pigttsee sea!

aang eB

ity

List of Fascicles issued to 22nd November 1930 sence —

Part IV. CoLropTerA.

Fase. 1. Gee By H. E. see a tere figures. Dynes. By A.

Zimmermann. 4 A iees Staphylinide. M. Cameron, MB. 2 text-

gures. Hydrophilide. Sedo. 1 text-6 sure Clavicornia and

Lamellicornia. By G. J. le 13 text-figures. Pp. 1-66. 1927, 4to. 3s.

Fasc. 2._ Heteromera, Bostrychoidea, ee ane Bupreatids By K.G.

Blair, B.Sc, text-figures. ae ae H. van Zwaluwenberg. 10

text-figures. Monae (encase ze). By EA Fleutiaux. Cerambycide. By

Chr. pew late. Brenthide. By R. Kleine. 4 text-figures.

Anthribide. ade ordan, Ph.D. 11 text-figures. Proterhinide. By

RICH: Mitim DS. -R.S. Pp. 67-174. 1928, 4to. 5s.

Fasc. 3._ Throscide. By K. G. Blair, B.Sc. 1 text-figure. Chrysomelide.

By S. Maulik, M.A. 18 text-figures. Pp. 175-215. 1929, 4to. 2s. 6d.

Fasc. 4. Platypodidae and Scolytidae. By C. F. C. Beeson, D.Sc. 13 text-

figures. Pp. 217-248. 1929, 4to. 2s. 6d.

Part V. HyMENOPTERA.

ons 1. Apoidea, Sphecoidea, and ees By R. C. L. Perkins, D.Sc,

and L. ees Cheesman, F.E.S., F.Z.S. 12 text-figures. Larride.

By Francis X. Williams. 12 text-figures. Formicide. By Dr. F. Santschi.

9 text-figures. Pp. 1-58. 1928, 4to. 5s.

Part VI. DurtTera.

Fasc. 1. Streblide and Nycteribiide. By L.Falcoz. 7 text-figures. Hippo-

boscide. By G. F. Ferris. 6 text-figures. Pp. 1-21. 1927, 4to. 2s. 6d.

Fase. 2. altaat By F.W. Edwards, M.A. auberiarne Cecidomyiine.

By arnes, B ‘A. PhD D. 4 text-figures. Pp. 23-108. 1928, 4to. 5s.

’ Fase. Bi eee STabanide: and pee By Gea Ricardo. 6 text-

gures. Larve of Bayes P. A Burton M.A. 2 text-figures.

Dolicho nine By C sib, 8 tk text-figures. Sarcophagide, By

PivA: , MA os reunion Muscide. By J. R. Malloct,.

Pp. 109-175. 1909, 4to. 5s.

fe . aoe ididae ue Pi Bese By J. Collin, 7 text-figures.

Syrphidae. Bye Frank M. Hull 2 text- gures. Footie (Heteroneuridae)

ane Sspromyzdae By R. Malle 6 text-figures. Pp. 177-213. 1929,

to 5. Od.

Fasc. 5. Ortalidae. By J: R. Malloch. 6 text-figures. Calliphoridae. By

J.R. Malloch. Pp. 215-237. 1930, 4to. 2s.

Fasc. 6. Lonchaeidae, Chloropidae and Piophilidae. By J. R. Malloch. 3

text-figures. Pp. 239-251. 1930, 4to. Is.

Part VII. Oruer Orpers oF INsEcTs.

Fasc. 1. Isoptera: Femi a ae By Gerald F. ve 14 text-figures and

| plate. onata. By Lt.-Col. Cl feaeer. I.MS., F.E.S. 5 text-figures.

Pp. 1-44. 1927, 4to. 2s. 6d.

Fasc. 2. Plectoptera. By a J. Ae ean D. (Cantab) ERS, and J. A.

Lestage. 2 text-figures. Siphonaptera. ysano} bie

By Richard. Bagrall, FREE ELS, Gtxctgues, Pp 45-76) 1928

Ane - Bee By J. ates tan DSc. 2 thane Anoplura, By

MA

Buxton, Tric hoptera. y artin E. Mose y. 1 figure,

Newrontera By P. Esben-Petersen. 1 text-figure and 2 ae Apterygota.

By George H. Carpenter,D.Sc. 32text-figures. Pp.77-I1 1928, 4to. 2s. 6d.

Part VIII. TrrrestriAL ARTHROPODA OTHER THAN INSECTS.

Fasc. 1, Isopoda Terrestria. By Harold G. Jackson, DSc. 2 plates. Scor-~

pionoidea. By P. A. Buxton, M.A. Pseudo-scorpiones. By A. Kastner. 11

po ee Acarina. By Stanley Hirst. 2 text-figures. ae 1-27. 1927,

4to 6d.

Fasc, 2. Myricpoden (Myriopoda). By C. Attems. 4 text-figures. Araignées

(Araneida). By Dr. Lucien Berland. 79 text-figures. Pp. 29-78. 1929, 4to. 2s. 6d.

Part IX. Summary AND INDEX.

Fasc. 1. Description of the Environment. By P. A. Buxton, M.R.CS. 2

text-figures and 6 plates. Pp. 1-31. 1930, rl 2s. 6d.

Date Issued.

19th December, 1927.

25th February, 1928.)

23rd February, 1929.

22nd June, 1929.

25th February, 1928:

23rd July, 1927.

23rd June, 1928.

Ith May, 1929.

27th July, 1929.

22nd March, 1930,

22nd November, 1930.

28th May, 1927.

23rd June, 1928.

28th July, 1928.

23rd July, 1927.

22nd June, 1929.

22nd November, 1930.

BRITISH MUSEUM (NATURAL HISTORY)

INSECTS OF SAMOA

AND OTHER SAMOAN TERRESTRIAL

ARTHROPODA

PART IX. SUMMARY |

FASC. 2. Pp. 33-104

By Professor P. A. BUXTON —

LONDON SCHOOD OF HYGIENE AND TROPICAL MEDICINE

LONDON ; :

PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM

SOLD AT

Tor Bertish Mustum (Narugant History), CRomwELL Roap, S.W.7

AND BY

B. Quartron, Lrp.; Doxrau & Co., Lrp.; anp THe Oxrorp UNIversrry Press.

1935

Issued 8th June, 1935.] foes [Price Five Shillings

INSECTS OF SAMOA

AND OTHER SAMOAN TERRESTRIAL |

ARTHROPODA

Although a monograph, or series of papers, dealing comprehensively with

the land arthropod fauna of any group of islands in the South Pacific may be

expected to yield valuable results, in connection with distribution, modification

due to isolation, and other problems, no such work is at present in existence.

In order in some measure to remedy this deficiency, and in view of benefits

directly accruing to the National Collections, the Trustees of the British

Museum have undertaken the publication of an account of the Insects and other

Terrestrial Arthropoda collected in the Samoan Islands, in 1924-1925, by

Professor P. A. Buxton and Mr. G. H. E. Hopkins, during the Expedition of

the London School of Hygiene and Tropical Medicine to the South Pacific.

Advantage has been taken of the opportunity thus afforded, to make the studies

as complete as possible by including in them all Samoan material of the groups

concerned in both the British Museum (Natural History) and (by courtesy of

the authorities of that mstitution) the Bishop Museum, Honolulu.

It is not intended that contributors to the text shall be confined to the

Museum Staff or to any one nation, but, so far as possible, the assistance of the

leading authorities on all groups to be dealt with has been obtained.

The work is divided into nine “ Parts” (see p. 3 of wrapper), of which

each is subdivided into “ Fascicles.”’ Each of the latter, which appear as

ready in any order, consists of one or more contributions. On the completion

of the systematic portion of the work it is intended to issue a general survey

(Part IX), summarising the whole and drawing from it We conclusions as

may be warranted.

A list of Fascicles already issued will be found on pp. 3 and 4 of this wrapper.

N. D. RILEY,

Keeper of Entomology.

British Museum (Natura History),

CromweELt Roap, S.W.7.

UNSsiers OF SAMOA

Part IX, Fasc. 2

SUMMARY

By P. A. Buxton

TABLE OF CONTENTS

I. THE INSECT FAUNA: ITS COMPOSITION

INTRODUCTORY

THe MATERIAL

THE CENSUS

Il. THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS .

BiotocicaL Norrs

General .

Effects of season

Effects of altitude

Effects of wind

Migrant insects

Insects of special Eronmonts

Insects and human agency

Taxonomic Notes

The endemic insects

The absence of certain groups

III. THE INSECT FAUNA: ITS ZOOGEOGRAPHICAL RELATIONS .

HieH anp Low ISLANDS CONTRASTED

RELATIONS wiTH INDO-MALAaya

Examples

Anomalies j

RELATIONS WITH HawallA AND THE MARQUESAS

Hawaiia : ;

Marquesas and ‘‘ Palaeonesia ” . :

RELATIONS WITH AUSTRALIA AND AMERICA

Australia

America : Wegener’s Eothers

IV. SUMMARY OF CONCLUSIONS

V. REFERENCES , ; i :

ip,

PAGE

34 INSECTS OF SAMOA.

I. THE INSECT FAUNA: ITS COMPOSITION

INTRODUCTORY

WE have already (ix, 19) discussed zoogeography, excluding the insects. The

evidence shows clearly that the fauna and flora reached Samoa from the west.

It appears that there is little evidence of the former existence of a large land

mass in the area now occupied by tropical Polynesia.

It now remains to discuss the evidence that may be derived from a study

of the Samoan insects, and it may be said that the other terrestrial Arthropoda

appear to give little help in this connection. Even with the insects what can

be done is limited, for we have so little knowledge of other archipelagos with

which to compare the Samoan fauna. Two considerable groups of islands, both

of them containing “ high’ volcanic islands with a rich flora and fauna, lie

relatively near to Samoa. To the west is Fiji: to the south and south west,

Tonga. The insect fauna of Fiji is fairly well investigated but only so far as

some groups are concerned. There is, for instance, the book by Bezzi (1928) on

the Diptera, excluding the Nematocera, and there are a number of papers by

Kirkaldy and others on the Heteroptera ; but no comprehensive study of the

Fijian insects has ever been attempted. As to Tonga, much less collecting has

been done, and our knowledge is extremely imperfect and fragmentary ; indeed,

I do not know any important group of Tongan insects which has been thoroughly

investigated. What is known about other archipelagos which lie farther from

Samoa is of less assistance to us. The Pacific Entomological Survey has made

large collections in the Marquesas, and the results are now coming from the

press as Bulletins of the Bishop Museum in Honolulu. A preliminary general

description of the Marquesas has been published by Mumford and Adamson

(1934), from which it appears probable that the relation between Marquesan

and Samoan insects is not close. It is, of course, well known that the insects

of the Hawaiian group have been very fully investigated. Apart from the

classical Fauna Hawaiiensis, there are a large number of papers in the Proceedings

of the Hawaiian Entomological Society ; Bryan (1934) has recently published a

census of the fauna, showing the number of endemic and other species, order by

order. But Hawaiia is so remote from the rest of Polynesia, and its fauna is

so peculiar that direct comparisons between it and Samoan insects cannot be

made.

As there is so little with which our fauna can be compared, the present

THE INSECT FAUNA: ITS COMPOSITION. 35

summary is not final but provocative: it is my intention to suggest subjects

which demand study in the field, which is preferable to the collection of more

insects.

THE MATERIAL

The Insects of Samoa is based upon all material which was available. The

nucleus was a collection formed by G. H. H. Hopkins and myself during our stay

in Samoa, which lasted throughout 1924 and 1925. We were working there for

the London School of Tropical Medicine, as it was then called, and we were only

able to work in the field as opportunity arose. But we collected steadily on

the island of Upolu, from the coast at Apia to the rain forest at about 2,000 feet

round the rest-house at Malololelei. Our collections in this part of Upolu were

made at all seasons of the year, but the work was done on occasional days and

half days, when we could spare the time. We also made small collections on

Savaii and Tutuila, and on several of the small islands which lie off the east coast

of Upolu. Our collection was thought to amount to about 20,000 specimens.

Where specimens are enumerated in the systematic parts of this work, and no

collector's name is given, it is to be assumed that they were collected by

Mr. Hopkins and myself. Our friend, Dr. J. S. Armstrong, made a collection

principally on Upolu between the years 1921 and 1926. All types and unique

specimens in his collection and ours have been placed in the British Museum

(Natural History).

The material which I have mentioned has been supplemented by all the

Samoan material in the Bernice P. Bishop Museum, Honolulu, which was put at

our disposal by the authorities of that institution: about 6,000 specimens were

received from Honolulu. This material included collections made by O. Swezey

and G. F. Wilder, who collected at Pago Pago, Tutuila and, to a less extent, in

other parts of Samoa in 1923. With their material were the insects collected

by E. H. Bryan in 1924. Mr. Bryan worked on all the three main islands, and

his collection is especially valuable because he reached a height of about 5,000

feet on Savaii: no other Samoan material from any altitude beyond 2,000 feet

has been available. Mr. A. F. Judd has also collected in Samoa, particularly

in the Manua Group, and his specimens were also forwarded from Honolulu :

no other considerable collections have been made in the Manua Islands. A

collection made by Dr. H. C. Kellers on Tutuila and other parts of the group in

1918 was sent to us by the Experiment Station of the Hawaiian Sugar Planters

36 INSECTS OF SAMOA.

Association by the kindness of the late Mr. Muir; I understand that this material

has now been incorporated in the Bishop Museum’s collection. Several of the

collectors from Honolulu were expert in obtaining the obscure insects found in

the wet mountain forests of those islands ; this fact made their Samoan collections

particularly valuable. On the other hand it should be recorded that no one

collector spent much time on several islands, which is regretted because the

skill and experience of the individual count for so much. The collections

from different islands in Samoa are not therefore quite comparable.

Types and unique specimens received from Honolulu have been returned

to the Bishop Museum. There has been a free interchange between that in-

stitution and the British Museum of duplicate material, and both institutions

have benefited greatly from this arrangement.

The British Museum (Natural History) contains a number of small collections

made in Samoa, some of them many years ago. Among them one may mention

Lepidoptera sent home by the Rev. C. Whitmee, P. de la Garde, J. Lister, and

G. F. Mathew: Dr. H. Swale’s beetles ; and a number of insects of several orders

collected by Dr. F. W. O’Connor. The authors of certain portions of this work

have made use of other Samoan material. Specimens from Hamburg, Vienna,

Tring, Oxford, and several American collections have been examined and

utilized.

It will be observed that, owing to friendly collaboration with the Bishop

Museum, a large amount of material has been available from all the principal

islands of the Samoan Group, with the exception of Rose Atoll. This exception

is of no material importance, for the atoll is utterly unlike the high, fertile,

voleanic islands which compose the rest of Samoa, and only four sorts of flowering

plants grow upon it (Mayor, 1924). We may, therefore, assume that its land

fauna is negligible : except in name it 1s no part of Samoa.

The collections were made on all the Samoan islands, and at all seasons ;

it may well be asked whether they afford a fair sample of all orders of insects.

In this matter I can speak definitely for Mr. Hopkins and myself, and we

accumulated the greater part of the material. It was our fixed intention to collect

insects of all orders, trusting that some channel of publication would be dis-

covered if an adequate general collection were brought home. But in point of

fact, two men, working only in their spare time, cannot devote attention to all

insects, even in an oceanic island ; one tends to see and collect those groups of

insects with which one is specially acquainted, or those about which some

THE INSECT FAUNA: ITS COMPOSITION. 37

correspondent at home is frequently writing. I should say that Hopkins and

myself were least successful in obtaining Coleoptera, Thysanoptera and parasitic

Hymenoptera ; we obtained fair but not good collections of most families of

Hemiptera, Hymenoptera (excluding Parasitics), Diptera (excluding Nemato-

cera), and most of the small orders of insects and of the terrestrial Arthropods.

In the following groups we obtained our best collections :—Orthoptera (especially

Gryllidae), Fulgoroidea, Lepidoptera (including Microlepidoptera *), Nematocera,

Isoptera and Neuroptera ; the Odonata were carefully collected by Armstrong.

Among the other classes of terrestrial Arthropoda, we made a large collection of

Araneida. The collections of Myriapoda and terrestrial Isopoda were not good ;

no collection of Acarina was made.

The fact that a considerable number of specimens were collected by other

entomologists has been of the greatest service, and I believe that the material as a

whole is a very fair sample of the insect fauna of the Samoan Islands. If further

exploration becomes possible, I think that particular attention should be devoted

to the Coleoptera t, the Diptera Athericera, and the insects inhabiting soil (at

present unknown 4).

We have been so fortunate as to find experts who were willing to deal with

all the material which was collected, except only the parasitic Hymenoptera.

Even in that group it would have been possible to identify certain parts, but it

appeared that in the present state of knowledge it was better that these insects

should be monographed by genera or families rather than on a geographical basis.

It is at least certain that, had we dealt with them in a necessarily imperfect

manner, no general conclusions could have been drawn from the work. The

parasitic Hymenoptera collected by ourselves have been presented to the

British Museum ; those collected by the workers from Honolulu, and this is the

greater part of the material, are in the Bishop Museum.

THE CENsuUS

The total number of species of insects dealt with in this work is 1,603. In

this figure are included not only the 1,568 species, specimens of which occurred

in the collections under consideration, but also 35 others which are definitely

* The term is convenient if not scientific.

{ Holdhaus (1934) observes, with justice, that the number of beetles recorded from Samoa

is under 400, but that an expert collector spending some years in the islands, and devoting due

attention to the insects of the soil, might raise the figure to 1,500 or 2,000.

t For example, no Japygid is recorded in this work. But Silvestri stopped a few hours in

Pago Pago and collected the type of a new species (Silvestri, 1930).

38 INSECTS OF SAMOA.

recorded in earlier papers but not represented in these collections. In addition

to the above, a few species have been mentioned by authors in the text, on the

eround that they will eventually be found to occur in Samoa ; there are also 39

which have been referred to a genus, the specific identity being left uncertain ;

the species which fall in both these categories are additional to the total of 1,603

species, which are definitely and precisely known to form a part of the insect

fauna of Samoa.

A table, showing the number of Samoan species in each order of insects,

and the number and percentage at present supposed to be endemic, is given on

page 67 below.

The number of new insect genera described in this work is 80. The number

of new species is 669 ; in this total we have not included a few species collected

from Tonga, etc., but not from Samoa, and described in this work because they

were closely related to Samoan insects.

With regard to the terrestrial Arthropoda other than insects, a total of 113

species is dealt with in the text, but it must be remembered that no collection

of mites (Acarina) was made. Of this total 52 were apparently endemic. There

were 15 new species and one new genus among them.

As a result of this labour in the field and the museum, it may be stated with

confidence that the insects of Samoa are better known than those of any group

of islands in Melanesia or tropical Polynesia, excluding the Hawaiian Islands.

Extensive collections have also been made in the Marquesas, and a short review

of the insects has appeared (Mumford and Adamson, 1934); a more compre-

hensive report on it is awaited with great interest. Though our knowledge of

many groups of Samoan insects is still very imperfect (page 37), the collections

on which the Insects of Samoa is based have resulted in a very great increase in

knowledge. This is evident when it is seen that 1,603 insects are known to

occur in Samoa, of which 669 are described as new in the present work. Even

in the Orthoptera, which were carefully collected by Rechinger (Holdhaus, 1908),

the total number of Samoan species has been raised from 48 to 78. The weevils

(Curculionidae) had also been previously investigated: Marshall's * earhier

* Tn referring to work published in the Insects of Samoa, I have generally thought it sufficient

to say “‘ Marshall states” or ‘‘ Edwards holds the view that,’ and to give no further reference.

But when it seems necessary the distinction is made between an author’s paper published elsewhere

(Marshall (1921) in the present instance) and his contribution to this work (iv, 249, referring to

part iv, p. 249).

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 39

paper (1921) included 22 species (15 genera), and his contribution published in

this work (iv, 249) deals with 86 species, which belong to 55 genera. The

increase in our knowledge of Fulgoroids is similar: Muir (1921) dealt with 26

species which belonged to 19 genera: the same author has now (11, 1) made a

list of 51 species and 27 genera. There are other groups, previously almost

unknown. We now know 103 Samoan Heteroptera, instead of 6 * ; 137 Micro-

lepidoptera instead of 6 ; 36 Scolytidae instead of one ; and 100 Nematocera of

9 families instead of 11 of 2 families.

It. THE INSECT FAUNA: ITS PECULIARITIES AND

CHARACTERISTICS

Briotocicat Notes

GENERAL.—If we wish to think of the Samoan insects in their natural

environment, we must bear in mind the considerable body of information which

has been set out in an earlier part of this work (ix,1-19). Briefly, we may say that

Samoa lies within the tropics, and that the islands are steep and volcanic. There

is a sharp distinction between the forest-covered hills and the coastal belt,

which is less rough and in many places cultivated. The climate is warm, moist

and uniform, and this is true of every part of the archipelago. Two climatic

events show some seasonal difference ; there is a time of heavy rain ; also there

are considerable differences in insolation, though the influence of this upon

the temperature of the air is masked by the thermostatic effect of the ocean.

The mountain forest is luxuriant, and epiphytic plants and Cryptograms are

abundant: a high proportion of the flora is endemic. The fauna, apart from

insects, is limited, and many groups are totally absent: the proportion of

endemic forms is high. The people are probably Asiatic in origin. In the old

days they made use of a great variety of plants which they must have carried

about the Pacific on their remarkable voyages, and doubtless they carried a

number of small animals. The influence of European shipping has been felt

* The figure is given as 5 by China (ii, 81). Since his paper was published he has called

my attention to an additional Reduviid, Adricomius annulatus described from Samoa by Distant

(Ann. Ent. Soc. Belg., vol. 47, pp. 60-61, 1903).

40 INSECTS OF SAMOA.

for a little more than a century, and a number of recently introduced plants

and insects flourish.

In this home the collectors obtained 1,603 species of insects, of which 787

appear to be endemic to Samoa. It would perhaps be fair to say that the

majority of these forms of life exist in a rather generalized environment. By this

T mean that they live and breed in dead moist leaves, or in rotten wood or in soil.

Collecting in the hills, one does not easily realize the abundance of insects

which occur ; indeed Scott’s (1933) description of the insect life of the Seychelles

would be almost true of Samoa. One walks along a path through the forest,

but sees very few insects, even if one sweeps the herbage. But by devoting

attention to a particular tuft of ferns on a tree trunk, or to a mass of the climbing

Pandanid, Freycinetia, one discovers a number of minute Psocids, Tineina and

sometimes Fulgoroids. Rotten trunks of different sorts of trees certainly have

their particular fauna of Scolytid and other beetles. Our rather short stay in

Samoa did not enable us to acquire knowledge of the particular environmental

niches which are occupied by the native insects ; a glance through the pages of

the present work will reveal how little is known, in comparison with the records

which Perkins accumulated in Hawaiia. The matter is commended to those who

may be able to give it the attention it deserves.

Collecting at night, particularly on the verandah at Malololelei, did not

often produce large numbers of insects. There is, however, one night which

stands out in the memory, that of 24th February, 1924; on this occasion I

collected 350 moths out of perhaps ten times as many on the verandah, and

noticed the relative scarcity of insects of other orders. There were a dozen

Cerambycids, and a few Scolytids, all apparently of one species ; there was also

one Pentatomid and three Cicadids (Baeturia exhausta); besides these there

were hundreds of Tipulidae (Trentepohlia).

