>7'-

A NATURALIST ON
LAKE VICTORIA

WITH AN ACCOUNT OF
SLEEPING SICKNESS
AND THE TSE-TSE FLY

Digitized by tine Internet Archive

in 2010 with funding from

Open Knowledge Commons and Harvard Medical School

http://www.archive.org/details/naturalistonlake1920carp

PLATE I.

Acraeinv models NyinpIuxUrui rrvunics .

lO

E.C Kmlht d'^l.e% f hronio Huth imp.

MODEL SPECIES OF PLANEMA WITH MIMETIC FORMS OF

PSE UDACRAEA ■ EURYTUS. ^/s ruxt s iz e .

HeprodiLced, hy perrru.sswny of the EnJx/malocjicaJi Soa^ety of London,.

(Ft 0 njti^p i&oe^J

tr-5^ T ■<- ^T'* T T

o

a\k Xjl jl^ jl

VICTORIA

" LEEPING

1 aTAJ^ -^

.IX6X ,1117 SI ,Bho*oi7 .J .Bajslal emoH , ? zwXhi jiwswaS^ .1

it^f SaaUl £^im(I ,'$\.4'^'^a4'' idiot' zwi^ws »5)»toJiSlwa^^ .S

.0161 tiirr 51 ,9cld8itaa , t) »Un»a»«v ,V\ .6

.8X61 tin I jbnfilel ^msW j t* V^ai^oA mioi awi^xija .a^ .9
.:tIGI ,17 es .BriBlei 9mo3 , $ JoUnjoo^iW A^ .Y

CB. Met ,iil Si ibxtBlgl o§jBlua! , $ ^a'^'^o^^ .5^ .Q

;i:XGX ,iiiv Zl , ? as^ji-oa'^'Qoc^ mioi ^w^^^•m^ .21 .01
inedaea ,no§l3 .dM lo dooi daaw laea avsald .A .8 ^d ddguaO

1 AV

•^L> ^wU^VirAlM 1

iil\ UE, NEW YORK

PLATE t

Planema tellus ? , Kome Island, L. Victoria, 12 viii, 1914.
Caught by native.

2. Pseudacraea eurytus form terra ? , Damba Island, 1-15
IX, 1911.

3. PL epaea form paragea ? , Bugalla Island, 24 ii, 1912.

4. Ps. eurytus form obscura $ , Bugalla Island, 18 ii, 1912.

5. PI. macarista 3" » Entebbe, 15 viii, 1910.

Caught by one of Dr. Wiggins' natives*

6. Ps. eurytus form hobleyi ^ , Wema Island, 1 lii, 1918.

7. PI macarista $ , Kome Islatid, 29 Vt, 1914.

8. Ps. eurytus form tirikertsis $ , Bugalla Island, 6 xil, 1912* J

9. PI. poggei ? , Bulago Islalid, 16 lii, 1914*

lO. Ps, eurytus form poggeoides $ , 13 Vitij 19lli

Caught by S. A. Neave near west foot of Mt* Elgon, Easterri
Uganda^

A NATURALIST
ON LAKE VICTORIA

WITH AN ACCOUNT OF SLEEPING
SICKNESS AND THE TSE-TSE FLY

y^ BY ^_^

c^'d! hale carpenter

D.M., B.Ch. (Oxon), Uganda Medical Service

Fellow of the Linnean, Entomological, and Zoological Societies

of London

WITH 2 COLOURED PLATES, A MAP,
CHARTS, AND 87 ILLUSTRATIONS

E. P. BUTTON AND COMPANY
68 1 FIFTH AVENUE, NEW YORK

0£^C 14/929

i^^hAK V

i^irs^ publisJied in 1920

^^^/S^^'^-^

■z?

(i4W ^-j^Afe reserved)

PRINTBD IN ORBAT BRITAIN

TO THE
MEMORY OF MY FATHER

P. HERBERT CARPENTER, D.Sc, P.R.S.

AND TO MY DEAR MOTHER

THIS, MY FIRST BOOK,

IS DEDICATED

" It appears to me the doing what little we can
to increase the general stock of knowledge is as
respectable an object of life as one can in any
likelihood ptirsue."

Letter from Charles Darwin to his sister, 1833.

mTRODUCTION

At the commencement of 1910, when studying Tropical
Medicine at the London School, I was asked if I would
undertake an investigation into the bionomics of the
Tse-tse fly, Glossina palpalis, in Uganda, as the Tropical
Diseases Committee of the Royal Society had come to
the conclusion that a greater knowledge of this fly was
essential for the successful dealing with the disease.

I left England in June 1910, and during the second half
of this year and beginning of 1911 worked at Jinja,
on the north shore of the Victoria Nyanja,^ where the
Nile takes origin, and endeavoured to famiharize my-
self with the novel surroundings and with the fly.

It soon appeared that residence on the completely
depopulated islands known as the Sesse Isles, in the
north-west part of the lake, would afford ideal conditions
for studying Glossiyia unaltered by the presence of man-
kind and his surroundings ; accordingly, in February
1911, I went to Nsadzi Isle, which, lying south of Entebbe
and within two or three hours of it by canoe, had been
recommended by the P.M.O. as suitable for a beginning.
This island, however, did not furnish all the conditions
required for the investigations, and camp was moved,
after a fortnight, to the large Island of Damba to the
east and on the Equator, where the rest of the year was
spent.

^ Nyanja, not Nyanza. The former is Luganda for lake, the latter
means nothing, and is erroneously copied from one book to another.

X INTRODUCTION

In January 1912, a new site was chosen on Bugalla,
the largest island of the Sesse group, lying in the north-
west corner of the lake, and fourteen months of unin-
terrupted residence were spent there, after which I went
home on leave, returning in December 1913. On arrival
at Entebbe I met Mr. W. F. Fiske, from the States,
who had been deputed by the Tropical Diseases Bureau
of London to investigate Glossina morsitans, and had
decided to first make the acquaintance of palpalis,
the problems connected with which seemed more
readily soluble, and by their solution might throw light
on morsitans. He had decided, as a result of several
months' previous work, to undertake a tour of the chain
of islands lying parallel with the north shore of the lake,
in order to compare one with the other, hoping that the
presence or absence of Glossina might be found to be
correlated with definite factors. He proposed that
I should join him, and we made the tour together, very
greatly to my benefit not only regarding Glossina in
particular, but in field work in general, his wide experi-
ence of field entomology and acute powers of observation
showing me how much I had missed when working by
myself. At the close of a tour of about two and a half
months Fiske left for another part of Uganda, and I
settled down on the western spur of Kome Isle to carry
out a year's continuous cycle of experiment and observa-
tion on the smaller islets within working distance
(Ngamba, Nsadzi, Kizima, Tavu, Bulago, Kimmi).
Unfortunately this was brought to an end by the out-
break of war ; for in August 1914 I was called off for
duty with the troops on the Uganda-German frontier,
and subsequently with the Belgian Northern column
under Colonel Molitor, which invaded German East
Africa via Kigezi and Ruanda, to Tabora, after which
my duties took me to various parts of German East
Africa and Portuguese East Africa. When released from

INTRODUCTION xi

active service in November 1918, I returned to Uganda,
and had a further most interesting tour among the same
islands that were visited with Mr. Fiske in 1914.

This book is an attempt to give an account of the
life on the islands of the Victoria Nyanja : and very
nearly aU the examples mentioned were met with there,
between February 1911 and August 1914. During active
service, however, other examples were met with, an
account of which seemed not to be out of place in this
book because of their bearing on certain of the points
discussed. The fascinating question of the Colouration of
Insects has been treated at some length, but mainly from
the point of view of the bearing on it of examples met
with on the islands. Consequently this chapter must
not be taken as a discussion of the relative merits of
rival theories from an abstract, or " arm-chair," point
of view. Similarly the chapter on Glossina must not
be taken as a complete summary of all that is known,
since this book deals almost entirely with some of
my own observations and work.

The chapters on Mammals, Birds and Reptiles, I am
aware, are dreadfully scanty. In defence, I can only
urge that I am primarily an entomologist ; and my
knowledge of field botany is a minus quantity!

I am indebted to the Royal Society and the Controller
of His Majesty's Stationery Office for permission to
reproduce illustrations and other matter that have
appeared in my reports to the Sleeping Sickness Com-
mission.

The beautiful photograph of Glossina is the work of
Mr. A. Robinson, of the Oxford University Museum.

Grateful acknowledgment is also due to the Council
of the Entomological Society of London, for permission

xii INTRODUCTION

to reproduce the two plates of butterflies which have
appeared in the Transactions for 1920. Many of the
observations on Insects have already appeared in the
publications of the Society, others in Bedrock, and some
of those on the Fossorial Hymenoptera in the Journal of
the East Africa and Uganda Natural History Society.

Mr. A. L. Shephard, of the Tropical Diseases Bureau,
very kindly gave me references which were of great help
in the preparation of the chapters on Sleeping Sickness,
and thanks are also due to the Director, Dr. A. G. Bagshawe,
C.M.G., in this matter. Dr. H. Eltringham most kindly
complied with a request for the use of his talent as a
draughtsman, and provided an original and humorous
drawing of an imaginary example of mammalian mimicry.

To my friend and mentor, E. B. Poulton, F.R.S.,
Hope Professor of Zoology at Oxford, I owe more than
I can say for his ever ready advice and encouragement,
for his kindness in writing the Preface, and in
particular for reading the MS. of Chapter XI, and
suggesting improvements therein. To him and Dr. S.
A. Neave are due the selection and arrangement of
specimens illustrating by two plates forms of Pseudacraea
eurytus and their models. Dr. Karl Jordan, of the Tring
Museum, was most kind in showing me Lord Rothschild's
collection of this wonderful species, including a unique
specimen of a new form which shows the geographical
range to extend further southwards than has been
known hitherto.

G. D. H. C.

PEEFACE

BY

PROFESSOR POULTON

Every reader of this book will, I feel sure, recognize
that it contains a really wonderful body of observations ;
especially so if the brief time in which they were made
be borne in mind — from February 1911 until the day
in August 1914 when, without a word of preparation or
a hint of warning, the author arrived in Entebbe to find
in full progress, and to bear his part in, the fiercest struggle
of human history. And even this short period of " island
life " was broken into by nearly a year's leave.

The observations recorded, whether dealing, now and
then, with men, or, throughout the book, with other
animals, are both loving and accurate. The two qualities
are closely associated, for the love of living beings renders
the patient study of them an unceasing fascination and
delight.

Apart from the chapters on Sleeping Sickness and its
carrier, the Tse-tse fly, the most important discoveries
are those which add to our knowledge of Papilio dardanus,
and, above all, the breeding experiments which brought
final confirmation to the conclusions of Dr. Karl Jordan
upon that wonderful series of mimetic butterflies here
proved beyond doubt to be forms of a single species —
Pseudacraea eurytus of Linnaeus.

xiv PREFACE

It is a curious fact that the African butterflies — far
less numerous in species than those of the tropical East,
and outnumbered even more by the American — should
include what are probably the two most remarkable
examples of mimicry in the world — Papilio dardanus
and Ps. eurytus ; furthermore, that they should illustrate
such different aspects of the subject.

The Swallowtail, dardanus, whose non-mimetic male
is accompanied by a very similar non-mimetic female in
Madagascar, the Comoro Islands, Somaliland, and Abys-
sinia (although two single but different mimetic females
were once taken in this last locality), is represented,
wherever it occurs on other parts of the African continent,
by females entirely different from males, and either
mimetic of other butterflies, or, far more rarely, inter-
mediates between these mimetic forms, or persistent
stages enabling us to retrace the history of their origin.
Until 1903 the mimetic females were all believed to
resemble the commonest Danaine butterflies of each
locality — often mimicking in a single area three different
species with widely different patterns.

This interesting and complicated story became still
further involved, when, on October 7, 1903, the late Mr.
Roland Trimen brought before the Entomological Society
an account of a new and entirely different female, mimick-
ing not a Danaine, but an Acraeine model, belonging to
the genus Planema. For some years the new mimic,
appropriately named planemoides, was only by strong
inference assumed to be a form of dardanus, but definite
proof was soon afforded by the study of a specimen
captured by Mr. T. T. Behrens, R.E., near the old southern
boundary of Uganda on the west shore of the Victoria
Nyanja. This single example was, on the left side, a
gynandromorph, viz. one of those rare individuals in
which the patterns of the two sexes occur combined.
In this case parts of the yellow and black non -mimetic

PREFACE XV

pattern of the male are, as it were, let into the mimetic
female pattern, clearly proving that the two belonged
to the same species. The specimen, now in the Hope
Collection of the Oxford University Museum, is figured
in the Transactions of the Entomological Society for 1906
(PL XVIII, Fig. 4).

The evidence of this specimen was very satisfactory,
but how much better was the proof by breeding from a
known jDarent, obtained by the author in Bugalla island,
Sesse Archipelago, when he captured on December 1, 1912,
a planemoides female, and reared from its eggs three
females like the parent, seven of the black-and-white
hippocoon females, and twelve non-mimetic males. And
later on, in 1915, when Medical Officer to the forces acting
on the southern frontier of Uganda, he reared one plane-
moides female, one trophonissa female (similar to hippocoon,
but with the main white areas replaced by orange), and
five males from the eggs of a captured female combining
the patterns of her two daughters. These are the only
occasions on which the planemoides female has been bred,
although rare forms which evidently represent an imper-
fect planemoides, many hundreds of miles away from its
model, have been bred by Mr. C. F. M. Swynnerton in
S.E. Rhodesia, and by Mr. G. F. Leigh, in the neigh-
bourhood of Durban.

The Pseudacraea mimics, belonging to the Nymphaline
sub-family, and allied to our own White Admiral {Limenitis
sybilla), illustrate the second great group of mimetic
butterflies in which the males resemble models as well
as the females. The models of the Ps. eurytus series,
with which this book deals, aU belong to the Acraeine
genus, Planema, but in some of them the sexes differ,
while in others they are alike. In Uganda there is one
common example of the first, mimicked by a Pseudacraea
whose male resembles the male and female the female,

xvi PREFACE

and also two examples of the second, each mimicked by
a Pseudacraea with sexes alike ; and when, in 1910,
Dr. Jordan announced that these three Pseudacraeas could
not be separated structurally from each other, from the
Liimean West African eurytus, or from other mimics west,
east, and south, which had been described as distinct
species, I felt that the obvious conclusion that they were
in fact all of them forms of one and the same species
ought to be confirmed by breeding experiments before
it gained acceptance. And it was the occurrence, in the
same area, e.g. Uganda, and ex hypothesi interbreeding,
of forms with sexes different and with sexes alike, which
raised the greatest difficulty, and at the same time
would prove of the deepest interest if the conclusion
were confirmed. In order to test it I wrote to
every naturalist known to me who was in a position
to undertake the work, pointing out its exceptional
interest and importance. I even tried to persuade the
late Mr. A. D. MiUar who had just been very successful
in breeding the Natal form of the Pseudacraea, to under-
take the journey to Uganda and apply his experience
there.

The complete success finally attained by the author
on Bugalla may be gathered from the results fully set
forth in Chapter XI. Dr. Jordan's conclusions were
thoroughly confirmed, and the large group of conspecific
forms, now proved to interbreed in the same areas, and
interbreeding, it may be legitimately inferred, with those
of other areas when they meet on the boundaries, the
group of Pseudacraea eurytus, took its place beside that
of Papilio dardanus as one of the greatest examples of
mimicry in the world.

This discovery was completed by another of equal
importance described in the same chapter, the evidence
that the mimetic patterns were only kept up to the mark
in islands where and when their models were relatively

PREFACE xvii

abundant. As these became scarce, so did the mimetic
Pseudacraeas run into each other more and more
completely through the intermediation of an increasing
number of transitional forms.

In concluding I should wish- to take this opportunity
of thanking all those naturalists who, during the past
quarter of a century, in distant lands, have helped me
in my work and have given me the great pleasure of helping
them ; and, among all, especially to thank the author,
Dr. G. D. Hale Carpenter, D,M., not only for the mutual
help which may be gathered from this book, but for
much further help in S.W. Uganda, and ex-German East
Africa ; the late C, O. Farquharson, in S. Nigeria, whose
death in the sinking of the Burutu almost at the end of
the war, was so great a loss to Natural History ; W. A.
Lamborn, in S. Nigeria, ex-G.E,A., and now in the
Federated Malay States ; Dr. G. A. K. Marshall, D.Sc,
C.M.G., in Natal and S. Rhodesia ; the Rev. Canon K.
St. Aubyn Rogers, M.A., in British and ex-G.E.A., the
late R. W. C. Shelf ord, M.A., in Borneo, another grievous
loss to science ; C. F. M. Swynnerton, in S.E. Rhodesia ;
and C. A. Wiggins, P.M.O., Uganda, in this country
and B.E.A. To all these and many others I offer most
grateful thanks for some of the greatest happiness I
have known.

EDWARD B. POULTON.
August 3, 1920.

1*

CONTEXTS

PAOB

Introduction . . . . . . . ix

Pkbface. By Professor E. B. Poulton, F.R.S. . . xiii

I. Sleeping Sickness . . . . .1

Symptoms — Ancient description — History of the
introduction of the Uganda epidemic — The
Rhodesian disease

II. Natural History of Sleeping Sickness . . 11
Trypanosomes, Tse-tse flies, and game

III. Natural History of Gloss ina palpalis . . 30

IV. The Lake . . . . . .67

Geographical data — Scenery — Weather — A storm —
Weather prophets — Noises of the night

V. Canoes and a Voyage . , . .84

Description of the Sesse canoe, and a voyage to
the southernmost islands of Sesse

XX CONTENTS

CHAPTER PAGE

VI. A Tour of Investigation among the Noethern

Islands . . . . . ,94

A number of islands compared as to their fauna

VII. Mammals . . . . . .126

Monkeys — Hippos — Situtunga — Otters — Bats —
Mongoose

VIII. Birds . . . . . . .145

IX. Reptiles and Fish ..... 180

X. The Colouration of Insects . . . 195

Procryptic — Anticryptic — Aposematic — Directive
marks — The habit of grouping together more
common in aposematic than procryptic species
— Enemies of protected insects — Sexual colouring
and courtship of butterflies — Mimicry, Batesian
and Mullerian — -Do birds eat butterflies ? — Longley's
revised hypothesis of mimicry

XI. PSEUDACRAEA EURYTUS . . . . 242

Some of its forms and varieties and their bearing
on the interpretation of mimicry by natural
selection

XII. Hymenoptera ..... 276

Ants — Wasps — Fossors or Sand-wasps — Bees

XIII. Sundry Insects ..... 305

Beetles — Dragon-flies — Mantis'pa — Mantidae —
Grasshoppers — Cockroaches — Earwigs — Flies —
Spiders — Myriapods — Crustacea — Molluscs

LIST OF ILLUSTRATIONS

Plate I. Model species of Planema with mimetic forms of Pseuda

craea eurytus (coloured plate) .... Frontispiece

FACING PAGE

Plate II. Model species of Planema with mimetic forms of Pseitda

craea eurytus (coloured plate)
Fly beach on Damba Isle .
Fly beach on Bugalla Isle
On Damba Isle ; not a good fly shore
A fly-boy at work ....
An Ambatch tree ....
Fly beach on Kimmi
Glossina palpalis and pupae

Uprooted tree on Kimmi Isle forming ideal shelter for pupae
Searching for pupae among grass tussocks on south point, Bulago

Isle

Tree trunk on Bugalla Isle under which many Olossina pupae could

always be found .......

Breeding place of Glossina at hollowed base of tree on Bugalla
" Enzibaziba " bushes on Damba Isle forming breeding ground
Shingle and sand ixnder sheltering Enzibaziba bushes where flies

have been seen to deposit larvae
The first artificial breeding ground, on Ngamba Isle
Artificial breeding ground on Tavu Isle
The Kagera river ....
The head of Napoleon gulf where the Nile arises
The eastern cataract of the Ripon falls
The birth of the Nile
Young Raphia palm ....
Wild date palm ....

A tangled mass of Ipomaea in flower
A bed of bracken on Bugalla
Abrupt margin of forest belt on Bugalla
Park land on Bugalla , .

from

the Lake

244
24
24
30
30
48
48
54
54

54

56
56
58

58
62
62
68
68
68
68
70
70
70
70
72
72

XXll

LIST OF ILLUSTRATIONS

My settlement on Bugalla Isle, 1912-13

Aerial roots dipping to ground, Bugalla forest

Giant twisted stem of creeper, Bugalla forest

Canoe returning with load of firewood

A canoe race .....

A halt on the east coast of Bugalla .

Pitching camp on the isthmus of Kirugu Isle

The western point of Kirugu Isle

On Kirugu Isle

On Kirugu Isle

The high ridge of Kuye Isle

South end of Buguye Isle

South end of Nkose Isle .

Canoe off Kirugu Isle

Bulago Isle, north bay and forested north point

The open grass land interior of Bulago

Large fig tree on Ngamba Isle

Nsadzi Isle, south beach .

Dwasendwe from Dwanga Mto,

Kisigalla Islet .

South beach of Nsadzi Isle

South-west end of Wema Isle

View South-east from Bugalla camp .

Wee Man choosing a grasshopper

Wee Man and Tommy taking grasshoppers

Wee Man being rude to a Tortoise

Tommy and Wee Man in an interval of a game

Puffin " hating " a Tortoise

A favourite haunt of Situtunga on Bugalla

A bat caught on tent at Jinja

Nest of Egyptian goose

Nest of stone ciu-lew

Nest of stone curlew

Nest of Osprey

View from camp on Kome

View towards Buvumira from Bugalla camp

Nesting place of kingfishers

Fly-catcher

Nest of crocodile

Canoe-man with Enswa-swa

A small Enswa-swa .

A very abundant tree frog

Fly-boy and Mamba

Semutundu . . .

LIST OF ILLUSTRATIONS

XXlll

Male ....

Caterpillar of Aterica galene

Pupae of Papilio dardanus

Moth at rest on low foliage

Aposematic caterpillars

Diagram of ant and typical spider

Imaginary case of Mammalian mimicry

" Toby " as a caterpillar {Pseudaeraea eurytus)

" Toby " as a chrysalis {Pseudaeraea eurytus)

View to the West from Bugalla camp

House on Bugalla, 1912-13

Interior of Bugalla house .

Back of house on Kome Isle, 1914

Front of Kome house

Fly-boys' camp on Kome Isle, 1914

Servants' huts on Kome Isle, 1914

Cook's hut and kitchen on Bugalla

Personal boys and their wives .

FACING PAGE

192
196
198
198
206
228
228
262
262
278
282
282
292
292
304
304
314
314

CHARTS

I. Comparison between climates of Entebbe and Bugalla Isle . 40
II. Influence of climate on flies at Jinja ..... 42

III. Influence of climate on the fly . . . . . .44

IV. Influence of climate on rate of reproduction of the fly . 46

V. Nimaber of female flies on different islands
VI. Number of male flies on difierent islands .
VII. Niimber of crocodiles and flies on Tavu island

At end behind map

TABLES

PAGE

Parallelism of Nagana and Sleeping Sickness .... 25
Prolonged catching of flies increases proportion of females caught 31
Proof that the fly roams along the coast . . . . .33

Sovirces of non-mammalian blood foimd in wild flies ... 37
Flies watched biting captive crocodile ..... 38

Number of butterflies on various islands ..... 125

Food of Bee-eaters . . 172

Numbers of eggs in nests of crocodiles . . . . .182

Degrees of relationship between model and mimic . . . 233

xxiv IjIST of illustrations

. PAGE

Some forms of Pseudacraea eurytus, their models, and distribution 254

Pseudacraea eurytus on Damba Island ..... 257

Forms of Ps. eurytus obtained by breeding frora known parent . 264

Models, mimics, and intermediates in varying proportions . . 265

The same shown graphically as a chart ..... 267

Models and mimics in different localities ..... 274

MAP

The Sesse Archipelago ....... At the end

A NATURALIST ON
LAKE VICTORIA

CHAPTER I

SLEEPING SICKNESS

In this chapter a general account will be given of the
history and symptoms of Sleeping Sickness as it was
known during the epidemic in Uganda. It may be said
at once that the disease is entirely confined to tropical
Africa, nor does there seem any reason, as will be
explained later, to fear its spread beyond Africa.

In a few words it may be said that the cause of the
disease is a minute unicellular creature, called a Trypano-
some, belonging to the lowest order of animal life, which
is as it were inoculated by the bite of a blood-sucking
fly, the " Tse-tse," or Glossina.

A common history given by patients who suffer from
Sleeping Sickness is that they have been in a country
where they were much bitten by Tse-tse flies, and that
after a few days a painful swelling has appeared on the
neck, accompanied by high fever. The swelling may
appear to be on the point of becoming an abscess, but does
not do so, and gradually subsides. It is probable that
this represents the site where the fly which was the
cause of the infection actually bit. The fever may
subside in a few days, and recurs at irregular intervals
lasting weeks or months : it often reaches a high point
attended with delirium.

2 1

2 SLEEPING SICKNESS

There is a good deal of headache, debility and languor,
and vague pains in legs. An interesting feature is the ap-
pearance of an erythematous rash, mainly on the chest
and back. There is great wasting and enlargement
of the glands of the neck. Another interesting feature is
puffy swelling of parts of the face and body : this oede-
matous swelling is a particular feature of the diseases
of animals, which are, as will be seen, so very closely
connected with Sleeping Sickness.

This condition may go on for years, and has been
known to disappear altogether with an apparent cure ;
it is known as " Trypanosomiasis." Next comes the
stage to which the term *' Sleeping Sickness " more
properly applies. The drowsiness becomes accentuated,
so that the subject takes no interest in his surroundings
and does not trouble to eat, though he will eat food
if it is brought him, and he is fed. The fever continues
irregularly, the eyes become more puffy, the lips and
tongue tremulous, the wasting more and more pro-
nounced, until death finally supervenes with the patient
in a state of coma. At the last there may be mania
and convulsions.

It has been noticed that natives suffering from Sleeping
Sickness appear to feel the cold very acutely, and will
often sleep so near to a fire that they inflict severe burns
on themselves. This also shows how the senses are dulled,
so that one can conclude that they cannot suffer
much.

Death may occur in a few months after the initial
fever, but more usually after one to three years.

The earliest account of Sleeping Sickness which is
known in print dates from 1742, and is of considerable
interest because it reflects the current medical opinion
of the day.

Dr. E. D. Whittle, in the Malay Medical Journal
for April 1911, drew attention to a book by a naval

EARLIEST ACCOUNT OF SLEEPING SICKNESS 3

surgeon named John Atkins, called Physical Observa-
tions on the Coast of Guiney,

The following extract is given in The Sleeping Sickness
Bulletin :

** The Sleepy Distemper (common among negroes)
gives no other previous Notice, than a Want of Appetite
two or three Days before : Their Sleeps are sound, and
Sense of Feeling very little ; for pulling, drubbing, or
whipping, will scarce stir up Sense and Power enough
to move ; and the moment you cease beating, the Smart
is forgot, and down they fall again into a state of Insen-
sibility, driveling constantly from the Mouth, as if in a
deep Salivation ; breath slowly, but not unequally, nor
snort.

"Young People are more subject to it than the Old ;
and the Judgement generally pronounced is Death,
the prognostick seldom failing. If now and then one
of them recovers, he certainly loses the little Reason he
had and turns Ideot.

4< « 9!: He <:

" The immediate Cause of this deadly Sleepiness in
the Slaves, is evidently a Super-abundance of Phlegm
or Serum extravated in the Brain, which obstructs the
Irradiation of the Nerves ; but what the procatartick
Causes are, that exert to this Production, eclipsing the
Light of the Senses, is not so easily assigned.

" We find sometimes in Europe that Enormities in
the Non-Naturals, Surfeiting and Drunkenness do gradu-
ally, as Age and Custom advance, weaken the Tone of
the Brain, to the Admission of serous and extrementitioua
Humours, including Sleepiness, etc. But here the Case
is different, they being young People that are generally
afflicted, and who have been destitute of the Means of
Surfeiting.

" I shall ascribe the Cause to catching Cold, and their
Immaturity ; to Diet and Way of Living ; and to the natural

4 SLEEPING SICKNESS

Weakness of the Brain ; some or all of these Causes co-
operating to it,

" First. In Immaturity, or Childhood, it is a common
and true Observation, that more of Phlegm and recre-
mentitious Humour is bred, than at Manhood ; because
the Fibres, and consequently the Faculties resulting
from, their Constitution, have not attained their due
Spring and Perfection ; and it is only supposing the
Africans continue longer Children than the Europeans.

" Secondly. Promoted here by their Diet and Way of
Living. At Home it is mostly on Roots, Fruits, and
Herbage, greedily devouring such as are wild and uncul-
tured ; which, together with the intolerable Heats of the
Sun, weakening the concoctive Faculty, together with
their Inactivity, render a very recrementitious Nutriment :
Their Indolence is such (when shipped on Board for Slaves)
as to be entirely dispassionate at parting with Wives,
Children, Friends, and Country, and are scarcely touched
with any other Sense or Appetite, than that of Hunger ;
and even in this, for want of Custom or Instinct, they
cannot distinguish proper Food, nor know when to leave
off, voraciously eating though Victuals be never so dirtily
cook'd ; and whether the Flesh be raw or dressed, whether
of the Guts or a Sirloin ; a Practice also that may sometime,
by over-stretching the Fibres of the Stomach, occasion
Crudity and Indigestion. By their Sloth and Idleness
the Blood becomes more depauperated ; and those recre-
mentitious Humours bred from it, that Exercise would
throw off through the proper secretory Organs, are here
disposed towards the weakest Part, which in the Gener-
ality of Negroe Slaves I take to be the Brain.

" Thirdly, The natural Weakness of the Brain, I am apt
to think the principal Cause of the Distemper. Doubt-
less that Part gains Strength by Exercise, i.e. by the
Employment of our rational Faculties, as well as the
Muscles and external Fibres of the Body by Labour;

EARLIEST ACCOUNT OF SLEEPING SICKNESS 5

and since the Africans are hereditarily ignorant, destitute
of all Art and Science, or any mechanical Knowledge to
exercise the Brain, it consequently grows weaker in its
inward Structure and Recesses ; and falls together with
the Judgement and Passions.

" The Cure is attempted by whatever rouzes the Spirits ;
bleeding in the Jugular, quick Purges, Sternutories,
Vesicatories, Acu-Puncture, Seton, Fontanels, and sudden
Plunges into the Sea, the latter is most effectual when
the Distemper is new, and the Patient not yet driveling
at Mouth and nose." ^

Noteworthy in this account are the number of causes
assigned to this mysterious disease, and the naive way
in which the author assigns the conditions to " some
or all of these causes " !

The essential characteristics of the disease in an advanced
stage are all noted ; the " Indolence," " State of Insensi-
bility," "Hunger," "Sense of Feeling very little,"
but the preliminary stages were not noted, for Atkins
says it gives no other notice than a want of appetite two
or three days before !

With the aid of a little imagination one can see a
terrible picture of these wretched, somnolent natives
being subjected to " pulling, drubbing or whipping " in
the endeavour to give them sufficient " Exercise " to
" throw o£E the recrementitious Humours " and afterwards
" voraciously eating though Victuals be never so dirtily
cook'd " !

The " Sudden Plunges into the Sea " are not used
for curing Sleeping Sickness at the present day !

^ The Navy Surgeon : or Practical System of Surgery with a Disserta-
tion on Cold and Hot Mineral Springs and Physical Observations on . the
Coast of Guiney, by John Atkins, Surgeon. London : Printed for J.
Hodges, at the Looking Glass, over against St. Magnus Church, London
Bridge. MDCCXLII. pp. 364-7.

6 SLEEPING SICKNESS

It was, I believe, well known to the African slave
traders that natives with enlarged glands were of no use
to them, as they always died and could not be made to
work. So they were careful not to take any in this
condition, but nevertheless took, without knowing it,
many cases who had the earliest stages of the disease,
and the American slave-owners used to find their slaves
dying of this peculiar drowsiness.

Yet it was not infectious, that is, it was not communi-
cated from one to another : the reason for this will be
entered into more fully further on.

In 1890, a French doctor discovered in the blood of
a patient from Africa who was suffering from a peculiar
fever, a microscropic organism which he recognized as
a Trypanosome, but he did not estabhsh any relation
between this and the fever, and the credit of first dis-
covering this belongs to an Englishman. In 1901 Dr.
Forde, in the Gambia colony on the West Coast, had
an English patient with a peculiar fever of a chronic
and irregular type. He found in the blood a curious
organism whose nature he could not recognize, and
called in the late Dr. Dutton, who at once recognized
the creature as a Trypanosome ; ^ the first one to be
discovered in man, though one had been described as
the cause of a disease in horses in India some ten years
before.

At first, and for some time. Sleeping Sickness was only
known to occur in West Africa, but when equatorial
Africa was gradually opened up the disease found its
way into Uganda with disastrous results.

This is believed to be due to Stanley's expedition for
the relief of Emin Pasha, which in 1888 travelled from the
Congo to the Lake Albert Nyanza. In 1891 the Sudanese
soldiers of Emin's force were brought into South Toro

^ British Medical Journal, 1902, January 4th, p. 42, and November 29th,
p. 1741.

THE UGANDA EPIDEMIC 7

with their followers, and eventually were brought into
Uganda itself to be under control.

In July 1901, Dr. Albert Cook of the Church Missionary
Society at Mengo (Kampala) noted eight cases of a mys-
terious disease, and six months later reported that on
Buvuma Island over two hundred natives had died and
thousands appeared to be infected. The mortality
became appalling, and the Government were at their
wits' end, for it seemed as if the whole population was
doomed.

In July 1 902, the first Royal Society Commission arrived
in Uganda, composed of Drs. Low, Christie, and Castellani,
and Colonel Sir David Bruce arrived in February 1903.
On April 28th, it was announced that the disease was
caused by a Trypanosome and conveyed by a Tse-tse fly,
Olossina palpaUs.^

It was suggested at once that as the haunts of this
fly were strictly limited, it would be easy to check the
disease by removing the population ; ^ but the natives,
with their characteristic fatalism, refused to leave their
villages along the shores of the lake. In the mean-
time the disease raged unchecked, and by the end of 1903
the number of deaths had reached over 90,000 ; whole
villages were being depopulated, and great tracts of
highly cultivated country relapsed into scrub and forest.

In March 1905 Lieutenant Tulloch, R.A.M.C, who
had been sent out by the Royal Society to help in the
investigations, became infected with the disease in its
virulent form, and died a few months later.

By November 1904 the epidemic had appeared on
the shores of Lake Albert in North-West Uganda, and a
survey of Uganda by six specially appointed medical
officers in 1905 showed that the banks of the lakes and

^ Reports of Sleeping Sickness Commission of the Royal Society,
1903, Nos. I, IV.
* See Bulletin of Sleeping Sickness Bureau, vol. 4, pp. 241-2.

8 SLEEPING SICKNESS

watercourses throughout Uganda were infested with
the Tse-tse fly.

Statistics furnished in a dispatch by the Governor
showed that " during the last five years the total mor-
tality from this scourge in this Protectorate has consider-
ably exceeded 200,000."i

Sir Hesketh Bell also reported " the natives have been
almost completely wiped out everywhere along the lake
shore, and in the islands the mortality has been even more
appalling. Buvuma, for instance, which a few years ago
was one of the most thickly populated and prosperous of all
the islands, counted over 30,000 inhabitants. There are
now barely 14,000. Some of the Sesse group have lost
every soul ; while in others a few moribund natives,
crawling about in the last stages of the disease, are all
that are left to represent a once teeming population."

In November 1906, it was again suggested that the
only way to save the people was to remove them into
fly-free areas, and segregate the infected natives into
camps. The aid of the chiefs was sought and the matter
fully explained to them, compensation was made to the
heads of evicted families, they were given land away
from the infected areas, and by degrees not only the
mainland shores of the lake, but the islands also, were
cleared of their population, so that by 1909 all these were
deserted and going back to the wild state. ^

Great difficulty was experienced in preventing the
natives from returning to their homes, and some managed
to obtain canoes and cross back to Buvuma and Damba
Islands, but at length the evacuation was finally completed,
and at the present day the whole of the fertile and valuable
island territory is abandoned to the Tse-tse fly.

But the lake shore can only be kept in this condition
by stringent regulations and penalties, and a few natives

1 Dispatch No. 218 from Sir Hesketh Bell, November 1906.

" See Bulletin of the Sleeping Sickness Bureau, vol. 4, pp. 241-2.

THE RHODESIAN DISEASE 9

are frequently discovered in the forbidden areas by the
patrolling canoes.

By these means the inhabitants of Uganda were saved,
and at the present day there are very few deaths a year ;
from 1905-1917 there were just over 30,000 deaths for
the whole of the Uganda Protectorate.

At the end of 1909 considerable alarm was caused
by the discovery in Nyassaland of Sleeping Sickness
(or rather, Acute Trypanosomiasis) in a native there,
and since then a number of cases have been found, some
of them Europeans, in Nyassaland, North-East Rhodesia,
and Portuguese East Africa : in the case of some natives,
they had certainly never left their homes. This was
very interesting from a scientific point of view, because
Glossina palpalis, the species of Tse-tse which carries
Sleeping Sickness in Uganda and the West Coast, does
not exist in those countries.

It was soon found that the carrier was another species,
namely, the very one which has been so long known to
travellers in Africa as the cause of Tse-tse fly disease or
" Nagana " of horses, cattle and dogs.

This species is known as Glossina morsitans^

As this new form of human Trypanosomiasis appeared
to be very much more acute than the form known as
Sleeping Sickness, the discovery was disconcerting. It
is, however, possible that our ideas of the severity of
this form of Trypanosomiasis will need to be modified
in the light of further knowledge, for during the campaign
in East Africa a number of natives were found to have
Trypanosomes in their blood while under treatment
for other complaints, and appeared to be little the worse
for their presence ; these natives had not been in the area
of Glossina palpalis, so that either the Trypanosome was
gambiense carried by morsitans, or else it was rhodesiense

^ Kinghorn.and Yorke, Annals oj Tropical Medicine and Parasitology,
1912, vol. 6, pp. 1-23.

10 SLEEPING SICKNESS

in a non-lethal form. It is of interest to note here that
a case of true Rhodesian Trypanosomiasis has recently
been recorded as the first to be cured by treatment ;
hitherto this form had always been regarded as fatal. ^

^ A case of T. rhodesiense infection which recovered. Daniels and
Newham, British Medical Journal, 1919, November 8th, p, 829.

CHAPTER II

NATURAL HISTORY OF SLEEPING SICKNESS

I PROPOSE now to deal at greater length first with the
Trypanosomes and next with the Tse-tse flies, then the
inter-relation of the two will be considered and the relation
of both to the wild animals of the countries in which
they occur.

Trypanosomes are not ** microbes " in the accepted
sense of the word as commonly used ; that is, they are not
micro-fungi, but belong to the Protozoa, the lowest members
of the animal kingdom. The Protozoa can be divided
into four main groups, in each of which are found species
causing disease in man and other vertebrates. The
Sarcodina is exemplified by the well-known Amoeba and
the species known as Entomoeba, which causes dysentery.

The next group, Mastigophora, is the one with which
we are most concerned at the moment, since to it belong
the Trypanosomes and their allies. The name, which
means " Whip Bearers," refers to the possession of one
or more lashes or Flagella, which by their rapid move-
ments draw or push the animal through the fluid in
which it lives.

The third great group of Protozoa is the Sporozoa, which

is only too familiar through the various species that

cause malaria. To the last group Infusoria belong the

myriad forms which are the delight of the amateur

microscopist and may be seen so easily in stagnant water.

One species, Balantidium, is a cause of dysentery.

11

12 NATURAL HISTORY OF SLEEPING SICKNESS

Let us now look a little more closely at the Mastigophora.
The group is divisible into six sub-groups,^ but only that
which includes the Trypanosomes need be mentioned
here. This is named Protomonadina, and includes the
following genera :

1. Trypanosoma. — There is a single flagellum arising

near the posterior nucleus and continued forwards
as a marginal flagellum of an undulating mem-
brane ; usually continued into a fi:ee flagellum.
The species are found mostly in blood and in the
digestive tracts of invertebrates, but are also forms
in the life cycle of species wholly parasitic in
insects.

2. Trypanoplasma. — There are two flagella, the posterior

one united to the body by an undulating membrane
along most of its length. According to their
mode of life they fall into three groups.

1. Parasites in the blood of fresh- water fish
and in the digestive tract of leeches.

2. Parasitic in the digestive tract of marine
fishes.

3. Parasitic in invertebrates.

3. CritTiidia. — A single flagellum arises about the middle

of the body and runs forwards to form the marginal
flagellum of a short or rudimentary undulating
membrane, and is continued beyond as a free
flagellum.

Crithidia occur as parasites in the gut of insects
or as a stage in the life cycle of a Try-
panosome.

4. Leptomonas or Herpetomonas. — There is a single

anterior flagellum but no undulating membrane.
These are parasitic on invertebrates, chiefly
insects, but also occur as a form of the next genus

^ This classification is taken from Minchin.

TRYPANOSOMES 18

in invertebrate hosts or in artificial cultures. It
is also of great interest that they are found in the
milky juice of certain plants {Eujphorbiaceae).
5. Leishmania. — There is no flagellum, but there are
two nuclei as in the other genera. These animals
have become specialized to live in the tissue cells
of vertebrates and so have no need of organs of
locomotion. It is of much interest that in arti-
ficial cultures Leishmania develops into forms like
the two preceding genera or like Trypanosomes.
The type species causes a fatal disease in Asia
known as " Kala-azar."

Now let us consider a little more fuUy the Trypanosomes
themselves. They are elongated bodies with a pointed
posterior extremity where is a small nucleus, and at the
anterior extremity the whip -like flagellum which by vigor-
ous movements drags the Trypanosome along. The
undulating membrane pursues a wavy course along the
body and is responsible for the name, which is derived
from rpviravov, meaning a carpenter's tool : the allusion
is probably to the spiral thread on an auger. In the
middle of the body is a large nucleus. T. gambiense
in the blood is several times as long as the diameter of a
red corpuscle, but much narrower. The different species
that may be found in the blood vary very greatly in size
and activity ; gambiense is not very active and merely
wriggles, but a species such as vivax in goats can hardly
be kept in the field of a microscope. The first one to
be seen was found in the blood of a frog, but the
first reliable description was not given until 1841, when
a specimen was described from the blood of a trout. ^
Not until 1879 was one found in a mammal, and this was
Trypanosoma lewisi, seen in the blood of the rat. It is

1 By Valentin. Recent work by IVIlle. M. Gauthier, however, places
this species in the genus Trypanoplasma.

14 NATURAL HISTORY OF SLEEPING SICKNESS

now known that in nearly every place rats are infected
with this species, often to the extent of 50 per cent., but
it is not harmful to them unless present in great numbers.
The first disease-producing Trypanosome described was
T. evansi, which kills numbers of horses, camels, elephants
and dogs in Asia and North Africa by causing a disease
known as " Surra " ; it was discovered in 1880. Not
until 1895 was the next pathogenic species discovered,
and this was an extremely important one, the know-
ledge of which was of the utmost help to the subsequent
investigations into Sleeping Sickness.

This species was named T. brucei, after Sir David
Bruce, who proved it to be the cause of " Tse-tse fly
disease " or " Nagana " in South Africa. ^

In 1901 another species was described as T. equiperdum,
the cause of " Dourine " in stallions, brood mares, and
donkeys in America and North Africa. Also in 1901
was discovered T. garnbiense, the first Trypanosome to
be known for a cause of disease in man.^

In 1902 T. equinum was shown to be the cause of " Mai
de Caderas," a fatal disease of horses in Brazil and the
Argentine Republic. Another human Trypanosome was
discovered in 1909,^ by Chagas ; it causes in children,
and adults who have not become immune in childhood,
chronic fever, enlargement of the thyroid gland, and
puffiness of the face, and may rarely result in death.
This species was found to differ sufficiently in its life-
history to be put in a new genus, and has been named
Schizotrypanum cruzi. In 1910 it was shown that the
cause of the acute Trypanosomiasis of Rhodesia was
differentiated, not by its morphology, but by its effects

* Preliminary Report on the Tse-tse fly disease or Nagana in Zulu-
land, 1895.

2 British Medical Journal, 1902, January 4th, p. 42.

' Chagas, Brazil Medico, 1909, April 22nd. See Bull. Inst. Pasteur,
1909, May 30th, pp. 453-4. Also Bulletin of the Sleeping Sickness
Bureau, vol. 1, p. 252; vol. 2, p. 117.

TRYPANOSOMES 15

upon animals, from its close allies T. gambiense and T,
brucei, and it is now known as T. rhodesiense.^

In 1913 another species was found in man, causing
in natives of Nigeria a chronic swelling of the glands in
the neck very like the earliest stages of Sleeping Sickness,
but very rarely going beyond that ; it has been named
T. nigeriense.^

It must not be thought that every Trypanosoma is a
cause of disease. This is very far from being the case ;
the production of disease is a mere accident from the
Trypanosome's point of view, an occurrence of fatal
import for itself as well as the host should the latter
die!

Although only a few of the pathogenic species have
been mentioned, all of them together are a very small
fraction of the total number of Trypanosomes and their
allies that are known. They occur in the blood of fishes,
reptiles, amphibia, birds and mammals ; in the vital fluids
of molluscs, in the alimentary canal of insects and other
invertebrates.

By cultural methods it has been found that normal
English cattle have Trypanosomes in their blood, and
Bruce first found that even the species that cause disease
in domestic animals lives harmlessly in the indigenous
wild animals. Dr. Duke has shown that the " Situtunga "
antelope {Tragelaphus speJcei) carries in its blood the
deadly Trypanosome gambiense without any harm to itself.
It must be remembered that Trypanosomes are not like
the bacteria such as that of anthrax, which form highly
resistent spores, and by the death of their host and its
disintegration are disseminated more widely in a condi-
tion in which life may be maintained indefinitely.

The position seems to be this. Many insects, for instance

^ Stephens and Fantham, Proc. Royal Soc, Series B, 1910, No. 561,
pp. 28-32.

2 J. W. Scott Macfie, Annals of Tropical Medicine and Parasitology,
1913, August, vol, 7, No, 3a, pp. 339-56.

16 NATURAL HISTORY OF SLEEPING SICKNESS

the house fly, contain in the alimentary canal Flagellates
of a type closely allied to Trypanosomes, which live a
natural parasitic life in the fly only.

When the ancestors of the present blood-suckers became
addicted to this habit, probably a development of the
habit of feeding on the fluid exuding from sores, it may
well have happened that some of these internal parasites
were inoculated into the blood of the animal. Those
that survived may well have caused disease as a result
of their vital activities ; if so, by the process of natural
selection, those which soonest became adapted to the
new conditions and no longer brought about a condition
unfavourable to the host, on whose life they would now
depend, would have most chance of surviving.

This theory is supported by some very interesting
work done by Fantham and Porter, who showed that
Herpetomonas jaculum, a normal inhabitant of the gut
of the " Water scorpion " {Nepa cinerea), when introduced
into the peritoneal cavity of mice or even taken by them
in food eventually finds its way into the blood and causes
symptoms like those of " Kala-azar " in man, from which
the mouse dies.^ But if the mouse had not died so
quickly there might have been developed such immunity
to the new substances circulating in its blood as the result
of the vital activities of the Herpetomonas that friendly
relations between the two would result and mammal
and parasite would become'''mutually immune to each
other, as appears to be the case with wild game and
T. brucei.

When man with his non-immune domestic animals
comes into relation with this equilibrised system the
equilibrium is disturbed ; the Trypanosome finds itself
introduced into the circulation of hosts to which it is
a novelty and " disease " results.

1 Fantham and Porter, Proc. Camb. Phil. Soc, 1915, vol. 18, pp. 39-50,
137-48.

TRYPANOSOMES AND THEIR HOSTS 17

But even man himself shows a gradual acquirement
of immunity. In West Africa Sleeping Sickness is not
generally so virulent as in the Uganda epidemic, and
the disease has been known in that area for a very much
longer time. The species of Trypanosome known as
T. nigeriense causes only a mild form of disease which
is very rarely fatal, so it would appear to be the species
which has been longest in contact with man. On the
other hand, T. rhodesiense is very much more virulent than
any other human Trypanosome known to us, and it
appears to have arisen as a sudden " mutation " from
the animal-inhabiting species T. brucei. It is only ten
years since this undesirable addition to the list of " human "
Trypanosomes made its presence known to us by causing
a disease previously unknown in that part, or at least
unrecognized as distinct from malaria.

Trypanosomes are transferred from one host to another
by the agency of various invertebrates, which are as
necessary for its existence as are the animals in whose
blood they also live.

In the blood they only multiply by fission, asexually
or vegetatively ; but in the invertebrate host they go
through a sexual form of reproduction. Development
generally commences in the alimentary canal, whence the
Trypanosomes find their way into the " saKvary glands,"
so that they are inoculated into the new vertebrate host
when the blood-sucker injects the irritating fluid secreted
by those glands in order to produce a free flow of blood.
The effect of this fluid is familiar to all who have been
" bitten " by mosquitoes. Such an alternation of sexual
with asexual methods of reproduction is called a " cycle " ;
the complicated life history of the malaria parasite is a
more familiar example.

I think the fact that the sexual process takes place
in the invertebrate host points to that being the original
host for the Trypanosomes ; the life in the blood is as it

3

18 NATURAL HISTORY OF SLEEPING SICKNESS

were an accident, and many close allies of Trypanosomes
can do perfectly well without it.

Trypanosomes of fishes are carried by leeches ^ ; of birds,
by mosquitoes ; those of mammals by various blood-
sucking insects. Thus, the rat Trypanosome is trans-
mitted by the rat flea, the species causing *' Surra "
in animals by the large flies often called " Cleggs "
{Tabanidce), and those causing " Nagana " in animals
and Sleeping Sickness in man are carried by Tse-tse
flies {Glossina).

The Trypanosome of " Dourine " is particularly interest-
ing, because it seems to have found that it can do without
an intermediate insect host, and is transmitted directly
from male to female animal, thus having severed all
relations with the ancestral home !

It may be pointed out that there is a close analogy
shown by Spirochsetal diseases of man, some of which
are carried by an intermediate host while others are not ;
the pathology of the two classes also shows much in
common, which is not surprising, considering how closely
allied are Trypanosomes and Spirochsetes.

Lastly, the very interesting species which causes in-
flammation of the thyroid gland with fever in Brazil, is
carried by an insect {Conorhinus) of the order Hemiptera
or " bugs," an order which numbers extremely few
blood-suckers among its ranks, although all are adapted
for obtaining food by suction. Darwin made special
mention of this bug in his journal, and remarks that
" one feast kept it fat during four whole months." ^

The Tse-tse flies will now be dealt with more fully,
but in this chapter I will only give a general account,
reserving a full account of the natural history of Glossina
palpalis for another chapter.

1 See Miss M. Robertson's paper in Phil. Trans. Royal So6., 1911,
Series B, vol. 202, pp. 29-50.

2 Voyage round the World, edition 1890, p. 316.

TSE-TSE FLIES 19

Some sixteen species exist, which are entilrey
Ethiopian in their distribution ; for though one species
{tachinoides) is found in the south-west corner of Arabia,
that area is faunistically part of the Ethiopian region.
Tse-tse do not occur in Africa north of the Sahara nor
south of St. Lucia Bay in Zululand. It is remarkably
interesting, therefore, that fossil flies have been found
in Colorado referable to this group^^ and it has been
suggested by Osborn that they were responsible at least
in part for the extinction of many of the large mammals
which abounded in Cainozoic times and by their
migrations came into contact with blood-sucking flies to
which, and their associated flagellates, they had not yet
become habituated. ^

The first Tse-tse which I wish to mention particularly
is Glossina morsitans, known to travellers in Africa as
"The Fly," or, collectively, as "Fly." According to
Austen ^ the exact origin of the name is uncertain ; it
is believed to be a corruption of " Nsi-Nsi," said to be
the name given to blood-sucking flies by natives of some
parts of Africa. A passage in the Old Testament * possibly
refers to Glossina, although it may also apply to other
flies with the same habits : " And it shall come to pass
in that day, that the Lord shall hiss for the fly that is in
the uttermost part of the rivers of Egypt, and they shall
come, and shall rest all of them in the desolate valleys,
and in the holes of the rocks, and upon all thorns, and
upon all bushes."

The early African travellers of the middle of the nine-
teenth century were, naturally, much impressed by the
fly which was so well known to the natives for inflicting
a " poisonous bite " upon their cattle.

* Cockerell, Proc. of the United States National Museum, vol. liv. 1918,
p. 308.

2 Letter from Professor Cockerell in Nature, 1919, June 5, vol. ciii.
p. 265. 3 Monograph on the Tse-tse Flies, 1908, p. 32.

• Isaiah, chapter vii. verses 18, 19,

20 NATURAL HISTORY OF SLEEPING SICKNESS

Thus Gordon Gumming in 1850 wrote as follows :^
" When under the mountains on the south bank of the
Limpopo river I met with this famous fly whose bite is
certain death to oxen and horses. This fly is similar to
the fly in Scotland called ' Kleg,' but a little smaller.
They are very quick and active, and storm a horse like
a swarm of bees, alighting on him in hundreds and
drinking his blood. The animal thus bitten pines away
and dies at periods varying from a week to three months."

And again : " The next day one of my steeds died of
Tse-tse. He had been bitten under the mountain range
lying to the south of this fountain. The head and body
of the poor animal swelled up in a most distressing
manner before he died. His eyes were so swollen that
he could not see, and in darkness he neighed for his com-
rades who stood feeding beside him."

Again, Livingstone in 1857 wrote as follows : ^ "The
peculiar buzz when once heard can never be forgotten
by the traveller, for it is well known that the bite of
this poisonous insect is certain death to ox, horse or dog.
In this journey we lost forty-three oxen by its bite.
We watched the animals carefully, and believe that not
a score of flies were ever upon them.

" A most remarkable feature is the perfect harmlessness
of the bite to man and wild animals." The italics are
mine, to emphasize that the scientific mind of the great
explorer had noticed this pregnant fact, although the full
significance was not made plain until Bruce's work was
published. " We never experienced the slightest injury
from them ourselves, although we lived two months in
their habitat, which was in this case as sharply defined
as in many others ; for the south bank of the river was
infested by them, and the north bank, where our cattle

1 Five Years of a Hunter's Life in the Far Interior of South Africa,
vol. ii. p. 210, etc.

^ Missionary Travels and Researches in South Africa, p. 79, etc.

LIVINGSTONE'S ACCOUNT OF THE TSE-TSE 21

were placed, only fifty yards distant, contained not a
single specimen. This was the more remarkable as we
often saw natives carrying over raw meat to the opposite
bank, with many Tse-tses settled upon it."

This account makes clear one of the most ..remarkable
points in the natural history of this species of Olossina,
namely, the very sharply marked areas which it inhabits ;
this must depend upon the presence or absence of certain
factors in its environment, but has not yet been thoroughly
explained.

After noting the symptoms of "fly disease " Livingstone
continues : " These symptoms seem to indicate a poison
in the blood, the germ of which enters when the proboscis
is inserted to draw blood. The poison germ, contained
in a bulb at the root of the proboscis, seems capable, although
very minute in quantity, of reproducing itself." The words
which I have put in italics were written by Livingstone.
about fifty years before the discovery of the Trypanosome
by Bruce, and some years before the first discovery of
micro-organisms of disease by Louis Pasteur ! ^

The genus Glossina was founded in 1830, when Wiede-
mann described a new species of fly from Sierra Leone,
and in the same year another species was described from
the Congo by Robineau-Desvoidy : this is the one which
is now known as Glossina palpalis. It is interesting
that the first scientific description of a Tse-tse was not
that of the one so long known to natives in South
Africa ; this was not named until 1850, when Gordon
Cumming's travels made the fly well known in England.

Glossina palpalis frequents forested and humid country,
and is not found so far south as its more widely distri-
buted relative morsitans. The great river courses and
lake shores in the tropics furnish the shade and humidity

^ In 1857 Pasteur first showed that fermentation was due to micro-
organisms ; in 1865 he showed that silkworm disease was due to micro-
organisms. In 1876 the first " bacterium " (of anthrax) was isolated
by Koch.

22 NATURAL HISTORY OF SLEEPING SICKNESS

that it requires, hence the Congo basin, the upper Nile,
and the shores of Lake Victoria abound with it.

Tse-tse flies are on the whole diurnal, though one or two
species are known to bite at night ; but the traveller is
safe from both morsitans and palpalis after dark.

The life history of the Tse-tse flies is very remarkable,
and almost unique among insects. No eggs are laid, but
a single egg is hatched within the abdomen of the mother
fly and the larva, a white grub, is fed by the secretion of
. special glands. When the larva is full grown the mother
fly seeks a suitable spot and the larva is extruded. It is
a very active little creature, and crawls about seeking
for a spot where it can burrow into the soil. It is helped
to do this by two curious bosses at the posterior end of
the body which are of hard chitin and give the larva a
firm purchase when it begins to burrow. As soon as
it has buried itself its skin hardens and it becomes a dark
brown oval pupa with the two bosses at the posterior
end which were seen in the larva, and differ in shape
according to the species of fly.

After a period of very varied duration the perfect fly
emeiges by cracking the pupa-case or *' puparium "
at one end and by means of an extrusible bladder on
the front of its head works its way up through the soil
covering it and creeps up a stem in a great hurry to let
the wings hang down and expand before they have hard-
ened. I have seen many palpalis thus emerge from the
ground, and they always give the impression that it is
a matter of the utmost importance that they should lose
no time !

Let us now glance at the development of our knowledge of
Tse-tse flies and disease. It was quite clear to the earliest
travellers in South Africa that " The Fly," as Glossina
morsitans was called, was a cause of the fatal disease of
cattle and horses to which allusions have been made.

Sir David Bruce in 1894 was able to report that the

NAGANA AND SLEEPING SICKNESS 23

actual cause of the disease " Nagana " was a Trypanosome
which was named brucei after its discoverer, and was taken
up by a Tse-tse when it fed on the blood of an animal
suffering from nagana. It was believed that the fly
inoculated the Trypanosome into a fresh animal when it
fed again in the same manner that a vaccinator introduces
the lymph on the point of his lancet, that is to say, by
direct inoculation. Thus in the case of Nagana we
have started with the carrier of the disease and found
the germ which it carried.

When the problem of Sleeping Sickness became so
acute in Uganda, the Royal Society sent out a Commission
in investigate the disease, and the members arrived in
Uganda in July 1902. But the first step towards the
elucidation had already been made in West Africa. In
1901 an Englishman in charge of a steamer on the Gambia
river was admitted to hospital at Bathurst for " fever,"
and Dr. Forde found in his blood peculiar organisms
whose nature was unknown to him. The patient was
sent to Liverpool, and Dr. Dutton recognized the new
organism to be a Trypanosome : the patient eventually
died at the commencement of 1903.

In 1902 Drs. Dutton and Todd found Trypanosomes
in the blood of several West African negroes suffering
from the early stages of what we now know to be Sleeping
Sickness, but in those days the condition had not been
recognized as connected with that well known disease.
This early stage of fever now came to be known as
" Trypanosome fever," or " Trypanosomiasis," and in
Marph 1903 Dr. Baker found the Trypanosome in a
case in Uganda, ^ though he did not recognize the full
importance of this fact. The next development was the
finding by Dr. Castellani of Trypanosomes in the cerebro-
spinal fluid of a case of Sleeping Sickness in April 1903,*

» British Medical Journal, 1903, May 30th, p. 1254.
2 Proc. Boy. Soc, 1903, vol. Ixxi. pp. 501-8.

24 NATURAL HISTORY OF SLEEPING SICKNESS

and afterwards by Bruce and Nabarro in every case of
Sleeping Sickness examined. Thus was made clear the
fact that *' Trypanosome fever " is the early stage of
the fatal disease, and the next step was to find out how
the disease was transmitted.

At this step a comparison of what was known about
Nagana with what has been found out about Trypano-
somiasis aids in the understanding of further developments.
In the former case investigation commenced with strong
presumptive evidence that a Tse-tse fly was the agent
through which the disease was acquired ; in the latter
case the germ was discovered and it became necessary
to ascertain how it was transmitted. Since Nagana had
been proved to be due to a Trypanosome carried by the
blood-sucking fly Olossina morsitans, and Sleeping Sickness
had now been shown to be due to another species of
Trypanosome, evidence pointed to a blood-sucking fly
as the carrier of this new species Trypanosoma gambiense,
and requests were made to officials in the Sleeping Sick-
ness areas to send specimens of all biting flies in the
neighbourhood to the laboratory in Entebbe. It very
soon became clear that one fly was found throughout
the areas being ravaged by the disease, that is the shaded
margins of the great lakes and rivers ; this fly was
Olossina palpalis. Accordingly experiments were made
to test whether the fly can be the carrier, and specimens
captured on the shores of the lake were fed upon monkeys
whose blood was examined daily, the procedure being
the same as in Bruce's classical work on Nagana. Con-
clusive proof was obtained when the Trypanosome was
found in the blood of the monkeys, and the discovery
was announced by the Commission in 1903.^ Further
proof was obtained by feeding bred flies on monkeys
which had been already infected by wild flies, and then
making them feed upon another monkey ; the second

^ Reports of the Sleeping Sickness Commission, No. IV, pp. 66-66.

FLY BEACH ON DAMBA ISLE.

A favourite breeding ground is under the bushes at the gap on the right.

By permission of the Royal Society and the Controller, H.M. Stationery Office.

FLY BEACH ON BUGALLA ISLE.
By permission of the Royal Society and the Controller, H.M. Stationery Office.

LIFE CYCLE OF TRYPANOSOMES

25

monkey in a proportion of cases also showed the Trypano-
some in its blood. Control experiments showed that a
freshly hatched fly contains no Trypanosomes, so that
those in the above experiments must have acquired them
from the first monkeys fed upon.

The two diseases Nagana and Sleeping Sickness are
thus entirely parallel, as is shown by the following table :

Disease.

Subject.

Cause.

Natural Host.

Carrier.

Nagana

Sleeping
Sickness

Domestic
animals

Man

Trypanosoma
brucei

Trypanosoma
gambiense

Big Game

The antelope

Tragelaphus

spekei

Olossina
morsitans

Qlossina
palpalis

It has been said that the transmission of the Trypano-
some of Nagana was believed to be entirely mechanical,
and for some years this was also thought to be the case
with T. gambiense. With further knowledge, however,
it became clear that there was a period after the fly
had fed during which the Trypanosome could not be
transmitted to a fresh animal. It had previously been
supposed that this was because the Trypanosome was no
longer alive in the fly, but Kleine, working in German
East Africa, showed in 1908 that the non-infectivity of
the fly after a few days did not mean the death of the
Trypanosome, but that it was going through a cycle of
development in the alimentary canal of the fly, and was
not in an infective condition. For when the develop-
ment was complete Kleine found that the fly could
convey the disease fifty days after it had acquired the
Trypanosome. These most important results were fully
confirmed in Uganda in 1909 ^ and it was found that the
time required for the cycle of development in the fly

^ Reports of the Sleeping Sickness Commission, vol. x."p. 46, etc. ;
vol. xi. p. 12, etc.

26 NATURAL HISTORY OF SLEEPING SICKNESS

varied from eighteen to forty-five days, after which
time a fly will remain infective and able to introduce
the Trypanosome into every animal it bites for as long
as seventy-five days !

The complete life cycle of the Trypanosome in the fly
was worked out fully in Uganda by Miss Robertson
in 1913.^ The Trypanosome multiplies in the gut of the
fly, but the forms of multiplication are not those which
will live in the blood stream of the vertebrate host, and
a somewhat different form is developed which finds its
way into the " salivary " glands of the fly and is
injected with the irritant secretion of the gland which
presumably is intended to produce a free flow of blood
in the site of puncture. We have now arrived at the
most important fact that when once it has acquired
the Trypanosome the Tse-tse fly can infect for the rest
of its life.

We must now consider the relations between Tse-tse,
Trypanosome, and the " alternative hosts " of the latter
from which the fly acquires it.

It was known from Bruce's researches on Nagana ^
that Trypanosoma hrucei is a natural and harmless inhabi-
tant of the blood of various species of big game in the
** fly areas," and as soon as Sleeping Sickness was
proved to be due to another Trypanosome, efforts were
made to discover its natural host or reservoir. The
Commission in Uganda made a series of experiments
with the blood of such animals and birds that, inhabiting
the lake shore, might be justly suspected of harbouring
the trypanosome, but with negative results. One impor-
tant species of antelope, however, was not at this time
(1908-10) examined. The next step was to infect captive
antelope and native cattle by feeding upon them flies

^ Reports of the Sleeping Sickness Commission, No. XIII, p. 119, etc.
^ Appendix to Further Report on Tse-tse Fly Disease or Nagana in
Zululand, 1903, p. ,8. London, Harrison & Sons.

RESERVOIRS OF TRYPANOSOMES 27

caught in the Sleeping Sickness area, and it was found
that this could be readily done ; waterbuck, bushbuck
and reedbuck in captivity could all be made reservoirs
of the Trypanosome without any harm to themselves,
and could continue to infect bred flies fed upon them for
more than twelve months after they had been artificially
infected.! Finally, in March 1912, Dr. Duke armounced
that the antelope known to the natives as " Enjobe "
was a natural reservoir of Trypanosoma gambiense, and
the chain was thus completed. ^

It had seemed for some time probable that there must
be a natural host, although it had not been found. Four
years and a half after the natives had been removed from
the islands the fly there was still infective, and it was
impossible to suppose that the same flies were still alive
that had been the cause of the epidemic. With one
exception the animals and birds and reptiles within
reach of the fly had been sufficiently examined to make
it almost certain that they were not incriminated, and
the one large antelope living on the islands was yet to be
excluded. Accordingly, in the latter part of 1911 Dr.
Duke came over to my camp on Damba Island to shoot
and investigate the very abundant " Enjobe " there.
This antelope, the Situtunga {Tragelaphus spekei), lives in
most intimate association with the fly among the shaded
forests at the water side.

A number were shot and their blood was injected
into monkeys, one of which, injected on November 5th
and 6th, showed Trypanosomes in its blood on the 18th.
It was taken to the laboratory on the mainland and the
nature of the Trypanosome investigated in every possible
way, and Dr. Duke considered that there was no doubt
that he had at last found the source from which the

* Reports of the Sleeping Sickness Commission, No. XI, p. 71, etc.
' Ibid. No. XII, p. 117 et seq. See also article by Duke in British
Medical Journal, 1914, February 7th.

28 NATURAL HISTORY OF SLEEPING SICKNESS

flies on the island acquired the cause of Sleeping
Sickness.

This discovery made the question of the return of the
natives to the lake shore a far more difficult one than
was anticipated when they were removed during the
height of the epidemic. It was then thought that as the
fly was merely a mechanical carrier of the Trypanosome,
an interval during which all known sources of infection
(the natives) were kept away from the fly would allow
the disease to die out, and when the life of the infected
flies had come to an end, the natives who were free from
Trypanosomes could return without danger. But now
that it is known that there is a " vicious circle," the
fly acquiring the Trypanosome from the antelope and
in turn inoculating it into fresh animals, the islands and
mainland shore of the lake are still dangerous. It is
true that some consider that it is possible that the
Trypanosome has been for so long away from man's
blood that it may no longer be pathogenic to him,
but against this is the fact that as a result of living on
the islands with me in 1911-12 three of my native
employees were found to be infected when I examined
them before returning to the islands in 1914, and one at
least is reported to have died.

So far as we know at present, the fly would be harmless
without the antelope and the antelope without the fly,
and to eliminate the disease from the most fertile and
beautiful part of Uganda these two must be kept apart ;
that is to say, one of them must be exterminated.

In 1914 I obtained most interesting confirmation of
Dr. Duke's results. The Island of Nsadzi, lying opposite
to Entebbe and south of it, was well populated in the old
days, and there was very little refuge for Enjobe there ;
I was told by my canoe-men that the antelope was not
to be found on the island in those days. In 1911 the
flies on Nsadzi were tested and were found to be free

THE SITUTUNGA AND SLEEPING SICKNESS 29

from infection, for as many as 5,765 failed to cause infec-
tion in a monkey. In 1914, however, I frequently saw
footprints of Enjobe on the island, which had presumably
swum across the narrow channel between Nsadzi and the
large isle Kome to the east where Enjobe abound. The
flies were again tested and found to be infected, for after
2,076 had fed upon a monkey it showed the Trypanosome
in its blood. ^ Less than half the number of flies that
did not produce an infection in 1911 produced an infec-
tion in 1914, and this is associated with the arrival of
the buck in the continued absence of the population.

It seems hardly possible to entertain the idea of de-
stroying the Enjobe, since when hard pressed it takes
refuge in dense papyrus swamps, and even if it could once
be eliminated from the islands would soon find its way
back by swimming from the mainland, and would resume
its former relations with the native. Is it, then, possible
to eliminate the fly ?

On this question I shall have something to say in
the chapter devoted to the natural history of Olossina
palpalis.

1 Eeports of the Sleeping Sickness Commiasion, No. xvii, 1919, p. 71-74.

CHAPTER III

THE NATURAL HISTORY OF GLOSSINA
PALPALIS

This species is pre-eminently a frequenter of the shady
forests on the shore of the lake, where it may be found
in great abundance.

The method of estimating its numbers is to set trained
natives to catch the flies with small nets from off the
herbage or rocks on which they alight, or from their
own clothing when the flies come to feed. At the end of
half an hour the boys are whistled in and the catch
counted, but, for reasons which will be given later,
only the males are used. An average figure is obtained
representing the number caught per boy per hour,
commonly alluded to as the " male-boy-hour " figure,
by means of which one locality can be compared with
another. The highest figure I have yet recorded was
obtained in 1919 on the north coast of the west of Kome
Isle : four boys working for half an hour obtained the
average of 125 male flies per boy per hour. If it be
considered that each fly is caught in a net, from which
it has to be taken out by hand and put into a bottle,
and that besides the males there were at least a quarter
as many females caught at the same time, it must be
admitted that the fly can be very plentiful !

One of the first things to be realized is that the number
of flies at a locality has little relation to the frequency
with which one is bitten, and therefore an impression

30

ON DAMBA ISLE.
Not a good dy shore, as it was too marshy for a breeding ground.

Pf-s

A FLY BOY (JAKOBO) AT WORK.

The white net in his right hand scarcely shows against his white clothes. He was found to
be infected with T. g,anibiense in 1914.

To face p. 30.

PROPORTION OF SEXES

81

derived from a large number of bites gives a very erroneous
estimate. This fact has arisen from Fiske's investigations
into the reasons for the very varying proportions of the
sexes at different localities.

It has been found that the females are much less inquisi-
tive than the males, and do not come to man in any great
numbers unless wishing to feed, whereas the male roams
about and investigates any new object without necessarily
being hungry. The percentage of females is therefore
an indication of the hunger of the flies. If, however,
a prolonged catch is made in one spot, the percentage of
females rises, because the more inquisitive males are
first caught ; the number of females per boy per hour,
however, remains fairly constant, since only the hungry
individuals are taken.

An experiment such as the following illustrates this
point clearly. Flies were caught continuously at one
point on the small island of Tavu on January 13-16, 1914,
from early morning until the afternoon, and it is clear
that the number of males was soon affected, while the
catch of females remained constant. The percentage of
females, however, rose, owing to the males being caught
off ; not in this case because there were more females
eager to feed.

Date.

Jan.

15,

1914

»>

9>

»>

»>

Jan.

16,

1914

>>

Time.

7.45-8.45

9-10

10.15-11.15

11.30-12.30

1.30-2.30

7.45-8.45

1.25-2.25

2.45-3.45

Total

Catch.

Males per
Boy-Hour.

Females
per Boy-
Hour.

473

62.5

7.5

303

36.7

4.3

235

23.5

6.5

216

20.3

6.8

222

20.3

7.5

275

33.0

5.7

241

21.0

11.4

128

9 = 8

7.1

Percentage

of

Females.

10.7
10.5
21.7
25.0
27.0
14.7
35.2
42.0

The relation between the percentage of female flies in a

32 NATURAL HISTORY OF GLOSSINA PALPALIS

catch and the hunger of the flies was bome out by the
behaviour of flies on different islands visited by Fiske
and myself in 1914. (See Chapter VI.) Certain islands,
notably Kimmi and Yempata, were remarkable for the
avidity with which the flies bit ; on Kimmi they even
flew out to the canoe before it reached the shore. On
both these isles the percentage of females was higher
than usual ; on Kimmi it averaged 53 per cent., on
Yempata 37 per cent.

The curiosity and spirit of investigation of the male
fly was shown by a little experiment on Tavu Island.
I noted that during the passage through some thick
Kinsambwe bushes many leaves became reversed so
as to show the silvery under surface. Male Tse-tse
often settled on them as if attracted by the novelty.
I therefore spent some time catching every fly that settled
on my white net, held out conspicuously, and in one hour
thus caught twenty-six males, but only six females.
I then caught from off my clothes those flies that had
settled and lowered the proboscis intending to feed ;
there were ten females to eight males. Male flies may
very often be seen which seem to have no desire to feed,
but merely to rest in a conspicuous position, often on a
rock hot with the sun ; possibly they are on the watch
for females. There can be no doubt that they use their
eyes in the search for food : it has long been known in
the case of G. morsitans that a man on horseback or cycle
moving wiU attract flies where none have been noticed
while he was stationary. Palpalis would often alight on
the canoe drawn up on the beach ; presumably the large
black object was thought at first to be a hippopotamus !

The fact that the fly will often sit on a rock hot with
the sun is of interest, seeing that they are apparently
dependent upon shade, and are susceptible to variations
in the humidity of the atmosphere. I am convinced,
however, that so long as palpalis has a shady base it will

ROAMING OF THE FLY

33

make long excursions from it, perhaps in search of food
or breeding places, and may be found in places where
there is practically no shade and where it could not abide
permanently. Thus, the western end of Nsadzi Island
consists of a high treeless grassy plateau sloping down
steeply to a rocky shore along which is a scanty fringe
of sparse bushes throwing very little shade over the bare
rocks. Yet if one descends to the shore fly is met with,
and one is immediately investigated.

An experiment was made to test whether the fly does
roam along this kind of coast. At one point there was
a small copse about thirty yards square, furnishing ample
shade to provide a base for the fly, and ten fly boys were
set to catch flies continuously here on June 10, 1914.

Time.

Number of Flies
Caught.

Males per
Boy-Hour.

Females

per
Boy-Hour.

Percentage

Male.

Female.

