Barberton

and all districts to the north of these), Zululand, Natal (coastal districts as far south

as Durban, also Weenen district). On account of its inconspicuous flowers probably

often overlooked, but frequent in many places.

The nomenclature of this species is fairly complicated. The name Convolvulus

calycinus Roxb. was the basis of two binomials, viz., Aniseia calycina (Roxb.) Choisy

and Ipomoea calycina (Roxb.) Benth. ex C. B. Clarke. (N.B.: Clarke cites “ Benth.,

Gen. PL II, 872 ” as the author of the combination, but Bentham only mentioned that

in his opinion the genus Aniseia Choisy had to be incorporated in Ipomoea and he did

730

not mention Aniseia calycina specifically, so that he did not actually make the combi-

nation). The genus Aniseia was reinstated by Hallier, but only to include species with

smooth pollen. The species under discussion has spinulose pollen and is a true Ipomoea.

The specific epithet “ calycina ” cannot be used for this species, because it is preoccupied

in Ipomoea on account of the earlier I. calycina (H.B. et K.) Meisn. in FI. Bras. 7 (1869),

p. 260, a South American species. The name Convolvulus hardwickii Spreng. (1827)

is illegitimate because it was superflous in the genus Convolvulus, being based on the

same type as C. calycinus Roxb. and purely a new name for Roxburgh’s species. It

cannot be taken up as a “ new ” name in Ipomoea either because other epithets are

available. The name Ipomoea cardiosepala Hochst. remained a nomen nudum till

its publication with a description in FI. Cap. (1904), but, in Ipomoea, the specific epithet

had already been used by Meissner in FI. Bras. 7 (1869), p. 265 for a different (South

American) species.

The name Ipomoea blepharosepala Hochst. ex A. Rich, was based on a specimen

collected in Abyssinia and was validly published. The type specimen was studied

in the Paris herbarium by Dr. Verdcourt, who reported (in litt.) that the above mentioned

synonymy applies. Mr. de Winter, who upon my request compared many African

specimens with Indian ones, reported that there are slight and possibly constant

differences in the shape of the leaves and calyx. Hallier (1893) on the other hand

reduced I. blepharosepala to a synonym of I. calycina (Roxb.) C. B. Clarke. There can

accordingly be very little doubt that I. blepharosepala and I. sinensis are closely related,

but they are not identical, so that the former is better regarded as a distinct subspecies.

The specific epithet “ sinensis ” had already been used in Ipomoea by Fischer

in Hort. Gorenk. ed 2 (1812), p. 28, but this name is a nomen nudum and does not

invalidate the combination I. sinensis (Desr.) Choisy.

8. I. crassipes Hook, in Bot. Magt. 4068 (1844); Hall. f. in Bull. Herb. Boiss. 7 (1899),

p. 44; Rendle in Jl. Bot. 39 (1901), p. 14; Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 56; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 147.

I. calystegioides E. Mey. ex Drege, Zwei Pflanz. geog. Doc. (1843), p. 145, 153,

nomen nudum. Aniseia calystegioides (E. Mey. ex) Choisy in DC., Prodr. 9

(1845), p. 131. Ipomoea calystegioides [(E. Mey. ex) Choisy] Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 127. I. adumbrata Rendle and Britten in Jl. Bot 32 (1894), p. 173;

Baker and Rendle, op. cit., p. 145. I. greenstockii Rendle in Jl. Bot 34 (1896),

p. 38, and 39 (1901), p. 14: Baker & Wright, op. cit., p. 51. I. sarmentacea Rendle

in Jl. Bot. 39 (1901), p. 15; Baker and Wright, op. cit., p. 57. I. bellecomans

Rendle, in Jl. Bot. 39 (1901), p. 15; Baker and Wright, op. cit., p. 55.

Type: The actual type is the plate in Bot. Mag. t. 4068, but as Hooker mentioned

that the seeds were collected by Burke, “ Macalis Berg ”, I consider Burke 177 in herb.

Kew (from Magaliesberg) to be equivalent to a type specimen.

Perennial. Taproot tuberous, fusiform, rather thick (up to about 10 cm. in diam.),

black. Stems several to many from the base, annual, wiry, suffruticose or herbaceous,

suberect or usually all ultimately prostrate, up to 75 cm. long, slender, terete, more

or less densely and softly hairy to pilose or occasionally nearly glabrous, as are all

other vegetative parts, peduncles, pedicels, bracts and sepals. Leaves very variable in

shape and size, usually broadly lanceolate or ovate-lanceolate to ovate, sometimes

linear or linear-lanceolate, and, if so, often somewhat panduriform, or oblong, broadly

ovate to almost reniform, 1 • 5-8 cm. long and (3-) 8-28 mm. wide, rarely wider,

herbaceous, entire, acute, gradually acuminate or obtuse at the apex, truncate, rounded

or sometimes more or less cuneate or subcordate, rarely cordate or narrow and tapering

at the base, more or less penninerved, lateral nerves in all except very narrow leaves

731

4-6 on either side, usually 2 of them ascending from the base, a short one and a long

one, reaching the margin near the middle; all prominent beneath but considerably

thinner than the prominent midrib; petioles up to about 10 mm. long, but usually

shorter, very rarely up to about 3 cm. long, rather stout to very slender. Peduncles

usually 1 -flowered, less often cymosely few-flowered, longer or shorter than the leaves,

usually rather slender, terete; bracteoles usually remote from the calyx, variable in

size and shape but usally leafy, ovate-lanceolate or lanceolate-oblong and 0-5-1 -5

times as long as the sepals, sometimes narrower or smaller; pedicels somewhat

thickened, usually under 2 cm., often under 1 cm. long. Sepals unequal (10-) 16-22 (-28)

mm. long, the outer ones in well-developed specimens ovate, more or less triangular

or lanceolate, 7-11 mm. wide at the broad, often subcordate base, with distinct midrib,

tapering into the subacute to acuminate apex, in some specimens narrower, lanceolate;

inner ones much narrower and not subcordate at the base, often long- and rather abruptly

cuspidate from on oblong basal portion. Corolla funnel-shaped, purple, magenta or

occasionally mauve with lighter midpetaline areas, or occasionally pale mauve, pink or

creamy white with darker purple centre, 30-45 mm. long and 35-60 mm. in diam.;

midpetaline areas distinctly margined by raised veins, shortly hairy towards the apex.

Ovary glabrous. Capsule ovoid-globose, apiculate, glabrous, 8-10 mm. in diam.

Seeds glabrous or velvety-pubescent, about 6 mm. long.

Southern Africa from the Eastern Cape, Orange Free State, Griqualand West,

South West Africa northwards and extending into Angola and tropical East Africa.

Rare in S. W. Africa.

Recorded from the Eastern Cape (Bathurst, Somerset East, Albany, Queenstown,

Peddie, East London, Komgha, Kentani, Elliotdale, Umtata, Lusikisiki); Basutoland;

Orange Free State (Rouxville, Ficksburg, Bethlehem); Natal (from Durban to Estcourt

and Northwards, including Zululand as far North as at least Nongoma); Swaziland;

Transvaal (recorded from practically all districts except a few in the extreme S.W.);

Griqualand West (one record from Vryburg); S. W. Africa (4 records: Ekuja, Ozond-

jache, Oukongo, Okahandja); Bechuanaland (one record). In addition Southern

Rhodesia, Angola, northern part of Portuguese East Africa, etc.

The following specimens are of special interest: Burke 177 (equivalent to the type,

in BOL) from Magaliesberg. Burke s.n. from Somerset East (PRE) and Burke 348

(BOL), Gilfillan in herb. Galpin 6157 (PRE), three numbers cited under Ipomoea

greenstockii in FI. Cap.; photographs (in PRE ex BM) of the types of /. greenstockii

(Creenstock s.n.. Pilgrim’s Rest), /. bellecomans ( Zeyher 1213, from Pretoria) and /.

sarmentacea ( Greenstock s.n., Pilgrim’s Rest).

An extremely variable species of which the extremes are very different in habit,

size and shape of leaves, colour of flower, pubescence, etc. Even the seeds can be

quite glabrous or pubescent. The various forms are all united by intermediate specimens,

so that it is not practicable to distinguish varieties as was done by Hallier and by several

others. All these forms hang together by the following characters: very unequal

sepals (outer ones broad and usually subcordate at the base, inner ones much narrower

to subulate from a somewhat broader base), the usually conspicuous bracteoles which

are generally remote from the calyx and the usually 1-flowered peduncles. So far

as I have seen, the plants are either prostrate or suberect. Climbing forms have been

mentioned (e.g., in FI. Cap.), but I think this must be a mistake.

Ipomoea greenstockii is an erect form described as a different species, but many

plants form first erect stems and subsequently prostrate stems. I. bellecomans is another

erect form. /. sarmentacea is a form with small flowers. Mr. de Winter confirmed

this synonymy after having seen the types. Occasionally the peduncles are cymosely

few-flowered. I. hewittioides, which was recorded from Angola, is possibly such a

732

form, and similar specimens are also found elsewhere (Transvaal, Natal, etc.), although

according to a note from Mr. Killick none of the South African material quite matches

the isotype of 7. hewittioid.es in BM, so that 7. hewittioides may be distinct. Ipomoea

crassipes can easily be distinguished from similar prostrate species plants, such as

Turbina oblongata , Ipomoea ommaneyi, I. atherstonei and 7. pellita but it sometimes

closely resembles specimens of the tropical African species Ipomoea asperifolia.

However, the sepals in 7. asperifolia are not so conspicuously unequal, narrower and

usually subobtuse; the calyx as a whole is much narrower at the base than that of

7. crassipes and more oblong (usually more or less conical in 7. crassipes). In addition,

the petioles are generally longer than in 7. crassipes, the bracteoles are minute, linear

(usually conspicuous, ovate or oblong, foliaceous in 7. crassipes ), and, finally, the corolla

in 7. asperifolia is somewhat abruptly narrowed and more or less tubular at the base,

whereas the corolla in 7. crassipes is gradually narrowed and funnel-shaped. As far

as I could see, 7. asperifolia has not been recorded from South Africa (distribution:

Angola, Rhodesia).

9. I. pellita Hall, f in Engl. Bot. Jb. 18 (1893), p. 130. 7. ovata E. Mey. ex Drege,

Zw. Pfl. geog. Doc. (1843), p. 154, 195, nomen tantum, and ex Rendle in Jl. Bot.

39 (1901), p. 19 descr. ; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 54, inclus.

var. pellita (Hall, f.) Baker.

Type: Drege 4905 (according to Hallier in Herb. Geneva-Deless. and Vienna).

Perennial. Rootstock thick, fusiform. Stems prostrate, several to many from the

base, herbaceous but firm, robust, often somewhat angular, densely covered with

patent bristly hairs, up to 2 m. long. Leaves ovate to oblong, entire, rather firm, obtuse,

rounded or subemarginate to subacute at the apex, often mucronate; rounded, truncate

or shallowly cordate at the base, 4-10 cm. long and 2-6-5 cm. wide; when young

very densely covered with yellowish bristly hairs, more thinly hairy to glabrescent

when old, the margin with a yellowish dense fringe of hairs; petioles hairy like the

stems, always shorter, usually much shorter than the blades. Peduncles 3-15 cm.

long, bristly, bearing the flowers in a terminal dense few-flowered head or very rarely

a cyme; bracteoles linear or linear-subulate, bristly, 12-25 mm. long, or very rarely

in a cymose inflorescence the lowermost more leafy, up to 35 by 5 mm. ; pedicels 0

or in the cymose inflorescences up to 1 cm. long. Sepals subequal, very similar to

the bracteoles, very bristly, 18-25 mm. long, 1-2-5 mm. wide, gradually tapering into

an acute point, in fruit up to 6 mm. wide. Corolla funnelshaped, magenta, 4-7 cm.

long and in diam., slightly hairy on the midpetaline areas. Ovary glabrous. Capsule

subglobose, 12-15 mm. in diam., apiculate, glabrous. Seeds about 1 cm. long, dark

brown glabrous.

Endemic.

Recorded from: Eastern Cape (Komgha, Queenstown, Willowvale, Efliottdale,

Mt. Currie, Umzimkulu); Basutoland; Natal [Inanda, Camperdown, Pietermaritzburg,

Richmond, Umvoti, Lions River, Estcourt, “ Tugela ” (prob. Weenen), Nqutu, Vryheid,

Newcastle]; Transvaal (Wakkerstroom, Ermelo, Bethal, Witbank).

When Rendle validated Ipomoea ovata E. Mey. he was of the opinion that his

specimens, including Drege specimens (but different gatherings), differed from 7. pellita

Hall. f. However, when Baker & Rendle united the two species, they retained the name

7. ovata E. Mey. ex Drege, being the oldest name (although published as a nomen

nudum), as was common practice in those days. Under the present rules the name

7. ovata E. Mey. ex Rendle is invalidated by 7. pellita Hall, f., because the latter is the

oldest validly published epithet.

733

I. pellita closely resembles Turbina oblongata E. Mey. ex Choisy and, when

not in fruit, can sometimes only be distinguished from the latter with difficulty. The

strictly linear bracteoles and sepals of I. pellita are the only constant distinguishing

character (in T. oblongata the sepals are never strictly linear and the bracteoles very

rarely). As a rule I. pellita has denser inflorescences on long peduncles, long petioles

and setose bracts and sepals, whereas T. oblongata usually has 1-2-flowered inflorescences

on shorter peduncles, short petioles and pubescent, but not so bristly, bracts and sepals.

These characters break down occasionally.

The section Dasychaetia as proposed by Hallier does not seem very homogeneous

to me. Ipomoea pellita belongs rather in the section Calycanthemum near I. blepharo -

phylla, I. asperifolia and I. crassipes, although the usually capitate and fairly dense

inflorescences have apparently no counterpart in Calycanthemum. I. pellita could

equally well be placed in the section Pharbitis subsect. Cephalanthae on account of

the dense, capitate inflorescence but its prostrate habit does not agree with the climbing

habit of this section.

I did not see the actual type, but some of the specimens referred to this species

in FI. Cap. and by Rendle, viz., Wood 806 (NH, PRE), 3460 (NH), 3974 (NH), 4999

(BOL, SAM), and the descriptions leave no doubt about the identity of this species.

In FI. Cap. the corolla is erroneously reported to be glabrous but both Hallier

and Rendle state that it is “ sparse pilosis ” which is quite correct.

10. I. nil (L.) Roth, Cat. Bot. I (1797), p. 36; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 136;

House in Proc. New York Acad. Sci. 18 (1908), p. 203; Ooststr. in Blumae 3

(1940), p. 497, and in Steenis, FI. Males, Ser. I. 4-4 (1953), p. 465. Convolvulus

nil L., Sp. PI. Ed. 2 (1762), p. 219. Ipomoea hederacea Auct. non I. hederacea

(L.) Jacq., Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 159; Hutch.

& Dalz., FI. W. Trop. Afr. 2 (1931), p. 218.

Type: The specimen of Convolvulus nil in the Linnean Herbarium is not the same

species as the one usually considered to be I. nil. The original description and citations

by Linne (1762) do not seem to apply to this specimen (see under no. 11 I. purpurea).

Herbaceous annual. Stems usually twining, retrorsely hirsute. Leaves (broadly)

ovate to orbicular in outline, entire or 3-lobed; middle-lobe ovate to oblong, acuminate,

lateral lobes obliquely ovate to broadly falcate, acuminate; the base broadly cordate,

margin entire; both surfaces thinly to rather densely pubescent with adpressed hairs;

length 4-14 cm., width 3-12 cm.; petioles retrorsely hirsute, 3-16 cm. long. Inflores-

cence a 1- to few-flowered umbellate cyme; peduncles thicker than the petioles of

the subtending leaves, hirsute like the stems, 2-12 cm. long; bracteoles linear to

filiform, 5-8 mm. long; pedicels 5-10 mm. long, retrosely hairy. Sepals subequal,

17-28 mm., later to 35 mm. long, patently hirsute mainly in the basal portion; outer

ones with lanceolate, inner ones with narrowly lanceolate base, all with a long and

linear acumen. Corolla funnel-shaped, glabrous, pale to bright blue turning purple,

or reddish magenta, rarely entirely white, 5-9 cm. long and 4-7 cm. in diam. ; the

tube and limb always paler outside. Ovary glabrous. Capsule ovoid to globose,

apiculate by the style-base, glabrous, 3-celled, 8-15 mm. long and in diam. Seeds

6 or less, black when ripe, grey-puberulous, 5-8 mm. long.

Circumtropical, probably originally indigenous in Africa, often planted as an

ornamental and running wild.

1332796-7

734

This species has often been confused with I. hederacea (L.) Jacq., an American

species. Hallier [in Jb. Hamb. Wiss. Anst. 16, 3. Beih. (1899), p. 42], and later House

(l.c.) have pointed out that these two species are very similar, but not identical.

The main differences are the following: Corolla: 2-5-3 cm. long in I. hederacea,

(3-) 5-9 cm. long in I. nil. Calyx: Sepals dilated at the base in I. hederacea, not

dilated in I. nil. Leaves: Lobes of leaf usually contracted below in I. hederacea,

usually not contracted below in I. nil.

I. nil I have seen from South Africa only as cultivated specimens from Durban

and Pretoria, but it is likely that it may occur as an escape from culture elsewhere,

especially because the specimens under discussion produced a great quantity of

viable seed.

As in I. purpurea, the ovary and capsule are normally 3-celled and 6-ovuled and

6- (or less) -seeded, respectively.

11. I. purpurea ( L .) Roth., Bot. Abh. (1787), p. 27, and Cat. Bot. 1 (1797), p. 36; Hall. f.

in Engl. Bot. Jb. 18 (1893), p. 137; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904),

p. 59 (pro parte); Ooststr. in Blumea 3 (1940), p. 496, and in Steenis, FI. Males.,

Ser. I, 4-4 (1952), p. 465. Convolvulus purpureus L., Spec. PI. Ed. 2 (1762),

p. 219. Ipomoea gerrardiana Rendle in J. Bot. 39 (1901), p. 21; Baker & Wright,

op. cit., p. 54.

Type: The specimen in the Linnean Herbarium does not seem to represent the

type and does not agree with the original description by Linne. It may have been

acquired by Linne after 1762. Dr. Verdcourt has asked Mr. Stearn to study the

nomenclature of I. nil and I. purpurea', for the time being the status quo is maintained.

Herbaceous annual. Stems twining, terete, with short hairs mixed with longer

retrorse bristles. Leaves broadly ovate to suborbicular in outline, entire or (in South

African specimens very rarely) 3-lobed; the apex acuminate, the base cordate with

broadly rounded auricles; with short bristly hairs on both surfaces, 4-15 cm. by

2-5-12 cm.; petioles retrorsely hirsute, 2-15 cm. long. Inflorescences axillary, cymosely

1- to few-flowered; peduncles retrorsely hairy, 3-18 cm. long; bracteoles linear to

filiform, up to 7 mm. long; pedicels shortly hairy or with a few bristles, 8-15 mm.

long, recurved in bud, afterwards erect, in fruit up to 20 mm. long, reflexed again and

thickened towards the apex. Sepals unequal, herbaceous, 10-15 mm., in fruit up to

20 mm. long; outer ones oblong, acute with bristly patent hairs in basal portion,

glabrous towards the apex, inner ones with narrow scarious margins, linear-oblong

to linear, acute, witu a few bristles near the base. Corolla funnel-shaped, glabrous,

purplish blue with reddish midpetaline areas and paler tube, reddish purple or magenta-

pink to white, in some cultivated forms sometimes variegated with blue, purple or pink

dots or strips, 5-6 cm. long and about 6 cm. in diam. or in depauperated specimens

considerably smaller. Capsule globose, glabrous, 3-celled. Seeds glabrous or pilose

at the hilum.

Originally a native of America, from New Mexico and Virginia to the Argentine

and Uruguay, often cultivated in the tropics and running wild; in South Africa in

several places becoming a noxious weed, recorded from most districts of the Transvaal,

Northern Natal and Zululand, Orange Free State and occasionally in the Eastern

Cape. Frequently cultivated (“ morning glory ”)■ Not recorded from South West

Africa or Bechuanaland.

This species has often been confused with /. congesta R.Br. (for details see under

the latter).

735

The flowers of I. purpurea vary from white, pink or pale purple to deep bluish

purple; the midpetaline areas are reddish in the blue-flowered specimens and the

outside, especially the tube, is frequently paler. In cultivated, but also in wild specimens

in South Africa, variegated flowers are often encountered, which have various combina-

tions of colours in spots, zones or bands.

The size of the corolla varies a great deal. This may be connected with ecological

conditions. At any rate, Ipomoea gerrardiana Rendle, of which I have seen an isotype

(Gerrard 620 in NH) is nothing but a small-flowered (depauperate?) form of I. purpurea.

Similar plants can be found in any locality where I. purpurea is common, especially

late in autumn.

As in /. nil, the ovary and capsule are normally 3-celled with 6 ovules and 6 (or less)

seeds, respectively.

12. I. congesta R. Br., Prodr. FI. Nov. Holl. Ed. 1 (1810), p. 485; Choisy in DC.,

Prodr. 9 (1845), p. 369; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 137, Wood, Natal

PI. 1 (1899), p. 75, t. 93; Ooststr. in Blumea 3 (1940), p. 500 and in Steen., FI.

Males. Ser. I, 4-4 (1953), p. 465. I. purpurea , Baker & Wright in Dyer, FI. Cap.

4, 2 (1904), p. 59, pro parte, non Roth.

Type: In herb. Banks, teste R. Brown (not seen).

Herbaceous climber, probably perennial or at least sometimes perennial. Stems

only occasionally prostrate (and then sometimes rooting at the nodes), terete or somewhat

angular, more or less densely retrorsely pilose, up to several metres long. Leaves

cordate, broadly ovate to orbicular in outline, entire (in most wild S. African specimens)

or some or all 3-lobed nearly to the middle, 4-10 (-17) cm. long and 3-10 (-16) cm.

wide: the entire leaves with acuminate, mucronate apex and broadly rounded basal

lobes, 3-lobed leaves with ovate or oblong, acuminate middle lobe and oblique-ovate

to falcate, subacute to long-acuminate lateral lobes; margin entire, lower surface

often densely, upper surface less densely hairy with soft short adpressed hairs, sometimes

lower surface sericeo-tomentose; petioles retrorsely hairy, 2-7 (-18) cm. long. Peduncles

retrorsely pilose (0- 5-) 4-20 cm. long, bearing one to several flowers in a dense umbellate

cyme with very short branches: bracteoles linear to filiform to occasionally foliaceous.

Sepals unequal, mainly towards the base with rather soft adpressed hairs to nearly

glabrous; outer sepals with a lanceolate or elliptic base, inner ones narrower, all

gradually long linear-acuminate, 14-22 mm. long. Corolla funnel-shaped, 5-8 cm.

long and as much in diam., glabrous, bluish purple, later more reddish purple turning

red; the limb darker than the tube. Ovary glabrous. Capsule in South African

specimens not seen.

Circumtropical, often cultivated and run wild, for its distribution in South Africa

see notes.

This species has often been confused with I. purpurea (L.) Roth, but is quite distinct

on account of the long-acuminate sepals with adpressed pubescence (acute, with

patently hirsute pubescence at the base in I. purpurea).

It is not always possible to distinguish between cultivated and “ wild ” specimens

in the herbaria. In South Africa I. congesta seems to grow wild in the Eastern Cape

and in Natal; in other areas it occurs either cultivated or subspontaneous. Found in

the coastal districts from Bathurst (Port Alfred) to Natal and Zululand, E. Transvaal

(occasionally); cultivated and subspontaneous elsewhere (e.g., Johannesburg, Pretoria).

The cultivated specimens are often referred to as I. learii Paxt. (Paxt., Mag. 6 (1839),

p. 267), but they are not essentially different from luxuriant specimens of I. congesta.

736

13. I. arachnosperma Welw., Apont. Phyto-geogr. (1858), p. 588; A. Meeuse in R. A.

Dyer, Flow. PL Afr. 31 (1956), pi. 1203. Convolvulus bicolor Desr. in Lamk.,

Encycl. Method 3 (1789), p. 564, non Vahl. (1794). C. pilosus Roxb., Hort. Beng.

(1814), p. 14, nomen tantum, and FI. Ind., Ed. Carey et Wall. 2 (1824), p. 55,

descr., non Rottler (1803). C. dichrous Roem. et Schult., Syst. Yeg. 4 (1819),

p. 263, nomen illeg. Ipomoea pilosa (Roxb.) Sweet, Hort. Brit. Ed. I (1827),

p. 289; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 161; Hutch. &

Dalz., FI. W. Trop. Afr. 2 (1931), p. 218, non I. pilosa Houtt. (1777), nec Cav.

(1797). I. calophylla Fenzl. in Flora 27 (1844), p. 312, nomen nudum. I. dichroa

(Roem. et Schult.) Hochst. ex Choisy in DC., Prodr. 9 (1845), p. 364; Hall. f.

in Engl. Bot. Jb. 18 (1893), p. 136, nomen illeg.

Type: Welwitsch 6243 and 6244 from Loanda, Angola are apparently the syntypes

(in BM); no. 6243 also in COL

Annual. Stems several from the base, twining or occasionally prostrate, rather

stout and firm, up to several metres long, covered with bulbous-based patent bristly

hairs. Leaves broadly cordate-ovate in outline, up to 10 cm. long and wide, digitately

3-lobed to about the middle, green and thinly hairy above, densely silvery-white,

cobwebly-tomentose beneath except on the green pilose veins; central lobe usually

somewhat narrowed to the base, ovate to suborbicular, cuspidate with terminal mucro ;

lateral lobes oblique, rounded at the base, with a triangular subacute to acute apical

portion; basal sinus acuminately reaching the insertion of the petiole; margin entire,

ciliate; petioles rather stout, terete, hairy like the stems as are peduncles, pedicels,

bracteoles and sepals, up to 5 cm., sometimes 7 cm. long. Inflorescences starting as

a perfect 3-flowered dichasium, ultimately 7-11 -flowered, the branches usually becoming

monochasial; peduncles terete, 3-6 cm. long; secondary and tertiary cyme-branches

gradually thinner, 2- 5-0- 5 cm. long; bracteoles in opposite pairs, ovate-lanceolate

to lanceolate with a broad base, long-acuminate to aristate, 9-12 mm. long or the

ultimate ones shorter; pedicels subclavate, somewhat flattened, up to 1 cm. long,

Sepals subequal, firm, green, smooth and glabrous inside, lanceolate, long-acuminate

to aristate with spreading or reflexed tips, in flower 10-13 mm. long and 2-3 mm.

wide; in fruit up to 16 by 5 mm. Corolla funnel-shaped; pale mauve with darker

mauve centre, 14-20 mm. long and as wide across, glabrous except on the midpetaline

areas near the tips of the corolla-lobes. Capsule globose, glabrous, apiculate by the

style-base, about 8 mm. in diam. Seeds fawn-coloured, very shortly velvety, about

4 mm. long.

Tropical Africa, India. Extends into South West Africa, Bechuanaland and the

Transvaal.

Recorded from the following areas: South West Africa (Northern part, as far south

as Otavi and Grootfontein); Transvaal (Waterberg, Petersburg, Lydenburg, Zout-

pansberg).

The oldest specific epithet bicolor cannot be used on account of Ipomoea bicolor

Lamk. (= I. nil ) and I. bicolor (Vahl) Sweet (= Hewittia sublobata). The names

Convolvulus pilosus Roxb. and C. dichrous R. et S. are later synonyms of C. bicolor

Desr., the first is, in addition, antedated by Convolvulus pilosus Rottl.; the second

name is based on the same type as C. bicolor Desr. (Roemer and Schultes evidently

renamed C. bicolor Desr. on account of C. bicolor Vahl, which latter name they

retained).

The identity of Convolvulus pilosus Rottl. (in Ges. Naturf. Fr. Neue Schr. 4 (1803),

p. 196) is not certain (it may be the same as Ipomoea purpurea) but even if it is the same

as Roxburgh’s C. pilosus it cannot be used for the species under discussion because

the epithet “ pilosa ” is pre-occupied in Ipomoea: I. pilosa Houtt. (1777) = Merremia

737

umbellata (L.) Hall. f. and I. pilosa Cav. (1797) = Ipomoea pentaphylla (L.) Jacq. =

Merremia aegyptia (L.) Urb.

Combinations based on the epithets “bicolor”, “ dichrous ” and “pilosus”

are in Ipomoea, strictly speaking, illegitimate and could only be retained as “ new

names ” in Ipomoea if no other valid epithet is available. Fenzl published the name

I. calophylla for this species in 1844 but this name is a nomen nudum.

The name I. arachnosperma Welw. (1858) is available for this species. It was based

on specimens collected by Welwitsch in Angola. A specimen of Welwitsch 6243 in

COI is undoubtedly the same as the plants found in southern Africa and there can be

very little doubt that the synonymy indicated above is correct because Hallier and Baker

and Rendle associated the Welwitsch number(s) 6243 (and 6244), and specimens collected

by Sehimper and Kotschy, with the name “ Ipomoea pilosa Sweet ”, so that they were

obviously convinced of the identity of the Indian plants with the African specimens.

14. I. wightii {Wall.) Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 470, and in DC.,

Prodr. 9 (1845), p. 364; Wight, Icon. (1848), t. 1364; C. B. Clarke in Hook, f.,

FI. Br. Ind. 4 (1883), p. 203; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 133, and 28

(1899), p. 32; Baker and Wright in Dyer, FI. Cap. 4, 2 (1905), p. 157; Brenan

in Mem. New York Bot. Garden 9 (1954), p. 7. Convolvulus wightii Wall., PI.

Asiat. Rar. (1831), p. 55, t. 171. Ipomoea arachnoidea Boj., Hort. Maurit. (1837),

p. 228, nomen tantum, and ex Choisy in DC., Prodr. 9 (1845), p. 364, pro. parte;

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 133.

Type: From India (the plate in Wallich’s PI. Asiat. Rar. was taken to be represen-

tative).