Krrects or Season.—In making our collections, Hopkins and I started

with the knowledge that we were at work in a tropical archipelago with an equable

climate ; we wondered whether we should find seasonal forms of insects, or any

evidence of prevalence at particular times of year. With regard to seasonal

forms, one may say at once that no author has observed them; even Prout,

after examining a series of 260 Cleora samoana taken in every month of the year,

concluded that there was no relation between the season and the occurrence of

any form of this extremely variable moth. Among the butterflies there are one

or two species which have “dry” and “ wet” forms in other parts of their

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 41

distribution, but in Samoa they are consistently “wet”: Melanitis leda

solandra * is the clearest example of this.

As to evidence of seasonal prevalence of particular species, little is yet known.

From what has been said of the climate (ix, 10-19), it might be expected that

most of the insects would be active throughout the year, and this is generally

true ; Chopard, for instance, has remarked upon it in discussing the Orthoptera.

But it is known that some at least of the trees flower at particular times of year

(ix, 25), and we may suppose that this determines a seasonal prevalence of the

insects which are dependent upon them for pollen, or nectar, or fruit. The only

clear case of seasonal prevalence which has yet been noticed is in the butterfly

Precis villida samoénsis ; in the year 1924 Hopkins noticed that this butterfly

was common in Apia from September till December, and that it was rare in the

other months. In 1925 we gave particular attention to the matter, and the same

seasonal prevalence was again apparent. On our occasional visits to Aleipata, at

the east end of Upolu, or to Savai, we found the butterfly abundant at the same

time of year. These facts cannot be explained ; the feeding habits are not known.

Our principal task in Samoa was the detailed study of the biology of a

mosquito, Aédes variegatus, which is of considerable importance as the vector

of the local filarial diseases of man. This insect lays its eggs in small rot-holes,

half coconut shells, and similar places, and in order to understand why these

places were chosen we carried out an extensive series of experiments with

trap breeding places. We found, among other things, that continuous heavy

rain reduces the number of eggs laid in the traps, and that dry spells of two or

three weeks had a still greater effect (Buxton and Hopkins, 1927). These

differences in the size of the population of this insect may perhaps be called

seasonal, depending as they do on climatic events, but it is to be understood

that they are not of regular occurrence.

In the systematic parts of this work care has been taken to publish the dates

on which specimens were captured. We do not yet know whether many of

the insects have seasons of prevalence or not, but we are confident that our

facts will be of value to some one who may care to inquire further into the

matter.

* So referred to by Hopkins. Poulton and Riley (1928) refer specimens from the Society

Islands and Tonga to this form, and remark that all their specimens are of the wet-season form,

They find consistent peculiarities in Samoan material, which they refer to a new subspecies,

M. l. hopkinsv.

42 INSECTS OF SAMOA.

Errects or ALtirupE.—In considering range in altitude, we must remember

that if an insect extends its range from the coast to the hills, it will reach a lower

temperature and also, according to Angenheister (1924), a greater daily range of

temperature ; it will also be exposed to more rain (ix, 17). On the other hand,

it will escape from many introduced weeds and other plants, and from some of the

introduced animals. As conditions are so different, it is not to be wondered at

that the greater part of the Samoan insects have been collected either in the hills,

for instance at Malololelei, or else near the coast; the proportion of species

which have been found in both is low. But as the facts are all accessible in the

systematic Parts of this work, it seems unnecessary to quote examples. Hopkins

mentions an interesting example of distribution limited by some factor unknown.

The butterfly, [ssoria sinha bowdenia, was found in all the Samoan Islands,

and it was common throughout the year in the coastal region. But it was never

seen above 1,000 feet, though its food-plant, a small native tree, grew up to

2,000 feet. The limitation of the insect to the lower levels cannot be due to the

distribution of its food ; it is not easy to ascribe it to the action of imported

enemies (which would readily explain a reverse distribution, with absence from

the low ground) ; and it is not probably due to temperature, for the same race

of the insect occurs in Tonga, which is a good deal cooler than Samoa.

Errects of Wr1np.—It will be remembered (ix, 18) that winds in Samoa

tend to blow from the east or south-east at all times of year, and that for months

on end they are very steady ; this regular south-easterly wind is the trade wind.

The surface drift of the ocean is from the same quarter. It therefore seems that

we are confronted with an anomaly, for the flora and fauna is unquestionably

derived from the west but winds and currents come up from the east or south-

east. But the difficulty is not so great as it seems to be, for it has been shown

(Thomson, 1925) that the trade winds are shallow—no more than 2 to 44 miles

deep ; over them lies a much greater stream of air, the anti-trade, which generally

blows at a greater velocity than the trades; its direction is approximately

opposite to that of the trade wind. Neither can we forget the hurricanes, or

cyclones: Visher (1925) has pointed out that these winds strip leaves and

branches from the trees, and may therefore introduce living things into the upper

air. They also produce occasional floods, so that rafts of vegetation, following

a hurricane, may drift far out to sea. Even though hurricanes are rare in Samoa,

they are common enough a few degrees to the west, and it is possible that forms

of life, swept up in Melanesia, have occasionally landed in Samoa.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 48

The carriage of insects over great distances by wind, and in particular by

the currents of the upper air, may be of great importance ; it is now becoming

possible to investigate it, and a considerable body of fact already exists. It is,

of course, well known that insects, including many whose powers of flight are

limited, circulate in the wind near the ground ; Fulton and Chamberlin (1931)

have described traps which operate 20 to 30 feet above the ground, and which

catch large numbers of small Homoptera, flies, moths, beetles, etc.; Ball’s

(1918) records of the insects which reach the lighthouse on the Rebecca Shoals

105 miles from Florida are well known. But it is easy to see that if insects are

circulating over the face of the earth, they may easily be carried upwards,

perhaps by convection currents (an agency which would not be easily detected

by the ordinary methods of meteorology, though the modern sport of gliding

tends to call our attention to them). Moreover, we have a few actual records

of insects collected at altitudes of a few thousand feet: applied entomologists

have begun to use trapping devices attached to aeroplanes, and a large number

of minute insects have been collected at these heights. Coad (1931), for instance,

describes an apparatus which can be brought into use at any time during a

flight, so that the level at which the insects actually occurred can be known.

Within 2,000-3,000 feet of the ground, a great variety of insects may be found,

including moths and other insects of strong flight. Higher up—and the in-

vestigators (and the insects) penetrated to 14,000 feet—the insects were nearly

all small and weak, and many wingless forms were taken *. A number of other

records of insects which are distributed by wind, and a general discussion of the

subject, is given by Felt (1928): he calls attention to the fact that the winds of

the upper air are swift and very constant in direction.

To me it appears that a number of facts about Samoan insects can best be

explained by supposing that they have reached the islands through the upper

air. One can mention a number of minute insects, found in Samoa and in other

places very far away. The Ceratopogonid, Forcipomyia mgrami, occurs in Fiji,

Samoa and West Africa; F. inornatipennis, a common West African species,

has been recently recorded from the Society Islands (Macfie, 1933) ; several of

the Samoan Psychodidae are apparently almost cosmopolitan. The mosquito,

Aédes vexans, occurs in the Palaearctic and Nearctic Regions, including many

* A very recent paper by Berland (Ann. Soc. ent. France, Vol. 104, pp. 73-96, 1935)

should also be consulted. The author, working in France from an aeroplane, has collected a

variety of small insects, some of them alive, up to 2,300 metres.

44 INSECTS OF SAMOA.

places which are quite cold; it is also known from a number of parts of Asia,

and it extends eastwards as far as Samoa; it is even known to occur on some

Polynesian atolls. Wherever it occurs it tends to breed in shallow, temporary

pools, and to disappear from a locality after existing for a generation or two ;

there is no reason for thinking that it is introduced by man, and it seems likely

that 1t owes its wide and erratic distribution to winds. Edwards has called

attention to the fact that the Samoan Nematocera tend to be small and light,

and that the larger Bibionidae and Tipulidae are absent.* This again seems to

point to aerial distribution.

The examples quoted are all Diptera, but examples from other groups of

insects can be mentioned : for instance, the Collembolan, Onychiurus fimetarius, —

is known from Europe, North America, Africa and Sumatra, as well as Samoa ;

but as it occurs in soil, it perhaps owes its distribution to some factor other than

wind. Osborn records several minute Cicadellidae previously known from

India or Ceylon : but whether spread with food-plants or by wind who can say ?

The small Zygopterid dragon-fly, I[schnura aurora, is clearly an insect which

owes its very wide distribution to carriage in the upper air; Fraser himself has

seen the newly emerged insects rise high into the air until they were lost to sight,

and it seems that its very weakness has allowed this insect to spread not only

from India to Samoa, but even out to the barren atolls of the Ellice group

(vil, 23) and the Tuamotu Archipelago (Cheesman, 1927).

There are also several minute Coleoptera, the distribution of which may

perhaps be due to wind, though a more intimate knowledge of their biology is

desirable. The minute (1:3 mm.) Anthribid, Scirtetinus pacificus, endemic to

Samoa, but congeneric with species found only in the Seychelles, may be wind-

borne: compare also the Scolytidae known only from Samoa and Ceylon,

especially Scolytomamus, a genus containing only two species, one endemic in

each island; also the small (2 mm.) Colydiid, Hystricones vagans, of Samoa,

representing a genus of which the only other species is Central American. It is

interesting to see that Scott (1933, p. 328) records several species of minute

beetles, of different families, each one known only from the Seychelles and

some remote island (Hawaiia in two instances).

It is possible that the distribution of these beetles and other insects may be

more normal than it appears to be: further research may show that they occur

* And compare Edwards’ (1926a) observation that in the Philippines the large Tipulids

tend to have a restricted distribution, the smaller Limnobiidae to be more widely spread.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 45

in many intermediate countries : the fact that they seem to be confined to islands

may be due to the care of individual collectors, or because obscure and minute

forms are more readily discovered in a limited fauna. But to me it seems

probable that they have been distributed by wind, particularly by the currents

of the upper air, and that it is to their minute size that they owe their immense,

rather erratic, distribution.

But we must admit that there are certain groups which seem not to have

been distributed by wind, though they are small. The fact that all the Aphids,

and nearly all the Coccids, belong to species of wide distribution may perhaps be

due in part to wind-dispersion, though many of the species are pests of introduced

agricultural plants, brought in by Huropeans. If carriage by wind had been

effective in the case of these insects, we should surely have expected to find

endemic species descended from those which arrived in the remote past. The

complete absence of native plume moths, Pterophoridae (ix, 67), which seem

so well adapted to wind carriage, is also most remarkable.

It appears that the distribution of spiders in Samoa and other remote

islands may perhaps give evidence of carriage by wind, for it is familiar that

young spiders of many sorts throw out gossamer by which they are carried high

into the air. It will be remembered that the Samoan spiders are rather well

known, and that of 81 species, 46 are endemic and 25 are found elsewhere in the

Pacific, one or two of them extending so far as New Caledonia and the Malay

Islands. The remaining 10 are cosmopolites, many of them of recent intro-

duction. The 71 species which may be presumed to be native are most unevenly

distributed among the families, for there are no Mygalomorphs or Drassidae, and

certain other families are very slightly represented. In these particulars,

there is general resemblance between the spider fauna of Samoa and of the

Hawaiian Islands. The Samoan spiders were dealt with by Berland (viii, 38),

and the same author (1934a and b) has discussed our present knowledge of the

spiders of the Pacific region. It appears that the Mygalomorphs are known to

occur in all the lands which form the western border of Oceania, the Philippines,

Japan, the New Hebrides, Loyalty Islands, New Caledonia and New Zealand :

and that they extend eastwards to Fiji, where two small species occur; but

there are none in the archipelagos farther to the east. It might therefore be

said that the Mygalomorphs have a continental distribution. Berland (1926)

has also discussed the species which possess a wide tropical distribution. He is of

opinion that all these owe their spread to human agency, though it is not neces-

46 INSECTS OF SAMOA.

sarily recent, and that carriage by air has had little to do with the geographical

distribution of spiders. He holds that, inasmuch as the spiders of islands are

largely endemic, and since there is a great difference between the spiders of

Madagascar and Africa, distribution as aeronauts cannot be effective.

The matter has also been studied by Bristowe, who has published a general

account of the distribution and dispersal of spiders, considering the effects of

various climatic factors including wind, also of soil, plants and human activities

(Bristowe, 1930). He has also collected on Krakatau and discussed its spiders

and the method by which they may be supposed to have re-colonized it since the

eruption of 1883, which is believed to have sterilized that island completely

(Bristowe, 1931). Bristowe does not question the importance of human agency

in distributing spiders, and it is clear that certain species which are widely spread

are domestic. But he clearly holds that distribution by air also occurs, and that

it is important. He finds little difficulty in meeting Berland’s objections to this

view, and produces a large body of evidence which appears to show convincingly

that dispersal by air is common and that it explains much of the geographical

distribution of spiders. In particular, he discusses the absence of Mygalomorphs,

Drassidae and Tetrasticta from Samoa, Hawaiia and other islands, and he makes

it appear probable that their absence is due to the fact that the young spiders

produce no threads and are not aeronauts.

It is hardly possible to express an opinion on a matter which has proved

puzzling to experts, but Bristowe’s view appears to be founded on a large body

of fact, and it provides an explanation of what is characteristic of the spiders of

Samoa and many other archipelagos.

Lack of wings, or of power to use them, is characteristic of insects inhabiting

certain islands. But I know of no evidence that this is so among Samoan insects :

I think that none of the systematists has discovered a high proportion of

apterous forms in his material. In one or two groups it is certain that there is

not a high proportion of wingless forms ; for instance there are about six native

Blattidae, all of them winged ; in the Gryllidae there are 32 native species, of

which 20 are winged, the remainder wingless (but two of these have winged males,

wingless females). The matter is mentioned here, for it is generally held that it is

the exposure to wind which causes insects on certain islands to be apterous.

Micrant Insecrs.—At this point it is appropriate to consider the migrant

insects, though it is not certain that their distribution or flight have any relation

to wind.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 47

Among the butterflies, the most interesting is Danaida archippus ; its history

in Samoa is dealt with by Hopkins (iii, 7), and the story of its spread in Oceania

is known in detail, thanks to Walker (1914), Collenette (1925a), and Poulton

and Riley (1928). It is remarkable that it occurs on a number of the smallest

and most remote islands, including the barren atolls of the Ellice group. A

general account of its flights in America and elsewhere is given by Williams

(1930, pp. 141-156). So far as Samoa is concerned, the only known food-plant

is the weed, Asclepias curassavica, which must owe its introduction to Huropean

shipping, and may well have come in as seed, with hay or straw. But the

butterfly could not have reached Samoa in dry material, at any stage of its

existence, and we must suppose that it came by flying. But if its distribution

is by flight and if it feeds only on a plant of recent introduction, then members

of this species must have been flying about the Pacific, seeking to colonize the

islands, for countless ages. The earliest record of the butterfly in Samoa is in

1869. A record which is similar in certain respects is given by Wood-Jones

(1910, p. 352); the moth, Plusza chalcites, a notorious migrant, was apparently

absent from Cocos-Keeling, until tomatoes were grown, when it appeared on the

plants: the tomato seed came by post from England, so that one can be con-

fident that the insect was not introduced in that way. The conclusion to which

Wood-Jones came is that the moth arrived by flight, and established itself

because the tomatoes were growing: inasmuch as it arrived soon after the

plants began to be grown, it seemed highly probable that it had repeatedly reached

Cocos-Keeling, but failed to establish itself.

There is some evidence which suggests that another Danaine butterfly,

Hypolimnas bolina, owes its distribution in the Pacific to recent flights from

island to island. It was apparently absent from Pitcairn Island in 1883, but it

has been recently found there (Poulton and Riley, 1928) *; it is improbable

that it had been introduced in one of the early stages on living plants. But,

as to Samoa, we may be certain that this butterfly is not of recent arrival, for it

has developed peculiar forms, which are different on different islands in that

archipelago, though they are not confined to Samoa.

Among the moths the following Samoan species are known to be migrant,

at least in some part of their range: Herse convolvuli, Hippotion celerio, Heliothis

* Poulton and Riley’s paper states that they have examined specimens received from Mr.

C. J. Grist, but they do not quote the date of collection. Mr. Grist has since informed Prof.

Poulton that the specimens were collected in 1923 and subsequent years.

48 INSECTS OF SAMOA.

armagera, Prodenia litura and Plutella maculipennis.* But all these insects are

known to feed on many sorts of plants, including economic plants of which they

are frequently pests: it seems impossible, therefore, to discover how much of

their range is due to flight and how much to artificial introduction. Achaea

janata falls in the same group. It is known to be a pest of many crops (ili, 175),

and may owe its occurrence in Samoa, and its wide distribution in the tropics

of the Old World and in Polynesia, to artificial introduction. On the other

hand, there is a record (under the name A. melzcerta) of a single specimen flying

aboard a yacht 450 miles from Ducie Island and much farther from any large

island (Collenette, 1925b). It is possible that the wide-ranging Arctiid, Utetheisa

pulchelloides, is a migrant, as the nearly related U. pulchella is known to be.

On the other hand, there are several notorious migrant Lepidoptera which

are not known to occur in Samoa. The small moth, Nomophila noctuella, was

not found, and it appears that it is not known from Fiji and Tonga; this is

remarkable, for it is very nearly cosmopolitan, and it occurs in the Malay Islands,

Formosa, Australia and Hawatia (W. H. T. Tams, by letter). The absence of

“ Painted Lady ” butterflies (Vanessa cardut) is also remarkable, as there is an

endemic race of this insect in Australia, Tasmania and New Zealand (V.c. kershawt).

There is a record of the capture of one specimen of this race in very worn condition

at Suva, Fiji (Williams, 1930, p. 204); in coming from its breeding place it

must have crossed about 1,000 miles of ocean, but in spite of this individual

occurrence, the species has failed to establish itself in Oceania.

It seems certain that the majority of migrant insects in the Samoan fauna

are Lepidoptera, but it is probable that several of the large dragon-flies

(Anisoptera) are also migrants, though I do not think that any records of actual

movement have been observed in Samoa. Among them it is probable that

Diplacodes bipunctata, Tramea limbata and Pantala flavescens reached the islands

from time to time as active migrants: it is quite likely also that representatives

of Anax, Henucordulia, Lathrecista and Orthetrum, all of them widely distributed

in the tropics, fall in the same group. Fraser, indeed, has put forward the view

that all the Samoan Anisoptera are migrants with the exception of three endemic

species of Hemicordulia and Gynacantha. The interesting Zygopteron, [schnura

aurora, Which appears to have been distributed by air currents rather than

actively as a migrant, has already been discussed (p. 44).

* This insect was certainly present in Samoa, and material was collected. The specimens

seem to have been lost, and they were not included in the collection dealt with by Meyrick.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 49

InsECTS OF SPECIAL ENVIRONMENTS.—In Samoa we collected in a few places

of particular interest ; it is clear that the highly specialized creatures which inhabit

these spots must have encountered particular obstacles in reaching their present

homes ; these would not have been eritirely avoided had they possessed the

power of calling land-bridges from the vasty deep. The particular environments

which we shall describe shortly are mountain gullies, lowland swamps, mangroves,

the strand and the caves.

(1) Mountain Gullies.—The steep mountains of Samoa, which

receive such abundant rain, are scored by innumerable parallel gullies, the sides

of which are often nearly vertical ; the water in these gullies rushes down in a

series of rapids and waterfalls (ix, 5-8). At Malololelei, Upolu, it was easy to

collect insects in two of these ravines ; at this point one was about 2,000 feet

above sea-level, and near the heads of the streams. The type of scenery and

vegetation is shown in Plates I to II, and V and VI (Part ix). A very similar

environment exists even as low as Vailima (600 feet), though here the streams

tend to be larger. One might indeed say that forest full of ravines and rapids

is typical of nearly all parts of Samoa, except the actual coast, and a zone of a

few hundred feet above it.

On the bank of such a stream, often on a patch of wet sand, one would find

the little grasshopper, Paratettix compactus. In the stream itself were great

black boulders of basalt ; on the down-stream side of a boulder, close to the

water, Anaxipha hopkinsz, an endemic cricket, used to sit. Owing to its colour

it was most difficult to collect it, especially as it flies freely and can jump to the

surface of the rushing water upon which it swims easily. We noticed that this

insect seemed to be confined to the boulders which stood in the water ; it was not

found on those on the bank, even when the stream was no more than a dozen

yards across. Another interesting insect, almost equally specialized in its habit,

‘is the Anthomyid fly Limnophora immaculiventris ; this fly selects a place where

the surface of the torrent is relatively still, under the protection of a mass of

rock, and there it skates over the surface of the water. It is difficult to see, and

even more difficult to catch, and it may easily be mistaken for one of the Gerrid

or Veliid bugs which have the same habit. Nothing is known of the breeding

place of this fly, but it is understood that most members of the genus are asso-

ciated with water. :

The most interesting insects which may be found in the ravines near Malolo-

lelei are the Ischnuran dragon-flies. Eight members of this group are at present

Ix 2

50 INSECTS OF SAMOA.

known to exist in Samoa. Ischnura aurora has an enormous tropical dispersion,

and does not at the moment concern us (see p. 44, above). But of the remaining

7 species, 4 (I. buxtoni, I. haemastigma, Amorphostigma auricolor, Pacificagrion

lacrimosa) have never been collected except in one gully at Malololelei; one

(A. armstrong?) is known from that gully, and from another place in the hills

of the same island; one (J. albistigma) has been collected in the same gully

and also in Tutuila ; one species (1. chromostigma) is known only from a place

in Tutuila. When the innumerable other gullies in the Samoan forests are

explored, it may be that the seemingly extreme localization of these insects will

disappear, and that they will be found widely dispersed in one or several islands.

But it is at least possible that the group of species collected by Armstrong and

others in the one gully is only a small sample of the Ischnuran fauna of Samoa.

The Samoan Ischnurans form an interesting parallel to the dragon-flies of the

genus Nesobasis in Fiji, though I understand that Ischnura is not closely related

to Nesobasis apart from the fact that they are both Agrionids. According to

Tillyard (1924) the genus Nesobasis is at present thought to be confined to Fiji,

save for the existence of a species on Bivak Island, 8.W. New Guinea. From

Fiji 13 species are already known, of which 10 were collected by Simmonds

along the Waidoi River, Viti Levu; it is not recorded whether the 10 species

occurred together, or whether they were localized in different parts of the river’s

course. Here again, we do not know whether further search on Viti Levu,

and elsewhere in Fiji, will show that these species have a relatively wide distribu-

tion, or whether enormous numbers of endemic and highly localized species

remain to be revealed.

In these gullies and in other dark places in the forest one may find that

curious complex of Tipulids and Pholcid spiders, more fully described elsewhere

(vi, 95). The Tipulids (Z'rentepohlia spp.) are extremely abundant, congregating

in large numbers among buttress roots or under overhanging rocks ; here they

stand swaying on their long legs. The Pholcids (Pholcus and Smeringopus)

sway in a similar manner on their webs in places of the same type, and their

legs have pale joints which resemble those of the Tipulids. The resemblance was

noticeable in the field. In the same places we occasionally got a Reduvud

(Gardena) shaped almost like a Tipulid with a very slender body and long legs,

but it was so rare that we have no knowledge of its habits.