9-45-11-15

11-20-11-50

12 -12-30

12-30-1

1-10-1-401

1-45-2-15

2-15-2-45

51
16
12
16

17
9

18

24

15

6

24

17

3-4
3-2
2-4
3-2

3-4

1-8

1-2

4-8

3

1-2

4-8
3-4

26
60
55-5

27

58-5
65

121

104

—

—

—

The fact that the number of males caught per boy per
hour was almost unaltered until the end shows that the
121 flies could not have been in the copse at the com-
mencement, but that a continual influx occurred. Had
the flies all been there at first ten boys would soon have
diminished the number. Also, after an interval spent
in looking for pupae there was a decided increase in the
total number of flies caught.

* One empty puparium found.
4

34 NATURAL HISTORY OF GLOSSINA PALPALIS

It is of some importance, for reasons that will appear
later, to know whether palpalis does roam out of the
shade. There is no proof whatever that it will ever cross
the open water from one island to another. The little
islet Sanga, off the east end of Kome, has everything that
the fly can want, but for some reason it is unable to
establish itself there, and none can be found. Yet the
adjacent shores of the larger Kome abound with fly, and
it could very easily find its way across the few hundred
yards of water to Sanga, yet it does not.

Palpalis, like morsitans, is at times a great nuisance,
for it will quietly sit on men's backs and thus be carried
a long way. Indeed, I have sometimes unwittingly taken
them back to my tent, where they remained, and rudely
bit me while having a bath ! On certain days, when it
is somewhat overcast but the sun shines through thin
clouds, they are perhaps at their worst, and are very
eager to bite, and will cling to one for long distances. A
very favourite trick is to alight on the under side of the
brim of one's hat and to quietly sidle round until the first
piece of skin free from hair is met with, when the proboscis
is lowered from its sheath and a sharp prick is felt on one's
temple. It is a curious fact that sometimes one is quite
unaware of being bitten, and may suddenly hear the
unmistakable heavy buzz of a bloated fly leaving after
a heavy feed ; at other times the prick is felt so sharply
that it might be produced by a needle. Presumably the
difference depends upon whether the fine point of the
proboscis finds a nerve end or not !

When a fly alights to feed it sidles about over the
skin and appears to be feeling for a soft place : when it
is satisfied the legs are slightly straddled out, and the
proboscis lowered from its sheath into a vertical position,
after which it gradually penetrates up to the bulb at its
base. It is then withdrawn slightly and the body of the
fly begins to swell with red blood. Almost at once the

FOOD OF THE FLY 85

excess of fluid is exuded as a bead at the posterior end of
the body, so that the fly is enabled as it were to have a
larger feed of corpuscles than if it were compelled to retain
all the fluid part. After a full feed the abdomen becomes
as large as a ripe red-currant, and the red blood shows
through the distended skin ; the gorged insect flies heavily
away to a neighbouring leaf and rests awhile to digest
the meal !

On clear hot days, when a strong wind may be blowing,
palpalis is not so eager to bite ; nor on any day during
the hottest hours. But about 4 p.m. it is always ready
again, and I have been bitten quite in the dusk of the
late afternoon : at night, however, it is harmless, nor
does it come to light as the big East African species
brevipalpis is known to do. This habit of biting again
more freely in the evening is of interest, for it is at this
time that the Situtunga antelope, which is a source of
food, comes out from the dense forests and wanders
along the edge browsing on the bushes, where it meets
palpalis.

The question of the food of Tse-tse flies is of the highest
importance, and was one of those set me by the Tropical
Diseases Committee of the Royal Society when I first
went out in 1910. In the case of morsitans certain ob-
servers, notably the late F. C. Selous, held that the blood
of buffaloes was essential to it, and that it died out or
migrated if none was available. Few observations had
been made on Glossina in a state of nature, but from
an examination of the blood taken from the stomachs
of wild flies the Commission in Uganda in 1910 ^ had
deduced that " the blood in the majority of the flies had
been obtained from birds or reptiles, and of these the
reptilian blood was twice as frequent as the blood of birds."
It must however be confessed that the method employed

^ Report XI of the Sleeping Sickness Commiasion, 1911, p. 112,
etc.

36 NATURAL HISTORY OF GLOSSINA PALPALIS

of distinguishing between avian and reptilian blood was
crude and open to a wide margin of error.

I repeated this work with a greater number of flies and
art greater length, and using as standards such sources
of food as the flies might meet with, viz. cormorant, croco-
dile, Varanus, python, frog and lizard.

At Jinja I had found that the proportion of mammalian
to non-mammalian blood in over 12,000 wild flies was
indicated by

M : N : : 31-5 : 68*5.

On Damba Island the proportions in over 6,000 flies were

M : N : : 21 : 79.

Now in both these places antelope's blood was readily
obtainable, for Situtunga abounded on Damba and bush-
buck were constantly seen in the fly area at Jinja ; hippo-
potami were of course available at both localities. In
1914 further work was done and blood was examined
from flies taken on islands where there were no Situtunga,
and others such as Kome, where Situtunga abounded.

In flies from five small islands not inhabited by Situ-
tunga the proportions of mammalian to non-mammalian
blood were :

M : N : : 4 : 96.

On two isles inhabited by the buck the figures were :

M : N : : 25 : 75.

Now on all islands hippopotamus blood is obtainable ;
indeed, on those of the first group it was the only possible
mammalian blood, and accounts for the 4 per cent, of
cases.

But even where Situtunga are present only about one
quarter of the food is obtained from them, and it appears
that Glossina palpalis much prefers non-mammalian
blood.

ANIMALS FED UPON

37

The question then arises, does this come from birds
or reptiles ? In 1911-12 careful measurements were
made in forty-six cases, that is, from blood taken from
flies that had fed recently enough to allow the corpuscles
to be measured unaltered by digestion.

In each case fifty corpuscles were traced with the camera
lucida, and the average measurements obtained for com-
parison with known types. Although in some cases a
given specimen was with difficulty allotted to any one
type, the following result was arrived at :

Source.

Percentage.

Actual
Number.

Source.

Percentage.

Actual
Number.

Avian
Reptilian

4-3

95-7

V

Lizards

Crocodile

Tortoise

56-8

40-9

2-3

25

18

1

—

100

46

—

100 44

In two cases birds appear to have been the source. These
would in aU probability have been cormorants or darters,
which so often sit on a branch at the edge of the water
with the wings spread open ; but the herons and ibis which
frequent the shore may also be bitten.

Nearly 57 per cent, of the blood, derived from reptiles
seems to have come from lizards, probably entirely from
the great monitor or Varanus, but measurements of the
corpuscles from it and one of the common small brown
lizards did not allow any distinction to be drawn.

There is no possible doubt that Varanus is a most
important source of food. I have several times seen one
slowly sculling along the surface of the water near the
shore, with the top of its head black with Glossina which
were evidently so troublesome that at intervals the
reptile submerged its head beneath the surface, only

38 NATURAL HISTORY OF GLOSSINA PALPALIS

to be reattacked the moment its head was raised. On
one occasion when sitting on the beach of Damba I saw
a Varanus come out from the forest and walk slowly
past me. Tse-tse were around me in considerable num-
bers, but the Varanus was evidently equally attractive, for
several went to it and began to feed, as was evidenced
by the repeated movements it made to scratch them off
with its hind legs.

Crocodiles are also an extremely important source of
food, and their habits favour this, for like Varanus they
often lie for long periods at the water's edge. Measurements
of the blood corpuscles in wild flies seemed to indicate
that nearly 41 per cent, of the reptilian blood is from
this source. It appears at first sight rather remarkable
that the. horny plates should prove no impediment to
the proboscis of the fly ; yet Fiske and I on Kimmi in
1914 watched closely for half an hour a small crocodile
which the natives had caught and was tethered near the
water, and between 8- 35-9- 45 a.m. forty flies had fed
fully ! This figure did not include those that were
disturbed and did not feed to repletion.

The sites chosen by the fly were carefully noted in each
case save two, and are indicated below :

Eyelid

. 12

Hind leg . .

9

Neck

6

Flank

2

Fore leg . .

2

Nose

2

Shoulder . .

Nostril

Angle of mouth .

Hind foot . .

Back

, ,

38

It is seen that the fly was even able to penetrate through
the back !

ANIMALS FED UPON 39

On Ngamba Isle I was able to see two flies upon a large
crocodile that was resting among the bushes and had
not run away at my approach ; one of them had certainly
fed.

In a single case out of the forty-four in which the
blood in wild flies was recognizable the corpuscles agreed
well with those of a tortoise. Olossina seems to recog-
nize even this as a source of food, for I saw a tortoise
one day with a fly on its back making vigorous attempts
to find a way for its proboscis through the shell. But
though a crocodile's plates are not insuperable, I fancy
that the carapace of a tortoise is proof, and the fly would
have to penetrate a leg or the neck.

No example of blood that could be attributed to snakes
was found among the forty-four cases, but the python
certainly must be counted as a source of food. On
Kimmi in 1914 I came upon a small specimen lying among
marshy grass close to the water, and two flies were on
it, one of which had the swollen red body indicative of
a good feed. Other snakes also seem to be attractive,
for on Ngamba in 1914 I saw a black species lying on
rocks about to shed its skin and in the usual lethargic
condition. Two male flies were upon it and were
obviously attracted by it, though they did not seem
desirous of feeding.

Lastly, the lung fish, Protopferus (the " Mamba "),
has been suggested as a source of food, though I do not
understand how it could serve as such, seeing that it only
appears at the surface of the water for a brief interval
to obtain a supply of air. Examination of its blood showed
that it is a physical impossibility for the fly to feed upon
it, for the corpuscles are so enormous that only one at
a time could pass up the proboscis, which would require
a very great effort of suction on the part of the fly !

The figures that have been given represent the result
of examination of many thousands of wild flies. For

40 NATURAL HISTORY OF GLOSSINA PALPALIS

out of a large number caught on a given day only a few
would contain fresh blood, and in these cases a proportion
would be mammalian blood. In the other cases many
would be merely recognizable as nucleated non-mammalian
blood, and in only a very small proportion would the
corpuscles be sufficiently unaltered to allow of measure-
ments being taken.

During this work note was taken of the presence or
absence of Trypanosomes in the gut of the fly, and in the
proboscis. Certain Trypanosomes (e.g. T. vivax, which
kills goats very quickly) only develop in part of the
proboscis and are not found in the gut ; but others, such
as gambiense, may be found in both.

Of 600 flies examined at Jinja, 11 per cent, contained
Trypanosomes in the gut ; of 695 on Damba, 3-4 per
cent, had Trypanosomes ; and on Bugalla 1,000 flies
showed Trypanosomes in the gut in 1*7 per cent, of the
total. Out of 638 cases in which the proboscis was
examined on Bugalla, 14 had Trypanosomes in the
proboscis alone ; i.e. 2-2 per cent, of the wild flies appear
to be infected with Trypanosoma vivax. ^

It must not be supposed that all the Trypanosomes
found in wild flies are gambiense : some are devived from
the crocodile {T. grayi), and so far as is known are harm-
less.

Besides Trypanosomes bacilli are often found in count-
less numbers in the gut of the fly, but in a different part.
Nevertheless there appears to be some inverse relation
between the two, for out of six hundred flies at Jinja,
in only 3*4 per cent, of flies containing bacilli were
Trypanosomes also found, and in only 6-1 per cent, of
flies containing Trypanosomes were bacilli found.
Bacilli were present in 19*3 per cent, and Trypanosomes

1 Dr. Duke has shown that the Situtunga is a natural host of T. vivax
as well as of 2\ gambiense. Report No. XII of the Sleeping Sickness
Commission, p. 122.

^ n

Hb

O 5

H

<

a

="•0

To face p. 40.

BACILLI AND TRYPANOSOMES IN THE FLY 41

in 11 per cent, of the wild flies. Thus there is marked
incompatibility between the two.

Since I have found bacilli in the gut of freshly hatched
flies, and even in pupae, their presence in the fly may have
something to do with the fact that only a few out of a
batch of flies fed upon an infective animal at the same
time will prove suitable hosts and will subsequently be
found to contain Trypanosomes. It may be that the
presence of bacilli in numbers is inimical to the Trypano-
some, or merely that they are present in flies which for
some other reason are physiologically unsuited to the
development in them of the Trypanosome. It would
be interesting to note the presence or absence of bacilli
in flies which contain T. vivax only in the proboscis.

One of the points set me to be investigated was whether
palpalis can feed on anything except blood. I have not
been able to obtain any evidence that vegetable juices
are sucked up save that on Damba Isle I several times
found grains of banana starch in preparations made from
the gut of flies. I am inclined to think, however, that
this was accidental contamination of the preparation,
perhaps from a cloth used to wipe the slides, for it is
difficult to see how such material as banana could be
sucked up the narrow proboscis of Glossina. Small
fragments of vegetable tissue were sometimes met with
in preparations from the gut of a fly, but I am disposed
to think that they had been sucked up in water, for
pieces of alga were also found.

Flies paid no attention to juicy papai fruit taken
down to the shore, although I have seen a mosquito
greedily feeding thereon.

There seems more evidence that palpalis imbibes
water, but though they have often been seen sitting on
wet mud, no fly has ever been seen with its proboscis
lowered into the water.

Examination of the contents of the gut of many

42 NATURAL HISTORY OF GLOSSINA PALPALIS

hundreds of flies on Damba and Bugalla Islands, as well
as at Jinja on the mainland, has shown various forms
of algae and other low vegetable organisms that must
have come from water ; and once a very minute ostracod
crustacean, only about three times the length of a
human red blood corpuscle, was found. Figures of
these are given in my first and second reports.

Experiments were made to see if flies lived longer when
able to obtain juice from fruit than when starved, but
access to fruit made no difference. The longest time for
which a wild fly has been kept without food or with only
fruit juices available is nine days ; a freshly hatched
fly will live for eleven days. After a single feed of blood
a fly lived for fifteen days (the longest period). This
is a remarkably short period, and if it is so in natural
conditions is rather surprising. Animals such as ticks
appear to be able to live for months after a single feed of
blood; the fly, however, is a more active creature, and
lives at a higher rate.

What is the natural length of life of palpalis ? This
was one of the first things to be settled when I went
out in 1910. For it was then not at all clear why the
flies on the lake shore continued to be infective several
years after the presumed sources of Trypanosoma gambiense
(i.e. the natives) had been removed from reach of the fly.
There were two possibilities — either the fly was acquiring
infection from some wild animal, or the same flies were
still alive on the shore that had been the cause of the
epidemic some years before. It was not until 1911 that
Duke showed what the reservoir of gambiense was, and I
have showed that the fly probably does not live for more
than one year. The drier months of the year cause a large
falling off in numbers, so that a fly which has been living
for some months probably succumbs when the trying
period of diminished humidity comes at the beginning
of the year.

:^/0 /kia.

Se/i

CHART
t. Oct. /<7of.

II

i^ff, -

Jan.

o
73

76

1

!

75 ■

- '<.

-

_

74-

"^s

_]

7^

N

percentaffe

A

7Z

"^.

' \

?Z

7/

V

4.

/

\

7/

70

k

\

—

fo

69

\

/

\

5J)

68

^s

,/

\

Zp

f,7

1 '

\

67

66

\

65

66

V-

Zp

Bi-

\

_5^

63

\

63

62

Tfi/ierdtUrt
'fa/irenhf/t)-

il^

74

(

74

73

_ -— pfcs

J^

73

7Z_

^

r'

■^^

4-f

-^ -

__

n

72.

7/_
70

-A

70
69

f

/o

9
8

7

6
5

4_
SO
J9

69

y^'

_j

48.

^€l

^

//

10

periioy
per hour

!^

^__^_

v

g

/

~?

-A

8

«> —

H w

/

7

N

6

^^ ,

/

s

■"k.

/

4

■~">

-./

.€0

■59

\

/\

56

'- \

/ v

s

58

57

\

firrcentage
of/ema/es
Xaught

/

\

37

J6

\

-4

s

V

Sd

35

\

/

\,

55

-54

\

^^

,/

54

JJ

\

A

V.

--,

J

53

SZ

\^^

sr

-

52

5/

^__

^^ ^j

5/

INFLUENCE OF CLIMATE ON NUMBERS AND PROPORTIONS OF

FLIES AT JINJA, ON MAINLAND.

. By permission of the Royal Society and the Controller, H.M. Stationery Office.

To face p. 42.

DURATION OF LIFE 43

A chart made at Jinja in 1910 (see Chart II)
shows very clearly the direct relation between relative
humidity of the atmosphere and the number of flies that
could be caught.

In order to estimate the length of life of the wild fly,
the method adopted was one that had been used by
Dr. Bagshawe to ascertain how far the fly would follow
canoes or men.

Large numbers of flies were captured, a portion of one
leg was amputated through the middle of either femur or
tibia (limbs may be lost at a joint naturally, but never
in this way), and then the fly was liberated. By con-
tinually catching flies at the same point day after day
it could be ascertained for how long a marked fly had
lived.

Since a fly has six legs and each can be marked in two
ways, twelve markings are possible : one method of marking
was kept for the flies of each week. At Jinja, between
July 27th and October 15, 1910, some eight thousand
flies were so marked, and eight markings were used, an
average of a thousand to each marking.

Between August 3rd and February 11th, careful exami-
nation was made of each fly caught by the boys, who
worked every day for four or five hours. On January
31st, 1911, a female fly was taken that had been marked
on one of the days July 27 to August 2, 1910, and on
December 29th a male was taken that had been marked
on one of the days July 27 to August 2. This gives
a period of life of at least 182 days for the female and
149 days for the male, but it is of course impossible to
say for how long before the marking the flies had been
alive : probably they were hatched fairly recently.

On Bugalla Island in 1912 this experiment was repeated,
and between March 18th and April 26th nine thousand
flies were marked in six different ways. The last marked
flies to be caught were noted on January 17, 1913 : two

44 NATURAL HISTORY OF GLOSSINA PALPALIS

males that had been marked between April 1st and 6th.
This gives a minimum duration of life of 247-253 days
for the male. A female % caught again on August 17th
had been marked between April 8th to 13th, so it had
lived at least 126 to 131 days. This was not quite so
long as the longest Jinja period (182 days). Thus it was
established that palpalis can live for several months, but
the probability is that it does not survive a second dry
season. In both experiments the last flies were caught
at a time when the relative humidity of the air was
low.

Both at Jinja in 1910 and on Bugalla in 1911-12 care-
ful meteorological data were kept, and the influence of
climatic conditions upon the total number of flies that
were caught per boy per hour was found to be very
interesting (see Chart III).

In those days the male catch alone was not used as a
means of estimation.

The number of flies caught at Jinja on the mainland
varied directly with the relative humidity of the atmosphere,
which itself varied during the period from the beginning
of August to the end of January from 63 per cent, to
76 per cent, (see Chart II).

On the other hand, on Bugalla Isle no such relationship
could be made out, but here the relative humidity only
varied from 77 per cent, to 72 per cent, during a continuous
period of twelve months. On the mainland there was a
very decided inverse relation between the number of
flies caught and the temperature. In the hotter (and drier)
months there were fewer flies. This did not seem to be
so markedly the case on Bugalla, probably because the
humidity of the atmosphere was so much more constant
on the island that a higher temperature did not necessarily
kill off the flies.

In this fact may be the explanation of the occurrence
of flies on Nsadzi at the west end, where there was so

iua.Al.?.^;^?-r-i11c^ g g 5 g : e 3:^ g:S 3 i3 ;3 3 rig g--^ J! t'

a

""

;s

M

a:

O

4.

o

>

"p

M

■*->

H

JS

?;

<!

<u

c>

S

n

i-l

"*-

u

S

"2

c

o

"rt

?s

u

fi

•??

r )

u

^

K

o

C/2

D

«

1-1

B/l

f^

«

OS

H °5

2" >"■ ?^~?-"?- r- P-

To face p. 44.

INFLUENCE OF CLIMATE 45

little shade : the humidity of the island atmosphere
saved them from the heat that might have been fatal
on the mainland.

Secondly, the hunger of females, as estimated by the
numbers caught, is very dependent upon the temperature.
Fiske proved to me in 1914 that the female percentage
is not an indication of the actual numbers but of the
hunger of the fly. Since the percentage figure varies
inversely each month with the temperature, as is shown
by the charts for Bugalla and Jinja, it follows that the
hotter the month the less the fly is inclined to bite.
Indeed, I have noticed on individual days that the fly
bites less during the hottest hours, and less on a brilliant
sunny day than when the sun's heat is tempered by thin
cloud.

A further effect of climate was noted on Bugalla,
namely, upon the number of pupae that could be found
each month (see Chart IV).

At the end of every month the boys visited certain
well known sites much liked by palpalis for its pupae,
and an average figure was obtained called " pupae per
boy " to indicate empirically the number found per
month. It is obvious that the total number found at
the end of each month must depend in part upon the
number of female flies, which is indicated, at least as
regards the number of hungry flies, by the " female-boy-
hour " figure for the month.

If the " pupae-boy " figure be divided by the latter
figure, the quotient may be used as a means of comparing
from one month to another the rate of reproduction,
bearing in mind that the times when the flies are hungriest
will thus appear as months when the reproduction figure
" pupae per female " is lowest.

Thus, the average number of pupae found per boy in
the given time at the end of February was 187-3 ; the
average number of females caught per boy per hour during

46 NATURAL HISTORY OF GLOSSINA PALPALIS

187-3
that month was 6*3. = 29*7, which is the " pupae

D* O

per female " figure for the month.

If these figures are plotted out for the twelve months,
it is seen that the curve thus formed is inverse to the curve
of relative humidity. This may be a definite adaptation
to enable Glossina to tide over the drier months, which
are adverse to the adult fly. That is, during the drier
months (of lessened humidity) more pupae are deposited,
the pupa being resistant to a greater degree to adverse
climatic influences than is the perfect fly.

Experiments were performed to test the vitality of pupae
exposed to various conditions of dampness or drought.
Batches of pupae were submerged in water daily for differ-
ent lengths of time, and it was found that daily submersion
for twelve hours on twelve successive days only destroyed
36 per cent, of the pupae, and that only 33 per cent, were
destroyed by four submersions of twenty-four hours,
with an interval between each of twelve hours. The
effects of continuous submersion were tried, and it was
found that a period of 108-120 hours is necessary to
destroy all the pupae in a batch.

Flotation on the surface can be borne for eight days,
but ten days are fatal.

Pupae were also put in a smaU wooden box covered
with one inch of earth, and the box was sunk in the ground
flush with the surrounding surface and left exposed
to the sun during the hottest hours of the day. A single
exposure up to six hours does not harm the pupae, but
if uncovered by earth a single hour is fatal. The effect of
repeated exposures on successive days was then tested :
two hours daily for three days made very little difference
to the pupae, but three hours daily for three days resulted
in only 2 per cent, of flies emerging ; six hours daily for
two days killed all. Lastly, alternate submersion and
exposure are fatal ; although a single submersion of

•3IEUI3J J3d aedntj

To face p. 46,

FLUCTUATIONS IN NUMBERS 47

twelve hours, or four hours exposure to sun have, alone,
very little effect, if the pupae were submerged one night
and exposed to sun next day only one per cent, developed
into flies !

Besides variations in numbers of the fly due to climatic
conditions, I found in 1914 interesting variations that
were very puzzling. The small islands Bulago, Kimmi,
Tavu and Ngamba were each visited weekly from the
camp on Kome, and catches were made of flies at the
same spots around the coast.

It soon became obvious that the numbers of flies (as
estimated by the number of males caught per boy per
hour) on each of the four islands did not vary concurrently,
and that variations did not affect the sexes in the same
way on the same island (see Charts V and VI).

On Bulago, for instance, on the south point and west
shore the two sexes on the whole varied together, but
on the shore of the north bay they varied inversely.
In the forest of the north point the variations in number
of the males were very great, while the number of females
kept very constant. On Kimmi it was noted that the
catches from two localities on the north and north-west
shores of the island, taken together, varied inversely to
the catches from the south and south-east shores, though
this was less marked after the beginning of July. This
must mean that at certain times the male flies found the
north shore more congenial, whereas at others they
congregated on the south side of the isle.

By charting together the figures representing the
average catch per boy per hour of males and females for
each of the four islands, a curious and interesting point
was brought to light.

On the island where the number of males shows the
greatest fluctuations the female figure shows least, and
vice versa, and the same was found to hold good for
individual localities on Bulago Island. On Tavu Island

48 NATURAL HISTORY OF GLOSSINA PALPALIS

the curve for male flies shows great variation, the female
curve little, but on Kimmi the female curve is the
greater variant and the male curve more constant. It
may be that the supply of food on Tavu is so good that
the females are well fed and do not come in any numbers
to be caught by the fly boys, whereas on Kimmi the food
supply is more erratic : it must be said, however, that
there is no direct evidence for this.

A very interesting relation was made out on Tavu
between the number of flies caught at the weekly visits
and the number of crocodiles seen when the island was
circumnavigated in the canoe before a landing was made
at each visit. There was a large flat topped rock on the
east coast forming a little plateau that was always in use
by crocodiles as a basking place ; the number on it
varied from one to four, and others were seen in the water.
On one visit as many as fifteen large crocodiles were
seen on or. about this small island, and on that day the
number of flies was at the rate of ninety-three per boy
per hour ! The lowest figure was fifty -six per boy-
hour when only two crocodiles were seen. The accom-
panying chart shows well the concurrent variation in
numbers. It must be confessed that the explanation of
this is not at all clear. One would have rather expected
that when there was an abundant supply of food on
the island the flies would have been less hungry, so that
fewer would have been attracted by men, and the catch
would have been less instead of greater the more crocodiles
there were within reach. On the other hand, had this
relation been noticed at one spot on a continuous length
of mainland shore it could quite well be explained by
the wanderings of flies in search of food which, finding
abundance at one spot, remained there.

But, as has been jpointed out, there is no ground for
believing that flies cross to Tavu over the open water
separating it from the larger Kome and Bulago.

AN AM BATCH TREE.

FLY BEACH ON KIMMI.
Behind a row of Ambatch trees on the right.

To face p. 48,

ITS ENEMIES 49

It is noteworthy also that this explanation, were it
possible, would be opposed by the fact that the
number of females caught on Tavu varied very little,
unlike the male figure, whereas if the flies had been more
attracted to Tavu on days when the crocodiles were
numerous, there should have been greater variations in
the female catch ; perhaps an inverse relation to the
number of crocodiles.

The explanation, however, may be connected with
the pairing of the flies. Male flies flock to the coast from
out of the forests when crocodiles are there in numbers
in order to meet the females, which go to feed on the
crocodiles. When crocodiles are fewer on the coastline
the male flies may confine themselves more to the forested
centre of the island : the boys caught always on the
east and west shore and not in the central forest. It
will be remembered that on Bulago and Kimmi the flies
seemed to frequent different parts at different times.

Having considered two influences that affect the num-
bers of the fly, viz. food and climate, it will be interesting
now to deal with the natural enemies of Glossina
palpalis in all its stages.

Considerable attention was paid to this question, and
I have spent very many hours observing, in the haunts
of the fly, birds and insects that might be expected to
destroy the fly.

Very many insects suffer great mortality in their im-
mature stages of egg, larva and pupa, but it will be
obvious from what has been said about the reproduction
of Glossina that it is hardly likely that any great destruc-
tion can be wrought upon egg or larva while within the
body of the adult fly. It seemed possible, however, that
the fat grub in the fly might be " stung " by some
parasite such as the Hymenopterous " Ichneumon flies,"
and that after it had been extruded the adult parasites
might emerge from the pupa.

5

ISTORY OF GLOSSINA PALPALIS

ide some efforts to obtain the larvae
js captured when the larva was far
^th and kept in a cage, fed at regular
the larva had been extruded. But
premature expulsion of the larva, ending in its death,
took place so often that the attempt was given up.
. The next opportunity afforded to enemies is the brief
moment after the larva has been extruded and is scramb-
ling over the surface of the sand in the endeavour to find
a spot where it can penetrate and hide itself. I have
watched this happen ; the birth of the larva in the eight
cases witnessed took place between 10"45 a.m. and I p.m.
It is very active, but being white with black bosses
at the posterior extremity, is quite conspicuous as it
wriggles over the surface. On several occasions I have
placed the larva in front of a species of ant, Paltothyreus
tarsatus, and seen the ant pick up the larva, give
it several nips with its large mandibles, and carry it
off. Once, on arrival at the breeding ground, among
some bushes where I used to lie and watch for events,
I saw one of these ants wandering about with a larva
in its mandibles.

This ant was very often seen on the beaches where the
fly breeds, wandering about and evidently searching for
food, and no doubt it does occasionally destroy a larva,
as was once seen. But it cannot often enough happen
to meet a larva before it disappears into the ground to
be of value as a controller of the numbers of the fly. Birds
also may occasionally pick up the larva while it is thus
exposed, but it can only happen rarely.

The greatest destruction of Olossina by enemies probably
takes place during the several weeks that it lies just below
the surface of the soil as an inert pupa, but I have obtained
no evidence of any vertebrate enemy. No traces have
ever been seen on a breeding ground such as would be
left by a bird scratching the surface as do game birds.

ENEMIES OF THE PUPA 51

It will be said that there were no guinea fowl or
francolins (save a single forest francolin) on the islands.
Nor have any evidences of the activities of such an
animal as a shrew been found on the breeding
grounds. The most important enemies, therefore, will
be insects, and accordingly efforts were made to breed
out from the pupae such enemies as the Chalcididae,
Hymenopterous parasites or " Ichneumon flies " of
minute size which lay their eggs in the immature stages
of other insects, particularly the eggs and pupae, as butter-
fly breeders know to their cost !

On Damba Island five thousand six hundred pupae
were kept in boxes with closely fitting glass lids, but
not a single one yielded any Chalcids, although from the
pupa of another species of fiy obtained with Tse-tse
pupae a species of Chalcid did emerge.

However, I felt certain that Glossina pupae were attacked,
for occasionally an empty pupa case was found with the
minute circular hole in the shell through which the Chalcids
had emerged. At last, on Wema Isle in 1914, when Fiske
and I were opening numbers of pupae to observe the stage
of development, he found one filled with rows of little
white Chalcid pupae, looking like mummies with black
eyes. I obtained more pupae from that locality and
succeeded in rearing some of the Chalcids, which proved
to be a new species that has been named Syntomosphyrum
glossinae.

It has been mentioned that some Glossina pupae were
found with the minute round hole made by the Chalcid,
but others are quite commonly found with holes of much
larger size and of jagged outline. I have never been
able to ascertain what is the insect responsible for this,
but in the case of the allied species Olossina morsitans,
a species of Mutilla (ant-like, wingless, fossorial Hymenop-
tera) has been found to be responsible. The female
Mutilla deposits her egg in the Glossina pupa, and its

52 NATURAL HISTORY OF GLOSSINA PALPALIS

larva destroys the Glossina. It has since been found out
that the Chalcid Syntomosphyrum destroys not the
Glossina but the Mutilla, which has itself destroyed the
Glossina, so that this Chalcid works on the side of and
not against Glossina.

When the adult fly has broken through the end of the
pupa case and pushed its way up through the loose soil
covering it, it seeks for a stem up which to crawl so that
it may rest with the wings hanging vertically downwards
while they stretch and dry. At this time the fly is a
soft white juicy morsel, and any ant that happened upon
it would find a pleasant meal. But the fly seems to
realize its danger, and is very much on the alert. More-
over, its legs are surprisingly firm, although it appears
so weak : I have several times watched newly hatched
specimens resting on a pebble until the wings had har-
dened, and. if an ant came near enough to be dangerous
they would quickly sidle away.

An interesting insect, the " ant lion," makes its conical
pits in the dry loose sand used as breeding grounds by
Glossina, and lies in wait at the bottom ready to seize
such insects as fall into the pit. Since these pits are often
very numerous in the soil over which the freshly emerged
fly has to scramble, the helpless insect must sometimes
fall into one of these pits and be devoured before the
wings have expanded, and I have actually seen one as
it scrambled along fall down a pit. But the " ant lion "
had either pupated or was not hungry, for it paid no
attention and the fly scrambled out again.

These pits are often so abundant in the very spots
selected by the fly for its larvae to burrow in that I think
it cannot be uncommon for the freshly hatched flies to
fall in and be devoured.

Much time has been spent in observational work on
possible enemies of the adult fly. On Damba Island
forty-four insectivorous birds feeding in the fly belt were

ENEMIES OF THE FLY 53

examined, and the stomachs searched for the wings of
Glossina, easily recognizable by their peculiar venation,
but none were ever found to have swallowed Glossina.
Twenty more birds were shot on Bugalla with similar
results.

A common green tree frog was examined, and in fifty-
three cases no Glossinae were found in the stomachs. This
species spends the day among the vegetation on the
shore, but I think only feeds at night, so that it is not
likely to meet with Glossina.

A large Agamid lizard haunts trees, around which there
is a certain space of open ground, and I shot five of them,
but their stomachs were found to contain only ants.

It does not seem probable, in view of the above evidence,
that vertebrates play a great part in keeping down the
numbers of the adult fly.

Of invertebrate enemies of the fly there is no doubt
that species of Bembex, sand-wasps or Fossors, are of
some value, but their distribution is very local.

A full account of these most fascinating insects will
be found in the chapter on Hymenoptera : it suffices to
say here that two species ^ have been seen to store up
Glossina in their underground burrows, and that they seem
to know that fat full-fed flies, which they prefer for the
food of their larvae, are to be obtained from man and
other animals. I have seen the same in the case of a
species (unidentified) which preys upon Glossina morsitans :
as one walked along, pursued and harassed by the Tse-
tse, the Bemhex would accompany one, flying round and
round searching for a fat fly.

A species of Dragon fly {Cacergates leucosticta) which
is very common on the lake shore has discovered the
same thing, and individuals have several times been seen
to chase Glossina, and once or twice I have actually seen
them catch and devour the fly. On one occasion I was

1 B. forcipata, B. capensis.

54 NATURAL HISTORY OF GLOSSINA PALPALIS

able with glasses to see Cacergates in attendance on a
hippo that was grazing on Bugalla fly ground, evidently
on the look out for flies full of hippo blood !

Predaceous two-winged flies of the family Asilidae
have once or twice been seen to be devouring Olossinae,
but I do not think that the Asilid is of any importance
as a regular enemy.

On the whole, it seems that the chief enemies of Olossina
are among the Hymenoptera, but I am inclined to think
that great loss of life does not result. Olossina has
such an extraordinarily slow rate of reproduction that
it can have few enemies : the greater the number of
offspring the more must be destroyed by enemies if the
species is to be kept within bounds, and vice versa. Clim-
atic conditions, in my opinion, together with facilities
for breeding, are the most important influences affecting
the numbers of the fly.

We now come to the question of the " breeding grounds "
of palpalis, on which a great deal of work has been done.

The very peculiar method of reproduction of the genus
Olossina was reported by Bruce in his papers on Nagana ;
hitherto Olossina morsitans had been supposed to deposit
its eggs in buffalo droppings, and this was held to account
for the supposed fact that this Tse-tse could not live
apart from the buffalo.

When palpalis was proved to be the carrier of Sleeping
Sickness efforts were made to discover its pupae, with a
view to the possibility of destroying it in large numbers
when it could be easily reached.

Dr. A. G. Bagshawe ^ in Uganda was the first to find
the pupae of palpalis : they were deposited around the
bases of stems of banana plants. Later he found them
sparsely under sundry bushes and masses of creepers,
but never in such great numbers as can now be found.

Dr. C. Marshall and Lieutenant A. D. Fraser, R.A.M.C.,

1 Reports of the Sleeping Sickness Commission, 1908, IX, p. 48.

GLOSSINA PALPALIS AND PUPAE.
Slightly less than natural size.
Photograph by A. Robinson.-

•UPROOTED TREE ON KIMMI ISLE

FORMING IDEAL SHELTER FOR

PUPAE.

SEARCHING FOR PUPAE AMONG

GRASS TUSSOCKS ON SOUTH

POINT, BULAGO ISLE.

To face p. 54.

BREEDING PLACES 55

were the first to find large numbers of pupae, ^ and reported
as follows : " They were most readily found on the
shore within a yard or two of the edge of the forest. It
has since been found that the nearer the forest the more
plentiful they are . . . and that the most favourable
place is close to the undergrowth that edges the forests
— in the loose dry sand near the roots of ferns, etc., and
shaded to some extent by the tall forest trees." This
excellent account may now be enlarged upon as a result
of many months' work and thought devoted to the breed-
ing places of palpalis.

It may be pointed out firstly that there are two broad
distinctions possible — between large areas of shore suit-
able, for reasons to be indicated later, along much of its
length for the pupae, and isolated spots at the bases
of trees, individual rocks, etc., such as the spots first
described by Dr. Bagshawe.

The former large areas may be termed " breeding
grounds," the latter I have termed " loci," and they will
be considered in detail.

The requirements of the pupae of Glossina palpalis
may be summarized as follows : " Loose dry soil, well
shaded, but with the surface thoroughly ventilated ;
within a few yards from the water but beyond its reach." ^
It is obvious that the soil must be loose, else the larva
would be unable to burrow down. Hence, however
suitable a spot may be in many other ways, if it is hard
and bound down by rootlets of grass it will not be a
good place for larvae to burrow into the soil.

The character of the soil is important ; that it must
be dry and well ventilated has been proved over and over
again. A very interesting observation on Tavu Island
in 1914 showed this very well. Close to a very favourite

1 Progress Report on the Uganda Sleeping Sickness Camps, by Dr.
A. D. P. Hodges, 1909, Appendix C, " The breeding grounds of Olossina
palpalis.'''' The Sleeping Sickness Btireau, 1909.

2 See my Fifth Report, p. 91.

56 NATURAL HISTORY OF GLOSSINA PALPALIS

breeding ground on a ridge was an area of low-lying sand
covered by dense vegetation, the leaves of which lying
on the sand were mouldy and kept it damp. My fly
boys searched for pupae there unsuccessfully, and in so
doing of course cleared away the dead leaves and opened
up the vegetation, so that a small area of sand about
two feet in diameter was much more freely ventilated,
though still shaded.

On a subsequent visit a week later I found, in that very
spot, about two dozen fresh pupae which had been deposited
there since my former visit. Incidentally this observa-
tion shows a very important point which will be alluded
to later, that the female fly searches carefully for the
most suitable spot for the pupae.

The soil itself may be of very diverse kinds : fine
pebbles, pebbles mixed with coarse sand, coarse brown
sand, coarse white sand, fine white sand, light friable
earth, vegetable humus and debris, and the very fine dry
dust found in caves and probably mostly derived from
disintegrated droppings of bats. Coarse sand with or
without pebbles has been found to yield the greatest
number of pupae : the very fine white sand such as com^
poses the southern beach of Nsadzi (see Fig.) is not so
suitable, because I think it holds moisture too much.
The very fine dry dust of caves does not seem very attrac-
tive to the flies ; possibly it is in its turn too dry.

Vegetable humus and debris are most important, for on
long stretches of shore they often form the only material
suitable for pupae, elsewhere there being only bare rocks.
Little pockets of humus and dried-up debris under angles
of rocks or at the bases of trees in such areas form the
only spots in which the fly can deposit its larvae, and
these are what I have termed " loci." Other " loci "
are sometimes formed in a forest by loose sand up against
the base of a tree trunk not far from the water, often
indicated as suitable for palpalis pupae by the presence

• :'^'*''rf . '■

Jv^^vt /

TREE TRUNK ON BUGALI.A ISLE UNDER WHICH MANY GLOSSINA PUPAE

COULD ALWAYS BE FOUND.

By permission of the Royal Society and the Controller, H.M. Stationery Office.

BREEDING PLACE OF GLOSSIN'A AT HOLLOWED BASE OF TREE ON BUGALLA.
By permission of the Royal Society and the Controller, H.M. Stationery Office.

To face p. 56.

BREEDING PLACES 57

of the pits of " ant lions," which have much the same
requirements.

On the little Isle of Lula, where fly was found In 1918-19
to be so numerous that the catch was at the rate of 56-5
males per boy-hour, there is nowhere for the fly to deposit
larvae except in " loci " — there is no sand or gravel beach
anywhere on this rocky little island.

This figure however is exceptional, but I have no doubt
in my own mind that palpaUs can exist in considerable
number on a coast line where there are only scattered
*' loci " in which to place its larvae.

The next point to be considered is shade, which is of
great importance. The pupa is dependent upon protec-
tion from the sun, although, as I have shown above, it
can survive a certain amount of exposure.

Shade for pupae may be either permanent or variable.
Permanent shade is afforded by (1) caves or large rocks
undercut at the base ; (2) prostrate tree trunks arching
over the ground ; (3) hollows at the base of or among
roots of living trees ; (4) the pent-house roof formed by the
disc of earth torn up by the roots of an overturned tree ;
(5) thick bushes.

I have not found caves to be nearly as productive as
I had expected; on Kimmi Isle, where flies were caught
at the rate of 33 males per boy-hour, pupae could only be
found at the rate of 6-7 per boy-hour. On the other
hand, at a locality on Wema where the fly rate was 59*5
per boy -hour, pupae were found at the base of an under-
cut rock at the rate of 72-4 per boy-hour. I think the
soil in caves is almost too dry.

Prostrate tree trunks arching low over the ground
form ideal sites for pupae, and the one illustrated was a
favourite collecting ground on Bugalla. On the same
beach on Bugalla and close to it was a tree much hollowed
at its base, which is figured ; it proved to be a very good
collecting ground for pupae.

58 NATURAL HISTORY OF GLOSSINA PALPALIS

When a tree is torn up by the wind and, unable to
fall, is held up by its neighbours, the roots and earth
embraced by them form an ideal shelter for pupae if the
tree is near enough to water. One such was found on
Kimmi in 1918-19 and is here illustrated. In the loose
dry earth forming an area of a few feet square under the
roots, four boys in half an hour secured 309 pupae, giving
a rate of 154-5 per boy-hour, which is extremely high.
Search was made for pupae in other localities near this
tree, but the result was only 28 per boy-hour !

Lastly, thick bushes, especially the " Oluzibaziba "
(Alchornea), provide shade all the year round, and if the
soil and other conditions are suitable, pupae may always
be looked for there with success.

The illustration shows a very favourite collecting ground
on the fly beach of Damba Island : under the bushes
close by but out of reach of the water.

Variable shade is formed by low undergrowth or creepers
or dense young growth of bushes near the ground. I
have been much struck with repeated evidence of the fly's
care in selecting spots where new young growth is rapidly
forming dense shade, and the quickness with which such
shelter is seized upon by the mother as suitable.

It was a common occurrence to find a cluster of some
creeper forming a dense tangle of fresh green that provided
admirable shade for the pupae, and investigation of the
pupae found there showed that they were in a very early
stage of development ; that is to say, had been recently
placed there. On Wema a very thick tangle of young
sprouts of a species of Polygonaceae was found on a ridge
of sand, projecting so as to arch over the surface below :
one boy found here eighty-eight pupae, of which 91 per
cent, had only recently been deposited, for they were
in a very early stage of development.

Moreover, only six empty cases were found from which
the adult fly had emerged. These facts show that the

'" ■' / '

ENZIBAZIBA LUSHES ij.N DAMBA IsLE FOR.MING BREEDIXt; (GROUND.
By permission of flie Royal Society and the Controller, H.M. Stationery Office.

SHINGLE AND SAND UNDER SHELTERING ENZIBAZIBA BUSHES WHERE FLIES
HAVE BEEN SEEN TO DEPOSIT LARVAE.

Foot-rule set upright shows scale.

To face p. 58.

BREEDING PLACES 59

green growth had only just reached the condition deemed
suitable by the fly as cover for the pupae. An opposite
condition was found on Bulago. A tree had large horizon-
tal branches near the surface of the sand arching over
which was a canopy of half dead creeper, throwing very
poor shade, but with some fresh green shoots springing
out so that in a very little while it would form a good
shelter. Forty-five pupae were found here, of which
twenty-seven were dead ; in three the adult fly was almost
ready to hatch, and the other fifteen were in the earliest
stage of development.

These facts show that the herbage had been at one
time dense enough to throw good shade, but having died,
many pupae beneath it had been killed by exposure to
the sun's heat. Now that fresh green shoots were growing
again, the fly considered the shelter good enough, and had
begun to use the spot again as a nursery.

At another time a boy searched for pupae at the base
of a papilionaceous tree, which having thin foliage throws
poor shade, and in this case had recently died. Here
were found 130 shells from which the adult fly had emerged,
and five pupae, all dead. Owing to the death of the tree,
and the poor condition of rambling plants around its
base which had at one time thrown good shade, the fly
no longer considered the spot a suitable one, and all the
pupae that had been deposited there when the shade was
good had either hatched out or died.

Lastly, among the requirements of the pupae is proximity
to water. The pupa seems to require this, although its
actual surroundings must be dry. Now, inasmuch as
the optimum conditions of soil and shade are provided
by the raised beaches of sand or gravel that have been
left by the subsidence of the lake from a former higher
level, these raised beaches being naturally near the water
it might be argued that the pupae were found here not
because the beaches were near the water, but because on

60 NATURAL HISTORY OF GLOSSINA PALPALIS

the beaches were found the optimum conditions. But a
recent observation, I think, disposes of that argument.
On Kimmi Isle in 1914 was a broad flat strip of low lying
marshy land, soft and waterlogged, that obviously owed
its origin to the level of the lake having fallen and the
water having receded from the real shore marked at
the edge of the forest by the usual raised beach, which
was thus separated from the water by a hundred yards
of marshy grass land.

This raised beach satisfied all the requirements of pupae
except proximity to the open water, but no pupae were
found in 1914. In December 1918, however, there having
been an exceptionally wet season in 1916, the water
had returned to a former height and had covered the
former marsh land, so that it came right up to the old
beach, which thus formed a typical " fly beach." It
was now used by the fly as a breeding ground, and pupae
were found there.

It may be said, then, that the ideal requirements of
the fly for its pupae are found on the beaches near the
water, well shaded, and composed of dry sand or gravel,
and that where these occur palpalis is found in maximum
numbers. But in their absence the fly makes use of such
scattered nooks as can be found, so that it can exist
in some number in the absence of " breeding grounds"
on rocky coasts. I consider this to be very important,
for it has been thought that it would be possible to
destroy the fly by merely cutting down and clearing the
low shade producing vegetation on the beaches so as to
spoil the breeding grounds. This would doubtless lessen
the numbers of the fly, perhaps by half, but would not
exterminate it.

In order to exterminate Sleeping Sickness two animals
must be kept from each other — the Situtunga antelope
from which the fly obtains the Trypanosome, and the
fly, which inoculates the Situtunga with the Trypanosome.

GLOSSINA OR SITUTUNGA ? 61

Each without the other is harmless, for the Trypanosome
cannot multiply indefinitely in the fly, and requires to
live for a cycle in mammalian blood.

Can the Situtunga be exterminated ? This would
be an extraordinarily difficult matter. It might be de-
stroyed, temporarily, on the islands, for there are very
few localities where it would be out of reach. But on
the mainland there are vast areas of papyrus swamps
in which it finds sure refuge and where it would be quite
beyond reach. Now it is well known that the antelope
swims readily, and I have given an example of its wander-
ings on to islands where in the old days it was kept down
by natives. Hence it would easily swim out again to
the islands from the mainland papyrus swamps, and from
one part of the mainland to the other. If, however, a
price were put upon its head and natives were allowed
to destroy it, it might conceivably be kept to the papyrus
swamps, where, beyond the reach of Glossina, which is
never found in such localities, it would be harmless.

As to the other side of the question, whether Glossina
can be destroyed, or at least so diminished in numbers
as to be harmless, there is a good deal to be said. It must
be remembered that when the fly is only present in small
numbers the chances of any one being infective are minute.
For every fly does not feed from a buck, and every buck
that is fed upon does not contain Trypanosomes. More-
over, as .Miss Robertson has shown, if the buck does
contain Trypanosomes they are not always in a condition
ready to multiply in the fly when it bites, and lastly,
every fly that takes in Trypanosomes is apparently not a
suitable medium for their further development.

When an undesirable insect is to be destroyed, it is
often found that the larval stage is easy to deal with, or
the next stage, when the insect is a quiet pupa or
chrysalis.

But since there is no free larval stage in Glossina,

62 NATURAL HISTORY OF GLOSSINA PALPALIS

the pupa is all that remains to us. Can we then
destroy it ?

This is theoretically easy since shade is a requisite. If
all the low bushes and creeper growing on dry sand near
the water were cut down, the " breeding grounds " would
be destroyed and the numbers of the fly might be reduced
by half.

This would leave, however, all the " loci " that have
been discussed, and I see nothing for it but to clear all
the shore of low bush for at least fifty yards back from
the water, and probably a great deal further. This would
be a colossal task with a lake of such a size. But a better
way occurred to me in 1914, and an observation in 1918
on Kimmi that has been already quoted made it seem
very likely to succeed.

The method is to provide for the fly all the conditions
that it requires in a concentrated form, in such a way
that it will deposit its pupae in the place chosen for it rather
than elsewhere. The pupae can then be easily collected
and destroyed. The conditions required by the fly
have been dealt with above, and they are provided by
building a low roof of thatch covered with green creeper
on a bank of sand or gravel near the water. In 1914,
I had commenced experimental work on these lines on
Tavu and Ngamba, which was cut short by the call of
active service, but I had satisfied myself that these
shelters were attractive to the fly. I also found that
although, as has been said before, a thick layer of dead
leaves on the surface is inimical to the pupae, yet the fly
prefers the surface to be lightly strewn with flakes of bark,
sticks or dead leaves. In the preliminary tests one half
of the sheltered area was strewn with bark or leaves and
the other half left bare, but there were always more
pupae found in the former half.

The accompanying figures show the nature of the
" artificial breeding ground " thus provided in the test

THE RAISED BEACH OF NGAMBA CLEARED OF VEGETATION UP TO THE

EDGE OF THE FOREST BEHIND.

Two boys are searching for pupae under the shelter.

By pennission of the Royal Society and the Controller, H.M. Stationery Office.

THE SHELTER ON THE RAISED BEACH OF TAVU, AFTER CLEARING.

The water lies beyond the reeds on the right hand.

By permission of the Royal Society and the Controller, H.M. Stationery Office.

To face p. 62

PROPOSED NEW METHOD OF DESTRUCTION 63

experiments on Tavu and Ngamba in 1914. When I
returned to the islands after the war this question was
much in my mind. It had seemed probable that in
order to prevent the fly making use of natural breeding
grounds and loci it would be necessary to clear them,
but a most enlightening observation on Kimmi in
December 1918 made it seem probable that clearing
would be unnecessary. On a sand bank or raised beach
the boys were set to find pupae, but found only a few
except at one spot. This was under a little pent-house
roof formed by roots of a tree that had been torn up by
the wind but had been prevented from falling by its
neighbours. The earth embraced by the roots formed a
roof at an angle of forty-five degrees protecting from the
sun and rain loose dry soil, and in an area a few feet square
one boy found pupae at the rate of 104-5 per boy-hour.
This proves that although there may be a large area of
well sheltered sand apparently suitable for pupae, the
fly will carefully search for the best spot and deposit
pupae there in numbers rather than scatter them along
the larger area.

I also found numerous other examples of the same thing :
moreover, an experiment has been cited earlier which
shows that the female fly does circulate round an island,
partly no doubt in search of food but also no doubt to
look for the best place for her young. Indeed, I feel sure
that the mother fly searches as carefully as does a parent
butterfly for the right plant on which to lay her eggs.
That being so, there is no need to cut down the bush,
*/ the artificial breeding place is made tempting enough.
The construction is very simple. All roots are pulled
up out of the loose dry soil. The roof covers an area
some twelve by four feet ; at its back it is just high enough
to allow a boy to creep under in search of pupae, in front
it just clears the ground. There are no walls, to allow
the free ventilation required. Creepers such as the

64 NATURAL HISTORY OF GLOSSINA PALPALIS

Ipomaea and others which grow abundantly in these
localities are trained so as to make a thick mass of young
greenery on the roof, and the soil is lightly strewn with
bark, dead sticks or leaves. At regular intervals of a
week or ten days these shelters can be visited and all
pupae easily collected by boys searching in the sand.

Should the method prove feasible as a means of destruc-
tion on a large scale, it might be possible to put sieves
or perforated trays in the sand under the shelter of a
kind that would allow the pupae to be quickly sifted
out. Or another method might be tried of destroying
the pupae by the sun's heat. The roof could be made in
two halves, and at regular periods one half could be lifted
off so that the pupae beneath would be killed by prolonged
exposure to the sun's heat, while next time that half
would remain covered and the other half be taken off.
The rainy season, however, would introduce complica-
tions into this method, for often there are days with very
little sun, and if the sand was allowed to get wet, the
flies would no longer deposit pupae therein. The method
is a simple and inexpensive one, and should at least
greatly reduce the numbers of the fly. There is no doubt
whatever that these shelters are very attractive, and
approved of by the fly as suitable for their young. Be-
fore I came away from the islands in 1919 for leave, a
number of artificial breeding grounds were prepared
on Bulago and Kimmi and are now being tested. The
pupae are collected weekly from each, and also from an
ideal natural breeding ground on each island for com-
parison, that on Kimmi being the upturned tree roots as
described. From figures that have reached me so far,
it is clear that my artificial shelters are just as attractive
as, or sometimes more than, the natural shelters, for on
some occasions not so many pupae are found under the
tree as under one of the shelters close by. -During July
over two thousand pupae were destroyed on Kimmi by
this method, Kimmi being barely a mile in diameter.

PROPOSED NEW METHOD OF DESTRUCTION 65

Seeing that the fly is such an abnormally slow breeder,
and that this method destroys generations yet to come,
it does seem possible that it will result in very substantial
reduction in the number of flies. Of course a few pupae
will continue to be placed in other spots, scattered " loci,"
but with the reduction in number of flies the number
of these would be reduced in far greater proportion,
until when there are only very few flies about it seems
reasonable to suppose that the number of pupae not
deposited in the shelters would be so small that the fly
would be exterminated altogether. This would probably
take some years, but the decline would be a steady
one.

The method could be made applicable to Olossina
morsitans also, only in this case the shelters would need
to be placed not near the water, but near some well
marked game track ; for morsitans seems to place its
pupae naturally near game tracks.

This chapter on the Tse-tse fly may fittingly be con-
cluded with the following verses :

TO GLOSSINA : A CURSE !

Thou dipterous hexapod. (by which I mean

Thou six-legged creature borne on pinions twain),

Thou tracheate arthropod, — in short, thou FLY,

With buzz importunate arousing my

Just wrath ; may fiends innitmerable rend

Thy quivering form in twain, and, ruthless, send

Its disunited fragments to the place

Where bad flies go, whate'er their name or race !

May my sanguineous fluid, y'clept blood,

Which thou desirest to extract for food,

Plunging thy sharp proboscis through my skin

To reach capillaries enshi-ined within,

Flow freely till thou bvirstest !

Or, again,
Lest this should cause thee insufficient pain,
May justice serve to thee another fate
That red corpiiscles ne'er corpusculate
In thy capacious proventriculus,
So that, attenuate, ridiculous,
6

(56 NATURAL HISTORY OF GLOSSINA PALPALIS

Thou'rt gnawed to death by cruel starvation's fangs !

Yet do I wish for thee severer pangs

Since none of the above are adequate

Or quite sufficient to express my hate !

May Bembex wasps (those skilful hunters) thrust

Into thy side the lancets which, I trust.

Will fail to paralyse thy sense of pain

While rendering thee inert, that sod, in vain

Thou strivest to escape from himgry jaws

Of Bembex' child, devouring without pause

Thy non-essential tissues ; till, at length

When it shall have attained its perfect strength,

Thy shrinking vitals are devoured by it.

And there is left of thee no little bit

Whereof a man might say, " This was a FLY,

A noble creature that knew how to die " !

So that there shall not e'en be left of thee

So vain a thing as a fond memory !

Thus would I wreak my vengeance on thee, pest,
That oft hath bitten me with fiendish zest !

Addendum.

The first observations on the use of artificial breeding
places were made in 1914, but were interrupted by the
war.

During the war my friend, Dr. W. A. Lamborn, in
Nyasaland, unknown to me, was experimenting on his
own lines by cutting down trees to make artificial breeding
places for O. morsitans. (See Bulletin of Entomological
Research, May, 1916, VII, p. 38.) This method had been
developed by him independently, for my report, delayed
in publication by the war, was not published untU 1919.
(See Reports of the Sleeping Sickness Commission of the
Royal Society, 1919, No. XVII, pp. 67-71.) I did not
hear< of Dr. Lamborn's work until after I had been
able to develop and make first use of my scheme early
in 1919.

CHAPTER IV

THE LAKE

Some facts as to the great Lake Victoria will probably
be found of interest before reading a general account of
the scenery and life thereon.

The following data are taken from the eleventh edition
of the Encyclopcedia Britannica :

Lake was discovei^ed by J. H. Speke on August 3, 1858 ; the Ripon falls,
the only outlet, discovered by him on July 28, 1862.

Area over 26,000 square miles, only exceeded by Lake Superior.

Greatest length from north to south, 250 miles.

Greatest breadth, 200 miles.

Coast line of over 2,000 miles.

Height above sea level, about 3,720 feet.

Greatest known depth, 270 feet.

Largest islands —

Ukerewe, near Muanza on south coast, whose greatest diameters

are 25 X 12 miles.
Buvuma, in the Napoleon Gulf on north coast, which has an area
of 160 square miles. (The Isle of Wight has 147 square miles.)

The Sesse Archipelago contains 62 islands, of which 42 were inhabited.

Area drained by the lake, including lake itself, 92,240 square miles.

Rainfall over whole area averages 50 inches yearly.

Annual rise and fall from 1-3 feet : maximum height in July.

By far the greater part of the water that enters is lost by evaporation.

The largest affluent is the River Kagera, which enters
on the west coast, and was crossed by Speke on January
16, 1862, on his way to the discovery of the Ripon falls.

He says of it : " Once over I looked down on the noble
stream with considerable pride. About eighty yards
broad, it was sunk down a considerable depth below the
surface of the land, like a huge canal, and is so deep, it
could not be poled by the canoe-men ; while it runs at a
velocity of from three to four knots an hour."

This fine river, reckoned as 430 miles long, derives its
waters ultimately from the south-eastern slopes of the

67

68 THE LAKE

Mfumbiro range north of Lake Kivu, and from the country
between Lake Kivu and the north end of Lake Tanganyika.

It is formed by the union of several other rivers, and
at first flows northwards and slightly westwards and then
turns abruptly eastwards, to flow eventually into the
lake at about the middle of the western shore, a little
north of Bukoba. It is of considerable geographical
interest that whereas the main affluent of the Victoria
Nyanja, and thus the chief ultimate source of the Nile,
arises from the south-eastern slopes of the Mfumbiro
Mountains, the other side of the same mountains gives
origin to streams of M^hich the most important is the
Rutchuru, flowing into Lake Albert Edward.

From this lake the Semliki river flows into Lake Albert,
adding its waters to those brought from Lake Victoria
by the first part of the Nile, via Lake Kioga.

So a very large part of the waters of the Nile above
Khartum have passed through Lake Albert.

The numerous islands among which I worked vary
much in size, from Bugalla and Kome down to mere rocks
or shoals of sand. The commonest type is composed
of centrpJ grassy uplands rising a few hundred feet, with
a belt of forest along the water of breadth varying accord-
ing to the lie of the land. But some islands have practi-
cally no forest growth ; thus, Kiuwa is conspicuous as
being merely a conical, grass clad hill, with a thin fringe
of bushes near the water. On the other hand, Damba
is quite flat and almost entirely covered with dense
forest. Wema has one peculiarity ; one of its four
hills is crowned with forest, the base and the surrounding
terrain being grass clad. On the largest isles there is
much park-like country ; that is to say, the grass land is
dotted with fine trees, or clumps of small trees and bush
forming copses, and the grass being thin and only knee
high, this is delightful country for walking. The shores
of the islands vary much in appearance, but are rarely

THE KAGERA RIVER AT KIFUMBIRO (KYAKA FORT), ABOUT
FORTY MILES FROM THE LAKE.

THE HEAD OF NAPOLEON GULF WHERE THE NILE ARISES FROM THE LAKE.
To the left of the tree a faint haze indicates the site of the Ripon falls.

To face p. 68 (first page).

THE EASTERN CATARACT OF THE RIPON FALLS IN I9IO.

THE BIRTH OF THE NILE.

Ripon Falls from below.

To face p. 68 (second page)-

SCENERY 69

anything but beautiful. The most picturesque scenery
is the narrow channel between two islands where the
blue water is bounded on each side by dense vegetation ;
for example, the lovely passage between Damba and Kome,
only a few hundred yards wide.

Forest lies on either side, and in front a narrow belt
of bright green papyrus, growing at the water's edge.
A belt of reeds springs from the slightly deeper water,
and in front of this again the still blue surface is carpeted
with leaves of water-lilies, whose blue or pink flowers
stand up a little out of the water.

An open shore, with beach of finest white sand or
pebbles, is not uncommon, with a forest belt terminating
abruptly some yards away from the water. Here there
is always evidence of the subsidence of the lake level
in the shape of raised beaches at the edge of the forest.
Such a strip of shore is termed a " fly beach," since it
provides ideal conditions of food and shelter for the
Tse-tse.

In some places the shore is formed of bare rock of a
spongy texture and red brown in colour, shelving gently
into the water, or sometimes forming small cliffs. Here
and there a stream of lava makes a strip of lighter colour,
embedded in which are fragments of quite different
rock, like plums in a duff.

A small headland on which was pitched my camp on
Damba ended in a face of grey rock perhaps 20 feet
high, in which was a crevice filled with water- worn
pebbles at least 12 feet above the present highest level
of the lake, shovv^ing how the v/ater had receded.

On a flat rocky coastline the forest often comes quite
to the water's edge, the trees actually overhanging the
water, and there are usually abundant ferns. Long
stretches of very low lying marshy shore are common,
and are more or less under water during the high level
of the lake in the heavy rains. Lush grass and reeds

70 THE LAKE

make these spots very attractive to hippos, and the
soft soil is churned up along their regular tracks. . A
very curious plant flourishes here, for all the world like
a small lettuce floating on top of the water. It propagates
very rapidly, probably by budding, so that many square
yards of water may be covered by it, and patches often
are driven out to sea by a land breeze. The curious
ambatch tree, with its short thick trunk of extraordinarily
light, pithy wood, scanty leaves and yellow papilionace-
ous flowers, flourishes on marshy shores, and the graceful,
tall, smooth " Makindu " (wild date) palm is plentiful.
Only one other species of palm is met with on the islands,
a species called " Ekibo " by the natives, having a short
stem and very large dark green leaves, whose mid-ribs
are often used as rafters for houses, or, when yor ig,
furnish the fibre used in sewing the planks of the canoes.
On low lying sandy or muddy shores a species of cane
growing to a height of 10 feet sometimes forms thickets
very unpleasant for one to penetrate owing to the sharp
points of the narrow, hard leaves. Several bushes are
very characteristic of the lake shore ; one ^ is known to
the Baganda as " Oluzibaziba," and forms dense thickets ;
it has leaves very much like those of a poplar in sha^De
and setting, which are the favourite food of the Situtunga
antelope. Another, called " Kinsambwe," has harsh,
hairy leaves, with rather a rank odour ; and a third, ^
with leaves much like ambatch, was called by the natives
" Ornuvuvumye." Several flowering plants sometimes
made patches of colour ; one like a small herbaceous
sunflower grows in banks of bright yellow ; another
{Ipomaea) rambles over bushes and spreads wide its purple
convolvulus-like flowers, while a third, called " Anyam-
berege " by the natives, sends up tall shoots on which grow
flowers much reminding one of single hollyhocks ; bright
yellow, with a velvety patch of purple brown at the

1 Euphorbiaceae, Alchornea chordata. * Papilionaceae.

YOUNG RAPHIA PALM (" KIBO ") ON THE RIDGE OF KIRUGU ISLE.

""^ 'ittk

WILD DATE PALM ON DAMBA ISLE.

To face p. 70 (first page).

A TANGLED MASS OF IPOMAEA IN FLOWER.

# .i-^m ri^!^

A BED OF BRACKEN ON BUGALLA IN FRONT OF THE FOREST BELT.

To face p. 70 (second page).

VEGETATION 71

base of each petal. One other characteristic plant may
here be noted, known to the natives as " E'tungulu " ^
It grows in dense patches, throwing up straight green
stems as much as 8 feet high, furnished with alternating
leaves of very thin texture, with a smooth shiny surface,
bright green in colour, elongated, and with pointed tip.
The flowers are borne on short stalks coming up separately
from the leaf stems ; they are in shape somewhat like
that of a pentstemon, in colour whitish or dull pink
according to the species. The fruits are pear shaped,
attached by their broad bases in a cluster, and are of a
very bright shiny red. If the tough skin is torn open
the interior is found to consist of a large number of
round black seeds embedded in a white pulp with delicious
acid taste ; hence these fruits are much esteemed by
the natives, and by monkeys. The fruit of the white
flowered species is poor compared to that from the red
flowers.

To return to the forest belt. A characteristic feature
of its landward side is the abrupt margin ; one suddenly
emerges from forest to open grass land as though the edge
had been cut by a landscape gardener.

It is possible that this edge owes its origin to former
systematic burning of the grass by natives, so that no
outlying, straggling fringes survived. Now that the
population is no longer there, however, the forest edge
seems in places to be gradually advancing over the grass,
the vanguard formed by rambling bushes of low growth.
The significance of this had not occurred to me until
Fiske pointed it out in 1914.

On the smaller islets, if very rocky, the only trees are
figs, which seem especially to thrive close to the water,
and may sometimes be seen perched on a rock, which their
descending roots embrace on all sides. These doubtless
owe their origin to seeds dropped by some pigeon.

1 Anonaceae.

72 THE LAKE

Owing to the nicist climate the foliage is always green,
and at no time are the trees as a whole bare of leaves :
it is quite commxon to see one tree shedding all its leaves
while another near by is in the full glory of new foliage.

Flowers on the islands are disappointing ; they do not
enter into the landscape, but there are many creepers
which ramble over the tops of the forest trees and are
covered with sweet scented white or yellow flowers :
many trees and shrubs also have fragrant blossoms,
som.e of which are beautiful, often of waxy appearance.

The colouration in the bright sunlight during one of
the clear days characteristic of the heavy rains is really
wonderful in its brilliancy. From high ground one
looks over the top of vividly green forest towards distant
purple islands set in a sparkling deep blue lake, which
is stirred into white-capped waves by the prevailing
south-east breeze. So clear is the atmosphere at this
time, especially in the evenings, that from Bugalla Island
some of the individual houses at Entebbe on the main-
land, twenty-five miles away, could be distinguished with
the naked eye, and the flagstaff at Government House
could be seen . v/ith glasses, though I could never make
out the flag itself.

The clim^ate of the islands is more equable than at
Entebbe, the water exercising its usual influence.
Consequently, although the average mean temperature
is degrees higher than at Entebbe, the maximum is
lower, and the nights warmicr. The rainfall also is more
equably distributed than on the mainland, although it
is greater. There was no month without some rain during
the twelve of which I kept meteorological records on
Bugalla, but the greatest fall came at the same two
periods as on the mainland. (See Chart I.)

A striking feature is that, with extrem^ely rare excep-
tions, the rain falls between midnight and noon, often
between midnight and daybreak, so that afternoon rain

ABRUPT MARGIN OF FOREST BF.LT ON BUGALLA.
A flowering bush of Haronga at the angle was much frequented by butterflies.

PARK LAND ON BUGALLA ISLE.

To face p. 72,

A THUNDERSTORM 73

which occurs at Kampala on the mainland, a few miles
from the lake shore, was unknown during my time on the
islands. Another characteristic is that the rain is nearly
always accompanied by thunder ; but though this is so
frequent one never experiences the oppressive heaviness
of atmosphere which precedes a thunderstorm in England.
The following is a description of a typical, or as I used
to call it a formal, thunderstorm, such as occurs with
varying intensity every few days from mid-March to
mid-June during the period of heaviest rains. The day
has been brilliantly fine, with a hard blue horizon and
vivid colouring, and an appearance, of nearness of distant
objects. There has been a steady breeze all day long,
dying away towards sunset, which has perhaps shown
for a brief spell a glory of flame -tinted cloudlets. After
dark not a cloud is to be seen anywhere, and the stars
shine brilliantly. Before turning in at about 8*30 one
takes a look round, but there is nothing to note except
a distant flickering of lightning to the south-east, which
is practically a nightly occurrence in the rainy season.

Having seen that tent pegs and ropes are secure, and
that the front of the tent can be quickly closed in an
emergency, for these storms often give one little time,
one turns in, to be not at all surprised at being awakened
in the small hours, usually between two and three,
by a rumbling of thunder. Stepping outside the tent
one sees magnificent piled up masses of inky cumuli
illuminated by almost incessant flickerings of lightning.
Sometimes a flash stretching in an arc right across the
heavens makes one start with the wonder and, almost,
fear of it.

In the momentary intervals between the rumblings of
thunder can be heard a faint musical sound, produced
by the downpour of rain on the water, as yet far away.
There is nothing to be done but to maka sure that the
tent is firmly secured, fasten up the door, enduring an

74 THE LAKE

" infinite torment " of mosquitoes which, aware of the
coming storm, have thronged into the tent for shelter,
and retire to bed again to await the onward march of the
storm, for no more sleep is possible until it has passed.
Overhead the stars are still shining brilliantly, and it is
absolutely calm. Presently a faint murmuring of wavelets
is heard ; heralds of the disturbance produced by the
distant storm, by the time they reach the island reduced
to mere ripples. Now a faint breath of wind can be felt,
which dies away, followed by another puff, and the sound
of the rain and thunder grows louder, while the light-
ning increases in intensity until it almost hurts one's
eyes, kept tightly shut. The ripples increase into waves,
and the noise of surf on the beach is added to the distant
sounds. A few fat drops of rain descend, and then with
a rush and a roar the storm arrives. A hurricane of
wind tears and shakes the tent until one expects every
minute to be carried away with it ; the roar of the sheets
of rain on the roof is almost loud enough to drown the
deafening thunder produced by flashes of lightning that
seem to split one's very head in twain. This pandemonium
lasts for about a quarter of an hour, when the wind moder-
ates a little, the rain increases in intensity, and the
thunder gradually passes away as the centre of the storm
moves off. After an hour or so the rain slackens, but
continues very heavily until shortly before daybreak,
though sometimes continuing steadily until the morning
is well advanced ; the evening being again fine and
clear. In one such storm two and a half inches of
rain fell between 6.30 and 9 a.m. Such is the formal
storm of the wet season, but quite frequently there
are dry thunderstorms, when the lightning apparently
discharges from one cloud to another, and there is
no rain. The thunder on these occasions has a hollow,
light quality, probably due to the rarified air of the
upper regions in which the storm occurs. The clouds

CLOUDS 75

are greatly disturbed, and one seems to be looking upwards
into a vortex. There is usually a central area to which
ragged pieces of cloud drift from all directions, only to
be torn into shreds and vanish as if they had gone up a
chimney. These overhead disturbances may be accom-
panied by no disturbance whatever at earth level, where
the sun shines and a fine weather breeze blows as usual.
Sometimes the edges of the clouds at the centre of these
storms may be wonderfully rainbow tinted, although no
rain reaches earth. On Damba Island in June 1911 was
seen one evening a wonderful cloudscape to the East.
Low down, over the mainland, were the usual lumpy
cumuli, and to the right these merged into nimbus.
Converging to a point on the horizon between these were
two broad filmy streaks, above and transverse to which
were bars of cirro-stratus. This cloud complex was the
forerunner of the highest wind I had experienced on the
islands up to that time ; it began at midnight and re-
peatedly awakened me : by mid-day the blue lake was
whitened with curling wave crests, and drifting purple
and green cloud shadows made lovely contrasts of colour.
Since I was very dependent on the state of the weather
for my work outdoors, I took some interest in the signs,
and an account of what was learnt may be of interest.

Firstly, the meteorological signs.

There were the usual general signs, such as cirrus or
cirro-stratus clouds and clearness of atmosphere, heralding
storms ; heat haze in fine weather, etc.

The approach of the rainy season is preceded by cloud
funnels, which appear in all stages from a nipple shaped
process at the lower edge of a ragged cloud, to a complete
pillar connecting cloud and lake, called by the natives
" Omusoke." From Entebbe, looking out over the
lake, a great number of these are seen, and they are
commonly called waterspouts, but I have never been
near enough to make out whether they correspond to
waterspouts at sea.

76 . THE LAKE

There is a great deal to be learnt from the wind, its
direction and force, and so on. The daily wind blows from
the south-east, being strongest in early afternoon and
dying away towards sunset. Sometimes it blew so strongly
when I was on Damba that canoes could not go out on
account of the white capped waves and the surf on the
shore. If this wind freshens again after sunset for a
little while, it usually implies bad weather next day,
especially if, during the night, a few fat drops of rain fall.
Such a night followed by a very calm, clear, warm, clouded
morning at daybreak invariably portends a storm before
mid-day, and one would not think of going out in a canoe.

On the contrary, an early morning cold northerly breeze
from the mainland, with thick haze, foretells a very fine
hot day. This north breeze blows quite violently until
8 or 9 a.m., when it suddenly drops ; and, as suddenly,
the south-east wind rises and impresses its mark on the
lake before the waves due to the north wind have subsided ;
so that one sees ripples travelling in a northerly direction
superimposed on the swell running southwards. During
the rainy season, when there is a calm, curious map-
like areas of irregular shape and varied tints appear on
the still water, and I came to the conclusion that they
preceded a storm.

Occasionally during this time of year the sun appears
surrounded by a misty ring at a considerable distance
from it ; but this was not common, and I never was able
to m.ake out that it heralded any particular type of weather.

In the driest months at the end of one year and com-
mencement of the next the sun's heat is powerful enough
to cause grass to take light spontaneously. Fires seen
on the islands are usually considered to be evidence of
the presence of natives in the forbidden area, but I am
quite satisfied, from the following instance, that they
can be of " spontaneous " origin.

On January 29, 1919, I was examining Nyenda Island,

SPONTANEOUS GRASS FIRE 77

my camp being on the southern side of Wema. As we
returned in the afternoon the grass all round the camp
was seen to be burning, and I found when I got back
to camp that it had had a narrow escape. The boys
said that about 2 p.m., while in their houses, they had
heard a sound like a gun-shot, and going out, saw a
cloud of smoke arising to windward, and heard flames
crackling. They only just had time enough to strike the
tent and get my things further away from the edge of the
long grass, and had there not been a space cleared for
the camp, the huts would all have been burnt.

They showed me afterwards the spot where the fire
had commenced. Among much bare red rock there was
thin dry grass on light soil, and at one place was a hole
looking as if the earth had fallen in over a termite's
burrow. Here was an appearance that suggested that a
quantity of dry debris lying at the bottom of the open pit
had suddenly taken fire with explosive violence, for the
edges of the hole had obviously just been disturbed.

Such fires are well known to the natives, who call one
" Nakibengeyi " ; they pointed out the red rock to me
as a kind that is particularly associated with these
spontaneous fires. It is of a spongy texture, apparently
of igneous origin, and certainly does absorb and radiate
a great amount of heat from the sun.

