gESH5H5H5E5E5H52525H5HE5H5HSE.

c

(

1 From the
1

iH525E5H5H5E5H5HSS5H5H5HSH5H5E5^

iiiDFary oi |

a Presenfed S

1 ¥7am the compiiimenis s

3 of fke

auittlior i

a ancl flie

pnaoiaslieF B

1 /

1

SH5ESH5E5HS25H5HSH5HSE5ESH5E5E5S

RECENT ADVANCES IN ENTOMOLOGY

A TEXTBOOK OF PLANT VIRUS DISEASES

By Kenneth Smith, D.Sc, Ph.U. loi Illustrations.
By the same Authwr

RECENT ADVANCES IN THE STUDY OF PLANT
VIRUSES

I Coloured Plate and 67 Jext-tigures.

RECENT ADVANCES IN CYTOLOGY

By C. D. Darlington, D.Sc, Ph.D. Sccnnd Edition, ih Plates, 160
Text-tigures and 81 Tables.

RECENT ADVANCES IN AGRICULTURAL
PLANT BREEDING

By H. Hunter, Hon. M.A.(Cantab.), D.Sc. (Leeds), and H. Martin
Leake, M.A., Sc.D.(Cantab.). 16 Plates.

RECENT ADVANCES IN PLANT GENETICS

By F. W. Sansome, Ph.D., F.L.S., F.R.S.E., and J. Philp, B.Sc,
F.L.S. 56 Illustrations.

RECENT ADVANCES IN BOTANY

By E. C. Barton-Wright, M.Sc.fLond.), F.R.S.E. 60 Illustrations.

AN INTRODUCTION TO COMPARATIVE
ZOOLOGY

A Text-book for Medical and Science Students.

By F. G. Sarel Whitfield, F.R.E.S., F.R.M.S., and A. H. Wood,

M.A. i}i Illustrations.

THE MICROTOMIST'S VADE-MECUM
(Bolles Lee)

Edited by J. Bronte Gatenby, Ph.D., D.Sc, and Theophilus
Painter, .A..M., Ph.D. Tenth Edition, ir Illustrations.

RECENT ADVANCES IN
ENTOMOLOGY

BY

A. D. IMMS

M.A., D.Sc, F.R.S.
Reader in Entomology^ University of Cambridge

SECOND EDITION

With 94 Illustrations

PHILADELPHIA

BLAKISTON'S SON & CO. INC.

1012 WALNUT STREET
1937

Printed in Great Britain

■ ... ©

PREFACE TO THE SECOND EDITION

In preparing a second edition of this book two reasons have
led to the decision to confine its scope within the original limits.
Firstly, in any attempt to discuss recent advances in entomology,
at all exhaustively, several volumes, each the size of the present
one, would be needed. Even to include adequate extra chapters
the size of the book would need to be increased a good deal beyond
its prescribed limits. Secondly, those other aspects of entomology
that commend themselves for inclusion are, to a large extent,
provided for elsewhere. In this connection, insect physiology,
insects and climate, general morphology, insects and plant viruses,
for example, have already formed the subjects of recent books or
monographs. The writers of these several books have discussed
such subjects far more adequately than could be expressed within
the limits of single chapters.

Certain aspects of entomology, naturally, have required more
revision than others. Under morphology, while current views on
Wing Venation have undergone but little change. Head
Segmentation and the Genitalia have needed considerable revision.
Also, recent studies of the musculature have led to revision of
views respecting the homologies of certain of the appendages
and their parts. As regards Metamorphosis, discoveries in
regard to the presence of hormones are discussed. Palaeontology
has developed so rapidly as to require very full revision of the
original chapter : several new illustrations, taken from recent
papers by Dr. F. M. Carpenter, have been included in this chapter
with the author's permission. In dealing with the Sense Organs
the subject of response to visual stimuli has been extended so as
to include the results of a good deal of recent investigation. In
this connection valuable help has been given by Dr. M. Hertz,
which is here gratefully acknowledged. The subject of Stimulatory
Organs has also been included. With regard to Biological Races,

vi PREFACE TO THE SECOND EDITION

the incorporation of new material has amphfied this section,
wliile Locusts and the Phase Problem form a subject which has
been rewritten and expanded. Finally, the two chapters on
Biological Control have required extensive revision in order to
bring them more nearly up to date. Minor additions and
alterations, which are numerous, do not require individual
mention.

It will become apparent, on comparing the present edition with
its predecessor, that certain sections of the book are reprinted
in toto or with only slight modifications. This does not necessarily
imply that no recent work has aj^peared with reference to the
sections in question. On the other hand, it may be taken that the
results of such work do not materially alter the conclusions or
better exemplify the facts or phenomena concerned.

As compared with the first edition, this book has been enlarged
to the extent of 56 pages. Of the illustrations, twenty-seven are
new, while eighteen of the earlier figures no longer appear. It
needs to be reiterated that the book is for students and teachers,
and does not cater for specialists in the particular fields that
are covered. The necessity for a second edition is regarded as an
indication that the book has, on the whole, fulfilled its purpose.

A. D. IMMS.

Zoological Laboratory,
Cambridge.

PREFACE TO THE FIRST EDITION

During the past fifteen years entomology has accumulated a
multitude of facts and has advanced along many paths of inquiry.
It attracts to-day a larger band of investigators and devotees than
any other province of zoology. The day of the general entomo-
logist is passing by, and it is becoming increasingly difficult for any
one individual to view the subject in full perspective. The tracing
of its many-sided recent developments, not only in the field of
purely scientific research but also in more direct relation to human
activities, has become a task of considerable magnitude. This
book makes no claim to fulfil so ambitious a scheme. It attempts
to review and discuss certain aspects of the subject along which
recent advances have been fertile in new facts and ideas. Short-
comings will inevitably assert themselves in this connection, and
specialists will perhaps maintain that the treatment meted out to
their particular spheres is lacking in detail, or that other aspects
of entomology have been omitted altogether. In answer to just
criticism of this kind it can only be urged that the dimensions of
the book necessarily limit its scope. Also, it does not presume to
instruct the specialist in his own field. Further, the theory and
practice of insecticides, along with certain other aspects of applied
entomology, have been omitted for the reason that they have
already been adequately treated in more than one very recent
text-book.

It is hoped that this book will enable the student to become
better acquainted with some of the more recent developments in
entomology. Particular stress has been laid upon aspects of the
subject that are but little discussed in modern text-books. A
general elementary knowledge of the subject on the part of the
reader is assumed, and the bibliographies provided at the ends of
the chapters will serve as a guide to the sources of further informa-
tion, In addition to students of entomology, there is a growing

viii PREFACE TO THE FIRST EDITION

body of more advanced workers interested in its general problems.
Some approach the subject from the standpoint of unfettered
research ; many others are practical entomologists whose occupa-
tion is largely directed towards insect control ; and there are those
who teach entomology and whose outlook requires them to com-
bine those of the others just mentioned. If this volume fulfils
their several requirements, to the extent that they derive some
profit by reading it, the author will feel amply repaid.

Opportunity is taken here of expressing indebtedness to
Dr. F. Muir for reading and criticising parts of the book when in
manuscript, and to all those whose writings have been drawn
upon in preparing it. M. d'Emmerez de Charmoy has kindly sup-
plied information with respect to biological control measures
in Mauritius. Dr. Anton Handlirsch has generously consented to
the use of several illustrations contained in his article " Ueber
die fossilen Insekten," published in the Proceedings of the First
International Entomological Congress. Mr. Arthur Gibson,
Dominion Entomologist for Canada, has loaned several original
drawings pertaining to publications issued by his Division. Dr.
R. J. Tillyard has also granted permission to use several illustra-
tions from his papers. Indebtedness is also expressed to Sir Guy
Marshall for the use of Fig. 94, which is taken from the Bulletin of
Entomological Research ; to Mr. Clyde W. Mason for three figures
taken from his articles dealing with structural colours in insects ;
to Dr. R. E. Snodgrass and Dr. H. E. Ewing for the use of several
illustrations taken from their respective publications issued by the
Smithsonian Institution. Thanks are also due to the latter body
for its consent to their adoption. A number of other figures have
been borrowed, or adapted, from various sources and acknow-
ledgments are made to the authors concerned in the respective
legends.

A. D. IMMS.

July, 1930.

CONTENTS

I'AQE

Preface to the Second Edition v

Preface to the First Edition ...... vii

CHAPTER I
Some Aspects of Morphology ...... 1

Segmentationof the Head. Wing Venation. The Metameric
Appendages.

CHAPTER II

Some Aspects of Morphology (continued) .41

The Metameric Appendages {continued). The Genitaha.
Literature.

CHAPTER III

Metamorphosis 51

Berlese's Theory. Types of Larv?e. General Remarks on
Insect Metamorphosis. Literature.

CHAPTER IV

Palaeontology 72

Introduction. Orders only known as Fossils. The Geo-
logical History of Insects. The Phylogeny of Recent and
Fossil Orders of Insects. Literature.

CHAPTER V
The Sense Organs and Reflex Behaviour .108

Introductory Remarks. The Eyes and Light Perception.
Chemical Stimuli and their Receptor Organs.

CHAPTER VI
The Sense Organs and Reflex Behaviour (continued) . .139
The Tactile Sense. Chordotonal Organs and Reactions to
Vibrations. General Stimulatory Organs. Reflex Behaviour
and Practical Entomology. Literature.

CHAPTER VII
The Fundamental Aspect of Coloration .172

Structural Colours. Pigmentary Colours. Combination
Colours. Further Remarks on Insect Coloration.
Literature.

R.A. ENTOMOLOGY. ix b

489 >>t^

PAGE

X CONTENTS

CHAPTER VIII

Some Aspects of Ecology 205

Temperature.

CHAPTER IX

Some Aspects of Ecology (continued) ..... 228

Humidity. Light. Atmospheric Pressure. Food. Climate.
Literature.

CHAPTER X
The Practical Application of Ecology . .262

Regulation of the Growth of the Crop. Resistant Varieties.

CHAPTER XI
The Practical Application of Ecology (continued) . . . 286

Host Selection and Biological Races. Ecological Aspects
of the Locust Problem. Literature.

CHAPTER XII
Parasitism 315

Introductory Remarks. Host Selection. I*hases of
Parasitism.

CHAPTER XIII

Parasitism (continued) ....... 330

Types of Life-cycle. Host Relations of Parasites. Poly-
embryony and Related Phenomena. The Behaviour of
Adult Parasites. Literature.

CHAPTER XIV
Biological Control 368

Biological Control of Insect Pests. I. Parasite Intro-
ductions.

CHAPTER XV

Biological Control (continued) 391

II. Factors Governing Parasite Introductions. HI. The
Utilisation of Indigenous Parasites. Biological Control of
Noxious Weeds. Literature.

Index of Authors ........ 421

General Index 426

RECENT ADVANCES IN
ENTOMOLOGY

CHAPTER I
SOME ASPECTS OF MORPHOLOGY

1. Segmentation of the Head, p. 4. 2. Wing Venation, p. 16.
3. The Metameric Appendages, p. 21. A. Thoracic Appendages,
p. 23 ; B. The Pleura, p. 28 ; C. The Primitive Insect Leg, p. 31 ;
D. The Cephalic Appendages, p. 32.

The importance of an adequate and exact knowledge of
morphology has become increasingly evident in connection with
many aspects of entomology. Two notable text-books have
appeared within the last four years — one by Weber (1933) in
Germany, and the other by Snodgrass (1935) in the United
States. A perusal of these two works will make it evident to the
reader that the subject is emerging on a more exact and scientific
footing than hitherto. While morphology is essentially a
descriptive subject, it is becoming less divorced from function.
In dealing with exoskeletal organs and parts, not so much emphasis
is being laid upon sclerites and sutures alone, as primary criteria,
and more recognition is being accorded to the underlying
musculature. The morphology of the labium, for example, when
studied in conjunction with the musculature, is becoming better
understood and placed upon a surer basis. The requirements of
taxonomy, alone, will long provide a field for the application of
morphological knowledge, and advances in the latter will gradually
influence the systematic side of entomology. Progress in this
direction is naturally very slow, since it is only the established, or
more or less stabilised, morphological conclusions that concern

K.A. ENTOMOLOGY. 1 1

2 SOME ASPECTS OF MORPHOLOGY

the taxonomist. The difficult subject of the genitaUa, especially
when taken in conjunction with the facts of development, is
becoming more clarified and we are beginning to see an underlying
fundamental similarity in these parts. It is to be deplored
that morphology, as a whole, is becoming involved in a con-
fusing terminology and a consistent nomenclature will sooner
or later have to be adopted if a complex synonomy is to be
avoided.

The most significant change of attitude to-day is seen in the
increasing appreciation of the importance of insect physiology. A
vast literature on this subject already exists ; most of the older
work is now of little value in view of the great progress made in
experimental technique. Some idea of the recent growth of this
subject may be gathered from a valuable article by Hoskins and
Craig (1935). These authors mention having considered some
1,200 papers published during the five years 1931-35, and they
cite 444 of these in the useful bibliography which is appended.
We may pause for a moment and consider the causes contributing
to the rapid growth of this type of investigation. For one thing,
the applied entomologist, in his efforts to find a surer foundation
than purely empirical methods for the control of noxious insects,
is devoting attention to the possibilities of insect physiology.
He is realising that the most promising line of approach lies in a
better knowledge of the inner working of insects themselves. The
growing importance of insects as carriers of disease organisms of
man and domestic animals is likewise creating a demand for similar
knowledge. Applied entomology, therefore, as a whole, in its
need for information of this kind, is focussing the attention of
many workers on physiological problems. It needs also to be
remembered that the rapid growth of the experimental viewpoint
among general zoologists is another contributing influence, since
an increasing number of zoologists is being attracted to the study
of problems of a physiological nature afforded by the insect
world. Already more than one academic body has been led to
provide for some measure of teaching and research in insect
physiology. A very large literature on the subject in question is
appearing in an extraordinary diversity of periodicals, many pf

SOME ASPECTS OF MORPHOLOGY 3

which it has not been previously necessary to consult in connection
with entomology at all. The future burden of the worker in
insect physiology of keeping pace with this literary growth seems
likely to become increasingly difficult for this reason. One
result from this increase in physiological knowledge seems to be
that the morphologist will become better able to develop his
subject, since it will be in the light of increased acquaintance
with basic functions. Growth of physiological knowledge, also,
is likely to involve the re-examination of many features of the
minute structure of insects, with possibilities of interpretations
being made from another angle of vision. In connection with the
foregoing remarks, reference needs to be made to an introductory
manual of insect physiology by Wigglesworth (1934). It
formulates what is virtually a new subject and co-ordinates
much scattered information not readily accessible or easily
evaluated.

In order properly to appreciate the morphological significance
of a good deal of insect structure it is necessary to look outside the
limits of entomology. While insects, as a class, show a funda-
mental organisation common to the Arthropoda, their own basic
characters have been evolved by a long and complex process
of modification. The morphological interpretation of these
characters is essentially a study of homologies, which is hampered
owing to lack of evidence as to the course taken by insects in
their evolution. While a number of writers claim that the
origin of these creatures was either in common with that of the
Myriapoda, or from some individual class among that assemblage,
other writers bring forward reasons suggesting derivation from
an early crustacean type. On the whole, both anatomically and
physiologically, insects appear to show a balance of characters
favouring the idea of descent frorn a myriapodan rather than
from a crustacean prototype (Imms, 1936). This contention
forms the basis upon which some of the more general morphological
deductions made in this chapter are founded.

In the present chapter the following aspects of morphology are
discussed: (1) the segmentation of the head; (2) the wing
venation ; and (3) the metameric appendages.

1—2

4 SOME ASPECTS OF MORPHOLOGY

1. Segmentation of the Head

Phylogenetic studies have led to the conclusion that arthropods
were originally descendants of annelids. The reader may be
reminded that in the latter animals the only differentiated head
is the prostomium, which projects in front of the mouth as an out-
growth of the first body-segment. Situated within the prostomium
is a pair of nerve ganglia forming the archicerebrum or primitive
'■-'-•.vv, •• ■■-•^ brain. Each of the post-oral regions or segments
: 'y.' * of a Poly chate worm carries a pair of appendages :

' ^' *^ sT ' * each segment is innervated by its own nerve

centre, and its cavity is formed by the coalescence
of a pair of embryonic ccelom sacs. It was
pointed out by Lankester in 1881, and by
Goodrich in 1898, that the arthropod head is
° derivable from that of the annelid by the fusion
together of the prostomium along with a definite
\:; number of the original body-segments which

Fig 1 An early ^^^^^^^ ^^ • ^^ other words, it is a syncephalon. It
embryo of follows, therefore, that those head-segments
hfJ'dTvrsion^nTo ^^^ ^^^^^ appendages which are pre-oral in
A, the proto- arthropods, must have secondarily acquired that
cephalic or^pn^ position. The arthropod brain is similarly a
region, and B syncerebrum, formed by the fusion of the annelid
orprimI?y trlTnk archiccrcbrum with the nerve centres of certain
region. (From of the segments behind.

Heymons.) Without following the evolution of the

syncephalon in different groups of arthropods, it may be mentioned
that among insects the embryo becomes differentiated at an early
period into a greatly expanded protocephalic or primary head
region, and a protocormic or primary trunk region (Fig. 1). The
primary head, or protocephalon, is represented by the large
cephalic lobes in connection with which the acron and first three
head-segments develop. The primary head, as thus constituted,
is an embryonic and presumably recapitulatory phase only, and
the completed head of an insect is formed by the coalescence of the
first three protocormic segments with those of the protocephalon.

SEGMENTATION OF THE HEAD

On the basis of these contentions it will be observed that six
segments enter into the composition of the insect head, and this
conclusion is now widely accepted. Embryology affords three
fundamental criteria which provide the surest evidence with regard
to segmentation, viz., the existence of paired appendages, of
neuromeres and of primitive coelom sacs. Taking these three
criteria as the basis, the composition of the head will be discussed
in the light of recent investigations. A noteworthy addition to
knowledge of insect embryology is the work of Leuzinger, Wiesmann
and Lehman (1926) on the stick insect {Carausius morosus). It
is Wiesmann's contribution in this memoir that is of special
importance in the present connection. Besides confirming much
of what was previously known regarding the composition of the
insect head, this author brings forward new facts of considerable
theoretical significance. According to him the head of Carausius
is composed of an acron and six, or possibly seven, embryonic
segments, and his findings are expressed in tabular form below.

Acron.

1. Labral segment

2. Pre-antennal segment

3. Antennal segment

4. Intercalary segment .

5. Mandibular segment .

6. 1st Maxillary segment

7. 2nd Maxillary segment

Without coelom sacs.

Coelom sacs rudimentary,

55 55

, , present.

,, rudimentary,

,, present.

Without appendages.

Appendages ?

Appendages rudimentary.

Antennae.

Appendages rudimentary.

Mandibles.

1st Maxillae.

2nd Maxillae.

Wiesmann's discovery of rudimentary coelom sacs in the labral
region (Fig. 2, B) was wholly unexpected, and this author has
suggested that a true somite, in front of the protocerebral segment,
is present. The labrum itself develops from paired lobes which he
believed may represent the appendages of this segment since they
are closely associated with the coelom sacs just mentioned.
Eastham (1930) states that in Pieris the labrum similarly arises
from a pair of lobes : these outgrowths fuse at their bases, pass
backwards with the stomodaeum and become connected with the
epipharynx. The pre-oral mesoderm becomes secondarily divided
into a principal anterior part, which passes into the hollow labral

6

SOME ASPECTS OF MORPHOLOGY

rudiments, and a smaller posterior part, which comes to lie at the
sides of the epipharynx. Cavities of doubtful significance may
appear in both these mesoderm masses, but, Eastham remarks, they
are so irregular and so unlike those of the well-defined somites of
the body that it is doubtful whether they have any metameric
importance. This author, in discussing the general problem of

n.

ow

Pr'A . coel.

Fig. 2. A. Anterior extremity of an embryo of the Stick Insect,
Carausius morosus, showing rudiments of developing appendages.
A, antenna ; H, hypopharynx ; L, labrum ; Md, mandible ;
Mci', maxilla ; Pr.A, pre-antenna. B. Transverse section across
a similar embryo, passing through the pre-antennal rudiments
(PrA). PrA.coel, pre-antennal coelom sac ; S, stomodaeum ;
Yl, layer of yolk cells. (From Wiesmann.)

head-segmentation, supports the established opinion that six
somites are involved, but concludes that the paired labral rudiments
represent the appendages of the protocerebral segment. Very
recently the occurrence of coelom sacs in close association with
paired labral rudiments has been described and figured in two
other insects : viz., by Mellanby (1936) in the Reduviid bug
Rhodnius prolixus and by Roonwal (1937) in the African Migratory
Locust (Locusta migratoria, subsp. migratorioides, Fig. 3). Both

SEGMENTATION OF THE HEAD 7

these writers conclude that the labrum represents the first true
cephaHc segment.

Whatever may be the true interpretation of the foregoing
observations, it needs to be recollected that in the Collembola and
in Lepisma, Campodea and Japyx the labrum develops as an
unpaired rudiment from its first stage of growth. The same
condition is found in Scolopendra and also in Blatella, Mantis,

Fig. 3. Portion of a longitudinal vertical section of a fifty-nine
hours' old embryo of Locusta, showing the labral ccelom.
X 375. a, amnion ; I, labrum ; Ic, labral coelom ; m, mesoderm ;
s, stomodaeum. (After Roonwal.)

Gryllus and Forficula. On the other hand, the labrum develops
from paired lobes in Carausius, Locusta, Stenobothus, Rhodnius
and in Lepidoptera and Coleoptera. Among Hymenoptera its
rudiments may be either paired (Chalicodoma) or unpaired (^pf^).
The labrum, therefore, may differ as to its mode or origin within
the limits of a single order of insects. We are thus faced with the
difficulty of deciding which of these two conditions is to be
regarded as the original or primary state. Since the labrum
develops from unpaired rudiments in the Chilopoda and

8 SOME ASPECTS OF MORPHOLOGY

Apterygota, besides in some of the Orthoptera, it might be claimed
with some justification that this condition is primitive. In those
examples where it is initiated in the form of paired lobes it might
be assumed that a secondary division occurs during development
and that the unpaired condition is re-acquired later. Without
pursuing this matter further it needs to be taken into account
that the process of the amalgamation of the cephalic segments is
phylogenetically an extremely ancient phase in arthropod
evolution : secondary changes have almost certainly supervened
to the extent of partially obliterating the original condition.
Furthermore, since the complete arthropod head is believed to
have been formed by the successive incorporation of additional
postcephalic segments with the primitive cephalon, the most
anterior segments are the most ancient in the process of cephalisa-
tion. It is these anterior segments, therefore, that are most likely
to have become so transformed, in the lengthy process of evolution,
as to betray but little evidence of their individuality. If what is
known of the segmental constitution of the head in all classes of
Arthropoda be taken into account, it appears unlikely that two
pre-antennary segments (protocerebral and labral) should occur
in the Insect a and in no other class. The subject is discussed
by Manton (1928), who doubts whether the labral coelom sacs are
independent of those of the protocerebral segment. If this
conclusion be correct it may be inferred that only a single pre-
antennary segment is involved and that it has undergone secondary
division. Further discussion of this problem, however, is scarcely
profitable until more evidence is available, andManton's hypothesis
is to be regarded as a tentative effort to reconcile conflicting
evidence.

Up to this point the embryological interpretation of the
constitution of the insect head has been considered. Hanstrom
(1937-30), from a comparative study of the structure of the
cephalic ganglia of Annelida and Arthropoda, arrived at a
different conclusion. While his theoretical deductions have
not gained wide acceptance, his work, as a whole, forms a
valuable contribution to invertebrate neurology. The arthropod
protocerebrum and deutocerebrum, he claims, are secondary

SEGMENTATION OF THE HEAD 9

developments from a single ganglionic centre, representing the
highly modified annelidan archicerebrum. The tritocerebrum,
according to Hanstrom, is the ganglion of the first true head-
segment. The arthropod head, on the basis of his theory, is
composed of four segments only, viz., those represented by the
second antennae, mandibles, maxillae and labium, together with a
pre-oral or prostominal region bearing the labrum, eyes and first
antennae. The coelomic cavities present in front of the trito-
cerebrum are interpreted by him as being of secondary origin and
devoid of metameric significance. The first antennae and pre-
antennae come under the category of prostomial tentacles and are
consequently not regarded as being true appendages. Hanstrom's
views are accepted by Snodgrass (1935), and without going into
details it needs to be stressed that great importance is given to the
general occurrence of corpora pedunculata, or mushroom bodies, as
indicating cerebral homologies. These structures constitute the
largest and most important association centres in the insectan
brain. They are, furthermore, present in all arthropods and are
also of common occurrence in the archicerebrum of annelids.
Relatively slightly developed in the Polychaete family Hesionidae
they become more highly differentiated structures in the families
Aphroditidae and Nereidae and ultimately attain their greatest
development among the Insecta. The evidence brought forward
by Hanstrom leads to the conclusion, at any rate, that the
annelidan archicerebrum forms a large part of the arthropodan
protocerebrum. Hanstrom also maintains that the optic centres
or arthropods have been evolved from homologous tracts located
in the optic lobes of the archicerebrum, as seen in the Polychaete
family Emiicidae. While these primitive optic centres in the
Polychaeta have a very simple structure, they appear to grade into
the more elaborate, but still elementaTy, optic centres found in the
Branchiopod Crustacea. While the value of Hanstrom's studies
is not likely to be disputed, his conclusions appear to be the result
of too much stress being laid upon brain structure alone. While
it may be granted that evidence of the existence of pre-antennae
in insects is scanty, it is difficult to accept his conclusions relative
to the deutocerebrum with its undoubted ccelom sacs and well-

10

SOME ASPECTS OF MORPHOLOGY

defined appendages (antennae). To disregard the evidence of
metamerism afforded by this segment and to accept similar
evidence betrayed by the segments that follow renders his theory
untenable in so far as morphological values are at present
interpreted.

A third view as to the segmental composition of the insect head
is based upon the conclusions of Hansen. In 1893 this well-known
Danish morphologist claimed the existence of a pair of reduced
but true appendages situated between the mandibles and the first
maxillae in most Apterygota and some of the more primitive
pterygote insects. He termed these appendages maxillulae, since
it appeared probable that they are homologous with the first
maxillas (maxillulae) of the Crustacea. In his most recent studies
Hansen (1930) confirms his earlier opinions and brings forward
more detailed evidence in their support. If this contention be
well founded it follows that the insect head would have to be
interpreted as being composed of seven segments, as shown below.

Segment.

Appendages.

Ccelom sacs.

Neuromere.

1

Embryonic

Rudimentary

Protocerebrum

2

Antennae

Present

Deutocerebrum

3

Embryonic

Rudimentary

Tritocerebrum

4

Mandibles

Present

Mandibular

5

Maxillulae

Absent

Absent

6

1st Maxillae

Present

Maxillary

7

2nd Maxillae

Present

Labial

Comparison of the above table with that given on p. 5 will make
it evident that these seven suggested segments do not entirely
correspond with those recognised by Wiesmann. In the Collembola,
for example, Hansen claims that the maxillulae have hitherto been
identified as maxillary palpi notwithstanding the fact they are a
perfectly free pair of mouth-parts with separate articulations.
In Japyx the maxillulae likewise have separate articulations and

SEGMENTATION OF THE HEAD

11

labrum ;
maxilla
palp ; s

are only connected with the maxillae by membrane : hitherto they
have been interpreted as being the palpi and galeae of the maxillae.
On anatomical data only,
Hansen's conclusions appear
to have much in their
favour. He does not lay
much stress on the value of
embryological evidence and
only briefly discusses the
work of Folsom (1900) on
the development of
Anurida. In this connec-
tion he states that the pair
of developing lobes (_p. in
Figs. 4 and 5) associated
with the first maxillae are
the fundaments of the
maxillulae. Folsom, he
maintains, has misinter-
preted them as being maxillary palpi,
receive no support from the embryological studies of Philiptschenko

(1912) and of Silvestri (1932).
These observers find no traces
whatever of a maxillulary
segment either in the form of a
neuromere, appendages or coelom
sacs. The lobes identified as
maxillary palpi by Folsom are
shown by Silvestri to develop
into those parts as stated by
the first-mentioned writer.
Cramp ton (1921) made a rather
superficial study of the whole
question and claimed that the
maxillulae of Hansen are homo-
logous with the paragnaths of the higher Crustacea which are
non-appendicular in character. Hansen, in discussing Crampton's

Fig. 4. Ventral aspect of the head region
of an embryo of Anurida, showing
developing appendages. a, antenna ;
/, oral fold ; h, hypopharynx ; I,
le, leg ; m, mandible ; m\
m", labium ; p, maxillary
superlingua. (From Folsom.)

Hansen's views, however,

Fig. 5. Head and first thoracic
segments of embryo (about 1-20
mm. long) of Japyx major.
Lettering as in Fig. 4. Adapted
from Silvestri.

12 SOME ASPECTS OF MORPHOLOGY

conclusions, showed that the pair of small plates (superlinguae of
Folsom) overlying the hypopharynx in Collembola and Thysanura
Entognatha are probably homologous with paragnaths. On the
other hand, he contended that true maxillul^e co-exist with them
in the groups just named and therefore cannot be their homologues
as Crampton thought. In summing up the position it may be
said that non-appendicular plates or superlinguas (paragnaths)
undoubtedly exist among the Apterygota, but evidence of the
occurrence of maxillulae is very unconvincing and points to the
contrary.

The available evidence relative to the segmentation of the insect
head leads to the conclusion that it has resulted from the
amalgamation of at least six embryonic somites. The present
status of knowledge of the problem may be summarised as follows :

1. The pre-antennal or ocular segment is usually regarded as
being the first true cephalic segment, its neuromere being the
protocerebrum or first division of the brain. The protocerebrum
is probably composite since it is believed to comprise also the
primitive archicerebrum. The existence of rudimentary coelom
sacs and very small evanescent appendages pertaining to the pre-
antennal segment in the Phasmid Carausius (Fig. 2B) has not so
far been confirmed in any other insect. It is now tolerably well
established that coelom sacs, in close association with labrum,
occur in Carausius, Locusta and Rhodnius. It appears unlikely
that two pre-antennal segments are exhibited in the Insecta and no
other Arthropoda and the possiblity that the labral coelom is
part of that pertaining to the ocular segment cannot be overlooked.
Evidence that the labrum itself represents a pair of true appendages
is very hypothetical, especially in view of the fact that this organ
arises from an unpaired fundament in the Apterygota and other
insects.

2. The antennary segment is innervated by the deutocerebrum
or second neuromere of the brain : its appendages are the antennae
and coelom sacs are present in the embryo.

3. The intercalary segment is vestigial, but evidence of its
existence is found in the embryos of most insects. Rudimentary
coelom sacs are retained in some cases, and vestigial evanescent

SEGMENTATION OF THE HEAD 13

appendages (homologous with the second antennae of Crustacea)
occur in certain Apterygota, as well as in the cockroach, silkworm
and honey bee among Pterygota. In Campodea (Uzel, 1897) and
the grasshopper Dissosteira (Snodgrass, 1928) rudiments, which
appear to represent these appendages, are discernible in the
adult insect. Its nerve centre, or tritocerebrum, forms the third
neuromere of the brain, but it subsequently becomes largely
incorporated with the deutocerebrum.

4, 5, 6. The mandibular, first maxillary and second maxillary
segments are all well developed, and bear corresponding
appendages. Coelom sacs are present in the embryo, and the
neuromeres coalesce to form the infra-oesophageal ganglion.

The completed head-capsule is so highly modified and con-
solidated that it no longer betrays obvious evidence of its
segmental composition. It becomes, therefore, a difficult and
largely speculative problem to assign the regions, or parts of the
head, to their original segments. The whole subject is admirably
discussed from a comparative standpoint by Snodgrass (1928), who
points out that since the segments of the protocephalic region are
never distinct, even in the earliest embryonic stages, it seems
fruitless to hazard what areas of the adult head are to be attributed
to them individually. The original protocephalic region must at
least be the area occupied by the labrum, clypeus and frons,
together with the compound eyes and antennae. Also, since the
muscles of the three pairs of gnathal appendages have their
origins in the posterior part of the head, Snodgrass concludes that
the areas upon which these muscles arise are parts of the walls of
those segments that have been added to the protocephalic region.
It seems probable that the frontal suture marks off the original
prostomial area from that derived from the segmental elements,
and, if this be correct, the labrum,. clypeus and frons represent
differentiations of the original prostomium. In the primitive
Machilidse the posterior part of the epicranium is crossed by a
prominent suture just behind the eyes. This suture extends
downwards to a point between the bases of the mandible and
maxilla. Snodgrass advances reasons which suggest that this
suture separates the mandibular segment from that of the maxilla

14

SOME ASPECTS OF MORPHOLOGY

on the cranial wall. Further back on the cranial wall lies the
post-occipital suture, which is very constant in its occurrence and

,^Hpwy

Fig. 6. The hypopharyngeal apophyses and the tentorium. A,
Under surface of head of Lithobius, mandibles and maxilla;
removed, showing suspensorial plates (HS) of hypopharynx
suspended from points (d) on margins of head, and hypo-
pharyngeal apophyses (HA) invaginated from their inner ends
and connected by ligamentous bridge (j) beneath pharynx.

B, Head of Scolopeiidra, ventral, maxillae and right half of
cranium removed, showing attachment of mandibular adductors
(KL, KL) on ligament uniting the hypopharyngeal apodemes.

C, Scutigera forceps, ventral view of hypopharynx {Hj)hy),
suspensorial plates {HS), their apodemes {HA) and uniting
ligament {j). D, Heterojapyx gallardi, ventral view of right
maxilla, hypopharynx {Hyph), and hypopharyngeal apodemes
{HA), upon which arise the muscles of the maxilla {cidmx), the
labium {Ibmcl), and the mandibles (not shown). E, Nesomachilis
maoricus, posterior view of unconnected anterior and posterior
arms of tentorium {HA, PT), part of the head wall with clypeus
{Clp) and labrum {Lm), base of maxilla {Mx), and mandible
(Md). F, Ephemerid nymph, ventral view of tentorium and
part of left side of head, showing anterior tentorial arms {AT)
arising from ventral margin of gena {Ge). (After Snodgrass.)

appears to be of definite morphological significance. Internally
it forms the post-occipital ridge, which appears to be a true inter-

SEGMENTATION OF THE HEAD 15

segmental phragma between the maxillary and labial segments.
The hypothesis that this suture marks the division between those
two segments is borne out by the positions of the maxillary and
labial articulations with reference to it.

After the gnathal segments became added to the protocephalon,
it appears probable that their sternal parts lost their identities.
From the fact that the ventral adductor muscles of all the mouth-
parts have their origins on a pair of apophyses, arising from the
base of the hypopharynx in Chilopods and most Apterygota,
Snodgrass concludes that at least part of each of their sterna
has become incorporated into the hypopharyngeal region — a
conclusion which is supported to some extent on embryological
grounds. From a comparative study of Crustacea, Myriapoda
and Apterygota he has advanced an ingenious theoretical inter-
pretation of the origin and evolution of the insect tentorium
(Fig. 6). Without going fully into his lines of reasoning, it may
be said that the beginnings of that structure are to be seen in the
hypopharyngeal apophyses previously alluded to. From their
original position, at the sides of the hypopharynx, it appears that
they have migrated outward in the ventral"^wall of the head to
the lateral edges of the cranium, but retain the original muscle
attachments. In Odonata and Ephemeroptera these apophyses
come to lie in the sub-genal sutures, but in all other Pterygota
they have proceeded further to take up their usual position on
either side of the clypeo-frontal suture. The hypopharyngeal
apophyses have, according to this interpretation, become the
anterior arms of the tentorium. The posterior arms are invagina-
tions in the lower extremities of the post-occipital suture, which
unite with the anterior arms to form the typical four-branched
skeletal structure : they are undeveloped, it may be added, in
Myriapoda and most Apterygota. The dorsal arms arise as
processes of the anterior pair, and may subsequently acquire
secondary connection with the cranial wall. The condition of the
tentorium in Machilis appears to be largely intermediate between
that found in most Apterygota and Pterygota, since the anterior
and posterior elements are present, but still independent of each
other.

16 SOME ASPECTS OF MORPHOLOGY

The morphology of the neck region and the cervical sclerites is
extremely difficult to interpret, and many speculative views have
been advanced to explain it. It is obvious that if the post-
occipital suture marks the division between the maxillary and
labial segments, the tergal region of the latter segment is repre-
sented by the narrow band-like area behind this suture, and
forms the greater part of the boundary of the occipital foramen.
The fact that the dorsal muscles of the prothorax pass through
the neck and are attached anteriorly — not to the posterior
margin of the labial segment, but to the post-occipital ridge
beyond — appears to be a morphological fact of importance.
This relationship has led some morphologists to conclude that an
intersegmental region has become deleted somewhere in the
neck. The cervical sclerites are consequently regarded as dis-
membered portions either of the labial segment or of the prothorax.
Other morphologists maintain that the neck itself is the greatly
enlarged intersegmental region, and that the cervical sclerites
are sclerotised areas which have developed in the membrane of
that region.

A.

2. Wing Venation

The determination of the origin and homologies of the wing
veins has, in the past, been embarrassed by the various systems
of nomenclature employed by different workers. As a rule, the
terminology of an author was only applicable within the limits of
the particular group or order of insects which he studied. The
establishment of these different systems has been the inevitable
outcome of specialised studies, uninfluenced by modern conceptions
of evolution. These defects could scarcely be remedied at the
time because no criterion had yet been explored upon which
homologies might have been based. Such deficiencies have been
realised for more than forty years, and it was in 1886 that
Redtenbacher published an important and well-illustrated
memoir in which he proposed a uniform system of venational
nomenclature ; he named the veins costa, subcosta, radius,
media, cubital and anal respectively. Redtenbacher, furthermore,
reaffirmed the importance of the fact that the main veins primi-

WING VENATION

17

tively alternate as convex and concave veins, a feature previously
noted by Adolph in 1879. He also claimed that the oldest insects
exhibit the richest venation, which disappeared to a marked extent,
among recent forms, by reduction. Subsequent researches by
him in conjunction with Brauer, and also by Spuler, considerably
extended what little had been previously known of the relations
of venation to the antecedent tracheation. It was not, however,
until 1898 that any comprehensive ontogenetic studies were
undertaken. In that year Comstock and Needham began their
investigations of the homologies and development of the wing
veins in all orders of insects and, in 1918, Comstock extended and

Fig. 7. Hypothetical scheme of wing venation according to Comstock
and Needham. C, costa ; Sc, sub-costa ; R, radius ; Rs, radial
sector ; M, media ; Cu, cubitus ; lA, 2 A, 3 A, anals.

consolidated this work in his well-known text-book on the subject.
It may be said, very briefly, that, in the Comstock-Needham
system, the original vein nomenclature of Redtenbacher is followed,
but with many modifications with regard to his theoretical
deductions. By means of an extensive study of the tracheae
which precede and, in a general sense, determine the courses of
the veins, these authors came to the conclusion that existing types
of venation (living and fossil) are all derivable from a common
hypothetical ancestor, and that the veins in different orders of
insects can consequently be homologised with one another. They
further concluded that deviations from the hypothetical ancestral
type resulted, either by the addition or by the reduction of veins,
as the case may be. No great importance was attached to
Redtenbacher 's theory that convex and concave veins alternate

18

SOME ASPECTS OF MORPHOLOGY

with one another, probably because this condition is regularly
apparent in certain orders only. Also the theory that the
ancestral winged insects had the richest venation, from which
later types were derived by reduction, was not accepted. While
we owe to Comstook and Needham what is one of the greatest
advances ever made in the study of insects, the importance of the
earlier fundamental work of Redtenbacher should not be over-
looked. The Comstock-Needham system (Fig. 7) is now well
known and, for this reason, it is not necessary to recapitulate here
what has often been described since in text-books. It will be
necessary, however, to discuss certain more recent investigations
on wing venation and their bearing upon the general theory.

The contributions of Tillyard, dating from 1919, have greatly
extended our knowledge of the subject in many orders, and this

Cii,

CU-TA

Fig. 8. Diagram of the branching of the cubital vein ; A, according
to Comstock and Needham ; B, according to Tillyard and later
authorities. Cu, cubital vein ; \A, first anal vein.

author's detailed studies of early fossil types shed further light
on venation from the palseontological aspect. In his papers on
the Panorpoid complex, ^ Tillyard arrived at the conclusion that
Comstock and Needham's hypothetical type is not entirely
applicable as one from which the venation, of the orders forming
this complex, was presumably derived. The most important
feature brought to light by Tillyard 's earlier discoveries refers to
the composition of the cubital vein. He was able to show on
ontogenetic grounds that this vein is primarily three-branched,
and not two-branched as Comstock and Needham believed
(Fig. 8). The vein which the latter authorities regarded as the
first anal, which had become secondarily associated with the

^ The Panorpoid Complex includes the Mecoptera, Neuroptera, Trichoptera,
Lepidoptera and Diptera, together with related extinct orders only known as
fossils.

WING VENATION

19

cubital stem, Tillyard maintains is the branch Cug. In other
words, the primary bifurcation of the cubital vein takes place
much nearer the base, and the forked cubital vein of Comstock
and Needham represents its anterior branch only. The primi-
tively three-branched cubital vein is present in the lower orders,
as well as among Holometabola, and it is evident that it is to be
regarded as a fundamental character in any hypothetical venation
scheme for insects.

The views of Redtenbacher regarding the significance of the
alternation of convex and concave veins are supported in general
by modern researches. If, for example, two convex veins occur

Fig. 9. Hypothetical scheme of wing venation according to Lameere.

in proximity to each other on an insect wing, a strong presumption
arises that an intervening concave vein has either been merged
with one or other of the convex veins, or has been suppressed.
Reasoning along these lines has helped to establish the identity
of veins in both fossil and recent insects. The important
researches of Lameere (1922) are largely based upon deductions
of this kind, drawn from a study of both recent and fossil insects.
He concludes that no living insect possesses as complete a venation
as that found in the early fossils of the coal measures. In this
contention he affords strong support to the views of Redtenbacher,
which were not accepted by Comstock (1918). Lameere's study
of the Palaeodictyoptera led him to stress the importance of
following the alternation of convex and concave veins, and,
according to him, there are six principal veins on the primordial

20 SOME ASPECTS OF MORPHOLOGY

insect wing (Fig. 9). Each vein bifurcates into an anterior or
convex branch ( + ), and a posterior or concave branch (or
sector, — ). The concave subcostal vein, he beUeves, represents the
posterior branch of the costal, while his conception of the radius
does not differ from that of Comstock. The division of the media
into an anterior convex branch (MA) and a posterior concave
branch (MP) is a conclusion of prime importance. The logical
deduction to be drawn is that the media of Comstock, being a
concave vein, is the counterpart of the posterior media (MP) of
Lameere. This interpretation, however, was not advanced by
Lameere, who believed that MP is wanting in Holometabola with
the exception of a rudiment which had been previously regarded
by Tillyard (1919) as representing M5. Since, however, M5
according to Tillyard (1925) is convex in character, Lameere's
homology is improbable. It needs to be pointed out, however,
that Tillyard stated definitely in his earlier writings that M5 is a
concave vein and that Lameere based his conclusions upon this
assertion.^ With regard to the cubitus, Lameere's anterior
branch corresponds with Cuj (convex), and his posterior branch
with Cug (concave). His interpretation of the anal veins differs
from those of his predecessors in that he claims that two bifurcate
veins are present in the anal area. This conclusion can scarcely
be reconciled with the more usually accepted interpretation
(Fig. 7), and the theoretical assumption, that every wing vein
conforms to a regular basic system, remains to be proved.

When Lameere's hypothetical scheme is applied to recent
insects several important modifications become apparent, more
especially with regard to the media. It would seem that this
vein is only rarely represented by both of its branches — MA and
MP. This condition is retained among Ephemeroptera and is
certainly present in the Orthoptera. A comparative study of the
venation of both recent and fossil representatives of the last-

^ The oblique vein M5 is a characteristic feature of the orders forming the
Panorpoid Complex, and is evident in their Upper Permian fossil representatives
also. Tillyard (1919) regarded it as a basal concave vein connecting M with
Cui to form the cubito-median Y-vein. In a later paper (1925) the results
of further study led him to conclude that it is little more than a convex strut,
which has arisen as a secondary development, in order to provide rigidity to
the base of Cu^.

THE METAMERIC APPENDAGES 21

mentioned order is necessary, however, before it is possible to
generalise in this immediate connection. Among Odonata,
Tillyard has brought forward evidence indicating that the anterior
media (MA) only is retained in this order and it is probable that
this same condition will be obtained among Plecoptera. In
other orders, excepting the Diptera, it is very doubtful whether
MA is ever present. Among Diptera the so-called vena spuria
appears to be the vestigial MA, and occurs in the Ptychopterinae
and in various members of the Syrphidae.

Notwithstanding the large amount of study that has been
concentrated upon the problem of venational homologies, our
conceptions of the subject are not, as yet, by any means stabilised.
The results of palaeontological investigations have been to give
increased importance to the significance of convexity and concavity
of veins as the criterion upon which their homologies largely rest.
The fact that the tracheation so often does not precede and,
therefore, determine the courses of the veins obviously limits its
application. The reader who is desirous of following recent
theoretical views should consult papers by Vignon (1929, 1932)
and by Vignon and Seguy (1929), whose work has special reference
to Diptera. The hypothetical generalised scheme of venation
promulgated by Vignon (1929) involves heterodox views which
require much wider study before the truth or falsity of his ideas
can be determined.

3. The Metameric Appendages

The reader may be reminded that the structure of Arthropoda
presents a unity of plan rendering it highly improbable that these
animals had a polyphylectic origin. Many years ago Lankester
claimed that arthropods have developed from a single line of
primitive Gnathopoda which arose^ by modification of parapodia-
bearing Annelida, and this same view holds to-day. It therefore
follows that the annelid parapodia and the arthropod limbs are
genetically identical structures. Among the most primitive
classes, i.e.^ Crustacea and Trilobita, a biramous type of limb is
very general, but in all other Arthropoda morphologically uni-
ramous limbs prevail. Evidence of the occurrence of biramous

22

SOME ASPECTS OF MORPHOLOGY

limbs in any class other than the two aforementioned cannot be
accepted as proven. In view of the remoteness of affinity between
insects and Crustacea, the structural differences in the appendages
of the two classes are very marked. Since it has long been
customary to regard the crustacean biramous limb as the standard
for comparison, discussion of arthropodan
limb-morphology usually has reference to
that type of appendage. While advantage
may be derived from a uniform standard
for comparison, it involves homologies which
require due precaution before being accepted
as being proven.

A complete biramous appendage (Fig. 10)
consists of a basal region or protopodite
which bears the main shaft or endopodite
and a usually smaller shaft or exopodite.
The protopodite is divisible into a basal
sclerite or precoxa, an intermediate sclerite
or coxa, together with a distal element or
basis. The endopodite consists of six clearly
defined segments, each of which has received
an individual name, viz., preischium, ischium,
merus, carpus, propodus and dactylus (Hansen,
1930). The exopodite is composed of a
variable number of divisions or sub-seg-
ments, while the coxa often bears an outer
lobe or exite (commonly termed an epipodite).
Both coxa and basis may ea*"!! develop an
inwardly directed masticatory process or
eiidite (often known as a gnathohase).
The recent works of Hansen (1930), Snodgrass (1935) and of
Ewing (1928) have contributed to the better understanding of the
morphology of insect appendages. Much confusion has prevailed
in the past owing to the neglect of evidence afforded by the
musculature, with the result that there has been a great diversity
of opinion with respect to the homologies of many of the parts
concerned. The extensive researches of Hansen form a valuable

Fig. 10. Diagram of a
complete biramous
crustacean appen-
dage, p, precoxa ; c,
coxa ; 6, basis ; i,
preischium ; i' , is-
chium; m, merus ;
ca, carpus ; pr, pro-
podus ; d, dactylus ;
en, en, endites ; ep,
epipodite ; ex, exo-
podite.

THE THORACIC APPENDAGES

23

contribution to the morphology of arthropod appendages in
general, but this author, unlike Snodgrass and Ewing, has not
concerned himself, other than incidentally, with their musculature.
A. Thoracic Appendages. In attempting to trace the homologies
of the parts of the insect leg, the best starting point is the thoracic
limb of a primitive Thysanuran such as Japyx or Campodea
(Fig. 11). The leg is seen to be a uniramous appendage and it is

Fig. 11. Second leg of Campodea. c, coxa ; e, extensor of tibia ; et,
extensor of femur ; /, femur ; fc, flexor of claws ; ff, flexor of
femur ; ft, flexor of tarsus ; fti, flexor of tibia ; It, levator of
trochanter ; p, pretarsus ; r, r, rotators of coxa ; s, subcoxa;
t, tarsus ; ti, tibia ; tr, trochanter. (From Ewing.)

generally agreed that it represents the protopodite and endopodite
of a generalised crustacean limb. Examining such an appendage
at its distal extremity, it will be observed that the apex of the
tarsus bears a somewhat complex- structure or pretarsus (trans-
tarsus of Hansen) which is composed of the claws and their
associated basal parts. The pretarsus is regarded as the
homologue of the dactylus which has become highly modified
so as no longer to show the character of a simple, claw-like
segment. The variations in form and structure of the claws,
and other parts of the pretarsus, found among insects and their

24 SOME ASPECTS OF MORPHOLOGY

larvae, are familiar to entomologists. Most Thysanura exhibit a
biungulate condition, the two claws being secondary formations
situated on the base of the pretarsus. The claw of the original
dactylus is retained in Japyx in a greatly reduced condition,
while in Campodea it has either disappeared or is represented by
a plumose arolium. In Lepisma, on the other hand, the feet are
three-clawed owing to the retention of the claw of the dactylus
between the paired or secondary claws. A simple, dactylus-like
claw is retained in most Myriapoda, some Protura and among
many insect larvae. In the great majority of insects the original
dactylus is represented only by the base of the pretarsus, which is
partially invaginated within the tarsus. Among both insects and
myriapods only a single muscle is associated with the pretarsus.
This muscle, the flexor (depressor) of the claws, arises from the
tibia, or from the tibia and femur, and, traversing the length of the
tarsus, is inserted on a long tendon which is attached to the
tractor plate of the pretarsus. The extensor (levator) muscle is
never developed in these two classes, whereas in Crustacea both
flexor and extensor muscles are evident, as well as in most of the
Arachnida.

The tarsus, in its simplest form, is an undivided segment, well
seen in the Symphyla, Japyx, Protura and many larvae. There is
little doubt that it is the homologue of the propodus of the
biramous limb. Its division into segments, so characteristic of
the Pterygota, is evidently a secondary condition possibly acquired
to ensure greater flexibility. In whatever condition the tarsus is
present, it never possesses intrinsic muscles : it is movable solely
by means of either a flexor muscle, or by both flexor and extensor
muscles, attached to its proximal border. The origins of these
muscles are variable ; in the Thysanura and the Symphyla they
are located in the tibia, or in the femur and the trochanter, whereas
in the Pterygota they commonly originate in the tibia.

The tihia or carpus is similarly movable by flexor and extensor
muscles which arise within the femur to be inserted in association
with the margin of the base of the tibia. The femur or merus
varies greatly in its degree of development. Thus in polypodous
forms, like the Symphyla and other Myriapoda, it is commonly

THE THORACIC APPENDAGES

25

t

a small segment only, but, with the development of the hexapodous
condition, it acquires greater relative size and importance. In
the Symphyla and other of the lower arthropods the femur is
movable by muscles originating within the trochanter or coxa.
Among the Insecta it is usually only provided with a reductor
muscle whose origin is in the trochanter : in cases where there
is no movement between the femur and trochanter this muscle is
undeveloped.

The homologies of the next three leg-segments, viz., trochanter,
coxa and subcoxa, have been productive
of much diversity of opinion. In some
insects, e.g., MachiUs, certain Coleopterous
larvae such as Dytiscus and Philonthus
(Fig. 12), together with the nymphs and
imagines of Odonata (Fig. 13, C), the
trochanters are double. The condition in
the Odonata has been investigated by
Snodgrass (1927), who has pointed out
that the reductor femoris muscle arises in
the second trochanter and is attached to
the base of the femur. It is evident,
therefore, that the first trochanter is with-
out muscles and simply functions as the
articulatory sclerite between the coxa and
the second trochanter. Hansen has shown
that the reductor femoris muscle is similarly
located in the second trochanter in the larva
of Dytiscus. The few writers who have

studied these parts, viz., Verhoeff (1903), Snodgrass (1927) and
Hansen (1930), see in the divided trochanters a primitive condition.
Hansen homologises them with the preischium and ischium of
such Crustacea as the Pericarda and Anaspidacea. He further
comments upon the fact that these two parts are fused in the
majority of the Insecta, as in most Decapod Crustacea. In view
of the remoteness of the affinity between insects and crustaceans,
Hansen's conclusions are unlikely to fyid general acceptance. It
is extremely unlikely that preischium and ischium would retain

Fig. 12. A. Dytiscus
sp. larva. Part of
first left leg showing
divided trochanter, t.
B. Philonthus ceneus
larva. Part of third
left leg showing
divided trochanter, t.
c, apex of coxa. (From
Hansen, 1930.)

26

SOME ASPECTS OF MORPHOLOGY

their individuality in such a highly evolved class as the insects,
especially in view of the fact that they are only developed as
separate sclerites among some of the more generalised Mala-
costraca.

On turning to the Chilopoda it will be noted that two segments
similarly intervene between the coxa and femur — these are the
trochanter and prsefemur of Hansen or the first and second

F F ^ ,5

Fig. 13. Structure and musculature of the coxa, trochanter and
base of femur. A, diagram of typical musculature of coxa,
trochanter, and base of femur ; B, trochanter and base of femur
of an ichneumonid (Megarhyssa), showing basal subdivision
(F^) of the femur (F) ; C, trochanter and base of femur of
middle leg of dragon-fly nymph (Aeschnidae), showing double
structure of the trochanter (iTr, 2Tr). Be, basicosta of coxa ;
be, basicostal suture ; Bex, basicoxite ; Cx, coxa ; F, femur ;
F^, basal subdivision of femur ; /, posterior coxo-trochanteral
articulation ; g, basicosta of trochanter ; h, ridge between
subdivisions of trochanter ; i, basicosta of femur ; j, femoral
ridge setting off basal subdivision of femur ; O, levator muscle of
trochanter ; P, thoracic branch of depressor of trochanter ;
Q, coxal branch of depressor of trochanter ; R, reductor muscle
of femur ; *S', levator of tibia ; T, depressor of tibia ; Tr, tro-
chanter ; ITr, 2Tr, first and second subdivisions of trochanter.
(After Snodgrass.)

trochanters of other authors. The possibility that these parts
are homologous with the divided trochanters of the Odonata
should not be overlooked, and in Lithobius the first trochanter is
greatly reduced, as in the Insecta.

In many of the parasitic Hymenoptera it is well known that the
trochanter appears to consist of two segments, but comparative
study over a series of genera reveals the fact that the second
segment is merely the separately differentiated proximal end of the

THE THORACIC APPENDAGES 27

femur. In many cases the separation is complete, but in numerous
other instances the so-called second trochanter is only demarcated
from the rest of the femur by a weak suture or a slight constriction
(Fig. 14). This conclusion is borne out by the relations of the
reductor femoris muscle which is inserted on an inflexion from the
base of the second segment of the trochanter : furthermore, the
depressor muscle of the tibia passes through this segment and is
attached anteriorly close to the insertion of the muscle just
mentioned (Fig. 13, B).

The homologies of the coxa now require consideration. Snodgrass
believes that the insect coxa is the homologue of the part bearing
the same name in Crustacea and that the first trochanter of the

Fig. 14. Basal segments of fore leg of a Cynipid, Rhoptromeris eucera,
female, showing rudimentary second trochanter (<i). c, coxa ;
t, true trochanter ; /, base of femur.

Odonata represents the basis which is otherwise wanting among
insects. Hansen sees in the insectan coxa the counterpart of the
basis, while the trochanter he believes corresponds with the true
or primitive coxa of the biramous limb. The term subcoxa is
not employed by Hansen, who uses the name pretrochantin
(precoxa) for the sclerite in question. A great difficulty in this
connection is the complex and extremely variable condition
presented by the small basal leg-«clerites in different insects.
The subject is discussed by Snodgrass (1927, 1935), who maintains
that in the primitive insect leg a single proximal sclerite the
subcoxa (praecoxa of Hansen) intervened between the coxa and the
thoracic wall. Further discussion of this subject is given in the
next section dealing with the pleura. All that need be stressed
here is that while the homologies of the four distal segments of

28

80ME ASPECTS OF MORPHOLOGY

v>

tc

^^

^

i

'>.

5

1

C

ce

V

Q

^

;S

2

T!

CC

to

o

=J

-3

P<

03

o

kl

u,

ce

^

H

^

73

1

.2

cS

S

5

2

H

H

^

j^

s

3

(4

S

l2

«

s

S

^

^

(H

u

«^

o

^

s

C

s

te

ce

_s

j=

o

"o

2

o

o

o

Ui

^

H

H

•d

t3

(M

M

t4

<D

■f^

c

cj

^

O

1

£

H

^

3

j_^

SJ

-t^

C

rt

§

g

ffl

ce

H

o

O

^

fi

eS

c

><

<A

p

^

o3

X

O

H

o

*^

1
1

i

cS

o

2
1

1

1

Pi

'^

w

tc

J5

H

O

CO

-<

the insect leg are tolerably clear, those of the
proximal parts are highly involved. It is by no
means certain whether any exact serial comparison
with the generalised biramous appendage is possible
in view of the evolutionary changes that have
supervened. In the present stage of knowledge the
interpretation of Snodgrass is in closer accord with
the observed facts than that of other morphologists.
A comparison of the views of Hansen and Snodgrass
is given in tabular form on this page.

B. The Pleura. The origin and differentiation of
the pleuron have been much discussed by morpho-
logists in recent years, and, in this connection, the
names of Heymons (1899), Prell (1913), Crampton
(1914, 1926), Bonier (1921), Becker (1923, 1924),
Weber (1924) and Snodgrass (1929) may be specially
mentioned. It is probable that Heymons' general
contention that the pleural sclerites are in part, or
mainly, derived from the subcoxa is substantially
correct. Among the Apterygota the pleural sclerites
are so variably developed that they exhibit no
obvious indications of having been evolved along
any common line of development. As Ewing(1928)
has pointed out, in Campodea, Scutigerella and
certain Collembola they are represented by a simple
subcoxal or basal segment of the leg (Fig. 11). In
Machilis a single, although reduced, sclerite is also
present. In other cases there is no defined subcoxa,
and its place is taken by two or more sclerites
which present diverse phases of development. It
is possible, however, to group these sclerites into a
proximal, and usually arched, plate or trochantin
over the base of the coxa, and a distal plate or series
of plates forming the eupleuron. Single proximal
and distal sclerites are evident, for example, in
Acerentulus (Ewing), and in Isotoma and other
Collembola (Fig. 15). In Lepisma, Snodgrass (1927)

THE P LEU RON

29

teit

has shown that the same general arrangement holds good, except
that the trochantin itself has apparently undergone division into
two pieces. In Eosentomon germanicum, Prell (1913) has
described eight small sclerites in the eupleural region, together
with a typically arched band-like trochantin. The ill-defined
and fragmented condition of the pleurites in this species renders
it scarcely possible to homologise the eupleural plates individually
with those of other insects.

In the ^ Pterygota, notwithstanding the
diversity of the sclerites that are developed
in the pleural region, they appear to have
been evolved to some extent along a definite
line in response to the need for rigidity
occasioned by the presence of wings. As
Snodgrass remarks, they serve to brace the
tergum against the downward pull exerted
by the tergosternal muscles which function
as elevators of the wings. He advances
theoretical conclusions which suggest that
the Pterygote pleuron was derived from the j^j^. ^5
undivided subcoxa (Sex) which became
incorporated into the side wall of the body
(Fig. 16). According to his theory the
ventral part of the subcoxa has either
become membranous or has united with the
sternum {S), while the remainder has given
rise to the pleuron. In the early Pterygota
the subcoxa became divided into a narrow, arched trochantin
(Tn), bearing a dorsal articulation (a) with the coxa, and a
larger sclerite or eupleuron (Epl). The whole pleuron sub-
sequently became strengthened by an internal ridge formed by
a lateral inflexion of its wall, thus giving rise to the pleural
suture (P.S.). The areas lying anteriorly and posteriorly to
this suture become the episternum (E) and epimeron (Ep)
respectively. In the more primitive insects the trochantin
remains as a separate sclerite in front of the suture, while,
posteriorly, it becomes merged into the epimeron. The

Prothorax
and leg of a Col-
lembolan (Isotoma).
ex, coxa ; fe, femur ;
h, base of head ; pt^
pretarsus ; sex, sub-
coxa ; terg, tergum ;
ti, tibia ; tr, tro-
chanter. (After
Ewing.)

30

SOME ASPECTS OF MORPHOLOGY

sclerotised parts of the subcoxa, situated anteriorly and
posteriorly to the coxa (Cx), form the pre- and post-coxal bridges
(P.B., P'.B.) and the ventral part of that sclerite may form, with
the lateral edge of the sternum, a defined element termed the
laterosternite (Ls). This, in few words, expresses the salient points
in Snodgrass's views. In order to instance how hypothetical the
problem is, it may be mentioned that Hansen (1930) rejects the
theory as a whole and homologises the eupleuron and trochantin
of Snodgrass with the subcoxa (precoxa) and coxa (qoxopodite)
of Malacostraca ; he further maintains that the insect coxa is the

Scx^

ABC

Fig. 16. Simplified diagrams representing the possible origin of the
Pterygote thoracic pleurites from the subcoxa, according to
Snodgrass (1929). A, hypothetical primitive condition ; B,
intermediate condition ; C, stage with a differentiated pleuron ;
a and h, dorsal and ventral articulations of coxa (Cx) ; c,
secondary articulation with trochantin ; W.P., pleural wing
process. For explanation of other lettering, vide pp. 29 and 30,

homologue of the basipodite. The subcoxal origin of the pleuron
is dismissed by him, and it would appear that he regards the
episternum and epimeron as independently developed sclerites.

The foregoing conception of the origin of the insect pleuron is
substantiated by following the development of the parts concerned
among the lower Pterygota. Thus in the cicada Tihicina the
original subcoxa shows in the nymph differentiation into rudi-
mentary pleurites with the beginning of a demarcation with a
small basal sclerite or trochantin. Roonwal (1937) has followed
the development of the parts in question in Locusta and has
demonstrated that the subcoxa is the fundament from which the
future pleuron is derived.

THE PRIMITIVE LEG

31

C. The Primitive Insect Leg. It will be gathered Iroiii the
foregoing remarks that the primitive inseet leg is composed of the
following parts (Fig. 17). An undivided basal segment or subcoxa
articulated distally with the coxa, the latter segment being hinged
to the femur by means of the trochanter. Distally the femur
articulated with a slender tibia, which in
its turn carried an un jointed tarsus. At
its apex the tarsus terminated in the
pretarsus, which included a single median
claw\ From this primitive type it appears
probable that the legs of all insects have
been derived by various processes of modi-
fication. In certain of the Thysanura the
primitive type is retained in an almost
unaltered condition ; in others the subcoxa
becomes reduced and divided, while the
tarsus may exhibit indications of imperfect
segmentation. In the Collembola the most
important features are the functional sub-
coxa, a double trochanter in certain genera
and the general absence of a tarsus.
Among the Pterygota the subcoxa has
disappeared as a definite entity and has
become modified and incorporated in the
thoracic wall to form the pleuron. The
coxa, on the other hand, has for the most
part assumed increased importance, but
among Plecoptera it may be greatly
reduced or even vestigial. The trochanter may undergo secondary
division (Odonata) or receive a spurious additional joint formed
from the proximal end of the femur (parasitic Hymenoptera).
The femur remains as an independent segment, but in most
coleopterous larvae the tibia fuses with the tarsus. The last-
mentioned segment is extremely variable, and most commonly
becomes divided into five subsegments. The claws are frequently
single in the larvae, but paired in adult insects.

The more or less elongated femora and the small trochanters of

Fig. 17. The segments
of a primitive insect
leg. C, coxa ; F,
femur ; PT, pre-
tarsus ; S, subcoxa ;
T, trochanter ; TA,
tarsus : TI, tibia.

32 SOME ASPECTS OF MORPHOLOGY

the Insecta are features which seem to be correlated with
hexapodous locomotion. In the Symphyla (Fig. 18) the same
components of the leg are present as in the Insecta, but, on the
other hand, in the first-mentioned class the femora are very
short and the trochanters large and elongated. These same two
features are also betrayed in Lithobius, Scolopendra and other
Chilopoda, where they are correlated with polypodous locomotion
in these animals. Once an hexapodous condition is assumed it

dc

Fig. 18. Antepenultimate leg of Scutigerella. d, depressor of
trochanter ; dc, depressor of coxa ; I, levator of coxa ; st, stylus.
Other lettering as in Fig. 11. (From Ewing.)

would appear that the trochanters lose some of their functional
importance, while the femora become larger and more powerful.

D. The Cephalic Appendages, (a) The First Maxillae. The
generalised cephalic appendages are the first maxilla,^ and conse-
quently the morphology of these latter organs is of primary
significance in the interpretation of the homologies of the
remaining gnathal appendages. The first maxilla is divisible into
a basal region of complex structure bearing a serially segmented
distal organ or palpus. It is generally agreed that the palpus
represents the ramus or telopodite of a locomotory appendage.