At least two groups of insects, which one might have expected to find in

the torrents and waterfalls, were apparently absent ; I refer to flies of the families

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 5]

Simuliidae and Blepharoceridae, the early stages of which are so often associated

with water of this type. The absence of Simuliids is particularly curious, for

these insects occur farther to the east in Tahiti and the Marquesas (p. 87).

One can hardly think of the insects found in torrents without wondering

whether similar evidence might be derived from specialized plants; but this

is not the case, for the Director of the Royal Botanic Gardens, Kew, tells me that

there are two families of higher plants, the Podestemonaceae and Tristichaceae,

which are characteristic of torrents in the tropics; neither occurs farther east

than Java, and no help can therefore be derived from them.

(2) Lowland Swamps.—In contrast to the torrential mountain

streams are the coastal swamps, in which the water moves slowly, though it is

not generally stagnant. These occur in many places, and we collected in them

at Apia, Laulu, Mulifanua and elsewhere. The great majority of these swamps

have been enlarged and cleared by the Samoans, who grow taro in them. The

characteristic dragon-flies are the large Anisoptera, the majority of which have

an enormous range in the tropics of the Old World. The species of Dytiscidae

which were obtained in these swamps were also of wide distribution ; the same

is true of the Gelastocorid, and of one of the Notonectids. There were also a

few Chironomidae, but Hdwards remarks on the paucity of the water-breeding

Nematocera, compared with those which breed in rotten wood or leaves. One

interesting Culicid is found, Aédes kocht samoana ; this insect breeds almost

exclusively in the axils of the taro plant, Colocasia, which is planted in these

swamps. From what has been said it will be seen that the fauna of the coastal

swamps has not provided much of interest ; I fear that it may have been still

further reduced by our successful introduction of a “top minnow,” Poecila

sphenops, from Panama.

(3) Mangroves.—aAt the places where the rivers and streams eventually

reach the sea, some of them have small muddy estuaries, overgrown with man-

groves. ‘The fauna of a mangrove swamp, which is exposed alternately to salt

and fresh water as the tide rises and falls, is not rich, and is mostly marine ;_ but

in the mangrove trees one finds a few insects, Aédes vartegatus breeding in pro-

fusion in holes and cavities in the trunks, and the small, endemic, cricket

Apteronemobius longipes, running on the trunks and hiding in little cavities ;

this particular insect was never found except among mangroves, on which its

dark colour makes it difficult to detect. It is wingless, but covered with fine

hairs, which prevent it being wetted (even by alcohol), and it swims well though

52 INSECTS OF SAMOA.

it does mot appear to take readily to water. When the tide is out the crickets

descend from the trees and run about on the black mud among crabs, gobies and

other marine creatures.

(4) The Strand —But most of the shores of Samoa are not fringed with

mangroves. ‘There are very many sandy bays, separated from one another by

rocky headlands. The white sand, which is really pounded coral and shell, is

crossed by long strands of Ipomoea and one or two Leguminous creepers. There

are also a few trees and bushes in this “strand vegetation,’ all of them very

widely distributed in the Pacific and Indian Oceans, and strictly limited to this

littoral zone (Guppy, 1906; Setchell, 1924, p. 15). Many of the plants have

widely distributed insects associated with them. Thus the yellow creeper

Vigna lutea (marina) is the food-plant of the Lycaenid, Jamides argentina, the

Samoan representative of the widely spread Indo-Malayan J. bochus (ii, 52).

One of the most characteristic of the trees is Tournefortia argentea (Bora-

ginaceae). This tree was the food-plant of the Arctiid, Utetheisa pulchellordes :

the same insect has been recorded as feeding on this tree in the Tuamotus

(Collenette, 1928) and even in certain remote coral islands, Chagos, Amirantes,

etc., in the western Indian Ocean (Scott, 1933, p. 338 footnote). This wide-

spread association of plant and insect is interesting, but it tells us nothing of the

means by which the moth reached the many isolated spots in which it 1s now

found. We suggest above (p. 48), that it is probably a migrant, but it seems

clear that tropicopolitan insects, associated with strand plants, deserve much

more experimental study than they have received: in particular we desire to

know whether there 1s any stage at which they resist sea water, and whether the

egg stage 1s ever so prolonged as to permit the insect to travel on drift wood.

The insects of the strand appear unattractive, for they are so widely distributed,

but they are every bit as interesting as the strand plants which were the object

of the researches of Guppy. That author showed long ago (Guppy, 1906) that

Tournefortia argentea has a vast Indo-Pacific distribution, occurring on the strand

of the smallest and most isolated islands ; that the small seeds have a coating of

buoyant corky material ; that they are common in drift material ; and that they

will float for a year in sea water. Hopkins has described the aggregation of

large numbers of males of the butterfly Huploea s. schmeltzi round dead or

withering branches and fruit of Tournefortia argentea. A similar habit has been

observed in many other species of Huploea in a number of different parts of the

Pacific (for references see Hopkins, ii, 15; also Lever, 1934; Buxton, 1927 ;

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 53

Poulton, 1932). In each place it is males only which are attracted, and where

several species occur on the same island they may all be seen together ; it is

only the dead and withered parts of this particular tree which are visited ; the

visits are observed by day ; assembling for sleep at night, as so many species

of Huploea do, is a distinct habit. The suggestion has been made by Hopkins

that dead pieces of Towrnefortia may give out an odour which resembles that of

the female butterfly. It is not easy to prove or disprove the suggestion, but it

would be of interest to test whether an infusion is also attractive, and whether

the boughs remain attractive 1f concealed with muslin. _

(5) Caves.—Another specialized environment is the cave, of which a

number exist, though only that near Malololelei was visited. It will be remem-

bered (ix, 7; Plate IV, B) that this cave is a lava tunnel, and that one can pene-

trate it for about 200 yards ; it is relatively straight, and a stream enters it at the

upper end, but later runs beneath the floor, so that from that point the cave is

much less wet. The temperature inside the cave in May was 22-5° C., and one

may assume that there is very little variation from this figure throughout the

year. As the cave is generally straight, and as the stream runs into it, one would

expect to find animals which do not in any sense belong to the real cave fauna ;

in this group we should probably put a dragon-fly (Henucordulia sp.) which

visited our lanterns on one occasion, also several undetermined spiders which were

found at the far end of the cave, and the Atyid prawns (Caridina) which were

common in the stream, both outside and also inside the cave even in

total darkness.

Bats and swifts form an important part of the fauna of the cave. The only

species of bat (Hmballonura semicaudata) is a small insectivorous one, hundreds

of which live in the cave, especially in the more remote part, at all times of year.

They were infested by the Streblid, Nycteribosca buxtom. The swifts (Collocala

spodopygia) are an endemic Samoan species; they construct minute nests of

moss and liverwort on shelves upon the walls of the cave ; we were surprised to

find them breeding even at the far end of the cave, in spots which were totally

dark. It appeared that these birds haunted the cave at all times of year. In

spite of the fact that we collected a considerable number, we obtained no para-

sites from them.

These two vertebrates, which live in the cave and feed in the forest, bring

in considerable quantities of material. We made no search in the nests of the

swifts, but the large heaps of bats’ dung, which accumulate on the floor of the

54 INSECTS OF SAMOA.

cave wherever it is not covered by the stream, provided interesting material.

In it we found larvae of a fly, of which we failed to breed adults ; in the same

place we discovered the Pseudoscorpion, Chelifer buxtont. On the heaps of bats’

dung one could always find specimens of a most interesting Locustid, Rhaphi-

dophora rechingert. These are insects of amazing agility, and it is not easy to

catch them by lantern light, as they leap about among boulders and heaps of

bats’ dung; even when they fall into water, they are capable of swimming

across the surface. This insect was described by Holdhaus (1908), from material

which Rechinger collected in Upolu and Savau; Holdhaus gives no notes on

biology and no precise locality. The species is believed to be endemic to Samoa,

and its existence raises two very interesting questions. The first of these is

biological: R. rechingeri lives, so far as we know, only in a cave, but if it left

the cave it could find most of the necessary conditions in the surrounding forests ;

it would find darkness and damp beneath stones and logs, or in heaps of fallen

leaves, where many other creatures already live; if it requires an extremely

equable temperature, this also exists in the forest.* In spite of this, we never

found this insect outside the cave, and the Rhaphidophorinae are, in general,

cave insects, in all the countries in which they occur. One might even say they

are specialized cave-dwellers, lacking eyes and wings. On the other hand, I

have a letter from Dr. R. J. Tillyard, in which he says :

“ T have found that Rhaphidophorinae occur fairly commonly hidden under

piles of loose rocks or stones in New Zealand, particularly in the stony beds of

large, swift rivers, such as are common in New Zealand, and I think it highly

probable that this is the more ancient habitat of the two, and that the caverni-

colous habit has been assumed later. When, therefore, you say that the Samoan

species has no other habit but that of living in caves, I can only reply that this

is contrary to my general experience in New Zealand. It is true that many

collectors in New Zealand would tell you that Cave Wetas only live in caves,

but it is due to lack of observation on their part. Even the famous cave-dwelling

genus Pachyrhamma is to be found sometimes under big, loose rocks; and,

indeed, ecologically, one might ask, what 7s the difference, to an insect with a

body one inch long, and hind legs and antennae both many times that length,

between a true cave, and the large, irregular, dark interspaces between the

tumbled rocks of a retreating glacier-face or the piled-up boulders in the bed

* Tt will be remembered that in the open garden at Malololelei, the mean daily range of

shade temperature was 8° C., and in the edge of the forest, in a shelter hung in a tree, it was 4° C.

One may assume that on the forest floor, among decaying vegetation, the range would be less,

and that it might be immeasurable (ix, 15).

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 55

of a steep mountain torrent? LHcologically, I conceive that the cave is only

the end-term of a series of habitats in which the factors of darkness and security

increase in a sort of arithmetical progression.”

This is Dr. Tillyard’s experience in New Zealand, and it is possible that R.

rechingert might be found in the type of place which he describes, if search were

made in Samoa ; but I do not hold that this is probable, for we devoted particular

attention to the Orthoptera, and were not unsuccessful in obtaining obscure

Gryllids, ete. (i, 9-58). The second question, relating to this insect, arises from

the first and relates to distribution. Numerous species of Rhaphidophorinae

occur in New Zealand, and there are a few in Australia (Tillyard, 1926, p. 96).

They also occur through the Malay Peninsula, and the islands as far east as New

Guinea ; none are yet recorded from Melanesia, Fiji or any part of Polynesia

except Samoa (Karny, 1929), though perhaps their apparent absence from these

archipelagos is due to our ignorance. But even if these insects are subsequently

found in Melanesia and Fiji, and admitting that they are not absolutely confined

to caves, their existence in Samoa is difficult to reconcile with the view that the

fauna is “ oceanic.”

(6) Certain Parasitic Relationships.—Another type of

specialist is the insect which is closely dependent on some other species of animal

or plant. This dependence limits its distribution: for instance the Nycteribiid

genus Cyclopodia, which is found in the fur of fruit-bats (Pteropus), occurs in

Samoa, and presumably in Tonga, where these animals are found; but the

absence of Pteropus from the Cook Islands, and the Society Group, must deter-

mine also the absence of the parasite. or a similar reason, as there are no wild

mammals except bats and rats, Samoa contains none of the Hippoboscids which

infest mammals. But this general relation does not always hold ; for instance,

the Hippoboscids found on birds are not specific. Ferris mentions that Orna-

thoica promiscua is found in California on Passerines and a falcon, and in the

Philippine Islands on a woodpecker and a kingfisher ; in Samoa it was collected

from a fly-catcher and a heron. In a similar way, Ornitheza metallica has been

collected in Samoa on a starling, and in the New Hebrides on a kingfisher ; in

Kurope it has been found on the jay and the heron, and in the Philippine Islands

on a bunting. It is clear that parasites such as these are heedless alike of the

canons of systematic ornithology and of the “ lines ” of the zodgeographers.

Among other parasitic groups, one notices that the order Strepsiptera,

and the Conopidae (Diptera), are not represented in the Samoan collections.

56 INSECTS OF SAMOA.

We may suppose that the conditions which allowed certain families of Hymen-

optera to colonize Samoa did not let their parasites do so; it is clear that any

random method of colonization which is perilous to a host species is much more

so to a parasite, for 1t occurs in the body of no more than a small proportion

of hosts. But the validity of this argument is weakened by the fact that two

species of Pipunculidae are known to exist in Samoa. Enquiry along these

lines might be extended. If, for instance, it were shown that gall-forming insects

were rare, though common in the more “ continental” islands to the west, the

fact would support the view that Samoa is “ oceanic.” But the absence of facts

relating to these insects and also the parasitic Hymenoptera precludes discussion

at present.

INSECTS AND Human Acency.—The endemic insects and other Arthropods

are discussed below. It appears that of a total of 1,603 species of insects, 787

(49°%) are peculiar to Samoa, so far as present knowledge goes. The remainder

are either native, occurring in other countries, but having reached Samoa by

natural means, or introduced, either by Polynesians or more recently by Huropean

ships.

It would be a matter of great interest could we divide all the insects into

the three groups endemic, native and introduced, and at one time | attempted to

do so. But I have now reached the conclusion that this is not possible, except

perhaps for an expert in a particular group. It will be observed that Perkins

(1913, p. xlii) discussed this grouping, originally suggested by Blackburn and

Sharp (1885), but recognized the impossibility of placing many of the insects

in the correct category. To me it appears best to distinguish only the endemic

species from the others; this is done, for each order of Insects separately, in

Table 1, p. 67. But though one cannot divide the whole of the Samoan insects

into these categories, there are many individual cases in which this is possible.

In certain circumstances one may go farther and show that it is extremely

probable that introduction was due to the Polynesians, or alternatively that

it is almost certainly recent and due to European shipping. For in dealing with

the fauna of Oceania, one must never forget that the Polynesians made long

and frequent voyages, and that they carried many useful plants with them

(ix, 26). Let us therefore endeavour to distinguish the species spread by

Polynesians from the more recent introductions for which the white man is

responsible, and to distinguish both these introduced groups from the native

fauna.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. DT

(1) Polynesian Introductions.—lIf an insect has a wide dis-

tribution in the Pacific, and is associated with a plant which was useful to the

Polynesians in the days before the Kuropean arrived, one may assume that it

was carried about in the canoes and class it as an early introduction. There

are two excellent examples among the Fulgoroids: Megamelus proserpina,

which feeds on the leaves of taro (Colocasia, Araceae) and is known to

occur in northern Australia, many Malayan islands, Fiji and Niue*; and

Perkinsiella vitiensis, which attacks sugar cane and is recorded from the New

Hebrides (Muir, 1931, p. 72), Fiji, Samoa and Niue * (ui, 11). In the same group

is the Aleyrodid, Alewroplatus samoanus, described by Laing as endemic, but

since found in Tahiti (Dozier, 1928); this insect feeds upon cultivated Croton

(Codiaeum), a plant widely used in Polynesia and Melanesia for decorating

graves and in other ritualistic practices. The relation of two Ortalidae, Scholastes

emctus and bimaculatus, to the life of Polynesians is rather different. These

flies breed in the rotten meat of coconuts which have been cut for drinking

while they were green, and the larvae can exist for many weeks without food

if they consume all the rotten material in an unripe nut. They are therefore

well suited for travelling the ocean with Polynesians. I may add that they do

not breed in the kernel of ripe nuts, or in copra, so that their distribution cannot

be related to Huropean commerce. Cassida strigula, associated with an old

food-plant, the sweet potato, may also be an ancient introduction, though

Maulik attributes it to modern commerce: the same may be said of the sweet-

potato weevil, Cylas formicarius. But one must be cautious of attributing

all pests of old economic plants to human introduction ; for instance, the hawk

moth, Hippotion celerio, is a creature of enormous powers of flight and very

wide range ; it has many unrelated food-plants ; and though it is a pest of

taro in Samoa we may well think that it came to the archipelago by natural

means.

These are also a few parasitic insects, of man himself or of animals which the

early Samoans are known to have kept; in all probability these insects were

brought to Samoa by canoe, though one must be cautious, for evidence suggests

that bedbugs and human fleas were brought to these peaceful isles by the

* The occurrence of these insects in Niue is significant ; this island is small and remote,

and landing from ships is not easy; the island has no economic importance to the European,

and is only visited a very few times in a year. This confirms the view that the insects are Poly-

nesian introductions,

58 INSECTS OF SAMOA.

European. To native introduction one may perhaps attribute Pediculus

humanus and Phthirus pubis of man; Ctenocephalus felis, a common parasite

of dogs in the tropics; also three species of Mallophaga from domestic hens.

I have already referrred to the pig-louse (vu, 86). At the time of writing it

was held that the specimens belonged to a variety (Haematopinus suas adventicius)

characteristic of the wild swine of south-eastern Asia. This appeared to give

interesting and unexpected corroboration of the view that the Polynesians are

derived from that area. But the recent and very careful investigations of

Ferris (1933) show that H. sws is extremely variable, but lend no support to

the theory.

In considering whether a particular insect is probably of recent introduction

or not, one may sometimes gain assistance from the systematic position of its

host plant. European food-plants are predominantly Cruciferae, Leguminosae

or Gramineae, but the Samoans did not eat members of any of these three

great Natural Orders. To this there is one apparent exception, for they grew

sugar-cane, though mainly as a thatch plant. Pests of sugar, such as Rhabdoc-

nemis obscura, may, therefore, have been introduced by the Polynesians or the

European. But pests of stored cereals, for instance Calandra oryzae, are

clearly due to European agency. So, in all probability, are the pests of growing

bean plants, for instance Maruca testulalis and Nacoleia diemenalis ; both

these Pyraustine moths have a very wide distribution through India, Malaya

and Melanesia, and attack growing Leguminous crops.

Furthermore, there are certain insects associated with the coconut, about

which it is difficult to feel any certainty. There is, of course, no doubt that the

palm was used and carried about by natives of Polynesia and Melanesia, indeed

to many of them it was the staff of life. Moreover, they used it and carried it in

a variety of states, for they drank from young green nuts, and used the ripe

kernels, and they also ate the large spongy embryo inside nuts which had already

sprouted: there were doubtless facilities for the carriage of insects as well as

the plants. This is all well known, but we do not know whether the coconut

is truly native of Samoa, having come floating over the sea and established itself

before ever man reached the archipelago. For what it is worth, the evidence to

be derived from insects suggests that coconuts are ancient natives of Oceania, as

local endemic insects have developed, apparently, on this particular palm.

For instance, there is a genus of Hispine beetles, Promecotheca, distributed through

the Malay Islands and the Philippines to New Guinea, North Australia, Melanesia

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 59

and parts of Polynesia. A number of species are known, each one confined to a

part of the area. The Samoan species (P. reichet), which is the most easterly

in its range, is found in Fiji, Samoa and Tonga. All the species are attached to

the one host plant, the coconut, except one from Zingiberaceae in the Solomon

Islands. These facts appear to suggest that the palm is truly native in the

islands mentioned above ; were it only a cultivated plant, carried about by man,

it would surely be attacked by a single, widely spread species of Promecotheca

(for facts, see Maulik, iv, 210; also Maulik, 1929a and b). Another coconut

insect is the Phasmid, Graeffea crouanz, but it is of little value in the present

discussion, because the systematics of the species which attack the coconut are

obscure. Many other insects are known from coconut palms in Samoa, but the

majority are probably of recent introduction (Laing, ii, 37; also Doane and

Ferris (1916) ; also Swezey (1924)).

The insects of the banana and other plants might be studied in a similar

manner.

The distribution of races of two species of Aédes (Culicidae) shows anomalies

which appear to be due to the voyages of the Polynesians ; the cases are im-

portant, as we have so much knowledge about mosquitoes in Oceania. Aédes

scutellaris * is known to breed in small collections of water, including rain-water

which has accumulated in a canoe, and its egg is resistant to drying. It has an

enormous distribution in Indo-Malaya and also in Oceania, including many atolls ;

and it has developed a number of insular races, differing from one another both

as adults and also as larvae.t In the area which is our concern, we find the

typical form in New Guinea, the Bismarcks and the Solomons; var. hebrideus

occurs in the New Hebrides, and probably in the small archipelagos to the north ;

var. pseudoscutellaris is known from Fijit and the Ellice group, and thence

extends through Samoa to the Tuamotus, that is to say, to the eastern limit of

Polynesia ; the range of this last variety is partly interrupted by that of var.

* It appears that the insect referred to elsewhere in this work (vi, 44) as Aédes (Stegomyia)

variegatus Dol. must now be called scwellaris, Walker. The change is regrettable, for scwtellaris

Theo. is a synonym which has been used for a different species, albopictus (for synonymy see

Edwards, 1932a).

+ The distribution of the forms of this interesting insect and of Aédes kochi was shown on

maps by myself and Hopkins (1927); we also summarized what was known at the time of our

publication, and gave a list of references. For recent additional information, see Paine and

Edwards (1929) ; also Edwards (1929a).

{ Specimens from Rotuma appear to combine the characters of this variety and of hebrideus

(Edwards, 1929a, p. 338).

60 INSECTS OF SAMOA.

tongae, which occurs in Tonga, including the northerly outlying islands of the

croup. Tongae also occurs in the island of Sikiana, im the Solomon Islands,

so that its range is interrupted by an interval of 1,500 miles, in which two other

varieties (pseudoscutellaris and hebrideus) occur. This astonishing anomaly is,

I think, explained by the fact that the Sikiana people are not related to other

Solomon islanders, for they are purely Polynesian. There is evidence to support

the view that, like the people of Tikopia and a few other islands on the easterly

fringe of Melanesia, their ancestors drifted down wind from Polynesia and

in their canoes they brought the Tongan race of this mosquito. It seems probable

that the other races, and particularly var. pseudoscutellaris, owe their wide

extension in the Pacific to carriage in native craft. The egg and the larva are

well suited to such voyages, and until the European broke up Polynesian society

there was much coming and going in great canoes which were capable of long

ocean voyages.

In the same genus there is a second species, Aédes (Finlaya) kochi, the

distribution of which is probably partly due to human interference. Two forms

of this insect are known, distinguishable in the adult, the larva and to some

extent the breeding place ; the typical form is found in New Guinea, the Bismarck

and the Solomon Islands ; it seems to be absent from the New Hebrides, where

I searched for it on many islands, but it occurs again in Fiji; in the Bismarcks

it breeds in empty coconut shells, but in New Guinea, according to Taylor

(1934b), it breeds almost exclusively in the axils of Pandanus, and occasionally

taro (Colocasia). The variety samoana occurs in Samoa and Tonga. It breeds

almost exclusively in the axils of taro, an aroid which was and is an important

article of food; even to this day, the roots, with the leaf bases on them, are

frequently carried on sea voyages by Samoans, and they are transported raw

to prevent decomposition : and as the egg of the insect is resistant to drying,

everything favours distribution by human agency. This form has also been

recently sent from Roviana, Solomon Islands, by Dr. HE. G. Sayers. At Roviana

the mosquito was captured on the verandah of a mission hospital. The mission

at that place was established 40 years before, and a considerable number of

Tongan teachers had been employed there, particularly in the early days. It is

easy to think that the Tongan mission teachers brought taro plants and eggs of

A, kocht var. samoana to Roviana.