Perennial herbaceous climber, only prostrate when no support is available. Stems

few from the base, often firm and becoming stout, terete, retrorsely to patently pilose

with yellowish or brownish hairs, up to several metres long. Leaves (ovate-) cordate

in outline, entire to 3-lobed to about the middle, thinly herbaceous, green and more

or less densly strigose above, densely (less densely in very old ones) covered with a

floccose-cobwebby tomentum beneath except on the veins which are covered with

adpressed yellowish hairs and thus clearly marked out, 3-12 (-20) cm. long, and 2-5-10

(-18) cm. wide; the apex acuminate, acute or obtuse, shortly cuspidate; basal sinus

usually rather deep and rather narrow, basal lobes broadly rounded; margin entire to

distinctly repand; apical lobe of 3-lobed leaves broadly ovate, usually somewhat

constricted at the base; lateral lobes rounded, semi-orbicular to obliquely ovate-falcate;

petioles rather stout, patently or retrorsely pilose, 2-10 (-15) cm. long. Inflorescence

a dense, headlike pedunculate cyme; peduncles stoutish, hairy like stems and petioles

but hairs usually pointing upwards, rarely retrorse, 3-13 cm. long; bracteoles lanceolate,

acuminate, hirsute and with shorter glandular hairs on the sides and margins, 10-12

mm. long; cyme-branches very short, pedicels wanting or nearly so. Sepals equal,

linear-lanceolate, acuminate-aristate, hirsute and with short glandular hairs on the

sides and margins, 8-15 mm. long. Corolla funnel-shaped, described as rose-red,

purple, mauve or magenta, 20-40 mm. long and the limb as much in diam., glabrous

or sparsely hairy on the well-defined mid-petaline areas. Capsule subglobose or broadly

ovoid, 8-10 mm. long and 7-9 mm. in diam.; valves almost papery, thin, usually

thinly and shortly hairy mainly towards the apex. Seeds dark brown, glabrous, about

3 mm. long.

India, Ceylon, Madagascar, East Africa from Uganda and Zanzibar to Natal,

Swaziland and Transvaal.

738

Recorded from the following districts: Transvaal (Barberton, Nelspruit, Letaba>

Pietersburg, Potgietersrust, Sibasa, Zoutpansberg) ; Swaziland; Natal (Eshowe) and

Zululand (Ngoye, Umhlatuzi Valley).

I. wight ii is closely related to I. ficifolia Lindl. and I. arachnosperma Welw., but

was placed in a different sub-section of § Pharbitis by Hallier, viz., in the group Cepha-

lanthae. In my opinion it should be transferred to the subsection Chorisanthae and

inserted next to its closest allies.

Depauperate specimens of 1. wightii sometimes closely resemble I. arachnosperma.

Well-developed specimens are quite distinct by their dense head of flowers. The

glandular bracts and sepals distinguish it in doubtful cases. The flowers of I. ficifolia

are almost invariably larger than those of I. wightii, and usually fewer in number

(1-5 together); with very narrow eglandular sepals.

15. I. ficifolia Lindl. in Bot. Reg. 26 (1840), Misc. 90, and 27 (1841), t. 13; Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 135 and 28 (1899), p. 35; Baker and Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 64. Convolvulus trilobus Thunb., Prodr. FI. Cap. (1794),

p. 35. I. holosericea E. Mey. ex Drege in Zw. Pfl. geog. Doc. (1843), p. 132, 195,

nomen tantum; ex Choisy in DC., Prodr. 9 (1845), p. 364. I. angulata E. Mey.

ex Drege, op. cit., p. 135, 195 nomen tantum. I. vitifolia E. Mey. ex Drege, op.

cit., p. 158, 195, nomen tantum. I. arachnoidea Choisy, op. cit., p. 364, pro. parte,

quoad spec. Drege, non Bojer. I. aitoni Choisy, op. cit., p. 363, non Lindl.

Type: The plate in Bot. Reg. 27 (1841) t. 13 was taken to be representative of

the species.

Perennial. Root tuberous. Stems climbing, occasionally prostrate, usually slender,

herbaceous, tending to become firm or woody, up to at least 2 m. long, terete or faintly

angular, often longitudinally striate, sparsely, rarely densely, pilose, hirsute or hispidulous

turning scabridulous in older parts. Leaves thinly herbaceous, ovate-cordate to sub-

orbicular-reniform in outline, frequently (partly) 3-lobed, more rarely all leaves

undivided, 2-9 cm. long and lJr-8 cm. wide; the apex acute or acuminate, more rarely

obtuse, mucronate; the base cordate; basal sinus usually broad and shallow; basal

lobes rounded; lobed leaves divided to the middle or less deeply so, with an ovate to

obovate, rhomboid-ovate or broadly ovate-triangular terminal lobe which is usually

(at least in leaves divided to about the middle) distinctly constricted at the base, and

obliquely ovate to somewhat falcate or rounded, semi-orbicular basal lobes which

vary from shortly acuminate to broadly rounded and are shortly mucronulate: the

upper surface green, more or less thinly covered with adpressed hairs; lower surface

in most cases very distinctly floccosely or cobwebby-tomentose with soft white hairs

when young except on the veins which are thus clearly marked out, usually less distinctly

floccose to glabrescent when older but almost invariably showing vestiges of the

tomentum as scattered floccose tufts of hairs often adherring to the slightly raised

veins; petioles slender, usually striate, hairy like the stems, 1-5 (-7) cm. long.

Inflorescences cymosely 1-5 (-many-) flowered; peduncles rather slender, pilose like

stems and petioles, 2-15 cm. long; bracteoles linear-lanceolate, 7-11 mm. long,

acuminate-aristate, pilose or hirsute as are the sepals; cyme-branches usually short,

but occasionally up to 7 cm. long, gradually thinner upwards; pedicels short, 0-10

(-15) mm. long. Sepals equal, lanceolate, gradually acuminate into a narrow point,

10-15 mm. long. Corolla funnel-shaped, 4-6 cm. long and the limb as much in diam.,

pale magenta, pink or purple, glabrous or with a few pilose hairs on the well-defined

mid-petaline areas; the limb shallowly 5-lobed. Capsule subglobose, 7-9 mm. in

diam., glabrous in all specimens seen (according to FI. Cap. sometimes pilose). Seeds

bearing small tufts of short hairs on the back and often with very long cottony hairs

attached to the edges near the top (according to FI. Cap. sometimes glabrous), 3-4

mm. long.

739

Cape Province, Natal, Zululand and Portuguese East Africa (see also Notes).

Recorded from the following districts: Cape: Uitenhage, Alexandria, Albany,

Queenstown, Bathurst, King William’s Town, East London, Komgha, Kentani, Willow-

vale, Port St. Johns; Natal: Port Shepstone, Umzinto, Durban, Pietermaritzburg,

Inanda, Eshowe, Weenen, Lower Umfolosi; Portuguese E. Africa: Inhaka (or Inyack)

Island nr. Lourengo Marques.

There is very little doubt that Convolvulus trilobus Thunb. is the oldest name for

this species, but the specific epithet triloba cannot be used in Ipomoea on account of

/. triloba L., Sp. PI. (1753), p. 161.

In FI. Trop. Afr. 4, 2 (1905), p. 161, some plants are referred to a variety of

I. ficifolia (var. laxiflora Hall. f.). This would extend the range of the species to

Somaliland; however, the typical form seems to be restricted to South Africa and the

southern part of Portuguese East Africa.

Thunberg’s type is reported to have been collected in the Humansdorp district.

No other specimen has since been collected farther west than Uitenhage and either

the species has become extinct west of Uitenhage or Thunberg’s locality was not

correctly recorded.

I. ficifolia can easily be distinguished from the related species I. wightii and

I. arachnosperma by its larger flowers, and from the latter also by its few-flowered

inflorescences. I. ficifolia has also occasionally been confused in the herbaria with

I. congesta, but the tomentose lower surface of the young leaves and the narrow, shorter,

gradually acuminate and densely hirsute sepals distinguish it at once from the latter.

In FI. Cap. the capsules of the species are said to be either glabrous with seeds

having long cottony hairs, or pilose with glabrous seeds. As far as can be ascertained

the capsules are always glabrous, so that the statement in FI. Cap. may be due to

confusion with the capsules and seeds of I. wightii which has pubescent capsules and

glabrous seeds.

Lindley mentioned South America (Buenos Aires) as the possible country o

origin of this species, but his plate undoubtedly represents the South African plant.

16. I. chloroneura Hall. f. in Engl. Bot Jb. 18 (1893), p. 132; Hiern. Cat. Welw. Afr.

PI. 1, 3 (1898), p. 734; Baker and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 52:

Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 153; Dinter in Fedde,

Repert. 18 (1922), p. 430.

Type: Hallier mentions only Welwitsch 6181 from Angola in herb. Berlin and

DC. (Geneva). The Berlin specimen having been destroyed, the specimen in the

Geneva herbarium must be taken to represent the type.

Annual. Main stem erect, branched from the base, with branches erect to decum-

bent, often branched again, densely covered with adpressed short white hairs and

spreading golden-yellow hairs when young (as are petioles and peduncles), less

densely so when older, up to 25 cm., rarely up to 60 cm. long. Leaves herbaceous,

elliptic, oblong or lanceolate to oblanceolate or elliptic-obovate, entire, obtuse to

acuminate, narrowed into a cuneate base, 2-10 cm. long and |-3-5 cm. wide; upper

surface with adpressed white hairs, lower surface much more densely so except on

the veins which are covered with golden-yellow hairs and thus clearly marked out;

the margin lined with yellowish hairs; petioles 5-20 (-30) mm. long. Inflorescence

a few-flowered pedunculate head; peduncle terete, erect or spreading 2-5 cm., in

fruit up to 7 cm. long; outer bracts foliaceous, often rather large, resembling the young

leaves, often 10-20 mm. long, enlarging up to 30 mm. but occasionally up to 40 mm.

740

in fruit, including the petiole ; inner bracts smaller. Sepals sub-equal, elliptic-lanceolate,

lanceolate or elliptic, about 6 mm. long; lower half with narrow glabrous edges, central

portion densely hairy, thicker and prolonged into a densely hairy tail, accrescent in

fruit. Corolla funnel-shaped, pale yellow to cream or almost white, a little longer

to twice as long as the calyx; midpetaline areas with white hairs which project as tufts

beyond the tips of the corolla-lobes. Capsule globose or ellipsoid, glabrous, 6-8 mm.

long and 5-7 mm. in diam. ; valves finely and longitudinally striate. Seeds with an

adpressed dense silky-villous, pale fawn or drab pubescence, 3-4 mm. long.

South West Africa, Angola, Bechuanaland, extending into Transvaal and Barotse-

land (N. Rhodesia), also in East Africa (teste Verdcourt).

S. W. Africa. — Between Cunene Riv. and Eunda: Barnard H. no. 32319 (SAM);

Auros (Otavi): Dinter 5778 (BOL, NH, PRE, SAM); Tsumeb: Dinter 7576 (BOL,

PRE); Gaub: Dinter 2436 (SAM); Okahandja: Bradfield 384 (PRE); Upper

Swakop nr. Okaharmi: Dinter 3276 (PRE, SAM).

Bechuanaland. — Mochudi: Harbor in herb. Rogers 6557 (BOL) = ? 492 (KMG);

N’gamiland: Lugard and Lugard 190 (GRA).

Transvaal. — Waterberg, Naboomspruit: Galpin M. 235 (PRE).

Angola. — Welwitsch 6181 (iso-type!), 6132 (COI); Baum 755 (COI).

N. Rhodesia. — Barotseland, Sesheke: Borle s.n. (PRE).

I was able to examine an isotype ( Welwitsch 6181 in COI) and a few other specimens

referred to this species by Hallier and others ( Welwitsch 6132, Baum 755, Lugard and

Lugard 190).

17. Ipomoea ommaneyi Rendle in Jl. Bot. 40 (1902), p. 190 (“ ommanei”); Baker

and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 53.

Type: Ommaney s.n. from Johannesburg (BM, photograph of type in PRE).

Perennial from a thick fusiform tuberous taproot which attains a length of at least

1 m. and a thickness of 25 cm. Stems several, annual, trailing, herbaceous, up to 2 m.

long or over and in the older parts up to about 1 cm. in diam., terete, densely pubescent

at least in the younger parts. Leaves secund, ovate-lanceolate or oblong-lanceolate,

up to 30 cm. long and 12 cm. wide, rounded or subcordate at the base, subobtuse,

penninerved, crisped, densely ciliate and yellow-edged on the margin, densely clothed

on both surfaces with silky silvery- white hairs, later somewhat glabrescent; midrib

stout, very prominent beneath; lateral nerves 9-12, prominent beneath; petioles

stout, 6-8 mm. in diam., terete, somewhat flattened and grooved above, densely hairy,

5-30 (-50) mm. long. Peduncles much shorter than the leaves, up to about 10 cm.

long, densely hairy. Flowers several together in a dense head; outer bracts ovate or

ovate-subspathulate, acuminate-cuspidate, 25-30 mm. long and about 12 mm. wide,

densely silky as are the inner bracts and sepals; inner bracts shorter and narrower.

Sepals longer than the bracts, 30-35 mm. long, the outer lanceolate, acuminate-

cuspidate, the inner much narrower, linear-lanceolate. Corolla funnel-shaped, rose-

magenta with distinct midpetaline areas which are white and densely silky outside;

(30-) 40-50 mm. long; anthers sagittate. Ovary glabrous. Capsule globose, completely

enclosed in the calyx, about 1 cm. in diam., with a thin, papery wall and 4 dull black,

blabrous seeds.

Transvaal and some districts of Northern Cape, extends into Southern Rhodesia

and Bechuanaland.

741

Recorded from Griqualand West (Barkly West, Vryburg); Transvaal (Marico,

Wolmaransstad, Ventersdorp, Rustenburg, Krugersdorp, Pretoria, Johannesburg and

Rand districts to Bethal, Heidelberg, Witbank, Middelburg, Ermelo, Carolina, Belfast,

Pilgrims Rest, Pietersburg, Potgietersrust, Waterberg); Beclmana/ancl (Lobatsi);

Southern Rhodesia (only recorded from IJmtali).

Mr. de Winter compared a few specimens with the type and, in addition, I was

able to study a photograph of the type ( Ommaney s.n. in BM) and some specimens

quoted in FI. Cap., viz. Gilfillan in herb. Galpin 6158 (PRE, GRA) and Wood 7189

(NH).

I. ommaneyi is a very distinct species and resembles only 1. atherstonei (for the

difference between these two species, see under the latter). Turbina oblongata sometimes

has several flowers in a head, but can be distinguished by the shape and pubescence

of the leaves and has usually smaller bracts and sepals (they are usually 25-35 mm.

long in I. ommaneyi and mostly under 25 mm. long in T. oblongata).

N.B. — A specimen Hutton 432, according to the label from Natal, was probably

wrongly labelled and possibly collected near Johannesburg (Florida).

18. I. atherstonei Baker in Dyer, FI. Cap. 4, 2 (1904), p. 53; Baker & Rendle in Dyer,

FI. Trop. Afr. 4, 2 (1905), p. 154, ex descr.

Type: On account of its specific name, the specimen Atherstone s.n. from

Middelburg (“ Nazareth ”) must be taken to be the type. However, Mr. de Winter

reported that the original specimens could not be traced in the Kew Herbarium, nor

could they be found in the collections of the British Museum.

Prostrate perennial with fusiform tuberous root. Stems annual, firm, stout, often

angular, up to 2 m. long, usually densely, sometimes thinly covered with short stiff

hairs as are petioles, peduncles, bracts and pedicels. Leaves oblong to sometimes

ovate-elliptic, entire, herbaceous, obtuse, emarginate or rounded and mucronate,

occasionally acute, at the apex, broadly cuneate, truncate, rounded to shallowly cordate

at the base, thinly covered on both surfaces with adpressed or somewhat spreading

short stiff hairs, sometimes very densely so, 6-14 cm. long and 4-7-5 cm. wide; petioles

1-3 cm., occasionally up to 5 cm. long. Inflorescence a pedunculate, dense, few-flowered

capitate cyme, occasionally reduced to a single flower; peduncles terete, rather slender

to rather stout, 3-8 (-16) cm. long; bracteoles lanceolate or elliptic-lanceolate, or

narrowly ovate-lanceolate, usually narrowed at the base, subacute to acuminate-aristate

to the apex, 20-35 mm. long hairy; pedicels very short or 0. Sepals unequal; outer

ones ovate-lanceolate, to oblong, inner ones narrower, all usually acute to long-

acuminate; (20-) 25-35 mm. long, hairy. Corolla funnel-shaped, magenta (often

described as “purple”), 4-5-7 cm. long, and as much in diam., midpetaline areas

well-defined, silky-pilose, usually densely so. Capsule and seeds unknown.

Eastern and South Eastern Transvaal, North West Natal, Southern Rhodesia,

also in Bechuanaland and S.W. Africa.

Transvaal. — Belfast near Machadodorp: Codd, 8255, 8260 (PRE). Ermelo:

Leendertz H. no. 7783 (PRE), Walker 99 (PRE), Pupils of Convent 152 (PRE), id.,

Spitskop: Pott 5001 (BOL). Lydenburg, foothills of Steenkampsberg: Codd 8203

(PRE). Piet Retief: Jenkins H. no. 11989 (PRE); Kretschmar s.n. (PRE); Compton

22338 (NBG); Iswepe: Sidey 1577 (PRE).

Natal. — Utrecht: Thode A 335 (PRE).

742

South West Africa.— “ latitude 23° Chapman & Baines (K).

S. Rhodesia. — Salisbury: Eyles 904 (SAM, SRGH).

This species is closely related to I. onvnaneyi, but differs in several points. The

bracts and sepals are less acuminate, the leaves have a different shape (more oblong

and less tapering towards the apex) and have more slender petioles, and the corolla

is longer than in I. ommaneyi. It also resembles some extreme forms of Turbina

oblongata , but differs from the latter in that the flowers are usually much more numerous

in the heads, the bracts usually wider and longer (often linear in T. oblongata ) and the

pubescence not so bristly. The leaves are usually larger than in T. oblongata.

No actual type being available, the identification of the specimens cited above

with I. atherstonei is done ex description. The specimens cited above are the only

ones which fit the description and resemble I. ommaneyi. Moreover, the type was

reported to come from near Middelburg and the specimens I refer to this species are

mainly from South East Transvaal, not so far from the type locality. On the other

hand, I have not seen a specimen from Middelburg, but at the time when Atherstone

collected his specimens the districts were much larger than they are now, and the towns

and villages fewer in number so that the localities of old collectors are often less accurate

than they are at present and the type of I. atherstonei may well have been collected

in a neighbouring area.

19. I. magnusiana Schinz in Verh. Bot. Ver. Brandenb. 30 (1888), p. 272; Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 135; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904),

p. 65; Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 162; N. E. Brown in Kew

Bull. 1909, p. 123; Dinter in Fedde, Repert. 18 (1922), p. 431. I. lugardi N. E.

Brown var. parviflora Rendle, op. cit., p. 163. I. otjikangensis Pilger et Dinter

in Engl. Bot. Jb. 41 (1908), p. 296; Dinter in Fedde, Repert. 18 (1922), p. 431.

Two varieties can be distinguished (for a discussion see below): —

19a. Ipomcea magnusiana Schinz. var. magnusiana.

(For synonymy see above).

Type: Schinz 752 from South West Africa in herb. Zurich.

Perennial, forming several to many annual stems from a thin woody taproot.

Stems twining or prostrate or young ones suberect, slender, pilose, up to about 2 m.

long; Leaves palmately 3- or 5-lobed nearly to the base, green or yellowish-green

and with rather thin adpressed pilose hairs above, densely covered with a white

cobwebby tomentum beneath except on the main nerves and main veins which are

covered with yellowish or brownish stiff, adpressed to patent hairs and thus clearly

marked out, 2-6 (-9) cm. in diam., lobes varying from obovate to narrowly elliptic,

lanceolate, oblanceolate, linear-lanceolate or ovate-lanceolate except the basal ones in

5-lobed leaves which are shorter and relatively broader; lateral lobes in 3-lobed leaves

with a lateral or basal rounded lobe or auricle; apex of central and first pair of lobes

subacute to acuminate, often cuspidate, of basal lobes sometimes obtuse or rounded,

base of lobes narrowed and confluent; basal sinus of leaf rounded or obtuse, sinuses

between the lobes acute or obtuse or sometimes rounded; margins entire or subentire,

more or less distinctly ciliate ; petioles a little longer to a little shorter than the leaves,

usually slender, pilose. Inflorescence a dense, few-flowered pedunculate hea4, rarely

by reduction flowers solitary; peduncles usually slender, pilose like stems and petioles,

shorter or longer than the subtending leaf; bracts short or up to 15 mm. long, linear

743

or lanceolate, hairy; pedicels very short. Sepals somewhat unequal, lanceolate or

broadly lanceolate, acute, 6-15 mm. long, hairy outside, somewhat accrescent in fruit.

Corolla funnel-shaped, magenta-purple, mauve or pale mauve or cream with darker

magenta centre, occasionally almost completely white, 12-20 mm. long and the spreading

limb as much in diam. ; midpetaline areas well-defined, pilose. Ovary glabrous. Capsule

subglobose, 6-8 mm. in diam., glabrous. Seeds pubescent, 4-5 mm. long, sometimes

also with long white hairs on the angles in upper half.

Recorded from South West Africa as far South as Okahandja; Bechuanalana

(Mochudi); Cape: Mafeking, Hay; Transvaal: Wolmaransstad, Potchefstroom,

Brits, Pretoria, Groblersdal, Waterberg, Rustenburg, Potgietersrust, Zoutpansberg,

Letaba, Lydenburg, Barberton, Swaziland; Orange Free State: Vredefort; Portuguese

E. Africa: between Komatipoort and Lourengo Marques; S. Rhodesia: Livingstone.

Ipomoea otjikangensis, of which I saw isotypes ( Dinter 517 in GRA, PRE and

SAM), is indistinguishable from I. magnusiana var. magnusiana.

19b. I. magnusiana Schinz var. eenii (Rend/e) A. Meeuse in R. A. Dyer, Flow. PI. Afr. 31

(1956), pi. 1201. I. eenii Rendle in J. Bot. (London) 39 (1901), p. 21, and in FI.

Trop. Afr. 4, 2 (1905), p. 163, exclus. var. parviflora. I. lugardi N. E. Brown

apud Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 163 and in Kew Bull.

1909, p. 124.

Type of variety: Een s.n. in herb. Brit. Mus. (from Damaraland), photograph

of the type in PRE.

Steins usually climbing, up to several metres long, generally stouter than in the

var. magnusiana. Leaves usually dark green above, 4-11 cm. long and 4-14 cm wide;

petioles stouter than in var. magnusiana as a rule, 3-1 1 cm. long. Peduncles usually

rather stout, 3-12 cm. long; outer bracts lanceolate, 13-15 mm. long; inner bracts

usually shorter. Sepals ovate-lanceolate or lanceolate, 12-13 mm. in fruit to 19 mm.

long. Corolla magenta (always?), paler outside, 20-25 mm. long and 30-40 mm. in

diam. Capsule 7-8 mm. in diam. Seeds pubescent (hairs often arranged in tufts in

two subparallel longitudinal lines) and usually with long white hairs on the angles

in upper portion. Otherwise as var. magnusiana.

Almost the same distribution as that of the var. magnusiana, but not yet recorded

from Griqualand West or Orange Free State; one record from Lusikisiki (E. Cape); in

the Transvaal so far recorded from Rustenburg, Pretoria, Groblersdal, Waterberg,

Lydenburg, Soutpansberg, Letaba, Nelspruit, Barberton. In addition recorded from

Southern Rhodesia, Birchenough Bridge.

The distinction between I. magnusiana and I. eenii (= I. lugardii) is very difficult.

The only character that can be used is the size of the corolla (up to 15, rarely up to 20

mm. long in I. magnusiana and 20-25 mm. or longer in dried specimens, in I. eenii).

The shape of the lobes of the leaf varies and also the number of lobes (3-lobed and 5-lobed

leaves often occur on one specimen). /. lugardii is identical with I. eenii because the

number of lobes of the leaves is not a satisfactory distinguishing character. Mr. de

Winter compared the original specimens of /. lugardii and the type of I. eenii, as well

as an isotype of Schinz 752 and reported that the first two are identical and the last

is very close. I was able to study isotypes of Schinz 752 (in GRA, BOL). The best

solution I can see is to regard I. eenii (= I. lugardii) as a variety of I. magnusiana,

although I am inclined to think that they are hardly worth varietal rank. Fruiting

specimens cannot be distinguished with certainty, and the vegetative characters are

not sufficiently distinct to separate these two varieties in every case if flowers are lacking.

744

20. I. pes-tigridis L., Sp. PL Ed. 1 (1753), p. 162; Choisy in DC., Prodr. 9 (1845),

p. 363; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 134 and 28 (1899), p. 34; and in Bull.

Herb. Boiss. 6 (1898), p. 539; Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), p. 158; Hutch, and Dalz., FI. W. Trop. Afr. 2 (1931), p. 218; Ooststr.

in Blumea 3 (1940), p. 504 and in Steen., FI. Males. Ser. I, 4-4 (1953), p. 467;

Brenan in Mem. New York Bot. Garden 9 (1954), p. 7.

Type: Linnaeus based I. pes-tigridis on the figures in Herm. Lugdb. t. 187, Dill.

Hort. Elth. t. 318, f. 11 and Rheede, Hort. Mai. 2, t. 59. The species is also represented

in the Linnaean Herbarium and the specimen may be proposed as the lecto-type.

Annual. Stems slender, twining or occasionally prostrate, hirsute with long

spreading hairs as are petioles and peduncles, 0-5-2 m. long. Leaves orbicular to

somewhat reniform in outline, palmately-pedately (5-) 7-9 (-11) lobed, 3-9 cm. long

and 4-12 cm. wide; broadly cordate at the base; segments lanceolate-elliptic to elliptic,

ovate-lanceolate or oblong, attenuate to acuminate and acute or subacute, minutely

mucronate at the apex, narrowed and confluent at the base, lateral ones vary gradually

smaller, lowermost somewhat oblique to falcate, often obtuse, all rather thinly pilose

on both surfaces with adpressed to patent hairs; sinuses between the lobes rounded;

petioles rather slender, 1-5-10 cm. long. Inflorescence a pedunculate, involucrate,

few-flowered head; peduncle 2-12 (-18) cm. long; outer bracts foliaceous, linear-

oblong to oblong, 1 • 5-3 cm. long, often broad, subcordate to subauriculate at the

base, densely hirsute-pilose, inner bracts smaller. Sepals lanceolate or the inner ones

somewhat narrower, pilose-hirsute like the bracts, 7-12 mm. long. Corolla funnel-

shaped, mauve or pale purple, 3-4 (-5-5) cm. and the limb as much in diam., sparsely

pilose with stiff, adpressed hairs on the midpetaline areas. Ovary glabrous. Capsule

ovoid, glabrous, about 8 mm. long. Seeds brown with a white, sparse, short villous-

tomentose pubescence, about 4 mm. long.

Tropical Africa, as far south as Angola and S. Rhodesia and Portuguese East

Africa, one record from Transvaal; in addition Mascarenes, tropical Asia and Malaysia.

Transvaal, Nelspruit, between lower Sabie and Skukuza: v.d. Schijff 1775 (PRE);

also collected recently by B. de Winter in the Okavango, South West Africa.

Hallier distinguished several varieties and subvarieties; some of them have been

upheld in FI. Trop. Afr. Anyhow, the African specimens seem to be slightly different

from the Asiatic ones in that the flowers are said to be “ pink ”, mauve or “ purple ”

(probably they are always mauve) instead of white and fewer in number in the inflores-

cence, and the leaves have usually more segments (7-11 against 3-7, rarely up to 9

in the Asiatic form). There seem to be sufficient grounds to distinguish the African

specimens as I. pes-tigridis L. var. africana Hall. f.

21. I. involucrata P. Beauv., FI. Owar. 2 (1817), p. 52, t. 89; Choisy in DC., Prodr.

9 (1845), p. 365; Hall. f. in Engl. Bot Jb. 18 (1893), p. 135, ex parte; Baker and

Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 150; Hutch, and Dalz., FI. W. Trop.

Afr. 2 (1931), p. 218.

Type: Not seen (Choisy, l.c., mentions a specimen in herb. Beauv. nunc Deless.,

which is evidently the type specimen). The plate in FI. Owar. is quite adequate to

recognise this species.

Annual? Stems herbaceous, slender, twining, finely and usually retrorsely hairy

to glabrescent, Leaves herbaceous, cordate-ovate, entire, attenuate-acuminate, subacute,

obtuse and mucronate at the apex, thinly hairy on both surfaces with adpressed hairs,

sometimes more densely so beneath, with an obtuse or rounded basal sinus and broadly

rounded basal lobes, 2-8 cm. long and 1-5-7 cm. wide; petioles slender, retrorsely

745

pilose like the stem and peduncles, 1-5-10 cm. long. Inflorescence a pedunculate

involucrate head; peduncle usually slender, terete 2-12 cm. long; outer bracts connate

into one large, hairy boat-shaped structure 3-6 cm. in diam. with 2 cusps; inner bracts

smaller, bluntly obovate or oblanceolate to linear-oblong. Sepals : outer ones lanceolate,

acuminate, about 12 mm. long, inner ones shorter and ovate, glabrescent or sparsely

hairy on the back and setose along the margin. Corolla funnel-shaped, its colour

described as purple, mauve or bright rose-red, 3-5 cm. long and as much in diam.,

midpetaline areas well-defined, minutely pilose. Capsile small, globose, glabrous.

Seeds shortly pubescent or glabrous.

Throughout tropical Africa, extending into Angola, Southern Rhodesia, Portuguese

East Africa and the Northern Transvaal.

Transvaal. — Soutpansberg: Entabeni Forest Station near Louis Trichardt:

Galpin 9463 (PRE).

Hallier united I. involucrata and I. pileata Roxb. but these species are quite distinct.

They can easily be distinguished by the characters mentioned in the key, even if corollas

are wanting, but vegetative specimens are indistinguishable.