The signs of weather afforded by the island life are
abundant, and doubtless many more than here recorded.
Several birds give useful indications. A curious booming,
hollow sound is sometimes heard during the rainy season
preceding a storm, and it was a long while before I finally
satisfied myself that it emanated from the crowned
crane. The natives insisted that it was produced by the
puff adder !

The large black and white hornbill is always much
more vociferous on one of the clear blue evenings before
rain, and gulls on the rocks scream and chatter much

78 THE LAKE

more vigorously. When terns appear off the coasts of
islands, one knows that windy weather is coming, for
under other conditions they were never seen ; presumably
they then frequent the open lake.

The glorious fish eagle is much more vociferous, accord-
ing to the natives, when the wind is rough, but I never
satisfied myself of this, and think he screams more on
account of brilliantly fine weather, which, of course, is
often accompanied by a fresh breeze. One of the large,
pale blue, insectivorous kingfishers, which haunts thick
bush or forest, utters his characteristic cheerful cry with
greater persistence when the early morning is particularly
fine and brilliant. A species of woodpecker, like the
" yaffle " at home, is only heard to cry before rain.

Frogs, of course, croak before rain, especially one
species which inhabits burrows in the earth ; indeed,
it is never heard to croak except in the evening, before
rain, unless it is breeding. The green tree-frog, so
abundant on herbage of the lake shore, sends forth its
high-pitched, scraping note when it thinks there is
rain about.

I had an amusing experience one day on the shore.
It was about noon on a very fine day ; a small cumulus
cloud slowly formed and emitted a single crack of thunder
without in any way dimming the brightness of the day.
This clap of thunder was immediately followed by a chorus
of satisfied quackings from a host of concealed frogs,
sounding ridiculously Hke applause of a well-meant effort !
There was no rain, however, so the prophets were
forsworn.

Fish also afford signs for a weather prophet to interpret.
Thus, in my journal for January 30, 1914, is noted :
" Off the south-west point of Ngamba Island little fish
were leaping about in the water and also in shore, close
under the bushes. About a hundred yards away from the
island the glassy surface of the water was broken by a

AERIAL ROOTS DIPPING TO GROUND, BUGALLA FOREST.

GIANT TWISTED STEM OF CREEPER, BUGALLA FOREST.

To face p. 78.

WEATHER PROPHETS 79

collection of the ' male ' fish (see Chapter IX, p. 193)
tumbling and wriggling about. The natives said they
were mating." The next day was noted as " a beastly
cold, stormy day, with black clouds, after a lot of thunder
and rain at 1 a.m." On Feburary 21st the same occur-
rence was seen off Wema Island ; the small fish hopping
out of the water, and male tumbling about. The
following da,j, however, was also fine and hot, and rain
did not come until the twenty-third.

I think this behaviour of the fish is not of any immediate
significance, but occurs just before the rains commence
in earnest, in March.

It was again noticed very often when voyaging in
January and February 1919 ; sometimes the canoe
would be so close that the men could almost have netted
the fish.

Insects, as might be expected, furnish many signs of
coming rain.

The most obvious is the appearance of immense numbers
of a non-biting gnat, of the family Chironomidae, called
by the natives " E'sami " (pronounced Sammee). They
appear as brown clouds drifting over the lake, along the
horizon, like smoke from a distant steamer. Eventually
they are blown ashore, where they seek shelter among
trees. These clouds are most often seen just before
and during the heavy rains. If Sami happen to arrive
at one's camp after dark they swarm around the lamp and
drive one nearly desperate. Other insects also come to
the lamp before rain, particularly Ephemeridae of different
sizes. One, about the size of the common English " may-
fly," once came to the camp fire in such numbers, and
their burnt bodies made such an appalling stench, that
it was too much even for the natives, who put out the
fire and retreated ! A smaller, paler species has the
extremely annoying habit of alighting upside down on
one's table or writing paper, and flopping miserably about

80 THE LAKE

until he dies ! Other still smaller species of two-winged
flies have an extraordinary liking for singing their high-
pitched song inside one's ears, until one can bear it no
longer and retires beneath the mosquito net.

To these nuisances may be added an enormous male
ant (Dorylus) with shiny, hard, light brown body over an
inch long, who bangs himself against the lamp, and returns
however often he be thrown outside with contumely,
until one is forced to go against one's scruples, and bottle
him. I have dispatched as many as twenty on one evening.
They are a very reliable sign of coming rain if they are
seen in any number : the workers also are much more
often seen on the march before rainy weather. Termites,
likewise, are much more harmful in wet weather, and
cover the timbers of one's house with earth with sur-
prising rapidity at such times.

The biting flies known as " buffalo gnats " (Simulium)
are especially eager to bite in the early morning or evening
before rain on the islands. Thus on- July 26, 1914,
on Kome Island, is noted in my journal : " I went down
to the rocky shore before breakfast, and was set upon by
a swarm of viciously biting Simulium. In the evening
many Ephemeridae came to the lamp." Next morning,
after daybreak, there was the worst thunderstorm I
had ever experienced on Kome.

The night time on the islands, apart from thunder-
storms, has charms of its own. A number of sounds
are to be heard from sunset until the small hours, when
for a while there is silence until the first glimmerings of
dawn. The thunderous snortings of hippos, the muffled
bark of the Situtunga, break in upon the continuous shrill
tinkling sound, curiously suggesting sleigh bells, produced
by thousands of small frogs along the shore. Crickets
chirp all round and in the house, and during the rains one
enormous species, sitting just inside the mouth of its
burrow, makes the earth resound with a continuous high-

BIRD SOUNDS 81

pitched buzzing sound, which if at all near seems to
penetrate through and through one's head, and renders
sleep impossible until the periormer has been silenced.

The chattering and quarrelling of gulls on the rocks
mingles with the plashing of wavelets and the gaggling
of Egyptian geese, which are restless at night.

Perhaps there passes overhead a flight of some kind of
plover, whose cry has earned for them the native name
" Empunya," and in the distance, over the open grass
land, a hunting owl emits its drawn out quavering call,
and a nightjar its " Tok-tok-tok-tok-tok " for long periods
without a pause.

Occasionally an ibis, startled by something, disturbs
the night with its raucous alarm cry, " Aa-aa-aa," or an
awakened monkey in the forest says what it thinks of
the disturber before settling down to sleep again.

With the first glimmerings of day the chorus of birds
begins, and the forest resounds with their cries and music,
for there are many songsters of merit. Soon the troops
of monkeys awake and commence their hunting, and the
loud " kubba-kubba " cry shows where they travel among
the tree-tops ; and for the rest of the da^y the forest is
never silent. A species of forest francolin, living in the
densest thickets, has a loud ringing, laughing cry which,
taken up by one after the other, often resounds through
the forest during the heat of the day when other birds
are relatively quiet. Towards evening the hornbill
waxes vociferous, and his " Nga-nga " cry, from which
he gets his name, seems characteristic of that time ;
flocks of grey parrots flpng home from the forest shriek
in the familiar discordant way, and snow white egrets
sedately wend their way with hoarse croakings across
the blue water to their accustomed roosting places. The
sunset chorus of songsters has hardly subsided before
the frogs and crickets take up the strain, and with the,
rising of the moon owls add their melodious cry.

7

82 THE LAKE

Life on the islands was soon reduced to a routine only
interfered with by the weather. Arising shortly before
the sun one pottered about for half an hour before a
bath, or a swim, and breakfast.

At eight o'clock one set off for the fly ground. On
Bugalla this simply meant a walk of about half a mile
down from the open grass clad hill on which camp was
placed, through the forest belt, and out on to the sandy
beach.

When on Damba we cut across a shallow bay by canoe
to the beach about a mile away ; from Kome camp we
went by canoe to the other islets, and very pleasant going
it was against the fresh north breeze and dancing waves,
taking an hour or more. From this one returned at any
hour between 2-5"30 p.m., after which there was a certain
amount of work to be done with the microscope.

On Bugalla the afternoon was always spent over the
microscope ; and a walk over the highlands in the evening
gave one exercise and beautiful views.

Sundays were always given up to butterfly hunting
and other changes of occupation. On Saturdays the big
canoe with crew of fourteen men started off for Entebbe
taking mails, returning on Tuesdays or Wednesdays
according to weather, with mails and the week's supply
of food for those of us who always remained in camp.
Sometimes this canoe had very rough passages, and I
well remember one night near the end of my residence
on Bugalla. It was one of the exceptional occasions when
there was a storm after sunset coming from the north-
east. The crew as usual had announced their approach
by blowing on a horn of the Situtunga antelope, and I
was rather anxious, as a storm was threatening from the
same direction.

GradualljT' the wind freshened until a gale was blowing,
and the house rocked. I lit a lantern and went down
to the shore to mark the landing place, as it was pitch

AN UNUSUAL STORM 83

dark. Above the roar of the surf on the coast could be
heard the shouts of the paddlers furiously urging on the
canoe in front of the storm, but they were not able to make
the shore where I was on account of breakers. By great
good fortune they were enabled safely to round a headland
and land in comparative calm. Had the storm been from
the usual south-east direction, they could not have faced
it, but would have had to turn and run before it ; but
these early night storms came from a direction different
from that of the formal storms described previously.

CHAPTER V

CANOES AND A VOYAGE

Unless one is very fond of the water and a good sailor,
travelling by canoe becomes tedious or even unpleasant.
If merely a passenger one sits right forward in the bows
and, when the lake is anything else than smooth, very
soon gets well splashed. I always took the steering
paddle and found it great fun ; a good deal of skill is
required to keep a heavy canoe true to its course when a
strong beam wind persistently blows the bows round.

The canoe (" Eryato ") made by the Basesse is of a
peculiarly interesting type, since in the jfirst place it forms
a link between a " dug-out " and a built vessel, and in
the second place shows how the keel of a modern boat
represents all that is left of an original dug-out.^ In the
old days of warfare on the lake between the Basesse of
the islands and the Baganda of the mainland, these canoes
were of very large size, and I have heard it said that they
could accommodate fifty paddlers.

The most important part is the keel (" Omugongo "),
which is hewn from a tree of a particular species, found
by experience to provide the best wood for the purpose.
It is very thick and solid, with a rounded bottom and
hoUowed-out upper surface, so that in transverse section
it is concavo-convex ; it is, in other words, a "dug-
out." Seen from above, it is broader in the middle than
at the ends.

1 Mr. Henry Balfour pointed this out when I gave him a model for
the Pitt Rivers collection at Oxford.

84

STRUCTURE OF A CANOE 85

The after end is 'shaped so that the stern post
(" Ekiwenda ") can be fitted to it, and, similarly, a stem
post is fitted near to the forward end of the hollowed-out
keel, of which the extreme anterior pointed end has been
left solid, to project beyond the stem post.

These stem and stern posts are inclined forward and
aft at half a right angle, and are fastened at the bottom
to the keel, whose hollow is shaped to embrace them
slightly by palm fibre passed through holes (" Endagire ")
bored with a hot iron. The next step is the fitting of the
lower side planks ("Amabasi"), one on each side, stretching
from stem post to stern post, and secured to them and the
keel by palm fibre (" Ensinga "). These planks are made
from wood different from that chosen for the keel ; the
latter has to be rigid, but a measure of pliability is required
for the planks. The upper edge of the lower plank is
neatly bevelled off to fit the lower edge of the upper plank
(" Olwero "). This, again of a different wood, for it has
to stand exposure to hot sun in between occasional wetting
by waves without cracking, is made of three pieces, for
it has to take a considerable curve. The main piece runs
along the greater length of the canoe, but a small length
(" Eryungo ") is required at bow and stern to take the
curvature from the end of the main piece to stem or stern
post. Where these pieces overlap they are neatly fined
down so that the thickness of the two is no greater than
the thickness of a single piece. At the extreme ends,
bow and stern, the pieces of opposite sides are firmly
pulled together by stout lashings passing across and across
the interior of the canoe. The lines of junction
(" Entabiro ") between upper and lower side planks,
and lower planks and keel, are covered by laths of wood
cut by splitting sticks longitudinally ; a strip (" Olu-
wamba ") is laid with flat surface next the junction,
and lashings of " Ensinga " pass through the planks on
each side of it, and over its rounded surface.

86 CANOES AND A VOYAGE

These strips cover the lines of junction on inner and
outer surfaces of the canoe, and all crevices are thoroughly
plugged with fibrous material.

The thwarts (" Amanga ") on which the paddlers sit
are fastened in a peculiarly ingenious manner, so that they
also serve to brace the sides.

Each rests on the upper margin of the lower side plank,
which is notched to receive it, as is also the lower margin
of the plank above ; the thwart itself is grooved, not
quite at the extremity, so that its actual end is visible
as a knob (" Empumi ") on the outside of the canoe.
The grooves in the thwart and notches in the planks all
correspond, so that everything fits as closely as possible,
and leakage is prevented by plugging. All the holes
through which pass the fibres fastening the planks are
tightly plugged, and as the plugging swells in the water
a good canoe leaks very little, although no resinous or
tarry material of any kind is used.

A peculiarity of these canoes is that right forward a
long spike projects out from each side, made by a thwart
(" E'gami ") fitted exactly as are the paddlers' seats,
with which it is serially homologous ; but its outer ends
are produced into spikes instead of being rounded off
into mere knobs. This projection is of great use when a
heavy canoe has to be hauled up on the shore, but I do
not know if that is why it was designed in the first place.

Another feature is the upturned prow (" Ensanda ").
It was said that the part of the keel left projecting beyond
the stem post is not hollowed out, and gradually tapers
off. Over this is fitted one arm of a large right-angled
piece of wood, whose other arm therefore projects
upwards into the air for two or three feet, and several
feet forward of the real bow of the canoe, and makes an
excellent breakwater against waves. The projecting keel
gives an extra lift to the bow of the canoe, so that they
are excellent sea boats, though the prominent prow and

r'''»«ku: , 'I^IUiyMillli^

A CANOE RACE.

A HALT ON THE EAST COAST OF FUGALLA.

To face p. 8e.

FITTINGS OF A CANOE 87

spikes manage to distribute a good deal of water over
any one in the bows, unless there is a calm.

On top of the " Ensanda " is placed a decorative tuft
of feathers, known as " Enkuli," and a string stretched
between it and the top of the stem post (" Ekiwenda ")
has a series of plaits dangling from it vertically. This
ornament is called " Akasenso." The top of the stern
post is often also ornamented with a tuft of grass
("Engudi").

When the canoe is sufficiently loaded, the ends of
the thwarts are just on the water line. The paddles
(" Enkasi ") are quite simply made, with no attempt at
decoration ; the blade is broad at the base and tapers
to a point, its surfaces not being absolutely plane, for one
side is very slightly concave, the other equally slightly
convex. An essential part of the equipment of a canoe
is the baler (" Olutiba "), made by hollowing out a
rectangular block of wood into a shallow basin. One of
the crew amidships makes it his business to free the canoe
at intervals from the water that collects in the hollow of
the keel (" Ekyuwo ") ; if the canoe is an old one so that
the edges of the planks and keel are getting soft, it cannot
be properly plugged, and one man is needed to bale con-
tinuously.

After two or three years the wood of the keel rots and
the planks crack, so that the life of the canoe is finished ;
the natives require two or three months to make a new
one.

With a good crew, each accustomed always to take the
same thwart, and paddling well together to a rhythmic
song, the motion is very pleasant, and considerable speed
is attained. Sometimes each man, at the height of the
stroke, knocks the haft of his paddle against the side
of the canoe, and a good test of whether the men are well
together is afforded by the noise : it should be as sharp
and well defined as the rattle of eight oars in the rowlocks,

\
88 CANOES AND A VOYAGE

and a paddler who is very much out of time with the
rest of the crew is greeted with yells of derision.

The time is set by the two bow paddlers (" Balambi "),
usually picked men. In my big canoe one was nearly
blind from smallpox ; he was a native of the Sesse Isles,
as are all the best canoe men.

The steersman (" Omugoba ") also is one of the best
paddlers, and in a high beam wind has all his work cut
out to prevent the canoe from falling off her course, as
I have often experienced ! For any slight correction
he uses a peculiar stroke ; putting the paddle into the
water as far away from the canoe as he can reach, with
powerful jerking movements, all his body and arms rigid,
he pulls the paddle through the water towards him with
a succession of rapid strokes. In a very heavy wind and
sea, however, v/ith a large canoe, this may not be sufficient,
and the steersman then has to make a rudder of his paddle
by using the projecting end of the thwart (" Empumi ")
as a fulcrum, whereby great leverage can be obtained.
The bowmen also help to pull the canoe's head round
into the proper course by putting the paddle well away
from the canoe and making the stroke towards the canoe,
on whichever side it is necessary.

On March 1, 1913, I started o£E on a " joy ride " to
see Nkosi, the southernmost isle of all the Sesse group,
which I had long wanted to see. The party consisted of
twelve paddlers, three boys and myself, with food for
four days and camp equipment in one canoe. At 7*30 we
got away, on a glorious, calm, sunny morning, in good
spirits at the prospect of a change after over a year spent
in the same camp on Bugalla. As we passed the north-
east cape of Bugalla there was a fine Situtunga buck on
the crest of the hill, but I could not see its horns, which
the men said were of a good size. We went southwards
down the channel between Bugalla and Bunyama, and
it was so calm that one saw well the lung fish that came

riTCHING CAMP ON THE ISTHMUS OF KIRUGU ISLK.

THE WESTERN I'OINT OF KIRUGU ISLE AND 'IKE ISTHMUS CONxNECTING IT
WITH THE CENTRAL RIDGE.

To face p.

A VOYAGE TO KIRUGU 89

up to the surface for a breath of air. They put their
mouths right out of water for a second or two, and then
turn on one side and go down with a peculiar roll, nearly
always flourishing one of the filamentous pectoral fins out
of the water.

Bunyama sends out a long point of rushes towards
Bugalla, so that the channel is very constricted, and the
scenery extremely beautiful. Bugalla here rises steeply
from the water, the slopes being forest clad, as is the coast
of Bunyama. Having passed this point we saw Fumve
in front of us, appearing as a steep grassy hill rising rather
abruptly about 350 feet, with forest at its base. We
went straight south between Bubembe and the prominent
south-east peninsula of Bugalla, and then between Fumve
and the little islet of Banda, off the bay on its west coast.
Here the scenery was lovely ; the densely forested shore
of Fumve was reflected in the blue glassy water of the
bay, and Banda was like an enormous rockery.

At 1 p.m. I thought a halt for lunch was indicated,
as we had been going for five and a half hours, so we landed
on the south end of Fumve and spent about an hour
there.

We then made straight for the northern end of Kirugu
islet, which we reached after another two hours paddling.
There was a lovely place for camping : a sandy isthmus
well sheltered by trees under which the men could sleep,
and providing an excellent place for the tent. This
narrow neck joined on to the main part of the island a
little plateau with cliff -like sides. The main part of the
island is a long narrow ridge of grass, falling steeply to
the water by cliffs or precipitous slopes, clothed by dense
bushes and trees wherever they could cling : the island
rises 180 feet above the water.'

Our choice of the sandy isthmus for camping was much
resented by some hippos, who evidently had their resting
places there, and grunted and snorted impatiently in the

90 CANOES AND A VOYAGE

water, whither they had been forced to retire on our
arrival. Eggshells of crocodiles were scattered about
over the sand, so that it was also a favourite resort, or
at any rate a resting place, of these reptiles, but none
were seen, which was fortunate, as a bathe in the clear
water was found most refreshing in the evening. The
little sandy neck was so narrow that one could choose
calm or rough water to bathe in, according to the time
of day. In the early morning the usual northerly land
breeze blew, and the water on the south side of the neck
was absolutely calm, to be ruffled during the heat of the
day by the southerly wind.

Next day I was awakened at 4 a.m. by the sound of
waves breaking on the south side of the beach, the sign
of a distant storm, which reached us about seven : the
usual burst of thunder and heavy rain. At about nine
the sky began to clear, and eventually the day became
brilliant, but with such a steady strong wind that it was
impossible to go anywhere in the canoe. I spent the
morning taking photographs, and getting a view of the
spacious distances from the top of the ridge of the island.
The views were superb ; it was wonderful to look south
and see nothing but bluest water with white capped waves,
and to think that there were ten score miles of it between
oneself and the mainland.

In the near distance lay Nkosi, the southernmost of
all the isles of Sesse, and to the east the long high ridges
of Kuye and Buguye, with some smaller isles to the
immediate west and north-west.

In the evening the strength of the wind abated a little
and the white caps were no longer seen, so that I was able
to struggle down the more sheltered west side of the
isle and see the waves breaking on the point, although the
wind was still sufficiently strong to make it very hard
work for both paddlers and myself steering. The sky at
sunset was full of wind and promise of rain to come,

■.■i>fr-.j,.'-^ ■ ;,;-,w»ii;,s.'w/"o«i>'.'jaif.gin»g5'<y?.'^-.

ON KIRUGU ISLE.

ON KIRUGU ISLE.

To face p. 90 (first page).

THE HIGH RIDGE OF KUYE ISLE.

SOUTH END OF BUGUYE ISLE WITH LULA AND LUVUNGU ISLETS.

To face p. 90 (second page).

DOWN SOUTH TO NKOSI 91

all the islands to the north being seen to stand out very
sharply and clearly, and there being a thick smoky haze
to south and west.

Next day at sunrise huge masses of black thunder
cloud were seen over Bugalla, Avhich evidently was getting
a heavy storm : our camp got a sharp squall at seven
o'clock, which nearly blew the tent away and split to
ribbons the outer fly.

For a few minutes the rain fell in sheets, but soon passed,
and by 8-30 a breeze cleared it up. The sun came out,
and the day eventually turned out gloriously, with
sufficient wind and waves to make the lake interesting.
We had a stiff fight southwards against the breeze, and
stopped for a brief rest among the pretty grottoes on
the rocky fern clad islet Lukiusa, eventually reaching
Nkosi at 10-30.

We first coasted along the south-east side, fully exposed
to the breakers coming up from the open lake, so could
not approach the shore very closely, especially as it was
very rocky, and no risk could be run of staving in the canoe.
Nkosi is a little over a mile long and a few hundred yards
broad ; it is low, flat and bush covered, but the trees do
not grow high, presumably owing to exposure to the
constant winds from the lake. We rounded a tail of rocks
at the south end, and got under the lea of the island into
sheltered water, almost at once finding a most picturesque
little grassy dell, with the lush green grass close cropped
as if by rabbits. This was the work of Situtunga antelopes,
of which one was soon seen quietly grazing. Their presence
on this small, very isolated and exposed island is of
considerable interest ; it is so covered with thick bush
that open spaces are few and small,, and they can only
traverse the bush by definite paths. Since there can be
no check on their increase save the amount of food
available, disease and the physiological effects of in-
breeding, the number on the island must long ago have

92 CANOES AND A VOYAGE

readied its maximum, and the pressure is presumably-
relieved by migration of the animals from Nkosi by
swimming across to Buguye and Kirugu Islands, and then
by secondary dispersion to the large islands further
north. 1

It is an interesting problem how far this antelope can
swim — for from the little islet Lukiusa, which could be
used as a stepping-stone, it is at least two miles to the
nearest neighbour, and open rough v/ater has to be crossed.

Even more interesting is the possible presence on Nkosi
of a species of Cercopithecus monkey ; for as Ave gently
drifted before the wind up the west side of the isle the
voice of a monkey was heard among the trees. I much
regret that I did not see the animal, for it is of course
possible that the noise may have been made by a parrot,
since I have certainly heard one imitate a monkey on
Kisigalla Isle.

The return journey from Nkosi to camp was easy, for
wind and sea followed and helped us along, and I got
back at 4* 15 after a most interesting day.

Next morning it was calm and clouded, and we broke
camp and got away at 6*45, having an uneventful run
as far as Fumve, taking a course this time along the east
coast, which is very picturesque.

The v/hole side of the island forms a sort of huge
amphitheatre backed by wooded slopes, and in the centre
of the foreground rises extremely steeply from the water a
flat topped, grassy hill, almost as high as the rest of the
island, about three hundred feet, mxaking one think of a
great altar in the centre of the amphitheatre. As we
passed Fumve a thunderstorm came up behind us from
the south-east, and I had to alter course, for it had been
my intention to return along the east side of Bubembe,

^ Major R. Meinertzhagen, F.Z.S., who visited Nkosi in one of the
lake steamers during the war, estimated that there must be two hundred
Sitiitunga on the island, and that thoy were sufficisntlj' different from the
Bugalla forms to be called a sub-species. (See Chapter VII.)

OUTH END OF .\KOSE ISLE.
Grass cropped by Situtunga.

CANOE SEEN " BOWS ON," KIRUGU ISLE.

To face p. 91

BACK HOME 93

pass between it and Bugaba, and then to the east of
Bunyama. The rough weather, however, now made it
impossible to pass along the exposed sides of the islands,
and we had to turn westwards and run between Bwiggi
and the north end of Fumve.

As it was, we had quite sufficient wind and waves to
make steering the canoe very hard work, for the wind
coming from behind and on the side blew the canoe's
head round, and it required continual effort to keep her
more or less before the wind.

However, when we had passed through the narrow
channel between Bunyama and Bugalla we got into
sheltered water and were able to land on Bunyama for
a rest and lunch, mid-day.

The storm soon passed, and the rest of the day was
bright and sunny, and we eventually got home again at
2 p.m., where I found that only -.'q inch rain had fallen,
showing that the storm was very local.

We had made the return journey in very good time,
having travelled about forty miles between 6*45 and 2,
with half an hour's rest.

CHAPTER VI

A TOUR OF INVESTIGATION AMONG THE
NORTHERN ISLANDS

The tour now to be described was made with Fiske at
the commencement of 1914, its object being a comparative
investigation of the chain of islands lying parallel to the
north coast of the lake, so far as could be done from
Entebbe as a source of food supply.

Besides our two selves and our four servants there
were thirty-nine canoe men, nine fly boys, and a minor
chief to act as headman, with his attendant. All these,
with belongings, were stowed away in two large canoes
and a small one holding five or six men only, which we
subsequently found very useful. We left Entebbe at
9" 16 p.m. on January 12th, having chosen the night time
because of the high wind during the day, which would
have made heavy work of paddling with well-laden
canoes. It was a beautifully calm moonlight night, and
we reached Bulago at 2*15 a.m., and pitched camp there,
as it was to be our base for some days. This is an
interesting island of curious shape, having a long forest
clad peninsula jutting out from the north-east corner,
and a large bay on the north shore, along which is a belt
of forest. The east coast is almost bare of trees or bushes,
the west has a fringe of bush, and the south point ends
in a sand-spit, much to the liking of crocodiles. Open
grass land forms the centre of the island, and rather on
the western side two rounded hills rise a couple of hundred
feet.

94

MS>^\

BULAGO ISLE, NORTH BAY AND FORESTED NORTH POINT.
Damba Isle on horizon.

THE OPEN GRASS LAND INTERIOR OF BULAGO, WITH FOREST BELT ALONG
SHORE OF NORTH BAY, AND ONE OF THE GRASSY HILLS.

To face p. 94.

$

BULAGO, KIZIMA 95

We saw a good deal of Bulago, and studied it
thoroughly, but found nothing very characteristic in its
fauna or flora. A notable tree was one called by the
natives " Musali," whose fruits were like loquats : I
had seen small specimens of this on Bugalla, but here
they were good-sized trees.

A colony of the fossorial wasp Bembex was noted on
crocodile point, where the sand was eminently suitable
for their burrows.

We had decided to take careful note of the butter-
flies to be found on the several islands, as some sort
of index to the general suitability of an island for
insect life.

On this first visit fifty-two species were noted, and on
the return journey in March twenty- two more were added
to the list. It was on Bulago that I first met with the
Acraeine species Planema aganice and PI. alcinoe, neither
of which occurred on Bugalla or Damba, though Damba's
immediate neighbour Kome was subsequently found to
have these interesting butterflies.

On January 14th we visited the islet of Kizima. This
is only about 300 yards in diameter, and rises out of the
water as a rocky, flat topped, grass clad eminence, with
shrubs and small trees along the steep coast line, and one
little bay with sandy beach on the west. A tangle of
" Oluzibaziba " bushes sheltered a hippo ; a large and a
small crocodile and two large monitor lizards {Varanus,
called " Enswa-swa " by the natives) were seen ; the
basking places of the latter on the rocks were very
numerous.

Among the branches of a bush I found a small elapine
snake, light grey with dark ' brown blotches, which had
very obviously dined both wisely and well ! Rodents,
about the size of a water vole, were abundant on this
isle ; their runs were very noticeable, and also the debris
of grass stems left where they had been feeding. Kizima

96 A TOUR AMONG THE ISLANDS

was therefore alluded to as a " rat island " ; several others
will be described later.

The little hill of Kizima seems to act as a chimney
directing a continual current of air upwards, for whenever
it v/as visited there were always numbers of kites soaring
round and round in narrow circles just over the island.
Around it are fragments of rocks that have broken away
and now form resting places for many cormorants and
darters ; snow-white egrets with black bills and legs,
Egyptian geese, and a giant heron were also there, with
a sacred ibis, not to be seen on every island.

Flycatchers were seen, and swallows were fairly plentiful.

Butterflies were few (see p. 125) : Olossina was repre-
sented by an average catch per fly boy of 29 per hour
(both sexes). No termite hills were seen ; indeed, ter-
mites seemed altogether absent.

The Island of Tavu was first visited on January
13th. It is about a mile in greatest diameter, being
low and flat and covered thickly with vegetation.
The northern two-thirds are marshy, and only ambatch
grows there, with a dense fringe of rushes ; the southern
end is very sandy and covered with cane, " E'tungulu,"
and bushes, with rushes growing in the shallow water.
On the west coast is a large raised beach of sand, which
was found to provide admirable breeding places for
Glossina. The rest of the island has forest growth, and
was at one time cultivated. Tavu proved to be a favourite
haunt of crocodiles, of which nine were seen ; some on
a rock on the east coast, others in the shallow water off
the sandy south point, on which they lay their eggs.
No " Enswa-swa " were seen, and subsequent visits
always showed them to be scarce.

Of snakes, the beautiful horned puff adder {Bitis
nasicornis) was seen, and a large black elapine was rather
common.

Birds abounded ; giant herons, and herons of other

TAVU— RESUVU 97

species, glossy ibis, the open-billed stork, egrets and a
stone curlew were seen, while cormorants and darters
were numerous in the ambatch trees, on which they nest.

In the thick jungle, where we very easily lost our sense
of direction, the abundance of insect life was shown by
bee-eaters, flycatchers, coucals and other insectivorous
birds.

Butterflies were numerous (see table, p. 125), and forest
species such as Planema macarista, two species of Charaxes,
and a large Eur alia were noted. Three species of Tabanidae
were noted on the shore, possibly the low lying and marshy
ground is suitable for the larvae ; a common Tabanus
[T, variatus), a Haematopota and Chrysops brucei were
the species seen, the latter quite plentiful, and one was
seen in the grip of a predaceous Asilid fly. Termites
abounded, and the Dorylus ant was seen, while Culicine
mosquitoes bit unpleasantly in the jungle.

Large Achatina snails were plentiful.

Glossina was very abundant, the island furnishing
ideal conditions of food, shelter and breeding grounds.
The average catch per boy hour was 39' 2.

On January 17th I visited the sand bank south of
Tavu, named Resuvu. Its north half is submerged at
all times, and there is a dense growth of ambatch on
which cormorants and darters nest. The south half, at
this time, was above water, and measured about 50 by
20 yards, but is probably almost entirely submerged when
the lake level is high. Notwithstanding, I found larvae
of a very common Noctuid moth ; in the sand a small
black ant was found, and I saw' a nest of the Belonogaster
wasp, which feeds its larvae on caterpillars.

On January 24th we moved camp to Nsadzi Island,
pausing on the way to examine Kimmi, which we found
extremely interesting, since, though quite small and flat,
it contains portions of almost every kind of condition met
with on the lake shore. We landed first on the south-

8

98 A TOUR AMONG THE ISLANDS

west side, whose eastern tip is formed of sand left by the
receding lake at the foot of a low cli£E in which are caves,
full of bats. Further westwards the shore is low, of flat
rocks, with many holes kept full of water by waves ; at
a higher level is a sand beach now grown over by cane
and bushes, and used as a nesting ground by crocodiles.
The north-west shore is flat and marshy, with a thick
fringe of rushes and bushes of " E'tungulu " and ambatch ;
the north-east part of the island is forested, but near the
water is a large patch of open grass land, with very sandy
soil, in which " Enswa-swa " burrow and lay their eggs.
The south-east shore is marshy, with lush grass much
frequented by hippos ; this type of shore was dubbed
" Hippo Grazing Ground."

The vegetation of Kimmi seemed to me much more
varied than on Tavu, which is about the same size. A
Chrysojphyllum tree (Sapotaceae), known to the Basesse as
" Omugalati," which was not seen on either Bulago or
Tavu, an anonaceous shrub scarce on the other two isles
but abundant here, and a shrub with clusters of fragrant
white flowers, which I had not seen before, were noted.
Butterflies were fairly abundant, a forest Papilio, with
several forest Planema, were noted ; the numbers counted
will be found in the table (p. 125).

The most striking feature of Kimmi was the voracHy
of Glossina, which came off in a swarm to the canoe before
we had touched land, and bit ferociously. It was found
that more than half the number caught were females,
which is quite exceptional ; the significance of this has
been discussed in the chapter on Glossina. We comforted
ourselves with the recollection that former experiments
had never shown the flies from Kimmi to be infective,
and put up with the inconvenience as best we could.
The average catch per boy-hour for this island reached
the high figure of 60*9 : on one day alone it was 81 !

The hunger of the fly tallied with the fact that we found

NGAMBA 99

no crocodiles or " Enswa-swa," the favourite source of
food ; although hippos are fond of the island I do not
consider them of much importance as food for the fly,
which has therefore to depend for food, on Kimmi, mainly
on birds which frequent the shore.

On January 30th we visited Ngamba, which lies between
Kome and Nsadzi. This island, like Tavu, is covered
with forest growth, but is more rocky and rises higher ;
it is roughly square, and of about the same size as Kimmi.
The forest everywhere comes close to the water's edge
save on the east side, where part of the shore is marshy,
and at the rocky north-east point. The shore, on the
whole, is either rocky or pebbly, and only at the north-
west corner is there a raised beach of sand. Along the
south coast the rock is grey, but at the north-east
point is of the more common spongy red brown
formation.

As usual we first circumnavigated the isle, and counted
six crocodiles and three "Enswa-swa; " two others were
also seen, and bushes of " Oluzibaziba " bore evidence
that Situtunga antelope had been browsing on them :
probably these had swum across from Kome, where they
live, for none were seen on Ngamba. The forest trees were
noted to be varied ; one very fine fig tree with huge
leaves, also occurring on Kimmi, was conspicuous here :
the natives called it " Omululu." Another noticeable tree
had sprays of greenish flowers like those of Portuguese
laurel : the natives called it " Omuziru " ; and another,
with large white flowers, was called " Omukoba."

Here and there were beds of a magnificent herb which
I took to be at least allied to Salvia, with spikes of large
blue flowers, growing up to a height of four or five feet.
This was subsequently often found growing in pebbly
places near the water, and was quite a feature of some
islands. Of birds, a sacred ibis and giant heron were
noted, but no other notes were made. There were some

100 A TOUR AMONG THE ISLANDS

features of interest in the insect fauna : a D. chrysippus
was seen crossing the channel between Ngamba and Nsadzi,
and nests of the interesting ant (Ecophylla smaragdina,
and the fierce little black ant called by the natives
" Obusaji-saji," were found on leaves of trees ; they are
not by any means universally distributed among the
islands.

The tubular cases of larvae of some species of Psychid
moth were seen on the tree trunks in great numbers, and
afforded the first example of the fact that a small circum-
scribed area may support only a few species, but that some
of these may flourish much more abundantly than in larger
areas where they are exposed to greater competition.
During the course of this tour many other examples were
noted, and will be mentioned as instances of " insularity "
in the fauna or flora.

Olossina was scarce on Ngamba, the average catch being
9" 9 per boy-hour. We remained at Nsadzi camp until
February 5th, but spent only a little time over the island
itself, since it was too big to be thoroughly investigated
on a short visit. It is about three miles long, but very
narrow : four-fifths consist of a steep grassy ridge almost
bare of trees but with innumerable termite hills, closer
together than I ever saw anywhere else. At the centre
of the south side of the island cliffs descend steeply to a
sandy bay, with a strip of forest at the foot of the cliffs.
There is another patch of forest on the north side, but
elsewhere the shore has only a very narrow fringe of bush.
The eastern fifth of the island is low and flat, and sharply
marked off by the escarpment of the end of the main
ridge ; it is covered with forest, full of birds. On Nsadzi
we found most interesting evidence of insularity. The
grass on the uplands had been burnt off by Fiske on a
previous visit some weeks before, and the new green blades
were growing up among the burnt black stems of the
tussocks. Over an area of a great many square yards

NSADZI TO KOME 101

every blade had been eaten by caterpillars of a Noctuid
mothji whose green and black colours harmonized abso-
lutely with those of the grass. These caterpillars were
hurrying about in all directions looking for more food,
and numbers of a brown " Ichneumon " were seen
depositing eggs in them. While piercing the larva with
its ovipositor the ichneumon grasped it firmly in its
mandibles, a habit which is shared by the stinging fossors,
who have adopted a somewhat more advanced kind of
parasitism.

On Nsadzi were found, firstly by Fiske, the footprints
of a large Situtunga buck ; the importance of this has
been alluded to in the chapter on Glossina.

We moved from Nsadzi on February 5th to a new
base on the east end of Kome, whence we intended to
examine the small islands of that neighbourhood.

It was a most lovely voyage on a calm and clear day,
with much cloud of varied type and deep purple shadows.
We left Nsadzi at 7 a.m., and travelled past the tips of
the several peninsulae jutting out from the north shore of
Kome, and then down the beautiful channel between
Damba and Kome, where blue, pink and white water
lilies were in fuU flower. A small cloud funnel was seen,
giving warning of a short storm which came from
the east at about 4 p.m., and burst on us while we were
on Sanga Islet, but only lasted for half an hour. I was
much interested in this, because while I had been encamped
on Damba in 1911, only a few miles from Sanga, which
could be plainly seen (see map), I had never had an
afternoon storm there, but had noticed that the end of
Kome often caught rain which missed my camp.

Sanga Islet was examined this day, and was found to
be amazingly interesting. It only measures about 100
by 50 yards, is very rocky and well covered with forest,
and shows abundant evidence of insularity.

^ Laphygma frugiperda {Xenohianae).

102 A TOUR AMONG THE ISLANDS

The shore is composed of broken rocks, and the forest
comes to the water, but there is very little undergrowth
of bushes, so that we were able to walk about freely.

Olossina was absent : the other arthropod life was of
great interest. The first thing noticed was the abundance
of very large, long-legged black and yellow spiders
(Nephila), which spin webs several feet in diameter.
These great webs formed sheets, often in planes one behind
another when the branches offered suitable support,
so closely placed that the spiders on the central sheets
could hardly get any insects on their webs, and looked
starved. Sanga, therefore, was known as a " spider
island," and others will be described later. The next
feature noted was the unusual abundance of predaceous
Mantidae, and this was associated with absence of insecti-
vorous birds ; a very interesting example of the " balance
of nature."

At the time of this visit, and we remained on the island
several hours searching it thoroughly, I neither saw nor
heard a single flycatcher, bee-eater, coucal, or any of
the common small warblers ; parrots and weavers were
the only birds noted in the forest. At a second visit,
however, one " Kunguvu " flycatcher was noted, and the
call of a common cuckoo was heard. Perhaps correlated
with this absence of insectivorous birds was the fact,
which Fiske and I noted independently, that the butter-
flies, especially the mimetic Eur alia and Aterica galene
which I have always found very difficult to catch on account
of their shyness, were here much less shy and easier to
catch. Fiske thought that it was on account of the
multitude of spiders' webs, which had resulted in the
butterflies flying more slowly and cautiously to avoid
them ; but I noticed that when on the ground the mimics
were more easily approached than is usually the case.

The numbers of butterflies found on Sanga are given in
the table (p. 125) ; the most noteworthy feature was unusual

SANGA— KISIGALLA 103

abundance of the large Hesperid Rhopalocampta cTialybe,
whose peacock green and blue wings flashed in the sun as
these beautiful butterflies darted about in the sunlight
in typical " skipper " fashion. Acraeines were represented
by two species among the tree tops out of reach, and one
specimen of A. egina ; Amauris niavius was the only
Danaine seen ; of Nymphalines, Euralia dubia, E.
dindrcha, and a forest Gharaxes were noted ; while the
large and beautiful iridescent Salamis, its pale green
wings suffused with a pink mother of pearl lustre, was
very plentiful. Lastly, the caterpillars of an aposematic
moth, Aletis erici (see description, p. 214), were very
abundant on Sanga, which interested me more than any
other island, because it was the first visited of those on
which the great spiders were so abundant, and showed
so many other examples of insularity.

On February 11th we visited the little isle Kisigalla,
lying close to Sanga, and, like it, rocky and covered
with trees, but with one piece of sandy shore. It is
about fifty yards in diameter, and, though so close
to Sanga, differs from it in many respects. It was
characterized by almost entire absence of the Nephila
spiders so abundant on Sanga ; by abundance of a large
black Scoliid, of cocoons of the gregarious caterpillars of
Lasiocampid and Eupterotid moths, and also of a very
large fawn coloured slug.

Birds were not so scarce as on the spider-haunted
Sanga ; the " Kunguvu " flycatcher, sunbirds, weavers,
pigeons, and the beautiful blue and crimson plaintain-
eater were all noted.

Butterflies were quite plentiful, the most interesting
being the large mimetic Lycaenid Mimacraea poultoni,
and the Nymphaline mimics Pseudacraea eurytus forms
hobleyi and tirikensis, A very fine Satyrine that was new
to me escaped capture. Olossina was scarce, the catch
per boy-hour reaching only the average of 2* 3,

104 A TOUR AMONG THE ISLANDS

The islet Lula, close to Kisigalla, was visited on the same
day, and was found to differ considerably from its neigh-
bour. It consists of a small rocky ridge covered by bushes
and not by large trees, and is much frequented by crocodiles,
whose basking places are numerous. Of birds, a black
and white flycatcher, not met with on any other of the
five islets off the east end of Kome, was noted to be
relatively plentiful. Neither the spiders of Sanga nor
the slugs of Kisigalla were found on Lula. Butterflies were
interesting (see p. 125) ; Acraeines were more numerous
than any other group, forming seven-fifteenths of the total
number of species found ; A. pe7itapolis was noteworthy,
as it is not found on every isle. The Nymphalines were
rather surprising : a huge Euralia, looking very out of
place among bushes on a tiny islet, Pseudacraea eurytus
form obscura, Ps. lucretia and Salamis made up the total.
The only Lycaenid found was a brown, black spotted
Pentila, whose Acraeine appearance fitted in well with the
prevailing group of the island.

Colonies of Eupterotid larvae were seen on Lula as on
Kisigalla.

The average catch of Olossina was 9' 1 per boy per hour.

On February 6th we worked Dwanga Mkuru, the
largest of three closely associated islands lying off the
eastern point of Kome. This is about three miles long,
and very narrow ; generally speaking, its features are
of the type associated with a clear, rocky shore.

The centre is a grass topped ridge sloping steeply on
the east down to the rocky or pebbly shore, where at one
point a dyke of red, spongy volcanic rock, stuffed with
fragments of grey rock as a pudding is stuffed with plums,
runs out across the pebbly beach. On the west side is a
bold cliff, at the sheltered foot of which is a strip of thick
fertile forest, with some huge " Omuvule " (African teak)
trees, and an overgrown banana plantation. All along
the shore on the east are thick beds of fern and Salvia^

DWANGA MKURU AND MTO 105

backed by thin and poor looking forest trees, swept by
the prevailing south-east winds. Many of them were
covered with spiders' webs, for this island is markedly a
" spider island " ; as on Sanga, this is correlated with
great scarcity of insectivorous birds and abundance of
Mantidae.

Butterflies were extremely abundant : the beautiful
green, pinkly iridescent Salamis was more numerous
than at any other locality visited, four might be
seen at once. Of Acraeines, Planema poggei was of
interest, and other Nymphalines besides Salamis were
Euralia dubia and Pseudacraea lucretia, which latter
Fiske, until it had been caught, thought must be a
Neptis from its flight. I had often noticed that its
flight was much more like that of a Neptis than the
Danaine (Amauris), which it somewhat resembles in
pattern.

Glossina was completely absent from Dwanga Mkuru.

DwANGA Mto, lying alongside the last named, was
visited on February 9th. It is more or less covered by
jungle, but there are some open patches, and it does not
rise so high as its larger neighbour. The shores are mostly
of pebble or rock, with much fern and Salvia, and the
usual common bushes.

A species of bush was seen by the water which had not
been met with before ; it had very large heart shaped
leaves and spikes of inconspicuous flowers : the natives
called it " Omukwakula." Other features of the flora
were a single specimen of the large " Ekibo " palm and a
small patch of elephant grass (" Ekisagazi "), which is
very unusual on the small islands.

Birds were not so noticeably scarce as on Dwanga
Mkuru, for sunbirds and bulbuls were plentiful and the
" Kunguvu " flycatcher was there.

Being markedly a " spider island," Dwanga Mto had
abundant Mantidae', the large grasshopper called "E'jansi"

106 A TOUR AMONG THE ISLANDS

by the natives was much more plentiful than on other
islands hitherto visited.

Regarding butterflies, the most noteworthy feature was
the abundance of the large Hesperid Rhopalocampta
forestan, hitherto not encountered on the small islands ;
its near ally, the beautiful Rh. chalybe of Sanga Isle, was
not seen here. The single Papilionine (see table, p. 125)
was P. leo7iidas, whose Danaine model Melinda petiverana
was also seen. Acraeines were so extraordinarily scarce
that I only caught one specimen, and only one other
species was seen. Nymphalines included Salamis, which
was very abundant, Pseudacraea eurytus forms hobleyi
and terra, Ps. lucretia, Euralia dubia, two Charaxes,
Si, Neptis, and H. misippus and C. cardui, neither of which
occurred on other islets of this group.

A curious Lycaenid was found here, a large grey species
with angulated wings. ^ A feature of the island fauna was
the number of long narrow cases of the larvae of a Psychid
moth, attached to tree trunks.

Olossina was almost absent ; the nine fly boys could only
catch four males and a female.

The third island of this group, Dwasendwb, was
visited on February 10th.

It is covered everywhere with forest or overgrown
plantations ; almost everywhere the shores are rocky ; on
the north-east there is a dense bank of ferns. The main
insular feature of Dwasendwe was the great abundance
of the Aletis larva first met with on Sanga ; these had
defoliated the food plants over a large area.

Of butterflies, Acraeines were again very scarce, one
Planema and one small Acraea only were seen ; of Hes-
peridae the large Gamia bucholzi, with mottled underside,
was noted : a Lycaenid was found which was never met
with on any other island, and was apparently associated
with the " Omukwakula " bushes on whose flowers it seemed
^ Aslauga purpurascens, Lipteninae.

, -^^■^mm^gi&s^^m

"^^^m^m^

\f fc'%;;>.i.t-;;;i--.

sM^yJfeflSfV

ilUHUUlBttriR^fKSE

MAUNGWE 107

to be ovipositing. Although this bush was abundant on
the neighbouring island, Dwanga Mto, this Lycaenid had
not been "seen there. ^ It is a large white species
with fine hair-like wavy lines on the under surface. On
this island was found in a spider's web a second specimen
of the West African Lycaenid, first caught on Dwanga
Mto. Dwasendwe was markedly a " spider island " ; and
Glossina was absent.

On February 13th we moved camp to Kerenge Isle,
one of a group lying further eastwards. Fiske and I
started at 6-30 with four paddlers in the small canoe, as
we intended to visit sundry islets en route, leaving the big
canoes with kit and men to find their more laborious way
in their own time.

At daybreak there were very threatening clouds, with
thunder to the south-east, and it seemed as if we must
get caught by a storm. The canoe men paddled furiously
to get to our first islet, and the small canoe seemed to
leap out of the water at every stroke, so that the un-
fortunate Fiske in the bows got very splashed, as there
was some swell from the distant storm. After one and a
half hours' paddling for the distance of six miles, we safely
arrived at Maungwe, a rock spit only about 80 by 30
yards in size, thickly covered with ferns, herbs, low bushes
and Ipomaea creepers with a few fig trees, on a fork of
one of which was an enormous nest, possibly of a heron
or a bird of prey. A very common weed, with
dull mauve flowers {Erlangea tomentosa, called by the
natives " Obutwatwa "), grew very tall and in dense
masses which sheltered a small bird with the habits of a
flycatcher, brown, with white throat and belly and brown
chest, which was subsequently found on minute islets
only, nearly always in masses of the " Obutwatwa."
It had a very sweet voice, and its song resembled part of
a nightingale's repertoire, so I dubbed it for reference

^ Spalgis pilos.

108 A TOUR AMONG THE ISLANDS

sake " Nightingale flycatcher." Swallows and weavers
feeding on the " Sami " gnats were also plentiful, and
small lizards were seen.

No Lepidoptera were noted, although a Belonogaster
wasp was seen, which is known to feed its grubs on cater-
pillars. Two species of carpenter bees which bore into
wood, and small ants, were seen, but there were no Glossina
on Maungwe.

By the time we had finished with this little island
all threatenings of storm had passed off, and we went
on happily to Enktjsa Islet, merely a reef of red spongy
rock for the most part broken into pebbles, with a
dense growth of ambatch bushes at one end, inhabited
by weaver birds. There was a great growth of a large
yellow vetch, 1 which seems to prefer pebble beaches in
the vicinity of water. One small and feeble Lycaenid
lived on this islet, keeping close to the ground.

From Enkusa we went to Ziro, going round Isentwa
and through a break in a reef which almost joins the two,
into a very secluded bay dubbed by Fiske "the Pirates'
Haven." Here we lunched, and afterwards went up on to
the central grassy hill to get a broad idea of this island
and its neighbours.

The general appearance of this group was not at first
sight inviting, after the beautiful, forested and friendly
islands of the Kome group. They have grassy hills, but
much less true forest growth : the pyramidal grass clad
peak of Kiuwa was very striking, apparently rising straight
out of the water, treeless, as high as the highest point of
any island of the neighbourhood.

We did no work on Ziro this day, but visited it subse-
quently. The only notable feature about its coast is a
rocky headland at the south-east point, with caves of
weird fashioning quite suited for pirates' lairs !

I noted that the flora of the isle in some respects rather

1 Crotalaria striata.

ZIRO— SPIDERS' WEBS 109

differed from that of the group we had just examined,
for several flowers were seen that had not been previously
noted but which had been noted at Jinja, on the main-
land, in 1910. Two instances of butterflies similarly
occurred ; and these facts suggested that these isles had
not been connected with the former group, but only with
that part of the coast immediately to the north of them,
the others perhaps having derived their inhabitants from
the mainland more to the west. (See map.)

As regards birds, several giant herons, a dappled grey
heron which I had not seen before, and open-billed storks
in great numbers were seen.

Flycatchers were seen and heard in fair number ;
Enswa-swa lizards seemed fairly numerous, for nine were
noted.

Ziro is markedly a " spider island," but Mantidae were
not noticeably abundant. Of butterflies, Acraeines were
very abundant ; A, encedon in several forms was note-
worthy. A Lycaenid ^ was taken which I found at Jinja
in 1910, but had not been met with on any of the
islands : this is notable after what has been said about
the flora.

Glossina was far from abundant, the average being
6' 8 per boy-hour.

We remained in our camp on Keeenge from February
13th-20th, using it as a base from which to visit neigh-
bouring isles. There was nothing very characteristic
about it ; it is flat and has a good deal of open grass land,
and there are remains of former banana plantations, now
full of the great spiders' webs to a disgusting degree :
to get one's face and hat, or bare arms, covered with the
sticky clinging web was extremely unpleasant. So strong
is the silk that a sunbird was seen in a web, having become
so entangled that it could not free itself ; we managed to
reach up to it with a long stick and set it free. The webs

* Lachnocnema bibulus^.

110 A TOUR AMONG THE ISLANDS

were so close that the spiders on those between others
were quite starved and their bodies very shrunken.

Mantidae were not noticeably conspicuous ; the large
" Amajansi " grasshoppers were, and the beautiful king-
fishers which feed on them were also abundant. Acraeine
butterflies were not so abundant as on Ziro, indeed here
they were almost absent ; Lycaenidae were relatively so
abundant as to be the main feature of the butterfly fauna.
Of Hesperidae the large Gamia bucholzi was noted, and
Fiske found a shady bank where large numbers of a
Hesperid new to me were to be seen ; ^ it seemed to be
associated with a plant, probably of the Arum family,
which carpeted the ground here. Glossina only averaged
6*2 per boy-hour.

We visited Lukalu Isle on February 16th, and found it
very interesting. About half a mile long, but narrower,
its greater part is a grassy plateau rising slightly in the
centre, with large beds of bracken on the slopes ; at the
east end are some high rocks with caves, and here are a
few big trees. One of these was like the common tree ^
called " Ekirikiti " by the Baganda, but the spiky flowers
instead of being scarlet were of a rosy pink tint which I
had not seen before.

From the eastern end of the isle stretches along the
south shore a narrow strip of dense bush ; the north shore,
however, is very different, there being a belt of thick cane
growing on sand,- on which were marks showing that
crocodiles had lain there, and hippos had made tunnels
through the cane thicket.

Five Enswa-swa we^e seen. The only birds that attracted
attention were a coucal and a grey plaintain eater ; the
*' Kunguvu " flycatcher was not seen. No Nephila spiders
were seen ; of butterflies Lycaenidae predominated ;
Lachnocnema bibulus was caught ; Acraea encedon was
in great variety ; the only Nymphaline seen was Precis
^ Andronymus leander. ^ Erythrina totnentosa.

LUKALU— ISENTWA 111

clelia. Glossina averaged 10' 6 per boy-hour. There is a
small accessory islet to the east of Lukalu, separated
from the main isle by about 50 yards of water ; it measures
about 100 by 50 yards, and is composed of red rock broken
up into pebbles on its surface. There was a variety
of flowering plants and low bushes at one end, and
many ambatch trees, among which a species of moor-
hen was noted. Weaver birds, sunbirds and wagtails
were plentiful.

The only butterflies noted on this accessory islet were
the Lycaenid T. telicanus, which was evidently much
attacked by wagtails, for a large proportion had a
A-shaped piece taken symmetrically out of the two hind
wings.

On February 16th we visited Isentwa Isle, which is
roughly triangular, having sides about three hundred
yards long.

It is composed almost entirely of grass land, with a
narrow fringe of bush or cane at those parts of the coast
where the rock is not too steep. Enswa-swa were noted
abundantly, but no crocodiles were seen ; indeed, as a
whole the islands of this group had very few crocodiles.

Big spiders were absent from Isentwa, and also Glossina ;
butterflies were very few, the feature being abundance
of the woolly-legged Lycaenid Lachnocnema bibulus, of
which at least half a dozen individuals arose together
from one plant as I passed. The only Nymphaline seen
was Precis archesia.

KiuwA Island was visited on February 19th, and proved
to be remarkably interesting from the point of view of
the bionomics of Glossina, which was found to the extent
of 18-2 per boy-hour, and yet we could not make out where
were the pupae which kept the numbers at this level.

Search as we would, no more than an average of 3- 3
pupae could be found per boy -hour, in spite of rewards
offered.

112 A TOUR AMONG THE ISLANDS

As has been said, the island consists mainly of a steeply
rising grass-clad conical hill, a little over a mile long and
half as broad ; its north-east end slopes less steeply,
and here is evidence of former banana cultivation, the
shambas now overgrown with bush and creepers ; some
of the rose flowered " Ekirikiti " trees first noted on
Lukalu were seen. There were also thick bushes of the
usual kinds, and beds of the Salvia-like plant. Along the
south-east shore are bold cliffs, at the foot of which are
two little bays with beaches of pebble ; the south-west
shore is formed of unbroken rocky strata rising steeply
from deep water ; the west shore slopes less steeply and
is fringed by bushes, so that, on the whole, the island
seems unfavourable for Glossina, which yet was numerous.
Three crocodiles and one Enswa-swa were seen, but no
traces of hippo. Fruit eating bats were abundant, and
small rodents were very numerous in the tussocky grass,
so that Kiuwa was called a " rat island,"

In the old banana plantations sunbirds and weavers
were very plentiful ; flycatchers were scarce, and none
of the black and white species were seen. A thrush was
heard singing, which, though plentiful at Entebbe on the
mainland, I had never heard before on an island.

Only one single specimen of Nephila spider was
seen ; no termite hills were seen anywhere on the
island. 1 "

The butterflies of Kiuwa were rather interesting (see
table, p. 125) ; three species of small brown Acraea were ex-
tremely abundant, but encedon was not seen. Nymphalines
were very scarce ; a Pierine scarcely ever met with on the
islands was quite abundant on the beaches on the south-
east shore ; two Lycaenids not common everywhere
were noted as abundant ; Hesperidae were very scarce in
individuals, and only four species were seen.

On February 20th we moved camp from Kerenge to the

^ Cp. Kizima, also a " rat island."

ZIGUNGA—WEMA 118

south side of Wema Isle, visiting Kibibi en route, and also
the Islet of Zigunga.

This consists of two small islets joined by an isthmus,
so that the south-west coast forms a bay : it is composed
of red rock, covered with grass and rising steeply ; there
are a few fig trees.

A feature of the island was a plant with thick, sword
shaped leaves, from which the natives obtain fibre. One
crocodile and two Enswa-swa were seen.

Swallows, weavers and sunbirds were quite common,
and it was interesting to find again the " Nightingale
flycatcher " frequenting thick clumps of the plant
Erlangea {" Ohutwshtwa, ") as it did on Maungwe Isle.
No Nephila were found, nor were Glossina seen ; a single
Acraea and a long-tailed Lycaenid, abundant on Kiuwa,
were the only butterflies noted.

After visiting Zigunga we went on to Wema, a large
island on whose south shore our new camp was pitched,
and used it as a base untU March 2nd. It was much
too big to be thoroughly examined in the time at our
disposal, so we confined ourselves to the south shore.

The views of Wema itself obtained from the summit of
the hill behind the camp were quite the most beautiful
of any I have seen on an island.

There are four hUls on the island, of which the one above
alluded to, at the south-west corner, is grassy and flat
topped ; another, at the north-west corner, is flat, with a
patch of forest running nearly to the top ; on the con-
necting ridge between these two is more forest, which
also runs up the third hill, the central one of the isle,
grass clad on top. The fourth hill at the north-east
corner has a grassy base and a cap of forest on top. The
south-east coast of the island has a fine fly beach with
forest at the back, and the eastern point shows a red cliff
visible from the camp. From the top of the hill behind
the camp a lovely panoramic view was seen on the first

9

114 A TOUR AMONG THE ISLANDS

evening ; looking eastwards across Yempata and Bus'rl
one saw distant Buvuma, lying in the mouth of the gulf
which gradually narrows down to the origin of the Nile at
Jinja (see map) ; to the west, Damba lay long and low on
the horizon, flooded with golden light from a lovely sunset.

The part of Wema that was investigated proved most
interesting ; Glossina was extremely abundant, averaging
55 per boy-hour, and at one spot the catch averaged 101*5
per boy-hour ; one boy caught 78 in half an hour !

Monkeys were seen for the first time since we left Kome,
and I was struck by the abundance of a black and white
crow which flew about in flocks like rooks. Wema proved
to be very markedly* a " spider island," and as regards
butterflies it was richer than any of the other islands
investigated ; I think that it has probably as many
species as the much larger Damba or Bugalla, and
several Nymphalines were seen there which had not
been met with on the larger isles.

Of Papilioninae, P. dardanus and P. ugandae, with several
other forest species, were noted ; also, P. leonidas and
P. poUcenes were very common. Of Pierinae, several
species of Terias and one Teracola were noted. The
following species of Nymphalines were especially noted :
Pseudacraea boisduvalli, lucretia, and eurytus forms hobleyi
and tirikensis, six species of Neptis, together with Nepti-
dopsis, nine of Charaxes ; Salamis was fairly abundant,
while no Euralia was seen. Danaines were poorly repre-
sented by a few D. chrysippus, Amauris echeria, A. niavius
and A. psyttalea. Only a few specimens of each were seen.
Acraeines were surprisingly scarce ; very few individuals,
belonging to nine species, were found ; noteworthy were
Planema alcinoe, macarista and poggei. The Hesperidae
included the peacock blue Rhopalocampta chalybe ;
among the Lycaenidae was an extremely beautiful large
blue species, with long tails of snowy white hue (Zeltut
aniifaunus).

MARIDA— WAVUZIWA 115

Perhaps the most interesting butterfly of the islands
was a Lihythea caught by Fiske ; no species of this family
had been met with previously on any isle. A total of
124 species was found on this brief visit to Wema, and
probably half as many again would be added by further
investigation (see p. 125).

On February 21st we visited two of the small islets
lying east of Wema, going first to Marida, which consists
of two flat, grassy portions connected by an isthmus
along which is a pebbly beach. There is no forest growth,
but low bushes and fig trees form a fringe along parts of
the shore. A coniferous tree, Podocarpus, called by the
Basesse " OmusSnene," was here noted for the first time.
A crocodile v/as seen, and a nest of eggs found, and two
Enswa-swa were found on the island.

The only point noticed about bird life was that bee-
eaters were there in some number, and they do not frequent
every island ; sunbirds also were quite common on the
isthmus, and the " Kunguvu " flycatcher was seen.

Marida is very decidedly a " spider island " ; Glossina
was represented by an average catch of 4*6 per boy-hour.
Of butterflies, it was remarkable that no Acraeines were
eeen, and while individuals of other species were scarce,
a fair number of species was seen (see p. 125).

The feature of Marida was the abundance of a slug,
broadly resembling the English L. agrestis, but with a
little fleshy spike on the end of the body reminding one
of the caudal horn of a Sphingid larva ; this species had
not been met with before.

From Marida we went on to Wavuziwa, a rocky islet
with very little shade except from fig trees ; ferns were
abundant. One crocodile and one Enswa-swa were seen ;
of birds, weavers and sunbirds were plentiful, also the
" Nightingale flycatcher," and one coucal was seen.

Glossina was very scarce, for all the fly boys together
in an hour only caught seven ; Nephila was not seen at

116 A TOUR AMONG THE ISLANDS

all ; of butterflies, Acraea zetes, of rather Eastern type,
was notable, and no Nymphalines were seen. Insular
features of Wavuziwa were two ; a Lymantrid caterpillar
new to me ^ was found in abundance feeding on the
" Anyamberege " plant, on which plant I had never
previously found any caterpillar, and fluffy white masses
of some species of Coccidae, which thickly infested bushes
and undergrowth.

On February 26th, Sege was visited ; a not very
interesting island of the same size as the last named,
rocky, covered with grass and without any beach. Rather
conspicuous on this islet was the fibrous leaved plant
found previously on Zigunga, and an aloe whose spikes
of scarlet blooms arose from rosettes of fleshy, thorny
edged leaves.

Two Enswa-swa were found on the island ; one, the
biggest I ever saw, was much less timorous than usual,
and did not retreat into the water until I approached
within a few yards of it, after which it remained close at
hand with its head above water.

Of bird life the only notable feature was the " Nightingale
flycatcher " ; Nephila was absent, Glossina very scarce, for
the nine boys in an hour only caught twenty-seven, all
males ; of butterflies, ten species were noted.

Waitwe was visited on the 26th ; it has a central
grass hill rising above the level of any of the neighbours,
and sloping precipitously to the water on the west side.
It is partly clothed with bracken, and the shores every-
where are steep and rocky ; there is a fair variety of
trees and, in contrast with Sege, abundance of ferns.
Waitwe, like Kiuwa, is pre-eminently a " rat island," the
tussocky grass being penetrated in all directions by their
runs. Perhaps on account of this a small python was
found on the island, in a cave.

Three Enswa-swa were seen ; of birds, the " Kunguvu "

^ Eucoma atricoata.

YEMPATA 117

flycatcher was present, and the familiar notes of a common
bulbul made one realize that it had not been found on
every island. Nephila was absent ; ten species of butter-
flies were noted ; Glossina very scarce, for the nine boys
only caught six in half an hour.

On February 24th we visited Yempata Isle, going first
to a minute accessory islet lying ofi its southern end.
This was noteworthy by the scarcity of the " Oluzibaziba "
bush ; there were some fig trees and, on the eastern side,
" Ekinsambwe " bushes. No ferns were noted, but
Ipomaea was abundant, and the whole centre of the islet
was covered with masses of a thick leaved aromatic,
labiate herb with spikes of purple flowers, and the mauve
flowered " Obutwatwa " {Erlangea tomentosa), in which,
as on other small islets, the sweet song of the
** Nightingale flycatcher " was heard.

Birds abounded on this islet ; on the rocks surrounding
it were cormorants, gulls, egrets and giant heron ; moor-
hens were among the bushes on the eastern shore, and
many swallows and weaver birds were noted. Nephila
and Glossina were absent ; the only butterfly was a
Lycaenid ; bees were seen on the flowers.

The main island of Yempata is rather flat, but there is a
small amount of open grass land rising slightly at about
the centre ; the rest is well covered with forest or the
overgrown remains of former plantations.

For at least half the length of the western shore there is
a pebbly beach, backed by dense thickets of elephant
grass (" Ekisagazi "), an unusual feature on any but the
largest islands. The remainder of the west coast has a
dense wall of bushes ; along the north coast there are
rushes and papyrus, and on the east side papyrus, ambatch,
and other bushes right up to the water's edge, forming
excellent shelter for the Situtunga antelope which the
• natives assert lives on the island. Along the southern
shore is pebbly beach.

118 A TOUR AMONG THE ISLANDS

The old banana plantations bore evidence that monkeys
live on Yempata, but they were not seen. Hippos, of
course, visited Yempata, for old plantations are favourite
resorts of theirs ; only one crocodile and one Enswa-swa
were seen. Nothing is recorded in my journal about the
birds of Yempata. Butterflies were numerous, but nothing
of particular interest was seen, and in one visit to an island
of this size no estimate could be made of their numbers.
Nephila spiders were not seen ; Glossina was extremely
abundant and as hungry as on Kimmi, and, as at Kimmi,
came off to the canoe before we reached the shore. The
average for the island was '67-6 per boy-hour; but for
the western beach only the figure was 79" 5 !

On February 28th I visited Sindiro alone ; an island
about the size of Bulago, with a low grassy hill at the
north end and a rather flat southern point, the shores
all round having a belt of very thick forest coming down
like a wall almost to the edge of the water. In front of
the forest are thickets of cane, e'tungulu and papyrus,
in places so dense that landing was impossible, aftd we
had to go round to the eastern side of the south point ;
here we found a hippo pathway up which we crept
ashore and found ourselves in a large open sandy space,
in which were numerous burrows of Enswa-swa, presumably
nesting places, as on Kimmi.

Sindiro proved to be very decidedly a " spider island,"
and Glossina was represented by an average of 9-2 per boy-
hour, in both of which respects this island shows interest-
ing differences from its neighbour Yempata. Cicadas
singing on the trees seemed to be more numerous than
had been noted on other islands ; of butterflies, the
Hesperidae seemed especially numerous in proportion to
the rest.

On March 2nd we moved camp, since food supplies were
running low and Wema was too far from Entebbe for
conveniently replenishing. »So, to my regret, we left

S 5

o 5

as CO

o a;

KIBIBI 119

beautiful Wema and returned to the former camp on
Kerenge, visiting Kibibi en route.

This island, about a mile square, was too big to be
thoroughly examined in the time at our disposal ; it has
a sugar-loaf, grassy hill rising to the usual height, and a
good deal of park land with trees scattered singly or in
clumps ; on the north point is some really beautiful old
forest, which is bordered at the water's edge by the tallest
thicket of Salvia that had been seen on the islands.
Crocodiles and Enswa-swa were both scarce on Kibibi.
Song thrushes were heard singing sweetly in some number,
and the fine blue and red plaintain-eater (" Fulungu ")
was noted ; this is usually not found on the smaller
islands, though one was seen on Kiuwa.

Only a single Nephila spider was seen. It was not
possible to estimate the butterflies, but it was interesting
to find a species of Byblia^ which had not been seen on
any other island, though it was found abundantly at
Jinja on the mainland in 1910.

On March 3rd, Kiuwa was again visited, and again
worked very thoroughly without success in finding
breeding grounds for the Glossina. I found three pupae,
which were more than any one else found.

On this visit was noted another butterfly which had not
been seen since 1910 at Jinja, a very beautiful milky
blue Lycaenid which was quite abundant. ^ Next day,
owing to supplies having run very low, we returned
westwards to a camp on the north shore of the large,
flat, forested island of Damba. The camp was pitched in
one of the few open spaces, and was memorable for the
myriads of mosquitoes.

En route we examined the Isle of Dyavodemu, lying
in a bay near the mainland, and I visited it again on the
6th. It is flat and without forest ; at one time it had been
much cultivated, but the shambas were quite overgrown.

^ See p. 109. ^ Castalius isis.

120 A TOUR AMONG THE ISLANDS

The shores generally are bordered by a thin growth of
bushes ; on the south side is a pretty bay backed by a
small cliff, at the foot of which bushes have apparently
been used as a nesting place by egrets for many years.
Birds were few on Dyavodemu ; no flycatchers, bee-eaters
or coucals were found, and especially noteworthy was the
complete absence of weavers, which had been found on
almost all islands examined hitherto.

Abundant on this island was a lizard which had not
been seen in such number before : a short-legged, grey
species, longitudinally striped, with a bluish tint at the
base of the tail.

Dyavodemu was emphatically a " spider island," not
only the large Nephila being abundant, but also a smaller,
reddish, thorny backed species with shorter legs being
much more abundant than on other islands where it had
been seen. Olossina was represented by an average of
10* 1 per boy-hour.

Butterflies were interesting (see p. 125). Papilio demo-
docus was seen ; Pierines were much more abundant in
individuals than on other isles. Catopsilia and a Belenois
were plentiful, and the latter could be seen in numbers
crossing over from the mainland. Only two Nymphalines
were found. Precis clelia and archesia ; Acraeines were
abundant in species as well as individuals ; the most
conspicuous, and so unusually plentiful as to be a feature
of the fauna, was the rosy pink A. natalica pseudegina ;
A. zetes was scarce, egina was not seen, encedon was in
some variety. Of Danaines, only D. chrysippus was met
with ; Lycaenids and Hesperids were few, both in species
and individuals. Other insular features were the great
numbers of Asilid flies, and of two species of beetles
(Cassididae). It is possible that abundance of the
predaceous flies, in the presence of multitudes of spiders,
corresponded with abundance of Mantidae on other
spider islands, for Mantidae were not found to be
numerous on Dyavodemu.

MTIOMU ISLET 121

The beetles were on the Ipornaea creeper, much of which
was stripped of leaves by the activities of them and their
larvae. One was brown in colour with rough surface ; ^
the other smooth light brown, with very variable and
irregular black spots, ^

On March 5th Fiske went to Entebbe on business, and
planned to meet me with supplies at the old camp on
Bulago on the 7th.

I went to visit the small islets lying off the coast of
Damba, close to the camp. The first, Mtiomtj, of spongy
brown rock, measures about 100 by 20 yards, and has a
fine sand spit at the southern end ; in the water at the
north end is a great grove of ambatch, which sheltered
abundant cormorants and weavers. For such a small
island the vegetation was extraordinarily varied ; except
for " Oluzibaziba " all the familiar bushes and flowering
plants of the lake shore were represented, so that several
species of butterflies might have been expected ; but
only T. telicanus, the cosmopolitan Lycaenid, was found.
Sunbirds and wagtails were seen, and particularly interest-
ing was the presence of the " Nightingale flycatcher." A
small brown lizard was very abundant. Nephila spiders
were absent, also Olossina.

Of Hymenoptera, several species of bees and Scoliidae were
seen, and also the black and yellow Sceliphron spirifex.

The insular feature of Mtiomu Islet was the great
abundance of two large, conspicuous, aposematic beetles ;
one a black and canary yellow species of Cetoniidae,^
evil smelling, and freely exposed on the flowers, the other
a Longicorn,* black with transverse red bands, which
fed on the green bark of sundry shrubs.

Both of these are common insects, but were never
found elsewhere in such numbers.

The adjoining Islet Tokwi is much smaller, consisting

^ Aspidomorpha near moufietti. ^ Aspidomorpha near signatipennis
^ Pachnoda sinuata. * Ceroplesis signata {Lamiidae),

122 A TOUR AMONG THE ISLANDS

merely of a few rocks and sand, with scanty vegetation.
Except for one Enswa-swa and weaver birds and the
" Nightingale flycatcher," there was little to note about
the fauna.

The tiny islet named Kawaga was visited next ; merely
a flat topped rock about ten yards in diameter, rising
sheer out of the water and about three feet above it.
All round the edge grows the " Omuvuvumye " shrub,
but nothing else save grass. Even this tiny islet had
its entomological feature ; for, feeding on the leaves of
the only kind of bush were great numbers of black larvae
of a small Chrysomelid ^ beetle, brown and black. More
interesting still was their complete absence from a pre-
cisely similar but slightly smaller rock about twenty
yards away, and from a third a little further away still.
I visited Mtiomu again, but could not find any of these
beetles or their larvae on the " Omuvuvumye " shrub there.

On March 6th I visited Masovwi, a diminutive neighbour
of Dyavodemu, which, measuring about 200 by 50 yards
and lying very low, is formed of grey rock and has a good
deal of sand on it. The southern shore is marshy, and
fringed for half its length with a dense thicket of ambatch,
the other half is open, with lush grass and flowers like a
small sunflower. The centre of the island is piled with sand,
on which grow thick bushes of " Ekinsambwe," and in
which a batch of crocodile's eggs were found, the crocodile
also being seen. On the whole, the vegetation was not
varied, and there was no " Oluzibaziba."

Birds were represented by a heron, many cormorants,
and moorhens, with weavers, sunbirds and, since the islet
was very small, the " Nightingale flycatcher." Nephila was
absent, and Glossina, rather to my surprise, was caught at
the rate of 2- 1 per boy-hour. Pupae were found, so that
Masovwi is the smallest isle on which Glossina is known to
breed.

* Mesoplatys ochroptera.

MOVING CAMP IN A STORM 123

On March 7th I moved camp to Bulago ; the dawn
was very stormy looking and there were spots of rain
until 9 a.m. It then seemed as if danger of a storm was
over, and I set ofi in the small canoe, heavily laden and
low in the water, but hardly had we got well away from
Damba when we were caught in a formal thunderstorm,
coming, fortunately, more or less from astern. The small
canoe plunged heavily and took a good deal of water over
the bows, and it was necessary to shift some of the load
further aft, rather a difficult procedure, as the canoe was
full and far from steady. However, after this she
rode better, and we proceeded on our way to Bulago,
getting, of course, thoroughly soaked with the heavy
rain.