If one studies the distribution of many of the insects which were probably

transported by early man, curious anomalies appear. For instance, neither of

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 61

the Fulgoroids associated with taro and sugar (Megamelus proserpma and

Perkinsiella vittensis, p. 57 above) has reached the Hawaiian Islands; had

they done so, they could not have escaped the economic entomologists of

those islands, and in particular the late Mr. F. Muir, the authority on this group

of insects. Rather similarly Aedes kocht samoana, which breeds in water in

the axils of taro, seems to be absent from the Society Islands and the Marquesas,

in both of which it has been looked for: its absence from atolls (Ellice, Tokelau,

etc.) is not surprising, for the taro which grows in their miserable soil is a different

type, which does net hold water in its axils.

(2) Introduced Insects of Uncertain Status.—lI have

defined a group of insects which were probably introduced by the early Poly-

nesians ; they are obviously associated with the native life, and many of them

show curious anomalies in their distribution. There remains a large group of

insects, which may be truly native or may have been introduced either by

Samoans or by Europeans. The difficulty of deciding may be illustrated by

the cockroaches. Apart from the endemic species there are several cosmopolites

(Blattella germanca, Periplaneta australasiae, P. brunnea, Pycnoscelus surma-

mensis) ; it is easy to attribute their presence to European ships. But how

are we to treat Hoblatta notulata, Cutilia soror and C. nitida ? These species

have a wide distribution in the Malay Islands, and extend thence into the tropical

Pacific: they may have reached Samoa by natural means, or with the

Polynesians, themselves native of 5.1. Asia.

The same difficulty is frequently encountered in other groups, for instance,

the earwigs and the many Coccids ; it is also impossible to feel any certainty

about the means of arrival of the two species of scorpions, both of them widely

distributed in the South Pacific. There are several small moths, such as

Setomorpha rutella, which we bred from chicken dung and also from bread-fruit

which had rotted and then dried, and which are known to breed in many sorts

of dry refuse ; such a species may be native, or may have reached Samoa by

artificial means at any period. In the same ambiguous group one must place

Pyralis pictalis and manthotalis, both bred from refuse in fowl-houses. One

encounters the same difficulty among the Drosophilidae : several (Drosophila

anpelophila and errans, Spinulophila nasuta) have a very wide tropical distribu-

tion in both hemispheres. One is inclined to attribute it to Huropean shipping

till one notices that other members of the family (Mycodrosophila gratiosa,

Hirtodrosophila semingra) seem to be native, occurrig in Malaya and in the

62 INSECTS OF SAMOA.

Samoan hills: but if these arrived by natural means, why not the other, tropico-

politan, species ?

The dung-breeding insects are also of doubtful status. At first sight it

seems likely that they arrived in ships, particularly if live cattle or other domestic

animals were carried, and many of them are probably of recent European

introduction. But others may have accompanied Polynesian travellers, and

others again may be truly native: one must remember that dung of one sort

or another must always have been available in Samoa. However they came,

the dung-breeders are numerous and represent many groups of insects, e.g.

several Sphaeroceridae ; Malichiella lutecpenmis ; that glorious Syrphid, Volucella

obesa; many Muscidae, etc. (Musca, Stomoxys, Sarcophaga, Chrysomyia) ;

also the small chafer Aphodius.

(3) European Introductions.—tThe presence of many sorts of

insects in Samoa is almost certainly due to European commerce. This is

particularly clear if the species attacks some commodity which was not used by

Polynesians in the old days. As instances one might quote the beetles eating

dried foods : Triboluum castaneum, Araecerus fasciculatus, Carpophilus hemipterus,

C. dimadiatus, Tenebrordes mauritianicus ; also such moths as Corcyra cephalonica,

Ephestia cautella; the wood-eating beetles, Lyctus brunneus and Muinthea

rugicollis ; also beetles which are predatory on some of the above, e.g. three

species of Cleridae. It also seems probable that insects which attack fabrics,

feathers and skins have been recently brought in: such are Dermestes vulpinus

and Tinea pachyspila. Another group are the parasitic insects, mostly itro-

duced with Huropean domestic animals or foreign rodents : Nenopsylla cheoprs,

from foreign species of Rattus and from Rk. (Mus) exulans, the native rat, an

animal which was carried in canoes and which appears to have no fleas of its

own in Samoa; the human flea, Pulex irritans, apparently not present till the

days of European shipping; a sucking louse, Polyplax spinulosa, from ships’

rats; the tropical bedbug, Comex hemipterus (11, 160 ; also Buxton and Hopkins,

1927, p. 54). The Lygaeid, Clerada apicicornis, which has a wide tropical

distribution and is found associated with rats’ nests, may fall in the same group ;

but it is equally likely that 1t toured the Pacific in canoes with Rattus erulans.

The hive-bee, Apis mellifica, is now common in Samoa, Tonga and other

sroups. As I can find no reference to bee-keeping among the Polynesians in

the old days, I assume it to be a recent introduction.

There are also many agricultural pests, clearly of recent introduction. Of

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 63

these the most interesting is Oryctes rhinoceros, because the story of its spread is

so fully told by Jepson (1912), Friederichs (1913) and Friederichs and Demandt

(1922). It seems that larvae were introduced to Samoa, from Ceylon, in boxes

of soil in which rubber trees were growing ; the introduction occurred at Apia in

November, 1910. Extension of the insect’s range was very rapid, and within

little more than a year it had occupied the coast of Upolu westwards, for about

60 miles ; eastwards, that is to say in the direction from which the Trade Wind

blows, it had only spread about 8 miles in the same period. The beetle became

an extremely serious pest of the coconut, the source of the agricultural wealth of

native and European alike, and so it still is, though partial control was achieved

by hand-collecting. The system has already broken down more than once, at

times of epidemic disease or political unrest. The introduction of Scolidae,

with a view to natural control, has been discussed many times since 1912, but I

believe that nothing has actually been done.

‘Many of the other pests of crops are certainly of recent introduction. — It

seems unnecessary to name them here, for Hopkins (1927) has dealt with many

of these insects and given a full list of earlier papers.

I cannot leave the subject of introduced pests without calling attention to

the grave danger that many other insects will reach the Samoan Islands. When

we lived in Samoa, in 1924 and 1925, every incoming boat used to bring fresh fruit,

some of it grown in New Zealand, but most of it from New South Wales. Most

of the fruit could well have been grown locally, and the trade was small and only

met the needs of the more luxurious Kuropeans. But though I pointed out

the danger, the Administration was too timorous to stop this importation, and I

felt little doubt that the island would be colonized by the Mediterranean fruit

fly (Ceratitis capitata) and other Trypetidae. Several troublesome species of

Dacus have already been introduced. One should also mention the danger of

bringing in timber-pests ; at the time when we were in Samoa, it seemed that there

were no foreign termites, and that the indigenous species were well-nigh harmless ;

there were also no tropicopolitan Cerambycids among the 35 species in the

collections. Another insect which menaces the islands is Anopheles punctulatus.

This species, which is a well-known carrier of malaria, occurs in Melanesia, as

far east as 170° Hi. ; it is absent from Fiji and all Polynesia (Buxton and Hopkins,

1927, p. 67). There is a risk, though it is not great, that it may be introduced

into Samoa, where its presence would presumably cause a great epidemic of

malaria. In the event of its reaching Apia, it would find that everything possible

64 INSECTS OF SAMOA.

had been done to ensure its establishment, for the streams which run through the

town have been dammed to make swamps for taro. It may be expected that

many other troublesome insects will eventually arrive. Piophila casei 1s not

yet known to occur, though it is recorded from Fiji by Bezzi (1928). The two

species of Dysdercus which are recorded are only known to occur in Fiji and

Samoa; in spite of the fact that some cotton and kapok is grown none of the

more widely distributed and harmful species has yet been found in Samoa.

At the time when we were in Samoa it was useless to point out these dangers,

for every one said that if the pests could have come they would have done so

long ago; also that if they are coming, nothing would stop them ; also that no

one really knew anything till he had lived 30 years in the islands (the converse

being more nearly true). I feel confident that the Samoan Islands will eventually

possess all the cosmopolite pests which may to-day be seen in Oahu or in New

Caledonia.

The insects mentioned above have come in with European trade or agri-

culture, but the white man has brought in many others quite casually. Among

them are all the Samoan Vespids (Polistes macaensis, and two species of

Odynerus) ; also perhaps the bee, Lithurgus scabrosus. The two bees of the genus

Megachile are more probably native, though members of this genus are so well

known on ships; for both the Samoan species have developed endemic races,

in some of the Samoan Islands and also in other parts of the Pacific. The pre-

datory ants, Pheidole megacephala and P. oceanica, are also casual introductions,

as are the mosquitoes Culex fatigans and Aédes argenteus, the history of which

in Oceania is described elsewhere (Buxton and Hopkins, 1927, pp. 83-85, 113-

115). The dung breeders, many of which are probably recent immigrants, have

already been considered. Berland (vill, 35) mentioned two widely distributed

species of spider which were not recorded from Samoa by two earlier writers,

but which occurred in our material. These, as he says, are doubtless very recent

introductions, and he mentions another species which may yet be expected to

follow them.

(4) Foreign Predators.—tThe general impression that the visitor

carries away is that, in the hills, the native flora and fauna of Upolu and the other

Samoan Islands is flourishing and in little danger of extermination. But round

the coast, and particularly in the neighbourhood of Apia, which has always been

the main harbour, the situation is different. For one thing, the Samoans live,

and always have lived, along the beach, where they have their villages and most

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 65

of their plantations. The Europeans and half-castes live in the same zone,

principally round Apia. Here, therefore, settlement has dispossessed the native

flora, and the insects dependent on it. . But it seemed that introduced insects

were also playing a part, for even when one could find patches of forest near the

coast, as in Luutuanuu, Aleipata and elsewhere, the insects were few and un-

interesting. This may be due to the introduced wasp, Polistes macaensis, and

to various ants, of which perhaps the most important are Pheidole oceanica and

P. megacephala ; both these insects were collected as high as 2,000 feet at

Malololelei, on Upolu. It will be remembered that Perkins (1913) found the

latter species to be the most abundant of foreign ants in Hawaiia, and blamed

it for the disappearance of much of the native fauna ; he said, “ save for a few

forms that can resist, or are tolerated by the ants, the endemic fauna is entirely

exterminated. This native fauna, especially the beetles, appears as if by magic,

the moment the limit of range of Phezdole is reached.” In Samoa we did not form

the impression that the case was so desperate, but it 1s impossible to be certain.

There seems to be only one insect about whose disappearance definite information

exists: the Geometrid moth, Cleora samoana, the extreme variability of which

makes it one of the most interesting insects in the country (111, 158 and Plate V)

was collected by Rechinger in 1905 at Malifa and Motootua, which are just outside

the town of Apia, within a couple of hundred feet of sea-level (Rebel, 1910).

Hopkins and I lived and worked for two years in the same neighbourhood,

in which we never observed this moth. The lowest point at which we obtained

it, in 1924 and 1925, was Vailima, 600 feet above sea-level. Here and also at

Malololelei, 2,000 feet, it was common ; it is to be hoped that it may remain

so, or at least that we may from time to time learn more of its status.

In the systematic part of this work, particular attention is given to precise

localities and dates of capture, because these may eventually give information

about the status and fate of the native fauna.

Taxonomic NOTES

Tue Enpemic Insects.—The naturalist who works on an island remote

from large areas of land has at least one question always in his mind: is the

fauna around him continental or oceanic ? He may obtain a partial answer to

this question by remembering that there are two characters of insular faunas :

they show a high proportion of endemic species and frequently of endemic genera ;

Ix 2 3

66 INSECTS OF SAMOA.

they are also remarkable for the absence of larger or smaller groups widely distri-

buted in the world and found in neighbouring continents.

(1) The Proportion of Endemic Species—lIt appears

from Table I that 1,603 species are known to occur, of which 787 (= 49%) are

endemic. It is clear that the proportion of endemic species is high in the insects

of all orders : this suggests that the Samoan fauna 1s oceanic, a conclusion already

reached after considering the flora, and the fauna in general (1x, 24).

It must be realized that when our knowledge of the insects of the south

Pacific is more complete, a number of species now thought to be endemic in

Samoa will be found to occur in Fiji, Tonga or elsewhere. On the other hand,

further collecting in Samoa will certainly increase the list of the endemic fauna,

many of the members of which are small and difficult to obtain. We may

perhaps conclude that the proportion of endemic forms will not be greatly altered

as knowledge increases. This has actually occurred in the Hawaiian Islands.

Perkins (1913) gave a list of 3,325 species known to occur, of which 82% were

thought to be endemic. With the passage of years the total number has been

raised a full third, to 4,620, but the endemic proportion is 81° (Bryan, 1934).

The proportion of endemic species in the whole insect fauna of other tropical

archipelagos is not known, except for Seychelles: according to Scott (1933),

2,090 species of insect are known of which 65% (1,366) are endemic. It would,

of course, be possible to gather together a large body of fact about the proportion

of endemics in particular orders or families in insular faunas; for instance,

of the 239 Brachycera and Athericera of Fiji, 60-5°% (144) were endemic (Bezzi,

1928). But I have not put myself to the trouble of obtaining such figures.

But the proportion of endemic species among the insects as a whole, or

even in particular orders, is a crude figure, for the different families, etc., can

hardly be compared with one another. For instance, we find (Table 1) that

58%, of the Diptera are endemic, and among the 100 Nematocera (vi, 24) the

proportion is almost identical (60°,) But if individual families are considered,

one finds that only 1 out of 7 Culicidae is endemic, whereas 23 out of 33 Tipulidae

(70%) and 10 of the 16 Ceratopogonidae (60°%) are unknown from other parts

of the world. The same holds good of the Neuroptera (vu, 89). Of the 5

Samoan species of Hemerobiidae, 4 are endemic and belong to endemic genera.

In contrast there are 9 Chrysopidae, of which 1 is the sole representative of an

endemic genus. The 8 others are referred to the widely distributed genus

Chrysopa ; 2 of the 8 seem to be endemic, and 2 others seem to be peculiar to

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 67

TABLE 1.

Showing Orders of Insects recorded from Samoa; the Number of Samoan Species in each ;

and the Number of Species which at present appear to be endemic.

ENDEMIC ENDEMICS

ORDER TotaL SPECIES

SPECIES Per cent.

Thysanura . : 2 : 5 4 2 —

Collembola . : : ; ; 4 1 —

Orthoptera . : : : ; ae 35 45

Dermaptera 5 3 : ‘ 12 +) —

Isoptera j : 3 : : is 6* —

Psocoptera. : : 11 7 —

Anoplura (with Mallophaga) : : 14 1 —

Ephemeroptera ; : 1 1 —

Odonata 5 ; 5 ‘ ’ 28 12 43

Thysanoptera ; , : ‘ 3 3 —

Hemiptera. 5 : j : 247 127 51

Neuroptera . : : : : 15 7 —

Trichoptera . : i . ; it 1 —

Lepidoptera . , : : : 442 191 43

Coleoptera. ‘ ; 385 202 52

Hymenoptera (not Parasitic) : j 57 15 26

Diptera 5 5 : ; ‘ 292 171 58

Aphaniptera . : : : ‘ 3 0 —

Total : : 3 ; ; 1,603 787 49

* One undescribed.

the Central Pacific: the other 4 have a wide distribution, and several occur in

Australia. Among the Myrmeleonidae, very few occur in Polynesia, but 2

of them have a very wide distribution. The fact that species of Chrysopa and

Exdoleon range so widely, and that they appear to be able to cross great stretches

of ocean is unexplained ; indeed, I do not think that attention has previously

been called to it. Another group of creatures in which endemicity is high,

but not equally so in all families, is the Araneida, already discussed (p. 45

under “‘ wind ’’).

In certain groups of insects, and particularly in those which possess great

powers of flight, the proportion of endemic species is low. This is well seen in

the Odonata. Ten of the 15 large and powerful Anisoptera have a wide distribu-

tion outside Samoa, and several of them are known to occur on other remote

68 INSECTS OF SAMOA.

islands : but among the more delicate Zygoptera, 10 of the 13 species are peculiar

to Samoa. Moreover, in several families of Diptera Brachycera and Athericera,

especially those which contain robust species of powerful flight, the proportion

of endemic species islow. This is illustrated by the following numbers :—

APPARENTLY KNOWN FROM

TOTAL.

ENDEMIC. ELSEWHERE.

Stratiomyidae 7 3 4

Tabanidae . 1 1 0)

Asilidae : ; : , : 4 2 2

Syrphidae : : : ; . 10 1 9

Total . : : : ; k 22 a 15

(2) Production of Endemic Complexes.—Among the insects

found in oceanic islands, two related phenomena have occasionally been noticed.

There may be very great variability within the limits of a single species on one

island: there is also a tendency for the production of groups of closely related

species, some or most of them peculiar to single islands within an archipelago.

We have seen that the proportion of endemic species is high among the

Samoan insects, and it would be interesting to find out whether either of these

other phenomena is observable. It seems certain that examples of great

variability are few, with the exception of that wonderful Geometrid, Cleora

samoana ; this insect exhibits an extreme degree of variation in colour, pattern

and size; moreover, teratological varieties appear to be unusually common

(Prout, ii, 158). Though much attention was devoted to the variation of this

species, it was not found that it was associated with season or with place : indeed,

all the insects figured on Plate V, Part ili, were taken on one night in one house.

It also appears that little if any of the variation is due to sexual dimorphism.

Two other examples of variability are commented upon in the text: the Lycid,

Samoaneros acuticollis, and the Chrysomelids of the genus Stygnobia. Maulik

(iv, 191), after studying a long series of Stygnobia, calls attention to the fact that

in an insular fauna a species may be more plastic than in a continental area.

The phenomenon has been observed in several groups of Hawaiian insects and

is mentioned in the summary of the Fauna Hawaiiensis (Perkins, 1913). It is

possible that in the rather simple fauna of an island some type of competition is

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 69

less than it would be in a continental area, permitting species to develop a high

degree of variability. In Hawaua, but not apparently in Samoa, variability

is associated with the formation of groups of closely related species.

As to the other characteristic of island faunas—the production of large

groups of closely related, endemic, species—the Samoan fauna provides few

examples. For instance, among the Microlepidoptera (in which two-thirds of

the total species are endemic), the 90 endemic species are distributed among 54

genera, so that the number of species per genus is low : moreover, the 8 endemic

genera are all monotypic, and they belong to 6 families. There are, however,

2 genera which show some development of endemic species ; 15 of the 17 Labdia,

and 7 of the 9 Decadarchis, are endemic. The figures for the Nematocera are not

dissimilar: the number of species per genus is under 2: there are only 2*

endemic genera, each of them monotypic. It should, however, be mentioned that

there are 9 species of Tvrentepohlia (Tipulidae), 8 of them endemic. Similar

figures could be given for other groups of which we have adequate knowledge.

In the Curculionidae the 86 species are distributed among 55 genera. One

genus, Trigonopterus, has 9 species, another, Cossonus, has 4 ; all the species in

both genera are endemic.

There are two or three groups of endemic species which are sufficiently

striking to call for mention, though the number of species is not very great in

any of them. The smaller dragon-flies were carefully collected at least in a part

of Upolu; among the Coenagrionidae there is an interesting development in

Ischnura and related genera (Amorphostigma, Pacificagrion). Ischnura contains

5 Samoan species, 1 (J. aurora) which is known to be distributed by wind (p. 44)

and ranging widely, and four which are endemic. Amorphostigma contains

2 species, both endemic, and Pacificagrion 1, also endemic. The interest of the

group is increased, for Fraser says that certain archaic features are discoverable

in them. The rather similar development of the Agrionid Nesobasis in Fiji

should be remarked (Tillyard, 1923). The Samoan Halictine bees are interesting

inarather similar way. As at present classified, there are 4 closely related endemic

species in Halictus, and 2 in the endemic genus Echthralictus. These particularly

interesting insects are discussed elsewhere by Perkins and Cheesman. Here it

seems sufficient to call attention to the need for collecting them in Fiji and

* When Edwards’ account of the group was written there were 3 endemic genera. One,

Pontomyia, has since been found in Japan (Tokunaga, 1932).

70 INSECTS OF SAMOA.

Polynesia, and to the very interesting biological question which is suggested by

the structural characters of Echthralictus. Another small group of endemic

species 1s found in Mordellistena (Mordellidae): in Samoa there are 2 species

which range widely in Fiji or Melanesia, and 5 endemics, each of them already

known from Upolu and Tutuila.

It seems then that complexes of closely related endemic species have not

developed in Samoa. But the Samoan fauna contains representatives of groups

which have developed endemically in central Polynesia and are characteristic

of that area. One might quote the Cicadellid genus Jassoidula, described in

the present work as new, to which Osborn attributes 8 species: 4 are endemic

to Samoa (1 Upolu; 1 Upolu and Savaii; 2 Tutuila), 3 are endemic to Tonga

(1 each on Vavau, Haapai and Nukualofa), and 1 to Niue. Numerous other

examples can be found in the pages of this fauna; for instance, the Lycid

genera Melaneros and Samoaneros ; also the Hydrocampinae, Piletocera and

Bradina.

(3) Peculiar and Anomalous Genera.—An island fauna

might demonstrate its antiquity by preserving very ancient types, or by develop-

ing endemic and highly peculiar forms of life. We may say at once that there

are no living fossils among the Samoan insects, or indeed among the other animals

or the plants. But several peculiar endemic genera have been discovered in the

collections. Much the most remarkable is the submarine Chironomid fly Pon-

tomyia. The anatomical characters of this insect were fully described and figured

by Edwards (1926b) ; briefly, the male has greatly reduced paddle-like wings,

with what appears to be a hinge across the middle of them; it shows several

other peculiarities, any one of which would render it almost unique in its family.

The female is cylindrical and entirely without mouthparts, antennae, front legs

or wings; the middle and posterior legs are greatly reduced ; the insect is as

much simplified as any known Dipteron. We were so lucky as to collect a large

number of adults, and also what appear to be the corresponding larvae and pupae.

The early stages were found among beds of Halophila, a flowering plant (Hydro-

charitaceae) which grew in sandy patches in the shallow sea inside the coral

reef. The adults were obtained by tow nets, after dark, and appeared to exhibit

lunar periodicity. Such biological observations as I could make, and a discussion

of the failure of the Insecta to colonize the sea, have been published elsewhere

(Buxton, 1926). Since that date a very similar member of the genus, till then

supposed to be entirely peculiar to Samoa, has been discovered by Tokunaga

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 71

(1932) in Japan. This author has been able to add greatly to what is known

of the structure and habits of Pontomyia. He confirms my observation of the

existence of lunar periodicity, and says that copulation occurs on the surface

of the water, the female laying all her eggs at once and dying within an hour of

emergence. On several points Tokunaga’s observations cannot be reconciled

with my own, which is remarkable, for the anatomical specialization of both

Species 1s so extreme that one would expect them to live nearly identical lives.