Ipomoea operosa C. H. Wright (in Kew Bull. 1897, p. 275) = I. involucrata var.

operosa (C. H. Wright) Hall. f. in Engl. Bot. Jb. 30 (1901), p. 387, does not seem to be

more than a hairy form of I. involucrata. I have not seen any authentic specimens

of I. operosa, but at any rate the Galpin specimen is typical I. involucrata.

22. I. pileata Roxb., FI. Ind. ed. Carey et Wall. 2 (1824), p. 94, and id., ed. Carey,

1 (1832), p. 504; Choisy in DC., Prodr. 9 (1845), p. 365; Baker and Wright in

Dyer, FI. Cap. 4, 2 (1904), p. 53; Baker and Rendle in Dyer, FI. Trop. Afr. 4,

2 (1905), p. 151; Ooststr. in Blumea 3 (1940) p. 507 and in Steenis, FI. Males.,

Ser. I, 4-4 (1953), p. 467. I. involucrata Hall. f. in Engl. Bot. Jb. 18 (1893), p. 135,

ex parte, exclus. type.

Type: Not seen.

Usually described as an annual, but a specimen in PRE (Levy 33 from Wankie,

S. Rhodesia) shows a fusiform thick taproot and seems to indicate a perennial habit.

Vegetative characters as I. involucrata (see no. 21). Inflorescence a pedunculate,

involucrate head, peduncle retrorsely pilose, 2-5 cm. long; outer bracts connate into

a large, boat-shaped structure 2-5-4 cm. long with two cusps; inner bracts much

smaller, oblong or elliptic, obtuse. Sepals herbaceous, 3 outer ones oblong-spathulate

to oblong, obtuse, about 10 mm. long, inner ones much narrower, lanceolate with a

long and slender point, 9 mm. long, all hairy outside and inside. Corolla pink with

darker centre or purple, hypocrateriform; tube about 2 cm. long, rather slender,

glabrous except near the top; limb spreading, 1-5-3 cm. in diam., midpetaline areas

sparsely pilose, mucronate. Ovary glabrous. Capsule globose, small. Seeds glabrous

or thinly pubescent.

East Tropical Africa, extending into Southern Rhodesia, Portuguse East Africa

and Transvaal. Also in the Mascarene Islands, India to China and Malaysia.

Transvaal. — Barberton: Galpin 882 (PRE, BOL); Thorncroft 800 (NH).

See also under I. involucrata. The African specimens agree in every respect with

the Asiatic ones.

746

23. I. batatas (L.) Lam., Tabl. Encycl. I (1791), p. 465; Hall. f. in Engl. Bot. Jb. 18

(1893), p. 138; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 175; Hutch,

and Dalz., FI. W. Trop. Afr. 2 (1931), p. 216; Ooststr. in Blumea 3 (1940), p. 512,

and in Steenis, FI. Males., Ser. I, 4-4 (1953), p. 469. Convolvulus batatas L., Sp.

PI. Ed. 1 (1753), p. 154. Batatas edulis (Thunb.) Choisy in DC., Prodr. 9 (1845),

p. 338.

Type: Linnaeus originally based his species on pre-Linnaean descriptions and

plates. This species is also represented in the Linnean Herbarium and the specimen

(or one of the specimens) may be proposed as the lecto-type.

Perennial with ellipsoid to fusiform, yellow or reddish tubers. Stems herbaceous,

containing a milky juice, prostrate or ascending, occasionally twining, up to 5 m. long,

much branched, terete or angular, rooting at the nodes, glabrous or hairy. Leaves

broadly ovate to orbicular in outline, acute to obtuse, mucronulate, broadly cordate

to truncate at the base, 4-14 cm. long and 4-11 cm. wide, entire or more or less deeply

palmately 3-5 (-7)-lobed; the lobes broad or narrow, broadly ovate to linear-oblong,

both surfaces glabrous or thinly hairy; nerves green or purple; petioles glabrous or

hairy, 4-20 cm. long. Inflorescences cymosely 1- to several-flowered; peduncles stout,

angular, glabrous or hairy, 3-18 cm. long; bracteoles minute, narrow, acute, 2-3 mm.

long, early deciduous; pedicels 3-12 mm. long. Sepals subcoriaceous; outer ones

oblong or elliptic-oblong; inner ones elliptic-oblong to ovate-oblong, all glabrous or

pilose on the back and fimbriate, acute or subacute and ending in a mucro, subequal

in length or inner ones longer, 7-8 mm. long or sometimes the inner ones 9-12 mm.

long. Corolla pale mauve, campanulate-funnelshaped, glabrous, 3-4-5 cm. long.

Ovary hairy or occasionally glabrous. Capsules rare or absent in cultivated specimens,

ovoid, 4- or less-celled. Seeds glabrous.

Probably originally a native of America, but now widely cultivated in the tropics,

occasionally found run wild as a culture-relic.

A few specimens in the herbaria which were not reported to be cultivated ones

may indicate that in Southern Africa also specimens are occasionally found as culture-

relics and that is why this species is included here.

24. I. obscura (L.) Ker-Gawl., Bot. Reg. 3 (1817), t. 239; Choisy in DC., Prodr. 9

(1845), p. 370; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 140, and 28 (1899), p. 38;

Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 62 ex parte; Baker and Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 164; Ooststr. in Blumea 3 (1940), p. 519,

and in Steen., FI. Males. Ser. I, 4-4 (1953), p. 471. Convolvulus obscurus L., Sp.

PI. Ed. 2 (1762), p. 220.

Type: As there seems to be no recognised type in the Linnaean Herbarium, the

plate and description in Dill., Hort. Elth., p. 98, t. 83, f. 95 were taken as the type.

The plate and description are sufficient for recognition and post-Linnaean authors

all seem to agree on its identity.

Perennial. Taproot not thicker than a finger, 50 cm. long and over, with yellowish

bark. Stems several to many from the apex of the root, prostrate or twining, 1-2 m.

long and over, slender, terete, glabrous or patently hairy, sometimes almost lanate,

green and herbaceous when young, older portions with a longitudinally grooved and

transversely split bark and tending to become woody. Leaves often secund on prostrate

stems, herbaceous, ovate-cordate to oblong-cordate, cordate-sagittate to broadly

cordate or occasionally cordate-reniform, acute to obtuse and mucronate, with broadly

rounded basal lobes and a narrow or wide sinus, entire or slightly undulate and often

ciliate along the margin, paler beneath, thinly pubescent on both sides or glabrous,

sometimes densely hairy, 2-5 (-10) cm. long and 2-4-5 (-9) cm. wide: petioles erect

747

on creeping stems, slender, glabrous or hairy like the stems, 1 • 5-4 (-9) cm. long. In-

florescences 1 -flowered to cymosely few-flowered; peduncles slender, 1-14 cm., but usually

3-8 cm. long, glabrous or shortly hairy; bracteoles minute, narrow, acute; pedicels

usually 1-2 cm., minutely verrucose, Ihortly hairy or glabrous, at first erect but in fruit

reflexed and thickened towards the apex. Sepals subequal or the outer ones a little

shorter, subcoriaceous, much imbricate, subacute, mucronulate 3-4 mm. long, shortly

pubescent or occasionally glabrous: outer ones ovate with narrow, white margins, the

middle portion thicker, inner ones broadly ovate, thinner; in fruit all somewhat accres-

cent, turning brown and ultimately often spreading or reflexed. Corolla funnel-shaped,

pale yellow to white with purple centre in the var. obscura, pale yellow and concolorous

in the var. fragilis, glabrous or the midpetaline areas thinly hairy towards the apices

and the adjoining parts of the limb finely ciliate, usually 2-3 cm. long and the limb

as much in diam., but not infrequently smaller, only 12-20 mm. long; the limb spreading,

shallowly 5-lobed-5-angled: midpetaline areas conspicuous. Ovary glabrous. Capsule

broadly ovoid, apiculate, straw-coloured, 7-9 mm. long and 6-8 mm. in diam. Seeds

brown but appearing greyish or drab from the very dense, short, adpressed velvety,

and shiny tomentum, 4-5-5 mm. long.

Two varieties can be distinguished:

24a. I. obscura var. obscura. Flowers white or pale yellow with a dark purple

centre.

This is a rare form in South Africa and only recorded from the coast of Natal

and south east Transvaal. Its distribution seems to be mainly East African; also

in the Mascarenes, tropical Asia, Malaysia to northern Australia and Fiji.

Transvaal. — Nelspruit, Pretoriuskop: v.d. Schijff 2638 (PRE). Barberton,

Umvoti Creek: Galpin 657 (PRE).

Natal. — Nr. Durban: Wood 728 (SAM), 3861 (BOL, PRE); Umhlanga: Wood

1424 (BOL); Melmoth Road: Lawn 725 (NH).

24b. I. obscura var. fragilis ( Choisy ) A. Meeuse in R. A. Dyer, Flow. PI. Afr. 31

(1956), pi. 1222. I. fragilis Choisy in DC., Prodr. 9 (1845), p. 372; Hall. f. in

Bull. Herb. Boiss. 7 (1899), p. 50; Baker and Rendle, op. cit., p. 165; Hutch,

and Dalz., FI. W. Trop. Afr. 2 (1931), p. 216. I. tenuis E. Mey. ex Drege, Zw.

Pfi. geog. Doc. (1843), p. 139, 144, 156, 159, nomen tantum; Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 140, nomen nudum. I. longipes Engl, in Engl. Bot. Jb. 10

^1888), p. 246, non Garcke (1849). I. demissa Hall. f. in Engl. Bot. Jb. 18 (1893),

p. 129 and 28 (1899), p. 38: Baker & Rendle, op. cit., p. 140. I. obscura Baker

& Wright, op. cit., pro majora parte.

Flowers pale yellow, concolorous.

Type of the variety: Choisy mentioned a Drege specimen (“ I. tenuis E. Mey.”)

and Burchell 2362. Mr. de Winter compared both numbers with material from South

Africa.

This is the form which is very common in a large area of South Africa and also

in Angola, Bechuanaland, Southern Rhodesia and West Africa, but occurs in East

Africa as well; apparently not found outside the African continent.

The var. fragilis has been recorded from South West Africa (wide-spread);

Bechuanaland; Griqualand-West (Vryburg, Kuruman, Hay, Barkly West, Kimberley);

Transvaal (wide-spread, not recorded from Christiana, Bloemhof, Schweizer Reineke,

Heidelberg, Witbank, Middelburg, Bethal, Standerton); Swaziland; Natal and Zululand;

also recorded from the Hoopstad, Bloemfontein, Fort Beaufort and Komgha districts.

The following specimen is of special interest:

748

Cape Province. — Barkly West, Boetsap: Marloth 981 (isotype of Ipomoea

longipes Engl., PRE).

I cannot distinguish the var. fragilis from the typical form apart from the difference

in the corolla. Both forms can be twining or prostrate, glabrous or more or less densely

hairy, 1 -flowered or cymosely few (2-7)-flowered. The only constant distinguishing

character seems to be the colour of the corolla-tube (either with a dark purple spot

at the base inside, or concolorous) so that the two forms are treated as varieties here.

Ipomoea demissa Hall, f., although first referred to the section Calycanthemum

by Hallier (and also by Baker & Rendle), was later correctly placed in the affinity of

I. obscura and I. fragilis by the same author (Hallier 1899). However, apart from the

smaller corolla, there is not one character to distinguish I. demissa from I. fragilis.

Specimens with small flowers corresponding with I. demissa are not infrequently found

(e.g. in South West Africa, Transvaal); they are either depauperate plants, or flowered

late in the season. Sometimes the corolla is only 12 mm. long. These specimens all

link up with the normal form in which the corolla is 20-30 mm. long.

25. I. transvaalensis A. Meeuse , nom. nov. — /. convolvuloides Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 140; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 60, non

Schinz (1888).

Type: Zeyher no. 1216 in herb. DC. (Geneva).

Perennial forming annual stems from a long fusiform tuberous rootstock. Stems

prostrate and herbaceous, up to 1 m. long, or occasionally suffruticose at the base,

erect and up to 30 cm. high, usually densely and rather shortly pilose as are petioles,

leaves and peduncles, sometimes very densely or only sparsely so. Leaves herbaceous,

entire, varying from broadly cordate-suborbicular to narrowly deltoid-cordate, up to

about 4 cm. long rarely longer, usually distinctly cordate at the base, usually obtuse

to subacute, minutely mucronate; veins usually somewhat raised and more densely

hairy beneath; petioles up to 12 mm., rarely up to 20 mm. long. Peduncles 1-2-flowered,

usually shorter than the leaves; bracts minute, lanceolate; pedicels usually under

15 mm. long; less hairy than the peduncles, thickening upwards, distinctly thickened

and subclavate in fruit. Sepals herbaceous, subequal, ovate-lanceolate, lanceolate or

ovate-oblong, inner ones with thinner pale margins, acute, minutely mucro-tipped,

usually thinly and softly hairy outside, rarely quite glabrous, in fruit not or hardly

accrescent but glabrescent. Corolla rose-coloured (deep-pink) or magenta-pink to

reddish purple, funnel-shaped, 3-5 cm. long and the spreading limb as much in diam.,

glabrous except a few short hairs towards the tops of the midpetaline areas and dense

short tufts of hairs protruding from the tips. Capsule subglobose or somewhat ovoid,

glabrous, about 8 mm. in diam. Seeds densely and shortly velutinous with drab or

liver-coloured hairs, about 5 mm. long.

Transvaal, but related forms, possibly of varietal status, in Southern Rhodesia

and East Africa.

Transvaal. — Rustenburg, Groenkloof: Van Dam H. no. 10982 (PRE). Water-

berg, nr. Warmbaths: Bolus 12163 (BOL); Smuts & Gillett 3085 (PRE); Codd 2251

(PRE); Meeuse 9018 (PRE, L); id. nr. Nylstroom: Van Dam H. no. 23378 (PRE).

id. Geelhoutkop: Breyer H. no. 25229 (PRE). Potgietersrust, Vlakfontein (about

18 miles due S. of Grass Valley: Meeuse 9611 (PRE). Brits, Silkaatsnek: Bottomly

s.n. ; Acocks 12426; Codd 737, 2616 (all PRE). Brits or Rustenburg, Castle Gorge

(Kasteelpoort) : Meeuse 9370 (PRE); Vermeulen s.n. (PRE). Pretoria, nr. Rust de

Winter Dam: Codd 3494. id. nr. Premier Mine: Menzies 1; Rogers 25031 (all PRE).

Bronkhorstspruit, Kameelpoort: Meeuse 9535 (PRE); without precise locality (but

prob. Brits distr.), “ Magaliesberg ”: Zeyher 1216 (BOL, isotype!); Burke 166 (SAM).

749

This very distinct species, characterised, among other things, by the tufts of hairs

at the tips of the midpetaline areas which are especially conspicuous in the late bud

stage, was described as /. convolvuloides by Hallier, but this name is illegitimate on

account of /. convolvuloides Schinz (— Merremia tridentata ssp. angustifolia). So far

as can be ascertained, no new name has ever been suggested before, so that I propose

to name it I. transvaalensis , the species probably being the only endemic species of

Ipomoea occurring in the Transvaal.

The occurrence of almost sufifruticose, erect forms next to prostrate ones which

can also be observed in several other species (e.g., Ipomoea crassipes — the erect form

was described as /. greenstockii because it was considered to be a different species)

is probably to be attributed to the prevailing ecological conditions.

I. transvaalensis is often (always?) found on rocky slopes (Magaliesberg, Waterberg)

in either exposed or shaded places and this may well account for the considerable

variation in the habit, shape of leaves and degree of pubescence.

An isotype (Zeyher 1216) is present in BOL and, in addition, Mr. de Winter

compared several specimens with an isotype at Kew. Burke 166, collected at the same

time as Zeyher 1216 (and practically an isotype) is represented by a duplicate in SAM.

26. I. bathycolpos Hall, f in Engl. Bot. Jb. 18 (1893), p. 144; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 61.

Type: Zeyher 1218 in herb. DC. (Geneva), isotypes in BOL and SAM; the

equivalent gathering Burke 175 in PRE and SAM.

Perennial , with a thin woody taproot producing several annual prostrate stems.

Stems terete or angular up to about 2 m. long; scabrid with rough raised points as

are petioles, peduncles, pedicels and, more thinly so, the main veins of the young leaves

below. Leaves subcoriaceous, usually secund, more or less peltately attached to the

petiole, cordate, cordate-oblong, cordate-sagittate to triangular-cordate, sometimes

sagittate or cordate-reniform, acute to broadly rounded, sometimes acuminate, usually

mucronate at the apex, with a deep basal sinus and incurved or spreading usually

oblong basal lobes, 1-5-4 -5 cm. long and 0-75-3-5 cm. wide (up to 5-5 by 4 cm.

in the var. sinuatodentata)\ upper surface glabrous or nearly so (except sometimes the

midrib which can be scabrous), rough, finely netted-veined, lower surface similar but

with somewhat raised veins; margin thickened, cartilaginous, scabrid, subentire to

sinuous or shallowly dentate or with 1 or 2 coarse teeth (with large irregular teeth all

round in var. sinuatoclentata)\ petioles usually shorter than or about as long as the

leaves, rarely distinctly longer. Peduncles 1 -flowered, rarely 2-flowered, equalling or

somewhat exceeding the leaves; bracteoles minute, lanceolate or ovate; pedicels

thickened (much more so in fruit), 0-5-2 cm. long. Sepals unequal, chartaceous,

outer ones obtuse, about 11 mm. long, inner ones gradually longer and more acute,

the innermost up to nearly 20 mm. long, much imbricate, dark green, olive green to

brown or deep purplish green, scabridulous in lower half. Corolla funnel-shaped

with very spreading limb, pale mauve to almost white with darker, mauve-magenta

centre, 3-5-5 cm. long (-7 cm. in the var. sinuatodentata ) and the limb as much in

diam., glabrous or nearly so; midpetaline areas conspicuous. Capsule globose or

depressed-globose, glabrous, about 15 mm. in diam. Seeds about 10 mm. long, with

fawn villous tomentum.

Endemic.

1332796-8

750

26a. I. batbycolpos var. bathycolpos.

Recorded from the following districts:

Cape Province. — Mafeking.

Orange Free State. — Hoopstad.

Transvaal. — Marico, Rustenburg, Lichtenburg, Wolmaransstad, Vereeniging,

Heidelberg, Krugersdorp, Johannesburg, Germiston, Benoni, Pretoria, Middelburg,

Belfast, Carolina, Ermelo, Pilgrim’s Rest, Nelspruit, Barberton; one record from each:

Waterberg, Potgietersrust, and Pietersburg. (A specimen in GRA leg. Hutton no. 434,

labelled: “ Howick, Natal ” is most probably wrongly labelled and came presumably

from Johannesburg.)

26b. Ipomoea bathycolpos Hall. f. var. sinuatodentata Hall. f. [in Bull. Herb. Boiss. 7

(1899), p. 53] is a form with larger leaves which have large irregular teeth, and a

narrow basal sinus, and usually with larger flowers. The type, Wilms 988 in herb.

Berlin was collected near Lydenburg and this form seems to be restricted to the

Lydenburg district, but it is doubtful if this variety is worth maintaining.

Transvaal. — Lydenburg: Wilms 988 (PRE, L, JE, isotynes of variety); Marais

69 (PRE, L).

27. I. papilio Hall.f. in Bull. Herb. Boiss. 6 (1898), p. 543; Rendle in Jl. Bot. 39 (1901),

p. 56; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 63; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 167. I. papilio Hall. f. forma pluriflora Merxm.

in Mitteil. botan. Staatssamml. Munchen H.6 (1953), p. 204.

Type: Not designated by Hallier, who quoted three Rehmann gatherings and

Galpin 624. The specimen Gal pin 624 in K (from Barberton, Transvaal) is proposed

here as the type specimen (isotypes in PRE and Z).

Perennial. Stems slender, firm in texture, trailing or sometimes also climbing,

puberulous when very young, glabrous, obscurely pubescent or scabridulous when

older, up to 3 m. long. Leaves herbaceous, broadly cordate or cordate-reniform to

cordate-ovate, usually abruptly acuminate into a triangular, usually entire, obtuse to

very acute, mucronate apical portion, coarsely and irregularly few-toothed in the lower

portion, but sometimes more gradually narrowed towards the apex and/or toothed

or somewhat sinuous in upper portion, 2-6-5 cm. long and 2-7 cm. wide; basal sinus

always broad and rounded but usually rather shallow; the blade glabrous or nearly

so on both surfaces when old except for the minutely and obscurely ciliate margin;

petioles slender, generally shorter than the blades, minutely hispidulous or scabrid

like the stem. Peduncles 1 -flowered or cymosely 2-5 (-7)-flowered, shorter or longer

than the leaves, slender, hispidulous or scabrid like stems and petioles or minutely

pubescent; bracts ovate, minute, usually scabrid; pedicels somewhat thickening

upwards, 5-12 mm. long, minutely hispidulous, scabrid or pubescent. Sepals unequal,

imbricate, thinly coriaceous, glabrous or minutely and thinly pubescent; outer ones

oblong, elliptic or ovate to somewhat spathulate, obtuse, 5-6 mm. long, inner ones

considerably longer, obtuse to almost truncate or faintly emarginate, minutely mucronate,

7-9 mm. long; in fruit hardly accrescent but somewhat spreading. Corolla funnel-

shaped with horizontally spreading limb, light magenta or “rose-red”, 2-3-5 cm.

long, 3-3-5 cm. in diam., glabrous, hardly lobed; mid-petaline areas well-defined.

Capsule globose, or broadly ovoid-conical, 6-9 mm. in diam., glabrous. Seeds brown,

3-4 mm. long, thinly ashy-pubescent and with a dense tuft of white or yellowish short

hairs round the hilum.

751

Transvaal, Swaziland, Southern Rhodesia, extending into Northern Rhodesia

and Portuguese East Africa. In the Transvaal recorded from Marico, Rustenburg,

Pretoria, Middelburg, Belfast, Lydenburg, Barberton, Waterberg, Pietersburg, Pot-

gietersrust, Zoutpansberg. In Southern Rhodesia wide-spread, from Bulawayo to

Umtali (and extending eastward into Portuguese East Africa) and from the Transvaal

border to the Zambesi (extending into Northern Rhodesia at least as far North as

Mazabuka).

Merxmiiller recently described a forma pluriflora, which is supposed to be distinct

in having 2-5-flowered peduncles. The inflorescences are often 1- or few-flowered on

a single specimen and this character is absolutely useless to distinguish “ forms ”,

Hallier, in his original diagnosis of I. papilio, already mentioned 1-3-flowered inflores-

cences and it is a well-known fact that in this family many species have normally

1-flowered inflorescences, which sometimes are few-flowered, and others have normally

few-flowered peduncles, which occasionally (by reduction) are 1-flowered.

28. I. crispa ( Thunb .) Hall. f. in Engl. Bot. Jb. 18 (1893), p. 143. Convolvulus crispus

Thunb., FI. Cap. Ed. 2 (1818), p. 15, and Ed. Schultes (1823), p. 168. Ipomoea

contorta Choisy in DC., Prodr. 9 (1845), p. 350; Baker & Wright in Dyer, FI.

Cap. 4, 2 (1904), p. 58. I. undulata Baker ex Baker & Wright, op. cit., p. 60.

Type: In herb. Thunberg, Uppsala (not seen).

Perennial . forming annual stems from the crown of a thick taproot. Stems slender,

prostrate, covered with short bristly hairs when young, glabrescent or persistently

pubescent when older, up to at least 75 cm. long. Leaves herbaceous, entire, ovate

or oblong, rounded to acute, minutely mucronate at the apex, rounded, truncate,

subcordate or broadly cuneate at the base, with usually distinctly crisped margin,

1 -5-3 (-4) cm. long: thinly covered with adpressed bristly hairs, especially underneath,

sometimes nearly glabrous except on the nerves below; petioles slender, 5-10 mm.

long, hairy like the stems. Peduncles 1-flowered, shorter or longer than the leaves,

usually slender, hairy like the stems or scabrous when glabrescent; bracteoles 2, small,

linear or lanceolate up to about 1 cm. long; pedicels very short; peduncles and pedicels

thickening in fruit; pedicels lengthening and reaching about 10 mm. Sepals subequal,

oblong or lanceolate-oblong, acute or obtuse, 8-13 mm. long, hairy outside, in fruit

enlarged, indurate, becoming 14-17 mm. long and 3-5 mm. wide, glabrescent. Corolla

funnel-shaped, “ purple ” (probably magenta), 3-5 cm. long, nearly glabrous outside

to silky on the midpetaline areas. Capsule globose, 8-9 mm. in diam., light brown,

glabrous, apiculate. Seeds about 5 mm. long, densely greyish-velutinous.

Cape Province. — Alexandria, Zwart Hoogte: Burke (PRE). Albany, nr. Grahams-

town: Ecklon &Zeyher 26 - 1 = no. 6 (SAM); McOwan s.n. (GRA); Lotsy & Goddijn

s.n. (L); Daly 665 (GRA, PRE); Dalv & Cherry 880 (GRA, BOL); Schonland 682

(GRA); Dyer 1 1 17, 1205 (GRA); 1432 (GRA, PRE); Koonap Heights: Britten 2025

(GRA); between Grahamstown and King William’s Town: Smuts and Gillett 2504

(BOL). Bedford, nr. Bedford: Dyer 2328 (GRA, PRE), Comins 741 (PRE). Stocken-

strom, Fort Armstrong: Martin 154 (BOL, NBG). Victoria-East, Alice, Breakfast

Vley: Barker 2805 (NBG). Peddie: 15 m. from Peddie on E. London Rd. : Barker

3973 (NBG). In FI. Cap. also recorded from Fort Beaufort.

The type in the Thunberg herbarium was studied by Hallier and by N. E. Brown.

In addition. Baker & Wright cite a specimen collected by Thunberg, and the above-

mentioned identity is, therefore, certain [see also Juel, Plantae Thunbergianae (1917),

p. 383].

752

/. crispa can easily be distinguished from related plants (such as I. pellita, Turbina

oblongata ) by the shorter sepals, smaller crisped leaves with recurved apical mucro,

the slender 5—10 mm. long petioles and the velutinous seeds.

The midpetaline areas are not glabrous outside as stated in FI. Cap., but usually

thinly hairy. The leaves can be subcordate so that there is not one character to

distinguish 1. undulata Baker from I. crispa.

29. I. lapathifolia Hall.f. in Engl. Bot. Jb. 18 (1893), p. 142; Baker & Rendle in Dyer,

FI. Trop. Afr. 4, 2 (1905), p. 168; A. Meeuse in R. A. Dyer, Flow. PI. Afr. 31

(1956), pi. 1209. I. zambesiaca Baker in Kew Bull. 1894, p. 70, non Britten.

I. hellebarda Schweinf. ex Hall. f. var. lapathifolia (Hall, f.) Hall. f. in Engl. Bot.

Jb. 28 (1899), p. 44.

Type: Stuhlmann 109 (from Quelimane, P.E.A.) in herb. Hamburg (not seen).

Perennial. Steins several from a thin taproot, twining or occasionally prostrate,

at first slender, herbaceous, terete, glabrous, puberulous or sometimes hirsute, the

older parts becoming somewhat woody and producing a thin wrinkled brown bark,

ultimately becoming 2-3 m. long. Leaves broadly ovate to elliptic or oblong, or

occasionally lanceolate, herbaceous slightly fleshy drying papery, green above, paler

below, glabrous or thinly hairy, 4-10 cm. long and 2-5-7 cm. wide; the apex obtuse to

subacute, rarely acute or acuminate (in lanceolate leaves), shortly mucronate; the base

rounded, truncate or cuneate; the margin entire or somewhat sinuous; petioles minutely

scabridulous to hirsute, 1-4 cm. long. Inflorescences cymose, pedunculate, capituliform,

3-12-flowered or occasionally (mainly the first formed on young shoots) reduced to

a single flower; peduncles stouter than the petiole of the subtending leaf, terete, 1-15

cm. long; bracteoles triangular to lanceolate-subulate, erect, often keeled and concave,

acute, 1-3 mm. long, early deciduous; pedicels up to 10 mm. occasionally up to 20

mm. long, terete or somewhat 4-angled, subclavate. Sepals unequal, much imbricate,

subcoriaceous ; outer ones green, triangular to oblong or lanceolate-oblong from a

broad base, sub-acute minutely mucronate, finely muriculate or verrucose on the back,

6-7 mm. long and about 3 mm. wide; inner ones longer and with a rather broad

hyaline membranous edge, oblong to ovate, smooth, 7-8 mm. long and 4-5 mm. wide;

all slightly accrescent, at first closely adpressed to the fruit but later patent to reflexed,

turning brown. Corolla hypocrateriform, glabrous; tube subcylindric very pale purple

with fine purple striations outside, magenta inside except a white zone below the

insertion of the stamens, 23-28 mm. long and about 4 mm. in diam. in upper part;

the throat magenta inside; the limb pure white with yellowish green, well-defined

midpetaline areas, horizontally spreading, 25-35 mm. in diam., shallowly 5-lobed-5-

angled. Stamens not hairy at the base. Ovary glabrous; stigma pale mauve. Capsule

globose, ultimately pale greyish brown, glabrous, apiculate, about 8 mm. in diam.

Seeds usually 4, dark brown, glabrous or minutely puberulous, 4-5 mm. long.

East tropical Africa, northern part of Bechuanaland (N’gamiland), Belgian Congo.

Transvaal. — Barberton, Komatipoort: Rogers s.n. (PRE); Codd 7791 (PRE).

Nelspruit: Acocks 16633 (PRE). Pilgrims Rest or Nelspruit: van der Schijff 1521

(PRE).

S. Rhodesia. — Salisbury: Wild 1045 (SRGH); Kerr h. no. 45074 (SRGH, PRE).

“ Premier Mine ” : Martineau 255 (SRGH). “ Batoka Plateau ” : Allen 441 (SRGH).

Nr. Victoria Falls: Rogers 5007 (GRA).

Bechuanaland. — N’gamiland: Curson 410 (PRE).

Belgian Congo. — Ruzizi plains: Germain 5543 (PRE) — this record extends the

range considerably.