THE FIRST MAXILLA

33

but very diverse views have been advanced to explain the
morphology of the basal region of the maxilla. In recent years
three interpretations have been put forward, viz., those of Borner
(1909, 1921), Crampton (1922) and Snodgrass (1928). These are
briefly summarised below.

1 . According to Borner the cardo of the maxilla is the counter-
part of the subcoxa. and the stipes the homologue of the coxa,

Fig. 19. Maxilla of Periplaneta. A, left maxilla, posterior (ventral)
surface. B, internal surface of cardo. C, right maxilla, anterior
(dorsal) view, showing muscles. Cd, cardo ; e, articulation of
cardo with cranium ; fga, flexor of galea ; flee, cranial flexor of
lacinia ; flcs, stipital flexor of lacinia ; Ga, galea ; /, promotor
of cardo ; KLed, adductor of cardo ; KLst, adductor of stipes ;
Le, lacinia ; O, levator of palpus ; Pip, palpus ; Q, depressor
of palpus ; q, sub-marginal suture (and internal ridge) near
inner margin of stipes ; r, internal ridge of cardo ; St, stipes ;
T, depressor of fourth segment of palpus ; V, depressor of
fifth segment of palpus. (After Snodgrass.)

while the galea and lacinia are coxal endites. The palpifer is a
subdivision of that part of the stipes ^which bears the palpus and
galea ; it is, therefore, not a true segment.

2. According to Crampton the cardo represents the coxa, and
the stipes is the homologue of the basis, while the galea and
lacinia are endites belonging respectively to the stipes and palpifer.
In this interpretation the palpifer is a true segment and represents
the first segment (or ischium) of the palpus.

R.A. ENTOMOLOGY, 2

34 SOME ASPECTS OF MORPHOLOGY

3. The theory of Snodgrass is more in accordance with that of
Borner than of Crampton. According to him the cardo and
stipes, however, are not separate segments, but represent the
secondary division of a single original basal segment. The galea
and lacinia are endites of this basal segment and the palpifer is a
secondarily demarpated portion of the stipes as maintained by
Borner.

The theoretical views of Snodgrass are largely formulated from
a study of the basal musculature of the maxillae (Fig. 19), and in
this connection it needs to be borne in mind that such muscles
may not necessarily retain their original functions or their former
dispositions. In accordance with alterations in the basal
articulation of an appendage, and in the number of its constituent
segments, the muscles concerned may shift their positions or
become segregated into secondary groups of fibres. Comparative
studies of progressive changes in the muscles are greatly needed
before we are in a position to ascribe to such criteria their due
morphological values. Snodgrass 's contribution is important in
that it discountenances attempts to determine homologies solely
upon the basis of a study of the sclerotised parts, without reference
to other considerations.

Since the margin of the basal cavity of the maxilla in the more
primitive insects surrounds the bases of the cardo, stipes and
lacinia, and the tergal and sternal muscles of the maxilla are
distributed to these parts, Snodgrass concludes that the latter are
all differentiations of a single primitive basal segment. The galea
and lacinia he regards as subdivisions of a single endite carried by
the stipes, since they are flexed by muscles having their origin in
the latter sclerite. In supporting Borner's contention that the
palpifer does not represent an independent segment, he lays stress
on evidence afforded by the basal palpal muscles. These muscles,
which move the palp as a whole, have their origin within the
stipes and pass directly through the palpifer to be inserted with
the proximal segment of the palpus ; furthermore, no muscles
arise within the palpifer or are inserted upon it.

While the study of the completed maxillae and their musculature
affords scope for more exhaustive investigation, the differentiation

THE LABIUM

35

of the parts involved during the later phases of development
remains to a large extent unexplored. Evidence afforded by the
musculature is unquestionably important, but, for reasons already
stated, too exclusive reliance upon this one criterion is undesirable.
It is more especially on this account that Borner's contention that
the cardo is homologous with the thoracic subcoxa remains an
open question. Furthermore, additional evidence is necessary,
perhaps more especially from the
developmental aspect, before
Crampton's view that the palpifer is
a reduced segment bearing the galea
as an endite can be regarded as
untenable.

(h) The Labium (Second Maxillce).
Evidence, afforded by development
and by comparative morphology,
provides proof that the labium is
formed by the fusion of the second
maxillae. The correspondence of the
free distal parts of each half of the
labium with similar parts of the first
maxillse is undisputed. Consequently
it is agreed that the palpi are homo-
logous in the two cases, and that the
paraglossae and glossae are the respec-
tive counterparts of the galea and
lacinia of the first maxillae. In so
far as the basal labial sclerites are
concerned, the homologies are prob-
lematical, and it has been too
readily assumed that the mentum and submentum are formed
by the simple fusion of the stipites and cardines respectively.

A comparative study of primitive types points to the conclusion
that the labium is composed primarily of a distal region or
prementum which is demarcated by means of the labial suture
from a proximal or basal region — the postmentum (Fig. 21). The
prementum may be defined as the region bearing the ligula and

2—2

pm<

Fig. 20. hahium of Mastotermes.
pm, postmentum. For ex-
planation of other lettering,
see Fig. 21.

36

SOME ASPECTS OF MORPHOLOGY

palpi, and from which the muscles of these parts take their origin.
In its most generalised condition it is a paired structure, as is
seen, for example, in the Machilidse, Blattida?, Plecoptera and
Isoptera (Fig. 20). It is probable that these paired components
are the equivalents of the stipites of the first maxillae. Further-
more, the muscles of the prementum appear to be directly
comparable with those of the palpi and the terminal lobes of the

.Is

t

Fig. 21. Diagram expressing the origin of the labium from a pair of
united maxillae. Left, generalised labium. Right, theoretical
union of second maxillae incorporating, in part, the labial
sternum, c, cardo ; //, flexor of lacinia ; fl^, flexor of glossa ;
fg, flexor of galea ; fg', flexor of paraglossa ; g, galea ; g', para-
glossa ; /, lacinia ; /', glossa ; Is, labial suture ; o, levator of
palpus ; J), palpiger ; jaZ, palpus ; q, depressor of palpus ; r,
median or stipital refractors ; ,s', labial sternum. The broken
line (Is) leading to the labial suture divides the labium into
a proximal element or postmentum and a distal element or
prementum. (Adapted from Snodgrass, 1931.)

first maxillae. The postmentum is, in many insects, subdivided
in mentum and suhmentum : much importance, from the morpho-
logical standpoint, has been given to these two subdivisions in
the past, but it now appears probable that their development is
solely due to a secondary division of an original single and entire
postmentum. The only muscles arising from the postmentum
are the median retractors of the labium : neither the mentum
nor submentum develops any muscles other than those alread}^
evident in relation with the postmentum. These muscles, further-

THE LABIUM 37

more, undergo no modification whether the latter region be in its
undivided or subdivided condition.

The homologies of the basal region of the labium (postmentum
or mentum plus submentum, as the case may be) have given rise
to considerable discussion. The lateral articulation of the basal
angles of this plate with the margin of the cranium, in Orthopteroid
insects, shows considerable similarity with the articulations of
the cardines of the maxillae. Snodgrass (1935) maintains that the
postmentum corresponds with the cardines of the first maxillae
which have become completely amalgamated with the primitive
sternum of the labial segment (Fig. 21). Other writers have
claimed that the postmentum, or at least the submentum, is
wholly a sternal derivative. The recent studies of Roonwal
(1937), on Locusta, show that the postmentum contains no sternal
component. If this conclusion be substantiated it would appear
that the postmentum has been formed entirely by the fusion of the
originally separate parts comparable with the cardines of the
first maxillae. Those who wish to follow recent work on the
morphology of the labium should consult the writings of Snodgrass
(loc. cit.) and Walker (1931).

Reference needs to be made to the gula, which is most usually
developed in insects in which the heads are prognathous (Walker,
1932). In such cases the ventral (or, morphologically, the posterior),
side of the head becomes lengthened by an extension of the post-
genal areas. The result is that articulations of the maxillae and
labium become carried forwards so as to become separated by a
more or less variable distance from the occipital foramen (Fig. 22).
It is in the intervening area thus formed that the gula is developed.
The researches of American morphologists, notably Crampton
(1921, 1928), Stickney (1923) and Snodgrass (1928), have done
much to establish the morphological identity of the gula upon a
sound basis. It is more frequently developed in Coleoptera than
in any other order, but it also occurs in Neuropterous larvae.
The gula is a sclerite which has become demarcated by the forward
migration of the tentorial pits, or mouths, of the invaginations
of the posterior arms of the tentorium (Fig. 22). According to
Snodgrass, the ventral parts of the post-occipital sutures terminate

38

SOME ASPECTS OF MORPHOLOGY

around the tentorial pits, but when the latter pass forwards the
sutures just named are carried with them, and are termed the
gular sutures. The median cranial area, thus delimited by these
sutures, is the gula, and it partly or wholly occupies the region
between the submentum and the occipital foramen. A graduated

series of stages, from the

B

Fig. 22. Diagrammatic figures of the
ventral region of the head illustrating
the origin and development of the
gula among Coleoptera. Excepting
the submentum, the mouth-parts are
not shown. A, generalised condition ;
B, rudimentary gula ; C, complete,
well-developed gula ; D, condition
found in weevils and a few other
Coleoptera in which the gular sutures
are confluent. e, compound eye ;
g, gula ; gp, tentorial (or gular) pit ;
gs, gular suture ; o, occipital foramen ;
OS, post-occipital suture ; s, sub-
mentum.

beginnings of the formation
of the gula up to its evolution
as a large chin plate, can
be traced among different
Coleoptera. When it is
fused with the submentum,
or its frontal margin is ill-
defined, the anterior limit of
the gula can be generally
fixed by an imaginary line
between the two tentorial
pits. Among termites, the
postmentum and gula are
fused, in the soldiers, to
form one continuous and
greatly elongated ventral
plate. In the Rhynchophora
(Fig. 22, D) and a few other
beetles, the gula is suppressed
owing to the meeting of the
two gular sutures in the
mid-ventral line of the head.
It is by no means certain,
however, whether the gula is

invariably formed in the manner thus described. Thus in the
Embioptera there is no evidence to show how the gula region in
those insects has been formed, since there are no traces of gula
sutures present {vide Walker, 1933). The so-called gula region
of Hemiptera requires investigation, especially in view of the
classificatory importance attached to it by some systematists.
(c) The Mandibles. In the Apterygota the prevailing type of

THE MANDIBLES

39

mandibles differs in important features from that of the higher
insects {vide Snodgrass, 1931, 1935). In their musculature the
mandibles of Japyx^ for example, show a very evident resemblance
to that displayed in the jaws of the Chilopoda and some of the
lower Crustacea. The mandible of Japyx is an elongated organ
which is hinged to the head by means of a single dorsal articulation.
It is moved by means of four muscles, two being dorsal and two
ventral (Fig. 23, A). The dorsal muscles are the promotor and

—ta

^za

Fig. 23. Diagram of typical A, apterygote, and B, pterygote
mandibles, a, primary articulation with cranium ; a-c,
secondary longitudinal axis of movement on cranium ; ha,
hypopharyngeal apophysis ; p, dorsal promotor ; r, dorsal
remotor ; ta, tentorial adductor ; za, zygomatic adductor.
(After Snodgrass, 1931.)

remotor muscles, while the ventral muscles comprise two groups
of fibres that are, functionally, adductors — the zygomatic and
tentorial adductors of Snodgrass. As the latter author has pointed
out, in the Pterygota (Fig. 23, B) each mandible has developed a
long axis of attachment and instead of having a single articulation
it has developed two. One of these is the condyle which represents
the single articulation of the Apterygota, while the other, the
facet or ginglymus, is a secondary acquisition. Owing to this
change in the articulation, the muscles undergo correlated modifica-
tions. The promotor and remotor muscles become respectively

40 SOME ASPECTS OF MORPHOLOGY

the abductor and adductor muscles. The ventral adductor
muscles of the Apterygota either disappear or are represented by
a greatly reduced tentorial adductor which is found in the
Orthoptera, Isoptera and the nymphs of some of the lower orders.
Since the mandibles perform their hardest work by means of
inward movement, the adductor muscles are more powerfully
developed than the abductors.

(d) The Antennce. That the insect antenna is the homologue
of the crustacean antennule is generally accepted. This con-
clusion receives support from the fact that in many of the lower
insects the tritocerebrum bear evanescent appendages which are
to be regarded as the last traces of the crustacean antennae. It
naturally follows that at some remote stage in their evolution the
insects and the myriapods, along with them, were derived from
some early crustacean form, otherwise the occurrence of these
tritocerebral appendages is left unexplained and cannot be
accounted for.

The scape of the insect antenna is the protopodite. and in the
majority of insects the muscles which move the flagellum as a
whole are located in this segment. In the Entognatha, however,
researches by the present writer, not at present published, show
that in the Entognatha among Thysanura a very different con-
dition prevails. The moniliform antennae of these primitive
insects bear muscles in each individual segment and a similar
arrangement prevails in the Symphyla and certain other myriapods.
The relations of the multiannulate antenna of Machilis and
Blatta, for example, with the segmented antenna of Japijx is
being studied, but insufficient data are at present available to
justify further discussion on the subject for the time being.

CHAPTER II

SOME ASPECTS OF MORFUOLOGY—continued

E. The Abdominal Appendages, p. 41 . 4. The Genitalia, p. 43. The
Female, p. 43 ; The Male, p. 45 ; Literature, p. 49.

E. The Abdominal Appendages. The frequent retention of
abdominal appendages in insect embryos is interpreted as a
survival of primitive organs inherited from polypodous ancestors.
In the larvae of Lepidoptera, and of some other orders, certain
of these appendages develop into functional organs, while the
remainder become resorbed. In the imagines a variable number
of abdominal appendages are retained in the Apterygota, but
among the Pterygota their sole
survivals are represented in parts of
the genitalia (vide p. 43) and by the
cerci.

In the Machilida?, among
Thysanura, the abdominal sterna are
more or less triangular shields, while
in Nicoletia each is divided into an
anterior and posterior s c 1 e r i t e
(Fig. 24). In close contact with each
sternum is a pair of lateral plates
or coxites which frequently meet along the median line. The
latter plates are believed to represent the basal parts of true
appendages whose distal parts have atrophied. They are termed
coxites for the reason that it is maintained they represent greatly
flattened coxae, but, in some cases at least, possibly other of the
basal segments may have become incorporated with them. In
Japyx and related Thysanura, the coxites are completely fused with
the sternum, and the entire plate thus constituted is a coxosternum.

41

Fig. 24. Nicoletia (Anelpistina)
wheeleri. Third abdominal
sternum s, s^ ; c, coxite ; st,
stylus. (From Silvestri.)

42

SOME ASPECTS OF MORPHOLOGY

In the Machilidae the second to ninth pairs of coxites each bear
styU, but in other famiUes the styh are more restricted, and in the
Lepismidse they may be present only in relation with the eighth
and ninth pairs of coxites ; in the Japygidae they are borne on
the first to seventh coxosterna.

In the Protura the first three abdominal segments bear greatly
reduced appendages, which, in the Eosentomonidae, are two-
jointed. These appendages are comparable with coxites which,
in the two-jointed condition, include the rudiment of a second
leg-segment (Fig. 25). The appendages of the first abdominal
segment in the Campodeidae appear to be of a comparable nature.
In the CoUembola true abdominal appendages are represented by

the ventral tube, the hamula
and the furcula which are
derived by direct conversion
from rudiments borne on the
corresponding segments in the
embryo.

The morphological signifi-
cance of the styli has been so
Fig 25. Eosentomon. Third abdominal ^^^^^ discussed that only brief
sternum s, showmg| two-segmented /

appendages^; c, coxite ; c^ second reference to the subject is

segment of appendage.
Berlese.)

(From required here. It appears
probable that the styli which
are borne upon a variable number of the abdominal segments in
Thysanura (Fig. 24) are homologous with similar organs present
in the relation with the limb bases in Symphyla (Fig. 18). In
two groups these styli are in similar close relations with eversible
sacs. It follows, therefore, from this homology that the styli of
Thysanura are not to be regarded as vestiges of true limbs as has
often been contended. The styli borne on the ninth sternum of
the males of the Isoptera, Ephemeroptera and of many Orthoptera
are evidently homodynamous with those just referred to.

As regards the cerci, it is well established on embryological
grounds that they are the persistent true appendages of the
eleventh abdominal segment. Each cercus is borne on the
membrane between the epiproct and paraprocts, the two latter

THE GENITALIA 43

sclerites being the remains of the sternum of the eleventh segment
(Fig. 26). In cases when the cerci apparently arise from the
tenth abdominal segment, the eleventh segment having atrophied,
examples are afforded of Lankester's sixth law of metamerism.
The atrophy of the eleventh segment has resulted in its appendages
becoming transferred to the tenth segment. The terminal pair of
abdominal appendages found on the ninth or tenth abdominal
segments of various larvae are of a problematical nature, especially
those which are devoid of articulation and arise from the dorsum
of their segment. They are probably best referred to under the
general term of cercoids in reference to their frequent resemblance
to cerci.

4. The Genitalia

During recent years much attention has been given to the
genitalia of insects, and more especially to the male organs, in view
of their importance in taxonomy. The whole subject has become
greatly involved owing to the lack of any convenient, uniform
terminology, and to uncertainties which exist with respect to the
homologies of the parts concerned in different orders. The
greater part of the literature deals with the completed organs and
especially those of the male. Ontogenetic studies have not been
pursued with the same enthusiasm. They are dependent upon
obtaining supplies of material in the proper successive stages
which are often difficult to provide. This fact, coupled with the
obscurity which so often attends the initial growth phases of the
genitalia and the difficulties of interpretation that are involved,
has resulted in a large field, so far, being inadequately explored.
Nevertheless, the writings of Singh Pruthi (1924, 1925), Mehta
(1933), Metcalfe (1932) and others have advanced our knowledge
of the developmental aspects of the subject.

The Female. It is well known that a typical, completely
formed insect ovipositor is composed of three pairs of valves
which are located in relation with the eighth and ninth abdominal
segments (Fig. 26). These valves arise in connection with the
basal plates or coxites of their respective segments, so that the
coxites of the eighth segment bear one pair of such valves, while

44

SOME ASPECTS OF MORPHOLOGY

the coxites of the ninth segment bear two pairs. The anterior
and posterior valves are outgrowths of the medio-lateral margins
of their respective coxites and are regarded as secondarily
developed formations known as the gonapophyses. The lateral
pair of valves borne by the ninth segment on the other hand
represent the greatly drawn out coxites and are therefore the
modified bases of an original pair of abdominal appendages. The
styli, however, are very rarely retained : they are present, for

Vlll IX

Fig. 26. Diagram illustrating structure of abdomen and morphology
of the ovipositor, c, cercus ; e, epiproct; g, g^, first and
second gonapophyses ; /, limb base (coxite) ; la, lamina supra-
analis ; la^, lamina subanalis ; p, paraproct ; s, sternum ;
sp, spiracle ; st, stylus ; t, tergum ; vii-xi, seventh to eleventh
terga. (From Snodgrass, 1931.)

example, in the rudimentary ovipositor of Mastotermes (Fig. 27)
and in the more highly developed organs of some of the Odonata.
It is well known that in Machilis the anterior and posterior
valves develop alike in connection with their respective segments
and form the ovipositor. The lateral valves are undeveloped as
such and remain as unmodified coxites, bearing long styli. While
the homologies of these lateral valves have been believed to
remain constant in different orders, observations by Nel (1929),
on the development of the ovipositor in Locustana, Colemania
and Blatella, appear to indicate that the anterior valves are
modified coxites like the lateral pair, and are consequently not

MALE GENITALIA

45

homologous with the inner valves. Also, Walker's figures (1919)

of the developing ovipositor of the grasshopper Conocephalus are

suggestive of a similar homology. On the other hand, careful

ontogenetic studies are needed from the first instar onwards in

order to ascertain whether or not any change in the relative

positions of the valves in question takes place during growth.

The study by Denny (1897) of the ovipositor of Blatta orientalis

shows that the anterior and posterior valves are homologous

parts, the lateral valves alone

being derived from coxites. It

would seem remarkable if, in

such closely allied genera as

Blatta and Blatella, the anterior

valves of the ovipositor should

develop in a different manner in

the two cases, and it is evident

that further research is necessary

before any conclusions to the

effect that the lateral ovipositor

valves are of different origin

among insects belonging even to

the same major group.

The Male. In the males of
those insects whose genitalia are
fully developed two sets of
organs, forming the completed
parts, are present. These organs
arise from rudiments which
develop in relation with the ninth sternum of the abdomen. The
outer organs are known as claspers, harpagones or valves and are
the functional organs used for grasping the abdomen of the female
during coitus. These organs will be considered first. They are
commonly regarded as developments of the coxites of their
segments, but Snodgrass (1935, 1936) has claimed that the balance
of evidence is in favour of their being regarded as modified styli
rather than coxites (Fig. 28). As this morphologist points out,
the claspers are movable lobes which, in some cases, are supported

Fig. 27. Ventral terminal abdominal
segments of a female soldier of
Mastotermes. c, coxite ; c', base of
cercus ; g, ventral or anterior valve
of ovipositor ; g\ intermediate or
posterior valve of ovipositor ; pa,
paraproct ; S^-S^, seventh to
ninth sterna ; s', stylus or dorsal
(lateral) valve or ovipositor ; tj^^,
tenth tergum. (Adapted from
Crampton.)

46

SOME ASPECTS OF MORPHOLOGY

on definite basal plates, or coxites, while in others the coxites are
amalgamated with the ninth sternum to form a compound plate
or coxosternum. The claspers are each provided with muscles
arising either from the coxites or from the coxosternum, as the
case may be, and inserted into their bases. It appears probable
that in cases where the coxites are fused with the adjacent sternum
the styli alone form the claspers, as Snodgrass maintains. In
other cases where the coxites are free, they themselves may be
movable by muscles arising from the sternum of their segment,

Fig. 28. Morphology of male external genitalia. A, generalised
condition. B, specialised condition (penis and parameres
united), a, aedeagus ; c, coxite or limb base ; c^, fused coxites ;
m, muscle of stylus ; p, parameres ; pe, penis ; s, ninth sternum ;
st, stylus. (From Snodgrass, 1931.)

and in such cases the functional clasping organs are formed by
the coxites together with their styli.

Among Orthoptera the claspers are not developed. In
Grylloblatta, Walker (1922) has shown that the coxites are free
and bear but little modified styli at their apices (Fig. 29). This
condition is evidently a primitive one and does not differ in
essentials from that prevalent in the Machilidge among the
Thysanura. In other Orthoptera the coxites are fused with the
sternum to form a coxosternum : in the Blattidae and some
other families unmodified styli are retained, but they perform no
copulatory function. The same condition is found among
Isoptera. Among the Ephemeroptera, Mecoptera and Trichoptera
the coxites are large and free and these plates, along with their
styli, are to be regarded as forming the functional clasping organs.

MALE GENITALIA

47

In most of the Lepidoptera and in some of the Diptera the only-
movable organs are formed by the modified styli. In the
Lepidoptera the coxosternum is termed the vinculum, while the
claspers are often known as harpes {vide Mehta, 1933). Among
Diptera, Snodgrass has shown that in the lower forms (Tipulidse)
the coxites and styli are both separate and distinct, but more
commonly the coxites are fused with their sternum. In the
Homoptera, Singh Pruthi (1924,
1925), George (1929) and
Metcalfe (1932) have shown
that claspers are undeveloped
as such since their rudiments
come to form the subgenital
plates. In Heteroptera, Snod-
grass (1935) states that claspers
are present, but the ontogenetic
studies of Christophers (1922)
seem to point to the conclusion
that the organs in question are
parameres. The true condition
of affairs in the Hemiptera as a
whole requires thorough investi-
gation and generalisations either
from the study of the completed Fig. 29.
organs alone or from an onto

Grylloblatta, male ; ventral
view of terminal segments, c, c,
coxites ; e, eversible sac ; r, I, right
genetic study of one or two and left lobes of paramere ; s, s,

types only are not trustworthy. f^y^i ' ^9' ^1"*^ abdominal sternum.

^ (From Walker.)

In Coleoptera, claspers are

wanting : in Tenebrio, a pair of evanescent papillae are recognised

by Singh Pruthi (1924a) as being vestiges of these appendages.

Among the Hymenoptera claspers are usually regarded as being

present, although Snodgrass concludes that they are seldom to be

found in that order.

The penis and its associated parts form the inner organs of the

male genitalia. A number of workers have shown that a pair of

lobes arise in a median position in regard to the ninth sternum.

These lobes develop as apparent outgrowths near the hind border of

48

SOME ASPECTS OF MORPHOLOGY

that segment or within a genital cavity or pocket which surrounds
the gonopore. In either case, the lobes in question ultimately
fuse to form the penis. This relatively simple condition has
been described by Metcalfe (1932a) in Gastroidea and Anihonomus
among Coleoptera and by Mehta (1933) in Pieris and some other
Lepidoptera. In other cases these lobes undergo subdivision,
thus giving rise to a pair of outer components or parameres and an
inner pair which unite to form the penis. This condition has been

Fig. 30. Penis (p) and parameres (pa) of A, Pedetonius palcearcticus
Silv. (Machilidae). (After Silvestri.) B, Nicoletia wheeleri Silv.
(Lepismidae). (After Silvestri.) C, Ephemeroptera (Callibcetis).
(After Walker.) D, Dermaptera (Anisolabis). (After Walker.)
(c, coxite ; s, stylus ; st, ninth sternum.)

shown to obtain in certain Coleoptera (Singh Pruthi, 1924 ;
Metcalfe, 1932a). Earlier observers have also described the same
state of affairs in some Lepidoptera as well as in Orthoptera,
Hymenoptera and Diptera. It is generally accepted that the
penis was originally a paired organ, and this generalised state of
affairs is retained among Ephemeroptera (Fig. 30, C) and
Dermaptera (Fig. 30, D). Among generalised members of the
last-mentioned order the parameres, however, instead of being
free, have become mere outgrowths of the penes. In the Thysanura

LITERATURE 49

Nicoletia (Fig. 30, B) the penis is slightly bifurcate distally and
apparently thus still retains traces of its former double origin.

With regard to the homologies of the genitalia in the two sexes,
the present state of knowledge does not shed much light on the
subject. The claspers of the male, however, appear in a general
way to be the counterpart of the lateral valves of the female
ovipositor. When the coxites actually participate in the formation
of the claspers they are homologies with the ovipositor valves
just referred to. On the other hand, the styli, which are so
important a part of the male genitalia, are, except in a few rare
cases, no longer evident in the female. With regard to the inner
parts of the male genitalia, it is not possible, on the evidence at
present available, to discuss the subject to any advantage or
arrive at a definite conclusion respecting their homologies.

Literature

I. General

HosKiNS and Craig, 1935, Physiol. Rev., XV., 525.

Imms, 1936. Trans. Soc. Brit. EnU, III. Also " Nature," March 9th, 1937.

Snodgrass, 1935. '' Principles of Insect Morphology." New York and

London.
Weber, 1933. " Handbuch der Entomologie." Jena.
Wiggles WORTH, 1934. " Insect Physiology." London.

II. Segmentation of the Head

Eastham, 1930. Phil. Trans. Roy. Soc. B., CCXIX., 1.
FoLSOM, 1930. Bull. Mus. Comp. Zool. Harvard, XXXVI., 87.
Hansen, 1930. '* Studies on Arthropoda," III. Copenhagen.
Hanstrom, 1927-30. Zeitschr. Morph. Okol. Tiere, MIL, 543; XL, 152;
XIII., 329 ; XIX., 732.

1928. " Vergleichende Anatomic des Nervensystems der wirbellosen
Tiere." Berlin.
Manton, 1928. Phil. Trans. Roy. Soc. B., CCXVI.
Mellanby, 1936. Quart. Journ. Mic. Sci., LXXIX., 1.
Philiptschenko, 1912. Zeits. wiss. Zool., CIIL, 519.
Roonwal, 1937. Phil. Trans. Roy. Soc. B., CCXXVII.
Silvestri, 1932. V'^ Cong. Internat. d'Ent.^ Paris, 329.
Snodgrass, 1928. Smithsonian Misc. Coll., 80, No. 1.

1935. " Principles of Insect Morphology." New York and London.
Uzel, 1897. Zool. Anz., XX., 232.
Wiesmann, 1926. Inaug. Diss. Jena.

III. Wing Venation

CoMSTOCK, 1918. " The Wings of Insects." Ithaca, N.Y.

CoMSTOCK and Needham, 1898-99. Amer. Nat., XXXII., 43 ; XXXIII., 118.

Handlirsch, 1925. In Schroder's Handbuch der Entomologie, III., 117.

50 SOME ASPECTS OF MORPHOLOGY

Lameere, 1922. Bull. Clas. des Sci., Acad. Roy. Belg., Ser. 5, VIII., 138.
Redtenbacher, 1886. Ann. des K.K. nat. Hofmus., Bd. I., 153.
Tilly ARD, 1919. Proc. Linn. Soc. N.S.W., XLIV., 533.

1925. Anier. Journ. Sci., IX., 328.

1926. " Insects of Australia and New Zealand." Sydney.
ViGNON, 1929. Arch. Mus. Hist. Nat. Paris (6), IV., 89.

1932. EncycU Ent., Diptera, VI., 133.
ViGNON and Seguy, 1929. C.R. Acad. Sci., CLXXXVIIL, 1699 ; Bull. Soc.
Ent. France, XIV., 227.

IV. Metameric Appendages

Becker, 1923-24. Zool. Anz., LVIL, 137 ; LX., 169.
Borner, 1909. Zool. Anz., XXXIV., 100.

1921. In Lang's Ilandbuch der wirbellosen Tiere, IV., 649.
Crampton, 1914. Zool. Anz., XLIV., 56.

1921. Ann. Ent. Soc. Amer., XIV., 65.

1922. Proc. Ent. Soc. Washington, XXIV., 65.
1926. Trans. Amer. Ent. Soc, LIIL, 199.
1928. Journ. Ent. and Zool., XXL, 1.

EwiNG, 1928. Smithsonian Misc. Coll., 80, No. 11.

Hansen, 1930. " Studies on Arthropoda," III. Copenhagen.

Heymons, 1899. Nova Acta Abh. Kaiserl. Deut. Akad. Natiirf., LXIV., 349.

Prell, 1913. Zoologica, XXV., Heft 64.

RooNWAL, 1937. Phil. Trans. Roy. Soc. B., CCXXVII.

Snodgrass, 1928. Smithsonian Misc. Coll., 80, No. 1.

1928. Ibid., 81, No. 3.

1929. Ibid., 82, No. 2.

1931. Ibid., 85, No. 6.

1935. " Principles of Insect Morphology." New York and London.
Stickney, 1923. Illinois Biol. Monog., VIIL, No. 1.

Verhoeff, 1903. Zool. Anz., XXVL, 25.

1904. Nova Acta Abh. Kaiserl. Deut. Akad. Naturf., LXXXL, 211.
Walker, 1931. Canad. Ent., LXIIL, 75.

1932. Ibid., LXIV., 223.
Weber, 1924. Zool. Anz., LX., 17, 57.

V. Genitalia

Christophers, 1922. Indian Journ. Med. Res., X., 530.
Denny, 1897. " Commemoration Volume," 169. Sheffield.
George, 1929. Quart. Journ. Mic. Sci., LXXIL, 447.
Mehta, 1933. Ibid., LXXVL, 35.
Metcalfe, 1932. Ibid., LXXV., 49.

1932a. Ibid., LXXV., 467.
Nel, 1929. Ibid., LXXIIL, 25.
Singh Pruthi, 1924. Ibid., LXIX., 59.

1924a. Proc. Zool. Soc, 857.

1925. Trans. Ent. Soc. London, 127.
Snodgrass, 1935. " Principles of Insect Morphology." New York and
London.

1936. Smithsonian Misc. Coll., 95, No. 14, 61.
Walker, 1919. Ann. Ent. Soc. Am., XII., 267.

1922. Ibid., XV., ^.

CHAPTER III
METAMORPHOSIS

Berlese's Theory, p. 51. Types of Larv^, p. 56. 1. The Protopod
Type, p. 56 ; 2. The Polypod Type, p. 58 ; 3. The Oligopod Type,
p. 60 ; 4. The Apodous Type, p. 60. General Remarks on Insect
Metamorphosis, p. 61. Nymphs and Pupce, p. 63 ; Hormones and
Metamorphosis, p. 64. Literature, p. 70.

The general features of insect metamorphosis are well known
and consequently do not require recounting. The present chapter,
therefore, will be confined to a discussion of the more recent
aspects of the subject, and the theories that have been advanced
to account for some of the phenomena associated therewith. For
many years past the compodeiform type of larva has been claimed
to be the most primitive form of immature insect, while departures
from this condition have been explained as being the result of
secondary and adaptive modifications. While there is considerable
truth in this idea, it fails as a complete explanation of insect
transformation, and leaves many problems unsolved.

Berlese's Theory

A comprehensive and largely new interpretation of insect meta-
morphosis was advanced by the late Antonio Berlese in 1913, and
was subsequently incorporated by him in his well-known treatise,
" Gli Insetti." His views attracted little or no attention until
1925, when the present writer adopted certain of Berlese's main
conclusions as a general theoretical explanation of the diverse
facts of metamorphosis. During the five years that have followed,
the writings of de Gryse (1926), James (1928), Eastham (1929),
Chrystal (1930), and others have contributed facts bearing upon
the theory, but the latter has not received the recognition it
merits.

In the first place, Berlese's theory takes into consideration the

51

52

METAMORPHOSIS

progressive development of metamerism in insect embryos. In
a broad general way, segmentation commences anteriorly and
extends backwards as development proceeds. The head and
thorax, for example, are usually clearly defined, along with their
developing segmental appendages, before the abdomen has
undergone any marked differentiation. Although embryonic
development is a continuous process, and none of its phases are
sharply marked off from another, three important stages or
" landmarks " in the process can be recognised in many insects.
These are termed by Berlese the protopod, polypod and oligopod

phases respectively.

In the protopod phase (Fig. 31, A)
the embryo, as a whole, is as yet
incompletely differentiated both
externally and internally. Although
the head and thorax bear evident
limb rudiments, the abdomen as a
rule is imperfectly segmented at
this stage, and is always devoid of
appendages. In the earliest aspect
of the phase the abdomen may
betray no evidences of metamerism,
but as development advances a
greater or smaller number of its
segments become defined : in
Berlese's terminology, growth pro-
ceeds from an oligomerous to a polymerous condition. The
internal organs are not fully differentiated, and even their
fundaments may be, as yet, very rudimentary, particularly as
regards the digestive, circulatory and nervous systems, while the
tracheal invaginations have yet to be formed.

The polypod phase (Fig. 31, B) follows upon the protopod, and its
chief characteristic is the completely segmented abdomen, each
somite bearing a pair of rudimentary limbs. The tracheal system
now becomes evident by the presence of spiracular invaginations,
while the digestive, nervous and circulatory organs are clearly
defined and more or less completely formed.

Fig. 31. Embryonic
according to Berlese.

topod
oligopod.

B, polypod

phases
A, pro-

BERLESE'S THEORY 53

The oligopod phase (Fig. 31, C) is derived from the polypod by a
process of further differentiation. The thoracic hmbs have
markedly increased in size ; the mouth parts have attained fuller
growth ; the fundaments of the tracheal system become evident,
while the abdominal appendages, excepting those destined to form
genitalia and cerci, have undergone resorption.

Comparison of the embryonic development of various insects
shows that the above-mentioned three phases can be identified in
varying degrees of definition in different groups. The reader is
referred to the standard embryological memoirs, dealing with the
more important types, in order to follow the growth changes
affecting these phases and the sequence in the ontogeny. The
typical protopod phase, for example, is clearly evident in the
embryos of Campodea (Uzel), Lepisma (Uzel), (Ecanthus (Wheeler),
Xiphidium (Wheeler), Mantis (Viallanes), etc. In these instances
the abdomen is at first unsegmented and its metamerism is subse-
quently acquired. In other cases the abdomen appears to undergo
segmentation simultaneously with the rest of the body, as in
Coleoptera. The polypod phase, which immediately follows, is
probably an ontogenetic feature common to all Hemimetabola and
to many of the Holometabola. In certain of the more specialised
members of the latter division all the abdominal appendages
may be suppressed from the ontogeny, as occurs in Diptera and,
according to Nelson (1915), in the hive bee ; or, the appendages
of the first, or first and second segments, alone are developed, as in
Donacia (Hirschler), Dytiscus (Th. Bang), and certain other
Coleoptera. When the development of embryonic abdominal
limbs is thus suppressed it is obvious that a polypod phase, in the
strict implication of the term, can scarcely be said to exist. The
incidence of the phase can only be inferred from the occurrence of
other correlated features such as the presence of spiracles, and
the general advance in development following the preceding phase.
The oligopod phase is the final embryonic stage among Coleoptera
and many Neuroptera ; it probably also occurs in the Diptera and
in the higher Hymenoptera, but in a greatly modified and masked
condition which renders its recognition by no means obvious.

According to Berlese the moment of birth, or eclosion from the

54 METAMORPHOSIS

egg, is a definite crisis of paramount significance in the ontogeny
of all insects. The theory maintains that, in all Hemimetabola,
the three phases alluded to are passed through in the egg, and
eclosion takes place when the insect has assumed a post-oligopod
phase, termed the nymph. In the Holometabola, on the other
hand, the young insects (or larvce) invariably issue from the
egg at any earlier ontogenetic stage than the nymphs of the
Hemimetabola. The stage in organisation in which eclosion
takes place determines in a large measure both the form and
subsequent development of any given insect. The more advanced
the condition at the time of emergence, the lesser are growth
changes required for an insect to attain its ultimate organisation.
In the Holometabola, the stage in which eclosion takes place
corresponds more or less closely with one or other of the embryonic
phases alluded to. Some groups emerge as veritable embryos ;
others may be polypod larvae or oligopod forms ; or they may
be apodous and have lost all traces of trunk limbs.

The factors which determine the moment of eclosion from
the egg are obscure. Even William Harvey, as long ago as 1651,
considered that premature birth in insects is due to paucity of
food reserves in the egg. There is little doubt that the main
influence is physiological, and depends to a considerable extent
upon the amount of yolk available in the egg, in proportion to
the size of the embryo. This aspect of embryology has, however,
been very little studied, and information respecting relative
amounts of yolk in the eggs of diff'erent insects, and its comparative
nutritive values, is a practically unexplored field. Precocious
emergence is also intimately related with the immediate environ-
ment of the newly-born larvae, and the egg-laying instincts of
the female parent. If, for example, the eggs are deposited in
situations providing the young larvae with an abundance of food
immediately at hand the early hatched, and more or less inert,
offspring would have chances of survival. On the other hand,
if the young larvae have to seek out their food survival would be
impossible, unless emergence be delayed until their development
had advanced to the stage where they had become endowed with
the necessary locomotory and sensory organs.

NYMPHS AND LARVJS 55

Nymphs and Larvae. It will be convenient at this stage to
emphasise the fundamental differences between nymphs and larvae.
In the strict zoological sense such differences scarcely hold good
and, whenever the young differ fundamentally from their parents,
the term larva becomes applicable. It has, however, long been
both customary and convenient to distinguish between these two
types of immature insect.

An insect undergoing incomplete metamorphosis leaves the egg
in a relatively advanced condition of morphological development.
In its general structure and body-form it prefigures the imago.
Furthermore, it adopts a mode of life similar to that of its parents,
it frequents the same habitat, and feeds upon similar food. Its
mouth-parts exhibit but slight, if any, structural differences
from the final condition assumed by those organs, and compound
eyes, usually accompanied by dorsal ocelli, are the functional visual
receptors. Rudiments of wings and genitalia develop gradually
as external rudiments, and the final transformation comprises
little or no marked morphological change other than the acquisition
of those organs in their fully developed condition. Physiologically
the main change is involved in the increasing development of the
sexual organs up to maturity. Immature insects exhibiting these
gradual growth changes are termed nymphs.

It is noteworthy that in the Plecoptera, Ephemeroptera and
Odonata the nymphs are exceptional in being aquatic, while their
imagines adopt an aerial mode of life. In the possession of special
respiratory organs and of other adaptive modifications, which fit
them for an aquatic existence, the nymphs of these three orders
exhibit more or less evident coenogenetic development. The
final transformation into the adult consequently involves more
drastic morphological change than usually obtains with typical
incomplete metamorphosis. This feature has been emphasised
by Comstock (1918), who proposed the term naiad for the
immature insect in each of the three orders named, and
adopted the expression in his subsequent text-book, " An
Introduction to Entomology " (1924). The distinctions between
nymphs and naiads, therefore, are of a purely adaptive nature
and are not absolute. In certain Plecoptera, for example,

56 METAMORPHOSIS

the adaptations are so slight that the term naiad is scarcely
applicable.

Larvye may be defined as immature insects which leave the
egg in a relatively early stage of morphological development.
In general, they differ fundamentally in form, structure and
behaviour from the imagines. Their mouth-parts and other
appendages usually exhibit marked or even profound differences
of form and structure, while lateral (or adaptive) ocelli, with only
very rare exceptions, are the functional visual organs. Throughout
larval life the reproductive system is in a very rudimentary
condition. Coenogenetic development has become so highly
specialised that transformation into the imago has involved the
intercalation of a quiescent pupal instar in the ontogeny. Prior
to the assumption of the pupal condition, the developing buds
of the genitalia, wings and other appendages lie concealed beneath
the body-wall.

In the present discussion the term Hemimetabola has been used to
include both the Hemimetabola stnsu stricto and the Paurometabola of
many authorities. The differences exhibited by the transformations in
the two cases are essentially those of degree.

In Giard's interpretation, the Hemimetabola {sens, kit.) undergo
transformation as opposed to true metamorphosis which is featured
only in the Holometabola.

Types of Larvae

Insect larvae may be grouped into three principal types depend-
ing upon the stage in their ontogeny in which eclosion from the
egg takes place. In addition to these a fourth type, derived from
the third, has to be recognised.

1. The Protopod Type. This type is characteristic of the
primary larva? of certain species of endoparasitic Hymenoptera.
The eggs in such species are either almost devoid of yolk, or but
poorly supplied with nutrient material, and the young larvae
consequently issue at a very early ontogenetic stage of develop-
ment. Their survival is rendered possible from the fact that they
are enclosed either in the eggs or bodies of their hosts, where
they develop immersed in an abundance of easily assimilated
food, usually in the form of yolk or of ha^molymph.

THE P ROTO POD LARVA

57

Some of the most typical examples of protopod larvae are found
in the Platygasteridae, which are very prevalent parasites of the
larvae of Cecidomyidse. The primary larva of Platygaster herricki
is represented in Fig. 32, and is shown in comparison with the
embryo of Xiphidium in the protopod phase of development.
The extremely close morphological similarity in the two cases is
particularly striking, and it is evident that the Platygaster larva is

Fig. 32. A, Protopod phase of embryo of Xiphidium. (From
Wheeler.) B, Protopod larva of Platygaster. (From Kulagin.)
Adapted from Berlese. a, Abdomen (unsegmented) ; at,
antenna ; m, mandible ; m^, maxillae ; o, stomodaeum ; t,
thoracic legs.

little more than a precociously emerged embryo. The abdomen
in this stage has not acquired segmentation, the thoracic
appendages are merely rudimentary outgrowths ; the nervous
and respiratory systems are, as yet, undeveloped, and the digestive
organs are still largely in an embryonic condition. Another
prevalent type of protopod larva is the cyclopoid primary larva
(Fig. 33) found in other members of the same family. Its general
features are well known through the researches of Marchal (1906).
The large inflated cephalothorax bears head-appendages and a

58

METAMORPHOSIS

single pair of thoracic limbs, while the abdomen is segmented.
The mandibles are specially large and adapted to hold on to the
host-tissues. The digestive system is represented by an extensive
sac-like mesenteron and the proctoda^um is, as yet, hardly evident.
The nervous system does not exist as such and its fundaments
are undifferentiated from the ectoderm ; the circulatory and
tracheal systems are likewise undeveloped. The so-called
eucoiliform larva (Fig. 34) present in the Cynipid family Figitidse
(vide James, 1928), whose members are common parasites of
Dipterous larvae, is evidently a specialised
protopod type. Both thorax and abdomen
are segmented, and the limbs of the
first-mentioned region are often greatly
developed, possibly functioning as a
means of rupturing the enclosing em-
bryonic membrane. Other types of
protopod larvae without thoracic limbs
occur in the Mymaridae, Scelionidae
and certain other families of parasitic
Hymenoptera. Alongside the primitive
embryonic characters of protopod larvae
there exist also specialised adaptive
of Synopeas rhanis{Flaty - features. The great development of the

Fig. 33. Protopod (cyclo-
poid) first instar larva

gasteridae), lateral view
mm, Mandibular muscle
other lettering as in Fig
32. (From Marchal.)

thoracic limbs in certain of the eucoili-
form larvae and the caudal prolongation,
often bifurcated in the Platygasteridae,
can only be interpreted as adaptional developments whose
functional significance is obscure.

2. The Polypod Type. Typical examples of polypod larvae are
afforded by the caterpillars of the Lepidoptera and the
Tenthredinidae and by the very similar larvae of the scorpion-flies
or Panorpidae. Their essential features are the retention of
abdominal limbs and the presence of a peripneustic tracheal
system. The presence of a polypod instar in the larval develop-
ment of the parasitic Hymenoptera is a recent discovery. It
occurs among the Cynipoidea in certain genera of Figitidae (James,
1928) and in Ibalia, which is a parasite of Sirex (Chrystal, 1930).

THE POLY POD LARVA

59

It has also been described by Eastham (1929) in the Proctotrypid
Phcenoserphus viator which parasitises larvae of the beetle
Pterostichus niger. In the Figitidae the polypod stage follows
the protopod or eucoiliform stage, but in Ibalia and Phceiioserphus

Fig. 34. Figites anthomyiarum. Above, protopod larva x 140 ;
below, polypod larva x 90. ^A'^., Anus ; C.E., endo-skeleton
of head ; O.P., oral papillae ; R.A., rudimentary appendages ;
S.O., sensory organ. (After James (reduced).)

the protopod stage is presumably passed within the egg, since the
first observed instar is polypodeiform. During the polypod stage
these parasitic forms live immersed in the blood of their hosts,
and, as is the rule in such cases, respiration is cutaneous. In
consequence of this adaptation the typical peripneustic condition
is not acquired until a later instar. The number of pairs of

60 METAMORPHOSIS

abdominal appendages j)resent varies from eight in Phcenoserphus,
to ten in the Figitidai (Fig. 34), and twelve in Ihalia.

3. The Oligopod Type. The oligopod type, in its most highly
developed condition, is represented by the so-called campodeiform
larva of many Coleoptera and Neuroptera. It is eminently active,
with highly developed locomotory and sensory organs, and has,
in the past, usually been regarded as the most primitive type of
insect larva. Almost every stage of adaptive modification and
degeneration is exhibited, among different insects, between this
highly developed type and the legless phase dealt with in the next
paragraph.

4. The Apodous Type. The apodous type, which is devoid of
all true locomotory appendages, appears to have been derived, in
the majority of cases, from the oligopod phase as the result of
degeneration. Larvae of this type are characteristic of Diptera
and all the Hymenoptera Apocrita ; they are also prevalent in
certain families of Coleoptera, besides occurring as a rare condition
in leaf- mining Lepidoptera.

Among Diptera the general suppression of body-appendages,
from all stages in their ontogeny, has resulted in the antecedent
protopod and polypod phases becoming largely obliterated.
Keilin (1915) has shown that three pairs of sensory papillae are
present in the larvae of this order, and are in direct relation with
the developing imaginal leg-buds, thus occupying the positions of
ancestral thoracic legs. These sensory papillae appear to be the
transformed vestiges of former thoracic limbs, and their presence
points to the conclusion that Dipterous larvae are to be regarded
as highly specialised derivatives from the oligopod type.

In all the higher Hymenoptera the prevalent larval type is
apodous. Among the Aculeata it persists through all instars,
while, as has already been shown, in certain of the Parasitica it is
preceded by earlier phases. The apodous type in this order is
most probably a derivative of the oligopod phase, as in Diptera.
In support of the contention it may be mentioned that vestiges,
in the form of papillae, of what appear to be the remains of thoracic
appendages have been detected in Polysphincta and certain other
of the Ichneumonoidea. Furthermore, in the Tenthredinoidea a

THE GENERAL SUBJECT 61

long series of transitional forms are traceable from typical poly pod
types, through various phases of reduction, to oligopod types,
until a completely apodous condition is attained. Thus, in the
Xyelidae their polypod larvae retain nine pairs of abdominal
appendages, while among the Tenthredinidae (sens, str.) this
number suffers reduction. In the Pamphiliidae, Cephidas and
Siricidse, abdominal appendages have atrophied and their larvae
are oligopod ; the thoracic limbs are never well developed and
exhibit varying degrees of reduction towards the apodous con-
dition. In Oryssus the latter phase becomes evident, the thoracic
limb-rudiments having atrophied. The legless condition of the
larva in this genus is correlated with the unique parasitic life that
is unknown elsewhere among the Symphyta.

Among Coleoptera an apodous larval type has been indepen-
dently acquired among different families. It is characteristic of
the majority of the Rhynchophora, besides occurring in certain of
the Buprestidae and Cerambycidae ; in Cercyon, in the Elaterid
sub-family Eucneminae and in the Bruchidae. In the last-
mentioned family the apodous condition may be preceded by
oligopod instars, while among the Curculionidae the last remnants
of thoracic limbs are retained as sensory protuberances in Hypera
(Phytonomus) according to Perez (1911). In Coleoptera, therefore,
evidence points to the apodous type being . derived from the
oligopod, and almost every transition between the two may be
found, sometimes in members of the same family. In Lepidoptera
the leaf-mining genera Phyllocnistis and Eriocrania are apodous,
but, unlike the preceding examples, the legless condition appears
to have been directly derived from the polypod phase.

General Remarks on Insect Metamorphosis

It will be apparent that Berlese's theory is in accordance with
the general facts of ontogeny and phylogeny. Criticisms, however,
may be raised wherever a given larval type does not appear to
conform with his premises and deductions. Due reflection will
show that it is obviously improbable that every form of larva that
issues from the egg will be in a stage exactly comparable with one
or other of the embryonic phases alluded to. Intermediates are

62 METAMORPHOSIS

bound to occur, and, at the same time, the ancestral form of any
given larva is liable to be masked by superimposed adaptive or
secondary modifications. Embryological development, although
recapitulatory in the main, is well known to be subject to abbrevia-
tion and to exhibit secondary features which in no way represent
ancestral characters. This is exemplified in the suppression of
ancestral appendages from the trunk segments in the embryos of
Diptera. In so highly specialised an order many primitive
features, common to more generalised insects, are no longer
reproduced in the ontogeny. Developmental features of this
kind, in their turn, modify the morphological characters of the
resulting larvae. There are, consequently, potent influences
affecting the forms assumed by insect larvae other than those of
ancestry. This fact is abundantly evident in the remarkable
adaptations to special modes of life exhibited by many insect
larvae, which frequently mask the actual ontogenetic phases which
such larvae represent. In the past the tendency has been to
overstress the influence of adaptation in determining the evolution
of larval forms. The perennial difficulty of determining which
characters are ancestral and which are purely secondary thus
presents itself. Berlese's theory maintains that insect larvae
represent, primarily, ontogenetic phases in the history of the
species concerned. They are subject to the same laws affecting
other organisms, and, consequently, the particular ontogenetic
phase exhibited in the species of one group may be largely obscured
by secondary features, and yet be plainly evident in another group.
In cases of hypermetamorphosis, where there is a definite
sequence of larval types in the ontogeny of a single species,
further support is afforded to Berlese's theory ; the sequence
follows that of the corresponding embryonic phases of normal
insect development. This has been shown to be the case in the
hypermetamorphoses of certain parasitic Hymenoptera (Fig. 34).
Furthermore, in those Coleoptera where hypermetamorphosis
prevails, and in the Mantispidae among Neuroptera, the first
instar is a typical campodeiform oligopod larva. In the second
instar, change of habit has led to its transformation into an inert
maggot-like phase with vestigial thoracic limbs and other

NYMPHS AND PTJP^ 63

correlated modifications. This instar, in its turn, may be followed
by an apodous stage which is well exhibited in the Meloidae. In
the Strepsiptera the campodeiform first larvae instar is directly
followed by an apodous stage.

In a few words, it may be said that Berlese's theory co-ordinates
and explains the significance of diverse larval types which can
only, otherwise, be interpreted as intercalated and purely adaptive
phases.

Nymphs and Pupae. The general characters of the nymphs of
the Hemimetabola have already been referred to. Nymphs
share many features in common with campodeiform larvae, but,
on the other hand, they exhibit in their visual organs and in the
developing external rudiments of wings and genitalia features
which appear only in the pupal instar among Holometabola. In
other words, nymphs have attained a more advanced stage of
ontogenetic development than larvae, and the protopod, polypod
and the true oligopod phases are passed through in the egg. In
the Hemimetabola the nymphal instars are preparatory to the
development of the imago, while in the Holometabola the larval
instars are preparatory to the development of the pupa. We may,
therefore, conclude that the pupa, along with the antecedent
prepupa, are the ontogenetic counterparts of the nymphal instars
of the lower orders of insects. It seems that tachy genesis, or
shortening of development, has resulted to the extent that the
whole series of nymphal instars has become concentrated into
these two stages, with the elimination of intervening ecdyses.
In support of this contention we may pause to consider certain
groups of Hemimetabola, which already exhibit well-defined
tendencies towards the holometabolous condition. Thus, among
the Coccidae definite evidence of a pupal instar occurs in the males.
Furthermore, in this family the insect leaves the egg not as a
nymph, but as an oligopod larva. In the sub-family Diaspinae
an apodous instar follows, and this, in its turn, is succeeded by
three quiescent nymphal instars delimited by separate ecdyses.
These nymphal instars are unquestionably incipient pupal stages,
and much the same condition prevails in Thysanoptera, except
that there are but two such stages. Among the Holometabola

64 METAMORPHOSIS

the prepupa is the sole vestige of the existence of more than a
single pupal instar, and even the ecdysis between the last larval
instar and this phase has been eliminated, except in very few
instances. In the metamorphoses of the fly Rhagoletis, for
example, Snodgrass (1924) showed that an extra-cuticular coat
intervenes between that of the pupa and the wall of the puparium.
Evidence of a prepupal ecdysis has also been noted by Raff (1934)
in the case of sawflies of the genus Perga, where the exuvias in
question are located near the caudal extremity in the pupal
cocoon. Such instances, however, are few, but the matter has
attracted very little attention.

The pupal instar represents, therefore, the abbreviated recapitu-
lation of ancestral nymphal stages. Change of function, however,
has resulted in the evolutionary process and an acquired quiescent
condition has been assumed. This condition became necessary
in order to allow the larval body and its component organs to
become remodelled, in order to adapt them to the requirements
of the future imago.

It is well known that, during the usual process of ecdysis between
any two nymphal or larval instars of a given insect the old cuticle is
cast off. Prior to the consummation of this event, the formative cells
beneath not only secrete a new cuticle, but also remodel the outer
configuration of the body and the appendages into the new form to be
assumed. This process goes on gradually from instar to instar, and in
many insects the changes involved are but slight, but in others they
are considerable. In the Anoplura, for example, they are so trivial as
to be scarcely evident, while in many of the Heteroptera remarkable
changes in body-form and the appendages may result. In the
Holometabola this same formative process is continued in the pupa, but
it lies concealed, and drastic changes are involved. The descendants of
the ectodermal formative cells, which modelled the larval appendages
from instar to instar, give rise to the imaginal buds, and it is from these
rudiments that the corresponding imaginal parts are derived. It
is too commonly implied that there is entire discontinuity during
pupation, and that the imaginal appendages are new developments,
apart from those of the larva. If a graded series of pupal types be
studied, it will be seen the change from the larval appendages into those
of the adult is a matter of degree only — there is no real discontinuity
in the process.

Hormones and Metamorphosis. One of the most notable results
of recent work in experimental entomology has been the discovery

HORMONES

65

of what are comparable with endocrine organs. Facts are being
brought to hght which lead to the conclusion that ecdysis, meta-
morphosis, and also the full maturing of the gonads, are controlled
through the intervention of certain secretions. Since the evidence
all points to the fact that these secretions are discharged into the
blood, they are therefore comparable with hormones of the
endocrine type. As in Vertebrata, they appear to be produced
by ductless glands and have a very definite effect upon other
organs and tissues. In insects, a pair of small bodies, known as

Fig. 35. Dorsal view of brain and associated parts of a winged
Termite, al, antennary lobe ; ca, corpus allatum (left) ; fg,
frontal ganglion ; o, oesophagus ; ol, optic lobe ; pn, posterior
sympathetic nerve (left) ; rn, recurrent nerve ; sg, oesophageal
sympathetic ganglion (right). Original.

the corpora allata (Fig. 35), found in association with the sym-
pathetic nervous system, a short distance behind the brain, are
almost certainly endocrine organs. The corpora allata arise in
the embryo as ectodermal invaginations, usually of the mandibular
segment. Subsequently they lose all connection with the exterior
and finally assume the form of small, ovoid, whitish bodies which
were formerly mistaken for a pair of posterior sympathetic
ganglia. Histologically, they differ from nervous tissue and have
rather the features of glandular organs. They occur in all orders
of insects and their structure was studied in considerable detail

R.A. ENTOMOLOGY. 3

66 METAMORPHOSIS

by Nabert in 1913. Wigglesworth (1934) showed that these
organs exhibit cychc changes in that they pass through a secretory
stage associated with moulting, which is followed by a resting
condition. In the blood-sucking Reduviid bug Rhodnius,
Wigglesworth states that ecdysis takes place at a definite interval
after feeding, only one meal being necessary during each instar.
There is a " critical period " in the moulting cycle and decapitation
before the advent of this period prevents moulting. The critical
period synchronises with the beginning of mitosis in the cells of
the hypodermis. He further states that insects which have
passed the critical period contain a hormone in their blood. If
blood from an insect in this stage be transferred into an insect
which has been decapitated before the critical period, moulting,
which otherwise does not take place, is induced. (The critical
period referred to is about seven days after feeding in the fifth
instar and about four days after in the earlier nymphs.) That the
corpora allata are the organs involved is suggested by the fact
that the cells of these bodies show their greatest secretory activity
during the critical period. In a later communication,
Wigglesworth (1936) showed that the moulting hormone from
Rhodnius will induce moulting in the allied genus Triatoma and
in the bed-bug Civiex. While the available evidence points to
the corpora allata as the seat of the hormone involved, comjjlete
proof is, as yet, not available : for one thing, their extirpation
is extremely difficult to carry out. Rather indirect evidence,
however, is afforded by some other experiments described by
Wigglesworth as follows : " fourth-stage nymphs of Rhodnius
in two groups (A and A') were decapitated seven days after
feeding and joined to groups of fourth-stage nymphs (B and B')
twenty-four hours after these had fed. Of these, B were transected
through the prothorax, the corpus allatum being removed, while
B' were transected through the posterior part of the head behind
the brain but in front of the corpus allatum.

" Groups B and B' both showed that moulting was beginning ^

^ At the onset of moulting, when the epidermal cells separate from the
cuticle and divide, there are visible changes in the abdominal tergites of the
living insect (see Wigglesworth, 1934).

HORMONES 67

at the end of five days. The pairs were then separated and B and
B' were joined to further groups of fourth-stage nymphs (C and C)
transected through the prothorax twenty-four hours after feeding.
At the end of eight days the group C, i.e., insects whose partners
retained the corpus allatum, were obviously beginning to moult ;
the group C was probably just starting to moult. The pairs
were then separated once more and joined to groups D and D',
cut through the prothorax twenty-four hours after feeding. At
the end of seven days D' were clearly beginning to moult ; but
D died several weeks later without showing any signs of moulting
beginning.

" Thus in the series A, B, C, D, in which no corpus allatum was
present, the moulting hormone appeared to diminish in con-
centration until it ceased to induce moulting at all. Whereas in
the series A', B', C, D', in which B' retained the corpus allatum
(and was presumably able, therefore, to supply an adequate
amount of the hormone to C), there was no sign of diminished
activity in the moulting hormone — ^the group D' being caused to
show signs of moulting as soon after joining to C as B' did after
joining to A'."

Wigglesworth has also brought forward evidence which appears
to indicate that metamorphosis is controlled by an " inhibiting
hormone " which is also secreted by the corpora allata. Thus it
was found that if a fourth instar nymph had only the apex of the
head removed six days after feeding and was then joined to a
fifth instar nymph decapitated twenty-four hours after feeding,
the latter moulted into an extra nymphal stage instead of becoming
an adult. The head is, therefore, necessary for the production
of the inhibiting hormone. Also, fifth instar nymphs, twenty-
four hours after feeding, had the corpora allata from third or
fourth instar nymphs (removed a ^day or two after feeding)
implanted within them. These fifth instar nymphs moulted to
" sixth instar " nymphs. As the result of similar experiments,
" sixth instar " nymphs may give rise to " seventh instar "
nymphs when they moult again. In some cases the resulting
characters were intermediate between those of adults and nymphs :
in others they were wholly nymphal in type. From these and

3—2

68 METAMORPHOSIS

other experiments it is concluded that the corpora allata control
the characters of the various instars and also secrete an
" inhibitory hormone " which prevents metamorphosis taking
place until the insect attains its fifth instar. In this phenomenon
of " inhibition of metamorphosis " there seem to be two phases
involved : (i.) deposition of new cuticle following rapidly upon
the initiation of growth and hence the arrestation of adult char-
acters ; (ii.) so long as differentiation is arrested in this way the
cells are capable of new growth and will respond to the moulting
hormone. Whether the latter hormone is biochemically distinct
with that concerned with the inhibition of metamorphosis has
not been proved.

In contrast with Wigglesworth's experiments it is noteworthy
that Weed (1936a), in an experimental study of moulting in the
grasshopper Melanoplus differentialis, obtained different results.
The account, however, of these experiments is very short and
they are stated to be preliminary only. It appears that in this
work the corpora allata themselves were removed from grasshopper
nymphs one to nine days after they had entered the last stadium.
In each case moulting resulted, and it was concluded that the
corpora allata do not influence this process. It is necessary,
how^ever, to await more extensive experiments before drawing
definite conclusions.

With regard to the Holometabola, it has been known for some
time that removal of the brain prevents jDupation. The earlier
observers did not associate the corpora allata with any significant
role in the metamorphosis, and it is extremely probable that their
removal took place along with the brain in, at any rate, most of
the experiments recorded. Thus Kopec (1922) removed the brain
from caterpillars of Lymantria disjpar in their last instar. If the
brain be excised ten days before the last ecdysis, or later, the
caterpillars pupated, but if the experiment be performed seven
days after that moult, jjupation did not occur. Kopec concluded
from these, and other experiments, that a hormone is produced
in the brain and that this secretion induces pupation. Quite
recently Caspari and Plagge (1935) obtained very similar results
with larvae of Sphinx ligustri and Deilephila eupJiorbiw. Bounhoil

HORMONES 69

(1936), on the other hand, obtained very different results with
silkworms (Bo77ibyx inori). The corpora allata were removed (by
a surgical method to be described by him in a later paper) from
larvae in their last instar at times varying from thirty-six hours up
to sixteen days after the last moult. As the result of these
experiments he concludes that the corpora allata do not influence
pupation. Out of forty-two larvae operated upon, twenty-four
produced pupae : of the latter, thirteen gave rise to imagines.
The time of onset of the " critical period," however, is not stated.
The author states that considerable mortality ensued owing to
unfavourable weather prevailing at the time ; given better
circumstances still more convincing results, he believes, could be
obtained. These results appear to be in accordance with experi-
ments by Kiihn (1936), who used larvae of Ephestia. By grafting
brains of larvae into individuals from which the brains had already
been excised, a very high death-rate supervened. Only eight
examples out of 150 pupated, and from among these individuals
five of them had no corpora allata in the graft. Kiihn, by a
series of ligaturing experiments, had previously determined the
incidence of a " critical period," but his experiments obviously
need repeating owing to the high mortality which prevailed.

Fraenkel (1935) has brought convincing evidence that it is the
action of a hormone which brings about pupation in the fly
Calliphom. It is secreted about sixteen hours before pupation at
a temperature of 20° C. Fraenkel was unable to locate the actual
hormone-producing organ, but showed that it was either the main
ganglionic centre or in its immediate neighbourhood. After the
hormone is discharged into the blood he found that pupation takes
place without the co-operation of the central ganglia.

In the present state of knowledge, of a very recent line of
inquiry, it wo«ld be obviously premature to draw hard and fast
conclusions from rather conflicting results. One thing, however,
emerges, and that is the existence of hormones which control
both ecdysis and metamorphosis. While there is strong evidence
in Rhodnius that the corpora allata are the organs secreting the
hormones concerned, proof that these bodies perform a similar
function in the Holometabola is less convincing. It needs also to

70 METAMORPHOSIS

be borne in mind that the oenocytes may also perform a significant
function in connection with ecdysis. They appear to be gland
cells of a specialised type which discharge their secretion into the
blood. Furthermore, they show a very marked cycle of secretory
activity which synchronises with the process of moulting (Albro,
1930 ; Wigglesworth, 1933). Roller (1929) has suggested that
they secrete a hormone which induces moulting, but Wigglesworth
does not consider that this is likely and is inclined to believe that
they are concerned with the synthesis of some of the non-chitinous
constituents of the cuticle during ecdysis.