In the evening there was great dissatisfaction among
the men on account of the small amount of food available,
each man's portion being only about half a dozen bananas.
While ruefully trying to make two and two into five,
to my great relief I saw Fiske's canoe in the distance
returning from Entebbe, and all was well, for he brought
abundant supplies and a huge mail.

March 8th and 9th were occupied with work on Bulago,
on March 10th we went to Tavu, and on the 11th to Kizima,
where I heard the " NightingaJe flycatcher," not previously
noted there.

The 12th- 15th were spent on Bulago, and on the 16th
we moved to our former camp on Nsadzi, visiting Kimmi
en route and working there again on the 17th. Ngamba
was re-visited on the 18th, and Nsadzi itself was worked
on the 19th.

The results of these second visits to the above islands
were most interesting, as the numbers of Glossina were
found to have greatly increased.

On March 21st I visited Rijmfua Island, which lies
off the south coast of Kome. A rocky, bush covered
isle, about half a mile long, but very narrow, the shore

124 A TOUR AMONG THE ISLANDS

formed of boulders, except for a shingle beach at the
northern end.

In the preliminary circumnavigation five Enswa-swa were
seen, but no crocodiles. Rumfua is decidedly a "spider
island," and the masses of their webs among the dense
bush prevented one from walking over it. A great many
hippo lairs were seen, so that the island seems to be a
favourite sleeping place for them. Birds were scarce ;
there were no weavers or sunbirds, and only one " Kung-
uvu " flycatcher was seen. A large flock of egrets was
noted on the shore. Egg capsules of Mantidae were
particularly noticeable, though the insects themselves
were not abundant at the time.

The only butterflies noted were three Lycaenids and a
Nymphaline, viz. the common T. telicanus and an equally
common Lycaenesthes, and the rare Aphnaeus orcas ; the
solitary Nymphaline was a fine Euralia.

On March 22nd, with Fiske's departure, our joint tour
came to ^n end. Apart from the results obtained bearing
on the bionomics of Glossina, many extremely interesting
facts had been noted regarding the predominance on
certain islands of one or another type of insect ; and it
was very interesting to see how, when the great spiders
abounded, insectivorous birds were absent and the " balance
of nature " was maintained by an abundance of Mantidae
or Asilidae keeping down insects which might otherwise
have been devoured by birds.

•aupiiadsaii

aBpiuaBD^I

•uamjCqil

•avmiA'\vs

•aBtniBqdttiX^

©fo looi— ii~i»a»o»o<MiOfo I— <'<^i«oo-H lo

fO'^eO^'MOOOOClCDlOOSlOt-lOOCMOO-^COOCO

i-iffO|iOi-<| I— (•<i<fO-*'-<rH| I |iO| I (<M

oO'-t |oeofooJOMMio?ooor-ir-(-!i<o-<*H iMcq-n

•ecmaBiov

n |(NO5t^»Of0^CD(MiMt^C0iO IICCTJ IfO |«0

•aBUi«UB(I

n I IMMMj— ItNM-^Oi-^C^I I IpHi^I'-Hi-h Ii— i

•aBouajj

T)*i— ic<>Tj(Oi-HeO(NTt<eococoe<5iM leoMi-HcHi-Hfo

•OBuraonidBj

■moi

U5IN ■^ThiMCO(MCOO-*CO'^-H (Mo^iM^F-HtN

'~' CD
O 73

is

s *

C8 ^ r"

f= -?o ^ ^ ^

bX) 3

(11 Co

* ce « S
5 R C > ®

2 ® § c3 faC

CHAPTER VII

MAMMALS

The Hippopotamus.

An animal of such large size as the hippo, called by the
Baganda " Envubu," must naturally come before one's
notice from time to time, but one really did not see much
of him. My first acquaintance with him dates from
Jinja in 1910, where a number could be seen in a large
pool about a couple of miles down the river below the
falls. The steep banks leading down to the river were
scored by the constant passage of these heavy animals
over the same track, until lanes had been worn in the soft
soil several feet deep, of the width of the hippo's body,
bearing an absurd resemblance to miniature Devonshire
lanes.

The bush on the banks was so thick that one could only
get down to the water by using a hippo track, and wondered
when so doing what a hippo would do if, when ascending,
we happened to meet. For he certainly could not have
turned. When on the islands one sometimes found the
tunnels made through dense growth of reeds quite useful
for getting about, though it was very fatiguing to walk
half doubled up, and the soft mud was often unpleasantly
odoriferous.

During the day time the hippo spends his time resting
in shallow water, or else sleeps in his favourite quarters,
for he seems to be very much attached to particular
places, and one would usually find a hippo off certain

126

NOISES— SLEEPING PLACES 127

low lying peninsulae, among the rushes. When passing
such resorts in a canoe the men always paddled very
quickly, but my experience was that an unmolested
hippo showed no resentment at being disturbed, and I
never had any trouble and never shot at them. On the
approach of the canoe the hippo would gradually sink
until only the top of the head, with ears and eyes and the
nostrils, were above water, when, in spite of the seeming
absurdity, there was a certain horsiness about him ! He
would watch the canoe suspiciously until it came too
close, and then disappear altogether, when the canoe-men
would paddle furiously until we were past what they
considered a danger zone. When a, frightened hippo
comes up to breathe he expires the used up air with only
a faint sound, but in the night their loud noises can be
heard some miles away,, and are probably used as
a call. It is impossible to describe the noise, neither
snorting nor grunting are completely descriptive ; it can
be imitated by making a noise at the back of the throat
while indrawing a breath.

These caUs at night have a fascination all their own ;
sometimes, however, could be heard other much louder and
less pleasing noises that seemed to indicate a battle royal,
perhaps between rival suitors.

Hippos have regular sleeping places in the forest or
bush on the islands, from which they may sometimes be
disturbed in the day time.

On Kimmi Island, when with Fiske one day, I heard
him say, '* Well, isn't this just cunning," and, coming up,
found h'm looking at the sleeping places of a mother and
baby hippo, where could plainly be seen the imprint of
head, body and legs on the dead leaves, showing how they
had slept side by side. The little rocky, bush covered
islet Rumfua, off the south coast of Kome Isle, was
evidently a favourite dormitory, for there were several bed
places among the thick bushes and ferns ; since there

128 MAMMALS

was nothing particularly tempting for them to eat they
must have gone there expressly for sleeping.

I once found among some reeds near the water's edge
a well marked bed place, with a mud wallow beside it ;
one was reminded of a house agent's advertisement,
" bath room adjoining " !

Hippos wander about over the smallest islands ; thus,
the central conical grass clad hill of Isentwa Isle, rising
steeply from the water to a height of a hundred feet or
more, had a definite track across the top ; but why they
should visit an island apparently so unprofitable from a
gastronomic point of view it is difficult to say.

The wealth of vegetation that springs up in deserted
banana plantations on the islands is very attractive to
these animals, whose tracks always lead to the plant-
ations. On a very sandy stretch of shore on Damba,
I saw tracks of a large hippo with a baby, leading from
the water into the forest, and quite close by were the same
tracks leading back into the water. It was quite obvious
that on the return journey the poor baby had been very
tired, for when he got on to the soft sand he had not lifted
his feet sufficiently to clear it, and a furrow between each
footprint showed how his weary toes had dragged over the
sand. However, he very soon got into the water for a
rest. It was quite possibly this same mother that I saw
one morning when on the way to the day's work on the
fly beach. As the canoe drew near we saw, lying off
shore, a large hippo lying on one flank in the water, with
the whole of the other flank exposed.

We naturally concluded it was a corpse, but as the
canoe came nearer there was much difference of opinion
among the men ; finally, when we came closer still, the
hippo proved to be very much alive, for she rolled over
and sank, and then we saw the heads of two babies also
in the water. None of us had ever before seen a hippo
take up this attitude in the water ; it seems possible
that she was giving her babies milk.

MOTHER AND BABY 129

One day in January 1919, I went across to Damba fly
beach from the east end of Kome Isle, but as a high wind
had raised a considerable sea, it was necessary to find a
sheltered spot for landing, and then walk round to the
exposed beach, where the noise of the waves and of the
wind in the belt of forest behind the beach drowned the
sound of footsteps. Consequently a hippo mother with
her baby that was sleeping in the forest did not take alarm
at first. I was standing on the beach, observing the
changes that had occurred since the last visit in 1914,
when suddenly a loud snorting was heard, and the huge
beast rushed past me into the water, closely followed by
the tiniest baby hippo I had ever seen. It bleated in a
most comical manner, rather reminding one of a toy,
and was so smaU that it was still quite weak, and stumbled
on to its knees several times before it reached the water.
Though so young it was not pink, but dark brownish
grey. I was so taken by surprise by their sudden
appearance that the chance of catching the baby was lost :
but it remained in shallow water just beyond the breaking
waves, putting up its nose at short intervals for a breath
of air. The mother was further away, evidently rather
anxiously watching the fate of her little one, and though
the boys with me wanted to go into the water and catch
the baby, I thought it better to say no, lest the mother
should attack.

This was the same part of Damba coast as that where
a mother was seen in the water apparently suckling her
young, as -described ; perhaps it is a favourite nursery.

One night on Damba I woke up hearing a curious noise
which, in my sleepy condition, I could not define at first.
Presently it resolved itself into the grinding of teeth and
I realized that a large hippo was grazing on the choice
grass just by the tent. As I was afraid he might blunder
into the ropes and bring down the tent, I went out with a
towel and waved it vigorously, and the huge beast moved

10

130 MAMMALS

off ; it was a dark night, and only a black shape could be
seen. Hippos are very fond of lush grass which grows
along low marshy or sandy stretches of shore, and these
were dubbed " grazing grounds," for the grass was always
kept short. A magnificent opportunity of photographing
a grazing hippo was lost because a camera was not to hand.
In company with the fly boys I was going down to the
morning's work on the fly beach of Bugalla. As we
emerged from the bordering strip of forest there was a
hippo on the sand, grazing quite openly on the grass.
I walked along the shore towards him to see how close he
would allow me to come ; he had his back towards me but
turned round at once, saw me, stared, and went on grazing.
I continued to approach, thinking it would be good fun
to pick up a stick and give him an unexpected blow from
behind. But his guardian angel was true to his duty,
and when I was within about thirty yards an ibis suddenly
took to wing with its raucous alarm cry, " Aa-aa-aa."
The hippo turned round in a fright, made an angry face
at me, and rushed into the water, leaving me lamenting.
I should like to conclude my remarks on the hippo by the
following. A new monthly journal was founded to deal
with Colonial interests, and a specimen copy of the first
number reached me on the islands. In it was an account
of a lecture given by some one on Sleeping Sickness at
a meeting of a photographic society, wherein were found
these astonishing statements. The lecturer said it had been
proved that the Tse-tse could not exist without a feed of
crocodile's blood (!), and that it was well known that the
hippo had an inveterate habit of destroying crocodiles'
eggs (or was he supposed to eat them ? My memory
fails on this point). Here, then, said the lecturer, is the
solution of the problem of Sleeping Sickness in a nutshell.
Preserve the hippo, and he does the rest ! One can
picture the establishment of " hippo hatcheries " all
round the lake, and schools where youngsters will be

MONKEYS 131

taught the difference in appearance between the eggs of
crocodiles and fowls !

But perhaps the lecturer was a humorist ? . . .

Monkeys.

One species of monkey, a Cercopithecus, lives on the
islands, but only on the largest : they were found on
Kome, Damba, Wema and Yempata out of those that
were visited in the group lying off the north shore of the
lake, and on Bugalla and the larger isles of the Sesse
group. A very curious fact is their presence on the Isle
of Nkosi, which is the southernmost of all the Sesse isles,
quite small and in a very isolated and exposed position ;
when I visited this islet in 1913 I distinctly heard a
monkey's voice, though he was not visible (see p. 92).
The species is of a common type : greenish grey in
colour, with black face and a white band across the fore-
head. It has been of very great service in the investigation
of Sleeping Sickness, since the Trypanosomes which cause
that disease are also pathogenic to the monkey, which
can thus be used as a test for infected flies. Hence I had
one with me on the islands for this purpose, and also
two others as pets. The " official " one was dubbed
Tommy, and, of course, had to be kept tied up, as also
was the older of the two pets, named Wee Man. The
third, Puffin, was a mere baby, and had not yet become
sufficiently mischievous for it to be necessary to tie him
up. When I first had Wee Man he was also allowed to
run loose, but became so mischievous that when I moved
to Bugalla he was tied up.

Being in very constant association with them, one soon
became thoroughly familiar with their behaviour, their
moods, facial expressions and language, and soon found
oneself able to distinguish the meanings of their various
utterances.

No less than fourteen sounds are used, although some

132 MAMMALS

of them would not -be recognized as distinct by any one
who had not carefully observed the monkeys, for they are
often of the nature of an alteration in inflection, or accent,
rather than a different word.

1. May be called a " General retnarh,^' and is the com-
monest expression ; used, seemingly, when a monkey is
at a loss what to do, or when his attention is attracted
by something, or when another monkey comes to him,
etc., etc.

2. Becognition. — A slight modification of No. 1 was used
by Wee Man whenever he caught sight of my servant,
of whom he seemed particularly fond. Often I would
hear this expression and, looking out, would see the man
walking about a little way away, and so in time came to
know what this sound meant, for it was never used on
any other occasion. It may be expressed by " Wok."

3. Eagerness. — Another modification of No. 1 indicated
eagerness ; as, for example, when the monkey saw some
one bringing a grasshopper which he particularly desired.

4. Alarm was expressed by still another modification of
No. 1, and had two forms :

{a) For a bird of prey overhead, very unmistakable

and emphatic,
(6) For a thunderstorm, or a bush fire, which they

dread very much.

5. Excitement, as when a monkey sees a boy chasing
a fowl, or two boys in play chasing one another. This is
also derived from No. 1. I have heard one of the monkeys
repeatedly make this noise when he saw a fish eagle chasing
another away from its private fishing ground. He took
the greatest interest in the occurrence, watching the eagles
as they soared around, and being unable to restrain his
excitement when one swooped down on the other.

6. Rage. — A quite unmistakable sound, possibly con-
nected with 5.

WEE MAN CHOOSING A GRASSHOPPER FROM THE BOX.
Puffin, behind, has had his.

WEE MAN (right) AND TOMMY TAKING GRASSHOPPERS.

To (ace p. 132,

MONKEY LANGUAGE 133

7. Pain. — A kind of squeak.

8. Cry for help. — A, high-pitched squeak. Wee Man used
to make this as a youngster when he became inextricably
entangled in his rope.

9. Melancholy. — A very distinct, long-drawn wail,
sometimes heard in the forest, presumably indicating that
the monkey had become separated from the troop. A
monkey seen in the act of crying thus has a most lugu-
brious appearance ; the mouth is held in a peculiar
fashion and one quite expects to see tears rolling down
the cheeks.

10. Hunting Call. — One of the most distinct sounds in
the monkey tongue. When a troop is searching the trees
for food in the forest an old male sits in a very conspicuous
tree top and utters a series of barking noises which can be
most nearly imitated by repeating very rapidly " Kubba-
kubba-kubba." To this the junior members reply by
high-pitched squeaks, and the whole troop is thus enabled
to keep together, as the total amount of noise produced
is considerable.

11. Dislike. — A short, expressive word, which may be
represented by saying beneath, one's breath the first two
letters of the word " come."

12. Intense dislike and fear. — The last mentioned sound
is repeated very rapidly and with great energy when a
monkey sees a snake, or anything that seems to savour of
a reptile.

13. A *' hahy " noise, only made by young monkeys
when they have been frightened or hurt and run for comfort
to be cuddled by their mothers or friends. It can be
represented by the noise " Qurra-qurra-qurra-qurra "
repeated beneath one's breath.

, Having been for months in close association with the
pet monkeys, I found their different natures extremely
interesting, and they were most charming companions,

134 MAMMALS

but it was necessary to keep them tied up owing to their
mischievous dispositions.

At one time, when very young, Wee Man had been
allowed to run loose, and whenever I left the house he
would be unable to resist the temptation of working havoc
among papers or anything else that took his fancy which
he could destroy, although he knew quite well he would
be punished, and showed it by his guilty demeanour when
I returned. At this time he would frolic around the
huts of the canoe-men when they were away, and make
little holes through the grass walls, so that he could, if chased
when the time came to tie him up for the night, dodge
in and out of the huts. It was a great game every evening
to try and find him, and if he managed to hide himself
away and go to sleep before he was found, it was considered
that he had won the game ; and he did, quite often !
Mischief was a very marked feature of the monkeys' games
with each other, which were delightful to watch. One
would be sitting on the bar, a couple of feet above the
ground, his tail hanging down, and another, with a
broad grin on his face, would steal up and, seizing the tail
with both hands, tug furiously in the endeavour to bring
down the other, who, quite appreciating the joke and also
grinning, would cling on with all his might.

They were of a very affectionate disposition, and if
one was beaten or threatened the other would do its
utmost to bite the enemy, shrieking with rage the while.
The baby Puffin, on Bugalla, was not tied up, and
occasionally got into mischief in the house, when I
chased him out with threatening gestures. He would
at once run to the others, who would receive him
anxiously and carefully examine him to see what damage
had been sustained, while he made the baby noise to which
they responded.

When the three settled down to sleep for the night,
all cuddled together, they would give each other a long
kiss, lip to lip !

WEE MAN BEING RUDE TO A TORTOISE.

TOMMY (left) and WEE MAN IN AN INTERVAL OF A GAME.

To face p. 13).

MONKEY AND TORTOISE 135

But their affection for each other never went to the
lengths of unselfishness as regards food ; if choice were
given, a monkey would of course take the largest piece,
but very often afterwards would drop that and endeavour
to take from the other the piece which he had himself
passed over. In fact, their maxim seems to be " the
other's piece is always better than mine " !

Wee Man was the most intelligent of the three, and had
a more capable-looking head. When his rope became
very much entangled with that of Tommy, he would look
at the tangle for a while and then deliberately try to
disentangle it by picking up a loop and, as it were, un-
threading himself by walking through it. Often, if the
tangle was a simple one, he would succeed, but if the first
method was not successful he would try walking through
the loop in a reverse direction. Tommy, however, though
he might make futile attempts, never succeeded, and seemed
to do it mechanically because he had seen Wee Man do
it. This is the only instance I ever saw of the reputed
imitativeness of a monkey. A certain intelligence was
also shown in the recognition of zoological affinity. The
noise made for a snake has already been described, and
the same was also made for a fish ; and, one day, having
found a tortoise, I brought it back to see what Wee Man
would say to it.

At first the tortoise withdrew itself within the shell :
Wee Man came down to look very cautiously at it,
intensely curious and interested, but when it put its head
out he hastily retreated to his perch with loud utterance
of the snake noise. When it began to walk he could not
restrain his excitement, and danced round and round it,
chattering, though keeping at a safe distance. I am quite
sure that this was the first tortoise he had ever met, for
he had been a captive ever since, as a baby, he was clinging
to his mother.

On Bugalla Island, where there was much open grass

136 MAMMALS

land as well as clumps of bush and continuous forest,
when walking in the evening I often came across a troop
of monkeys among the long grass hunting for grasshoppers,
which are a great delicacy. A monkey would walk slowly
and intently through the grass until a grasshopper took
to flight just in front of him, when with a quick snatch
he would catch it in his hands. The monkeys always bit
off the head first, and then the powerful hind legs, if the
grasshopper kicked much, and the rest was eaten at leisure.
Even the largest Acrididae, three or four inches long, with
formidable spines on the hind legs, were eaten with gusto.
It always made me think of a man eating a live lobster,
shell, legs and all, and not one only, but half a dozen in
succession.

I once saw a small young monkey in a patch of papyrus
reed, apparently hunting for insects. He climbed up one
of the tall stems, but just as he reached the top it slowly
bent over with his weight and he disappeared from view.
Fortunately for him the papyrus was not growing in
water.

In Damba forest I came across a troop, on the
ground, very busily engaged in turning over dead leaves,
looking for insects. So pre-occupied were they that I was
enabled to creep up quite close and watch them before
one looked up and gave the alarm.

The captive monkeys were always ready to eat insects,
and if a species was offered of which they were afra d,
such as a large bee or powerful ground beetle, they would
paw it on the ground with rapid strokes of one hand after
another until it was disabled, when it would be quickly
picked up and nipped between the teeth, to be subsequently
eaten at leisure, if desirable.

One evening on Bugalla I surprised a troop hunting in
the long grass some little distance away from the nearest
trees, to which they hastily retreated. A tiny youngster
who had been put down by his mother was unable to

A BABY MONKEY 137

escape and took refuge in a small bush, where I easily
caught him — a dear little fluffy beast, all head and tail,
that sat easily on the palm of my hand. Unfortunately
for him the bush was occupied by a species of very power-
fully biting ant, and when I got him the poor little monkey
was being severely bitten by the ants deeply embedded
in his fur. So I took him on my knee and soothed him
while I picked out the ants with forceps, the youngster
sitting quietly as if my attempts at the " cuddling noise "
were quite intelligible. His parents and friends on the
neighbouring trees were naturally much alarmed for his
safety, and danced up and down in impotent rage, with
shrieks of defiance that could plainly be interpreted as
*' Hurt him if you dare ! " When I had freed him from
the ants and smoothed him down, I showed him his friends
on the trees and put him down in the grass ; he ran off
more or less in the right direction, and I could see him
thoroughly overhauled by his parents, anxious to see
what damage had been inflicted, and then conveyed by
the whole party into the forest. These very young
monkeys are carried about beneath the belly of the mother,
their limbs embracing her body, and hands and feet firmly
grasping the hair on her back.

The Situtxjnga.

The only antelope on the islands which I visited is the
Situtunga {Limnotragus or Tragelaphus speJcei), known to
the natives of the islands as " Enjobe." This interesting
buck was discovered by J. H. Speke in 1862 in the Karagwe
district, south of the Kagera river. He gives it the native
name of " Nzoe " and says of it : "It proved to be closely
allied to a water-boc found by Livingstone on the Ngami
Lake, but instead of being striped was very faintly
spotted, and so long were its toes it could hardly walk on
the dry ground ; whilst its coat, also well adapted to the
moist element it lived in, was long, and of such excellent

138 MAMMALS

quality that the natives prize it for wearing almost more
then any other of the antelope tribe. The only food it
would eat were the tops of the tall papyrus rushes ; but
though it ate and drank freely, and lay down very
quietly, it always charged with ferocity any person who
went near it."

The Enjobe is about the size of a donkey. The male has
beautiful horns, long and spirally twisted, with almost
translucent tips ; the females are hornless. The colour
varies considerably ; the young males are bright foxy
red, marked with white, adult males are darker brown,
but I have seen an adult female of the same colour,
although typically they are more brightly coloured like
the young male.

This antelope with its elongate hooves is well known to
haunt swamps for which its deeply cleft feet are suitable,
in the same manner as are those of the reindeer for snow
or of the camel for sand. «

It is mentioned in books as spending its days immersed
in water among papyrus, etc., and is apparently considered
as an animal confined to swamps. On the islands, however,
it is a creature of different habits, possibly owing to the
entire absence of enemies, human or otherwise, from which
on the mainland it takes refuge in swamps. Indeed, it now
behaves much like its close ally the bushbuck, called
" Engabi " by the natives. On some of the islands which
it inhabits there are no papyrus swamps, and during the
day time I often disturbed Enjobe from their resting
places in the forest belt near the water.

In the evening they could be seen coming out on to the
open grass covered spaces and browsing on bushes,
especially " Oluzibaziba," the Euphorbiaceous Alchornea
chordata, of which they eat the young tips.

The bushbuck may well have had similar habits on the
larger islands in former days until it was exterminated.
The Situtunga, on the other hand, could have taken refuge

SITUTUNGA 139

in the water, as it now does on the mainland. Since the
removal of the islanders, however, which was completed
in 1909, its habits have become changed, and it is multi-
plying very quickly in the absence of all enemies.

Major R. Meinertzhagen, F.Z.S., visited Bugalla from a
steamer during the war in October 1915, and examined
several Situtunga, finding interesting differences from the
mainland type ; the horns diverged more widely. But
the specimens from the much smaller and more isolated
islet of Nkosi showed further deviation to such an extent
that they are given sub-specific rank.^

The feet show a slight return from the specialized
narrow hoof of the type, for in the Nkosi animals the
hooves are proportionately broader, and in a young animal
were " no longer than one would expect to find in a young
bushbuck of his age."

The horns also are more like those of a bushbuck, and
the colour of the adult male was noted as " of a uniform
dull mouse colour and not a dark brown " ; the animals
are also larger than the typical mainland forms.

It is remarkably interesting that this change in structure
can be noted together with the change of habits ; the
Nkosi animals are losing the high specialization, correlated
with the peculiar mode of life among swamps, which
characterizes the Situtunga discovered by Speke.

The Situtunga could frequently be heard barking at
night, the call being somewhat like that of the bushbuck
but distinct from it, rather more muffled and less dog-
like.

The following may be cited as evidence of the effect of
the removal of all inhabitants from the islands.

Nsadzi, the long narrow isle lying due south of Entebbe,
was free from this antelope when inhabited, according to
native testimony, although the densely forested eastern
end provides ample shelter. When I visited it in 1914 with

* Proceedings oj the Zoological Society, June, 1916, pp. 375-81.

140 - MAMMALS

Fiske he found footprints, and later on I found abundant
evidence that Situtunga at least visit Nsadzi. I have
also found bushes browsed by them on Ngamba, between
Nsadzi and Kome, on which latter large island they abound,
so that it is probable they had crossed to Nsadzi from
Kome, using Ngamba as a stepping stone, for it is well
known that they swim readily. The importance of these
movements from the point of view of Sleeping Sick-
ness has already been discussed in Chapter II, pp.
28-29.

The smallest island on which I have seen Enjobe is the
previously mentioned Nkosi, lying to the extreme south
of the Sesse archipelago, and very exposed. Several were
seen there on a visit (see p. 91), and areas of grass
were kept close cropped by their grazing. It seems
almost certain that they must swim to and from this
tiny island from the larger isles to the north.

Otters.

Otters abound in the lake, as might be expected, and
one is glad to feel that their numbers are steadily
increasing now that they can no longer be destroyed for
their skins by the islanders. There can be few animals
more beautiful than otters in the water ; they are so
lithe and graceful and sleek, and so full of the joie de
vivre. I remember well one very hot day when I was
visiting the little islet Kizima. Looking down from the
top of a little cliff into the water I saw below some half
dozen otters sporting in the water and tying themselves
into knots, but in a very leisurely way, as if it was too hot
for much exertion. On another occasion off Tavu Isle
there were about a dozen, fishing and pliaying with their
catches in a very delightful sportive mood ; throwing
the fish up into the air and catching it again. One lay
on his back in the water in a most ludicrous attitude,

OTTERS— LEOPARD 141

with a fish in his mouth and hind feet and tail sticking
up out of the water.

The size of fish that otters can tackle is very surprising.
On Ziro Isle, in 1919, I saw the remains of a very large
Silurid fish, which had evidently been dragged up on to a
rock high above water by one or more otters which had
eaten all except the shoulders and the large flat bony head,
measuring 12 by 9 inches. This part that remained was
taken to camp, and was found to weigh twenty pounds !
So the fish, which the natives called " Akasonzi," would
probably have weighed forty pounds at least. It was
the largest one the natives had ever seen.

The otters in Kingsley's Water Babies are described as
making a queer chorus of squealing noises, but I never
heard a sound that could be ascribed to an otter, even
at night, save when one comes up to breathe close by,
and emits a short grunt before diving again.

Leopard.

The larger carnivora were not met with on the islands
save for one exception. On Kibibi, in January 1919, fresh
spoor of a leopard was seen on the wet sand on two succes-
sive visits nearly a week apart. The leopard must have
swum from the coast of the mainland across at least 2| miles
of open water. (See map.) The spoor was seen very fresh
on both January 17th and 22nd, there having been heavy
rain between the two visits. So it would seem that the
animal was there all the time, although there cannot have
been much food for it. On Kibibi there are neither buck
nor monkeys, and there are no game birds (save an
occasional bustard, quail, or forest francolin) on any island
that I have visited. Presumably the leopard managed to
catch fish or ate crabs that it found on the beach !

Mongoose.
Animals of the civet cat tribe ( Viverridae) were not often

142 MAMMALS

met with on the islands, and then only on the very largest.
A pair of genets was once seen on Kome, their long snaky
bodies, with spotted coats, and short legs being quite
unmistakable, even though they were only seen darting
away in alarm.

On Damba one day in 1911,1 had a delightful experience.
I was sitting quietly by a clump of bushes, watching for
bee-eaters at the edge of the forest, some 30 yards away,
and keeping as motionless as possible. A beautiful
mongoose, of a rich red-brownish green colour, with tail
all fluffed out, walked across a little grassy bay in the edge
of the forest, very full of vigour, but not in a hurry, and
disappeared again among the bushes. Shortly afterwards
there was heard an agitated quacking and cackling from
the jungle on my left, and an Egyptian goose came across
the grass in front of me, and only a few yards away,
followed by the mongoose.

The goose fluttered along making a great fuss, apparently
just keeping out of reach of the pursuer^ It seemed to
me, however, that the mongoose was not exerting himself
very much, as if he thought that the obviously disabled
bird in front must very soon fall an easy prey. The two
passed into the jungle on my right, and the deluded
mongoose lost his prey. Shortly afterwards I saw him
again, walking across the little bay, but the ruse had
been successful ; the goose had drawn him away from a
brood of babies, and while he had been chasing one parent
the other had led the goslings away in another direction,
and I could see them swimming away to a marshy bit of
land where they were safe for the time. But a few days
later I saw that the number of goslings was much
diminished, so the mongoose may have been more success-
ful a second time, knowing the trick that had been played
on him.

It was very interesting to see this old, old trick actually
succeed.

A FAVOURITE HAUNT OF SITUTUNGA ON BUGALLA.

'They feed in the evening on the bushes at the edge of the forest belt which slopes steeply to

the water.

By permission of the Royal Society and the Controller, H.M. Stationery Office.

A BAT CAUGHT ON TENT AT JINJA.

To face p. ■.f?.

BATS 143

Bats.

Bats abound among the dense vegetation of the old
banana plantations, and may often be seen in the day
time hanging from the midrib of a banana leaf. The
caves found on Kimmi, Damba and other isles shelter
numbers of the smaller species, named " Akawundo " by
the natives. I shot a few of these one morning in order
to investigate the contents of their stomachs, but although
it was only about 9 o'clock, their stomachs were quite
empty. The noise reverberating in the cave was so
distasteful to the bats that for a long while they did not
return ; possibly their instinct warned them that such
a noise could only, in the experience of the race, have
been caused by rock falls making the cave unsafe.

I had a curious experience at Jinja one evening in 1910.
Hearing an animal scrambling about over the roof of the
tent and being unable to drive it away by repeatedly
striking the canvas, I went out with a lantern and found to
my surprise that it was not a mouse but a bat, whose
photograph is reproduced (p. 142). It was a more active
walker than others that I have encountered, and seemed
quite unwilling to fly. It had possibly been crawling
about on the roof after insects that were attracted to the
light.

The much larger frugivorous bats, commonly called
" Flying Foxes," and by the natives " Ekinyira,"
occasionally gathered together in great numbers, possibly
for breeding purposes. This was first noticed on Bunyama
in January 1912, and, later, on Ngamba in March and
April 1914.

They assemble among dense creepers, and may be heard
chattering, and also smelt, before a landing is made.

When disturbed they made a prodigious noise, scrambling
about and squeaking loudly before taking to flight and
circling round looking for a new resting place.

144 MAMMALS

Although I visited Ngamba every week between March
and August 1914, I did not find these bats congregated
after April.

Since the expanse of wings is about two feet, they look
large creatures in flight, and the beating of the wings
makes a considerable noise when one disturbs a large
congregation.

At night the " Ekinyira " in the forest keeps up a most
insistent loud noise at intervals of about a second. For a
long while I thought it was produced by a large frog.
It is a sharp note, between a high-pitched croak and a
grunt, with something of a bark in it. It may be heard
from shortly after sunset for hours on end, and, if one
awakes in the middle of the night, may still be heard
coming from the same spot !

CHAPTER VIII

BIRDS

Ornithology is a subject to which I have devoted less
time and observation than entomology, so that I have not
so much to say about birds because of my ignorance of
their identity.

In writing this chapter I have been much indebted to
Mr. C. F. Belcher, of Uganda, who has kindly given me
the scientific names of such species as are dealt with,
and also some of the native names which I had not found
out.

The arrangement and nomenclature is according to
Reichenow,

Perhaps the most striking ornithological feature of the
islands is the difference of their fauna from that of the
adjoining coast of the mainland.

I am, of course, only speaking of the islands that I
know well, those lying parallel to the north coast of the
lake between Entebbe and Jinja, and Bugalla in the
north-west corner of the lake. The differences that have
been noted will be mentioned as the various groups are
alluded to.

Gulls are some of the first birds noticed by the traveller
on the lake, and while I was on Damba Island in 1911
I saw a good deal of them. Some low rocks in the water
lay off the point on which I was camped, and for some
reason these were a favourite resort of gulls and many
other birds, so that there was a bigger collection there
than at any other place I know among the islands.

11 i«

146 BIRDS

At least two species of gull (Larus fuscus and cirrho-
cephalus), one, larger, with black back, the other grey,
were always chattering and screaming there, together
with terns, cormorants, darters, pratincoles, sandpipers,
goliath herons and other herons, amongst them a
magnificent snow white species seldom seen, but there in
some numbers (? Herodias alba), white egrets and the
white egret-like buff backed heron, the open-bill and
another stork, Egyptian geese, and sometimes ibises,
either the black and white sacred species or the dark
greenish black Hagedash ibis.

The gulls are much more noisy before stormy weather,
and were quite useful as weather prophets. It seemed
to me, also, that one did not see much of the terns
except in windy weather, when they were noticeable flying
about close to the land. Gulls are known to the natives
as " Enkobyo-kobyo," and terns {Hydrochelidon leucoptera)
by a name that sounded like " Akalerwe," but I would
not be certain of its correctness. Two species were noted.
It often seemed curious to me that I never met with the
nests of gulls on the coasts of the islands ; possibly they
do not breed close to the water.

They feed on dead fish and other flotsam on the surface
of the lake, but one has been seen in close attention on a
number of cormorants that were fishing in shallow water.
When one came up to the surface with a wriggling fish
in its mouth the gull swooped down upon it, and in self
defence the cormorant had to dive again with the fish
unswallowed. This haiDpened repeatedly until the passing
of the canoe disturbed the birds, so that I could not see
whether the gull had made the cormorant give up its
prize.

Cormorants naturally abound among the islands, and
soon become familiar sights. One is much puzzled by the
variety of markings of the larger birds (Phalacrocorax
lucidus). They may be black without any white, or with

CORMORANTS 147

snow white throat and breast, with or without a round
white patch on the side just behind the angle of the
wing.

A smaller species (Phalacrocorax africanus) is all black,
and nearly always very ragged looking ; indeed, one cannot
feel any affection for the cormorants, they are smelly
and unattractive. Natives call the smallest specimens
" Semirindi," the larger ones, without much white,
" Ensogwe," and the largest white breasted specimens are
named " Engadala," but most natives class them all,
together with the darters, as " Ensogwe."

These birds have their regular " rookeries " where they
breed year after year, so that the ambatch trees on which
they continually perch and build their loose nests of
sticks become flattened down and often look very unhappy.
The birds when sitting on the nests have a curious habit
of rapidly moving in and out the loose skin of the " chin,"
in somewhat the same way as a fowl when gasping for
breath in very hot weather. It is possibly due to the
same cause in both cases, as the cormorants sit on the
nests exposed to the full blaze of the sun.

I first noticed this when watching the nesting birds on
the islands of the Nile at Jinja, below the falls. Cor-
morants abound here, and swim fearlessly about amidst
the turbulent eddies immediately below the actual cataract,
where one would think they must be beaten against the
rocks.

About sunset large numbers of these birds return from
the lake, flying up the narrow gulf until they reach the
falls where the gulf becomes the river. Here, in 1910,
I often noticed a very curious trick of flight.

When the leading bird of the flock arrived over the
falls it would suddenly close the right wing so that it
fell down sideways for some twenty feet or so, to recover
with perfect ease and resume its former steady flight.
Often several would do this in succession, as if playing

148 BIRDS

foUow-my-leader, but I have never seen this trick played
anywhere else.

Occasionally a single bird winging its way home utters
a rhythmic croak at intervals, strongly suggesting a man
loudly panting for breath. They are not exactly songsters ;
their noises are more of the nature of quackings than
anything else. Except for the trick above mentioned one
does not see much sign of the joie de vivre in cormorants ;
they seem to find life a serious business, and have to work
hard for a living. When resting, and possibly digesting,
on the branches of their favourite trees, they often spread
their wings out on each side of the body, as if they felt
the heat, or were holding them out to dry, although so
oily are their feathers that they cannot really get wet.
It always amused the canoe-men, v/hen we were going
along, to see one of these birds suddenly appear on the
surface, often with a wriggling fish in its beak, hastily
look round, and either dive again or take to fiight. Being
heavy birds, they have to paddle their feet along the
surface for some distance before their wings have developed
enough momentum to lift them up.

Closely allied to the cormorants, and grouped with
them by the natives under the name " Ensogwe," are the
snaky-necked Darters {Anhinga rufa), often erroneously
called Divers, which name belongs properly to the
marine Colymbi.

These are not nearly so abundant as the cormorants,
from which they are easily distinguished by the much
smaller head and very long slender neck, bent upon itself
with a curious kink. Like the cormorants, these birds
take life very seriously, and are equally unattractive,
though their quaintness is interesting. They are certainly
most efficient water birds and fly well, but are ungraceful
and malodorou — at any rate in quantity. They are,
however, of somewhat less dingy plumage, and in the
breeding season become almost brightly coloured with
yellowish and brown tints.

GEESE 149

Pelicans have only been seen in the narrow channel
between Bugalla Island and Bukakata port on the Buddu
coast of the west of the lake.

They may also be seen in Kavirondo gulf, which is very
shallow and muddy, but why they do not occur elsewhere
on the sheltered waters among the islands it is difficult
to explain.

Geese are, of course, numerous on Lake Victoria, but
I confess not so numerous as I expected to find them,
and ducks were very rarely seen. I never saw on Lake
Victoria anything approaching the hosts of brown ducks
seen in 1916 on Lake Bunyonyi in the Kigezi district of
the south-west corner of Uganda, or in 1918 on Singidda
Lake at the southern end of the East African rift valley,
or on the marshy land along the central railway of ex-
German East Africa, near Dodoma, in 1917. The native
name " Embata " applies to both geese and ducks. The
Egyptian, or Zambesi {Chenalopex oegypfiacus) goose, well
known over so large a part of Africa, and on ornamental
waters at home, is the species most often met with, and is
usually found in little companies of two or three, or some-
times half a dozen. It prefers especially an open shore of
flat rock or sandy beach— the white sand of Nsadzi beach
was a very favoured spot, and one could always reckon on
finding at least a pair there.

They are handsome birds with their mottled brown and
grey plumage, and green and white speculum on the wings,
and walk easily, carrying themselves well without waddling.
As one moves along the beach towards them their agitation
shows itself by loud cacklings, until they finally take to
flight and make a half circle over the water, return ng to
shore a little further away. A wounded one that had
been shot and was pursued by a canoe endeavoured to
save itself by diving, but I think this bird only does so
in its utmost extremity.

A nest was found on July 25, 1914, at the foot of a tree

150 BIRDS

behind a bank of bushes and reeds on Kimmi Island ;
excavated in a bed of dead dry leaves and beautifully
lined with warm down, it contained four white eggs.
About a week later, however, I found only two eggs in
the nest, so that possibly some had been devoured by
Varanus or a snake.

I have related elsewhere how a mongoose that ardently
desired a gosling was frustrated by an old, old trick
played upon it by one of the parents.

The large black and white spur- winged goose .{Plectrop-
terus gamhensis) is not so plentiful as the Egyptian goose,
but is a much better bird for the table. I saw it so seldom
that I have no notes about it.

Similarly with the brightly coloured dainty little pigmy
goose (Nettopus auritus), which is found sparsely along
rush fringed stretches of shore, such as the north coast
of Bugalla and the channel between Damba and Kome.

Rather curious little birds are Pratincoles {Galacto-
chrysea emini), which may be seen sitting on small rock
islets. They seem very sluggish birds, and are not very
interesting or attractive. Short legged, with rather slender
bodies of grey colour, the rounded heads bearing a short
beak, the pratincole is called " Akasalu " by the natives.

Of the Plovers and their allies the most noticeable species
on the islands is a stone curlew {Oedicnemus vermiculatus),
known to the natives as " Mutunwa." Almost every
beach of gravel or sand or flat rocky shore has a pair or
more of these birds, which lay their eggs there, placing
the two in a slight hollow, without any attempt at a nest.

If one interferes with eggs or young they are apparently
removed by the parents. I first noticed this on Damba
beach, where I found a pair of eggs lying on the shingle
and examined them ; but next day they had disappeared.
Again, on Ngamba, I came across a young bird, with the
other just beginning to make its way through the shell.
I helped it out, a poor little bedraggled feebly chirping

NEST OF EGYPTIAN GOOSE UNDER TREE.

NEST OF STO\E CURLEW [OEDICNEMUS).
Two e^gs lie in front of the stick Ij'ing obliquely across the left centre of the picture.

To face p. 150,

STONE CURLEW

151

object, and left it there ; but a few days later both had
disappeared. It is, of course, possible that they had
really been eaten by a Varanus lizard ; I have elsewhere
described how one seized a bird I had shot. The Mutunwa
has a most curious habit of, as it were, shrugging its
shoulders, a habit shared by many of the kingfishers, of
which the meaning is quite obscure. They are certainly
decidedly procryptic birds, and this unnecessary move-
ment helps to reveal them when they might otherwise
escape notice.

Their characteristic cry is often heard at night, especially
in bright moonlight. It begins with a single high-
pitched note, sharply accentuated, and then a series
beginning lower than the first, rising to the same level
and falling again — a very mournful cadence. It may be
represented graphically thus ;

llliliiu

One stone curlew that I shot on Nsadzi in order to in-
vestigate its stomach contents had eaten a cricket and
apparently a smaU frog, so far as I could teU from the
slender bones that were all that was left.

The common Sandpiper (Tringoides hypoleucos) is
abundant on beaches or flat rocks, very often singly,
sometimes in small flocks.

It may be seen at the Ripon falls paddling about in
the water at the very edge of the cascade, occasionally
having to leap away from an unusually large patter of
spray or a rush of water. Possibly it feeds upon the
Simulium larvae which adhere to the rocks in abundance
in such localities.

152 BIRDS

I once saw on Damba some extraordinarily long legged
birds that presumably were Stilts, but they are far from
common. The locality was a very marshy piece of shore,
also frequented by snipe.

On Kome, in 1914, were seen some plovers which had
not been met with previously, nor have I seen them since,
although there are large areas of marshy shore similar to
the locality where they were seen.

The colouration was roughly as follows : Back grey,
belly white, chest and back of neck black, face and front
of neck white. When flying the bird shows black and white
wings. Subsequently, in ex- German East Africa, a bird
of something the same type was seen commonly on open
grassy plains. This was termed " Kibiki " by South
Africans.

Another species of the plover group was heard flying
overhead at night time on Kome. Its cry became very
familiar, and was apparently responsible for the name
" Empunya " given to the bird by the natives. Since
the meaning of this word would be something connected
with the word for " to smell, or emit an odour," it is
probably merely onomatopoeic, and is quite descriptive of
the cry.

Cranes are represented in Uganda by one species of
crowned crane {Balearica regulorum), a most orna-
mental and graceful bird which would be a splendid
addition to the peacocks and pheasants often kept to
ornament a park. The soft brown and grey and yellow
hues, with velvety black head furnished with scarlet
wattles and erectile crest, make it a most lovely bird,
and its quaint antics when with others of its kind give it
additional interest.

Unfortunately they are very seldom met with on the
smaller islands, and even on Kome and Bugalla are rare ;
and I never saw a flock such as may be seen at Entebbe
or Jinja. Their cry has earned for them the Luganda

BUSTARDS— LILY TROTTER— IBIS 153

name "E'ngali." The booming noise which they make at
night, before rain, has been already alluded to, and the
curious ignorance of the natives, who assign the noise to
the puff adder !

Next come the Bustards, of which a single species
{Otis melanog aster) known as " Olunyonkante " occurs on
islands where -there is a sufficiently large open grassy
area. This species shows black and white wings when
flying, and has a most ridiculous little call note. One hears
first a very short low whistle, followed after a few seconds
by a single sharp but very quiet note like a bubble
bursting. The peculiar native name would seem to indicate
that the native considers it to suck cows, so that this bird
holds a place in folklore analogous to our " goat sucker "
in England.

Birds of the Rail family were plentiful among the reeds
and rushes of the lake shore. One, looking much like
our English moorhen, was responsible for making a most
extraordinary noise among thick cover — a sort of grunting,
squealing and whining, quite indescribable and very
unlike a bird noise. The pretty brown, yellow and black
Lily trotter may conveniently be placed here, " Akatassa "
(Actophilus africanus). It is seen where there are enough
water lily leaves or thick vegetation in the water for it
to run over, its very long toes supporting it by distributing
its weight over a large area. It is a very quiet bird, and
one rarely heard it make any kind of noise unless it was
angrily scolding at and driving away a trespasser on its
private hunting ground.

We next come to the Ibises, of which I only know two
species on the islands, the most conspicuous being the
dark greenish black Hagedash ibis, known to the natives
as " Empavana," the word being an attempt to reproduce
its very characteristic cry, which can be indicated by
" Aa-aa-aa." This is a very annoying bird if one is
anxious to approach an animal without being noticed,

154 BIRDS

for it is very nervous and quickly takes to flight with noisy
beating of wings and raucous lamenting cry, frightening
everything within earshot, so that it must be a great
nuisance to sportsmen. I have already related how one
of them frustrated my hoped for interview with a hippo.
Sometimes these birds perch on a tree top, at other times
sit among the rushes and bushes on the shore, but wher-
ever they are their wailing cry betrays them. When
flushed they make a great noise with the hurried strokes
of their wings and have a characteristic flight, interposing
an extra quick beat here and there until they feel steady
in the air. Presumably the main object is to rise quickly
from the ground, and not to place a great horizontal
distance between themselves and the disturber of the
peace.

The other species is the handsome sacred ibis {Ibis
oethiopica), whose plumage, half black, half white, makes
it easy to recognize. It is very far from common, and is
only to be seen occasionally here and there, so that it is
unfamiliar to the natives, and I could obtain no native
name for it. During 1914 there was usually to be seen
one on a rock o& Kizima Islet, and it was noted to be a
very silent bird, thus contrasting strangely with its
relative. I did not notice the hurried flight, with inter-
posed beat, so characteristic of the Hagedash ibis.

Of Storks, several species are noticeable, particularly the
curious " open-billed " species {Anastomus lamelligerus),
the two halves of whose beak do not meet along the middle
of the bill, giving the bird an unusual appearance. It is
rather an uncouth and ungraceful black bird, met with
along the shore, where it is said to feed on the large
Ampullaria " water snails," its curious bill being adapted
to enable it to deal with their strong shells. The natives
call it " Enkonamasonko," which shows that they know
its habits, " E'sonko " (plural Masonko) meaning " shell,"
and " Oku-kona " meaning " To knock."

STORKS 155

It is quite a common bird, but only at certain places,
usually an open rocky shore where the Ampullariae are
easily found.

Another all black stork of about the same size {Ciconia
nigra) is often mistaken by the natives for the " Mpavana,"
so that at first when one asked its name one always got
the name for the ibis. However, a more observant
native distinguished it as " Sombabyuma," a curious
name which implies that this bird collects pieces of
iron !

Another stork of about the same size, but black and
white {Abdimia abdimii), appears from time to time in
large flocks that, after soaring round and round high up
in the air, descend to open grass land and eat grasshoppers.

The natives call it by the name " Enunda," which is
apparently their equivalent for the English " stork,"
for they applied the same name to a specim^en of the fine
saddle-billed stork {Ephippiorhynchus senegalensis), which
I once, and only once, saw on the shore of Buvu, in
the Sesse Archipelago, in 1912, It is a fine bird, with
blue-black and glossy bottle-green and white plumage,
the huge bill being crimson and black with a " saddle "
of yellow. The European stork, curiously enough, was
never met with on the islands, and I had never seen it
until I found it numerous in the Singidda district of ex-
German East Africa, at the southern end of the great
East African Rift valley — this was in February 1918.
The remarkable Nilotic whale headed stork {Balceniceps
rex) was also absent from the islands, perhaps because
there was not a large enough area of papyrus swamp.

The Hammer-head, or Tufted Umbre {Scopus umbretta),
is not uncommon along the shore, although one never
sees more than one at a time, for it appears to be a solitary
bird. It is of a dull dark brown and has a large head,
tufted as its name implies, with a heavy bill. It is found
standing in the shallow water of a sheltered pool or bay,

156 BIRDS

apparently watching for fish as do the herons, but possibly
also dozing, for I think it is more crepuscular or nocturnal
than diurnal.

Flamingoes, somewhat to my surprise, were never met
with on the Victoria Nyanja, but I cannot say why.
The lake abounds in shallow bays and quiet sheltered
waters, where one would think flamingoes would find all
they desire.

Herons and their allies are numerous on the islands,
as might be expected. The finest bird to be seen on the
lake shore is the goliath heron, which, like many other
species of the Uganda fauna, extends its distribution to
the west coast.

This magnificent bird is known to the natives as
" Kimbala," ^ and may be seen standing at the edge of the
water or on rocks on every island. I should think
there may be one to every two miles of coast line. Such
large birds would, of course, require extensive fishing
grounds, and I have twice seen disputes between a goliath
heron and a fish eagle.

On the first occasion I was examining the coast of
Buvumira Isle, in 1912, when one of these great birds was
seen flying over the surface of the water with a fish eagle
in hot pursuit. When the eagle came uncomfortably near
the heron dropped on to the surface of the water and sat
there with his beak in a defensive position ready to jab
at the eagle should it be unwise enough to attack, and
turning round as so to keep facing the eagle from whatever
direction it threatened, erecting a crest of feathers on top
of its head when the eagle came near. The eagle did
not venture to attack, and drew off a little, whereupon
the heron took to flight again, its enormous wings beating
the air with slow strokes. Seeing its enemy again on the
wing the eagle made another dash at it, whereupon the

^ Soma less observant natives call this bird " Balwa," but I think this
name belongs to another heron, possibly the Ardea purpurea.

GOLIATH HERON AND EAGLE 157

heron repeated its defensive tactics, dropped on to the
water, and took up the same attitude as before.

As soon as the eagle drew off the heron took to wing
again and this time was not pursued ; the eagle, being
satisfied that the trespasser had been driven from the private
fishing grounds, returned victorious to its favourite perch,
while the heron slowly disappeared into the distance.

On the second occasion when the two birds were seen
in conflict, the heron had the best of it. This was in
1913, when I was going down to see Nkosi Islet, the
southernmost of all the Sesse group. While passing down
the channel between Kirugu and Buguye, I saw, on the
islet Lula, a fish eagle swoop down upon a goliath heron
that was standing on the shore, but presumably trespassing
on the eagle's preserves.

There was some rough and tumble business, and then the
eagle was seen on the ground, and whenever it moved the
heron jabbed viciously at it with its formidable beak,
and, I think, had one foot on the eagle to hold it down.
A considerable noise was made, but I think the heron was
chiefly responsible for the very savage snarling and grunting
sounds that came from the pair. I confess my sympathies
were with the eagle, who seemed to be getting much the
worst of it, but eventually managed to disengage and
fly away, and the heron did not pursue.

It would seem as if the eagle was superior on the wing,
and the heron knew it, but was in the stronger position
on the ground. In the case first described the heron
was at first seen on the ground, but took to wing at the
approach of the canoe, when the eagle gave chase ; presum-
ably it had not dared to attack before. The hoarse,
raucous cry of the Eambala is quite startling when one
suddenly gets up, frightened by a gunshot, from a thick
bed of rushes where it had been standing unperceived.
It is a handsome bird, coloured with various shades of
brown and purplish grey. It is curious that I have never

158 BIRDS

seen one actually feeding, but presumably it catches fish
and frogs like other herons.

Several other species are to be seen. A grey one is
known as " Sekanyolya " {Ardea melanocephala), but I
have nothing to say about them.

An extremely beautiful but seldom seen species was
noted in 1918 on rocks of! Damba Island, a locality that
has been alluded to before. There were seen, among many
other birds, half a dozen great white herons {Herodias
alba), collected on one small low rock, their pure white
plumage and graceful build arousing intense admiration.
Subsequently, some were seen on other islands, but they
are by no means abundant, and the natives have no
name for them other than " Enyange nene," i.e. large
egrets.

But the Egret {Egretta garzetta) is more slender, smaller,
and has a black bill and legs. It is quite common on the
islands, and may constantly be seen paddling at the edge
of the water on the beach or on flat rocks where the waves
are breaking, picking up small fish from the water, or
investigating small pools for frogs, its pure white plumage
making it very conspicuous.

Although they usually fish by themselves, egrets may
often be seen at sunset in small flocks wending their way
home, flying slowly over the surface of the blue water
with hoarse croaks.

They nest in much the same situations as are chosen
by darters and cormorants, namely, ambatch trees growing
in shallow water. In January 1919 I found some young
almost ready to leave the nests.

A bird that is often mistaken for the true egret is
the white buff-backed heron {Bubulcus ibis), which is
extremely abundant at Jinja or Entebbe, where flocks
may be seen in attendance on cattle, ready to pounce on
the grasshoppers and other prey disturbed by the move-
ments of the grazing beasts. But they are not so slenderly

PIGEONS— FRANCOLINS 159

built as the more graceful egret, and their shorter, stouter
bills are yellow instead of black, as also are their legs.

Curiously enough they are, on the islands, no more
plentiful than the true egret, perhaps because there are
no cattle to collect them together in flocks.

Doves were common, but several notable species found
on the mainland were never seen on the islands — for
example, the small, long tailed Mna capensis, seen at Jinja.
A forest species, whose long drawn out call was heard
coming from the tops of tall trees when I examined patches
of forest on the Kyagwe coast, was never heard in the
island forests. The beautiful green fruit pigeon (Vinago),
with red bill, was quite common, as might be expected
from the abundance of fig trees growing on the rocky
shores. It has a long and rather complicated call, which
one soon got to know, but found impossible to imitate.

Game Birds are chiefly remarkable for their absence.
No guinea fowls were ever seen or heard, nor any of the
larger " bush-partridge " (species of Francolin). It is
possible that they had been exterminated by natives
trapping them in the old days before the outbreak of
Sleeping Sickness.

A certain loud, ringing cry was so frequently heard
from the depths of the forest belt on Bugalla that I soon
became thoroughly familiar with it. But it was long
before I found out what was the bird responsible for it.
One day in the forest a bird that at the time was noted
to be of the partridge type, ran out from a clump of thick
undergrowth, but seeing me almost at once dived back
again. It was undoubtedly the bird that was responsible
for the well-known cry that had been coming from that
spot. Further knowledge of birds made it obvious that
this must have been one of the forest Francolins.

I believe that this bird has much increased its range
on the islands since those days. Between December 3rd,
1918, and mid-February 1919, I examined thoroughly a

160 BIRDS

chain of islands which I had visited in 1914 with Mr.
Fiske. On the second tour this forest Francolin was
heard calling on every isle that had forest to shelter it,
where it was not noticed in 1914.

The only other game birds of the islands are Quails,
of which two species were met with ; but save that one is
much larger than the other there is nothing to be said
about them.

Birds of Prey. — The most conspicuous bird of prey is,
of course, the glorious white breasted Fish Eagle or
" Empungu " (Haliaetus vocifer), which is literally quite a
feature of the landscape. Its pure white head, neck and
breast, set off by rich red brown and black on body and
wings, render it visible afar when perched on a tree
top or prominent branch hanging over the water. Its
joyous scream, uttered when soaring round and round
high in the air, or when perched on the tree, is a most
delightful noise, full of life and vigour. When screaming
the bird throws back its head until it almost touches the
back, and the loud cry is responded to by its mate, for
they seem to live together in couples.

The noise made by the rush of air between the pinions
when one of these superb birds sees a fish at the surface
and stoops to secure it can be heard before one has actually
seen the bird, and amounts to a loud roar. Sometimes
the bird chooses a fish of such size that it cannot lift it.
An ofiicer of the Uganda Marine told me that he had seen
one taken under water by a fish in which its talons were
presumably inextricably fixed. It was probably one of
the large lung fishes which had come to the surface for
a breath of air.

So far as I can remember, however, the eagle does not
usually plunge quite into the water, but checks itself
before the impact, neatly picking up the wriggling prey
with its feet, and carrying it off to a favourite perch.

During brilliantly fine weather, when there is usually

NEST OF OSPREY (" MAKWANZl ")

To face p. i6o

FISH EAGLES - 161

a steady southerly or south-easterly breeze, the eagles
soar round and round high up in the air and scream
frequently and loudly.

A native told me that this was a sign of high wind
coming, but I do not think it is of any more significance
than an indication of the fine breezy weather as afore-
said.

The nest of the fish eagle is a conspicuous structure in
the fork of a large tree. In October 1910 I saw a bird
carrying in its talons a mass of building material up to
the chosen nesting place.

A certain large " Muvule " tree at Jinja was in use as
a nesting place in 1914, and I found it still in use in
February 1919, so that these birds would seem to choose
the same spot year after year.

The fully fledged young bird has a rather untidy
appearance, its dark blackish plumage being irregularly
spotted with dull white, and it only gradually acquires
the magnificent snowy breast of its parents. The young,
two in number, remain at the nest long after they have
begun to fly, and are apparently driven away by the
parents to look after themselves. At least this is the
interpretation I put on what was seen on Kerenge Island
in January 1919. Two adult birds were soaring round and
round, and a third, in immature plumage, was apparently
objectionable to them, for they kept swooping down and
obviously trying to drive it away, yet not really en-
deavouring to hurt it, and desisting from their attacks
when it went a sufficient distance away.

Another Ush eating bird of prey is the common Osprey,
but I have unfortunately nothing particular to say about
it. The natives call it " Makwanzi," and a nest photo-
graphed on the small Islet of Wavuziwa in January 1919
was attributed to this bird.

After the fish eagle the most conspicuous bird of prey
on the islands is a brown, yellow billed Kite, " Akamunyi "

12

162 BIRDS

[Milvus cegyptius parasitus), which is really extremely
abundant. It was noticeable that the hills on certain
islands, such as Kizima, always had a number of kites
soaring over them, and it seems probable that the small
area of heated lands surrounded by cooler water caused
a continual up-current of air on which it was easier for
them to soar.

Their shrill, quavering cry is very characteristic and
quite pleasant to hear, I think there are in proportion to
area many more kites among the islands than on the main-
land, at any rate the cry is not nearly so often heard at
Entebbe or Jinja. Possibly this is because they are in a
manner scavengers of the surface of the lake, and will
swoop down and pick up small fish, or dead fish mangled
by gulls, or other objects floating on the surface. But
their appetites are all embracing. A grass fire will attract
numbers of them, as well as of other birds, who sail round
and round ready to pounce on grasshoppers and other
insects fleeing before the flames.

When winged termites swarm out of the ground in
times of rain, kites speedily collect and circle round and
round, catching the feebly flying termites in their feet.
It is pretty to see one while on the wing stretch his foot
forward and take the food from it with his beak in an
easy, leisurely manner. The kite is not above trying to
rob other birds of" their prey, and will swoop down upon
one in the endeavour to make it drop the fish which it
has picked up for itself.

Another smaller kite, known as " Ma'ga " by the
natives {Elanus coeruleus), is quite common on certain open
grassy islands. It is of a soft dove-grey and white. Its
habit of hovering vertically over one spot with rapidly
beating wings deceived me into thinking it was a species
of kestrel, but I was put right by Mr. C. F. Belcher. While
in camp at Jinja in 1910 I used to watch these birds
carry the mice which they had caught on to a certain

HAWK AND CHAMELEON 163

tree in front of the tent, and noted that they usually
pulled out the viscera and let them drop to the ground
below.

A fairly large grey bird that may have been a buzzard
was noted at Jinja under rather peculiar circumstances.

I was walking along a path thickly bordered with
castor oil bushes, when I heard a great fluttering coming
from one, and saw a large grey hawk hanging by its feet
from something yellowish on a stem of the bush. It
appeared as if the bird had caught its feet in a tangle of
creeper, and I went up to release it, wondering how I
was going to avoid its sharp bill.

When I approached within a few feet the hawk suddenly
flew away, leaving a yellow object hanging from the
branch. Only when a bright red drop slowly trickled
down did I realize that the object was a chamseleon.

The hawk had pounced on it and was trying to drag it
from the branch, but the chamaeleon had such a firm grasp
with its prehensile tail that all the flutterings of the bird
had not succeeded in dislodging it. However, the grip
of the talons had so injured the chamseleon that it died
almost immediately, so that the bird would have been
able to carry it off in the end.

I examined the stomach contents of the chamseleon and
found a small snail and remains of orthopterous insects,
probably tree dwelling Locustidae.

A projpos this observation, it may be said that the small
brown lizards that run about on the grass roofs of native
huts appear to be a desirable food, for one often sees
birds of prey swooping down on them.

I once saw a large cock peck at one till it was disabled
and then swallow it whole, a disgusting sight !

Owls. — The cries of two species of owl are very familiar.
One is a melodious quaver, heard on moonlit nights,
the other a deep soft " ZToo-hoo," probably uttered by a
big " long eared " owl {Bubo lacfeus). During the day

164 BIRDS

time one often hears quite a different noise proceeding
from a thick shady tree, obviously emanating from an
owl, but whether from the latter species or another I
cannot say.

As with other groups of birds, notable absentees from the
islands may be mentioned among birds of prey. Vultures
were never seen, but this is perhaps what might be expected
from the absence of game. The beautiful Bateleur eagle
{Helotarsus ecaudatus), so noticeable on the mainland, was
either absent or so scarce on the islands that I cannot
remember having seen it there.

Another fine species, the Black and White Crested
Eagle {Spizaetus coronatus), known to the Baganda as
" Wonzi," was only very occasionally seen. I first became
acquainted with this handsome, fierce looking bird at
Jinja, where its rather weird scream was often heard,
and the bird was often seen to perch on the summit
of a dead tree.

The next family to be mentioned is that of the Parrots,
about which I have not much to say, since I have only
met with one species on the islands. This is curious,
because at Entebbe and Bukakata and Jinja may be seen
quite commonly one of the small green parrots with yellow
shoulders.

The well known grey parrot with red tail occurs on
many of the larger forested islands, and its discordant
shrieks were familiar. At daybreak and eventide they
fly to and from their favourite feeding grounds in small
flocks, and, like no other bird, chatter and whistle con-
tinually as they go. Often one hears the well known
noises high overhead before the birds have come into
view.

On Sanga Islet one day, there were some parrots
feeding in the trees which took alarm and flew off as
I walked underneath. One of them gave a call which
seemed to be an imitation of the monkey's alarm, the

PLAINTAIN EATERS 165

only example I can cite of a wild parrot copying other
sounds.

After the parrots come the Plaintain Eaters {Muso-
phagidae), one of which {Musophaga rossae), known to the
natives as " Fulungu," is one of the loveliest birds of
Uganda.

Somewhere about the size of a rook, but with much
larger tail, this bird has deep purplish blue plumage
with bright crimson patches on the ' wings, making it
very conspicuous when it flies. It also makes itself
conspicuous by its voice, for it is a noisy bird. A company
of them will be heard apparently working themselves up
to concert pitch, with short cacklings and gurglings,
and then they suddenly all burst out together into loud,
not unmusical cooing in a most pleasing manner.

These birds are typically forest species, and are only
seen where the forests are of some size, so they do not
occur on the smaller islets.

Another species of very different appearance has much
less gaudy plumage, being ashy grey, and showing patches
of black and white on wings and tail when flying : probably
it is SchizorMs. I often used to see a pair on Damba,
where a marshy piece of shore had ambatch trees in flower
that were highly attractive to these birds, which were
frequently seen pulling off the bright yellow flowers and
devouring them.

They, like the Musophaga, are noisy birds, and attract
attention by their vociferous habits.

One, presumably the cock bird, used to do what airmen
call " stunts " in the air, shooting vertically upwards and
then dropping headforemost to Join the other in the
ambatch bush, with the accompaniment of loud cackling
and crowing noises. Though so frequently seen in the
ambatch bushes they always came from the forest, to
which they retired when alarmed.

The Lark-heeled Cuckoos or Coucals (Centropidae) abound

166 BIRDS

among the forests and patches of bush on the islands, but
there are differences from the mainland in the propor-
tions of the species. Their liquid gurgling notes, descend-
ing and ascending again, have earned for them the native
name " E'tutuma " and the English " Bottle Bird," the
latter name because of the resemblance to the noise
made by water being poured out of a bottle with long
narrow neck.

Generally speaking, they are skulking birds, avoiding
the open and concealing themselves among bushes, under-
growth or papyrus, whence they fly heavily away with
broad tail outspread.

Heavily built, with large stout beaks, they cannot be
considered attractive, though their rich brown, black
and white or yellowish plumage is often beautiful.

A species that was unusually abundant on Bugalla
(Ceuthmochares oeneus) was of a dark greenish black all
over, with the large bill bright yellow. Its note also was
somewhat different from that of most coucals, being a
harsh screech. I saw one fly away from a clump of bush
with a butterfly held in its beak, but another that I shot
had fed largely on hairy caterpillars, as is the cuckoo
custom. So much less common is this species elsewhere
that I cannot remember having noticed it on other islands
or the mainland. On the other hand, a certain large
species, whose deep booming call is heard in mainland
forests, was absent from the islands.

Next to be mentioned are the Cuckoos. I am familiar
with the notes of only two species of cuckoo on the islands,
one of which {Cuculus soUtarius) is very common and from
its cry has been named *' Sekoko " by the natives. The
frequently reiterated call is a descending sequence of
three notes, and during the rainy season is very noticeable,
the bird sitting perched in a conspicuous place and calling
again and again as does our English species.

Though its colouring is different from that of the latter

43

VIEW FROM CAMP ON KOME, LOOKING EASTWARDS.
Damba Isle in far distance.

fi.i w

VIEW TOWARDS BUVUMIRA FROM BUGALLA CAMP.

To face p. i66.

CUCKOOS— BARBETS 167

bird, the Sekoko is very obviously a cuckoo, both when
seen perching and on the wing. It seems to call only
during the two periods of greatest rainfuU, when it is
presumably breeding.

The other cuckoo whose note is familiar is the
*' Ekirimululu," {Chrysococcyx caprius or C. klaasi), one
of the bronze cuckoos, a small species, metallic green
above and whitish beneath. The native name, like so
many others, is doubtless onomatopoeic.

Another group, the Barhets, is strangely unfamiliar on
the islands. One of the commonest of all, the little Copper-
smith {Barbatula leucolaima), has never been heard on
any island, although its monotonous " Tonk-tonk-tonk "
can be heard all day long in the Entebbe gardens. Their
absence from the islands was made more noticeable when
in January 1919 I went across to the Kyagwe coast from
the camps on Kerenge and Wema. Almost immediately
one landed on the mainland one heard the well known
metallic note in the forest, together with the notes of
other birds not present on the islands.

Larger, red breasted barbets have not been noticed
either there, but they do not force themselves upon one's
attention as does the persistent coppersmith.

The next family, CoUidae, is easily dealt with, for they
are absent from islands that have been visited, and since
I was quite familiar with them at Jinja before I first went
on to the islands, they would certainly have been seen or
heard, for I know the twittering sound they make very
well ; it is constantly heard in the gardens at Entebbe.

We now come to the Rollers and their allies. Rollers
themselves may be dismissed in a few words, for I never
saw one on the islands. I had seen them at Jinja in 1910,
and thus knew their appearance, and, having seen them
in numbers during the East African campaign, was on the
look out for them when I returned to the islands in Decem-
ber 1918, but the result was the same.

168 BIRDS

This seems very curious, and I should like a practised
ornithologist to confirm or contradict the statement.

Other members of the sub -order now being considered
are the Hornbills, one of which (Bycanistes suhcylindricus)
is one of the most characteristic birds of the Uganda
forests, and abounds on the islands. This great black
and white bird with its curious bill is known to the natives
as " E'nga-nga," from its nasal, raucous cry. It is a noisy
and conspicuous creature, clumsy and grotesque in its
habits, and is perceived both by ear and eye from a
considerable distance away. The flight is heavy, and the
beating of the wings for a few strokes is followed by a
pause during which the bird sails through the air with
wings outspread, making a loud roaring noise presumably
owing to the arrangement of the feathers. When an
" E'nga-nga " flies past, I am always reminded of the noise
made by a puffing locomotive as one hears it coming
from a distance, the noise rising in pitch and getting louder
as the engine approaches, falling and dying away as it
disappears in the distance. The bird also frequently
screams while flying, as if to let all the world know where
it is.

I believe that there is an explanation of this noisy
flight as aposematic — the conspicuous bird makes itself
perceived by ear as well as eye in order that enemies may
recognize it. It is of interest that Dr. G. A. K. Marshall
records that the flesh of an allied species was found dis-
tasteful by a mongoose, and that natives look on it as so
poisonous that it will render water unfit for drinking if
the carcase falls into a stream.^ The subject of aposematic
flight will be brought up again in the chapter on
the Hymenoptera in connection with the fossorial Pom-
pilidae.

Foolish looking as is the " E'nga-nga," it is surpassed
in that respect by a more slenderly built smaller species,

1 Proc. Ent. Soc. Land., 1902, pp. .378-9.

KINGFISHERS 169

without any casque on the bill, whose appearance always
fills me with resentment that it should exist at all
{Lophoceros melanoleucus). It has a floppy flight, rising
and. falling as it flaps its way along v/ith inane, shrill
cries. Its dull whitish and black plumage has not the pure
tints and arresting appearance of the " E'nga-nga," and
indeed I find it difficult to say anything nice about this
bird.

No other hornbills than these two species have been
noticed on the islands.

Next to be mentioned are Kingfishers, and it is curious
that the natives have the same nam^e for one very common
black and grey spotted species as for Bee-eaters (" Mujolo "),
but other kingfishers are known as " Akasimagizi,"

The " Mujolo " kingfisher {Ceryle rudis) is extremely
common along the lake shore, where shallow water with
sandy bottom provides good fishing grounds. Its habits
ar?i somewhat different from those of other species, such
as our familiar one at home, which plunges into the water
from some commanding perch. Ceryle rudis hovers over
the water, head hanging down with beak pointing vertically
to the surface, wings rapidly beating like those of a kestrel.
Often it has to wait a long time, and change its position
till some unwary little fish comes near enough to the
surface, when the kingfisher plunges in head foremost and
secures it. This bird is rather vociferous, and its twittering
cry is very often heard as it flies across the water, keeping
close to the surface. Another species of Ceryle (C. maxima)
is scarce, and of solitary habits, very much larger than
any other kingfisher on the islands. It lurks among
bushes overhanging the water, and is usually first noticed
flying away when it has been disturbed, uttering a loud
unmusical bleating cry.

Beautiful little species of the genus Ispidina (? picta)
are often seen among the rushes growing in shallow water
or on low twigs of ambatch trees just above the water,

170 BIRDS

whence they plunge in to secure their prey. Their feeble
little piping note is much like that of our English king-
fisher, but though very soft it often directs one's attention
to these lovely little birds as they dart past.

About the size of a sparrow, but, of course, of different
appearance, with short tail and enormous bright red
beak, these birds have lovely orange and deep blue
plumage.

Natives say that if one comes near their house it is an
omen that some one is going to die ; and it was a curious
coincidence on Bugalla that when one of my canoe-men
was ill with pneumonia one of these birds was frequently
noticed in the camp, although I had not seen it there
before.

I did not know the superstition until some time after
this. Fortunately the canoe-man recovered ; perhaps if
he had known the bird of evil omen was there the issue
would have been less satisfactory !

Another kingfisher, of different habits, is Halcyon
senegalensis, whose large beak and legs of crimson contrast
beautifully with its sky blue wings and back and grey
under surface. This is a forest or bush frequenting species
which has given up a fish diet in favour of large grass-
hoppers.

Sitting on a branch it watches the grass for movements
and then pounces down on a " Janzi " grasshopper which
it carries back to the perch. Here the unfortunate bulky
insect is banged against the branch until it is in the right
condition to be swallowed whole. This kingfisher has a
loud and cheerful cry. It commences with a sharp high-
pitched note which is followed by a churring descent down
the scale, and on a bright sunny morning this familiar
sound may be heard over and over again.

I saw an interesting little incident on Kerenge in 1914.
One of these birds was angrily scolding on a perch which he
considered to be his, but the justice of the claim was

KINGFISHERS— BEE-EATERS 171

disputed by one of the Paradise flycatchers, which kept
trying to get at the kingfisher from behind. The other,
however, spun round on the perch just as quickly so as
always to face the angry flycatcher.

But at length another flycatcher came up, and the
kingfisher made off, thinking discretion the better part of
valour.

Another species of Halcyon may here be mentioned on
account of its remarkable absence from the islands. This
dull mottled greyish bird, H. chelicuti, is abundant both
at Jinja and Entebbe, where its " cheer-oh " cry may often
be heard in chorus, and yet I never heard it on the islands,
although Van Someren gives " Sesse Isles " as a locality
for this species in his check list.^

The absence of the insectivorous Rollers seems to be
compensated on the islands by great abundance of their
allies the Bee-eaters. These graceful birds, known to the
natives as " Mujolo," soon attract attention from their
habit of perching on the tip of a dead bough in a com-
manding position and swooping down on their prey,
returning with it to the same perch. Two species, Merops
superciliosus and Melittophagus meridionalis, became very
weU known, and the liquid twitter of the former was one
of the most familiar bird sounds.

When a blustery wind is bringing up storm clouds, this
bird often soars overhead in numbers, braving the buffeting
of the gusts of wind and possibly catching the insects that
are driven before the storm.

Melittophagus is found more often in grassy places, or
at the edge of the forest perched on an especially tall
stem of grass or rush. It keeps nearer to the ground than
does its larger, long-tailed relative. Neither of these birds
is brilliantly coloured, but their green, yellowish, and
brown hues are very pleasing, and their long biUs and
graceful shapes give them a very attractive appearance.