753

The identification of I. zambesiaca Baker, based on Kirk and Scott specimens,

with /. lapathifolia, was made by Hallier (1899). Although I have not seen the type

specimen, there can be little doubt as regards the identity of the cited specimens with

I. lapathifolia-, at any rate, they were compared by Mr. de Winter with the original

specimens of /. zambesiaca Baker, which were cited by Hallier under /. hellebarda

var. lapathifolia.

Although I. lapathifolia is indeed closely related to I. hellebarda, I agree with

Rendle (in FI. Trop. Afr.) that the former should be regarded as specifically distinct

from the latter. I. lapathifolia has leaves with cuneate base, 2-4 cm. long flowers and

seeds which are glabrous or subpuberulous, sometimes with a ring of hairs around the

hilum; I. hellebarda has leaves with sagittate, hastate or cordate base, larger flowers

and velvety-pubescent seeds. At any rate, the name I. hellebarda was only validly

published by Hallier in 1899 and cannot replace I. lapathifolia (1893).

There is some variation in the pubescence, the calyces and the length of the pedicels.

The stems are usually glabrous, but sometimes they are thinly hispid with yellowish

hairs (at least when young). The leaves are usually tapering from the base into the

apex, covered with short bristly hairs on the nerves below, thinly strigose above and

below between the nerves, ciliate, sometimes almost completely glabrous except the

ciliate margin. The petioles are almost invariably shortly and rather densely hirsute.

The flowers usually open during the night or very early morning and on all but

very overcast days close before 9 a.m.

30. I. aquatica Forsk., FI. Aegyp.-Arab. (1775), p. 44; Baker & Rendle in Dyer, FI.

Trop Afr. 4, 2 (1905), p. 170; Ooststr. in Blumea 3 (1940), p. 528 and in Steenis,

FI. Males. Ser. I, 4-4 (1953), p. 473; Brenan in Mem. New York Bot. Gardens

9 (1954), p. 7. I. reptans Poir. in Lamk., Encycl., Suppl. 3 (1814), p. 460, non

Convolvulus reptans L. ; Choisy in DC., Prodr. 9 (1845), p. 349; Hutch, and Dalz.,

FI. W. Trop. Afr. 2 (1931), p. 215. I. sagittaefolia Hochreut. in Candollea 5

(1934), p. 186, non Burm. f. I. dinteri Schulze-Menz in Notizbl. Bot. Garten

Berlin-Dahlem 15 (1941), p. 457, e descr. I. natans Dinter et Suesseng. in Mitteil.

bot. Staatssamml. Miinchen H. 4 (1952), p. 112.

Type: Not seen. Dr. B. Verdcourt mentioned (in litt.) that the specimen in

the Forskahl herbarium does not agree with the description, but an isotype in BM

does. The specimen in C must have been wrongly labelled.

Herbaceous perennial (but sometimes annual in unfavourable habitats). Stems

several to numerous from a stout woody base, thick, terete or striate, hollow or spongy,

rooting at the nodes, trailing on moist soil or mud or floating on water, up to 2-3 m.

long, entirely glabrous or hairy at the nodes; no subterranean tubers (the species is

easily propagated by cuttings). Leaves very variable in shape and size: ovate, triangular,

ovate-oblong, lanceolate or linear, acut or obtuse to retuse and mucronulate at the

apex, truncate or rounded at the base in the narrower leaves, but more often cordate

to sagittate or hastate, with rounded or with acutish to acute, entire or dentate auricles,

3-15 cm. long and 1-9 cm. wide; margin above the auricles entire or coarsely dentate;

petiole thick, glabrous, shorter or longer than the blade, 3-20 cm. long. Peduncles

axillary, glabrous, thinner than the petiole, 1-12 cm. long, cymosely 1 -few-flowered,

pedicels longer than the calyx, glabrous, 20-65 mm. long; bracts minute, narrow, acute,

1 -5-2 mm. long. Sepals thinly coriaceous with thin, pale margins, glabrous, sometimes

verrucose, equal in length or the outer a little shorter, the latter ovate-oblong, obtuse,

minutely mucronulate or blunt, 7-8 mm. long; the inner ones ovate-elliptic, obtuse,

minutely mucronulate, about 8 mm. long. Corolla funnel-shaped, pink or mauve,

often with a purple eye, rarely entirely white, 3-5 cm. long with a 4-5-5 cm. broad

limb, or in depauperated specimens only 2-5 cm. long, glabrous; the tube slightly

754

constricted at the place of insertion of the filaments. Ovary glabrous; style filiform,

glabrous; stigma biglobular, papillose. Capsule ovoid, glabrous, 8-10 mm. long.

Seeds densely pubescent.

Circum tropical.

S. W. Africa. — -Okavango: Dinter 7236 (BOL, PRE, isotypes of I. natans Dinter

et Suesseng.); Schoenfelder 40 (PRE); nr. Angolan border, Kachipu and Cunene

river banks; Barnard h. nos. 32322, 33135 (SAM).

Bechuanaland Prot. — N’gamiland: Curson 407 (PRE). Tsotsorogo Pan (in

N. part): Van Son s.n. (PRE).

Also in Angola, Southern Rhodesia and Portuguese East Africa.

Hallier [in Meded. Rijksherbarium Leiden I (1910), p. 20] has pointed out that

Convolvulus reptans L., the type of which he was able to examine, is the same species

as Merremia caespitosa (Roxb.) Hall. f. [= M. hirta (L.) Merrill], The name Ipomoea

reptans Poir. refers to the plant usually known as I. aquatica Forsk. and not to Linnaeus’s

species. It is, therefore, not a new combination, but a “ new name ” and is consequently

of more recent date than /. aquatica Forsk.

Hochreutiner found a specimen in the herb. Burman which is, according to

Van Ooststroom (1953), indeed Ipomoea aquatica , but the latter pointed out that this

specimen cannot be the type of Ipomoea sagittaefolia Burm. f., because it does not bear

the name “ Ipomoea sagittaefolia ” in Burman’s handwriting (it is labelled “ Convolvulus

sagittaefolius Burm.” in Houttuyn’s handwriting) and it does not resemble the plate

in Burm. f., FI. Ind. (1768), t. 18, f. 2. The actual type specimen of /. sagittaefolia

Burm. f. could not be traced at Geneva and Van Ooststroom could only tentatively

identity Burman’s plate with Ipomoea maxima (Linn, f.) Don ex Sweet. At any rate

Burman’s name cannot replace I. aquatica Forsk., as was erroneously concluded by

Hochreutiner.

Ipomoea natans Dinter & Suesseng. is a perfectly normal specimen of I. aquatica

Forsk. The specimen Dinter 7236 on which the former was based is represented in

PRE and BOL by duplicates and was compared by Mr. de Winter with the African

material of I. aquatica in the Kew herbarium. The habit of describing new species

on a single specimen without sufficient checking of more material is not to be recom-

mended, because these “ new ” species mostly have to be sunk later and the only gain

is a number of unnecessary synonyms.

I. aquatica is often cultivated or kept in a semicultivated state in many tropical

countries, because the leaves are used as a vegetable. However, the specimens found

in Southern Africa appear to be perfectly wiid and have not, or at least not recently,

been introduced by man.

31. I. pes-caprae (L.), Sweet , Hort. Suburb. Londin. (1818), p. 35; Roth, Nov. PI.

Spec. (1821), p. 109; Choisy in DC., Prodr. 9 (1845), p. 349; Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 145, and in Bull. Herb. Boiss. 5 (1897), p. 376; Hutch, and

Dalz., FI. W. Trop. Afr. 2 (1931), p. 215; Ooststr. in Blumea 3 (1940), p. 532,

and in Steenis, FI. Males., Ser. 1, 4-4 (1953), p. 475. Convolvulus pes-caprae

L., Spec. PI. Ed. 1 (1753), p. 159. C. brasiliensis L., Sp. PI. Ed. 1 (1753), p. 159.

Ipomoea biloba Forsk., FI. Aegypt.-Arab. (1775), p. 44; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 52; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p.

172; Marloth, FI. S. Afr. 3 (1932), p. 106, t. 27, fig. B.

Type: Linnaeus based his Convolvulus pes-caprae originally on the plates “ Herm.

lugdb. 174 t. 175” and “ Rheede, mal. 11, p. 117, t. 57” and his C. brasiliensis on

“Plum. amer. 89, t. 104”.

755

Perennial. Stems from a thick woody base, long-trailing and rooting at the nodes,

stout, firm, often hollow, terete or angular, often flattened, laticiferous, 5-30 m. long.

Leaves often secund, firm, sub-coriaceous, sub-orbicular, obovate, elliptic or transversely

elliptic, emarginate or sometimes truncate, rarely rounded at the apex, mucronate,

glabrous, 3-10 cm. long, 3- to 10-5 cm. wide, midrib below with 2 glands at the leaf-

base; petioles glabrous, up to 12 (-17) cm. long. Peduncles secund, stout, angular

or flattened, 1- or cymosely few-flowered, glabrous, 3-16 cm. long; primary cyme-

branches 1-7 cm. long; bracts or bracteoles small, ovate-lanceolate, 3-3-5 mm. long;

pedicels 12-30 mm. in fruit to 45 mm. long. Sepals subequal or outer ones a little

shorter, subcoriaceous, glabrous, obtuse, mucronate, outer ones distinctly 3-5-nerved,

ovate to elliptic, 5-8 mm. long, inner ones broader to orbicular, very concave, 7-11

mm. long. Corolla funnel-shaped, glabrous, pink, magenta or purple, darker inside

at the baser, rarely entirely white, 3-5 cm. long. Capsule subglobose or ovoid, 12-15

mm. long, glabrous. Seeds black, densely brownish tomentose-villous, 6-7 mm. long.

Circumtropical; in S. Africa from Knysna, Plettenberg Bay (Cape Province),

eastwards along the coast in all suitable localities (sandy beaches) to Natal and

Zululand.

This species was divided by Van Ooststroom (l.c.) into two subspecies. The

South African specimens seem to belong to the ssp. brasiliensis (L.) Ooststr. For full

synonymy, cf. Van Ooststroom (1940).

I. pes-caprae is mainly found on sandy beaches, but occasionally it occurs more

inland, e.g. along the shores of Lake Nyassa.

32. I. simplex Tliunb ., Prodr. FI. Cap. (1794), p. 36 and in FI. Cap. Ed. Schultes (1823),

p. 170; Wood & Evans, Natal PI. 1 (1898), p. 15, t. 15; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 48. Convolvulus plantagineus Choisy in DC., Prodr. 9

(1845), p. 405. Ipomoea plantaginea (Choisy) Hall. f. in Engl. Bot. Jb. 18 (1893),

p. 147 and in Bull. Herb. Boiss. 7 (1899), p. 53.

Type: leg. Thunberg from Uitenhage (in herb. Thunberg, Uppsala; photograph

of type in PRE).

Glabrous perennial. Rootstock tuberous, obovoid to fusiform, usually 2-4 cm.

in diam. and 4-10 cm. long. Stem or stems erect, slender, woody, usually partly

underground (up to 9 cm.) and gnarled, very rough, above the ground very short or

occasionally up to about 10 cm. long, angular and sulcate, rather stout to slender.

Leaves usually approximate, linear to lanceolate, entire or with a few pinnately arranged

teeth or lobes up to 15 mm. long, acuminate or acute at the apex, 3—10 cm. long and

1-6 (-12) mm. wide, long-tapering and almost sessile or sometimes rounded with

a distinct petiole at the base. Flowers solitary, axillary, white (in only one case reported

to be pale pink, perhaps they turn pinkish after wilting as in some species of Convol-

vulus)-, peduncles always short, bracteoles linear or linear-subulate, variable in length,

those of one pair sometimes unequal in length; pedicels short. Sepals lanceolate or

oblong-lanceolate to ovate-lanceolate, subequal or the inner ones longer, acute, 9-15

mm. long. Corolla funnel-shaped, 20-35 mm. long and 20-25 mm. in diam., shallowly

lobed. Capsule globose or somewhat ovoid, glabrous, 7-9 mm. in diam. Seeds dark

brown with a brownish, very short velutinous tomentum, 4-5 mm. long.

Cape Province. — Uitenhage, Alexandria, Bedford, Queenstown, Albany, Bathurst,

East London, King William’s Town, Komgha, Kentani, Umtata.

756

Orange Free State. — Ventersburg, Rouxville.

Basutoland.

Natal. — Inanda, Umvoti, Nkandhla, Zululand (one record).

Transvaal. — Lichtenburg, Krugersdorp, Johannesburg and adjoining Rand

districts, Pretoria, Heidelberg, Middelburg, Belfast, Ermelo and one record from

Pietersburg (Haenertsburg).

For the differences between I. simplex and I. bolusiana see under the latter.

Ipomoea plantaginea (Choisy) Hall. f. was based on a specimen collected by Drege

in the Zuurberg Mountains. This gathering proves to be conspecific with Thunberg’s

plant, but Hallier for some time mistook /. simplex for a different species (viz., /.

bolusiana ) until he had seen the type specimen in the Thunberg herbarium. On the

type sheet of I. simplex is a label attached by Hallier in 1909: “ Ipomoea (Leiocalyx)

simplex Thunb. = I. plantaginea Hall. f. 1893 ”.

The plant referred to I. simplex by Rendle in Jl. Bot. 39 (1901), p. 56 (and in FI.

Trop. Afr. 4, 2, p. 174 as I. simplex var. obtusisepala Rendle), viz., Rand 272, does not

belong here. It is a slightly depauperate specimen of 1. bolusiana (see under this

species).

33. I. welwitschii Vutke ex Hall. f. in Engl. Bot. Jb. 18 (1893), p. 146; Hiern. Cat.

Welw. Afr. PI. I, 3 (1898), p. 739; Rendle in Jl. Bot. 39 (1901), p. 57; Baker

6 Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 174. I. welwitschii Vatke ex

Hall, f., var. latifolia Britten in Jl. Bot. 32 (1894), p. 85; Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1905), p. 175. I. hystrix Hall, f., op. cit., p. 146, and in Bull. Herb. Boiss

7 (1899), p. 53. I. aspericaulis Baker in Kew Bull. 1894, p. 70. /. inamoena Pilger

in Engl. Bot. Jb. 45 (1910), p. 221; Dinter, Neue u. wenig bek. Pfl. S.W.Afr.

(1914), p. 36, fig. 54, inch var. trisecta Dinter, and in Fedde, Repert. 18 (1922),

p. 431. I. semisecta Merxm. in Trans. Rhodesia Sci. Assoc. 43 (1951), p. 41.

Type: Hopfner 39 from Okahandja, S.W.Afr. in JE, as this is the only specimen

on which Vatke wrote “ Ipomoea welwitschii Vatke ”.

Perennial , with subterranean tuber attaining the size of a man’s head (Dinter),

which has a dark bark, producing several annual stems, which are up to 50 cm., but

usually under 30 cm. long (Dinter), suberect to prostrate, stiff, rather stout, 2-4 mm.

thick, rarely slender, glabrous or minutely hirsute. Internodes mostly 1-3 cm. long.

Leaves usually secund, linear to linear-lanceolate or ovate-lanceolate to oblong, often

somewhat falcate or oblanceolate, usually tapering, acute to acuminate or apiculate

at the apex, or bilobed with an apiculus in the notch, obtuse, rounded or cuneate

at the base, up to 12 cm. long by 4 cm. wide, but usually 4-8 by 2-3 cm. or narrower

and only 4-20 mm. wide, rigid, glabrous or with short bristles on the nerves and margins,

usually conspicuously nerved with 5-8 ascending, long, rather straight nerves on

either side and often with distinct reticulate nervation; petioles short, stout, 5-10

mm. long, canaliculate above, glabrous or minutely hirsute like the stem; in some

specimens the leaves are more or less deeply incised (= I. semisecta Merxm. !) or

completely trisect (-5-sect), to palmately trifid (/. inamoena var. trisecta Dinter!); the

leaf-segments of deeply dissected leaves are generally under 1 cm. wide, the middle

one linear or lanceolate the lateral ones linear or falcate. Peduncles 1 -flowered or

occasionally few-flowered, either very short and under 1 cm. long, or sometimes

attaining 5 cm., stout, glabrous; bracts minute, lanceolate; pedicels subclavate, from

757

very short up to about 12 mm. long. Calyx glabrous; sepals subequal, ovate-oblong

to ovate-lanceolate, obtuse to subacute (occasionally acute), mucronate (mucro not

terminal), with narrow base which sometimes produces a subcordate appearance,

(8-) 10-12 mm. long, the outer ones slightly shorter than the inner ones. Corolla

(5-) 6-8 cm. long, pale pink or pale mauve, the tube usually darker inside, glabrous;

the limb 5-angled with distinct midpetaline areas ending in mucronate-aristate points

which are not always conspicuous in dried specimens. Ovary glabrous. Capsule

globose (-ovoid), glabrous, apiculate, 10-12 mm. long, about 10 mm. in diam. ; valves

brown outside, pale straw-coloured inside, rather coriaceous. Seeds about 4 mm.

long, glabrous with a basal tuft of hairs near the hilum.

South West Africa, Angola, Southern Rhodesia, extends into N. Rhodesia, Tangan-

yika, Nyasaland; one record from Bechuanaland.

Angola. — Alta Catumbela, Ganda: Faulkner 43 (PRE). Cume: Faulkner 74

(PRE). S. Bie, Kamundongo: Pocock 834 (BOL). Between Gwelei and Luarivi:

Pocock 770 (BOL). Between Gwelei and Kamundongo: Pocock 890 (BOL). Gwelei-

Cumboio: Pocock 881 (BOL).

S. W. Africa. — Otavi: Dinter 5258 (BOL, NH, PRE, SAM). Gobabis-Oas:

Dinter 2716 (SAM). Aitsas: Dinter 836 (SAM, Isotype of I. inamoena Pilger). Ozond-

jache: Dinter 1844 (SAM). Gaub: Dinter 2429 (SAM, PRE, one of the original

numbers quoted by Dinter as var. trisecta\ ). Okatjimane: Dinter 3302 (SAM, PRE,

also quoted by Dinter as var. trisectal). Eahero: Dinter s.n. (SAM). Babi-babi:

Wilman h. no. 1614 (KMG, also BOL).

Bechuanaland Prot. — Molepolole: Codd 8922 (PRE).

S. Rhodesia. — Umvuma, Mtao: Eyles 7555 (SRGH) and N.N. herb. No. 4290

(SRGH). Hartley, Poole: Hornby 3120 (SRGH, PRE). Salisbury: Eyles 4595

(SRGH); Wild 61 1 (SRGH). Shamva: Leviseur herb. no. 32905 (PRE). Felixburg:

Mainwaring in herb. Eyles no. 2805 (PRE, SRGH). Marandellas: Rattray 1378;

Corby 548; Dehn 533 (all SRGH). Rusapi: Dehn “ 193a” (a regathering of the

type of I. semisecta, SRGH). Umtali: Chase 1945 (SRGH). Manica: Teague 560

(BOL).

N. Rhodesia. — Munshiwemba: Stohr 398 (BOL).

A variable species, showing in some specimens narrow leaves, in others much

broader leaves or more or less completely dissected to trisect (or 5-sect) leaves.

Accordingly, the species was described several times under different names. After

having seen many specimens I feel confident that the above-mentioned reductions are

correct although I have not seen all the actual types. The only plants collected near

the type locality to which the description is applicable are those redescribed as

I. inamoena.

The identity of /. inamoena Pilger was already suggested by Dinter (1922) and was

put beyond doubt after I had seen an isotype (Dinter 836, in SAM).

I. semisecta Merxm. (of which I saw a regathering Dehn 193a in SRGH) agrees

with several specimens with more or less dissected leaves 1 have seen from S. W. Africa

and from Southern Rhodesia. Some specimens with partly entire, dissected leaves

(such as Dehn “ 193a ”) link up Dinter’s var. trisecta with the typical form and there

is a considerable variation in the width of the leaves, so that I prefer to distinguish

only one variable species without distinct varieties.

758

For the difference between /. welwitschii and I bolusiana see under the latter. It

must be emphasized that although there are several good distinguishing characters

(seeds, colour of flowers) some individual specimens may be difficult to assess to the

one species or the other.

34. Ipomoea bolusiana Schinz in Verh. Bot. Ver. Brandenb. 30 (29th Sept. 1888), p. 271;

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 147 and in Bull. Herb. Boiss. 7 (1899), p. 53;

Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 175. Ipomoea angustisecta

Engl, in Engl. Bot. Jb. 10 (9th Oct. 1888), p. 245, t. 7, fig. A: Baker & Wright

in Dyer, FI. Cap. 4, 2 (1904), p. 49. Ipomoea simplex Hook, in Bot. Mag. 72

(1846), t. 4206; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 146, non Thunb. 1. mesen-

terioides Hall. f. in Bull. Herb. Boiss. 7 (1898), p. 544; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 50. Ipomoea praetennissa Rendle in J. Bot. (1901), p. 56;

Baker & Wright, op cit. , p. 48. I. simplex Thunb. var. obtusisepala Rendle in

Dyer, FI. Trop. Afr. 4, 2 (1905), p. 174.

Type: Schinz 101 from S. W. Africa in herb. Zurich.

Perennial. Root tuberous, subglobose or somewhat fusiform, usually 5-10 cm.

in diam. Stems one or several, erect or prostrate; erect ones up to about 30 cm. long,

sometimes very short, prostrate ones up to at least 3 m. long, herbaceous or erect ones

woody at the base, glabrous, terete, usually slender. Leaves either palmately 3-7 (-9)

sect (mainly in creeping specimens) with very narrow, linear or filiform segments 2-7

cm. long and 0-5-3 mm. wide (sometimes somewhat pinnate to distinctly pinnate if

3 terminal segments are partly fused to form a common rhachis), or simple, linear,

sometimes linear spathulate and emarginate to bilobed at the apex, 4-10 (-15) cm.

long and 2-4 (-7) mm. wide; petioles of dissected leaves distinct, up to about 2 cm.

long, of simple leaves sometimes inconspicuous because the leaf is much narrowed

at the base. Peduncles 1 -flowered, very short, rarely longer and up to 5 cm. long;

bracteoles small, lanceolate, often deciduous; pedicels short, thickened. Sepals ovate-

lanceolate, lanceolate or oblong-lanceolate, sometimes ovate or elliptic, usually acute

or acuminate, glabrous, equal or unequal in length (7-) 9-16 (-20) mm. long. Corolla,

funnelshaped, said to be “ mauve ”, “ bright purple ” or “ rosy-purple ”, but probably

nearly always bright magenta-pink, somewhat lighter outside and darker in the centre

and on the midpetaline areas (see the plate in Bot. Mag. t. 4206), only once reported

to be white, glabrous, 4-7 cm. long and 4-6 cm. in diam. Capsule subglobose-conical,

often depressed at the apex, glabrous, apiculate by the style-base, 10-12 mm. long

and in diam. Seeds normally 4, either (immature?) with a short dense velutinous

shiny pubescence or with shiny fawn hairs longer than the 5-7 mm. long seed.

South Africa, extending into Angola, Bechuanaland and southern tropical Africa;

also in Madagascar.

Recorded from: South West Africa (wide-spread in northern part); Bechuanaland;

Griqualand West as far south as Kuruman and Hopetown; Orange Free State as far

south as Bloemfontein; practically the whole Transvaal; Swaziland; in Natal only

in a few northern districts and occasionally in Zululand. Common and frequent in

many places.

A variable species. The flowers usually are considerably longer (4-7 cm. long)

than those of the white-flowered I. simplex Thunb. (2-3-5 cm. long), which has also

relatively shorter and broader, linear-lanceolate leaves which are usually entire and

never palmatisect, greener, straighter and flatter than those of I. bolusiana. I. bolusiana

can be distinguished from the forms of I. welwitschii with palmately 3-sect leaves by

its usually narrower, indistinctly nerved leaf-segments (of which there are usually

759

5-9), its usually thinner stems and the perfectly glabrous stems and leaves (in 1. welwit-

schii the vegetative parts are often more or less scabrous to subhirsute, especially the

stem). The flowers of I. welwitschii are paler than those of I. bolusiana (pale pink or

pale mauve).

Specimens of both Schinz 101 (in GRA) and Marloth 111 (in PRE, isotype of

/. angustisecta ) and a specimen Rand 272 (in GRA, isotype of I. simplex var. obtusisepala )

were compared and they all represent one species. The type of I. mesenteroides Hall,

f. ( Rehmann 5267, from “ Klippan ”, i.e. near Grass Valley, Potgietersrust distr.,

Transvaal) was kindly sent on loan by the herbarium Zurich. The specimen is poor,

but a very similar plant was received from Rust-de-Winter, Transvaal, which shows

the same bilobed leaf-apices and Rehmann 5267 is, to my mind, referable to I. bolusiana.

I. praetermissa Rendle is represented in PRE by a photograph of the type and appears

to be the form of /. bolusiana with undivided, lanceolate leaves. The type of I. praeter-

missa (Zeyher 1214) was said to be without locality, but it is very likely that it was

collected in the Transvaal.

Hallier was of the opinion that “ I. simplex ” of Bot. Mag. (non Thunb.) was

different from /. bolusiana (= I. angustisecta ), probably because the leaves of the first

are undivided and of the latter dissected. However, these two extremes are linked up

by many specimens with divided and dissected leaves on one specimen. Hallier, when

he had seen Thunberg’s type of I. simplex, rectified his original mistake (see under

I. simplex), but he apparently never published his findings.

35. Ipomoea tricolor Cav., Icon. 3 (1794), p. 5, t. 208; Choisy in DC., Prodr. 9 (1845),

p. 359: Van Ooststr. in Van Steenis, FI. Males. Ser. 1, 4-4 (1953), p. 478. /.

violacea L., Sp. PI. Ed. I. v 1 753), p. 161 ; House in Ann., N.Y. Acad. Sci. 18 (1908),

p. 259; Van Ooststr. in Blumea 3 (1940), p. 541. L rubro-caerulea Hook, in Bot.

Mag. (1834), t. 3297; Choisy, op. cit., p. 375.

Type: The plate in Cav., Icon., was taken to be representative.

Herbaceous, glabrous twiner with perennial rootstock. Stems in S. Africa annual,

terete, up to 3 m. long, hollow. Leaves ovate, cordate at the base, long acuminate,

somewhat succulent drying thin and papery, 3-5-7 cm. long and 2-5-6 cm. wide;

petioles thin, 1-5-6 cm. long. Peduncles cymosely few-flowered, terete; fistulose, as

thick as the stems, 3-9 cm. long; bracteoles minute, triangular; pedicels 15-18 mm.

long, in fruit up to 25 mm. Sepals subequal, green with white margin, carinate along

the midrib, narrowly triangular to ovate-lanceolate, gradually narrowed towards the

apex, 4-5-6 mm. long. Corolla violet-blue or purple, with a white tube, but in

S. Africa usually the form with bright light blue flowers is seen (known as “ Heavenly

Blue ”), funnelshaped, 4-6 cm. long and as much in diam., glabrous. Capsule ovoid,

mucronate by the style-base, 8-10 mm. long. Seeds black, about 5 mm. long, minutely

puberulent.

A native of tropical America (extending into Mexico), widely cultivated in many

tropical countries and occasionally run wild, e.g. in Malesia (cf. Van Ooststroom 1953).

This species is cultivated in several places in the Union, such as Durban, Johannes-

burg, Pretoria, Irene, Warmbaths. There are two specimens among those I have seen

from Southern Africa which were not expressly stated to be cultivated and they may

be escapes from culture. These two specimens are: Boss s.n. from the Angolan border.

South West Africa (PRE) and Torre 1410 from Nampula, Niassa, Portuguese East

Africa (COI). The label on the specimen Torre 1410 suggests that the plant was actually

growing as a wild plant and was not cultivated.

760

36. Ipomoea coptica (L.) Roth ex R. et S., Syst. Veg. 4 (1819), p. 208; Roth. Nov.

PI. Spec. (1821), p. 110; Choisy in DC., Prodr. 9 (1845), p. 384; Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 147, id. 28 (1899), p. 45; Van Ooststr. in Blumea 3 (1940),

p. 544, and in Van Steenis, FI. Males., Ser. I, 4-4 (1953), p. 479; A. Meeuse in

R. A. Dyer, Flow. PI. Afr. 31 (1956), pi. 1217a. Convolvulus copticus L., Mant. 2,

App. (1771), p. 559. Ipomoea dissecta Willd., Phytogr. (1794), p. 5, t. 2; Baker

and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 67 ; Baker and Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1905), p. 176; Hutch, and Dalz., FI. W. Trop. Afr. 2 (1931), p. 218.

Type: One of the specimens in the Linnaean Herbarium is to be taken as the lecto-

type (photographs in PRE).

Annual quite glabrous except the very base of the stamens. Stems several from

the base, herbaceous, prostrate or climbing, up to 1-50 m. long, 4-angled, longitudinally

striate and often finely muriculate especially on the angles. Leaves pseudostipulate

by the leaves of developing or suppressed axillary shoots, orbicular to ovate in outline,

2-8 cm. long and wide, digitately compound with 5 (-7) ovate to oblong, lanceolate or

oblanceolate, coarsely serrate to deeply and irregularly (in some forms twice) pinnatilobed,

to pinnatisect segments; the middle lobe longest, lateral lobes gradually smaller, all

lobes more or less petioluled, acute or subacute at the apex, herbaceous, slightly fleshy

drying thin and papery; pseudostipules resembling the leaves, but smaller, up to 2 cm.

in diam.; petioles resembling the stems but usually a little thinner, 1-5 cm. long.

Peduncles cymosely 1-3-flowered, resembling the stems, 1-4 cm., in fruit up to 6 cm.

long; bracteoles setose to lanceolate, minute, acute, entire, 1-3 mm. long, or in some

specimens larger and palmately laciniate like the leaves; pedicels at first erect, in fruit

bent downwards, thickening upwards, 4-8 mm. long. Sepals subequal, oblong or

elliptic, sub-acute, minutely apiculate, 4-5 mm. long and about 1 • 5 mm. wide, thinly

coriaceous, muricate in 3 (outer ones) to 1 (inner ones) vertical lines, accrescent to

6 mm. long and 3 mm. wide in fruit. Corolla funnelshaped, dull white or pale cream

tinged with pale mauve on the tube outside, about 12 mm. long, the spreading limb

5-lobed-5-angled with mucronate lobes, 10-12 mm. in diam. Capsule depressed-globose,

brown when ripe, 6-7 mm. in diam., minutely apiculate, 3-celled and dehiscing by

(4-) 5-6 valves. Seeds 6 or less by abortion, triquetrous, dark brown, covered with

a grey or silvery-white silky tomentum except on the black circular hilum, 2-3 mm. long.