Although not directly concerned with the metamorphosis, the
functions of the corpora allata in adult insects may also be referred
to here. Wigglesworth (1936) has shown that these organs are
necessary for the production of ripe eggs in Rhodnius. -Further-
# more, it was found that the corpora allata of Triatoma females
will induce egg-development in Rhodnius. In the absence of the
secretion from the corj^ora allata the oocytes continue to grow
so long as they are connected with the nurse cells, but when
their nutrition is taken over by the follicular cells they die and
are absorbed. Further evidence shows that the moulting hormone
of the nymphs will not induce egg-development in the adult
female ; nor will the egg-forming hormone bring about moulting.
In the male the corpora allata are necessary for the full activity
of the accessory glands. Their secretion will also induce egg-
development in the female and similarly the secretion of the
female will activate the male accessory glands. It is noteworthy
that Weed (1936) has obtained very similar results with the
grasshopper Melanoplus differentialis. By removing the corpora
allata, in a series of preliminary experiments, it was concluded
that these bodies are necessary for the normal development of the
eggs and that, further, secretion in the glandular portion of the
oviducts is also influenced by the corpora allata.

Literature

Albro, 1930. Journ. MorphoL, L., 527.
Berlese, 1913. Redia, IX., 121.
BouNHOiL, 1936. C.R. Acad. Sci., CCIII., 388.
Caspari and Plagge, 1935. Naiurwiss., XXIIL, 751.

LITERATURE 71

Chrystal, 1930. Oxford Forestry Memoirs , XI.
CoMSTOCK, 1918. Ann. Ent. Sac. Am., II., 222.

1924. " An Introduction to Entomology." Ithaca, 179.
Eastham, 1929. Parasitology, XXL, 1.
Fraenkel, 1935. Proc. Roy. Soc. B., CXVIII., 1.
De Gryse, 1926. Trans. Roy. Soc. Canada, XX., Sect. V., 483.
Imms, 1925. '' General Textbook of Entomology." 1st Edition. London,

179.
James, 1928. Ann. Appl. Biol., XV., 287.
Keilin, 1915. Bull. Sci. Fr. et Belg., XLIX., 1G6.
KoLLER, 1929. Biol. Rev., IV., 269.
Kopec, 1922. Biol. Bull., XLIL, 323.

KuHN, 1936. Gesell. iviss. Gottingen ; Nach. aus der Biol., XL, 141.
Marchal, 1906. Archiv. Zool. Exp. et Gen., IV., Ser. IV., 485.
Nabert, 1913. Zeits. wiss. Zool, CIV., 181.

Nelson, 1915. " The Embryology of the Honey Bee." Princeton.
Perez, 1911. Comp. Rendu Soc. Biol. Paris, LXXL, 498.
Weed, 1936. Proc. Soc. Exp. Biol, and Med., XXXIV., 883.

1936a. Ibid., 885.
WiGGLESWORTH, 1933. Quart. Journ. Mic. Sci.. LXXVL, 269.

1934. Ibid., LXXVIL, 191.

1936. Ibid., LXXIX., 1.

CHAPTER IV
PALEONTOLOGY

Introduction, p. 72. Orders only Known as Fossils, p. 74
Pakeodiciyoj^tera, p. 74 ; Protorthoj)tera, p. 75 ; Megasecoptera, p. 77 ;
Protodonata, p. 79 ; Proioj^erlaria, p. 80 ; Protohcmiptem, p. 81 ;
Protocoleoptera, p. 82 ; Protelytroptera, p. 83 ; Incertce Sedis, p. 84.
The GeologiCxVL History of Insects, p. 86. Devonian, p. 86 ;
Carboniferous, p. 86 ; Permian, p. 87 ; Trias, p. 90 ; Jiirassie, p. 90 ;
Cretaceous, p. 92 ; Tertiary, p. 92. The Phylogeny of Recent and
Fossil Orders of Insects, p. 94. Literature, p. 102.

Introduction

In the fossil condition insects are only found in certain geological
strata especially favourable for their preservation. Under such
conditions they are often abundant, and this appears to be due to
the probability that most of the specimens got drowned and were
quickly covered with silt or other deposit before their remains
had time to decay completely. Some of the richest stores of fossil
insects are consequently found either in vegetable deposits like
coal, lignite and peat, or in ancient freshwater basins. A great
many insects have also been found in amber, where they became
entangled in the resin which speedily enveloped them before
they had any chance of freeing themselves. Specimens sealed
up in and impregnated by this natural medium are retained in a
wonderfully perfect state of preservation.

For the most part, insect fossils are represented by wings
only. It appears that under the conditions of fossilisation decay
speedily sets in, with the result that the bodies of the insects soon
disintegrated and became separated from the wings. The
lightness of the latter caused them to float away and to accumulate
in certain beds where their more durable structure enabled them
to resist destruction. The natural result has been that insect
palaeontology is mainly a study of venation, and knowledge of

72

PALEONTOLOGY 73

the bodies, mouth-parts, legs, and other parts of most fossils is
of a more fragmentary character. It is only when we come to
the amber fossils of Miocene age that complete whole insects
prevail, but owing to their comparatively recent date they lack
the importance and exceptional interest associated with the
earlier and more imperfect remains. Altogether about 11,500
species of fossil insects are known, and of these approximately
one-fifth belong to orders or families no longer existing to-day.
Recent advances in insect palaeontology have, to a marked degree,
filled in conspicuous gaps in our knowledge of the evolution of
most of the main orders, but we are still faced with want of direct
evidence from fossils with regard to the two most important
problems of insect phylogeny, viz., the origin of insects as a class,
and the origin of wings. The oldest winged insects are of Upper
Carboniferous age (North America), but the fact that these forms
are both abundant and in some cases already specialised, leads to
the conclusion that the earliest Pterygota must have developed
not later than Devonian or Lower Carboniferous times. Among
arthropod remains found in flakes of Rhynie Chert from the Old
Red Sandstone of Scotland are fragments of what appear to be
Apterygota, and, if this conclusion be correct, these remnants
constitute the oldest known fossil insects. The oldest living insects
are the Blattidae, which have persisted through the ages, with
comparatively slight evolutionary changes, from Carboniferous
times to the present day.

The strata where the richest and most important recent dis-
coveries in fossil insects have been made are of Permian and
Triassic age. The fossils disclosed in these rocks have revealed
so remarkable an insect fauna that our whole outlook has under-
gone marked changes. In a few words, it may be said that these
discoveries have resulted in the Permian and Triassic records of
insect life being transformed from among the least known to
almost the best known geological epochs with respect to these
animals. The strata in question are (1) the Upper Triassic Beds
of Ipswich, Queensland : these were discovered many years ago,
but have been more carefully explored recently by B. Dunstan,
Queensland Goventment Geologist ; (2) the U}:)))er Permian Beds

74 PALEONTOLOGY

of Belmont and Newcastle, New South Wales : the discovery and
exploration of these rocks is largely due to J. Mitchell ; (3) the
lower Permian Beds of Kansas, U.S.A. : originally discovered by
A. C. Sellards of Yale, they have been explored by the Yale
University Expedition under C. Dunbar and more recently by
F. M. Carpenter of Harvard ; (4) the Upper Permian Beds of
North Russia : these have been explored during 1926-28 by
A. B. Martynov and M. Edemsky. Of these four series the most
important are those of Kansas, and many of the insect remains
preserved in these beds are among the most perfect so far known.
A number of the specimens show the colour-pattern and the finest
details of the trichiation on the veins and membrane of the wings.
In addition to the foregoing, mention needs to be made also of
studies by Martynov on the Jurassic insects of Turkestan ; by
Ping on the Cretaceous insects of China ; and by Tillyard on
Liassic forms from Europe.

Orders only Known as Fossils

The palaeozoic and earlier mesozoic fossil insects include
representatives of orders which no longer exist at the present
day. The chief extinct orders, including those brought to light
by recent research, are the following.

Palaeodictyoptera Gold. The oldest and most primitive order of
winged insects and its members are characterised by the following
features. The head is rounded and bears setaceous antennae.
The prothorax is generally provided with conspicuous lateral notal
outgrowths resembling vestigial wings. The mesothorax and
metathorax are subequal in size and subquadrate in form. The
wings are elongate and attached to the thorax by broad bases :
the two pairs are closely alike in form and venation, the hind
wings being without the fan-like anal lobe characteristic of many
early groups. The venation is more complete than in any other
insects (tide Lameere, p. 19), and its outstanding features are
(1) the greatly developed branching of most of the principal
veins ; (2) the presence of well-developed branches MA and MP to
the media ; (3) the retention of at least three, and usually more
than three, anal veins or branches ; and (4) the presence of an

FOSSIL ORDERS

75

irregular network, or archedictyon, of fine veinlets over the
whole wing membrane (Fig. 36). The abdomen is elongate and
10-segmented, each segment often with conspicuous side-lobes of
a similar nature to those of the prothorax, although much less
pronounced. At the apex of the abdomen is a pair of slender and
often greatly elongated cerci, and, in some genera, a median caudal
filament is also present.

The order is almost exclusively confmed to the lower and
middle strata of the Upper Carboniferous, and about 160 species

Fig. 36. Stenodictya lohata Brongn., restoration. (After Handlirsch.)

are known. It appears to have declined comparatively abruptly,
since there are very few representatives in the later Carboniferous
and Permian rocks. Handlirsch (1925) recognises twenty-four
families, and among the more important genera are the following :
Stenodictya Brongn. (Fig. 36), Commentry (France) ; Stilbocrocis
Handl., Germany ; Mecynoptera Handl., Belgium ; Lithomantis
Woodward, Scotland; Homoloptera Brongn., Commentry (France);
Homaloneura Brongn., Commentry (France) ; Styne Handl.,
Germany ; D unbar ia Till., Kansas ; and many others.

Protorthoptera Handl. This order comprises a miscellaneous

76

PALAEONTOLOGY

assemblage of Ortliopteroid forms ancestral to some of the existing
cmsorial and saltatorial groups. It is an early offshoot of the
Pakeodictyoptera, whose members retain many of the venational
characteristics of the latter, but are distinguished by the following
features. The pro thoracic notal expansions, so prominent among
Palfeodictyoptera, are present in certain forms only ; the prothorax

Fig. .'J7. Dieconeura arcuatu Scud.
(After Handlirsch.)

restoration.

is more specialised, lacing markedly elongated in some forms and
broader and more shield-like in others. The hind wings are
broader than the fore wings, owing largely to the development of
a fan-like anal lobe (Fig. 37). The venation resembles that of the
more specialised of the PalcCodictyo]:)tera, the median vein exhibits
the branches MA and MP and the primitive archedictyon is
retained. Handlirsch's order Protoblattoidea (Fig. 37) which
comprises ancestral cockroach-like forms does not appear to merit

MEGASECOPTERA

11

separation as an order distinct from the Protorthoptera and is
here regarded as a superfamily of the latter. The tarsi are
5- jointed and there is often a well-developed ovipositor.

The Protorthoptera occurred from the Upper Carboniferous
until the Trias, and number about 200 species. They comprise
about eighteen families, and at least another twelve are included
in the Protoblattoidea. Representative genera include Spaniodera
Handl.. Dieconeura Scud. (Fig. 37), and Gerarus Scud., from

Fig. 38. Protophasma dumasi Broiign., restoration. (After Handlirsch.)

Illinois, U.S.A., and Ischnoneura Brongn., Cnemidolestes Handl.,
Stenoneura Brongn., Protophasma Brongn. (Fig. 38), from
Commentry, France, and Euccenus Scud., Illinois.

Megasecoptera Brongn. Related to the Protodonata, but
distinguished from that order by the elongate cerci, narrowed
or subpetiolate wing-bases and the single anal vein, with or
without descending branches (Fig. 39). The dense network of
transverse veinlets characteristic of Protodonata is wanting in
this order. Remains of these insects occur in the Upper
Carboniferous of Commentry (France), England and North

78

PALEONTOLOGY

America, and the Lower Permian of North America (Kansas).
Seven famihes are recognised by Handhrsch, and representative
genera include Corydaloides Brongn., Campyloptera Brongn.,
Brodia Scud. (Fig. 46, A), Mischoptera Brongn. (Fig. 39), Kuloja
Mart., Protohymen Till., and Asthenohymen Till. The two last-

FiG. 39.

Mischoptera woodwardi Brongn.,
(After Handlirsch.)

restoration.

named genera were placed by Tillyard (1924a) in a separate order
which he termed Protohymenoptera. This group it was claimed
was ancestral to the Hymenoptera. Carpenter (1930) and others
maintain that this conclusion is erroneous and that the
" Protohymenoptera " are specialised Megasecoptera without
direct affinity with the Hymenoptera. The presence of an

PROTODONATA

'9

ovipositor in Asthenohymen (vide Tillyard, 1936a) seems to be the
only character which might suggest affinity with Hymenoptera.

Protodonata Handl. These insects are readily separated from
Odonata by the absence of a true nodus and pterostigma from

U:^

YTJ^^x^ —

r^X>

^ U— ^■^'' —

^^^^^^^

1^

-2-^ / / /sc^^-^^'

Fig. 40. Reconstruction of Lemmatophora typa Sellards.
(After Carpenter.)

the wings. Their venation is distinguished from that of the
Palseodictyoptera by (a) the network of cross-veinlets being at
right angles to the main veins and their branches, instead of
forming an irregular and obliquely distributed meshwork ; and

80

PALEONTOLOGY

{b) by the single anal vein (Fig. 41). In the latter feature they
approach the Megasecoptera, but the broad wing bases and the
close meshwork of cross- veins readily separate them. The
Meganeuridse include the largest of all fossil insects with a wing-
expanse in some species which exceeds 2 feet. The order may be
divided into two families.

1. Fam. Protagriidce. Both the main branches of M and Cu
present. Protagrion Brongn, Upper Carboniferous of Commentry,
France. Calvertiella Till., Lower Permian of Kansas.

2. Fam. Meganeuridce. Veins MP and CUj vestigial or absent
Meganeura Brongn., Upper Carboniferous of Commentry, France ;
Lower Permian of Germany and North America. Megatypus
Till., and Typiis Sell., Lower Permian of Kansas. Boltonites.

Fig. 41. Wing of Protagrion audouini Brongn. Restoration of wing ;
cross-veins omitted (length about 00 mm.). (After Tillyard,
Rec. Ind. Mus., XXX., 1928.)

Upper Carboniferous of Somerset, England. Paralogus Scud.,
Upper Carboniferous of North America.

Protoperlaria Till. (Mio^noptera, jmrtini Martynov). This order
differs from the Plecoptera (or Pcrlaria) in the presence of 5-
segmented tarsi, well-developed j)ronotal expansions to the
prothorax and of both main branches (MA and MP) to the media
(Fig. 40). Its limits are as yet imperfectly understood, and
probably certain types relegated to other groups belong here.
So far representatives have only been found in the Permian rocks
of Kansas and of North Russia. The Kansas examples were
originally placed by Sellards in new families of Protorthoptera
and subsequently Martynov (1927) erected the order Miomoptera
for their reception along with the family Palteomantidte Handl.,
and new forms discovered in North Russia. The IMiomojotera as

PROTOPERLARIA

81

defined by JMartynov appear to be an unnatural assemblage of
diverse forms, including, for example, the Delopterida?, which are
relegated by Tillyard to the Psocoptera. In 1928, Tillyard,
from a study of abundant Kansas material, showed that
Letmnatophora and its allies, which were referred by Sellards
to the Protorthoptera, exhibit relationships with the recent
order Plecoptera, and proposed the new order Protoperlaria for
their inclusion. Members of this order agree with the most
primitive living Plecoptera in their general facies, including the
presence of cerci and the absence of a
median caudal prolongation, in the
general shape of the wings and the
possession of a large fan-like anal
area to the hind pair, and in the
partial fusion in the hind wings of Rs
and MA. The well-developed anal
lobe and the absence of an irregular
network of obliquely arranged veinlets
easily separate this order from the
Palaeodictyoptera.

The discovery by Carpenter (1935)
of nymphs belonging to the Proto-
perlaria is of great interest. They
possessed nine pairs of lateral
abdominal gills which still remain as

vestiges seen in the imagines (Fig. 42). Fig. 42. Nymph of Proto-
T ., p . ^1 11 perlaria. (After Carpenter.)

In these leatures they are closely

paralleled by the Eustheniidye, which form the most primitive

family of living Plecoptera.

The order comprises only the single family Lemmatophoridce,
which includes the following genera : Lenwiatophora Sell.,
Paraphrisca Handl., Leconium Sell., Artinska Sell., and Sellardsia
Till., along with certain forms from North Russia (Martynov,
1928a).

Protohemiptera Handl. The existence of this order is based
upon well-i^reserved but fragmentary remains of two genera,
probably belonging to separate families, viz., Eugereon Dohrn.

82

PALAEONTOLOGY

(Fig. 43), from the Lower Permian of Germany, and Mesotitan
Till, from the Middle and Upper Trias of New South Wales.
The Protohemiptera are among the largest fossil insects so far
discovered, and the wing-expanse of Eugereon boeckingi Dohrn.
is estimated by Handlirsch at 16 cm. Except for the discovery of
the head and prothorax of this remarkable fossil palaeontologists
would have been faced with a puzzling problem in settling its
affinities from the wing remains only.

The Protohemiptera are characterised by pronotal expansions
very similar to those of the Palaeodictyoptera and Protoperlaria.
The head is rather small, with projecting mouth-parts of the

Fig. 43. Eugereon hoeckingii Dohrn. Reproduced by permission of
the Trustees of the British Museum.

piercing and suctorial type. The whole of the wing is covered
with a meshwork of delicate and very closely set transverse
veinlets. The venation is markedly reduced with R^ and Cuj
simple unbranched convex veins, but Sc is well develoj^ed and
gives off numerous costal veinlets. M is represented by MA only
and Cu2 is pectinately branched.

Protocoleoptera Till. The order Protocoleoptera was founded
by Tillyard (1924) for an example of an elytron (or tegmen), about
24 mm. long, discovered in the Upper Permian of Belmont,
N.S.W. This specimen, Protocoleus mitchelU Till., appears to point
to the existence of a primitive order ancestral to the Coleoptera,
but having flattened teguminous elytra with straight sutural
margin and a complete system of venation. Tillyard, in 1928, also

PROTEL YTROPTERA

83

referred to this order Permofidgor Till., which he had originally
placed in the Homoptera. In this genus there is a complete,
although narrow, sutural margin and a distinct alula, resembling
that of the Hydrophilidae, situated at the base of the elytron. In
Permofulgor the venation is more reduced than in Protocoleus, and
the genus appears to lead to the extinct true beetles of the family
Permophilidjc, whose remains occur
contemporaneously in the same strata.

The venation of these early Coleop-
teroid types (Fig. 44) is totally vmlike
that of the elytron of any known true
Coleoptera, since in the latter insects
the main veins are simple and un-
branched throughout their courses and
the anal area much reduced. In 1928,
Forbes pointed out that the deep notch
at the articulation and the basally
extended anal area are common Orthop-
teroid features, together with the richly
branched condition of the main veins.
On these and other features he refers
Protocoleus to the group Gryllacridiidae
of the Orthoptera. The resemblance
between this genus and a typical
Gryllacrid, however, is not very marked,
the venation being very different in
the two cases and the absence of the
characteristic cross- veinlets of Gryllacris is particularly noticeable.

Protelytroptera Till. Some of the most remarkable insect
remains found in the Elmo limestone of the Lower Permian rocks
of Kansas are those to which Tillyard (1931) gave the ordinal name
of Protelytroptera. The discovery of further material has enabled
these fossils to be studied more fully by Carpenter (1933). While
a certain amount of detail is known with regard to their general
external form and structure, no light has, so far, been shed as
regards the character of their mouth-parts. In the more
generalised types such as Protelytron the fore wings were convex

Fig. 44. Protocoleus mitchelli
Till, elytron x 3. (After
Tillyard, Proc. Linn. Soc.
N.S.W., XLIX., 1924.)

84 PALEONTOLOGY

elytra (Fig. 45) with a widened costal area : II was 2-branched,
M single-branched, while Cu divided basally in Cuj and Ci^.
There were three simple anal veins and a general absence of cross
veins. In other genera the venation was more specialised by
reduction. The hind wings were evidently adapted for folding
and, as in Dermaptera, the anterior veins are much modified and

Fig. 4..5. Protelytron permianum Till. Reconstruction.
(From Carpenter.)

there is a large anal fan formed by a number oF radiating anal
veins. The prothorax is large, the tarsi 5-segmented and the
antenna? composed of about twenty- two joints. The order is
divisible into five families which take their names from the
following genera. Protelytron Till., Megelytron Till., Blattelytron
Till.. Archehjtron Carp., and Ehjtroneiira Carp.

Incertae Sedis. In addition to the foregoing there are a number

MECOPTEROID GROUPS 85

of smaller groups, many of them of very uncertain affinities,
which have been given ordinal rank by Handlirsch. For the most
part they are based upon anomalous fossil remains, usually few
in number, which have been made the types of provisional orders
more especially to emphasise their essential differences as compared
with other groups of insects. The orders Paramecoptera, Para-
trichoptera and Protodiptera of Tillyard are also only doubtfully
separate and distinct. No parts other than wings have so far
been discovered, and from their venational characters it is presumed
that their nearest living relatives are the Mecoptera. In his
later communications Tillyard (1935, 1937) regards them as being
sub-orders of this originally extensive order (the Mecoptera).
They are all characterised by a primitive venation, which was
essentially similar in both fore and hind wings, and by the single-
branched Cu^.

The Paramecoptera are represented by the genera Belmontia
Till, and Parahelmontia Till, from the Upper Permian of Belmont,
N.S.W. They are characterised by the rather extensive dichotomy
of Rs and M, and in Belmontia there is a short distal fork to Cu^.
In so far as the last-named genus is concerned, its features betray
no insuperable barrier to its being regarded as a remnant of a.
group which was ancestral to the Trichoptera and Lepidoptera.

The Paratrichoptera are characterised by both Rs and M
having four branches only, arranged dichotomously. They
comprise Mesopsyche Till., Neuropsyche Till., Aristopsyche Till.,
and Triassopsyche Till., from the Upper Trias of Queensland :
Pseudopolycentropus Handl., Lias of Europe and LiassopMla
Till., Upper Lias of England.

The Protodiptera are based upon a dipterous type of wing to
which Tillyard (1929) gave the name of Permotipula. This
discovery was followed several years later by the finding of a
four-winged insect with very similar venation in both pairs of
wings (Tillyard, 1937). Both these fossils were found in Upper
Permian rocks of Warner's Bay, New South Wales. The second
example is clearly that of a four-winged insect whose wings are
of a type still to be found in the most primitive living families of
Diptera. The wings bear a marked dipterous facies in their

86 PALEONTOLOGY

narrow bases and correspondingly reduced anal area. The
venation is characterised by Rs with three or four branches ;
M with four branches ; Cug reduced and approximated to Cu^,
and the anal veins reduced to lA or with a vestige of 2A also.

The Protodiptera are undoubtedly the most distinct of the
above supposed sub -orders, but so far no annectant forms have
been found between them and the true Diptera.

It is probable that future discoveries may result in bringing
to light forms connecting the Paramecoptera {Belmontia) with
existing Protomecoptera, justifying the merging of the former
with the latter. Whether Parahehnofitia is really closely connected
with Behnontia seems doubtful.

The Geological History of Insects

The sequence of the appearance of insect remains in geological
times is displayed in the accompanying table (p. 101), and the chief
facts in the past history of these animals will be briefly discussed
in the light of recent discoveries. It may be pointed out that no
insect fossils have so far been found in any pre-Devonian rocks.

Devonian. The presumed insect remains of this period are
extremely small and fragmentary, and have been found in flakes of
Rhynie Chert from the. Middle Old Red Sandstone of Scotland.
The most complete remains consist of portions of the head-
capsule, mouth-parts and antenntje of the species Rhyniella
prcecursor Hirst and Maulik, which is regarded by Tilly ard
(1928a) as being an early Collembolan. The other fossils consist
of mandibles only, and the species to which they have been
relegated, viz., Rhyniognatha hirsti Till., may possibly have been
a Thysanuran. If, as apj^ears very probable, Rhyniella is a
species of true Apterygota, it is remarkable that no evidence of
the existence of such insects has so far been revealed in inter-
vening rocks until the Oligocene period.

Carboniferous. Remains of insect life are found in comparative
abundance during this period, and the earliest undoubted examples
of the class occur in the lower layers of the Upper Carboniferous
of North America. This early fauna consisted almost entirely
of Palseodictyoptera, Protorthoptera (including Protoblattoidea),

PERMIAN 87

Protodonata, Megasecoptera and Blattidac. With the exception
of the last mentioned, all these groups are extinct to-day — they
are ancestral forms from which more recent types have been
derived. None of these early orders gives any clue as to the origin
of insects as a class, while their abundance, together with their
relatively high grade of specialisation in some cases, clearly
indicate that more ancient relics of insect life still await disclosure
in antecedent rocks.

The primordial and heterogenous order Palasodictyoptera
appears to have already attained its maximum development
in the middle strata of the Upper Carboniferous and the famous
beds of Commentry in France, those of Mazon Creek in North
America and other beds of the same age in England and Belgium
have yielded many remarkable forms. This order, together with
the Megasecoptera, appears to have declined relatively quickly,
and towards the end of the Carboniferous age their remains
become very scanty. In addition to the orders mentioned, a
number of isolated fossils of doubtful affinities occur during this
age, and have been raised by Handlirsch to the rank of separate
orders. Among them are the j3rimordial mayfly, Triphlosoba
Handl.j which forms the sole representative of Handlirsch's order
Protephemeroidea, and the Hadentomoidea, which he regards as
forerunners of the Embioptera.

Many of the Carboniferous insects were of large size, and
included some of the giants of their class. For the most part they
appear to have been slow and weak fliers, a feature which may
have contributed to their early extinction. The Blattidse, however,
are an exception, since they have retained their general facies,
with comparatively slight modifications, down to the present day.

Permian. Recent discoveries, previously alluded to, have
revealed a rich and highly interesting fauna in Permian times.
In rocks of Lower Permian date are found the earliest known
Holometabola which are represented by several families of
Mecoptera (i;i<i^ Tillyard, 1935 ; Carpenter, 1930a) and Neuroptera
(Tillyard, 1932). Of the Mecoptera the family Permochoristida;
has thirteen species in five genera, while the single species of
Anomochorista gives its name to a second family. It is remarkable

88 PALEONTOLOGY

that even these early fossils already exhibit in the Aiioiuoehoristidte
a venation specialised by reduction. Among the Neuroptera are
foiind representatives of the existing Raphidiodea, in the genus
Permoraphidia. The SiaHodea are also met with in Marty novia
and Choristosialis. while the earliest known Planipennia appear
in Permoberotha. It is noteworthy that all these early Neuroptera
are comparatively highly specialised types. The curious beetle-
like Protelytroptera are confined to the various rocks of this
period, which also contain an abundant series of Megasecoptera.
In the same strata there also occur the earliest known true mayfly
(Protereisma) and a number of generalised Perlid-like forms which
constitute the order Protoperlaria. In the lower Rothliegende
of Birkenfeld. in Germany, there is to be found the remarkable
fossil EugereoH Dohrn., belonging to the order Protohemiptera.
It is noteworthy that the earliest undoubted fossil records of true
Hemiptera are of approximately the same age, and occur in the
Lower Permian of Kansas. They comprise Homoptera only, and
appear to indicate that both the divisions Sternorrhyncha and
Auchernorrhyncha were already clearly differentiated at that
period. True dragon-flies also appear in Lower Permian times, but
they differ sufficiently from living forms to form a separate sub-
order, the Protozygoptera. The remarkable genus Kennedya
is the most important type (Fig. 46, B), and it shows that in these
early dragon-flies the specialised discoidal cell and the sub-nodus
of recent forms were undeveloped and, at the same time, they
retained a vestige of vein Cu^ within the petiole of the wings
(Tillyard, 1925). The Psocoptera were also of great antiquity, and
recent researches (Tillyard, 1926c, 1928b, 1935b) have brought to
light examples of species which appear to represent two new and
extinct sub-orders- the Permopsocida and the Embiopsocida. In
the latter group is the genus Permemhia Till., which combines
apparent Embiid affinities with those of the Psocoptera.

Coming to Upper Permian times, rocks of this age in Australia
and Northern Russia have disclosed abundant records of insect
life which have been the subject of recent studies by Tillyard
and by Martynov. In Australia the Belmont beds of New
South Wales have afforded further evidence of the existence of

PERMIAN 89

Hemiptera-Homoptera , and among the latter the family
Scytinopteridse was dominant. The earliest true Plecoptera
make their appearance at this period. They are represented by
StenoperUdiuin (Tillyard, 1935a), a member of the existing family
Eustheniidse. The Paramecoptera, possible ancestors of both
the Trichoptera and Lepidoptera, make their appearance, and
along with them are the earliest known fossil Coleoptera. The
latter consist of two families, the Permophilidse and Permosynidae.
Of these, the first mentioned appear to be the direct ancestors of
the existing Hydrophilidae, while the Permosynidje evidently lead
on to genera existing in the Upper Triassic rocks. Remains of
what appear to be even more generalised Coleopteroid types occur
in the same strata as these ancient beetles, and they are regarded
by Tillyard as forming his order Protocoleoptera. Neuroptera
also occur during the Upper Permian age and are represented by
four genera forming the family Permithonidse.

From Russia Martynov (1928a) has described some early
examples of the Megaloptera in his genus Permosialis : althougli
related to living Sialidae its more specialised features have led
him to place it in a separate family — the Permosialidse. The
Planipennia are represented by the new family Palseomerobiidae
which appears to be allied to, but more j^rimitive than, the
Australian Permithonidae. Homoptera were apparently a
dominant group and the early families Prosbolida? and
Scytinopteridae were well rej^resented. Remains of what appear
to be May-flies allied to Protereisma, Mecoptera and Psocoptera
are also recorded, together with a number of forms which are
referred to his order Miomoptera (see p. 80).

All the ancient Carboniferous orders extended into Permian
times. Protorthoptera and Blattidae were abundant, but the
Megaseco2:)tera, although common in^the early Permian, appeared
to have died out at the end of this period : the genus Kuloja
Mart., from the Upper Permian of North Russia, appears to
be the last-known survivor, but the remains of this fossil are
too scanty for ascertaining their affinities with certainty.
Palaiodictyoptera, already much on the decline, are represented
by Dunharia Till, and Permoneura Carp, in the Lower Permian

90 PALEONTOLOGY

of Kansas, and possibly by some wing fragments from the Upper
Permian of Russia. Protodonata likewise were on the decline,
and among their last representatives are Calvertiella Till, and
Megatypus Till, from the Kansas beds.

Trias. Recent discoveries among Australian rocks of Triassic
age have revealed a rich insect fauna. In the Ipswich beds of
New South Wales Hemiptera show a greater development and
specialisation as compared with the previous epoch. In the family
Dunstaniidse Pentatomid affinities appear to be present, while the
Triassocoridse are possibly ancestral to the Notonectidse and
related aquatic forms. Homoptera were a dominant group and
were represented by a great variety of forms, including the large
Mesogereonida?, which were evidently ancestors of the Cicadidae,
numerous Fulgoroid^c and some early Jassidse. It is noteworthy
that in the discovery of Mesotitan Till, from these same beds
we have the second known genus of Protohemiptera. Among
Mecoptera, the sub-order Protomecoptera is represented by
Archipanorpa in a remarkably perfect condition of preservation,
while Aristopsyche and its allies, which were regarded by Tillyard
as forming a separate order (Paratrichoptera), also occur in the
Ipswich beds. Neuroptera of this age include the family
Prohemerobiidse, together with Triassopsy chops, which is a
forerunner of the existing Psychopsidae ; in the Trias of Europe
are found some remains of alder-flies. Among dragon-flies are
the genera T Has sole stes, a true Zygopteron ; and Triassagrion, a
member of the sub-order Archizygoptera which became extinct
in Liassic times. True Coleoptera were also abundant in the
form of elytra and include already specialised types, some of
which have been referred to the families Chrysomelida^ and
Curculionidai. It is also noteworthy that the Carboniferous
order Protorthoptera extends into the Trias, and some of the
Triassic Blattidae differ remarkably little from those present
during the Palaeozoic era. Accompanying these primitive
Orthopteroid types were true Acridiidae and forms related to the
Mantidae and Tettigoniidac.

Jurassic. The Jurassic period is notable from the fact that
all the main orders of insects, with the exception of the

JURASSIC 91

Lepidoptera, had come into existence before it terminated.
The earhest true Hymenoptera occur in Upper Jurassic strata
and include both Symphyta and Parasitica. In the Lower
Purbeck beds of England, but more especially in the contem-
poraneous Solenhofen slates of Bavaria, there are abundant
remains of primitive wood-wasps of the genus Pseudosirex. From
rocks of similar age at Montsech in Spain there has been found an
Ichneumonoid genus, Ephialtites, which is regarded by Handlirsch
as intermediate in character between the Symphyta and Apocrita.
The recent discoveries of Martynov (1925) in Turkestan have
brought to light (a) Paroryssus, a genus which appears to be
annectant between the living Oryssidae and the most primitive
Ichneumonoidea ; (h) Anaccyela, a form antecedent to Xiphidria ;
(c) Mesaulacinus, which is referred to the Evaniidae ; and {d) the
Proctotrypoid Mesohelorus.

True Diptera, belonging to the orthorrhaphous series, appear
for the first time in the Upper Lias of Europe, together with the
earliest known fossil Trichoptera. The Diptera include Rhyphidae,
Bibionidae and other Nematocerous families from the Upper Lias
of Mecklenberg and the Upper Jurassic of England. The remains
of Upper Rhaetic crane-flies, described by Wieland in 1925, from
South America, have since been referred by Tillyard to the
Homopterous family Scytinopteridae. In the Upper Lias of
England are also found early Diptera of the family Architipulidae
(Tillyard, 1933). The Trichoptera are referred to the family
Necrotaulidae, which has close affinities with the existing
Rhyacophilidae. In Liassic times in Europe dragon-flies of the
now almost extinct sub-order Anisozygoptera were abundant and
true Anisoptera first appear in the Upper Jurassic.. In rocks of
Middle Jurassic date a few wings and nymphs of some early
Plecoptera have been found in Siberia. Neuroptera are well
represented in the Lias and Upper Jurassic of Europe by the
Prohemerobiidae, the large and striking Calligrammatidae, together
with allied forms. Other dominant Jurassic insects include
Coleoptera, which are plentiful in rocks of this age in Europe,
while both sub-orders of the Hemiptera are represented by well-
preserved fossil remains. Several recent families of Heteroptera,

92 PALEONTOLOGY

including the Reduviidse, Lygaeida?, Hydrometridae and
Belostomatidae, existed at this period. The Pala^ontinidse from
the Solenhofen slates, which are regarded by Handlirsch as the
earliest known Lepidoptera, are referred by Tillyard to the
Homoptera and appear to have been Cicada-like forms. It is
also noteworthy that Mecoptera occur plentifully in the Lower
Lias of England. Among the genera are Orthophlebia, which
appears to be ancestral to the living Panorpidse, and Probittacus,
which connects the recent Bittacidse with the Permochoristidae,
The sub-order Paratrichoptera also extends into the Lower Lias
where it is represented by a fore wing of Liassophila from rocks
in Warwickshire (Tillyard, 1933).

Cretaceous. Only scanty and fragmentary remains of insects
occur in rocks of this date. On the whole, Coleoptera pre-
dominated in the Lower Cretaceous of England, and the Upper
Cretaceous of Bohemia. Fragments referable to other orders
have also been disclosed in the last-mentioned strata, but they
are, for the most part, too 230orly preserved for accurate deter-
mination. The most considerable recent discoveries among
Cretaceous insects are those of Ping (1929) in China, where a
wide and very scattered series of forms has been disclosed
ranging, among the higher orders, from Paraulicus (Evaniidse)
and Chironomidse to Perlidae and Blattidae. All the examples
disclosed in China appear to belong to existing families. Taken
as a whole the insect remains during Cretaceous times are scarce,
but this is not to be attributed to actual paucity of those animals
during that period, but rather to the absence of suitable fresh-
water deposits wherein examples might have accumulated and
become fossilised. The comparatively sudden appearance of
highly organised Lepidoptera, Aculeata and cyclorrhaphous
Diptera, in the early Tertiary period, suggests that these groups
must have been represented by forerunners in Cretaceous times.

Tertiary. Many of the Tertiary deposits are extremely rich in
insects and they have yielded nearly three-fourths of the known
fossil species, almost all modern orders being represented. Taken
as a whole the fauna of these times did not differ markedly in
composition from that of to-day — even during the Oligocene,

TERTIARY 93

which is relatively early in the Tertiary period, the insects were
very much the same as those which prevail at present.

Among the strata richest in insect remains are those of the
Green River of Wyoming, of Eocene date ; the famous Baltic
amber beds of Oligocene age, and the White River deposits of
Colorado ; the Miocene deposits of Florissant in Colorado, together
with those of Oeningen and Radaboj in Europe, and others.

Collembola and Thysanura are well represented in Baltic Amber,
and excepting the Devonian remains previously referred to (p. 86)
they are unknown from pre-Tertiary deposits. The Collembola,
which have been revised by Handschin (1926), are representatives
of living genera pertaining both to the Arthropleona and
Symphypleona, while the Thysanura comprise Machilidse,
Lepismidae and a species of Campodea.

The earliest undoubted Isoptera occurred in Eocene times and
became more abundant in the Oligocene and Miocene. The most
interesting forms are Mastotennes and Archotermopsis, which are
the most primitive known genera and are each rej^resented by a
single species at the present day. Specialised genera such as
Euterm.es and Termes were also living in Oligocene and Miocene
times.

Lepidoptera are present in rocks of Eocene age onwards, but it
is practically certain that the order is of pre-Tertiary origin, and
it may be added that certain leaf mines of Upper Cretaceous age
are very possibly the work of their larvae. In Lower Oligocene
times a few butterflies and more numerous moths, referable to
living families, occurred in Baltic amber, and some further
remains of the order have been disclosed in the deposits of Miocene
date. Fossil Lepidoptera, however, give no information relative
to the evolution of the order, and we have still to seek for the
existence of Homoneura in geological times. According to
Professor Cockerell, the Micro^jterygidae are represented by a
specimen in Burmese amber (in the British Museum) and this
seems to be the only record.

Coleoptera are extremely abundant in Tertiary rocks. A
number of modern genera were existing in Eocene times and more
than 430 species of beetles are known from Baltic amber, while

94 PALEONTOLOGY

an even larger number have been collected from the Miocene of
Florissant.

The earliest known Aculeate Hymenoptera date from the
Eocene j^eriod, and it is evident that they were one of the latest
of the great groups to be evolved. Both Parasitica and Aculeata
are abundant as fossils ; Chalcids of the genus Eurytoma are
described from the Eocene of Green River, and all the great
groups of the parasitic forms were existing in the Lower Oligocene.
Considerably over 300 species of ants are known from Tertiary
strata, and among them male, female and worker castes were
differentiated much as to-day. Wheeler has remarked upon the
absence of polymorphism among the worker caste, and it appears
that differences which separate individuals into true workers and
soldiers were not evident until the Pleistocene.

The higher Diptera occur as fossils in Eocene rocks and became
abundant in the Oligocene. Syrphidse are especially prevalent as
fossils, and this family, together with the Tachinida^, were already
existing in the Eocene period of the Green River in Wyoming.

Little needs to be said of the remaining orders, since most are
represented in Tertiary times, including the smaller groups
Embiojitera, Strepsiptera and Aphaniptera.

The Phylogeny of Recent and Fossil Orders of Insects

The earliest fossil remains of insects found in Carboniferous
rocks were exopterygote forms endowed with large and tolerably
efficient wings. Along with them are fragments of contem-
poraneous nymphs, thus indicating that these insects underwent
incomplete metamorphosis. Many of these ancient insects are
grouped to form the order PaLneodictyoptera which comprises a
large assemblage of generalised forms divisible into a number of
families. New genera are frequently being discovered and
knowledge of the order has greatly developed in recent years. A
revisional monograph is much needed to-day wherein the vena-
tional features of the earlier described forms are re-studied in the
light of more recent knowledge. Much the same condition of
affairs prevails with regard to other Palaeozoic insects and the
limits of some of the orders require re-definition. There is

PHYLOGENY

95

undoubtedly a good deal of intergrading among different groups
wherein differentiation has not proceeded very far, and many
forms which have to be regarded as of generic status display
characters which would be classed as of family, or higher,
significance, among recent forms.

The Protorthoptera exhibit definite advance on the
Palaeodictyoptera ; for example, the structure of their thorax
indicates that they had developed some capacity for running,
whereas the ancestral order probably had feeble powers of
terrestrial locomotion and resorted to short spontaneous flights.
These early Orthopteroid forms show increased development of
the hind wings, and the expanded anal area of those organs is a

Cu, lA *,♦,

Fig. 46. ' A, Brodia priscotincta Scud, wing (58 mm. long), with
cross- veins omitted. B, Kennedya rnirabilis Till, wing (44 mm.
long). (Adapted from Tillyard, Rec. Ind. Mus., XXX., 1928.)

distinct advance upon the closely similar fore and hind wings of
the Palaeodictyoptera. The Protoblattoidea were early cockroach-
like forms which connected the Protorthoptera with the true
Blattidge that co-existed with them. Other of the Protorthoptera
exhibit modifications in different directions, and (Edischia Brongn.,
for example, has the unmistakable facies of a primordial long-
horned grasshopper. Some of tKe Protorthopterans carried
pronotal expansions and thereby resembled the Palaeodictyoptera.
The presence of these expansions, however, is sometimes coupled
with evidences of an ovipositor — an organ which is not developed
in the Palaeodictyoptera.

The Megasecoptera were an evidently specialised order : the
narrowing of their wing-bases and the reduction of their venation

96 PALEONTOLOGY

indicate that they must have been a special offshoot from the
Pahrodictyoptera. In the genus Brodia Send. Tillyard (1928b)
claims that we have a type ancestral to the Protodonata and
Odonata and that these two orders are collateral developments
from the Megasecoptera (Fig. 46). Great stress has been laid
upon the disco\ cry of the primitive true dragon-fly Kemiedya
Till., and its bearing on the evolution of the Odonata. This
theory of Tillyard involves the derivation of the broad-winged
Anisopterous forms from narrow-winged Protozygoptera which
has been disputed by Carpenter (1931). As this writer points
out, Tillyard's grounds for his conclusions rest on a slender basis,
and, in view of more recent discoveries, his theory will have to be
abandoned. The discovery of the ancient sub-order Protanisoptera,

a

Fig. 47. \Ving of Ditaxineura anomalostigma Till.
(From Carpenter.)

represented by Ditaxineura (Fig. 47) from the Lower Permian,
shows that this type extends equally far back in geological time
as the Protozygoptera, represented by Kennedya. Without
going into a long and complex argument it may be affirmed that
no group ancestral to these two ancient sub-orders of dragon-flies
has yet been discovered. In Kennedya Cu.^ + lA combine to
form a long stalk traversing the petiole of the wing, whereas in
Ditaxineura there is no petiole and these two veins remain separate.
It would seem clear, therefore, that on this feature alone the
last-named genus is the more primitive. Tillyard's belief that the
Protozygoptera were derived from the Megasecoptera has very
slight evidence for its support. Carpenter sees no close affinity
between the last-named order and either the Protanisoptera or
the Protodonata. In our present state of knowledge we liave to

PHYLOGENY 97

recognise that the Odonata probably arose along two separate
lines of modification from common ancestors. These ancestors
were evidently related to the Protodonata, but the relationships
of the Protodonata with the Megasecoptera are obscure.

Arising apparently as another offshoot of the Pala^odictyoptera
are the Protoperlaria, which exhibit a type of venation fore-
shadowing the recent Plecoptera. The more advanced types of
Protoperlaria, such as Artinska Sell., show features ancestral to
the more primitive of the living Perloids and indicate that the
latter were probably descendants of the former. The presence of
prothoracic lobes and 5 -segmented tarsi in the Protoperlaria affords
characters which are well known to be wanting in all recent
Plecoptera. The discovery of a true Plecopteron, viz.,
Ste7ioperlidium (Tillyard, 1935a), which clearly belongs to the
living family Eustheniidse, is one of the more striking disclosures
in insect palaeontology. The fact that a representative of a living
family of insects should be preserved in Palaeozoic rocks (Upper
Permian) is only paralleled among the Blattida:.

The Ephemeroptera reveal evident Pala^odictyopteroid
characters in their primitive venation and their long cerci. The
Upper Carboniferous genus Triphlosoba. which constitutes the
order Protoephemeroidea of Handlirsch, appears to be related
to the genus Homaloneura of the Palaeodictyoptera. These
forms lead to the true May-flies of the Lower Permian epoch
which belong to the genus Protereisjua. In the latter insects
the fore and hind wings were closely alike in size and venation,
and the archaic living family of the Siphlonuridae has not diverged
very markedly from these early fossils.

Another ancient order — the Protohemiptera (Fig. 43)— had a
type of wing venation which suggests that it may have arisen
from a specialised family of the Palaeodictyoptera, but it combines
primitive venational characters with specialised Hemipterous
mouth-parts. Up to the present the fossil record has afforded
no indication as to how the latter organs were derived from
generalised biting jaws.

The Protohemiptera may, or may not, have been on the direct
line of descent of modern Hemiptera, but they indicate the

R.A. ENTOMOLOGY. 4

98 PALEONTOLOGY

probability that the latter order had a common ancestry with
the other Exopterygota mentioned. The most primitive true
Hemiptcra occurred as early as Lower Permian times, and their
most ancient families are referable to the sub-order Homoptera.
The oldest of the Heteroptera are the Dunstaniidse, from the
Upper Trias, but palaeontology has not so far shed much light on
their derivation and relationships with the Homoptera. The only
evidence so far revealed is the discovery of an upper wing in the
Trias of Queensland belonging apparently to the Homopterous
family Scytinopteridte (Auchenorrhyncha). According to Tillyard
(1936) this specimen shows evidence of transformation into a
true Heteropterous tegmen.

Discoveries of Lower Permian Homoptera, belonging to the
extinct family Archescytinidae, have demonstrated venational
affinities with early types of Psocoptera found along with them.
It appears likely that the Hemiptcra and Psocoptera are divergent
developments from a common stock and that, later on in geological
time, the Anoplura were evolved as descendants of wingless
Psocids which had become adapted for a parasitic life upon
warm-blooded vertebrates.

Mention has been previously made of the resemblance between
certain of these early Psocids and the Embioptera. If this
reveals anything more than parallelism, or convergence, we may
have to seek for the ancestry of the last-mentioned order in this
direction. The palseontological history of the Embioptera,
however, is so scanty that their ancestry remains enveloped in
obscurity, and no undoubted evidences of the order occur until
the Tertiary period. The Isoptera similarly have left no certain
indications of their ancestry. Structurally, they betray very
evident Blattoid characters and there is little doubt that they
arose from cockroach-like forms which subsequently developed a
complex social organisation. In Tertiary times their genera were
very much as w^e find them to-day, and it is evident that ancestral
forms will have to be sought in rocks of considerably earlier date.
Palaeontology likewise reveals nothing with respect to the affinities
of the Thysanoptera, but, from a study of recent forms, we are
led to conclude that they ha\'e more in common with the Hemiptcra

PHYLOGENY

99

than with any other order. Possibly they arose as offshoots of
some early Psocopterous group, but evidence of this contention is
not forthcoming.

The oldest order of the Holometabola are the Mecoptera which
were already well represented in Lower Permian times. The
venation of these early Mecoptera (Fig. 48, A), however, is so
fundamentally different from either that of the Palaeodictyoptera,
or of the Protorthoptera, that we are led to the conclusion that

/•

z::^^^-

li^^rffc

rs^v-

\.' /

r — ^■"'

L- / .'

ryM

2-5

Fig. 48. A, Permochorista jucunda Till, fore wing. x 6-5.
B, Belmontia mitchelli. Till, x 2-5. (Adapted from Tillyard.)

a long series of transitional forms must have existed in early
Carboniferous times, of which we have as yet no evidence.
Although its ancestry is, therefore, obscure, the order itself is
of great phylogenetic importance, since all the main groups of
the Holometabola (excepting the Coleoptera and Hymenoptera)
are perhaps referable to a Mecopterous ancestry The extinct
sub-order Protomecoptera, with its tendency to a more profusely
branched venation and the widened costal space, included Merope-
like forms which may well have been ancestral to the Neuroptera.
The sub-order Paramecoptera (Fig. 48, B) of the Upper Permian

100 PALEONTOLOGY

exhibit features ancestral to the Trichoptera and Lepidoptera on
the one hand, and which ally them with the Mecoptera on
the other. The Triassic Mesopsychida^ (forming the order
Paratrichoptera of Tillyard) combine characters which suggest
that they are on the Hue of descent of the Protodiptera. The
true Diptera, however, liave no known annectant forms con-
necting them with the Protodiptera.

Among Hymenoptera the venation is specialised to a degree
wliich proclaims the singularly isolated position of the order.
Even among the most primitive of the saw-flies there is no indication
of apparent relationship with any other orders.

The Protelytroptera (Tillyard, 1931) are a tegmen-bearing
order whose affinities are far from clear. That they bear an
apparently close resemblance to the Dermaptera as regards the
hind wing is obvious : furthermore, pentamerous tarsi are also
found in the Jurassic earwig Protdiplatys Mart, from Turkestan.
It is probable that they are to be regarded as early forerunners
of the earwigs, and according to Carpenter (1933) the venation
of the tegmina suggests their derivation from certain Blattids of
Upper Carboniferous date. The venation of both the tegmen and
hind wings is more specialised than that of the Protocoleoptera
and Tillyard's suggestion that the last-named order may prove a
remnant of the Protelytroptera — at a stage when the elytra were
still tegmen-like and flattened — seems to be in harmony with the
facts at present available. The scanty remains of Protocoleoptera
so far discovered, however, scarcely warrant any but very tentative
conclusions being drawn.

The conclusions to be derived from recent advances in insect
palaeontology may be briefly summarised as follows : —

1. At least nine recent orders are known to have existed from
Palaeozoic times. Of the remainder, the Trichoptera,
Hymenoptera and Diptera are known from Jurassic strata, while
the Thysanura, Isoptera and Lepidoptera have, so far, only
been traced back to the Tertiary period (Table I, p. 101).

2. Three recent families have persisted apparently from
Palaeozoic times, viz., the Podurida? (Collembola) from Middle
Devonian ; the Blattidae sen. lat. (Orthoptera) from Upper

PHYLOGENY
Table I.

101

i

UJ

1

L SU

<

U

m

in
<

3

>•

a

<

1-

>- 1-

Is

Is

^1

ORDERS ARRANGED ACCORDING
TO THE SEQUENCE OF THEIR

APPEARANCE AS FOSSILS
EXTINCT ORDERS IN ITALICS

—

COLLEMBOLA

P/JL /lEODIC T YOPTER/J
PROTODON/lTA,MEGASECOPTER/1

PROTORTHOPTER/I

ORTHOPTERA (SENSU L/^T.)

ODONATA.EPHEMEROPTERA.PSOCOPTERA
HEMIPTERA.MECOPTERA.NEUROPTERA

PROTOHEMIPTER/i

—

PROTOPERLARl/i, PROTELYTFZOPTERA

PROTOCOLEOPTER/i

PLECOPTERA

DERMAPTERA.THYSANOPTERA,
OIPTERA TRICHOPTERA.HYMENOPTERA

THYSANURA, ISOPTERA
EMBIOPTERA , LEPIDOPTERA

Carboniferous ; and the Eustheniidae (Plecoptera) from Upper
Permian.

3. The Palaeodictyoptera come nearest to the theoretical
conception of ancestral winged insects. The Protodonata.
Megasecoptera and Protorthoptera were contemporaneous with
them in Upper Carboniferous times. Since these three orders
were already sharply differentiated from the Palseodictyoptera it
is concluded that actual ancestral insects remain undisclosed in
still earlier strata.

4. The existing Ephemeroptera appear to be nearest related
to the Palaeodictyoptera (Dictyoneuridse), while the Orthoptera
and Plecoptera are clearly derivations of the Protorthoptera.
The Odonata apparently arose in common with the Protodonata,
but the ancestry of the Hemiptera and Psocoptera is obscure.

5. The majority of the Holometabalous orders appear to have

102 PALEONTOLOGY

arisen from Mecopteroid ancestors, and the origin of the latter
may have been from a speciahsed Palaiodictyopterid stock (Paolia
or alhes). The relationships of the Hymenoptera are unknown,
while those of the Coleoptera are very uncertain, but may have
been, through Protocoleoptera, from the Protorthoptera.

Literature

Carpenter, 1980. Psyche, XXXVII., 343.

1930a. JJull. Mus. Comp. Zool. Harvard., LXX., 09.

1931. Am. Jouru. Sci., XXL, 97.

1933. Proc. Am. Acad. Sci. and Arts, LXVIIL, 411.

1935. Ibid., LXX., 103.
Forbes, 1928. Psijchc, XXXV., 32.
Handlirsch, 1908. " Fossilen Insekten." Leipzig.

1925. '• Palajontologie," in Schroder's Handbuch der Entomologie, III.,
117.
Handschin, 192G. Ent. Mitt., Berlin, 161, 211, 330 ; also Zool. Anz., LXV.,

179.
Lameere, 1922. Bull. Acad. Roy. Belg., 138.
Martynov, 1925. Bull. Acad. Sci. U.K.S.S., XIX., 233, 569, 753.

1927. Zool. Anz., LXXIL, 99.

1928. Ann. Mag. Nat. Hist., Ser. X., I., 319.
1928a. Trav. Mus. Gcol. Leningrad, IV., 1.

1930. Ibid., VI., 69.

1931. Ibid., VIIL, 202.

Ping, 1928. Palceon. Sinica, XIIL, 1.

Tillyard, 1924. Proc. Linn. Soc. N.S.W., XLIX., 429.

1924a. Amer. Journ. Sci., VIIL, 111.

1925. Ibid., X., 41.

1925a. Ibid., X., 315.

1928. Amer. Journ, Sci., XVI., 185, 313.
1928a. Trans. Ent. Soc. London, 65.
1928b. Rec. Ind. Mus., XXX., 151.

1929. Nature, May 18th,

1931. Am. Journ. Sci., XXL, 232.

1932. Ibid., XXIIL, 1.

1933. "The Panorpoid Complex in the English Rhfetic and Lias."
British Museum (N.H.).

1935. Ann. Ent. Soc. Am., XXVIIL, 1.
1935a. Proc. Linn. Soc, N.S.W., LX., 374.
1935b. Ibid., LX., 265.

1936. Nature, October 24th, 719.
1936a. Amer. Journ. Sci., XXXIL, 435.

1937. Nature, January 9th, 66.

CHAPTER V
THE SENSE ORGANS AND REFLEX BEHAVIOUR

1. Introductory Remarks. The Sensory Neurons, p. 103;
Trichoid Sensillce, p. 104 ; Other Types of Sensillce, p. 104 ; The
Functions and Classification of Sensillce, p. 106. 2. The Eyes and Light
Perception. Compound Eyes, p. 107 ; Ocelli, p. 109 ; Differential
Response to Light Rays, p. 113 ; General Reflex Behaviour to Light, p. 121.
3. Chemical Stimuli and Their Receptor Organs, p. 123. The
Olfactory Sense, p. 124 ; The Gustatory Sense, p. 132.

1. Introductory Remarks
The Sensory Neurons. Sensory perception in insects is achieved
by means of a variety of receptors or sensilla^. It is possible that
the majority of types of sensillse are derivable originally from
primitive clothing hairs, differing from their fellows by having
acquired relations with the central nervous system. The
innervation is brought about by means of a specialised hypo-
dermal cell, at the base of the hair, which is connected with a nerve
axon. The hypodermal cell thus becomes a sense cell, but the
origin of its connection with the axon is an open question. When
the sensory nerves of insects are traced inwards they terminate
within the ganglia of the central nervous system as fine fibrillae
which form the sensory neuropiles. The only cells so far detected
which can be regarded as sensory cytons are those, just mentioned,
that lie at the peripheral terminations of the axons. If this con-
tention be correct, we are led to conclude that the axons develop
as inwardly growing processes from the hypodermal sense cells.
There appears to be no direct evidence, however, that this actually
takes place, and, on the other hand, it may be that the true
sensory cytons yet remain to be discovered. This latter view is
maintained by Vogel (1923), who followed the development of the
antennal nerve in the larva of the wasp. According to him the
nerve axons grow outwards and become secondarily connected

103

104 THE SENSE ORGANS AND REFLEX BEHAVIOUR

with the hypodermal sense cells. In the light of this finding he
concluded that the cytons related with this nerve are located in
the deiitocerebrum, but no investigator has yet revealed the
existence of any sensory cytons within the central ganglia. The
subject is more fully discussed by Snodgrass (1926), to whose
paper the reader is referred.

Trichoid Sensillae. The simplest type of sensilla is hair-like in
form, it is secreted by a special trichogenous cell and innervated

by a bipolar sense cell (Fig. 49). The
sense cell, it will be observed, lies within
the hypodermis, and its distal process,
after penetrating the trichogenous cell,
,sb passes to the base of the hair ; both
the sense cell and its distal process
are invested by neurilemma. Trichoid
"^ sensillge of this type are well described
by Sihler (1924) in his paper on the
sense organs of the cerci of various
insects. In many cases the tricho-

^ ^ o- , 1 • I J ojenous cell contains a large central

Fig. 49. Simple trichoid sen- *=" ^

silla from the cercus of vacuole in which is lodged the sensory

Gryllus campestris (integu- process of the nerve cell. It would

mental parts deep black). ^ i ^ . .

c, chitogenous cell ; h, hypo- appear that the vacuolar fluid extends

dermis ;>i, neurilemma ; ?w ^^-^^^ |.j^g cavity of the hair, and certain
nerve hbre ; sc, sense cell n - i- ■ i • i i

and its process sp ; t, base kmds of stmiuli received outside the

S-hwV '' ^''^"''^*'* ^^^""^ latter would be transmitted through

the fluid to the sensory process. More
usually a triclioid sensilla exhibits a more elaborate structure in
that the hair is attached to its socket by an articular membrane
which is the product of a special membrane cell, distinct from
the principal trichogenous cell. Membrane cells, which are little
more than individualised hypodermal cells, may also be detected
in association with ordinary covering hairs of a non-sensory
character.

Other Types of Sensillae. In the well-known classification of
Schenk most of the prevailing types of sensillae are grouped
according to characters afforded by their external cuticular parts.

SENSILL^

105

They differ in various ways from the simple trichoid type just
described, and may become either scale-Hke, peg-hke or conical ;
or, the hair shaft may be wanting and the cuticular structure
represented by plates, papillae, discs, or flattened domes. In
other cases, their only external manifestation is marked by a pit
or a mere cuticular thickening. Whatever their external form
may be, they are all j^resumably derived from originally simple
hairs. Furthermore, they all exhibit the same general similarity
of internal structure in that they consist of a trichogenous cell, a
membrane cell and a sense cell. In many cases the distal process
of the latter becomes differentiated at
its apex into a sense-rod or scolopale,
which is directly attached to the inner
side of the cuticular portion of the
sensilla. The exact nature of this
structure is not entirely clear, but its
staining reactions and optical appear-
ance seem to indicate that it is
cuticular in nature. This belief receives
support from Sihler's observation that
the scolopales in connection with
trichoid sensilla; on the cerci of the
Acridian Gomphocerus rufiis are cast at
each ecdysis. Structurally, scolopales
exhibit considerable diversity ; they
may be mere caps or cones over the
apices of the sensory processes, or they become elongate with
their walls thickened by a variable number of longitudinal ribs
(I'ig. 50). The extremity of such a rod often contains a deeply
staining apical body, which is connected with an axil fibre
extending through the sensory process into the nerve cell itself.
In other cases, there is no apical body, and the axial fibre
terminates at the extremity of the scolopale.

Most sensillae, in their simpler forms, pertain to the general
type just described. The chief exception is met with in the
visual receptors whose evolution is probably traceable originally
from mere localised areas of light-sensitive pigment. Visual

Fig. 50. Scolopalae or ter-
minations of sensory pro-
cesses of A, a trichoid sen-
silla of Gryllus (from
Sihler), and B, a campani-
form sensilla of Dytiscus
(from Hochreiither). a.
Apical body ; «/, axial
fibre ; r, rib ; sd, sense
dome or cuticular covering
of sensilla.

106 THE SENSE ORGANS AND REFLEX BEHAVIOUR

receptors, therefore, bear no relation to trichoid sensillse or their
derivatives.

The Functions and Classification of Sensillae. Recent research,
employing modern refined methods of technique, has added much
to our knowledge of the histology of different types of sensillse.
Experimental studies have extended our acquaintance with the
reactions of insects to various stimuli. Notwithstanding know-
ledge thus gained, it has to be admitted that the fundamental
aspects of the subject are, as yet, but little explored. We are not
in a position to classify many of the prevailing kinds of sensory
receptors on a functional basis, and are forced, in consequence, to
attempt to argue function from structural characters alone.
Data afforded by the structure of sensillae and the positions they
occupy on the body are valuable up to a point, but the ultimate
test is obviously the experimental one. Experiments upon the
living insect enable general ideas to be gained with respect to
its behaviour in relation to different stimuli. Amputation of
particular appendages or of defined regions of the body, their
cauterisation, or the coating of such parts with varnishes, wax,
or other material, enables the positions of specific receptors to
be located. Histological examination of the same parts may
enable the existence of a particular type of receptor organs to be
correlated with a specific kind of response of the individual.
When, however, the receptors are of a heterogenous character, the
method immediately becomes encompassed with difficulties. It
would appear, therefore, that the main stumbling block is the
paucity of exact experimental evidence. This, in its turn, is
bound up with the difficulties attending the investigations of
organs of minute size, often of scattered distribution and often
intermingled with those of a different type. These few remarks
will serve to emphasise the general character of the problems
involved, and we may now proceed to discuss certain of the
more outstanding facts and principles of reflex behaviour among
insects. It will be convenient for present purposes to deal with
the subject under the headings of the different kinds of external
stimuli ; to discuss the organs accredited to the reception of such
stimuli and to examine the nature of the responses involved.

COMPOUND EYES 107

2. The Eyes and Light Perception

The eyes, or photo-receptors of insects, vary greatly in
structure and degree of development. Almost every condition
of differentiation of these organs is met with among various
insects or their larvae. In their simplest phase the eyes consist of
single visual elements, or sensillai, sensitive mainly to variations
in light intensity. The most complex and highly developed
organs consist of aggregations of many thousands of sensillse,
and are capable of forming images of external objects. It has
been customary to group the different types of visual organs of
insects into two categories, viz., compound eyes and ocelli, and
certain features respecting them will be briefly considered.

Compound Eyes. Compound eyes are composed of variable
numbers of individual sensillse, or ommatidia, closely apposed to
one another. In certain aberrant or degenerate forms only a few
ommatidia, or even a single ommatidium, may compose the eye,
but, as a rule, these elements are much more numerous, and 20,000
or more may be involved. The well-developed compound eye is
the highest type of visual organ found in insects. Notwithstanding
its complex structure, it is a less perfect optical instrument than
the vertebrate eye, for the reason that no general focussing
mechanism is known to be present. In the dragon-flies, Vigier
(1904) states that a means of accommodation, consisting of
myofibrils and tracheal branches, is present between the
ommatidia. By the contraction of these fibres compression of
the crystalline cones is presumed to result, while the action of
air pressure on the tracheae may allow of a certain amount of
elongation. This interesting observation has attracted but little
attention, and the subject is one requiring fuller exploration. In
the general absence of any focussing mechanism, it follows that
in the vast majority of insects the capability of compound eyes
for image formation can only be effective in producing a clear
impression within a very limited focal distance. The distinctness
of the image, therefore, depends partly upon the nearness of the
object concerned. This, however, is not the only factor, since
the number and size of the ommatidia are also important, for a
larger number of small ommatidia will, on the mosaic theory,

108 THE SENSE ORGANS AND REFLEX BEHAVIOUR

produce a sharper and more detailed impression than if the same
visual area be occupied by a smaller number of larger visual
elements. The unknown photo-receptive capacity of the nervous
centres within the optic lobes must also be involved.

It is well known that compound eyes are divisible into two main
types, namely, those giving an apposition image, and those giving
a superposition image. In the first type the vision is essentially
diurnal, and the images are composed of a large number of apposed
points of light : such images are true mosaics as implied by the
theory of mosaic vision. It has been generally believed that the
direct image so formed is an inverted one, but apparently this
is not always the case, since Eltringham (1919) has observed a
direct image with the eucone eye of a butterfly. Compound eyes
giving a superposition image differ from the previous type in the
disposition of the components of their ommatidia. The rhabdoms
are not in contact with the apices of the crystalline cones, and the
two elements are separated by a space filled with a transparent
medium. The ommatidia in this kind of eye are consequently
greatly elongated. The distribution of the pigment surrounding
these elements is capable of adjustment according to the amount
of light available, and consequently this type of eye is most highly
developed in nocturnal or crepuscular species. At night time the
pigment congregates far forward, so as to expose a larger area of
the apices of the crystalline cones, with the result that rays of
light entering adjacent ommatidia traverse the space already
alluded to, and contribute to those entering the same retinula. In
this type of vision there is an overlapping of points of light, and
the image thus formed is termed a superposition image. It will be
evident that, in an eye of this description, a limited amount of light
will produce a better image than in an eye giving an apposed image
where so much of the light is absorbed by pigment. The eyes of
nocturnal insects are adapted, therefore, lor the perception of
movement and general form only, and, as already stated, they are
able to take advantage of varying degrees of light. In the presence
of daylight the pigment moves backwards like a dark sleeve
cutting off more and more of the peripheral rays, thus tending to
give a sharper image, and one more resembling the apposition type.