1 The Ibis, 1916, p, 244.

172

BIRDS

I noticed particularly that Dragon-flies are a very favourite
food of both species, as well as the bees and other
Hymenoptera on which they are well known to feed,
which were constantly found in the stomachs when I
was searching for evidence that the Tse-tse fly was eaten
by birds. (See vol. xiv of Sleeping Sickness Reports of
the Royal Society, 1913, pp. 15-16.)

•Stomach Contents of Bee-eatebs.

Merops superciliosus.

1. Mainly dragon flies.

2. Dragon flies and honey bees.

3. Dragon flies and honey bees.

4. Dragon flies and honey bees.

5. Dragon flies and honey bees.

6. Dragon flies and honey bees

7.
8.
9.

10. Dragon flies and honey bees. *10

Dragon flies and honey bees.
Dragon flies and honej^ bees.
Dragon flies and honey bees.

Melittophagus meridionalis.

1. Mainly dragon flies.

2. Dragon flies and honey bees.

3. Dragon flies.

4. Dragon flies.

5. Dragon flies and small beetle

(Lamellicorn).

6. Winged ants and one small
dragon fly.

7. Dragon flies and bees.
"8. Winged ants only.
^9. Winged ants and one Agrionid

dragon fly.
Winged ants and a small species
of Chrysididae.
*11. Winged ants.

These were all shot on the same
day(lM2-ll).

11. Dragon flies and honey bees.

12. Dragon flies, bees, and a

Belonogaster wasp.

27. Dragon flies.

28. Dragon flies.

29. Dragon flies and bees.

30. Dragon flies and bees.

31. Dragon flies and bees.

32. Dragon flies.

33. Dragon flies and bees.

34. Caterpillars and young Mantidae.

35. Eleven bees.

36. Twelve bees, one wasp, one

dragon fly.

37. Twelve bees, one male Stink

ant (Paltothyreus).

Frequently one would hear a loud tapping noise, and
looking up would see a bee-eater with a large dragon fly
in its beak banging it against the perch on which it sat,
finding the long body and wings rather unmanageable,
for they swallow the dragon flies wings and all.

NIGHTJARS ITS

Butterflies appear to be eaten by Melittophagus, for
I have found on the ground, under a grass stem on which
one had been sitting, v/ings of the little yellow Terias
butterflies, although I have not actually witnessed the
capture in this case.

Another group of birds allied to the Rollers may be
dismissed in a few words, namely, the Wood hoopoes
(Irrisorinae), for I have never met with them on the
islands, and first saw them when on active service in
German East Africa.

Such noisy and conspicuous birds could hardly have
escaped notice had they been on the islands, and when I
returned in 1918 they were watched for, but never seen.

The family of Goatsuckers, or Nightjars, is represented
on the islands by a species of Caprimulgus, whose native
name is " Olubugabuga," presumably an attempt to
reproduce one of the noises that it makes. On moonlit
nights, when owls are also heard, these birds keep up a
continual " Tok-tok-tok-tok-tok," varied occasionally by
the melodious cry from which their name is derived.

The bird itself is often flushed during day time from its
resting place in places that may sometimes be quite open
and fully exposed to the hot sun, or sometimes shaded and
among bracken. There is nothing about the bird to call
for remark, unlike the East African species with long
floating plumes, which has not been met with on the
islands.

It is, of course, highly probable that there is more than
one nightjar on the islands, but I am familiar with only
one cry.

We now come to the great order of Passeriformes,
embracing large numbers of small birds, which I must
perforce pass over in silence, not being sufficiently familiar
with the comparatively inconspicuous species.

Certain species, however, call for notice, and first
among them are the Flycatchers (Muscicapidae), of which

174 BIRDS

two species particularly attract attention. The first,
Tchitrea emini, is illustrated from a photograph which
shows well a very beautiful feature, namely, the long
snow white tail plumes assumed by the cock bird in
the breeding seasons. The rich chestnut brown, black
and white plumage of body and wings is set o£E by
slaty blue black crested head, and it is a delight to
watch the vivacious bird, as it is for ever flitting among
the branches and chasing another from perch to perch.
The insistent call, quite impossible to describe, loud and
ringing, was the first bird note to be heard in the early
morning from the forest around the camp on Bugalla.
As soon as the darkness began to lighten these delightful
birds were awake and filled the forest with their call.
The native name for this bird is " Kunguvu." When I
left Bugalla I went home, but the memory of the call
there has persisted clearly, and I am sure that the Bugalla
race of the Kunguvu has its own call, i.e. a geographical
race is being developed. The bird is plentiful enough on
many other islands ; indeed, it is one of the species which
is far more abundant on islands than on the mainland,
but nowhere has its call quite the same character and
finish as on Bugalla. I particularly noticed this during
the tours of 1914 and 1918-19, during the latter of which
several forests on the mainland coast were visited.

It is not surprising that the birds on a large island such
as Bugalla should be forming a local race — they are
confined to the forest and do not take long flights.

One was never tired of seeing the cock bird pursuing
the hen among the branches, excitedly calling to her,
with snow white plumes trailing in the air as he flitted
from branch to branch. The nest is a small shallow cup-
shaped structure ; I have seen it fastened to the end of a
dependent creeper, quite exposed in the open, but in such
a position that nothing without wings could reach it.
The hen bird was fully exposed as she sat.

"kunguvu" fly-catcher [tchitrea EMim).

Foot-rule on right.

NEST OK CROCODILE.
The marks of the mother's body and tail on the sand can be plainly seen.

To face p. 174.

FLYCATCHERS— SHRIKES 175

Another delightful little bird that soon became very
familiar on Bugalla was the black and white flycatcher,
with bright red wattle over the eye, probably Platysteira
jacksoni, whose call note is very characteristic. Two
birds were often heard calling one to another. The first,
with high-pitched pipe, would sing " How axe you ? " to
which the other responded, " I'm pret-ty well, thanks."

By whistling the first phrase I could often get the
answer and bring the bird near. Like Tchitrea this bird
seems to be forming a local race on Bugalla, for the call
elsewhere is not exactly the same, although one can recog-
nize it as coming from the same bird.

It is not so much a forest frequenter as Tchitrea,
being found more among the clumps of bush and copses
that make the park like country on the larger islands.

A third very pretty little flycatcher {Elminia longicauda)
is worth noticing here, because I have not seen it on the
islands, save once on Kizima, though it is often seen in
the gardens of Entebbe. It is a light blue bird which has
a habit of spreading out its tail fanwise, and hence attracts
one's attention.

Shrikes are noteworthy because several species familiar
in the gardens of Entebbe and Jinja have never been
seen on the islands. Conspicuous among these is the
beautiful black bird with scarlet breast {Laniarius
erythrogaster), pairs of which haunt Hibiscus bushes, whose
red flowers exactly correspond with the breast of the bird.

They have a call note which may be crudely represented
by " What-ho." This is uttered by one, and simultaneously
the other of the pair makes a harsh scolding or churring
noise, not at all musical. It is curious that many shrikes
have this habit of uttering a joint call ; for instance, the
long tailed black and white species, whose very musical
notes I have often heard on mainland but never on an
island.

A brown shrike (? Pomaforrhynchus) that was shot on

176 BIRDS

Damba Isle was found to have eaten only grasshoppers
in two cases, in the third grasshoppers and one beetle.

Very noticeable absentees from the islands are the black,
fork-tailed, insectivorous Drongos {Dicruridae) , which were
never met with there. The first ones I ever saw were on
the frontier between German East Africa and Uganda
in the early days of the war, and later in various parts of
German East Africa they were seen abundantly, so that
I became thoroughly familiar with them. When I returned
to the islands in 1918 for about three months, I should
therefore certainly have seen Drongos had they been
present. It has seemed to me during my wanderings
that where Drongos were abundant Bee-eaters were scarce,
and since Bee-eaters appear to be unusually plentiful on
the islands and Drongos are absent, it is possible that these
birds of similar habits take each other's place to a certain
extent.

The Starling family is also notable for the absence of
conspicuous species from the islands, viz. the Glossy
Starlings, which certainly could not escape notice. These
birds have the typical shape and noisy habits of our
home species, but their dark plumage is resplendent with
metallic purple and green. Their harsh voices are very
familiar at Entebbe among the tall trees, and their noisy
flight also renders them conspicuous to the ear. Directly
I entered the forests on the Kyagwe coast opposite the
islands where I was working, these noisy birds attracted
attention, so that I am quite sure they do not exist on any
island I have visited.

Weaver Birds {" Endegeya ") are abundant among the
ambatches which overhang the water ; but some species,
that make their nests in colonies on large trees on the
mainland, so that the noise of them at a little distance is
like the sound of a waterfall, have not been met with
on the islands.

Neither have I seen the Bishop and Widow finches,

FINCHES— SUNBIRDS 177

beautiful brightly coloured birds, which could hardly
escape notice.

The Fringillidae, or Finches, would doubtless be of much
interest to an ornithologist, but I have little knowledge,
and less to say, about them. Noteworthy varieties on the
islands are Sparrows, " Enkazalugya " [Passer griseus), of
which very few were ever seen ; possibly the absence of
mankind has something to do with this. Another notable
absentee is a curious rotund little grass finch, which is
abundant at Jinja and Entebbe. The cock bird when
adult is black and white, and has very long floating tail
feathers which appear to hinder his flight. He flits for a
short distance over the grass with an up and down motion,
uttering feeble chirps, accompanied by a flock of brown,
insignificant looking individuals, which I always took for
females. Mr. C. F. Belcher, however, tells me that most
of them are immature males.

A very common and pretty little finch is often seen
hopping about on the ground at Entebbe and Jinja ;
of a reddish pink or old rose colour, it is often nicknamed
the " animated plum." It is about the size of a red plum,
and agrees with one very well in hue.

Another species, much like it in habits and shape, is
grey and blue. Neither of these have been met with on
the islands.

The pretty little Sunbirds (Nectariniidae) are abundant
on the islands, though I know not how many species are
found there. On certain very small islets they and wagtails
are the chief part of the bird fauna, but this depends upon
whether their favourite food plant is there. This is a
labiate, which sends up tall stems bearing scarlet flowers
arising from a knob. These knobs are set at intervals
up the stem, and when the flowers are withered and .the
seeds develop are unpleasantly spiky to knock against.

Wherever a cluster of these plants grows, a sunbird
is sure to be seen clinging on to the stem and thrusting

13

178 BIRDS

its slender bill down the tubes of the flowers one after
another.

There is no doubt, however, that they also eat insects,
or at any rate spiders. I have seen one hovering in front
of a web spun amongst tall stems of grass and delicately
picking off the spider that sat in the centre of it.

On the other hand, a sunbird was once found hopelessly
entangled in the strong sticky threads of a web of a Nephila
spider, the sheets of whose webs are so unpleasantly
abundant on many of the islands. Fortunately, I was
able with a stick to reach up and set the bird free. The
cock bird is a beautiful little creature, the prevailing hue
being rich metallic green, but different species have the
throat and breast of brilliant scarlet or other bright
colours ; the hen bird is dull brown or grey. As is usually
the case with brilliantly coloured birds, the song is not
musical, but it is quite a pleasing, lively little twitter.

They are pugnacious birds, and may often be seen
chasing each other with angry scoldings, and I have
watched them chasing away the huge Carpenter bees from
the attractive flowers of a bush.

The Warblers must be passed over, since I have no
ornithological knowledge of them ; but there is one
species that is very familiar, since it inhabits the low
bushes in the forest and its sharp chirp is heard every-
where. It has the appearance and manner of a wren,
with short upturned tail, and is constantly in evidence
climbing about among the branches on which it hunts for
sedentary insects, or angrily chasing away another which
is trespassing on its feeding grounds. It is, I think,
more abundant on the islands than on the mainland.

The Thrush family deserves mention on account of the
fine songster Cossypha, (possibly heuglini,) or Robin Chat
(E'nyonza), which abounds in the forests. It is a handsome
bird, but one seldom seen, as it is of retiring habits and
rarely leaves the sheltering bushes. Occasionally one gets

COSSYPHA 179

a glimpse of its bright orange under side, with dark back
and white streak on the head. Its mellow voice reminds
one in a way of the blackbird ; but it has a unique method
of singing as if for its own enjoyment, quietly and in a
thoughtful way; there is no "fine careless rapture." It
is an extraordinarily clever mimic, and in the evenings
one can hear it copying the notes of fish eagle, owl, blue
kingfisher, cuckoo, and doubtless many other birds as
well, introducing the well known notes into its own steady
flow of music.

Song thrushes, obviously closely allied to the English
songster, are abundant at Entebbe, where their music
brings up memories of home ; but they are only occasion-
ally heard on a few of the islands. Kibibi, Kiuwa and
Kizima were noted as having these birds.

CHAPTER IX

REPTILES AND FISH

A BOOK about life on the Victoria Nyanja may reasonably
be expected at least to mention the crocodile, known to
the natives as " E'gonya."

His scarcity was what one noticed most, for which a
reason will be suggested later. One had read so much
about crocodiles in tropical rivers lying thickly on
the mud, and in great numbers in the water, that one
had expected to see the same in the lake. This was far
from being the case, and it soon became clear that one
or two could usually be seen on any day frequenting
the same haunts, but that it was possible to go long dis-
tances and never see a crocodile, and young specimens
are rarely seen.

Their favourite haunts are of two kinds — a stretch of
coastline fringed with rushes and reeds growing in shallow
water, wherein they lie with only the top of the head
and scaly ridge of the back and tail showing, or an open
shore of rock or sand on which they warily bask. Near
my camp on Bugalla the northern bay furnished a haunt
of the first kind, and a large crocodile could nearly always
be seen there. Ngamba and Tavu Islands provided
rocky basking places, and a favourite large rock on the
east side of Tavu was seen, whenever visited, to have
one or more crocodiles basking on it, and on one occasion
seven were seen there, which was the greatest number
ever seen together. It was amusing to see these huge

180

CROCODILES 181

antiquated reptiles take alarm as soon as the canoe drew
near, and plunge headlong into the water ; they were
always off before one got near enough to take a photo-
graph.

It seems strange that such a huge and well defended
creature should be so nervous. On the only occasion
when I have got within camera range, of course I had
no camera with me. On the northern coast of Ngamba
I was pushing my way through rocks and dense bushes
when I suddenly heard, apparently right in front of my
feet, a horrid noise, half hiss, half snarl. Hastily stumbling
back, I saw I had nearly stepped on the head of a large
crocodile lying on the ground. I remained still and
watched, noting Tse-tse fly feeding on him, and wondered
why he did not, as usual, rush headlong into the water.
Before going away I threw a lump of rock on to his back,
but that did not move him ; he merely hollowed his back,
raising head and tail in the air, and with mouth wide
agape gave a sort of bellow. Possibly he was sick or
was exhausted from fighting, for there was an ulcerated
wound at the base of his tail. On the same island at
a later date I saw another large crocodile with a similar
wound, but on the opposite side, and wondered if this
had been the partner of the duel.

Crocodiles deposit their eggs in a hole scooped in a
dry sandy beach ; these are afterwards carefully covered
over to a depth of about a foot, and one often sees the
imprint of the body of the mother on the soft sand, even
the pattern of the scales on the belly being clearly shown,
so that it appears that the parent returns to visit the
nest, though she does not seem to be very successful
in protecting the eggs, as wiU be seen later.

During 1914 numbers of nests were found, of which
a list is given below : the number of eggs varies little,
averaging sixty in one nest. No evidence was obtained
that more than one crocodile lays eggs in one nest.

182

REPTILES AND FISH

The period required for development is stated to be
twelve weeks, ^

Where Found.

Date.

Number
of Eggs.

State of Development.

1. Marida . .

Feb. 21

68

^33 almost ready to hatch.
(^25 .shrivelled and undeveloped.

2. Masovwi

March 6

59

Embryos about ^ grown.

3. Kizima .

March 11

60

Embryos about J grown.

4. Bulago .

March 14

64

Embryos about } grown.

5. Tavu .

Jtme 17

60

Freshly laid.

6. Kimmi .

June 30

59

Laid between June 23-30.

7. Kimmi .

July 6

69

Laid between June 30-July 7.

8. Bulago .

July 24

66

1

9. Bulago .

July 24

57

>Laid between July 17-24.

10. Bulago .

July 24

76

J

11. Tavu

August 18

65

]

12. Tavu .

August 18

48

sLaidbetween July 30- August 18.

13. Tavu .

August 18

38

J

Note. — Only in the first case was every egg examined, and the number
that had failed to develop is surprising. In view of this fact it is to be
regretted that subsequently only a few from each nest were broken open.

The Monitor Lizard.

By far the most notable reptile on the islands is the
enormous Monitor Lizard (Varanus niloticus), called by
the natives " Enswa-swa." It is of a duU grey colour
speckled with yellow, and its long tail is flattened from
side to side for swimming. This fine creature may
measure almost six feet from nose to tail tip. It
is never found far from the water, into which it rushes
headlong when disturbed, so that its panic-stricken flight
is very often heard as one pushes through the dense
vegetation bordering the shore. In the water the Enswa-
swa swims with vigorous strokes of the flattened tail,
and may often be seen lazily sculling itself along with the
top of the head just above the surface. When alarmed
it hastily dives, and can remain under water for a

^ Gadow, Cambridge Natural History, vol. Reptiles, p. 463.

A CANOE-MAN AND AN " ENSWA-SWA " JUST CAUGHT.

■■■- i'-^.kl -JiM-Hrihlir

VARANUS 183

surprisingly long time. Like the crocodile, Varanus has
favourite basking places on the rocks, choosing preferably
a place where ferns grow, on which it lies ; but curiously
enough if one is tethered in the sun it dies, apparently
from exposure, in an hour or so.

Crocodiles are, as has been said, scarce among the
islands, but Varanus is very abundant, and must have
increased in number very greatly since the removal from
the islands of the natives, among whom skins of Enswa-
swa are in great request for making the long narrow
'* Engabi " drums, which are beaten by the hand at con-
vivial gatherings or weddings. But though the Enswa-
swa are abundant enough, one does not often meet with
their nesting places ; the only three that I ever saw were
a sandy patch of open ground on the north side of Kimmi
Island, where burrows and fragments of old eggshells
were found, a regular warren in sandy soil on Busiri,
and a similar place on the south end of Sindiro Isle.
Occasionally I saw very young Enswa-swa close to the
water sunning themselves on little branches like any
other lizard, but I do not think the adult climbs up trees ;
it appears to be an instance of the losing by an adult
of an ancestral habit retained in youth. A curious
inverted ratio was found to exist between the members
of crocodiles and Enswa-swa on any particular islet : during
the tour at the commencement of 1914 it was noted that if
many Enswa-swa were seen there were very few crocodiles,
and vice versa. Thus, on Tavu Island, where more
crocodiles were seen than on any other island, Enswa-
swa were exceptionally scarce ; this was verified by many
visits and work on the island. On the other hand, when
Rumfua was visited, during the examination of its shores
from a canoe five Varanus but no crocodiles were seen.
It must be admitted, however, that Rumfua was only
visited once. Isentwa was noted to have abundant
Enswa-swa, but no crocodiles were seen there. Although

184 REPTILES AND FISH

it wiU be shown below that Varanus harS a profound
inlSuence upon the number of crocodiles, yet it is not
clear how the reverse can be the case. Varanus seems
to enjoy a varied diet. The stomach of one was found
to contain two slugs and the shell of an operculate fresh-
water mollusc, and collections of shells of the large water
snail Ampullaria are often seen on the shore broken on
a rock as garden snails are broken by thrushes. ^

I have seen a young Varanus biting off the juicy lower
part of the stems of rushes ; natives say they eat fish,
and on one occasion when my boys were walking a little
ahead on the shore, one of them saw an Enswa-swa run
away from a piece of dead fish. On one occasion at
Jinja a small bird that I had shot fell down almost on
top of an unperceived Enswa-swa, who at once picked
it up and ran away with it, shaking the struggling bird
as a terrier shakes a rat.

The most important food of Varanus, so far as man is
concerned, is the egg of the crocodile, and there is no
doubt whatever in my mind that the increase of Varanus
on the islands since the depopulation must eventually have
a powerful effect in reducing the number of crocodiles
there. Thus on July 30, 1914, on Tavu Island, two
crocodiles' nests previously noted there were found to
have been disturbed ; holes had been dug and a crushed
egg was lying on the sand. On Bulago Island, on
August 1st, two nests previously noted were found
ravaged and crushed eggs lay all round. At another point
a nest of seventy-six eggs that had been found, freshly laid,
on July 24th, had been absolutely cleared out and not
one was left. In seven days eggs had been accounted
for at the rate of eleven a day, which would seem to be
more than one Enswa-swa could accomplish ! In fact,
no nest of crocodiles' eggs which I visited a second time

^ This may, however, be entirely the work of the open-bill stork. See
Chapter VIII.

VARANUS AND CROCODILE'S EGGS 185

was found to have been left undisturbed. Though I have
not actually seen Varanus devouring the eggs, there is
no doubt of it from the tracks they leave in the sand
all round and over the nest, and my friend Dr. Lyndhurst
Duke told me that he had actually seen one digging
away at a nest. It therefore appears that strict protec-
tion of Varanus against slaughter by natives should be an
effectual method of keeping down the numbers of croco-
diles. The importance of Varanus as a food supply of
Glossina has been alluded to in the chapter on the Tse-tse.

A fine large lizard of the family Agamidae is known
as " E'konkomi " by the natives. It has a brightly
coloured male partly sky blue, the female being grey.
They particularly frequent open places, but prefer to
have a large tree in the neighbourhood, up which they
run when alarmed. The E'konkomi seems to feed very
largely on ants, judging by the remains found in its
excreta, and from the contents of the stomachs examined
in several specimens. It has to a marked degree
the habit that may be seen in many lizards, alternately
raising and lowering the body between the fore limbs, as
is done by a man lying prone on the ground in a common
gymnastic exercise. I could never make out the sig-
nificance of this habit, though I think it is developed
to a greater degree in the male sex.

Smaller, more slender lizards (Lacertidae) played on
the roof of my tent, and from inside I could watch their
shadows as they scuttled about. It was most amusing
to see them playing : one would stealthily creep up
behind another and bite his tail, and then there would
be a pursuit and a rushing about all over the tent.

They would dispose themselves for the night some-
where among the folds of the tied-back doorway, and
on one occasion I was awakened by hearing their well
known pattering feet on the roof. Realizing that some-
thing must have turned them out from their sleeping

186 REPTILES AND FISH

place, I lit a lamp and found that a column of the " Safari
ant " (Dorylus) had directed its attention to that part
of the tent, but fortunately for my night's rest they did
not investigate my bed.

The little grey Gecko, with discs on its toes enabling
it to run upside down on ceilings and walls, was not seen
in my huts on the islands, although common enough on
the mainland ; but I do not mean to state that it did not
exist on the islands.

Tortoises.

Like crocodiles, tortoises are less plentiful than might
be expected. The numerous isles of varying size furnish
abundant areas with marshy shore and shallow, rush-
grown water, which appear eminently suitable for water
tortoises, yet it was many months before I saw one — a
flat, black, short tailed species that wanders about near
the water's edge. One species of land tortoise is occasion-
ally seen ; the natives call it " Enfudu," a name that
seems to carry with it a suggestion of blundering stupidity
not altogether foreign to a tortoise. It is somewhat
larger than the European species kept in gardens, and
has a high domed shell, with the anterior and posterior
portions of the ventral plate hinged so as to be capable
of a little movement. I once found a specimen that had
apparently been dropped from a height ; it was a large
one, and the highest part of the dome of the shell had
been fractured, a portion being depressed. It had probably
been destined to be eaten by some bird of prey, which had
been frightened away at my approach, for the blood on
the wound was still fresh. So there may be, after all,
some truth in the old legend which ascribes the death
of a Greek sage to the impact of a tortoise dropped upon
his bald head by a bird of prey, which mistook the shining
cranium for a stone !

SNAKES 187

Snakes.

During one's wanderings through thick grass and
among bush and rocks one frequently came across snakes
hastily retreating as one approached.

It seems to be a fixed idea with some people that a
snake's main object in life is to find a human being and
bite him. I take the opposite view, that so far as man is
concerned the snake is chiefly occupied with avoiding
him lest he do the snake harm !

From the snake's point of view, what possible object
can there be in wasting valuable poison by biting a huge
creature who cannot be eaten, and could, even when
mortally wounded, crush the life out of a sna-ke before
he dies ? So, hearing some heavy footstep come crashing
through the grass, the wise snake departs as quickly as
possible lest he be trodden upon.

This is where a booted man has the advantage of a
silent-footed native, who gives no warning, so that a
snake half asleep, or languorous after a meal, or stupid
and half blind owing to an approaching shedding of
skin, suddenly finds a large animal almost on top of him,
and strikes in a panic of self-defence.

Once only I very nearly trod upon a large black snake,
a species common on the islands, and reaching up to six
feet in length. I was walking through very thick, high
tussocks of grass, which made it necessary to go slowly
and to lift the feet high. In the middle of one step,
while my foot was in the air, the snake crawled out from
a very dense tussock across the very spot for which my
foot was destined. It was fortunately just not too late
to put the foot down elsewhere, and all was well, for
the big snake vanished as suddenly as it had appeared,
obviously bent on escaping being trodden upon.

Puff adders, being very sluggish, will not get out of
the way until forced to do so. I have literally pushed

188 REPTILES AND FISH

one off the track with a stick because it would not move ;
a lovely species, the glossy skin marked with carpet
pattern of different shades of purplish greys and
pinks, soft browns and creams, harmonizing exquisitely
with the lights and shades among the dead leaves
surrounding it.

This was Bitis nasicornis, the puff adder of forests,
with a little horn on each side of the nose formed by
clusters of upright scales. Beautiful as is its colouring,
the shape is so thick and squat that this species loses
considerably in beauty from this cause. The much less
ornamental Bitis arietans, well known in South Africa,
was once found on Damba Island by my canoe-men
hidden among the rocks at the very place where the
canoe was kept. Somehow they managed to get a rope
round it, and dragged it up in triumph to my house.
It was 4 feet 6 inches long, and one could not but be
glad to be rid of so dangerous a neighbour for the bare-
footed men.

Mention has been made of the booming noise uttered
by the crowned crane during the wet season. My boys,
when asked what made the noise, always replied, " E'sa-
lambwa " (puff adder). Some puffing !

The python is another fine snake, handsomely marked,
and when in good condition his glossy skin is really beauti-
ful. I only met with him once, on Kimmi Isle, when I
came upon a small specimen about six feet long lying
on a marshy bit of shore. I watched him for some time,
noting two Glossina fly up from him, one obviously full
of blood. After a while he became aware of me, and
moved slowly away, absolutely silently, with the dignity
befitting a king of snakes.

On Ngamba Island a place was found where the dead
leaves in a circular area were flattened, and there was a
trail of flattened leaves leading to and from it. The
natives said this was the resting place of an " E'timba "

SNAKES 189

(python), which was probably the one previously seen
on Kimmi, for they swim well, as do most snakes.

One day when going round Tavu Isle in the canoe
preparatory to landing, I happened to look over the side
and saw a snake swimming very gracefully, with the
top of its head just above water. It soon took fright,
dived, and was lost to sight in the deep water. It appeared
to be a young python about two feet long.

Another snake which on account of its beauty particularly
calls for mention is a harmless, slender, grass green species
often seen on the lake shore hunting for frogs, whose
pitiful cry when once heard is always recognizable again.
The brilliant grass green colouration and very graceful
movements of this snake always evoked one's admiration.

A similar bright green species might sometimes be seen
high up among the branches of trees, sometimes travelling
rapidly and almost leaping across gaps too big to be
spanned. The manner in which it seemed to spring
across from one tree to another is very interesting, seeing
that in the Malay Archipelago certain snakes have carried
the process further, and take gliding flights from a high
to a lower branch. ^

The last snake to be mentioned is a very curious
and interesting species, which was once found deeply
buried in the base of a termite hill which was being
demolished. It was about twelve inches long, greenish
grey in colour, very smooth, and could move back-
wards or forwards equally well. The eyeless head was
so small that the slightly swollen posterior extremity,
ending in a blunt spike, was larger. When handled the
snake pressed this blunt spike forcibly against one's hand ;
doubtless an ignorant person would have dropped it,
thinking it was biting or " stinging." These snakes are
entirely subterranean, and are found both in Asia and
Africa ; presumably they live on worms and burrowing

1 Shelford, A Naturalist in Borneo, pp. 79-82.

190 REPTILES AND FISH

insects, possibly on termites also. This specimen much
resembled one of the legless lizards found under stones
or logs of wood, which the Baganda call " Namugoya,"
and the English " blind worm " or " slow worm," in spite
of their bright eyes and often quite active movements.
The " Namugoya " is, of course, regarded by the natives
as a snake, and therefore on no account to be touched.

Frogs and Toads.

Frogs have already been mentioned as contributing to
the sounds heard on the islands. The species most often
seen on the shore much resembles the European " Green
tree frog," and may be found freely exposed to the sUn
on leaves or branches. It adapts its colouration to the
intensity of the illumination, and in very light surroundings
becomes very light golden green or even milk white.
Among dark surroundings it is greenish black, but does
not assume brown tints. Another tree frog, very much
smaller, seems to be responsible for the shrill tinkling
noise, like sleigh bells, always heard among reeds and
rushes after sunset. The green frogs are devoured by
the green grass snake previously described, and their cry,
when caught, is pitiful, and surprisingly like that of a
child.

Another species — ^possibly a toad — which is also much
to the liking of snakes is larger, and has a rough brown
skin ; it lives by day in burrows which it digs where the
soil is suitably light. A green snake was seen one day
with its head down one of these burrows attempting to
swallow the occupant, and as the mouth of the burrow
was large and the snake thin, I was able to look past
the snake and see how the toad had distended itself with
air to such an extent that the snake, with'^mouth wide
agape, could not get a grip on the spherical surface.
Though it persevered for an hour, with jaws so widely
apart that upper and lower jaws were a' most in a straight

k L

•<i«

A SMALL "ENSWA-SWA."

A VERY ABUNDANT TREE FROG.

Tofa;e p. 190.

SNAKES AND FROGS 191

line, it could do no more than merely apply itself to the
toad's side, and eventually had to give up the attempt
to swallow such an unmanageable victim ! I got the toad
out, and save for a few scratches made by the snake's
teeth it was none the worse, and, having deflated itself,
crawled away quite calmly as if this were no unusual
experience !

It was interesting to see the toad actually save its life
by its power of increasing its size, and one wonders why
the snake had not caught hold of a limb. Had the snake
been a poisonous species, the prey would have been killed
by contact with the fangs, and even , had it been able
first to inflate itself, its subsequent death would have
ensured its return to normal size.

This brings up a point which had always been a difficulty
to me until I meditated over this and other observations.
If the poison has been evolved by slow degrees from the
ordinary non-poisonous saUva, how could this have been
brought about ? What use could slightly venomous
saliva be to a snake ? For a creature thus bitten might
still be able to go away a long distance and die later,
where the snake would not find it. A clue was given me
during active service in German East Africa, where I
first met with the handsome snake called " Boomslang "
by the Boers, and was discussing the debated question
whether or no it is poisonous. One old Boer told me he
had seen one bite a man's arm, which afterwards swelled
up dreadfully, although the effects were not serious ; but
many Boers said it was not poisonous. This snake is
an " Opisthogljrph," i.e. it has not the long fangs of a
viperine snake, but some of the back teeth are grooved
in part of their length, and a fold of mucous membrane
forms a pouch round the groove to conduct the slightly
poisonous saliva while the snake mouths its prey, for it
does not strike and let go. Now it is easy to see what an
advantage it would be to a snake if, whilst it was hold-

192 REPTILES AND FISH

ing on to struggling prey, the latter should be weakened,
even only a little, by injection of slightly poisonous saliva,
for the prey would then be more easily swallowed. Having
started thus one can see how the poisonous quality could
be developed. For the more quickly the struggling
animal succumbed, the less time would the snake need
to hold on to it, and the less chance would there be of
its escaping, until, with the increasing strength of the
venom, a stage would be reached when the snake would
only need to strike once, and quietly wait a few minutes
until the prey, at once seriously poisoned, falls helpless
and dies.

An extremely interesting observation is here quoted
from a letter in Nature by Professor Poulton, dated
January 2, 1918. Professor Poulton quotes the observation
for another purpose, but it is so exactly illustrative of my
point that it is here repeated. Dr. G. A. K. Marshall
says : " When happening to look over a low stone wall
near Estcourt, Natal, in 1897, I chanced to observe a
small snake in the very act of striking a frog. After being
bitten the latter hopped off at a great pace, and I was
rather surprised to see that the snake made no attempt
at pursuit, but merely followed in a very leisurely manner "
(the italics are mine). " Seeing that the frog had come to
a standstill at a considerable distance off, I crept along
under the wall so as not to disturb the snake, and on
getting near the frog I looked oautiously over the wall
to see the end of the tragedy. The snake was still
some way behind, approaching steadily, and on reaching
its victim stood watching it for some moments with
its head raised, the frog meanwhile sitting trembling
in front of it. At last the snake seized its prey and suc-
ceeded in s war owing it after hut feeble resistance. It
seemed clear that the trembling and inability to escape
on the part of the frog were simply due to the action of
the poison injected at the first bite."

ONE OF THE FLY BOYS HOLDING A SMALL " MAMBA " (lUNG FISH)

SEMUTUNDU.
Foot-rule to show scale.

Mi

iM'

. i- f;

** MALE."

To face p. 192,

MAMBA AND MALE 193

The only remark I would pass on this most apt observa-
tion is that we have no statement from Dr. Marshall
whether the snake really was a poisonous one : it is only
presumed to be so from the effects of the bite.

Fish.

The most noteworthy fish of the lake is the Lung Fish
(Protopterus), known to the Baganda as " E'mamba," and
a favourite article of food with them. It attains a large
size, up to five feet long, and has a flattened body with
broad tapering tail, furnished with a wavy fin along both
dorsal and ventral edges. The Baganda spear the mamba
in shallow water, but it may frequently be met with in
deep, open water, coming up to the surface to breathe
in the same way as does a tadpole. On one occasion
a mamba appearing just behind the stern of the canoe
in which I was sitting emitted the used-up air from its
lungs with such a loud grunt, that I turned round expecting
to see a hippopotamus !

Having taken a new breath, the mamba turns on its
side and gives a peculiar heave of its flat tail as it dives
down again, but, being very conspicuous, it is quickly
perceived by a soaring fish eagle, who stoops to secure it.

Probably they spawn during the rainy season in shallow
water, for shortly after the heavy rains the young may
be found there. I did not find the tough flesh of the
mamba worth eating, but that of a species of Siluridaey
called " E'male " by the natives, is as good as any fish
I know. It is commonly called " Mud-fish " by the
English, but specimens caught in the open waters around
the islands had no suspicion of muddiness, and the firm,
boneless white flesh was most excellent eating. The Male
has a flat, broad head covered with bony plates (see Fig.),
and has long barbels round the mouth. The body is
somewhat cylindrical ; it may measure two feet or a little
over.

14

194 REPTILES AND FISH

The Baganda are fond of eating a small, very bony-
fish that haunts shallow water ; they call it " E'nkeje,"
and fish for it with rod and line, and then spike a number
on a sharp stick and dry them in the sun. This fish is
also much relished by kingfishers ; it may be a species
of perch. I saw one in shallow water off a sandy shore
very busily excavating a funnel shaped pit in the sand.
With great vigour it would push its snout along the bottom
until its mouth was filled with sand, and then swim a
little way away and discharge the load. Whether this
was destined for a spawning place I know not, but the
fish furiously drove away any too inquisitive neighbours.

A very curious fish is termed " E'mbegede " by the
natives. It has a tubular snout turned down at right
angles, and presumably obtains its food from mud. I
saw the remains of one on Kimmi that had probably
been caught by an otter.

Other fish well known to the natives are the " Semu-
tundu," a fish with long barbels, smaller than the Male,
but much the nicest to eat ; a very large Silurid, called
" Akasonzi " ; a species commonly sold in the. market,
and much eaten by the English population at Entebbe,
called " Ensoga," which somewhat resembles the roach ;
the beautiful but bony '* Ekisinja," a barbel of deep olive-
green colour, usually caught with rod and line in shallow
water where ambatch grows ; and a small, very beautiful
silvery species netted in shallow water, called " Omukene "
— it is about the size of a minnow.

CHAPTER X

THE COLOURATION OF INSECTS

The study of the colouration of insects has been attracting
more and more attention from evolutionists, whether of
the Darwinian or Mutationist school, since Bates ^ and
Fritz MuUer 2 in South America, Wallace ^ in the Malay
Archipelago, and Trimen 4 in South Africa gave the first
explanation of the phenomena now known as Mimicry,
and interpreted by these naturalists on the lines of Dar-
win's hypothesis.

In this chapter examples will be given which have
come within my own experience on the Sesse Islands,
and an attempt will be made to show how the Darwinian
explanation is more satisfactory than that of the Muta-
tionist.

In 1890 Poulton classified the colours of animals in
one scheme, embracing Apatetic or " Deceitful " colours ;
Sematic or " Warning " colours; and Epigamic or " Court-
ship " colours.^ The first two headings only will be dis-
cussed.

Apatetic colours are divided into Cryptic and Pseudo-

1 " Contributions to an Insect Fauna of the Amazons Valley." H. W.
Bates, Trans. Linn. Soc, vol. xxiii, 1862, Pt. III.

2 Proc. Ent. Soc. Lond., 1897, p. xx.

3 " On the Phenomena of Variation and Geographical Distribution
as illustrated by the Papilionidae of the Malayan Region." . A. R. Wallace,
Trans. Linn. Soc, vol. xxv, Part I, 1865.

4 Roland Trimen, " On Some Remarkable Mimetic Analogies among
African Butterflies." Trans. Linn. Soc. xxvi, 1870, Part III, 1869.

5 The Colours of Animals, 1890.

195

196 THE COLOURATION OF INSECTS

sematic groups. Cryptic colours are divided by Poulton
into Procryptic and Anticryptic. Procryptic colouring
conceals its wearer from danger, causing it to resemble
either the general surroundings or some particular part
thereof (Special Procrypsis). . Instances of the former,
such as a green grasshopper among grass, or a mottled
grey-brown weevil on earth, are so numerous and well
known that no further examples need be mentioned.

Special procryptic colouring never fails to arouse
wonder from its extreme perfection ; indeed, it has been
said that these resemblances are so minutely perfect that
Natural Selection cannot possibly have produced them.

This complimentary doctrine has been termed Hypertely.

Even in the British Isles stick-like caterpillars abound,
which are quite indistinguishable until they move. Here
one must urge a point which is often overlooked. Natural
Selection, in producing special procrypsis, has not only
altered the shape and colour of the insect, but has pro-
duced deeply seated changes in the neuro-muscular system.
A stick-like caterpillar needs to maintain itself motionless
for hours in an attitude very unusual for the majority
of caterpillars, and one which must require great develop-
men of muscle for that special purpose.

As an example of another attitude, very unusual, but
associated with an especial scheme of colouring, the
following is of interest.

In 1910 I found at Jinja, on the north shore of Lake
Victoria, a moth,^ allied to our English Lucania, among
some very dry dependent spikes of grass flowers, of a
light silvery grey tint. The moth was hanging in an
inverted position with wings brought together over the
back ; their under surfaces thus exposed were of a silvery
grey hue closely corresponding with that of the dried
glumes of grass. But this colour was only shown on those
surfaces that were visible ; the part of the fore wing

* A species of Cirphis.

JULL FED CATERPILLAR OF THE NYMPHALINE BUTTERFLY ATERICA GALBNB

HANGING FROM THE TIP OF A FERN FROND JUST BEFORE PUPATION.

Photographed in Daniba Island Forest.

To face p. 196.

SPECIAL PROCRYPSIS 197

concealed by the hind wing was of a more usual dull
grey-brown colour. The moth when disturbed flew away
to another similar tuft of grass and adopted the same
attitude again, which was most certainly procryptic, and
very unusual in a moth of this group, the members of which
usually rest with the wings brought closely against "the
body. The fact that the silver grey colour was confined
to those parts of the wings where it was necessary for the
concealment of the moth by procrypsis is a very important
point. Many other examples of this nature exist, and are
arguments against such protective colouration being the
result of mere chance. Natural selection produces no
more than is necessary : this is contrary to the doctrine
of Hypertely. Exactly similar limitations are to be met
with when studying mimicry.^

Another beautiful instance of special procrypsis is
afforded by a rare Notodontid moth,^ of which I found
a specimen at rest oh a bush on Damba Isle. The only
specimen known up till then was the type in the British
Museum.

This moth so closely resembled the tube formed by a
dead, dry, roUed-up leaf, that I was for long in doubt
as to its nature, and finally had to pluck the twig on
which it sat and examine the object most minutely before
I could decide.

The wings were closely folded round the body ; the
inner margins of the fore wings meeting over the back
were of slightly darker tint than the rest of the wing,
so as to resemble the thick midrib of a dried leaf, whose
petiole was represented by a curved tuft of long hairs
projecting upwards from the top of the head.

The colour of the wings was the shining brown of a
crisp, dry leaf, and on each side there were several lines
representing venules, and three doubly ringed markings

^ Essays on Evolution, Poulton, p. 240.
^ Scalmicauda niveiplaga.

198 THE COLOURATION OF INSECTS

like fungus spots. Near the tip was an absolutely pure
white marking, shining, which probably represented a
silvery patch such as is often seen on a dried leaf. The
most realistic portion of the resemblance was the deception
produced by the very dark brown of the anterior surface of
the head, which appeared to be the dark shadow of the
interior of the tube formed by the supposed dead and
rolled-up dry leaf.

The completeness of the likeness cannot be realized
when the moth is seen set in the cabinet, but it was one
of the most perfect examples of special procrypsis that I
have ever seen.

Precisely the same attitude is made use of by the " Buff-
tip moth," but in this case the colouring causes it to
resemble a short piece of very dry stick with light grey
bark, broken off square at one end and obliquely at the
other. In the case of this moth there is again to be noted
the profound change necessary for the complete success
of the deceptive resemblance. The deeply rooted instinct
to escape when alarmed has been modified into a stronger
instinct to remain motionless at all costs in order to
maintain the appearance of a dead leaf, for the slightest
movement would attract the eye of an enemy, who would
certainly " investigate the object more closely and perhaps
handle it roughly. The Mutationist who believes that this
deceptive appearance, so minutely perfect, was produced
more or less complete in one step, will have to admit
that the sudden modification of instinct, by a strange
coincidence, took place at the same time, or that at least
it occurred in connection with the other. The Darwinian,
on the other hand, believes that both modifications took
place by equally slow gradations.

Very interesting examples of special procryptic colour-
ation are the resemblances of insects to bird droppings
on leaves. There is a beautiful example in England —
the little moth Cilix glaucata at rest fully exposed on a

- o

c

fe; z
^ <:

a; ^

^§

o P

RESEMBLANCE TO BIRD DROPPINGS 199

leaf.^ The resemblance may be either to a dropping
which, fallen from a height, has spread out to form a white
and mottled grey patch on a leaf, or to one which has
kept its cylindrical shape, and appears dark grey or
grey-brown and white.

The young caterpillars of Papilio dardanus much
resemble the latter object, and their very sluggish habits
contribute considerably towards the illusion.

The former class is well exemplified by a beautiful
Geometrid moth ^ which rests with outspread wings fully
exposed on the surface of a leaf. The wings are of a chalky
white hue, with wavy darker lines, and in the centre of
the fore win^ is a dark, irregular patch of black and brown,
glistening here and there with silver. The whole produces
an effective likeness, which must be seen in its natural
surroundings to be believed. If the moth is frightened
it flies away, but takes up the same exposed position on
another leaf.

A propos the silver marking, it is curious how rare
silver is on insects ; gold is perhaps commoner. Besides
the above case, the under surfaces of Fritillary and
some other butterflies, and the marks on some Plusia
moths, I have only seen one other insect with silver — a
large Saturnid caterpillar, green, with humps on the
back covered on one side with pure silver. The attempt
to rear this beautiful caterpillar was unsuccessful, so the
name cannot be given.

Certain insects which resemble bark, and are to be
found on tree trunks, are very wonderful examples of
procrypsis, though whether it is special or general pro-
crypsis is a matter of words only.

Many Lasiocampidae have caterpillars which are seen
with difficulty when at rest, so closely do they lie against

^ First recorded by Sidgwiek in the Journal of the Rughy School Natural
History Society.

2 Problepsis cegrettUi

200 THE COLOURATION OF INSECTS

the bark which they resemble. On Damba Isle one of
my natives came to tell me he had found some caterpillars
on a tree trunk, and when I went to look he showed me
three Lasiocampids, lying flat on the bark at a point
where its surface was somewhat irregular. While looking
at them and marvelling at the minute detail of the resem-
blance, I gradually became aware that the immediate
surroundings of these caterpillars were not bark but more
caterpillars, and much to the native's amazement, nine
of them were found lying closely packed side by side, and
utterly indistinguishable from the bark except by the
very closest inspection. These caterpillars fed at night
only,^ crawling up the trunk when the sun set and
returning to the same resting place next day. The
perfection of the concealment owes something to attitude. -
The claspers are spread out sideways, so that the body
lies flat against the bark. The surface of the skin is
roughened by numerous little excrescences, and the gap
between the edge of the body and the bark is partly
obliterated by fleshy lappets, and often by short,
soft hairs.*

It is worth pointing out that a very large insect may
be concealed in a way that c nnot be believed, unless
it is actually seen in its natural surroundings. It is well
known that the caterpillar of the Privet Hawk Moth,
with its lilac and white oblique stripes, is not so conspicuous

* A colony of Eupterotidae caterpillars {Chrysopsyche varia) was found
on Bugalla Isle resting by day on a tree trunk. Wishing to obtain the
perfect moth I took some of the caterpillars and kept them in a large
circular glass topped box in which they rested quietly during the day.
When evening came they formed the usual procession and attempted
to climb upwards, which of course they could not do. The result was
extremely absurd, for the caterpillars formed an endless chain, nose to
tail one behind another, which ceaselessly promenaded around the cir-
cumference of the box in the endeavour to find a path upwards. Although
there was plenty of food in the box they would not eat, since the proper
preliminary of marching up the tree trunk had not been performed, and
I had to put the caterpillars back on the tree and hope to secure the
cocoons later. The limitations of instinct are here well shown.

* See diagram in The Colours of Animals, p. 30.

CONCEALMENT OF A LARGE INSECT 201

as would be expected among the leaves and varying lights
on a bushji but the following is a more remarkable example.
On Kimmi Islet I was looking at a bu^h thinly covered
with small leaves, and thought I saw a small brown
" micro " moth sitting on one of the leaves. But gradually
it became obvious that what I was looking at was not a
small moth, but part of the ventral aspect of an enormous
green Saturnid caterpillar, which was on the lower surface
of a twig, its anterior segments being at an angle with
the twig, as in the common attitude of many large cater^
pillars, such as those of our English hawk-moths. The
closely folded brown legs on the first three segments had
made the brown area which, against the green of the rest
of the caterpillar, had been mistaken for a small moth
sitting on a leaf. There was a whitish-green stripe running
along the mid-ventral line of the caterpillar, which must
have served to break up the otherwise uniform green
area, and helped to render the huge form difficult to be
seen. The curious point was that when the insect had
once been visualized, I could not un-see it again, and found
great difficulty in realizing how I had at first failed to
discern it. Some of the natives were called up and asked
if they saw anything on the bush, and their surprise was
as great as mine when I pointed out the caterpillar. The
moth was reared, and proved to be Bunaea phaedusa.

When procryptic colouring is dealt with, place must
be found for mention of Seasonal Variation. Certain
butterflies have been known for some years to have
appearances differing greatly in the wet and dry seasons.
So marked are the differences in the Nymphaline genus
Precis, for example, that two forms of P. sesamus, one
salmon-pink above, the other deep blue, with extreme
difference of pattern and colour on the under surfaces,
were thought to be distinct species until G. A. K. Marshall
bred one from eggs laid by the other. ^

^ The Colours of Animals, p. 42.

' Trans. Ent. Sac, 1902, part iii, p. 414 et seq.

202 THE COLOURATION OF INSECTS

Poulton, from careful study of this genus, concluded
that " dry " forms are more procryptic, and that this
is due to the operation of Natural Selection during the
greater stress of the dry season, when insects are so scarce
that insectivorous animals need to work harder to obtain
food, and the risk to any particular insect is proportionately
greater.! Now the climate of the islands of Lake Victoria
has been shown to be more uniform than that of the main-
land,^ and it is very interesting and significant that in
the case of the commonest Precis there (P. archesia), the
only form met with was the " wet " one ; and of P. sesamus,
the salmon-pink " wet " form was so much the commonest
that the appearance of a " dry " one was quite a memorable
event. During the campaign in German East Africa, I
first became familiar with the phenomena of a typical
dry season, and was soon impressed by the much greater
procrypsis of the " dry " Precis, and also of such Pierines
as Teracolus, Belenois, Pinacopteryx. When the grass is
quite dead and dried up it becomes brown, tinted with
pink or grey, and the dry season Pierines, when at rest,
harmonize most admirably with their surroundings on
account of the suffusion of the under surface with brown,
grey or pinkish scales, so that they are often exceedingly
difficult to detect after they have alighted. It is note-
worthy that, just as in the case of the moth above men-
tioned, this especially procryptic colouring is only found
where it is needed on the under surface, viz. on the whole
of the hind wing, but only on the tip of the fore wing,
which is not covered by the hind wing in the position
of complete rest. The remainder of the under surface
of the fore wing is as in the " wet " form.

The above examples of concealing colouration are all
procryptic ; that is to say, they are species, found highly
desirable by vertebrate enemies, which have acquired

1 Trans. Ent. Soc, 1902, part iii, pp. 424-443.
^ See chart, Chapter iii.

ANTICRYPSIS 208

colours, shape and habits concealing them from dis-
covery by enemies.

The class of anticrj-ptic colouration has not so many
examples. In it are to be found creatures whose con-
cealment aids them, primarily, not to escape their enemies
but to obtain their prey.

On many occasions I have seen a butterfly apparently
caught by a flower which it had visited, and at first imagined
that the proboscis had become entangled. On examining
the flower it was found that the butterfly had been caught
by a spider, which had lain in wait on a flower which it
so closely resembled that, until one had had this experience
several times, one still imagined it was part of the flower.

Two species, or possibly two forms of the same species,
have been frequently met with that thus caught their
prey, both of the " crab-spider " type. One was found
on a yellow flower like a small sunflower, the other was
on a milk-white flower of a small low-growing herb. In
each case the spider was of precisely the same tint as the
flower on which it caught its prey.

The curious and interesting Mantidae are often quoted
as instances of anticrypsis. A species Pseudocreobotra
ocellata was abundant on the islands, and, like the spider,
varied in hue according to the colour of the flowers which
it frequented in order to catch its prey. One form, green
and yellow, was found on the small " sunflower," the other,
green and mauve, sat among the spikes of purple flowers
of an abundant aromatic labiate herb. Very young speci-
mens found on the latter were of a uniform purple or
mauve tint, and it was extremely difficult to see them.

In all these cases, however, the procryptic element
cannot be definitely set aside, since birds are weU known
to eat spiders, and Mantidae are so universally cryptic
that they must be supposed to be highly edible.^

^ In 1916-17 I had a tame Cercopithecus monkey which, taken out
to hunt for itself, would devour Mantidae with the utmost avidity.

204 THE COLOURATION OF INSECTS

But the following example is more probably anticryptic.

On Nsadzi Isle there is a fine sandy beach frequented
by Egyptian geese and wading birds, whose droppings
attract little blue butterflies and Skippers, which settle
on them to feed. I found one day, sitting on a patch
of bird dropping and sucking the juices of a captured
Lycaenid butterfly, a small flat bug, Mononyx grandicollis,
of the group Cryptocerata, whose tints accorded admirably
with those of the dropping on the wet sand.

Hitherto we have been considering examples of insects
concealed by their resemblance to their surroundings.
But a large number of species are extremely conspicuous,
and many of them seem to court attention. I use these
words deUberately, and as a result of several years' obser-
vation in the field. But a school has arisen in America
following the teaching of the distinguished artist-naturalist
Thayer,! vrhich believes that all creatures are concealed
by their resemblance to their surroundings, no matter
how brilliant and startling their colouration appears to
be in the cabinet. Many butterflies are supposed to be
like the flowery part of their surroundings. ^ In some
cases this is certainly true ; the greenish-yellow Pierines
of genus Terias, when feeding from yellow flowers among
herbage, are very well in harmony with them, and quite
well concealed. So also the under surfaces of the wings
of other Pierines tone very well with flowers or grasses ;
our own orange-tip is a v/ell known instance.

But these are all what a Darwinian calls procryptic,
so that here there is no difference between him and the
follower of Thayer. When the species which Poulton
called Aposematic (i.e. with " warning " colouration) are
considered, it is difficult to accept the doctrine that they
really harmonize with their environment or with any part
of it. Take for example the typical habitat on the islands

^ See Concealing Coloration in the Animal Kingdom.
2 Loc. cit., pp. 228-229, New edition, 1918.

THAYER'S VIEWS 205

of a typically aposematic butterfly, Acraea egina. This
brilliant black and scarlet species could be seen in great
numbers on the open grass land near the well defined border
of the dense forest which, as has been described, ends
abruptly, presenting a dense wall of greenery among
which one searches in vain for scarlet flowers or any patch
of scarlet with which the Acraea could harmonize. Near
the edge of the forest, but often as much as a hundred
yards away, in the open grass land, flourish clumps of
the plant Erlangea tomentosa, which bears fine heads
of lavender coloured flowers. These are extraordinarily
attractive to Acraeines, which crowd together on them,
making a brilliant picture. Not only the black and
scarlet Acraea, such as egina, zetes, perenna and the
rarer pharsalus, together with smaller species similarly
coloured, but black and white females of Planema macarista,
P. alcinoe and P. aganice, the brown males of the two
latter, and black, orange and white males of P. macarista,
with both sexes of the similarly coloured P. poggei, all
congregate in various proportions, and being freely exposed
on the lavender heads of flower, make a very brilliant
and conspicuous assembly which by no stretch of imagin-
ation could be considered to harmonize with the surround-
ings.

It is true that one tree with brilliant scarlet flowers,
known to the Baganda as " Ekirikiti " {Erythrina tomentosa),
is plentiful on the islands, but butterflies do not visit
it, and it usually has no leaves at the time of its flowering,
so that the isolated bunches of flowers high up in the air
can hardly be considered to form a background against
which scarlet butterflies would be invisible. Moreover,
the tree usually stands by itself in open places.

It must, I think, be admitted that such insects as the
above are conspicuous in their natural surroundings, and
the explanation, first suggested by Bates, is that these
are relatively distasteful species, to whose advantage it

206 THE COLOURATION OF INSECTS

is that they should be as conspicuous as possible in order
that their vertebrate enemies may readily recognize them
as things to be avoided. The colours of such creatures
are therefore known, in the terminology used by Poulton,^
as " warning colours " or " aposemes."

Obviously it wiU be to the advantage of such insects
that their aposemes should be of simple type to aid recog-
nition and memory on the part of the enemy. Hence
one finds schemes such as black and yellow alternating
rings in a wasp, or, in a butterfly, the wings of one colour
except for strongly contrasted patches of another colour,
especially at the apex of the fore wing. It is an interesting
point that this apical or sub-apical patch of contrasting
colour usually has its long axis at right angles to the long
axis of the wing. Presumably the transverse direction
makes the patch more conspicuous during the movements
of the wing than if it were parallel to the long axis of the
wing. This point is well shown in the illustrations of
the various Planema (see Plates).

Before dealing with typically aposematic species, it
may be said that quite a number of proeryptic insects
are furnished with an aposeme which they display if
an inquiring enemy comes unpleasantly near. The
caterpillars of Lasiocampidae, previously alluded to,
show, if disturbed, a cleft across the second or third
segments, or both, filled with brightly coloured fine
sharp spines, which very readily become detached
and penetrate one's skin, and would prove very un-
pleasant in the mouth of an enemy tasting such a
caterpillar for the first time. The colour of the spines
varies in different species. It may be orange, or pink,
or steely blue-black, but is always such as to make the
spines very conspicuous. When the caterpillar is quietly
at rest and concealed by its procrypsis, the cleft is closed,
but if the caterpillar is disturbed or touched it makes

1 The Colours of Animals, pp. 336-337.

BLACK AND WHITE APOSEMATIC CATERPILLARS OF THE LARGE NOCTUIU
MOTH AMBLYPRORA MAGNIFICA.

PROCRYPSIS COMBINED WITH AN APOSEME 207

the cleft gape widely open, so that the brightly coloured
spines are fully displayed. This may be seen in England
in the caterpillar of the " Lappet " moth.

In the case of these caterpillars it must be remembered
that the presence of such an aposeme does not necessarily
imply that the caterpillar is intrinsically distasteful, but
that it possesses some unpleasant quality on account of
which it had better be left alone. I have a certain amount
of evidence that Hymenoptera, such as the stinging ants,
bees and wasps, are not avoided so much on account of
inedibility as on account of their powers of defence, which
they advertise by their conspicuous scheme of colouring,
or their habits.

So that the combination of procrypsis with an aposeme
is not so inconsistent with the theory of natural selection
as it might appear to be at first. Another interesting
example of this combination is a fine large Noctuid moth,
abundant on the islands [Ophideres). It was usually first
seen on the wing after it had been disturbed from its
resting place, when its orange hind wings bordered with
black were very conspicuous. If it was followed up it
was seen to settle head downwards on the bark of a tree,
expose fully for a moment its orange hind wings, and then
suddenly close the wings with a snap and become invisible
owing to the mottled grey-green and brown tints of the
fore wings closely resembling the surface of the bark.
I am inclined to think that the conspicuous hind wings
are aposematic, else why should the moth take such pains
to display them for a moment, on alighting, before covering
them with the procryptic fore wings ? I suggest that the
moth endeavours to dissuade a pursuing enemy by freely
displaying its warning colours, and then settles and be-
comes invisible to any new enemy that had not seen it
when on the wing.

On the other hand, it is possible that the colour of the
hind wings is of the nature of directive markings,^ which

1 Colours oj Animals f p. 204 et seq.

208 THE COLOURATION OF INSECTS

entice the enemy to seize an insect at some particular
part which is especially brittle, and, being seized, breaks
away, so that the insect escapes unharmed save for the
wholly immaterial loss of some part of the wing. Such
special directive marks often take the form of conspicuous
eye-like spots at the angle of the hind wing or along its
margin, or of tails, and every collector must have noticed
how often in Papilio, CharaxeSy Satyrinae, Lycaenidae such
parts are missing, and a gap in the margin of the hind
wing shows where the enemy has secured nothing but a
mouthful of dry wing tissue. These gaps in the wing
are just in the position where the wings are least likely
to bo damaged by contact with objects during flight.
Moreover, it is just these pieces of the wing which break
off when the butterfly is vigorously fluttering in the folds
of a net ; the tails of Papilionidae are notorious in this
respect.

On one minute islet that I visited there were numbers
of black and white wagtails, and little Lycaenid butterflies
were plentiful. It was noticed that a large proportion
had a A-shaped piece missing from both hind wings
symmetrically as if a wagtail had seized the butterfly
by the hind wings when it was at rest, and the butterfly
had escaped owing to that portion of the wing breaking
away. After leaving the islands in 1915 I was fortunate
enough to see this actually happen, and also to see
numbers of Lycaenidae actually eaten by wagtails.^ The
question of directive markings has been here dealt with
because it seems possible that certain instances of very
simple, conspicuous patches of colour, which in the dead
specimen look like aposemes, are possibly directive mark-
ings. I refer especially to the orange, crimson, or purple
tips of the fore wings of the Pierine genus Teracolus.
These lovely butterflies are conspicuous on the wing, but
shy and wary. When one is fluttering in the net the

1 Vide p. 238

BEHAVIOUR OF APOSEMATIC SPECIES 209

orange tips very frequently are broken, which is an un-
usual occurrence with butterflies, the anterior part of the
wing usually being the strongest (c/. supra). As has
been said, the lower surfaces of these butterflies are de-
cidedly procryptic, but it must be acknowledged that,
as in our English " orange-tip," the bright colours of
the Teracolus are more developed in the male, so that
sexual selection may have been a factor in producing
them {vide infra).

We must return to the subject of aposemes, having
digressed somewhat.

Just as cryptic insects must behave in accordance with
the resemblance to their surroundings, so aposematic
species have a behaviour in keeping with their colouring.
It would be of Httle use for a gaudy, distasteful insect
to retreat timidly or endeavour to hide itself on approach
of an enemy which, only partially seeing it, might grab
at it and damage it fatally before the mistake was found
out. Everything is to be gained by flaunting the banner
in the face of the enemy, so that there can be no possibility
of a mistake. Consequently, aposematic insects are of the
boldest demeanour, and are the species most frequently
seen by the casual observer. If endeavours are made
to catch one, for instance a Planema butterfly, it merely
flits out of reach, and returns to the same spot apparently
unalarmed.

A large grasshopper {Dictyophora laticincta), of family
Acrididae, was frequently met with on the islands fully
exposed on pathways : a bloated creature with heavy
body, whose wings are so atrophied that it cannot possibly
fly. So sluggish is it that it scarcely troubles to move
out of the way, or gives a feeble hop merely carrying it
for a distance about equal to the length of its own body.
In colour it is grey-black, and the short wings are scarlet,
and there is some of the same colour along the sides of
the fat abdomen. If persistently annoyed it will raise

15

210 THE COLOURATION OF INSECTS

the wing covers to show the red wings underneath. Such
an insect, a priori, must be distasteful, and proved to be
so when offered to my young monkey, who would eat
greedily until he was sick equally large, but procryptic,
grasshoppers.

Since I offered this grasshopper to Wee Man he naturally
thought it would be edible, and at once seized it, though
he did not straightway bite it as he would have done
had it been the ordinary procryptic species he was accus-
tomed to eat. While he held it in his hand the grasshopper
emitted copious bubbles of strongly smelling yellow froth
from the thoracic spiracles, forcing it out by first dis-
tending and then strongly contracting the abdomen, so
that a hissing sound was produced, audible several yards
away. At the same time the wing covers were raised
so a3 to display the bright red, black bordered wings.

Wee Man was obviously much interested in this very
curious and, to him, novel phenomenon, and tasted the
froth, but though he obviously did not like it, he persisted
in trying to eat the insect, pulling it to pieces and tasting
each. But none of it was actually eaten I

Many other insects, none of which are procryptic, emit
strongly smelling, bright coloured fluid when roughly
handled. This is well known to all who have collected
Acraeine butterflies.

The following is quoted from my article in Bedrock ^ :

" A most typically aposematic Arctiid moth {Rhodo-
gastria leucoptera) was found resting fully exposed on low
herbage. Its wings were of a pure hard shining white,
but not very thickly scaled, so that when they were brought
together over the body of the moth, the abdomen, which
was of a bright rose pink, was distinctly visible. The
thorax was pure white, spotted with black. The legs
were of the same bright rose as the abdomen. When
the moth was disturbed it separated its wings and spread

* Vol. ii, 1913, "Notes on the Struggle for Existence in Tropical Africa."

BEHAVIOUR OF APOSEMATIC SPECIES 211

out the legs so as to display the bright pink (a typical
aposeme), and emitted from the thorax just behind the
head a copious yellow froth, till a mass of yellow bubbles,
with a very strong acrid odour and taste, projected on
each side."

Equally remarkable is another specialized grasshopper,
known to the Baganda as " Semukutu." ^ It is a liarge
heavy insect of dull greenish-grey colour, and though
not in any way aposematic in colouration, yet has all
the characteristics of such an insect. It has no trace
of wings or wing covers, and the thorax is set with spines
around its margin. The abdomen is fat and bloated,
and the sluggish insect makes no attempt to evade an
approaching enemy. I could never induce a monkey to
tackle one of them. When handled, a Semukutu ejects
with some force from the sides of the thorax a yellow
fluid of strongly acrid odour, which would prove extremely
unpleasant in the mouth ; and, indeed, the Baganda will
not even handle the insect, for they say the fluid produces
sores on the skin.^ But I never found that it had any
effect on my skin, and I have handled a great many
Semukutu. This is an interesting case, for here we have
an insect which has acquired the habits and attributes
of an aposematic species, yet not the colouration. Indeed,
a young specimen is rather procryptic, being mottled with
green, but it has not then the power of exuding the
acrid fluid. One might, however, suggest that the appear-
ance of this stout, spiny, wingless grasshopper is so unlike
that of others that it can be very easily recognized, and
the need for aposematic colours may not have arisen.

The above observations on the monkey and grasshoppers
have exemplifled the deliberate use of an aposeme by an
insect when hard pressed. This is again brought out
very well by an experiment with fowls, the subject being

^ Enyaliopsis, possibly durandi.

2 See Bulletin of the Entomological Research Committee, vol. i, p. 227.

212 THE COLOURATION OF INSECTS

Phymateus viridipes, another large grasshopper of a hard
green colour, with spiny thorax, and fully developed blue
and crimson wings, so that when flying it is extremely
conspicuous.

One of these was put on the ground some little distance
away rom a group of fowls, one of which at once ran
up to it. The grasshopper remained where it was, but,
when the fowl came dangerously near, raised the wings
and wing covers perpendicularly and opened out the
former to show their bright red colouring. The fowl
halted, looked at it, turned round and walked away,
nor could it be induced to touch the insect. After a
similar result had been obtained on another occasion,
the grasshopper was killed and laid down near the fowls,
the wings being of course hidden beneath the covers.
A half-grown fowl at once ran up and pecked at it, and,
being pursued by another, ran off with the grasshopper,
put it down and pecked at it again, but certainly did
not seem to relish it, finding it very tough. The fowl at
length picked off the legs, but made no impression on the
body.

Just as it was about to leave it, a second half-grown
fowl rushed up, took it away and pecked vigorously at
it, pulled off the head but did not eat it, and finally
walked away and left it. These observations show,
firstly, the value of the aposeme in warning off dangerous
enemies, and secondly, the distastefulness of the owner.

It will have been noted that the fowl made no impression
on the tough body of the grasshopper, and this brings
out another point. Should an aposematic insect have
the misfortune to be one of those that are tasted by an
inexperienced young foe, it may quite possibly survive
the tasting, and be able to propagate its kind if it is of
sufficient toughness. Consequently one finds that apose-
matic species, in strong contrast to procryptic species,
are endowed with astonishing powers of resisting injury,

HAIRY CATERPILLARS 213

at any rate in the case of Lepidoptera. It is quite impos-
sible to kill a Danaine or Acraeine butterfly by the mild
pinch on the thorax sufficient for a non-aposematic Pierine
or Satyrine, for their tissues are so elastic that but little
impression is made on them. Further, such insects have
remarkable powers of resistance to chemical poisons.^
It takes a very long time to kill an Acraea, or a Burnet
moth, in a cyanide bottle, and I have had Planema pupae
all night in a kilHng bottle, which when taken out in the
morning were unharmed, and showed their characteristic
movements, although that very bottle was in daily use
for other insects, which very quickly succumbed.

It has been pointed out above that a disagreeable taste
or smell is not the only quality associated with aposemes,
as in the case of the sting of a wasp. The well known
irritating qualities of fine hairs from Arctiid, Lymantrid,
or Lasiocampid caterpillars, or the spines of Apodid or
Saturnid caterpillars are all found to be associated with
typically aposematic colouration. In this case the aposeme
has no bearing on the actual taste of the insect, but only
refers to the presence of some irritant or penetrating
hairs ; indeed, many such caterpillars, as in Lymantridae,
become typically procryptic, and presumably edible moths.
But the rule cannot be stated that when the aposeme is
due to adventitious hairs, etc., the insect is not distasteful
in itself, for many Arctiid caterpillars renowned for their
hairiness, such as the " Woolly Bear," become typically
aposematic moths, such as our English " Tiger-moth."
It would, however, be very interesting to test the edibility
of Lymantrid and Arctiid pupae removed from the hairy
cocoon. On the other hand, when a caterpillar has
aposemes that are, as it were, intrinsic, that is, not de-
pending upon hairs or spines, it seems to be the rule hat

^ It must be acknowledged, however, that certain very procryptic
weevils, found on the ground which they closely resemble, have equal
powers of resistance to chemical poisons, but they are exceptional in
this respect among procryptic species.

214 THE COLOURATION OF INSECTS

its distasteful qualities are carried through all stages
to the perfect insect. I do not know of a case of an
aposematic larva of this type becoming a procryptic,
and therefore edible, adult. A most typical instance is
the Hypsid moth, pactolicus, which is equally conspicuous
in all its stages. The caterpillar is ringed alternately
with dead black and pure Chinese white, the head being
crimson. It feeds fully exposed on a yellow flowered
vetch, 1 on which it is visible from far away. When full
fed it stretches a few silk threads across and across to
make an open hammock, in which it becomes an equally
conspicuous chrysalis, fully exposing its light yellow
colouring with black markings. The large adult moth
is bright orange with steely blue-black blotches, and is
one of the most conspicuous insects ; it is of sluggish
habit, and rests freely exposed on low herbage, or flies
slowly and heavily. The moths of this family are some-
times copied by other less favoured insects, and there is
little doubt that they are relatively distasteful.

On the other hand, cases of a procryptic caterpillar
becoming an aposematic adult must, I think, be exceptional.
I have not yet met with one on the islands.

A very interesting species is the typically aposematic
genus Aletis. These moths are sometimes mimicked by
butterflies (seep. 233). One species, A. erici, was found
in abundance on Sanga Island in such numbers that the
bushes had been completely stripped of their leaves
(see p. 103). The caterpillar was tawny orange with black
blotches, and was extremely conspicuous ; since many
grouped themselves together, hanging head downwards in
clusters from the" bare twigs, their attitudes accentuated
the conspicuous colouring — they had the typical shape
of all Geometridae. Inasmuch as the vast majority of
Geometrid larvae are procryptic, the ancestral forms were
probably so, and the Aletis has been transformed from a

^ Crotalaria striata.

APOSEMATIC INSECTS CROWD TOGETHER 215

procryptic to an aposematic type, the larvae still retaining
the characteristic shape and the attitude which follows
from the absence of the first three pairs of claspers, but
acquiring the habit, most unusual among Geometrid larvae,
of grouping themselves so that the effect of the warning
colours is accentuated — a habit not at all uncommon
among aposema.tic insects.

The Pentatomidae, highly odonieronsHemiptera, resplend-
ent in green or blue and gold, commonly mass together.
In 1917 I met with a very marked instance of this
aggregation. An Acridian grasshopper,^ conspicuously
coloured green, orange and black, and of typically apose-
matic habits, was proved very definitely to be distasteful
to a young monkey. When very young it is coal black,
speckled with yellow, and on several occasions large
numbers were found closely crowded together to form a
black mass at the end of a spray of herbage, which attracted
my attention, whereas a single individual from its small
size would easily have escaped notice. Some of these
black youngsters were given to a monkey, who would
not even taste them, although hungry, as was proved
by his subsequently eating other insects. Although the
habit of collecting in a mass, whereby the conspicuousness
of an individual is much exaggerated, is most commonly
made use of for aposematic purposes, yet I have met
with an instance where a full effect of procrypsis was
only produced when a number of individuals collected
together in a certain way. The effect produced was a
likeness to a bird-dropping on a leaf, and could not so
well have been produced by a single individual.

The species concerned is the Bombycid moth, Trilocha
ohliquissima, of which the type specimen, from Angola,
was the only one in the British Museum collection, A
company of very young caterpillars was found on Bugalla
Island on a leaf of a Sapotaceous tree, Chrysophyllum.

^ Zonocerus elegans.

216 THE COLOURATION OF INSECTS

The general colouration was a mottled stony grey. After
the first moult the lighter parts of the markings became
chalky white, and as the caterpillars always lay close
together in a mass, and fed on the flat surface of the
leaf from the upper parenchyma only, the resemblance
to a bird dropping was very close. At the next moult
the caterpillars became darker, and as they grew larger,
after the third moult, were of a rich brown. They now
had different habits, feeding from the edge of the leaf
in the manner of all large caterpillars, and no longer
massing together.

When full grown they were extraordinarily procryptic,
being of the same rich brown as the leaf stalks, with green-
ish and purpUsh mottlings, the body being swollen here
and there with numerous scale-like excrescences accurately
resembling scales at the bases of the leaf stalks, and
rugosities of the bark. The caudal horn, projecting
forward over the back, also helped in this likeness. The
caterpillars now rested along twigs and leaf stalks, and
adopted a special attitude, which rendered them more
procryptic.

An instance has been given of the use of an aposeme
on certain occasions only. The following is a very re-
markable instance of a structure useful at one stage of
development being apparently made use of to protect
a subsequent stage.

The larva of the Lasiocampid moth, Chrysopsyche varia,
is quite conspicuous, being coloured a rich chestnut brown
with large pale blue spots. When full grown it has in
addition on the dorsum of each segment from 4-10
a dense patch of very short, closely set fine hairs of
glistening ;w^hite.

It has been already mentioned as showing peculiar
limitations of instinct (p. 200). When full fed, it spins a
firm brown cocoon of the type usual among Lasiocampidae.
The point of interest is that at one end of the cocoon,

USE OF LARVAL APOSEME BY PUPA 217

opposite to the posterior extremity of the occupant, a
weak spot, almost a hole, is left. "\ATien the caterpillar
becomes a chrysalis the discarded brightly coloured larval
skin is pushed out so that it partly projects through this
hole. It is naturally to be expected that the adult moth
would emerge through this hole, but it does not : it pushes
its way out in the usual manner through the other end of
the cocoon. Xow if the object of the weak spot is to
allow the discarded larval skin to be extruded so as not
to take up space inside the cocoon, one would expect
it to be got rid of completely, but this is not the case ;
it remains partly inside, partly outside,^ its bright colours
freely visible. Apparently the pupa makes use of the
discarded aposematic skin to protect itself. I have not
met with another case of this partial extrusion of the
larval skin through an orifice especially provided.

Insects furnished with well marked aposemes are
commonly said to be protected by the possession of sting,
distastefulness, spines, hairs, irritant fluid, exceptional
hardness, etc. It must be understood that this immunity
from attacks by birds or mammals is not by any means
claimed to be absolute at all times, but is entirely relative,
depending upon the abundance of insect food generally,
of that species in particular, and the state of hunger of
the enemy, who, when food is difficult to obtain, will eat
species which he would pass by when more edible insects
could be easily found. ^ Also it must be remembered that
certain creatures appear to be specialized for devouring
prey which others pass over.

"^ In a certain number of cases the skin is completely extruded —
e\'idently the habit, if it is of protective value, has not yet been carried
to its fullest efficacy.