Among other matters Tonkunaga says that the male of the Japanese species

uses its wings only for gliding on water, and he adduces evidence which appears

to show that its unique structure fits it for this purpose and no other ; whereas

I was never able to see the Samoan species use its wings, though I observed that

the long first and third legs were used for propelling the insect beneath the water.

Moreover, the Samoan species was caught alive, in tow nets, and it was frequently

found in the lagoon when the tidal current was so strong that one could not

swim against it ; it was associated with patches of Halophila, in which the female

pupae were discovered. Whereas the Japanese species appears to have no

association with the plant, and to live in pools between tide-marks. It seems,

therefore, that in spite of similarity in structure, there are considerable differences

in biology between Pontomyia natans in Samoa and P. pacifica in Japan, and that

the former is the more oceanic or planktonic in its habits. It is perhaps a little

disappointing that Pontomyia, which appeared to be the peculiar glory of Samoa,

should be re-discovered in Japan: it is presumably a widely spread insect,

which has escaped detection because of its peculiar habits.

The present work contains descriptions of several other insects which are

remarkable in structure or puzzling in their systematic position. For instance,

the Fulgoroid genus Buxtoniella, described by Muir (ii, 23) to contain 2 endemic

species, was placed in the Lophopidae ; but its position was an enigma, the

external characters suggesting an Issid, the genitalia of both sexes showing

Lophopid characters. Subsequently, the same author (Muir, 1931, pp. 78, 79)

described the genus Painella, containing 2 species from the Solomons: this

genus resembles Buatoniella, but the affinities remain obscure. Several other

Hemiptera of obscure affinities have been discovered in Samoa; for instance,

the Lygaeid Bryanella, which appears to be intermediate between the Myodochini

and the Rhyparochromini; and Moana, a Cicadid of which the venation is

extremely aberrant.

One might quote other genera, the systematic position of which is obscure,

72 INSECTS OF SAMOA.

But none of the specialists who have contributed to our knowledge of the fauna

has found it necessary to erect a new tribe or subfamily for the reception of a

Samoan insect: they describe many endemic species and genera, but nothing

of higher taxonomic rank.

(4) Species confined .to .Parts of the Samoan

Archipelago.—lt is tempting to endeavour to assess the age of the Samoan

fauna by discovering how many of the endemic species are widely spread within

the archipelago, and how many are peculiar to one or two islands. But at present

this is difficult, for only in a few instances do we possess enough evidence. It has

to be remembered that most of the collecting on different islands was done by

different men, and the apparent absence of an insect may often be due to this

personal element. One observes, for instance, that the Nematocera and Micro-

lepidoptera were thoroughly worked by Mr. Hopkins and myself on Upolu,

very little on other islands. Generally speaking, one can attach most importance

to absence from Upolu, for its insects were collected by many men, and at all

times of year.

But there is at least one group of insects, the butterflies, on which we may

feel that conclusions are solidly based. This is partly because good material

is available from all the larger Samoan Islands and from other archipelagos,

and partly because the principal collector of them, Mr. Hopkins, is also the author

of the systematic paper. From his work (Table 2), it appears that 3 of the 28

TABLE 2.

Showing Distribution of Races of three Species of Butterfly, within the Samoan Archipelago.

The + sign signifies that the race is present, the 0 that it is very definitely absent.

Upo.uv. Savall. TUTUILA. Manva.

Danaida melissa melittula : ++ +--+ 0 0

Pe 1 tutuilae , 0) 0 ++ 2

Hypolimnas bolina inconstans . + ++ 0 0

A a pallescens . 0 0) ++ +

Precis villida samoensis . ; ++ ++ 0 0

» villida ; : 0 0 +--+ +

species of butterflies which occur in Samoa have produced races endemic to

certain islands, In each case the race in Upolu and Savaii (Western Samoa) is

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 73

the same, and differs from that in Tutuila * ; the Tutuilan race is already known

to occur in Manua in two of the three species. Danaida melissa has produced two

endemic races in Western Samoa and Tutuila: Hypolimnas bolina and Precis

villida have developed endemic races in Western Samoa, but the forms in

American Samoa are found in Tonga and elsewhere. We must return to the

Samoan butterflies when we discuss the geographical relations of the insect

fauna. For the moment it suffices to remark on these three species, which

have developed island endemicity.

In the Heterocera there are one or two endemic species which occur only in

a part of the Samoan Archipelago. The conspicuous Lithosiid, Chrysaeglia

samoensis, is common on Upolu, and has been collected on Savau; but it has

never been taken on Tutuila, and it seems highly probable that it does not occur

in American Samoa. It appears almost certain t that the two species of

Deilemera (Arctiidae) have a similar distribution. D. alba is a white day-flying

moth, abundant and very familiar on Upolu and Savaii. In the course of one’s

first walk on Tutuila, one remarks the absence of the white insect, and the

abundance of the black and white D. mundipicta samoensis, which also occurs in

the Manua group, but is unknown in Western Samoa. In the field the difference

is extremely striking.

In order to show that species endemic to single islands have not frequently

developed, Table 3 has been prepared: I have taken some trouble to collect

these examples. At first sight the table appears to show much evidence of

island endemicity, but the reader will see that many insects which appear to be

confined to one island have been collected on a single occasion only. With

increased knowledge the proportion of island endemics would probably become

less. Among the insects shown in the table it seems that good evidence of island

endemicity is only shown by Stygnobia aenescens, Trigonopterus submetallicus,

Megachile scutellata, Eucarobius spp. The table might be extended to include

the following and others :—Anaxipha (Gryllidae ; four species on Upolu, Savaii

and Tutuila, five species from a single island: it should be observed that the

* Upolu and Savaii, with Apolima and other small islets, lie within one 100-fathom line.

But east of Upolu the bottom falls to very great depths, nearly all the interval between Upolu and

Tutuila being deeper than 1,000 fathoms. There is a similar deep channel between Tutuila and

Manua. Contrast Fiji, almost the whole of which is within one 1,000 fathom contour; Tonga is

substantially similar.

{ There is a single record (iii, 193) of D. alba from Ofu, Manua group. I am personally

convinced that this must be due to erroneous labelling, either in the field or the museum.

74 INSECTS OF SAMOA.

TABLE 3.

Showing the Distribution within Samoa of certain Groups of Species which exhibit island

Endemicity. The Distribution is given under the separate Islands, +--+ indicating that a Species

was collected on many Occasions, + on several Occasions, and 1 once only.

Upouv. SAVAIL. TUTUILA. MANUA.

CHRYSOMELIDAE

Stygnobia aenescens : : : +--+

a cauta : P : : +--+ 1 ++ 1

5 variabilis : : f +--+ +--+ ++

a minua . ; F j + +

CURCULIONIDAE

Sphaerorrhinus puncticollis * + 1

Acicnemis eludens + : : : 1 ae

Trigonopterus crinvpes—. ‘ ‘ +4. --

- submetallicus : +

zs 5» mendax . ++ 1

FA caesipes —. : ; 1

5 aeneoniveus ; 5 qe 1

. bicolor F : ; +

a serratipes . : : 1

si bryant A : A 1

samoanus . : : il

i binotatus : . 1

Cossonus platyrrhinus . : j 1

53 dentipes . ; : : + 1

es limbaticollis . é : 1

- quaerens . : : : 1 1

os » afonus 3 ; 1

MEGACHILIDAE

Megachile diligens armstrong. . 5 Sn

a » edlens t : ; 1

a » buxtons t

re scutellata wilmattae . : +-+

>; s4 tutuilae ; ; +

HEMEROBIIDAE

Eucarobius fasciatus F : : +--+

~ graeffer : ; ; 1 1

i oblongus ‘ : ; --+-+ +

* The three specimens from Manua differ from all those from Tutuila in colour of scales.

{ The single specimen from Upolu differs conspicuously from the three from Tutuila ; but

on this material one cannot say whether the variation is random or insular.

{ The distribution of these forms leaves much to the explorer of the future. The record

of calens from Tutuila rests on a single specimen, and though it seems absent from other Samoan

islands it is recorded from Fiji and Ellice Islands. Buztoni is described from a single male from

Manono, but is probably not confined to this small island which hes within the main coral reef

surrounding Upolu.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 75

Gryllidae of American Samoa are most imperfectly known): certain Droso-

philidae (endemic genera, Samoaia and Upolumyia, and endemic species in

Drosophila and other genera): Prosoplus (Cerambycidae ; one species frequent

on Upolu, another on five occasions on Tutuila): certain other Cerambycidae

(see table of distribution, iv, 135).

In contrast with these there are several genera which contain a considerable

number of Samoan species, not one of which is confined to a single island: for

instance, Mordellistena (Mordellidae), contains 7 Samoan species, 2 of which are

known from other archipelagos: every one of the 7 occurs both in Western

and American Samoa.

It is necessary to be extremely cautious in accepting evidence that species

are confined to particular islands in Samoa. There are several insects which

appear to be locally distributed, though they probably exist in all parts of the

archipelago. One might mention two Lygaeidae: Nysius pacificus was collected

on Tutuwila by three collectors on at least five dates, but never found in Upolu

or Savau, though it is known from Efate, New Hebrides: Orthaea limbata was

collected on Savaii on ten different dates, and once on Tutuila, but there is no

record from Upolu: this is the more remarkable as this bug is recorded from

Fiji and Nive. It is impossible to believe that these insects, so widely spread in

the Pacific, are absent from Upolu.

One might summarize this section by stating that only a few insects have

developed separate races or species in different parts of Samoa. But when this

has occurred, one race or species occurs in Upolu and Savaii, and another in

Tutuila (and frequently Manua). In the slight development of island endemicity

Samoa appears to resemble the Seychelles (Scott, 1933). In Hawaiia, on the

other hand (Perkins, 1913), and also in the Marquesas (Mumford and Adamson,

1934), many genera have produced a wealth of species confined to single

islands.

THE ABSENCE OF CERTAIN GRoups.—It is a general characteristic of an

insular fauna that certain groups of organisms are absent, though they occur

on the continents in the same part of the world ; other groups, which are repre-

sented on the island, comprise a disproportionate part of the fauna. The Samoan

material seems to be adequate, and it is possible to examine the fauna and call

attention to those groups which are absent, or which are present in disproportion.

The subject is of interest because it may throw light on the origin of the land

fauna of Polynesia.

76 INSECTS OF SAMOA.

(1) Absence and disproportionate Representation

of Orders.—If we consider first the representation of the orders of insects,

it will be observed (Table 1, p. 67) that 18 of them are represented in Samoa.

Five orders * appear to be absent. With regard to 2 of these, the Protura and

Strepsiptera, one must admit that they may be present and unrecorded, because

the necessary specialized collecting was not undertaken. The other 3 orders,

the Plecoptera, Embioptera and Mecoptera, are not represented in the material,

in spite of the considerable collections of insects which were made in places

where these insects would be expected to occur. The absence of the Nemourid

stone-flies is remarkable, having regard to their wide distribution in the world

and their existence in Australia, New Zealand and South America. It appears

to me to be unlikely that representatives of these 3 orders occur in Samoa. A

sixth order, the Siphonaptera, contains several species which are of recent in-

troduction to Samoa, but only one which may be native.

Apart from the total absence of certain orders, the disproportionate repre-

sentation of others is remarkable. Of the 18 orders which have been found in

Samoa, 2, the Trichoptera and the Plectoptera, are there represented by a single

species. With regard to the second, one notices that it seems to be the most

easterly outlier of a genus (Chloéon) well represented in south-east Asia.

One may therefore say that out of 23 recognized orders of insects, 5 are

absent from Samoa so far as we know ; a sixth is not represented by any species

which is certainly native: 2 others are represented only by single, but endemic,

species. The great bulk of the fauna is therefore contributed by 15 of the orders.

It is interesting to compare the representation of certain orders in Samoa

with what is known of the Seychelles (Scott, 1933), Hawaiia (Perkins, 1913 ;

Bryan, 1934) and the Marquesas (Mumford and Adamson, 1934), information

about which is not yet completely available.

SAMOA. SEYCHELLES. Hawalia. MaRQUESAS.

Protura . ’ . | No record Unknown when | No record No record

material col-

lected.

Strepsiptera. . | No record One endemic Two species, No record

One endemic

Plecoptera : . | Absent Absent Absent Absent

* For the sake of uniformity we adopt the 23 orders given by Imms (1934).

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. chr

SAMOA. SEYCHELLES. Hawalta. Marquesas.

Embioptera Absent One One, endemic One

Mecoptera Absent Absent Absent Absent

Siphonaptera . Several, appar- | One, introduced | Five, intro- Three, intro-

ently intro- duced,endemic| duced

duced

Plectoptera . | One, endemic One, endemic Absent Absent

(Ephemeroptera)

Trichoptera . | One, endemic Six, endemic Absent Absent

The general resemblance between the faunas of Samoa and these other

archipelagos is remarkable : no simple geographical explanation seems to cover

the facts.

(2) Disproportionate Representation among

Families and Other Groups —wWith regard to other groups of

insects which are absent, a complete list for any order could be made by referring

to the systematic Parts of this work, having due regard to the fact that the

collections in certain orders are admitted to be very incomplete (p. 37). It is

sufficient here to point out some of the more obvious and interesting absentees.

Of the Orthoptera, it seems that the Mantidae are absent, though there is

an old record of Tenodera based on a specimen which was stated to have been

collected on Upolu ; Chopard disregards this and says that none of these insects

occurs east of the Solomon Islands, though it seems that two genera (but not

Tenodera) have been recorded from Fiji (Holdhaus, 1929, pp. 680-681). The

Gryllacridae afford another interesting problem. According to Karny (1929),

their distribution in the Malay Islands and Oceania is extremely interesting. It

is therefore much to be regretted that the Samoan record is doubtful. The

species, Gryllacris oceanicus, was collected by Le Guillou in 1841; neither

Rechinger nor any subsequent collector obtained material, and since Le Guillou

collected insects on many groups of islands, I am inclined to suggest that the

Samoan record is due to an error in labelling. On the other hand, both Chopard

and also Karny, to whom the matter has been referred, point out that these are

insects of specialized habits and express the view that the species may eventually

be rediscovered. Moreover, according to Karny, oceanica falls in the dubia

group, which is represented in Fiji and the New Hebrides, so that its occurrence

in Samoa would be in no sense anomalous. Another Orthopteran which is

78 INSECTS OF SAMOA.

absent from Samoa is Gryllotalpa ; it appears that representatives of this genus

occur in Australia and New Zealand, also in the Malay Islands, including New

Guinea, also in New Britain and New Caledonia; but these insects are not

known to occur in the Solomons, and New Hebrides, also Fiji and all parts of

Polynesia (Chopard, letter).

Even among the Orthopterous families which are represented in Samoa

there is considerable disproportion. The Phasmids are represented by two species

of Graeffea, a genus which may owe its present distribution to native journeys,

as 1t is associated with the coconut, and by a species of Nisyrus. Members

of this latter genus have a wide distribution in Melanesia, and have been recorded

from Fiji and Tonga; as the female is wingless, it has been held that this

distribution is evidence of former land connections (Holdhaus, 1929, pp. 653,

681). The Acridians are represented by only 7 species, the Blattidae by 22,

of which about 6 seem to be native; all or nearly all the others have been

introduced by man, but not necessarily within recent years. In contrast with

these families the Gryllidae number 34 species, most of them endemic, and many

of them brachypterous or wingless: the Gryllidae make up 44% of all the

Orthoptera, excluding the Dermaptera. Chopard expresses the view that it is

the hygrophile Orthoptera which are best represented in the fauna of Samoa.

It is perhaps for the same reason that the Gryllidae are relatively numerous in

the Seychelles (Scott, 1933) and Hawaiia.

From the order Hemiptera there are certain very remarkable absentees.

It appears that of the 33 species of Coccidae which have been collected, only three

can be regarded as endemic. In this respect and in many others, the Samoan

insects present a contrast with those of Australia and New Zealand. In those

countries, and especially in Australia, native species are numerous, and contain

representatives of all the subfamilies into which the Coccidae are divided

(Tillyard, 1926). Moreoever, there are no native species of Aphididae of Aleyro-

didae in Samoa. Among the Heteroptera, the Samoan fauna includes repre-

sentatives of all the more widely spread families, though 31 of the recognized

48 families are not recorded from these islands (11, 82). The absence of those

powerful insects the Belostomatidae is remarkable, especially as Lethocerus

wmdicus is distributed from India to Australia (Tillyard, 1926, p. 157).

It appears that few important families of Lepidoptera are unrepresented

in Samoa. Indigenous Pterophoridae are absent, as they are from Fiji, and

eastern Polynesia (Meyrick, 1928); Meyrick (ii, 67) calls attention to their

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 79

absence, which is remarkable because the adults appear to be constructed so

that they can be carried by wind, and the larvae of some of the species are

polyphagous. The absence of the Oecophoridae is still more curious. These

small moths are numerous in the countries which surround the Pacific, but no

representative of the family occurs in Samoa (though we collected 113 species

another is recorded from the Society Islands, but there are grounds for thinking

that it may have been carried there by man (Meyrick, 1928). Among the

Geometrid moths there is a curious example of disproportion. In most parts

of the world the Geometrinae form 30% to 50% of this family, but in Samoa

there are only 4 out of a total of 30 species. It appears that a similar disparity

has been observed in Melanesia and Polynesia generally, with the exception of

Hawaiia, where the ratio is disturbed by the great development of a single genus.

There are several other interesting absentees among the Lepidoptera. Members

of the genus Scoparia (Pyralidae) are well represented in the faunas of islands,

in many parts of the world. Meyrick informs me that he has described a species

from Papua, but it seems that none are known from Melanesia, Fiji or Samoa.

This might possibly be attributed to imperfect collecting, were it not that we

failed to find Scoparia in Samoa in spite of much successful collecting of small

moths ; indeed, there are no Scopariinae from Samoa, though 119 species of

Pyralidae are recorded. It must not be overlooked that Scoparia has reached

Hawaiia, where it has developed a wealth of endemic species. Members of the

genus are also recorded from the Marquesas, Tahiti and Rapa (endemic species

in each, Meyrick, 1929).

The curious distribution of the butterfly, Libythea, absent from Samoa and

present on one island in the Marquesas, is discussed below.

It is unnecessary to refer in detail to the distribution of families, subfamilies

and species of large Heterocera, for Tams tabulates their occurrence in Samoa,

Fiji, Australia and New Guinea. But one family, the Zygaenidae, may be

mentioned, because certain of its members which are pests of coconut have been

intensively studied. Several are known from Java and the great Malay Islands,

others from New Guinea, the Bismarcks and the Solomons ; none has yet been

found in the New Hebrides, Fiji or Polynesia (Tothill, Taylor and Paine, 1930).

A considerable number of families of Coleoptera, including some of wide

distribution, are not known to occur in Samoa. This may be in part due to in-

sufficient collecting ; but though neither Hopkins nor I have an adequate know-

80 INSECTS OF SAMOA.

ledge of these insects, several of the collectors from the Bishop Museum made

satisfactory collections of beetles. Let it suffice for the moment to call attention

to the absence of the Cicindellidae. Of these insects an endemic species has

been recorded from Fiji, and others exist in the Loyalty Islands and the New

Hebrides, as my friend Dr. W. Horn tells me by letter: none are known from

Samoa. Several authors have called attention to the disproportionate repre-

sentation of some of the families. Among the Lamellicorns, the Lucanidae are

well represented, but there is a dearth of Scarabaeidae ; as the former breed in

rotten wood and the latter in soil, it is suggested that the facts point to coloniza-

tion by drifting logs rather than by land continuity (iv, 35). Among the weevils,

of which a considerable collection was made, the subfamilies which frequent

bark are well represented ; of those which eat leaves, very few species were

collected, in spite of the fact that we devoted much time to beating the branches

of trees and shrubs.

The paucity of Hymenoptera * in Samoa strikes one at once, and a reference

to Part v of this work will show what a large number of families are not repre-

sented. It seems that there are no Chrysididae, in spite of the occurrence of

insects which they might parasitize. The entire absence of saw-flies (Chalastro-

gastra) of all families is remarkable. According to Kuznetsov-Ugamsky (1926),

the area which includes Australia, New Guinea and some of the adjacent islands

is inhabited by a small but interesting fauna of saw-flies, some of which belong

to endemic subfamilies. The eastern limit to which these insects extend does

not appear to be defined, but it seems that Samoa is outside it. Saw-flies are

also unknown on the Marquesas (Mumford and Adamson, 1934).

Among the Diptera Nematocera, the absence of the Bibionidae, Simuliidae

and Blepharoceridae is remarkable, especially as extensive collections were made

of other Nematocerous families. Dr. F. W. Edwards tells me by letter that he

knows of no Bibionids in Fiji, but that Bebio and Plecia are represented in the

New Hebrides. He calls attention (vi, 26) to the fact that a high proportion

of Samoan Nematocera breed in decaying organic matter: he finds evidence

of this in several families. He remarks also on the absence of larger and more

robust Nematocera, not only Bibionids, but also Tipulini. Of other crane-flies

(Tipulidae) a large collection was made, so that the absence of the large Tipulini

is significant ; I understand that they are numerous in the Oriental region, reach-

ing their eastward limit in Fiji (Ctenacroscelis, three species). A Blepharocerid,

* The parasitic families have been excluded from this work.

THE INSECT FAUNA: ITS PECULIARITIES AND CHARACTERISTICS. 81

Apistomyia, is known from Queensland, but there are apparently no records of

any member of this family from New Guinea or Melanesia. The apparent

absence of Simuliidae is remarkable, in view of the care with which we collected

in one or two ravines and of the fact that several members of this family occur

in Tahiti and the Marquesas : the matter is discussed on p. 87.

Among the higher flies, it will be noticed that the Anthomyinae are absent

from Samoa, and also from Fiji ; it is understood that these flies are not numerous

in the Indo-Malayan and Australasian areas.

The Odonata of Samoa are interesting, and they were carefully collected,

particularly by Armstrong. Fifteen species of Anisoptera are known, belonging

to the Libellulidae and Aeschnidae; the Gomphidae and Cordulegastridae

appear to be absent. Among the Zygoptera all the 13 species belong to the

Coenagrionidae, 8 of them being referred to [schnura and closely related genera ;

there are no Agrionidae.

Esben-Petersen points out that the Neuropterous families Ascalaphidae,

Osmylidae and Mantispidae are absent from Samoa and from all parts of Poly-

nesia, in spite of their wide range in the world. Only one Myrmeleonid is recorded ;

another species, Hidoleon (Distoleon) bistrigatus, was originally described from

Tahiti, and is known to occur also in Raiatea and the Tuamotus, also in Hawaiia,

Fiji and Australia. Its absence from Samoa is remarkable (vii, 91: also Chees-

man, 1927).

Our list of Apterygota includes no member of the Japygidae, but an endemic

species has been collected and recently described by Silvestri (1930).

The Samoan spiders are rather well known, and show absence of certain

families and great development of others. The possible relation of this to

dispersal by air is discussed elsewhere (p. 45).