Tropical Africa, extending into South West Africa and the Transvaal, Sudan,

also in tropical Asia and North Australia.

S. W. Africa. — Recorded from the N. part as far south as Okahandja.

Bechuanaland.

Transvaal. — Zoutpansberg, Pietersburg, Waterberg, Pretoria, Groblersdal, Kruger

National Park, Barberton. On account of its small flowers possibly often overlooked;

this plant is for instance common around Warmbaths and in the Pretoria district,

but is represented in all the S. African herbaria by not more than 35 gatherings.

37. Ipomoea dasysperma Jacq., Eclog. I (1811-1816), p. 132, t. 89; Choisy in DC.,

Prodr. 9 (1845), p. 386; C. B. Clarke in Hook. f„ FI. Br. Ind. 4 (1883), p. 215;

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 148; Baker & Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1905), p. 179; Skan in Bot. Mag. 145 (1919), t. 8788; Ooststr. in Blumea

3 (1940), p. 580. Convolvulus pedatus Roxb., Hort. Beng. (1813), p. 14, nomen

nudum, and in FI. Ind., Ed. Carey, I (1832), p. 478. Ipomoea odontosepala Baker

in Kew Bull. 1894, p. 73; Baker & Rendle op. cit., p. 180. /. saccata Hall. f.

in Engl. Bot. Jb. 28 (1899), p. 48; Baker & Rendle, op. cit., p. 180, ex descr.

/. calcarata N. E. Brown apud Baker & Rendle, op. cit., p. 180, and in Kew Bull.

1909, p. 124.

Type: Jacquin’s plate was taken to be representative.

761

Annual glabrous herbaceous climber. Stems slender, attaining 2 m. or more.

Leaves herbaceous, broadly-cordate-ovate in outline, 3-10 cm. in diam., pedately

tripartite, to twice tripartite, terminal segment 3-sect, lateral ones 2- or 3-sect or

occasionally entire; segments lanceolate, linear, lanceolate or elliptic, acute to acuminate,

entire, 0-4-2 cm. wide; petioles 2-6 cm. long, often pseudostipulate by the leaves of

young or suppressed axillary shoots. Peduncles 1-8 cm. long, 1—3, occasionally many-

flowered, bracteoles minute, pedicels 0-5-3 cm. long, subclavate, thicker than the

peduncle. Sepals coriaceous with thinner margins, elliptic to ovate or oblong, obtuse,

7-10 mm. long and 5-6 mm. wide, outer ones sometimes a little shorter than the inner

ones and with 2 or 1 saccate or calcarate or gibbous basal processes, sometimes almost

spurred. Corolla hypocrateriform-funnel-shaped, with spreading subentire limb; the

tube pale mauve outside, darker mauve or bright magenta inside, 2-4 cm. long sometimes

longer; the limb bright yellow, paler outside, up to 7 (-10) cm. in diam.; midpetaline

areas distinct, somewhat greenish. Capsule subglobose, about 1 cm. in diam., glabrous.

Seeds about 6 mm. long, densely villous and often also with long cottony hairs on the

angles.

Widely distributed, but according to Skan (l.c.) very probably not native in some

of the localities from which it is recorded. Judging by some of the African localities

(S.W. Africa, Bechuanaland, Nubia, Abyssinia) where this plant was collected 50-100

years ago, it seems to be indigenous to Africa. In addition it is found in India and

Ceylon. There is no satisfactory evidence that it is native in China or Australia, and

Van Ooststroom (l.c.) reported that the records from Java all referred to specimens

formerl> cultivated in the Botanical Garden of Buitenzorg. In Africa recorded from

Nubia, Abyssinia, E. Africa, S. Rhodesia, Bechuanaland, S.W. Africa.

S.W. Africa. — Grootfontein, Tsumeb: Dinter 2506, 3019 (SAM), 7451 (BOL);

Namutoni: Barnard 223 (SAM); without precise locality, but probably nr. Groot-

fontein: Schoenfelder S835 (PRE).

S. Rhodesia. — Nr. Victoria Falls: Flanagan 3293 (BOL).

Most of the above-mentioned identities were already established by Skan. The

type of I. saccata Hall. f. ( Stuhlmann 210) was probably destroyed, but I feel very

confident about its reduction to I. dasysperma. There is some variation in the inflores-

cence (flowers solitary or in cymes), in the size of the flowers and in the development

of the pouches at the base of the outer sepals, but there is no reason to distinguish

more than one species.

Skan (l.c.) mentions Calonyction trichospermum var. diversifolium as a synonym

of I. dasysperma , but van Ooststr. (op. cit., p. 578) retains Ipomoea trichosperma

Bl. as a distinct species, occurring in Java, Celebes and the lesser Sunda Islands.

For the differences between I. cairicu, I. heptaphylla and I. dasysperma see under

I. heptaphylla.

38. Ipomoea cairica (L.) Sweet, Hort. Britt. Ed. 1 (1827), p. 287; Hall. f. in Engl. Bot.

Jb. 18 (1893), p. 148; Hutch, and Dalz., FI. W. Trop. Afr. 2 (1931), p. 216;

Ooststr. in Blumea 3 (1940), p. 542, and in Steenis, FI. Males., Ser. I, 4-4 (1953),

p. 479; Brenan in Mem. New York Bot. Garden 9 (1954), p. 8. Convolvulus

cairicus L., S>st. Ed. 10 (1759), p. 922. Ipomoea palmata Forsk., FI. Aeg.-Arab.

(1775), p. 43; Choisy in DC., Prodr. 9 (1845), p. 386; Baker & Wright in Dyer,

FI. Cap. 4, 2 (1904), p. 66; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 178.

Type: Most probably the specimen in the Linnean herbarium listed as “I.

cairensis ”, which agrees with the conception of this species by all authors.

762

Glabrous twiner with tuberous root. Stems smooth or tuberculate, high climbing

or accasionally prostrate. Leaves ovate to orbicular in outline, herbaceous, 3-10 cm.

in diam., deeply palmately dissected into 5-7 lanceolate, elliptic, ovate-elliptic, ovate-

lanceolate or somewhat oblanceolate segments, obtuse or acute, mucronulate, narrowed

at the base; basal lobes often bilobed in 5-lobed leaves; petioles 2-6 cm. long, usually

pseudostipulate by small leaves of developing or suppressed axillary shoots; pseudo-

stipules resembling the leaves but smaller. Peduncles 1- to few-flowered; 0-5-7 cm.

long; bracteoles minute; pedicels 12-20 mm. long. Sepals subequal or the outer

ones a little shorter, thick, green with pellucid dots and with pale, scarious margins,

often minutely tuberculate outside, 4-6 • 5 mm. long ; outer ones ovate, obtuse to subacute,

mucronulate, inner ones broader, obtuse, mucronulate. Corolla broadly funnel-shaped,

with the tube contracted near the base at the place of insertion of the stamens, mauve,

paler outside and with darker magenta centre (rarely entirely white), 3-5 cm. long and

4-6 cm. in diam. Capsule subglobose, glabrous, 8-12 mm. in diam. Seeds 5-6 mm.

long, densely and shortly tomentose and with white, up to 9 mm. long silky hairs along

the edges.

Africa, tropical Asia, naturalised elsewhere.

Recorded from the Cape Province (Uitenhage, Pt. Elizabeth, East London, Komgha,

Kentani, Port St. Johns); Natal (Durban, Pinetown, Pietermaritzburg, Inanda, Mtunzini,

Nongoma and several other localities in Zululand); Transvaal (Barberton, L>denburg,

Zoutpansberg; cultivated elsewhere, e.g. in Pretoria distr.). In addition Portuguese

East Africa, Southern Rhodesia, Angola, etc. Although South West Africa is men-

tioned among the localities in FI. Trop. Afr., I have not seen a single specimen from that

area. Frequently cultivated (in S. Africa as “ Messina creeper ”) and some localities

may refer to cultivated or naturalised specimens.

/. cairica has often been confused with “ I. pulchella Roth.” (= I. heptaplivlla )

which has smaller flowers, but can easily be distinguished from related species by

the characters given in the key. See also under /. heptaphylla.

39. J. hochstetteri House in Ann. New York Acad. Sci. 18 (1908), p. 223; A. Meeuse

in R. A. Dyer, Flow. PI. Afr. 30 (1955), pi. 1189. I. quinquefolia Hochst. ex Hall,

f. in Engl. Bot. Jb. 18 (1893), p. 147, and in Bull. Herb. Boiss. 6 (1898), p. 545

(as var. albiflora Hall, f.); Baker and Wright in Dyer, FI. Cap. 4, 2 (1904), p. 66,

excl. var. pubescens Baker; Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905),

p. 177, non L. (1753). I. quinquefolia Hochst. ex Hall. f. var. purpurea Hall. f.

in Bull. Herb. Boiss. 6 (1898), p. 546; Baker and Wright in Dyer, FI. Cap. 4, 2

(1904), p. 66; Baker and Rendle, op. cit., p. 178. I. kwebensis N.E.Br. in Kew

Bull. 1909, p. 123.

Type: Hallier (1893) quotes “ Ipomoea quinquefolia Hochst. mss. in Herb. Boiss.”,

so that the specimen on which Hochstetter had written this name can only be the

specimen Schimper 321 in herb. Geneva and this must be the type. (Hallier, l.c., only

cited one additional specimen, Steudner 954, in Herb. Berlin, which was destroyed.)

Annual , glabrous, herbaceous climber, occasionally prostrate when no support is

available. Stems several from the base, terete, smooth or finely striate, occasionally

somewhat muriculate, up to several metres long. Leaves digitatelv 5-lobed, 4-10 (-12)

cm. in diam.; lobes oblong or elliptic to ovate-lanceolate or lanceolate, attenuated

at both ends, obtuse to acuminate, mucronate, the base much narrowed and subpetioluled,

bright green, herbaceous and somewhat succulent drying papery, entire or one or

both of the basal ones with a lateral lobule up to 2-5 cm. long and leaf apparently

6-7-lobed; central lobe up to 7 by 3 cm., lateral ones gradually smaller, petioles

pseudostipulate at the base by young leaves of developing axillary shoots, often somewhat

muriculate, up to 7 cm. long. Inflorescences cymosely 3-6-flowered or occasionally

763

by reduction 1-flowered; peduncles usually 2-5 cm. long, sometimes shorter, rarely

longer; bracteoles minute, subulate, deciduous; pedicels stoutish, subclavate, 0-5-3

cm. long. Calyx densely and finely muriculate, 7-8 mm. long; sepals erect, sub-equal,

ovate-oblong or lanceolate, acute and with a fine mucro; inner ones with hyaline edges.

Corolla purplish mauve or white, funnel-shaped, 20-25 mm. long and about 25 mm.

in diam., the limb spreading, slightly 5-lobed; midpetaline areas well-defined, smooth.

Capsule tightly enclosed by the persistent calyx, globose, very shortly apiculate, muri-

culate when young, less distinctly so when ripe, 8-9 mm. in diam. Seeds dark brown,

velvety pubescent and often with very long, soft silky hairs on the edges.

From Abyssinia to S.W. Africa, Transvaal and Natal. According to Mr. de Winter

also in India.

S.W. Africa. — Grootfontein: Schoenfelder S 570 (PRE), prob. Grootfontein:

Schoenjelder S 848 (PRE); Namutoni-Sandup: Barnard 216 (SAM). Okahandja:

Dinter 131 (SAM); Otjisaza-road: Dinter 2585 (SAM). Otjiwarongo, Waterberg:

Bradfield 405 (PRE).

Cape Province. — Nr. Mafeking (Vryburg): Bolus 6460 (BOL).

Transvaal. — Pretoria, subspontaneous or spontaneous in garden of Div. of

Botany: Meeuse 9074, 9075, 9076; Roodeplaat Exp. Farm: Vermeulen s.n.; nr.

Hammanskraal : Codd 2740, Meeuse 9512; nr. Pienaars River Station: Codd 4040.

Warmbaths, Warmbaths: Meeuse 9077, 9078 (PRE). Waterberg. Naboomspruit,

Mosdene; Galpin M 238. Potgietersrust, nr. Grass Valley: Meeuse 9562 (all in PRE).

Pietersburg: Royers 14136 (BOL); Chunies Poort: Codd & Dyer 7756 (PRE). Zout-

pansberg, between Louis Trichardt and Pietersburg: Schweickerdt and Verdoorn 665

(PRE); Wylies Poort: Rodin 4229 (PRE). Letaba, The Gorge: v. d. Schijff 3025

(PRE). Pilgrims Rest, Satara: v. d. Schijff 2254 (PRE). Nelspruit, Skukuza: v. d.

Schijff 2737 (PRE).

Natal. — “ Thorns ”: Wood 4428 (NH, BOL; one of the original numbers quoted

by Hallier under this var. purpurea). Tugela Burns: Evans H. no. 19958 (NH).

Bechuanaland. — Mochudi: Harbor in Herb. Rogers no. 6519 (BOL); prob.

Mochudi: Rogers 6381 (BOL).

S. Rhodesia. — Bulawayo: Rogers- 5753 (BOL); Martineau 337 (SRGH). Hartley:

Hornby 3137 (SRGH).

Cultivated Specimen. — In Nat. Bot. Gardens Kirstenbosch ex Potgietersrust,

N.B.G. no. ^ (BOL).

House renamed Ipomoea quinquefolia Hochst. ex Hall, f., on account of /. quin-

quefolia L.

The two “ varieties ” distinguished by Hallier in 1898 (var. albiflora and var.

purpurea, respectively) cannot be maintained. White- and purple-flowered specimens

are often found growing together, as well as prostrate and climbing plants. N. E.

Brown, who raised the var. purpurea to specific rank under the name /. kwebensis

stated : “ (I. kwebensis differs from /. quinquefolia Hochst.) in that it is probably perennial

with very long twining stems, nearly or quite glandless petioles and purple flowers.

I. quinquefolia Hochst. has annual, short prostrate stems that show no tendency to

twine, glandular-tubercular petioles, white flowers and has hitherto only been found

in Abyssinia ”. However, the twining plants are also annuals, not perennials, and

can be white- or purple-flowered. Purple-flowered ones can be prostrate. Mr. de Winter,

who saw an isotype of I. hochstetteri and various authentic specimens referred to the

var. purpurea by Hallier or to I. kwebensis by N.E.Br., reported that they can hardly

be different and I fully agree with him.

764

I. hochstetteri is much more common in S. Africa than I. heptaphylla.

I. quinquefolia var. pubescens Baker is Merremia verecunda Rendle (q.v.).

40. Ipomoea heptaphylla ( Rottl . et Willd.) Voigt, Hort. Suburb. Calc. (1845), p. 360.

Convolvulus heptaphyllus Rottl. et Willd. in Ges. Naturf. Fr., Neue Schr. 4 (1803),

p. 196. ? Ipomoea pulchella Roth, Nov. PI. Spec. (1821), p. 115; Choisy in DC.,

Prodr. 9 (1845), p. 386; Hall. f. in Bull. Herb. Boiss. 7 (1899), p. 55, and in Engl.

Bot. Jb. 28 (1899), p. 48 (inclus. var. arachnosperma Hall, f.); Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 177; House in Ann. New York Acad. Sci.

18 (1908), p. 222: Ooststr. in Blumea 3 (1940), p. 544, in obs. and in Steen., FI.

Males. Ser. I. 4-4 (1953), p. 479, in nota.

Type: Rottler in herb. Willdenow no. 3721 (Berlin).

Herbaceous, glabrous, probably annual, climber. Stems very slender, subterete,

up to at least 1 m. long. Leaves pseudostipulate by the leaves of developing or suppressed

axillary shoots, herbaceous, drying thin, orbicular in outline, 2-6 cm. in diam., deeply

palmately 5-lobed, the lowermost lobes usually with a small lobe on the outside; lobes

elliptic to lanceolate, usually obtuse, minutely mucronate, narrowed and subpetioluled

at the base, entire; petioles usually a little longer than the blade, sometimes somewhat

muriculate; pseudostipules resembling the leaves but much smaller. Peduncles usually

shorter than the leaves and often shorter than the petiole of the subtending leaf, very

slender, almost filiform, terete, 1 -flowered to cymosely 2-3-flowered; bracteoles minute:

pedicels shorter and (much) thicker than the peduncle, somewhat subclavate. Sepals

thinly coriaceous, ovate to orbicular, subequal with membranous edges, usually obtuse

or rounded, sometimes muriculate near the base, 3-4 mm. long. Corolla funnel-shaped

with subcylindric tube, and spreading limb, purplish mauve, 12-18 mm. long and as

much in diam. Capsule subglobose or ovoid, 8-10 mm. long and 6-8 mm. in diam.,

glabrous. Seeds pubescent and usually also with long white hairs on the edges, 5-6

mm. long.

Tropical America and West Indies, tropical Africa, India.

Transvaal. — Zoutpansberg, Messina: Young s.n. = Moss 14607 (J, PRE).

Letaba, the Gorge: Smuts & Gillett 3535 (PRE).

Portuguese E. Africa. — Sul do Save, Guija: Myre 31 (LM, PRE); nr. Sabi

River, Meringua’s Kraal: Chase 2548 (SRGH).

Angola. — nr. Coroca Riv. : Baum 11 (COI).

S. Rhodesia. — Wankie, Deka River: Eyles 7968 (SRGH).

Tanganyika. — Central Province, Lake Kimagai: Hornby & Hornby 811 (PRE).

Hallier (Bull. Herb. Boiss. 7) has indicated that the type is still extant. Hallier

accepted its identity with I. pulchella Roth ex descriptione; this identity had already

been suggested b> Choisy. The name Convolvulus heptaphyllus Rottl. et Willd. is the

oldest name and is also the first use of the specific epithet in Convolvulus.

Mr. de Winter reported that he had seen an original sheet at Kew annotated by

Voigt. This sheet contains a mixture of I. heptaphylla and of I. hochstetteri, which

is irrelevant as far as the combination I. heptaphylla (Rottl. et Willd.) Voigt is

concerned, but may throw some doubt on the identity of Ipomoea pulchella Roth,

because obviously both /. heptaphylla and I. hochstetteri occur in India and without

a type specimen it is impossible to decide which form was described by Roth. If the

plant described as I. pulchella by this author was the same as I. heptaphylla Voigt,

it is merely a synonym, but if it was the same plant as I. hochstetteri the latter name

has to be replaced by /. pulchella Roth. I prefer to regard I. pulchella as a synonym

765

of I. heptaphylla Voigt, as was done by Choisy, Hallier and others and retain the name

/. hochstetteri for the other species. Both I. heptaphylla and I. hochstetteri are

represented by types and, therefore, well defined, whereas I. pulchella Roth is dubious.

Although I have not seen the type specimen of I. heptaphylla , I saw the specimen

Baum 1 1 referred to this species by Hallier. Other specimens referred to I. pulchella

by Hallier and quoted in FI. Trop. Afr. make it perfectly clear that Convolvulus hepta-

phyllus Rottl. et Willd. is the species with the filiform peduncles (see below).

The differences between /. heptaphylla and I. hochstetteri seem to be constant,

but the best distinguishing character is the shape of the peduncle, which is very thin

and filiform in I. heptaphylla and much stouter in I. hochstetteri. In addition, the

corolla seems to be always mauvish-purple in I. heptaphylla (sometimes white in I.

hochstetteri).

I. heptaphylla has, according to Van Ooststroom, often been confused with /.

cairica. The following table shows the differences between I. heptaphylla, I. hoch-

stetteri, I. cairica and I. dasysperma : —

41. I. alba L., Sp. PI. ed. 1 (1753), p. 161; Hall. f. in Meded. Rijksherb. Leiden 1

(1911), p. 25, and 46 (1922), p. 19; Ooststr. in Blumea 3 (1940), p. 547 and in

Steenis, FI. Males. Ser. I. 4, 4 (1953), p. 480. Convolvulus aculeatus L., Sp. PI.

ed. 1 (1753), p. 155. Ipomoea bona-nox L., Sp. PI. ed. 2 (1762), p. 228. Calonyction

speciosum Choisy in Mem. Soc. Phys. Geneve 6 (1833), p. 441, t. I, f. 4, excl. var.

/9, and in DC., Prodr. 9 (1845), p. 345, excl. var. y ex parte et var. S; Hall. f. in

Engl. Bot. Jb. 18 (1893), p. 153; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2

(1905), 117. Calyonyction bona-nox (L.) Boj., Hort. Maurit. (1837), p. 227; Hall,

f. in Bull. Herb. Boiss. 5 (1897), p. 379, 1028. Calonyction aculeatum (L.) House

in Bull. Torrey Bot. Cl. 31 (1904), p. 590; Hutch. & Dalz., FI. W. Trop. Afr.

2 (1931), p. 213.

Type: Linnaeus based I. alba on “ Rheede, mal. II, 101, t. 49 ”, Apparently

not represented in the Linnean herbarium.

Glabrous (or rarely pubescent) perennial twiner. Stems herbaceous up to 4 mm.

thick, laticiferous, terete, smooth or sometimes muriculate, up to 5 m. long. Leaves

thin, herbaceous, ovate to orbicular in outline to oblong or ovate-oblong, entire or

1332796-9

766

3-lobed (often on the same plant), 6-20 cm. long and 5-16 cm. wide; margin entire,

apex acuminate with acute or obtuse, mucronulate acumen, base cordate with broad

or narrow rounded sinus and broadly rounded to angular basal auricles; petioles

slender, 5-20 cm. long. Inflorescence 1- to several-flowered, cincinnate or dichasial:

peduncle stout, terete, 1-24 cm. long; bracteoles small, deciduous; pedicels 7-15

mm. long, in fruit much thickend and clavate and up to 30 mm. long. Sepals coriaceous,

elliptic, glabrous, unequal, 2 or 3 outer ones shorter, 5-10 mm. long, and with a long,'

thick, recurved or patent, 4-8 mm. long awn; inner ones 8-15 mm. long, mucronulate

with a much shorter and thinner 2-3 mm. long apiculum; all often reflexed in fruit.

Corolla opening at night, fragrant, hypocrateriform, white; the tube 7-12 cm. long,

the limb 1 1-14 cm. in diam. when fully expanded. Capsule ovoid, mucronate, glabrous,.

2-5-3 cm. long. Seeds brown or black, glabrous, smooth, 10-12 mm. long, 7-8 mm.

wide.

Circumtropical, originally from tropical America [cf. Hallier (1922), p. 19]. In

S. Africa cultivated and occasionally found as an escape (Queenstown, Coast of Natal).

42. I. riparia G. Don, Gen. Syst. 4 (Febr. 1838), p. 265; Choisy in DC., Prodr. 9 (1845),

p. 153; Exell, Catal. Vase. PI. S. Thome (1943), p. 251; Van Ooststr. in Van

Steenis, FI. Males., Ser. I. 4-4 (1953), p. 484; Brenan in Mem. New York Bot.

Garden 9 (1954), p. 8. /. lilacina Bl., Bijdr. FI. Ned. Ind. (1825), p. 716, non Schrank

(1822); Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 187; Hutch. & Dalz.,

FI. W. Trop. Afr. 2 (1931), p. 215. I. baclei Choisy in Mem. Soc. Phys. Geneve

8 (1838), p. 60, t. 2. Pharbitis fragrans Boj., Hort. Maurit. (1837), p. 227, nomen

nudum, ex Choisy in DC., Prodr. 9 (1845), p. 341 (descr.) I. fragrans (Boj. ex

Choisy) Boj. ex Choisy, op. cit. (1845), p. 341, in syn., et p. 393; Hall. f. in Engl.

Bot. Jb. 18 (1893), p. 153; Van Ooststr. in Blumea 3 (1940), p. 564. I. parkeri

Choisy, op. cit. (1845), p. 381. I. parkeri Choisy var. subsericea Meissn. in Mart.

FI. Bras. 7 (1869), p. 284; Van Ooststr. in Pulle, FI. Surinam 4 (1932), p. 92.

Type: Don s.n. in BM (fide Exell, l.c.).

A perennial twiner, mostly occuring in marshes, along rivers and in marshy forests.

Stems terete, finely striate when dry, densely short-pilose with soft white hairs. Leaves

broadly ovate to orbicular, acuminate at the apex with obtuse or acute mucronulate

point, broadly cordate at the base, densely pilose beneath, much more sparsely so and

ultimately glabrescent above, 5-12 cm. long and 4-10 cm. wide; nerves 7-9 on each

side; petiole slender, 3-10 (-12) cm. long, pilose like the stems. Peduncles pilose like

the stem but often glabrous or nearly so in the lower portion, terete, 2-12 cm. long,

cymosely 1 -few-flowered with very short cyme branches and consequently flowers

subumbellate; pedicels pilose, mostly longer than the calyx, 7-14 mm. long; bracts

ovate, minute. Sepals equal in length, 7-10 mm. long; the two outer ones elliptic-

oblong, acute, shortly pilose; the inner ones broader, ovate-elliptic, less acute. Corolla

funnel-shaped, mauve with darker centre, 4-5 cm. long; mid-petaline areas with

sericeous hairs outside. Capsule globose, glabrous, about 12 mm. in diam.; valves

brown outside, whitish inside. Seeds 6 mm. long, white-villous.

Distribution. — Tropical America (known as I. parkeri until Van Ooststroom

established its identity with I. lilacina Bl.), tropical Africa, Madagascar, Malesia.

Extending into the northern part of S.W. Africa (Okavango marshes) and Bechuana-

land.

South West Africa. — Niangana, in reeds on the Okavango: Dinter 72C5 (BOL,

PRE); Okavango Reserve: Maguire 1653 (NBG).

767

Bechuanaland Prot. — Okavango, among reeds: Schoenfelder S 176 (PRE) r

Story 4768 (PRE). Chobe River, Kasane: Van Son h. no. 28686 (PRE).

As regards the synonymy, the oldest name /. lilacina Bl. is illegitimate on account

of /. lilacina Schrank. The names following in priority are I. riparia G. Don and

I. baclei Choisy, both published in 1838, but Exell (l.c.) has pointed out that /. riparia,

publisued in February 1838, is almost certainly older than Choisy’s name, which was

used for the first time by Choisy in a paper read for the Soc. Phys. at Geneva by the

end of January 1838. This paper was, therefore, probably not published in the

“ Memoires ” of that society as early as February of the same year.

According to Van Ooststroom (private communication) the African specimens are

in every respect identical with the Asiatic ones.

I. riparia is a real marsh plant and is always reported on collectors’ labels as

growing “ among reeds ”, “ near edges of pools ”, “ along rivers ”, “ in marshes ”, etc.

43. I. digitata L., Syst. Nat. ed. 10 (1759), p. 924; Choisy in DC., Prodr. 9 (1845),

p. 389; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 64; Baker & Rendle

in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 189; Flutch. & Dalz., FI. W. Trop. Afr. 2

(1931), p. 216; Van Ooststr. in Blumea 3 (1940), p. 558, and in Van Steenis, FI.

Males., Ser. 1 .4.4 (1953), p. 483. Convolvulus paniculatus L., Sp. PI. Ed. 1 (1753),

p. 156. Ipomoea paniculata (L) R. Br., Prodr. FI. Nov. Holl. ed. 1 (1810), p. 486;

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 149 and in Bull. Herb. Boiss. 5 (1897), p. 393,

non /. paniculata Burm. f. (1768).

Type: (Not represented in the Linnean Herbarium).

Large glabrous perennial twiner, occasionally prostrate. Roots tuberous. Stems

terete, glabrous, hollow, becoming woody. Leaves orbicular in outline, palmately

dissected to or mostly beyond the middle (or, in the var. eriosperma ovate-cordate,

entire or shallowly lobed) cordate or truncate at the base, 6-14 cm. long and 6-15 cm.

wide, segments (3-) 5-7 (-9), lanceolate to ovate, entire, acuminate to obtuse, minutely

mucronate; middle lobe 5-9 cm. long and 1-5-3 cm. wide, lateral lobes shorter;

petioles smooth or muriculate 3-10 cm. long. Peduncles terete but often angular

near the apex, cymosely branched near the apex, few-flowered, 2-5-20 cm. long;

pedicels terete, 9-25 mm. long. Sepals equal or the outer ones shorter, all suborbicular

or the outer ones narrower, oblong to broadly elliptic, obtuse, concave, much imbricate,

coriaceous, 6-11 mm. long. Corolla funnel-shaped with spreading limb, mauve with

darker mauve centre, 5-6 cm. long and 6-7 cm. in diam., glabrous. Capsule globose

or ovoid, obtuse, glabrous, 8-10 mm. in diam., 12-14 mm. long. Seeds black, with

long wooly-sericeous, easily detachable whitish hairs, 5-7 mm. long.

Circumtropical.

Natal. — Durban, near Durban: Krauss 94 (BOL), Wood 10319 (NH, PRE).

Lower Tugela, Kearsney: Milner H. no. 23403 (NH). Hlabisa: Codd 7005 (PRE).

Ingwavuma: Ward 2025 (PRE).

In addition Angola, Portuguese East Africa and West and East tropical Africa

(apparently not in Southern Rhodesia or Bechuanaland).

Ipomoea digitata L. var. eriosperma ( P . Beauv.) Rendle in Dyer FI. Trop. Afr. 4, 2

(1905), p. 190 (sphalm. “var. eriocarpa ”); Ooststr., op. cit. (1953), p. 484.

This variety differs from typical I. digitata var. digitata in that the leaves are entire

to shallowly lobed instead of palmately divided to or below the middle.

768

It has not been recorded from South Africa, but a specimen leg. Myre et Carvalho

(no. 100) from Maputo, Sul do Save, Portuguese East Africa (PRE) was collected

dose to the Zululand border.