2 See Poulton, E.^^B., The Colours cf Animuls, pp. 180, 181. Marshall,
G. A. K., The Bionomics oj South African Insects. Trans. Ent. Soc.
Lond., 1902, part iii. Swyimerton, C. F. M., Ptoc. Ent. Soc. Lond.,
Februarj' 3, 1915; and in "Experiments and Observations bearing on
the Explanation of Form and Colouring." Journal Linn. Soc, 1919,
Zoology, sxxiii.

218 THE COLOURATION OF INSECTS

Thus a Roller which was examined at Jinja was found
to have devoured numbers of a bright green ball-rolling
dung beetle, a member of a family whose habits render
them nauseous to monkeys, and which emit a very foul
smelling fluid when handled.

Bee-eaters (Merops) feed on bees and other stinging
Hymenoptera,^ cuckoos appear to feed largely on very
hairy caterpillars. Now it is well known that every living
thing produces very many more offspring than can possibly
survive, and Wallace pointed out, in 1858,^ that, of the
offspring of one pair, on the average all but one must die
or at least fail to produce young. So that any one who
refuses to accept the explanation that aposematic colours
and habits are produced by Natural Selection is quite
justified in saying, " If you claim that these species escape
being devoured by vertebrate enemies, the burden is laid
on you of proving by what means they are prevented
from overrunning the earth."

In order to answer this, all the restraining factors must
be considered.

Firstly, there are the vertebrate enemies, birds and
beasts, against which it is claimed aposematic insects
are protected. Secondly, predaceous insects, such as
dragon flies, carnivorous beetles and bugs, and Asilid
flies, which eat or suck juices of other insects. Spiders
are not here included, for they appear to be indiscriminate
feeders on protected and unprotected alike. For the
same reason fossorial Hymenoptera, which feed their
young on other insects, are not here included. Thirdly,
there are parasitic Hymenoptera and Diptera, which lay
their eggs on other insects ; and fourthly, the micro-
organisms of disease.

^ See Reports of Sleeping Sickness Commission, vol. xiv, 1913, pp. 16,
16, and table extracted therefrom in Chap, viii, p. 172.

^ " On the Tendency of Varieties to depart indefinitely from the Original
Type." Essay reprinted in Natural Selection and Tropical Nature, 1895
edition, pp. 24-5.

ENEMIES OF PROTECTED INSECTS 219

Let the number of offspring of any two parents destroyed
by each of these agencies be represented, in order as
above, by V, P (with above described Hmitations), p and
M : then if X be taken to represent the total number
of offspring, X — 1 represents the number destroyed by
all agencies together. So the equation can be written :

X — l = V+P+p + M.

And the number destroyed by any one agent alone

V = (X — 1) — (P + 2? + M).

It is obvious, therefore, that if, as is claimed, V is negligible
for protected species, V-\-p-\-M. must be correspondingly
greater, and there is some reason for thinking that this
is actually the case, although much statistical evidence
is required.^ For instance, in the case of P, those who
have paid attention to, or observed in the field, predaceous
insects with their prey,^ have probably had the same
experience as myself, that the prey is very often among
precisely those species which are believed to escape
molestation by vertebrate enemies. Thus I found a com-
pany of spiny black and yellow larvae of Acraea perenna,
of which many had already been sucked dry by a bug,^
which was found in the act of sucking one. This one
bug might very easily have destroyed seriatim the whole
of the brood of these aposematic larvae. In the case
of p we need especially comparative statistics showing
the relative proportions of aposematic and cryptic species
that are destroyed by parasites ; the following bears on
the subject. Acraea zetes, a typically aposematic scarlet
and black butterfly, abounded on Bagalla Island, and its

1 See Trans. Ent. Soc. Lond., 1902, part iii, pp. 328-38.

2 See collected data by Poulton, Trans. Ent. Soc. Lond., 1906, part iii,
pp. 323-409.

^ Damarius splendidulus.

220 THE COLOURATION OF INSECTS

larvae, ringed alternately with orange and slate grey,
with black spines and crimson head, could be found in
numbers when full fed wandering over the grass land in
search of suitable places for pupation. Judging by the
situations in which the pupae were found, the larva chooses
as conspicuous a site as possible, such as a tall grass stem
fully exposed and unsheltered, quite out in the open.
Numbers of these full fed larvae and of the equally con-
spicuous pupae were collected, and out of seventy, 77 per
cent, were destroyed by parasitic Diptera {Tachinidae)
and Hymenoptera {Chalcididae, Braconidae).

Lastly, in the case of M. The very peculiar and highly
specialized legless caterpillars of the Apodidae are often
brilliantly coloured, and furnished with intensely sharp
strong spines set in clusters, often said to be "poisonous,"
so irritating are their qualities. The Baganda have a
great dread of these stinging caterpillars, which they
brought me with the utmost caution. I was very seldom
able to rear them, for they seemed to be extraordinarily
susceptible to some disease which very rapidly killed
them. Possibly it was of the same nature as the
" pebrine " which Pasteur investigated in silkworms.^

A very interesting special case of aposematic colouring
is the development of " Terrifying " marks, and just as
with the cases previously described, the colouration is
associated with a special attitude, whose purpose is to
make the specialized areas of colouring suggest some-
thing that frightens or discourages the enemy. In many
cases the attitude is far more important than the colouring.
I would refer the reader to Poulton's book,^ and wish here
merely to describe an example from my own experience.
The large and handsome moths of the family Saturnidae
very often, like our " Emperor moth " of the heather

^ Comparative statistics of the number of caterpillars, procryptic and
aposematic, destroyed by p and M would be exceedingly valuable.
'^ The Colours of Anitnals, p, 258 et scq.

TERRIFYING MARKS AND ATTITUDE 221

moors, have large eye-like markings on the hind wings,
which are concealed by the fore wings when the moth
is at rest, but under certain circumstances are exposed.
One of these, a comparatively dull coloured species, was
offered to my pet monkeys on Bugalla Island. The
following account is quoted from my article in Bedrock.^

" The moth was a large yellow species with well marked
eye-like spots on the hind wings. When alarmed it bent
the body ventrally into a strong curve, and held the wings
in a very curious and unusual fashion — almost upright,
with the upper surface directed forwards so that the
eye-like markings were extremely conspicuous ; indeed,
the attitude was obviously intended to display those
* eyes.' The moth thus looked curiously weird and un-
mothlike, and the monkeys were afraid even to touch it.
It was not merely the size of which they were afraid,
because they caught and readily devoured large and
protectively coloured moths (Sphingomorpha) often found
about the house."

Finally, it must not be concluded .hat all bright colours
in insects are aposematic. In the case of butterflies both
surfaces of the wings must be considered ; only the lower
surface is displayed when the insect is at rest and likely
to be caught unawares, for when on the wing a butterfly
is alert and can escape by flight. Consequently a large
number of species, the upper surfaces of whose wings are
vividly coloured, are really procryptic if the lower surfaces
be considered. I need only mention our English " Red
Admiral," and the famous leaf butterflies of Asia and
Africa {Kallima).

Darwin concluded that the colours of butterflies owed
much to Sexual Selection, and those who have seen butter-
flies courting will probably agree that vision does enter
into the question. I have seen a male Papilio dardanus

* " Notes on the Struggle for Existence in Tropical Africa," Bedrock,
1913, vol. ii, p. 366.

222 THE COLOURATION OF INSECTS

in headlong flight through the forest suddenly arrest its
course and dally awhile with an Amauris niavius, and it
was impossible to resist the conclusion that the swallow-
tail had for a moment mistaken the Amauris for its own
female of the form hippocoon, which closely resembles
Amauris. Eltringham ^ has described an elaborate appar-
atus in male Amauris which I was fortunate enough to
observe being actually used in courtship by A. psyttalea,
on Bugalla Isle, in July 1912. A male was flying about
after a female, which presently alighted on a dead flower
spike of a common herb about two feet high. She sat
almost at the top, vertical, with head upwards and wings
outspread, and remained perfectly still while the male
hovered a few inches above her head with a peculiar
flutter causing him to rise and fall a little. Every now
and then the flaps at the extremity of the body were widely
everted at right angles to the body, and a large white
brush-like structure was most energetically protruded and
as rapidly withdrawn. I watched this for a minute or
two, and then, to my surprise, for I had made no move-
ment, the female suddenly flew away as if the performance
had not appealed to her, and the male followed. ^

The importance of the scent emitted by male butterflies
was first recognized by Fritz Muller,^ who described certain
scales on the wings especially formed for producing scent.

A very curious and unusual occurrence was noted on
Bugalla on October 25, 1912. A male of a small and
abundant Syntomid moth {Epifoxis albicincta) was on a
grass stem, and a male of the common butterfly Acraea
terpsichore form ventura, in a state of great excitement,
was endeavouring to effect union with the moth, passing

1 Trans. Ent. Soc. Lond., 1913, part ii, p. 399 et seq. ; also 1915, part i,
p. 152 et seq.

^ For observations showing the relation of these anal tufts to the
" brands " on the hind wing, see Lamborn, Proc. Ent. Soc. Lond., 1911,
p. xlvi ; 1912, p. xxxiv ; 1918, p. clxxii.

3 Jen. Zeit., vol. xi. p. 99; Trans. Ent. Soc, 1878, p. 211.

COURTSHIP OF BUTTERFLIES 223

the tip of its abdomen repeatedly over that of the moth.
It was the more curious that there were numbers of freshly
emerged Acraea females in the neighbourhood with which
the male butterfly could have mated. Professor Poulton
suggested that possibly it was a case of accidental likeness
between the odours of two species very far removed from
each other, which implies that the male seeks for the odour
of the female as well as trying to charm her with his own.

White butterflies of the genus Belenois have often been
observed courting. A male for a long time flutters round
and pursues a female, who, perhaps wearied by his im-
portunities, settles at last with wings expanded. The
male alights behind her and commences vigorously to
"paw" her hind wings with quick movements of his
anterior legs, the action being as if he was trying to walk
up a very slippery surface. Refusal on the part of the
female seems to be indicated by a curious attitude. The
wings are spread out flat, but not quite at right angles
to the axis of the body, for the fore wings are directed
slightly backwards. The abdomen is directed vertically
upwards at right angles to the thorax.

In 1917 I witnessed the consummation of a courtship
of a Pierine, the initial stages of which had not been
observed. A male and female Pinacopteryx simana were
fluttering in the air and the female suddenly settled,
quite quietly, and sat with wings closed. The male stood
over her with wings widely expanded so as completely
to conceal her ; the union was the act of a moment, and
then they flew away, the female hanging passively with
wings closed, the male carrying her.

Nymphalines such as Precis and Byblia, however, adopt
different methods of courtship. The female, with wings
closed over her back, faces the male, who endeavours
to induce her to allow the union by butting against her
closed wings with the anterior margin of his own, brought
together over his back. It appears as if the male's

224. THE COLOURATION OF INSECTS

endeavour is to approach from the flank or rear, and the
female, unless willing, turns so as always to face him.

In the case of Acraeines little time seems to be wasted
on persuasion. A male, in the air, makes for the female,
who at once falls fluttering to the ground, where the two
seem to struggle together. If the female is willing, union
is. almost immediate ; if not, the male soon flies away.

On August 2, 1918, in Portuguese East Africa, I saw a
very good illustration of this. A fresh specimen of Acraea
natalica was flitting aimlessly about, and I was awaiting
an opportunity of catching it, when an old and very worn
female came by. The male at once went to her, and she
fell straight to the' ground. I watched for several minutes.
The male was very excited, and gripped the female with
his legs round the roots of the wings, but she lay motionless
on the ground adopting an attitude of passive resistance,
the wings folded against her body and pressed closely,
so as to prevent the male thrusting his abdomen down to
meet hers, which he struggled to do. After a little he
managed to pass his abdomen between her wings, but
she made no response, and he suddenly gave up and flew
away. I caught him, and then found the female still
lying motionless on the ground, and secured her too.

In connection with the pairing of butterflies, the following
is of interest to evolutionists.

During the nuptial flight it is known that in the case
of some species the male carries the female, in other cases
the positions are reversed, the one that is carried remaining
absolutely passive with wings closed. In a discussion on
Sexual Selection among insects in the Descent of Man,^
Darwin remarks on some cases where the male is *' less
bright " than the female,^ and points out that in the
English Pieris, Colias and Epinephele janira the duller male
is supported during the nuptial flight by the female, " so

^ Chapter xi.

^ The I^ycaenid butterfly Laehnocnema bibulus has a dull brown male
which is pursued in courtship by the brighter blue-marked female.

NUPTIAL FLIGHT OF BUTTERFLIES 225

that the part which the two sexes play is reversed, as is
their relative beauty," and Darwin goes on to say that
it may be supposed that the female takes the more active
part in the wooing. Since the middle of 1917 I have paid
particular attention to this point, and a large number
of pairs have been noted on the wing. In the Pierines,^
in all cases save one, the male has carried the female
as above described. So that it would appear that Darwin
was misinformed regarding Pierines, unless our English
species have different habits. My few observations of
Nymphalines show that the female carries the male.
This appears also to be the case with Lycaenidae.
Acraeines show no definite rule. In the case of Satyrines
the female seems to carry the male, and I have seen
Papilio dardanus form hippocoon carrying the male.

We now come to the consideration of Mimicry, one of
the most fascinating subjects for study that insects can
provide, and one on which a large amount of work has
been done in recent years, from the point of view of the
Darwinian hypothesis, thanks to the influence of Professor
Poulton, of the Hope Department at Oxford.

The study of mimicry was first taken up by Bates,
who, on the Amazons, found that a large number of
abundant butterflies were of the type previously alluded
to as aposematic, but that others, although superficially
resembling the former, belonged to groups anatomically
very different from them.^

He put forward the explanation that these latter lived
on the unsavoury reputations of their associates, or, in
other words, were Mimics of more distasteful models.
He found that whereas the models were abundant, easy
to catch, and avoided by birds,^ the mimics were less

1 Proc. Ent. Soc, 1918, p. clii. * Loc. cit.

^ Bearing on this is the fact that Acrseines and Danaines rarely show
evidence of attacks by birds which is so common among Lycaenidae,
unprotected Pierines or NjnnphaKnes, Satyrines, etc., in which a ^»
shaped piece is synunetrically cut out from both hind wings, showing
the nip of a bird's beak.

16

226 THE COLOURATION OF INSECTS

abundant, shy, and easily alarmed, and there was no
evidence that they also were distasteful. On reflection
it is quite obvious that a mimic must be less abundant
than its model, for if, of a series of conspicuous insects
met with some are edible and others distasteful, an enemy
might find it worth while to catch every one he saw on
the chance of it being edible. If, however, the proportion
of edible to distasteful is very low, the chance of finding
an edible specimen is too small to make it worth while
catching large numbers that will not be eaten.

The mimic, as has been said, is easily alarmed, and it
is not difficult to understand the reason. It has been
pointed out above that aposematic species, which serve
as models, are extremely resistant to harm, and even
if caught and tasted do not of necessity suffer vital
injury. Mimics, however, come from a stock which has
in the past relied for safety upon concealment or rapidity
of flight, and have not developed the toughness of their
models ; hence a nip by a bird's beak or the rough handling
of a monkey may cause fatal damage.

So that one finds by experience that so long as mimics
are not alarmed, they have a great resemblance in move-
ment and habits to the aposematic species with which they
are associated. If a mimic is frightened it takes to flight
and dashes off at a great speed, whereas the model
will merely move a little way, and often come back
to the same spot, or may not even deign to move away
at all.

Here again is well exemplified the action of natural
selection in producing changes, not only in colouration
and shape, but in the nervous system also, yet only so
far as such changes are required. Thus in certain mimetic
South American Pierines, a small part of the fore wing
that is normally concealed by the hind wing retains the
hue from which this group has derived the name of
" Whites," while the visible areas of the wings are mimetic

IMPORTANCE OF ATTITUDE IN MIMICRY 227

of a brightly coloured model. ^ Again, instincts are
modified so long as the assumed attitudes and move-
ments are needed to correspond with the false appear-
ance of inedibility, but when these are no longer helpful,
rather a danger, the recently acquired instinct of deceit
is abandoned, and the more deeply seated instinct of
flight comes into action.

Attitude is of verj?^ real importance in producing
mimetic resemblance, and living specimens often appear
mimetic when in the cabinet they do not. So that the
ignorant remark, which may sometimes be seen in
print, that mimicry is the product of the imagination
of an " arm-chair philosopher " need not be seriously
considered.

One day, on Damba Island, I was looking at a nest
of the tree ant, JEcophylla, among the leaves of a bush.
This species, by means of silk spun by a larva held for
the purpose by an ant, fastens together leaves to make
a more or less globular nest about the size of a cocoanut.
Wishing to obtain a few specimens I proceeded to box
them, when, to my great astonishment, one of them
jumped, thus revealing itself to be no ant, but a mimetic
spider,^ which had been running about unperceived
among the ants !

The wonderful thing about this case is the great
anatomical difference between ant and spider, both
internal and external. The model has head, thorax,
and abdomen separated from each other by well marked
constrictions, but the spider has head and thorax fused
into a cephalo -thoracic mass, separated from the abdomen.
The ant has three pairs of legs, the spider four ; the ant
has a pair of long mobile antennae, the spider none.
Diagrammatically the difference is shown overleaf.

^ Poulton, Essays on Evolution, p. 239.

" Myrmarachne Joenissex. Another spider mimic of the same model ia
described and figured byShelford in A Naturalist in Borneo, pp. 230-1.

228 THE COLOURATION OF INSECTS

These differences are sufficiently obvious in the dead
specimens, and the quick nervous movements of a live
ant with waving antennae are as different as possible
from those of a typical spider.

Nevertheless in this case the spider completely deceived
me. How was it done ? Firstly, the shape is unusual
for a spider, the cephalo -thorax and abdomen being thin
and prolonged, and a constriction in the cephalo-thorax
of the spider represents the division between head and
thorax of the ant. The spider did not use the first pair
of legs for walking, but they were held up in the air and
waved about to copy the movements of the antennae
of the model. It is obvious that this functional modi-
fication necessitates a profound alteration in the neuro-
muscular system of the spider. It is worth noting, too,
how in the last extremity the spider disclosed its true
nature in its endeavours to escape by jumping, a habit
common to the family to which it belongs {Salticidae),
but suppressed by the new mimetic instinct until an
emergency arose. This spider was not the only mimic
of jEcophylla, for running among them, and almost as
closely resembling them as did the spider, was a small
bug of the family Capsidae,^ which proved to be not only
an undescribed species, but sufficiently distinct to be
placed in a new genus.

It is of course easier for a bug than a spider to be brought
to resemble an ant, since both are insects, and natural
selection has the same basis to work upon. It might
be supposed that the wings and wing covers of the mimic
would prove a difficulty, since they lie over the abdomen,
but in Xenetomorpha there is a constriction near the
base of the wing covers which, seen from above, matches
the constriction between thorax and abdomen of the
ant. The bug lives among its models, probably sucking
the juices from the stems of the leaves among which
^ Xenetomorpha carpenteri.

0^.

QSj-a-crmi-n:

mx

?c4*

SPIDER .

IMAGINARY CASE OF MAMMALIAN MIMICRY.
Drawn by Dr. H. Eltringhani.

To face p. 228,

IMAGINARY MIMETIC RABBIT 229

the nest is made. The spider catches flies, and on one
occasion was actually seen to run into the outer spaces
of the nest with its prey. The wonderful nature of this
mimicry of an ant by a spider may be brought out by
an imaginary, but parallel, example for the benefit of
those unaccustomed to deal with insects, and unfamiliar
with the important differences between ant and spider.
It is probable that tortoises are not eaten by those
carnivorous animals which prey upon rabbits. Let us
imagine a farmer in the country seeing what he thought
was a tortoise, and idly watching it crawling about in
the characteristic tortoise fashion. He goes up to it and
frightens it, when, to his astonishment, it bounds away
with the typical rabbit gait. He shoots it, and on ex-
amination finds all the essential features of a rabbit :
it has rodent teeth, but the external ears are so small
as to be unnoticeable. The main resemblance to a tortoise
is produced by a matting together of the hair on the
back to form a carapace-like structure, ^ and on the legs
to look like scales, while the legs themselves are much
reduced in length, but very thick.

A species which " assumes a virtue though it has it
not " is a true mimic, and is said to be Pseud-aposematic,
or to have false warning colours, for it appears in the
guise of another more fortunate than itself in the possession
of distasteful qualities. But when the mimetic association
first described by Bates was more fully investigated,
some of the members were found to be quite as abundant
as the species which they resembled, and to belong to
groups which could be claimed to be as well protected as
the models. It was pointed out previously that a mimic
must be less abundant than its model. Fritz Muller.
first pointed out how a distasteful and abundant species
can gain by resembling another equally distasteful and

^ Compare the so-called horn of a rhinoceros, which ia structurally
merely agglomerated hair.

230 THE COLOURATION OF INSECTS

abundant. In a brief preliminary statement in 1878 ^
he says : " What advantage can it be to a creature pro-
tected by repellant odour to resemble another similarly
protected species ? If their foes avoid protected species
by ' instinct,' none at all ; but if, on the contrary, as
appears so much more probable, the foes have to learn
their unpalatability by experience, then the benefit is all
the greater the less numerous the species. The advantage
gained by two unpalatable species by their resemblance
is in inverse ratio to the square of their numbers."

This quotation may be amplified a little. A certain
number of lives of any species must be sacrificed in
teaching inexperienced enemies what to avoid. But
if the loss of insects of a certain definite aposematic
pattern, instead of being borne by one species only, could
be distributed over several resembling each other, the
loss borne by each species would be only a small pro-
portion of the total loss, while the lesson would be equally
well learnt. Moreover, it will be of further advantage
in that there will be fewer patterns to tax the enemy's
memory.

So now we have to consider a second type of resemblance,
characterized by the presence of Common Warning Colours,
or, to use Poulton's term, Syn-aposematic. Although the
term " Mimicry," first used by Bates, should refer to
an edible species masquerading as inedible, it is loosely
used to cover the likeness between two distasteful species,
which is a matter of a different order. Properly speaking,
Mimicry is Pseud- and not /S^i/w-aposematic.
, Instances of syn-aposematic resemblance may be seen
in England, such as the yellow and black bands of different
kinds of wasps, or the red and green of the Burnet moths.
These insects being so closely related, it may with reason
be argued that it is not surprising that Burnet moths
of a single genus should have a close similarity.

1 See Proc. Ent. Soc. Lond., 1915, pp. xxii, xxiii.

SYNAPOSEMATIC ASSOCIATIONS 281

As a better instance may be cited a number of Hymen-
optera of different groups v/hich I found on the islands,
characterized by dark grey or dull black colouring with
a segment of the antennae and the tip of the abdomen
conspicuously white, and the transparent wings partly
clouded to form a pattern. Several species of Fossorial
and Parasitic Hymenoptera were collected that belonged
to this syn-aposematic association ; in one case, a Scoliid,
the white on the antennae was produced by white hairs,
and was not, as in the other species, the colour of the
integument. This is an interesting example of a point
with which Professor Poulton has dealt at some length,^
that natural selection has brought about the same effect
in different ways on different subjects, according to the
material offered for selection. Natural selection can
originate nothing ; it can only work with and modify
material offered to it by variation. This point can never
be kept too much in prominence.

The best known syn-aposematic association centres
round the undoubtedly distasteful and conspicuous Mala-
coderm beetles of the family Lycidae — such insects are
very conveniently spoken of as " Lycoid." The Lycid
aposeme is a general colouration of bright orange-brown,
with the extremity of the abdomen and elytra black.
The antennae and limbs are black or black and orange.
The beetles are found collected in numbers ^ on flowering
bushes, are of sluggish habit, slow, heavy flight, unafraid,
and often exude a droplet of yellow fluid when handled,
and have been proved to be distasteful to vertebrate
enemies.^ G. A. K. Marshall figured 4 a large number
of Lycoid insects belonging to many orders of insects.
In beetles and bugs (Hemiptera) the species are coloured

^ See Essays on Evolution, pp. 2G4-6 ; also Punnett, Mhnicry in
Butterflies, pp. 40-2.

2 Proc. Ent. Soc. Land., 1917, p. Ivii.

» Trans. Ent. Soc. Lond., 1902, Part III, pp. 347, 391.

* Ibid.

232 THE COLOURATION OF INSECTS

in the same way as in the Lycidae. In the clear winged
Hymenoptera the wings are tinted with orange, and, at
their tips, with black, so that when brought together
over the body they produce the Lycoid effect. Lycoid
beetles are numerous, I have met with them among
the families Cerambycidae, Chrysomelidae, Cetoniidae, and
their closer Malacoderm relations, Lagriidae. Of Hymen-
optera, the fossorial Sphegidae and Pompilidae, true
wasps, and parasitic Braconidae, all had Lycoid repre-
sentatives on the islands ; also Hemiptera.

A small moth, of family Lithosiidae, abounded among
long grass on Bugalla Isle, and as it rested with its wings
superimposed upon its back, the resemblance to a Lycid
was quite striking. This was less so in the case of a new
species of Zygaenid moth ^ which I was fortunate enough
to find. The wings were laid along the side of the body
in the usual Zygsenid manner, so that the flat appearance
of the model beetle was lacking. Many of these Lycoids
could be found at the same time as the Lycid beetles
on the flowers of Haronga madagascariensis, a tree or
shrub very abundant on the islands, and known as the
" Gamboge " tree.

The Lycid aposeme is not confined to Africa. Shelf ord^
gives a long list of Lycoid insects met with in Borneo,
some of which are syn-aposematic, while others are
probably pseud-aposematic.

Mimicry, using the term in its widest sense, it to be
found with every degree of relationship between the
insects concerned. I have drawn up a table, from examples
met on the islands, showing this graphically. At the top
of the column of species will be found the pair in which
model (uppermost) and mimic are as closely related as
possible, being members of the same genus. At the
bottom of the column model and mimic are as far apart
as members of the Phylum Arthropoda can be. Between

^ Saliunca egeria. ^ A Naturalist in Borneo, p. 241.

MODEL AND MIMIC

233

■lysd qoBa
JO !}sonij3(Idn Sujaq ispom aq^ ; pa^^possB jSiiBoifjGmira Eapsde jo ei\vi

02

a

ll

l=! "d

zetea
boisduvalli

petiverana
leonidas

1 i

1 1
^ ft

o a

03

i 1

' V

1

9' =3

%

1 1

' 1
P

05

s

« -§

^ 03

& a

1

Y

«<1

a

a s
•S ft

2 a

» .a
=« a
.a °

a 3

(3 S^

1 1

W Pi
<1 3

1 1
1

1-^

2 5

O c3

1 i

1
1

!2;

II

a n

c3 a>

a i

o u
O 1^

05

C«

05 "^

c3 '3

1 §

2 03

•3 -a

a :s

o 13

pR CO

i

O

n
p

Y

o
o

1
o

c3

C3 05

■2 "to
3 a

1 1
1

o
-I!

o

Y

ft

1
■ft

05
Hi

1

S ft
H 5

s

a

O 03

a c3

a i

V

"8

05
I

(8

2

1 j

g

H
P4

Y

o
ft

1

-"I

§a

|S>?

M <8t3

Ph

O m

05 05

03

m a .

tj rt no
05 a: n

§a

Kan

— TOOJj peqoeps iSioiratw jo saidui«xa

234 THE COLOURATION OF INSECTS

these two is a series of pairs, model uppermost, showing
every grade of relationship.

Mimicry, looked at from one point of view, is merely
an example of resemblance to an especial part of the
environment, and is strictly analogous to special pro-
cryptic resemblance. Both are forms of Apatetic coloura-
tion.^ In one case the insect is like a leaf or twig which
is of no interest to an insectivorous enemy ; in the other
it resembles another species which the enemy rarely
eats, knowing it to be distasteful. There seems no logical
reason for supposing that the latter example is beyond
the power of natural selection any more than the former,
yet in spite of much recent work on these lines by the
Hope Department, and also by naturalists in the field,
it is held by some that the resemblances are fortuitous,
or the result of causes which together or separately are
not able to explain facts as are the Batesian and MuUerian
theories, based on natural selection.

An objection is sometimes urged somewhat as follows.
Since the potential numbers of patterns are limited, and
the number of species large, it is quite possible that the
same pattern may have been accidentally produced in
two species without there being any meaning to be attached
to the resemblance, except mere coincidence.

Dr. F. A. Dixey dealt with this argument at the
meeting of the British Association at Birmingham in
1913.2 He showed a number of examples of Lepidoptera
from different continents with approximately the same
colouration and pattern, and to these cases it would
of course be absurd to ascribe a mimetic significance.
But it is important to note that such coincidence has
gone no further than to produce the crudest of patterns
and contrasts, such as a light subapical bar on a dark
fore wing, and the likeness even then is not so close as

^ Colours oj Animals. See classification at end.
2 See Report, p. 518.

THE SELECTIVE AGENTS 235

between species of Planema and their mimics, forms of
Pseudacraea.

Coincidence has proved itself unable to produce such
cases as those in South America described by Bates,
where the colouration is a comparatively complicated
mixture of red and yellow on a dark background. These
cases, indeed, almost seem to be an exception to the rule
stated above, that aposematic patterns are extremely
simple.

The geographical variations of a mimetic species,
according to the different species serving as models in
the various subdivisions of its habitat, such as in the
case of Acraea alciope ^ or Pseudacraea eurytus (see next
chapter), with the transitional forms at the meeting
points of two areas, are extremely difficult to explain,
except by the theory of a causal connection between
model and mimic.

One of the great difficulties felt by many to militate
against acceptance of the theory of mimetic resemblance
produced by natural selection is lack of definite evidence
of a selective agent. It is claimed that it has not yet
been proved that birds eat butterflies in sufficient numbers
to produce selective action, and, moreover, that they have
in certain cases been seen to devour just those species
which are claimed to be exempt. ^ This latter point has
been already dealt with. I think that few would claim,
for instance, that the sting of a bee is of no protection
to it because the Bee-eaters appear to be especially fond
of them. A mimetist would account for such cases as
the ashy wood swallow devouring Danaines in this way.

Regarding the general question, there is plenty of
indirect evidence of attacks by birds upon butterffies. AU
collectors must have noted the ^-shaped gaps in the hind
wings, symmetrical on both sides, that could only have

1 Bedrock, No. 1, April 1912, pp. 57-64.

2 Punnett, Bedrock, vol. ii, July 1913, pp. 169-60.

236 THE COLOURATION OF INSECTS

been produced by a portion of the wings enclosed in a
bird's bill breaking away, as I have . previously described.
As for direct evidence, G. A. K. Marshall first collected a
number of records of attacks by birds actually witnessed,^
but C. F. M. Swynnerton has recently produced such
overwhelming evidence,^ that there can no longer be any
doubt about the matter. His many careful observations
have shown that attacks on butterflies may very easily
be overlooked unless an observer is especially on the look
out for them.

For instance, the late F. C. Selous said : " I have never
once seen a bird eat a butterfly in Africa." [African
Nature Notes and Reminiscences, p. 9.) But I do not
see that the negative evidence of a hunter, who did not
pay so much attention to small life as to big game, weighs
against the positive evidence brought forward by Swynner-
ton and others, who have paid special attention to birds
and butterflies from this point of view.

In fact, Selous's statement only shows how much an
observer may fail to see outside his particular sphere
of interest !

It appears to me that observation of one class of life
renders it difficult to observe another ; indeed, there are
three definite fields, requiring radii of observation of
different lengths. First, and nearest, are insects whose
small size renders it necessary to look for them near at
hand. I find that I instinctively keep my vision on objects
within about ten yards' radius when I am out in the field,
insects being my natural prey. Big game can hardly
be expected to be seen within this area ! If it is desired
to observe birds, the radius is longer, and one looks not
on the ground or low herbage near at hand, but at bushes
and trees at least twice as far away, and for mammals

1 Trans. Ent. Soc. Lond., 1902, Part III, pp. 353-71.

2 Proc. Ent. Soc. Lond., 1915, pp. xxxii-xliii ; Journ. Linn. Soc,
1919 ; Zoology, vol. xxxiii.

BIRDS AND BUTTERFLIES 287

the radius is much increased. I have often noticed that
when out for insects I notice very little of the bird life,
and if I want to see birds have to determine to ignore
insects. How much the more is this necessary for
mammals !

If an entomologist fails to observe distant game,
equally so must the hunter, who is looking out for distant
animals, be oblivious to much that goes on near at hand.
The habitual gaze of such a practised hunter as Selous
would be fixed beyond the shorter radius necessary to
observe small birds and their prey. This seems a possible
explanation of the failure of a great hunter to see what
I have seen on several occasions in a fraction of the time
that he spent in the field. I think, speaking from my
own experience, that one is too apt only to notice such
birds which, perched on conspicuous twigs or tree tops,
look as if they were going to catch butterflies, such as
Bee-eaters, Drongos, etc. At any rate, that has been
my error on the islands, and I never saw a Bee-eater
attack a butterfly, although I found several pairs of wings
of Terias lying on the ground under a stem where Melit-
tojphagus meridionalis had just been sitting. Had I read
Swynnerton's work before I collected butterflies, I should
certainly have seen, as Swynnerton says, " a sudden
sharp movement at the back of a flower head or the
quick dash of a bird over the top of a pannicle on which
butterflies and Hymenoptera may be feeding together ; "
for birds often flew away at my approach to the flowering
bushes of Haronga, on which great numbers of butterflies
fed.

Since I left the islands I have seen a bush frequenting
bird, probably Campephaga nigra, the black cuckoo-shrike,
very quietly eating a Pierine butterfly {Belenois), but
had I not heard the quick flutter of wings and the snap
of a bill as it caught the butterfly, my attention would
not have been drawn to this case.

238 THE COLOURATION OF INSECTS

In 1915, while on active service in German East Africa,
I saw two wagtails at puddles eat eleven butterflies in
five minutes, and on another occasion saw one eat eleven
butterflies in three minutes, definitely selecting for its
food Lycaenidae, Belenois, Atella, from among a crowd
of butterfiies assembled on the mud, and paying no atten-
tion to a number of the typically aposematic Danaine
Amauris echeria, or the black and scarlet Papilio
ridleyanus, mimetic of Acraea egina.^

Lastly, it may be said, when dwelling on this question
whether birds do or do not by eating butterflies cause
mimetic resemblances, is it not too often forgotten that
butterflies are not the only mimics ? No one will deny
that spiders are eaten by birds, or flies, or beetles, or
grasshoppers, yet numbers of mimetic members of all
these groups are known, all explicable by the same theory
as mimicry among butterflies, and all must stand or fall
together.

Dr. W. H. Longley has suggested a modification of
Thayer's views on concealment, supposing that all
species have been forced to assume colour combinations
which most effectually conceal them in their normal en-
vironment ; and that in a few cases patterns have appeared
which have been sufficiently alike to deceive enemies
which discriminate in their choice of food.^ This hypo-
thesis thus evades the difficulty of understanding the
gradual building up of a mimetic likeness while acknow-
ledging that mimicry does exist and is of value to the
mimic, and that different degrees of edibility exist, but
it affords no explanation of the difference in behaviour
between model and mimic. If both Planema tellus and
Pseudacraea eurytus form terra, are equally well concealed,
why should the former be so bold that it can be plucked

^ Proc. Ent. Soc, 1915, pp. Ixix.-lxxv.

2 " A Revised Working Hypothesis of Mimicry," The American Natural'
ist, vol. li, May 1917, p. 276.

LONGLEY'S HYPOTHESIS 239

from a flower by the fingers, while the utmost caution
is necessary if one wishes to catch the mimic ? Dr.
Longley lays stress on the assertion that " colour and
habit are associated variables," which hardly seems to
be consistent with the great differences in habit between
model and mimic of the same colour and pattern.^

Again, I would point out that it is difficult in the
extreme to think of all larvae, for example, as concealed
by likeness to their surroundings. On our common
hawthorn bush, for instance, caterpillars of the gold
tail moth, or of the " Figure of eight " moth {Diloha
caeruleocephala) thrust themselves upon one's notice,
and side by side on the same bush highly modified
Geometrid larvae are only to be discovered by careful
search, so similar are they to twigs. A large Hypsid
larva (pactolicus) abounded on a papilionaceous plant
growing on sandy shores of the islands of Lake Victoria.
It was marked with alternate rings of dead black and
purest Chinese white, with red head and legs and long
black or white hairs, and was visible clearly from as
far away as a creature of that size could be visible. It
developed into a gorgeous orange moth, with blue-black
blotches on the fore wings and a black border to the hind
wings, which was a brilliant and conspicuous object at
rest or on the wing. It freely exposed itself, and was
of sluggish habits and slow, heavy flight, as it would be
expected to be on the Darwinian explanation. Yet
Dr. Longley would have us believe that it is really con-
cealed, and offers no consistent explanation of its habits.

The great attention that is nowadays being paid to
the principles of heredity as expounded first by Mendel,
and to the theory of " Mutations," has resulted in
attempts to account for cases of mimetic likeness by sup-

^ In the case of the Lycidae the larvae are carnivorous and Uve in the.
open ; those of a very close Longicorn mimic live in dead wood. At
the stages when the future similar colours of the adults are being pre-
pared the habits of model and mimic are as dissimilar as possible.

240 THE COLOURATION OF INSECTS

posing that a mimic was produced suddenly in the likeness
of some other species, by a large variation known as
a mutation, and that the Mendelian principle perpetuated
this unchanged, the pattern, etc., being due to presence
or absence of certain " factors," or of factors inhibiting
others.

If mimics are produced by mutations, it is remarkable
that not only superficial aspect, but movements and habits
should be produced that, quite fortuitously, are extra-
ordinarily like those of some other species in the same
neighbourhood, and it seems highly remarkable that such
mutations should resemble species of the type described
as aposematic. How is it that mutation does not pro-
duce new forms resembling procryptic species ? Why
should all the females of P. dardanus be modified by
mutation to resemble different species of conspicuous,
relatively distasteful butterflies of genera belonging to
two different sub -families ? Why has not a form of the
excessively variable species Pseudacraea eurytus (see next
chapter) been produced that deceptively resembles some
procryptic species ? This most wonderful genus has
only two out of over a score of forms that are not close
copies of some other butterfly, and yet the species
resembled are all aposematic ! It seems incredible that
mutation can produce only pseud-aposematic variation.
Where, then, are the other less conspicuous forms ? No
other conclusions seem possible than that they have
been destroyed by enemies before they could establish
themselves. If this be so, the argument that insectivorous
enemies do not destroy butterflies in sufficient numbers
to have a selective influence falls to the ground.

As a matter of fact, a few cases are known of such a
variety differing very considerably from its parents to
a> degree that could be claimed as an instance of the
sudden appearance of a mutation. ^ But they are ex-

1 Bedrock, vol. i, pp. 63-4 — an example of a rare large variation in
Acraea alciope.

INHERITANCE OF SMALL VARIATIONS 241

cessively rare compared with the numerous finely gradated
forms that may be found connecting a mimetic form
with a non-mimetic or transitional between two mimetic
forms.

That birds do exercise discrimination and do destroy
large numbers of butterflies there can no longer be doubt,
thanks to Swynnerton. It seems difficult to avoid the
conclusion that the conspicuous mimics are preserved
by selection, and it is on the question how they arose
that Darwinians and mutationists join issue.

The study of a mimic of wide distribution and changing
form, such as Acraea alciope, Pseudacraea eurytus, Charaxes
etheodes, shows by the transition stages that a mimic
is not suddenly turned out complete. Evidence may be
found in the publications of the Entomological Society ^
that the small, often quite small, variations on which
a Darwinian depends, are heritable, whereas, according
to the mutationist, such " fluctuations " are not handed
down. On the other hand, the mutationist claims that
his large variations are irrevocably fixed, and, when once
formed, do not retrogress towards the parent form. Thus,
for instance, the wonderful forms of Pseudacraea eurytus,
mimicking different species of Planema according to the
locality, should, if the model ceased to exert its influence,
stUl continue to show the same great degree of resemblance
without the occurrence of any intermediate variations.
The next chapter, however, shows that this is not the
case.

^ Poulton, E. B., "Heredity in Six Families of P. dardanus, subsp.
ceneay Trans. Ent. Soc, 1908, Carpenter, G. D. Hale, " The Inherit-
ance of Small Variations in the Pattern of Papilio dardanus." Trans.
Ent. Soc, 1914.

17

CHAPTER XI

PSEUDACRAEA EURYTUS

The Nymphaline genus Pseudacraea is closely allied to
the more widely spread Limenitis, but is confined to
the Ethiopian regions, including Madagascar. The name
is an excellent one ; for out of a considerable number
of species, although the number has been much reduced
during recent years owing to the fusion of many " species "
into one, only two are non-mimetic. The remainder
resemble for the most part members of the Acraeine genus,
Planema, a few mimic Acraea itself, one mimics the Danaine
genus Danaida and others Amauris.

One of the most remarkable points in the resemblances
is. the fidelity with which a single polymorphic species
mimics sundry species of models of very different appear-
ances in different localities ; and where a model is sexually
dimorphic the sexes of the mimic faithfully copy the
corresponding sex of the model. Yet in the very locality
where this dimorphism of model and mimic exists other
species of monomorphic models are closely copied by
monomorphic forms of the same species of mimic.

The closeness of the resemblance is most remarkable,
and is indicated by names such as deceptor, simulator,
imitator, etc. Indeed, as Eltringham says in his Mimicry
in African Butterflies, the deception caused Hewitson
himself, a violent opponent of Bates's explanation of
these resemblances, to comment upon the strange simil-
arity between Acraeine and Nymphaline.

242

THE FLIGHT OF PSEUDACRAEA 243

I have many times been deceived by these mimics
in the forests on the islands, until long association and
close familiarity enabled me to differentiate. But even
then, after a pause of a few weeks during which they
had not been seen, it was by no means easy to say
at a glance whether a specimen seen suddenly in the
forest was Planema or Pseudacraea. Certainly the mimics
are not so bold as Planema, and if really alarmed will
show it in their hasty flight, whereas the model is much
more stoical, and may be easily picked off a flower with
the fingers.

The flight of Pseudacraea is often of the " floating "
nature, especially when several of these butterflies are
flying rather high up, in the sun, and round about a tree.
By this the mimic can be distinguished, but some species
of Pseudacraea, e.g. poggei, do not show the " floating "
flight to the same degree, and thus poggei bears an even
closer resemblance to its model, Danaida chrysippus,
whose flight is much more of the " flapping " type.

Professor Poulton has written as follows about this
flight in a letter dated September 5, 1912 : "I am very
interested in the non-floating flight of the most distaste-
ful butterflies. But I have seen a kind of floating flight
in D. plexippus in America. Still this is not like our
Limenitis, and this latter I suspect is like Pseudacraea.
A floating flight for display, accompanied by alertness
and activity when alarmed, seems to be characteristic
of the second category of distasteful insects in MuUerian
combinations ; it is also true of the true Heliconinae in
South America, mimics of the Ithomiinae.^^ Again, in
another letter, written March 6, 1912, Professor Poulton
said : "It is an extraordinarily interesting genus ; its
habits, I should think, are rather like those of our
Limenitis, and the degree of distastefulness I should
think about the same."

The genus Pseudacraea, therefore, is to be regarded as

244 PSEUDACRAEA EURYTUS

an example of the " transition from Warning to Mimetic
colours '^ described by Poulton in the Colours of Animals,
pp. 221-223, that is, these butterflies possess a certain
degree of distastefulness compared to some others, but less
than others which they resemble, and so benefit by the
more unpleasant reputation of these latter. They are
Syn-aposematic and not Pseud-aposematic : MuUerian
and not Batesian mimics.

I have noticed a peculiar sheen on the wings, especi-
ally on the under surface, that has been found very
useful in enabling me to distinguish the nature of a
specimen seen sitting far out of reach on a leaf in the sun.

The forms with which we are concerned in this chapter
frequent forest, and are not to be found outside. But
flowering shrubs at the edge of a forest will often prove
extremely attractive, so that Pseudacraeas will collect
thereon in numbers, coming from the shady forest to
the bright sun at the edge, when they may be found
on the flower heads until shortly before sunset. On
Bugalla Isle, in 1912-13, I had a favourite walk along
the edge of the forest between 4*30-6 p.m., visiting the
gamboge trees {Haronga madagascariensis), which, when
in full bloom, supplied many fine Pseudacraeas, Planemas,
and insects of many other kinds.

It is not proposed to mention here more than the
forms of eurytus which occur in Uganda, together with
the first known form, now known as eurytus itself, in
West Africa, and the Eastern and South African forms.
The valuable paper by Dr. Karl Jordan at the first
International Congress of Entomology ^ and the beauti-
fully illustrated book by Dr. H. Eltringham on African
Mimetic Butterflies, will enable the reader to gain some
idea of the most wonderful polymorphism of this species.

Dr. Jordan (loc. cit.) figures thirteen West African

males " considered to belong at least to seven distinct

species," but says that " as a result of my inves-

1 lere Congr6a International d'Entomologie, Bruxellea, AoAt, 9110,
pp. 385-404. Plates xxi.-xxiv.

PLATE il ''^^^

1. Planema aganice form montana 3", Ngamba Island, L.

Victoria, 11 vii, 1914.

2. Psei^daeraea eurytus form rogersi ^ (the type specimen)

Caught by Eev. St. A. Rogers near Mombasa, November, 1906.

3. PI. aganice form montana ? , Tavu Island, 1 vii, 1914

4. Ps. ewytus form rogersi $ (the type specimen).

Caught near Mombasa, 28 Tii, 1906.

5. PI. aganice form nyascB ^ , 13 vi, 13.

Caught by S. A. Neave on Mt. Mlanje, Nyasaland.

6. Ps. eurytus form mlanjensis ^ , 11 xii, 13.

Caught by S. A. Neave on Mt. Mlanje, Nyasaland.

7. PI. aganice form nyasoe $ , 10 vi, 13.

Caught by S. A. Neave on Mt. Mlanje, Nyasaland.

6. Ps. eurytus form mlanjensis 5 , 17 iii, 13.
Caught by S. A. Neave on Mt. Mlanje, Nyasaland.

9. PI. aganice 3" , 13 iv, 1897.

Caught by G. A. K. Marshall, Malvern, Natal.

10. Ps. eurytus form imitator S ,1 v, 1910.

Bred by the late A. D. Millar, near DurbaUj Natal.

11. PI. agardce ? , 22 iii, 1896. . ••'^-

Oaught by G. A. K. Marshall, Malvern, Natal.

12. Ps. eurytus form imitator $ , 20 iv, 1910.

Bred by the late A. D. Millar, near Durban, NataL

244 PSEUDACRAEA EURYTC

nsition from V
Poulton in the '
. ;^hese butterfliet j..
Iness compared to Bome othr
thati others which they ff^'r^^'^x. ^^^ ^^ benefit- by ;
raore u!. "' "^t repiUaffo..; -EA^ff^se latter. They • .' -

and not Batesian mimics. JIGI ,iiv XI ,;Biioioi7

.aOGI ,117 8S ,J3aficfmol/[ iboxi id^ueO
M ,iT 81 , ^ s)gAHvjv ra-ioi ar,*^»tt« .S^ .a

.81 ,iiz 11 , % iszn^^ii^SiW caiai z^^mi^ 'ai\a

the bright sun at •h.gi :i^Oi ^^^^^.^i^'imro^^oi^^S vtf
on^ the flower •fifl>Bkgij{K,,.a[:iiiiM iM^o.ov^aPi'.A .a ^d d^a^io '' ' ^
BugaUa Isle, i- .€t\iii tl /? emva^iwikw mioUiviw-vwa ^^ S'

■^^^■'- <''■ ^^ < '& a^wf nQ» .x^ fe"

.Ifi*£M ,aio7f£M ,Ueife:f£M .3 .A .O ^d W§xr^6'' \'''"^-
.Oiei ,7 Y , ^ toi»ikw mioi ^wi^wa .2^ OX
deim ^a^ii-sua Men .T^niM a A aiai 3di jd W/fl' ' ^- •

.6681 jii 22 , $ aowm^to'!J%^J^l'^
^lJ '^^^^^ ,m9vl«15 :nfid3iBM .3 .A .0 y:d iji§0^el f , i)

Wtx-.v X. . •- . ; -OXei.viOfi, ? ■toiwim^ jcaiol 8A5i'^'vwV:8g: M''
The valuable p??:'^'^'*^^ '^^« '^^"^^ •« -A ai^i add ^d fiaia
International Congress of Entor;

';• '^ the render to ^ain Bome I

PLATE II.
AcraeJjxe TnodjUs. N ympJioJjirue rrdmics .

T..C Knight del. et chromo

12.

Hut.h imp

MODEL SPECIES OF PLANEMA WITH MIMETIC FORMS OF
PSEUDACRAEA EURYTUS. ^fsixaub. svze.

Rp^rodjLced- by permzssiorh of ihe, ErctOTnological Sodety of Londoru.

I To face- Page- ZA-6).

THE TYPE 245

tigations I must regard these insects as forms of only
one single species, Pseudacraea eurytus (Linnaeus, 1758)."
Along with these forms of one species are figured thirteen
di-fferent species of Planema, each a model for a form of
eurytus ; and on another plate " five white banded female
specimens of Pseudacraea eurytus, together with the
females of five species of Planema.'"'

Pseudacraea eurytus ^ itself was named by Linnseus in
1758. It is sexually dimorphic. The male is tawny
orange and blackish brown, the female black and
white, with the white areas roughly corresponding to
the pattern of the orange in the male, but being more
contracted on the hind wing. It is a West African form,
mimicking the West African dimorphic species Planema
epaea. I have, however, taken on Bugalla Island a single
female closely approximating to the typical eurytus ;
it is intermediate between two forms described below,
namely tirikensis and terra, having the fore wing pattern
of terra, but the black and white colour of tirikensis, while
the pattern of the hind wing closely approximates to that
of eurytus and thus differs from the pattern of tirikensis.

In 1919, in a small isolated patch of forest on the
Kyagwe (mainland) north coast of the lake, I took a
fine male which is not essentially different from the
typical male eurytus.

All the Pseudacraeas mentioned in this chapter,
except kuenowi, are now known to be forms of eurytus,
using the name in its widest sense. It has a distribution
throughout the tropical and sjibtropical forests of Africa,
from west to east, and from northern Uganda to Pondo-
land, south of Natal. It is, however, very scarce in
East Africa and Rhodesia.

The form imitator was described from Natal by Roland

Trimen in 1873 : it mimics Planema aganice and, like

the model, is sexually dimorphic, although the difference

in colour between male and female is not so marked as

^ See the two coloured plates for illxistrations of models and mimetic
forms of this species.

246 PSEUDACRAEA EURYTUS

in the first named form. Colonel Bowker, in a letter
to Trimen, said it was " quite impossible to distinguish
between this butterfly and aganice either when settled or
on the wing." (Plate II, figs. 9-12.)

The male and female have the same pattern on a
black ground, the pattern being cream coloured in the
male, while in the female it may be either cream coloured
or white. The female may thus resemble either the
male or female of the model, which, so far as I am aware,
does not occur among other forms of eurytus, in which,
if the model is dimorphic, the male is copied by the male
of the mimic and the female by the female. The white
areas of the female differ quite noticeably in this form
from the pattern of the female eurytus, there being only
a trace of white, instead of a large, well defined area
on the inner margin of the fore wing. This species
possesses in a very high degree a most conspicuous aposeme
on the underside of the base of the hind wing which,
as in the model, must be of great value when the insect
is at complete rest with the wings brought together over
the back. This aposeme takes the form of a well defined
patch of bright purplish or reddish* brown, extending
outwards from the base along the costal margin.

The form imitator, southern in distribution, is con-
nected with the northern forms shortly to be described
by forms found in Nyasaland quite plentifully, and others
more rarely in East Africa.

Very recently has been discovered a single specimen
of a most remarkably interesting form which extends
our knowledge of the distribution of Pseudacraea eurytus
south of Natal to West Pondoland. The specimen
referred to is in the museum at Tring, and I am greatly
indebted to Lord Rothschild and Dr. Karl Jordan for
the loan of it that I might compare it with others. It
was taken on April 16, 1915, by H. H. Swinny, at Port
St. John's, and is of quite a distinct reddish orange tint,

A UNIQUE SOUTHERN SPECIMEN 247

which renders it markedly different from any other form
of eurytus yet known. It seems to combine the features
of the western eurytus male with the imitator of Natal.
The subapical orange area of the fore wing is small and
narrow, as is the corresponding pale area of imitator,
and the inner marginal area on the fore wing is Hke that
of eurytus. The hind wings have a very large orange
area with only a narrow black border, as in the western
eurytus, but at the base they show the very la,rge purplish
brown aposeme extending along the costal margin to
the tip as in imitator. This most interesting specimen
is a male, and one awaits the discovery of its female
with enthusiasm : probably it will be black and white.
It presumably mimics a local form of Planema aganice
which has yet to be discovered. ■

In 1904, S. A. Neave described four new Pseudacraeas
from Uganda, which are now known to be all forms of
eurytus, using the name in its wide sense. The form
hobleyi was the first to be named, and the others make
with it a group inhabiting Uganda, although some at
least extend their range out to the West Coast. Hobleyi
is a male form only, mimicking the male of Planema
macarista.^ On a ground of blackish brown it has a bright
orange bar bent at an angle, crossing the fore wing, and
a white bar crossing the middle of the hind wing. This
latter bar, however, as in the model macarista, is often
bordered with orange-brown and, in examples brought
by Neave from Western Uganda, becomes wholly orange.
This form appropriately mimics the local form of Planema,
in which the white bar of macarista is replaced by orange.
This is Planema pseudeuryta, very closely allied to
macarista. The form hobleyi has a basal aposeme on
the under surface of the hind wings, but this differs
slightly from that of imitator, being more concentrated
into a basal triangle, and of a more reddish and less

1 Plate I, figs. 5, 6.

248 PSEUDACRAEA EURYTUS

purplish brown. Needless to say, this aposematic triangle
is a very marked feature in the model macarista.

The shape of the orange bar crossing the fore wing
of hohleyi varies considerably, and in some cases bears
a resemblance to the rather different shape of the bar
in a very different species of Pseudacraea, namely, Ps.
kuenowi, which has the same general scheme of colouring.
It seems possible that these forms of hohleyi show the
secondary resemblance of one species to another, both,
however, mimicking the same model, which is found
and has been described by Dr. Dixey and Professor
Poulton in other combinations. It would be necessary,
however, before accepting this for a fact, to ascertain
whether the proportion of hohleyi with the bar like that
of kuenowi is greater in the areas where both are found
together than in areas where kuenowi does not occur.

Another form of eurytus, described at the same time by
Neave as a different species, is a black and white female
known as tirikensis. Later this was found to be the
female of the male described as hohleyi, and the name
tirikensis fell into abeyance. Now, however, that both
are known to be not a distinct species but only forms of
eurytus, the name tirikensis again holds good as a mimetic
form name. This female form has a black ground ; on
the fore wing is a white bar bent at an angle, but the
posterior end of this beyond the elbow is often so suffused
with black as to reduce the bar to a large subapical area
only. Such a pattern very closely resembles that of the
female Planema macarista, also found in Uganda flying with
the mimic. The hind wing is black and white, and shows
well the umber triangle at the base on the under side as
in the model. (Plate I/figs. 7, 8.)

This form is very variable, and graduates into others.
Thus a form from Gaboon, copying the female Planema
excisa, has the fore wings like those of tirikensis, but the
hinder end of the white bar is tinted dull yellow, and the

FORMS FROM NYASALAND 249

hind wings are pale yellowish, approaching those of such
a form as terra, to be described below.

Another specimen in the Tring museum, from Fort
Anderson, Nyasaland, although it might be termed a
tirikensis, has a much larger area of white on the hind
wing, whereby it approaches to the East African form
rogersi, yet to be considered.

Dr. S. A. Neave has discovered on Mount Mlanje in
Nyasaland a form of Planema aganice differing markedly
from the more southern Natal form by the greater size
of the pale areas on all wings in both male and female.
It is copied closely by a corresponding form of Pseudacraea
showing, the same differences from the Natfil form,^
which is most interesting, for the shape and size of the
subapical white bar on the fore wing in the female comes
near to the form of tirikensis whose white transverse
bar is suppressed at the posterior end, while on the hind
wing the white area is as in the female form rogersi of
East Africa. At the base of the hind wing beneath
there is a large brown aposematic patch, which is exactly
intermediate between the long narrow purplish brown
patch of the southern imitator and the more triangular,
umber-brown patch of tirikensis.

This new form in Nyasaland is thus beautifully tran-
sitional between the more northern and the eastern forms
tirikensis and rogersi and the more southern imitator,
and agrees as closely with its local model form of Planema
as do other Pseudacraeas already mentioned.

At the time when hobleyi and tirikensis were looked
on as male and female of a distinct species of Pseudacraea,
a very interesting form was described by Poulton in
his paper on the Wiggins collection ^ as a " female pos-
sessing the colour and to a large extent the pattern
of the male." That is to say, it has an angular orange
bar crossing the fore wing, but the direction and shape

1 Plate II, figa. 5-8.

2 lere Congr^s International d'Entomologie, Bruxelles, Aovlt, 1910.

250 PSEUDACRAEA EURYTUS

of the bar is that of the white bar of the female form
tirikensis in its most complete development : the colour,
however, is not the rather deep orange of macarista, but
the paler yellowish orange of the different species Planema
poggei, of which both sexes are alike, and in which the
direction of the orange fore wing bar is much more oblique
than in the male Planema macarista.^ The form of
Pseudacrdea now under discussion is thus brought to be
a mimic of poggei rather than of the male macarista both
in colour and direction of the bar across the fore wing,
the obliquity being given to the bar because it is of the
shape of that in the female form tirikensis, and not of
the male hohleyi.

It is now incorrect to speak of a " female hohleyi
with male colouring " with this difference in shade
in the orange bar, and this form has been named
poggeoides by Professor Poulton.^ It is remarkably
interesting from the point of view of geographical
distribution. The models Planema macarista ^, and
PI. poggei ^ $, occur around Entebbe, but the
former is there more abundant and Pseudacraea eurytus
form poggeoides is very scarce, the females being almost
all of the form tirikensis mimicking the black and white
female macarista. But around Mount Elgon, in the
north-east of Uganda, Planema poggei completely replaces
macarista, which is not as yet known to occur east of
the Nile, and the black and white female forms tirikensis
are outnumbered by the form poggeoides with orange
bar acToss the fore wing like the model. ^ Since the male
of the mimic (hohleyi) has the bar on the fore wing orange,
the change in proportion of the two model Planemas
does not affect its appearance.

Thus, in the case of the female forms poggeoides and
tirikensis, there is an exact parallel to the case of the

1 Plate I. figs. 9, 10. - Proc. Ent. Soc, 1912, p. cxvii.

' Ibid., pp. Ixx.-lxxi.

FORM TERRA 251

male form hobleyi already discussed. In the centre of
Uganda a few forms are found which, in the case of the
male, become the predominant form to the west, in the
case of the female become the predominant form to the
east. In each case the alteration between the proportions
of the forms coincides with the dijfferent proportions of
the Planema models on the west and east sides of Uganda.

Another form that was described by Neave in 1904
from the collection made by C. A. Wiggins in the environs
of Entebbe is known as terra. (Plate I, figs. 1, 2.)

This form is monomorphic, male and female alike
showing a rich orange pattern on a brownish-black ground,
mimicking to an extraordinary degree of perfection the
abundant Planema tellus eumelis, which is also mono-
morphic, although its West African form has a black and
white form of female as well. The pattern of the form
terra is much like that of the male eurytus, save that
the subapical and inner marginal orange areas on the
fore wing are more extensive : the colour, however, is
a much lighter orange. But forms of terra with con-
tracted fore wing areas are quite common on the islands,
and one such, mentioned previously, that was taken in
1919 on the Kyagwe coast in an isofated area of forest,
had such dark brownish orange colour that it was prac-
tically indistinguishable from a typical male eurytus from
the West Coast. In terra there is no basal aposematic
umber-brown triangle on the under side of the hind wing,
and this is absent from the model Planema tellus also.
The form terra is found from Uganda to the West Coast.

I consider it to be the most perfect mimic of all the
forms of eurytus that I have seen alive : in pattern and
colouring it approximates so closely to the model tellus
that I have been deceived by it over and over again.

On the islands many variations have been found,
showing some very distinct patterns and transitions to
other forms. A common variation is reduction in breadth

252 PSEUDACRAEA EURYTUS

of the black bar separating the orange areas on the fore
wing, and when this is completely eliminated the form
described by K. Griinberg ^ as a distinct species
*' impleta " is obtained. It would not require a great
change to develop the male form of rogersi ^ from impleta.

Another form known as ohscura was also described
by Neave in 1904, and is one of the group from Uganda
coming under the subhead of hohleyi. This is also a
monomorphic form, male and female being alike, as are
the models, the eastern form paragea of Planema epaea,
a common West Coast Acraeine.^ The model is rather
rare, and the mimic ohscura is also rare. It has pale
creamy markings, easily derivable from those of terra,
but much smaller, on a dark greyish brown background.
The pattern of these creamy markings does not follow
so exactly the pattern of the model as does that of terra,
but the general effect is decidedly that of resemblance.
As regards flight, there is much more difference between
model and mimic in this case than with the other members
of the hohleyi group, for paragea the model has rather
a weak fluttering flight, and looks a feebler insect than
the robust and powerful Pseudacraea. I have never
mistaken ohscura for its model. Both show a certain
amount of reddish brown at the base of the under surface
of the hind wing, but this is of a paler tint than the rich
umber that makes the characteristic basal triangle of
hohleyi, tirikensis, poggeoides and their models, and is
not so sharply defined.

The form ohscura seems to be the least fixed of the
Uganda forms of eurytus comprising the hohleyi group ;
at any rate, on the islands it was quite difficult to obtain
two that were closely alike and did not show transition
to one of the other forms.

Lastly, there is the extremely rare and little known
form from near Mombasa, British East Africa, described

^ Sitzungsber, d. Oes. NaturJ, Freunde, Nr. 4, 1910.
2 Plate II, fig. 2. 8 Plate I, fig3. 3, 4.

FORM ROGERSI 253

as rogersi after its captor, by Trimen in 1908.^ The form
is dimorphic, the male having the wings blackish brown
and orange, but the orange is of a slightly more reddish
tint than in other forms, and approaches the hue of the
unique specimen previously mentioned from West Pondo-
land. There is an irregular area on the fore wing like
that in some of the varieties of terra in which the black
bar is broken through.

The black and white female has a pattern after the
style of the typical female of the western eurytua, except
that the white area on the hind wing is very much larger,
and sharply outlined by a narrow black border. There
is a certain amount of reddish suffusion on the under
surface of the base of the hind wing.

This form, rogersi, is represented in collections by a
single male and female at Oxford, the types. When I
was recently looking through Lord Rothschild's magni-
ficent collection of Pseudacraea eurytus at Tring, I found
there were three females bearing labels " Mombasa "
that appear to be of this form, but I do not know of any
other males.

The only Planema of abundance in the locality from
which this form is known is Planema aganice, in its
eastern and northern form montana.^ In this form the
sexes are different and much more unlike than those
of the southern aganice ^ which, be it remembered, served
as model in Natal for the form of Pseudacraea eurytus
known as imitator. In aganice montana the male, instead
of having a creamy yeUow pattern, has the same areas
of a rich orange. In the form rogersi now under dis-
cussion, the pattern is not quite like that of Planema
aganice montana, and is much nearer that of another
species of East African Planema, viz. adrasta. This
species has, in the male, a complete orange bar across
the fore wing, and the female has the very large sharply
marked white area on the hind wing. Specimens of

1 Plate II, figs. 2, 4. " Plate II, figs. 1, 3. ^ Plate II, figs. 9, 11.

254

PSEUDACRAEA EURYTUS

adrasta in the Tring museum come from Usambara, but
it appears at present too rare to be claimed as a model
for rogersi, although the latter resembles it much more
closely than it does montana. In colouring, however,
the male rogersi approaches closely enough to the male
of its model, and the black and white female is a close
mimic of the female aganice montana.

The following table shows at a glance the models and
mimics hitherto dealt with, arranged in geographical order
from the west to the east coast and then southwards, each
type of colouring and pattern having a row to itself.

Species of Planema.

Forms of Pseudacraea
eurytus.

Distribution (roughly)

1.

epaea ^

eurytus ^

W. Africa

2.

epaea ?

eurytus $

W. Africa

3.

telltis eumelis ^ $

terra ^ $

W. coast to Uganda

4.

macarista ^

hobleyi ^

W. coast to Uganda

5.

macarista $

tirikensis ?

W. coast to Uganda

6.

macarista pseud-
euryta ^

form of Jiobleyi

(?

W. border of Uganda

7.

epaea paragea ^ ?

obscura ^ $

Uganda

8.

poggei ^ ?

poggeoides $

Uganda, especiallyto
its E.

9.

aganice montana (J

rogersi ^

Mombasa, E. coast

10.

aganice m,ontana $

rogersi $

Mombasa, E. coast

11.

aganice nyasoe ^ cj

mlanjensis c?

Mt. Mlanje, Nyasaland

12.

aganice nyasoe ^ $

mlanjensis ?

Mt.Mlanje, Nyasaland

13.

aganice ^

imitator (J

Natal

14.

aganice $

imitator ?

Natal

15.

aganice ^ (?)

form of imitator ^

W. Pondoland

We have here fifteen forms of one species of Pseudacraea
mimicking eight different species or subspecies of Planema,
Specific names were also given in the past to a large
number of other forms of Pseudacraea eurytus on the
west coast of Africa differing from the above, although
easily derivable from them by transitional varieties
known to exist, and these bear equally close mimetic

^ The neimes nyasce and mlanjensis have been chosen to distinguish
these geographical races of model and mimic.

EARLY EVIDENCE OF CONSPECIFITY 255

resemblance to different species of West African PlanemU.
(See Dr. Jordan's paper, already cited.)

It will be interesting now to trace the development
of our knowledge of this most fascinating polymorphic
species.

Professor Poulton, in a letter to Nature, August 28,
1912, stated : "A little more than two years ago Dr.
Karl Jordan informed me that he had been studying
the male genital armature of the Pseudacraeas, and that
he could not find any difference between the * species *
of a large group made up of Linne's eurytus and its
numerous allies on the West Coast, of Neave's hobleyi
terra and ohscura in Uganda, of Trimen's rogersi of the
Mombasa district and his imitator of Natal."

The provisional conclusion drawn from this fact was
that these butterflies are conspecific, and this was brought
before the first International Congress of Entomology
at Brussels by Dr. Jordan in August 1910.

At the same congress Professor Poulton gave a statis-
tical study of the large collection made near Entebbe in
Uganda by Dr. C. A. Wiggins, and stated that in this
were two specimens which tended to confirm Jordan's
anatomical findings. One was " a male Ps. terra with
a pattern approaching the male of Ps. hobleyi,^' the
other *' a female Ps. hobleyi bearing the mimetic colours
of its own male " (this is the form now known as
poggeoides). (Plate I, fig. 10.)

Quotations from letters that I received from Professor
Poulton after work was commenced on the islands in 1911
will show how keenly interested he was in the problem,
and how important he considered it.

I first met with Pseudacraea eurytus on Damba Island
in 1911, towards the end of my stay there, and, not know-
ing anything about it, found this mimetic butterfly very
puzzling. Not only was it difficult at first to distinguish
the forms from their models, but they showed such varia-

256 PSEUDACRAEA EURYTUS

tion that it was difficult to assign specimens to any
particular form, for a single specimen often appeared to
be transitional between terra, hobleyi and ohscura. The
model species of Planema were extremely scarce on
Damba, and this point proved to be of great significance.

A letter from Professor Poulton, dated October 30, 1911,
referring to the arrival at Oxford of the first consignment
of these intermediate Pseudacraeas, said : " They looked
most interesting, and perhaps here your island forms
will be different and throw light on the polymorphism
of this set of forms."

November 9, 1911. — " You have the Pseudacraeas abun-
dantly evidently, and it is of the utmost importance
to breed them from known parents. They tend to be
far more intermediate than on the mainland. I suspect
that selection of them is somewhat relaxed on the island,
and they at once tend to mix. You have terra and
hobleyi 2 (I've not seen 3" yet) and another form,
obscura, and transitions of the most beautiful kind
between obscura and terra, and between obscura and
hobleyi ? . I've only seen now among your specimens
the models of hobleyi [Planema macarista 3" and ? and
PL poggei), but no models as yet of the other two (terra
and obscura). Now if these models are rare or absent
perhaps we have the cause of the varieties of obscura
and terra, viz. a consequence of freedom from rigid
selection. It is an exciting problem, and you are
evidently in a most interesting locality."

November 12, 1911. — " This is the exciting thing ;
there is no doubt that the Pseudacraeas tend to lose their
distinctness in the absence, or rarity, of their models.
It is most interesting and entirely supports Dr. Jordan's
views. I cannot tell you how pleased I am."

These specimens from Damba Island were exhibited
at a meeting of the Entomological Society of London on
December 6, 1911, by Professor Poulton, who said :

EARLY CAPTURES ON DAMBA ISLE 257

" I suggest that, in* an area where these mimetic
patterns are less strongly selected, there is a tendency,
checked elsewhere, for them to run into each other, and
also to move in the direction of the western eurytus
forms, from which there can be little doubt that the
mimetic Pseudacraeas of Uganda originally developed.
It is to be hoped that Dr. Carpenter may be able to
obtain the material by breeding, as well as by capture,
by which this hypothesis will be confirmed or refuted." ^

The following table gives at a glance the forms and
varieties of eurytus, with their models, that were taken
on Damba :

Models.

Planema poggei . .
„ macarista
„ epaea paragea .
„ tellus eumelia .

Models

^ 9

2 1

1 1

c? ? 4 •
(? ? 7

16

Mimetic forma of eurytus.

poggeoides . , . . — 1

hobleyi 6 —

tirikenais . . . . — 4

obscura 2 1

terra 6 9

Mimica 28

Intermediates . . 10

Total eurytus . . 38

It is seen that the mimics are almost twice as numerous
as the models, and there are in addition more than one-
third again of forms intermediate between the mimics.
The total number of Pseudacraeas is more than twice
that of the Planemas. Concerning these specimens Pro-
fessor Poulton wrote January 1, 1912 : " I am most inter-
ested to hear that you did not distinguish the Planemas
and Pseudacraeas (I do not wonder at it) : it makes the
results more conclusive. There must be, I conclude and
argue in this little paper, something unfavourable to
Planemas in the conditions of Damba Islands ; and in
their absence or scarcity selection is damped down and

1 Proc. Ent. Soc, 1911, p. xciv.
18

258 PSEUDACRAEA EURYTUS

the forms are beginning to run into each other and also
to assume entirely different proportions from those of the
mainland. It is most interesting and indeed exciting."

Early in 1912 I moved camp to the north-east corner
of Bugalla Island, the largest of the Sesse archipelago in
the north-west corner of the Lake. The north-west
promontory of this island almost touches the mainland
coastline at the port of Bukakata (see map).

It very soon became apparent that all the combinations
which were of so great interest at Entebbe and on Damba
Island were present here, and on March 16, 1912, Pro-
fessor Poulton wrote :

" I am very glad that Sesse is turning out to be so
interesting, and I quite think now that it will be even
more interesting to compare these two groups of islands
together rather than to get evidence from only one. You
will realize the utmost importance of catching all the
models and mimics you can secure on given days without
any selection, so that we can estimate the proportions
of models and mimics and see whether, as in Damba,
the mimics are relatively far more abundant than on the
mainland, and whether, as also in Damba, the inter-
mediates are specially prevalent.

If the results in Damba are supported by a large col-
lection on Sesse, I think the matter will really be estab-
lished, and will constitute the strongest argument I
know for the origin and maintenance of mimicry by
natural selection ; for directly the models are reduced
the mimics begin to run into each other. I think, in
your case, when you are necessarily pressed with other
work, that it would be best to concentrate attention on
the most important problems, and those certainly are
the proportions of models and mimics' and the compo-
sition of the mimetic Pseudacraea group, and breeding
the latter to prove that they are all one."

April 6, 1912. — " How splendid that the problem is

VARIOUS FORMS COURT EACH OTHER 259

more illuminated at Sesse than Damba. I never thought
that I should get such evidence of the effect of a model
on its mimic. ... It is very important that the Wiggins
Pseudacraea mimics come from several localities near
Entebbe and yet never approach your island forms in
being intermediate, or resemble their proportions to each
other or to the models."

I began to try to obtain ova from captured females
placed in a box in the forest, but although I knew the
genus of tree on which they would probably oviposit
{Chrysophyllum, Sapotaceae), it was some time before I
discovered it in the forest. It may be said here that
Miss Fountaine has published ^ beautifully coloured draw-
ings of the early stages of the form imitator, which is the
only representative in Natal.

But breeding this form has thrown no light on its
genesis, or on the other forms. For, since there is only
one Planema to act as a model, imitator appears to have
become more fixed than the forms of hobleyi in Uganda,
and varies very little.

Meanwhile, I had often seen the various forms in
Uganda jflirting with each other in the manner character-
istic of male and female of the same species. The follow-
ing letter to Professor Poulton is quoted from the Proc.
Ent. Soc, June 5, 1912, p. Ixxxv : "I have already told
you that I have seen male hobleyi flirting with female
terra and vice versa, both hovering flutteringly in the
air. Since then I have seen a male obscura paying court
to a female terra in the same way. This makes the
observations complete. ... So far I have not succeeded
in getting eggs, though I have kept four females full
of ova : three have died without result, the fourth I
have had for a week and it is still living, though it has
hardly any wings left."

However, prolonged observation in the forests at

1 Trans. Ent. Soc, 1911, pp. 57-9.

260 PSEUDACRAEA EURYTUS

length met with success, as told in a letter dated June
16th (and partially printed in the Proc. Ent. Soc, 1912,
p. cxv) :

" I feel I can almost say, as did Charles Kingsley,
' At last ! ' To-day (Sunday) being a sunny morning
after many wet mornings, I went butterflying. I saw
a few freshly emerged Pseudacraeas, and, just as I was
coming away, saw a beautiful obscura, whose very large
pale areas indicated more than a touch of the female
hohleyi. It was fluttering about from bush to bush,
and was too shy to let me get near to catch it. At last
it settled, and hung from the underside of a leaf, and I
was able to see it had a fairly distinct basal triangle
(indicating admixture with the form hohleyi). It remained
motionless a few seconds, and though that attitude is
exceptional for Pseudacraea (they always rest on the
upper side of a leaf with wings usually expanded) it never
struck me what was up. I tried to catch it, but it flew
off before I got within striking distance. It then occurred
to me to look at the leaf, and to my inexpressible joy
and excitement, there was a freshly laid egg on the middle
of the under surface, still moist with the secretion which
fastened it to the leaf. So, if this egg produces a terra
(and the chances are in favour of this, as terra is much
the commonest here), you will have the proof you so
ardently desire, seeing that the parent was a mixture
of hohleyi and ohscura ! Anyhow, now that I know and
have found the food plant, I may have better luck in
getting a captive Pseudacraea to lay. There is just time
for the egg, larva, and pupa to develop before the con-
gress at Oxford is over, so that, should the offspring be
terra or hohleyi I will let you know. As of course there will
be no time to write, I will cable just the one word, either
terra or hohleyi. If it is obscura I will not cable, but
will, of course, write. I feel that it will be such a splendid
opportunity for making this result known when you

PROOF BY BREEDING 261

will be showing the Pseudacraeas with special intent to
prove their conspecificity by the intermediate forms."