Two general conclusions emerge from the consideration of insects which

are absent from Samoa. Many widely distributed families of several orders

are not represented, as is typical in an oceanic fauna. In almost every case the

family or order exists in the lands to the west and is not found farther east in

Polynesia. Some of the groups which are absent in Samoa have extended

eastwards to Fiji but no farther: others reach their limit in the New Hebrides

and the Solomon Islands: others again have not extended even into Melanesia.

One may say, therefore, that a study of the absentees shows clearly, as does a

study of the forms known to exist in Samoa (p. 83), that the fauna has reached

these islands from the west.

Ibe. G2 4

82 INSECTS OF SAMOA.

Ill. THE INSECT FAUNA: ITS ZOOGEOGRAPHICAL

RELATIONS

Hicu anp Low Istanps CoNTRASTED

Before we study the zoogeographical evidence which is provided by the

Samoan insects, it 1s well to recall several facts of a general nature which could

not be ascertained from a study of maps alone. The naturalist in the Pacific

learns at once that he must distinguish high from low islands. The high islands

are volcanic and very fertile: they are covered with a dense flora and fauna.

The low islands or atolls are very small; they barely rise above the level of the

sea, so that the subsoil water is generally brackish. In these places the whole

flora belongs to the strand association, and the island only contains a very few

species of plants and animals, all or nearly all of them of the widest possible

distribution.

It seems that Holdhaus (1929, 1934) hardly gave sufficient attention to

this distinction between high and low islands when he defined the subregions of

his Australian Region. His Melanesian subregion stretches eastwards to

include Fiji, Samoa and Tonga, high islands being characteristic of these three

archipelagos. His Polynesian subregion includes a rather heterogeneous collec-

tion of other islands, and stretches from Jaluit and the Gilbert and Marshall

Islands through Tahiti and Hastern Polynesia to Haster Island. Many of the

islands in this group are atolls, for which reason their fauna and flora is very

poor. The contrast between the richness of Holdhaus’ Melanesian and the

poverty of the Polynesian subregion is rather due to the type of island than

to zoogeographical causes, and one doubts whether the division into subregions

is justifiable. A rather similar case has been discussed by Scott (1933) ; in the

area which he considered, the Seychelles Islands were high and fertile with a

large endemic flora and fauna ; the other islands—Aldabra, the Chagos Islands,

etc.—consist of coral a little elevated above sea-level and carrying a poor flora

and fauna and few endemic species. But Scott, perhaps because he was per-

sonally familiar with the area, realized the essential unity of his islands, in spite

of their superficial differences.

But even in those Pacific archipelagos which contain high islands, consider-

able geological differences exist. If we are discussing the insects which inhabit

Fiji, Samoa and Tonga, we should not forget that the archipelago of Fiji contains

THE INSECT FAUNA : ITS ZOOGHEOGRAPHICAL RELATIONS. 83

much the largest islands with some ancient rocks. This archipelago, moreover,

lies nearer to the mainlands on the west. For both these reasons it may be

expected that the flora and fauna will be found to be the richest when it has been

more thoroughly studied. Samoa (ix, 1-10) contains volcanic rocks and volcanoes

of different ages and supports a large and endemic flora and fauna. In Tonga

some of the islands are volcanic and probably not dissimilar to those in Samoa,

though the insects remain almost unknown. But some of the islands, for

instance, Tongatabu, consist of elevated coral, on which it may be supposed that

the number of plants and animals will not be very great.

RELATIONS witH InpDo-MALAYA.

EXAMPLES.—One might say with truth that every writer who studied the

insects of Samoa before the publication of the present work recognized that

they were related to forms living in Melanesia and Malaya (see, for instance,

Rebel, 1910), and this is consistent with what is known of the other animals

and the plants (ix, 19 and 26). Those who have contributed to the present

work are of the same opinion, or passively accept the view that the Samoan

insects are Indo-Malayan in origin. Meyrick, for instance (ill, 67), expresses

the view that the Microlepidoptera of Samoa are Indo-Malayan, and that they

reached the central Pacific by way of Melanesia ; indeed, he goes farther than

that and specifically says that there is no element derived from Australia and

New Zealand or from America. Meyrick’s views command particular respect

because of his wide knowledge of the Microlepidoptera of Australasia and Oceania.

The same view is expressed in the pages of this work by Prout, Beeson, Edwards,

Berland and several other authorities. Moreover, all those contributors

who have made a table * of the distribution of Samoan insects have shown

clearly their origin from the west.

In the simplest case one finds that as a group of insects extends itself from

west to east in the Pacific, some species or larger unit ceases to appear as each

sea gap is passed. In studying the groups of insects which are absent from Samoa

* The following tables showing the distribution of insects in Samoa and other archipelagos

have been published in this work :—Fulgoroidea, ui, 2 (Muir); most families of Heteroptera, 1,

83 (China) ; Miridae and Anthocoridae, ui, 194 (Knight); Rhopalocera, Samoa and Tonga only,

ii, 6 (Hopkins); Heterocera (most families), iii, 173 and 176 (Tams); Heteromera, Bostry-

choidea, Malacodermata, Buprestidae, iv, 68 (Blair); Cerambycidae, iv, 135 (Aurivillius) ;

Curculionidae, iv, 249 (Marshall) ; Nematocera, vi, 24 (Edwards). There is also a table of dis-

tribution of Araneida, viii, 35 (Berland).

84 INSECTS OF SAMOA.

(p. 81) a number of examples were quoted ; but the insects which have reached

Samoa show the same phenomenon, for instance, the genus Papilio extends

eastwards through Melanesia and Fiji to Samoa where a single species (P.

godeffroyr) 1s found: east of this the genus is unrepresented. There are other

representatives among the moths: Mnesiloba, which is a subgenus of Hupithecia

(Geometridae) is found from the Nilgiris and Ceylon eastwards to Samoa and

Tonga. The genera Scopula and Sterrha (Geometridae) are very widely distri-

buted, though absent from the south of South America, New Zealand and

Hawaia; in Malaya and Melanesia there are several species, and in Samoa one

only. Among other orders of insects many examples could be quoted, but two

from the Diptera may suffice (Tabanidae and Culicidae). In the Tabanidae

there are a number of genera and species in the Malay Islands and several

species of Tabanus are known from Melanesia and Fiji; in Samoa there is a single

endemic species (7'. samoensis), but no representative of the genus is known from

Tonga, eastern Polynesia or Hawana (Ferguson, 1927).

The mosquitoes (Culicidae) exhibit the same type of distribution. These

insects provide admirable material for the study of zoogeography, for they

were never collected until 30 to 40 years ago, so that there are few erroneous

records, and since that date they have become well known and been studied

more thoroughly than any other group of insects in the Pacific. The geographical

distribution of some of the species within Oceania was fully discussed a few years

ago (Buxton and Hopkins, 1927) ; since then much new work has been published,

which has been summarized by Kdwards (1932a) and Taylor (1934) *: it is

therefore unnecessary to give detailed references. In Table 4 I have set out the

geographical distribution of all those sub-genera of Culicinae which are recorded

from any islands east of New Guinea. The subgenera are those accepted by

Edwards (1932a). The islands are arranged from west to east ; the subgenera

have been arranged to show their extension eastwards, those with the most

* 'Taylor’s list is so useful that I must be pardoned for calling attention to one or two omissions

and statements which appear to be erroneous. Mansonioides uniforms is recorded from Solomon

Islands (Kdwards, 1926c), and there are female specimens in the London School of Hygiene and

Tropical Medicine, from Rabaul, New Guinea (Backhouse). <Aédes (F.) kochi, type form, record

from Tonga quoted by Taylor refers to var. samoana : type form occurs in Fiji (Edwards 1926c).

Aédimorphus vexans is known from New Hebrides, Tonga, Ellice Islands (and Cook Islands doubt-

fully), see Buxton and Hopkins (1927). The record of Anopheles (Myzomyia) punctulatus from

New Caledonia is an error, and rather an important one ; it seems to have originated in Edwards

(1932a) and been copied by Taylor.

THE INSECT FAUNA: ITS ZOOGEOGRAPHICAL RELATIONS. 85

TABLE 4,

Giving the distribution of all subgenera of mosquitoes (Culicinae) known to occur east of New

Guinea.

SUBGENUS.

SoLomons.

New HeEpripes.

New CaLEponia.

Fst.

ELLICE.

Society Is.

+ | Coox Is.

+ | MARQUESAS.

Stegomyia

Culex .

Aédimor phus

Finlaya

Trvpteroides .

Mimetomyia

Uranotaenia

Coquillettidia

Aédes .

Ochlerotatus . : : : :

Mucidus. ; ‘ : é : ae

Myzomyia ,

Lophoceratomyia . 3 : : :

Armigeres, Lutzia, Rachisoura, Geoskusea,

Mansonioides, Culiciomyia, Mochtho-

genes P

Megarhinus *

Hodgesia

++

+++

(?)

+ +++

aa

opp

++++ +4

+

eee ae eee ae | BIsMARCKS.

ste amas ole ates oe te

-

a

++

|

b+

+++

extended distribution in Oceania heading the list. I have omitted the distri-

bution of Culex fatigans and Aédes aegypti (argenteus), for we know that these

are species of recent artificial introduction. It will be seen that mosquitoes are

an excellent illustration of the general rule that the fauna of the Pacific spreads

out from the west towards the east. The farther one goes towards the east,

the thinner is the fauna, and there is no evidence of the development of im-

portant endemic faunas in the eastern Pacific: moreover, there are very few

instances in which a subgenus is found in the eastern Pacific, but is absent from

islands more to the west. It must, however, be admitted that much fine detail

* The record of Megarhinus inornata from Funafuti, Ellice Islands (Rainbow, 1897), need

not be seriously considered.

86 ' INSECTS OF SAMOA.

is lost by tabulation under subgenera. The table, for instance, fails to show

that in Samoa and in Tahiti, endemic species of Culex (C. samoaénsis and atriceps

respectively) have developed : we do not know from what member of the genus

Culex these endemic species have originated. There are one or two gaps in the

table (e.g. in Mucidus, Ochlerotatus) which may be due to the imperfection

of our present knowledge ; some of them, for instance, the absence of Finlaya

from the New Hebrides, are probably genuine, for I failed to find Aédes (F.)

kochi during several months spent in that archipelago. The irregularities in the

distribution of the subgenera of Mansonia are probably also genuine and not

due to present ignorance, for the distribution of these mosquitoes may be

assumed to depend on some aquatic plant.

As the fauna has spread in this simple manner from west to east, there is

much similarity between the insects of Samoa and those of Fiji and Tonga :

for instance, Tams names 5 Samoan moths which are represented by subspecies

in Fiji and 2 which are represented by subspecies in Tonga: he adds a list of

other Samoan moths which have very close relations in Fiji or Tonga. Prout

mentions a number of Geometridae which appear to be common to these three

archipelagos, and not found outside them ; other examples could be discovered

in other orders of insects. It would not be difficult to extend the circle of study,

and name insects found in these three archipelagos, and also in Tahiti, or New

Caledonia or the New Hebrides.

A “ Central Polynesian ” distribution has been observed in the Arachnida :

among 81 species known from Samoa, no less than 25 show relationship with

Fiji, Tonga and other island groups in Polynesia. This resemblance is what

one would expect from the proximity of these archipelagos to one another and

appears to be easily explained by the existing geography. Berland (vi, 41),

however, has expressed the view that we have here an indication of an ancient

land mass of which Samoa, Fiji and Tonga are the remains. More recently

(Berland, 1934b) he has examined collections from several other parts of Poly-

nesia, which have led him to the conclusion that the fauna of the whole of

Polynesia, including the Hawaiian Islands, is so uniform as to suggest that

these islands are but the fragments of a single mass of land. Berland’s view,

derived from a study of spiders, is similar to Pilsbry’s, founded on Mollusca

(ix, 23). But I feel that perhaps Berland has hardly considered the difficulties

which are revealed by a general study of the distribution of plants and animals,

of the ocean depths, and of the geology.

THE INSECT FAUNA: ITS ZOOGEKEOGRAPHICAL RELATIONS. 87

ANnomALIEs.—We find then that the majority of Samoan insects have

reached their present home from the west, and that their resemblance to the

insects of Fiji and Tonga is accounted for by their geographical proximity to

one another. But there are certain exceptions and anomalies: a number of

insects occur in Samoa though they appear to be absent from Fiji and the islands

to the west: there are also insects which are not known to occur in Samoa,

but which are found in islands to the east, particularly in the Marquesas.

There are certain butterflies absent in Fiji and the islands to the west,

but present in Samoa: at least three species are so distributed. Two of them,

Hypolimnas errabunda and Atella exulans, are hill species in Samoa, with closest

relatives in New Guinea (similar Hypolimnas being known also from the

Solomons). It is possible that collecting-in the mountains of Fiji or the New

Hebrides might show their existence in these intervening areas. But the

third butterfly, Huploea schmeltzi, occurs in the lowlands and is absent from the

area between Western Samoa and the Loyalty Islands. Moreover, within

the Samoan archipelago it occurs commonly in Upolu and Savaii but not in

Tutuila or the Manua Group, where its place is taken by another member of the

genus not found in Upolu and Savau. The distribution of EH. schmeltz is

extremely difficult to understand, for it has actually been observed flying

out to sea to the small island of Nuutele, which lies about a mile from the

coast of Upolu, though it has never been discovered on Tutuila, which is

only 45 miles away. There are similar anomalies of distribution among those

butterflies which have developed endemic races (Table 2, p. 72); here

again a sharp distinction between Western Samoa and Tutuila is evident

in three different species. The view has been put forward (Hopkins, iii, 5) that

the form found in Tutuila has been derived from Tonga and Fiji, but that the

form in Western Samoa shows some relationship with islands as far away as

New Guinea.

Certain insects which appear to be absent from Samoa but which are known

to occur in eastern Polynesia, particularly in the Marquesas, are of great interest

because their distribution raises difficult questions of zoogeography. Of those

which occur in the Marquesas the small Simuliid midges are the most remarkable.

It appears (Edwards, 1932b, 1933a and b) that there is a single species of Simu-

lwum which is widely distributed in the New Hebrides and Fiji: no representative

of the genus is known from Samoa, Tonga (or the Hawaiian Islands). In Tahiti

there are three endemic species, and in the Marquesas there are three different

88 INSECTS OF SAMOA.

endemic species. These six species form a compact group within the widely

distributed subgenus Husimuliwm, and it is definitely Edwards’ opinion that

they are derived from the west and that they have not affinities with Simuhidae

in America. It seems then that these midges have spread eastwards across the

Pacific but have in some way failed to colonize Samoa. An alternative is

possible. It may be that members of the genus still remain undiscovered in

the Samoan Islands in some of the hundreds of gullies in the forests, many of

which have never been entered by a European. In Savaii, particularly, the

forest is almost unknown except round the coast. But against this view that

there may be Samoan species of very limited distribution is the fact that those

species which occur elsewhere in the Pacific are not extremely localized, as can

be seen from the large number of localities from which they have been collected.

While I admit the possibility that Simuliwn may occur in Samoa and remain

to be detected later, I incline to the view that it is really absent.

One may also call attention to the absence of the butterfly Libythea from

Samoa. I understand that a single species (L. geoffroyz) is widely distributed

from the Malay Peninsula to the Solomon Islands and the Loyalty Islands

(Riley, letter). But eastward of that, in Fiji and Polynesia, there is no record

of any species, though the butterflies of several archipelagos appear to be well

known, except L. collenettei, which occurs only on Nuka Hiva in the Marquesas

(Poulton and Riley, 1929). L. collenettei appears to be related, but not closely,

to geoffroyi, but to present a number of unusual and perhaps archaic

features.

Reference has been made elsewhere (p. 79) to the curious distribution of

the Pyralid genus Scoparia. These moths appear to be absent from the whole

of Melanesia, Fiji and Samoa; it may certainly be claimed that the small

moths of Samoa have been carefully collected and studied. A few endemic

species are known to occur in the Marquesas, Tahiti, Rapa, and in the Hawaiian

Islands there are many endemic species. The similar anomalous distribu-

tion of an ant-lion, Hidoleon bistrigatus, has been mentioned (p. 81); the insect

has a very wide distribution which embraces Tahiti, Raiatea, the Tuamotus,

the Hawaiian Islands, Fiji and Australia, but it appears to be absent from

Samoa. The distribution of Rhyncogonus, a genus of weevils, is rather similar.

There are 33 species in the Hawaiian Islands, some of them having been found

even on the outlying islands (Laysan, Wake, etc.) : in the Marquesas there are

6, in Rapa 1, and the Kermadecs 1 (Van Dyke, 1932). These weevils are not

THE INSECT FAUNA : ITS ZOOGEOGRAPHICAL RELATIONS. 89

known from Fiji, Samoa, ete., in spite of the considerable collection made in

Samoa.

These curious geographical facts are capable of several interpretations, but

we cannot discuss them without considering what relations there may be between

the insects of Samoa and Hawaiia, for it is now apparent that the Hawaiian

and Marquesan faunas have notable points of resemblance.

RELATIONS WITH HAWAIIA AND THE MARQUESAS.

HawataA.—The remarkable and highly endemic fauna of the Hawaiian

Islands was the subject of the classical ‘“ Fauna Hawaiiensis,’ published im

the early years of this century and summarized by Perkins in 19138. Since those

days, investigations in other parts of the Pacific have tended in the main to

confirm the conclusion that the fauna of Hawaiia stands quite by itself in the

world. But it is now known that one or two groups of insects, formerly believed

to be endemic to the Hawaiian Islands, have a wider distribution. A considerable

number of them are found also in the Marquesas (Mumford and Adamson,

1934). So far as Samoa is concerned, the most interesting among these insects

is Proterhinus, a genus of minute beetles distantly related to the weevils. It was

at one time supposed that the members of this genus were confined to the

Hawaiian archipelago, where more than 100 species are known to exist; but

within recent years 1 species from Samoa, 1 from Enderbury Island (Phoenix

Group) and 2 from the Marquesas have been described. Moreover, the view

is now held that Proterhinus is not so peculiar as to justify the maintenance

of the family Proterhinidae, and it has recently been sunk in the Aglycyderidae,

the other members of which inhabit New Zealand and the Canary Islands

(Perkins, 1932): Proterhinus is one more example of the category of minute

beetles widely distributed in some tropical islands, and in this particular case

greatly developed in Hawaiia. Elsewhere in this work I have called attention

to other examples of this type of distribution, which is perhaps due to the

effect of wind (p. 43). In this connection I observe that Arrow has recorded

from Samoa 5 minute beetles belonging to the families Corylophidae, Cucujidae,

Mycetophagidae and Dermestidae, previously thought to be peculiar to

Hawaua,

A few other examples of groups of insects now known to occur in Samoa

though previously regarded as endemic to the Hawaiian Islands have been found.

90 INSECTS OF SAMOA.

Perhaps the most interesting is the genus Austrochrysa (Chrysopidae), with one

endemic Samoan species, and a Hawaiian species which is perhaps referable to

this genus. Mention may also be made of Libnotes perkinsi, a Tipulid originally

described from Hawaiia and since discovered in Tahiti, Fiji and Samoa. In

view of its distribution it is not impossible that the early stages of this insect

have been carried about the Pacific by Polynesians among the roots of growing

plants. The distribution of the woodlouse, Ligia perkinsi, should also be

mentioned: in the Hawaiian Islands it occurs high in the mountains, where

it was supposed to be endemic until we discovered it on rocks on the shore of the

islet of Namua, off the coast of Upolu. But such an anomalous distribution

as this must surely be due to our present ignorance.

We have mentioned a few points of resemblance between the Samoan and

the Hawaiian Islands, but the points of difference are almost innumerable :

indeed, one may say that with the exception of Proterhinus, none of the genera

which have developed great numbers of endemic species in the Hawanan Islands

are represented in Samoa. One may take a single example. Tams has observed

that of the 52 moths which have been recorded in the Hawaiian Islands but

are not endemic there, 6 only are known from Samoa: all of them are either

cosmopolites or pests of agricultural crops. There are also very great differences

among the families and higher groups. From a consideration of all the facts,

I think we may conclude that the points of positive resemblance between the

Hawaiian Islands and Samoa are very few. There are points of negative

resemblance, such as the absence of certain orders (p. 76) from both, but these

are explained by saying that both archipelagos have a typical oceanic fauna.

Almost the same degree of negative resemblance might be found between the

fauna of Samoa and that of the Seychelles, between which no zoogeographical

connection can be supposed to exist.

MarQuEsas AND ‘‘ PALAEONESIA.”—We can now consider the relation

of the Samoan fauna to that of the Marquesas, which in turn has certain points

of resemblance to the fauna of the Hawaiian Islands. As we have seen, it

is almost certain that a number of insects—Simuliwm, Scoparia, Libythea and

others—are really absent from Samoa though they occur farther east in the

Marquesas ; and yet competent authorities tell us that the forms found in the

Marquesas have certainly been derived from the west and not from the American

shores of the Pacific. It will be observed that none of these genera is endemic

in Polynesia and that in none of them, except perhaps the Libythea, are the

THE INSECT FAUNA: ITS ZOOGEKOGRAPHICAL RELATIONS. 91

Marquesan species very peculiar or primitive. It should also be remembered

that certain of these insects, for instance Scoparia, extend their range to the

Hawaiian Islands. But there is little relation between Hawaiia and Samoa,

though the genus Proterhinus, formerly thought to be endemic in Hawaiia,

has now been found in several other archipelagos, including Samoa.

It would not be fitting to give much space to the zoogeographical problems

of eastern Polynesia, which are not strictly relevant here: but it is necessary

to make reference to the theory of Palaeonesia. Meyrick (1926) investigated

a collection of small moths from Rapa, an island so small and remote that he

could not believe that the moths had reached it by air. He therefore suggested

that there must have been a much larger land mass during the Eocene, and he

postulated an elevation of 12,000 feet of the floor of the Pacific which would

give an island 400 miles long. Such an island might help to account for the

presence in Rapa of insects showing affinities with Fiji and eastern Australia.

At a later date the same author (Meyrick, 1928) examined a collection of Micro-

lepidoptera from the Marquesas, Society Islands, Tuamotus and Austral Islands.

This led to some modification and extension of his view, for he found that he

required a much larger land mass, to which he then gave the name Palaeonesia.

He also supposed that a different elevation of the floor of the Pacific had occurred

farther west, which greatly increased the area of land in that region. But in

spite of this, the Palaeonesian genera are thought to have reached their present

home by some route which avoided those large islands, the hypothetical position

of which seems to be directly in their hypothetical path. Indeed, it seems to

be an essential part of Meyrick’s view that there is a fundamental dissimilarity

between the faunas of western and eastern Polynesia, so that in this second paper

the connection with Fiji is no longer sought for. Subsequently, Meyrick (1929)

examined a collection of Pyrales from the same area in eastern Polynesia. In their

distribution he found confirmation of the previous existence of Palaeonesia, and in

particular he referred Scoparia to the “ Palaeonesian period,” not yet defined.