COMPOUND EYES 109

It will be obvious that in conducting experiments relative to
insect vision it is important to take into account the type of com-
pound eye present in the subjects utihsed. So far no comparative
studies have been made with respect to the photic reactions in the
two types mentioned, and a wide field for enquiry remains for
future investigators. Apparently, most observations have been
made with insects whose eyes give apposition images. Such types
are essentially adapted for diurnal vision only. Their focal range
as regards form-perception is very limited, and, in the case of
Vmiessa urticce, Eltringham gives reasons for concluding that it
does not exceed 3 feet. Perception of movement, it would appear,
may be effective over a greater distance, and, in the case of dragon-
flies, Tillyard has observed that the species Petalvra gigantea can
be frightened away b}' waving a net at a distance of 20 yards.
In this connection the observations of Vigier with regard to the
existence of a capacity for focal accommodation may be significant.
While appreciation of form is probably of great importance in the
pairing of sjiecies, there is a considerable evidence that compound
eyes are endowed more especially for the perception of movement,
and that insects rely upon this capacity to escape from their
enemies. With eyes adapted for daylight, the movement of even
small objects in the field of vision would be registered as a fleeting
impression on the retinal cells, and it becomes transmitted to the
brain of the insect, which responds accordingly. It is often possible
to approach an insect without occasioning its alarm. If the
approach be sufficiently slov^ and gradual the change of position
passes apparently unnoticed. The explanation appears to be that
an object as large as a human being affects all the ommatidia more
or less equally and simultaneously. Since the movement occurs
very slowly, the change of position entails only very slight and
gradual changes in the character of the retinal image as a whole.
A sudden movement of a part of the body of the person, however,
is sufficient to bring about immediate departure of the insect, for
the reason that it has caused an abrupt change in the retinal
image by affecting a number of ommatidia in rapid succession.

Ocelli. Ocelli present so many different types that they are
not definable upon any common structural basis. In some few

110 THE SENSE ORGANS AND BE FLEX BEHAVIOUR

instances they appear to be the degenerate vestiges of compound
eyes, but for the majority they have no direct relations with the
latter organs and, if there is any community of ancestry in the
two cases, evidence of such is not forthcoming within the class
Insecta. Ocelli may be grouped into two classes, viz., dorsal or
primitive ocelli and lateral or adaptive ocelli.

Dorsal ocelli are characteristic of nymphs and adults and occur
in all the chief orders of Pterygote insects, but are not present in
larvae. They share the common structural feature of being
composed of a greater or smaller number of retinulae grouped
together in relation with a single l^ns. In the compound eye it
will be recalled that there is a separate lens, or facet, in direct
association with each visual element or ommatidium. Dorsal
ocelli differ from compound eyes as regards their innervation.
Whereas the nerves supplying the compound eyes are in connection
with the optic lobes, the ocellar nerves are traceable to ocellar
lobes located in the protocerebrum, between the mushroom bodies.

The typical number of dorsal ocelli is three : they are disposed
as a triangle on the head-capsule, with the apex median in position
and indicated by the unpaired ocellus of the group. There is a
certain amount of evidence in favour of the view that the median
ocellus represents the fusian product of original paired organs,
and that the primitive number of dorsal ocelli was four. This
conclusion is based upon the fact that among Hymenoptera the
nerve supply of the median ocellus is frequently double. Further-
more, as Patten showed a number of years ago, the median ocellus
in the wasp is represented in the prepupa or late larva by a pair
of pit-like rudiments which ultimately coalesce. Evidence of
paired structure is also present in the median ocellus of Odonata
and of Bomhus, while Wheeler (1936) has recorded a high
percentage of double anterior ocelli in the ant Atta cephalotes.

Glaser (1925) records an example of the grasshopper Melanoplus
differ eniialis in which the usual median ocellus is replaced by paired
organs. These latter, however, are of smaller size and their total
superficial area scarcely exceeded that of a normal unpaired ocellus.
He also refers to records by Blackman of a similar phenomenon in
another species of the same genus, and by Smulyan in a species of
saw-fly. That these tetralogical examples may have a phylogenetic

OCELLI 111

significance, and represent reversions to an ancestral condition, will
immediately suggest itself in the light of foregoing remarks. Such an
explanation, however plausible, remains an open question in the absence
of any knowledge of the directing causes involved.

Little is known with respect to the functions of the dorsal
ocelli. The fact that the lens is strongly biconvex indicates that
their visual powers are limited to the perception of very near
objects. The small number of visual elements which compose
an ocellus is strongly suggestive that any image-forming capacity
it may possess would be of a crude and indefinite kind. The
perfection and complexity of the structure of ocelH in many
insects argue that they are functionally important organs, but
it needs to be emphasised that the power of perceiving even only
crude images apparently necessitates a somewhat complex visual
organ.

It has been suggested that the dorsal ocelli are adapted for the
perception of very near objects and for vision in darkness or
subdued light. This view^ seems, however, to have no proper
evidence based upon actual experiment. In so far as ants and
bees are concerned, it has been shown by Miiller (1931) that when
the compound eyes of these are painted over the creatures no
longer exhibit the capacity of orientating themselves in relation
to light. A growing opinion favours the idea that dorsal ocelli
are general stimulatory organs which have the property of " toning
up " the nervous system and increasing general sensitivity of the
brain. It is now believed that apart from any imperfect photo-
tactic function the ocelli exercise a photokinetic influence in, at
any rate, day-flying insects, by helping to maintain the sensitivity
of the visual centre of the brain to light stimuli. It is possible,
therefore, that the dorsal ocelli perform this function as imjDortant
accessories to the compound eyes. Here again the experimental
evidence is rather meagre, but Bozler (1925) has shown that in
Drosophila the response to light stimuli, received through the
compound eyes, is more acute and lasting w^hen the dorsal ocelli
are in their normal condition than when painted over with an
opaque substance. A similar function has been ascribed to the
ocelli of Lepidoptera by Friedrich (1931).

112 THE SENSE ORGANS AND REFLEX BEHAVIOUR

Lateral ocelli present no general uniformity of structure. They
are the functional visual organs of insect larvae and are antecedent
to the future compound eyes of the imago. It has already been
explained in an earlier chapter that insect larvae represent phylo-
genetically earlier phases of development than either nymphs or
imagines. They issue from the egg at a stage before the adult eyes
appear, and consequently have acquired provisional visual organs
that are functional only so long as the larval period lasts.

In their very simple
form, as seen in larvae
of Lu cilia, for example,
the ocelli consist merely
of groups of bipolar
sensory cells, devoid of
pigment, while the
overlying cuticle has
been modified into a
lens-like structure for
the condensation and
transmission of incident
light rays (Fig. 51).
According to Ellsworth
(1933) these larvae are
extremely sensitive to
light, and by localising
the latter into a small
" pencil " he was able
to show that this
sensitivity is confined to the organs just mentioned. In larva*
of the Lepidoptera' certain Coleoptera and Neuroptera the
lateral ocelli occur, as a rule, in a group on either side of the
head and in close association with the site where the imaginal
eyes are to be developed. In these orders each ocellus is usually
separate from its fellows : it is provided with a cuticular lens
with differentiated visual cells and sensory rods beneath. In
larvae of the saw-flies there is a single lateral ocellus on either
side which exhibits a structure very similar to a typical dorsal

Fig. 51. Section tlirough a pair of ocelli of a
Lucilia larva. x 450. c, collar ; cy,
cuticular cylinder ; e, outer cuticle ; e^,
inner cuticle ; /, nerve fibrils from retinal
cells, r ; ^, hypodermis ; /, cuticular lens ;
m, limiting membrane ; n, nerve ; r,
retinal cell ; r^, nucleus of retinal cell ;
.V, sheath, within which the lens moves
outward or inward ; v, prcretinal membrane.
(From Ellsworth. 1933.)

RESPONSE TO LIGHT 113

ocellus. The Collembola are exceptional in that their so-called
ocelli are the functional adult eyes : they occur in a group com-
prising up to eight such organs on either side of the head, and each
ocellus has the general structure of an ommatidium. It will be
obvious that the visual properties of ocelli of these diverse types
will vary more or less directly in relation with their structure.
Mere groups of pigmentary sense cells can appreciate little beyond
variations in light intensity. The more highly developed lateral
ocelli appear to have visual properties adaj^ted for near vision
only, but their image-forming capacity can only be of a rudi-
mentary kind. Certain experiments conducted with lepidopterous
larva:* appear to show that they are only able to perceive objects
at a distance of about 2 cm.

Differential Response to Light Rays. Recent research on
insect vision substantiates the conclusion that colour perception
exists in insects but that it does not apply to the whole range of
the spectrum visible to human beings and, at the same time,
involves the application of rays entirely invisible to man. In
his well-known experiments von Frisch (1914, 1924) has shown
that the honey bee, which has been the chief subject of colour-
vision experiments in insects, can be conditioned so as to associate
certain colours with food. A number of pieces of paper of diverse
colours, but of the same shape and size, were disposed in a sort
of chequer-board fashion. A watch-glass containing an inodorous
sugar solution is placed over one of the colours. After this has
been repeated a number of times the food-solution is omitted,
but it was found that the bees still went straight to the colour
upon which the watch-glass had previously been placed. A
series of fifteen shades from white through various tints of grey
up to black were laid out, and a blue paper introduced. An
empty watch-glass was then placed over each piece of paper and
the whole covered with a sheet of glass. Bees previously con-
ditioned to blue at once gathered over that colour, notwithstanding
repeated changes in its position, when they searched for the
non-existent food. Von Frisch showed that bees conditioned on
scarlet or black failed to discriminate between those two colours,
or dark grey. He therefore concluded that bees are blind to red.

114 THE SENSE ORGANS AND REFLEX BEHAVIOUR

He also found that bees are capable of being conditioned to
yellow, but that they failed to discriminate between that colour
and orange or yellow-green : similarly, bees conditioned to blue
turned indiscriminately to violet or purple. Von Frisch failed to
train bees to visit any particular shade of grey in preference to
any other shade. This important result affords a strong argument
against the objection that a highly developed response, on the
part of the insects, to different degrees of brightness might account
for their supposed differential colour responses. Von Frisch's
conclusions proved to be sound when training experiments with
bees were repeated subsequently, by using monochromatic light,
by Kiihn (1927) in Gottingen. In addition to the two main
colours (yellow and blue), which von Frisch found to be responded
to by the bee, Kuhn was able to include blue-green and ultra-
violet. It appears, therefore, that four regions of the spectrum
exercise separate and definite stimulus qualities with respect to
the bee, i.e., 650 i to 500 /x/x, 500 to 480 /x/x, 480 to 400 /x/x, 400 to
313 /x/Lt. By training insects along lines very similar to those
adopted by von Frisch, Knoll (1921) found that the Humming
Bird Moth (Macroglossa stellatarum) showed good capacity for
differentiating between yellow and blue, but was likewise unable
to discriminate between red and black. It must not, however,
be assumed that all insects are red blind, since Kugler (1930)
succeeded in training Bombus to a pure red Hering paper and Use
(1928) found that in Pieris and related butterflies responses are
mainly to red and to purple papers, while in the case of Vanessa
they were to yellow and to violet. Both Knoll (1921) and Use
(1928) state that a predilection for definite colours is inborn in
butterflies since the latter visited certain coloured but inodorous
surfaces, in preference to all shades of brightness, just after
emergence from pupa?.

Further important experiments, bearing upon colour vision in
insects, have been carried out by means of a screen which revolves
round a circular horizontal platform covered with white paper

1 According to Bertholf (1931) the upper limit for the bee extends to at
least wavelength 677 ^i^ in the red end of the spectrum ; a difference in the
intensity of the spectral light used probably explains why these results do not
conform with those obtained by Kiihn.

RESPONSE TO LIGHT 115

upon which the insects are placed. SchUeper (1927) first showed
by means of experiments, in which the wall of the screen was
lined with adjacent vertical stripes of different shades of grey and
of various colours, that insects do not respond to the movement
of the pattern if the grey and coloured stripes appear of equal
brightness for the eyes of the creatures (an analogous effect is
known in human sense physiology). By this means it was possible
to determine, not only the relative brightness of different coloured
papers, but also to discover different colours which appear of
equal brightness for a given insect. If the insect responds to the
movement of a pattern, which consists only of stripes of equal
brightness, but of different colour, it becomes evident that it is
able to differentiate colours. If under the same circumstances it

•600 700 WO 500 -400 -300 -ZOO 100

^ — ii Mi i li " '1^ J ^ ^

I ^ I I I ! !

Fig. 52. Approximate limits of visible spectrum for man {a-h) and
the bee {c-d).

does not respond, the insect may be regarded as being colour-blind.
By using this method Schlegtendahl (1934) showed that several
beetles, including Geotrupes, the fly Fannia canalicularis, and also
some night-flying Noctuidae, all show evidence of colour vision.

Little is known concerning the actual physiological nature of
colour vision in insects. Kuhn (1927) made the suggestion that
blue and yellow in the one case and blue-green and ultra-violet in
the other might be complementary colours. At present, however,
there is very little evidence available in connection with this
suggestion. Kuhn, however, carried out experiments wherein he
trained bees to visit blue surfaces and made sure that there was
no interfering tendency to visit the yellow. He then exposed a
series of square pieces of card of different shades of grey and
placed one single yellow card among them . Over each card was a
paper ring of a contrasting shade of grey. He noted that the bees

116 THE SENSE ORGANS AND REFLEX BEHAVIOUR

visited in numbers the grey ring on the yellow background, but
the other grey cards received no visits. The conclusion seems to
be that where grey is surrounded by yellow, in the above manner,
it produces a physiological effect of blue in the eye of the bee.
The experiment also affords further evidence that mere brightness
of shade is not involved or the bees would have visited some or
other of the grey cards.

The experiments of Hertz (1931-35) on the recognition of form
and pattern by bees are significant. It was shown that, as a
whole, bees tend to visit those patterns which show the largest

Fig. 53. The amount of outline displayed in patterns a and h is
approximately the same, and bees can be trained to visit either.
Nevertheless, after short training bees showed a much higher
preference for a, while b becomes neglected when both patterns
are exposed simultaneously. The preference is for complex
patterns of many small units in variable directions, as compared
with regular patterns of few units and simple lines. (After
Hertz.)

amoimt of contours or outlines within a given dimension. Hertz
distributed in various ways several black stripes of equal length
and breadth in areas of similar size. There resulted different
patterns with the same amount of contrast and of black and
white. The behaviour of the bees proved that patterns with the
most crossing-over and isolated units were much more attractive
than those which mainly showed regular parallel or concentric lines.
Thus in Fig. 53, for example, the pattern of crosses was visited in
preference to the concentric circles. In a similar way it was
found that bees could be conditioned to visit either patterns Pi 5
and F12, or G5 and P20 in Fig. 54, but not Pl5 and G5 or Fl2

VISION

117

and P20. It is easy to condition bees so that they visit a pattern
composed of relatively many contours or outhnes such as crosses
or a chessboard and to avoid a plain circle or square of the same
colour. So far, however, it has not proved possible to condition
bees to show the opposite preference. Hertz's conclusion that a
predilection for patterns with many contrasting outlines is present
in bees, before any training to \'isit specific patterns has taken
place, has been confirmed by Zerrhan (1933). These san.e results

Pl5

ri2

G5

P20

Fig. 54. In this figure all the patterns extend over equal sized
superficial areas. G5 shows the largest amount of outlines
(2,000 mm.), but, because of its uniformity, it is less attractive
to bees than P20, which is composed of numerous small units
with a greatly reduced amount of outlines (570 mm.). F12
(1,600), however, is much more attractive than either of the
foregoing patterns ; it is only eclipsed in attractiveness by one
composed of a very large number of small units P15, but with
reduced amount of outlines (1,026 mm.). An experiment of
this kind forms an infrequent example in which bees can be
trained both ways with regard to a pair of patterns. Thus, P20
is more attractive than G5 but less so than F12. (After Hertz.)

have also been obtained for other insects. Use (1932) used the
same patterns as those employed by Hertz in connection with
experiments with butterflies. She found that if the colour of the
pattern were very attractive to the insect that pattern which had
the largest area of colour and less outlines was preferred to a
pattern of smaller area but greater amount of outlines. Where,
however, the colour used only exercised a weak stimulus, the
behaviour of the butterflies was of the reverse kind and the same
response was obtained as that described by Hertz. The last-

118 THE SENSE ORGANS AND REFLEX BEHAVIOUR

mentioned observer (Hertz, 1931) carried out further experiments,
using three-dimensional objects fashioned out of white paper and
placed on a white background. These different objects were
characterised only by their form and the distribution of shadows.
An overwhehning response was obtained, showing preference for
the more differentiated forms over the simple ones. Thus in

Fig. 55. In this figure a, c, d are three-dimensional objects and b is a
Hat object ; all are made of the same material as the background
(white paper). Bees can be trained to visit any of these objects
when exposed singly. When they are exposed simultaneously
high preference is shown for the object with the most complex
form (a) which is due to the large amount of contrasting outline
and shadows. (After Hertz.)

Fig. 55 that simulating a flower {a) was preferred to the plain
flat object (b). It was further noted that the pattern of the
shadows exercises an important influence. The more complicated
the shadows of an object are the more attractive the object
becomes : also, the deeper and darker the shadows on the objects
the more attractive they are. Certain experiments were carried
out by von Buddenbrock (1935) by using other insects such as

VISION 119

Coccinella and Eristalis. The insects were placed at some distance
from a light towards which they walked by a straight path
flanked by cardboard walls bearing a pattern of black and white
stripes. The insects became regularly attracted off the direct
path to the wall which bore a pattern or, if the pattern were on
both walls, to that wall which bore the pattern containing the
most detail or outlines. There was no difference in effect between
vertical and horizontal stripes.

This capacity for a primitive discrimination of pattern seems to
be intimately connected with the responses of insects to visible
movements. Hertz (1933) showed that so long as a bee is moving
the image of the surrounding objects is moving on the retinulae
of the compound eyes. Similarly, a type of pattern with many
outlines involves a high frequency of alternating stimuli on the
retinulae. Great emphasis is laid on this feature by Wolf (1933),
who, as the result of his own experiments with intermittent
illumination, suggests that the whole effect of pattern on the
insect eye is nothing but that of flickering.

The two principal qualities of pattern, i.e., the amount and
effectiveness of the distribution of outline, also play an important
role in the reaction of insects to visible movement. In experiments
where an insect is in the middle of a small table around which
a cylindrical screen, with black and white vertical stripes, is
revolving the animal moves in the same direction. Up to a
certain limit the effect increases with the velocity of the movement
and with the number of stripes {i.e., the amount of outlines) :
this has been shown by Wolf (1933) for the bee and by Gaffron
(1934) for flies. Hertz (1934) found that for flies where there
are a number of stripes an equidistant distribution of outlines is
more effective than an alternation of narrower and broader
intervals between the stripes. The best effect to be obtained
with the minimum amount of outlines is by means of small spots
regularly distributed on the revolving screen. As a consequence
of the foregoing experiments, the question arises as to whether the
insect is able to differentiate between the appearance of movement
resulting from a change of position to surrounding objects and
from the movements of the animal itself in relation to such

120 THE SENSE ORGANS AND REFLEX BEHAVIOUR

objects. On the whole it appears quite improbable, but experi-
ments dealing with problems of this kind are very few. Gaffron
raised the (question as to whether stroboscopic movement and
after-images of movement occur in insects also. The answer
appears to be partly in the positive and partly in the negative.
The nymph of the dragon-fly Mschna responds to the movement
of any small object and will never respond to any kind of food
which is stationary. This creature can, however, be made to do
so by previous or simultaneous strong movements in the surround-
ing field, such as a screen which revolves around the receptacle in
which the animal is contained. When such movement has been
actuated for about ten minutes, the after-effect lasts for about
half an hour after such movement has ceased. Flies which are
responding very regularly to a striped revolving screen cease to
move if this real movement is replaced by an irregular appearance
and disappearance of the stripes at different points on the screen
(which produce a semblance to movement in the human eye).
Thus, there appears to be no appreciation of stroboscopic move-
ment in these insects at any rate. On the other hand, insects
like Coccinella, which are known to display light-compass reactions,
do respond to the interrupted appearance and disappearance of a
pattern if the change be not too fast. Experiments of this kind
show that it is necessary to differentiate between movement
responses and light-compass reactions which have sometimes been
confused with them.

In connection with the subject of light intensity and its wave-
length, it may be noted that the most modern work of this kind
has been carried out by Sander (1933) in Gottingen. In his
experiments the light energy of twelve different spectral lines,
including ultra-violet (366 fifi), was made equal to that of the
weakest line, the absolute amount of energy of any line being
436-10-^ watt per cm.^ Each one of the twelve different rays
was compared with another in the two windows of a small dark
cage in which a group of bees was confined. The number of bees
passing, during repeated exposures, at the two windows was
counted and results plotted in twelve curves. The curves showed
two maxima — in the yellow and blue ; and three minima — in

ULTRA-VIOLET RAYS 121

the orange, green and ultra-violet. Rays of a wavelength longer
than 630 /x/x were not visited at all. By raising the energy of a
less effective colour its stimulatory influence could be made equal
to that of a more effective colour. Sander further determined that
with different wavelengths the effects do not increase in the same
degree with the energy. The increase proved to be relatively
slow with the green rays and faster near the ends of the spectrum.
It thus appeared that the effects of light of different wavelengths
of equal energy may be very different for different amounts of
absolute energy. Thus it may be that the effect of green is
greatest in very low and that of ultra-violet in very high intensity.

A relatively strong stimulatory effect of ultra-violet has been
observed by Peterson and Heussler (1928) in their experiments
on the photic responses of the Oriental Peach Moth (Laspeyresia
lyiolesta) and the Codling Moth (Cydia pomonella). Bertholf
(1931) also obtained a strong response to ultra-violet in the case
of both Drosophila and the hive bee. In the case of the last-
named insect he found that there are two zones of maximum
stimulative efficiency — one in the yellow-green at wavelength
553 /A/x and the other in the ultra-violet at 365 /xjla — the latter
being about four and one-half times as great as the former. In
view of the work of Sander, which stresses the unequal increase of
effect in different wavelengths, it is possible that Bertholf's con-
clusions may require modification. Bertholf, it may be added,
used lights of unequal energies and determined the relative effect
of equal energies by calculation. It is noteworthy that Priebatsch
(1933) found that the relative responses to different wavelengths
alter when the amount of energy is altered. With Carausius
rrwrosus it was found that this insect becomes black when resting
on a black background and becomes darker in proportion to the
amount of illumination. The effect, however, does not depend
on the energy alone. If monochromatic light of equal and
relatively high energy is used the effect is most pronounced in
ultra-violet light, whereas with weak illumination the effect of
green is the most pronounced.

General Reflex Behaviour to Light. The responses of insects
to rays of specific colours have already been discussed, and there

122 THE SENSE ORGANS AND REFLEX BEHAVIOUR

remains for consideration their reactions to ordinary white Hght
and its absence. Positive and negative responses to Hght have
been explained by Loeb (1918) on the basis of tropisms. His
theory states that when a bilaterally symmetrical animal responds
positively to a given stimulus, it approaches that stimulus owing
to the continuous action of the latter in provoking symmetrical
muscular activity. If the stimulus does not fall equally on the
two sides of the body, it provokes asymmetrical muscular activity,
which results in the animal orienting itself so that its long axis
comes directed towards the source of the stimulus. If the response
be negative, the muscles of the side nearest the source of the
stimulus are less active, which has the result of turning the
animal away from the stimulus. Although many facts relative
to the behaviour of insects appear to be in accord with this
general theory, a larger number seem to imply the existence of
differential sensitivity to the stimulus, rather than a compulsory
and irresistible response. In such cases orientation is not wholly
of the mechanical nature implied by the theory of tropisms.
Thus Urban (1932) found that the hive bee can be so conditioned
that it will choose one of two equal light sources and move directly
toward it. If the eye of one side be blinded, it may still be
conditioned to pursue a straight course toward the light. If one
or more legs on either side be removed the bee will still go to the
light in a straightforward course.

There are also insects which move at a constant angle to a
source of light, this type of response being termed menotaxis.
The light-compass orientation shown by ants and bees has been
studied more recently by von Buddenbrock (1931, 1935) and his
pupils. With certain Coleoptera it was found that when walking
on a plain surface these insects maintain a straight course at a
definite angle towards a light some metres distant. If the position
of the light be only slightly altered, corresponding to the opening
angle of a single ommatidium, the insect reacts by means of a slight
change in its direction of walking. Beetles, kept in the dark
between experiments, maintain the same angle towards the light
for hours and even days. They turn into the exact angle required
by the shortest possible movement.

CHEMICAL STIMULI 123

Heliotropism or response to the stimulus of light, for example,
is frequently correlated with differential sensitiveness to variations
of luminous intensity. This results in a less direct form of
orientation towards, or away from, the stimulus, the orientation
being the resultant reflex response to a combination of heliotropism
and differential sensitivity. Thus, Mclndoo (1929) found that larvae
of the Codling Moth are weakly positively heliotropic, but they
were not found to move and orient themselves with the precision
implied by the theory of tropisms. They pursued sinuous paths
towards the light, and larvae in the second to fourth instars proved
to be positive to weak light and indifferent to strong light. When
about to spin their cocoons they appeared to lose all differential
sensitiveness and became strongly photonegative. More defined
examples of differential sensitivity are afforded by the behaviour
of many butterflies which are strongly positive to strong sunlight.
They become inactive immediately the sun becomes obscured,
and, when resting, some species direct the long axis of the body so
that the head is turned away from the source of light. Certain
species, when alarmed, no longer maintain their flight in the full
sunlight, but dart away and seek concealment in shade. In such
examples differential sensitiveness to light is a predominant feature.

It is a matter of common observation that most insects are
active either only by day or only by night. We are led to the con-
clusion that this kind of behaviour is activated, not by any
tropism, but by a periodic rhythm dependent upon some internal
factor operating through the nervous system. This rhythm is
correlated with the regular diurnal alternation of daylight and
darkness, and cannot be explained as being solely due to the
stimulus of variations of light intensity. Many nocturnal insects
will remain quiet and motionless in a dark room throughout the
day and, without any change in illumination of the chamber, will
become active at a regular time each night.

3. Chemical Stimuli and their Receptor Organs

The reception of chemical stimuli is believed to take place in
insects by means of a variety of types of sensillae, whose functions
are largely deduced from their structure and their location.

124 THE SENSE ORGANS AND REFLEX BEHAVIOUR

Chemical receptors are those of smell and taste ; while smell is
the perception of such stimuli acting from a distance, with taste
the distance factor hardly enters into account. It is probable
that among insects the two kinds of sensation are often not
sharply demarcated and that such a distinction may not be
necessarily valid. The antennae and palpi, for example, are used
to " explore " volatile substances, both at a distance and by
actual contact — a fact that suggests that their sensillae may
function both as distance (smell) and contact (taste) receptors.
If this be true the sensations appreciated may be merely those
of degree rather than different impressions. As Wheeler has
observed, the words " taste " and " smell " are charged with
anthropomorphism, and as the stimuli in both cases are chemical
in nature, it is better to refer to the receptors involved simply as
chemoreceptors. While admitting the force of this argument, it
appears desirable to recognise, at any rate tentatively, a distinction
between olfactory and gustatory organs. It must be remembered
that we have no evidence that sensilhc located within, or in close
association with, the oral cavity can function otherwise than as
gustatory or chemical contact receptors.

The Olfactory Sense. It has long been known that insects are
highly scnsiti\'e to volatile substances, but the results of many
different experimenters are not in complete accord with regard
to the location of the olfactory sensillae. Various types of sensillae
occur distributed over the insect body and its appendages which
may conceivably function as olfactory receptors, but since two or
more types may occur grouped together, it often becomes a
difiicult matter for the experimenter to determine which particular
type may be concerned with the sense in question. It is easy to
prove that a given insect is definitely attracted or repelled by
certain volatile compounds, and there is now an extensive
literature on the purely empirical aspect of this subject.
Frequently it will be found that the olfactory sense is developed
more or less in inverse proportion to that of vision. Odonata and
Ephemeroptera, with their highly developed compound eyes and
dorsal ocelli, have their antennae reduced to small setiform
appendages. Chironomidae and Paussidae, with their highly

THE OLFACTORY SENSE 125

developed antemiiie, have only moderate-sized compound eyes,
and ocelli are usually wanting. Many dipterous larvae are devoid
of eyes of any kind, but their olfactory sense is highly developed,
as for example in Callip/wra. Among many of the higher
Hymenoptera and Diptera, however, both senses are acutely
developed in the adults.

For many years it has been believed that the true seat of the
olfactory organs is the antennae. It will serve no useful purpose
to discuss the extensive literature on this subject, and it will be
sufficient to stress certain features lending sui3port to the belief
mentioned. Elongated, movable organs, projecting outwards
from the body like antennae, are especially favourable as situations
for the location of olfactory sensilla^. As Forel has pointed out,
the position of such sensillae, externally on the antennae, gives to
the insect better olfactory powers at a distance than if the
odoriferous particles had to be drawn into an olfactory chamber,
as occurs in terrestrial vertebrates. It is a matter of common
observation that the antennae of many insects betray, by their
agitation and differential movements, an apparent acute sensitivity
to odours in their vicinity. Furthermore, elimination of their
functions by coating the antennae with impervious material or
by the amputation of those appendages has shown that certain
insects so treated no longer respond to odours, or do so much
more tardily. Mclndoo rejects the view that the antennae in
general are the seat of the olfactory sensillae. In a long series
of papers he has investigated the structure of supposed olfactory
sensillae and described the results of a number of experiments.
He criticises experiments involving amputation of the antennae
in that it induces abnormal reactions, yet, nevertheless, this
operation was carried out in some of his own experiments ! There
is no doubt that amputation should be avoided wherever possible,
and more natural reactions are likely to result when these
appendages are put out of action by coating over only. Mclndoo
maintains that the olfactory sensillae occur scattered over various
regions of the body, particularly on the bases of the legs and
wings. When these sensillae are varnished over a much slower
response results than when the antennae are amputated. Without

126 THE SENSE ORGANS AND REFLEX BEHAVIOUR

going into the details of his experiments, it may be said that
Mclndoo has given considerable evidence that the olfactory
receptors do occur on regions of the body other than the antennae.
He has not, however, advanced sufficiently convincing proof to
discount the accumulated evidence that the latter organs play a
definite part, and in many insects a predominating part, in
olfactory sensation.

Olfactory Sensillae. It will be convenient at this point to
consider the general structure of presumed olfactory sensillae.
Berlese (1909) maintained that a chemoreceptor contains both
glandular and sensory cells. The glandular secretion serves to
keep the cuticular part of the sensillae moist, and therefore in a
condition for the reception of stimuli. It is presumed that the
secretion fills the cavity of the organ and reaches the exterior
either by filtration through the covering membrane, or by means
of minute pores in the latter. The belief that a liquid, or mucous
layer, is a necessary component of such a type of sensilla implies
that the particles of a volatile compound must either become
soluble in that liquid, or enter into some form of combination with
it, before any stimulus will be appreciated by the nerve-ending.
Obviously, such a conception is based upon analogy with the
vertebrate olfactory organ, but Berlese considers that what
evidence there is points to the same process occurring with insects.
He quotes observations of Erichson made in 1847 and of Saulcy
in 1891, that the antennae of insects are covered with a delicate
film of liquid. Recent investigators have advanced little support
to Berlese's contention, since they have found no clear evidence of
gland cells in connection with any type of presumed olfactory
sensilla. On the other hand, Snodgrass (1926), in his very clear
account of insect sense-organs, suggests that the vacuole which
surrounds the process of the sense-cell in many of these organs
may be a possible source of a solvent liquid. It is noteworthy,
as he also points out, that the cuticular walls of insect sensillae
are often extremely delicate, in many cases not more than half a
micron in thickness. They are frequently so thin that in section
they do not show a double border when viewed under the highest
magnification. There is good reason to believe that such

OLFACTORY SENSILL^ 127

membranes may be permeable to liquids, and the existence of
the merest film bathing the exterior of a sense-organ would be
extremely diffieult to deny or affirm. The alternative hypothesis
is that the membrane is of a texture permeable to many types of
odoriferous particles, which pass into solution immediately upon
coming into contact with the vacuolar fluid surrounding the
nerve-ending of the sense-cell within.

In the study of chemoreceptors we are met with a baffling
variety of sensillae often widely distributed over the insect body.
As Wheeler remarks (1928), the confusion is increased by the
difference of opinion prevailing in regard to their structure in
relation to their possible functions. It grades from sensillae which
may be either tactile or olfactory, through a variety of presumably
olfactory and gustatory to the Hicksian or campaniform sensillae,
which have been variously interpreted as organs of pressure,
temperature, humidity, or vibration.

The usual course has been to determine by suitable experiments
which appendages or regions of the body afford evidence of
response to olfactory stimuli. The parts in question are then
subjected to histological examination, and those sensillae present
which appear structurally adapted for the reception of the
stimulus in question are regarded as olfactory receptors. Thus
placoid sensillae are abundant upon the antennae of certain
Homoptera, many Coleoptera and most Hymenoptera. In many
of the Parasitic Hymenoptera they are elongated structures
sometimes termed rhinaria ; in Aphididae they are commonly
circular in form, while in the hive bee they are more ovoid. As a
general rule placoid sensillae are more abundant in the male than
in the female ; in aphides they are most numerous in the alate
forms. In the hive bee, for example, Vogel computed that in the
drone each antenna bears about 30,^00 of these sensillae ; in the
worker about 6,000 ; and in the queen between 2,000 and 3,000.
The general conclusion is that the greater number of these organs in
the male is correlated with the more highly developed olfactory
powers in that sex, which come into play in seeking the female.
Furthermore, sexual odour emanated by the female, acting upon
the great number of receptors in the male, stimulates the male

128 THE SENSE ORGANS AND REFLEX BEHAVIOUR

for their presumed function

l^-.k

-sc

C —

to a degree of excitability which leads to the consummation of
the sexual act. Whether the placoid sensillae are actually the
receptors involved has not been absolutely proved. Structurally,
rhinaria appear to be well adapted for this function, since they
are invested by an extremely thin cuticle and are innervated by a
multinucleate sense-cell {vide Chrystal, 1930), but in the hive bee,
for example, the placoid sensill?e appear to be but poorly adapted

As Snodgrass points out, the
plates themselves are about 1-5 /x
in thickness, but the peripheral
membrane of the surrounding
groove is not more than 0-5 /x in
thickness. It would seem, there-
fore, that it would be necessary
to assume that a liquid filters
through the thin membrane of the
groove and spreads itself over the
surface of the plate. There is as
yet, hov»'ever, no proof that this
actually takes place. ^Iclndoo
regards the campaniform sensillse
as being the true olfactory
receptors. He termed these
structures " olfactory pores " for
the reason that he claims that the
covering membrane of these
minute dome-like structures is perforated by a central pore which
thus allows the process of the sense-cell concerned to come in
direct contact with odoriferous particles. Notwithstanding the
large amount of histological study that has been devoted to these
structures, no other observer has detected the presence of the
central pore which he claimed to be evident. The most recent
workers, including Sihler (1924) and Newton (1931), state that the
covering membrane of the sensilla is continuous and devoid of any
kind of pore. It seems probable that what Mclndoo described
is merely an optical effect caused by a thin central area of cuticle
wherein the apex of the scolopala is lodged.

niA

Fig. 56. Campaniform sensilla from
the cercus of Blatta orientalis. c,
Chitogenous cell ; d, dome-like
cuticiilar covering ; h, hypo-
dermis ; n, neurilemma ; iw,
nerve fibre ; s, scolopala ; sc,
sense-cell ; v, vacuole ; (integu-
ment deep black). (From Sihler.)

CAMPANIFORM SENSILLM

129

The delicacy of the coveriii^^ iiieiiibrane of the campaniform

sensilla suggests that, in many cases, these organs may function

as olfactory receptors. They occur distributed over the body

and on the wings, halteres, legs, antenna?, palpi, sting, cerci and

other parts in different orders of insects. This very wide range

of location has rather argued in the past against any idea that

they are of olfactory significance. They have been assumed

by some authorities to be receptors of vibrations or of changes

in atmospheric pressure, but the only experimental evidence

relative to their function is afforded by Mclndoo. While it

appears from the results of his extensive

observations that these sensilla? function as

olfactory receptors, his data are insufficient

to invalidate the conclusions of other skilled

experimenters that olfactory sensilhie, of

different types, are also present and are

chiefly located on the antenniE. Certain

experiments by Minnich (1924) may be

quoted here, since they appear to afford an

example of how the two points of view

may be reconciled. He showed that by

coating one antenna of Pieris rapcu either I^'ig. 57. Diagram of

.,, ,. .., . . p «. an olfactory (cam-

with vasehne or with a mixture ot pararhn paniform) ' sensilla,

and vaseline, or by directly amputating the according to Mclndoo.

1 lo J. ' .• . 1 • • c, Cuticle ;/>, olfactory

appendage, oliactory reaction to apple juice p^^.^ . ^^ sense-cell. '

was only affected by 6 per cent. When

both antenna? were treated in a similar manner the response was

reduced by 58 per cent. Minnich states that the treatment

described resulted in no general abnormal behaviour ; even with

amputation of the antennae little disturbance was noted, and

clear-cut results were obtained. Although the histology of the

sensilla? involved was not investigated, Minnich concludes that

the abundance of these structures on the antenna?, taken in

conjunction with the reactions mentioned, indicates that these

appendages represent probably the most important olfactory

region of the body. On the other hand, he points out that the

responses of the insect show that a considerable proportion of

K.A. ENTOMOLOGY. 5

130 THE SENSE ORGANS AND REFLEX BEHAVIOUR

the receptors must be located on some other part of the body,
since after all antennal function was eliminated the insects were
still responsive to an extent of 42 per cent.

The series of experiments conducted by von Frisch (1921) with
regard to the olfactory responses of the hive bee showed that
this insect is just as insensible to odours when the eight terminal
segments of the antennae are excised as when the entire appendages
are amputated. He emphasises the fact that bees trained to
respond to a specific odour no longer respond to that odour after
amputation of their antennae. The absence of the olfactory
reaction, he maintains, is due to loss of the organs of smell, and not
to shock caused by the operation. In support of this conclusion he
points out that bees conditioned to respond to a certain colour still
seek that colour, and no other, after deprivation of their antennae.
It is further noteworthy that the most conspicuous antennal
sensillae are of the placoid type, and these are located on the last
eight flagellar segments, ^'on Frisch concluded on these grounds
that the olfactory sense resides in the antennae and that the
placoid sensillae are the chief receptors involved. Mclndoo
(1914), on the other hand, came to the conclusion that the antennae
are organs bearing little or no relation to olfactory sensation and
that the true organs concerned are the so-called olfactory pores
which are distributed over various regions of the body.

As Snodgrass has pointed out, Mclndoo's experiments were
conducted under confined conditions, whereas those of von Frisch
related to bees free in the apiary. Furthermore, the latter
observer experimented with floral extracts, whereas Mclndoo
used more powerfully smelling volatile oils. There is, therefore,
a very evident difference in the experiments in the two cases.
Mclndoo's observations only showed that antennaless bees can
perceive strong odours when in close proximity to them, while von
Frisch demonstrated that bees so treated are unable to sense the
presumably much milder floral odours when in a state of natural
freedom around the apiary. It appears likely therefore that, in
all experiments dealing with olfactory responses, the use of
powerful or irritant compounds, especially those which do not
enter into the normal experience of a given insect, needs special

OLFACTORY SENSILL^ 131

precautions. Unless applied in very dilute concentrations, or
used at a sufficient distance to ensure adequate dilution by the
intervening air, erroneous results may ensue. It is probable that
substances of this type may produce merely irritant sensations
capable of affecting almost any type of receptor organ, after the
analogy of pungent vapours which directly affect the nose, eyes
and throat in human beings. It will be obvious that a very
erroneous conception may arise with regard to the location of the
true olfactory organs of insects, unless adequate precautions are
taken to exclude the possibility of a different kind of reaction
being involved at the same time.

In his most recent experiments, using certain kinds of " blow-
flies," Mclndoo (1933, 1934) reasserts his previous conclusions
regarding the location of the olfactory sense being elsewhere
than on the antennae. Among recent workers, Glaser (1927)
and Valentine (1931) arrived at very different opinions with
respect to other insects. Glaser concluded that olfactory
chemoreceptors in Periplaneta are present not only on the antennae,
but also on the labial and maxillary palpi. The antennae, he
claims, are capable of longer distance olfactory perception than
either of the two pairs of palpi. Valentine carried out a number
of experiments with reference to sex responses in Tenehrio molitor
(the meal-worm beetle). He found that individuals of both sexes
responded to a glass rod smeared with gland secretion pertaining
to the opposite sex, provided the antennas were intact. Males
without antennae were unable to discover the females and they
betrayed no responses when brought into close proximity to
individuals of the opposite sex. Similar results, it may be added,
were obtained when using food stimuli (aqueous bran extract).
He was further able to show that olfactory perception was located
in the apical four segments of the antennae. When employing
essential oils of clove, bergamot and thyme, he found that they
exercised an irritating effect, and he concluded that they reacted
upon other senses besides that of olfaction and also affected the
respiratory system. A summary of the subject of olfactory
receptors in insects is given by Marshall (1935), to which the
reader is referred for a fuller reference to the literature. At the

5—3

132 THE SENSE ORGANS AND REFLEX BEHAVIOUR

present time certain of the results obtained by Mclndoo cannot
be reconciled with those of other workers, and it is evident that the
subject requires further exploration under carefully standardised
conditions.

The Gustatory Sense. The probability that, in many cases,
insects do not experience " taste " and " smell " as separate and
distinct sensations has been previously referred to. Nevertheless,
the positions of certain sensillae in relation with the oral cavity
strongly suggest that such organs are adapted to perceive chemical
stinmli by actual contact only. Proof of such a contention is
extremely difficult to attain, since the location and minute size
of the organs concerned do not render their functions amenable
to determination by exact experiment. Sensillae situated on the
lining membrane of the epipharynx of mandibulate insects, the
two groups of sensory hairs mentioned by Deegener as situated on
the floor of the pharynx in Lepidoptera, and the elaborate
pharyngeal organ of Hemiptera appear to be almost certainly
gustatory in function. The organ in Hemiptera (Fig. 60, A) lies in
the clypeal region of the head and is located on the roof of the
pharyngeal duct. As described by Bugnion and Popoff (1911), the
cuticular lining of the duct at this point is differentiated into a
cribriform plate whose pores appear to be occupied by processes of
sense-cells. The organ has a definite nerve-supply, but its whole
structure, and its range of development in different families of
Hemiptera, need further elucidation. In some species the pores arc
few in number, but they range up to about 600 in others. This
organ, from its structure and location, appears to be adapted for
testing the qualities of the plant-sap passing up the pharyngeal
duct and before entry into the cavity of the pharynx ; there
seems little doubt, therefore, that it can only function as a contact
(taste) organ. The blood-sucking bug, Rhodnius, will seldom
imbibe salt solution through a membrane unless a small amount
of haemoglobin be added (Wigglesworth). It is probable in
this instance that the gustatory organ, just referred to, is
involved in testing any fluid taken up.

The various sensillae that occur distributed on the mouth-parts
of different insects have also been regarded as gustatory organs.

GUSTATORY SENSILL^

133

They include tlie styloconic sensillte found abundantly on the
proboscis in many Lepidoptera, and several types of hair-like
organs found on the mandibles, maxillae, palpi, etc., of different
insects. They are abundant, for example, on the mouth-parts of
the hive bee, but Mclndoo (1916) concludes that the thickness of
their cuticular covering precludes their performing any gustatory
function, and that they are more probably tactile organs. This
observer maintains that the only sensillae present in the bee which
function as chemoreceptors are the so-called " olfactory pores."
These, he states, are olfactory in function, as their name implies,

1^

■I ^ T-- I

\2^^

P ^

B

Fig. 58. A, cross-section of apparatus used by Miniiich. B, The
arrows 1, 2, 3 represent the positions in which the insects were
tested, the walking legs being indicated by crossbars. Explana-
tion of lettering in the text. (Adapted from Minnich.)

and he finds no evidence of true gustatory organs being present.
There is the perennial difficulty, however, of advancing sufficient
experimental proof that sensillae located on the mouth-parts
perceive taste and smell as separate impressions. Opinion is
inclining to the view that they are probably chemoreceptors, in
the broad sense, which are capable of appreciating stimuli both by
actual contact (taste) and at a distance (smell).

The experiments of Minnich (1921) indicate that certain butter-
Hies (Pyrameis atalanta, Vanessa antiopa) are able to respond to
non-volatile substances by means of the tarsi of the middle and
hind pairs of legs. The apparatus (Fig. 58) used consisted of a
square wooden platform iv, bearing two cross-pieces iX, y, which
supported a tightly stretched wire screen s. The dimensions of

134 THE SENSE ORGANS AND REFLEX BEHAVIOUR

this construction were such that a Petri dish j), 15 cm. in diameter
and 1 cm. deep, could just be passed beneath the screen without
coming in contact with it. Within the Petri dish were mounted
two tin receptacles (a and h), 4 cm. long, 2 cm. wide, and of such
a height that their upper edges were on a level with the rim of the
Petri dish. Two rectangular openings were cut in the screen, so
that when the Petri dish was in position they were directly over
the tin receptacles. Owing to the fact that the insects in question
are well known to feed upon the juices of fallen or decayed fruit,
apple juice was used in the experiments and poured into the Petri
dish. Two packs of cheese cloth were folded to fit the pans :
one was saturated with apple juice (in a), and the other with
distilled w^ater (in h). An individual butterfly was held by a
specially devised arrangement which gripped the closed wings,
and was tested in the apparatus in three positions. In the first
position (Fig. 58, B) it was placed as near as possible to the edge of
the receptacle a with the antennic reaching out over the cloth
saturated with apple juice, but with the four walking legs upon
the screen. In the second position the insect was placed in relation
to receptacle h, so that the middle tarsi w ere in contact with the
distilled water : in the third position the same experiment was
performed in relation to receptacle a.

The conditions of chemical stimulation in the three positions
may be summarised as follows : —

Position 1. Distance stimuli from apple juice.

Position 2. Distance stimuli from apple juice in the surrounding
Petri dish plus contact stimulus of distilled water on the
tarsi.

Position 3. Distance and contact stimuli from apple juice.

The experiments were carried out with starved individuals, and
the response was judged by the coiling and uncoiling of the
proboscis. The maximum duration of each trial was one minute,
and if no visible uncoiling of the proboscis occurred during that
interval " no response " was recorded. With Pyrameis 29 per
cent, of the individuals responded in position 1, only 17 per cent,
in position 2, and 100 per cent, in position 3. After removal of
the antenna?, palpi and vestigial fore legs the experiments were

TARSAL PERCEPTION 135

repeated and the responses in the three positions were 4 per cent.,
17 per cent., and 65 per cent, respectively. The responses with
Vanessa were in some respects less striking, but they were similar
in nature. In a further series of experiments it was shown that
Pyrameis, by means of its tarsi alone, is able to distinguish IM
saccharose solution from distilled water and such solutions as
IM HCl, M/600 quinine sulphate, and IM NaCl from either IM
saccharose or distilled water. Distilled water alone is a very
effective stimulus when applied to the tarsi. Minnich gives
reasons for concluding that chemoreceptors are located in all four
tarsi of the walking legs of Pyrameis, but gives no information
with respect to the structure and exact location of the sensillai
involved. Since perception takes place by means of actual
contact with the stimulus, he regards the organs involved as being
gustatory in function.

In 1922 Minnich repeated his former tests and obtained very
similar results, and, according to the scheme adopted in this
second paper, the total responses of all the individuals tested
were 100 per cent, to IM saccharose, 84-7 per cent, to M/10 quinine
hydrochloride, and 51-6 per cent, to 2M NaCl. An intensive
study of the differences of response showed that Pyrameis could
differentiate between almost all, or probably all, these solutions.

In a further series of experiments (Minnich, 1922a) it was
shown that contact of the four ambulatory tarsi with IM
saccharose solution will always effect an extension of the
proboscis, irrespective of the nutritional condition of the insect.
After a prolonged period of inanition with respect to saccharose
the tarsal sensitivity may be so delicate as to appreciate as low
as M/12,800, and may thus be as much as 256 times that of the
human tongue. This highly developed sensitivity is doubtlessly
correlated with the fact that sugars form the chief food constituent
in Pyrameis.

Additional observations and experiments were carried out by
Minnich (1926) with regard to the responses of three species of
common Muscid flies (Phormia regina, P. terrce-novce and Lucilia
serricata). The results obtained were very similar to those
derived from the experiments with Nymphalid butterflies. When

136 THE SENSE ORGANS AND REFLEX BEHAVIOUR

the tarsi are in contact with distilled water 95 to 100 per cent,
response ensued, while without tarsal contact only 1-4 to 13 per
cent, response was observed. With paraffin oil the response was
very small, while with IM saccharose solution the response varied
from 80 to 100 per cent., depending upon the nutritional state of
the insects. By means of their tarsi they were able to discriminate
between water and saccharose solution and between water
and paraffin oil. The oral lobes of the proboscis appear also
to be equipped with taste sensillae which are more sensitive to
saccharose than are the tarsi. In a subsequent paper (1926a)
Minnich states that gustatory hairs are present on the proboscis
of Calliphora. A fly, after being abundantly supplied with water,
but otherwise starved, does not exhibit any response when
distilled water is brought into contact with these hairs. When
they are touched with a minute brush moistened with sugar
solution the proboscis becomes quickly extended. The sensitivity
of these hairs is stated to be so great that a single one, when
stimulated in this way, produces the proboscis response. It is,
therefore, concluded that a true gustatory response is involved.
Minnich subsequently extended these observations while working
in the laboratory of Professor K. von Frisch in Munich. Thus,
he determined the range of chemical sensitivity of the legs of the
blow-fly {Calliphora vomitoria) to Narious sugar solutions (Minnich.
1929). In 1931 he published the results of studies on the reactions
of the mouth-parts of that same insect. He found that the
aboral surfaces of the oral lobes of the proboscis are clothed with
marginal hairs which appear to be gustatory in function. The
minimum concentration of saccharose which, when applied by a
fine brush to the hairs in question, will produce proboscis extension,
is about 1/100 M and much higher than that for the legs. The
threshold for glucose is at least four to eight times higher than
that for saccharose, while lactose was the least effective of the
sugars that were tested. The fact that saccharose is sixteen to
sixty-four times more effective than lactose suggests that the
hair-receptors concerned are chemical sense organs rather than
tactile or osmotic in function, since equimolar solutions of these
two sugars are similar as regards their viscosity and osmotic

TARSAL PERCEPTION

137

pressure. From Minnich's data it seems that the general order

of effectiveness for the four sugars tested is — saccharose and

maltose >glucose>lactose. In other words, it is essentially the

same as that observed both by von

Frisch (1927, 1928) and by Kunze

(1927) for the mouth-parts of the hive

bee.

In 1932 Minnich published the
results of experiments with the hive
bee, the responses being judged by
the extension of the proboscis. The
bee is able, by means of the antennae
or the fore leg, clearly to distinguish
between 64/100 M saccharose solution
from an equimolar solution of lactose
and from water. The antennae are
well known to function as distance
receptors, and it appears that they
are used as contact receptors also.
Although it seems clear from these
experiments that the apex of the
antennae bears contact chemoreceptors,
it is by no means certain which type,
of the five kinds of sensoria present,
actually function in the j^rocess.

Other investigators, including Abbott
(1928), Anderson (1932), Crow (1932) Fig. 59. Chemoreceptor

and Weis (1930), have confirmed and

extended Minnich's chief results and

conclusions, with the outcome that it

appears quite certain that the legs of

certain butterflies and flies, at any

rate, are highly sensitive to chemical

stimuli. In 1934, Mclndoo contested certain conclusions of

Minnich with respect to blow-flies and claims that it is not

necessary, in order to explain the proboscis response, to assume

that the tarsi bear gustatory organs. He describes experiments

Pyrameis atalanta, highly
magnified. t, trichogen ;
c, probable cap-cell ; s,
sensory cells ; t", tube of
chemoreceptor ; «, dark-
staining end of tube ;
p, papilla. (From

Eltringham.)

138 THE SENSE ORGANS AND REFLEX BEHAVIOUR

wherein the flies are easily able to distinguish between chemically
pure saccharose water and distilled water when the tarsi are
about 3 mm. distant from these liquids. The act of touching
the tarsi by means of a brush, moistened with the substance to
be tested, produces the initial stimulus and brings the liquid
almost in contact with certain campaniform sensillae in the tarsi.
He further states that a high response resulted by merely
touching the tarsi with a needle, the proboscis being protruded
when the sugar solution was 3 mm. away. It is possible,
therefore, that the tarsi of the blow-fly, like the antennae of the
bee, bear both contact and distance receptors. For other insects,
tarsal sensoria are described by Eltringham (1933) as being
extremely slender, thin-walled tubes in the butterfly Pyrameis
atalanta (Fig. 59).

The work of Minnich opened up a new field of experimentation
in so far as insects are concerned. It is noteworthy that several
previous investigators have shown that the walking legs in the
higher Crustacea are sensitive to contact with foods and other
chemical stimuli. B. M. Patten has also demonstrated that the
modified anterior legs of the whip-tail scorpion are likewise
sensitive to water and other chemical stimulation. In the
light of the foregoing evidence, Minnich considers that contact
chemoreceptors, located in the distal extremities of the thoracic
legs, are of general occurrence among Arthropods.

CHAPTER VI

THE SENSE ORGANS AND REFLEX BEHAVIOUR^

continued

4. The Tactile Sense, p. 139. 5. Chordotonal Organs and
Reactions to Vibrations, p. 140. Chordotonal Sensillce, p. 141 ;
Simple Ligamentous Chordotonal Organs, p. 142 ; Johnston's Organs,
p. 143 ; Tympanal Organs, p. 143 ; The Functions of the Chordotonal
Organs, p. 145. 6. General Stimulatory Organs, p. 153. 7. Reflex
Behaviour and Practical Entomology, p. 155. Reactions to
Chemical Stimuli, p. 155 ; Reactions to Light, p. 164 ; Concluding
Remarks, p. 167. Literature, p. 168.

4. The Tactile Sense

The simplest kinds of sensillse present in insects are innervated
hairs, and several types of these structures have been regarded as
being tactile in function. Among the more recent investigators,
Mclndoo (1926) has described sensory hairs, bristles and peg-like
organs, which are believed to be tactile in function, on various
regions of the body in the Cotton Boll Weevil. They are present
in the head, antennae, mouth-parts, thorax, legs, wings, abdomen
and genitalia. Histologically, they are regarded as being too
thick-walled to function as chemoreceptors ; they are innervated
by delicate nerve terminations, and movably articulated with the
general integument. Such sensillse appear to be only adapted for
tactile sensation, and even very slight pressure exercised by any
external object would seem to be appreciated by organs of this
nature, whereas the general cuticle of the body would fail to
register such stimuli.

It is well known that many insects betray what is regarded as
a positive contact reflex (often termed thigomotropism). Such
species ensconce themselves in crevices, beneath bark, under
stones or logs, etc. The same kind of reaction is betrayed when
oviposition occurs in fissures, crevices, etc., and beneath the

139

140 THE SENSE ORGANS AND REFLEX BEHAVIOUR

soil, the highly sensitive apex of the abdomen being provided
with sensillae that are presumably tactile in function. Many
lepidopterous larvse, when about to pupate, seek fissures or dark
crevices or pass beneath the soil at that period. It would seem
that at this vulnerable phase in their life such responses are
brought about by a change in their reflex behaviour. Prior to
pupation many species live openly on foliage and exhibit different
behaviour. It has been argued that these responses are due to
negative phototropism and not to thigomotropism, but in certain
cases, at least, the light factor appears to be of minor significance.
Thus, Mclndoo (1929) states that when larvae of Cydia pomo7iella
are about to spin their cocoons they seek concealment in some
dark, tight situation, but if a dark crevice be not available, they do
not hesitate to adopt a well-lighted place instead. Loeb also came
to the conclusion that in the case of certain moths which rest in
fissures and crannies in the bark of trees, contact stinuilus alone is
involved. There is, however, very little experimental evidence
available with reference to contact responses, and it is possible
that, in many cases, a complex kind of differential sensitivity is
brought into play, involving negative response to light and
positive responses to gravity and to contact with external objects.

5. Chordotonal Organs and Reactions to Vibrations

Insects belonging to very diverse groups are known to respond
to aerial vibrations of different degrees. In some cases these
vibrations are imperceptible to the human senses, and, in others,
vibrations which produce a very 2:)ronounced auditory sensation
in man appear to exercise little or no stimulatory effect upon
certain insects. In a large number of instances there is no
evidence that the perception by insects of aerial vibrations is of
the nature of hearing, in the anthropomorphic meaning of the
word. On the other hand, the power of sound production and the
presence of highly developed tympanal organs in certain families,
together with the behaviour of the insects possessing these organs,
indicate that in these instances a sense closely akin to hearing is
present.

Before discussing this subject further, it is necessary to consider

CHORDOTONAL SENSILLJE

141

briefly the nature of the sensory organs coneerned. These may
be grouped under the general term of chordotonal organs, which

Fig. 60. A. Section through presumed gustatory organ of an
Hemipteron {Graphosoma lineatum). (Adapted from Bugnion
and Popoff.) B. Diagram of a simple chordotonal organ con-
sisting of a single sensilla. C. Diagrammatic longitudinal
section through the second antennal segment of a Geometrid
moth (left side), showing Johnston's organ, (Adapted from
Eggers.) a, articular membrane ; ab, apical body ; al, axial
fibre ; c, cap-cell ; cu, cuticle ; e, envelope-cell ; h, hypo
dermis ; /, ligament ; n, antennary nerve and branches
n^, n^, to Johnston's organ ; p, cu^icular pore ; s, sense-cell
sc, scolopala ; i', vacuole.

imphes the presence of chords or filaments sensitive to tones or
vibrations.

Chordotonal Sensillae. The sensillae composing chordotonal
organs may occur singly, but more often in small groups, or larger
numbers may be combined to form more elaborate structures.

142 THE SENSE ORGANS AND REFLEX BEHAVIOUR

The simplest type of chordotonal organ consists of a single sensilla
which exhibits the following structure (Fig. 60, B). An ovoid or
fusiform sensory cell is provided with a distal prolongation
terminating in a sense-rod or scolopale. The scolopale is
strengthened by a variable number of internal ribs and is usually
capped by a deeply staining apical body. A delicate axial fibre
extends from the sense-cell and traverses the axis of the scolopale
to terminate in the apical body. The prolongation of the sensory
cell, together with the scolopale, are enclosed within an envelope-
cell. The latter, in its turn, is closely associated with an elongated
cap-cell, which serves to connect the sensilla with the integument.
The scolopale itself is often connected with the integument by
means of a terminal fibre which traverses the envelope-cell. At
the point where the distal extremity of the sensilla is fixed to the
cuticula, there may be a pit or a thickened disc, but there is no
special receptive structure. Proximally, the sensilla is attached
to the integument by a special ligament which is inserted into the
base of the sense-cell. It will be observed, therefore, that a
typical chordotonal sensilla is suspended between two points of
the integument. In some cases, it may be added, the scolopale
appears to be hollow, and its cavity is connected basally with a
vacuole filled with fluid. The scolopale is immersed in the semi-
fluid protoplasm of the envelope-cell, within which it is free to
vibrate.

Chordotonal organs may be grouped into three classes arranged
in order of increasing complexity of structure : (1) simple liga-
mentous organs ; (2) Johnston's organs ; and (3) tympanal organs.
These three types are only briefly dealt with in the following
sections, since their structural features are well described in
modern text-books. A very full and admirably illustrated treatise
on the subject will be found in the recent work by Eggers (1928).

Simple Ligamentous Chordotonal Organs. Organs of this type,
comprising small groups of one or more sensillse, occur in a great
many different groups of insects and their larv?e. Among larval
insects they are commonly segmentally arranged with an organ on
either side of each of the first seven or eight abdominal segments
(Fig. 60, B). Structures of this type are well known from Graber's

TYMPANAL ORGANS 143

familiar studies in larvae of the Diptera, Lepidoptera, Coleoptera
and Tenthredinidae.

In adult insects simple ehordotonal organs are likewise very
widely spread among different orders, where they occur chiefly
on the antennae, wings, halteres and legs. They are especially
constant in relation with the bases of the wings, and in this
position they are known to be present in all the major orders
of Pterygota.

Johnston's Organs. Organs bearing this name are located in
the pedicel, or second segment, of the antenna, and are prevalent
in all the main orders of insects, and also in Lepisma among the
Thysanura. They are not the only ehordotonal organs present in
the antennae, and smaller groups of sensillae may occur in other of
the segments of those appendages. A Johnston's organ consists
of a group of usually very numerous ehordotonal sensillae disposed
in the form of a sheath around the antennal nerve, and when it is
highly developed, as in the Culicidae and Chironomidae, the pedicel
of the antenna is greatly enlarged. The distal extremities of
the sensillae are usually attached to the articulatory membrane
between the second and third antennal segments, and are indicated
by a series of pits or annuli in that membrane. Proximally, the
sensillae are directly connected by means of nerve fibres with the
main sensory nerve of the antenna (Fig. 60, C).

Tympanal Organs. As their name implies, these organs consist
of a sound amplifier or tympanum developed in close relation
with one or more defined groups of ehordotonal sensillae. They
are present in grasshoppers and cicadas, which are well known to
be capable of producing loud and often shrill sounds, as well as in
some other insects whose sound-producing capacity is far less
evident.

Some of the most highly specialised of these organs are found
in the saltatorial Orthoptera, but since their structure has so often
been described and figured, this aspect of the subject will not be
dilated upon here. It will be recalled that in the Acridiidae a tense
vibratory membrane or tympanum surrounded by a cuticular ring
is clearly visible externally on either side of the first abdominal
segment. In the Tettigoniidae and Gryllidae there is often a pair

144 THE SENSE ORGANS AND REFLEX BEHAVIOUR

of tympanal organs near the proximal extremity of the tibia of
each fore leg.

In the Cicadidfe tlie j)resume(l absence of any receptor organs
for sound vibrations had for many years been an outstanding
difficulty in interpreting the function of their highly de\'eloped
powers of sound production. It was not until 1923 that Vogel,
in Germany, proved on histological grounds that true tympanal
organs are present in both sexes of these insects. In the male the
cavity beneath the operculum of the sound-producing organ is

bounded posteriorly by a struc-
ture commonly termed the
mirror. The latter organ had
long been claimed to be a
resonator, and it remained for
V\)gel to show that it is the
tympanum ^ of an auditory organ.
The true })ercipient part of this
organ consists of a group of an
immense number (about 1,500)
Fig. G1. Hori/.ontal section across of chordotonal sensillae stretching
portions of thorax and abdomen of across the cavity of what may

thr°TgM Tnd- Tff'VSl be tern.ed the auditory capsule",

organs, at, Accessory tympanum ; At one of their extremities these
d, chitinous lamella ; /, ligament -ii attached bv Haa-

supporting chordotonal organ ; n, ^ensuia" are attacnea, D\ liga

chordotonal nerve ; t, tympanum ; mentous prolongations, to a

tc, right tympanic chamber. „„^- ,-]„„ ..^ocess connected with
(Adapted from Eggers.) cuticuiai process connected witn

the mirror or tympanum. At

their other extremities the sensillse are attached to the outer

wall of the auditory capsule. In the female the tympammi is

also well developed but somewhat reduced in size.

In many Lepidoptera highly developed tympanal organs are

also present. The comprehensive studies of Eggers (1919, 1928)

have shown that these structures are present on either side of the

^ Swinton, as long ago as 1877, gave reasons for concluding that the so-called
mirror was tlie tympanum of an auditory organ. As is pointed out by Myers
in his book on Cicadas (1929), this account had been long overlooked, and
Vogel only discovered Swinton's paper when his own detailed studies were
Hearing completion.

FUNCTIONS OF CHORDOTONAL ORGANS 145

thorax or the base of the abdomen in different families, while
they are totally wanting in many others. The tympanal organs
(Fig. 61) eonsist of a pair of vesicles— one on each side of the
body. Each vesicle is lined by tracheal epithelium and is lodged
in a cavity formed by the invagination of the integument of its
segment. It communicates with the exterior by means of an
opening on the side of the body, and this aperture is guarded by
an integumentary flap or shield. Just within the opening there
is found the glistening membrane of the true tympanum, and
closely associated with it there is a group of chordotonal sensillae
connected with a special tympanic nerve derived from the meta-
thoracic ganglion. Lying more deeply within the tympanic vesicle
there is, in many species, an accessory tympanum which is not,
however, associated with any sensillae.

Tympanal organs are also described by Hagemann (1910) in
Coricca and its allies. They are located on either side of the
metathorax and in close relation with the second pair of spiracles.