This egg, whose development is fully recorded else-
where,^ with descriptions of the larva at different stages
(which I dubbed " Toby," feeling that he was important
enough to deserve a name !) and of the pupa, eventually,
on August 16th, became a butterfly of the form terra,
somewhat transitional to obscura. Thus was obtained the
first proof by breeding that the forms of eurytus known
as tirikensis, hohleyi, terra and obscura are conspecific.
The cable bearing the single word " terra,^' probably the
first cable ever sent about a butterfly only, did not reach
Oxford until after the conclusion of the Entomological
Congress, but Professor Poulton announced the result
in the letter to Nature (September 12, 1912, p. 36), which
has been already partially quoted.

A letter that I received from him may be quoted here.

August 19, 1912. — " I must write at once now that
the great cable has come to-day. I will send a letter
to Nature next week, so that you may get the credit at
once of this most important discovery. . . .

" It is splendid to have this great disturbing question
settled at last. Your work is a splendid help towards
confirming the Darwinian theory of mimicry. ... I
must now conclude, but cannot do so without again con-
gratulating you on this great success. I have longed
for this to be done for two or three years, and have tried
my best to induce friends to do it. I even tried to persuade
Millar to journey from Durban on purpose to do it !
It is a splendid thing to have accomplished ; quite side
by side with Marshall's breeding of the seasonal forms
of Precis.^ ...

" P.S. — After your cable I imitated your example
and have just had a sleepless night ! But it's only

1 Trans. Ent. Soc, 1912, pp. 706-11.

2 IbU; 1902, pp. 417-58.

262 PSEUDACRAEA EURYTUS

natural, and to be expected after such exciting intelli-
gence."

Again, on October 26, 1912, Professor Poulton wrote :
" I am glad that you are realizing the great importance
of this material as a test for Natural Selection against
Mutation. It seems almost too good to be true. ..."
On November 22, 1912, he again wrote :
"I do not doubt that imitator in Natal is just as
successful as eurytus on the West Coast. Probably
nowhere in Africa are Pseudacraeas quite as successful
as they are in Uganda. But on the West Coast they
are very polymorphic, in Natal monomorphic, yet both
about equally successful. I take it that any variety in
Natal that might lead on to a new form is at once exter-
minated because there is only one Planema there, namely
aganice, but on the West Coast there are many models,
and any variety leading to one of them causes the varia-
tion to fall into the surviving percentage. If that model
disappeared the form would then drop out. We get all
this happening round Entebbe according as the models
change or disappear " (e.g. the change in hohleyi from white
banded to brown banded hind wing, and of the female
from tirikensis to poggeoides in different proportions in
different parts of Uganda). " Only to-day I saw the male
hohleyi from Neave's Western Uganda series with a brown
bar crossing the hind wing, and in the same lot was the
Planema, a form of macarista with a similar male." (This
form is known as pseudeuryta, and has been already
alluded to.)

" I do not believe that any of these mimics depend
for their existence on the model. The pattern depends
for its existence on the model, but not the species which
produces the pattern. That is to say, Pseudacraea
eurytus is a species which could stand by itself, and indeed
does so on certain islands where the protection afforded
by Planema models is non-existent, but the form

FAMILIES REARED 263

in which it appears is influenced by the varying pro-
portions of these models. Reasons have already been
given for regarding eurytus as a species at least partially
protected."

On November 23, 1912, Professor Poulton writes
again :

" The non- variation of imitator seems to me quite
explicable by selection, for in its locality there is only
one Planema. Rogersi may be looked upon as the con-
necting link between Uganda and Natal, and I have not
the least doubt that there are plenty of Pseudacraeas in
British East Africa and further south, but they are mis-
taken for Planema by collectors. ... I have no doubt
you are right about hobleyi being the most distinct form ;
that it should keep its sexual dimorphism so clearly is
further evidence of this, and I should suspect it does
follow the Mendelian relationship to one or both of the
others. It would be very interesting if this could be
tested : the main reason against it is the very great
abundance of intermediates. For this reason I should
very much doubt a Mendelian relationship between
terra and obscura."

Since " Toby " was reared I have been fortunate enough
to obtain eggs from females kept in captivity. They
were put with branches of the Chrysophyllum into a large
wooden box whose sides and top had been partially
replaced by mosquito net. The box was placed on tins
standing in water in a large basin which stood on a fallen
tree trunk in a small open place where gleams of sun-
shine could penetrate into the cage. I had no success
unless the butterfly was kept in the forest, since the
species seems to be easily killed by the less humid atmo-
sphere outside the forest, and the greater heat. Several
small families were reared from known mothers, and
may be seen with the parents in the Hope collection at
Oxford,

264

PSEUDACRAEA EURYTUS

Family. ? Parent.

1 obscura-tirikensis

2 obscura

3 obscura-tirikensis

5 terra

6 tirikensis

tirikensis

terra-tirikensis

Offspring.

1. $ terra-tirikensis

2. $ obscura-tirikensis

3. $ terra-tirikensis

1. ^ oftscMra-ierra

1. ? terra-tirikensis

2. 9 terra-obscura

3. ? terra-tirikensis

4. $ terra-obscura

5. (? obscura-terra

6. $ obscura-terra

7. (J obscura-terra

1. 9 tirikensis

2. (J hobleyi-terra-obscura

3. $ obscura-tirikensis

1. (J terra-hobleyi

1. ? tirikensis

2. (J hobleyi

3. ? tirikensis

4. ? tirikensis

5. ? tirikensis

6. $ tirikensis

1. (J obscura-terra

2. (J obscura-terra

3. (? hobleyi

4. $ obscura

5. (J obscura-terra

6. (? obscura-terra

7. (J hobleyi

8. $ obscura

1. (? terra-hobleyi

It seems as if the forms hobleyi and tirikensis were
predominant over the other two, for although ^erra and
obscura tainted with hobleyi or tirikensis abound (as shown
by the umbre triangle at the base of the hind wing), yet
I have never seen a specimen that could be described
as hobleyi tainted with ferra or obscura. So far, ferra
or obscura untainted with hobleyi or tirikensis have not
been bred from tirikensis parent, and one of the most
abundant intermediate forms is ^erm with more or less

MODELS, MIMICS AND INTERMEDIATES 265

m

E-i •

^S
Qg
W Q

I— I

r f^
O^
m<

CO P

g^

o
M

o

I— I

o

m
H

m

OJ

>..2

I

CO

00

l-H

05

m5

o

Qi

a .

o5

o o5

V

&^

o

lO

(N

1— t

OJ

2

(UM

M

U5

(M

Th

(M

a>

^

3t3

lO

-d a
a 03

(M

CO

O

OS

O

05

a-a

O

"-^

00

IH

OS'S

a

3 .

o ^

•*

O

OO

M

»b

rt<

T— 1

^A

■^

I>

00

tH

Ttl

C'i

c8

-v

Oi

<s>

O

t^

O

c!l

(N

>n

W

(N

CO

1-1

SW

1—1

eo

i-H

M

es

IH

•S«5

to

03

o

Oi

T-H

IS
ft

M

-H

IM

(M

(u-aao

:g|^^

•

•

o»

(M

CO

(N

■^

<N

O

-2s -s

»o

00

1>

<5 'SS <! 'So

(M

,

m

(D .

•

i

^ ■

43

a

ft

03

tn

O

<3 :

c3

c3

b

+3

o

O

<3

+3

T3 •

•

<D

3 •

A

(D

43

Qj CO

^..^

<D ©

-5

(0

in

w
o3

''d

1

in

el

4^

1

a

a
§

a1

^H IB
(3 ^

a

o

•F-«

o

tM O

•1-4
>

4^

^

o ^

in

(D

Os

& a

CO

cS

CS

S 3

a

O

o
^

c3

03

«i

^ r3

tJ

s

1

a,

SJ 05

1'

c8

(0

1
A4

1

266 PSEUDACRAEA EURYTUS

of the basal umbre indicating influence by hohleyi or its
female tirihensis. It would seem as if hohleyi and tiri-
kensis are the strongest and most stable forms, influencing
the other two.

These abundant forms of Pseudacraea eurytus that
have been captured on the islands of Lake Victoria have
a very important bearing on the explanation of mimicry
by Natural Selection. As is abundantly evident from
letters quoted, Professor Poulton believes that here is
strong evidence of the efiicacy of Natural Selection in
at least keeping the mimics up to the mark. When models
are scarce no one form of Pseudacraea has any particular
survival value over another, and all kinds of non-mimetic
transitional forms are preserved which, when the models
are superabundant, are presumably destroyed by enemies
so that they do not appear in collections.

It would be of the very greatest interest at the present
time to obtain offspring from the mimetic forms on the
mainland and ascertain whether they will produce transi-
tional offspring. Indeed, we know that they do produce
a certain proportion of these, for two examples have
been captured at Entebbe and have been alluded to
before.

The table on p. 265 shows by comparison the difference
between collections made without prejudice at localities
where Planema outnumbers Pseudacraea, and where the
latter are predominant.

The chart on p. 267 shows very graphically how the
figures vary and the manner in which the numbers of
Planemas and of intermediate Pseudacraeas vary inversely.

A most remarkable thing is the complete change in
the proportion of Planema to Pseudacraea on Kome
Island and its near neighbours to the west in the period
between 1914 and the end of 1918 and commencement
of 1919. When I was on Kome for several months in
1914, Planemas far outnumbered Pseudacraeas, which

MODELS, MIMICS AND INTERMEDIATES 267

showed very little variation, as may be seen from the
table.

But to my great astonishment, when I returned to
the same islands after the war, the situation had been
completely reversed, and the Pseudacraeas, which showed
much variation, greatly outnumbered the Planemas
(see table, p. 265). It must be supposed that the
Planemas had been found at the summit of a wave of
prosperity in 1914, but that their enemies (possibly
parasites) had increased so greatly that they in turn
reached the summit of a wave while the Planemas fell

Proportion of
P/anemas in /fie
hial captures
of Butterfliei.

90
80
70

Entebbe
1909

Dambal?

1911

Ruga Ha Is
1912-13

Kome group
1914

Kakindu.
19/5

Komegroup

/9/8-/9

Sundjyo^
Isles/9/819

9

9

1

1

\
\

\
\

P

1
1

1

O.60
5 50
4; 40

^■\

\

1

1

\ /
\

V

Proportion of
intermediate
to total number
of Pseudacraeas

50
20
10

y

^-.

-\

•■

/

/

•

into a trough. This would be analogous to what is known
to happen in England in the case of the Hop Aphis and
the Ladybirds which feed upon them, each in turn
fluctuating in number.

It is most interesting to note how the Pseudacraeas
changed in constancy, the intermediate forms having
been apparently destroyed while the models exercised
their protection to the mimetic forms in 1914, but
having an equal chance of survival in 1918-19 when
the protection of the models was in abeyance. Young
birds in 1914 had a different lesson to learn from

268 PSEUDACRAEA EURYTUS

those of 1918-19; and also those that had learnt their
lesson in 1914 must be supposed to have unlearnt it in
1918, when the models which they had known to be
distasteful were no longer to be seen. If this be so, then
it must be further supposed that birds need to refresh
their memories at intervals during adult life.

These numerous and beautifully graded transitional
forms of one -species are quite consistent with the explan-
ation of mimicry being produced and maintained by
the continuous operation of Natural Selection upon small
variations.

But at first sight they do not seem to be at all in con-
formity with the theory of the production of mimics by
Mutations — that is, sudden large variations which appear
complete all at once and never revert to the form of the
parent. Indeed, the writings of such an exponent of
the Mutation theory as Professor Punnett seem at first
sight to support this argument, that the intermediates
are a difficulty in the way of the Mutationist.

Thus, in the 5th edition of Mendelism, 1919, he dis-
cusses two forms of an interesting species of Euralia, a
Nymphaline of which the form wahlbergi mimics Arnauris
niavius dominicanus, and the form mima mimics Amauris
echeria. On p. 177 Professor Punnett says : " According
to Mendelian views, on the other hand, the dominicanus
pattern arose suddenly from the echeria pattern (or vice
versa), and similarly wahlbergi arose suddenly from mima.
. . . On the modern Darwinian view natural selection
gradually shapes mima into the wahlbergi form owing to
the presence of dominicanus ; on the Mendelian view
natural selection merely conserves the wahlbergi form
when once it has arisen. Now this case of mimicry is
one of especial interest, because we have experimental
evidence that the relation between mima and wahlbergi
is a simple Mendelian one, mima here being the dominant
an,d wahlbergi the recessive form. The two have been

INTERMEDIATES AND MUTATION THEORY 269

proved to occur in families bred from the same female
without the occurrence of any intermediates, and the
fact that the two segregate cleanly is strong evidence in
favour of the Mendelian view."

In the case of the theoretical origin of wahlbergi from
mima the change in pattern is a large one, and it may
well be asked what are the chances against the proba-
bility of the dominicanus-wahlhergi pattern arising sud-
denly, complete, in two cases, from the very different
echeria-mima pattern, and in such a way that at one
step the new pattern is the same in the two new butter-
flies and differs in the same way from the parent pattern !

Moreover, the possibility of the same large variation
suddenly occurring in these different butterflies is rendered
still more remote by the fact that while mima and wahlbergi
are forms of one species, echeria and dominicanus are
different species of one genus and by no means closely
related within that genus. The two examples given are
therefore not analogous.

Again, in Bedrock, vol. ii. Professor Punnett discusses
the polymorphic forms of the mimetic oriental " swallow-
tail " butterfly Papilio polytes. He says : " Mr. J. C. F.
Fryer has recently succeeded in carrying out an elaborate
series of breeding experiments with this species and has
shown that any form of female can produce any other
form provided that she mates with an appropriate male,
while in certain cases all three forms may appear in the
same brood. Even in such a case all the three forms
are sharply cut and clear, there being a complete absence
of intermediates or transitional forms."

Incidentally it may be remarked in passing that had
Mr. Fryer worked with the African Papilio dardanus
there would not have been a complete absence of tran-
sitional forms.

At first sight the Mendelian argument seems to be that
absence of intermediates is proof of origin by mutation.

270 PSEUDACRAEA EURYTUS

It would seem therefore that abundance of intermediates
as in the case of the Pseudacraeas is in favour of the
Darwinian explanation. But the Mendelian claims their
presence also to be explicable by the theory of mutations.
Thus in the same article in Bedrock Professor Punnett
discusses the various offspring of a cross between a
" Silky hen " and a " Brown Leghorn " cock which
show grades of transition, and says : *' Even if a completely
grading series " can be put together, it does not follow
" that the various forms have arisen through the accumu-
lation of minute variations." The series intergrading
between the two extremes of the cock and hen " can be
expressed in terms of two Mendelian factors. . . . Thus
an appearance of continuity in variation may be brought
about by the interaction of a small number of definite
factors upon one another." And in Mendelism, p. 162,
Professor Punnett says: "Neither the existence of such
a continuous series of intermediates, nor the fact that
some of them may breed true to the intermediate con-
dition, is incompatible with the Mendelian principle of
segregation."

In his book on Mimicry in Butterflies, he also says, on
p. 129 : " As the result of modern experimental breeding
work it is recognized that an intermediate form between
two definite varieties may be so because it is heterozygous
for a factor for which one variety is homozygous and
which is lacking in the other — because it has received
from only one parent what the two typical varieties
receive from both parents or from neither. Its germ
cells, however, are such as are produced by the two
typical forms, and the intermediate cannot be regarded
as a stage in the evolution of one variety from the
other. ... It is quite possible that the new mimetic
pattern appeared suddenly as a sport, and that the inter-
mediates arose when the new form bred with that which
was already in existence."

INTERMEDIATES AND MUTATION THEORY 271

The extraordinarily varied forms of P. dardanus, sub-
species polytrophus, on the high Kikuyu escarpment
(6,500-9,000 feet) in British East Africa do not support
the view that the mimetic patterns arose suddenly and
the intermediate forms from them by interbreeding.
Many of these transitional forms are abundant on the
escarpment, but are very rarely found elsewhere ; that
is, they have a certain peculiar geographical distribution.
" The Kikuyu escarpment is the only locality at present
known where these transitional forms make up a large
proportion of the females." i

Such a distribution is far more consistent with the
interpretation that these are an ancestral set of forms,
such as may have existed before the different named
forms became as distinct as they are to-day, which have
been left isolated on a mountain ridge.

One of the most varied of all these ancestral forms is
trimeni, which is not considered to be an intermediate
between two other forms of mimetic females, but to show
how the female first began to depart from the ancestral
type like the male. That is to say, trimeni is at one end
of the long series of female forms, and cannot be supposed
to be a stage between two others.

Thus it is seen that these intermediate forms are held
by both Darwinian and Mendelian to support the respec-
tive hypotheses of evolution, but the latter has perforce
to fine down his " large " variations or " sports " to such
a small size that they appear to be of the kind which
the Darwinian recognizes as " small " variations upon
which Natural Selection can, and apparently does, work.

The Mutationist claims that if these differences are
inherited according to Mendelian laws they must be
of the nature of " large " variations ; the Darwinian
claims that they are the " small " variations which he

* Poulton, " Mimicry, Mutation and Mendelism," Bedrock, vol. ii,
p. 49. A plate in this article shows some of these Kikuyu forma.

272 PSEUDACRAEA EURYTUS

believes to be acted upon by Natural Selection, and
adduces proof that they can be inherited !

One can quite appreciate the argument that the theo-
retical origin of Eur alia wahlbergi from a form of such
different appearance as mima would be an example of
origin by " Mutation " or ** Sport," but when it is neces-
sary to fine down the " largeness " as in the intermediate
fowls it becomes an apparent absurdity to allude to these
differences as " large," that is, as " sports." It seems
difficult to draw distinctions between the Mendelian's
" Mutation," which is inherited by segregation of some
very minutely differentiating " factor," and the Darwinian
" small " variation, which can be inherited, but which
the Mutationist calls a *' fluctuation " and says is non-
heritable, but if proved to be heritable claims as a
" mutation " or " large variation."

If segregation comes down to produce such minute
differences, the term " large variation, mutation or
sport " seems to be worn rather thin !

On this matter the work done by T. H. Morgan upon the
colour of the eye in the fruit fly Drosophila has a bearing.
The eye is normally red, but in the investigations upon
the inheritance of the eye colour it was found that there
were seven gradations between white and red, and further,
one of the grades has seven modifying factors, each of
which altera its intensity and gives rise to a secondary
grade. Professor Jennings remarks that " by means
of these graded changes one could obtain, by the
mutationist 's own statement, the continuously graded
results which selection actually gives. What more can
the selectionist ask ? " ^

A case strictly analogous to the case of Ps. eurytus,
with its abundance of insular variations in the absence
of the models, is noted by Professor Poulton in his presi-

^ See a summary of Jennings' papers by Professor Poulton in the
Proceedings of the Entomological Society of London, 1917, pp. Ixxxv-lxxxix.

INTERMEDIATES IN ABSENCE OF MODELS 273

dential address to the Linnean Society in 1916. Dr.
Lamborn bred Papilio dardanus on the west coast of
Africa (South Nigeria), where the black and white form
hippocoon is almost the only form known. This mimics
the abundant Amauris niavius. Another form of dardanus
female has been caught rarely all along the tropical west
coast of Africa. This has a fore wing pattern approaching
that of hippocoon, but with the bar between the subapical
and inner marginal white areas very ill defined, and the
hind wings are yellow instead of white. This rare form
is known as dionysus, and there is no model for it in
Nigeria.

Lamborn bred from a hippocoon parent nine female
forms like it, but also eight dionysus which has never
been bred before. Now these dionysus are in a position
similar to the Pseudacraeas which, in the absence of control
by the models on the islands, are producing large numbers
of transitional varieties, for no two of them are alike,
and they present a beautifully graded series, from a fore
wing pattern approaching that of hippocoon to a pattern
closely approaching that of the male. The nine hippocoon
sisters, however, mimicking the abundant Amauris, present
a very close resemblance to each other and to the hippocoon
parent.

It is noteworthy that in the case of both Papilio dardanus
and Pseudacraea eurytus the variation in the absence
of models is best shown in the shape, size, and sharpness
of outline of a dark bar separating the paler portions
of the fore wing into two main areas.

We now come to discuss another point of great interest.
If the collections of Pseudacraea eurytus from different
islands in the lake, sometimes very close together, be
compared, further evidence in favour of the maintenance
of mimetic resemblance by Natural Selection is obtained.

Although the islets of Kimmi and Ngamba are so
very close to each other and Kome (see map), yet the

19

274

PSEUDACRAEA EURYTUS

predominant form of Pseudacraea is not the same on
each island.

At each locality the predominant Pseudacraea is that
which mimics the species of Planema that is most
numerous in that locality, and this is further borne out
by the collection from Kakindu on the mainland to the
west of the lake.

Combina-
tion.

Models and Mimics.

Kome,
1914.

Ngamba.
1914.

Kimmi,
1914.

Kakindu,
1915.

I

PI. macarista ^ poggei c? ? ■ •
Ps. eurytus hohleyi <J

66

8

23

9

15

2

10

1

II

PI. macarista ?

aganice montana $
alcinoe camerunica ?

Ps. eurytus tirikensis $

8

23

7

9

2

5

III

PI. epaea paragea (J $
Ps. eurytus obscura (^ ?

2*

2

—

16

IV

PI. tellus eumelis (? $
Ps. eurytus terra cj ?

236
36

—

—

71
12

From the above table it is seen that while on Kome
the model of type of colouring IV {Planema tellus) was
most abundant, on Ngamba the species that predom-
inated were of types I and II, and on each island the
proportions of the forms of Pseudacraea are according
to the prevailing type of models. At Kakindu, again,
type IV predominated in the models, and all the Pseud-
acraeas save one were of type IV. The islands mentioned
are so close together that it would be absurd to attempt
to explain the differences in the Pseudacraea fauna by

VARIED FORMS ON DIFFERENT ISLES 275

climatic conditions ; the factor that varies, however, is
the proportion of species of Planema. The supposition
is that on Ngamba a form of Pseudacraea that does not
resemble male or female of Planema macarista has
less chance of surviving birds' attacks than a form which
is deceptively like macarista ; consequently the form terra,
which is so common on Kome Isle only a few hundred
yards away, is at a disadvantage on Ngamba.

This explanation, of course, presupposes that the
species of birds which are presumed to exercise selection
do not fly across from Ngamba to Kome or vice versa
(see map). Now, although Bee-eaters may constantly
be seen crossing open water in flocks, I do not
think that I have ever seen such birds as Flycatchers,
which abound on the islands and, in some cases at least,
only in the forest which is the haunt of Pseudacraeas,
crossing from one island to another. I have even noticed
that the very characteristic songs and call notes of species
such as the " Kunguvu " {Tchitrea emini) and a pretty
little black and white Platysteira (? jacksoni) varied on
different islands.

While on Bugalla in 1912-13, I became extremely
familiar with the call of these abundant birds, and
directly I visited the Kome group of isles in 1914 at
once noted a difference — yet the difference was slight
enough to make me feel certain that the species of
bird was the same. This seems to show that at any
rate on islands at this distance apart tl^ere is a tendency
for geographical races to be perpetuated, but I have
not been long enough among the islets of the Kome group
to know if the song of the flycatchers of Ngamba differs
from that of the same species on Kome. Yet, as has
been shown in Chapter VI, many islands quite close
together show distinct differences in their avian fauna.

CHAPTER XII
HYMENOPTERA

Ants.

The ant that most obtrusively calls for notice is the
well known Dorylus or " Safari ant," known to the
Baganda as " Ensanafu " and in Kiswahili called
*' Siafu." 1

This totally blind reddish ant varies exceedingly in
size, the largest individuals being half ah inch long and
several times larger than the smallest, but it is difficult
to make out any difference in their functions.

They live entirely by hunting live prey, and owing
to their incredible numbers must destroy very many
of the creatures that are unable to escape them. They
make temporary subterranean nests, which are easily
discovered on account of the piles of loose earth that
has been brought up from below grain by grain, and
the numerous large irregular holes leading down to the
burrows and chambers in which the queen lives and the
young are reared. The queen is a curious looking creature,
with abdomen much enlarged, as is the case with the
well known queen termite, but in the latter the
distension is very much greater.

Queen Dorylus are rare in museums, probably because
of the courage that is required to dig one out of the nest
among the hordes of ferociously biting workers ; a native
once brought me two, but I should have liked to see

^ Dorylus nigricans and glahratus.
276

■ ENSANAFU ANTS 277

him getting them, from a safe distance ! The adult
male has been elsewhere alluded to in connection with
signs of the weather. The worker Dorylus are often
seen travelling swiftly in a column several ants wide,
in a very orderly manner : hence the name Safari ant.
They prefer a sheltered route among and under leaves
or dense grass, but where they must cross an open smooth
area such as a well trodden path, they sometimes make
a tunnel for at least part of the way beneath the surface,
or a deep furrow. At other times, however, they protect
the travellers by forming a line on each side of the track,
the ants standing high on their legs with mandibles
wide agape as if seeking for something to bite. These
guardians are so intertwined that they make a living
wall which often bridges right across the ceaseless stream
hurrying below. If interfered with, or disturbed by
the tramp of a foot shaking the ground, they break loose
and run about with menacing jaws, but do not go far,
and if nothing is met within a yard of the column
the guardians return and re-form the protective wall.
These columns of hurrying ants fill one with amazement
on account of the immense numbers of individuals.

The first one I saw was crossing a road close to my
hut on the evening of July 18, 1910, and next morning
the stream was still showing no signs of abating. It
was about an inch wide, and 4-6 ants across. At noon
the ants were still pouring across the road in the same
direction, but by 5 p.m. those forming the protecting
walls had moved, and by sunset they had all gone across :
that is, for about twenty-four hours there had been a
continual stream, for there is no reason to suppose that
this column was exceptional, and had ceased to move
when night fell.

This column was apparently changing camp, for nearly
every ant carried a naked white pupa, although it is a
curious fact that not a single larva was seen in the pro-

278 HYMENOPTERA

cession. Indeed, it may here be said that I have never
seen Dorylus carrying its larvae on safari, only the pupae ;
nor does one ever see a queen traveUing. Is it possible
that a column such as is above described is analogous
to a swarm of bees, but that instead of taking a queen
with them the members of the swarm carry pupae, from
one of which a queen will emerge when they have ex-
cavated a new nest ? It is curious that where there are
large and small ants there should not be some differen-
tiation of function. Quite small individuals are seen
struggling along carrying enormous pupae, while lusty
fellows of the biggest size appear quite content whether
they are carrying a small individual or one of their own bulk.

Although, on the whole, the column is an orderly one,
occasionally considerable confusion is caused by indi-
viduals running back against the stream ; possibly they
are " returning empty " ! If they are of very small size,
they are often seen to be unable to make headway against
the hurrying multitude, and unless they run quite at
the edge are at length compelled to turn round and run
with the rest. A very large individual, however, since
it stands well above the small ones, can often run against
the tide, for the small ones can run under it, between
its legs.

But every column of Ensanafu that one sees is not
engaged in changing camp ; most often they are either
going out to hunt, or returning laden with spoils.

When they have reached the hunting ground the main
column breaks up into minor streams, and the ants swarm
over the ground, investigating every nook and cranny
and climbing up bushes.

If one is in the forest when they are hunting, one often
first becomes aware of it by hearing a peculiar pattering
noise, caused by the countless ants which have run up
branches to the extreme tip, and finding no prey there
drop off and fall on to the leaves below.

ENSANAFU ANTS 279

Their activities are also revealed by the great dis-
turbance they cause among the countless spiders and
insects of all kinds that have previously been concealed
among the dead leaves and ground cover. Before one
sees the hunters themselves one ineets spiders, harvest-
spiders (Phalangeridae), caterpillars and especially cock-
roaches scurrying away in great excitement. As one
draws nearer to the area being hunted over, one sees
little masses of Dorylus ants busily engaged in cutting
up some unfortunate insect that has failed to escape,
and then, standing just outside the area and keeping
careful watch around one's feet, one can see the hunters
running about and going into every nook and
cranny.

A cockroach takes fright and rushes out from under
some dead leaves, only to be seized by one or more ants
and overpowered by superior numbers. Sometimes, on
a sandy shore, where they have a chance of running away,
the cockroaches forfeit their lives by losing all presence
of mind and running wildly hither and thither, often
tumbling head over heels in their frantic endeavours to
escape, and so falling all the easier prey.

Sometimes, however, certain spiders by moving
cautiously first in one direction and then in another,
and, as it were, feeling their way, manage to evade the
awful fate of being eaten alive.

It is an unpleasant sight when a huge caterpillar is
seen writhing and twisting in the endeavour to free itself
from the numerous ants which have firmly fixed their
sharp jaws in its skin and will not leave go.

Slugs of large size are cut up piecemeal, although some
of the ants at least are overwhelmed and suffocated in
the slime that is secreted more abundantly when the
slug is attacked.

A certain snail that was common on Damba Isle
was able, as related in another chapter, to keep the ants

280 HYMENOPTERA

at a distance by surrounding itself with bubbles of foam
into which the ants could not penetrate.

A medium sized Plant bug {Hemiptera) was once seen
on a leaf of a bush over which Dorylus was swarming,
and I was much interested to note that although ants
would frequently seize hold of a limb or antenna of the
bug they always let go again, and no harm was done to
the bug. It is possible that its powerful odour may
have been disagreeable to the ants.

One night a large army of Ensanafu raided my grass
hut on Bugalla Isle ; fortunately I was sleeping in a
tent close by, and though a column ran through the
tent they were only on the way to the hut, and I was
not turned out. But a nestful of fledgling swallows in
one corner of the hut was cut up and carried away piece-
meal.

One day at Jinja I heard pitiful squeakings coming
from some long grass, and found a baby striped rat being
attacked and in danger of a horrible fate ; needless to
say he was freed, but very likely fell a victim subsequently.

During the campaign in German East Africa I was
twice turned out of the little tent in which I was sleeping
on the ground by an invasion of Ensanafu, and had to
bolt, and then, after picking off those ants which had
already attached themselves to various parts of my
person, make a dash for boots, and by frequent painful
visits gradually withdraw the bedding and shake down
elsewhere outside, while the ants proceeded along the
line to some one else's tent, to my secret gratification,
for no one likes to be the only one who is turned out in
the night !

On Damba Isle in July 1911, I saw a very interesting
thing while watching the unfortunate denizens of the
shady forest being turned out by Dorylus. Cockroaches
were scurrying about in all directions, and hovering
over them, and occasionally darting down upon them,

lENSANAFU AND STINK ANTS 281

was a small, thin bodied, long legged insect whose appear-
ance suggested a Syrphid fly, though one seemed to have
a short ovipositor, which would imply that it was an
" Ichneumon " fly.

After watching for some time I suspected that the
fly's object was to dart down and lay an egg upon a
cockroach before it was cut up by the ants. Presumably
the egg would thus be carried into the nest, where the
larva would find food that it required, and possibly frag-
ments of insects that had been brought in by the ants.

Bates, in his Naturalist on the Amazons, describes
something similar in the case of Stylogaster and the Eciton
ant, which has habits analogous to those of Dorylus.

In July 1914, I saw an interesting affair on Bulago
Island. A small column of Ensanafu was hunting, and
had discovered in a hollow,, broken cane stem the nest of
another ant, a large black species whose rotund abdomen
is covered with golden pubescence. The rightful owners
of the nest had found discretion the better part of valour,
and were no longer attempting to resist the raiders, who
had bitten through the stopper closing one end of the
stem. Some, inside, brought up the grubs in the nest,
which were at once seized by eager helpers outside, who
usually fell to the ground with their burden in their
excitement. It was- a busy scene, and curiously reminded
one of the unloading of a ship.

Another abundant ant is the giant Paltothyreus tarsatus,
known to natives as " Waka," and to Europeans as
" Stink ant," owing to the appalling smell of bad eggs
emitted by it when roughly handled. About an inch
long, coal black in colour, this species is often seen singly
slowly wandering about searching for food. It seems to
live on animal food — small insects or bits of dead ones
which it meets with on its wanderings — but is not active
in its movements. If handled it can inflict a severe sting.

The Waka nests underground, and in disused chambers

282 HYMENOPTERA

in the base of old termite hills : one such nest that I dug
out was perceptible to the nose from some yards away.
When the nest was broken open the ants and termites
running about of course met, and the termite invariably
seized hold of the ant. But the ant seemed to be too
hard for the termite to make any impression on it, and
if unable to free itself by struggling turned its abdomen
forwards and deliberately stung the termite, which at
once let go and seemed more or less paralysed. One ant
was tackled successiyely by two termites, and accounted
for each in turn, but it was never the aggressor, and
seemed reluctant to sting until other methods failed to
release it from the termite's grip. I have found a winged
male of this ant in the stomach of a Bee-eater.

Another large black ant that is often confused with
Paltothyreus is Megaponera fcetens, though it is con-
siderably smaller, has no odour, and is of very different
habits. The name fcetens should apply to Paltothyreus,
with which this ant, called " Enkolokoto " by the natives,
must have been confused when the name was given
to it.

I mention this ant because it is absent from the islands,
nor have I met with it on the Uganda shores of the
lake.

I first saw it when on active service on the southern
frontier of Uganda in 1914 (Kagera river), and subse-
quently became very familiar with it in German East
Africa. But the Kagera river, which forms the natural
boundary to Uganda, also seems to act as a barrier
to Megaponera in that part of the world, for though the
ant is often seen on the south side of the river, it was
apparently absent from the north side. I was informed
by a missionary, however, that he knew it well further
north in Uganda.

Megaponera feeds on termites, and marches out in
orderly bands of a hundred or more, and when inter-

HtJl'SK OX BUGALLA, I912-I3.
Barbed wire lightning conductor over supports on roof.

INTERIOR OF BUGALLA HOUSE.

To face p. 282.

SUNDRY ANTS 2^8

fered with produces a distinct stridulating or squeaking
noise. These characteristics make it easy to recognize.
Probably it is absent from the lake area because its
favourite prey is not found there. This is a species of
dark coloured termite which does not build hills, but
lives in burrows underground, from the open mouths
of which they come out in daylight to cut the segments
of grass blades, which they then drag down and pre-
sumably eat.

The remarkable ant (Ecophylla smaragdina has already
been alluded to on account of its being mimicked by a
spider and a bug. It is not found by any means on all
islands, but was common on the trees overhanging the
" fly beach " on Damba. The nest is a globular shell
formed by leaves attached together along their edges
with silk spun, as H. N. Ridley first pointed out, by a
larva held between the jaws of an ant for that especial
purpose, and applied first to the edge of one leaf and
then to another. It is an inoffensive species, its feeble
mandibles being barely able to pierce one's skin, and
being without a sting.

On Damba they fed largely on the thousands of small
" E'sami " lake flies (Chironomidae) which settle on the
leaves, but also from secretions of Aphidae on the stems
of the leaves. The island specimens were of a shining
light brown colour and form a distinct race, known as
longinoda. Some that I saw in Portuguese East Africa
in 1918 were more of a greyish brown tint.

Another ant that lives on leaves of trees is known
to the Baganda as " Obusaji-saji," but unfortunately no
specimens were obtained for identification.

It was very abundant on Bugalla and Damba, but does
not occur on many islands. It is a small, very active
long legged black species, which makes a frail nest of
some friable brown material on the backs of leaves or
between two leaves, which thus adhere together.

^84 HYMENOPTEiRA

If one is so unfortunate as to disturb these when pushing
between leafy branches, great numbers of the little ants
rush out, and, getting a firm grip with their man-
dibles, inflict a sting which leaves a burning irritation
for several hours. It is particularly unpleasant when
they find their way down one's back or into one's hair !

The most noticeable wasps are species of the genus
Belonogaster : large, dark, grey-brown insects with long
pedicle to the abdomen, known as " E'numba " by the
natives. Their nest is of papery substance, and formed
of a single tier of cells supported from a slender stalk,
with the open mouths downwards. It is attached to
the under side of an overhanging rock, or to the eaves
of a house, and in a favourite site numbers of nests are
seen close together.

Natives are very much afraid of the sting of these
wasps, which is certainly extremely painful, and do not
like going near their nesting places. It must be acknow-
ledged that there is reason for this, for sometimes a
wasp will fly off the nest and viciously sting any one
who it thinks has come too close, returning immedi-
ately to the nest. I noticed, however, that if one allows
a nest to grow up in a position constantly approached
or passed by, the wasps become accustomed to people,
and are inoffensive as a general rule.

I allowed several nests to be built from the underside
of the ridge of my tent, and the colonies grew to a large
size without any sign of aggressiveness on the part of
the wasps. But one day an officious individual, perhaps
a freshly hatched and inexperienced one who had not
been taught properly, flew off at me when I came into
the tent and inflicted a vicious sting. This sealed the
death-warrant, and all the nests were destroyed.

The grubs in their cells hang mouth downwards, the
anterior end of the body protruding slightly. They are
watched over by certain individuals who act as nurses,

BELONOGASTER WASPS 285

while others go out hunting for the smooth caterpillars
wherewith to feed them.

A wasp having found a caterpillar pounces on it and
stings it, and at once begins to pulp it with its large
mandibles, beginning at one end and passing systemati-
cally along the whole body. Finally, the mangled remains
are rolled up into a pellet about the size of a pea, with
which the successful hunter flies back to the nest. In
one case that was observed two wasps were on the nest
when the hunter (A) returned with a mashed-up insect
in her mandibles. This she gave to B, who appeared to
mince it up further and then shared it with C, each
going to one of the larger larvae in the cells and holding
the mass to its mouth. The larvae could be seen eating
it. A afterwards went round and fed the larvae with
fluid regurgitated : this process was accompanied by a
kind of violent shuddering motion. I could see the
drop of fluid appearing between the wasp's mandibles^
to be slowly sucked up by the larva.

Caterpillars are not the only prey of Belonogaster,
for one frequently sees them capturing and pulping
small brown Acraea butterflies, which is of interest
when considering the enemies of protected insects. The
Belonogaster itself may be considered a typical pro-
tected insect, but I have found it in the stomach of a
Bee-eater {Merops superciliosus). This subject is discussed
elsewhere.

Belonogaster has all the characteristics of a protected
species, and serves as a model for a most interesting
Neuropterous insect, Mantispa, which has been elsewhere
described.

Although it is not considered too formidable by the
Bee-eater, which seems to be specially adapted to feed
on stinging insects, there are probably few other birds,
and certainly no mammals, that would dare to catch
it, so that its enemies must be mainly other insects.

286 HYMENOPTERA

There are often to be seen in empty cells of the nest the
pupa cases of flies, one to a cell, probably Tachinidae,
but I have not actually seen a fly depositing its egg on
the wasp larva. These wasps belong to the Diploptera,
or true wasps, whose wings are doubled up longi-
tudinally when not in use, giving them a very narrow
appearance.

In the group of wasp-like insects now to be discussed,
known as the Fossores, the wings are not folded. While
some of them, such as Bembex, have a certain resemblance
to a true wasp, most of them are readily distinguished.

The group is remarkable from the habits whence the
name Fossores, which means *' diggers," is derived ;
they are also called " Sand wasps." The wasp Belonogaster
has been described above as feeding its young on insects
which it has first killed and pulped. It is obvious that
meat of such kind must be troublesome to prepare, and
must be supplied fresh at frequent intervals.

Some other wasps, which are of solitary habits {OdyneruSy
etc.), do not pulp the prey, but store them up whole in
their burrow. But if one examines the store that has
been laid up for the larva to feed upon, one finds that
in the case of small geometrid caterpillars, some of them
shrivel and dry up, and are not in very good condition to
form food.

One family of fossors, the Bembecidae, has slightly
improved upon this method. They feed their larvae
almost entirely upon whole flies, which they do not
kill, but sting so that they remain alive and juicy but
cannot move, since the central nervous system has been
paralysed ; probably, therefore, they are insensitive as
well. Regular supplies of these paralysed flies are taken
down into the burrow by the mother Bembex.

A further step in the saving of trouble is exemplified
by other fossors of the families Sphegidae and Pompilidae.
These lay up in a burrow a store of paralysed insects

FOSSORS 287

sufficient to feed one grub for the whole period of its
growth. Having stocked the burrow, the mother lays
an egg, affixing it to one of the living but helpless prey,
and having sealed up the burrow for good and all, she
goes off to make another. Sometimes this process in-
volves very heavy labour, for the inert grasshopper,
caterpillar or spider has to be taken to the burrow.
Usually it is too heavy to be carried in flight ; some-
times it can be carried walking, but often it has to be
dragged along the ground. In the latter case the fossor
first finds and stings the prey, and then excavates the
burrow in a suitable spot in the neighbourhood ; but if
she can carry her prey the burrow is excavated first and
the prey carried back to it, gripped by the mandibles
and carried between the legs. It is wonderful with what
accuracy the mother will find her way over sticks, stones,
bare patches of sand, or through thick grass, until the
burrow is reached ; sometimes she travels bacJctvards
with her load !

It is an obvious saving of this labour to find the prey
first, and then excavate the burrow where it lies.

Another family, however, the Scoliidae, has economized
labour even further. The larvae of these fossors feed
upon the subterranean larvae of large Lamellicorn beetles.
The mother burrows down to them and merely lays her
egg upon them in situ, presumably stinging them first
to render them inert, as is the custom in the Fossores.

Lastly, there are the peculiar Mutillidae, whose wingless
females, as they run swiftly over the ground, look like
brightly coloured ants. These are mainly parasitic upon
other Hymenoptera, and do not burrow, but pierce through
the walls of the nest and lay their egg upon the larva
inside, just as the typically parasitic " Ichneumon " fly
lays its egg upon a caterpillar. Thus one can trace all
stages between the somewhat crude method of the
Belonogaster wasp down to the labour saving parasitism

288 HYMENOPTERA

of the Mutilla and the well known " Ichneumon " fly,
parasites of caterpillars.

It may be remarked here that *' parasite " hardly
seems the correct term for an Ichneumon larva, whose
activities necessarily result in the death of its host, which
is devoured as completely as an antelope is by a lion.

The only difference is that at first the young Ichneumon
larva feeds only upon the juices of its prey, and the
essential tissues are not devoured until the larva is full
grown, when it has nothing to fear in the death of its
unwilling host.

It does occasionally happen, however, if the dispro-
portion in size is great between the host and the parasite,
or if the number of parasites is very small, that the host
is not killed. I once reared an Arctiid moth and a single
parasitic Tachinid fly from the same caterpillar !

The habits of Bembecidae are, like those of other fossors,
of absorbing interest. As has been pointed out, these
feed their larvae on flies, ^ but they do not lay up one store
for good and all as do the Sphegidae and Pomjnlidae.
Perhaps the reason is that the flies, not being large-bodied,
fleshy insects, might dry up before the Bembex larva
had eaten them all. So the mother attends to her larva,
bringing to it freshly stung flies at frequent intervals.

Since this necessitates constant work opening and
closing the burrow, it is dug in localities where the
soil is very loose and light, such as a sandy shore on the
lake, and I have found thriving colonies of Bembex in
such situations.

The first that I met with was on a dazzlingly white
beach of fine sand on the south coast of Nsadzi Island,
in March 1911, when I found numbers of B. forcipata
flying about over the hot sand, and busily excavating
their burrows, and spent many hours watching them,

^ Occasionally on other insects. I have watched one catch a " Skipper "
butterfly. See Proc. Ent. Soc, 1917, p. xli.

BEMBEX 289

for I soon saw that they were catching Tse-tse fly.
A Bembex would alight on the sand at a spot which
she evidently knew, although it was indistinguishable
from the surrounding bare area, and commence to dig
away the loose sand with her powerful toothed fore limbs,
specially adapted for this work by a fringe of stiff hairs.
So rapidly does she work that the sand is thrown behind
her in a continuous stream, passing underneath her body
and falling a couple of inches behind.

Without a pause she persists, and soon lays bare the
mouth of a pre-existing tunnel, at the bottom of which
lies her grub, hungrily devouring flies. As she works
in the loose sand it constantly trickles down from above,
but she removes it at such speed that it is thrown out
faster than it falls, and presently the burrow is clear and
she disappears from sight into it.

Sometimes she throws up behind her enough sand to
block the entrance ; if she does this she usually remains
below for some time, but what she does down there I
know not.

Presently up she comes, and, if she is careful, turns
away from the mouth of the burrow and throws enough
sand back into it to block the entrance. The reason for
this will be given later.

But quite often the careless worker flies away without
closing the burrow, leaving the helpless grub at the mercy
of any marauder that discovers the opening.

Now the busy mother has to seek food for her young,
and she flies round and round me as I sit on the sand,
apparently looking for a nice fat Tse-tse fly full of blood !
Naturally she finds none, and has to take the next best,
but searches very carefully, coming within an inch of
my face and under the brim of my hat. Sometimes a
Tse-tse that has been sitting on my clothes takes alarm
and darts away, but not quickly enough to escape its
enemy, who pounce^ on it in a flash.

20

290 HYMENOPTERA

There is a shrill buzz, like a scream, from the fly as
it is caught and stung, and then the Bembex is seen to
be holding the fly closely to her abdomen between the
thighs of her hind legs, and darts away to her burrow.

Sometimes the Bembex pounces on a fly as it sits on
my clothes, and I am able actually to see her sting it.

Arriving at her burrow, Bembex alights at the exact
spot, and without a moment's hesitation, still holding
her prey between her legs, opens up the burrow and
passes in.

Almost immediately she emerges again and repeats
the chase. On March 13th I thus watched one worker
for three and a quarter hours, during which she caught
twenty-nine Tse-tse and two other flies. The last four
Tse-tse had been caught and carried into the burrow in
five minutes. After this she finally closed the burrow
and flew away, and I, with great difficulty in the loose
sand, found the chamber at the end of the burrow and
took out all the flies that had been collected, and the
legless white grub, full grown.

There were thirty-one Tse-tse flies and one other fly.
About half a dozen of them were dried and some partially
eaten. Twenty-one were males and ten females, eight
of which were fat ones containing a larva ; this is a very
high percentage of pregnant flies, for out of ninety-six
female flies that had been caught by my boys that
morning only 22 per cent, contained large larvae.
So that the Bembex had definitely selected the fattest
flies she could find in default of full fed individuals, which
are heavier on the wing and easier to catch.

Another specimen was seen to catch six Tse-tse and
several other flies in an hour ; if, after close examination,
she found no suitable Tse-tse on me, she would fly away
and return with other flies, though once she came with
a Tse-tse caught elsewhere.

The burrow of another that was dug up, about eight

BEMBEX 291

inches deep, contained a nearly full grown larva and
twenty freshly caught flies, one of which was as large
as a ''blue-bottle," but the rest were all about the size
of the common house fly, and appeared to be of several
species — three of one species, nine of another, one of
another, two of another, one of another, and three of
another.

It is a curious fact that the mother does not know
when to cease provisioning the burrow, and gives her-
self much unnecessary labour. Thus, in the first case
described above, the larva was full grown and com-
menced spinning its pear shaped cocoon at once. On
another occasion I found that the Bemhex had taken down
flies to a larva that had already begun to spin its oval
cocoon ! This is formed of silk, with grains of sand
attached on the outer side, and measures about an inch
in length.

The mother Bemhex appears to pass the night in the
burrow underground, for quite often, if one visits the
colony early in the morning, a Bemhex will suddenly
emerge from the ground at a spot where the soil has not
yet been disturbed that morning ; and also late in the
evening they may be seen going down into the burrows
and closing them behind them. The male spends an
idle hfe basking on the hot sand or darting about in chase
of the females, or sipping honey from neighbouring flower
heads.

The Bemhex is a most deHghtful insect to watch, as
she works at such high pressure, and so elBficiently and
quietly.

She has her enemies, for wherever a colony is found
there can be seen numbers of httle brown insignificant
flies, by name Idia, who await a chance to go down the
burrow and deposit their own eggs on the flies laid up
as store by the Bemhex for her own offspring.

As was pointed out, the Bemhex may close her burrow

292 HYMENOPTERA

when she leaves it. The moment she has gone, up hurries
one of the Idia flies and attempts to get through the
barrier. Sitting almost upright on the end of the abdo-
men, the little fly throws the sand forwards with quick
movements of its legs, as if trying to burrow backwards
through the barrier. But I have never seen one accom-
plish more than making a slight dimple in the loose sand.
But quite often the careless Bembex has gone away leaving
the burrow open, and then the Idia has its chance. It
hurries up with such eagerness to get down the burrow
that it may be seen literally to tumble head over heels
as it scrambles down. There is need for hurry, for how
can it tell when the rightful owner will come back, and
then it may meet with very rough treatment ! But as
a matter of fact I have never seen Bembex take any
notice of the mean little intruder, though it may some-
times actually be met at the mouth of the hole by the
returning huntress. Yet sometimes it does fall a victim
to the sharp sting, for among flies dug up from a burrow
I have found specimens of Idia.

Small pink dipterous larvae that are found in the
burrows are probably Idia, though I was unsuccess-
ful in rearing them. It is, of course, possible that
they are not strictly enemies of Bembex, even in an
indirect manner, by devouring the larva's food ; they
may merely feed on debris of legs and portions of bodies
left by the more fastidious Bembex larva.

Every colony of Bembex, of those species that I have
seen, has had these little brown flies in attendance. Some
were sent to the British Museum, and I was informed
that the species was a new one, and this habit had not
previously been recorded for the genus.

The first Bembex that I ever saw at work was B. cap-
ensis, which I found at Jinja in 1910. One was seen
going into the burrow carrying a full fed Tse-tse, whose
shining red, bloated abdomen full of blood was quite

BACK OF HOUSE ON ROME ISLE, I914.

FRONT OF KOME HOUSE.

To face p. 292.

BEMBEX 298

unmistakable. Another burrow that was opened up
was found to contain the remains of several small flies
and of the Tabanid Chrysops hrucei, besides two Tahanua
thoracinus. The fact that Bembecidae prey upon Tabanidae
is well known, but not, I think, that they prey upon
Glossina, the Tse-tse fly.

In March 1912, I saw the following curious incident
on Bugalla Isle, where I had found a colony of Bembex
ugandensis. One was seen to come out of her burrow,
and soon returned with a small fly, went down, quickly
came out and flew away, leaving the burrow open.

There now appeared on the scene a medium sized
fossor of quite a different type, with the black and orange
** Lycoid " colouring. To my great surprise the stranger,
after hunting about, found the burrow and went straight
down ! Then she came out again and stood waiting at
the mouth on the heap of loose sand, but soon went in
again and stood with her head blocking up the entrance.

The Bembex now came back carrying a fly of some
kind, and attempted to go down the burrow, but on
meeting the intruder flew off in a great fluster, dropping
her prey. She soon came back, tried again, and again
flew off, nonplussed. The third time she managed to
get in, and I listened for the sound of an underground
battle, but heard nothing.

However, in a few seconds the stranger came out,
not at all hurriedly, and flew away, but soon came back
and went in again while the Bembex was still there,
then came out again and stood at the mouth. I then
incautiously frightened it away by moving.

The Bembex now came out and flew away, and the
intruder came back and stood inside the burrow again
with its head at the entrance. After a while it flew
away, but again came back and stood on the heap of
loose sand outside the entrance, then buzzed round for
a while and finally flew away.

294 HYMENOPTERA

I tried to dig up the contents of the burrow, but failed,
and am at a loss to know what is the explanation of the
intrusion. Unfortunately the Lycoid fossor was not
caught for identification.

Another family of Fossores is the Sphegidae, which
contains insects of very varied size and habits, although
they all agree in storing up paralysed insects or spiders
for their young. The species of prey chosen by Sphegids
differ greatly ; some choose spiders, many more grass-
hoppers or crickets, and many others the smooth bodied
caterpillars of Noctuid or Geometrid moths, but never
hairy caterpillars such as those of Arctiidae or Lymantridae,
though I have seen one species taking a pilose Lycaenid
caterpillar into its burrow. If the species selected is of
small size, a number must be collected, but very often
a single individual furnishes all the food that a Sphegid
larva requires.

Although they all belong to the natural group Fossores^
the Sphegidae do not all dig in the ground.

A dead bough that contained a nest of a large Carpenter
bee was also used by a large black Sphegid for its brood.
A tunnel about half an inch in diameter had been exca-
vated, and the mouth was stopped up by little pieces
of wood and lichen. It penetrated perpendicularly into
the wood for about an inch and then turned sharply to
the right and ran longitudinally about four inches. At
the end was a collection of the remains of some. half-
dozen small Locustidae, and a cocoon about an inch and
a half long, formed of an outer silky and an inner papery
layer. Inside the cocoon was the larva, and six small
larvae, probably Diptera, which were presumably indirect
parasites like Idia. Beyond the cocoon the passage was
directed towards the centre of the bough, then turned
to the left for two inches, and eventually opened to the
exterior below the original opening. This second orifice
was not stoppered. It is much to be regretted that

SPHEGIDAE 295

the species was not identified, since tliis method of nesting
is very unusual for a fossor.

Another very unusual nesting place was found to be
used by a specimen of Sphex (Isodonta) pelopaeiformis.
This was in the open end of the hollow stem of a broken
reed. The Sphex was seen going there carrying a small
piece of wood, and I found she had finished provisioning
her nest and was stopping the hole at the end of the stem.
There were two collections of stored up Locustidae, belong-
ing to species of very different types ; all were young,
and in some the wings had not begun to grow. There
were nine in one chamber and six in the other, the two
being separated by a partition partly of earth, partly
of some soft brown substance of almost woolly texture
whose origin was unknown. One Locustid of each col-
lection had the Sphex's egg affixed to it on the under
surface close behind the head. It is very interesting that
this fossor, while abjuring the earth as a nesting place,
still made use of it as material wherewith to form a
partition between the two collections of prey.

Another Sphegid with nesting habits unusual among
this group is the extremely common and widespread
Sceliphron spirifex, familiar in every house ; it is even
found in South Europe.

This rather elegant black and canary-yellow species^
with long limbs and narrow pedicle to the abdomen, is
classed as a " Mud wasp " by the annoyed housewife,
from its habit of building a collection of earthen cells
in any convenient nook, such as the fold of a curtain,
the corners of open boxes, the back of a book, and such-
like places. Like most Sphegidae, this energetic worker
emits a curious note when working — a quavering, high-
pitched buzz rising and falling slightly that is presumably
made by quick vibrations of the wings as they lie upon
the back. There are few houses in which this sound
cannot be heard coming from some obscure corner. The

296 HYI^IENOPTERA

completed cell is about an inch and a half long, thick
walled and smooth, with an internal calibre of an ordinary
lead pencil. It is stufifed full of small paralysed spiders,
on one of which the elongated egg has been deposited.

The Sceliphron obtains its building material from a
spot where the earth is damp, such as the edge of the
lake, and gathers up in her mandibles a pellet about the
size of a sweet-pea seed. With this she at once flies back
to the chosen spot, and plasters it on to one end of the
commenced cell, smoothing it with her mandibles very
deftly, singing the while her high-pitched working song.

No sooner has she used up the pellet than back she
goes for another, until the cell is completed, save for the
closing of one end, which is done when the stock of spiders
has been brought.

On Kerenge Isle, where spiders were exceedingly
abundant, Sceliphron was also plentiful, and my grass
hut was a very favourite nesting place, so that the high-
pitched song of the worker was heard all day long. While
sitting writing one morning I timed the journeys of one
individual, noting the moment when she reappeared with
a fresh pellet of moist building material, which she applied
without waste of time, flying away again immediately to
bring another pellet.

o The following are the times noted : 8*45, 8* 48, 8- 50^,
8-57i 91, 9-6, 9-8^ 912|, 917, 9-23|, 927, 930, 9-33|,
9-36|, 9-391, 9-411, 9Ul, 9-47. After this she did
not reappear for half an hour, and I went out.

The average time taken to fly from the nest to the
lake shore, collect a mud pellet, bring it back, and build
it on to the nest was 3-6 minutes, the shortest being two
minutes, the longest seven.

Although the cells are often built singly, they may
be aggregated, being built one alongside the other, so
that a single mass the size of a small fist may be formed,
comprising eight or ten cells, each stuffed with spiders.

SCELIPHRON AND ITS PREY

297

It is quite extraordinary to think that all that material,
weighing perhaps half a pound, has been brought by
one insect.

The destruction of spiders wrought bj'' one Sceliphron
in providing for a future generation is also surprising.

On Kerenge Isle I counted the number taken from
twenty -nine cells. In the following table a cell is indi-
cated by a serial number, and all cells found attached
together, i.e. built by one individual, are indicated by
the same letter.

lA
2A
3A
4B
6B
6C
7C
8D
9D
lOD

}

.. 6

IID

8

21G

7

12E

. 12

22G

8

13E

8

23H

..25

14F

8

24H

15F

8

25K

1

16P

. 5

26K

7

17F

4

27K

5

18F

6

28K

..12

19

. 12

29K

.. 14

20

6

7
9
11
14
12
15
11
11
19

The average number of spiders in a cell was 9-3, the
minimum one, the maximum nineteen. The variation
in the amount of food stored up is considerable, even in
cells constructed by one Sceliphron. Thus No. 6 had only
a single spider, very little larger than one of the seven
in cell No. 7. The spiders are usually of a number of
different species, but the twelve in cell No. 19 appeared
to be the same.

The destruction of spiders by Sceliphron must be con-
siderable. Supposing that ten cells are completed during
the lifetime of one individual, it will have accounted for
about ninety spiders.

Sceliphron itself, however, is victimized by a large
** Ichneumon " fly of the same black and canary -yellow
hues presenting a remarkable likeness to its victim,
especially when seen on the wing at a little distance.
Two-winged flies {Tachinidae) can also be reared from

298 HYMENOPTERA

Sceliphron cells, but they are probably indirect parasites,
their larvae feeding upon the spiders and not upon the
Sceliphron larvae, which thus die of hunger.

A smaller species of Sceliphron makes cells somewhat
similar to the mud nests built by S. spirifex, but they
are placed upon grass stems and are of more delicate
structure, being composed of some light friable material
mixed with fibrous substance. These nests were common
also on Kerenge Isle, and were filled with spiders.

The great majority of Sphegidae, however, are true
fossors — that is, they excavate burrows into which the
paralysed prey is put.

Since a Sphegid having once stocked the burrow pays
no more attention to it, there is no need for it to work
like Bembex in loose light soil ; indeed, the firmer it is
the better they seem to like it, and often may be seen
working on the hard surface of a trodden pathway.

The methods of a Sphegid differ somewhat from those
of a Bembex. The earth cannot be removed by mere
scratching, but needs to be excavated by " tooth and
nail."

An Ammophila, for instance, having found a suitable
site, commences to scrape together a little of the loose
soil on top, which she holds between her front feet and
head and, having walked backwards a few inches, deposits
it with a little buzz of satisfaction and returns for another
armful. Presently the soil is found firmer, and she sets
to work with her powerful mandibles to loosen it, some-
times bringing up portions in her jaws. The work pro-
ceeds with rapidity until the hole descends vertically to
such a depth that she disappears from view.

She seems to be on springs, so full of life and energy
is she ; it is often extremely curious to note how rapidly
she shoots out from the burrow backwards, drops her
load, and plunges headlong down the hole again as if
drawn by a piece of elastic attached to the bottom thereof.

AMMOPHILA 299

Occasionally a piece of stone defies removal, and a pro-
longed and angry buzzing from beneath the surface
attests her efforts to dislodge it with her mandibles, but
presently she works it loose and comes up with it.

At last all is ready, and she proceeds to cover the mouth
of the burrow before going off to seek the prey where-
with to fill it. Pieces of stone of small size are selected
and wedged together so as to block the entrance, and
with her back to it she scratches a little loose earth over
the stones and, after a final careful examination, flies
away.

Now you may see her quartering the ground for the
caterpillar, grasshopper or other insect which she needs ;
wings flicking and antennae quivering, she runs about
eagerly searching until — pounce ! — she has found one
and stung it, and it lies helpless on the ground, with
limbs slightly twitching.

Picking it up by her mandibles, unless it is too
large, she walks with it back to her hole, keeping a
marvellously straight course over all kinds of obstacles,
even climbing with it up perpendicular rock faces, until
after much effort she arrives at the site of the burrow.
This is quite indistinguishable from the surrounding soil,
but she knows it, for did she not herself take pains to
conceal it before she left it ? The precious burden is
put down, and the barricade that closes the burrow is
removed. Down she goes head first, and presumably,
after making sure that all is well, turns round in the
chamber at the bottom, for almost immediately her
head reappears and she seizes the prey and drags it down.
After a very brief interval, during which the egg is laid,
she reappears and at once sets to work to fill up the
burrow. Standing with her head away from the open
mouth, she vigorously scratches the loose earth back-
wards into it with sure aim. But to make a good job
of it something more is required, so she picks up pieces

800 HYMENOPTERA

of stone and puts them down the hole, and then proceeds
to ram it all down in a most workmanlike manner.

Firmly gripping the lips of the hole with her middle
and hind legs, she launches herself with great force against
the loose earth, using the front of her broad head as a
battering ram, and making the loud buzz wherewith a
great effort is signalized. At length the burrow is filled
up flush with the surface, and it only remains to con-
ceal the fact that the spot is in any way different from
the surrounding terrain. Small stones are scattered over
the surface, a little loose dust is scratched about, and
the energetic Ammophila flies off to refresh herself on
a neighbouring flower head before commencing to excavate
a fresh burrow elsewhere. Unlike the Bemhex, she will
never visit her burrow again ; it has been adequately
stocked with food once and for all.

The above is a general description of the methods
adopted by the majority of Sphegidae, and we will turn
now to another great family of fossors, the Pompilidue,
which is broadly characterized by rather longer limbs,
and by the fact that they hunt spiders only. In essentials
they work in the same manner as the Sphegidae, but so
far as I can remember they work quietly, and never
make the excited buzzing so characteristic of Sceliphron
spirifex and other Sphegidae.

But when it comes to the filling up of the stocked
burrow there is an important difference in the manner
of working, which I do not think has been sufficiently
appreciated.

As described above, a Sphegid uses the front of her
broad head as a battering ram wherewith to press down
the loose earth; but Pompilidae invariably use the end
of the abdomen.

Sometimes the insect stands in the mouth of the hole
and quietly presses down the soil in the burrow, but one
specimen that I saw threw her body into such quickly

POMPILIDAE 301

repeated movements that her outline became blurred,
and she reminded one of the rapid oscillations of a
pneumatic hammer.^

Perhaps Pompilidae do not use the front of the head
because it is less broad and flat than in Sphegidae, and
therefore does not form such an efficient rammer ; but
the difference is extremely remarkable.

The Pompilidae are most typically aposematic insects.
Many are of intense blue-black with shining bluish or
greenish black wings, often set off by canary-yellow
antennae. The sting inflicts a very painful wound, but
they do not use it unless molested, and may be watched with
perfect immunity so long as they are not roughly handled.

It is difiicult to conceive a more conspicuous insect
than one of these fine creatures as it sails through the
air with its long legs hanging down. Yet according to
some of the American naturalists these insects are con-
cealed by resemblance to their surroundings.

In that case it is difficult to explain one of the most
striking characteristics of these large Pompilidae, namely,
the extraordinarily noisy flight, which is often a loud
rattling or clicking noise audible before the insect is seen,
and reminding one of a badly made piece of clockwork.

If a Pompilid is really concealed, what is the meaning
of this noise which at once directs attention to it ?

If, on the other hand, these insects are really highly
conspicuous, as I think, the loud noise is merely an
adjunct to their aposematic colouration, and serves still
further to advertise their owner's unpleasantness.

Many times have I seen my pet monkeys' attention
caught by the noise of one of these formidable insects
on the wing, and they would turn their heads to the
direction from which it came, and watch the Pompilid
when it came into sight with the most careful attention,
being evidently anxious to avoid it when it drew near.
1- Batozonus fuUginosus. See Proc. Ent. Soc. Lond., 1917, p. Ix,

302 HYMENOPTERA

This " aposematic noise " is entirely in keeping with,
and explicable by the same means as, the conspicuous
appearance of the insect, which is thus enabled to appeal
to more senses than one. But if the Pompilid is con-
cealed, then one explanation is required for its colour
and another for its noisy flight, though it is difficult to
imagine what explanation of the latter could be given.
I have suggested elsewhere ^ that there is an analogous
example among birds in the case of the large black and
white " E'nga-nga " hornbill, whose wings make an extra-
ordinary roaring noise as it beats them, or sails with
them outspread through the air, often at the same time
calling attention to itself by its loud raucous cry.

It has been mentioned in the case of the Bemhecidae
that they have enemies in certain flies, and the same
holds good with Sphegidae and Pompilidae.

The golden moment for the enemies of these fossors is
when they have arrived with their burden at the mouth
of the burrow and leave it outside while they go down.
The attendant fly can then dart down and deposit its
own egg upon the prey while the fossor is busy, and that
it does so is evidenced by the fact that on grasshoppers
taken out of the burrows of the little Sphex marginatus
I found dipterous larvae which became flies of a species
unknown to the British Museum, but allied to the Milto-
gramma described by Fabre. It is rather interesting
that whereas most fossors seem to take little notice of
other insects while they are at work, Sphex marginatus
is extremely suspicious of any intruders on its privacy,
and charges fiercely with open mandibles at even the
smallest ant that draws near to its burrow, much to the
dismay of the quite inoffensive ant ! If the trespasser
does not at once withdraw, the Sphex makes threatening
movements as if about to sting, turning the tip of her
abdomen forwards between her legs.

1 p. 168.

MUTILLIDAE BEES 303

Another family of fossors, the Scoliidae, has been men-
tioned because they save themselves a great deal of
labour by searching out the food for their larvae and
laying an egg on it in situ.

They are interesting also from another point of view,
because, of all Hymenoptera, they have the quietest flight ;
even the largest specimens produce a barely audible hum,
thus contrasting greatly with Pompilidae. They are
usually rather solidly built, hairy insects, sometimes
black, sometimes reddish or yellowish in colour.

Lastly, there are the Mutillidae, which will only be
mentioned on account of their scarcity on the islands.
It seems that forested country and a damp climate are
not so suitable to them as the more open bush country
with regular dry seasons such as occurs in much of ex-
German East Africa, where I found Mutillidae far more
abundant than on the islands. They are remarkable
among Hymenoptera for the complete absence of wings
in the females, which are often mistaken for ants as they
run quickly over the ground. A species of Methoca was
obtained which was not in the British Museum collection :
the few African species that are known of this genus
had previously only been obtained from south of the
river Limpopo.

Regarding Bees I have little to say. The very hand-
some large " Carpenter bees " (Xylocopidae), which in
the tropics take the place of our " Bumble bees "
at home, are plentiful enough, and may often be seen
visiting the large yellow flowers of ambatch trees. A
particularly handsome large one is covered with bright
golden brown hair ; this is the male of a species whose
female is black and white. Their burrows in dead
trees are commonly seen. The natives call them
** Buvumira," which is quite descriptive of a large
buzzing insect.

A smaller bee {Crocisa meripes) presents a great con-

304 HYMENOPTERA

trast to these furry fellows, as its integument is hairless,
and brightly coloured sky blue and black.

I do not know where Crocisa nests, but it has a remark-
able habit of passing the night in the open on a grass
stem, its mandibles firmly clasped round the stem, and
its legs tucked up against the body, so that the insect,
supported only by the grasp of its mandibles, projects
obliquely from the stem. They may often be found dead
in this position.

Sometimes several will affix themselves thus, one
below the other, and the stem may be picked and carried
about without disturbing the bees. Another species
{Coeliaxn carinata) has the same resting habit. The
islands are singularly blest in the absence of minute,
stingless, but very annoying bees (Melipona), commonly
known as ** Sweat bees," from their habit of settling on
one's bare skin and licking up the perspiration. If they
would only settle quickly one would not object so strongly,
but they spend a long time dancing about in front of
one's face under the brim of one's hat ; when they do
settle they cause an intolerable tickling.

Though they have lost their sting they are very possibly
still protected, for if one is squashed by a blow it has
an unpleasant acrid odour. I first made their acquaint-
ance on active service, and soon learnt to give thanks
that the islands did not support them !

Like the Mutillidae they seem to prefer more open,
drier country ; hence the thirst which leads them so
eagerly to suck up moisture from one's skin.

FLY boys' camp ON KOME ISLE, I9I4.

servants' huts on KOME ISLE, I9I4.

To face p. 304,

CHAPTER XIII
SUNDRY INSECTS

COLEOPTERA.

Beetles were not systematically collected on the islands,
so that there is not so much to be said about them as
about Lepidoptera.

The first interesting specimen that was met with on
the lake shore was a new species of Coprid, which Mr.
Arrow informed me was quite unlike anything in the
British Museum. I was watching a column of the
" Safari " ant (Dorylus) moving from one camp to
another, carrying their pupae with them, and saw this
flat, black, highly polished beetle running in the column
among the ants, who took not the slightest notice of it.
That it did not belong to them was improbable, for it
took the greatest care to remain in the column, and
these ferocious ants most strongly resent the presence
of any stranger among them. It is interesting to find
that these dread hunters have their familiars. Possibly
the larva acts as a useful scavenger in their nests, as does
the larva of the Rose beetle in nests of other ants.

While at Jinja in 1910, I watched with interest the
behaviour of some large polished green Coprids engaged
in making their balls from the recent droppings of some
grazing animal.

When I first came up one had almost finished
making its ball. When it was finished the beetle pressed

21 305

806 SUNDRY INSECTS

the material together with its front legs and then very
cleverly rotated it on one spot, lying almost underneath
it, so as to get a firm coating of earth on it ; it then
trundled the ball away. Soon other beetles arrived on
the wing, dropping quite near to their objective and
eventually crawling to it, obviously guided by smell.
If one happened to meet another when both were search-
ing for the dropping, one always fell in and followed
behind as if quite certain that the other must know the
way ! I watched No. 2 make its ball. The desired
portion of the dropping was the outer part which had
hardened somewhat. The beetle stays at one spot and
reaches towards it with its powerful forelegs enough
to make a ball about the size of a very large cherry.
It frees it below by insinuating the sharp front margin
of the head between the ball and the mass, and making
powerful lifting movements of its head.

No. 2, after making the ball, did not for a long time
push it away, but kept pressing it together and patting
it, and finally gave it a very smooth coating of fine
earth, and then sat resting on it.

Meanwhile others had arrived (ten altogether came
while I watched), some smaller in size. One of them,
No. 3, apparently desired a share in No. 2's ball, but
was driven off, hurled away by jerks of the strong front
legs. Eventually it joined with another small one,
No. 4, and these two very quickly made a rather ill-
constructed ball and trundled it off, No. 2 still patting
at its own.

There were numerous fights between the workers, who
sometimes got mixed and each took the other's ball.

These beetles are eaten by a species of Roller, one
of which when shot was found to have several of them
in'^its stomach. Presumably its powerful bill enables it
to^feed on such horny beetles, whose large size and hard
coat of mail must deter smaller birds. Of the carnivorous

BEETLES 307

ground beetles I saw little on the islands, but two species
of Cicindelidae (Tiger beetles) were found, which were
very acceptable to the British Museum. One, a pretty
green species, not unlike the English G. campesfris, but
larger, was found abundantly running over termite hills
on Nsadzi Island. The other was a mottled grey species,
found, but very difficult to see, on a stretch of sand on
Bugalla that had at one time obviously been the lake
shore. The huge black Cardbidae,^ common in Africa,
were never seen on the islands ; though I do not assert
that they were not there, it seems curious that they
should never have been met with.

Carnivorous water beetles {Dytiscidae, Gyrinidae, etc.)
were of course abundant in the weedy shallow waters.
The readiness with which they take to the wing is well
known ; but it was quite amusing to note how, during
a storm of rain, several Gyrinidae appeared in the trench
which led away the rainfall from the roof and solemnly
executed their well known whirligig manoeuvres so long
as the water lasted, and when the storm was over they
vanished. Large Dytiscidae several times dropped on
to a shiny patch of wet ground during heavy rains,
evidently misled into thinking it was a pool.

The beautiful Longicorn beetles were often met with,
and one very abundant green species of medium size ^
had a very rank aromatic odour, and my pet monkey
would not eat it.

The most interesting Longicorns are the mimetic
species, and a very fine one ^ was abundant on the
flowering shrub Haronga, where it was associated with
the large Lycidae which it so much resembled. Other
Longicorns mimic the parasitic Hymenoptera of the family
Braconidae, many of which are typically aposematic in
appearance, and have a powerful, rank odour.

^ Genus Anthia. ^ Phrosyne brevicornis.

3 Amphidesmus analis.

308 SUNDRY INSECTS

A common colour scheme for large Braconids is orange
and black, and on Bugalla a Longicorn ^ beautifully
mimetic of these was found. Being long and narrow,
the beetle has a shape which affords a good basis for
the resemblance. The orange abdomen, however, does
not show the narrow waist of the Hymenopterous model,
to which the resemblance is produced by a portion of
the side of the base of the abdomen being of a glistening
white, contrasting strongly with the remainder, so that
that part in a high light suggests that it has been, to
use Professor Poulton's phrase, " painted out." The
first specimen that I found was a male which when
handled strongly curved the tip of the abdomen in such
a way as to suggest it was about to sting, and actually
protruded a flexible white viscus which it moved about
like a sting. Of course Braconidae are not strictly
stinging insects, yet when handled they will use the
ovipositor as such. The Hymenopteroid appearance of
the beetle with its false sting was so very striking that
although reason told me it was a beetle, instinct was
so strong that misgivings almost prevented me handling
it, and I feel certain the very great majority of people
would have dropped the beetle in a panic. On the
wing the resemblance is much greater ; both insects
have a slow, steady flight, and the long antennae are
extremely conspicuous. The wings of the beetle are
transparent and invisible during flight, but the orange,
black-tipped, wing covers reproduce the appearance of
the similarly coloured wings of the Braconid. Several
smaller species of Longicorns also very closely resemble
smaller black and yellow Braconids, so that even after
several years of field work one is still deceived and,
catching an insect which one has thought to be a Braconid,
finds a beetle in the net.*

^ I)irphya, species near princepa.

• This happened to me repeatedly in 1917 in German East Africa.
See Proc. Ent. Soc, 1918, pp. cxxxviii-cxlii

BEETLES 309

Many Braconids are coral pink and black, and they
also are mimicked by small Longicorns. On Bugalla
I obtained model ^ and mimic * at the same spot and
time, and also a small Capsid bug,' equally mimetic.