The present writer can claim no knowledge of the Microlepidoptera, but he

has examined the evidence put forward by Mr. Meyrick and endeavoured to

relate it to our general knowledge of Oceania. He is led to a less elaborate

hypothesis by the following considerations :

(1) In his first paper (1926), Mr. Meyrick stated that “a rise of 12,000 feet

in the bottom of the South Pacific is required.” In the second (1928) he defined

Palaeonesia as a “land area to include the whole of this region of the Pacific

92 INSECTS OF SAMOA.

from Rapa to the Marquesas, and from Pitcairn Island to the Society and Cook

Groups.” The student of maps will discover that this handsome piece of territory

extends about 20° by 30°, say 1,200 by 1,800 miles. No consideration seems

to have been given to the geophysical problems involved in raising this area

12,000 feet.

(2) Mr. Meyrick cannot believe that so remote and minute an island as

Rapa could have been colonized from the air, even over vast periods. This

is not a matter on which it is possible to dogmatize, but there is positive evidence

of the wind carriage of minute insects (p. 43), and there are a number of small

beetles and other insects which appear to be very widely distributed in the

tropics by some natural agency (p. 44). It is possible that the smallness and

fragility of the microlepidoptera might aid their dispersal.

(3) It is certainly true that the distribution of some of the Microlepidoptera

is difficult to explain in relation to existing geography ; Scoparia is an example

(p. 79). Another distributional anomaly is found in the Pyralid genus Aspithra,

of which 1 species is found in Ceylon and the other 4 in the Marquesas. Another

quite different problem is afforded by the Tortricid Dichelopa, with 14 endemic

species in eastern Polynesia, 6 in eastern Australia and none elsewhere. The

Tineid Gracillaria presents a difficulty of another type : its distribution is nearly

universal, species occurring in New Zealand, Rapa and Hawaii, but it appears

to be absent from Fiji and Samoa. The reader will observe that Palaeonesia

would help to explain the distribution of Scoparia, but it does not appear to

reduce these other anomalies.

(4) But though these difficulties occur, many of the Microlepidoptera

conform to the general rule that the fauna of Oceania has spread out from the

Asiatic side, becoming thinner and thinner as each sea gap is crossed. For

instance, the Tineid genus Labdia is abundant in the Indo-Australian region

and species, most of them endemic, have been collected in the New Hebrides,

Fiji, Samoa, Tonga and the Ellice Islands. That represents approximately the

eastern limit of Labdia (cf. the distribution of Papilio and other insects, p. 84).

On the other hand, Jmma and Decadarchis (Tineina) and Bradina and Margaronia

(Pyrales) show affinities with Indo-Malaya, but they have developed endemic

species not only in the west of the Pacific, but also in eastern Polynesia. In

the distribution of such genera one sees an essential similarity between the fauna

of those groups which Mr. Meyrick would include in Palaeonesia and that of

Samoa, Fiji and the islands to the west.

THE INSECT FAUNA: ITS ZOOGEOGRAPHICAL RELATIONS. 93

(5) To sum up, the Marquesan insects are undoubtedly peculiar and highly

endemic, as are those of Hawatia, but there is evidence among the Microlepido-

ptera and elsewhere indicating that the fauna came from the west and that it is

essentially like that of Fiji, Samoa, etc. I believe that the zoogeography

of Oceania, including eastern Polynesia, can be explained by reference to the

map as it 1s to-day. Certain anomalies undoubtedly exist and will perhaps

continue to trouble us, even when our knowledge of these faunas is much more

complete than it is at present. Some of these may be explained if we consider

the possibility that the insects of the Polynesian archipelagos arrived by different

routes. One must remember that in this part of the world the anti-trade wind

blows steadily from the west at a height of a few miles, while the trade winds,

which are doubtless more important because they are at ground level, blow from

the east or south-east for most of the year (ix, 18). Here then we have two natural

machines of great power and regularity tending to move animals and plants

in opposite directions. It is a matter of fact that most of the fauna has come

from the west, but it may be supposed that some of it has been carried back

again towards the west, perhaps over a different route.

RELATIONS WITH AUSTRALIA AND AMERICA.

AUSTRALIA.—We have made it clear that the insects of Samoa have come

from the west and that the fauna of Melanesia and at least the greater part of

Polynesia is derived from Asia. It is perhaps best to describe this fauna as

Indo-Malayan.” The word “ Indo-Australian” is sometimes used but 1s

misleading because it tends to suggest a resemblance to the fauna of the

Australian continent.

But though the relation between the animals of Oceania and continental

Australia is slight, one or two definite points of resemblance exist. We have

referred elsewhere (ix, 22) to the honeysuckers (Meliphagidae), a family of birds

which spreads outwards from Australia and Papua through Melanesia and much

of Polynesia. One or two examples of similar distribution among the insects

are known. They are important and interesting by reason of their rarity.

The most remarkable of these are the Tachinidae of the genera Rutilia and

Rhinomyiobia. In Samoa there are two endemic species of Rutilia and there is

one in Fiji. The members of the tribe Rutiliini are almost exclusively Australian,

though one is recorded from China and one or two from the Malayan area: there

are none in New Zealand. Apart from the fact that the Samoan and Fijian

94 INSECTS OF SAMOA.

species are endemic, so far as we know at present, it would be difficult to suppose

that they had been introduced by man, for their early stages are probably

parasitic on larvae of Coleoptera. On every ground, therefore, we conclude that

they are native. The other Tachinid, Rhinomyiobia plumifera, is rather similar

in distribution. The species is peculiar to Samoa and Fiji, and the other

members of the genus are mainly Australian. Several other Samoan insects

of different orders show a similar relation to the Australian continent. The

genus Phycomorpha (Copromorphidae, Tineina) contains at present two species,

one from Sydney, New South Wales, and the other from the hills of Upolu, Samoa.

The only Samoan Crambus (Crambidae, Pyrales) is a species known to occur in

Fiji and Australia. Tams has called attention to five species of moths which are

known only from Samoa and Australia, but he states that the material is in-

sufficient and that more careful and critical work is necessary before we can be

certain in four of these cases. The remaining moth is Phlegetoma fasciatria.

There are also points of resemblance between the Samoan Carabidae and those of

Australia: four of the fifteen Samoan species are known to occur in Australia,

but it should be remarked that little is known of the Carabidae of neighbouring

archipelagos, for instance, Fiji and Melanesia: for the same reason we must be

cautious in attaching importance to the distribution of Crambus and Phycomorpha.

The distribution of Chrysopa appears to stand rather by itself. It seems that

certain members of this genus have a very wide range, and Chrysopa itself is

nearly world-wide. Why this should be so is not known, and perhaps the facts

have not yet been sufficiently studied, but we observe that four out of the nine

species of Chrysopa which occur in Samoa are also recorded from Australia.

AMERICA: WEGENER’S Hyportuesis.—It is already clearly shown that

the insects of Samoa are derived in great part from the west ; to myself it seems

clear that the boundary of this Indo-Malayan element lies far to the east of

Samoa, and that the region embraces all Polynesia including the Marquesas.

But it might be asked whether there is no American, or eastern, element among

the native Samoan insects? From a consideration of the articles which

specialists contributed to this work, it seems that very few insects of American

origin could be quoted and that for the most of them the evidence is of doubtful

value. Arrow (iv, 36), for instance, recognizes an American element in the fauna,

but perhaps some of his examples are insects of very wide distribution at present

not sufficiently collected: one has probably been introduced into Samoa, and

one at least is very minute so that it may fall into the wind-carried group. Beeson

THE INSECT FAUNA: ITS ZOOGEOGRAPHICAL RELATIONS. 95

(iv, 245) shows that among the 34 Scolytidae, only one species, Xyleborus con-

fusus, is probably derived from America: but the evidence is of indefinite

value, for the insect, which has a very wide distribution in the tropics but is

apparently absent from the Indo-Malayan region, is known to feed on a

number of plants of economic importance (Cocos, Manihot, Hevea). Among

the Diptera a remarkable insect is Zygothrica samoaénsis (Drosophilidae).

This insect, endemic to Samoa, exhibits the sharply defined peculiarities of its

genus, all the other species of which are confined to South America. It is

possible that the anomalous distribution of this minute fly may be due to wind

currents. One or two other examples of insects apparently of American origin

have received notice in the text, but they are not convincing. There is, for

instance, an Asilid doubtfully referred to the Central American genus Atonia,

but represented in the Samoan collections by a fragmentary specimen. One

might also mention Sarcophaga peltata, but it may well have been recently

introduced ; moreover, its specific identification is not beyond dispute, and the

view is expressed by Malloch (1932) that the American S. peltata Aldrich does

not occur in the Marquesas and Tahiti (and by implication in Samoa), and that

the Polynesian specimens are referable to taztensis Schiner, which is probably

native to these archipelagos.

We conclude that very few Samoan insects show affinities with America.

It is perhaps appropriate to consider the theory of continental displacement

put forward by Wegener, but it is unnecessary to discuss it fully for its scope is

mainly geographical or geophysical ; its biological applications relate particularly

to the fauna and flora of those continents which now le round the Atlantic and

Indian Oceans. But so far as the Pacific is concerned, we observe that Wegener

upholds the view of the permanence of this ocean at least since the Upper

Carboniferous Period. He believes that since that time the Pacific has been

narrowed from two directions: Australia and New Guinea have broken away

from the east extension of the Antarctic region and come to occupy their present

position on the flank of Malaya and Melanesia (a movement with which we are

not now concerned); during the same period the American continents have

swung to the west from a fixed base in the Arctic regions so that at the Equator

the American coast of the Pacific is some 30° (say 1,800 miles) west of the

position which it occupied in the Upper Carboniferous Period. Nothing that

could be discovered in the fauna of Oceania is likely to prove or disprove this

theory, but we may say that what is known of the distribution of animals in

96 INSECTS OF SAMOA.

the Pacific is consistent with the view that that ocean is permanent. Moreoever,

if America once lay farther to the east than it does at present, that helps to explain

the absence of American plants and animals from Hawalia (Wegener, 1924,

p. 84) and from Polynesia in general.

IV. SUMMARY OF CONCLUSIONS

The collection of insects from Samoa is far from complete and not equally

representative of all orders (p. 37), but it is certainly a sufficient foundation

for general conclusions. But it must be remembered that these can only be

tentative, because so little is known of the fauna of other parts of the tropical

Pacific. It is a matter of regret that geological evidence is, and must perhaps

always be, almost negligible in Polynesia, so that we are compelled to rely to

a great extent on faunistic material.

The study of a large collection of animals or plants from an archipelago

may be expected to supply information on several points of interest. One should,

for instance, be able to decide whether the islands are oceanic or continental,

by which we mean whether they have always existed as small points of land

in a great ocean of water or whether at one time they were part of a large mass

of land, perhaps connected with existing continents. One should also be able

to discover from what direction the fauna reached the islands and perhaps

by what means or channel it came.

OcEANIC OR ConTINENTAL ?—Evidence of possible connection with con-

tinental masses of land may be obtained by considering what groups of animals

are absent as well as what are present. In the Samoan insect fauna a number of

remarkable and doubtless significant absentees have been observed. Of the

23 recognized orders of insects, 5 have not been found in the archipelago. Of

these, 2, the Protura and Strepsiptera, may perhaps be discovered when more

skilled and specialized collecting is undertaken, but we may probably accept

it that the other 8, the Plecoptera, Embioptera and Mecoptera, are absent.

In addition to these 5, the Siphonaptera are only represented in Samoa by

species which have apparently been introduced by human agency, and the

Trichoptera and Plectoptera (Ephemeroptera) are each represented by a single

endemic species. The very great bulk of the fauna is therefore made up of

representatives of 15 of the orders. Study of those orders which are absent

SUMMARY OF CONCLUSIONS. OU

reveals remarkable points of similarity to the faunas of the Hawaiian Islands

and the Marquesas: the resemblance to the fauna of the Seychelles is just as

striking and even more remarkable, for it cannot be due to any local cause

(p. 76). The absentees among the families and lesser groups of insects have

been discussed at some length. The facts lead to the general conclusion that

the Samoan Islands are oceanic ; since certain widely distributed families are

absent from Samoa, other families are over represented in proportion to the

whole fauna (p. 77).

Let us now make a study of the insects which are present, and particularly

of the endemic species. 1,603 species are known to occur in Samoa, of which

49°%, appear to be endemic. The corresponding figures for the Hawaiian

Islands are 4,620 species, 81° endemic, and for the Seychelles 2,090 species,

65°% endemic (p. 66). It must be admitted that the percentage of endemic

species in the Samoan fauna may be modified with an increase in knowledge,

but it appears to be unlikely that it will be greatly raised. One characteristic

of the fauna of such an archipelago as Hawaii is the development of complex

groups, many of them containing a very large number of closely related species.

In this respect the Samoan fauna is poorer than that of the Hawaiian Islands

and probably the Marquesas; in Samoa such groups are rather few, and not

generally very complex, but the phenomenon does occur and instances are

quoted among the Tineina, Tipulidae, Curculionidae, Zygoptera, etc. (p. 68).

Where such endemic complexes occur, it is generally found that many of the

species are found only on one island in the archipelago. But in this matter

evidence must be full before it can be convincing. If an insect has been collected

on only one island, we can only admit that it is absent from the others if it was

obtained on many occasions or was so familiar that its local distribution was

remarked by those working on the spot. Among the Samoan insects there are

only a few examples which are undoubtedly confined to part of the archipelago.

There are, for instance, three Rhopalocera which exhibit racial differences between

the form found in Western Samoa (Savaii and Upolu) and that found in American

Samoa (Tutuila and Manua). The familiar and conspicuous moths of the genera

Chrysaeglia and Deilemera are also certainly confined to particular islands. It

has been observed that nearly always, when island endemism exists, the line of

division between species or races is that separating Western from American

Samoa (p. 72). The endemic fauna of an oceanic island sometimes contains

animals, the systematic position of which is anomalous, but among the Samoan

IX 2 9)

98 INSECTS OF SAMOA.

insects no examples could be quoted. It should also be noticed that the

systematists who have dealt with the Samoan material have not found it

necessary to erect any new tribe or subfamily.

The study of the insects, both of those which are absent and of those which

are present, leads to the same conclusion, that the fauna of Samoa 1s oceanic.

But it is not so peculiar (and therefore perhaps not so ancient) as that of the

Hawaiian Islands, or probably of the Marquesas. But though I have no hesita-

tion in stating that the fauna of Samoa is typically oceanic, I must admit that

there are certain facts which are difficult to reconcile with this conclusion.

For instance, it will be remembered that we collected a peculiar Locustid,

Rhaphidophora rechingert, in a cave in Upolu. The insect lived in total darkness

upon heaps of bats’ dung, and was clearly specialized for existence in a cave,

lacking wings and eyes. Its affinities are with the west, and a number of other

and equally specialized and peculiar members of the group have been found in

Malaya, etc. It seems to be almost impossible to imagine how this insect,

specialized for life in a cave, could have reached the Samoan Islands by any

method of transport overseas. The same difficulty confronts us when we

consider the distribution of many other insects, of which the habits or breeding

places are narrowly specialized; for instance, the Anthomyid, Limnophora

mmmaculiventris and the Gryllid, Anaxipha hopkinsi, found together skating

over the surface of streams; parasitic insects, gall-makers, and all specialists

fall in the same group ; it is difficult to think how such insects or their ancestors

could have been brought across the ocean, presumably from an equally specialized

home in some archipelago to the west.

Ornicin.—It is clear, and generally admitted, that the fauna of Samoa is

essentially Indo-Malayan, and that the fauna of Oceania, or at least the greater

part of it, has spread out from the west, passing through Melanesia and Fiji

into Polynesia, and becoming poorer with the passage of each area of sea. There

is no one point at which the fauna becomes suddenly impoverished, and there

is no faunistic evidence to demarcate the edge of an old continent at any particular

point. It may be added that within the distribution of a genus or family, the

general rule is that the number of species may be considerable towards the west,

whereas only a few occur where the eastward limit is approached.

As would be expected, there are a number of points of similarity between the

insects of Fiji, Samoa and Tonga, owing to their geographical proximity, and

their nearly equal distance from the large Malay Islands ; but there are also many

SUMMARY OF CONCLUSIONS. 99

differences between them and it would not serve a useful purpose to delimit a sub-

region. I fail to find evidence supporting the view which has been enunciated

that these three archipelagos are the remains of an old land mass (p. 86).

The general rule that the fauna spreads evenly from the west to the east

is subject to certain exceptions. For instance, the butterfly, Huploea schmeltzi,

has an anomalous distribution, for it occurs in the Loyalty Islands and Western

Samoa but not in Fiji. Several insects are absent from Samoa (and not merely

unobserved) though they occur in islands to the east. The butterfly genus

Inbythea extends from Malaya to Melanesia, including the Loyalty Islands,

but it does not occur in Fiji or anywhere in Polynesia except that an endemic

species exists on one island in the Marquesas. The distribution of the Pyralid

genus Scoparia is perhaps even more remarkable, for it is absent from Melanesia,

Fiji and Samoa, but present in eastern Polynesia including the Hawaiian Islands.

Other examples have been quoted from among the Diptera and Coleoptera

(p. 87).

The existence of these insects in the Marquesas and of some of them in

Tahiti, Rapa, etc., and their absence from Samoa and generally from Fiji seemed

to Meyrick to indicate a fundamental difference between the fauna of western

Polynesia, which is Indo-Malayan, and that of eastern Polynesia. A considera-

tion of the distribution of the Microlepidoptera led him to suppose that a large

land area once existed in eastern Polynesia, to which he has given the name

Palaeonesia (p. 91). His hypothesis certainly helps to explain the distribution

of Scoparia, though it leaves a number of other riddles unsolved, and it seems

that the very great geological difficulties have not been considered. To the

present writer it appears that, though certain anomalies exist, the fauna of

eastern Polynesia resembles that of western Polynesia in essentials, though

certain groups are absent. It appears, also, that the insects may well have

reached their present home at a time when land and water were distributed

approximately as they now are (p. 93).

The recent increase in our knowledge of the insects of Samoa has shown

that a few groups of insects occur in these islands which at one time were thought

to be peculiar to Hawaiia, but in the main the fauna of the Hawaiian Islands

stands by itself and shows little resemblance to that of the rest of Oceania,

though there are certain notable points of resemblance to the fauna of the

Marquesas (p. 89).

Though the Samoan fauna is clearly Indo-Malayan, one or two characteristic

100 INSECTS OF SAMOA.

Australian groups have reached Samoa, apparently by natural means. The best

example is perhaps the Tachinid genus Rutilia, but there are one or two among

the Lepidoptera. These insects with Australian affinities are very few compared

with the rest of the fauna, but the relation to Australia is definite and extremely

interesting (p. 93).

In the systematic Parts of this work, several insects of American affinities

have been mentioned, but the importance of most of them is shght for their

distribution may be due to commerce (p. 94). One of the most interesting

examples is a Drosophilid (Zygothrica), but it is perhaps one of those small

insects with a very wide tropical distribution which may be due to carriage in

the upper air. It seems, therefore, that among the native Samoan insects the

American element is very small or perhaps non-existent ; this 1s consistent with

the views of Wegener, who believed that the Pacific has existed for very lone

geological periods, but that there has been a tendency for it to become narrower,

so that at earlier epochs Samoa was even farther from America than it is now

(p. 95). |

Admitting that the insect fauna of Polynesia is oceanic and derived from

Indo-Malaya through Melanesia, it is pertinent to inquire how the existing forms

of life or their ancestors crossed the oceans and reached their present homes,

so many of which are minute islands. These questions have puzzled zoo-

seographers since the days of Wallace, and we are not yet in a position to answer

them, though knowledge is advancing in certain particulars. We now know

that, though the surface currents and trade winds appear to hinder the spread

of insects from the Asiatic side into the Pacific, there is a great stream of anti-

trade wind at no great altitude. This blows and has blown all down the ages,

and might well carry insects and other forms of life eastwards (p. 18). Recent

work has shown that a considerable number of insects occur in the upper air

(p. 43) and that there are species whose minute size has apparently assisted

their carriage by wind so that they are distributed very widely in the tropics

(p. 43). It seems, therefore, that we may suppose that wind, and particularly

the currents of the upper air, are much more important agents in distributing

small insects than we have realized. I venture also to suggest that the student

of insect distribution has given insufficient attention to that interesting animal,

Man. It is clear in Oceania that the wanderings of peoples before the coming

of the Kuropean had a considerable effect on the distribution of plants and

insects, and a knowledge of the life of the Polynesians, and particularly of their

REFERENCES. 101

crops and of the use which they made of vegetable products, assists one to dis-

tinguish organisms which may have been distributed by them.

But even if we have discovered something of the importance of the upper

air and of human agency, it remains very difficult to understand how the insects

of Polynesia reached their island homes ; it is impossible to assess the importance

of the traditional floating log. But it appears that there are two ways in which

the problem might be approached. The investigator of the future might try

to discover whether the Samoan insects contain a particularly high proportion

of forms breeding in unspecialized habitats, for instance, damp soil, dung and

rotten wood, and whether those whose habits are specialized (insects inhabiting

caves, associated with particular plants, parasitic on other insects, etc.) are few.

An attempt should also be made to study the effect of sea water on all stages

of the insects which have a wide distribution in tropical islands. Special atten-

tion should be given to those which are characteristic of the strand and which

live in association with the strand plants whose seeds are known to survive

long exposure to sea water.

V. REFERENCES

ANGENHEISTER, G. (1924). A summary of the meteorological observations of the Samoan

Observatory (1890-1920). (Edited by KE. Marsden and D. M. Y. Somerville). Wellington.

Batt, 8. C. (1918). Migration of insects to Rebecca Shoal Lighthouse and Tortugas Island, with

special reference to mosquitos and flies. Carnegie Inst., Washington, Publ. No. 252 (papers

Jrom Dept. Marine Biol., vol. 12).

BeruanpD, L. (1926). Les Araignées ubiquistes, ou a large répartition, et leurs moyens de dis-

sémination. C.R. Soc. Biogeogr., No. 23, pp. 65-67.

Beruanp, L. (1934a). Les Araignées du Pacific. Contrib. a Vétude du peuplement zool. et bot.

des tiles du Pacifique. Soc. Biogeogr. Mém. 4, pp. 155-180. Paris.

BErRianD, L. (1934b). Araignées de Polynésie. Ann. Soc. Ent., France, vol. 103, pp. 321-336.

Bezzi, M. (1928). The Diptera Brachycera and Athericera of the Fiji Islands. British Museum,

Londen.

Biacksurn, T. and Suarp, D. (1885). Memoirs on the Coleoptera of the Hawaiian Islands.

Trans. Dublin Soc. (2), vol. 3, pp. 119-290.

BristowE, W. S. (1930). The distribution and dispersal of spiders. Proc. Zool. Soc. 1929,

pp. 633-657.

Bristowr, W. 8. (1931). A preliminary note on the spiders of Krakatau. Proc. Zool. Soc.,

1931, pp. 1387-1400.

Bryan, K. H. (1934). The contributions of Bishop Museum to Polynesian biogeography. Contrib.

a Vétude du peuplement zool. et bot. des iles du Pacifique. Soc. Biogeogr. Mém. 4, pp. 281-288.

Paris.

Buxton, P. A. (1926). The colonization of the sea by insects: with an account of the habits

of Pontomyza, the only known submarine insect. Proc. Zool. Soc. 1926, pp. 807-814.