The Functions of the Chordotonal Organs. No definite proof
exists with regard to the functions of the simple ligamentous
chordotonal organs, and the subject is highly problematical.
Snodgrass (1926) lays stress upon Eggers' view (1923) that they
function as " rhythmometers." The last-mentioned authority
claims that most of the movements of insects result in rhythms.
In the case of the wings, for example, the action of these organs
sets the whole body into rapid vibration. He points out that the
movements of the antennas and legs tend to become rhythmic,
and the same applies to the pulsations of the dorsal vessel and the
movements of respiration. The body-movements of locomotion
in certain aquatic insects, the motion of the gills and of other
parts likewise tend to become rhythmic. Eggers suggests, there-
fore, that the chordotonal sensillge are organs for the purpose
of co-ordinating and regulating muscular movements of this
character. Notwithstanding this hypothesis, we must conclude
that similar kinds of sensillae in association with a tympanum, or
other cuticular receptive organ or membrane, are capable of
perceiving vibrations in the surrounding medium wherein the
insects live. As will be mentioned later, evidence in favour of

146 THE SENSE ORGANS AND REFLEX BEHAVIOUR

tympanal organs being auditory in function is too well established
to admit of any other interpretation. Wherever chordotonal
sensillaj occur in relation with a cuticular membrane capable of
transmitting vibrations set up in the surrounding medium to the
underlying scolopaUe, the latter are presumably able to respond
on their own part by vibratory movement, since they are immersed
within the vacuolar fluid of the envelope-cells. It will, therefore,
appear possible that chordotonal sensillae are general muscle-
tension receptors which may function in relation to muscular
activity in accordance with Eggers' view and, in other cases, as
receptors of vibrations external to the insects.

The locations of the simpler types of chordotonal organs in
relation with many parts of the body suggest that they serve to
detect, or register, stresses or strains put on those parts by muscular
activity. This especially applies to the joints of the appendages
(Fig. 62). In the grasshopper Melanoplus, for example, there are,
according to Slifer (1936), at least seventy-six pairs of chordotonal
organs, each organ being composed of two or more scolopalia.
It is probable, also, in this connection, that they serve to co-
ordinate muscle action so as to adjust it on the two sides of the
body. Without some such perceptor organs there appears to be
no means by which an insect is enabled to orientate itself in
relation to the effects of gravitation — especially when in flight.
Hertweck (1931) also came to the conclusion that in insects
response to a gravitational field is by means of tension receptors
or kinesthetic sensitivity. The theory of geotropism developed
by Crozier and his co-workers has established that the direction
of an animal in a gravitational field is determined by the body-
weight acting upon tension receptors in the musculature of the
appendages. Barnes (1930) has show^n that the angle described
by an ant crawling up an inclined plane is a function of the sine
of the angle of inclination of the plane. He states that the insect
turns until the " pull " of the body-weight on both sides of the
symmetrical muscles is reduced below threshold value. This
author (1931) also showed that the slightest movement in the leg
of Periplaneta produced a discharge of sensory impulses which
were led off through electrodes and reproduced by an amplifier

CHORDOTONAL SENSILL^

147

wired to a loudspeaker. His observations are to be regarded as
giving further evidence supporting Crozier's theory of geotropic

Fig. 62. Third leg of first instar nymph of Pediculus. c, mono-
scolopal chordotonal organ ; cl, claw ; e, extensor muscle of
tibia ; /, flexor muscle of tibia ; fc, flexor muscle of claw ;
fe, femur ; n, nerve of leg ; t, tarsus ; ti, tibia. (From Keilin
and Nuttall.)

orientation, which involves the presence of tension receptors. So
far, no sensory nerve endings have been detected in insect muscle,

148 THE SENSE ORGANS AND REFLEX BEHAVIOUR

but it seems probable that in some way or other these afferent
impulses are registered by chordotonal organs.

The significance of Johnston's organs is equally problematical.
The constancy of their occurrence in so manj^ different groups of
insects and their essential similarity of structure, which only varies
in its complexity and the number of sensillae involved, suggests
that these organs play an important role in the lives of the insects
possessing them. Eggers holds the view that they register
antennal movements, which exercise varying stresses upon the
intersegmented membrane to which the component sensillse are
attached. It has also been suggested that wind pressure, air
currents, mechanical vibrations, etc., might also be perceived by
these organs. In this connection, it may be pointed out, the
general structure of the cuticle investing the antennal segment
bearing Johnston's organ does not exhibit any special adaptation
for the reception of such stimuli. In the case of Chironomidae
and Culicidac, however, where these organs are greatly developed,
the distal extremities of the sensilla^ are all attached to a special
conjunctival plate. Mayer (1874) claimed that in these insects
the whorls of antennal seta?, so highly developed in the males, are
caused to vibrate when stimulated by certain tones. It is believed
that in such cases the vibrations of these seta; are transmitted to
the conjunctival plate and hence to the sensilla^, and that insects
possessing this special type of organ are able to appreciate sound
waves of certain qualities. The mere fact, however, that certain
hairs of specific insects can be caused to vibrate by different
pitches of sound may be of little significance. It is well known
that a cord of a one- stringed instrument responds to that of
another instrument set in vibration in a similar key !

Pumphrey and Rawdon-Smith (1936) have shown that the long
sensory hairs on the cerci of the cricket {Gryllus domesticus) and
cockroach (Periplaneta americana) are sensitive to auditory
stimuli. The sensitivity of the cercus to low frequencies is much
greater than that of the human ear, a marked response being
obtained to a stimulus completely inaudible to man. It is
possible that the acoustic function of the cerci is merely incidental.
The difficulty of distinguishing between acoustic and vibratory

SOUND PERCEPTION IN LARV^ 149

stimuli for such an organ is stressed, but the relative insensitivity
of the cercus of the cricket to all but the lowest frequencies,
together with the fact that the organ is frequently in contact with
the ground, leads these authors to suppose that its main function
here is to mediate the detection of earth-borne vibrations. In
the cockroach, where the cerci are carried upwardly inclined, it
seems likely that their function as wind gauges may be equal in
importance to their function as acoustic organs. They found
that there is a close parallel between the response in the cereal
nerve and that in the mammalian eighth nerve — a fact of con-
siderable physiological interest.

It needs to be borne in mind that mechanical stimuli, appreciable
to man as touch and sound, probably intergrade as phases of
a single type of sensation in the majority of insects. Aerial
vibrations may be appreciated by those creatures through the
medium of generalised tactile receptors. In certain caterpillars
there is definite evidence that sound waves produce marked
responses when impinging upon their body-hairs. It has long-
been known that various lepidopterous larvae, and also those of
some Coleoptera, show evident indications of stimulation to
particular types of sounds. Minnich (1925), however, appears to
have been the first to locate the sense involved. He found that
larvae of Vanessa antiopa respond to appropriate sound stimuli
by throwing the anterior one-third of the body dorsally or dorso-
laterally. He used sounds produced by the human voice, piano,
organ, violin, Galton whistle, tone modulator, tuning forks and
by other means, and observed that the larvae responded in the
manner described in all cases except to the whistle and tone
modulator. The extent of the response varied according to the
intensity of the sound, and responses were obtained from vibrations
ranging from 32 to 1,024 per second. Decapitated bodies, and
fragments of bodies, will respond to sounds, and he concludes
that certain hairs, located chiefly in the anterior two-thirds of the
body, probably constitute the receptor organs involved. This
conclusion is based upon the fact that responsiveness to sounds
increases as the number of hairs increase in successive instars.
Partial destruction of the hairs by singeing either greatly reduces

150 THE SENSE ORGANS AND REFLEX BEHAVIOUR

or abolishes the response ; heavy loading of the hairs with line
droplets of water, or with flour particles, produces a similar effect.
During ecdysis, when the old hairs are more or less disconnected
from the underlying parts and the new hairs are not yet functional,
the response is either feeble or nil. Larvai subjected to a tone
stimulus of short duration, repeated at intervals of five seconds,
showed little or no fatigue in the course of 100 trials. Individuals
subjected to continuous tone stimulation, however, generally
became fatigued and ceased to respond after a period ranging
from five to sixty seconds. Minnich further observed that
continuous air currents inhibit the response to sounds. This is
due, in general, to one form of stimulation inhibiting the response
to another kind. When the air current is prolonged, however,
another factor, probably that of fatigue, also becomes involved,
Larvse fatigued to C (256 V/S) or G (384 V/S) will not respond to
C (512 V/S). When fatigued to C, however, larvae still gave
100 per cent, response to C and about 50 per cent, response to G.
These results suggest the tentative conclusion that hairs of
different lengths are affected by the various pitches of sound
indicated. Abbott (1927) carried out very similar observations
with larvae of the moth Datana (fam. Notodontidae). Definite
responses were noted to air currents, sudden jars, and to the two
notes only, viz., C" (512 V/S) and F sharp (728 V/S) ; larvae thus
stimulated elevated the anterior and posterior regions of the
body. During the experiments the larvae were protected from
air currents : they were several feet from the instruments used,
and vibrations from the substratum were eliminated. No
responses were observed when the body-hairs were coated with
water or shellac, or when the body-surface was anaesthetised with
a 2 per cent, solution of procain. Baier (1930) has also shown that
larvae of Pieris rapce react to certain acoustic stimuli in much the
same manner as those of Datana. His experiments, however,
were by no means extensive, but he showed that when stimulated
by a whistle or by vocal means the extent of the reaction varied
directly with the nearness of the stimulus, its intensity and its
suddenness. The fact that larvae whose body-hairs had been
entangled with cotton fibres responded very feebly, and when

FUNCTIONS OF TYMPANAL ORGANS 151

relieved of the fibres again responded normally, suggested that
the hairs are the receptors involved, as was found by the two
previously mentioned investigators. The results drawn from
foregoing experiments, with lepidopterous larvae, point to the
conclusion that these animals are sensitive to air currents and
to certain aerial vibrations. They obviously afford no evidence
that any sense of hearing in the anthropomorphic meaning is
involved, but lend support to the dictum already stated that,
in many insects, the tactile and auditory senses are very
questionably differentiated. No relation between the responses
mentioned and the segmentally arranged chordotonal sensillai,
known to occur in lepidopterous larvae, has been recorded, and
it appears improbable that they play any part in the process.

There now remains for consideration the functions of tympanal
organs. It is noteworthy in this connection that in the saltatorial
Orthoptera, the Cicadidae and Corixidae, sound-producing organs
are present. This fact in itself suggests that if these sounds are of
significance in the lives of the species concerned, some type of
perceptor organs is necessary. In species devoid of any capacity
for sound-production tympanal organs are wanting : this happens,
for example, in the wingless Tettigoniidae and in the primitive
soundless genus Tettigarcta among the Cicadidae. That the
tympanal organs are auditory in function is supported by the
ingenious and controlled experiments of Baier (1930) with field
crickets and other Orthoptera. Individuals of species whose sexes
were segregated, and previously conditioned to a life in cages, were
placed in different rooms sufficiently distant to be beyond the
range of sound perception. A microphone, placed in the cage
containing the males, was connected by telephone with a receiver
in the cage containing females of the same species. The females
responded to the stridulation of the ^nales by orienting themselves
when they perceived the sound : this they did by approaching the
receiver with actively moving antennae and by performing move-
ments indicative of search for the summoning partners. When the
current was temporarily interrupted, the females may move away
from the receiver, and in any case no longer betray their former
signs of attention. On restoring the circuit the females at once

152 THE SENSE ORGANS AND REFLEX BEHAVIOUR

approached the receiver again. The cages, it may be added, were
insulated so as to guard against soHd vibrational stimulation.
None of the responses came from females whose tympana were
excised, and a local anaesthetic (ethyl chloride) applied to these
organs had only the same effect just so long as its influence lasted,
normal response being ultimately restored. The use of gramo-
phone records of the male stridulatory notes had an effect similar
to that exercised by the living insects themselves.

In their investigation of auditory responses in Locusta,
Pumphrey and Rawdon-Smith (1936) deny any ability on the
part of this insect to discriminate frequency of vibrations by
means of its tympanic organ. They find no differentiation of
nervous response for different frequencies of stimulus and conclude
that the physical structure of the tympanal organ is such as would
not permit of discrimination of frequency by means of a
differentially resonant mechanism. They remark, however, that
the organ is well adapted to enable the insect to localise the
source of a sound whose frequency falls within its auditory
spectrum, e.g., the stridulation of another locust. The organ
resembles the human ear only in its relatively greater sensitivity
to the higher frequencies of sound waves. On this subject
references should also be made to the paper by Wever and
Bray (1933), which deals with nerve impulses originating in the
tympanal organs of crickets and katydids. These organs respond
to sound frequencies of from 250 to over 11,000 vibrations per
second : the quality of the response was found not to vary with
the frequency. Myers (1929) has collected a large amount of
circumstantial evidence relative to the " songs " of Cicadas and
their significance. The primary function of the notes of these
insects, according to this observer, is an assembling one, and it
may also serve to arouse the necessary stimulus leading to the
consummation of the sexual act. There does not appear to be
any conclusive experimental evidence with regard to sound
perception in these insects, but anatomical and observational
data, relating to their behaviour, appear to offer little reason to
doubt the auditory functions of their tympanal organs. The
tympanal organs of Lepidoptera likewise appear to be organs for

GENERAL STIMULATORY ORGANS 153

sound perception. This conclusion is in accordance with their
gross and minute structure and is also supported by analogy with
the better-known organs, of a somewhat similar nature, found in
the Acridiidic. Certain experiments carried out by Eggers afford
further support to this belief. This observer states that when the
tympana of both organs are destroyed, the moths so treated no
longer react to tests with sounds. When, however, a single
tympanum is destroyed, the insects responded to sound stimuli,
in more than half the cases tested, by flying. With both organs
left intact reaction to different sounds was expressed either by
flying or by elevating and moving the wings. Amputation of the
antennae did not interfere with response to sounds, while insects
whose wings had been cut off still responded, either by running
or by movements of the antennae or legs. The most puzzling
feature is the real significance of sound receptors in an order where
sound production appears to be confined to relatively few species.
In some cases, e.g., Vanessa butterflies, a faculty for sound pro-
duction prevails and yet tympanal organs are wanting. It is
possible that sound vibrations — inaudible to human senses in
many cases — may be produced by Lepidoptera and be perceived
by the organs in question. In this way individuals may perhaps
become aware of the proximity of other members of their species,
but it has to be admitted, however, that the suggestion is merely
hypothetical.

6. General Stimulatory Organs

Stimulatory organs are sensory organs whose functions are to
send continuous afferent stimuli to the central nervous system.
These stimuli, it is claimed, are essential to the normal and proper
working of the nerve-muscle system or of certain reflex arcs.
They are necessary for the production and maintenance of muscle
tonus and to increase the kinetic efficiency or the contractility
of the muscles. Most sensory organs, in addition to their normal
specific receptor ability, also probably exercise stimulatory
functions in keeping the nervous system in a reactive condition
[vide Wolsky, 1933).

In the modern interpretation the dorsal ocelli of adult insects

154 THE SENSE ORGANS AND REFLEX BEHAVIOUR

are considered to be photokinetic stimulatory organs (vide also
p. Ill), a view which has some evidence of an experimental
nature in its support : optic investigations of the ocelli are also
in conformity with this interpretation.

The halteres of Diptera, according to von Buddenbrock (1929),
probably come under the category of stimulatory organs. It
would seem on this interpretation that they produce afferent
stimuli owing to their swinging or vibratory movement when an
insect is in flight. The receptors of such stimuli are said to be
either the campaniform or chordotonal sensillae at their bases.
The continuous production of such stimuli, it is claimed, maintains
the requisite muscle tonus required during flight. It is doubtful,
however, whether this explanation covers their whole function,
since their extirpation may lead to complete loss of ability for
flight. Also, certain wingless Diptera still retain well-developed
halteres.

According to Wigglesworth (1934) the antennae in Rhodnius
exercise a stimulatory function. When both antennae are
amputated the insect falls in a condition of torpor from which
it is only aroused with difficulty, but even then its movements are
less active than in the iminjured insect. Normal insects are
highly responsive to slight air currents, but those without antennae
give no response even when blown upon strongly. The torpor,
it is claimed, is not a shock effect due to the amputation of the
antennae, since it is not apparent until an hour or two later.
Furthermore, it does not occur when the wings, legs or proboscis
are cut off. Wigglesworth concludes that an important function
of the antennae in Rhodnius is a kinetic one, which maintains the
nervous system in a responsive condition so that it will readily
react to stimuli.

It is possible that the campaniform sensillae, which have been
investigated for many years by Mclndoo, are stimulating organs.
Their extremely wide distribution over the insect body and
appendages in some insects is in accordance with this view, and
it may be that their function is called into activity by thermal
and olfactory stimuli. A tendency, however, to ascribe a
stimulatory function to sensillae whose functions are very

REACTIONS TO CHEMICAL STIMULI 155

inadequately understood does not advance knowledge of the
subject unless backed by proper experimental evidence.

7. Reflex Behaviour and Practical Entomology

The aim of the practical entomologist is insect control, and in
seeking to attain this object his first step is the investigation of
life-histories. If this study has been more or less thorough a
certain amount of data will probably have been collected with
respect to the behaviour of the species concerned. In the haste
to achieve practical results, further investigation of this kind will
probably be discarded, and the serious work of control embarked
upon. Unsatisfactory results so often follow this procedure that
it is coming to be realised that more basic studies are required,
and an increasing tendency is being shown to explore the wide
field of insect behaviour. It has come about that the sensory
reactions of insects now often form part of the programme in the
search for more adequate means of repression. The study of these
reactions is fundamental to many problems of applied entomology,
in that it affords possibilities that advantage may be taken of
specific kinds of behaviour in devising methods of control.

Reactions to Chemical Stimuli. The most promising field in the
practical study of insect behaviour is afforded by their reactions
to chemical stimuli, which have come to be included under the
general term of chemotropism. Since the first critical experiments
were conducted in this subject by Barrows in 1907, a very con-
siderable literature has accumulated. It will serve, however, no
useful purpose to review all these numerous publications, and for
bibliographical information reference may be made to papers by
Imms and Husain (1920) and by Mclndoo (1927). Almost all the
investigations so far conducted in this field are of a purely empirical
kind, and have been prompted by n^ed for discovering chemical
compounds attractive or, less often, repellent to specific insects.
The Aristotelian view, that an animal only moves for a purpose,
has been the guiding principle in investigations of this kind. The
habits of a particular species, for example, frequently suggest the
practical ultilisation of chemical compounds improbable to come
under notice, except fortuitously, in other ways. Alongside work

156 THE SENSE ORGANS AND BEFLEX BEHAVIOUR

prompted by such metaphysical ideas, there is a need for a more
scientific analysis of insect behaviour, and for the discovery of any
quantitative laws that may govern their reactions. In this con-
nection it may be pointed out that insects have been shown to be
attracted to chemical compounds that are neither associated
components of their food, nor occur in the natural environment
of the species concerned.

The fact that insects are powerfully attracted to certain kinds
of food materials is a matter of common observation. It has only
been during about the last twenty years, however, that economic
entomologists have made systematic efforts to take advantage of
this phase of behaviour in devising methods of repression. The
practice of exposing traps or baits, containing molasses as their
chief chemical constituent, has now become a widespread and
familiar measure. Little is definitely known with respect to the
specific attractive ingredients of this mixture of substances, but
all observations show that its chemotropic properties are greatly
enhanced in the presence of fermentation. It is especially
attractive to many species of moths, particularly to those of the
family Noctuidse, while numerous species of Diptera, whose larvae
inhabit decaying organic matter, likewise respond. Molasses
baits were at one time a common method of controlling grape-
vine moths in Europe, but in recent years it has fallen into
disrepute. At the present time much attention is being given
to chemotropic measures in the control of the Codling Moth
(Cydia pomonella) and the Oriental Peach Moth (C (Grapholitha)
molesta). Yothers (1927) has described the results of three years'
experiments in trapping the first-mentioned insect. A number of
substances were tested, but fermented molasses proved the most
efficient, and captured many thousands of moths per acre during a
season. About 55 to 60 per cent, of the insects captured were
females, and among them about 95 per cent, were gravid, having
laid none or very few of their eggs before capture. He stresses
the use of such baited traps for determining the beginning and
end of the season of flight of the insect, the beginning and end
of each generation, and the period of maximum abundance.
Information thus gained may be used to advantage in arranging

REACTIONS TO CHEMICAL STIMULI 157

dates of application of spraying. Temperature affected the
efficacy of the measure, since the majority of the insects responded
while the mean thermometer reading was 70° F. or above. A
sudden fall in temperature will practically stop all moths visiting
the baited traps. His conclusion is that a sufficiently attractive
lure may be discovered which will destroy a large enough number
of moths at a cost that may recommend its adoption as a supple-
mentary control measure. Fowler (1927) carried out very similar
experiments in South Australia, using fermented apple juice,
with and without the addition of 10 per cent, by weight of sugar^
and a solution of molasses to which a small amount of yeast
powder had been added. On the whole it was found that apple
juice proved the more attractive of the three mixtures. He
concludes that it would probably be profitable to maintain traps
during the maximum emergence periods of the insect rather than
for the whole flight season. Their greatest value lies in their
use in determining dates of maximum emergence, so as to make it
possible to fix the most advantageous times for spraying.

The control of the Oriental Peach Moth by chemotropic means
has been the subject of experiments by Peterson (1925), who found
that large mmibers of this insect responded to a molasses-yeast
mixture. He later (1927) tested about 250 aromatic chemicals
as possible attractants, but several fermenting sugar-producing
substances proved the most efficient. Fermenting fruits (dried
fruit in water) were also found to attract large numbers of the
moth. Several other investigators have found that the insect
responds readily to suitable chemical stimuli : Frost (1928)
believes that chemotropic methods may find a definite place in
the control of this insect, notwithstanding the fact that the
incidence of prolonged low temperatures or high precipitation
exercises an adverse influence upon their efficacy. The inexpen-
siveness of the method, the simplicity of the apparatus (pails only
being used) and the small amount of labour required, all point in
its favour. Stear (1928), on the other hand, using a 10 per cent,
solution of molasses in water, arrived at an opposite opinion. He
states that two years' experiments failed to give appreciable con-
trol. By marking moths and releasing them, and noting the

158 THE SENSE ORGANS AND REFLEX BEHAVIOUR

numbers of such insects that were caught in the traps, he found
that 30 per cent, was the maximum number caught over a period
of nine days after release. Assuming that this percentage is
approximately the percentage taken of the entire local population
of the species, the particular chemotropic method adopted is
ineffective, especially since many of the moths caught have
opportunity for oviposition before capture.

Lloyd (1920) recommends the use of chemotropic measures
against the moth Hadena (Polia) oleracea, whose larvse are
destructive to tomatoes grown under glass in England. He
found that a mixture of cane molasses, ale and 1 per cent, sodium
fluoride is an effective means of control, particularly when following
periods during which spraying against the larvae is impracticable.
The relatively constant environmental conditions afforded by
glasshouse cultivation greatly favours the use of this measure,
which is now regularly practised by growers.

In 1910 Verschaffelt conducted experiments in order to ascertain
the factors which determine the adoption of specific food-plants by
phytophagous larvae. He utilised caterpillars of Pieris which feed
upon certain of the Cruciferae along with Tropwolium and Reseda.
In these plants there occurs a group of glucosides — the mustard
oils — and Verschaffelt took a solution of sinigrin, which is a
constituent of black nuistard, and spread it uniformly over leaves
of plants the Pieris larvae had previously refused to eat. Leaves
thus treated were readily devoured, and he concluded that Pieris
larvae exhibit a strong positive reaction to mustard oils, and it is
their presence in the leaves of certain plants that determines the
selection of the latter by the larvae for their food. By similar
methods this same investigator showed that larvae of the saw-fly
Priophorus {Cladius) pyri, which feed upon certain Rosaceae, are
apparently attracted by the glucoside amygdalin.

Mclndoo (1926) has described a laboratory instrument which
he terms an " olfactometer " designed for the purpose of testing
the chemotropic responses of insects. His main object was to
prove that plants attract insects by the odours which they emit.
A small potted potato plant was placed in a special chamber of
the instrument and a gentle current of air was drawn over it into

REACTIONS TO CHEMICAL STIMULI 159

one fork of a Y-tube by means of a suction apparatus. Experi-
ments carried out with the Colorado potato beetle {Leptinotarsa
decemlineata) showed that when individuals were introduced
into the Y-tube an average of 62-7 entered the arm through which
air from the chamber containing the potato plant was being drawn.
The insects also responded to the odours of water extracts of
potato tubers and potato foliage ; odour from the steam distillate
of the tubers was only slightly attractive, while a more pronounced
response was obtained in the case of distillates of the potato-
foliage. Power and Chesnut (1925), acting upon the assumption
that the cotton boll weevil is attracted to its food-plant by some
odorous volatile substance, undertook to isolate this substance.
Among the twelve substances isolated three appeared to be of some
significance, viz., an essential oil, ammonia and trimethylamine.
The percentage of oil isolated appeared to be too small to be of
chemotropic significance, and, despite the fact that more ammonia
than trimethylamine was found, they concluded that if the cotton
plant contains a positive chemotropic constituent it is probably
trimethylamine. No very conclusive results appear to have been
derived from this work, and Mclndoo doubts whether it is possible
to reproduce accurately the odour or odours, which emanate from
a plant, by using certain constituents obtained from the plant by
chemical means. In 1927 Richmond, experimenting with various
essential oils with reference to the discovery of agents attractive to
the Japanese beetle {Popillia japonica), found sassafras oil to be
by far the most efficient in this respect. The constituent of this
oil which exercised the most powerful attraction to the insect
proved to be geraniol, which is an unsaturated alcohol belonging
to the class of terpene compounds. This alcohol appears to be
one of the most powerful specific chemotropic agents so far
discovered, and enormous numbers of the Popillia have been
caught, through its agency, in suitably designed traps. Although
geraniol occurs as a constituent of the apple, which is one of the
favourite food-plants of the beetle, it is by no means clear as to
what part it plays as an agent influencing the selection of this
and other plant species by the insect in question.

A considerable amount of investigation has been devoted to

IfiO THE SENSE ORGANS AND REFLEX BEHAVIOUR

the cheinotropic responses of Diptcra, especially those breeding
in decaying organic material of various kinds. Many species are
attracted to varions individual by-products produced as the result
of fermentation of different kinds. The early experiments of
Barrows dealt with the reactions of Drosophila melanogaster
{ainpelophila), which occurs in great numbers around cider presses,
packing sheds, orchards, and other situations where fermenting
fruit is present : within the latter it deposits its eggs and its larva?
develop. Barrows' experiments were made with various con-
stituents of fermenting fruit : ethyl alcohol, ethyl acetate and
acetic and lactic acids were tested separately and in mixture. The
insect was found to exhibit the maximum response to a mixture
of ethyl alcohol of 20 per cent, strength and acetic acid of 5 per
cent. It was further ascertained that cider vinegar, and fermented
cider, contain alcohol and acetic acid in percentages closely
approximating to those just mentioned. Subsequent investigators
have devoted a considerable amount of attention to the repression
of the house-fly by chemotropic means. Many highly complex
food materials have been advocated as baits for trapping this
insect, leaving the specific attractive constituents undetermined.
Among individual reagents tested, low concentrations of ethyl
alcohol (3 to 8 per cent.) and other alcohols are definitely attractive,
while the addition of sugars is stated to enhance this property.
Speyer (1920) found that alcohols, aldehydes and acids containing
the methyl group CH3, and whose molecular weight is 30 or over,
are positively chemotropic, especially when the methyl grouj) is
united to (CHg)^. He suggests that the molecular group
CH3(CH2)aj is the actual food stimulus concerned, but his experi-
ments were of too restricted a character to be regarded as other
than purely tentative. Imms and Husain (1920) tested a range
of substances by means of field traps. Ethyl alcohol alone
exhibited little or no chemotropic properties in various con-
centrations, but with the addition of small amounts of butyric,
valerianic or acetic acids it exercised a powerful stimulus. Since
dilute acids themselves did not prove attractive, the respective
esters were probably the agents involved. Both sexes of species
of Rhyphus, Sarcophaga, Calliphora and various Muscidse and

REACTIONS TO CHEMICAL STIMULI 161

Anthomyidse responded in large numbers. Roubaud and Veillon
(1922) experimented with the attraction of simple products of
fermentation with regard to nineteen species of flies. They
found that sugars, non-fermented sugar products, alcohols and
ammonia produced little or no positive reactions on the part of
the insects. Among the various compounds resulting from
organic fermentation the most attractive were ammonium
sulphohydrate, ammonium valerianate, valerianic acid and
trimethylamine. Peterson (1924) found that fermenting yeasts
were highly attractive to the flies Hylemyia cilicrura and H.
antiqua, probably, as he remarks, on account of the alcohols and
ethereal odours produced. Various alcohols, added to honey and
water, proved powerfully attractive, especially alyl alcohol and
isopropyl alcohol. A number of other paj^ers of a kindred nature
to the foregoing have appeared, and the results obtained indicate
the complexity of the subject. By no means all the substances
tested induce feeding reflexes, and in many cases the significance
of the responses is unknown — they appear to be non-purposeful
in that they are unconnected with any vital functions. There
are other instances where certain compounds have been shown to
exercise an influence in connection with oviposition. Both
Roubaud and Veillon, and Richardson, have pointed out that the
female house-fly, for example, is attracted to the manure heap
for purposes of egg-laying, feeding mainly taking place elsewhere.
The part performed by chemical stimuli in inducing oviposition
has been discussed by Richardson (1925), who gives reasons for
concluding that, in the house-fly, ammonia is one of the volatile
constituents of the manure heap that is important in this respect.
Crumb (1924) has published the results of experiments showing
that the addition of hydrogen sulphide, old yeast infusion, methane
or stale urine to water, in each case, t^nds to induce oviposition in
the mosquito Culexpipiens. Ordinary tap water, used as a control,
exercised no influence.

There are, again, other examples where specific compounds have
proved attractive to the male sex only. This was well shown in
the important experiments of Howlett (1912), dealing with the
reactions of certain Indian fruit -flies. He found that citronella oil

R.A. ENTOMOLOGY. 6

162 THE SENSE ORGANS AND REFLEX BEHAVIOUR

exercises a remarkable attraction for the males of Dacus diversus
and D. zo7iatus, and suggested that possibly this reagent is allied
to a secretion emitted by the females, and that the phenomenon
of the male alone being attracted is to be regarded as a repro-
duction, by artificial means, of a sexual attraction similar in kind
to that which operates in most cases of " assembling." In a later
paj^er (1915) Howlett conducted a further series of experiments,
confirming his previous results that certain odours are remarkably
attractive to male flies of the genus Dacus and that, by the
employment of attractive substances, the movements of the flies
can to a great extent be controlled in any direction. Three of the
common species {D. diversus, ferrugineus and zonatus) normally
breed respectively in (1) anthers of Cucurbitacese, (2) fruits of
Solanaceae and mango, and (3) peach, guava, mango and other
fruits. D. diversus (1) is most strongly attracted by iso-eugenol,
zonatus by methyl-eugenol, and ferrugineus (2) by both iso- and
methyl-eugenol. The odours of these substances have not yet
been identified with those of the plants which constitute the
normal breeding places, but male flies have been found attracted
to mango, papaya, a Cycad and Colocasia, plants with a very
characteristic smell similar to that of eugenol-derivatives. Females
have never been seen to frequent these places or to breed in them,
but more extended observation on this point is needed. Three
explanations suggest themselves, (a) That the smell is a direct
sexual guiding smell, emitted by the female, as previously suggested,
but, nevertheless, the young crushed females were not found to
attract males, {h) The smell is not emitted by the female, but may
be termed an " indirect " sexual guide to the plants where the
females are accustomed to congregate for breeding purposes.
Under these circumstances it is difficult to see why females should
not be attracted by the odoriferous chemicals, (c) The odour is a
food smell ; if this be so it can only be attractive to males. None
of these fundamental questions has so far been solved, but,
nevertheless, Howlett's researches have proved a landmark on
the applied side of the subject, and have indicated some of its
possibilities. Mention needs also to be made of the work of the
Severins (1914), who followed up the reactions of another species

REACTIONS TO CHEMICAL STIMULI 163

of fruit-fly, viz., Ceratitis capitata. They showed that certain
mineral oils, including kerosene, naphtha and benzene, similarly
exercise a powerful chemotropic attraction for the males of this
species. The significance of the phenomenon, however, remains
unexplained.

Practical measures designed to exploit the reactions of insects
to chemical stimuli afford abundant promise of extended
api^lication. Much of the data at present available, however,
is of a purely tentative character. It is too heterogeneous and
scanty to allow of any fundamental conclusions being drawn.
We know as yet, for example, very little with respect to the
relations between chemical constitution and the attractive
properties of different compounds. An indication of possibilities
in this direction, however, has been given by Cook (1926), but
his results are marred by the fact that, although about 50,000
insects were involved in his experiments, their specific identities
are unstated. The discovery of chemical attractants is only one
aspect of the problem. Due cognisance needs to be taken of the
fact that reactions to such substances are only the end-results of
a series of interacting stimuli. On the one side are external
factors, including temperature, humidity and light, and on the
other are the internal physiological states of the insects themselves.

The preliminary study of the reactions of insects in the
laboratory under properly controlled conditions is obviously an
important aspect of any work designed to exploit the subject
from an economic standpoint. A variety of instruments or
" olfactometers " has been used. One devised by Mclndoo
(1926) has already been referred to. When investigating blow-
flies this worker used an instrument of different pattern (Mclndoo,
1938) and partly made of wood. Other kinds have been described
by Barrows (1907) in connection with experiments with Drosophila :
by Ripley and Hepburn (1929) in their studies with the South
African fruit-fly Pterandus rosa : by Hoskins and Craig (1934)
with reference to Lucilia : and by Marshall (1935a) for the hive
bee. Some of these instruments show obvious disadvantages :
either they are too slow in operation or give inconclusive results,
or are not easily washed and rendered inodorous. Also, the

164 THE SENSE ORGANS AND REFLEX BEHAVIOUR

conditions under which the tests were carried out were not always
above criticism. The reader who is interested in this aspect of
the subject is advised to consult the papers by Hoskins and
Craig and by Marshall.

Reactions to Light. In many countries practical entomologists
have made use of light traps as a means for the quantitative
attraction and destruction of noxious species of Lepidoptera, and
in some cases of Coleoptera also. These traps are usually so
constructed that insects flying to the light fall into a surrounding
receptacle containing either oil, or water covered by a film of oil.
On the whole, the results derived from the application of this
method cannot be regarded as commensurate with the trouble
and expense involved. In temperate countries the success of such
a method, as with chemotropic traps, is considerably influenced
by meteorological conditions. Temperature, for example, is
significant in that when it falls below certain limits, varying
according to the time of the year and the species concerned,
phototropic responses almost cease. Such factors as precipitation,
general humidity, wind, etc., appear to exercise a variable effect,
but little has been determined with regard to their respective
influences. In the tropics, on the other hand, where much more
uniform conditions prevail, such variable factors are far less in
evidence. It is frequently claimed that the preponderance of
male insects that are attracted discounts the value of light traps.
A further disadvantage that has been emphasised is that the
female insects have, for the most part, already deposited a quota
of their eggs before being trapped. In some cases these drawbacks
are obviously well founded, but in others premature conclusions
have been draw n on account of insufficiency of experimental data.
Due allowance needs to be made for the fact that the males of a
number of species tend to appear earlier in the season than the
females. Also, cases are known where the males are on the wing
during one part of the night, and the females during another.
There seems, also, little doubt that more can be done towards
improving the efficiency of the light source itself. More attention
needs to be given to light rich in violet and ultra-violet rays (vide
p. 121). In this connection it may be added that very large

REACTIONS TO LIGHT 165

catches of moths have been recorded by Neustadt (1913), usmg
incandescent lamps containing mercury vapour, in which these
component rays were well represented.

More or less detailed analyses with respect to the species of
Lepidoptera attracted to light traps, the proportions of the sexes
represented and other data will be found in publications by
Willcocks (1916), Criddle (1918), Ainslie (1917), Shiraki (1917),
Turner (1918, 1920), Ballard (1923), Theobald (1926), Husain and
Khan (1934), and others. For the most part the results recorded
by these observers are not encouraging, but in some cases the
experiments were far too inadequate to be conclusive. Shiraki,
however, has obtained promising results from the use of light
traps with reference to the paddy borer moth {Schoenobius
incertellus) in Formosa. He records five years' observations, and
during this period many thousands ot individuals were captured
by the method in different localities. In practically all cases his
figures show a great preponderance of females, although at the
time of the commencement of the emergence of each generation
of moths, more males than females responded. Shiraki's observa-
tions are supported by those of Ballard (1923) in India, who
likewise demonstrated that an enormously greater number of
female moths of this species, as compared with males, are caught
by means of light traps. He also showed that out of 13,640
trapped females, 35-5 per cent, had previously laid their eggs
before being caught. Theobald (1926) has recorded the results of
two years' experiments in England, and concluded that acetylene
lamps are beneficial in destroying Tortrices in orchard districts
where these pests are serious. Unlike some observers, he fcund
that no objection could be raised against light traps on the grounds
that large numbers of beneficial insects were destroyed at the
same time.

Husain and Khan (1934) carried out investigations with light
traps in the Punjab with reference to the Pink Bollworm Moth
(Platyedra gossypiella), which is a well-known pest of cotton. After
enumerating certain disadvantages attendant upon the use of
this method, they concluded that a light trap, instead of minimising
attacks in the field where the trap was stationed, only concentrated

166 THE SENSE ORGANS AND REFLEX BEHAVIOUR

moths from adjoining fields and thus intensified the attack in the
area concerned.

The use of Hght traps for the destruction of injurious Coleoptera
has been principally applied against members of the Scaraba^idse,
more especially the genus Lachnosterna. Sanders and Fracker
(1916) record the great attraction artificial light has for various
species of this genus. During May and June, 1914, 440,000
examples were captured by this means in a single locality, while
forty light traps in five localities captured 1,036,400 of the beetles.
The males caught greatly exceeded the females in number, but in
some of the commoner species females formed a larger proportion
earlier in the season. In Porto Rico, Van Zwaluwenburg records
the attraction of large numbers of Lachnosterna to light, the
sexes responding in about equal numbers, and he estimates that
only about 17 per cent, of the females had completed oviposition
at the time of capture.

An investigation of the activity and frequency of insects based
upon those caught in light traps has been carried out for several
years by WilUams (1935, 1936) at the Rothamsted Experimental
Station. In the analysis of the figures obtained it was demon-
strated that, taking all insects together, the maximum number was
trapped at the beginning of the night with a steady decline towards
the morning. The Diptera, which composed about 76 per cent,
of the total, showed a maximum in the first part of the night and
the Lepidoptera, which formed 16 per cent, of the total, were
caught in largest numbers in the second to fourth periods — the
night being subdivided into eight periods of equal length. When
there is a marked difference in respect to the number of individuals
of each sex, the females appeared earlier in the light trap than the
males. The total captures of insects in two successive years
exceeded 100,000 specimens per year, so the numbers appear to
be sufficient to warrant the deductions drawn. An examination
of the insects captured during three successive years was made in
order to test the validity of the general belief that the nocturnal
activity of certain kinds of insects is reduced at full moon. The
analysis had special reference to the Noctuidae, and it was
considered that the lunar effect on the captures at a light trap

REACTIONS TO LIGHT 167

had been demonstrated. The opinion is advanced that a
physiological effect on the insects is involved and that a reduction
in the efficiency of the trap when the moon is shining is not the
explanation. Further experiments to test this conclusion are
being carried out by using traps not dependent upon light as the
source of their attracting power.

The use of light traps for determining the population density
of the Rice-Borer by Harukawa, Takato and Kumashiro (1935)
has not proved reliable over a period of years. The larvae may
become very abundant even when the moth population, as
determined by light traps, is not very large or, in other words,
the larval population is not proportional to that of the number
of moths captured. The light trap captures, therefore, are not a
trustworthy index of the abundance of larvae to be expected in the
next generation. The population density can be best determined
by collecting injured rice haulms and by direct counting of the
larvae found.

The reactions of insects to colour {vide also p. 113), apart from
questions of light intensity, have certain practical applications, and
merit more extended study than has been accorded to them.
Nuttall showed many years ago that mosquitoes have a marked
tendency to congregate and rest on a dark blue surface — a fact
which has led to certain practical developments. Lloyd (1921)
demonstrated that the white -fly Asterochiton vapor arioruin is
powerfully attracted to a yellow colour, while Folsom and Bondy
(1930) found that applications of calcium arsenate dust to cotton
plants are often followed by heavy infestations of Aphis gossypii.
One of the factors operating in this case was found to be due to
the positive phototropic reaction of the winged aphides to the
white deposit of calcium arsenate on the plants. Other white
dusts, such as calcium carbonate, starch or flours, were also found
to induce a similar response on the part of that insect. In tropical
Africa the method of trapping tsetse-flies, on tanglefoot screens
attached to moving automobiles, ajDpears to offer room for
exploration with respect to the colour surfaces most attractive
to those species.

Concluding Remarks. Reactions to chemical stimuH, and to

168 THE SENSE ORGANS AND REFLEX BEHAVIOUR

light rays, are the only reflex responses of insects that have been
found to lend themselves to practical utilisation to any consider-
able extent. It has been shown that chemotropic responses afford
a highly promising field for exploration, while, on the other hand,
the phototropic responses of insects have so far yielded less
encouraging results. It is unlikely that material advances will
come from the application of existing methods of utilising these,
or other phases of insect behaviour, until much more thorough
exploration of their fundamental aspects has been undertaken.
Up to the present so many of the experiments have neglected to
take into account the numerous interacting factors involved.
Until their influence can be adequately evaluated the empirical
methods, so far practised, are scarcely amenable to material
improvement. Properly controlled experiments, using the most
scientific equipment available, afford the only sure method of
obtaining knowledge capable of being turned to practical
advantage.

Literature

A. The General Subject

Berlese, 1909. " Gli Insetti." Milan.

VON Frisch, 1924. " Sinnesphysiologie und ' Sprache ' der Bienen." Berlin.

LoEB, 1918. " Forced Movements, Tropisms and Animal Conduct." London

and Philadelphia.
McIndoo, 1926. Journ. Agric. Res., XXXIIL, 1095.
1929. Smithsonian Misc. Coll., LXXXL, No. 10.
SiHLER, 1924. Zool. Jahrb., Abt. Anat. Ont., XLV., 519.
Snodgrass, 1926. Smithsonian Misc. Coll., LXXVIL, No. 8.
VoGEL, 1923. Zeits. wiss. Zool., CXX., 281.
Wheeler, 1928. " The Social Insects." London.

B. The Eyes and Light Perception

Bertholf, 1931. Journ. Agric. Res., XLIL, 379 ; XLIIL, 703.

1932. Zeits. vergl. Physiol,, XVIIL, 32.
BozLER, 1925. Ibid., IIL, 145.
Buddenbrock, 1931. Ibid., XV., 597.

BuDDENBROCK and ScHULTZ, 1933. Zool. Jahrb., Abt. Physiol., LIL, 513.

Ellsworth, 1933. Aim. Ent. Soc. Am., XXVL, 203.

Eltringham, 1919. Trans. Ent. Soc, 1.

Friedrich, 1931. Internat. Ent. Zeits., XXV., 326.

VON Frisch, 1914. Zool. Jahrb., Abt. Physiol., XXXV., 1.

Gaffron, 1934. Zeits. vergl. Physiol., XX.

Glaser, 1925. Pstjche, XXXIL, 285.

Hertz, 1931. Zeits. vergl. Physiol., XIV., 629.

1933. Biol. Zentralb., LIIL, 10.

1934. Zeits. vergl. Physiol., XXL, 579.

1935. Die Natiimiss,, XXXVL, 618.

LITERATURE 169

Ilse, 1928. Zeits. vergl. Physiol., VIII., 658.

1933. Ibid., XVII.
Knoll, 1921-26. " Insekten und Blumen." Wien.
KuGLER, 1930. Planta, X.
KuHN, 1927. Zeits. vergl. Physiol, V., 762.
MuLLER, 1931. Ibid., XIV., 348.

Peterson and Heussler, 1928. Ann. Ent. Soc. Am., XXL, 353.
Priebatsch, 1933. Zeits. vergl. Physiol., XIX., 453.
Sander, 1933. Ibid., XX., 267.
SCHLEGTENDAHL, 1934. IMd., XX., 545.
ScHLiEPER, 1927. Zeits. vergl. Physiol., XX., 267.
Urban, 1932. Zeits. wiss. Zool., CXL., 291.
ViGiER, 1904. C. R. Acad. Sci., CXXXVIII., 775.
Wheeler, 1936. Biol. Bull., LXX., 185.
Wolf, 1933. Zeits. vergl. Physiol, XX., 151.
Zerrahn, 1933. Zeits. wiss. Zool, XX., 117.

C. Chemical Stimuli and their Receptor Organs
Abbott, 1928. Psyche, XXXV., 201.

Anderson, 1932. Journ. Exp. Zool, LXIIL, 235.

BuGNiON and Popoff, 1911. Archiv. Zool Exp., 5 ser., VII., 664.

Chrystal, 1930. Oxford Forestry Memoirs, No. 11.

Crow, 1932. Physiol Zool, V., 16.

Eltringham, 1933. Trans. Roy. Ent. Soc, LXXXL, 33.

VON Frisch, 1921. Zool Jahrb., Zool u. Physiol, XXXVIII., 449.

1927. Naturwiss., XV., 321.

1928. Ibid., XVI., 307.

1930. Ibid., XVIII., 169.
Glaser, 1927. Psijche, XXXIV., 209.

KuNZE, 1927. Zool. Jahrb., Abt. Zool. u. Physiol, XLIV., 187.
Marshall, 1935. Trans. Roy. Ent. Soc, LXXXIIL, 49.

1935a. Journ. Exp. Biol, XII., 17.
McIndoo, 1914. Journ. Exp. Zool, XVI., 265.

1916. Smithsonian Misc. Coll., LXV., No. 14.

1933. Journ. Agric Res., XLVL, 607.

1934. Journ. Morph., LVL, 445.
MiNNiCH, 1921. Journ. Exp. Zool, XXXIII., 173.

1922. Ibid., XXXV., 57.

1922a. Ibid., XXXVI., 445.

1924. Ibid., XXXIX., .339.

1926. Biol Bull, LI., 166.

1926a. Anal Record., XXXIV., 126.

1929. Zeits. vergl Physiol, XL, 1.

1931. Journ. Exp. Zool, LX., 121.

1932. Ibid., LXL, 375.

Newton, 1931. Quart. Journ. Mic Sci., LXXIV., 647.

Ripley and Hepburn, 1929. DepL Agric S. Africa., Ent. Mem., VL, 55.

Valentine, 1931. Journ. Exp. Zool, LVIIL, 168.

Weis, 1930. Zeits. vergl Physiol, XII., 206.

D. Chordotonal Organs and Reactions to Vibrations

Abbott, 1927. Psyche, XXXIV., 129.
Baier, 1930. Zool Jahrb., Abt. Allg. Zool, XLVIL, 151.
Barnes, 1930. Journ. Gen. Psych., II., 318.
1931. Ann. Ent. Soc Am,, XXV., 824.

170 THE SENSE ORGANS AND REFLEX BEHAVIOUR

Eggers, 1919. Zool. Jahrb., Anat. Out., XLI., 272.

1923. Zool. Anz., LVIL, 224.

1928. " Die Stiffiihrenden Sinnesorgane." Zool. Bausteine, Bd. II.,
Heft 1. Berlin.
Hagemann, 1910. Zool. Jahrb., Anat., XXX., 394.
Hertweck, 1931. Zeits. wiss. Zool., CXXXIX., 559.
Hess, 1917. Ann. Ent. Soc. Am., X., 63.
Mayer, 1874. ^wcr. A'a/., VIII., 577.
MiNNiCH, 1925. Journ. Exp. Zool., XLIL, 443.
Myers, 1929. " Insect Singers." London.

PuMPHREY and Rawdon-Smith, 1936. Proc. Hoy. Soc. B., CXXI., 18.
Slifer, 1936. Ent. News, XLVII, 174.
VoGEL, 1923. Zeits. Anat. u. Entwicklungs., LXVIL, 190.
Wever and Bray, 1933. Journ. Cell. Comp. Physiol., IV., 79.

E. Stimulatory Organs

BozLER, 1926. Loc. cit. (p. 168).

Buddenbrock, 1919. Pfliiger's Archiv., CLXXV., 125.

1924-28. " Grundriss der vergleichenden Physiologic." Berlin.
WiGGLESwoRTH, 1934. Jourti. Exp. Biol., XL, 120.
WoLSKY, 1933. Biol. Rev., VIIL, 370.

F. Reflex Behaviour and Practical Entomology

AiNSLiE, 1917. Journ. Econ. Ent., X., 114,

Ballard, 1923. Mem. Agric. Dept. India, Ent. Series, VII., No. 10.

Barrows, 1907. Journ. Exp. Zool., IV., 515.

Cook, 1926. Journ. Agric. Res., XXXIL, 347,

Criddle, 1918. Canad. Ent., L., 73.

Crumb, 1924. Ent. News, XXV., 242.

FoLSOM and Bondy, 1930. U.S. Dept. Agric, Circ. 116.

Fowler, 1927. Journ. Dept. Agric. S. Australia, XXXL, 480.

Frost, 1928. Journ. Econ. Ent., XXL, 339.

Harukawa, Takato and Kumashiro, 1935. Berohara Inst. f. Landwirts.

Forsch. in Kuraschiki, VI., 1.
HowLETT, 1912. Trans. Ent. Soc. London, 412.

1915. Bull. Ent. Res., VI., 297.
HusAiN and Khan, 1934. Ind. Journ. Agric. Sci., IV., 261.
Imms and Husain, 1920. Ann. App. Biol., VL, 269.
Lloyd, 1920. lbid.,\U., m.

1921. Bull. Ent. Res., XII., 355.
McIndoo, 1926. Journ. Econ. Ent., XIX., 545.

1927. Journ. Agric. Res., XXXIIL, 1095.

1928. Journ. Econ. Ent., XXL, 903.
Neustadt, 1913. Journ. d'' Agric. troj)., January 31st.
Peterson, 1924. Journ. Econ. Ent., XVIL, 87.

1925. Ibid., XVIIL, 181.

1927. Ibid., XX., 174.
Power and Chesnut, 1925. Journ. Am. Chem. Soc, XLVIL, 1751.
Richardson, 1925. Bidl. 1324, U.S. Dept. Agric
Richmond, 1927. Proc Ent. Soc Washington, XXIX., 36.
RouBAUD and Veillon, 1922. Ann. Inst. Pasteur, XXXVL, 752.
Sanders and Fracker, 1916. Journ. Econ. Ent., IX., 253.
Severin and Severin, 1914. Journ. Anim. Behav., IV., 223.
Shiraki, 1917. " Paddy Borer, Shoenobius incertellus,^^ Taihoku Agric
Expt, Sta,

LITERATURE 171

Speyer, 1920. Ann. App. Biol., VII., 124.
Stear, 1928. Journ. Econ. Ent., XXL, 565.
Theobald, 1926. Journ. Roy. Ilortic. Soc, LI., 314.
Turner, 1918. Journ. Agric. Res., XIV., 135.

1920. Ibid., XVIIL, 475.
Verschaffelt, 1910. Amsterdam, Proc. Sci. k. Akad. Wet., XIIL, 536.

(English.)
WiLLCOCKS, 1916. " The Insect and Related Pests Injurious to the Cotton

Plant." Cairo. 314.
Williams, 1935. Trans. Roy. Ent. Soc, LXXXIIL, 523.

1936. Phil. Trans. Roy. Soc. B., CCXXVL, 357.
Yothers, 1927. Journ. Econ. Entom., XX., 567.

CHAPTER VII
THE FUNDAMENTAL ASPECT OF COLORATION

Structural Colours, p. 172. Diffraction, p. 174 ; Interference,
p. 175 ; Scattering of Light by Minute Particles, p. 179 ; Selective
Reflection, p. 180. Pigmentary Colours, p. 183. 1. Chlorophyll and
other Derived Pigments, p. 183 ; 2. Hcemoglobin and Allied Pigments,
p. 189 ; 3. Pigments of Protein Origin, p. 190 ; 4. Pigments zvith Purine
Bases, p. 193. Combination Colours, p. 194, Further Remarks on
Insect Coloration, p. 195. Literature, p. 203.

Insect coloration has attracted a large amount of attention
from biologists in connection with the relations of both the
individual and the race to the environment. Extensive recourse
has been made to the amazing variety of colour patterns and
markings of insects with the aim of adding fresh arguments
bearing upon the central doctrine of natural selection. The
relations of colour phenomena to current biological theories have
been repeatedly discussed, and consequently it is not intended to
deal with this aspect in the present chapter. Attention will be
more especially directed to recent research bearing upon the
fundamental phenomena of insect colours and pigments, their
derivation, and their significance in the j^hysiological processes
of the individual.

It is well known that insect colours fall very naturally into
three groups : (1) structural or physical colours, (2) pigmentary
or chemical colours, and (3) combination or physico-chemical
colours.

Structural Colours

Structural colours are often extremely difficult to elucidate, and
the findings of recent research have not in all cases been in general
agreement.

White is usually the result of the scattering of light by reflection

172

STRUCTURAL COLOURS 173

and refraction by minute structural details exhibited in the scales
or other cuticular parts. As Mason (1926) observes, whenever we
have practically colourless cuticle, in such a form as to present
many more or less unordered reflecting surfaces, whiteness is
produced. All whites in insects are ultimately explainable upon
a purely structural basis. Proof of this is afforded by the fact
that if the air in contact with the parts in question be replaced by
a colourless liquid, whose refractive index closely approximates to
that of chitin {i.e., n = 1-55 — 1-6), whiteness disappears ; after
removal of the liquid and drying the white appearance becomes
fully restored. In the white Pierine butterflies, Hopkins showed,
a number of years ago, that uric acid becomes stored in the wing
scales of certain of these insects, and he regarded the phenomenon
as an example of an excretory product functioning as a colour
producer. According to Mason, the white Pierine butterflies form
no exception to the rule that whiteness is a structural feature only.
Although uric acid is white, and might well function as a pigment
when finely divided, he states that it plays only a very minor
part in producing the white appearance of the butterfly scales.
He mentions that, after such solvents as dilute alkalies have
dissolved out the uric acid, the white remains undiminished after
the wings have been washed and dried. On the other hand, a white
Pierine wing becomes transparent when submerged in liquids of
the appropriate refractive index. He therefore concludes that, in
so far as whiteness is concerned, the presence or absence of uric
acid is of little importance.

A variety of possible causes has been regarded as responsible
for the production of " metallic " or iridescent colours, and the
subject has been treated at length by Onslow (1921), and more
recently by Suffert (1924), and by Mason (1927). The theories
which have been advanced fall under four categories.

1. Diffraction of light at the surface of a grooved structure or
" grating."

2. Interference of light at the surfaces of single or multiple
" thin films."

3. Scattering of shorter light- waves by minute particles smaller
in diameter than the waves themselves.

174 THE FUNDAMENTAL ASPECT OF COLORATION

4. Selective reflection of light from extremely opaque highly
reflecting surfaces.

The history of the subject need not detain us here, but those
who are specially interested should refer to Onslow's paper
previously mentioned. The phenomena involved are so diverse,
and the wealth of material for study so immense, that the tendency
to generalisation involving the application of any one single
theory to account for all cases is beset with pitfalls.

Diffraction. The scales of most insects, whether iridescent or
not, bear striations parallel to their long axes and numbering
almost anything between 4,000 to 16,000 to a centimetre. Onslow
has described a method of taking scale impressions on colourless
collodion films, by which the effect of any surface pattern can be
analysed and isolated from colours produced in other ways.
These replicas show that the upper surfaces of most scales give
good " gratings." Onslow's observation that all insects, the
impressions of whose scales yield spectra, do not themselves
necessarily exhibit iridescence makes it doubtful whether diffrac-
tion is ever a main source of colour, especially as he found that
the brilliant scales of the butterfly Morpho cypris, and others with
the same structure, give an almost flat impression. It is also
noteworthy that the changing colours of the unpigmented wings
of dragon-flies and other insects cannot be explained upon the
basis of a diffraction grating, since striae are entirely wanting in
these cases. The conclusions of Onslow, Suffert, and Mason are
contrary to any idea that diffraction plays much part in insect
coloration. That it may be responsible in certain special cases,
however, cannot be excluded. Mason (1927), who is a very
decided opponent to diffraction being an explanation of iridescence,
states that an exception occurs in the Lamellicorn beetle Sericea
sericea. The elytra of this insect show brilliant iridescence in
direct sunlight and much less in ordinary light. The variations
in iridescence under different tests have led him to look for some
oriented structure as the cause of the phenomenon. This was
found in the presence of fine striae running transversely to the
length of the elytra. The striae are shallow (0-5 — 1-0 /x) and
very evenly spaced apart (1 — 1-5 /x or about 20,000 per inch).

INTERFERENCE COLOURS 175

Collodion impressions retain the same pattern, and the iridescence
exhibited by these impressions is as good as that shown by the
insect itself. All the properties exhibited are in accord with
those of diffraction gratings and are not manifested by thin films
or any other colour-producing structure known.

Interference. The majority of recent investigators are in
favour of the interference theory as being the one which explains
by far the largest number of cases of iridescent colours. In the
wings of many insects colours are exhibited which recall the
iridescence of soap bubbles. It is obvious that these colours are
structural, since they prevail not only in entirely unpigmented
wings, but also in pigmented wings after subjection to bleaching.
When examined by transmitted light, an iridescent insect wing
exhibits only very pale hues which are complementary to those
shown in the same area by reflection. Immersion in liquids of
the correct refractive index largely destroys such colours,
which become restored after thorough washing and drying. The
iridescence is not destroyed by the separation of the upper and
lower wing lamellae, while collodion impressions of either surface
show no iridescence. Mason, who has paid some attention to
iridescent wing membranes, concludes that the phenomenon is
explainable on the basis that the colours are due to multiple thin
films separated by material of slightly different refractive indices.
He maintains that microscopical demonstration of an actual
laminated structure in the upper or lower cuticle is hardly to be
hoped for, since the thinness of the films as calculated (circa 0-2 /x)
must be about at the limit of resolution. Great difficulties would
be involved in making microscopical preparations sufficiently
reliable to avoid the possibility of spurious diffraction images.
He concludes, therefore, that optical tests are more reliable, in
this case, than histological study, and it may be added that the
tests as applied by him afford definite support to his conclusions.

The explanation of scale iridescence is not so obvious as in the
case of wing-membranes, and the subject has received a good deal
of attention by all writers on the structural colours of insects. The
striking resemblances between the optical properties of insect scales
and of thin, colour-producing, artificial films, can be substantiated

176 THE FUNDAMENTAL ASPECT OF COLORATION

by detailed comparisons. Also, the histological evidence afforded
by scale-structure is consistent with the optical phenomena
exhibited.

Mason points out that the colours of the scales are readily recognised
by comparison with artificial thin films, as corresponding to those of
Newton's series, in the second or third orders. This relationship is con-
firmed by the change in hue which takes place with increasing angle of
incidence. In every case this is towards a colour lower in order (change
towards violet, the colour lowest in the second order), and duplicates

Fig. 63. Scale of Urania type (schematic). A, thin vanes or ridges
comprising longitudinal striations. B, multiple films, the seat
of the colour. C, lower lamella. (After Mason.)

that of an artificial thin film held parallel to the reflecting surface of the
scale. The colour produced by swelling a scale is predictable in all cases
as that corresponding to an increase in order (thicker film), while by
compressing a scale the colour changes are predictable in all cases as
corresponding to a decrease in order (thinner film). The destruction
of colour where extreme pressure is applied to the scale corresponds to
bringing the lamellae into optical contact. The loss of iridescence and
metallic lustre, when a scale is permeated by a liquid of its own refractive
index, corresponds to that of thin films under similar conditions. The
fact that liquids markedly higher or lower in refractive index do not
cause total loss of colour is also consistent. The existence of lamellae,
more or less parallel to the plane of the scale, is borne out at the fractured
edges of scales : distorted sections, furthermore, may show the lamellae
separated. Penetration by a viscous liquid is not immediate, but tends

INTERFERENCE COLOURS

111

to proceed in layers, as if it were filling inter-lamellar spaces. The
dimensions of the colour-producing structure are consistent with the
foregoing data, the lamellae of which would only be about 0-3 to 0-5 \x
in maximum thickness : there is space for a number of such films in
any of the specimens examined.

Iridescent scales as studied by Mason are grouped by that
observer into three main types, each of which owes its colour to
multiple films separated by air, viz. : —

1. The Urania type, in which the colour-producing lamellae are

Fig. 64. Scale of Morpho type (schematic). A, longitudinal vanes,
consisting of inclined multiple films, the seat of the colour.
B, basal layer. (After Mason.)

parallel to the plane of the scale and may be overlaid by a ribbed
or meshed structure (Fig. 63). The multiple films may reside in
either the upper or the lower lamina of the scale ; their plane is
that of the scale itself, and their optical properties are consistent
with that of a system of multiple thin films. The superposition of
a ribbed (vaned) or meshed structure on the outer surface of the
upper lamella of the scale modifies the visibility of the film colours,
but does not prevent the recognition of their typical properties.
This type was recognised by Onslow, prior to Mason, and these
two authorities are in agreement as to its chief features.

178 THE FUNDAMENTAL ASPECT OF COLORATION

2. The Morpho type, in which the lamellae are situated in the
vanes (or ribs) and are inclined towards the base of the scale
(Fig. 64). This type of structure has been demonstrated very
convincingly by Suffert. A brilliant highly metallic blue is the
prevailing colour, and, according to the last-mentioned observer,
it is connected with the peculiar reflecting properties of the
individual ribs, since it shows its full intensity and lustre on an
isolated rib or on ribs which have become laterally twisted, and

Fig. 65. Scale of Entimus type (schematic). A, multiple thin films,
the seat of the colour. B, cuticle enclosing lamellar structure.
(After Mason.)

consequentl}^ does not depend on their arrangement or proximity
to other ribs on the scale. Onslow, who paid a large amount of
attention to scales of this type, concludes that in the arrangement
of the ribs they show a regular periodic structure of the correct
theoretical magnitude ; the thickness of the ribs and of the air-
films between them varies from one to three half wavelengths. In
other words, he claimed that the ribs and the air layers between
them behaved as multiple films placed on edge. He did not
apparently recognise the lamellar structure of the individual vanes
or that a single rib is capable of giving the full colour. Mason, who

INTERFERENCE COLOURS 179

is in agreement with Suffert, consequently does not accept Onslow's
explanation, and points out that the ribs, seen flatwise, produce
no colour, although in this position they should give maximum
brilliancy according to Onslow's theory.

3. The Entimus type (Fig. 65), where the lamellae fill the
interior of scale, and are inclined in different directions in sharply
defined areas, with the result that corresponding colour patches
are exhibited. This type is prevalent in the Diamond beetles
Entimus and Cyphus, and previous observers, including
Biedermann and Mallock, regarded the colour as being due to
the presence of thin films. Michelson (1911) concluded, on the
other hand, that since the colours exhibited by these scales are
so varied and bright, and change so rapidly with varying incidence,
the effect must be due to diffraction caused by a grating-like
structure on the interior surface of the scale. He calculated that
there were from 5,000 to 10,000 striae per centimetre, a figure which
agreed with actual counts made under a magnification of
1,000 diameters and oblique reflected illumination. Under these
conditions the striae appear clearly defined in small patches. It
was necessary, in order to satisfy his theory, to postulate that the
grating structure is of an asymmetric saw-tooth type, since all
the light is concentrated in one spectrum. Onslow comments
upon the fact that Michelson's explanation of the vivid colouring
of Entimus is supported by the similarity between these scales
and certain artificial saw-toothed gratings ; nevertheless, he does
not wholly accept the diffraction theory. He finds evidences of
stratification clearly displayed in some of his microscopical
sections in addition to the vertical striae already alluded to. He
then proceeds to discuss certain objections to Michelson's views,
and concludes that, although diffraction may participate in
producing the colour phenomenon observed, it is not the sole cause,
and interference by thin films also plays a part. Mason holds
that the multiple film theory alone affords the true explanation,
and for the theoretical and experimental evidence upon which his
conclusion is based the reader is referred to his original paper
(1927a).

Scattering of Light by Minute Particles. This familiar

180 THE FUNDAMENTAL ASPECT OF COLORATION

phenomenon, so well exhibited in the blue of the sky, and in other
common features of Nature, was first explained by Tindall in 1869,
when he showed that finely divided particles of a transparent
substance possess the property of scattering light. If white light
penetrates a system composed of minute particles, immersed in a
medium of different refractive index to their own, the scattered
light will be blue, provided the diameter of such particles is small
as compared with the wavelengths of light. Since such a system
promotes the scattering of the short waves, the light is blue, while
the longer red waves pass unhindered.

A number of investigators have regarded that blue produced
in this manner contributes to the phenomenon of iridescence,
especially in cases unexplainable by any of the more usual laws
of optics. Recent research, however, has revealed very few
cases that can be interpreted on this basis, and most blues are
attributable to other causes. According to Mason (1926), the
bluish " bloom " or efflorescence present on the bodies of certain
dragon-flies, and less frequently on the wings, consists of particles
in a fine state of subdivision. When wetted with a liquid of
refractive index near 1-5 the particles become invisible and the
dark integument, which they overlie, is revealed. Upon drying,
the original appearance is restored. In the few cases investigated
by him it was found that the bluish " powder " or efflorescence
satisfied in a general way the main criteria required when light is
scattered in the manner indicated.