Fireflies are interesting beetles, but, having read much
about the wonder of tropical fireflies, I was much dis-
appointed on the islands ; only a few were seen at a
time, belonging to a species looking much like the male
of the English glow-worm, and I have seen a much better
display in the spring on the Riviera.

Wood-boring beetles are a great nuisance in a house
constructed of freshly cut boughs and stems of trees ;
the rafters are soon penetrated in every direction by
tunnels, from the open mouths of which a continuous
stream of the finest sawdust falls and soon covers any-
thing left lying for long in one place.

Quite the most remarkable beetle met with on the
islands belongs to the aberrant family Lymexylonidae,^
it several times appeared in my hut in the evening,
flying with sonorous buzz, and the first time I saw it
puzzled me not a little, for it was like no beetle I had
ever seen or heard of at that time.

The very long, thin body was quite soft and flexible,
and a very short pair of wing covers entirely failed to
cover a long pair of wings with stout longitudinal
rays so that they folded like the wings of a grass-
hopper, but then projected far behind the covers. The
legs were long and thin, and the antennae short, with
broad flat segments.

Nexjroptera.

Dragon flies are, of course, abundant on the lake shore,
but I do not think they are any more noteworthy than

^ Not yet identified.

' Dirphya, species near pascoei, Lamiidae.

' KolopeUis bergrothi. * Atractocerus brevicornia.

310 SUNDRY INSECTS

our beautiful English species. One fine, abundant species
is blood red, and another is vividly coloured with cobalt
blue and red, but there are many dull species. One of
these,^ the commonest of all, may be seen on the wing
before sunrise and after sunset, and on one occasion on
Kerenge Island I saw a host of them about sunrise,
dancing up and down in the air just as Mayflies do
in England.

The large species of Dragon flies often prey on the
smaller, as well as on bees and other protected insects.
I have seen Cacergates preying on Glossina, and since
both large and small Dragon flies are favourite articles
of food of the bee-eater Merops superciliosus, the
bionomic relation of this bird to the Tse-tse fly is rather
a complicated one !

A curious Neuropterous insect named Bittacus was
attracted to light several times on the islands ; it is a
slender light brown species with narrow wings and very
long slender legs. On the first occasion I thought it
was a Tipulid, and did not take much notice until some-
thing about it made me look more closely, and I saw it
was not a " Daddy longlegs," but had four wings.

If there was any reason for supposing the Tipulidae
to be protected one might consider the Bittacus to mimic
the fly ; but there seems no reason for supposing that
a resemblance to a " Daddy longlegs " would be of any
benefit.

A remarkable genus of Neuroptera is named Mantispa.
These insects resemble Mantidae, and have the fore limbs
modified in the same manner for holding the prey, which
they devour alive. Several of them very deceptively
resemble Hymenopterous insects. On Kome, one day,
I saw on a twig an insect that at first appeared to be
one of the abundant, conspicuous and fiercely stinging
wasps of the genus Belonogaster. It was munching at

^ Cacergates leucoatieta.

MANTISPA 311

some insect it had caught precisely as the wasp chews
up caterpillars into pulp, and when disturbed flew with
it on to a neighbouring branch. But the thickness of
the abdomen caught my attention, and disclosed the
true nature of the insect. The question may well be
asked, How was it that the broad wings of a Neuropterous
insect did not attract attention, as differing from the
narrow folded dark wings of the wasp ?

When the Mantispa was at rest the fore wings
lay above the posterior wings and concealed them.
Their greater part was absolutely transparent and
hardly noticeable, but the appearance of the narrow
wings of the wasps was produced by a darkening of a
narrow strip along the anterior margin of the fore wings,
which dark strip roughly resembled in shape the whole
of the folded wings of the wasp.i The Mantispa at rest
held its wings directed backwards as does the wasp,
but not quite parallel to the body, so that they made a
slightly greater angle with each other than did the
wasp's wings.

In colour the Mantispa very closely resembled Belono-
gaster, and the resemblance, probably pseudaposematic
or true mimicry, was extraordinary, and would not be
believed if one had only seen the specimens in the cabinet.^

Orthopteea.

Mantidae. — One species, Pseudocreobotra ocellata, has
already been alluded to as an instance of Anticryptio
colouring ; other equally cryptic grass green species
were particularly abundant on certain islands, an account
of which was given in Chapter VI. Sometimes the
curious egg clusters were very abundant ; the eggs are
embedded in a substance somewhat resembling the crust

1 See also Trans. Ent. Soc. Lond., 1902, Part III, pp. 528-9.

2 Cp. G. A. K. Marshall, Trans. Ent. Soc, 1902, Part III, pp. 536-7.

312 SUNDRY INSECTS

of a meringue, covered by a substance resembling parch-
ment in texture, but produced by the Mantis as foam
at the time of oviposition. These egg masses vary much
in shape, size and texture according to the species — some
are pure white, spherical, and about the size of a bantam's
egg ; others, no larger than a hazel nut, have an amber
coloured, almost transparent covering, with a ridge along
one side looking like a seam where two halves meet.
These are very conspicuous, and may be seen at the tips
of twigs, etc. ; but there are other very cryptic forms
of egg masses, laid along a twig and closely adherent to
it, with rough opaque brown surface. The freshly hatched
Mantids may sometimes be found hanging from the
egg clusters by slender threads, but I do not know
for how long they remain thus. Very young specimens
are often black, and look extremely ant-like as they run
about.

Phasmidae. — Stick insects found on the islands are
small and not remarkable : " grass insects " would be
a better name, because they are found in long grass, the
stems of which their slender wingless bodies resemble
in size and colour. They frequently betray themselves
by a curious swaying from side to side, for which it is
difficult to account, for were they to remain quite still
they would certainly be invisible.

Locustidae. — The most noticeable insect of this family
was a slender active grasshopper known to the natives
as ** Ensenene " ; about two inches long, usually grass
green but sometimes light brown. These grasshoppers
used to appear suddenly in great numbers and could
be heard faintly chirping in the grass. As one walked
along they took to the wing in clouds, and their
pearly wings flashing in the sun produced a curious
resemblance to driven snowflakes. At these times the
natives would go out and catch large numbers for eating,
walking through the long grass and whistling through

ORTHOPTERA 813

their teeth to make a sibilant noise. They were prepared
for food by frying. My pet monkeys were also exces'
sively fond of them, and would eat them for as long as
they were offered, even until the overloaded stomach
rejected them.

A ghoulish species of Ephippigeridde was a great
nuisance in my hut on Damba, as it found its way into
every box of food and quite replaced the kitchen cock-
roach as a pest. It has a rotund short body with im-
mensely long limbs and antennae, and is offensively
familiar ! It would come on to the table while I was
dining in the evening and stand with the filamentous
antennae waving in the air, and was so quick that I
could never catch it before it leapt away. As it
was a great nuisance I was glad to find that the egg
cases, like little portmanteaux, were attacked by a
Chalcid parasite, which laid its own eggs inside the case.
One such, when opened, was found packed with rows
and rows of the minute white pupae of the parasite, white,
with shining black eyes, looking for all the world like
neatly arranged mummies. The male of this species has
imperfect wings, and has not been heard to make a
noise.

Another very large burrowing species during the rains
sits at the mouth of its burrow making a continuous
** shrilling " noise which, if at all close, is quite un-
bearable, as it seems to make one's whole head vibrate
in sympathy.

Another Locustid worthy of notice, of the same section,
is known to the natives as " Semukutu." It is disliked
very much by them, for they say it bites and makes sore
places, and often show sores which they say were caused
by a Semukutu. When handled this insect ejects with
some force from the side of the thorax a stream of
clear, yellow fluid with an acrid smell, but though I have
handled a large number in the endeavour to collect a

814 SUNDRY INSECTS

quantity of the fluid I have never suffered in any way
from it. No doubt it would be excessively unpleasant
in the mouth or eyes of an enemy. The Semukutu is
a large, fat, bloated creature destitute of any trace of
wings, and with spiny thorax. Young specimens are
often greenish, but the adult is dull brownish grey. It
is a very sluggish insect, and freely exposes itself as it
crawls heavily about.

Acrididae. — Large and powerful grasshoppers of this
family, species of Cyrtacanthacris, are known as " Ama-
jansi " to the Baganda. Some of them are several inches
long, and were apparently much esteemed as food by
the monkeys. They were not noticeably abundant on
all the small islands visited, but seemed to be so on
Kerenge, where the beautiful bright blue kingfisher that
feeds on them was much in evidence.

Blattidae. — Cockroaches of many species were common
among dead leaves in the forest, but one saw little of
them except when the *' Safari ant " was out hunting.
Then the unfortunate cockroaches, driven from their
lairs, rushed hither and thither in frantic desperation,
and in such a state of panic that they often tumbled
head over heels, and the more easily fell a prey to the
host of ants, which at once tore them to pieces. Some
curious species may be found half buried in loose dry
earth at the base of trees ; absolutely wingless, they
look much like huge woodlice.

Dermaptera.

The earwig family was not at all well represented on
the islands, much to my surprise, for there seemed to
be ideal conditions for them. I thought at the time
that perhaps this was the rule in equatorial Africa, but
later modified this view when on active service in German
East Africa, where earwigs were found extraordinarily

cook's hut (left) and kitchen on bugalla.

Cook on right. The water boy on the left developed sleeping sickness.

.-..- t.

HEAD BOV (A SWAHll.Ij AM) SECOND BOY (MUGANDA) ON RIGHT.

Their wives had come out for a brief visit.

To face p. 314.

FLIES 315

abundant in some parts of the country. So that the
islands would appear to be unfavourable to them.

One large species was found on Damba under a stone
at the very water's edge, so that the tiniest ripple must
have surrounded it.

DiPTERA.

The most important fly on the islands, namely Glossina
palpalis, has had a chapter to itself. Other biting flies
(Tabanidae) were less common on the islands than on
the mainland, but T. variabilis a small black and white
species, and Chrysops brucei were occasionally seen, with
one or two species of Hoematopota.

Mosquitoes, of course, were to be reckoned with, and
they were particularly abundant on Damba Island,
perhaps because it was covered with low lying forest.
But the Anophelines, carriers of malaria, were so
scarce that the numbers of times one was seen could
be counted on the fingers, so that the health of the party
was excellent. The Simuliidae, or " Buffalo gnats," have
been alluded to as particularly troublesome on certain
islands when rain was threatening ; they frequent rocky
open shores, and when abundant render such spots un-
approachable. The species found on the islands was
not the same as the one I had previously met at Jinja,
S. damnosum, called " Embwa " by the natives ; the
island SimuUum is smaller and rather prettily coloured
with golden pubescence. Both leave a severe burning
and itching sensation, but whereas S. damnosum attacks
particularly the lower extremities, the island species
makes for one's head and neck, and especially the ears.

The parasitic Tachinidae, which lay their eggs on the
surface of living insects, have been briefly alluded to as
enemies of " protected " insects.

On one occasion a species was observed in the act of
oviposition.

316 SUNDRY INSECTS

On the minute island Lula, several colonies of very-
hairy caterpillars, of the famUy Euptevotidae, were seen
on trunks of trees. Near one colony was a large Tachinid
fly, and the caterpillars, close together, seemed to be
aware of her presence, for they moved uneasily ; but
the fly moved as they moved, sidling about in a very
amusing manner so as always to face the caterpillars.
She endeavoured to find one that was not moving,
and then approached the head from in front. The long
hairs of the caterpillar projected from all parts except
the head, so as to form a chevaux de frise, but opposite
the head there was a small gap in the barrier. The fly,
having approached as close as she could, raised herself
up on her anterior legs and protruded forwards, beneath
herself, an enormous ovipositor whose tip projected in
front of her head. The egg was laid in a very brief instant
on or near the head of the caterpillar, but sometimes
the fly was unable to reach its head, and then had to
wait until the larva was near enough for her ovipositor
to reach between the long hairs so as to deposit an egg
on the flank of the larva.

It was a most interesting and instructive proceeding,
illustrative of the adaptation of one enemy (the fly) to
meet a condition in its victim probably directed against
other enemies (birds). It was noted that the fly laid
its eggs fortuitously on any larva that presented itself,
so that one received many eggs but others none. This
might quite well lead to some of the offspring of the fly
receiving insufficient nourishment and being stunted in
growth or failing to develop. One sometimes meets the
opposite condition, when too few ova have been deposited,
and the parasites have not damaged the vital parts
of their host. I once reared a single Tachinid fly from
an Arctiid larva which eventually produced a perfect
moth !

Asilidae or Robber flies were often met with at the very

FLIES 317

margin of the forest where it abuts on open grass land ;
others in the forest, and others on the grass land. Bobber
flies seems a poor name for these powerful, hairy, pre-
daceous flies, with long narrow bodies and strong legs.
Hawk flies would be a better name, seeing how they
pounce upon their prey ; in some six years of work in
the field I can only once remember seeing an Asilid
pounce on its prey sitting.

Emphasis has already been laid upon the importance
to the theory of mimicry of noting the prey of Asilidae.
It is, I think, quite certain that the fly injects poison
into the victim the moment it has been captured and
the proboscis has been plunged into it. The prey seems
to succumb at once before it can have been sucked dry,
and if one actually witnesses the capture and at once
catches both insects the prey is almost always found
to be dead, or feebly moving its legs only. Only once
have I met with an exception to this. An Asilid was
seen to catch a bug, and I struck at it with the net. The
fly escaped, but dropped the bug, which was found to be
apparently unharmed.

Another family of Diptera, Chironomidae or " gnats," is
worth mentioning here, owing to the abundance in which
they appear over the lake during and shortly before the
rainy season, resembling clouds of smoke from distant
steamers. I have seen, on a calm day, a large area of
water covered by the pupal skins of these flies, the pupa©
having come up from quite deep water, and the flies
rising in a cloud from the surface, which was of a brown
tint from the myriads of empty skins. When these flies,
called *' E'sami " by the natives, and looking more or less
like mosquitoes without the sucking proboscis, have
drifted in a cloud on to some land they find shelter from
the wind on the lee side of trees and bushes, where they
may be seen hovering in a cluster, the end of which,
furthest from the tree, is strung out and torn by the wind.

318 SUNDRY INSECTS

On a very calm and damp clouded day the air is filled
with the high piping of the Sami, hovering in myriads
in the shelter of bushes, and a sudden noise will cause
them all to rise suddenly upwards. It was very amusing
to sing a scale, for one note appeared especially to upset
the Sami, and when it was reached every member of
the hovering cloud would simultaneously leap an inch or
so upwards. The natives catch large numbers in baskets
like strawberry baskets made of plaited grass through
which a stick is passed ; the whole is vigorously waved
about in a cloud of Sami, the basket rotating around the
stick, which passes across its diameter. The catch is
compressed into a cake, but I do not know how it is
prepared for food.

Sami have an odour of the lake which it is difficult
to describe — a smell like weeds and fish from a muddy
pool ; it is a common saying at Entebbe that the arrival
of clouds of these gnats produces an outbreak of nasal
catarrh among the white inhabitants ; possibly this is
of the nature of " hay fever."

Spiders.

The most noticeable fact about spiders on the islands
has been already recorded in the account of the tour
in 1914, namely, the extraordinary abundance on certain
islands of the huge Nephila, and the sheets of their webs.
It was noteworthy that on some of the islands the spider
was present in normal numbers only, as on Kibibi. Spiders,
generally, are called " Nabubi " by the Baganda. A
curious habit has been noticed in the case of a species,
making the typical " orb web." Over part of the web
it would spin a piece of very conspicuous, opaque, glisten-
ing white silk, which was visible from some distance
away. The design in the same web would be changed
from time to time, for sometimes there would be an

SPIDERS 319

opaque bar right across the diameter of the web, or at
others a zig-zag between two of the radii only. The
reason for this is obscure.; I can only suggest that it
is analogous to the trapper's artifice of putting an object
across the track of some animal which, making a detour
to avoid what it imagines to be a trap, falls into the real
trap set at the side. An insect on the wing, supposedly,
sees the conspicuous part of the web, takes care to fly
past the side of it, and is caught in the inconspicuous
part which it has not seen.

On an earlier page attention was directed to silver
markings on insects ; a spider found on Dwaji Isle in
1919 was notable. It was of fair size and spun its web
among heads of dry grass ; the colour of the cephalo -thorax
and front part of the abdomen was pure silver, the rest
of the abdomen was dark, but speckled with yellow and
with a series of silver bands ; the legs were banded alter-
nately dark and speckled-yellow.

Of the smaller members of Arachnida, Ticks call for
notice, but only on account of their great scarcity on
the islands. This is presumably associated with the
absence of cattle and all buck except the Situtunga ;
and if the islands could be again inhabited it should be
possible to keep cattle there free from ticks and therefore
from the diseases carried by them.

The Varanus often has numbers of ticks on it, but
these are not of the species which feed on cattle ;
prettily decorated ticks ^ have also been found on the
Horned Puff Adder. The islands are also singularly
blest in the absence of Scorpions, which were not met
with ; possibly the climate is too continuously humid
for them.

The great group of Myriapoda calls for passing
notice, firstly because of the absence of the huge
centipedes found in some parts of Africa, and secondly

^ Aponomma Iceve.

320 SUNDRY INSECTS

on account of the great abundance of giant Millipedes
(" E'gongolo "). These fine creatures, with cylindrical
bodies some six to eight inches long, as thick as one's
finger, clothed in rings of polished black armour with
reddish legs, are really extremely handsome. They may
often be seen on the sandy beach feeding on decaying
rubbish cast up by the waves, and are especially notice-
able when rain is coming. They were not met with on
certain of the smaller islets, which were yet of sufiicient
size to be expected to support them.

I think that they are occasionally devoured by the
Enswa-swa, for rings of their armour may be seen in dried
excrement believed to be of that reptile. But I am
quite unaware what are the main checks on the increase
of these millipedes ; possibly the greatest loss is caused
by enemies that devour the eggs, or parasites of the eggs.
It is a curious fact that I cannot remember having seen
a specimen less than half grown.

Crustacea.

Small crabs are one of the factors of the lacustrine
fauna which suggest the sea, and they are abundant in
shallow water. Superficially they much resemble crabs
of a couple of inches in diameter that one finds in
seaside pools. The natives call them " Enjaba.^^ One
was found under a stone among damp debris about
twenty yards away from the water, a young and lively
specimen.

Crayfish were never met with, but shrimps, looking
much like the common marine species, abounded in
shallow weed grown waters.

The natives call them " Obuduli," and use them as
bait for fish, but I do not think they eat them.

Fish lice, of two species, were met with on the " Mamba "
and " Mal§ " fish already described, and specimens sent

MOLLUSCS 821

to the British Museum proved to be of some value and
interest. I

Woodlice, like other creatures such as earwigs, scor-
pions and centipedes, usually regarded as objectionable,
were exceptionally scarce on the islands, and this was
very surprising, for one would have expected a com-
paratively equable, damp climate to afford favourable
conditions.

Molluscs.

The lake shore abounded with univalves and bivalves
which had quite a marine appearance ; securely fastened
to rocks was a species resembling the common oyster.
Besides these were others very closely resembling our
English genera of fresh- water snails and mussels.

A curious slug, that . had very short tentacles
and seemed to have no " mantle " at all, was found
slowly crawling on wet sand, where also the large Ampul-
laria were occasionally thrown up by waves. Shells of
these large molluscs may be found on the lake shore,
apparently battered to pieces on a stone as our garden
snails are by thrushes. I strongly suspect this to be
the work of the " open-bill " stork.

On Damba Island there was found quite commonly a
snail of the type of our English Vitrina, but very much
larger, with a thin fragile shell practically enveloped by
extensions from the mantle. This was easily procured in
numbers when the " Ensanafu " ant had been hunting, for
its only means of escape was to crawl upwards. Should
it happen to come to the top of a stem it was unable to
descend again and face the ants, so defended itself by emit-
ting bubbles to form a mass of foam completely surround-
ing itself, which the ants could not penetrate, and if they
bit into it they could reach nothing solid. These white
masses of foam, like " Cuckoo spit," were very con-

1 Dolops ranarum and Argulus africanus.
22

322 SUNDRY INSECTS

spicuous in the forest after a raid by the Ensanafu. Speci-
mens sent to England were found by Lieutenant-Colonel
Godwin- Austen, F.R.S., to be a species hitherto un-
described. On Tavu Island, among dead leaves, I once
found a minute snail like a Vertigo, but lost it again.

A very large snail (Achatina) with pointed shell of
brown colour, the mouth tinted with purple, was common
on Damba, and I once disturbed an Enswa-swa which
had apparently been much interested in one that was
laying its eggs in the ground. It is possible that the
eggs were the attraction and not the snail ; they were
a little larger than a pea, but not spherical, with a firm
shell of canary-yeUow colour.

The general name for a snail in Luganda seems to
be '' E'sonkor

r^

\

\

/\

X

\

.- «.

K^ -^

^

^

• — »

•^

-_i>«-

^^-~~

U^a^t^'-

'-"*

^"

z^" """"" °CI^'°°'"" "'" °~'^J~' "'■>■■- "^"''^^^-'-^^"-"

/\

\

\

\

/

A

>

\/

\

^\

\

— '

y

\

y '

\

INDEX

Ahdimia abdimii, 155

Achatina, 97, 322

Acraea, prey of Belongaster, 285

alciope, variations in, 235, 240, 241

egim, 103, 120, 205
as model, 233, 238

encedon, 109, 110, 112, 120

jodutta, as mimic, 233

natalica, courtship of, 224
„ pseudegina, 120

pentapolis, 104

perenna, 205
as mimic, 233-
larvae of, prey of bug, 219

terpsicfiore and Syntomid moth, 222

zetes, 116, 120, 205
parasites of, 220
as model, 233
Acraeinae (see also Acraea, Planema),
103-5, 109, 110, 112-5, 120, 125,
205

emit distasteful fluid, 210

toughness of, 225, 213

courtship of, 224

nuptial flight of, 225

as models and mimics, 233, 242
Acrididae, 136, 314

aposematlc, 209-12, 215
Actophilus afrieamis, 153
/Em, 159
Agamidae, 185

(" Akalerwe ") (see Eydrochelidon), 146
(" Akamunyi ") (see Milvus), 161
(" Akasalu ") (see Galactochrysea), 150
" Akasonzi," 141, 194
(" Akatassa ") (see Lily Trotter), 153
(" Aka%vundo ") (see Bat), 143
Albert, Lake, 68
iAlehornea) (see " Oluzibaziba "), 70
Aletis erici, larvae of, 214

evidence of insularity, 103, 106
Aloe, evidence of insularity, 116
(" Amajansi ") (see Cyrtacanthacris), 314
Amauris, as models, 242, 268, 273

Mendelian relationship in, 268-9
Amauris echeria, 114

evidence of relative inedibility of, 238

niavius, 103, 114, 222

psyttalea, 114
courtship of, 222
Ambatch, 70. 96-8, 108, 111, 117, 121, 122,
147, 158, 176, 194, 303

flowers, 165, 303
Ammophila, 298-300
Amoeba, 11

Amphidesmus analis, 307
AmpuUaria, 154, 184, 321
Anal tufts of Amauris, in courtship, 222
ATMstomus lamelligerus, 97, 109, 146, 154

323

Androtiymus leander, 110
Anhinga rufa, 148

Ant (see also Dorylus, Megaponera, ' Obusa-
jisajl," (Ecophylla, PaUothyreus), 97,
108, 207, 276-84

enemies of, 172, 185, 282
(Antelope), and Trypanosomes (see Bush-
buck, Reedbuck, Situtimga, Water-
buck)
Anthia absent from islands, 307
Ant lion, and Olossina, 52
" Anyamberege," 70, 116
Aph7iaeus orcas, 124, 199
Apodidae, larvae of, 213, 220
Aponomma laeve, 319

Aposeme (see also Colouration, aposematic),
206

implies unpleasant not necessarily dis-
tasteful quality, 207, 213

must be simple, 206, 235

combined with procrypsis, 206-7

displayed when necessary, 206, 210-2

proved to be of value, 212

of larva, used by pupa, 216-7

Lycid, 231-2

of Planema-Pseudacraea association,
246-9, 251-3
Arachnida, 318-9

mimetic, 227, 233
Arctiidae, aposematic, 210-1

hairy larvae of, 213
Ardea melanoeephala, 158

purpurea, 156
Argulus africanus, 321
Arum, 110
Asilidae, 316-7

and Mantidae, 120, 124

prey of, 54, 97, 317

poison injected by, 317
Aslauga purpurascens, 106
Aspidomorpha, insular abundance of, 120-1
Atella, 238
Ateriea galene, 102
Atractocerus brevicornis, 309
Attitude of caterpillars, 200-1, 214

terrifying, 221

importance of, for procrypsis, 196-8
216
for mimicry, 227, 311
Austen, E. E., on name " Tse-tse," 19

Bacilli and Trypanosomes in Glossina, 40,

41
Bagshawe, Dr. a. G., 54
Baker, Dr., 23
Balceniceps rex, 155
Balance of nature, 102, 105, 110, 120, 124,

171, 176. 213-20, 316
Balantidium, 11

324

INDEX

Balearica regulorum, 162

booming noise of, 77
(" Balwa ") (see Ardea 'purpurea), 166
Banda ISIiE, 89
Barbatula leucolaima, 167
Barbel, 194

Barbet, scarce on islands, 167
Bates, H. W., 195, 205, 225, 229-30, 235,

281
Bats, 98, 112, 143-4
Bees, 117, 121, 303-4

prey of Bee-eaters, 172, 235
not intrinsically distasteful, 207
Carpenter, (see Xylocopidae), 108
Bee-eater, 97, 102, 115, 120, 171-3
food of, 172, 173, 218, 235, 237, 282, 285
Drongos & BoUers, interchangeable
function with, 171, 176
Beetle, 305-9
as evidence of insularity, 121, 122
distasteful to monkey eaten by Boiler,

218, 306
mimetic, 231-2, 308-9
wood-boring, 309
Belcher, C. F., 145, 162, 177
Belenois, 120
attacked by birds, 237-8
courtship of, 223
seasonal forms of, 202
Belongaster, 284-6
prey of, 97, 108, 285
eaten by Bee-eater, 172, 285
as model, 285, 310, 311
enemies of, 286
compared with Fossors, 287
Bembex, 95, 286, 288-93
and Glossina, 53, 288-93
enemies of, 291-293
capensis, 292-3
forcipata, 288-92
ugandensis, 293
Bird-droppings resembled by insects, 198,

199, 204, 216
Birds and Glossina, 37, 53

of islands, compared with mainland, 145,
164, 166-7, 171, 173, 175-7, 179
local races on islands, 275
iniectivoious, on " Spider islands," 102,
105, 120, 124
attacks of, on butterflies. 111, 173,

208, 225, 235-8, 267-8
attacks of, on butterflies, selective, 238
species of, interchangeable function of,
171, 176
Bitis arietans, 188

nasicornis, 96, 187-8
Bittacus, 310
Blattidae, 314

hunted by Dorylus, 279-80
(Blind worm) (see Slow Worm)
Boomslang, 191
(Bottle-bird) (see Coucal)
Bracken, 116
Braeonidae, 220
as models, 233, 307, 308-9
Lycoid, 232
Bruce, Sir David, 7, 14, 22, 24, 26, 54
bubbmbe isle, 89, 92
Bubo lacteus, 163
Buffalo and G. morsitans, 35, 54
(Buffalo-gnat) (see Simuliidae)
Buff tip moth, 198
BtJGABA Isle, 93

BUGALLA Isle, 89, 114, 145, 149, 166, 180,
215, 219, 221-2, 283, 308-9
local races on, 139, 174, 175, 275
Breeding places of Glossina on, 57
Pseudacraea eurytus on, 268-9

(Bug) (see Ecmiptcra)
Btjguye Isle, 90
Bulago Isle, 94-5, 182-4

Glossina on, 47

Breeding place of Glossina on, 59

butterflies on, 125
Bul-bul, 105, 117
Bunaea i^haedusa, 201
Bunyama Isle, 89, 93, 143
(Burnet moth) (see Zygaenidae)
Bushbuck and T. gambiense, 27

compared with Situtunga, 133
Busmi Isle, 183
(Bustard) (see Otis)

Butterflies, attacked by birds, 111, 173,
208, 225, 235-8, 267-8

courtship of, 221-5

numbers on difterent islands, 125

mimetic, less shy in absence of insectiv-
orous birds, 102
BuvuMA Isle, size of, 67

mortality from Sleeping Sickness on, 7-8
(" Buvumira ") (see Xylocopidae), 303
IJuzzard and Chamoeleon, 163
BwiGGi Isle, 93
Byblia, 119

courtship of, 223-4
Bycanistes subcvlindricus, 168

an aposematic bird, 302

cry of, 77, 81

Cacergates leucosticta, 310

preys on Glossina, 53, 54
Camvephaga nigra, attacking Butterflies,

237
Cane, 70, 96, 98, 110, 111, 118
Canoe, description of, and names of parts,

84-7
Caprimulgus, cry of, 81, 173
(Capsidae, mimetic) (see Eolopeltis, Xene-

tomorpha)
(Carabidae) (see Anthia)
Carpenter, Dr. G. D. H., on inheritance

of small variations, 241
(Cassididae) (see Aspidomorpha)
Castalius isis, 119
Castbllani, Dr. A., 23
Caterpillar, 199
insular abundance of, 100, 101, 103, 104,

106, 116
enemies of, 101, 172, 220, 288, 316
instincts of, 200

procryptic, resembling bird-dropping,
199, 216
rarely becomes aposematic adult, 214
with special defence, 206
aposematic, becomes aposematic adult,
214, 239
if hairy, may become procryptic adult,
213
haiiy, Irritant hairs of, 213-4

enemies of, 166, 218, 294, 316
very large, but concealed, 201
Catopsilia, 120
Caves, 98, 108

not good breeding places for Glossina, 57
Centipede, large, absent from islands, 319,

321
Centropidae, 102, 110, 116, 120, 165-6

(food of, see Ceuthmochares)
Cerambycidae, 121, 307-9

mimetic, 232, 233, 239, 307-9
(Cercopithecus) (see Monkey)
Ceroplesis signata, 121
Ceryle maxima, 169

rudis, 169
Cetoniidae, 121, 232
Ceuthmochares ceneus, 166

INDEX

825

Chagas, Dr., 14
Chalcididae, 51, 52, 220, 313
Champeleon and buzzard, 163
CMraxes, 97, 103, 106, 114
directive markings, 208
etheocles, transitional forms, 241
Chenalopex cegyptiacus, 81, 96, 146, 149
nest of, 150
and mongoose, 142
Chironomidae, 79, 317-8
food of (Ecophylla, 283
Chrysididue, prey of Bee-eater, 172
Chrysococcyx, 167
Chrysomelidae, insular abundance of, 122

Lycoid, 232
Chrysophyllum, 98, 215, 259, 263
Chrysops brucei, 315
enemies of, 97, 293
Chrysopsyche varia, cocoon of, 217

caterpillars of, 200, 216
Cicada, insular abundance of, 118
Cicindelidae, 307
Ciconia nigra, 155
Cilix glaucata, 198
Cirpkis, 196
Clegg, 18, 20

Climate of islands and mainland compared,
44, 72
storms, 73-5

effect upon G. palpalis, 42-6
Clouds, 75, 101
Coccidae, 116

(Cockroach) (see Blattidae)
Coeliaxis carinata, 304
Coincidence, and detailed mimicry, 234-5,

310
Colias, nuptial flight of, 224
Coliidae, absent from islands, 167
Colouration, classified, 195
held to be always concealing, 204, 238
and habit associated variables, 239
anticryptic, 203-4
apatetic, 195, 234

aposematic (see also Aposeme), 204-6,
209-16
and noisy flight, 168
versus Thayerism, 204-5
and bold demeanour, 209-12
proof of value of, 210-2
and resistance to injury, 212-3, 226
effect increased by massing, 215
carried through all stages, 214
cryptic, 195
epigamic, 195
procryptic, 196-202, 312
combined with special aposeme, 206-7
general, 196, 199
special, 196, 197-9, 216

analogous to mimicry, 234
effect increased by massing, 216
pseudaposematic, 229
pseudosematic, 195
synaposematic, 230-1
Concealment of all creatiu:es, theory of
(see Longley, Thayer)
by breaking up large area, 200-1
(Copper-smith) (see Barbatula)
Copridae, 305-6

eaten by Roller, 218
Conorhinus, 18
COOK, DR. A., 7

Cormorant, 96, 97, 117, 121, 122, 146-8
nesting places, 147, 158
robbed by gull, 146
Cossypha, 178-9
(Coucal) (see Centropidae)
Courtship of butterflies, 221-5
(Crane, crowned) (see Balearica)

Crayfish, not met with in lake, 820
Creepers, 72

make shelter for Glossina pupae, 58, 64
Crickets, 80
Crithidia, 12
Crocisa meripes, 303-4
Crocodile, 180-2

abundance on different islands, 94, 96,
99, 104, 111-3, 115, 118-9, 122,
124

relations with Varanus, 183-5

as food for O. palpalis, 38, 39, 181

Trypanosome of, 40

basking place of, 180

nest of, 90, 98, 115, 122, 181, 182
C'rotalaria striata, 108
Crow, 114
Crustacean, 320-1

Ostracod, found in gut of Glossina, 42
Cuckoo, 102, 166, 167

feeds on hairy caterpillars, 218

(bronze) (see Chrysococcyx)
(Cuckoo shrike) (see Campephaga)
Ciiculus solitarius, 166, 167
CuMMiNG, Gordon, 20
Cycle, in life history of Protozoa, 17, 25, 26
Cynthia cardui, 106
Cyrtacanthacris, 105, 170, 314

Damarius splendidulus, 219
Damba Isle, 8, 68, 114, 119

notes on fauna of, 131, 142-3, 145-6,
158, 165, 176, 188, 197, 200, 227
279, 280, 283, 315

Glossina on, 51, 58

Ps. eurytus on, 256-9
Danaida chrysippms, 114, 120

crossing water, 100

as model, 242-3

flight of, 243
Danaida plerippus, flight of, 243
Danainae (see also Amauris, Danaida)
103, 114, 120, 125

as models, 106, 233, 242

and birds, 225, 235, 238

toughness of, 213
Daniels and Newham, on acute Trypano-
somiasis, 10
Darter, 96, 97, 146, 147, 148, 158
Darwin, 18, 224, 225
Darwinism, and the Mutation theory, 195,

198, 241, 262, 268-9, 270-2
Depopulation of islands, a measure against
Sleeping Sickness, 7, 8, 28

effect on fauna and flora, 29, 71, 139, 184
Dermaptera, 314-5, 321
Dicruridae, 237

absent from islands, 176
Dictyophora laticincta, 209-10
Diloba caeruleocephala, caterpillar of, 239
Diploptera, 286

Diptera (see also Chironomidae, Glossina,
Simuliidae, Tabanidae), 315-8

parasitic (see also Tdia, Stylogaster,
Tachinidae), 294, 302
Dirphya, 233, 308-9

Disease, a distiu'bance of equilibrium, 16
Distasteful species, require to advertise, 206
DiXEY, Dr. F. a., on chance resemblances,
234

on Mullerian mimics of second degree, 248
Dolops ranarum, 321
Domestic animals and Trypanosomes, 16
Dorylus, 97, 186, 276-81

male comes to light, 80
Dourine, 14, 18
Dove, 159
Dragon fly, 309-10

326

INDEX

Dragon fly, prey of Bee-eater, 172

bionomlc relations with Olossina, 53-4,
310
(Dronso) (see Dicruridae)
Duck, 149

Duke, De. H. L., 15, 27, 41, 185
DUTTON, Dr , 6, 23
DWAJi Isle, 319
DWANGA Mkuru Isle, 104-S

butterflies on, 125
DwANGA Mto Isle, 106-6

butterflies on, 125
DWASENDWE Isle, 106-7

butterflies on, 125
Dyavodemu Jsle, 119-121

butterflies on, 125
Dysentery, 11
Bytisridae, 307

Eagle (Bateleur) (see He^otarsus)

(Crested) (see Spizaetus)

Fish (see also Haliaetus), 160-1
(Ear^vig) (see Termaptera)
East Africa, British, Ps. euri/tus in,
245-6, 249, 252-4, 263

Ex-German, 149, 152, 155, 173, 176,
238, 282, 302-3, 308

Portuguese, 9, 283
Eciton, 281

(" Egongolo ") (see Millipede), 320
(" E'gonya ") (see Crocodile), 180
(Egret) (see Egretta)

Egretta, 81, 96, 97, 117, 120, 124, 146, 158
(" E'jansi ") (see Gyrtacanthacris), 105
" Ekinsambwe," 70, 117, 122
(" Ekinyira ") (see Flying Fox)
(" Ekirikiti ") (see Erythrina)
(" Eklsagazi") (see Elephant grass)!
(" Eklsinja ") (see Barbel)
" E'konkomi," 185
Elanus cceruleus, 162
Elephant grass, 105, 117
Elgon, Mt., Ps. eurytus on, 250
Elminia longicauda, 175
Eltringham, Dr. H., 222, 242, 244
(" E'male ") (see Mud-fish), 79, 193
" E'mamba ") see Lung-flsh), 193
(" Embata ") (see Goose), 149
*' Embegede," 194

(" Empavana ") see Ibis, hagedash), 153
(" Empungu ") (see Eagle, fish), 160
(" Empunya ") (see Plover), 152
(" Endegeya ") (see Weaver bird), 176
(" Enfudu ") (see Tortoise), 186
" EngabI," 138, 183
(" Engadala ") (see Cormorant), 147
(" Engali ") (see Crowned Crane), 153
(" Enga-nga ") (see Hornbill), 168
(" Enjaba ") (see Crab), 320
(" Enjobe ") (see Situtunga), 137
(" Enkazalugya ") (see Passer), 177
" Enkeje," 194

(" Enkobyokobyo ") (see Gull), 146
(" Enkolokoto ") (see Megaponera), 282
(" Enkonamasonko ") (see Stork, open-

bUled), 154
Enkusa Isle, 108
(" Ensanafu ") (see Dorylus), 276
" Ensenene," 312
'* Ensoga," 194

(" Ensogwe ") (see Cormorant), 147
(" Enswa-swa ") (see Varanus), 182
Entamoeba, 11

Entebbe, birds at, 164, 167, 171, 175,
177, 179

Ps. eurytus at, 250-1, 258-9, 262. 266
(" E'numba ") (sec Belonogaster), 284
(" E'nunda ") (see Stork), 156

(" Envubu ") (see Hippopotamus), 126
Enyaliopsis, 211

emits irritant fluid, 313
(" Enyange ") (see Egretta), 158
(" Enyonza ") (see Cossyplia), 178
Ephemeridae, 79, 80
Ephippigeridae, 313
Ephippiorhynchus senegalensis, 155
Epigamic colouration, 195
Epinephele janira, nuptial flight of, 224
Epitoxis albicincta and Acraea, 222-3
Erlangea tomentosa, 107, 113, 117

attractive to butterflies, 205
(" Eryato ") (see Canoe), 84
Erythrina tomentosa, 11(), 112, 206
(" E'salambwa ") (see Bitis), 188
(" E'sami ") (see Chironomidae), 317
" E'sonko," 322
(" E'timba ") (see Python), 188
" E'tungulu," 71, 96, 118
Eucoma atricosta, evidence of insularity, 116
Euphorbiaceae and Protozoa, 13
Eupterotidae, caterpillars of, 200, 316

evidence of insularity, l()3-4
Euralia, 97, 102, 104, 114, 124

dinar cha, 103

dubia, 103, 105, 106
Mendellan relationship in, 268-9, 272.

Factor, Mendelian, 270, 272

Fantham and Porter, on Eerpetomotms, 16

Ferns, 104-7, 115-7

Fig trees, 71, 113, 115, 117, 159

Finches (see Fringillidae)

Bishop and Widow, absent from islands,
176-7
Fkeflies, 309
Fire, spontaneous, 76-7
Fish (see also E'male, E'mamba), 193-4

as weather prophets, 78-9
Fish lice, 320-1
FlSKE, W. F., 51, 71, 101, 108, 110, 115

121, 123, 124, 127, 140
Flagella, 11

Flagellates, relation with insects, 16
Flamingoes, absent from Lake, 156
Flight, aposematic, 168, 301-2

of model and mimic, 243, 252, 308
Flowers, 70-2
Fluid, emitted by distasteful insects, 210,

211, 231, 313
Fly beach, 69

Flycatcher (see also Tchitrea), 96, 97, 102,
104, 109, 112, 120, 173-6

nightingale-like, 107-8, 113, 115, 117,
121-3
Flying fox, 143-4
FORDE, Dr., 23
Forest, abrupt edge of, 71
Formicidae, 233
Fossor, 53, 95, 286-303

enemies of (see also Idia), 302

difterent stages in methods, 286-7

Lycoid, 232, 293
Fountaine, Miss, on Pseudacraea eurytus,

259
Fowls, and aposematic grasshopper, 212
Francolin, forest, 159, 160

call of, 81
Fraser, a. D., 54
Fringillidae, 177
Frogs, noises of, 80

as weather prophets, 78

Tree, 190

and snakes, 190-3
Fryer, J. C. F., 269
(" Fulungu ") (see Musophaga), 165
FUMVE Isle, 89, 92

INDEX

327

Gaboon, Pseudaeraea eurvtus from, 248
Oalactochrysea emini, 146, 150
(Gamboge tree) (see Earonga)
Game, Glossina and Trypanosomes (see

also Situtunea), 15, 25, 26, 35
Game birds, scarce on islands, 159-60
Oamia bucholzi, 106, 110
Gecko, absent from islands, 186
Genet. 142

Geographical distribution, of models and
mimics, 235, 241, 247, 250-1, 264,
262
Oeometridae, caterpillars of, 214, 239

moths, 199, 233
Olossina (see also Tse-Tse), 21

geographical distribution of genus, 19
number of species, 19
life history, 22, 50, 54
brevipalpis, comes to light, 35
moraitans, early accounts of, 19-21
and buffalo, 35, 54
breeding places, 65
enemies of, 51
and Trypanosomes, 9, 23
palpalii, 21

distribution of, 9, 24
numbers of, 30

and crocodiles, 48-9
proportions of sexes, 31-2, 45, 47-8
length of Ufe, 42-4
effects of climate upon, 42-6
dependant upon shade, 21, 32-4, 45
does not cross open water unless

carried, 34, 48
birth of larva of, 50
emergence from puparium, 22, 52
time and method of feeding, 34-5, 45
proved to carry Sleeping Sickness, 2

4,7
no inheritance of Trypanosomes, 25
duration of infectivity of one fly, 26
infective in absence of mankind, 42
breeding places of, first discovered, 54
characteristics of, 55-60
ideal, 58, 63
" Loci," 57

searched for by female, 56, 63
destmction of, not possible by only
clearing breeding grounds, 60
by artificial breeding grounds, 62-5
enemies, of larva, 49, 50
of pupa, 50-2
of fly, 52-4
food of, 35-42
mammalian, 36
non-mammalian, 36-9
vegetable, 41
water, 42

endurance of starvation, 42
larvae and papae ot, requirements of,
55-60, 62
resistance to adverse conditions, 46-7
numbers on different islands, 96-8,

100, 102-11, 113-8, 120-3
numbers of pupae on different
islands, 57, 58, 111, 119, 122
tachirwides, the only species outside
Africa, 19
(Goatsucker) (see Nightjar)
Gold markings on insects, 199
(Goose, Egyptian) (see Chenalopex)
(Pigmy) (see Nettopus)
(Spur wing) (see Plectropterus)
Orasshopper, as evidence of insularity, 105,
110
aposematic, 209-12, 215
Gregariousness, increases conspicuousness,
215 '

Gregariousness, rare In procryptic species,

215-6
GuU, 78, 81, 117, 145, 146
Oyrindae, 307

Haematopota, 97, 315

Hairs, of caterpillars, 206, 213

Haley an chelicuti, 171

senegalensi-s (see also Kingfisher), 170
Baliaetus vocifer, 160-1

cry of, 78

and prey, 193

and Goliath heron, 156-7
(Hammer head) (see Scopiis)
Haronqa madagascariensis, attractive to

insects, 232, 237, 244, 307
Heliconitme, 243
Eelotarsus ecaudatus, 164
Hemiptera (see also Capsidae)

carrier of disease, 18

anticryptic, 204

prey of, 204, 219

aposematic, 215

mimetic, 228, 231, 309

disliked by Dorylus, 280
Heredity of small variations, 271-2

in Papilio, 241, 269
Herodias alba, 146, 158
Herons, 96, 107, 109, 122, 166-8

Buff backed, 146, 158, 159

Goliath, 96, 99, 109, 117, 146, 166-8
and Fish-eagle, 156-7

(Great White) (see Herodias)
Eerpetomonas, 12, 13

of water scorpion pathogenic to mouse, 16
Eesperidat (see also Andronymus, Gamin,
Rhopalocampta), 106, 112, 114, 118,
120, 125

prey of Bembex, 288
Hippopotamus, 95, 110, 112, 118, 126-31,

grazing ground, 98, 130

noises of, 80, 89, 127

pathways, 118, 126, 128

sleeping places, 124, 126, 127

wallows, 128
(Hornbill) (see Bycanistes, Lophoceros)
Humidity, on islands and mainland, 44

effects upon Glossina, 44, 46
Eydroclielidon leucoptera, 78, 146
Hymenoptera, prey of Bee-eaters, 172, 218,
235, 282, 285

not always intrinsically distasteful, 207

parasitic, 220, 313

syn-aposematic, 231-2

as models, 233, 308-311
Hypertely, 196-7
Eypolimnas misippus, 106
Eypsidae, 214

Ibis aethiopica, 154
hagedash, 97, 146, 153-4

cry of, 81, 130
sacred, 96, 99, 146, 164
Ichneumon flies, 49, 61, 101, 287-8, 297
Idia, 291-2
Infusoria, 11
Insects, as weather prophets, 79-80

parasitic, 49, 51, 52, 101, 219, 220, 297-8,

288, 313, 316
predaceous, prey of, 204, 219, 310
protected, 217

enemies of, 218-220, 235, 285, 315
immunity only relative, 217
Instinct, limitations of, 200
in accordance with colouration, 196, 198,
209, 216, 227, 228, 308
Insularity, faunal, 100-6, 109, 110, 116-9,
120-2, 124, 174-5, 275

328

INDEX

Insularity, floral, 105, 107, 108, 110, 113,

116-7
Ipomaea, 64, 70, 107, 117, 121
Irrisorinae, absent from islands, 173
ISENTWA ISLE, HI, 128, 183

butterflies on, 125
Islands, 68, 69, 72, 89, 90, 108, 114

largest, 67

of Kome and Kerenge groups compared,
108-9

derivation of fauna and flora, 109, 119

smallest on which Olossina breeds, 122

with monkeys, 131

with Situtunga, 101, 117, 139-40

faunal peculiarities (see also Insularity),
145, 159, 164-7, 171, 173-7, 179,
237, 304, 319, 321
Ispidvm, 169

of evil omen, 170
Ithomiinae, 243

Jennings, Prof., 272

JiNJA, 126, 161, 162, 164, 167, 171, 175,

177, 184, 196, 218, 315
Jordan, Dr. Karl, on Ps. eurytus, 244-5,

255

Kagera, E,., 67-8, 137

barrier to Megaponera, 282
Kakindtj, Ps. eurytus at, 274
" Kala-azar," 13, 16
Kallima 221

(" Kasiniagizi ") (see Kingfisher), 169
Kawaga Islet, 122
KERENGE Isle, 161, 170, 109-10, 314

butterflies on, 125

Sceliphron on, 297-8
KIBIBI ISLE, 119, 179, 318

leopard on, 141
" Kibo," 70, 105

KiKTJTU Escarpment, P. dardanus on, 271
(" Kimbala ") (see Heron, Goliath), 156
KIMMI ISLE, 97-9, 182-3, 188, 194, 201

Glossina on, 32, 47-9, 98
breeding places for, 58, 60

Ps. eurytus on, 273-5

butterflies on, 125
Kingfisher, 110, 169-71

cry of, 78
(" Kirimululu ") (see Chrysococeyx), 167
KiRUGU Isle, 89
KisiGALLA Isle, 92, 103

butterflies on, 125
Kites, 96, 161-3
KiuwA Isle, 68, 108, 111-2, 116, 179

Glossina on, 119

butterflies on, 113, 125
KiziMA Isle, 95-6, 123, 154, 162, 175,
179, 182

butterflies on, 125
Kleine, Dr., 25
Kolopeltis bergrothi, 309
KoME Isle, notes on fauna, 131, 140, 142,
152, 275, 310

Ps. eurytus on, 266-8, 274-5
(" Kunguvu ") (see Tchitrea)
KuYE Isle, 90
Kyagwe coast, 167, 176, 245, 251

Lacertidae, 185

Lachnocnema bibulus, 109, 111
Lamborn, Dr. W. a., 66, 273
Lamiidae, 121, 233
Laniarius erythrogaster, 175
Laphygma frugiperda, evidence of insu-
larity, 101
Larus, 146
Lasiocampidae, 199, 200, 206, 213, 216

Lasiocampidae, evidence of insularity, 103

(Leaf butterfly) (see Eallima)

Leech, 18

Leishmania, 13

Leopard, 141

Leptaletis forbesi, 233

(Leptomonas) (see Eerpetomonas)

Libythea, 115, 125

(Lily Trotter) (see Aetophilu^), 153

Limenitis, 243

(Limnotragus) (see Tragelaphus)

Lipteniime, 107, 233

Lithosiidae, 232

Livingstone, 20, 21, 137

Lizard (see also Gecko), 108, 120-1, 186, 190

fed upon by G. palpalis, 37

prey of hawk, 163
Locustidae (see also Enyaliopsis), 312-4

enemies of, 163, 294-5
(Longicorn) (see Cerambycidae)
LONGLEY, Dr. W. H., theory of coloura-
tion, 238-9, 301
Lophoceros melanoleucus, 169
LuKALU Isle, 110-1

butterflies on, 125
LuKiusA Isle, 91-2
LULA Isle, 104

faunal characteristics, 316

butterflies on, 125
Lung fish, 88, 160, 193

and Olossina, 39
(" Lunyonkante ") (see Otis), 153
Lycaenidae (see also Castalius, Lachnoc-
nema, Tarucus), 103, 104, 106-8,
110-4, 117, 120, 124-5, 199, 204

attacked by birds. 111, 208, 225, 238

directive markings in, 208

larva stored by Sphegid, 294

mimetic, 103, 233

nuptial flight of, 225
Lycidae, and mimics, 231-2, 239, 307
Lymantridae, caterpillars of, 116, 213, 294
{Lymexylonidae) (see Atractocerus)

Madagascar, Pseudacraea on, 242

(" Ma'ga ") (see Elanus), 162

" Makindu," 70

(" Makwanzi ") (see Osprey), 161

Malacodermidae, 231-2

Malaria, 11, 17

" Mai de Caderas," 14

Mantidae, 311-2

on spider islands, 102, 105, 109, 110, 124

and Asilidae, interchangeable functions,
120, 124

enemies of, 172, 203

anticryptic, 203, 311
Mantispa, 285, 310, 311
Marida Isle, 115, 182

butterflies on, 125
Markings, directive, 207-9

terrifying, 220-1
Marshall, Dr. C, 54
Marshall, Dr. G. A. K,, on Hornbill, 168

frog and snake, 192

seasonal variations, 201, 261

relative edibiUty, 217

Lycoid insects, 231-2

birds and butterflies, 236

Mantispa, 311
Masovwi Isle, 122
Mastigophora, 11, 12
Maungwb Isle, 107-8, 113
Megaponera foetens, absent from islands,

282-3
Meinertzhagen, Maj. K.., on island

Situtunga, 92, 139
Melinda petiverana, 106, 233

J

INDEX

329

Melipona, absent from islands, 304
Melittophagus meridionalis (see also Bee-
eater), 171-2, 237
Mendelian relationships, Punnett on, 268-9
in Ps. eurytus, Poulton on, 263
and intermediate forms, 269-270
Merops superciliosus (see also Bee-eater),
171-2
Bionomic relations wjth Glossiim, 310
Mesoplatys ochroptera, 122
Meteorology, 76, 91
Methoca, 303
Mfumbiro Mts., 68
Millipede, 320
Miltogramma, 302
Milvus aegyptius, 162
Mimacraea poultoni, 103
Mimicry, 195, 225, 230
examples of (see also Ps. eurytus), 227-8,

231-3, 238, 307-11
and Natural Selection, 226, 234-5, 266,

268
analogous to special procrypsis, 234
depends upon experimental tasting, 230
modifications caused by, 228
imaginary mammalian case, 229
Mimics, 825
polymorphic, 242
occur with models, 228, 250, 251
numerical relation to models, 226, 229,

265
variations in absence of model, 256-9,

262, 273
pattern, not species, depends upon

model, 262-3, 273-5
may retain ancestral appearance m

parts not seen, 226-227
behaviour of, 102, 226, 243, 308
disregarded by wagtail, 238
origin by mutation or fluctuation, 241,
270, 271
Models, 225
numerical proportions to mimic, 226,

266, 267
behaviour and habits of 226, 238 239,
243
MoUuscs, 97, 103, 115, 184 .321-2

attacked by Dorylus, 279, 280, 321-2
Mombasa, Ps. eurytus at, 252-4
Mongoose, 168

tricked by goose, 142
(Monitor) (see Varanus)
Monkey, 114, 117, 131-7
on Nkosi Isle, 92
language, 132-3
intelligence, 135
imitativeness, 135
and Tortoise, 135

and grasshoppers, mantis, 136, 203
and alarming insects, 136, 221
Mononyx grandicollis, 204
Moorhen, 111, 117, 122, 153
MOEGAN, T. H., 272
Mosquitoes, 97, 119, 315
Moth, 196-9, 207, 213-4, 220, 222,

232-3, 239
Mtiomu Islet, 121-2
Mud-fish, 79, 193

(" Mujolo ") (see Bee-eater, Kingfisher), 169
MuLLER, Fritz, 195
on Common Warning Colours, 229-30
on Scent Scales, 222
Muscicapidae (see also Flycatcher), 173-5
Musophaga rossae, 119, 165
(Mutation) (see Variation)
Mutation theory, 239-40
difficulties of, 198, 240
and Darwinism, 198, 262, 268-72

Mutation, and intermediates, 269-71
mutations should not retrogress, 241
does not account for absence of pro-
cryptic mimics, 240

Mutilla, 51, 287, 303

Myriapoda, 319-20

Myrmarachne foenissex, 233, 227-8

(" Nabubi ") (see spider), 318
" Nagana," 9, 14, 18, 20, 21

compared with Sleeping Sickness, 24, 25
(" Nakibengeyi ") (see Fire), 77
•■ Namugoya," 190
Natal, Ps. eurytus in 245-6, 249, 254,

259, 262-3
Neave, Dr. S. a., 247-52
(Nect/iriniidae) (see Sun-bird)
Nepa cinerea, protozoal parasite of, patho-
genic to mouse, 16
Nephila, numbers of, on islands, 102, 103,

105, 107, 109-23, 318
Neptidopsis, 114
Neptis, 105-6, 114
Nest, Ant's, 227, 276, 282

Birds', 107, 150, 158, 161, 174

Crocodile's, 90, 98, 115, 122, 181-2

Fish's, 194

Fossor's, 288-90, 294-7

Varanus', 98, 118, 183

Wasp's, 284-6
Nettopus auritus, 150

Neuro-muscular system, modifications of,
in deceptive resemblance, 196-8,
226-8
Neuroptera, 309-11
Ngamba Isle, 99-100, 143, 180-1, 188

Glossina on, 47, 100

butterflies on, 125

Ps. eurytus on, 273-5

Situtunga on, 140
(Nightjar) (see Caprimulgus)
Nile, B,., 67-8
Nkosi Isle, 90-92, 131
Noctuidae, 101, 207
(Notodontidae) (see Scalmicauda)
Nsadzi Isle, 100-1, 204, 288, 307

Olossina on, 44
breeding places of, 56

Situtunga on, 28-9
Nuptial flight of butterflies, 224-5
Nyasaland, Trypanosomiasis in, 9

Ps. eurytus in, 246, 249, 254
Nymphalinae (see also Pseudacraea), 103-6,
110-2, 114, 116, 120, 125, 242

as mimics, 233

courtship and nuptial flight of, 223-5

attacked by birds, 225

(" Obuduli ") (see Shrimp), 320

" Obusaji-saji," 100, 283-4

(" Obutwa-twa ") (see Erlangea), 107

Odymrus, 286

(Ecophylla smaragdina, 100, 283

mimics of, 233, 227-8
(Oedictiemus) (see Stone Curlew)
" Oluzibaziba," 58, 70, 95, 117, 121, 122

food of Situtunga, 99, 138
(" Omugalati ") (see Chrysophyllum), 98
" Omukene," 194
" Omukoba," 99
" Omukwakula," 105
" Omululu," 99
" Omusali," 95

(" Omusenene ") (see Podocarpus), 115
(" Omusoke ") (see Waterspout), 75
(" Omutunwa ") (see Stone Curlew), 150
" Omuvuvumye," 70, 122
" Omuziru," 99

330

INDEX

Onomatopoeia, 152-3, 166-168, 173, 303

Ophideres, 207

Orthoptera, 311-4

OsBORN, Prof. H. F., 19

Osprey, 161

Otis melaiiogaster, 153

Otter, 140-1

Owl, 81, 163-4

Pachnoda sinuata, 121

pactolicus, 239

Palm, 70, 105

PaUothyreus tarsatus, 50, 172, 281-2

PapUionidae, 98, 106, 125

directive markings in, 208

as mimics, 233, 238
Papilio dardanus, 114
form dionysus, 273
hippocoon, 222, 273
trimeni, 271
variation in absence of model, 273
transitional variations, 269, 271
heredity of small variations, 241
(T deceived by mimicry, 221-2
? ¥ are pseudaposematic, 240
nuptial flight, 225

caterpillars resemble bird-droppings
199

demodocus, 120

leonidas, 106, 114, 233

polytes, 269

ridleyanus, 238

Uganda, 114
Papyrus, 69, 117, 118
(Paradise flycatcher) (see Tchitrea)
Park land, 68, 119
Parrot, 81, 92, 102, 164
Passer griseus, 177
Passeriformes, 173
Pasteur, 21, 220
Pelican, 149

Pentatomodiae (see also Damarius), 215
Pentila, 104

Phalacrocorax africanus (see Cormorant)
147

lucidus (see Cormorant), 146
Phasmidae, 312
Phrosyne brevicornis, 307
Phijmateus viridipes, 212
Pierinae, 112, 114, 120, 125, 204, 208, 213

attacked by birds, 225, 237-8

courtship and nuptial flight of, 223, 224-5

dry season forms, 202

mimetic, 226-7
Pigeon, 103, 159
Pinacopieryx, 202, 223
Plaintain eater, 103, 110, 119, 165
Planema, 98, 206, 235

boldness of demeanour, 209, 233-9

resistance to poison, 213

as models, 233, 242, 245-9, 250-4, 256

variation in numbers on Kome, 266-7

on sundry islands, 274-5

adrasta, 253-4

aganice, 245-7, 249, 253, 254

aganice motUana, 95, 205, 253-4, 274

alcinoe, 95, 114, 205, 274

epaea, 245, 254

epaea paragea, 252. 254, 257, 274

excisa, 248

macarista, 97, 114, 205, 247-8, 250, 254,
256-7, 274-5

macarista pseudeuryta, 247, 254, 262

poggei, 105, 114, 250, 254, 256-7, 274

tellus, 233, 238-9, 251, 254, 257, 274
Platysteira jacksoni, 175, 275
Plectropterus gambensis, 150
Plover, 81, 150-2

Plusia, 199
Podocarpus, 115
Pomatorrhynchus, 175-6
Pompilidae, 232, 286-7, 300-2
aposematic flight of, 301-2
PONDOLAND, Ps. eurytus in, 253-4, 245-7
(Porter) (see Fantham and Porter)
POULTON, Prof. E. B., classiflcation of
colouring, 195-6, 206, 230, 234
on limitations of Natural Selection, 197,

226-7
on seasonal variation, 202
on relative edibility, 217
on enemies of protected insects, 219
on terrifying marks, 220
on scent of butterflies, 223
on same effect produced by different

means, 231
on Inheritance of small variations, 241
on flight of Pseudacraea and Limenitis,

243
on Mullerian mimicry, 243, 248
on forms of Ps. eurytus, 249-250
on relaxation of selection in absence of

model, 256-9, 262-3, 266, 273
on Mutation theory versus Darwinism,

262, 271
on DrosopMla, 212.
on " painting out," 308
(Pratincole) (see Galactochrysea)
Precis, seasonal changes of, 201-2
courtship of, 223-4
archesia, 111, 120
clelia, 111, 120
sesamus, 201
Problepsis aegretta, 199
Protomo7iadi)ia, 12
Protopterus, 39, 88, 160, 193
Protozoa, classiflcation of, 11

causing disease, 11
Pseudacraea, allied to Limemtis, 243
distastefulness of, 243-4, 263
distribution of, 242
flight of, 243
habits of, 244
boisduvalH, 114, 233
deceptor, 242

eurytus, distribution of, 235, 241, 245
genital armature of, 255
life history of, 259-61
bred families of, 264
polymorphism of, 244-5, 255, 259,

261-2
transitional forms of 241, 245, 247-8,

252, 254-9, 260-1, 267
mimicry deceptive, 242-3, 246, 252,

255, 263
never mimetic of a cryptic species, 240
partially protected, 262-3
aposeme of, 246-9, 251-3
Mendelian relationships, 263
lorm eurytus, 244, 245, 247, 251, 253,
254-5, 262
the ancestral form, 257
form hobleyi, 103, 106, 114, 247, 249,
250-2, 254-7, 262, 274
a stable form, 264
a secondary mimic of kuenowi, 248
flirting with terra, 259
lorm imitator, 242, 245, 246, 247, 249,
254-5
a fixed form, 259, 262-3
life history of, 259
form impleta. 252
form obsoura, 104, 252, 254-7, 263, 274 -
an unstable form, 252, 264
inefficient mimic, 252
flirting with terra, 259

INDEX

331

Pseudacraea 'eurytus, form poggeoides, 250,
252, 254-5, 257. 262
form rogersi, 249, 252, 263, 254-5, 263
form terra, 106, 245, 249, 251, 252-7,
263, 274-5
an unstable form, 264
perfection of mimicry, 251
flirting with hobleyi, 259
beliavlour of, 239
form tirikensis, 103, 114, 245, 247, 248-9,
250, 254, 257, 262, 274
a stable form, 264
kuenowi, 245, 248
lucretia, 104, 106, 114

flight like that of Neptis, 105
poggei, flight of, 243
simulator, 242
Pseudocreobotra ocellata, 203, 311
Psychidae, evidence of insularity, 100, 106
(Puff-adder) (see Bitis)
Ptjnnett, Prof. R. C, on " Mimicry in
Butterflies," 231
on birds and butterflies, 235
on Mutation theory versus Darwinism,

268-9
on Mendelian heredity and intermediates,
270-1
Python, 116, 188-9
source of food for G. palpalis, 39

Quail, 160

(Race, insular) (see Islands, faunal pecu-
liarities)
Rat, abundant on certain isles, 95, 112, 116

Trypanosome of, 13, 18
Rails, 153

Reedbuck and T. gambiense, 27
Resuvu Islet, 97
Rhodesia, Trypanosomiasis in, 9

Ps. eurytus scarce in, 245
Rhodogastria leucoptera, 210-1
Rhopalocampta chalybe, 103, 106, 114

forestan, 106
RiDiEY, H. N., 283
Robertson, Miss, 26, 61
(Robin chat) (see Cossypha)
Roller, functions interchangeable with
Bee-eaters, 171

absent on islands, 167

eat Copridae, 218, 306
RUMFUA Isle, 123-4, 183

butterflies on, 125

(Safari ant) (see Dorylusy

Salamis, 103-6, 114

Saliunca egeria, 232

Salivary glands, of Glossitia, Trypanosomas

in, 17, 26
Salticidae, 228, 233
Salvia, 99, 105, 112, 119
(Sandpiper) (see Tringoides)
(Sand wasp) (see Fossor), 286
Sanga Isle, 101-3, 164, 214

Glossina absent from, 34, 102

butterflies on, 125
Sarcodiua, 11
Saturnidae, caterpillars of, 199, 201, 213

terrifying markings of, 220
Satyrinae, 103, 125, 213

attacked by birds, 225

directive markings of, 208

nuptial flight of, 225
Scalmicauda niveiplaqa, 197
Sceliphron spirifex, 295-8
Scent, produced by male butterflies, 222-3
Schizorkis, 165
Schizotrypanum cruzi, 14

Scoliidae, 121, 231, 287, 303

evidence of insularity, 103
Scopus umbretla, 155-6
Scorpion, scarce on islands, 321
Sbge Isle, 116

butterflies on, 125
Segregation, 270, 272
(" Sekanyolya ") (see Ardea melanocephala),

158
(" Sekoko ") (see Cuculus solitarius), 166
Selection, Natural, 196, 241, 262

produces no more than is necessary,

197, 202, 226-7
produces same effect by different

means, 231
produces more detailed likeness than

chance can, 234-5
originates nothing, 231
and the Mutation theory, 262, 268-9
of mimic relaxed in absence of model,
256-9, 262, 266, 273

Sexual, 209, 221-5
Selous, F. C, 35, 236
Sematic colouration, 195
(" Semirindi ") (see Cormorant), 147
(" Semukutu ") (see Enyaliopsis), 313
" Semutundu," 194
Shelford, R. W. C, on flying snake, 189

on mimetic spider, 227

on Lycoid insects, 232
Shrike, 175-6
Shrimp, 320

(" Siafu ") (see Borylus)
SiDGWICK, A., 199
SUuridae, 141, 193-4
Silver, markings, 199, 319
Simuliid-ae, 80, 151
Simulium damnosum, 315
SiNDiRO Isle, 118, 183
Situtunga, 88, 91-2, 101, 117, 137-40

and Trypanosomes, 15, 27-9, 41

and G. palpalis, 28-9, 36, 61

Meinertzhagen on, 139

Speke on, 137-8

swimming powers, 61

Food of, 99
(Skipper) (see Hesperidae)
Sleeping Sickness, confined to tropical
Africa, 1

symptoms of, 1-2

closely allied to animal diseases, 2, 24-5

earliest account of, 2-5

enlarged glands in, 6

first associated with Trypanosome, 6, 23

spread of, 6-9

Royal Society Commission on, 7, 23

cause and carrier of, 7, 23-4

checked by depopulation, 8

in absence of G. palpalis, 9

in W. Africa, 17

eradication of, 61-5
Slow worm, 190
Slugs, 321

as evidence of insularity, 103, 115

attacked by Dorylus, 279
Snails, attacked by Dorylus, 279
Snakes, 95-6, 116, 187-93

danger of, 187

evolution of poison of, 191-3

as food of G. palpalis, 39, 188
Snipe, 152

(" Sombabyuma ") (see Ciconia), 155
Spalgis pilos, 107
(Sparrow) (see Passer)
SPEKE, J. H., 67, 137, 139
Sphegidae, 286-7, 204-300

enemies of, 302

noise of, when working, 295, 298

832

INDEX

Sphegidae, methods compared with Pom-
pilidae, 300

unusual nesting place of, 295

prey of, 294

Lycoid, 232
Sphex marginatus, 302

pelopaeiformis, 295
Sphingomorpha, 221
Spider (see also Nephila), 120, 318-9

eaten by Sunbird, 177

and Dorylus, 279

anticryptic, 203

mimetic, 227-8

stored by Sceliphron, 297

curious web of, 318-9
Spider islands, 102, 105, 107, 109, 114-5,
118, 120, 124

scarcity of birds on, 102, 103, 105, 120,
124

predaceous insects on, 102, 105, 109, 120
124
Spines of caterpillars, 213
Spirochaetes, 18
Spizdetus coronatus, 164
Sporozoa, 11

Starling, glossy, absent from islands, 176
(Stick insect) (see Phasmidae)
Stilt, 152

(Stmk ant) (see PaUothyreus)
Stone Curlew, 97, 160-1
Stork, 146, 155

Common, 155

(open bil'.ed) (sei;* Anastomus)

(saddle billed) (see Ephippiorhynchus)

(whale headed) (see Balaeniceps)
Storm, 73-5, 82-3, 91-2, 101, 123
Stylogaster and Eciton, 281
Sunbird, 103, 105, 111-3, 115, 121-2, 124

caught in web of Nephila, 109
Sunflower, 70, 122
Surra, 14, 18
Swallow, 108, 113, 117
SwYNNERTON, C. F. M., on birds and
butterflies, 236-7

on relative edibility, 217
{Syntomidae (see Epitozis)
Syntomosphyrum glossince, 61-2

Tahanidae (see also Chrysops, Tabanus),

carriers of Trypanosomes, 18
Tabanus thoracinus, 293

variabilis, 315

vctfiQ/ttcs 97
Tachinida'e, 220, 286, 288, 297-8, 315, 316
Tarucus telicanus, 111, 121, 124
Tasting, experimental, 212-3
Tavtt Isle, 96-7. 180, 182-4, 189

Qlossina on, 32, 47-9
breeding places of, 55

butterflies on, 125
Tchitrea emini, 102, 103, 105, 110, 115 116
124, 171, 174

local island races of, 275

nest of, 174
Telipna nyanzm, 233
Teracoluis, 114

seasonal forms of, 202

bright colours of, 208-9
Terias, 114

colouration of, 204

attacked by bee-eater, 173, 237
Termites, 80, 97, 162

absent from some islands, 96, 112

prey of Megaponera, 283
(Tern) (see Hydrochelidon)
Thayer, A. H., on colouration, 204, 238
Thrush, 112, 119, 179
Ticks, scarce on islands, 319

Tiger moth, 213

Tipulidae, 310

ToKwi Isle, 121-2

Tortoise, 39, 135, 186

( Tragelaphus spekei) (see Situtunga)

Trees (see also " Ekirikisi," Fig-tree, Podo

carpus^, 70, 72, 95, 98-9, 104-5
Trilocha obHquissima, caterpillar of, 215-6
TRIMEN, K., 195, 245, 253
Tring museum. Ps. eurytus in, 246, 249

253-4
Tringoides hypoleucos, 146, 161
Trypanoplasma, 18, 13
Trypanosoma, 12-14
brucei, 14, 16
becomes rhodesiense by mutation, 17
believed to be mechanically trans-
mitted, 23, 25
carried by G. morsitam, 23
natural hosts of, 26
equinum, 14
eguiperdum, 14
evansi, 14

gambiense, 13, 14, 24
possibly carried by G. morsitans, 9
and natural hosts, 15, 26-9
life cycle, 25-6, 40
still pathogenic on depopulated
islands, 28
grayi, 40
letvisi, 13
nigeriense, 15, 17
rhodesiense, 10, 14, 17
vivax, 13, 40
Trypanosomes, and Sleeping Sickness, 7, 24
position in animal kingdom, 11
near relations, 12
general appearance, 13
human, 6, 9, 10, 14, 23
do not always cause disease, 15
hosts of, 15-17
life cycle of, 17, 25, 40
and bacilli, in gut of fly, 40-1
carriers of, 18

compared with Spirochaetes, 18
and direct inoculation, 23
no hereditary transmission of, 25
Trypanosomiasis, human, 2, 23
in Brazil, 14, 18
in Nigeria, 15
in Rhodesia, 9, 10
compared with Nagana, 24-5
acquired immunity in, 17
Tse-tse (see also Qlossina), in Uganda, 8

origin of name, 19
(Tse-tse fly disease) (see " Nagana ")
(Tufted Umbre) (see Scopus)
TtTLLOCH, Lt., 7
(" Tutuma ") (see Coucal), 166

Uganda, Ps. eurytus in, 245, 247-8, 250-2,
254-9, 262-3
Sleeping Sickness in, 7, 8
Ukerewb Isle, 67
Uranothauma, 110
USAMBAKA, Ps. eurytus from, 254

Van someren, Dr., 171
Varanus, 182-5

as food of G. palpalis, 37-8

nesting places, 98, 118, 183

basking places, 183

food of, 151, 184-5, 320

relations with crocodile, 183

numbers on different Isles, 96, 99
109-113, 115, 116, 118-9, 122, 124
Variations, continuity in, 270

geographical, 235, 241, 247, 262

4

INDEX

333

Variations, heritable fluctuations, 241,
268 271—2
Mutations, 240-1, 268-9, 270-2
of mimic in absence of model, 256-9,

262, 267, 273
seasonal, 201-2
Tiansitional forms, in P. dardanus,
269-71
absent in P. polytes, 269
in Ps. eurytus, proportions of, in

different localities, 265
survive in absence of models (see Ps.

eurytus), 266, 273
and Mutation theory, 26S-271
Vertigo, 322

Victoria Nyanja, 67, 69-71
climate of, 72
faU in level of, 69
signs of weather on, 76-80
sounds of life on, 80-81
storms on, 73-5, 82-3
Vinago, 159
Vitrina., 321
Viverridae, 141-2
Vulture, absent from islands, 164

Wagtail, 121

attacks on butterflies. 111, 208, 238
Waitwe Isle, 116, 117

faunal characteristics, 116

Glossina on, 117
(" Waka ") (see Paltoihyreus), 281
Wailace, 195

on struggle for existence, 218
Warbler, 102, 178
Warning colours, 195, 206, 230
Wasps, 97, 108, 207, 284-8

prey of Bee-eater, 172

syn-aposematic, 230, 232
Waterbuck and T, gambiense, 27
Water lilies. 69, 101

(Water scorpion) (see Nepa)
Waterspout, 75
Wavuziwa Isle, 115-6

faunal characteristics, 116

Glossina on, 115

butterflies on, 125
Weather prophets, 77-80
Weaver birds, 102-3, 108, 111-3, 115,

117, 120-2, 176
Weevil, 196
Wema Isle, 68, 113-5

Olossina on, 51, 57, 114

butterflies on, 114, 125

monkeys on, 131
West Africa, Ps. eurytus in, 245, 247-8.

251, 254, 262
Wiggins, C. A., collection of Ps. eurytus

255, 259
Wind (see also storms), 76, 90
(" Wonzi ") (see Spizdetus), 164
(Wood hoopoe (see Irrisorinae)
Woodlice, 321
Woodpecker, 78
Wood swallow, 235
Woolly bear, 213

Xenetomorpha carpenteri, 228
Xylocopidae, 303

Ybmpata Isle, 117-8
accessory isle, 117
Olossina on, 32, 118
monkeys on, 131

ZeUus antifaunus, 114
ZiGUNOA Isle, 113. 116

faunal characteristics, 113
ZiRO Isle, 108-9. 141
Zonocerus elegans, 215
Zygaenidae, 213, 230, 232

Printed in Great Britain by

tJNTVIN BROTHEBS, LIMITED
WOKING AND LONDON

i