44. I. albivenia ( Lindl .) Sweet, Hort. Brit., Ed. 2 (1830), p. 372; Choisy in DC., Prodr.

9 (1845), p. 379; Hall. f. in Engl. Bot. Jb. 18 (1893), p. 151; Wood, Natal PI. 1

(1899), p. 32, t. 38; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 59. Con-

volvulus albivenius Lindl. in Bot. Reg. 13 (1827), t. 1116. Ipomoea gerrardi Hook

f. in Bot. Mag. (1867), t. 5651; Hall, f., l.c. and in Engl. Bot. Jb. 28 (1899), p. 51,

in obs.

Type: As the description was made from a living specimen, the type actually

is the plate in Bot. Reg. (t. 1116). However, there may be a herbarium specimen extant

made of the same plant which could be taken as representing the type. The species

in question is so distinct that there can be no doubt about its identity, in spite of Lindley’s

description of the calyx as being glabrous (it is, in fact, always somewhat hairy, at least

near the base).

Perennial climber. Stems shrubby, when young herbaceous with white somewhat

flossose tomentum, soon glabrous, older ones becoming woody, attaining 10 m. and

over in length. Leaves broadly cordate or sub-orbicular-cordate to cordate-reniform,

occasionally ovate or oblong, 3-10 (-17) cm. long and 3-10 (-16) cm. wide, rather

firm in texture, shallowly and widely cordate to truncate at the base, rounded to

cuspidate at the apex, entire to sinuous or somewhat crenate, when very young covered

on both sides with a white floccose tomentum, both sides glabrescent but on the lower

side the tomentum persists longest as a reticulate tomentum on the main nerves and

veins, ultimately sometimes quite glabrous; petioles rather slender, white-tomentose,

usually persistently so, 1-6- (-8) cm. long. Peduncles very short, 1 -flowered, tomentose;

bracteoles usually longer than the calyx, linear-spathulate or oblong-spathulate, mem-

branous, brown, glabrous inside, thinly floccosely tomentose outside, early deciduous

(not minute as stated in FI. Cap.); pedicels short, tomentose. Sepals broadly oblong,

ovate-orbicular, ovate, elliptic or obovate-oblong, obtuse, chartaceous, much imbricate,

11-15 mm. long, originally densely floccosely tomentose, glabrescent, retaining the

pubescence longest at the base, in fruit ultimately spreading to reflexed. Corolla

funnel-shaped, white, 6-9 cm. long and the limb 6-8 cm. in diam., glabrous. Capsule

ellipsoid or somewhat ovoid, 16-22 mm. long and 12-16 mm. in diam., brown, with

glabrous, coriaceous valves, apiculate. Seeds brown, 7-9 mm. long, densely covered

with very long cottony white hairs, giving the dehisced capsule the appearance of a

ripe cotton ball.

South Africa (Natal, Transvaal), Portuguese East Africa and the eastern part of

Southern Rhodesia.

Recorded from Natal and Zululand: northern districts as far south as Weenen,

Greytown and Mt. Edgecombe; Transvaal: Barberton, Lydenburg, Pietersburg,

Potgietersrust and the districts N. and E. of these; sometimes locally abundant.

The name is often cited as “/. albivenia (Lindl.) G. Don., Gen. Syst. 4, p. 270”,

but the correct citation is the one given above.

Ipomoea gerrardi Hook. f. is a synonym. One of the differences given is that this

plant is more hairy than I. albivenia, but the differences are very slight. In herbarium

specimens the calyx of open flowers is very rarely glabrous and, in my opinion, the

specimen figured in Bot. Reg. may have been slightly atypical because it was grown

in a hothouse. Lindley obtained his seeds from Forster and quotes “ Algoa Bay ”

.as the locality. It is obvious that his should be “ Delagoa Bay ”.

769

As a rule this plant is very distinct on account of the typical reticulate pubescence

on the veins of the young leaves. The older leaves are glabrescent and this is a good

distinguishing character against /. vevbascoidea = I. dammarana (which has a persistent

tomentum on the lower surface of the leaves), even if young leaves are lacking. Other

related species are even more completely geographically separated from /. albivenia

than /. verbascoidea and differ in some other respects. I. marmorata Britten & Rendle,

I. grandiflora Lamk. and I. lapidosa Vatke have hypocrateriform corollas and glabrous

sepals.

45. I. verbascoidea Choisy in DC., Prodr. 9 (1845), p. 356; Hall. f. in Engl. Bot. Jb.

18 (1893), p. 151; Baker and Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 183.

I. dammarana Rendle in Jl. Bot. 34 (1896), p. 36; Baker & Rendle; op. cit.,

p. 183. I. conceiroi Rendle in Jl. Bot. 46 (1908), p. 182. I. seineri Pilger in Engl.

Bot. Jahrb. 41 (1908), p. 297; Dinter in Fedde, Repert. 18 (1922), p. 431, e descr.

Type: A specimen “ ex Angola ” in herb. Paris (photograph of type in PRE).

Suberect to climbing perennial shrub. Stems terete, woody, 1-2 m. long in suberect

specimens, often longer in climbing ones, densely covered with a whitish, somewhat

floccose tomentum as are the leaves, petioles, peduncles, bracts and sepals. Leaves

firm, varying from suborbicular to cordate-oblong, acute or obtuse to apiculate, cordate

or truncate at the base, entire, or somewhat sinuous or crenate, 4- 15 cm. long and

3-17 cm. wide, upper surface thinly, lower surface densely tomentose, both usually

somewhat glabrescent, but upper surface most, veins somewhat raised and reticulated;

petiole 1-5-14 cm. long, bearing a gland on each side of the insertion of the blade.

Peduncles 1- to few-flowered, usually under 3 cm. long; bracteoles linear-oblong to

oblong or oblanceolate, 14-20 mm. long, brownish, almost membranous, tomentose

outside; pedicels short. Sepals elliptic, chartaceous, very obtuse, equal, 12-16 mm.

long. Corolla funnel-shaped, described as purple or rosy-purple (probably mauve),

glabrous, 6-10 cm. long and the limb 5-7 cm. in diam. Capsule usually oblong-ovoid

or ellipsoid, 20-25 mm. long, 10-15 mm. in diam., rarely globose, glabrous; valves

coriaceous. Seeds brown, 6-8 mm. long, densely covered with long white or sometimes

fulvous cottony hairs, giving the dehisced fruit the appearance of an open cotton boll.

South West Africa, Angola, Bechuanaland, Southern Rhodesia and Northern

Rhodesia.

Recorded from S.W. Africa in many localities in the northern part as far south

as Okahandja; in Bechuanaland, one record from N’gamiland.

A constant character is the white persistent tomentum on the lower surface of the

leaves, the young ones having in addition a raised tomentum on the main veins. The

closely related I. albivenia has glabrescent leaves and the calyx is usually not so

completely' tomentose as in I. verbascoidea.

The capsules of I. verbascoidea are usually not globose (as stated in FI. Trop.

Afr.), but oblong-ovoid, up to 2-5 cm. long and 1-1-5 cm. in diam. The seeds bear

long woolly hairs which are white or sometimes fulvous.

The differences between I. verbascoidea and I. dammarana (of which I saw an

isotype, Rand 273, in GRA, and a photograph of the type in PRE) do not hold and

these two names are clearly synonymous. Several other described species (such as

I. grantii Oliv.) are very close and may eventually prove to be synonyms. In any case

the name I. verbascoidea is the oldest and must be maintained at least for the specimens

occurring in Angola and South West Africa.

Ipomoea conceiroi Rendle, based on Gossweiler 2443 (flowers) and 4703 (fruits)

is not distinct. The types are represented by duplicates in COI and by photographs

1332796-10

770

in PRE and these specimens fall within the range of variation of I. verbascoideci. The

differences mentioned by Rendle (leaf-shape and ovoid, instead of subglobose, capsules)

break down altogether.

Ipomoea seineri (type destroyed) is reduced here to I. verbascoidea, because the

description agrees very well. Dinter (l.c.) also suggested this identity.

46. I. shirambensis Baker in Kew Bull. 1894, p. 72; Hallier f. in Engl. Bot. Jb. 28 (1899),

p. 49 (“ schirambensis ”); Baker & Rendle in D>er, FI. Trop. Afr. 4, 2 (1905),

p. 186; Brenan in Mem. New York Bot. Garden 9 (1954), p. 8.

Type: Kirk 93 from Shiramba, Lower Zambesi in herb. Kew.

Tall perennial woody climber. Stems glabrous or sometimes pubescent when

young, glabrous, firm and woody with gre>ish or yellowish bark and raised longitudinal

ridges when old. Leaves deciduous, ovate to orbicular, entire, long-acuminate, cordate

to truncate at the base, up to 9 cm. long and 7 cm. wide, densely pubescent to glabrous;

petioles shorter than the blades, slender. Flowers appearing before the leaves on the

naked branches, solitary or in 2-14-flowered congested, fascicle-like cymes; peduncles

and cyme-branches very short; bracts minute; pedicels thickening upwards from a

slender base, 10-25 cm. long. Sepals coriaceous, much imbricate, ovate, mucronate,

9-15 mm. long, outer ones usually the shortest, glabrous. Corolla funnel-shaped,

white or pale mauve with mauve centre, 3-5-6 cm. long and the limb as much in diam.,

glabrous, midpetaline areas conspicuous. Capsule ovoid, glabrous. 15-17 mm. long

and about 11 mm. in diam. Seeds about 6 mm. long, with very long spreading fulvous

cottony hairs.

From the northern part of Portuguese East Africa westwards; apparently common

in the western part of Southern Rhodesia and northern Bechuanaland, extends into

Northern Rhodesia, Nyassaland, Tanganyika, Belgian Congo, Transvaal and probably

Eastern Angola.

Transvaal. — Zoutpansberg, 2 miles N.E. of Punda Maria: Codd & Dyer 4560

(PRE); Punda Maria: v. d. Schi/ff 970, 3025, 3189 (PRE).

Bechuanaland. — Chobe: Miller B / 1 088 (PRE); Robertson & Elffers 91 (PRE,

SRGH), N’gamiland: Curson 908 (PRE); nr. Ngoma: Codd 7581 (PRE); without

precise locality: Pole-Evans 4603 (PRE).

Mozambique. — Sena: Lea 22 (PRE); Baroma, Sisitso: Chase 2648 (SRGH);

Manica e Sofala, between Muatize and Tete: Barbosa & Carvalho 3249 (PRE);

Mocuba: Faulkner “ Kew 79 ” (COl); Niassa, Meconta: 7o/rf 883 (COI); Nampula:

Torre 657 (COI).

One Transvaal specimen was compared with the type by Miss Kies at Kew and

was reported to be a good match of Kirk 93, and some other specimens* (e.g., Pole

Evans 4603) were named at Kew by Mr. de Winter.

47. I. adenioides Schinz in Verhandl. Bot. Ver. Brandenb. 30 (29th Sept. 1888), p. 270;

Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 51; Baker & Rendle in Dyer,

FI. Trop. Afr. 4, 2 (1905), p. 195. /. marlothii Engl, in Engl. Bot. Jb. 10 (9th Oct.

1888), p. 244; Rivea adenioides (Schinz) Hall. f. in Engl. Bot. Jb. 18 (1893),

p. 156.

Type: Not designated, because Schinz mentioned three specimens (not seen).

However, the species is so characteristic that it can easil> be recognised from the

description and 1 have seen specimens referred to this species by Hallier. The type

of I. marlothii Engl., reduced to its synonymy by Hallier ( Marloth 1250) is represented

by an isotype in PRE.

771

Erect shrub up to about 1-20 m. high. Young branches covered with very short

silky hairs, older ones glabrous, greyish or yellowish, canescent. Leaves at the ends of

the young branches firm to subcoriaceous, varying from lanceolate or oblanceolate

to elliptic, obovate to obovate-orbicular, usually cuneate at the base, obtuse to acute,

often with a densely white-hairy mucro, entire, penninerved, 3-8 cm. long and 1-5 cm.

wide; green, glabrescent above, densely silky beneath mainly on the veins, ultimately

glabrescent; margin densely and shortly ciliate when viewed from upper surface;

petioles much shorter than the leaves, sometimes very short, shortly silky-pubescent.

Peduncles 1 -flowered, short, densely silky-pubescent as are bracteoles, pedicels and

calyx; bracteoles linear to linear-lanceolate, acute, 8-18 mm. long; pedicels very short.

Sepals subequal, lanceolate, long-acute or acuminate, 1 5-18 mm. long, in fruit accrescent,

up to 28 mm. long. Corolla hypocrateriform; tube narrowly cylindric, 7-10 cm. long,

greenish white outside, deep magenta inside, adpressed silky outside, glabrous inside;

limb spreading, white, 4-5-6 cm. in diam., silky on the midpetaline areas outside.

Capsule ovoid, thinly hairy to glabrous, 19-23 mm. long and 14-18 mm. in diam.,

valves coriaceous. Seeds about 1 cm. long, densely covered with very long, shiny,

fulvous-brown hairs.

South West Africa, Bechuanaland., Transvaal, extending into the western part of

Southern Rhodesia and possibly into Angola, also recorded from Somaliland.

Recorded from S.W. Africa (practically the whole area in sandveld where there

are no real deserts, as far south as the Great Karasberg; apparently common in many

places); Bechuanaland (N’gamiland) ; Transvaal: Zoutpansberg, Petersburg, Lydenburg

(one record only), Potgietersrust, Waterberg and Pretoria districts; S. Rhodesia:

Wankie.

Hallier referred this plant to Rivea sect. Poliothamnus , but as is pointed out under

Turbina , this section has to be referred to the genus Turbina Rafin. However, Ipomoea

adenioides has dehiscent 4-valved capsiles and long-pubescent seeds and belongs,

therefore, in Ipomoea sect. Eriospermum.

The flowers of /. adenioides are open only during the late afternoon and evening-

Excluded species of Ipomoea : —

Ipomoea angusti folia Jacq. = Merremic tridentata (L.) Hall. t. subsp. angustifoha

(Jacq.) Ooststr.

/. argyreoides Choisy = Turbina oenotheroides (Linn, f.) A. Meeuse.

/. bcrrettii Rendle = Turbina oenotheroides.

/. bowieana (Rendle) Baker (= Merremia bowieana Rendle) = Convolvulus capensis

Burm. f. ssp. bowieanus (Rendle) A. Meeuse.

I. curtoi Rendle = Turbina curtoi (Rendle) A. Meeuse.

I. holubii Baker = Turbina holubii (Baker) A. Meeuse.

/. lambtoniana Rendle = Turbina oblongata (E. Mey. ex Choisy) A. Meeuse.

I. malvaefolia (Rendle) Baker = Merremia malvaefolia Rendle.

I. oblongata E. Mey. ex Choisy = Turbina oblongata (E. Mey. ex Choisy) A. Meeuse.

I. oenotheroides (L.) Rafin. ex Hall. f. = Turbina oenotheroides.

I. pyramidalis Hall. f. = Turbina pyramidalis (Hall, f.) A. Meeuse.

I. quinque folia Hochst. ex Hall, f., non L., quoad var. pubescens Baker = Merremia

verecunda Rendle.

772

I. rhodesiana Rendle = Turbina holubii.

I. robertsiana Rendle = Turbina robertsiana (Rendle) A. Meeuse.

I. saundersiana Baker = Operculina turpethum (L.) S. Manso.

I. shirensis Oliv. — Turbina shirensis (Oliv.) A. Meeuse.

I. stenosiphon Hall. f. = Turbina stenosiphon (Hall, f.) A. Meeuse.

I. sublucens Rendle = Turbina suffruticosa (Burch.) A. Meeuse.

/. suffruticosa Burch. = Turbina suffruticosa (Burch.) A. Meeuse.

I. tetraptera Baker — Merremia pterygocaulos [(Steud. ex) Choisy] Hall. f.

/. uncinata Hutch. = Turbina robertsiana.

I. woodii N. E. Br. = Stictocardia woodii (N. E. Br.) Hall. f.

Doubtful record: —

Ipomoea stolonifera (Cyrill.) J. F. Gmel., Syst. Nat. Ed. 13, 2 (1791), p. 345. Thunberg

described his Convolvulus radicans from a specimen which he reported to be

collected in South Africa. In FI. Capensis 4, 2, p. 52, this species is included as

I. carnosa R. Br., but the only specimen quoted is Thunberg’s specimen (the type

of C. radicans).

Ipomoea stolonifera has never been collected in S. Africa again and most probably

Thunberg’s locality is erroneous. The nearest locality I know is Angola, so that

this species is not included here.

15. STICTOCARDIA

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 159; Hutch, and Dalz., FI. W. Trop. Afr. 4

(1931), p. 218; Van Ooststr. in Blumea 5 (1943), p. 346, and in Van Steen., FI.

Males. Ser. I, 4-4 (1943), p. 491. Argyreia Sect. Pomifera C. B. Clarke in Hook, f.,

FI. Br. Ind. 4 (1883), p. 184. Argyreia Baker and Rendle in Dyer, FI. Trop. Afr.,

4, 2 (1906), p. 200, type excluded. Ipomoea Auct., pro parte.

Type species: Convolvulus tiliifolius Desr. = Stictocardia tiliifolia (Desr.) Hall. f.

(first species mentioned by Hallier and also the oldest species described).

Herbaceous or woody perennial twiners. Leaves entire, ovate to orbicular, usually

rather large to large, almost invariably cordate at the base and with numerous minute

glands on lower surface (appearing as black dots in dried specimens). Flowers axillary,

pedunculate, solitary or in few- to many-flowered cymes; bracts small, deciduous.

Sepals elliptic to orbicular, obtuse or rounded to emarginate, equal or subequal, often

much imbricate, subcoriaceous, frequently with thinner margins, much accrescent in

fruit. Corolla large, funnel-shaped with shallowly-lobed to subentire limb, usually

purple, magenta or reddish; the midpetaline areas often somewhat hairy and with

minute glands like the leaves. Stamens inserted near the base of the corolla-tube;

pollen globose, spinulose. Disc annular, entire or somewhat 5-lobed. Ovary 4-eelled,

4-ovuled, glabrous: style 1, filiform; stigma biglobular. Fruit completely enclosed

by the accrescent calyx, globose; its dissipiments with 2 transverse wings at the surface

of the fruits, woody; the wall between these wings thin, detaching irregularly from

the dissipiments and their wings so that 4 openings are formed and the fruit becomes

lantern-shaped. Seeds ultimately exposed by the 4 openings in the fruit, dark brown

or black, pubescent.

773

6-7 (possibly more) species; one circumtropical, a few in Africa and a few in

Asia and Malaysia.

This genus is very similar to Ipomoea in its vegetative and floral characters, but is

quite distinct by its typical fruit and by the black dots (glands) on the lower leaf surface.

The genus Stictocardia is very homogeneous and the species are very similar. One

species in S. Africa.

Stictocardia woodii (IV. E. Br.) Hall. f. in Bull. Herb. Boiss. 6 (1898), p. 548. Ipomoea

woodii N. E. Br. in Kew Bull. 1894, p. 101; Baker and Wright in Dyer, FI. Cap.

4, 2 (1904), p. 60; Wood, Natal PI. 6 (1912), t. 557.

Type: Wood 4146 and Wood 4864 are quoted by N. E. Brown. Both were taken

from the same cultivated specimen in the Durban Botanical Garden, as were Wood

4806, 10065, 11579, 13064. They are for practical purposes all “type specimens”.

Perennial with a large tuberous rootstock (Wood). Stems stout, the older portions

lying on the ground and reaching a length of 7 m. or more, the younger portions creeping

or climbing, finely pubescent, glabrescent. Leaves orbicular-cordate with broad,

shallow basal sinus, up to 25 cm. long and wide, thinly hairy to quite glabrous, the

apex obtuse to acuminate, minutely emarginate and mucronate: petiole 5-15 cm.

long, stout, channelled above. Inflorescences cymosely 3-8-flowered. or in upper

portions of the stems 1-3-flowered; common peduncle 2-6 cm. long, stout; pedicels

up to about 3 cm. long; bracteoles early deciduous, minute. Calyx somewhat inflated,

glabrous; sepals subequal, strongly imbricate, concave, coriaceous, ovate-oblong,

obtuse or rounded at the apex, the outer ones slightly larger then the inner ones.

Corolla funnel-shaped, rosy-pink (Wood) but also reported to be yellow according

to Hallier, 6-5-8 cm. long, the limb as much in diam., spreading, shallowly 5-lobed,

the lobes rounded to emarginate; midpetaline areas very distinct, quite glabrous or

bearded at the apex. Stamens somewhat unequal, included, shorter than the style.

Capsule uhknown.

Zululand and S. part of Portuguese East Africa, in lowland forests.

Zululand. — Ngoye: Wood 10354 (NH).

Portuguese E. Africa. — Lourengo Marques (Delagoa Bay): Junod, according to

Hallier (1898).

Cultivated Specimens. — In Durban Bet. Garden, originally from Zululand:

Wood 4146 (NH), 4806 (PRE, SAM), 10065, 11579, 13064 (NH).

This species is not quite identical with the four species mentioned in FI. Trop.

Afr. under Argyreia and Hallier stated that it is distinct from all the other African

species he distinguished. It differs, among other things, from the other species in

its glabrous calyx. Only a monographer of this difficult genus will be able to decide

if S. woodii is to be upheld or not. For the time being it is treated as a separate taxon.

16. TURBINA

Rafin., FI. Tellur, 4, 1836 (1838), p. 81 ; Ooststr. in Steen., FI. Males. Ser. I, 4-4 (1953),

p. 493. Legendrea Webb, et Berth., Hist. Nat. lies Canar., Bot. 3, 2 (1844), p. 26,

t. 137; Choisy in DC., Prodr. 9 (1845), p. 328; Van Ooststr. in Blumea 5 (1943),

p. 355. Rivea Choisy emend. Hall. f. in Engl. Bot. Jb. 18 (1893), p. 155, pro parte,

non Rivea Choisy 1833, nec Rivea Hall. f. in Engl. Bot. Jl. 16 (1893), p. 584.

Ipomoea Auct. plur., pro parte.

Type species: Convolvulus corymbosus L. = Turbina corymbosa (L.) Rafin.

774

Amended generic description: Mostly woody or suffruticose, decumbent or ascending

to scandent, rarely erect, almost invariably more or less pubescent or tomentose

perennials, sometimes with large fusiform tuberous rootstock. Leaves often subcordate

or cordate. Flowers solitary or in (sometimes subumbellate) cymes, or compound

cymes arranged in a terminal panicle. Floral characters generally as in Ipomoea.

Sepals equal or sometimes unequal, lanceolate to oblong or occasionally orbicular,

in fruit usually slightly enlarged and more or less spreading or loosely enclosing the

capsule. Corolla funnel-shaped, rarely hypocrateriform. Pollen spinulose. Fruit

indehiscent, with hard, more or less woody or thin and leathery pericarp, usually

apiculate by the persistent style-base, often only 1- or 2-seeded. Seeds often puberulous.

According to Van Ooststroom (1953), there are at least two species in America,

one of which (T. corymbosa) has been introduced in the Old World and occasionally

become naturalized, but he expects that there are probably more species under Ipomoea.

There are indeed at least eight species, found in Southern and Central Africa, which

were described under Ipomoea and have to be transferred to this genus.

Van Ooststroom discussed the delimitations of the genera Argyreia Lour., Rivea

Choisy s.s. and Turbina (as Legendrea ) in Blumea 5 (1943), p. 353-355. For details

the reader is referred to this discussion. The essential points are: —

(a) that Hallier (1893) extended the generic limits of Rivea Choisy (1833) to

include many species which are partly referable to Ipomoea sect. Eriospermum,

partly to Argyreia, partly to Turbina;

( b ) that the genus Rivea has to be brought back to its original limits as indicated

by Choisy in 1833 [not as treated in DC., Prodr. 9 (1845)] and by Hallier

in Engl. Jb. 16, p. 559-560, 504, and comprises two South-East Asian species

only;

(c) that Turbina (= Legendrea ) had to be given generic rank again. The African

representatives of Hallier’s genus Rivea in its wider sense were suggested

to belong to Turbina or to distinct genera closely related to Turbina (if not

referable to Ipomoea sect. Eriospermum).

The inclusion of at least eight African species necessitates some amendments in the

generic description of Turbina. This is not surprising because the larger the genus, the

greater usually the variation in characters. The species described as Ipomoea shirensis

Oliv. is very similar to the type species Turbina corymbosa and undoubtedly belongs in the

same genus. The species Ipomoea pyramidalis Hall, and I. curtoi Rendle are obviously

closely related to Turbina shirensis. There are also distinct relationships between

T. pyramidalis, T. shirensis and T. holubii. T. holubii and T. pyramidalis are both erect

or suberect, so that species of Turbina need not always be climbing shrubs. The last

two species link up with Ipomoea suffruticosa, but the latter has a long-apiculate fruit.

Ipomoea robertsiana is closely related to T. suffruticosa. Finally, the last-mentioned

species is again closely related to /. oenotheroides, which is an erect shrub with narrow

linear leaves, and to Ipomoea oblongata, a prostrate perennial. I cannot find any

distinct discontinuity which would exclude any of these species, but they are all sharply

distinguished from Ipomoea by their indehiscent fruits. It is true that as far as I know

the fruit and seed of I. pyramidalis and /. curtoi have not yet been found, but these

species are so similar to T. shirensis and T. holubii that they must belong to the same

genus. The species described as Ipomoea stenosiphon (= Rivea stenosiphon ) differs

in its hypocrateriform corolla, but this cannot be a sufficient reason to separate it from

the other ones included here in Turbina because in the related genus Ipomoea both

funnel-shaped and hypocrateriform corollas occur. The vaguer generic characters make

the distinction between Turbina and Argyreia less easy, although the consistency of

the pericarp (pulpy, fleshy or thick and leathery in Argyreia, hard and woody or thin

and leathery in Turbina ), apart from the geographical separation, seems to be a satis-

factory distinguishing character.

775

Erect, much branched shrub, usually shortly silvery-tomentose, rarely strigose;

leaves linear or oblanceolate, sessile or occasionally shortly petiolate,

narrowed at the base; flowers solitary on very short peduncles, rarely

peduncles longer and/or 2-flowered; sepals lanceolate; seeds glabrous 1. T. oenotheroides.

Prostrate or climbing plants, or, if suberect to erect, leaves usually distinctly

peholate, rounded or subcordate at the base, or plant not silvery

pubescent ;

Flowers solitary or 2 together; sepals equal or sub-equal, lanceolate or

ovate-lanceolate to oblong; prostrate to suberect plants, corolla

funnel-shaped;

Vegetative parts (except upper surface of older leaves), peduncles

and calyces densely and shortly silvery-tomentose, sometimes

whitish-silky; leaves generally elliptic-oblong not distinctly

recurved-mucronate at the apex (Griqualand West, Bechuana-

land, S.W. Africa) 2 T. suffruticosa.

Vegetative parts, peduncles and calyces, variously hairy but not

densely and shortly silvery-tomentose;

Leaves up to 35 mm. long, quite glabrous above, often with

crisped margin, frequently with recurved mucronate apex,

on the lower surface covered like the stems, peduncles

and calyces with rather stiff, white, more or less spreading

hairs; peduncles slender; seeds shortly velutinous

(Transvaal) 3. T. robertsiana.

Leaves often much longer, usually more or less hairy above;

pubescence on leaves and other parts usually yellowish

or brownish, seeds glabrous (wide-spread) 4. T. oblongata.

Flowers either in cymes or panicles, or, if solitary, sepals distinctly unequal,

obtuse, rounded or emarginate, or plants erect or climbing:

Corolla funnel-shaped, pale mauve with magenta centre; sepals

broad, suborbicular, unequal, hairy, flowers solitary or in

few-flowered cymes; erect to spreading much branched shrub

(tips of twigs sometimes sinuous or twining) 5. T. holubii.

Corolla white or cream-coloured; sepals oblong or elliptic; flowers

solitary (sometimes peduncles fascicled) or in panicles:

Corolla funnel-shaped, 2-3 cm. long; sepals equal, at first

tomentose, later spreading and ultimately denuded;

inflorescence a lax panicle with cymose branches, fruit

subglobose, tomentose, not distinctly apiculate 6. T. shirensis.

Corolla hypocrateriform, its tube 7-13 cm. long; sepals unequal,

the outer ones shorter, coriaceous, glabrous, much accres-

cent and loosely enclosing the glabrous ovoid distinctly

apiculate fruit 7. T. stenosiphon.

1. Turbina oenotheroides {Linn, f.) A. Meeuse, comb. nov. Convolvulus oenotheroides

Linn, f., Suppl. (1781), p. 137. Ipomoea oenotheroides (Linn, f.) Rafin. [FI. Tellur.

4 (1838), p. 74] ex Hall. f. in Engl. Bot. ,1b. 18 (1893), p. 156, in syn. I. carta

E. Mey. ex Drege, Zw. Pfl. geog. Doc. (1843), p. 45, 54, 195, nomen tantum.

I. argyreoides Choisy in DC., Prodr. 9 (1845), p. 357; Rendle in J. Bot. 40 (1902),

p. 191 ; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 49. Rivea oenotheroides

(Linn, f.) Hall. f. in Engl. Bot. Jb. 18 (1893), p. 156. Ipomoea harrettii Rendle

op. cit., p. 190; Baker & Wright, op. cit., p. 50.

Type: A specimen leg. Sparrman (according to Linn, f.), but no type specimen

could be located in the Linnean Herbarium or in the Stockholm herbarium.* The

identification of Convolvulus oenotheroides with 7. argyreoides Choisy has been done by

Hallier and he was followed in FI. Cap. The description given by the younger Linnaeus

* Note during correction: Quite recently the Stockholm herbarium sent authentic specimens (leg.

Spacnnan). They agree with the type of Ipomola argyreoides.

776

in his Supplementum Plantarum agrees very well with the type of I. argyreoides, except

that the calyx is described as glabrous. However, the calyx is called “ canis ” and this

obviously refers to the often very short canescent pubescence on the outer sepals.

In addition there is a specimen in the Thunberg herbarium, which, was labelled

“ Convolvulus oenotheroides ” by Thunberg and Thunberg must have obtained his

identification from Linnaeus the younger. This specimen was examined by Hallier

and by N. E. Brown who were both satisfied that it is the same as Ipomoea argyreoides

Choisy. There is a note in the Kew Herbarium on a specimen Zeyher 1206 which says

“ Matches Convolvulus oenotheroides of Thunberg Herbarium ”. The identity of

Convolvulus oenotheroides, therefore, seems to be definitely established.