Selective Reflection. Onslow devoted considerable attention to
the highly metallic colours exhibited by many scaleless beetles
and other insects. He was concerned with the question as to
whether the colour is a surface effect or one due to interference.
In his experiments with certain highly metallic beetles he ascer-
tained that the colours do not disappear under pressure : on
immersion in highly refractive fluids the reflected colours still
persist : the transmitted colours, so far as they can be seen in
the thin surface layer, also persist : the colour-producing layer
appears to be so near the surface that no room is available for a
suitable structure that might produce the colours observed : the
colour changes, on reducing the thickness of the metallic layer by

SELECTIVE REFLECTION 181

polishing an elytron with carborundum powder, are very abrupt
and, in other respects, quite different from the colours of thin
films. Onslow points out that single films have been shown to
be very bright, but if the metallic colours of the insects in question
are due to such films, it is essential to suppose that they are of an
extraordinarily uniform thickness throughout, or the colour
would not be the same all over. Moreover, the thickness in all
individuals of a species must be the same, for the difference of a
fraction of a wavelength would cause a considerable alteration
of colour, and it appears very difficult to conceive of an organic
film complying with these requirements. He concludes that a
layer of organic material, which acts like a film of a dye or of a
metal, and gives rise to the metallic reflection, is a more con-
ceivable means of colour production. He is, however, faced
with the difficulty of not knowing how the surface layer is formed,
but concludes that it is difficult to imagine how a periodic structure
can arise in a homogeneous medium, about 0-5 /a in thickness,
during its extrusion to the exterior.

Mason (1927b) further discusses the subject of smooth iridescent
integuments of the kind just referred to, and recognises the
" metallic type," common in many insects, and the " enameUed
type," seen in certain Cetoniid beetles. He comments upon the
fact that no selectively reflecting material is known which does
not absorb light very strongly and show any intense colour in very
thin layers. The brilliant colour and lustre of specimens bleached
to a pale straw colour, when viewed by transmitted light, is utterly
at variance with the properties of substances which exhibit
selective reflection. The idea that a small number of thin films
is responsible for the observed effects is not inconsistent with the
thickness of the material concerned, since he finds that in most
specimens the colour layer exceeds^ 1 /x in thickness. As with
iridescent wing membranes, he concludes that lack of sufficient
differences between the refractive indices of the presumed lamellae
renders it hopeless to expect to obtain their resolution under the
microscope.

Notwithstanding their marked differences in lustre. Mason con-
cludes that " enamel-like " integuments similarly are explainable

182 THE FUNDAMENTAL ASPECT OF COLORATION

on the multiple film theory, only the film layer is thicker in these
cases. He remarks that the " colours and their changes, with angle
of incidence, pressure, swelling or ' faults,' the complementary
character of the transmission colours, and their low saturation,
the loss of colour on penetration, after treatment with HNO3, the
persistence after bleaching or in different light — all these criteria
are fully satisfied and they could not be met by any other known
type of structural colour." The existence of minute rod-like
structures, which penetrate the cuticula perpendicular to the
surface, modifies the properties of the multiple film layer : they
increase the lustre and may impart a bluish-white appearance
owing apparently to the scattering of light by the ends of the
rods (about 0-3 ju, in diameter) which function as individual
particles. Mason's conclusions are in accord with those of
Biedermann before him, but it may be said that neither the theory
of selective reflection, nor that of multiple films, can be regarded
as a fully established explanation of iridescent integuments. The
problem, so briefly outlined here, awaits further evidence for its
solution. Mason's contention that the existence of multiple
films has to be tested by optical methods rather than by micro-
scopical examination cannot be controverted. Nevertheless, it
must be remembered that microscopical evidence of the existence
of a given theoretical type of structure is of extreme importance.
No very large amount of study has been directed to the histological
aspects of the problem by way of applying refined methods of
technique. In this respect the work of Suffert is a definite
advance upon that of others, and he was able to show that definite
stratification occurs in the scale-wall in species of Urania, Papilio
and other genera, thereby satisfying theoretical requirements.
He was also the first to make the important discovery that in
scales of the Morpho type the ribs themselves actually show a
stratified composition — a fact brought out by observation upon
broken or split scale-ribs and by the use of polarised light. Not-
withstanding Mason's belief to the contrary, some indication of
lamination, if it exists, in iridescent insect integuments should
prove capable of detection by the application of suitable technique
to specially favourable types.

PIGMENTARY COLOURS 183

Pigmentary Colours

Pigmentary colours in insects are of a diverse character and
belong to very different groups exhibiting differences of origin and
of chemical constitution. It is not intended here to enter into
a detailed discussion of their nature and properties from the
biochemical standpoint, but to refer to advances in knowledge of
the more important pigments, and mainly from their biological
aspect. For this purpose it is convenient to group the pigments
of insects into four categories : (1) chlorophyll and other derived
pigments ; (2) haemoglobin and allied pigments ; (3) pigments
of protein origin ; and (4) pigments with purine bases. This
grouping does not include all insect pigments, and certain others,
either lesser known or of more restricted occurrence, have been
omitted. The general subject of pigments in animals, treated
both chemically and biologically, is well summarised by Verne
(1926). The reader is referred to this handbook for an elementary
treatise on the problems concerned and for a bibliography of their
extensive literature.

1. Chlorophyll and other Derived Pigments. Included under
this heading are those pigments which are absorbed from the food
without undergoing very marked changes in composition. They
are plant pigments which belong chemically to. two very different
groups. They include the carotins, anthocyanins and fiavones,
which have no nitrogen in their chromogenic nucleus, and the
chlorophylls which have a tetrapyrollic nucleus.

The first impetus given to the study of the fundamental nature
of insect pigments is due to Poulton (1893), whose experiments on
the colours of Lepidopterous larvae are familiar to entomologists.
His work suggests that a modified chlorophyll, derived from
plant chlorophyll, and xanthophyll occur in the blood and
integument of caterpillars, and form the basis of their prevailing
type of coloration. Spectroscopic examination of the blood, when
compared with that of an extract of the colouring matter of the
food-plant, showed an extremely close similarity in the two cases.
Considering the chemical processes to which these pigments are
subjected in the passage through the walls of the digestive canal

184 THE FUNDAMENTAL ASPECT OF COLORATION

to become incorporated with the protein constituents of the blood,
the closeness of the spectral resemblances is specially remarkable.
It is noteworthy that certain green caterpillars become colourless
when fed upon parts of plants devoid of chlorophyll, but become
green when access to green leaves is established.

In recent years great advances have been made with regard to
our knowledge of the composition of chlorophyll, more especially
through the researches of Willstatter and his co-workers. The
green pigment or true chlorophyll consists of two components,
both of which exhibit in solution a notable red fluorescence.
The one, chlorophyll a, forms about 72 per cent, of the mixture,
and is blue-green in alcoholic solution when viewed with trans-
mitted light. The other, chlorophyll b, forms the remaining
28 per cent. ; it is yellow-green when viewed by transmitted light,
and differs from chlorophyll a in having two atoms less of
hydrogen and an extra atom of oxygen. The yellow components
of chlorophyll comprise an orange-yellow pigment known as
carotin, which is a highly unsaturated hydrocarbon, and a second
yellow pigment, xanthophyll, which is its dioxide.

Recent researches lend support to Poulton's main conclusions
that many caterpillars are coloured by pigments derived from
their food, which are absorbed into the blood without undergoing
marked changes in their character. Whether the yellow pigment
referred to by Poulton under the general term xanthophyll is
true xanthophyll, or carotin, appears to be uncertain, and Palmer
(1922) advances reasons for suggesting that the chief carotinoid
pigment of Lepidopterous larvae may be carotin. This substance
is widely distributed among insects and is known to figure in the
coloration of beetles of the families Coccinellidae and Chrysomelidse
and in red-bugs or Pyrrhocoridae. Recent research has afforded
further evidence of its presence in insects. The work of Palmer
and Knight (1924a) indicates that the yellow and red coloration of
the Pentatomid bug Perillus bioculatus is hypodermal, and largely
due to the presence of secondarily absorbed carotin. This insect
is predaceous in habit, and the favourite food of the adults and
nymphs consists of the eggs and larvae of the potato beetle
(Leptinotarsa decemlineata) as well as the beetle itself. They

PIGMENTARY COLOURS 185

were able to demonstrate that in the larva of the latter insect
the blood is coloured deep orange by carotin derived from the
potato plant. In the fresh blood of the beetle this pigment
occurs in a concentration amounting to 0-0136 per cent., which is
as high as that found in fresh green leaves. It is remarkable that
no xanthophyll could be detected in the blood, since there can be
scarcely any doubt that the green leaves of the potato plant are
as well endowed with this pigment as other plants. It appears,
therefore, that the Perillus, in the process of imbibing the blood
of its prey, obtains the carotin which so largely contributes to its
characteristic coloration. In both this bug and in its prey the
yellow and red colours of the integument are due to differences in
the degree of concentration of the same pigment. Knight (1924)
has shown that temperature, by influencing the physiological
processes during the active life of Perillus, indirectly affects the
coloration of that insect. When reared at a temperature of
85° to 95° F. white forms are produced, while at temperatures
below 78° F. the red and yellow forms appear. When subject to
the higher temperatures carotin is not deposited in the hypo-
dermis, and the insects void chalky white excreta. Under these
circumstances it appears that the carotin becomes oxidised, a
fact which Palmer and others had previously noted to take place
in animals having a high rate of metabolism. If white examples
of Perillus be subjected to a relatively cool environment (70° to
75° F.) they discharge normal excreta — dark reddish, brown or
black in colour, and at the same time commence to accumulate
carotin in the hypodermal cells. Humidity, it appears, does not
behave as an individual factor in the process, since it does not
exercise notable effect upon the coloration of the insect. A
relative humidity of 100 per cent, probably inhibits metabolism
in various ways, particularly by slowjng down physical activity,
and in this way a greater deposition of carotin might be expected.

Evidence that xanthophyll plays a considerable part in insect
coloration is by no means established, and experiments by feeding
caterpillars on yellow maize, or other plants known to be rich in
this pigment, are greatly to be desired.

In 1921 Gerould described a larval colour mutation in the Pierid

186 THE FUNDAMENTAL ASPECT OF COLORATION

butterfly, Colias philodice, in which the usual yellow-green
coloration was replaced by blue-green. He concluded that some
factor was involved which caused the yellow components of
chlorophyll to become broken down or decolorised. By acting
locally from the nuclei of the cells of the intestinal epithelium it
changed the crude chlorophyll during its absorption into the blood.
The result has been that a pigment derived from the blue-green
component of chlorophyll, viz., chlorophyll a, is left intact and
becomes the persisting colouring material. The eggs of imagines
derived from blue-green caterpillars were pure alabaster-white,
not the normal cream-white, an effect due to the absence of
yellow pigment in the blood of the parent. The pupa is only
slightly less blue-green than the larva ; the exuviae are white
instead of the normal yellow. The wing colours of the adult, being
due to uric acid derivatives, are not affected, but the eye colour is
of a bluish-green rather than the normal apple green. A further
point of great interest is that the cocoon colour of its parasite,
Apanteles flaviconchce, is white instead of the usual yellow, and it
thus appears that the lack of yellow pigment in the host affects
the colour of the salivary secretion of its parasite. Gerould's work
suggests, in the light of Poulton's observations, that the blue-
green coloration in C philodice is derived from the clover upon
which the larva feeds, but actual proof by way of chemical or
spectroscopic examination remains to be established.

Przibram (1913) has contested the existence of chlorophyll in
insects, and has classified under the non-committal name of
" Tiergrun " a variety of green animal pigments. Spectroscopic
examination alone he claims is insufficient to establish their
identity, and chemical methods are necessary. An ether solution
of the green pigment in grasshoppers and Cantharis, when heated
with an equal quantity of alcohol solution of potassium hydroxide,
yields a yellow precipitate and the solution becomes clear. An
extract of plant pigment, on the other hand, remains green and
turbid. Other striking differences appear upon heating and
adding more of the hydroxide, while concentrated sulphuric
acid, when added to the ether solution, and also nitric acid, give
very different reactions when compared with their effects upon

PIGMENTARY COLOURS 187

plant chlorophyll. His argument does not appear very conclusive,
because insect chlorophyll has never been claimed to be absolutely
identical with plant chlorophyll. Furthermore, he apparently
did not investigate the pigments of Lepidopterous larvae. Gerould
(1927) criticises Przibram's conclusions and points out that the
reactions described are such as one would expect in cases where
the pigments are associated with proteins, as in the blood of
Lepidopterous larvae.

Faure (1932) has shown that in the solitary phase of the African
Migratory Locust (Locusta migratoria subsp. migratorioides R.
and F.) green individuals are of frequent occurrence. This same
feature has also been observed by Hertz and Imms (1937) : the
green coloration only develops in the presence of a relatively
high humidity and is independent of the coloration of the
immediate environment and of food. Faure finds that the green
pigment involved betrays no close affinity with chlorophyll. It
shows, for example, a single absorption band with its centre at
6,700 A. and there is no trace of the red fluorescence characteristic
of chlorophyll solutions. In the stick insect Carausius (Dixippus)
and many caterpillars the green pigment is also apparently
synthesised by the insect itself and is developed independently
of the nature of the food (Giersberg, 1928 ; Meyer, 1930). Accord-
ing to Meyer it is an oxidation product of protein metabolism. It
is evident, therefore, that due precaution must be exercised before
concluding that the green colour of many insects is a chlorophyll
derivative.

Another group of plant pigments, namely, the anthocyanins, has
recently been shown to be present in insects. According to
Hollande (1923) certain insects absorb pigments of this nature
from their food which become concentrated in the albumen granules
of the fat-body. Thus, the larva of the beetle Clonus olens, which
feeds upon the purple staminal hairs of Verbascum nigrum, con-
tains an anthocyanin derived from that plant ; the pigments
remain purple or become coloured either blue or red according to
the prevailing degree of alkalinity or acidity present. A blue
pigment of a similar nature was also found by him in larvae of the
saw-fly Athalia spinarum. In the bright vermilion-coloured aphid

188 THE FUNDAMENTAL ASPECT OF COLORATION

Tritogenaphis rudheckice Palmer and Knight (1924b) found that
the chief red pigment present is not soluble in fat solvents, but
is readily extracted by water or cold methyl-alcohol, yielding
vermilion-coloured solutions. The properties of this pigment,
as determined by them, are strongly suggestive of its being an
anthocyanin. Small quantities of carotin were also detected,
but were insufficient to have any real influence upon the general
coloration of the insect. It is noteworthy that Glaser (1917) had
previously shown that another aphid, viz., Pterocomma smithice,
which feeds upon willow, contains a red pigment which is
apparently located in the fat-body. From preliminary tests
Glaser concluded that the colouring matter appeared to be of the
nature of an anthocyanin. Since, however, no anthocyanin could
be detected in the shoots of the food-plant, he suggested that the
aphid absorbs hydroxyflavones along with the sap, and that these
become reduced to anthocyanin in the body of the aphid, and
ultimately converted into the red pigment.

Pigments of the flavone group have also been recently detected
in insects, and, according to Palmer and Knight (1924b), a red
colouring material obtained from the Coreid bug Leptocoris
trivitatus is of this nature. They also obtained a similar pigment
from other Hemiptera belonging to the families Capsidae, Lygaeidse
and Reduviidae. As to the origin of the flavone-like pigment in
these cases, it is presumed that it is derived from the food-plants,
but they did not pursue this subject further. In the Reduviidae,
which are predators, it would appear that the pigment would either
have to be derived indirectly through some plant-feeding prey, or
be synthesised by the insects themselves. More recently Thompson
(1926) has provided evidence that insect pigments of the nature
of flavones or flavonols are derived from the food-plant. In the
Marbled White butterfly (Melanargia galatea) a small amount of
a yellowish pigment is present in the wings, and gives no reaction
or spectrum, which supports the assumption that it is a carotinoid.
On the other hand, its properties ally it with a flavone. He was
able to show further that an identical pigment is present, both
free and in glucosidic combination, in cocksfoot grass {Dactylis
glomerata) which is one of the known food-plants of the larva of

HJEMOGLOBIN 189

this butterfly. It is probable that pigments of the flavone type
are widely spread among insects, but the subject has been very
little studied.

2. Haemoglobin and Allied Pigments. The chlorophylls (sensu
str.) are pigments with a tetrapyrollic nucleus and whose
integral metallic constituent is magnesium. Recent research
has demonstrated the chemical affinities between chlorophyll
and haemoglobin, each having as a base the substance termed
jDorphyrin, composed of four pyrrol groups in complex linkage.
It is possible, therefore, that haemoglobin is ultimately derivable
from chlorophyll, and that in the processes involved in animal
evolution the latter pigment has become profoundly transformed
so as to assume new functions, and its metallic constituent
magnesium has become replaced by iron {vide Fulton, 1922).
Among insects there are three types of pigments of this character
whose metallic element is iron, viz., haemoglobin, cytochrome and
certain red pigments of Lepidoptera.

Haemoglobin is of rare occurrence among insects and it is only
in certain Chironomid larvae that it plays any part in coloration.
Among these insects it is well known to be present in the blood
plasma, and not in the blood-cells, and since Chironomid larvae
possess a transparent integument it imparts to them their char-
acteristic red appearance. Haemoglobin also occurs under localised
conditions in a few other insects, notably in the peculiar tracheal
end-cells of the larvae of Gastropkilus ; Hungerford (1922) has
also shown that it is found in the Notonectid Buenoa, and Mrs.
Brindley has recently (1929) recorded its presence in the accessory
genital gland of the male of Macrocorixa geoffroyi. The important
work of Keilin (1925) on the intracellular respiratory pigment
cytochrome, which is apparently of general occurrence in the
animal kingdom, shows that in the possession of this pigment
insects already have the chief constituents from which haemoglobin
may arise. The existence of haemoglobin, therefore, in the isolated
instances just enumerated does not necessarily imply any pro-
found physiological difference between them and other insects
devoid of this pigment. Since cytochrome plays no direct part in
insect coloration it does require further consideration here.

190 THE FUNDAMENTAL ASPECT OF COLORATION

The red and yellowish hypodermal pigment present in the
wing-scales of certain butterflies, particularly Vanessa io and
V. urticce, plays a direct part in the coloration of those insects.
The researches of the Grafin von Linden (1905) indicate that it
is derived from the chlorophyll absorbed during larval life.
The transformation commences in the epithelial cells lining the
digestive canal, which contain at first green granules derived from
the chlorophyll of the food. So long as the digestive juices of
the larva remain alkaline little change supervenes, but when
transformation to the pupa is taking place the gut contents
become acid and the colouring matter changes to red. During the
process of change the red colour results from the combination of
the actual pigmentary material with an albuminoid substance. The
final colour assumed by the pigment in the hypodermis depends
essentially upon its degree of oxidation ; in its reduced condition
it is red-carmine, while oxidation transforms it into a dull greenish-
yellow. The pigment is soluble in hot or cold water, in neutral
salts and mineral acids, but is insoluble in organic solvents. By
the nature of its crystals, its spectrum and other properties it
shows itself closely allied to bilirubin. On the other hand, it
resembles haemoglobin in that it represents the combination of an
iron-containing pigment with an albuminoid substance, and in the
facility with which it forms unstable combinations with oxygen.
Hollande (1923) maintains that the origin of the red pigment
from chlorophyll is insufficiently established, and claims that its
formation is practically the reverse of that described, in that it is
primarily hypodermal in origin. When pupation is about to occur
disintegration of the larval hypodermis ensues, and the red
granules which it contains become liberated into the blood and
are then absorbed by the digestive canal. Any excess of this
material passes into the gut cavity and is ultimately excreted to
the exterior. One is inclined to agree with Verne (1926) that the
processes described by the Grafin are better substantiated than
the more hypothetical origin explained by Hollande.

3. Pigments of Protein Origin. The pigments coming under this
category are those of the melanin group. Knowledge of their
chemistry is still very restricted, but they have been shown to

MELANIN 191

be derivable from certain amino-acids, which result as the out-
come of katabolic changes in proteins. True melanin is produced
by the oxidation of an amino-acid, functioning as a colourless
chromogen, through the action of an enzyme. It is well known that
tyrosin is converted into melanin by the ferment tyrosinase and,
more recently, it has been shown that "dopa " (dioxyphenylalanine)
is also a source of this same pigment, under the influence of
tyrosinase or of dopaoxidase. Whether these two enzymes are
actually distinct appears to be doubtful, and opinions differ on the
subject.

Melanin is probably a very widely spread pigment among
insects, and the existence of tyrosinase in the blood, or other
tissues, has been detected by a number of investigators. This bare
fact, however, without evidence of the presence of a chromogen
susceptible to its influence, has only a partial bearing upon the
problem of melanisation. Evidence of the presence of tyrosin
has been found by Van Fiirth and Schneider in the blood of
Lepidoptera ; by Gessard in the larva of Lucilia ; and by Gortner
in the larva of Tenehrio. The findings of these earUer observers
have been confirmed by Onslow (1916), who has shown that the
black markings on the wings of Pieris hrassicce are due to the
presence of this pigment. By grinding pupae of this species,
before the black markings began to appear on the wings, and
extracting the haemolymph, it was shown that the latter blackened
at the surface when in contact with the air. When the immature
fore wings were dissected off the insects, and immersed in the
haemolymph, they became blackened, but more so at the seat of
the usual markings. A dense blackening effect was produced
all over when the wings were treated with a solution of tyrosin,
but when placed in tyrosinase solution the blacking was confined
to the actual markings. These experiments show that, although
tyrosinase is generally distributed, the black markings and their
position are determined by the localisation of the tyrosin in
these areas, and that the process of natural melanisation occurs
as soon as atmospheric oxygen has access to the developing
wings.

Subsequent researches by Hasebroek, published in numerous

192 THE FUNDAMENTAL ASPECT OF COLORATION

papers (1921, 1922, 1925-26), and by Schmalfuss and his col-
laborators (1930, 1933) indicate that the chromogen in some insects
is " dopa " (dioxyphenylalanine), as in the elytra of Melolontha or
chater beetles, or dioxyphenylactic acid, as the mealworm
(Tenebrio). The processes by which the antecedents of melanin
are produced are unknown, but it has been suggested that the
production of this black pigment involves means for the disposal
of phenolic compounds which apparently are necessary for its
elaboration. The amount of melanin produced in an insect has
been shown to be modifiable by changes of temperature. This
has been demonstrated in the Ichneumonoid Habrobracon and
the effects, it is claimed, may be transmitted up to the second
generation of the offsj^ring {vide Schlottke, 1926 ; Kaestner, 1931).
Schmalfuss finds that an enzyme which, in the presence of oxygen,
converts dopa into melanin occurs in insects pertaining to five
different orders. Hasebroek claims that the chromogen involved
in certain black and brown pigmentations of the wings of Vanessa
antiopa is dopa. Melanising ferments appear to be generally
distributed among Lepidoptera, both in melanic and non-melanic
species and, according to Hasebroek, it is the chromogen that
is absent in non-melanic forms. Whether there is more than
one enzyme, e.g., tyrosinase, involved in melanin production is
extremely uncertain. Hasebroek (1921, 1922) maintains that
a dopaoxidase exists in certain Lepidoptera at a stage before
tyrosinase betrays itself. Thus, he states that the haemolymph of
larvae and of extracts of the eggs blacken when treated with dopa,
and not with tyrosin ; at a later stage, haemolymph from the
pupae and imagines exhibit melanin formation when treated either
with dopa or tyrosin. According to Przibram (1922), melanin may
appear in the cocoons of the moths Eriogaster lanestris and Saturnia
pavonia without the intervention of any enzyme. In these cases
the chromogen is dopa and he states that it differs from tyrosin
in that it produces melanin by direct oxidation in the presence of
moisture, provided the medium be alkaline. It may be said,
therefore, that the evidence of melanin production in insects,
although well established, is of an extremely conflicting nature.
The researches of the English biochemists, Onslow and Robinson,

URIC ACID 193

and Raper, render it probable that dopa is the initial product of
the action of tyrosinase upon tyrosin.

4. Pigments with Purine Bases. Hopkins, in his well-known
researches on the chemical nature of the pigments of the Pieridse
(1896), showed that the white and yellow colours in the wing
scales in these butterflies are due to uric acid and its derivatives.
Subsequent workers have confirmed this general conclusion and,
so far as is known, the Pieridae are the only family of butterflies
which have the property of depositing uric acid in the wings.
Wigglesworth (1924) has made a quantitative study of this sub-
stance in Pieris hrassicce and finds that part of the uric acid,
contained in the fat-body of the pupa, becomes deposited in the
wings just before emergence of the perfect insect. More of it
becomes transferred to the alimentary canal, from which it is
voided to the exterior. Estimations of the amount of uric acid
present in the wings of the two sexes showed a markedly greater
quantity in the male. There appears to be no actual constant
difference in the uric-acid content of the two sexes, and the excess
present in the male is due to the larger size of the wing scales,
and hence their greater capacity for storing this substance.
Wigglesworth further investigated the same problem in Vanessa
urticce, a member of the family Nymphalidae. In this species it
was found that the total amount of uric acid present is approxi-
mately the same as in P. hrassicce, but that during development
it is, in part, passed to the digestive canal, but not to the wings.
Schopf and Wieland (1926) have further studied the white wing
pigment in Pieris hrassicce and in P. napi and, while agreeing in
its similarities to uric acid, claim that it differs from that substance
in certain of its reactions. It consists according to them of a
special compound which they term leucopterine. The same
authors (Wieland and Schopf) also examined the yellow pigment
of another Pierid, viz., Gonejpteryx rhamni, and gave the name
xanthopterine to the uric acid derivative, of which it is composed.
Tliis name is apparently given to the same pign\ent which Hopkins
termed lepidotic acid.

Uric acid and its derivatives are the end products of purine
metabolism in insects, and are formed as the result of the decom-

R.A. ENTOMOLOGY. 7

194 THE FUNDAMENTAL ASPECT OF COLORATION

position of nucleic acids which occur with a protein group to
form nucleoprotein. Pigments, therefore, with purine bases are
excretory products which, in the Pieridae, become deposited in
the wings. With respect to the white colour of these insects, it
has already been pointed out that, according to Mason, it is due
to structural features rather than to the presence of urates.

Combination Colours

Combination colours are much more prevalent among insects
than colours solely due to structural features. In other words,
pigment very commonly co-exists with structures that are, in
themselves, colour-producing. The structural features involved
produce iridescence which is due to one or more of the four causes
already discussed. Whenever a pigment is suspected to occur in
combination with a structural colour its effect can be eliminated
by bleaching (preferably with hydrogen peroxide) ; similarly,
the absence of pigment can be demonstrated by the same means,
when the original colour will reappear unaltered after the specimen
has been dried. Iridescent colours are very often devoid of the
presence of all pigment excepting black or dark brown. The dark
pigment acts as a background or absorbing screen, thus inter-
cepting white light which would otherwise reach the eye. The
amount of desaturation structural colours undergo is largely
proportional to the extent to which the dark pigment is developed.
As a rule a background of this nature is necessary to the j^roduction
of the brilliant and intense structural blues and metallic greens.
The pigmentary substance may be deposited in the same scales
which bear the special structural features concerned, or in a layer
of scales beneath them ; it may occur either in the scale cavity or
diffused in the cuticular substance of the scale.

Onslow's researches have brought to light interesting examples
of combination colours among insects, besides other cases where
the causes of the colours observed are entirely obscure. In the
butterfly Morpho cypris, for example, there are brilliant iridescent
blue scales and also white ones, and the white scales only differ
in the entire absence of brown or black pigment. The absence
of this absorbing screen causes areas composed of these scales

FUNCTIONS OF PIGMENTS 195

to appear white and their iridescence to be greatly reduced.
The blue scales, it may be added, can be made to appear similar
to the white ones by treatment with bleaching agents. In
the white and in dark purple areas of the wings of the male of
Hypolimnas bolina the same phenomenon also presents itself.
Onslow also describes the means of colour production in the case
of the brilliant green seen in the male of Ornithoptera poseidon.
This he finds is due to a structural blue combined with a yellow
pigment diffused in the cuticular parts of the scales themselves.
In Teracolous phlegyas a pigmentary red in the scale- wall, combined
with a structural purple, produces a magenta. Also in Callitcera
esmeralda a dull reddish-brown pigment alone is present, but, in
such areas of the wings where it occurs, structural features of the
scales combined with it produce vivid magenta red spots. The
actual colour is probably largely structural, the pigment behaving
more especially as an absorptive layer which serves to enhance the
iridescence. A number of other examples of combination colours
will be found in the literature already quoted, and do not require
mention here.

Further Remarks on Insect Coloration

Much has been written regarding the relations of the colours of
insects to the environment in which those animals live. It is
often comparatively easy to give an explanation of the way in
which a particular colour or type of coloration may be useful to a
given species. Entomological literature abounds with empirical
deductions of this kind. Such explanations are largely concerned
with superficial appearances and do not penetrate very far into
the fundamental nature of the problems involved. The more
difficult subject of the function and significance of pigments, in the
physiological processes of the individual, tends to be relegated to
the background in such cases. Possibly, more often than not,
paucity of biochemical knowledge is one of the contributing
causes. Colour, it may be said, arises as a necessary result of the
complex chemico-physical constitution of the organism ; it is the
outward expression, or the indicator, of a chain of processes within.
It is primarily determined by physiological causes and it remains

7—2

196 THE FUNDAMENTAL ASPECT OF COLORATION

for the more ardent selectionists to prove that it is forced upon
the organism by the stress of competition.

(a) There is reason to beheve that carotinoid pigments, for
example, are widely spread among insects, and their absorption
occurs incidentally along with the food. Insects store up fat
in their adipose tissue, and carotin, being fat-soluble, becomes
deposited at the same time. Palmer and Knight's observations
have shown that, in the Reduviid bug Perillus, carotin under
ordinary conditions becomes stored in the integument in an
unoxidised state. If the temperature be raised sufficiently it
becomes entirely eliminated from the body, presumably after
oxidation. We can safely say that many caterpillars, whose
digestive secretions contain no enzymes which destroy carotinoid
pigments, will assume a yellow coloration if they feed upon
parts of plants in which carotin is the predominating colouring
material.

Gerould, as has been previously mentioned, has described a
clear case of the inheritance of a factor which suppresses
carotinoid pigment in caterpillars of Colias philodice. The
recessive mutants in which this occurs are blue-green, whereas
the normal caterpillars are leaf-green, owing it would seem to an
admixture of carotinoid pigment. Gerould formulates the
conclusion that the nuclei of the intestinal epithelial cells, and
especially their chromosomes, produce an inheritable enzyme, or
recessive gene, capable of inhibiting or decolorising the carotinoid
pigments during the digestive processes. This mutant is therefore
essentially physico-chemical in nature and is apparently of negative
survival value. He mentions that an outdoor culture of larvae
comprising individuals of both colour types was exposed to the
attacks of sparrows for a period of twelve days. During that time
the conspicuous blue-green examples were almost all taken by
the birds, while many of the grass-green normally coloured larvai
were unmolested.

(b) With regard to the coloration of Lepidopterous larvae by
means of derived chlorophyll, the most significant fact is the
absence of any enzyme or other chemical constituent destructive
to that pigment. We have no evidence whether chlorophyll

FUNCTIONS OF PIGMENTS 197

performs any direct physiological function or not in the life of the
individual. It has been suggested by Gerould that it may play
some part in the elimination of CO 2 from the blood, but there is no
data either for or against this idea. It may well be, as appears to
be the case with carotin, that its presence is incidental and non-
injurious, and consequently much of it, instead of being eliminated
from the system, accumulates within the tissues. It occurs in
larvae which feed exposed during daylight upon plants, as well as
in night-feeders which lie so well concealed during the day that
they are hard to discover.

There are many leaf-mining dipterous larvae which feed upon
chlorophyll-bearing tissue and yet do not assume a green
coloration. In these cases the explanation may be found in the
biochemical nature of the secretion of the mid-intestine proving
to be destructive to that pigment. There is some evidence that
certain anthocyanins, derived from plants, may pass through the
walls of the digestive canal in a condition sufficiently unmodified
to impart their characteristic colour to specific insects. This has
been demonstrated by Hollande in the case of larvae of the beetle
Clonus olens, which feed upon the purple staminal hairs of
Verhascum. The presence or absence of derived pigments in
different insects is a biochemical indicator of certain features in
the constitution of these animals. The important role which such
pigments play in the ecology of the organisms would, therefore,
ap2:)ear to be of a secondary nature.

(c) It is well known that waste products, such as uric acid
and its derivatives, become deposited in the wings of Pierine
butterflies. Why this should be the case has not been solved,
and in the allied family Nymphalidae no such transfer takes
place. Evidence of a similar phenomenon outside the Pieridae is
very uncertain. We know that comparatively slight biochemical
differences in the constitution of these substances produce very
marked colour differences among different species.

(d) Much has been written with respect to melanism in insects,
mainly Lepidoptera. It has long been customary to refer to
all and sundry specially dark or black forms of species of this
order as being melanic. The term has been applied without

198 THE FUNDAMENTAL ASPECT OF COLORATION

reference as to whether true melanism is involved or not, but
from what is known of the occurrence of melanin in insects as a
group, it is extremely probable that the presence of this pigment
accounts for the phenomena just alluded to. Frequent reference
occurs in literature to the prevalence of melanic forms of
Lepidoptera in the neighbourhood of manufacturing towns and
urban areas. They have appeared in species among which they
tvere previously unknown, and the melanisation has in some
cases increased in prevalence. On the other hand, it should be
mentioned that melanic forms also occur in areas far removed
from the atmospheric conditions of industrial centres.

The incidence of melanism in several species appears in some
way to be bound up with the industrialisation of certain areas of
England. This suggests that the presence of metallic salts
contaminating the food-plants in such districts might not only
affect the germ-plasm, so as to produce an inherited melanism,
but also might stimulate it when once developed to attain its
maximum expression. It was on premises of this kind that
Harrison and Garrett (1926) and Harrison (1928) conducted a
series of experiments with certain Geometrid moths, using strains
from localities where melanism in the species involved is unknown.
In these exiDcriments the larval food was artificially charged with
lead nitrate, manganese sulphate or manganese chloride in the
different series. The most conclusive results apply to Selenia
bilunaria, material of which was obtained from both England and
Germany. A certain proportion of the progeny of these moths
proved to be melanic forms, of a type stated to be unknown in
Nature, and which did not appear in the controls. It was further
shown that the melanism, thus apparently induced, was inherited
as a mendelian recessive, the insects being reared upon untreated
food. The experiments were carried out on three different strains
of the insect, with the application of three different metallic
compounds, and melanic forms appeared in each of the series.
Results comparable with those of Harrison have not been obtained
by McKenny Hughes (1932), who similarly attempted to induce
melanism in Selenia bilunaria. Larvse were fed upon hawthorn
treated with lead nitrate and with manganese sulphate, and six

MELANISM 199

generations, numbering 3,265 moths, were reared. No melanic
examples appeared during the course of these experiments, either
in the case of individuals fed on the treated food or in the controls.
It would seem, therefore, that Harrison's explanation of the
appearance of melanics in his cultures as the result of the treat-
ment to which the larvae were subjected has not been conclusively
demonstrated. The reader is also referred to the paper by
Thomsen and Lemche (1933).

Hasebroek (1926) has studied induced melanism in industrial
areas, and arrived at the conclusion that it results through the
absorption of volatile substances via the tracheal system. He
discountenances the influence of compounds absorbed through
the alimentary canal. By exposing pupae of various species of
Lepidoptera to the influence of the vapour of pyridene,
sulphuretted hydrogen, both separately and in mixture, he was
able to produce a considerable degree of melanism. Also sewer
gas alone, ammonia, pyridin and chloroform in combination
caused melanic tendencies : methane and ammonia also produced
pronounced darkening. While his experiments are suggestive,
and of great interest, they are open to considerable criticism.
In the first place, he does not appear to have used stocks known
to be free from melanic tendencies, and, consequently, the
possibility that the melanism may have arisen independently of
the treatment cannot be wholly excluded. Secondly, most of the
species employed do not appear, in England, at any rate, to
exhibit any marked melanic tendency in industrial areas, so it
may be argued that his experiments do not necessarily bear upon
this local problem. Thirdly, no mention is made of any breeding
trials being carried out with a view to ascertaining whether the
induced melanism is inheritable.

{e) Most entomologists are familiar with the results of Poulton's
experiments (1887, 1892) with reference to the reactions of certain
colour-sensitive Lepidopterous larvae and pupae. It will be recalled
that with regard to Vanessa urticce and V. io, and also Pieris
hrassicce and P. rapw, he showed that when their larvae take up
their final resting position upon surfaces, preparatory to spinning
their silken attachment, they are in a condition highly sensitive

200 THE FUNDAMENTAL ASPECT OF COLORATION

to the differential effects of light rays. This sensibility is influenced
by certain constituents of diffused light reflected from surfaces in
the immediate environment, which affects the colour assumed by
the pupae. On black, deep red, blue and certain green backgrounds,
for example, darkly coloured pupse resulted. On the other hand,
orange, yellow or white backgrounds gave rise to green, or pale,
bright pupse, and the same effects were produced by surroundings
of green leaves and shoots. Two pigments, a green and a black,
bccur in the pupae, and it appears that since rays from the orange
and yellow zone of the spectrum are inimical to the superficial
black pigment, they conduce to the production of green or pale
pupae. It seems tolerably certain, therefore, that those same rays,
reflected from green surroundings in Nature, stimulate the pupae to
assume a green appearance. The blue rays, on the other hand,
tend to produce dark pupae. The idea that the light may act
through the larval eyes led Poulton to cover the latter with black
varnish, but no resultant influence upon pupal coloration was
observed, and it would appear that light produces its effects via
the general integument. These extensive researches, so briefly
outlined, have stimulated others to experiment along simiilar lines,
and their results support the general conclusion that a power of
colour adjustment, in relation to their immediate surroundings,
exists in certain Lepidoptera.

While it is generally agreed that the black superficial pigment
is melanin, the nature of the green colouring matter is prob-
lematical. Whether an enzyme promotes its formation as
suggested by Fraulein Brecher, or it is a modified derivative of
chlorophyll, is still unsettled. Experiments conducted by
Brecher (1921) with blood of larvae of Pieris hrassicce, in vitro,
appear to show that, during the sensitive period, yellow rays
increased the acidity of the blood and reduced the activity of
tyrosinase, while ultra-violet rays had the opposite effect. Her
more recent researches (1925) do not confirm this conclusion when
applied to an examination of the blood of both larvae and pupae
living under the influence of different light rays. When the
blood was drawn directly, its ^H reaction was found to be
remarkably constant (6-50 to 6-77). On the other hand, Fraulein

COLORATION IN PIERIS 201

Brecher advances evidence which suggests that the photochemical
action of black, and other dark surfaces, depends upon the ultra-
violet rays which they reflect stimulating melanin formation, while
infra-red rays tend to inhibit the development of that pigment.
She also found that tyrosinase was less active in the blood of green
pupae than that contained in the blood of darkly coloured pupae.
Her researches present a line of approach towards the fundamental
nature of the phenomena observed, but further investigation is
obviously needed, including the influence of temperature as well
as light rays.

Przibram (1922), by electrical cauterisation of the eyes of the
larvae, and also by decapitation, came to the conclusion that the
action of the light rays takes place through the eyes, since he
obtained the same results as in darkness, whatever the surround-
ings of the larvae were. Brecher (1924) confirmed these results
by painting the larval eyes with a transparent yellow varnish,
when the pupae became green in Pieris and golden in Vanessa ;
when blue varnish was used the pupae were darkly coloured, as
under blue surroundings. It would, therefore, appear that light
of different wavelengths exerts a differential stimulus through
the larval eyes on the nervous system, but the nature of the
phenomenon is unexplained.

Durken (1923) and Fraulein Brecher (1923) have further
experimented on the influence of orange light in inhibiting the
development of grey and black pupae in Pieris brassicce, thus
allowing the green pigment to become evident. Their results are
in conformity with those already quoted, and they further suggest
that the tendency to produce green pupae is an inheritable
property. Larvae descended from the orange glass cultures,
when allowed to pupate under ordinary light conditions, gave
rise to more green pupae than ocqurred among the controls.
Durken repeated the exposure to orange-coloured rays for two
generations, and when the larvae finally transformed under normal
non-coloured surroundings, almost all the pupae were green.
Experiments by Harrison (1928a) afford support to the conclusion
that the induced green coloration is inherited. In the species
Pieris napi a mixed lot of eggs was obtained from six captured

202 THE FUNDAMENTAL ASPECT OF COLORATION

females : forty-three control larvae reared in ordinary glazed cages
gave 20-9 per cent, green pupge, while forty-six larvae reared under
orange-coloured glass produced 93-4 per cent, green pupae. Larvae
derived from imagines emerging from these green pupae were, in
their turn, submitted to the same orange rays and, on transforming,
95-2 per cent, of the pupae were found to be green. The latter were
again used for rearing a third generation, but in this case the larvae
were reared under similar conditions to the controls. Only
thirty-one pupae were obtained, but all were green. The butter-
flies from these were segregated into two batches — early-emerged
and late-emerged. In the first batch their offspring were allowed
to pupate on a varied selection of surfaces, and 58 per cent, of
green pupae resulted (from a total of fifty individuals). The
offspring of the other batch of butterflies were reared in cages
whose glass walls were painted black, and 96-4 per cent., of the
forty-eight pupae which resulted, were green. The results are
presented as they stand, and Harrison does not attempt any
theoretical discussion of the factors involved.

(/) The phenomena of colour change in the stick insect Carausius
have attracted the attention of several recent workers in Germany,
notably Atzler (1930), Giersberg (1928) and Priebatsch (1933).
This insect is able to change its general coloration in response to
sudden changes in its immediate surroundings, becoming darker
or paler, as the case may be. Other stimuli, such as temperature
and mechanical pressure, elicit a similar response. This response
is brought about by movements of the pigment in the hypodermis.
The pigment granules either become aggregated or diffused,
according to whether the insect becomes paler or darker in its
general colour. The primary sensation of the background colour
is received by the eyes, which act upon the visual centre in the
brain, and apparently induce the secretion of a hormone into the
blood. It would appear, therefore, that it is this hormone which
activates the movements of the pigment granules. Destruction
of the tritocerebrum causes all colour response to cease. Carausius
is unique among insects in that its colour response seems to
operate in a manner similar to those animals which have true
chromatophores controlled by hormone secretion. It would.

LITERATURE 203

in fact, seem that in the stick insect the hypodermal cells are
virtually unspecialised chromatophores.

(g) Mottram and Cockayne (1920) and Cockayne (1924) have
shown that fluorescence is a common colour property among
Lepidoptera. By using a quartz mercury vapour lamp, whose
light was allowed to pass through a sheet of a special kind of black
glass, which is opaque to white light, a beam of invisible ultra-
violet radiation was obtained. When this beam is directed upon
insects in a dark room any fluorescence present reveals itself. It
was found that the phenomenon is widely spread among species
that are whitish or yellow, or have markings of such colour upon
their wings. The fluorescent colours vary very greatly, being
pale blue in many Geometridae, dull violet among Pieridse, bright
yellow in some Arctiidse, and bright green in the Danainse. The
phenomenon may be common to both sexes as in Geometridae,
while many Papilionidse have fluorescent males and non-fluorescent
females. It is interesting to speculate with regard to the possible
relationships of fluorescent pigments to one another, and this
property may prove to be a useful index of affinity as regards
chemical constitution. Whether fluorescence is of any functional
importance in the lives of insects whose colours exhibit the
phenomenon, or is solely an incidental feature, is problematical.

Literature

Atzler, 1930. Zeits. vergl. Physiol., XIII., 505.
Brecher, 1917. Arch. f. Entw. Mech., XLIIL, 88.

1921. Ibid., XLVIIL, 46.

1922. Ibid., L., 209.

1923. Zeit. Indukt. Abstammslehre, XXX., 291.

1924. Arch.f. FaUw. Mech., CO., 501.

1925. Zeits. vergl. Physiol., II., 691.
Brindley, 1929. Trans. Ent. Soc. London, LXXVIL, 5.
Cockayne, 1924. Ibid., LXXII., 1.

Durken, 1923. Arch.f. Mikros. Anal., XCIX., 222.
Faure, 1932. Bull. Ent. Res., XXIII., 334.
Fulton, 1922. Quart. Journ. Mic. ScL, LXVL, 339.
Gerould, 1921. Journ. Exp. ZooL, XXXIV., 385.

1926. Ibid., XLIIL, 413.

1927. Quart. Rev. Biol., II., 58.
Giersberg, 1928. Zeits. vergl. Physiol., VII., 657
Glaser, 1917. Psyche, XXIV., 30.
Harrison, 1928. Proc. Roy. Soc, B, CIL, 338.

1928a. Ibid., 347.

204 THE FUNDAMENTAL ASPECT OF COLORATION

Harrison and Garrett, 1926. Ibid., B, XCIX., 241.
Hasebroek, 1921. Biol. Centralb., XLL, 367.
1921-26. Fermentforschung, V., VII., VIII.

1922. Arch.f. Entw. Mech., LIL, 261.

Hertz and Imms, 1937. Proc. Roy. Sac, B, CXXIL, 281.
Hollande, 1913. Arch. Zool. exp. gen., 41, Notes, 53.

1923. Arch. Anat. micr.. XVIII., 85.

Hopkins, 1896. Phil. Trans. Roy. Soc, B, CLXXXVI., Pt. 2, 661.

Hungerford, 1922. Canad. Entom., LIV., 262.

Kaestner, 1931. Arch. Entw. Mech., CXXIV., 1.

Keilin, 1925. Proc. Roy. Soc, B, XCVIII., 312.

Knight, 1924. Ann. Ent. Soc. Amer., XVII., 258.

Linden, 1905. Ann. Sci. Nat. Zool., XX., 158.

Mason, 1926-27. Journ. Phys. Chem., XXX., 383 ; XXXI., A, 321 ; B, 1856.

McKenny Hughes, 1932. Proc. Roy. Soc, B, CX., 378.

Meyer, 1930. Zeits. vergl. Physiol., XL, 173.

MiCHELSON, 1911. Phil. Mag. (6), XXL, 564.

Mottram and Cockayne, 1920. Proc Ent. Soc. London, XXXVL

Onslow, 1916. Biochem. Journ., X., 26.

1921. Phil. Trans. Roy. Soc, B, 211, 1.

Palmer, 1922. " Carotinoids and Related Pigments." New York.
Palmer and Knight, 1924. Journ. Biol. Chem., LIX., A, 443 ; B, 451.
PouLTON, 1887. Phil. Trans. Roy. Soc, B, CLVIIL, 311.

1892. Trans. Ent. Soc London, 293.

1893. Proc Roy. Soc, B, LIV., 417.
Priebatsch, 1933. Zeits. vergl. Physiol., XIX., 452.
Przibram, 1913. Pfliiger's Arch. Physiol., CLIIL, 385.

1922. Biochem,. Zeitschr., CXXVIL,
1922a. Arch. f. Entw. Mech., L., 20S.

Schlottke, 1926. Zeits. vergl. Physiol., III., 692.

ScHMALFuss, 1924. Fcrmentf orschung, Vlll., 1, 86.

ScHMALFuss and Barthmeyer, 1930. Biochem. Zeits., CCXXIIL, 457.

ScHMALFuss, Herder and Winkelmann, 1933. Ibid., CCLVIL, 188.

ScHOPF and Wieland, 1926. Ber. deut. Chem. Ges., LIX., 2067.

SuFFERT, 1924. Zeit. Morph. Okol. Tiere, I. (II)., 172.

Thompson, 1926. Biochem. Journ., XX., 73, 1026.

Thomsen and Lemche, 1933. Biol. Zentralb., LIIL, 541.

Verne, 1926. " Les Pigments dans Torganisme Animal." Paris.

Wieland and Schopf, 1926. Ber. deut. Chem. Ges., LVIIL, 2178.

Wigglesworth, 1924. Proc Roy. Soc, B, XCVIL, 149.

CHAPTER VIII
SOME ASPECTS OF E^COLOGY

Temperature. The General Subject, p. 206 ; Fatal Temperatures :
A. Heat, p. 220 ; Fatal Temperatures : B. Cold, p. 221.

Ecology is primarily concerned with the relationship between
organisms and their environment. It is to be regarded as a branch
of general physiology which deals with the life-processes of
organisms as a whole, as distinct from the functions of their
individual organs. In certain of its phases ecology grades into
pure physiology, and no hard and fast line can be drawn between
the two. Plant ecology has made more rapid and more definite
progress than animal ecology, a fact which is doubtlessly owing
to their stationary habit and the greater facility with which
they can be studied in intimate relation with the environment.
Animal ecology has only in recent years become subject to exact
experimentation, and it has, as yet, scarcely found its feet; as
applied to insects, it is insufficiently advanced to admit of any
far-reaching co-ordination of data, and its foundations lie in an
adequate conception of their physiology which at present does
not exist.

Ecology has been divided into *' autecology," which treats of
the individual factors of the environment together with their
effects upon individual organisms ; and " synecology," which is
concerned with environments as consisting of combinations of
factors and their effects upon communities of organisms. In so
far as the present chapter is concerned, " autecology " alone is
implied, with the exception that combinations of environmental
factors are also briefly considered. These several agencies are
discussed more especially with reference to their influence on
metabolism and general development.

206

SOME ASPECTS OF ECOLOGY

50"

35'

>ZON£ OF F/iT/^L HIQH TEMPERyiTUIZE

'ZONE OF /N/^CT/\//TY

Temperature
The General Subject. It is well known that temperature
exercises a profound influence upon insects in diverse ways, and
their relations to this factor have been subject to more adequate

experimentation than any
QO^rr^-^M/^xiMUM F^Ty^L TEMPEftATuR£ othcr physical agcucy. The

most comprehensive recent
work on the subject is that
of Belehradek (1935), which
deals with it from the
general biological stand-
point.

The extremes of tem-
perature limit insect
activities both in space
and time. Their rates of
metabolism and, conse-
quently, those of growth,
reproduction and general
behaviour, are largely con-
trolled by temperature. A
given species of insect is
active within certain limits
of temperature which form
the zone of effective tem-
perature (Fig. 66). Above
the maximum effective
temperature there is a zone
of inactivity, or heat-
dormancy, from whose

Fig. 66. Temperature zones in relation to effects the insect recovers
the activities of the cotton-boll weevil. i j j.

(According to Hunter and Pierce.) ^hcn removed to more

favourable conditions.
Above this zone of inactivity, there follows a range oi fatal high
temperatures at which death supervenes after the lapse of a certain
interval of time, according to each temperature. At the maximum
fatal temperature death is rapid and almost immediate. On the

13-3*

•ZONE OF £FFECTI'/E T£MPEf^TU/^E

'^'.THRESHOLD OF DEVELOPMENT

>ZOHE OF /N/tCT/y/TY

4-4'

'■ZOME OF Fy^T/lL LOh^ TEMPE/^/^TURL
\Z'Q''Vr-'' MINIMUM F/imL TEMPER/^TURE

TEMPERATURE 207

descending scale, the minimum effective temperature is often
termed the threshold of develop7nent, and at which the metabohsm
of the insect borders on a condition of equihbrium, but with
sufficient excess of anabohsm over katabohsm to allow of growth
and development not being quite stationary. This point obviously
requires prolonged experimentation for its determination, and has
been accurately ascertained in very few instances. Below the
minimum effective temperature there follows a zone of inactivity,
due to rigor or cold-dormancy, but from whose influence recovery
takes place at a higher temperature. Still lower on the descending
scale is a zone of fatal low temperatures, corresponding with the
zone of fatal high temperatures, and, finally, the minimum fatal
temperature.

The observations of Hunter and Pierce (1912), with reference
to the cotton-boll weevil, may be taken as an illustration of
these temperature ranges. Their experiments, however, were not
strictly controlled, but the prevailing humidity only varied between
37 and 40 per cent. The zone of effective temperatures (Fig. 66)
for this insect lies between 13-5° ^ and 35°. Above 35° up to 50°
the insects become inactive, and from 50° to 60° (soil temperatures)
is the range of fatal high temperatures in which the weevils died
in from fifteen minutes to one second, according to the temperature.
The maximum fatal temperature was found, therefore, to be 60°.
The lower zone of inactivity, which is due to cold, ranges between
13-3° and — 4-4°, while below the latter temperature, and down to
— 13-8°, is the zone of fatal low temperatures, with — 13-8° as
the minimum. The above observations illustrate the fact that
such temperature zones exist, and are to be regarded as accurate
for the particular conditions under which they were carried out.
Within the zone of effective temperatures there exists an optimum
range, at which the greatest number of insects complete their
normal development. Its limits naturally vary when applied to
the testing of more specific phenomena ; the range of optimum
activity, for example, not coinciding with that of longevity.

^ Temperatures are given in Centigrade degrees throughout ; in many
instances Fahrenheit readings are given in the original sources of information,
and their conversion accounts for the quotation of decimal degrees.

208

SOME ASPECTS OF ECOLOGY

Peairs (1927) is one of the more recent workers to investigate
the relation of temperature to insect development. His experi-
ments were made on a considerable scale with a number of
different species of insects (larvae, pupa;, and, in one experiment,
eggs), which were divided into batches, each batch being subjected

50 iQo IS* 20" 25" 50" 550

Fig. 67. Velocity of development of Lucilia ccesar larvae. A, based
upon mean time of emergence. B, based upon average time of
first emergence for the lots at different temperatures (C°).
(From Peairs, West Virginia Agric. Exp. Sta. Bull., 208, 1927.)

to a different constant temperature. The relative humidity was
not constant, but was usually maintained near the point of
saturation. In presenting his data the reciprocals of the time
factor have been plotted against the temperatures, and the curves
consequently show the direct relation between increase in velocity
of development and rise in temperature (Fig. 67). The curves
in all cases approximate within certain limits to a straight line,

TEMPERATURE 209

which agrees with the previous findings of Krogh (1914) and
others. Peairs advanced reasons for concluding that the average
time of the first emergence of the insects, at different temperatures,
is probably a more reliable index than the mean time of emergence,
and both sets of data are plotted for comparison. When experi-
ments of this kind are carried out over a sufficiently wide range
of constant temperatures, the velocity curve is not a straight line

Fig. 68. Temperature-velocity curve of development of pupae of
Tenebrio. Curve T shows relation between incubation time and
temperature. The broken line represents the theoretical
velocity curve, with the zero of development at Z. (Adapted
from Krogh.)

throughout its course, but exhibits a definite lag phase at the
lower temperatures and inclines downwards at the higher.
Although the threshold of development should, theoretically, be
the zero of the velocity curve, or the point at which that curve
intersects the temperature axis, as a rule this is not actually the
case. The investigations of Krogh, Shelford (1927) and others
lend reasons for believing that the threshold of development lies
below the zero point of the velocity curve. There is a departure
from the theoretical curve, indicating a slowing down of the

210 SOME ASPECTS OF ECOLOGY

velocity of development at low temperatures, and that the
development of a stage commences at a point lower in the scale
than that of the intersection of the velocity curve with the
temperature axis (Fig. 68). Similarly, at the upper extremity of
the velocity curve there is a high temperature retardation effect ;
at points ranging from about 35°, or less, development slows down
and is rarely completed at 40° (Peairs). Subject to the two
deviations just mentioned, the developmental curve of an insect,
when plotted as the reciprocal of the time factor, is a straight line
which represents the rate of increase in the velocity of development.

The law of van't Hoff states that the velocity of a chemical
reaction increases with the temperature, and that the coefficient,
which expresses the rate of increase, is usually between two and
three times for each ten degrees of temperature. The relation
between temperatures and metabolism in cold-blooded animals
has been often stated to follow van't Hoff's law. It has been
shown by Krogh with reference to Tenebrio, and by many others,
however, that biological processes do not follow the principle of
a simple chemical reaction. The rate of increase of biological
processes, which accompanies a rise of temperature, cannot be
accurately expressed in terms of van't Hoff's formula, because the
value of the coefficient (commonly referred to as Q^^q) is not
constant enough to represent an approximation. The metabolism
of an insect, for example, is so complex and consists of a series of
inter-related processes that it is perhaps scarcely surprising that
it does not follow van't Hoff's rule.

Although the study of the effects of constant temperatures
yields important data relative to development and metabolism,
and most of the critical experiments refer to such conditions, it
has to be remembered that it is variable temperatures which
prevail in a state of nature. The experimental study of varying
temperatures, however, has not progressed very far in relation to
insects, and obviously much depends upon the length of exposure
to given temperatures and the ranges of the latter. There is
evidence that temperature fluctuations exert considerable influence
in accelerating development among insects. Thus Bodine (1925)
and Parker (1929-30) found that if the eggs of grasshoppers were

VARIABLE TEMPERATURES- 211

transferred to 0° immediately after they had been laid and were
later removed to a favourable temperature (27° to 37°) they
developed more rapidly than the controls which remained at the
higher temperature. This acceleration was progressively main-
tained up to a period of 240 days at the low temperature. Longer
exposure resulted in only slight further acceleration. Kept at
37°, the eggs hatched in forty-six days. When placed at 0° for
thirty days and subsequently transferred to 37° they hatched in
thirty-seven days. There was progressive acceleration up to 240
days at 0°, after which they hatched in eleven days at 37° (Table
II.). It might be argued that 0° is above the threshold of develop-
ment and that growth was going on all the time, but as a matter
of fact the threshold is above 8°, so that the result obtained is only
to be explained as acceleration.

Table II. Development of Grasshopper Eggs at Higher
Temperatures following an Exposure to 0° C. (from Parker).

Days taken for hatching at temperatures given

Days at ° C.

27°

32°

37°

0

26

32

46

30

25

31

37

60

23

29

81

120

20

21

17

180

17

18

15

240

16

12

11

500

15

10

10

The reason for the accelerative effect of exposure to low tempera-
tures is obscure, and several hypotheses have been advanced to
account for the phenomenon. For instance, it has been suggested
that definite protoplasmic rhythms occur as regards susceptibility
to their influence, which express themselves in subsequent
acceleration of growth. Also the view has been advanced that
protoplasm is highly adapted to variable conditions as a normal
environment, whereas constant temperatures are abnormal and
tend to retard development. As Peairs has remarked, this would
interpret differences in growth rate as a retardation due to
constant temperature rather than as acceleration resulting from
temperature variation. Carothers (1923) showed that the eggs of
the grasshopper Circotettix do not hatch normally under laboratory

212

SOME ASPECTS OF ECOLOGY

G

cC

4^

C

03

-M

C/1

G

O

O

4^

cd

^

s;

,14

c>

CUD

Q

-<*

u

^

c:

o

cc

u

U^

O

!<

^

C^

Cr,

O

<h

)^

o

G

ft.

«M

;-l

o

a

crt

G

•M

0^

H

hH

bX)

+i

2

4->

03

^^

G

(h

=^-i

QJ

o

+J

G

-11

o

+->

cd

(H

> ^o

C/3 S

t- o «5 o

3 ^ ^ 05

o "-1 CC O

-C 4^

^ 05

03

o -^ ceo

b< -^ l-r^

a C^ CO

j; O O CO O

23 g ob 6 6 't

§ ^ io:^^^

-M

-C 4^

.— c C3

C3

o -^ 9999

H :i3 'f (M >o o

K »0 -}< CO -t

o

o

>^- 9 9 <=> 9

:3

c3

O "-1 CO C^l C^l c^

>:;

o.

-^ +J

s -^

— , OJ

a<

a

H
be

o -§) ^^ 9 -? 9
H -^ J- ^ o o

s

K CO '0 -f o

t«

>,

t3

c« > -? « =P 9

O '—^ I— 1 I—I rH

O,

r/i

^^

3 oo

— , rt

o

rt

W

o -& ^' Tf* -? 9

H •- »0 QC OI o

K I- "(^ »o 00

> 00 CO "^ -O*

g ^ -^^'^'Q

IX>

-c ^

^ c«

OS

o -W) 9 '9 « 'S
H a t- t CO .^

K 05 I- >0 Q

.

-f

"o ocooopoo^'

01

■M (N(>iobobobcba>

C CO i-H CO ^ l^ CO ^r

«

^

.Sp2 o

(N CO 05 (M t^ (N b-

W o. "

i-H rH rH (M (N CO CO

e

E

H

I

VARIABLE TEMPERATURES

213

conditions unless previously subjected to low temperature, and
the same principle is well known to the commercial rearers of
eggs of univoltine silkworm moths.

In further illustration of this same phenomenon the work of
Cook (1927) may be cited. This experimenter used first instar
larvse of the Noctuid moth Porosagrotis orthogonia which were

•018

T

—I r

1

1

1

/<VV^--

v\

/.

■/y

•X\

•015

-

/* // y^^

\\

^ ^-..i

•012

M

& i

•009

'

Af

\

/

Constant

0 —

4

IcKours

© —

•006

5/

^yiouTS

■ —

y

^hourS

€)

•003

y

Zhoujs
1 hour

•

o

lf»

, '^ .

2^o _ ,,o

, r

, "i

10' 15' 20' Z5' 30- 35'

Fig. 69. Influence of temperature on the growth rate of Porosagrotis
orthogonia. (From Cook, Journ. Econ. Ent., XX.)

subjected to constant temperatures of 12°, 16°, 22°, 27°, 32° and
37° as controls, and the time which elapsed before the first ecdysis
took place was noted. Simultaneously other batches were sub-
jected to alternating temperatures. For high temperatures 22°,
27°, 32° and 37° C. were used, while 8° was the low temperature
in each instance. Four batches of larvse were placed at each
high temperature, one batch each for two, four, eight and sixteen
hours per day, while for the remainder of the twenty-four hours
they were placed at 8° C. (Table III.). The growth rate, or the

214

SOME ASPECTS OF ECOLOGY

reciprocal of the time required to complete the first instar, is
shown in Fig. 69. It must be pointed out that growth at a
temperature of 8° is exceedingly slow, but as the time spent under
this condition was large, in some cases a considerable amount of
growth must have occurred. Inspection of the curve labelled
constant in Fig. 69 indicates that a value 0-001 is the approximate
growth rate at 8°, and assuming this value Cook gives the following
sample calculation : —

Larvae exposed at 22° for sixteen hours daily underwent ecdysis after
97-6 hours at that temperature, followed by 40-8 hours at 8° (Table III.).
Total development . . . =100000

40-8 hours at 8° (value 0001) . . = 00408

By subtraction this leaves 0-9592 for the value of 97-6 hours at 22°,
and the value of one hour at that temperature is consequently 0 0099,
which is taken as the growth rate for the batch of larvae in question.

The growth rates are thus calculated for each lot in the series and
are shown in Table IV., together with those of larvae maintained
at constant temperatures.

Table IV. Growth Rate for First Instar of Porosagrotis
orthogonia at Constant and at Alternating Temperatures.
(From Cook.)

Hours per Day at
High Temperature.

24

16

8

4

2

1

High Temperature.
Degrees C.

Figures below give the Value of One Hour at
High Temperature.

12

0-0022

16

0-0047

19

0-0072

22

0-0084

0-0099

0-0015

0-0102

0-0077

27

0-0128

0-0130

0-0154

0-0150

0-0130

32

0-0152

0-0172

0-0174

0-0176

0-0177

0-0138

37

—

—

0-0107

0-0120

0-0112

00100

Cook also gives data with respect to the rate of COg production
by the nearly mature larvae of another Noctuid, Chorizagrotis
auxiliaris during similar combinations of temperatures. Both

VARIABLE TEMPERATURES 215

sets of experiments lead to the conclusion that length of time of
exposure to given temperatures is of prime importance, and that
there is an optimum combination of time and temperature for
the given species — the point at which the growth rate attains its
maximum value. It is further possible to secure a given growth
rate by several combinations of time and temperature.

Ludwig and Cable (1933) give a useful discussion of the subject
of alternating temperatures. They also describe the results of
their experiments with pupae of Drosophila which were exposed
to daily alternations between two temperatures. When one of
the temperatures is above the optimum for development, and the
other is between the a point (10°) and the optimum, development
appears to be retarded. If, on the other hand, both temperatures
lie between the a point and the optimum, the range of development
is unaffected as compared with controls at the optimum tem-
perature. If one temperature is between the a point and the
actual threshold of development the rate is accelerated. This
acceleration is due to development having taken place below the
a point. These authors further point out that alternating tem-
peratures provide a satisfactory method for determining the
development threshold. In these experiments, if the pupae of
Drosophila be exposed to a high and a low temperature, and the
developmental rate is the same as that of controls at the constant
high temperature ; it is obvious that no development has occurred
at the low temperature. If, however, the time at the high
temperature is less than is required in a constant exposure to
the same temperature, some development occurs at the low
temperature ; the latter, therefore, is above the developmental
threshold. By this method it was found that the a point (10°)
in the case of Drosophila pupge is considerably above the true
threshold, which lies between 7° and 8°.