102 INSECTS OF SAMOA.

Buxton, P. A. (1927). On the habit of Hwploea spp. males of frequenting dead twigs of Touwrne-

fortia. Proc. Ent. Soc. Lond., vol. 1 (1926), p. 36.

Buxton, P. A. and Hopkins, G. H. E. (1927). Researches in Polynesia and Melanesia. No. 1,

Mem. Series, London Sch. Hyg. & Trop. Med.

CurEsMaN, L. KH. (1927). A contribution towards the Insect Fauna of French Oceania: Part I.

Trans. Ent. Soc. Lond., vol. 75, pp. 147-161.

Coan, B. R. (1931). Insects captured by airplane. Yearbook, U.S. Dept. Agric., 1931, pp.

320-323.

CoLLENETTE, C. L. (1925a). The present status of Danaida plexippus L. in the Pacific Islands.

Lint. Mon. Magq., vol. 61, pp. 198-202.

CoLLENETTE, C. L. (1925b). A Noctuid moth in Mid-Pacific. Ent. Mon. Maq., vol. 61, p. 206.

CoLLENETTE, C. L. (1928). The Arctiidae, Noctuidae and Sphingidae of the “St. George ”

Expedition, from French Oceania. Trans. Ent. Soc. Lond., vol. 76, pp. 469-487.

Doanr, R. W. and Ferris, G. F. (1916). Notes on Samoan Coccidae with descriptions of new

species. Bull. Ent. Res., vol. 6, pp. 399-402.

Dozimr, H. L. (1928). Two new Aleyrodid (Citrus) pests from India and the South Pacific.

Journ. Agric. Res., vol. 36, pp. 1001-1005.

Epwarps, F. W. (1926a). Philippine Nematocerous Diptera. I. Tipulidae. Notulae Hnto-

mologicae, vol. 6, pp. 33-44.

Epwarps, F. W. (1926b). On marine Chironomidae (Diptera) ; with descriptions of a new genus

and 4 new species from Samoa. Proc. Zool. Soc., 1926, pp. 779-806.

Epwarps, F. W. (1926c). Mosquito notes, VI. Bull. Ent. Res., vol. 17, pp. 101-181.

Epwarps, F. W. (1929a). Mosquito notes, VIII. Bull. Ent. Res., vol. 20, pp. 321-343.

Epwarps, F. W. (1929b). British non-biting midges (Diptera, Chironomidae). Trans. Ent.

Soc. Lond., vol. 77, pp. 279-480.

Epwarps, F. W. (1932a). Diptera, family Culicidae. Genera Insectorum.

Epwarps, F. W. (1932b).. Marquesan Simuliidae. Pacif: Entom. Surv. Publ. No.1. Honolulu.

Epwarps, I. W. (1933a). Tahitian Simulidae. Paci. Entom. Surv. Publ. No. 6, Article 7.

Honolulu.

Epwarps, F. W. (1933b). Mycetophilidae, Culicidae and Chironomidae, and additional records

of Simuliidae, from the Marquesas Islands. Pacif. Entom. Surv. Publ. No. 7, Article 6.

Honolulu.

Fett, E. P. (1926). The physical basis of insect drift. Nature, vol. 117, p. 754.

Fett, E. P. (1928). Dispersal of insects by air currents. N.Y. State Mus. Bull. 274.

Frereuson, EH. W. (1927). Tabanidae of the Samoan Islands. Bull. Ent. Res., vol. 17, pp. 815-316.

Ferris, G. F. (1933). Contributions toward a monograph of the sucking lice, Part VI. Stanford

Univ. Pub. ; Biol. Scr., vol. 2, p. 30.

Friepericus, K. (1913). Uber den gegenwirtigen Stand der Bekimpfung des Nashornkifers

(Oryctes rhinoceros L.) in Samoa. Tropenpflanzer, Year 17, Nos. 10, 11 and 12.

Frieprericus, K. and Demanpt, H. (1922). Weiteres iiber den indischen Nashornkifer. Zezts.

angew. Entom., vol. 8, pp. 295-324.

Furton, F. A. and Coamperiin, J. C. (1931). A new automatic insect trap for the study of

insect dispersion and flight associations. Journ. Keon. Entom., vol. 24, pp. 757-761.

Guppy, H. B. (1906). Observations of a naturalist in the Pacific between 1896 and 1899. Mac-

millan & Co., London.

HoutpuHaus, K. (1908). Kritisches Verzeichniss der bisher von den Samoainseln bekannten

Orthopteren. Denkschr. math.-naturwiss. Kl. K. Akad. Wiss. Wien, vol. 84, pp. 537-562.

Hoipuaus, K. (1929). Die geographische Verbreitung der Insekten. Schréder’s Handb. der

Entom., vol. 2.

REFERENCES. 103

Houpuaus, K. (1934). Die Verbreitung der Insekten auf den Inseln des Pacifischen Oceans.

Contrib. a Vélude du peuplement zool. et bot. des tiles du Pacifique. Soc. Biogeogr. Mem. 4,

pp- 201-218. Paris.

Hopkins, G. H. H. (1927). Pests of economic plants in Samoa and other island groups. Bull.

Ent. Res., vol. 18, pp. 23-32.

Inns, A. D. (1934). Textbook of Entomology. London. Third Edition.

Jupson, I. P. (1912). The Rhinoceros beetle (Oryctes rhinoceros) in Samoa. Dept. Agric. £11,

Bull. No. 3.

Karny, H. H. (1929). On the geographical distribution of the Pacific Gryllacrids. 4th Pacif.

Sct. Congress, pp. 157-172.

Kuznetsov-Ucamsky, N. N. (1926). Beitrage zur Zoogeographie der Tenthredinoidea. Bull.

Umw. Asie Cent. Tashkent, vol. 13, pp. 87-98 ; vol. 14, pp. 79-101 ; vol. 15, pp. 131-161.

Lever, R. A. (1934). On a Euploeine association on a small island near Santa Isabel in the

Solomons ; also a further observation on Euploeas attracted by dead leaves of Touwrnefortia.

Proc. R. Ent. Soc. Lond., vol. 9, p. 57.

Macrin, J. W. 8. (1933). Ceratopogonidae from the Society Islands. Bishop Mus. Bull. 114.

Honolulu.

Mauuocg, J. R. (1932). New Species of Calliphora from the Marquesas, with notes on Sarcophaga

- taitensis Schiner. Pacif. Entom. Surv. Pub. I, Article 2. Honolulu.

MarsHatL, G. A. K. (1921). On the Curculionidae of the Samoan Islands (Coleoptera). Proc.

Hawai. Ent. Soc., vol. 4, pp. 585-600.

Mautik, 8. (1929a). New injurious Hispinae. Bull. Hint. Res., vol. 20, pp. 81-94.

Mautik, 8. (1929b). Injurious Hispinae from the Solomon Islands. Bull. Ent. Res., vol. 20,

pp. 233-239.

Mayor, A. G. (1924). Rose Atoll, American Samoa. Papers from Dept. Marine Biol., Carnegie

Inst., Washington, vol. 19, pp. 73-79.

Meyrick, H. (1926). On Micro-lepidoptera from the Galapagos Islands and Rapa. Z'rans. Ent.

Soc. Lond., vol. 74, pp. 269-278.

Meyrick, I. (1928). The Micro-Lepidoptera of the “St. George”? Expedition. Trans. Ent.

Soc. Lond., vol. 76, pp. 489-521.

Meyrick, H. (1929). Pacific Pyrales of the “St. George ’ Expedition. Z'rans. Ent. Soc. Lond.,

vol. 77, pp. 155-169.

Muir, F. (1921). On some Samoan Fulgorids (Homoptera). Proc. Hawaii. Ent. Soc., vol. 4,

pp- 564-584.

Muir, F. (1931). Descriptions and records of Fulgoroidea from Australia and the South Pacific

Islands, No.1. Rec. Austral. Mus., vol. 18, pp. 63-80.

Mumrorp, H. P. and Apamson, A. M. (1934). Entomological researches in the Marquesas

Islands. Contrib. a Vétude du peuplement bot. et zool. des tles du Pacifique. Soc. Biogeogr.

Mem 4. Paris.

Murpeny, R. C. (1929). Birds collected during the Whitney South Sea Expedition. IX. Zostero-

pidae from the Solomon Islands. X. On Plerodroma cookii and its allies. Amer. Mus. Nov.

365 and 370.

Paine, R. W. and Epwarps, F. W. (1929). Mosquitos from the Solomon Islands. Bull. Ent.

Res., vol. 20, pp. 803-316.

Perkins, R. C. L. (1913). Introduction, being a review of the land-fauna of Hawaua. Fawna

Hawaiiensis, vol. 1, Part 6.

PERKINS, R. C. L. (1932). On two new species of Proterhinus from the Marquesas and the inclusion

of this genus in the family Aglycyderidae. Pacif. Ent. Surv. Pub. 1, Article 3. Honolulu.

104 INSECTS OF SAMOA.

Poutron, E. B. (1932). Observations on . . . the association of male Euploeine butterflies

with . . . Tournefortia argenteus in Fiji. Proc. Ent. Soc. Lond., vol. 6 (1931), pp. 77-80.

Poutton, E. B. and Ritey, N. D. (1928). The Rhopalocera of the “‘ St. George.” Expedition

from French Oceania. Trans. Ent. Soc. Lond., vol. 76, pp. 453-468.

tarnpow, W. J. (1897). The insect fauna. The Atoll of Funafuti. Austral. Mus., Sydney,

Mem. 3, pp. 89-104.

Reset, H. (1910). Lepidopteren von den Samoainseln. Bot. und zool. Ergebnisse. Denkschr.

math-naturwiss. Kl. K. Akad. Wiss. Wien, vol. 85, pp. 412-482.

Recuincer, Kk. (Editor, 1907-1914). Botanische und zoologische Ergebnisse einer Wissen-

schaftlichen Forschungsreise nach den Samoainseln dem Neuguinea-archipel und den

Salomon-inseln. Denkschr. math.-naturwiss. Kl. Kk. Akad. Wiss. Wien, vol. 81, pp. 197-317 ;

vol. 84, pp. 885-562 ; vol. 85, pp. 175-432 ; vol. 88, pp. 1-65; vol. 89, pp. 443-708 ; vol. 91,

pp. 139-213.

Rerecke F, (1898). Die Flora der Samoa-inseln. Botan. Jahrb., vol. 23, pp. 237-368 ; vol. 25,

pp. 578-708.

Scorr, H. (1933). General conclusions regarding the insect fauna of the Seychelles and adjacent

islands. Trans. Linn. Soc. Lond., 2nd Ser. Zoology, vol. 19, pp. 307-391.

SeErcHELL, W. A. (1924). American Samoa; Part 1, vegetation of Tutuila Island. Part 2,

ethnobotany of the Samoans. Dept. Marine Biol., Carnegie Inst., Washington, vol. 20.

Sitvestri, F. (1930). Contribuzione alla conoscenza degli Japygidae (Thysanura) della regione

australiana. Boll. Lab. Zool. Portici, vol. 23, pp. 210-226.

Swuzpy, O. H. (1924). Notes of insect pests in Samoa. Proc. Hawaw. Hnt. Soc., vol. 5, pp.

385-393.

Taytor, F. H. (1934a). A check list of the Culicidae of the Australian Region. Service Publication

(School of Pub. Health & Trop. Med.), No. 1. Sydney.

Taytor, F. H. (1934b). The Diptera of the Territory of New Guinea. I. Family Culicidae.

Proc. Linn. Soc. N.S. Wales, vol. 59, pp. 229-236.

THomson, A. (1925). Upper air observations, 1923-1924. Apia Observatory. New Zealand.

Trttyarp, R. J. (1924). The dragon-flies (order Odonata) of Fiji. Tans. Ent. Soc. Lond., 1923,

pp. 305-346.

TrtLyarD, R. J. (1926). The insects of Australia and New Zealand. Sydney.

ToxunaGa, M. (1932). Morphological and biological studies on a new marine Chironomid fly,

Pontomyia pacifica, from Japan. Part 1. Morphology and taxonomy. Mem. Coll. Agric.

Kyoto Univ., No. 19, pp. 1-56.

Torui11, J. D., Tayior, T. H. C. and Painn, R. W. (1930). The coconut moth in Fiji. London,

Imp. Bureau of Entomology.

Van Dyke, EH. C. (1932). Microgonus, new genus, and Rhyncogonus, from the Marquesas.

Pacif. Ent. Surv. Pub. 1, Article 4. Honolulu.

VisHER, 8. 8. (1925). Tropical cyclones and the dispersal of life from island to island in the

Pacific. Amer. Nat., vol. 59, pp. 70-78.

Visuer, 8. 8. and Hopes, D. (1925). Australian hurricanes and related storms, with an appendix

on hurricanes in the South Pacific. Commonw. of Austral., Bureau of Meteorol. Bull. 16.

Waker, J. J. (1914). The geographical distribution of Danaida plexippus L. with special

reference to its recent migrations. Ent. Mon. Maq., vol. 50, pp. 181-193 and 224-237.

Wecener, A. (1924). The origin of continents and oceans. (Translated from 3rd German

edition). London, Methuen.

WiuuiaMs, C. B. (1930). The migration of butterflies. Edinburgh.

Woop-Jongs, F. (1910). Coral and atolls. London.

PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS, LTD., LONDON AND BECOLES.

INSECTS OF SAMOA

AND OTHER SAMOAN TERRESTRIAL

ARTHROPODA

LIST OF PARTS AND SYSTEM OF PUBLICATION :—

Part 1. Orthoptera and Dermapiera. (Complete.)

ns II. Hemiptera. (Complete.)

» III. Lepidoptera. (Complete.)

» IV. Coleoptera. (Complete.)

i. V. Hymenoptera. (Complete.)

MI. Diptera. (Complete.)

» WII. Other Orders of Insects. (Complete.)

» WIII. ‘Terrestrial Arthropoda other than Insects. (Complete.)

» LX. Summary and Index.

__. The work is published at intervals in the form of numbered fascicles.

Although individual fascicles may contain contributions by more than one

author, each fascicle is so arranged as to form an integral portion of one or

| other of the Parts specified above.

List of Fascicles issued to 8th June, 1935 :-—

Insects of Samoa and other Samoan Terrestrial Arthropoda. Maps | and Date Issued.

2 (in envelope). 1927, 4to. 6d. ; 26th February, 1927.

Part |. OrTHOPTERA AND DERMAPTERA.

Fasc. 1, Dermaptera. By Dr. Alfredo Borelli. Pp. 1-8. 1928, 4to. Is. 28th July, 1928.

Fasc.2. Orthoptera. By Dr. L. Chopard. 5] text-figures. Pp.9-58. 1929, 4to. 5s. 26th January, 1929.

Part II. Hemiptera.

Fasc. 1. Fulgoroidea. By F. Muir. 25 text-figures. Psyllide (Chermide). By

Prof. D. L.. Crawford. 4 text-figures. Coccide, Aphidide and Aleyrodide.

By F. Laing, M.A., B.Sc. 3 text-figures. Pp. 1-45. 1927, 4to. 2s. 6d. 25th June, 1927.

Fasc.2. Cercopide. By V.Lallemand, M.D. 10 text-figures. Cicadide. By

G. Myers, Sc.D. 22 text-figures. Aquatic and Semi-aquatic Heteroptera.

k, Prof. He Esaki. 6 tert figures, Pp. 47-80. 1928, dt. 2s. 6d. : 23rd June, 1928.

Fasc. 3. Heteroptera. By W. E. China, B.A. (Cantab.). 28 text-figures.

Pp. 81-162. 1930, dto. 5s. 26th July, 1930.

Fasc. 4. Cicadellide (Jasside). By Herbert Osborn, Ohio State University, Bion

Columbus, Ohio. 15 text-figures. Pp. 163-192. 1934, 4to. 2s. 27th January, 1934.

oa ope a Eee: By Harry H. Knight, Ph.D. 9 text-figures. Ded Fabri 1935,

Part II]. LepipoprTera.

Fasc. |. Butterflies of Sa d hbouring Island-gr By G. H.E.

TSM BES fa techn: al ones Pe Fou 1927 he. Se 9h Aprib 1927,

Fike, 2 Micro-Lepidoptera By Edward Mesmick, BA. FARO: Pa 65-116.

1927, 4to 28th May, 1927.

Pec aC: rticde By Lam BUPonEOELE S(O reer an ace :

TOMI doe Ba takr in He NOAH Mosck 1928.

Pa a pee Be OW Fans 10 een Ae iat ane

Pp, (69-290, 1935, 40. 1 D8 February, 1935,

Part IV. CoLEopTeRA.

Fasc. 1. Carabide. By H. E. Andrewes. 9 ree: Dytiscide. By A.

Zimmermann. 2 text-figures. Staphylinide. By M. Cameron, M.B. 2 text-

figures. Hydrophilide. By A. d’ Ble fies ire: Clavicormia and

Lamellicornia. By G. J. Arrow. 3 text-figures. Pp. ]-66. 1927, 4to. 3s. 19th December, 1927,

Fasc. 2. Heteromera, Bostrychoidea, praleced wats ae Buprestide. By K.G.

Blair, BSc. 14 text-figures. aia H. van Zwaluwenberg. 10 |

text-figures. Melaside (Eucnemide). yi RP Fleutiaux Cerambycide. By

Chr. Aurivillius. 1 plate. ae ide. y R. (Seer 4 text-figures.

Anthnbide. By Karl eee Ph.D. 11 text-figures. Proterhinide. By

R.C. L. Perkins, D.Sc., F.R.S Pp. 67-174, css Ato. 5s. 25th February, 1928.

List of Fascicles issued to 8th June, 1935 (continued) :—

Parr IV. CoLropTerRA—continued.

Fasc.3. Throscide. By K. G. Blair, B.Sc. 1 text-figure. Chrysomelide. Date Issued.

By S. Maulik, M.A. 18 text-figures. Pp. 175-215. 1929, 4to. 2s. 6d. 23rd February, 1929.

Fasc. 4. Platypodide and Scolytide. By C. F. C. Beeson, DSc. 13 text- / Ae

figures. Pp. 217-248. 1929, 4to. 2s. 6d. 22nd June, 1929.

F; I S at

age Bee ers Awe me Neha C.M.G., D.Sc., F.R.S. 31 text 25th April, 1931.

Part V. HyMENOPTERA.

Fasc. 1. Apoidea, Sphecoidea, and Vespoidea. By R. C. L. Perkins, D.Sc.,

FR, beZise

. and _L. Evelyn Cheesman, i 12 text-figures. Larride. UE ate

By Francis X. Williams. 12 text-figures. Formicide. By Dr. F. Santschi. —

9 text-fisures. Pp. 1-58. 1928, 4to. 5s. 25th February, 1928.

Part VI. Dyprera.

Fasc. |. Streblide and Nycteribiide. By L. Filcer a text-figures. Hippo- che Sue

boscide. By G.F. Ferris. 6 text-figures. Pp. 1-21. 1927, 4to. 2s. 6d. 23rd July, 1927.

Fasc.2. Nematocera. By F. W. Edwards, M.A. 20 text-figures. Cecidomyine.

By H. F. Barnes, B.A., Ph.D. 4 text-figures. Pp. 23-108. 1928, 4to. 5s. 23rd June, 1928.

Fasc. 3. Stratiomyide, Tabande: and Senin By Gertrude Ricardo. 6 text- a

figures. Larve of Stratiomyiide. P. A. Buxton, M.A. 2 text-figures.

Re epee, By C.G Col anke sb. 8 text-figures. Sarcophagide. By

A. B MA. 9 text-figures. Muscide. By J. R. Malloch. ana ut

Pp. 109-175. 1929, 4to. 5s. lth May, 1929.

Fasc. 4. Empidide ae Pi Ninebar By J. E. Collin. | 7 Fore nelnes: ees

Syn e. By F ie rae ae ie Cane ures rani eee ; on

text- iy Al hin H iets

doe Onn e PaO caiman 27th July, 1929.

F By Il 6 = iphoride. : Bip h

rR Met Bo. pide mS ait vee Ba figures. Calliphoride. By 22nd March, 1930.

F 6. L dz, Chl d Pi . R. Malloch.

ae Ree Pe O82 a0 de lee ne Zand Naternber 180.

Fasc. 7. Trypetide. By J.R. Malloch. | text-figure. Pp. 253-266. 1931,4to. Is. 28th November, 1931.

Fasc. 8. Drosophilide, Ephydride, Sphzrocerid d Milichiide. By J. R. at 5

Malloch. 16 vale forrest ps, "261-38. : 1934, te” 2s. hae Saeums 23rd June, 1934.

Fasc. 9. Phoride, Asromyzide, Micropezide, Tachinide and Sarcophagid es igo

“By [Re Malloch.” 15 texeheures:) Pr 329"366. 1935! 4to) OGii yt (237m Meta Fae

Part VII. Oruer Orpers or INsEcTSs. a ae a

Fasc. 1. Isoptera: Family Termitide. By Gerald F. Hill. 14 text-figures and

plate. Odonata. By Lt.-Col. F. C. Fraser, MS., F.ES. 5 text-figures. :

Pp. 1-44. 1927, 4to. 2s.6d. 28th May, 1927.

Fasc. 2. Plectoptera. By R. J. Dhee eer (Cantab.), F.R.S., and J. A. Wis,

Lestage. 2 text-figures. Siphonaptera. By P. A. Buxton, M.A. Thysanoptera.

ay Richard: Bagnall, FRSE. oe LS. Te rceRoe Pp. nani 1928, 4to.

3 :

23rd June, 1928. :

Fasc. 3. Mallophaga. By J. Waterston, D.Sc. 2 text-figures. Anoplura. By ne i as

Bs Buster, M.A. Trichoptera. By Martin E. Mosely. {1 figure. oe 4

Neuroptera. By P. Esben-Petersen. 1 text-figure and 2 plates. Apterygota. One a

By George H. Cee DSc. 32 text-figures. Pp. 77-116. 1928, 4to. 2s.6d. 28th July, 1928. :

Fase. 4. P ter By Dr. H. H. Kar 8 text-fieur Pp. 117-129.

1932 deol Mig me Bens ee arin Fi 27th February, 1932.

Part VIII. TerrestrisL ARTHROPODA OTHER THAN INSECTS.

Fasc. |. Isopoda Terrestria. By Harold G. Jackson, D.Sc. 2 plates. Scor-

pono eh P.A. Be ae panda cee os porcine

text- ft 1 text-; = ' =

fie Suid carina. By Stanley Hirs ext-figures. Pp 23rd July, 1927.

Fasc. 2. _Myriopoden (Myriopoda). By C. Attems. 4 text-figures. Araignées a

(Araneida). By Dr. [en Bee 79 text-figures. Pp. 29-78. 1929, 4to. 2s. 6d. 22nd June, 1929.

Part IX. SumMary AND INDEx. :

Fasc. 1. Descripti f the E t. By P. A. Buxton, M.R.CS 2 a

Tislecarcs Gad € ote (PEICa OA edt ae “22nd November, 1930

Fasc. 2. Summary. By Prof. P. A. Buxton, Pp. 33-104. 1935, 4to. 5s. 8th June, 1935.

Nay

fe ae

‘ tne ‘

sic,

Eso tre,

os

ecenrsn SS

eS ke