Erect, suffruticose to shrubby perennial, 30-100 cm. high, usually much branched

and forming a dense bush, with slender fusiform tuberous roots up to 1 • 25 m. long.

Stems stout and woody at the base, covered, like the young parts, petioles, lower surface

of leaves, peduncles, bracteoles, pedicels and calyces with a very short silvery adpressed

pubescence (which is very rarely strigose consisting of somewhat longer, not silvery

hairs), only glabrescent in the very old parts, subterete to somewhat angular, solid;

ultimate twigs slender, often angular, sometimes virgate, sometimes sinuous. Leaves

entire, usually linear to oblanceolate or lanceolate, 3-6 (-11) cm. long and 2-8 (-30)

mm. wide; the base usually decurrent into the short (0-6 mm., occasionally up to

18 mm. long) petiole; the apex acute or obtuse, mucronate. Peduncles 1-flowered,

usually very short, under 15 mm. long, but occasionally up to 7 cm., rather slender,

subterete or angular; bracteoles small, narrow, rather stiff-, pedicels always short,

under 1 cm. long. Sepals subequal, or the outer ones a little shorter, lanceolate, acute

or subacute, 8-16 (-25) mm. long. Corolla bright magenta or magenta-pink, funnel-

shaped, 3-5-7 cm. long and as much or more in diam., midpetaline areas densely

covered with short adpressed silvery hairs, rarely very sparsely so. Capsule ovoid or

ovoid-globose, abruptly apiculate and the acumen again crowned by the persistent

style-base, dark-brown, glabrous, 15-20 mm. long and 10-15 mm. in diam., 1-3-seeded,

pericarp leathery. Seeds dark brown to black, glabrous, very minutely areolate to

almost smooth, about 9 mm. long.

Recorded from the following districts: South West Africa (Auas Mts., Lichtenstein,

Kuisib Riv., Haris); Cape (Kenhardt, Phillipstown, De Aar, Kimberley, Graaff-Reinet,

Somerset East, Cradock, Middelburg, Colesberg, Albert, Queenstown, Cathcart,

Albany, Stockenstrom, Ft. Beaufort, Victoria East); Orange Free State (Fauresmith,

Bloemfontein, Senekal, Heilbron, Vredefort, Kroonstad); Natal (Klip Rivier, Estcourt,

Weenen); Transvaal (Bloemhof, Wolmaransstad, Klerksdorp, Potchefstroom, Ver-

eeniging, Heidelberg, Lichtenburg). %

The type of I. barretii Rendle was examined by Mr. de Winter who reported that

it is very similar to specimens named Ipomoea argyreoides Choisy. There is a photograph

of the type in PRE and I am of the opinion that it is indeed only a depauperate form

of T urbina oenotheroides. The reduction of I. barrettii to the latter species had already

been done by Hallier in Meded. Rijksherbarium Leiden no. 1 (1910), p. 25.

2. Turbina suffruticosa {Burch.) A. Meeuse, comb. nov. Ipomoea suffruticosa Burch.,

Trav. S. Afr. 2 (1824), p. 226; Choisy in DC., Prodr. 9 (1845), p. 357, exclus.

syn.; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 51. I. contorta Engl, in

Engl. Bot. Jb. 10 (1888), p. 244, non Choisy (1845). Rivea suffruticosa (Burch.)

Hall. f. in Engl. Bot. Jb. 18 (1893), p. 156. Ipomoea sublucens Rendle in J. Bot.

39 (1901), p. 17; Baker & Wright, op. cit., p. 57.

Type: Burchell 1838 in herb. Kew.

Ill

Perennial with thick fusiform taproot. Stems several, suffruticose, prostrate to

suberect, slender to stout, up to about 1 m. long; clothed with a fine silvery subtomentose

pubescence which is also present on all young parts, petioles, peduncles, pedicels and

calyces but is often sparser on the leaves. Leaves elliptic-oblong, ovate or lanceolate-

oblong, rather acute with usually recurved mucronate apex to obtuse or slightly retuse,

more or less densely hairy beneath, less densely to glabrescent and somewhat verrucose

above, rounded to truncate or subcordate, rarely cuneate, at the base, 2-5 (-7) cm.

long and 0-4-3 (-5) cm. wide; petioles from about 4 mm. up to 1 5 mm. long. Peduncles

1-5 cm. long, 1- or rarely 2-flowered, more or less distinctly widened and articulated

against the short, 2-10 mm. long pedicel; bracteoles lanceolate, oblong-lanceolate or

more or less oblanceolate, densely hairy on the outer surface, much less hairy on inner

surface, more or less distinctly ciliate, usually acute and usually distinctly narrowed

(subpetioled) at the base, 8-15 mm. long, 1-3 mm. wide. Sepals subequal, lanceolate

or oblong, acuminate, acute, 14-17 mm. long, 4-8 mm. wide; the inner ones sometimes

slightly shorter and less acute. Corolla funnel-shaped, described as rosy-purple

(probably a bright magenta), 4-5 cm. long and as much or a little more in diam. ;

midpetaline areas with rather short adpressed hairs. Fruit globose or ovoid-ellipsoid,

dark brown, glabrous, about 10 mm. long and 7-10 mm. in diam., abruptly apiculate

and crowned by the persistent stylebase, (always?) 1 -seeded. Seed minutely puberulous.

S. W. Africa. — Windhoek, N. outcrops of Auas Mts. : Dinter 1873 (SAM);

20 m. SSE of Windhoek: de Winter 2561 (PRE); Lichtenstein: Dinter 4307 (SAM).

Cape Province. — Kuruman: Le Grange 15 (PRE). Hay, Postmasburg, Wildeals-

put: Aucamp h. no. 496 (KMG, also BOL, PRE); Hay: Power h. no. 7403 (KMG,

also BOL, PRE). Vryburg, Armoedsvlakte : Sharpe h. no. 7389 (PRE). Barkly-W. :

Wilman A59 (GRA); Boetsap: Wilman s.n. (BOL); Marloth 978 (PRE, GRA); Koop-

mansfontein: Acocks 17855 (PRE).

Bechuanaland : Bryant s.n. (BOL).

This species sometimes resembles T. oenotheroides very much, because in some

specimens the leaves are narrowed to the base (e.g. in the specimens Power s.n. and

Aucamp s.n.). There can be no doubt, however, that the two species are quite distinct,

because the leaves are always distinctly petioled in T. suffruticosa, and, in addition,

the stems are usually prostrate and unbranched (erect, much branched in I. oeno-

theroides).

I have not seen the type, but the specimen Marloth 978, referred to this species

by Hallier and in FI. Cap., was taken to be representative.

3. Turbina robertsiana ( Rendle ) A. Meeuse, comb. nov. Ipomoea robertsiana Rendle

in Jl. Bot. 39 (1901), p. 18; Baker & Wright in Dyer, FI. Cap. 4, 2 (1904), p. 50.

I. uncinata Hutch., Botan. in South Afr. (1946), p. 337.

Type: Greenstock s.n. from Pilgrim’s Rest, Transvaal, in BM.

Perennial, forming several to many prostrate, suffruticose stems from a more or

less cylindrical or fusiform, 1-5 cm. thick rootstock. Stems either simple or with

many short axillary branches, up to 120 cm. long, terete or conspicuously angular,

solid, sparsely covered with rather stiff, white, more or less spreading hairs (the same

pubescence is found on the petioles, lower surface and margins of the leaves, peduncles,

bracts, pedicels and sepals). Leaves linear-lanceolate to ovate-lanceolate or ovate-

oblong, 10-35 mm. long and 3-18 mm. wide, often with crisped margin and folded

on the’ midrib, frequently with recurved, mucronate apex, narrowed to rounded at

the base, quite glabrous above; petioles 1-8 mm., rarely up to 13 mm. long. Peduncles

1 -flowered, slender, 5-30 (-40) mm. long, terete; bracts narrowly linear to almost

1332796-11

778

setaceous, 4-20 mm. long; pedicels very short to at most 3 mm. long. Sepals ovate-

lanceolate to linear-lanceolate, acute to acuminate to aristate, 13-21 mm. long, subequal.

Corolla funnel-shaped, magenta, 5-7 cm. long, usually sparsely hairy on the midpetaline

areas, more rarely quite glabrous. Fruit laxly enclosed by the glabrescent, subcoriaceous,

hardly accrescent sepals, subglobose, indehiscent, finely longitudinally striate, abruptly

apiculate and often the apiculus again apiculate by the persistent style-base, glabrous,

about 10 mm. long without the apiculus and about 10 mm. in diam., 1- or 2-celled and

1 - or 2-seeded ; pericarp leathery. Seed about 6 mm. long, densely and shortly velutinous

with greyish-brown hairs.

Transvaal. — Waterberg, Naboomspruit: Galpin M230 (PRE); nr. Naboom-

spruit: Meeuse 9448 (PRE); nr. Hermanusdoorns : Codd 8502 (PRE). Potgietersrust,

Potgietersrust : Leendertz h. no. 6611 (PRE); Eliovson h. no. 27002 (J. PRE); S.E.

of Moorddrift : Meeuse 9463 (PRE). Petersburg, near Pietersburg : Bolus 10924 (BOL) ;

Hutchinson 2285 (BOL, isotype of Ipomoea uncinata Hutch.); Smuts s.n. (PRE); Bruce

& Kies 46 (PRE); v.d. Merwe 2283 (PRE); Meeuse 9154 (PRE); Codd 7932 (PRE);

between Bandolierskop and Zoekmakaar : Meeuse 9154 a (PRE). Lydenburg, Sekukuni-

land: Barnard 121 (PRE). Pilgrimsrest : Greenstock s.n. (fide Rendle and FI. Cap.).

Barberton: Rogers 14032 (GRA, PRE).

Mr. de Winter compared several specimens with the type of Ipomoea robertsiana

Rendle (viz. Galpin M 230, v.d. Merwe 2283) and reported that they agree very well,

except that the leaves are wider than in the type. Leendertz H. no. 6611 agrees with

Galpin M 230, but has narrower leaves and must, therefore, be very similar to the

type. However, v.d. Merwe 2283, which has wider leaves, is not essentially different

from the isotype of I. uncinata ( Hutchinson 2285). All the specimens cited above form

a continuous series, from narrow-leaved ones to broader-leaved ones.

There are some variations in the characters of T. robertsiana, so that the descriptions

given by Rendle and by Hutchinson do not fit the more extreme specimens. For

example, the corolla was described by both authors as glabrous, whereas actually the

corolla is almost invariably thinly hairy on the midpetaline areas, especially when

still in bud. Even the flowers of Hutchinson’s own type are hairy on the midpetaline

areas in bud. The stems are not always unbranched as indicated by Hutchinson and

not always terete as stated by Rendle, etc.

T. robertsiana is very similar in appearance to some forms of Turbina oblongata.

T. robertsiana can be distinguished by a number of minor characters, such as white

pubescence (often yellowish in T. oblongata), glabrous upper leaf surface, smaller,

more herbaceous, often subuncinate and crisped leaves, more slender peduncles and

velutinous seeds (glabrous in T. oblongata). It is also very similar to Ipomoea crispa,

but the latter is hairy on the upper surface of the leaves and is, in addition, geographically,

separated.

4. Turbina oblongata ( E . Mey. ex Choisy) A. Meeuse, comb. nov. Ipomoea oblongata

E. Mey. ex Choisy in DC., Prodr. 9 (1845), p. 368; Hall. f. in Engl. Bot. Jb. 18

(1893), p. 127; Rendle in J. Bot. 39 (1901), p. 16 (var. hirsuta Rendle); Baker

& Wright in Dyer, FI. Cap. 4, 2 (1904), p. 57. I. oblongata E. Mey. ex Drege,

Zw. Pfl. geog. Doc. (1843), p. 46, 142, nomen tantum. I. lambtoniana Rendle,

op. cit., p. 16; Baker & Rendle, op. cit., p. 61. I. randii Rendle op. cit., p. 18;

Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 146.

Type: Drege, “in graminosis inter Schalumna et Buffelsrivier alt. 1500 ped.”

(Choisy), i.e., King William’s Town or East London distr. (in herb. Geneva). I have

not seen the type or any isotypes, but I have studied several specimens referred to

/. oblongata by Hallier and in FI. Cap. In addition, there is only one species of Ipomoea

779

or Turbina found near the type locality which can be identified with this species, other,

similar, ones occurring in that area (such as Ipomoea crassipes and I. crispa) being quite

distinct.

Perennial, forming several to many annual prostrate or, when still young, suberect

stems from a large, up to 1 m. long and 15 cm. thick fusiform tuberous root. Stems

often suffruticose at the base, often firm, stout, angular and somewhat zig-zag,

occasionally thinner, wiry and/or terete, usually, like the petioles, leaves, peduncles

and calyces, thinly or occasionally densely pubescent with stiff, usually yellowish or

brownish hairs, very rarely glabrous, 0-6-2 m. long. Leaves very variable in size and

shape, usually oblong or elliptic, varying to ovate or linear, entire, with usually rounded,

truncate or subcordate, occasionally broadly cuneate or cordate, obtuse or mucronate,

sometimes emarginate, acute or broadly rounded apex, and usually ciliate to sometimes

densely ciliate rimmed margin, 2-10 (-15) cm. long and (0 - 4-) 1-5-5 (-8) cm. wide;

upper surface thinly covered with strigose usually yellowish hairs, lower surface as

thinly or more densely so, occasionally almost sericeo-tomentose, rarely leaves quite

glabrous but, if so, stems, peduncles and calyces also quite glabrous; petioles usually

much shorter than the leaves but occasionally about as long. Peduncles usually

1 -flowered, sometimes 2-flowered, rarely 3- or 4-flowered, terete, usually shorter than

the leaves; bracteoles very variable but usually lanceolate, acute, hairy like the calyx,

a little shorter than the sepals, sometimes broadly oblong; pedicels usually very short

(and bracteoles more or less contiguous to the calyx), sometimes longer but rarely

exceeding 6 mm. Sepals generally lanceolate to ovate-lanceolate, subequal (inner

ones slightly wider) but more or less unequal in specimens with broader, oblong or

ovate sepals in which the inner ones are narrower; usually acute to acuminate with

very acute tips, rarely subobtuse; outer ones more or less densely covered with usually

stiff yellowish hairs but in very hairy plants almost sericeo-tomentose and in thinly

hairy plants subglabrous, inner ones less hairy and often with membranous edges

(12-) 16-22 (-25) mm. long. Corolla magenta, funnel-shaped, 3-5-7 cm. long and

about as much in diam., midpetaline areas usually thinly covered with silky adpressed

hairs, sometimes nearly, but very rarely quite glabrous, occasionally densely silky.

Fruits only rarely produced, subglobose, glabrous, dark brown, abruptly apiculate

and the apiculus again crowned by the persistent style-base, loosely enclosed by the

somewhat spreading, slightly accrescent (in length and width) and coriaceous sepals,

12-15 mm. in diam., 1-4-seeded; pericarp leathery. Seeds glabrous, grey, finely

punctate to smooth, about 7 mm. long.

From the eastern Cape to the Orange Free State and Transvaal to South West

Africa, Bechuanaland, Southern Rhodesia, also in north-west Natal and Portuguese

East Africa.

Recorded from: Eastern Cape Province (Alexandria, Albany, King William’s

Town, East London, Stutterheim, Cathcart, Queenstown, Kentani, Umtata, Mquanduli,

McLear, Mount Currie, Herschel, Aliwal North, Albert); Gricpialand-West (Kimberley,

Vryburg, Mafeking); Orange Free State (Rouxville, Bloemfontein, Winburg, Senekal,

Bethlehem, Kroonstad, Heilbron); Transvaal (practically all districts); Swaziland

(Mbabane); Bechuanaland (Kanye); South West Africa (Okahandja, Otjiwarongo,

Otavi, Grootfontein); Natal (Utrecht, Pietermaritzburg); Portuguese East Africa

(Sul do Save).

Some interesting specimens are the following: Ecklon & Zeyher 8-12 from

Klipplaatrivier near Shiloh (Queenstown), quoted by Hallier (GRA, SAM); Shiloh:

Baur 853 (SAM) = prob. Baur s.n. (in GRA, PRE), quoted in FI. Cap.; Zeyher 1208

(BOL) and Burke 179 (SAM), both from Magaliesberg (Transvaal), isotypes of the

var. hirsuta Rendle; Wilms 2152 (PRE) from Pietermaritzburg quoted under var.

hirsuta in FI. Cap.; Wood 3466 (NH), from Little Tugela River, quoted in FI. Cap.;

Rand 271 (GRA) from Bulawayo (isotype of 7. randii Rendle).

780

Interesting extreme forms are, among other ones, the following: With subcordate

to distinctly cordate leaves: Corby 555 (SRGH), from Marandellas, S. Rhodesia

(this specimen has also leafy bracts and few-flowered inflorescences), v.d. Merwe s.n.

from Carolina (Tvl.) in PRE (with leafy bracts and dense inflorences); with leafy

bracts: Galpin 14390 from Pilgrim’s Rest (PRE). Subglabrous to thinly hairy forms:

Barberton: Galpin 731 (PRE, BOL, GRA), Codd 7794 (PRE), Coningham 28 (PRE);

Nelspruit: Codd & de Winter 5109 (PRE), v.d. Schijff 2406 (PRE); Witbank: Repton

887 (PRE); Belfast: Smuts & Gillett 2211 (PRE). With narrow, linear leaves:

Galpin 731, Repton 887, Smuts & Gillett 2211. With softer, strigose-subtomentose

pubescence: Bradfield 220; Liebenberg 2497 (PRE), from Belfast; Galpin 14390.

After having seen many specimens, I have come to the conclusion that this species

has a very wide range and is extremely variable. It has, accordingly, been described

several times and the extreme forms are indeed rather different in pubescence, leaf-shape,

shape of petals, length of petioles and peduncles, and in the number of flowers per

peduncle. However, all these various forms are united by intermediate specimens.

They cannot even be grouped into distinct varieties.

Ipomoea randii, of which I saw a photograph of the type (PRE) and an isotype

(in GRA) and the type of which was compared with several specimens by Mr. de Winter,

is indistinguishable from specimens referable to the “ var. hirsuta ” of FI. Cap.

On account of the considerable variation, forms of T. oblongata resemble several

other species: those with long petioles and long peduncles resemble I. pellita (which

has more flowers per peduncle and usually longer and strictly linear, very bristly bracts

and sepals); those with several-flowered inflorescences show some resemblance to

I. atherstonei; small-leaved forms are very much like I. crispa and Turbina robertsiana

in appearance, but can always be distinguished by the characters given in the key to

the species of Ipomoea and Turbina (see under Ipomoea).

T. oblongata, though flowering freely, rarely seems to produce ripe capsules.

In the Transvaal area most flowers are eaten or damaged by beetles and the genitalia

are eaten away before the fruit has had time to set. A careful search in a large area

near Pretoria resulted in one mature capsule on hundreds of plants.

5. Turbina holubii {Baker) A. Meeuse, comb. nov. Ipomoea holubii Baker in Kew

Bull. 1894, p. 72; Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 188.

I. rhodesiana Rendle in Jl. Bot. 39 (1901), p. 57; Baker & Rendle, op. cit., p. 188.

Rivea (sect. Poliothamnus) holubii (Baker) Hall. f. in Meded. Rijksherb. Leiden I

(1910), p. 25. Ipomoea awasmontana Dinter m.s. on Dinter 4307 in BM.

Type: Not designated, because Baker mentions two specimens, Holub 572 and

969 from S. Rhodesia (herb. Kew).

Shrub, many-stemmed from the base, and much branched up to 2-50 m. high.

Stems covered with a light grey bark, slender, sinuous, erect, decumbent or climbing

at the tips. Innovations sericeo-tomentose; young stems more or less densely greyish-

pubescent, sometimes reddish. Leaves deciduous, ovate-cordate to orbicular-cordate,

sometimes oblong-cordate, entire, often folded along the midrib, usually with gradually

deflexed, obtuse or subacute, rarely acuminate, often mucronate apex, usually widely

and shallowly cordate, sometimes obtuse, truncate, or rounded, rarely subacute at

the base, more or less sericeous to tomentose on both surfaces, more densely so and

more silvery so beneath, sometimes fulvo-sericeous above and on the prominent,

conspicuous curving lateral veins beneath, very rarely almost glabrous ( Chase 730),

1-6 cm. long and 0-75-4 cm. wide, rarely up to 8 by 5, very rarely up to 12 by 7 cm.;

781

petioles rather slender, ascending, pubescent like the stem, 4-15 (-25) mm. long.

Inflorescences 1-3 (-5)-flowered, axillary, sometimes forming a sort of leafy panicle

at the ends of the branches; peduncles slender, sometimes nearly obsolete, only a few

mm. long, but usually 2-5-6 (-9) cm. long, hairy like stems and petioles as are the

elliptic or spathulate to linear-oblong, herbaceous, rather small (2-5-3 mm.) to foliaceous

(up to 15 by 6 mm. long) bracteoles and the 8-30 mm. long pedicels; bracteoles often

numerous and forming a sort of involucre at the base of the subumbellate cyme and

in this case one of them much larger, occasionally resembling a leaf and up to 4 by 2

cm. Calyx (6-) 8-12 mm. long, usually greyish-pubescent to tomentose (at least the

outer sepals), sometimes only hairy at the base and sepals conspicuously ciliate; sepals

firmly herbaceous, unequal, broadly elliptic to obovate or orbicular or obovate-

spathulate, much imbricate, obtuse and mucronate or apiculate, the two inner ones

conspicuously larger than the outer ones, in fruit all accrescent, becoming subcoriaceous-

chartaceous, glabrescent, brown, up to about 16 mm. long, ultimately spreading to

almost reflexed. Corolla funnel-shaped, pale mauve or pinkish with magenta centre,

4—5 cm. long with a horizontally spreading limb up to about 6 cm. in diam. ; midpetaline

areas thinly strigose with rather long adpressed hairs outside. Stamens very unequal.

Ovary glabrous. Fruit with thinly leathery pericarp, ellipsoid, apiculate, 8-10 mm.

long, and 3-5 mm. in diam., usually 1-celled and with a single seed. Seed ellipsoid,

pale yellowish brown or light brown, glabrous, very finely areolate, about 7 mm.

long and 3 mm. in diam.

S. W. Africa, (Bechuanaland?), S. Rhodesia, N. Rhodesia; also recorded from the

Waterberg and Pietersburg districts of the Transvaal.

Transvaal. — Waterberg, nr. Oslo: Codd 4002 (PRE); nr. Ellisras: Codd 8492

(PRE); 15 miles N.W. of Hermanusdoorns: Meeuse 9654 (PRE). Pietersburg, near

Ganspoort: Codd & Dyer 7745 (PRE, L).

S. W. Africa. — Otavi: Dinter 917 (SAM). Guchab: Dinter 1608 (PRE, SAM).

Rehoboth- Aub : Dinter 2246 (SAM), near Windhoek: de Winter 2352 (PRE). Awas

Mts. : Boss s.n. (PRE). Tigerschlucht: Boss s.n. (PRE). Grootfontein: Schoenfelder

S56 (PRE). Nosib: Schoenfelder 965 (PRE).

S. Rhodesia. — Bulalima-Mangwe, Greystones: Feiertag s.n. (SRGH, PRE).

Bulawayo: Rand 141 (photograph of isotypes of Ipomoea rhodesiana Rendle in PRE ex

BM); Bulawayo and Matopos: Rogers 5647 (PRE, NH); Kolbe 4078 (BOL). Sinoia:

Hopkins B 1473 (SRGH). Penhalonga: Chase 730 (SRGH). Gwaai: Davies 251

(SRGH), Allen 244 (SRGH, PRE). Fort Victoria: Rodin 4254 (PRE, SRGH). Salis-

bury: Eyles 3485 (BOL), 8940 (SRGH). Filabusi, Patrick’s Dam: Davies D 20‘

(SRGH).' Miami, K34 Expt. Farm: Wild 1804 (SRGH). Hillside: Martineau 866

(SRGH). Marandellas, Skipton: Collins 133 (SRGH, PRE). Umtali: Hopkins H. no.

7450 (SRGH). Mazoe, Umvukwe: Wild 3910 (SRGH, PRE). Victoria Falls, S. bank:

Rogers 13034 (PRE, GRA); without precise locality Brain no. 78914 (SRGH).

N. Rhodesia. — Livingstone: Grant 4504 (PRE, BOL).

There is considerable variation in the shape of the leaves (from distinctly cordate

to sub-cordate or rounded, broadly ovate to orbicular, sometimes cordate-oblong),

in the inflorescences (flowers solitary or in few-flowered cymes; peduncles short or

long, pedicels 8-30 mm. long, bracts small, linear-oblong and about 3 mm. long or

foliaceous, large, up to 30 mm. long and 20 mm. wide) and the pubescence (the plants

can be densely silky, sometimes shortly tomentose, or much more glabrous and the same

applies to the sepals). The variation in the leaf shape, inflorescence and bracts is

often present in a single specimen, so that I do not hesitate to include /. rhodesiana

Rendle.

782

Hallier, although he did not know the fruit or seed of this species, already noticed

the relationships between I. holubii, I, pyramidalis and I. suffruticosa and placed them

all in his section Poliothamnus of Rivea. The fruiting specimens I saw confirmed the

relationships and that is why I also include I. pyramidalis in Turbina.

Some specimens 1 have seen were compared at Kew and apart from variation in

the pubescence agree with the types of I. holubii and I. rhodesiana. After having seen

many specimens 1 came to the conclusion that the pubescence is not a constant

character and, therefore, include all the above-cited specimens.

6. Turbina shirensis ( Oliv .) A. Meeuse, comb. nov. Ipomoea shirensis Oliv. in Hook.,

Ic. Plant., Ser. Ill, 5 (1884), p. 58, t. 1474; Baker & Rendle in Dyer, FI. Trop.

Afr. 4, 2 (1905), p. 189. Rivea (Sect. Legendrea) shirensis (sphalm. “ schirensis ”)

(Oliv.) Hall. f. in Engl. Bot. Jb. 18 (1893), p. 157 and Jb. Hamburg. Wissensch.

Anst. 16, Beih. 3 (1898), p. 14. Porana subrotundifolia de Wild., Et. FI. Katanga I

(1902-1903), p. Ill, t. 5, figs. 8-18.

Type: Not designated, because Oliver mentions two specimens: Buchanan 262

and Kirk s.n. from Nyasaland (herb. Kew), but as Oliver used the name “ shirensis ”

and the Buchanan specimen came from the Shire highlands, the specimen Buchanan 262

must be taken as the lecto-type.

“ Tall, robust climbing shrub, covering bushes and small trees ” (v.d. Schijff,

Faulkner). Sterns woody, terete, covered with a short lax whitish tomentum of very

short adpressed hairs, more densely so when young. Leaves broadly cordate to orbicular-

ovate or cordate-ovate, entire, obtuse to acuminate or shortly and abruptly cuspidate

at the apex, often with terminal mucro, 3-14 cm long and wide; the basal sinus wide and

shallow to leaf-base almost truncate, upper surface green-drying brown, thinly pubescent

with short hairs above, more densely so when young, densely white or greyish shortly-

tomentose or silky-tomentose beneath; petioles rather slender, 1 -5-8 cm. long, densely

tomentose. Inflorescences cymose, axillary in the axils of the upper leaves and forming

a lax terminal panicle, peduncles patent or erecto-patent, rather slender, densely

tomentose like the young stems and petioles, few- to many-flowered, 3-9 cm. long;

bracteoles thinly papery or almost membranous, oblong-oblanceolate, oblong-obovate

or somewhat narrowly oblong-spathulate, much narrowed and subpetioled at the

base, 7-12 mm. long, pale yellowish brown when dry, hairy outside, glabrous inside,

very early deciduous and rarely preserved; cyme-branches short, usually under 2 cm.

long, pedicels usually under 1 cm. long, both densely tomentose. Sepals thin, almost

papery, subequal, oblong or obovate-oblong, obtuse, densely sericeo-tomentose outside,

9-1 1 mm. long, at first erect, much imbricate, later accrescent, glabrescent and much

spreading, ultimately papery, brittle, brown and often purplish outside, pale-straw-

coloured inside, 13-15 mm. long. Corolla white (not lilac as stated in FI. Trop. Afr.),

widely funnel-shaped, 15-20 mm. long; midpetaline areas sharply defined, densely

silky. Capsule broadly ellipsoid or somewhat obovoid, rounded-truncate to depressed

at the apex, very densely greyish sericeo-tomentose, 55-71 mm. long and 5-6-5 mm.

in diam., one-seeded; pericarp hard, woody. Seed light brown or fawnish, subglobose-

4-angled, about 4 mm. long and 3-5-4 mm. in diam., very shortly velutinuous or

puberulous.

N. Transvaal, S. Rhodesia, Portuguese E. Africa, Nyasaland, Belgian Congo,

N. Rhodesia.

Transvaal. — Sebasa, Kruger National Park: Codd 5978, v.d. Schijff 586, 647

(PRE), v.d. Schijff & Marais 3723 (PRE).

Portuguese East Africa.— Near Transvaal border near confluence of Limpopo,

Nuanetsi and Pafuri rivers: Smuts 2396, s.n. (PRE). Quelimane, Mocuba, Lugela:

Faulkner 96 (PRE); “ Kew ” 263 (PRE, SRGH). Niassa, Lalaua: Torre 1443 (COI).

783

S. Rhodesia.— Danga, Sabi-Lundi junction: Chase 2296 (SRGH). Lundi River:

Pole Evans 4825 (PRE). Umtali: Eyles 8425; Chase 1709 (SRGH). Umvumvumwe

Riv.: Chase 324 (SRGH). Belingwa: Harvie 6/51 (SRGH, PRE). Gwanda, Doddie-

burn Ranch: Davison s.n. (PRE).

Nyasaland. — Chikwakwa: Gerstner 7066 (PRE).

N. Rhodesia. — Without precise locality: N.N. in Govt. Herb. S. Rh. no. 3314

(PRE).