In computing the effects of variable temperatures, many
entomologists have used the number of " degree-days " for this
purpose. The total number of degree-days, for a given stage of
development of an insect, is arrived at by taking each day from
the commencement to the termination of the stage concerned,
and by computing the number of degrees by which each day's

216 SOME ASPECTS OF ECOLOGY

mean temperature exceeds that of the threshold of development,
and then summing the number of degrees so obtained for the
whole period under consideration. Further investigation has
showed that the idea of a constant total of effective temperatures,
or the thermal constant as it is often termed, necessary to com-
plete a given stage has proved to be approximately correct in
certain cases. The thermal constant, however, may exhibit wide
fluctuation, especially during exceptionally hot or exceptionally
cold seasons. This is owing to the fact that the influence of one
degree of temperature during a unit period of time is not the same
at different temperatures. As an example, it may be mentioned
that the beetle Sitodrepa panicea is stated to require 1,820 day-
degrees at 20°, and 3,638 at 17°. As a rule, for practical purposes,
temperature summations are based upon mean temperatures
over twenty-four-hour periods. In many cases the thresholds of
development for the species concerned are unknown, and an
arbitrary figure of say 55° (42° F.) has to be adopted. This
method is admittedly rough and ready, and its use can only be
qualified in cases where no experimental study of a given species
in relation to temperature has been carried out. Shelford (1927)
has come to the conclusion that temperatures cannot be summed
correctly for biological purposes unless readings are taken at
intervals of one or two hours, instead of daily, and corrected for
the effects of other factors. Such correction he terms temperature-
substitution, and is only possible through preliminary observation
or experimentation affording temperature and humidity data for
the defining of standard conditions. The temperature-substitution
method translates the observed conditions into terms of the
response of the organism into developmental units, which can be
summed for biological purposes. He claims, therefore, that a new
method is required which will take into account the effects of all
variations of the factors involved in small units of time. Hence
he adopts what he terms the developmental unit, which is defined
as the difference in amount of development resulting in one hour
from a difference of one degree of mean medial ^ temperature

^ Medial temperatures as implied by Shelford are those whose range lies
on the straight line portion of a reciprocal (velocity) curve.

TEMPERATURE

217

70

(other conditions being average), as shown by the difference in
time required to complete the stage. Shelford's work is with
special reference to the Codling moth, for which he gives tables of
developmental units under all combinations of temperature and
humidity, and based upon data
covering ten years of observation
and experimentation. Such a
method, however, is obviously
too laborious to allow of its ^„

60

general adoption.

The influence of temperature
on the duration of life has been 50
shown, in a general way, to
exercise a shortening effect in
proportion to its rise. The most ^^
exact study of this phenomenon
is found in the work of Alpatov

^ 30

and Pearl (1929) on Drosophila.
Their experiments showed that,
in this insect, length of life is
doubled in the 10° of temperature
reduction between 28° and 18°,
which were the temperatures
used. Females proved to be
longer lived than males, and
individuals of both sexes were
used in the double series of
experiments. In one series they
were reared from the egg at 18°
and in the other series at 28°.
The adult flies were then tested
as regards their viability at temperatures of 18

20

Fig. 70. Average duration of life in
days : males (two lower curves)
and females (upper curves) of
Drosophila. Black lines indicate
development at 18° and broken
lines at 28°. (From Alpatov and
Pearl.)

25° and 28°, and

the results are shown in Fig. 70. It will be noted that the survival
of individuals reared at 18° was marked longer than those at 28°.
Alpatov and Pearl's conclusion that, at the higher temperatures,
the increased rate of energy expenditure during growth and
imaginal life shortens survival is what would be expected.

218 SOME ASPECTS OF ECOLOGY

The influence of temperature upon insect metabolism can be
measured in the respiratory exchange by computing either the
carbon dioxide output or the oxygen consumption. The respira-
tory exchange is an extremely variable quantity, and the most
important influencing factor is functional activity. It is well
known that there is a certain minimum or basal metabolism
inseparable from the life of the individual. Basal metabolism
cannot be determined in the strict sense of the term, but an
approximation to it, corresponding to a minimum functional
activity, may be obtained, and is termed by Krogh (1916) standard
metabolism. The latter is regarded as the result of minimum
functional activity when voluntary muscular movements are in
abeyance, and no food is being taken or absorbed. In so far as
insects are concerned, the main source of error in many cases is
the difficulty of eliminating all muscular activity. Metabolism
due to the functional activity of the whole organism is an increased
metabolism over and above the standard. In many cases, in
dealing with adult insects or their larvge, standard conditions have
not been obtained, but such experiments require consideration
in the absence of the more exact data. Standard metabolism is
more nearly approached in the quiescent stages of the egg and
pupa, but the more rapid morphological changes undergone in
those stages are not characteristic of ordinary development. In
insects the respiratory exchange is high compared with other
animals, but, as Krogh remarks, since the experiments have been
made on a large number of insects at a time, the latter were often
probably restless, and consequently standard conditions wTre not
always obtained.

The literature on the subject of respiratory exchange in insects
has been fully reviewed by Sayle (1928), but the available data
admit of few general conclusions. ^lost observers studying the
COg output in different stages of an insect record a reduction in
the, pupa as compared with the larva or imago. The amount
produced during a given stage generally rises with the temperature
within certain limits. Several observers, including Krogh (1914),
Weinland (1906) and Frew (1929), have shown that the respiratory
exchange (COg production) of insect pupae is high at first, then

TEMPERATURE AND METABOLISM

219

falls for some days, and finally rises again, the final rise being
possibly partly due to muscular movement within the nearly
mature pupae themselves, which is preparatory to the emergence
of the imagines. The same general conclusions were arrived at
by Taylor (1927) with reference to the COj production of certain
dipterous and Lepidopterous pupae, but in some species a marked
depression was noted just before the time of emergence — a feature
not previously noted by other observers. Krogh determined the
CO 2 output of pupae of the beetle Tenehrio molitor at different
temj^eratures from 21° to 33° C. (Table V.), and concluded that
the total metabolism, as judged by this criterion, was constant
over the range of temperatures given for individuals of uniform

Table V. COg Production by Pupae of Tenehrio molitor. (From

Krogh.)

Temperature.

Duration of Pupal

Life.

Hours.

CO2 produced during

Life by 1 leg. of Pupae

(litres).

Average CO2

production

per hour c.c.

32-7°

137-9

59-3

427

27-25°

172-5

58-0

336

23-65°

234-1

59-1

252

20-9°

320

59-6

186

weight.^ Krogh also showed that the relation between tempera-
ture and COg output could not be satisfactorily expressed by
van't Hoff's coefficient, and that between 20-9° and 27-25° C.
the increment in COg production was proportional to the
temperature : above 27° it decreased, as did the velocity of
development. Peairs (1927) also advanced some evidence that
the amount of CO2 produced during the pupal instar of the fly
Calliphora vomitoria is near a constant, irrespective of temperature,
and he suggests that departures from the constant may represent
basal metabolism not directly associated with the developmental
processes. On the other hand, Northrop (1926) found that the
amount of COg produced by Drosophila is not constant for either

^ Similar observations made by him on the eggs of the water beetle Acilius
showed that a fixed amount of CO2 is produced during their incubation.

220 SOME ASPECTS OF ECOLOGY

larval or adult life, being greater at 15° than at 26° or 30° C,
but the factor of movement does not appear to have been taken
into account.

Fatal Temperatures. A. Heat. A large amount of data has
accumulated in recent years with reference to the thermal death
point for many insects, but the results are discordant, and admit
of only approximate comparisons. This is largely due to the
fact that the length of time different insects were subjected to
different temperatures varied within wide limits, and the same
applies with regard to the interval elapsing between transference
from one temperature to another, the conditions of humidity and
other factors. In a state of nature the highest temperatures
under which insects regularly live are those prevailing in hot
springs and in deserts. In the case of hot springs (vide Brues,
1927), the highest temperature supporting insect life appears to
be 49° to 50°, Chironomid larvae being abundant at the first-
mentioned temperature in hot pools in the Yellowstone Park,
U.S.A. Coleoptera of the genera Hydroscapha and Bidessus are
regularly found in hot springs in Europe and North America at
temperatures varying from 30° to 46°. Buxton (1924) mentions
that certain insects are active in the Palestine desert on bare
earth whose surface temperature reaches 55° to 62°. The lethal
temperature for insects living in situations of this kind does not
appear to have been investigated, and the results would yield
data of considerable interest.

With regard to other insects, it may be mentioned that Graham
found little difference in the thermal death points for the larvae
and adults of three species of beetles living in logs of timber, the
mean death point being between 46° and 48°, with a maximum
fatal temperature of 50° to 52°. In the case of the honey bee,
Pirsch (1923) found that the fatal temperature lay between 46°
and 48°. The grain weevils Calandra granaria and C. oryzw have
frequently been the subject of temperature experiments, and
Back and Cotton found that all stages were killed after one
hour's exposure to a temperature of 47-8° to 48-9°. Dendy and
Elkington (1920) found that three minutes' exposure to 48-9° was
the minimum time required to kill the adults of the first-mentioned

HEAT 221

species. On the other hand, these same observers ascertained
that five minutes' subjection to a temperature of 62-8° was required
to kill Rhizopertha dominica, also a grain-infesting beetle, but one
which appears to be especially resistant to a high temperature.
Clothes moth larvae are killed after thirty-one minutes at 53-3°
(Back), and the beetle Silvanus surinamensis in all its stages
succumbs at a temperature of 51 -T" (Back and Cotton). Pediculus
humanus, according to Nuttall, is killed in dry heat at 55° after
five minutes' exposure, while its eggs require ten minutes at
55° to 61°.

The practical application of fatal high temperatures, with
reference to insects infesting stored grain and analogous products,
has come very much to the fore in recent years, and there is a
considerable literature on the subject. Under such conditions it
is obviously essential to ensure that every part of the material
is heated to the temperature known to be lethal to the insects
concerned. Much depends, therefore, upon the nature of the
stored product that it is desired to sterilise, its moisture content
and its bulk. Recent experiments by Barber (1929), with
reference to the killing of caterpillars of the corn borer (Pyrausta
nubilalis) within the ears of maize, will serve to illustrate this
point. It appears that fifteen minutes' exposure to a temperature
of 54° is necessary to kill these larvae. On the other hand, to
attain this temperature within the corn cobs the latter require to
be subjected to a temperature of at least 60° for a period of eight
hours, or for a shorter time at higher temperatures. It is probable
that if stored food-products be heated uniformly to a temperature
of 62-8°, and maintained at that temperature for five minutes, they
will be completely sterilised as regards all insect life (Dendy and
Elkington). The same result, however, would be achieved by more
prolonged exposure at lower temperatures, and it appears from a
recent bulletin by Dean, Cotton and Wagner (1936) that a tempera-
ture of 49° to 52°, applied for a period of ten to twelve hours, is
sufficient to kill the various insect inhabitants of flour mills.
• Fatal Temperatures. B. Cold.^ When insects are subjected to

^ The greater part of this section is taken from a paper by A. D. Imms in
Annals App. Biol., XIX., 1932, p. 125.

222 SOME ASPECTS OF ECOLOGY

temperatures that descend below the threshold of development, a
point is reached where cold becomes fatal. This point varies
enormously in different species, but is lowest in those which
hibernate and is also dependent on several factors. Certain over-
wintering species are well capable of withstanding a temperature
of — 50°. Remarkable as this fact is, it seems to fade into
insignificance when compared with the resting phase in rotifers
and tardigrades which have been recorded by Rahm (1923) in
Germany to be able to withstand a temperature of 7° absolute
(—272°) for several hours. The researches of Robinson and
Payne in America, and of Sacharov in Russia, have shown that
cold resistance in insects is explainable on certain biophysical
principles. Robinson's experiments showed that cold hardiness
cannot be acquired by an insect rapidly — it requires a falling
temperature. As the temperature sinks below freezing-point the
colloid particles in the tissues and body fluids withdraw and bind
free water as films around themselves. This so-called bound or
adsorbed water ceases to behave as ordinary free water since it
does not freeze even at — 20°. It is the ratio of bound to free
water which, when sufficiently high, protects insects against
freezing temperatures. In pupae of the moth {Telea polyphemus)
Robinson found that the bound water increased 48 to 50 per
cent, when the insects had become fully cold hardened. Payne
(1929) has stressed the importance of humidity, and her numerous
experiments indicate that cold hardiness bears an inverse relation
to absolute humidity, regardless of temperature. For a given
temperature insects are acclimatised to, their survival at lower
temperatures depends upon the amount of humidity present, as
is shown in the accompanying table (Table VII.). In the case
of the beetle {Synchroa punctata), Payne, by using thermo-electric
methods, found that when the insects were placed in a desiccator
at 15° their freezing and undercooling points were greatly lowered.
Thus, in the normal larvae, these points were — 3-04° and — 6-71°
respectively, whereas in the desiccated larvae these same points
were — 71-4° and 23-26°. It therefore appears that the moisture
content of the insect itself is also an important factor in the process.
Now the water content of insects is known to have a direct relation

1

COLD HARDINESS

223

Table VI. Water Content of Various Insects and their Food.
(From Robinson, 1928.)

Water

Water

Species.

Food Plant.

Content of
Food.

Content of
Insects.

Per cent.

Per cent.

Granary weevil, adult

Stored wheat

9-11

46-47

(Sitophilus granarius)

Rice weevil, adult

Stored wheat

15-16

48-50

(Sitophilus oryza)

Locust borer, adult

Locust tree trunk

30-32

56-60

(Cyllene robinice)

Colorado potato beetle, adult

Potato plant, leaves and

70-74

62-66

(Leptinotarsa decemlitieata)

stem

White grub, larva

Wheat shoots and roots

64-67

73-82

(Phyllophaga spp.)

Mourning cloak, larva

Willow leaves

70-73

77-79

(Vanessa antiopa)

Willow saw-fly, larva

Willow leaves

70-73

79-82

(Cimhex americana)

Cutworm, larva

Lettuce

78-79

83-88

(Chorizagrotis)

Army- worm, larva

Corn leaves and stem

77-78

87-89

(Cirphis unipuncta)

Imported cabbage-worm

Cabbage

88-89

83-84

(Pontia rapes)

Polymephus moth, larva

Hazel leaves

71-73

90-92

(Telea polyphemus)

Table VII. Cold Hardiness in Tussock Moth Eggs.
(From Payne, 1929.)

Environmental

Survival

Relative Humidity

Absolute

Temperature ° C.

Temperature.

in per cent.

Humidity.

20

- 1-5 ± 0 5

100

17-5

- 61 ± 1-3

80

13-9

- 11-5 ± 2-4

50

90

15

- 70 ± 1-7

100

12-8

- 100 ± 2 25

80

10-2

- 13-5 ± 31

50

6-5

10

- 110 ± 2-3

100

9-2

- 13-3 ± 3 2

80

7-4

- 16-5 ± 5-6

50

4-6

0

- 16-5 ± 41

100

4-6

- 170 ± 30

80

3-5

- 200 ± 5-3

50

2-2

224 SOME ASPECTS OF ECOLOGY

to that of food (Table VI.). Grain weevils have a water content of
45 to 50 per cent. ; in the cabbage butterfly larvae it is 83 to 84
per cent. Payne also found that cold resistance is greater in
starved larvae and lower in those fed with moist food. Her
conclusions have been confirmed, and in some directions amplified,
by the work of Sacharov (1930). This experimenter paid
particular care to the technique employed. He avoided the
insertion of thermopile needles into the insects by employing the
dilatometer and criohydrate solutions. In this way surgical
shock, which causes a measurable rise in temperature, was avoided.
Sacharov's conclusions are apparently based upon careful technique,
and they may be briefly summarised as follows. Cold hardiness
depends upon two intrinsic factors : (1) the proportion of easily
freezable water to the total water content, and (2) the percentage
of fat present. Hibernating insects, he claims, fortify themselves
by reducing the one and increasing the other. He tested these
two factors in the caterpillars of the brown-tail moth taken
straight from hibernation and after feeding for several days in a
warm chamber. The differences are very striking, and his results
are quoted below (Table VIII.).

Sacharov further carried out a number of observations upon
different insects, taken from their natural habitats during winter,
and subjecting them to different temperatures. His results are
shown in abstract in the accompanying table (Table IX). With
chafer larvae he points out that the reason for their penetration to
greater depths in the soil during hibernation appears to be due to
their possessing but a limited capacity for cold hardening. They
would seem to have an exceptionally high w^ater content and even
at a temperature no lower than — 5-7° the amount of non-frozen
water w^as found to be only 25-7 per cent, of the total water. The
hive bee is well known to be susceptible to the effects of cold and
damp and requires well-insulated quarters for hibernation. In
this connection Sacharov's results are rather unexpected, since it
would seem that this insect is more resistant than chafer larvae.
The adult of the herald moth (Scoliojpteryx lihatrix) and the larvae
of the longicorn (Plagionotus arcuatus), are both highly resistant
to the cold. In the case of the herald moth, which hibernates

COLD HARDINESS

225

Table VIII. Freezable Water and Fat in Caterpillars of the
Brown-tail Moth, Euproctis chrysorrhcea L. (From Sacharov,
1930.)

Water.
Per cent.

Dry
Substance.
Per cent.

Ratio of
Frozen
Water
to Live
Weight.

Ratio of
Frozen
Water

to Total
Water.

Non-
frozen
Water.

Fat.

Temp.
" C.

Ratio
to Live
Weight.

Ratio
to Dry
Weight.

A. Caterp

illars from hibernating nests.

- 5-75

71-83

28-17

. —

. —

100

4-93

16-44

- 7-8

71-83

28-17

. —

. —

100

4-93

16-44

- Ill

71-83

28-17

3-63

5-06

94-94

4-93

16-44

- 17-35

71-83

28-17

10-93

15-22

84-78

4-93

16-44

1
B. Caterpillars fed for 3-

1 days.

- 3-9

82-94

17-06

. —

. — .

100

252

12-46

- 5-75

82-94

17-06

4-13

4-98

9502

2-52

12-46

- 7-8

82-94

1706

37-20

44-85

55-15

2-52

12-46

Note. In series A the caterpillars withstood a temperature of
while in series B 82 per cent, succumbed at — 7-8°.

17-35'

Table IX. Freezable Water and Fat in Various Insects.
(From Sacharov, 1930.)

Temp.
° C.

- 5-75

5-75
111
17-35

11-1
17-35

2-9
5-75
111

Water
Per cent.

79-16

48-65
48-65
48-65

54-12
5412

74-05
74 05
74-05

Dry
Substance.
Per cent.

Ratio of

Ratio of

Frozen

Frozen

Water

Water

to Live

to Total

W^eight.

W^ater.

Non-
frozen
Water.

Larvae of Melolontha hippocastani L.
20-84 I 59-13 I 74-27 I 25-73

Adults of Scoliopteryx libatrix L.

51-35
51-35
51-35

13-97
25-94

28-91
53-39

100
71-19
46-61

Larvae of Plagionotus arcuatus L.

45-88 — — 100

45-88 1-80 3-53 96-47

Adults of honey bee, Apis mellifera L.

96-52
36-03

25-95
25 95
25-95

2-58
27-36
54-72

3-48
26-97
73-92

2608

Fat.

Ratio
to Live
Weight.

607

1818
18-18
1818

14-36

2-66
2-66
2-66

Ratio
to Dry
Weight.

29-80

56-31
56-31
56-31

30-05

10-26
1026
10-26

ll.A. ENTOMOLOGY.

I

226 SOME ASPECTS OF ECOLOGY

under bark, in thatch buildings, etc., Sacharov mentions that it
has been found coated with ice and yet subsequently revived and
became active. It will be observed that its total water content
is very low and 71 per cent, of this water was non-frozen at — 11-1°,
while the ratio of fat to live weight appears to be high.

The work of the three observers mentioned has put us on the
track of understanding the way of solving the problem of cold
hardiness in insects and that it is linked up with the biophysics of
metabolic water in the living creatures. All that we know about
this water is that in cold hardy species a high proportion of it is
held in a peculiar phase that does not freeze until relatively very
low temperatures are attained. Whether this water is actually
bound or adsorbed in the strict physical sense will not be discussed
here.

In a state of nature we may picture to ourselves what appears
to happen to an insect when it enters into hibernation. It ceases
to feed, and wanders until it discovers an appropriate resting
place. At this period it has a liberal store of fat in its tissues that
has accumulated during the preceding period of activity. Its
water content is reduced and, as the prevailing temperature of its
environment gradually falls away during autumn and early
winter, the ratio of non-freezable to total water in its body increases
in relation to the fall in its freezing and supercooling points.
Cold dry weather, it would seem, aids the acquisition of the insect's
powers of resistance, whereas cold damp weather is much less
favourable. It is probable that vast numbers of insects perish
when the acquisition of cold hardening is interfered with, since
the fatal supercooling point is higher and more quickly reached.
The farmer, in cultivating his land during winter, disturbs the
winter quarters of immense numbers of insects. These are often
brought in the process into contact with much lower temperatures
than their hardening capacity enables them to resist, and large
numbers are killed.

For the most recent contribution to the subject of freezing
temperatures in relation to insect life, the reader is referred to a
paper by Salt (1936). In addition to giving a critical review of
the literature up to date, this author describes improved apparatus

COLD HARDINESS 227

for studying the undercooling and rebound points and for main-
taining a continuous accurate record of body-surface temperatures.
Many important aspects of the subject are dealt with, and among
them the effect of moisture in contact with the body-surface may
be specially mentioned. This is usually in the form of ice, and in
many insects it is found to result either in the reduction or com-
plete elimination of the capacity for undercooling. In some other
insects it seems that no effect in this connection results from a
relatively heavy coating of ice or water.

CHAPTER IX
SOME ASPECTS OF ECOLOGY— continued

Humidity, p. 228. Light, p. 235. Atmospheric Pressure, p. 237.
Air Currents, p. 238. Food. The General Subject, p. 239 ; Vitamins,
p. 243 ; Symbiotic Micro-organisms, p. 245. Climate. Geographical
Distribution, p. 251 ; Climate in Relation to Seasonal Prevalence, p. 257.
Literature, p. 259.

Humidity

The influence of humidity on insects has attracted a good deal
of attention in recent years. Before considering the effects of
humidity as an ecological factor it is necessary to refer briefly to
the subject of water balance and kindred problems in the insect
body which have a very definite bearing upon it. While the
water necessary for the maintenance of the insect organism is
derived in the main from the food, there is evidence that at least
some part of it may be obtained from other sources. The meal-
worm {Tenebrio molitor) appears to be one of the insects that is
known to be able to make use of water resulting from the oxidation
of its own reserve substances. According to Buxton (1930) this
insect is able to regulate the proportion of water and dry matter
in its body. This ratio, it was found, was constant in starved
insects kept at several relative humidities over a period of a
month at 23° C. Certain insects appear to be able to absorb
water from moist air. Here the best-known example is the meal-
worm. The actual mechanism by means of which this " hygro-
scopic property " functions is, however, unexplained. According
to Buxton (1930) meal-worms lose weight when kept at 80 per
cent, or lower relative humidity : but if kept at 90 per cent,
humidity and at a temperature of 23° C. or 30° C. they gain in
weight. The mean gain in weight, Buxton states, represents
water ; and during twenty-three days at 30° C. and 90 per cent,
relative humidity, from 57 to 65 per cent, of the total weight of

228

WATER LOSS 229

the insects. It is concluded, therefore, that meal-worms are able
to gain water from an atmosphere which is 90 per cent, saturated.
Other instances of a kindred nature are also quoted by Buxton in
his general review on humidity (1932). That the eggs of some
insects also absorb water is certain. Roonwal (1936), for example,
has recently shown that in the African migratory locust (Locusta
migratoria migratorioides) the water content of the egg increases
from about 52 per cent, of the net weight in the freshly laid egg to
82 per cent, in the fully developed egg. The water in this case is
most probably absorbed from the surrounding soil in which the
eggs are laid.

Both climatic conditions and metabolism affect the rate at
which water is lost from the bodies of insects. This rate of water
loss, when measured in air of different humidities, bears a general
relation to the saturation deficiency of the air. According to
Mellanby (1935) practically all the water evaporated from the
body of an insect is lost through the tracheal system. This
conclusion is disputed by Ramsay (1935b), who showed that in
Periplaneta a considerable proportion of the water given off is
evaporated from the body-surface and at 30° C. a sudden increase
in the rate of evaporation occurs. This phenomenon, he states,
results from the presence of a thin film of fatty substance on the
body-surface which undergoes a change of state at that temperature
in that it shows a greatly increased permeability to water. Many
insects can reduce their body temperature below that of sur-
rounding air when the latter is tolerably dry and hot, and this
temperature regulation is brought about by evaporation.

The existence of an insect in an atmosphere approaching
saturation depends on the excretion of water through the
Malpighian tubes and in the faeces. On the other hand, in an
insect living under extremely dry conditions, any loss of water
via the alimentary canal is practically nil. At the same time, by
excreting solid uric acid, which is insoluble, water is conserved
and any water present in the hind intestine appears to be resorbed
by means of the rectal papillae (Wigglesworth, 1932). The
regulation of the water content of the insect organism is, as
already mentioned (p. 222), also important in connection with its

X

230 SOME ASPECTS OF ECOLOGY

survival at freezing temperatures. In order to survive it must
avoid ice-formation in its blood and tissues. Water loss by
concentrating salts and other solutes in the body-fluid lowers the
freezing point. The adsorption or " binding " of much of the
remaining water by the colloids of the body still further lowers
the freezing point. The relation of " bound " to free water appears
to be highest in those insects which are able to withstand the
greatest degree of cold.

While certain insects, such as Rhodnius, are unaffected by
humidities ranging from 0 to 90 per cent., even with twenty-four
hours' exposure, with others such as flea larvae the death point is
at a lower temperature in dry air. The cockroach, however, can
survive a higher temperature in dry air — 10° higher — than in moist
air (Bodenheimer, 1927). The adult Calliphora and the beetle
Ejpilachna, on the other hand, survive best at moderate humidities
of 60 per cent, or over. Davies (1928) has shown that atracheate
species of Collembola require a saturated atmosphere in order to
survive even a few hours at a constant temperature of 25°. On
the other hand, the tracheate form, Sminthurus viridis, proved
more resistant and was able to live fifteen hours at 20 per cent,
humidity before all were killed off.

Experiments by Headlee (1917), on the effects of atmospheric
humidity on the bean weevil (Bruchus ohtectus) and the Angoumois
grain moth [Sitotroga cerealella), are noteworthy. The insects
were kept uniformly in the dark and at a constant temperature of
21-1°. In their late larval and pupal stages these two species
exhibited a marked reduction in rate of metabolism in relation
to increase of humidity. The pupal stage of the moth lasted
seventeen days in 100 per cent, humidity, and was reduced to
twelve days when the humidity was 21-8 per cent. ; in the case of
the weevil the corresponding periods were twenty-two days in 100
per cent, humidity, and fourteen days at 44-6 per cent, humidity.
With the adult insects the evidence pointed to decreased humidity
shortening the life of the moth, and lengthening that of the bean
weevil. In the case of the eggs of the latter insect, two days
longer were required for hatching when the humidity was 100
per cent., as compared with one of 23-6 per cent. Taking the life-

HUMIDITY 231

cycle as a whole, development was most rapid at about 90 per
cent, humidity, the shortening apparently occurring at the stages
when feeding was most active ; with a humidity of 25 per cent,
comparatively few of the insects attained maturity, and even
fewer at lower moisture conditions. The same author's earlier,
and well-known, experiments with the aphid Toxojptera graminum
show a much greater tolerance of variations of atmospheric
humidity. At a constant temperature of 26-6° the same develop-
mental rate was maintained at relative humidities varying from
37 to 100 per cent. It has been suggested that, with sap-sucking

PERCENT/JQE HUMIDITY

25

f

11 14

20

15

-i— -i . . ^~^~t" t

0 7-1 25-9 73'4- 100

Fig. 71. Relation between humidity percentage and length of pupal
instar in days for Protoparce quinqueniaciilatus and its parasite
Winthemia A^-pustulata (broken line). (Adapted from Hefley.)

insects, this range of tolerance is explainable on account of the
relative unimportance of atmospheric humidity, since the insect
is rendered largely independent of such conditions on account of
the moisture obtainable through its food. This conclusion is
supported in a general way by more recent observations by
Wadley(1931).

The experiments of Hefley (1928), with the pupae of the hawk
moth Protoparce quinquemaculatus, and its Tachinid parasite
Winthemia quadripustulata, will serve to illustrate the differential
effects of humidity upon two separate species, living in close
association. The host does not, it appears, exhibit a definite
optimum reaction to humidity, since the duration of its pupal

232

SOME ASPECTS OF ECOLOGY

PER.CENT^QE HUMIDITY

25-9

T3-4

Fig. 72. Relation between humidity percentage and percentage
viability (emergence from the pupa) for Protoparce quinque-
maculatiis and its parasite Winthemia 4<-pustulata (broken line).
(Adapted from Hefley.)

instar only varied 1-5 days over a range of atmospheric moisture
from 0 to 73*4 per cent., where the temperature remained nearly
constant at 27^. The optimum humidity for the parasite at the

HUMIDITY 233

same temperature is at 73-4 per cent., when the length of its pupal
instar is at its minimum ; between that percentage and 7-1 per
cent, humidity its developmental period showed a difference of
10-5 days. At 0 per cent, humidity and 27° temperature all the
parasites failed to complete their development, whereas all the
hosts emerged. This feature does not appear to be attributable
to variations in metabolic activity, but rather to differential
powers of resistance to desiccation in the two instances, since both
species of insect showed their highest rate of development at
about 73-4 per cent, humidity. Above the latter figure, up to
100 per cent, humidity, the duration of the pupal period of both
host and parasite showed a progressive lengthening. If viability
be judged by the percentage emergence from the pupae, it is note-
worthy that host and parasite behave very differently in this
respect, dryness favouring the former and high humidity the
latter {vide Figs. 71 and 72).

Griswold and Crowell (1936) discuss the influence of humidity
on the length of life-cycle in the webbing clothes moth (Tineola
hiselliella). At a constant temperature of 25° C. this insect
developed under relative humidities ranging from 20 to 93 per cent.
At 75 per cent, humidity development was more rapid than under
other moisture concentrations, mortality was lowest and adults
lived longest. A humidity of 93 per cent, proved more deleterious
to the insect than 20 or 30 per cent. In the case of the cockroach
(Blatta orientalis), Gunn (1934) states that the preferred tem-
perature range is 20° to 29° C, and that this zone is not affected
by changes in the humidity of the air. With tsetse flies (Glossina),
Buxton and Lewis (1934) found that the effects of humidity
are complex. Flies kept at 30° C. behaved under 44 per cent,
humidity in a way which suggests that this concentration is near
the optimum : at humidities above or below this figure the flies
fed less often, and at 88 per cent, humidity they nearly always
refused to feed and died. It was also observed that the adults
lived longest and bred best at 44 per cent, humidity. At higher
temperatures the effects of humidity are clearly shown, since the
insects are able to survive in drier air for a short period at 42° to
43° C. — temperatures which are fatal in moist air. Field observa-

234 SOME ASPECTS OF ECOLOGY

tions were found to be in good accord with those conducted in the
laboratory. Thus, at the end of the dry season, adults of Glossina
tachinoides are abundant and a high proportion of the females
are pregnant, and it was found that at this period the seasonal
climatic conditions were close to the optimum as defined in the
laboratory. For a detailed discussion of the relations of various
insects to humidity, reference should be made to the writings of
Uvarov (1931) and of Buxton (1932).

Readers interested in the various methods of technique relative
to the measurement and control of humidity should consult
papers by Buxton (1931) and Buxton and Mellanby (1934), which
are mainly concerned with the maintenance of required humidities
within closed vessels such as desiccators. In many cases it is
preferable to pass air, conditioned to the required humidity,
through the containers within which the insects are lodged :
such a method, for example, is described by Gunn (1934). The
purely physical problems, which are fundamental in attempts to
establish a relation between the loss of w^ater from an animal
and the evaporating power of the air, are discussed by Ramsay
(1935a).

It is customary in entomological experiments to express
humidity in relative terms — i.e., if the relative humidity, for
example, be 60 per cent, it means that the air holds that amount of
water vapour, as compared with the maximum, at the particular
temperature of the experiment. The loss of water from the insect
and the direct influence of humidity upon it are related to the
evaporating power of the air. The relative humidity may be the
same in two separate experiments under different temperatures,
yet the amount of water which the air would require to take up
before reaching the saturation point would be different in the two
cases. Relative humidity consequently does not exj^ress the
evaporating power of the air, and in cases where it is desirable to
measure this latter factor the saturation deficiency of the air is
the index required. The saturation deficiency, it may be added,
provides comparable figures for the evaporating capacity of the
air upon which loss of water depends. The rate of evaporation is
influenced by several factors, including temperature, air move-

LIGHT 235

ment and barometric pressure. Shelford, in America, has devoted
a good deal of attention to this subject and has adopted the
method of measuring evaporation from an exposed surface of
water by means of a porous cup atmometer. Several types of
the latter instrument have been devised and are described in
Shelford's recent work on ecology (1929). By use of an atmometer
he conducted experiments on the reactions of insects to air of
different evaporating powers. It was found that certain species
reacted to air of similar evaporating capacity irrespective of
whether humidity, air currents or temperature was the chief
controlling factor influencing the rate of evaporation. The
atmometer is probably the most convenient instrument so far
devised for giving a single measurement of the combination of
factors that are involved, but, at the same time, it provides little
or no information as regards any individual effects. When
critical analyses of the factors involved are desired it is necessary
to take measurements of wind, temperature and humidity at the
same time.

Light

Almost all experimental work dealing with the effects of light
on insects is in relation to their sensory behaviour and not to its
influence upon general metabolism and growth. Northrop (1926)
is one of the few experimenters who have investigated the effect of
light intensity in relation to growth ; he tested this factor with
reference to the duration of the larval and imaginal instars of
Drosophila, using cultures which had been previously kept in the
dark for 200 generations. The source of light used was a con-
centrated filament Mazda bulb immersed in a vessel of running
water, and its intensity was measured with a contrast photometer
against a standard Hefner amyl acetate lamp. At intensities
around 2,500 metre candles the duration of th- larval life was
slightly reduced, but it became increasingly lengthened at higher
intensities until they were killed at 7,000 to 10,000 metre candles ,
while the pupae were killed at 5,000 metre candles. Above 1,000
metre candles the duration of the imaginal life was rapidly
shortened. The actual effect of the action of the light is difficult

236 SOME ASPECTS OF ECOLOGY

to evaluate owing to its influence upon movement. A number of
observers have recorded an increase in CO2 production under its
influence as compared with darkness, the increment in most cases
being attributable to the stimulating effect of light upon move-
ment. In the case of the larvae of Drosophila-this same explanation
may hold good and excess of movement thus caused may result
in retarded development. It would seem desirable that critical
experiments, upon the influence of light, be carried out with
reference to eggs and pupae in which conditions of standard
metabolism prevail.

It is very questionable whether light, even when of high
intensity, is ever the cause of death in insects. Closely connected
is the problem of radiant heat, and there is good evidence that the
death of insects under direct sunlight is due to this cause, or to
its effect on water loss from the body.

The effects of the daily duration of light upon the production
of different types of reproductive forms in aphides have been
investigated by Marcovitch (1924), and by Davidson (1929).
They have shown that the incidence of the sexuales generations
appears to be influenced by length of day under the conditions of
their experiments. By reducing the normal amount of daylight
per diem the production of sexual forms was induced at times of
the year when they do not normally occur. Conversely, by
extending the duration of daylight, asexual viviparous
reproduction results during times of the year when only sexual
reproduction normally prevails. The influence of light in such
experiments appears to be important, but it is not a clearly
defined result, owing to its effect on the host plants upon which
the aphides were feeding. Also, the factor of temperature
cannot be entirely excluded.

Very similar results have been obtained more recently by
Wadley (1931), who found that in the aphid Toxoptera grarninum
sexual forms can be induced to appear under defined experimental
conditions. These conditions involved the shortening of the days
to less than twelve hours and temperatures averaging below 22° C.
No other conditions affecting their appearance were noted.

Early experiments, on the differential effects of limited parts of

ATMOSPHERIC PRESSURE 237

the spectrum, have shown that violet rays exercise an apparent
stimulating effect upon growth in the case of flesh-fly larvae and
silkworms and also with animals other than insects. This
subject, however, has not been followed up in recent years, but
is one which merits critical examination with modern technique.

Atmospheric Pressure

Observations on the influence of changes of atmospheric pressure
upon insects are too scanty to admit of any deductions to be
drawn. A few field observations have been recorded which
indicate that such changes do affect insect behaviour, and certain
early experimenters adduced evidence that insects are able to
survive in a partial vacuum even for many days. The evidence
appears to indicate, however, that ordinary barometric changes
which prevail in Nature probably exercise no very profound
influence upon insect life. Recent experimental investigations
on the subject are very few. According to Pictet (quoted from
Uvarov, 1929), a uniform reduced pressure of 710 to 728 mm.
maintained throughout the pupal instar of Pieris rapce greatly
shortens development, the duration of this stage being reduced
to nine days as compared with a normal period of fifteen days.
The emergence of the adults appears to take place mostly at times
of reduced barometric pressure, a diminution of only 1 mm. being
stated to be sufficient to induce eclosion, provided the insect be
in a condition ready for emergence. It would appear that a fall
in atmospheric pressure operates through creating a difference in
relation to the internal pressure within the insect itself.

Experiments conducted by Back and Cotton (1925) with
reference to the effects of vacuum treatment upon insects affecting
stored products revealed considerable differences in their powers
of resistance. Adults and pupae succumbed more quickly than
larvae when subjected to a vacuum of 28 to 29 inches in a bell- jar,
and of 24 to 28 inches in a concrete vault, at a temperature in
both cases of 15-5° to 21-0°. Three days' treatment in a concrete
vault killed not only all the adults of twelve out of eighteen
species, but also the pupae of all the species present in that stage,
and all the larvae of several species. It required seven days' treat-

238 SOME ASPECTS OF ECOLOGY

ment, however, to kill the surviving adults of every species, but
that duration proved insufficient for the complete extinction of
th<=^ larvae of Attagenus piceus, Tenebrio ohscurus and Trogoderma
tarsale. The last-mentioned species proved the most resistant
and suffered a mortality of only 30 per cent. These results are
not in accord with those carried out in the bell- jar experiments,
when complete mortality of the insects in all their stages occurred
within four days. The experiments prove, however, that
individuals of certain species of insects can withstand several
days' subjection to an almost complete vacuum. The treated
specimens, it may be added, were retained for two weeks after
the experiment and frequently examined to ascertain whether
recovery took place.

Air Currents

Within the last fifteen years or so some attention has been given
by Coad (1931) and others in America, and more recently by
Berland (1934) in France, to the carriage of insects by currents of
the upper air. By the adoption of trapping arrangements
attached to aeroplanes, small insects, including a good many
wingless forms, have been collected up to an altitude of at least
14,000 feet. Once such creatures become carried upwards by
convection or other movements it would appear that it is possible
for them to be transported long distances, since the upper air
currents are stated to be both strong and constant in direction.
Buxton (1935) in discussing the Samoan fauna, concludes that a
number of facts concerning the distribution of the insects can be
best explained on the supposition of their having reached the
islands via the upper air. It is obvious that when a more extended
study of insects occurring in the upper air has been undertaken
we may be in a position to be able to account for some of the
anomalies of insect distribution. It has also certain very obvious
economic implications regarding the entry of injurious species
into lands which they did not previously inhabit. Furthermore,
quarantine regulations naturally have to be considered in a new
light, and previously unaccountable cases of their inefficacy seem
now to be explained. The fortuitous transport of noxious

I

FOOD 239

insects by aeroplanes is scarcely any longer an entirely novel
subject. The possibility, for example, of mosquitoes infected
with the virus of yellow fever becoming stowaways in aeroplanes
and thus obtaining passages to territory free from that disease
is a subject of inquiry at the present time. In England an
investigation of insects of the upper air is being carried out under
the direction of Professor A. C. Hardy of University College, Hull.
By the use of suitable kites, with opening and closing nets attached,
the heights at which insects are obtained can be properly registered.
This method is an adaptation of the tow-net principle used in
sampling life in the depths of the sea. Professor Hardy informs
me that, so far, rather more than 2,000 feet is the maximum height
sampled. A series of nets, attached at different heights, to the
300 feet masts of a wireless station in Lincolnshire is another
experiment. It is providing information respecting the drift of
insects at such heights in relation to meteorological data on a
linear distance of approximately one mile. Similarly, by flying
kites with nets attached, while on board a fisheries research
vessel in the North Sea, Professor Hardy states that quite sur-
prising collections were obtained of insects drifting above the sea
up to a distance of 150 miles from the land. In these cases the
kites were flown up to 400 feet in the air, while other nets were
attached to the masthead of the vessel. The preparation of some
of the results of this work is now nearing completion for the i3ress.

Food
The General Subject. In dealing with the influence exercised
by food we arrive at the point where physical and biological
factors intergrade. A large amount of scattered and unco-
ordinated information exists with regard to the role of food and
its constituents upon insect life, and an annotated bibliography
of the subject, including nearly 600 titles, has been compiled by
Uvarov (1928). Much of this information belongs to the realm
of pure physiology, and a considerable proportion of it requires
revision in the light of improved methods of technique. From
the ecological standpoint, the effects of nutriment upon growth
and reproduction are fundamental, but the subject cannot be

240

SOME ASPECTS OF ECOLOGY

divorced from physiology. Recent research has been directed
towards analyses of the influence of individual constituents of
the food, and the part played by vitamins or substances allied to
them. Increasing attention is also being given to the behaviour
of micro-organisms in the nutritional process, but, for the most
part, we know as yet very little concerning their actual functions.
In illustration of the effects of different host-plants upon the
metabolism of Aphis rumicis, as expressed in the reproductive
capacity of this insect, certain experiments carried out by
Davidson (1921, etc.) at the Rothamsted Experimental Station
may be taken as an example. The figures given below represent
the average number of aphides (average of five plants in each
case) produced on one plant of the undermentioned hosts, by
single adult apterous viviparous females, during a fourteen-day
period. The aphides used were all the progeny of a single parent,
and the experiments were carried out simultaneously under
identical environmental conditions.

Table X.

Plant.

Longpod
Bean.

Sugar
Beet.

Red Beet.

Mangold.

Shirley
Poppy.

Number of aphides
produced

896

123

115

114

164

Similar experiments carried out with different varieties of broad
beans ( Vicia faba) showed that if the longpod bean be taken as
the standard or control, and the mean reproductive figure for the
aphid on that host-plant be represented by 100, the corresponding
figures for other varieties of beans may be grouped into classes
yielding reproductive figures ranging from 98 per cent, to 3 per
cent. The last-mentioned figure was obtained on Vicia narhonensis,
which some authorities regard as the prototype of all the cultivated
varieties of the broad bean. It appears that these differences, in
the reproductive capacity of the aphid, are correlated expressions
of the influence of the cell-sap, of different host-plants, upon the
metabolism of the insect. Certain further experiments by
Davidson lend support to this contention, since they show that

FOOD 241

factors influencing the nutrition of the host-plants are reflected
in corresponding deviations in the reproductive capacity of the
aphid as shown below.

Table XI.

Treatment of Plants.
(Longpod Broad Beans.)

Reproduction of Aphides.

Grown in sand, treated with tap water
only

The same, but watered with complete
nutritive solution ....

Grown in standard soil

The same treated with complete artificial
fertilisers .....

Control : taken as 100

225
140

200

These figures represent, as before, the mean number of aphides
produced (on five plants in each experiment) by a single apterous
viviparous female per plant over a fourteen-day period.

The familiar practice of pinching off the growing points, along
with the adjacent young leaves of bean plants, has also a very
marked influence on the number of progeny produced by Aphis
rumicis in a given time. In one series of five plants, in which this
treatment had been carried out, the mean reproductive rate of
the aphid was 104-8 (in fourteen days), while for plants that
remained intact the mean reproductive rate was more than three
times as high, viz., 335-2. It appears, from this experiment, that
the cell-sap of the young growing tissue of the host-plant is of a
very definitely higher nutritive value as food for the aphid than
the older tissues. Screening experiments, cutting off varying
amounts of daylight and thereby reducing the photosynthetic
activity of the host-plants, suppprt the conclusion that the
nutrition of the latter affects the metabolism of the insect, since
aphides on the most heavily screened plants produced the lowest
number of progeny in a given time.

In further illustration of the influence of food upon insect
metabolism as expressed by fecundity, and also by the duration
of life, experiments by Simmons and Ellington (1925) with regard

242

SOME ASPECTS OF ECOLOGY

to the beetle Necrobia rufipes are noteworthy. Thirteen pairs of
this beetle were kept under observation from the times of
emergence from their cocoons until they died, the average daily
mean temperature being 19-4°. During the whole period the
insects were fed entirely upon stale fat bacon, and the average
life of the female insects was 136 days with a maximum of 183
days. The average number of eggs laid per female was 137, with
312 as the maximum. In another experiment 76 pairs of the
same species were fed throughout their existence upon migrant
larvae of the fly Piophila casei. In twenty selected pairs of this
series the average life of the female beetle was 279 days, with a
maximum of 409 days. The average number of eggs laid per
female was 906, with 2,131 as the maximum. The average daily
mean temperature, from readings given by these two authors,
works out at about 20-5° during the experiment. It is evident
that there is an increase in the duration of the life of the individual
amounting to over 100 per cent., but this fact is apparently not
solely accountable for the high increase in fecundity recorded.
This conclusion is borne out by the fact that one female nourished
upon Piophila larvae lived only 121 days, but nevertheless laid
576 eggs.

With reference to oviposition by the Noctuid moth Hadena
(Polio) oleracea, Lloyd (1920) has shown that unfed imagines,
although prolific, produce a greatly reduced number of eggs as
compared with those allowed to feed, the food in this case being
broken ripe tomato fruits. Lloyd's results, which, however, are
based upon small numbers, are given below.

Table XII.

Fed.

Unfed.

Number of female moths

17

9

Average duration of life

19 davs

11 days

(9 to 29)

(6 to 16)

Batches of eggs laid ....

190

91

Average number of batches per moth

11-2

5-3

Average number of eggs per batch

78

55

Average total number of eggs per moth .

872

263

VITAMINS 243

Glaser (1923) has shown that the house-fly in captivity lives
only a few days and lays no eggs on a diet of proteins or products
of protein hydrolysis, or of raw starch. When fed on sucrose,
longevity is increased but no eggs are laid, while on a diet either
of sucrose and bouillon, sucrose and blood serum, glucose and
bouillon or of glucose and blood serum, longevity and oviposition
reach their maximum. Longevity is also high and eggs are laid
on a diet of soluble starch and bouillon or of hydrated starch and
bouillon. It appears, therefore, from his experiments that sugar
or assimilable starch, together with a solution of proteins or
products of protein hydrolysis, like bouillon or blood serum, are
necessary for oviposition.

Vitamins. Further and more detailed analyses of the nutritional
requirements of certain insects lend support to the conclusion that
they, similarly to the higher animals, exhibit certain vitamin
requirements in order to undergo normal growth. In some cases
complete growth is unobtainable unless certain accessory food-
substances are included in the diet, while in other cases such
requirements do not appear to exist. The latter conclusion is
always confronted with the difficulty that unless the insect be
reared aseptically from the egg, the possibility that growth-
promoting substances are contained in micro-organisms present
in the gut, or other organs of the insect, cannot be excluded. The
difficulties to be faced in rearing many species of insects under
absolutely aseptic conditions can scarcely be overcome, since in a
number of cases micro-organisms are known to be present in the
egg and are transmitted from generation to generation.

A considerable amount of research has been carried out with
respect to the influence of vitamins on the growth of Drosophila
and other insects, and short reviews of the subject are given by
Richardson (1926) and by Uvarov.(1928). With respect to the
species Drosophila melanogaster, the earlier work of Loeb and
Northrup, and of Northrup, showed that the larvae of this insect
require a growth-promoting substance present in yeast in order
to complete their development. Guyenot confirmed their findings,
and concluded that the substance present in yeast, which is
indispensable for growth, resembles in certain respects the water-

244 SOME ASPECTS OF ECOLOGY

soluble vitamin B, which is necessary for the higher animals.
Baumberger, two years later, advanced evidence which led him
to conclude that the substance necessary is yeast nucleoprotein.
Bacot and Harden (1922) grew Drosophila larvae aseptically, and
were able to confirm that this insect requires vitamin B for its
complete development, but not vitamin C ; at the same time, it
was able to develop in the presence of very small quantities of
vitamin A, and it was uncertain whether the latter was essential
or not.

Certain other insects have also been the subject of vitamin
investigation. The experiments of Wollman, for example, with
the blow-fly Calliphora and the cockroach Blatella appear to lead
to the conclusion that vitamins are not essential for the growth
of either insect. Notwithstanding sterilisation precautions, the
complete elimination of micro-organisms from within the eggs or
larvae cannot be ensured, and these alone may contain sufficient
vitamin to enable the insect to undergo development in sterilised
food. The careful experiments of Richardson (1926), with
reference to the Mediterranean flour moth (Ephestia kuhniella),
afford evidence that vitamins are necessary for the complete
development of that insect. Full growth occurs when the insect is
reared upon normal entire-wheat flour, but when the latter be
treated to prolonged extraction with chloroform, all the larvae,
when fed on the residue, die at an early age. If, however, the
extract containing the growth-promoting substance was added to
the wheat, residue growth of the insect took place at nearly normal
rate. Similarly, ether-extracted wheat allowed of but few larvae
to reach the pupal stage, but when the extract is added to the flour
residue practically normal growth results. It would appear that
the extract in these experiments contains a substance similar to,
or identical with, vitamin A. When an ether extract of egg-
yolk containing vitamin A was added to the wheat residues of the
previous experiments, the larvae grew normally. Richardson also
found that highly milled patent flour retarded larval growth and
resulted in moths of significantly lower weight. Since, however,
the growth value of this medium is considerably improved by the
addition of an alcoholic extract of yeast, it is believed that its

MICRO-ORGANISMS 245

deficiency lies at least partly in the low concentration of vitamin
B which appears to be present largely in the wheat kernel.

Chapman (1924), using another flour insect, the beetle Tribolium
confusum, came to the conclusion that the wheat embryo itself
is more satisfactory for growth and transformation than other
food substance. Vitamin B from the wheat embryo did not
appear to be able to supplement deficient diets, and it would seem
that some substance other than this vitamin may be involved.

The nutritional requirements of the larvae of the Sheep Blow-fly
(Lucilia sericata) have attracted a good deal of attention during
the last few years. This work has been in connection with the
importance of this species as an animal pest. The old idea that
bacteria play an essential role in the process of digestion is not
borne out. The extensive researches of Hobson, in a series of
ten papers pubhshed between 1931 and 1935, and of others, have
been reviewed by Evans (1936). It appears that the chief functions
played by bacteria are to supply a growth-promoting substance
and to liquefy the food. Other accessory growth factors necessary
for normal growth and development are cholesterol, vitamin B,
together with certain substances analogous to members of the
vitamin B complex.

Without further discussion of this subject it will be evident that
substances of the nature of vitamins are essential for the growth
of certain insects, but it should not be too readily assumed that
they are necessarily identical with those required by the higher
animals. For certain other insects, evidence that vitamins are
essential factors in the nutritional process is far from conclusive.

Symbiotic Micro-organisms. The literature dealing with the
association of micro-organisms with insects has assumed extensive
proportions, but restrictions of space will only permit of limited
reference to the subject. In a number of cases, of which the
malaria parasite and Nosema disease are familiar examples, such
organisms are parasitic in behaviour. In other instances their
relationship is of the nature of inquilines, or they are neutral,
in that they neither confer obvious benefit to their hosts nor
occasion deleterious consequences to the latter ; the intestinal
bacterial flora prevalent in many insects probably comes under

246 SOME ASPECTS OF ECOLOGY

this category. Further species of micro-organisms are beheved
to be definitely symbiotic in that they play an important part in
the digestion of the food of their hosts and, at the same time, meet
their own nutritive requirements in the process. Micro-organisms
of this type are more especially present in insects feeding upon
dead or decaying animal and vegetable material, or in plant-
sucking insects. A large number of cases of reputed symbiosis of
this kind have been brought to light in recent years, but the
relationship has more often been accredited from inference than
as the result of definite physiological proof. For example, the
intracellular and intestinal yeast-like, conidia-like and bacteroidal
bodies present in so many insects, and transmitted from generation
to generation through the eggs, are usually regarded as symbionts,
but until such micro-organisnis can be studied apart from the
hosts, definite proof of their physiological role would seem to be
unobtainable.

One of the best established cases of symbiosis between micro-
organisms and an insect host is afforded by the intestinal Protozoa
harboured by the lower termites. The present writer suggested
in 1919 that many of the peculiar forms of flagellate Protozoa
harboured by wood-feeding termites were not parasitic as
commonly stated. On the other hand, it appeared that they
derived their nutriment by ingestion of the wood-particles render-
ing the latter, in the process, assimilable as food by their hosts.
The important researches of Cleveland (1924, 1925) have proved
the truth of this contention and are of exceptional interest, since
they demonstrate that the principal constituent of the wood
utilised by the Protozoa is cellulose. He showed that the Protozoa
can be readily killed off, without injury to the termites, by
incubation at 36° C. for twenty-four hours, by starvation or by
increasing the oxygen pressure of the atmosphere. Termites thus
defaunated die within ten to twenty days if fed upon normal wood
diet, but when reinfested with Protozoa their ability to utilise
wood is regained, and they are able to live indefinitely. Cleveland
further demonstrated that termites are able to live and actively
multiply on a diet of nitrogen-free cellulose so long as Protozoa
are present. He, therefore, suggested that these insects must in

MICRO-ORGANISMS 247

some way be able to fix atmospheric nitrogen or transform
carbohydrate into protein. This faculty would appear to reside,
however, in the Protozoa, and it may be pointed out that since
the latter multiply and die in thousands within their hosts their
remains alone might furnish the nitrogen required by the latter.
Cleveland did not investigate the constituents extracted from the
wood by the Protozoa and what substances, if any, are thereby
released as food for the termites are undetermined. In a recent,
and very extensive, memoir by Cleveland and his collaborators
(1934) a rich Protozoan fauna, akin to that found in termites, is
described from the gut of the wingless, wood-feeding cockroach
Cryptocercus punctulatus of the United States. It was found that
the Protozoa of this insect contain the enzymes cellulase and
cellobiase, which produce dextrose in vitro. The cockroach, on
the other hand, does not possess cellulase, and probably not
cellobiase. It is, however, able to live longer without Protozoa
when dextrose is used as food, and it is concluded that this
carbohydrate is one of the important food constituents supplied
by the activities of the Protozoa. Trager (1932) has also
investigated the occurrence of cellulase in connection with the
above problem, and finds that it can be extracted from the Protozoa
living in the termite Termopsis, and also from those present in
Cryptocercus. In the defaunated hosts this enzyme was found
to be wanting.

Apart from termites, most other wood-inhabiting insects
harbour micro-organisms of various types in the alimentary
canal, its appendages or other organs, and the subject is admirably
discussed by Buchner (1928). The actual nature of the food of
such insects is, in many cases, quite obscure, and it has been too
readily assumed that the wood-particles excavated during the
tunnelling process serve as food. The regular presence of micro-
organisms, generally in localised situations, lends indirect support
to the conclusion that the latter are symbionts which perform an
important digestive function. Thus Sitodrepa panicea, and many
other wood-boring beetles, possess special intestinal cgeca harbour-
ing yeast-like organisms. In the larva of the beetle Potosia
cuprea, which feeds upon dead pine needles, Werner (1926) has

248 SOME ASPECTS OF ECOLOGY

shown that cellulose-digesting bacteria are constantly present in
relation with the intestine. Although it is held that the micro-
organisms referred to are symbionts, the part they actually play
in the physiology of nutrition is obscure, and knowledge of this
aspect of the problem is greatly needed.

Mansour and Mansour-Bek (1934) in their review of this subject
conclude that the intra-cellular micro-organisms of xylophagous
insects cannot be considered as playing an important part in the
digestion of wood. The work of Ripper (1930), and of others,
appears to indicate that many more insects than was previously
believed possess the ability to digest various components of wood
by means of the action of their own enzymes, particularly
cellulases. Extra-cellular micro-organisms found in the intestines
of various insects likewise play no part, they believe, in breaking
down cellulose for the use of the hosts. These organisms appear
to be used by the latter as a source of food. It is evident that the
subject is not one which it is possible to dogmatise over in the
present state of knowledge, and micro-organisms have been too
readily credited with performing functions of which there is
still a lack of conclusive proof of their actuality.

In many plant-sucking Hemiptera micro-organisms are likewise
constantly present, and they often bear close resemblance to yeasts
both in form and method of reproduction. As a rule they are
located in special " symbiont " cells termed mycetocytes, which
form a compact and very characteristic organ known as the
mycetome. The latter is lodged in the body cavity (usually in the
abdomen), and is unconnected with the digestive system by ducts
or other means. The physiological significance of these organisms
remains unsolved, and this subject, along with their origin and
incidence in different insects, is discussed by Buchner (1921) and
by Uichanco (1924). Infection of the host takes place through
the egg, and for the details of the process the recent papers by
Uichanco and by Schrader (1923) may be consulted. It is
generally believed that the micro-organisms profit by the
protection and nutritive material which they receive from their
hosts, and that they benefit the latter as absorbers of waste
products and excess food materials. Efforts have been made to

I

MICRO-ORGANISMS 249

cultivate these micro-organisms when removed from their hosts,
but the results are by no means conclusive. Berlese, in 1906, for
example, succeeded in growing the yeast-like organism, found in
association with the Coccid Ceroplastes rusci, on artificial media.
A fungus, which he termed Oospora saccardiana, resulted, and it
grew rapidly on gelatine and showed a powerful liquefactive action
on the latter. Pierantoni (1910) similarly was able to grow the
yeast-like organisms from the mycetocytes of Icerya purchasi on a
gelatine medium. More recently, Brues and Glaser (1921) reported
that they were successful in cultivating the symbionts of the
Coccid Pulvinaria innumerahilis which occur, not in a definite
mycetome, but diffused in the fat-body of the insect. The
yeast-like bodies developed on suitable media into a separate
fungus of undetermined affinities which had the property of
secreting proteolytic, lipolytic and diastatic enzymes. It is evident
that the hosts in some unexplained manner are able to inhibit any
excess of multiplication of their symbionts, as the number of
mycetocytes remains very constant and does not appear to increase
beyond certain limits. Whatever role these organisms may play,
it would appear to be exercised upon materials contained either
in the blood or the fat-body of their hosts, since, as has already
been pointed out, they do not occur in direct relation with the
digestive canal. Plant-sucking Hemiptera absorb large quantities
of soluble carbohydrates in the cell-sap, which are eliminated to a
considerable extent as honey-dew. Buchner's hypothesis, that the
symbionts are able to assimilate atmospheric nitrogen, appears
to be based upon the assumed deficiency of nitrogen in the food of
these insects. There is, however, no evidence that plant-sucking
insects do not obtain all the proteins they require from the cell-sap,
and it appears more probable that the symbionts are in some way
concerned in dealing with an excess _pf carbohydrates present. It
is evident, therefore, that more definite information is required as
to the digestive physiology of the hosts before the role of their
symbionts can be elucidated.

The cultivation of symbionts in vitro does not carry the problem
much nearer solution for the reason that their activities are often
very diverse, and no light is shed upon their essential functions

250 SOME ASPECTS OF ECOLOGY

in relation with their hosts. The most important work in recent
years is that of Aeschner and Ries (1933) and Aeschner (1932,
1934) with reference to the body louse, Pediculus. The symbionts
in this creature are localised in a small organ situated in the
mid-ventral line of the abdomen rather nearer the anal end of
the body. The complete life-cycle of these micro-organisms has
been worked out by Ries (1932). Aeschner (1932) and Aeschner
and Ries (1933) were able to show that extirpation of the mycetome
when filled with symbionts caused marked deficiency symptoms
in the louse. If, however, extirpation took place after the micro-
organisms had migrated to the oviduct it had no influence on the
well-being of the insect. From these results it was concluded
that it was not the operation, or the absence of the mycetome,
but rather the absence of the symbionts that induced the deficiency
symptoms. In his later work Aeschner (1934) studied the effect
of centrifugalisation on the eggs of the louse. It appears that
the symbionts are eliminated in the process and are wanting in
the young insects. The latter fed and developed normally until
the fifth or sixth day, when they died. This result was in complete
accord with that obtained by extirpating the mycetome, and
it is concluded that the relationship of micro-organisms and
host is one of true symbiosis. What role these organisms play
in Pediculus is unknown, but it is possible they are concerned
with the production of some " vitamin " or growth-promoting
substance.

Blood-sucking insects regularly contain intracellular micro-
organisms, and Wigglesworth (1936) describes a bacterial organism
which occurs constantly in Rhodnius proliccus. Hobson found
that sterile blood was a deficient diet for blow-fly larvae unless
" vitamin B " was added. This same investigator (quoted by
Wigglesworth) has carried out a few experiments adding the
bacteria from Rhodnius to sterile horse serum and then feeding
Lucilia larvae on the mixture. The results appear to point to
the conclusion that when blood is so infected it becomes an
adequate diet for insect growth. These observations lend support
to the view that symbiotic organisms, in exclusively blood-sucking
insects, provide an endogenous source of vitamin.

GEOGRAPHICAL DISTRIBUTION 251

Climate

Climate is the expression of a combination of physical factors,
and is of paramount importance in connection with the survival
of all forms of life. The effects of the more important of the
individual factors which collectively make up a climate have
already been referred to, and their combined influence now remains
for consideration. Climate may be defined as the complex of
meteorological factors prevailing throughout the year for a given
area, while the incidence of these same factors at some particular
point of time constitutes weather. Climate in the broader sense
influences the geographical and ecological distribution of species,
while weather influences their local prevalence by seasons : the
one operates in space, the other in time.

Geographical Distribution. It is well known that the vast
majority of insects are more or less restricted to definite areas.
Without taking into account past geological changes, the factor
governing such distribution at the present day is very largely
climatic. On the other hand, topographical barriers, such as
mountain ranges, deserts and oceans, play an extremely important
part. Their influence, it is true, is partly in relation to the climates
they engender, but, apart from such, they remain as obstacles
which few species can overcome. The tsetse fly, for example,
is unable to survive the ecological conditions presented by the
Sahara desert, and consequently does not penetrate north of that
vast area. Similarly, it has not spread into Asia, where the
Arabian desert presents a more formidable barrier than the
relatively narrow strip of water forming the Red Sea. The Atlantic
Ocean effectually precludes the spread of most elements of the
western European fauna into North America, and it has only been
due to the relatively recent facilities, afforded by maritime
communications, that certain species of insects have become
accidentally transported across that barrier. The latter event is
one of outstanding importance, since it demonstrates that once
topographical barriers are overcome, some species of insects find
other climatic areas at least as favourable for their colonisation as
their lands of origin. The significance of this fact is becoming
increasingly appreciated, and is witnessed in the rigidity with

252 SOME ASPECTS OF ECOLOGY

which quarantine regulations are being drawn up and enforced in
progressive states. While probably no quarantine regulations
can be devised that will wholly effect the purpose intended, their
maintenance has served to check or delay the introduction of
many noxious insects, which would have readily entered a country
in the absence of such restrictions.

The strict maritime quarantine upheld at San Francisco, and
other Californian ports, appears to be responsible for the effective
exclusion of the Mediterranean fruit-fly (Ceratitis capitata) from
that state, and the freedom which the whole North American
continent has enjoyed from this widespread pest must be largely
attributable to similar measures. In Aj^ril, 1929, an infestation
of the fruit-fly was discovered in part of Florida. Fortunately
the application of effective measures, involving the expenditure
of a large sum of money, resulted in its complete eradication.

The original home of the Mediterranean fruit-fly was probably
tropical Africa : from that region it has become very widely
spread, probably mainly through fortuitous agencies of commerce.
There appears to be no evidence that the insect will become a
serious pest in any country where the mean monthly temperature
falls to, or below, 10° for three or four consecutive months (Back
and Pemberton, 1918). As Verguin (1928) has pointed out, if all
the areas known to be infested by this insect be plotted on a
world map, they are found to lie almost entirely between the
January isotherms of 10° in the two hemispheres. Outside these
limits it appears to occur constantly only in territory bordering
on the northern Mediterranean coast. It has been reported as
causing damage in the environs of Paris, but without becoming
regularly established, and in this district it is stated to be capable
of passing through only two generations per year as compared
with four on the Mediterranean littoral. The available informa-
tion respecting this insect all points to temperature being the most
important of the climatic factors concerned : the optimum
conditions are a warm and more or less equable climate which, at
the same time, allows of a continuous supply of some or other of
the host-fruits being available throughout the year. In Europe
the insect is largely checked by winter temperatures which

1

GEOGRAPHICAL DISTRIBUTION 253

likewise preclude the fruiting of suitable hosts for long periods. It
would appear, therefore, from these remarks that the insect
exhibits a greater tolerance of heat than of cold, and that it is a
species incapable of hibernation in the usual sense of the term.
In the Hawaiian Islands, for example, the development of the
insect progresses most rapidly when the mean temperature
exceeds 23-8° to 26-1°. Back and Pemberton state that complete
mortality follows continued exposures to temperatures below 10°,
and that no immature stages can withstand subjection for two
weeks at freezing point. On the basis of the foregoing remarks,
inferences may be drawn with respect to the possibility of the
insect extending its range into lands free from its presence. The
recent incidence of the pest in Florida points to the liability to
infestation of all the southern states of North America where
suitable climatic conditions prevail, and also of Mexico and
Central America. The West Indies, East Indies and many of
the Pacific Islands, to mention but a few other areas, afford ideal
climatic conditions, and their greatest danger lies, perhaps, in the
unintentional transfer of infested fruits carried by travellers
enjoying the facilities now available for world tours.