Some specimens ( Codd 5978, Faulkner 96) were compared with the type at Kew,

so that the identity of the cited specimens is definitely established.

7. Turbina stenosiphon {Hall, f.) A. Meeuse, comb. nov. Ipomoea stenosiphon Hall. f.

in Sitz. ber. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 107 (1898), Abt. I, p. 50;

Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 192. Rivea stenosiphon

(Hall, f.) Hall. f. in Jl. Hamb. Wiss. Anstalt. 16 (1898), Beih. 3, p. 15.

Type: Pospischil s.n. from Taweta, foot of Kilimanjaro, Tanganyika in W

(not seen).

A tall climbing or rambling shrub, up to 15 m. long (Gerstner). Stems terete,

glabrous, some often sterile, slender, whip-like (as are the young sprouts); older parts

of stems woody with a yellowish-grey separable bark; flowering branches usually

more robust, straight and erect, densely leafy especially towards the apex and frequently

with rugose bark. Leaves cordate, cordate-orbicular or cordate-ovate, emarginate,

acute or more or less acuminate, mucronate, with a distinct and rather wide basal

sinus, glabrous or more or less pubescent, mainly on the nerves beneath, entire or

sometimes more or less subrepand, 3-11 cm. long and 2 -5-7 -5 cm. wide; lateral nerves

8-11 on either side of the flattened and (underneath) narrowly winged midrib; reticulate

nervation conspicuous beneath; lower surface, at least of the majority of the leaves

of each specimen, dotted with numerous black, small circular glands; petiole slender,

2-9 • 5 cm. long, nearly as long as the corresponding leaf, sometimes glandular on each

side a little below the blade. Peduncles usually on very short lateral branchlets up

to about 5 mm. long, very short or obsolete, 0-3 mm. long, articulated against the

solitary or 2-5-fascicled pedicels, the latter terete, very slender, sulcate when dry,

thickened at the base and widening into the calyx, glabrous, 2-5 cm. long; bracts very

early deciduous, more or less membranous, yellowish-brown, oblong-lanceolate, 2-3

mm. long, glabrous, acute or subacute. Calyx subcylindric, 14—17 mm. long, glabrous,

thinly coriaceous; sepals elliptic, oblong or more or less subspathulate-oblong, obtuse

or emarginate, sometimes subtruncate, much imbricate, with membranous subpellucid

margins; the outer ones distinctly and sometimes much shorter, all in fruit becoming

much enlarged, more or less brittle and finely and distinctly nerved, the outer ones

becoming about 2 cm. long and about 1 -5 cm. wide, the inner ones up to about 4 cm.,

long and about 2 cm. wide, all laxly enveloping the fruit to somewhat spreading.

Corolla white, cream or pale yellow to greenish-yellow, hypocrateriform; the tube

very slender (8-) 10-12 (-15) cm. long and 2-4 mm. in diam., subcylindrical or very

slightly widening towards the limb; the latter more or less spreading, cm. across

when fully expanded, about 2 cm. long when flattened or collapsed, with woolly

hairs near the apices of the midpetaline areas outside. Stamens and style long

vexerted. Ovary glabrous, 4-celled. Capsule narrowly conical, dirty brown, 2-2-5

cm. long and 0-1-5 mm. in diam., cuspidate by the thick, awn-shaped, persistent,

12-15 mm. long style-base; pericarp thin, not splitting into valves, 1 (-3)-celled by

abortion; septa, if present, thinly membranous, not separating from the pericarp.

Seed{s) oblong, 11-15 mm. long and 5-6 mm. wide, brownish, velvety, with longer

ochraceous hairs around the hilum.

784

E. Africa, S. Rhodesia, N. Transvaal, S. W. Africa(?), Belgian Congo.

Transvaal. — Zoutpansberg, Louis Trichardt: Gerstner 5919 (PRE); Msekwa’s

Poort: Gerstner 5897 (PRE); Wylie’s Poort: Gerstner 5776 (PRE).

S. Rhodesia. — Matopos: Eyles 58 (SRGH); Rogers 6900 (J). Fort Usher: Hopkins

s.n. H. no. 9903 (SRGH). Victoria: Monro 1037 (SRGH). Umtali: Chase 1373,

4286 (SRGH).

Nyasaland. — 20 miles S. of Lilongwe: Pole Evans & Erens 612 (PRE).

Tanganyika. — Kyimbila: Stolz 1422, 1426 (L). Usagara, Mlali: Stuhlmann

198 (L). Pangusi: Stuhlmann All (L).

Belgian Congo.— Kasai Province: Quarre 5460 (PRE).

?S. W. Africa. — Kaokoveld, Gauko (Kaoko) Otavi: Hall 448 (NBG).

Cultivated Specimen. — Seed collected near Matopos (leg. Steyt/er), flowered in

National Botanic Gardens, Kirstenbosch, in (BOL).

Apart from the excellent description of this very characteristic species, which is

sufficient for recognition, I saw duplicates of the specimens Stuhlmann 198 and 427,

referred to this species and annotated by Hallier himself (in L).

The leaves in this species can be densely hairy to glabrous.

The specimen P. Quarre 5460 consists of a twig with a few ripe fruits without leaves

or flowers and is possibly referable to a closely related species ( Turbina curtoi?), but

the fruiting calyx, fruit and seed agree in every respect with those of some of the other

specimens, so that it is tentatively referred to T. stenosiphon.

The specimen Hall 448 has very young fruits, no corollas and glabrous, somewhat

coriaceous leaves. Without complete flowers it is impossible to decide if this plant

represents a different species or is only an ecotype with glabrous, rather thick leaves.

The calyx and young fruit are exactly as in T. stenosiphon and that is why the plant is

tentatively referred to this species.

Apart from the above-mentioned species, the following have to be transferred to

Turbina : —

Turbina pyramidalis {Hall. /.) A. Meeuse, comb. nov. Ipomoea pyramidalis Hall. f.

in Engl. Bot. Jb. 18 (1893), p. 152; Hiern, Cat. Welw. Afr. PI. I, 3 (1898), p. 741;

Baker & Rendle in Dyer, FI. Trop. Afr. 4, 2 (1905), p. 193. Ipomoea megalo-

chlamys Baker in Kew Bull. 1894, p. 72. Rivea (Sect. Poliothamnus) pyramidalis

(Hall, f.) Hall. f. in Meded. Rijksherb. Leiden no. 1 (1910), p. 25.

Of this species which, as far as I know, is still only known from the original

Welwitsch gatherings 6113 and 6113b, I saw an isotype ( Welwitsch 6113) in COL

There is no doubt that this species is closely related to T. shirensis and T. holubii, although

the fruits are still unknown.

Turbina curtoi ( Rendle ) A. Meeuse, comb. nov. Ipomoea curtoi Rendle in Jl. Bot. 46

(1908), p. 182.

Of this species, apparently only known from the type gathering, I examined a

photograph of the type (in PRE) and an isotype ( Gossweiler 4275, in COI). As Rendle

pointed out already, this species is very closely related to Turbina shirensis (but it is

distinct in that it has an umbellate inflorescence and glabrous sepals), and it is, therefore,

indubitably a species of Turbina.

785

Index,

Aniseia Choisy

A. bi flora (L.) Choisy

A. calycina (Roxb.) Choisy

A. calystegioides Choisy

A. hackeliana Schinz

Argyreia Choisy

Asterochlaena Corda

Asterochlaena Garcke

Astrochlaena Hall f.

A. floccosa Hall. f.

A. lachnosperma (Choisy) Hall. f.

A. malvacea (Klotzsch) Hall. f.

A. rot undata Pilger

Astripomoea A. Meeuse

A. lachnosperma (Choisy) A. Meeuse

A. malvacea (Klotzsch) A. Meeuse

A. rotundata (Pilger) A. Meeuse

Batatas (Choisy) Griseb. (sect, of Ipomoea)

Batatas edulis (Thunb.) Choisy

Bonamia Thouars

B. capensis (E. Mey. ex Choisy) Burtt-Davy

B. madagascariensis Poir

B. poranoides Hall f.

B. schizantha (Hall, f.) A. Meeuse

B. suffruticosa (Schinz) Burtt-Davy

Bonamya Neck

Breweria R. Br

B. baccharoides Baker

B. capensis (E. Mey. ex Choisy) Baker

B. malvacea Klotzsch

B. sessiliflora Baker

B. suffruticosa Schinz

Calonyction Choisy

Calonyction (Choisy) Griseb. (sect, of Ipomoea)

C. aculeatum (L.) House

C. bona-nox Boj

C. speciosum Choisy

C. trichospermum var. diversifolium

Calycanthemum Klotzsch (sect, of Ipomoea )

Calycanthemum leucanthemum Klotzsch

Calystegia R. Br

C. sepium (L.) R. Br

C. soldanella (L.) R. Br. ex R. et S

Cephalanthae (Choisy) Hall. f. (subsect, of Ipomoea)

Chrorisanthae Hall. f. (subsect, of Ipomoea)

Convolvulus L

C. aculeatus L

C. albivenius Lindl

C. alceifolius Lam

C. althaeoides Thunb

C. angolensis Baker

C. argillicola Pilger

C. arvensis L

C. aschersonii Engl

C. batatas L

C. bicolor Desr

C. bicolor Vahl

C. bidentatus Bernh

C. biflorus L

C. boedeckerianus Peter

C. bullerianus Rendle

C. brasiliensis L

C. burmannii Choisy

PAGE

729, 730

723

]728,’ 729, 730

730

726

712, 772, 773, 774, 775

709

709, 710

709, 711

644, 709

709, 710

709, 711

713

746

643, 662,664

663

664

665

663, 664

664

662, 664

663, 664

663

710

663, 664

711, 712

712, 713

765

761

713,733,748

722

643, 696

697

697, 698

713, 733

713

643, 666, 694

765

768

692

690, 692

683, 684

667, 670

666, 667, 686, 695

668, 677, 678, 679, 680, 684, 695

746

698, 736

‘ ‘ ” 698, 736

.667, 684, 685, 686, 695

• ' ’ ' .623

666, 667, 668, 671, 673, 674, 675, 676, 679

666, 687, 689

754

696

Convolvulus L.

786

PAGE

C. cairicus L

C. calycinus E. Mey. ex Choisy

C. calycinus H.B.K

C. calycinus Roxb

C. capensis Burm. f

C. capensis Burm. f. var. plicatus (Desr.) Baker

C. copticus L

C. cordifolius Thunb

C. corymbosus L

C. crispus Thunb

C. dichrous R. et S

C. dinteri Pilger

C. dregeanus Choisy

C. falckia Thunb

C. falckia Jacq

C. farinosus L

C. filiformis Desr

C. filiformis Thunb

C. galpinii C. H. Wright

C. hardwickii Spreng

C. hastatus Desr

C. hastatus Forsk

C. hastatus Thunb

C. heptaphyllus

C. hirtellus Hall, f

C. hispidus Vahl

C. huillensis (Baker) Rendle

C. inconspicuus Hall. f.

C. kentrocaulos Steud

C. liniformis Rendle

C. longipedunculatus Dinter ms

C. massonii Dietr

C. mauretanicus L

C. mucronatus Engl

C. multifidus Hall. f.

C. multifidus Thunb

C. namaquensis Schltr. ms

C. natalensis Bernh

C. nil L

C. obscurus L

C. ocellatus Hook

C. ocellatus var. ocellatus

C. ocellatus var. ornatus (Engl.) A. Meeuse

C. oenotheroides Linn. f.

C. oenotherae Vatke

C. ornatus Engl

C. paniculatus L

C. pedatus Roxb

C. penicellatus Hall. f.

C. penicellatus A. Rich

C. pes-caprae L

C. petraeus Lee ex Choisy

C. phyllosepalus Hall. f.

C. pilosus Rottl

C. pilosus Roxb

C. plantagineus Choisy

C. plicatus Desr

C. pterygocaulos Steud

C. purpureus L

C. radicans Thunb

C. randii Randle

C. reptans L

C. rhynchophyllus Baker ex Engl

C. sagittatus Thunb. and vars

C. sepium L

C. sinensis Desr

C. soldanella L

761

687, 688, 689

687, 689

687, 689, 728, 729, 730

666, 669, 670, 676, 686, 691, 692-693, 694, 771

690, 691, 692, 693

760

684, 685

773

751

736

667, 673, 684

666, 667, 671, 695

692

666, 692, 693

668, 679, 684, 685

692

666, 692, 693, 695

669, 687, 696

728, 730

685, 686

679, 685, 686

764, 765

682, 684

722

683, 684

666, 691, 693, 694

704, 705

.666, 671

706

696

667

663, 664

673, 691

668, 673, 675, 676, 691

682

.667, 669, 670, 687-689, 691, 694

733

1 1 746

’.'.’.'.’.'.’.’.’.’.’.'.’.’.666, 667, 669, 670, 675, 676, 684

672

672, 673

775, 776

727

666, 673, 675

767

760

677

684, 685

754

696

681, 684

....736

;;;;;;;;;;;;;;;;; 755

.690, 691, 692, 693, 694

....702

..734

772

672

753, 754

.V ’!!!!! !!!"!” 678, 679

. . .667, 668, 669, 673, 677, 678, 679-683, 684, 686

'.".'.'.'.'.'.'.'.'.m

697, 698

787

Convulvulus L.

PAGE

C. steudneri Engl

C. sublobatus Linn, f

C. thompsonii Baker

C. thunbergii Hall, f

C. thunbergii R. et S

C. tiliifolius Desr

C. transvaalensis Schltr

C. trilobus Thunb

C. ulosepalus Hall. f.

C. wightii Wall

Cuscuta L

C. abyssinica

C. africana Willd

C. africana var. capensis Baker

C. alpestris Fourc

C. americana Thunb. non L

C. angulata Engelm

C. appendiculata Engelm

C. appendiculata var. macroflora Yunck

C. australis R. Br

C. balansae Boiss. et Reutt

C. balansae var. mossamedensis (Welw.) Yunck

C. bifurcata Yunck

C. brevistyla A. Braun

C. campestris Yunck

C. capensis Choisy

C. cassytoides Nees ex Engelm

C. cordofana (Engelm.) Yunck

C. cue ul lata Yunck

C. epithymum Murr. ex L

C. epitribulum Schinz

C. europaea L

C. gerrardii Baker

C. hyalina Roth

C. hyalina var. nubiana

C. kilimanjari Oliv

C. madagascarensis Yunck

C. madagascarensis var. schlechteri Yunck

C. medicaginis C. H. Wright

C. natalensis Baker

C. nitida E. Mey. ex Choisy

C. obtusiflora H. B. et K

C. planiflora Ten

C. planiflora var. madagascarensis (Yunck) A. Meeuse

C. planiflora var. mossamedensis Welw. ex Hiern

C. planiflora var. planiflora

C. suaveolens Ser

C. timorensis Decne

C. trifolii Babingt

Dasychaetia Hall. f. (sect, of Ipomoea)

Dichondra J. G. et R. Forst

D. repens J. G. et R. Forst

Evolvulus L

E. alsinoides (L.) L

E. alsinoides var. linifolius Baker

E. alsinoides var. glaber

E. capensis E. Mey. ex Choisy

E. hederaceus Burm. f

E. mummularius L

Falckia see Falkia.

Falkia Linn, f

F. abyssinica Engl

F. dichondroides Baker

F. diffusa (Choisy) Hall. f.

F. oblonga Bernh

F. repens Linn, f

F. villosa Hall, f

683

698

680

675

667, 669, 670, 689, 690, 691, 694

772

666, 687, 689

738 739

667,' 668, 678, 679, 684’ 685

737

641, 642, 643, 644

655

646, 652, 653, 655

652

646, 655

645, 649, 650

649

645, 647, 648

655

645, 648

655

644, 645, 647, 648, 657

652

645, 651

646

652

644, 647, 656

650

644

646, 652

646, 650

646

645, 650

655, 656

656

648

646, 653

646, 654, 655

646, 650

655

646, 656

647, 655

655

644, 646, 648

651

656

713, 733

643, 657

657

643, 661

661

662

663

700

641, 661

643, 658

660

659, 660

659, 660, 661

658, 648, 659, 660, 661

659

788

Hewittia W. et A

H. bicolor (Vahl) Wight

H. sublobata (Linn, f.) O. Ktze

Ipomoea L

I. adenioides Schinz

I. adumbrata Rendle et Britten

/. aitoni Choisy

I. alba L

I. albivenia (Lindl.) Sweet

I. angulata E. Mey. ex Drege

I. angustifolia Jacq

I. angustisecta Engl

I. aquatica Forsk

I. arachnoidea Boj

I. arachnoidea Choisy

I. arachnosperma Welw

I. argyreioides Choisy

/. aspericaulis Baker

I. asperifolia Hall, f.

I. atherstonei Baker

I. auxocalyx Pilger

I. awasmontana Dinter ms

I. baclei Choisy

I. barret tii Randle

I. batatas (L.) Lam

I. bathycolpos Hall, f

I. bellecomans Rendle

/. biflora (L.) Pers

I. biloba Forsk

/. bipinnipartita Engl

J. blepharophylla Hall. f.

I. blepharosepala Hochst. ex A. Rich. . .

I. bolusiana Schinz

I. bona-nox L

I. bowieana (Rendle) Baker

I. cairica (L.) Sweet

/. calcarata N. E. Br

I. calophylla Fenzl

I. calycina Benth. ex C. B. Clarke

I. calycnia Meissn

I. cana E. Mey. ex Drege

/. calystegioides (Choisy) Hall. f.

I. cardiosepala Hochst. ex FI. Cap

I. cardiosepala Meissn

I. carnosa R. Br

I. cecilae N. E. Br

I. chloroneura Hall. f.

I. clappertonii R. Br

I. commatophylla A. Rich

I. conceiroi Rendle

I. congesta R. Br

/. contorta Choisy

/. contorta Engl

I. convolvuloides Hall. f.

I. convolvuloides Schinz

I. coptica (L.) Roth

I. coscinosperma Hochst. ex Choisy

I. crassipes Hook

I. crispa (Thunb.) Hall. f.

/. curtoi Rendle

L. cynanchifolia C. B. Clarke

I. cynanchifolia Meissn

I. dammar ana Rendle

I. dasysperma Jacq

/. demissa Hall. f.

I. dichroa (R. et S.) Hochst. ex Choisy

I. digitata L

PAGE

643, 698

698

698, 736

644, 700, 701, 708, 711, 772, 773, 774

712, 720, 770, 771

730

738

720, 765

719, 768

738

706, 771

758, 759

716, 753, 754

737

738

714-5, 736, 737-9

771, 775-6

756

732-3

718, 732, 741, 780

728

780

766-7

771, 775-6

715, 720, 746

716, 749, 750

730, 731

723

754

705

733

729

715, 720, 756, 757, 758

765

692, 693, 771

720, 761, 765

760

736, 737

728, 729

775

730

729, 730

772

727, 728

714, 739

753

728

769

717, 734, 735, 738

751

776

748, 749

706, 749

720, 760

714, 721, 722

.715, 717, 718, 730, 732, 733, 749, 779

719, 751, 752, 778, 779, 780

771, 774, 784

723, 724, 725

723

769

720, 760, 761, 765

747, 748

736

719, 720, 767

789

Ipomoea L. PAGE

I. dinteri Schulze-Menz 7C-1

I. dissecta Willd '.'.’.'.'.'.‘.'.'.'.'.I 760

/. eenii Rendle 743

I. eriocarpa R. Br . 714,' 721,' 725

I. hcifolia Lindl 733 739

I. fragdis Choisy " ’ ’ ’ ’ 747

/. fragrans Boj . ....................... .166

I. geminiflora Welw 773 774 77 s

I. gerrardi Hook. f. .’.768

I. gerrardiana Rendle 734

I. gracilisepala Rendle 714; 725, 726

I. gracilisepala var. lyciifolia (Merxm.) A. Meeuse .126

I. greenstockii Rendle 730, 731 749

I. hackeliana (Schinz) Hall. f. .’.713, 726

I. hardwickii (Spreng.) Hemsl ’ 728

I. hederacea Jacq 733 734

I. hederacea Auct. non Jacq .’.733

I. hellebarda Schweinf. ex Hall. f. 752, 753

I. heptaphylla (Rottl. et Willd.) Voigt 720, 761, 762, 764’ 765

I. heterophylla R. Br. non Ortega 728

I. hewittioides Hall. f. 731, 732

/. hispida (Vahl.) R. et S .’.721

I. hispida Zucc 721

J. hochstetteri House 720, 762, 764, 765

I. holosericea E. Mey. ex Drege 738

I. holubii Baker 771, 780, 782

/. huillensis Baker 683

/. hystrix Hall. f. 756

I. inamoena Pilger 756, 757

I. involucrata P. Beauv 713, 744, 745

I. kentrocaulos Clarke 704, 705

I. kwebensis N. E. Br 762, 763

I. lachnosperma Choisy 709, 710

I. lambtoniana Rendle 771, 778

I. lapathifolia Hall. f. 720, 752, 753

I. learii Paxt 735

I. leucanthemum (Klotzsch) Hall. f. 714, 722

I. linosepala Hall. f. 726

I. lineariloba Chiov 727

I. lilacina B1 766, 767

I. longipes Garcke 747

I. longipes Engl 747, 748

/. lugardii N. E. Br 742, 743

I. lyciifolia Merxm 726

I. magnusiana Schinz 720, 742-743

I. malxae folia (Rendle) Baker 706, 771

I. marlothii Engl 770

I. maxima (L. f.) Don. ex Sweet 754

1. megalochlamys Baker 684

/. mesenteroides Hall. f. 758, 759

I. natans Dinter et Suesseng 753

I. nil (L.) Roth 717, 733, 734, 736

I. oblongata E. Mey. ex Choisy 771, 774, 778

I. obscura (L.) Ker 716, 746-747, 748

I. odontosepala Baker 760

I. oenotherae (Vatke) Hall. f. 715, 720, 727, 728

I. oenotheroides (Linn, f.) Rafin. ex Hall. f. 771, 774, 775

I. ommaneyi Rendle 718, 732, 740, 741, 742

I. operosa C. H. Wright 745

I. otjikangensis Dinter et Pilger 742, 743

I. ovata E. Mey. ex Rendle 732

I. pachypus Pilger 727, 728

I. palmata Forsk 761

I. paniculata (L.) R. Br 767

I. papilio Hall. f. 716, 717, 750, 751

1. par ker i Choisy 766

790

Ipomoea L.

PAGE

I. pellita Hall. f.

I. pentaphylla (L.) Jacq

I. pes-caprae (L.) Sweet

I. pes-trigridis L

I. petunoides Baker

I. pileata Roxb

/. pilosa Cav

I. pilosa Houtt

I. pilosa (Roxb.) Sweet

/. pinnata Hochst. ex Choisy

I. plantaginea (Choisy) Hall. f.

I. plebeia R. Br

I. plebeia var. africana A. Meeuse

I. polymorpha R. et S

1. praetermissa Rendle

I. pterygocaulos (Steud. ex) Choisy

I. pulchella Roth

I. purpurea (L.) Roth

1. pyramidalis Hall. f.

I. quinquefolia Hochst

I. quinquefolia (L.) Griseb

/. quinquefolia var. pubescens Baker

I. quinquefolia var. purpurea Hall. f.

I. randii Rendle

I. rep tans Poir

I. rhodesiana Rendle

I. riparia G. Don

I. robertsiana Rendle

1. rubro-coerulea Hook

I. saccata Hall. f.

I. sagittaefolia Burm. f.

/. sagittaefolia Hochr

I. sarmentacea Rendle

I. sounder siana Baker

I. seineri Pilger

I. semisect a Merxm

I. shirambensis Baker

I. shirensis Oliv

I. simplex Hook

I. simplex Thunb

I. sinensis (Desr.) Choisy

I. sinensis var. blepharosepala (A. Rich.) Verde.

I. stenosiphon Hall. f.

I. stolonifera (Cyrill.) J. F. Gmel

I. sublucens Rendle

7. suffruticosa Burch

I. tenuirostris Choisy

7. tenuis E. Mey. ex Dr6ge

7. tetraptera Baker

I. transvaalensis A. Meeuse

I. trichosperma B1

I. tricolor Cav

I. triloba L

7. tuberosa A. Rich, non L

7. uncinata Hutch

7. undulata Baker

I. verbascoidea Choisy

7. verecunda (Rendle) N. E. Br

I. verticillata Forsk

I. violacea L

7. vitifolia E. Mey. ex Drege

I. welwitschii Vatke

I. wightii (Wall) Choisy

7. woodii N. E. Br

7. xiphosepala Baker

7. zambesiaca Baker

696, 718, 732, 752, 780

737

714, 754, 755

711, 720, 744

727, 728

713, 745

736, 737

736

736, 737

707

696, 755, 756

714, 722, 723, 725

723

728

758, 759

702

762, 764, 765

717, 733, 734, 735, 736

771, 782, 784

762, 763

703, 764, 771

762

778

753, 754

772, 780, 781

717, 766, 767

772, 774, 777, 778

759

760, 761

753, 754

730, 731

708, 772

769, 770

756, 757

719, 770

771, 774, 782

758, 759

696, 715, 755, 756, 758, 759

689, 714, 723, 729, 730

724

772, 774, 783

772

776

772, 774, 776, 782

725

747

702, 772

716, 718, 748, 749

761

716, 759

739

705

772, 777, 778

751, 752

719, 769, 770

703

724, 725

759

738

.715, 718, 720, 756, 758, 759

714, 737, 738, 739

772, 773

725, 726

752

791

Jacquemontia Choisy

J. capitata (Desr.) G. Don

J. ferruginea (Steud.) Choisy

J. tamnifolia (L.) Griseb

J. thomensis Henriq

Legendrea Webb, et Berth

Leiocalyx Hall. f. (sect, of Ipomoea)...

Merremia Dennst. ex Hall. f.

M. aegyptia (L.) Urb

M. angustifolia (Jacq.) Hall. f.

M. bipinnipartita (Engl.) Hall. f.

M. bowieana Rendle

M. caespitosa (Roxb.) Hall. f.

M. convolvulacea Dennst

M. hederacea (Burm. f.) Hall. f.

M. hirta (L.) Merr

M. kentrocaulos (C. B. Clarke) Rendle ,

M. malvaefolia Rendle

M. palmata Hall. f.

M. pinnata (Hochst. ex Choisy) Hall. f.

M. pterygocaulos (Choisy) Hall. f.

M. quercifolia Hall, f

M. tetraptera (Baker) Hall. f.

M. tridentata (L.) Hall. f. ssp. angustifolia (Jacq.) Ooststr.

M. tuberosa (L.) Rendle

M. umbellata (L.) Hall. f.

M. verecunda Rendle

Metaporana N. E. Br

M. angolensis N. E. Br

M. densiflora (Hall, f.) N. E. Br

Operculina S. Manso

O. kentrocaulos Hall. f.

O. turpethum (L.) S. Manso

Pharbitis Choisy

Pharbitis (Choisy) Griseb. (Sect, of Ipomoea)

P. fragrans Boj. ex Choisy

Polyothamnus (sect, of Rivea sensu Hall, f.)

Porana Burm. f

P. densiflora Hall. f.

P. subrotundifolia De Wild

Quamoclit Moench ,

Rivea Choisy

R. adenioides (Schinz) Hall. f.

R. holubii (Baker) Hall. f.

R. oenotheroides (Linn, f.) Hall. f.

R. pyramidalis (Hall, f.) Hall. f.

R. shirensis (“ schirensis ”) (Oliv.) Hall. f.

R. stenosiphon (Hall, f.) Hall. f.

R. suff ruticosa (Burch.) Hall. f.

Schrebera schinoides L

Seddera Hochst

S. capensis (E. Mey. ex Choisy) Hall. f.

S. mucronata (Engl.) Hall. f.

S. schizantha Hall. f.

S. suffruticosa (Schinz) Hall. f.

S. virgata Hochst. et Steud

S. welwitschii Hall. f.

S. welwitschii var. bakeri Hiern

Shutereia Choisy

S. bicolor (Vahl) Choisy

S. sublobata (Linn, f.) House

Shuteria W. et A

Skinneria Choisy non Forst

Spiranthera Boj. non St. Hil

Streptandra Hall. f. (Sect, of Merremia)

PAGE

643, 699

700

699

700

721

773, 774

712, 713

643, 694, 700, 701, 708, 712

737

706

702, 705

666, 692, 693, 695, 771

754

700

754

701, 704, 705

702, 771

701, 704

701, 707

701. 702, 772

701

702

686, 702, 706, 749, 771

705, 708

736, 737

701, 703, 764, 771

664-5

. .665

643, 705, 708, 712

704-5

701, 708, 772

711-2

713, 733

766

..771, 782

665

782

712

711-2, 773

. .712, 770

780

775

784

782

783

776

652-3

. .643, 662

. .662, 663

663

665-6

. .663, 664

662

663-4

664

698

701

792

Stictocardia Hall. f.

S. tiliifolia (Desr.) Hall. f.

S. woodii (N. E. Br.) Hall, f

Turbina Rafin

T. corymbosa (L. Rafin

T. curtoi (Rendle) A. Meeuse

T. holubii (Baker) A. Meeuse

T. oblongata (E. Mey. ex Choisy) A. Meeuse.

T. oenotheroides (Linn, f.) A. Meeuse

T. pyramidalis (Hall, f.) A. Meeuse

T. robertsiana (Rendle) A. Meeuse

T. shirensis (Oliv.) A. Meeuse

T. stenosiphon (Hall, f.) A. Meeuse

T. suffruticosa (Burch.) A. Meeuse

Volvulopsis Roberty

Volvulus Medik

Xanthips (Criseb.) Hall. f. (Sect, of Merremia)..

PAGE

644, 712-3, 111

Ill

772, 773

644, 712-3, 715, 719, 771, 773, 774-5

773-4

771, 774-5, 784

717-9, 771, 774-5, 780, 782

715, 717-8, 732-3. 741-2, 752, 771, 774-5, 778

696, 771, 774, 775, 776-7

771, 774-5, 784

715, 719, 772, 775, 777, 780

720, 772, 774-5, 782, 784

720, 772, 774-5, 783

772, 774-5, 776

641

696

701

BOTHAEIA , vol • 6,