Zwolfer (1934), in Germany, has made a special study of the
temperature relations of the different instars of the nun moth
(Lymantria monacha), together with the regional distribution of
this insect over Europe. He has determined the thermal constants
in day-degrees and the threshold of development for the egg,
the six larval instars, pupa and imago. Since the complete
development within the egg occurs in the autumn, no growth
occurs during hibernation in this stage. Beginning with the first
larval instar and ending with the egg stage, the thermal constants
and thresholds of development (in brackets) are stated to be as
follows : 65(4-9) : 217(3-2) : 84(5-7) ; 84(7-2) : 90(7-6) : 132(7-8) :
197(6) : 130(8-4) : 240(6-8). The thermal constant for the total
development of this insect is given as 1,240 ^ 40 day-degrees.
Zwolfer further mentions that the stages up to the fourth larval
instar lend themselves best for control measures, and that their
appearance can be calculated by means of experimental data
coupled with the use of mean monthly temperatures. He states

254 SOME ASPECTS OF ECOLOGY

that a " nun moth temperature index " may be obtained for a
given locahty by dividing the sum of monthly temperature
indices for the months in which development has gone on by the
thermal constant for the total development. The monthly
indices are obtained by subtraction of the threshold of
development from the mean monthly temperature and by the
multiplication of the remainder by the mean number of days in
that month during which any particular stage of the insect occurs.
If the temperature index for the month is equal to or in excess
of 1, then the existence of the insect is, at any rate, theoretically
possible. If it be less than 1, then the total temperatures in a
given locality are below the limit required to allow of the develop-
ment of a complete generation. By means of this index Zwolfer
has provided maps showing the world range of L. monacha, and
these are in general accord with what is known relative to its
actual distribution. These maps further indicate where the
insect may be expected to occur in Asia, where its range is very
little known. Typical nun moth areas or regions, it may be
added, have a temperature index ranging from 1-1 to 1-4. Zwolfer's
ideas have so far been very little tested, and there are not many
species of insects for which the necessary data are available.
The codling moth, Mediterranean fruit-fly and the Dytiscus
beetle, however, may be indicated as being species with reference
to which his method might be applied.

Davidson (1934) has studied the distribution of the Collembolan
Sminthurus viridis in Australia in relation to climatic factors.
The regions favourable to the insect and where it may become
established are determined primarily by rainfall and temperature.
The range of mean monthly temperatures most favourable
to the insect lies between about 11° and 16°. This author
has shown that these temperature conditions are intimately
connected with those of moisture. The ratio of mean monthly
rainfall to evaporation as determined by the saturation deficiency
of the air is a guide to the months in which the moisture will be
favourable for population increase. Where the rainfall evaporation
ratio falls below a value of 1, fluctuating moisture conditions will
inhibit multiplication. When this ratio is much in excess of 1,

GEOGRAPHICAL DISTRIBUTION 255

the wetness will likewise restrict population increase. By
combining data respecting temperature and moisture in con-
junction with the prevalence of soils of the right type, regions
favourable to the incidence of the insect appear to be tolerably
clearly explained.

The life-history of the Colorado beetle {Leptinotarsa decemlineata)
affords an interesting ecological study. Its original home was
in the Rocky Mountains region, where its native food-plant is
the buffalo bur (Solanum rostratum). Following the westward
march of civilisation in North America came the potato, and the
insect found in this new food-plant one admirably suited to its
economy. The subsequent history of the species shows that it
has extended its range over a vast territory, which it did not
previously inhabit, owing to the presence of the new plant-host.
The insect was detected, for example, in Nebraska in 1859, and by
1874 it had reached the Atlantic coast. On the other hand, its
spread towards the Pacific slope has been very slow, owing to the
barrier of the Rocky Mountains : it did not occur in British
Columbia until 1919, and its presence in the states of Washington
and Oregon is likewise comparatively recent, and there does not
appear to be any evidence that it crossed the mountain range
by its own unaided activities. It would seem that the original
restricted distribution of this insect has to be ascribed largely to
biological causes and not to climatic influences. Once suitable
food-plants, such as the potato, tobacco and others, became
available in new areas, the impetus to spread was imparted. The
history of its diffusion shows that climatic factors have played a
relatively minor part in the process, since its range to-day is from
Mexico in the south to Quebec in the north. This territory lies
between the North American January isotherms of 10° and — 12°
and between the July isotherms of 30° and 20°. The annual
rainfall over the whole infested area varies from about 15 inches
in the north-west to some 60 inches in the south. It would appear
obvious, therefore, that the Colorado beetle can withstand a wide
range of climates before meeting such extremes as are likely to
become effective as limiting factors. In this connection its
accidental entry into France in 1922 is of particular importance.

256 SOME ASPECTS OF ECOLOGY

After becoming established first in the Gironde, it has steadily
extended its range, notwithstanding the application of measures
of repression. The result is that it has now spread over the
greater part of France and Belgium ; it has also reached the
Swiss frontier and entered the Saar region of Germany. In 1933
a small outbreak was discovered in the Tilbury-Gravesend district
of England, but measures taken by the Ministry of Agriculture
resulted in its eradication. It is, however, unlikely that England
will remain free from outbreaks indefinitely in view of the extent
to which the Colorado beetle is multiplying and spreading on the
continent of Europe. A species so prone to migrate will most
likely find a ready means of ingress afforded by cross-Channel
steamer traffic, even if it does not often succeed in crossing the
intervening strip of sea by flight. The insect now constitutes a
real and potential menace to potato cultivation over part of
Western Europe, and no climatic factor prevailing in this area
appears very likely to restrict its further dissemination (vide
Fryer, 1934, 1935).

The foregoing examples have been selected in order to illustrate
the relationship of climate to the spread of insect pests. Although
emphasis has been laid upon the temperature factor, it scarcely
requires pointing out that the range of a given insect is governed
by a complex of climatic and other environmental influences.
Such a combination of factors may act in unison over the greater
part of the range of a given species, but when any one of them
approaches the limit of toleration that factor will be the major
controlling influence.

With the advent of highly developed communications by land,
sea and air it has become no longer sufficient to take cognisance
only of the existing distribution of insects known to be pests :
the zones of their possible distribution cannot be wholly neglected,
at least by the applied entomologist. An investigation of the
climatic conditions prevailing over the area covered by the existing
range of a given species will enable certain general conclusions to
be drawn relative to its possible spread. The influence of the
individual factors upon that species can, to some extent, be
subjected to analysis by adequate laboratory experimentation.

CLIMATE AND WEATHER 257

The technique of the latter, it may be added, has greatly improved
in recent years, and is dealt with in the recent books by Chapman
(1931) and Shelford (1929). It will, therefore, become evident
that two parallel lines of inquiry are desirable, one applicable to
actual field conditions and the other to be prosecuted in the
laboratory. Taken in conjunction, they are more likely to advance
knowledge than if laboratory experimentation be pursued without
reference to field conditions ; in other words, the latter provide
the real problems which are to be taken to the laboratory for
analysis.

Climate and Weather in Relation to Seasonal Prevalence. The
relations of climate and weather to seasonal and periodic behaviour
of insects afford more difficult problems than those governing the
general distribution of those animals. In certain parts of the
tropics the climate is extremely uniform throughout the year,
with relatively small fluctuations of temperature around either
the daily or the annual mean. Under such conditions seasonal or
periodic behaviour of insects is but little pronounced. True
dormancy, for example, is almost a non-existent phenomenon,
and insects are able to pass through uninterrupted cycles of
development and activity from year to year. In other parts of
the tropics, where there are sharply contrasted dry and wet
seasons, the former with a prevailing high temperature and the
latter with a cooler temperature, these changes are directly
reflected upon the insect life. The incidence of special dry and
wet season forms, for example, is well known in Lepidoptera, but
the most marked phenomenon is the tendency of the life-cycle in
most insects to be interrupted during the dry season. Heat and
drought, often accompanied by leaf-fall, are prejudicial to insect
life in general, and it is at this period activity becomes arrested
and a condition of aestivation is entered upon. In temperate and
cold regions the most prejudicial period is the winter, when insect
activities attain their minimum and the phase of hibernation
supervenes. Hibernation and aestivation, in the true sense of
those terms, are so deeply impressed upon the species, as the result
of environmental influences affecting many thousands of
generations, that these phases have become definite rhythms

R.A. ENTOMOLOGY. 9

258 SOME ASPECTS OF ECOLOGY

expressing certain physiological conditions. Neither phase can be
readily engendered out of season by modification of temperature
or other physical influences, nor can they be easily dispelled and
activity reinduced by similar means. It is not proposed to deal
here with established rhythms in insect life, but to refer to climatic
factors in so far as they are correlated with fluctuations in the
abundance or scarcity of insects which recur at more or less
irregular intervals.

It is a matter of common observation that there is a more or less
strongly marked fluctuation in the prevalence of certain species
of insects taken over a period of years. Probably few species are
exempt, in temperate regions at any rate, from such rises and falls,
and the literature of entomology contains frequent reference to
such occurrences. In the case of noxious insects, upon which
most observation has been concentrated, the majority occasion
what may be described as average damage during ordinary
seasons, while now and again mass outbreaks of variable duration
occur. Diverse causes have been attributed to such outbreaks,
but it needs to be emphasised how difficult it is in many cases to
definitely prove that any observed factor is actually responsible.
Much exact and prolonged observation is required before it will
be possible to discriminate between the direct effects of weather
upon insect prevalence, as distinct from its indirect influence,
operating through such potent biological factors as food-plants,
parasites, predators and disease organisms of various kinds. In
certain instances tolerably close correlation has been observed
between weather conditions and the prevalence of individual
pests. Examples of such phenomena will be found in text-books
by Wardle and other writers, and in the published proceedings of
the Empire Meteorologists' Conference ^ held in London in 1929.
It has been recorded, for instance, that the prevalence of low
temperatures, unaccompanied by snowfall, has a marked effect
in reducing the prevalence of the Colorado beetle in Manitoba :
low temperatures following a previous mild spell in early spring

^ Conference of Empire Meteorologists, 1929. Agricultural Section.
I, Report ; II, Papers and Discussions. Each obtainable, price \s., post
free, from the Ministry of Agriculture and Fisheries, 10 Whitehall Place,
London, S.W.I.

I

LITERATURE 259

appear to be a potent factor in destroying large numbers of newly
emerged aphides : a temperature of 16-6° is said to totally inhibit
oviposition by the codling moth : in North America wet periods
during May and June tend to bring the pale western cutworm to
the surface of the soil and, by thus favouring its biological control,
is usually followed by a decreased infestation. It will thus be
evident how important it is to establish such correlations. Their
main desideratum is the accumulation of sufficient accurate
quantitative and qualitative data to enable the prediction of
insect outbreaks to be made with a degree of certainty that
would warrant the preparation, in advance, of appropriate control
measures.

The collection of information of this kind is possibly best
organised and most advanced in Germany. In a paper by
Schnauer (1929) a very large amount of information, systematically
collected, year by year, through the auspices of the German
Agricultural Society since 1893, is summarised and analysed.
The causes of the outbreaks of major pests are surveyed in con-
junction with data respecting the chief meteorological factors.
The information thus accumulated has enabled it to be possible
to trace the probable connection of fluctuations in the prevalence
of specific pests with particular weather conditions. The data is
now being utilised as the basis for future forecasting.

Literature

Aeschner, 1932. Naturwiss., XX., 50.

1934. Parasitology, XXV., 309.

and RiES, 1933. Zeits. Morph. Okol. Tiere, XXVI., 529.
Alpator and Pearl, 1929. Amer. Nat., LXIIL, 37.
Back and Cotton, 1925. Journ. Agric. Res., XI., 1035.

and Pemberton, 1918. U.S. Dept. Agric, Bull. 640.
Bacot and Harden, 1922. Biochem. Journ., XVI., 148.
Barber, 1929. U.S. Dept. Agric, Circ 71. ^

Belehradek, 1935. " Temperature and Living Matter." Berlin.
Berland, 1934. Soc Biolgeog., Mem. 4, 155.
Berlese, 1906. Redia, III., 8.
Bodenheimer, 1927. Zeits. angew. Ent., XII., 89.
Bodine, 1925. Journ. Exp. Zool., XLIL, 91.
Brues, 1927. Quart. Rev. Biol., II., 181.

and Glaser, 1921. Biol. Bull., XL., 299.
Buchner, 1921. " Tier und Pflanze in intrazellular Symbiose." Berlin.

1928. " Holznahrung und Symbiose." Berlin.

260 SOME ASPECTS OF ECOLOGY

Buxton, 1924. Proc. Roy. Soc, B., XCVL, 123.

1930. Ibid., CVL, r)(iO.

1931. Bull. Ent. lies., XXII., 431.
1935. '' Insects of Samoa," IX., 43.

and Lewis, 1935. Phil. Trans. Roy. Soc, B., CCXXIV., 175.

and Mellanby, 1934. Bull. Ent. Res., XXV., 171.
Carothers, 1923. 2'rans. Ent. Soc., XLIX., 7.
Chapman, 1924. Journ. Gen. Pfiys., VI., 5G5.

1931. " Animal Ecology." London.
Cleveland, 1924. Biol. Bidl., XLIV., 177.

1925. Ibid., XLVIIL, 282, 289, 295, 309, 455.

and others, 1934. Mem. Am. Acad. Sci. and Arts, XVII., 185.
CoAD, 1931. Yearbook U.S. Dept. Agric, 320.
Cook, 1927. Journ. Econ. Ent., XX., 769.
Davidson, 1934. Commonwealth of Australia, Dept. Sci. and Indus. Res.,

Bull. 79.
Davies, 1928. Brit. Journ. Exp. Biol., VI., 79.
Dean, Cotton and Wagner, 1936. U.S. Dept. Agric, Circ. 390.
Dendy and Elkington, 1920. Reports Grain Pest Comm. Roy. Soc, 7.

London.
Evans, 1936. Trans. Roy. Ent. Soc, LXXXV., 363.
Frew, 1929. Ibid., VI., 2.05.
Fryer, 1934. Journ. Min. Agric, XL., 907.

1935. Ibid., XLL, 1058, 1089.
Glaser, 1923. Journ. Exp. Zool., XXXVIII., 395.
Griswold and Crowell, 1936. Ibid., XVII., 241.
GuNN, 1934. Zeits. vergl. Phys., XX., 617.
Headlee, 1917. Journ. Econ. Ent., X., 31.

1921. i6irf.,XIV., 264.
Hefley, 1928. Ibid., XXL, 213.

Hunter and Pierce, 1912. Bur. Ent. U.S. Dept. Agric, Bull. 114,
Krogh, 1914. Zeits. fur allg. Physiol., XVI., 178.

1916. " Respiratory Exchange of Animals and Man." London.
Lloyd, 1920. Ann. App. Biol., VII., 66.
Ludwig and Cable, 1933. Physiol. Zool., VI., 493.
Mansour and Mansour-Bek, 1934. Biol. Rev., IX., 363.
Marcovitcii, 1924. Journ. Agric Res., XXVIL, 513.
Mellanby, 1935. Biol. Rev., X., 37.
Northrop, 1926. Journ. Gen. Phys., IX., 81.

1926a. Ibid., 319.
Payne, 1926. Quart. Rev. Biol., I., 270.

1929. Ann. Ent. Soc. Am., XXII., 601.

1930. Journ. Econ. Ent., XX., 80.
Peairs, 1927. Bull. 208, West Virginia Exp. Sta.
Pierantoui, 1910. Zool. Anz., XXXV., 96.
PiRSCH, 1923. Journ. Agric Res., XXIV., 275.
Ramsay, 1935a. Journ. Exp. Biol., XII., 355.

1935b. Ibid., 373.
Richardson, 1926. Journ. Agric. Res., XXXII., 895.
RiES, 1932. Zeits. Morph. Okol. Tiere, XX., 233.
Ripper, 1930. Zeits. vergl. Physiol., XIII., 313.
Robinson, 1926. Minnesota Agric Exp. Sta. Tech. Bull. 41.

1927. Journ. Econ. Ent., XX., 80.

1928. " Colloid Symposium Monograph," V., 199.
1928a. Journ. Agric Res., XXXVII., 749.

LITERATURE 261

RooNWAL, 1936. Bull. Ent. Res., XXII., 431.

Sacharov, 1930. Ecology, XL, 505.

Salt, 1936. Univ. Minnesota agric. Exp. Sta. Tech. Bull., 116.

Sayle, 1928. Quart. Rev. Biol., III., 542.

ScHNAUER, 1929. Zeits. angezv. Ent., XV., 565.

SCHRADER, 1923. Biol. Bull., XLV., 279.

Shelford, 1927. Bull. Illinois Nat. Hist. Survey, XVI., 311.

1929. " Laboratory and Field Ecology." London.
Simmons and Ellington, 1925. Journ. Agric. Res., XXX., 845.
Taylor, 1927. Journ. Morph., XLIV., 313.
Tracer, 1932. Biochem. Journ., XXVI., 1762.
Uichanco, 1924. Philipp. Journ. Sci., XXIV., 143.
UvAROv, 1928. Trans. Ent. Soc, LXXVIL, 255.

1931. Ibid., LXXIX., 1.
Verguin, 1928. Rev. Zool. Agric. et Appl., XXVIL, 141.
Wadley, 1931. Ann. Ent. Soc. Am.., XXIV., 325.
Weinland, 1906. Zeits. f. Biol., XLVIL, 186.
Werner, 1926. Zeits. Morph. Okol. Tiere, VI., 150.
Wigglesworth, 1932. Quart. Journ. Mic. Sci., LXXV., 131.

1936. Parasitology, XXVIIL, 284.
ZwoLFER, 1934. Zeits. angezv. Ent., XXL, 333.

CHAPTER X
THE PRACTICAL APPLICATION OF ECOLOGY

Regulation of the Growth of the Crop. Time of Sowing, p. 263 ;
Crop Nutrition and Soil Conditions, p. 267. Resistant Varieties,
p. 276.

There are many insect pests affecting agriculture which do
not lend themselves to direct control measures with any real
prospects of success. Or, the crop may be of such an extent or
character which renders such measures impossible from the
practical standpoint. Difficulties of this kind assert themselves
wherever agriculture is practised, but more especially in the vast
areas of newly developed country inhabited by backward races,
where insecticidal measures are usually out of the question. In
these contingencies, biological control may prove applicable in
certain cases, and this subject is fully discussed in Chapter XIV.
On the other hand, there are numerous instances where biological
control does not afford promise, or cannot be attempted. Under
such conditions control methods largely resolve themselves into
devising modifications of cultural practice as will place the pest
at some disadvantage and, at the same time, ensure a better crop
yield. In formulating such measures an acquaintance with the
chief ecological features in the life of a given pest in relation to
its plant-host is called for. Knowledge so gained, when translated
into practice, resolves itself into what may be termed cultural
methods of control.

Cultural methods of control have definite practical advantages
as contrasted with physico-chemical methods. In the first place,
they are, as a general rule, preventive, and serve to restrain losses
that would otherwise supervene. Further, in the case of most
crops, they usually incur little or no extra financial expenditure
upon the practical grower, they are generally easy to apply, and

GROWTH OF THE CROP 263

frequently involve no drastic disturbance of accepted routine.
It also has to be remembered that the present trend of agricultural
conditions in many countries is such as leaves a narrowing margin
of expenditure available for insecticidal treatment, or other more
or less costly measures. There is, in consequence, an increasing
tendency to study the relations between insect attack and the
conditions under which the crops are grown. This in its turn
involves the application of ecological methods, and in the present
chapter endeavour will be made to show that success, in this
direction, is dependent upon adequate knowledge of the biology
of the insect in relation to the growth phases of the crop which it
attacks.

In the present instance the application of ecological methods
will be discussed under four aspects. (1) The regulation of the
growth of the crop ; (2) the production of varieties of plants
immune, or highly resistant, to insect attacks ; (3) the bearing of
host selection by insects, and the evolution of biological races of
the latter ; and (4) the ecological aspects of the locust problem.
It is in relation to these four problems that ecology, in the broad
sense of the term, enters into closest relations with practical
entomology to-day.

Regulation of the Growth of the Crop

This procedure aims at regulating the growth of the crop in
such a way that it either escapes infestation or is subjected to
appreciably lighter attack. In a general way, two methods are
applicable. Firstly, the adoption of an early sowing date in
order to ensure a given crop is sufficiently advanced in growth to
withstand attack at the usual period of the latter. Secondly, the
investigation of the nutritional requirements of the crop in order
to accelerate its growth at its most vulnerable period, or to
promote its growth in such a manner that it becomes less
susceptible to infestation.

Time of Sowing. In the case of certain cereals time of sowing
of the crop has a very direct bearing upon their subsequent
liability in relation to insect attack. With the frit-fly (Oscinella
frit) the earlier the date of sowing in spring the better, providing

264 THE PRACTICAL APPLICATION OF ECOLOGY

cultural conditions and the prevailing weather admit of such a
procedure. Evidence gathered from many sources shows that
late sown spring crops suffer most of the attacks of this insect.
The frit-fly is so well known as a pest of oats, and also of other
cereal crops, that a detailed account of its biology is unnecessary.
In a few words, it may be said that, under English conditions, the
flies of the first generation oviposit on the leaves or stems of spring
oats and on various grasses ; they attain their greatest abundance
about May 26th.i The larvae bore into the shoots, causing the
death of the central leaves and growing point. Flies of the second
generation oviposit in largest numbers about July 15th ^ on the
ears of oats, and the larvae feed upon the spikelets and young
grain. Oviposition by the third generation of flies occurs during
August and September or later, the eggs being laid on the leaves
of winter cereals and various grasses ; the flies are most numerous
about August 22nd. ^ Winter is passed in the larval condition at
the bases of the shoots, which are ultimately destroyed, and the
over-wintering larvae give rise to the first generation of fl-ies of the
next year.

Oat plants in the shoot and panicle stages have definite phases
of growth within which they are most susceptible to attack.
The shoot is most susceptible when in the stage bearing two or
three leaves ; once it has attained the four-leaf stage, liability to
infestation declines rapidly (Cunliffe, Fryer and Gibson, 1925;
Cunliffe, 1928). The grain is most susceptible about the time of
fertilisation, and the period of greatest liability terminates before
the grain attains half its normal size (Fryer and Colhn, 1924).

With varieties of oats at present under cultivation the provision
of an adequate growing period, to ensure that the plants reach the
four-leaf stage sufficiently early, depends upon the date of sowing.
It is only practicable with spring oats by eliminating sowings
later than the end of February, or in some years the middle of
March. Adverse soils or climatic conditions may easily entail a
set-back in plant growth, and this necessitates a safety margin,

1 These dates appear to indicate a three-cycle rhythm in the annual life of
the insect, and were observed to hold good during six years' observations
{vide Cunliffe, 1929).

TIME OF SOWING

265

probably of about two weeks, which requires a total of ten weeks
from the sowing date to the appearance of the plants in the
resistant stage. Early sowing, furthermore, reflects its advantages
upon reduced grain infestation, since it tends to the early exsertion

22

-

• '

21

-

20

-

19

-

18

-

17

-

16

-

15

-

14

-

•

•

13

-

1 ^2

-

>

-•.

m 11
S 10

-

/

•^*^v •

o

/

° 9
8
7
6
5
4
3

■

•

/
/

/

/ •

•

2

-

i

1

-

#

0

1 I

1 1 1 1 t

1

6 13

20 27

6 13 20 28 4

10

17 24 1 8 15 23 29

Feb.

Mar.

Apr. May

Fig. 73. Graph showing correlation between dates of sowing of barley
and degrees of subsequent infestation -by gout-fly. (After Frew.)

of the panicles. Since the majority of the second swarm of flies
appear about mid-July, the risk of heavy grain attack is often
obviated, since by that time the grain should be half-way towards
maturity and therefore approaching its resistant stage {vide also
Cunliffe, 1929).

266 THtJ PRACTICAL APPLICATION OF ECOLOGY

Observations conducted by Frew (1924) at Rothamsted likewise
show that time of sowing has a direct bearing upon the infestation
of barley by the gout-fly (Chlorops tceniopus). The comparative
immunity of early sown spring barley from the attacks of this
insect has been noted by previous observers, but apparently
without any experimental work having been conducted. In
Frew's experiments an area as uniform as possible was divided
into a series of plots which were sown with barley at successively
later dates. Although Frew's experiments were hampered by
heavy soil conditions, they nevertheless demonstrate the
advantages to be derived from sowing of spring barley before the
end of March (Fig. 73). Reference to the graph shows that
exceptionally high infestation figures were obtained for three of
the plots, which appear to have been due to retarded plant growth,
owing to the caking of the soil on the particular areas concerned.

The ordinary culinary varieties of the broad bean ( Vicia faba)
are well known to suffer from heavy attacks of the aphid Aphis
rumicis over most parts of Europe. It has been previously
pointed out on p. 241 that this insect finds ideal conditions for
prolific agamic production on the growing, sappy, terminal shoots
of that plant. It has long been the practice of growers to sow the
seed as early as possible in the season, or even in the preceding
autumn. The advantage gained is in the advanced condition of
growth attained by the crop by the time the aphid becomes
prevalent. From among experiments conducted by Davidson at
Rothamsted, the reproduction of aphides on broad beans sown on
March 15th and April 27th was compared. In the former case
the mean number of young produced at the end of fourteen days
by single apterous females was 548, while on the later sown plants
the corresponding reproductive figure was 1,341.

In 1918 Hopkins, in the United States, advanced the hypothesis
now generally referred to as the Bioclimatic Law. His contention
has a phenological basis, and it formulates the rule that such
events as the opening of the buds of a particular plant, dates of
appearance of species of insects, first dates of flowering, and other
phenomena, occur in North America according to a more or less
definite law. Phenological events of this kind occur later in

TIME OF SOWING 261

spring and earlier in autumn if they be followed in northward, east-
ward or vertical directions. The law states that the variation in
time is at the rate of four days for each degree of latitude, for each
five degrees of longitude and for each 400 feet in altitude. The
hypothesis has a purely empirical basis, and was originally founded
upon data afforded by certain beetles destructive to forests.
Subsequent studies by Hopkins (1919) and other observers have
led to its application to other biological phenomena, the most
important being in relation to the time of incidence of the autumn
generation of the Hessian fly (Mayetiola destructor). In the case
of this insect time of sowing of wheat has a very definite bearing
in relation to subsequent attack. The method is to sow wheat
sufficiently late, so that it will appear above ground in the
susceptible stage, at a time when the second generation of the
Hessian fly has died out. At the same time the crop must be
sown early enough to have attained sufficient growth to enable it
to withstand the winter. Based on the bioclimatic law, the
critical dates for sowing range from about September 10th in the
north-east to October 25th in the south-west. It has been found
that the actual emergence dates of the insect, taken over a very
wide area, all appear to fall within one week of the calculated
time under the bioclimatic law. Flint and Larrimer (1928) have
examined the problem in some detail with respect to the State of
Illinois. They conclude that under normal weather conditions
the relatively " fly-free " date is September 15th in the northern-
most counties in the state, and October 9th for the southernmost.
They present experimental data which show, without exception,
that wheat heavily infested with Hessian fly gives appreciably
lower yields than wheat grown during the same year, and in the
same field, but sown later according to the law enunciated.

A " law " of so comprehensive a character has its limitations
and is based upon climate, and not upon weather. Due allowance
has to be made for local fluctuations of the latter, and exceptional
seasons. West of the Rocky Mountains it breaks down, and it
does not appear to hold good for the conditions prevailing over
continental Europe.

Crop Nutrition and Soil Conditions. The problem of regulating

268 THE PRACTICAL APPLICATION OF ECOLOGY

the nutrition of a given crop in relation to its susceptibility to
insect attack is one in which the entomologist becomes largely
dependent upon knowledge derived from other sciences. The
nutritional requirements of different crops form in themselves a
specialised branch of plant physiology, and they are intimately
associated with questions of soil conditions, rainfall, and other
factors. Recent research is bringing to light evidence which
points to the conclusion that excess or deficiency of certain food-
materials required by the plant betrays itself in the response the
plant exercises towards insect injury. Similarly, particular soil
conditions in some cases appear to conduce to heavier damage
resulting from insect attacks than when plants are grown under
different conditions. The future is likely to see increasing
attention being given to the cultural conditions under which
certain crops, particularly liable to insect infestation, are grown.
The possibility is, therefore, opening out of ensuring a measure of
insect control by regulating the nutrition of crops in such a way
that they become less liable to suffer severe injuries.

1. The problem of froghopper ^ damage to sugar cane in
Trinidad has become more one for the plant physiologist and the
agricultural specialist than for the entomologist. The injuries
caused by the sucking propensities of the insect appear to be
intimately connected with the internal physiological condition of
the cane ; the crop does not necessarily suffer the greatest damage
when the froghoppers have been most numerous in attack. The
important observation that there is a marked difference in the
degree of attack resulting on cane growing in red acid soil and that
growing in alkaline black-marl soil in certain localities, ^ in that the
former suffers severe injury whilst the latter does not, has opened
up the whole question of the nutrition of the cane-plant in this
connection. It was found that a positive correlation prevails
between the degree of saturation of soils with lime and cane
susceptibility. It appears that soils of areas free from severe
injury by the froghopper are absorbtively saturated by lime, and

1 The species involved is Monecphora (Tomaspis) saccharina, family
Cercopidae.

2 The greatest number of eggs also appears to be laid in heavy acid soils,
and the fewest in black marl soil.

CROP NUTRITION 269

those regularly affected by the insect are less than 60 per cent,
saturated. There seems to be no correlation between gross lime-
content and susceptibility to insect damage, and there is stated to
be little doubt that the soils of areas liable to severe froghopper
attacks are lacking in combined lime (exchangeable calcium), and
therefore usually give a definitely acid reaction. The reason why
damage prevails under these conditions has not so far been
elucidated. Both Williams (1921) and Withycombe (1926) have
advanced the theory that the severest injury is largely dependent
upon the water-relations of the crop. P'ield observations,
subsequently carried out, show that differences between the acidic
and alkaline areas are reflected in the moisture contents of the
leaves of the cane-plants growing on these areas. In the
unfavourable areas the transpiration rate of the canes is increased,
which in itself leads to rapid drying-out during the afternoon and
evening, which is the time when the froghoppers feed. It has
been further noted that a field survey of canes, growing in soils
known to favour froghopper damage, yielded evidence that
marked amounts of iron and aluminium accumulate in their
nodal tissues. Toxic aluminium, for example, often begins to
appear when the degree of lime saturation falls below 60 per cent.,
and this may markedly increase the plants' susceptibility to
insect attack.

The foregoing empirical observations serve to indicate how
closely susceptibility to froghopper damage is correlated with
factors relating to the soil and to plant nutrition ; for further
information the reader should consult the Froghopper Committee's
publications (1925 onwards).

2. The extensive observations conducted by Frew (1924), at
the Rothamsted Experimental Station, have a direct bearing on
the influence of the nutrition of barley upon attacks by the gout-fly
(Chlorops tamiopus). The flies commence to appear in May, and
deposit their eggs on the leaves of young spring-sown barley.
They are usually laid singly, on the upper side of a leaf, and it is
rarely that a shoot bears more than one egg. Upon eclosion the
larva bores its way downwards into the centre of the shoot, and
eats a groove along one side of the developing ear and to the

270 THE PRACTICAL APPLICATION OF ECOLOGY

internode below it. A typical attack, at this period, shows a stem
with two or three internodes, terminated by a swollen region,
consisting of the ear enclosed by its ensheathing leaves, from
which it never escapes, and therefore cannot ripen. A striking
feature in the behaviour of the insect is its instinct always to bore
downwards. Should the fly deposit its egg upon a leaf, whose base

Fig. 74. Diagram of barley shoots, showing C, " critical " leaves ;
|C, "half critical" leaves; E, ear; and NC, "non-critical"
leaves. (After Frew.)

comes off from the shoot helow the ear, the newly hatched larva
cannot reach the ear to feed ; such leaves may be termed
" non-critical leaves." In thirty-six instances dead larvai were
discovered by Frew in the ear-bearing internodes, and in all cases
the ear had remained untouched. In these examples the young
larvae entered the shoots below the ear, and, owing to the toughness
of the tissue and lack of suitable nourishment, they had perished.
If, on the other hand, an egg is laid on a leaf which arises from the
shoot above the base of the ear, the ensuing larva, by migrating

I

CROP NUTRITION 271

downwards, reaches the ear ; such leaves may, therefore, be
termed " critical leaves." A leaf arising from a shoot half-way
up the ear is to be regarded as a " half critical leaf," since the
larva would only be able, at most, to attack the base of the
ear (Fig. 74).

The permanent barley plots, on the Hoos Field at Rothamsted,
have yielded significant data indicating that the summer infesta-
tion by gout-fly shows a wide range of intensity, according to the
manurial treatment adopted. Frew's observations indicate that
manurial treatment affects infestation by altering the early growth
of stem and ear, and, consequently, by altering the number of
" critical leaves " present at the time the insect is egg-laying, and
during the period of incubation of eggs already deposited. Field
counts showed that the number of critical leaves per shoot was
lowest on plots least affected by the insect, and vice versa. Two
years' observations, conducted on the twenty-eight plots of the
Hoos Field at Rothamsted, involved the counting of over 60,000
separate barley shoots, and the noting of the numbers infested by
gout-fly in each case. It became evident, as the result, that plots
receiving superphosphate, farmyard manure or complete mineral
fertilisers,! were remarkably constant in showing the lowest
percentage of infestation. Untreated (control) plots were heavily
infested, while those treated with nitrogenous fertilisers indicated
that the latter, in some cases, exercise a beneficial effect, but that
heavy dressings do not conduce to reduction of infestation, and
may increase it through retarding the growth of the ear. The
application of nitrogenous fertilisers, in conjunction with either
superphosphate or complete minerals, however, proved markedly
beneficial in each case.

3. A further example of the influence of differential nutrition
of crops, in relation to insect damage, is afforded by the
experiments of Jepson and Gadd (1926) in Ceylon with reference
to the shot-hole borer (Xyleborus fornicatus) of the tea plant. This
insect injures about 80 to 90 per cent, of all tea bushes growing
below an elevation of approximately 4,000 feet. The insect does

1 These comprise sulphate of potash, superphosphate of lime, and sodium
and magnesium silicates.

272 THE PRACTICAL APPLICATION OF ECOLOGY

not limit its attacks to any special season, and the bushes, except
for a brief period of respite after each pruning, are liable to
continuous infestation from the time of planting onwards. The
injuries entailed are caused by the tunnelling of the beetles, and
their larvae, in the stems and branches of the bushes. Most
experimental work in the past has been devoted to applying
insecticidal measures of control ; some have had obvious practical
disadvantages, and most of them have proved costly, since
treatment requires to be directed against virtually every bush
on an infested estate.

An early observation made by Green, relative to the prevalence
of healed-over gallery entrances in well-cultivated bushes, led to
an investigation of the subject upon an experimental basis. The
most obvious damage caused to tea bushes by the attacks of the
shot-hole borer is the loss of branches, which break off at spots
weakened by the galleries of the insect. High winds, and the
movements of coolies among the bushes, are together responsible
for heavy destruction of this kind. It appeared feasible, therefore,
to discover whether it were possible to so improve the vigour of
the bushes as to induce rapid healing over of the gallery entrance
holes or " shot-holes," and so lessen the tendency of the branches
to break.

After preliminary trials, field experiments were carried out at
Peradeniya during the three years 1923-1925, some forty plots
being utilised for the purpose. The treatments included two
experiments with nitrogenous manures, two with potassic, two with
phosphoric acid, and also one with lime. There were, including
controls, eight treatments, each being repeated on five separate
plots. The main result of the experiments was shown in the
accelerated healing of the galleries caused by the shot-hole borer,
which was most marked in those plots treated with nitrogenous
fertilisers. Healing was completed on these plots in 2-9 months,
as compared with 3-75 months in the controls. The authors
conclude that suitable manuring of tea bushes offers a ready and
satisfactory means of diminishing the effects of attack by the insect
in question, and that it should aim at increasing the nitrogenous
content of the soil.

CROP NUTRITION 273

4. In connection with tea pests, mention needs to be made of
the now well-known experiment by Andrews (1923), who showed
that a high ratio of available potash to available phosphoric acid
in the soil, in which tea bushes are growing, greatly reduces attack
by Helopeltis theivora. The latter is a Capsid bug, and a persistent
enemy of tea, against which spraying appears to be an
impracticable proposition. Andrews found that where a constant
supply of soluble potash is directly applied to the roots of the
bushes, this substance is absorbed and the bushes are freed from
the pest, remaining free from attack the rest of the season. This
would lead to the obvious conclusion that the addition of a potassic
fertiliser to the soil was needed. Success, however, cannot be
regularly obtained by such treatment, for the reason that the
plants are apparently unable always to utilise the added potash.
The effect of the potash in manurial experiments is transient, and
some factor, operating to cause the original shortage in available
potash obtained by the plant, intervenes to prevent its continued
absorption. Andrews' inquiry showed that the composition of the
leaf of the tea plant shows differences corresponding with those
deduced from observation of the soils, and that comparative
immunity from attack accompanies an increase in the proportion
of potash, as compared with phosphoric acid, in the leaf. The
factors controlling this ratio, however, are not understood, since
the chemical composition of the leaf may be quite different when
grown in the same soil under different cultural conditions, with a
resulting difference in liability to attack by Helopeltis. It would
appear, therefore, that before practical application can be made of
this discovery, an inquiry into the behaviour of the substances
mentioned in the soil under different conditions is called for.

5. In South Australia the springtail Sminthurus viridis has,
during the last forty years, become one of the most important pests
of field crops in that country, particularly to lucerne and clovers.
The damage is caused by the insect gnawing the epidermis and
underlying mesophyll of the leaves of the young plants, giving a
field a scorched appearance as the result of a bad attack.

This insect has recently been the subject of a comprehensive
ecological study by Davidson (1934), while earlier contributions

274 THE PRACTICAL APPLICATION OF ECOLOGY

have been made by Holdaway (1927) and by Maclagan (1932).
It has been shown by these three observers that the soil reaction
has a marked effect upon the number of eggs laid. Soil moisture
is the dominant factor affecting oviposition, and this needs to be
considered in conjunction with pH values of the soil. Below
10 per cent, soil moisture the number of eggs laid decreases, and
where the moisture approaches saturation it also becomes
unfavourable. Davidson finds that above 12 per cent, moisture
normal egg batches are produced. The optimum pH value is
about 6-5, but the possible range for free oviposition extends
between about 5-5 to 7-0. Since the insects ingest considerable
quantities of soil, it seems that in passing through the gut the
soil would influence the normal ^^H of the digestive tract and so
affect general metabolism. In view of the economic relationship
between Sminthurus and leguminous crops, it may be pointed out
that lucerne is intolerant of acid conditions and prefers a pYL of
over 7-0, while clovers exhibit a greater tolerance to acidity than
lucerne. Whether the soil ^^H can be so altered from what appears
to be an optimum for the insect, and maintained within a range
favourable to the production of leguminous crops, appears to be
doubtful. Davidson remarks that, considering the lime require-
ment of the soils of many infested areas, it would be impracticable
to raise them to a pH value unfavourable for the Sminthurus.
Such a method, it appears, would have very restricted application
under Australian conditions. On the whole, it seems that pasture
management and the possibility of biological control by the
introduction of suitable predators offer better prospects for the
exploration of measures of control.

6. French viticulturists have observed that infestation of
vines by the Phylloxera is accompanied by the most severe injury
where the plants are growing on certain types of soils. In other
types of soil, even in the same vine-growing district, injuries
are commonly much less severe. The correspondence between
the two series of phenomena appeared to be sufficiently close to
be regarded as an axiom. Some years back, the French vine-
growers came to the conclusion that light-textured or sandy soils,
containing" up to 60 per cent, of silicate sand, contributed to the

SOIL CONDITIONS 275

resistance of the crop to attacks by the insect in question. More
recently the same subject has been re-investigated in CaUfornia
by Nougaret and Lapham (1928), since it appeared possible that
if certain soil types in that state are inimical to Phylloxera
propagation, a certain degree of control might be achieved by the
wider planting of vineyards under such conditions. These observers
found that infestation appears to be particularly frequent and
pronounced in flat or depressed localities, where inferior soil
drainage is aggravated by the occurrence of shallow, compact and
impervious subsoils which readily cake, giving rise to a condition
of shallow-rooted vines and stagnated soil drainage. Where
infestation occurs in deep friable soils, with porous subsoils,
destruction of the vines is less rapid than in more shallow soils.
The texture of the soils appears to be a factor which exerts itself
in a number of ways. It is suggested that in those of a heavy
compact type, which cake, shrink and crack when dry, conditions
are rendered more favourable for the development and migration
of the root-forms of the insect. Facilities for movement, it is
claimed, are afforded by intersecting cracks, and along the courses
of the roots, from which the soil tends to withdraw slightly in
shrinking. If forced to the highly heated surface of sandy soils,
which do not cake and crack, and which may be more uniformly
in close contact with the roots, the delicate and minute insect
would not long survive under the hot arid conditions to which
many of the interior valleys of California are subjected. Various
soil types in relation to Phylloxera infestation are discussed and
the whole area shown in a coloured map. Their bulletin is too
lengthy for detailed discussion here, but it may be said that the
authors conclude that the correlation between susceptibility to
Phylloxera infestation and soil types is a tolerably close one. In
certain areas of light-textured soils it is recommended that
planting can be extended with a wide margin of safety, while
in others, at present free from infestation, the soil types are such
that if the insect became established, immunity from attack
would not probably be wholly maintained. Loams and soils of
heavy texture are favourable to general infestation, and on such
lands areas of widespread and long-standing attacks prevail.

276 THE PRACTICAL APPLICATION OF ECOLOGY

The foregoing examples will serve to illustrate the influence
of plant nutrition, and soil conditions, in relation to the
susceptibility of crops to insect infestation. Recent research
emphasises the complexity of the subject, but, at the same time, it
indicates the importance of studying the ecological relations
between host and pests as an interconnected problem. Progress
of knowledge on this field is necessarily slow, and consequently
its application to cultural practice can only result very gradually.
We are only just beginning to realise that persistent heavy
attacks by insects may be the result of the particular cultural
conditions under which a given crop is grown. The possibility
is, therefore, opened out of so regulating the growth of the crop
that it will be rendered less prone to such attacks, and still yield
an adequate economic return.

Resistant Varieties

In the wild state many plants appear to suffer relatively little
injury from insect attacks as compared with their cultivated
representatives. They appear to have acquired, through many
thousands of generations, a degree of tolerance to such interference,
whereas cultivated varieties of these same plants suffer severely
from repeated infestation. This difference in plant response
appears attributable to two main causes. (1) The conditions
under which the cultivated varieties are grown which favour
infestation. (2) It would seem that during the process of selection
and breeding, which has led to the production of cultivated
varieties, certain properties, inherent in the wild state, have been
out-bred ; that the plants are rendered in this way, not only more
attractive to their respective insect enemies, but also less able
to sustain the injuries incurred.

The problem which presents itself is likewise two-fold. Firstly,
an ecological study of the conditions in which a given crop is
grown, in order to trace those conducive to infestation, which has
already been discussed. Secondly, the discovery of varieties or
strains of plants which betray high resistance to insect attack
and, at the same time, exhibit desirable properties from the stand-
point of the grower. In practice, strains which exhibit one or

VARIETAL RESISTANCE 277

other of these attributes are well known, but the real problem
involves their judicious selection and combination, which has, as
a rule, proved extremely difficult of attainment.

Much has been written with reference to the nature and meaning
of resistance. Absolutely resistant plant varieties are of great
rarity, and resistance, more often than not, implies little more
than a low degree of susceptibility to attack. The factors which
control, or influence, this susceptibility are extremely varied, and
have been classified in various categories. In many cases it is
probable that no single factor is in itself responsible for the
observed effect, and that multiple factors are involved. Among
these contributing causes, thickness of cuticle, development of
sclerenchyma or other mechanical tissue, general hairiness or its
absence, acidity or alkalinity of the cell-sap, silica content, presence
or absence of certain glucosides, date of maturity, power of
recuperation and other attributes have been regarded as account-
able in particular cases.

Another type of so-called resistance is brought out by the
work of Davidson (1922) at Rothamsted, which concerns the
reproductive capacity of Aphis rumicis on different varieties of
culinary and field beans {Vicia faha). Discrimination on the part
of the insect in the selection of one variety in preference to another
was eliminated in these experiments, which were conducted
throughout under identical conditions. His observations show
that the insect exhibits a wide range of fecundity on different hosts.
These differences are probably due to differences in the constitution
and nutritive values of the cell-sap in the respective varieties.
There appears to be a direct inhibiting effect of the cell-sap upon
the reproductive capacity of the aphid when grown on certain
varieties of beans, and the opposite effect when grown upon
other varieties, as is shown in the table on p. 278.

It is probable that varietal differences in the properties of the
cell-sap of many other species of plants are reflected, in the
manner described, in the relative abundance of numerous kinds
of sucking insects feeding upon them. Much the same kind of
reaction, it may be added, has been observed by Monzen with
reference to the fecundity of Eriosoma lanigera on different

278 THE PRACTICAL APPLICATION OF ECOLOGY

Table XIII

Varieties.

Mean Infestation and
Probable Error.

Degree of Susceptibility.
Per cent.

Prolific Longpod

Bohus (Swedish), Mazagan .

1,037 ± 51-8
1,012

Standard, taken as
100

98

English field bean

737 ± 43-7

71

Heliogland

Scotch var. Granton .
Winter bean (Dorset)
Winter bean (Somerset)

\

570 ± 19 2

55

Winter bean (Essex) .
Spring tick (Suffolk) .
Carse bean

Kilbride bean and five other
varieties

•

407 ± 11 1

39

Winter bean (Suffolk).

286 ± 27-2

27

Vicia narbonensis

37

3

Reproduction of Aphis rumicis upon different host varieties. The
infestation figures are the mean in each ease of five plants, each plant
being infected with a single apterous viviparous female which was
allowed to reproduce for a period of fourteen days.

varieties of apple {vide p. 280). There is evidence, both in the
case of varieties of beans and of apple, that those most desirable,
from the point of view of human consumption, appear to favour
and stimulate the highest reproductive rates in their respective
aphid enemies. This, in itself, lends support to the suggestion
already made that certain properties, contributing to the greater
" resistance " of the wild forms of plants, have become out-bred
in the course of repeated selection in cultivation.

It may be said that, up to the present, comparatively few
discoveries of insect-resistant varieties of plants, of proved
commercial importance, have come to light. The methods
involved in the search for such varieties are either those of selective
breeding, varietal hybridisation, or graft hybridisation. Selective
breeding entails either the continued selection of individual plants
of a crop, observed in the field to betray least injury ; or line

VARIETAL RESISTANCE 279

selection and propagation, which involves breeding for highly
resistant special strains. A good example of selective breeding is
afforded by the work done in South Africa with regard to jassid-
resistant strains of cottons. Varietal hybridisation is practised
with the object of combining desirable features present only in
separate and different strains. It is dependent for its success upon
the purity of the segregation of the resistance and other factors
involved. The method has been practised by Harland in Barbados
in the crossing of Sea Island and Upland cottons with resistant
native strains as a means of control against the leaf blister mite
{Eriophyes gossypi). In the Southern U.S.A. the crossing of
commercial sweet maize with native field varieties is stated
to have resulted in strains better able to resist attack by the
corn earworm (Heliothis ohsoleta). Inter-specific hybridisation
has resulted in the production of the Leconte and Keiffer varieties
of pears which are resistant to the San Jose scale. These varieties
are Fj hybrids between Pyrus communis and the Chinese pear
(Pyrus sinensis) ; both inherit the desirable qualities of P.
communis^ in combination with the resistance to scale injury
characteristic of P. sinensis. The need for discovering varieties
of spring oats resistant to infestation by frit-fly is referred to on
p. 282, and it appears that more promise of success is likely to
result by adopting varietal hybridisation, rather than by relying
upon any varieties, yielded by selective breeding, exhibiting the
desired combination of characters. Graft hybridisation, which is an
old method, involves the grafting of commercially valuable, but
highly susceptible, plants upon root-stocks of resistant varieties.
It has proved successful in the now classical cases of apple trees
infested by woolly aphis, and of vines attacked by the Phylloxera.

The foregoing remarks will serve as a brief outline of the nature
of the problems involved and, in further illustration of this phase
of insect control, certain examples will be taken in connection
with recent researches.

1. With regard to the susceptibility of different varieties of
apple trees to the attacks of the woolly aphis (Eriosoma lanigera),
an extensive literature exists on the subject, but no very con-
clusive deductions can be drawn. Some authorities maintain

280 THE PRACTICAL APPLICATION OF ECOLOGY

that resistance is dependent upon mechanical factors, while others
believe that physico-chemical influences are the most important.

It has been recognised as long ago as 1831 that certain sour
varieties are much less liable to attack by the aphis in question,
and this observation led to utilisation of sour apples as grafting
stocks as a preventive measure. In this way the resistant varieties
Northern Spy and Winter Majetin came into prominence, but
they have not proved an unqualified success in other ways, since
dwarf trees result, productivity is not high and they do not react
well to adverse local conditions. A number of other resistant
varieties have come to be recognised in different countries, but it
appears that varieties that prove to be resistant in one country
are not invariably so when grown under different conditions in
another country. Thiele (1902), for example, states that Northern
Spy becomes susceptible in certain regions of Germany, while
under English conditions it has been found to be immune to both
the root and branch phases of aphis attack. According to Monzen
(1926) the variety Jonathan is susceptible in Japan, whereas
Theobald showed that it was almost free from aphis attack in
England. Similarly, the last-mentioned authority found Cox's
Orange Pippin to be subject to attack in England, yet Misra
(1920) states that under Indian conditions it is immune.

It would appear that the resistance in these immune, or slightly
susceptible, varieties of trees to attack by the Eriosoma is to a
large extent an inherent property, but that this property can
either be intensified or reduced by cultural conditions.

According to Crane (1937) in the results of crossing immune
with susceptible root stocks the variety Northern Spy has a higher
value in respect to the transmission of immunity than any other
of the immune varieties used. The proportion of immune plants
obtained is often as large, and frequently much larger, in crosses
between susceptibles and immunes than that obtained in families
raised from immune crossed with immune. Some susceptibles
carry factors or genes concerned with immunity and some do not,
and there is a gradation in the inheritance of susceptibility. All
the immune stocks used were heterozygous. The results of
breeding investigations are not discussed in detail in the present

1

VARIETAL RESISTANCE

281

stage of the work, but they indicate that the hereditary behaviour
of immunity is complex. It seems to be determined by and
dependent upon a certain balance of genetical factors and governed
by a number of genes of somewhat varied action — in part com-
plementary and in part cumulative. Monzen (1926) lays stress
upon the physico-chemical nature of resistance, and gives certain
data lending support to his contention. He determined the ^H
values of the cell-sap of eleven varieties of apple by the colorimetric
method, and ascertained that it ranged between 4-4 and 5. The
variations, therefore, are very limited, and from these deter-
minations it appears that the ^H of susceptible varieties ranged
from 4-5 to 5, while that of the immune varieties (including
Northern Spy) was 4-4. The acidity of the immune varieties
he regards as being due to the larger amounts of malic acid
present. With the object of further testing this conclusion, he
carried out a series of experiments with apple trees grown in
carefully washed sand and nourished with Pfeiffer's solution.
The latter medium was rendered alkaline or more acid in different
degrees, and the fecundity of the aphis was determined during
a definite period on trees grown in the different solutions. The
results, shown below, indicate that acidity appears to markedly
reduce the fecundity of the aphid, while alkalinity favours it.
Monzen further adduces evidence that manuring with acid
fertilisers will tend to increase the resistance of the trees to
infestation.

Table XIV

pB. of
Solution.

Number of Progeny of the Aphid.

Trees.

Colony A. Colony B. Colony C.

Average.

I.

9-4

279

104

75

152-7

II.

8-8

131

65

No progeny

980

III.

6-8

27

No progeny

—

270

IV.

5-6

41

19

—

30 0

V.

3-6

92

61

68

73-7

VI.

2-8

42

37

■ —

39-5

It is improbable that relative acidity or alkalinity is the whole
explanation of the questions involved, and Monzen's conclusions

282 THE PRACTICAL APPLICATION OF ECOLOGY

with regard to the apphcation of acidic fertihsers are not in com-
plete accordance with opinions in Germany.

It may be added that the beHef that the acidity of the cell-sap is an
important factor in relation to the attacks of sucking insects is by no
means new. Petri in 1911 claimed that it played an important role
with regard to the infection of vines by the Phylloxera, and Comes in
1914 held that it presented means of defence of various plants, more
particularly against Coccidse.

The factor, or factors, responsible for immunity to woolly
aphis attack in Northern Spy and other varieties of apple are
discussed by Roach (1937). In the experiments described by this
author indications were obtained which point to the conclusion
that immunity is probably a biochemical problem and that it is
not due to mechanical or other properties rendering such varieties
more or less impenetrable to the stylets of the insect. The
woolly aphis was reared on artificial media prepared from the bark
of Northern Spy and of susceptible trees. It was found that the
development of the insects, when fed upon media of this kind,
corresponded to their development on living trees in the smaller
number and shorter life of the progeny associated with immune
extracts. The results seem to show that the cause of immunity
(or susceptibility) resides in some chemical substance insoluble in
alcohol or ether and only slightly soluble after prolonged boiling
in water. The behaviour of the insects on the media suggests
that, whatever the factors concerned may be, they are not
necessarily destroyed when the tissues containing them are
killed.

2. The problem of the relative degrees of resistance of different
varieties of oats to attacks of the frit-fly (Oscinella frit) has, in
recent years, received considerable attention in England (vide
Cunliffe and Fryer, 1924). As to what is exactly implied by
" resistance " in this case little is definitely known, and it is
gauged by the relative freedom from infestation of different
varieties grown under similar conditions. Direct resistance
implies that the plant can resist entry by the larva of this insect,
or that it provides conditions so unfavourable as to cause the
death of the larva before material injury is sustained. On the

VARIETAL RESISTANCE 283

other hand, it is possible that it may depend upon characters
associated with growth and which vary with the age of the plant.
Such characters might comprise silica content, hardness or thick-
ness of cuticle, hairiness, sap qualities, etc. Since it has been
shown (p. 264) that susceptibility appears to be correlated with
the age of the plant, one or more of these features may be involved.

In Cunliffe and Fryer's experiments twelve varieties of oats
were grown in two widely separated localities, viz., Sandford
(Oxon.) and Harpenden (Herts). The infestations of both stem
and panicles were noted in detail, and the results obtained were
subjected to statistical analysis. The correlation between stem
infestation of different varieties at the two localities was high,
but the grain infestation exhibited no correlation whatever. Of
the varieties submitted to experiment, Goldfinder and Supreme
are those least and most subject to infestation, respectively.
Consistent and significant differences in the extent of shoot
infestation, of the order of 20 per cent., were noted in favour of
Goldfinder. Unfortunately, correlated decrease of seed infestation
was not so marked ; Goldfinder was not in the group suffering
least infestation, but it suffered less than 5 per cent, heavier
attack than those varieties least attacked. The variety New
Abundance was found to be almost equal to Goldfinder in its
freedom from the stem infestation, but, on the other hand, it was
decidedly the heaviest sufferer from the grain attack.

It is not possible to say how far the results of these experiments
can be held to be of general application. The fact, however, that
such varietal differences do exist is hopeful for the future. The
great difficulty in many parts of England is to sow oats sufficiently
early to escape attack, owing to the fact that weather and soil
conditions are frequently adverse to this procedure. It seems,
therefore, that the chief hope for the future lies in the production
of varieties which combine resistance with good yield, since
heavy yielding strains for the most part occupy inferior positions
in the scale of resistance. Cunliffe (1936) describes a series of
experiments designed to ascertain whether it is possible to breed
resistance to attack by this insect into the best varieties of English
oats, which are highly susceptible. It appears that the most

284 THE PRACTICAL APPLICATION OF ECOLOGY

highly resistant variety yet found is " Sommar " (summer) from
Gotland, and this along with "Spet" from Sweden and "Hede"
from Denmark were used in hybridisation experiments. Many
varieties, such as "Victory," "Abundance," and "Star," were
used in conjunction with the aforementioned resistant types, some
seventy-five or eighty hybridisations being made. The outcome
of the trials points to the conclusion that resistance to attack is
an inheritable character (or complex of characters) that is not
incompatible with desirable agricultural qualities. The nature
of resistance is unknown, but it is suggested that it may be
connected with crude fibre production or deposition of silica in
such varieties.

3. The cotton jassid, Empoasca (Chlorita) fascialis, of East,
South and West Africa is one of the most important pests of the
cultivators of the crop concerned. It occurs in colonies on the
leaves of the cotton which, in bad attacks, become deformed and
are finally shed ; growth of the plant becomes arrested and it may
ultimately be killed. The possibility of discovering resistant
varieties has attracted a good deal of attention in South Africa,
and much valuable breeding work has been carried out at the
Empire Cotton Growing Corporation Station at Barberton. The
discovery that some strains of varieties and individual plants of
the American Upland Cotton (Gossypium hirsutum) resist jassid
attack, to a remarkable degree, gave impetus to the work. None,
however, has been found to be absolutely immune to the ravages
of the insect, and all sooner or later succumb and serve for its
breeding and multiplication. Worrall (1923, 1925) observed that
the resistant strains are markedly hairy, but not all the hairy
strains are necessarily highly resistant. The qualities which con-
stitute resistance are not understood, and it appears that hairiness
may be mainly the outward expression of such qualities. Mere
hairiness alone is not necessarily important and, in the resistant
varieties, this character also involves length of hair and density
of hairiness per unit area. Worrall expressed the opinion that
hairiness is one of the characters to be included by cotton-breeders
in their selection work, but that care will need to be exercised to
select only those plants with a good staple, because hairiness on

VARIETAL RESISTANCE 285

the whole appears to exhibit a tendency to be correlated with
short staple.

A further step was taken when the Cambodia variety of resistant
cotton, introduced into India a number of years ago from French
Indo-China, was tested out under South African conditions.
This variety has also the definite property of hairiness and,
according to Parnell (1925), it was found to show immunity — the
jassid does not appear to breed upon it or injure it in the least
degree. The large size of the plant, and its rather weak and
delicate stems, which render it susceptible to storm injury, make it
of doubtful promise as a field crop. Only further work will show
whether it is possible to obtain less rampant strains. At the same
time, research is being carried out with the object of selecting and
testing promising local varieties betraying evidence of jassid-
resistance in the field. In the latest available Report on the
subject, it is stated that remarkable progress has been made in
this direction and has resulted in the discovery of several strains,
of which those known as Z.l and U.4 are both highly resistant to
jassid attack, of good staple and prolific yielders. This has led
to supplies of their seed being sent to farmers, and to an increase
of acreage given over to cotton-growing. It would appear that
the discovery of these resistant varieties has brought the solution
of the problem within sight. Doubtlessly, much further work is
still necessary with regard to re-selection and improvement of
these types, and it remains for the future to show whether their
inherent properties will be retained under diverse local conditions.
Their discovery, under the aegis of the Empire Cotton Growing
Corporation, is one of great importance, and will lend encourage-
ment to further efforts in the same direction, with reference to
pests in other parts of the world.

CHAPTER XI

THE PRACTICAL APPLICATION OF ECOLOGY—

continued

Host-selection and Biological Races, p. 286. The Adoption of
New Food-plants, p. 294 ; Hybridisation Experiments, p. 297 ; The
Interpretation of Host-selection, p. 297. Ecological Aspects of
THE Locust Problem, p. 299. The Phase Theory, p. 300 ; Factors
Governing Swarming, p. 305 ; The Species of True Locusts, p. 306 ;
Phase Development in Grasshoppers, p. 311. Literature, p. 313.

Host-selection and Biological Races

The last few years have witnessed the pubHcation of a number
of papers bearing upon the theory enunciated by Hopkins (1916)
and now generally known as the " host-selection principle."
According to this authority, in a given insect species that is capable
of breeding in two or more hosts, individuals will normally con-
tinue to select the particular host species upon which their own
life-cycle was passed. In other words, this host predetermines
selection by the ovipositing female of the same plant species for
her offspring as that in which her own developmental stages took
place. The outcome of the principle is the evolution of special
physiological strains, or races, each betraying its own particular
host-preferences, but, at the same time, exhibiting little or
no indications of morphological differentiation. This general
phenomenon was recognised more than sixty years ago by
Benjamin Walsh, who maintained that we have, in such host-
preferences, the beginnings of species differentiation. Such
preferences, he claimed, become habituated and destined through
heredity to result, in course of time, in increasing divergence and
to lead ultimately to the formation of separate species.

Many cases of the apparent separation of polyphagous or
oligophagous species into different phytophagic or biological
races have been recorded {vide Thorpe, 1930). In some cases

286

HOST-SELECTION 287

slight structural differences, or differences in size or colour, are
involved, in other cases not. For many years experimental studies
of host-preferences of this kind, and of their possible repetition in
succeeding generations, was lacking. It is only comparatively
recently that the problem has received critical investigation, but
in all cases larger scale and more prolonged studies are still a
desideratum. The whole subject is not only of importance in
relation to the general question of the origin of species, but also
from the standpoint of applied entomology.

It is necessary here to refer briefly to the instance of host-selection
specifically stressed by Hopkins. His example concerns the moun-
tain pine beetle {Dendroctonus monticola), which attacks various
species of pines. According to him, in this insect, when it becomes
established in one species of pine through many generations, the
beetles upon emergence show a marked preference for that
species, and will not attack any other. Hopkins applied this
observation to control measures, and held the view that each
such " strain " can be dealt with as if it were a separate species.
In 1922 Craighead published the results of a number of years'
experiments with about a dozen species of Cerambycidse, which
afforded support to Hopkins' general contention. He found that,
in practically all the species studied, the adults exhibited marked
predilection for the host-species in which their larval existence
was passed, provided they were not deterred by such factors as
unfavourable conditions or the scarcity of their host. Con-
tinued breeding appeared to intensify the preference and, in some
species, transference to another species of host for only a year
was said to be sufficient to determine a preference for this host
by the resulting adult beetles. In 1927 Larson discussed the
host-selection principle very fully and tested it with reference to
the cowpea weevil (Bruchus quadrirnaculatus). His results were
not in accordance with those of Craighead, since he concluded that
the insect in question showed no marked preference for the host
in which it had been bred, and that continued breeding in any
one specific host did not appear to intensify preference for that
host. In 1928 Thompson and Parker examined the question
with regard to the corn borer (Pyrausta nuhilalis). This insect

288 THE PRACTICAL APPLICATION OF ECOLOGY

is notably polyphagous, but, in Europe, it chiefly attacks maize in
regions favourable to the cultivation of that crop ; in regions
north of the maize-growing areas it mainly attacks weeds and
more especially Artemisia vulgaris. In some regions, however,
it attacks both maize and Artemisia, but in others it favours only
one of these two hosts. Thus in the Department of Gers, in
south-western France, maize is practically its only food-plant,
while around Paris it lives principally upon Artemisia, and has no
opportunity of attacking maize owing to the absence of that crop.
It appeared likely, therefore, that material from these two areas,
where the conditions have remained relatively stable for a long
period of time, would be specially suitable for the study of the
host-selection principle. The conclusions derived from these
experiments show that, notwithstanding the long association of
the corn borer with Artemisia, which may indeed be its original
host, it exhibits no marked preference for this plant and laid the
majority of its eggs upon maize. In fact, the latter host is the
favoured one both by the Artemisia and maize strains of the insect,
as is shown by the following records of oviposition (in percentages) :

Table XV

strain.

Maize.

Masses. Eggs.

Artemisia.
Masses. Eggs.

On Woodworli, etc.
Masses. Eggs.

Artemisia
Maize .

58-4
69-6

55-6
701

38-8
13-2

41-2
11-5

2-8
17-2

32

18-4

It would appear, therefore, that the corn borer is a species
having a strong innate preference for maize, and that the Artemisia
strain exhibited a greater attraction for that host than for its
original host. It is true that this strain exhibited a stronger
preference for Artemisia than was evidenced by the maize strain —
in the latter case, in fact, the preference does not appear to be
significant, since even more eggs were deposited away from any
specific food-plant than upon Artemisia.

Experiments by Thorpe (1929, 1931), with the moth
Hyponomeuta padella, indicate that this insect is divisible into
two biological races which were formerly regarded as separate

HOST-SELECTION

289

species. One race is specially attached to apple as its food-
plant, and the other to hawthorn and blackthorn. The apple-
bred moths exhibit a great preponderance of forms with pure
white fore-wings, while those reared upon hawthorn seldom
exhibit this feature, the wings being most often suffused with
grey. The two forms also exhibit differences in cocoon texture
and methods of oviposition. The results obtained in these
experiments are shown in tabular form below.

Table XVI

Hawthorn Race.

Apple Race.

Hawthorn and
Blackthorn.

Apple.

Hawthorn and
Blackthorn.

Apple.

Number of egg-
batches .

26

7

10

87

Number of eggs

911 (79-3%)

237 (20-7%)

367 (9-75%)

3,395 (90-25%)

Average number of
eggs per batch

35

33-9

36-7

39

When the hawthorn race was reared upon apple, the resulting
females laid all their eggs upon hawthorn. On the other hand,
when the apple race was reared upon hawthorn, 69-2 of their eggs
were laid on apple and 30-8 per cent, on hawthorn. Experiments
carried out with hawthorn- and blackthorn-reared moths,
allowing them free choice of the two food-plants, indicated that
the hawthorn-blackthorn is subdivided in the same way. It
appears that out of 825 eggs laid by hawthorn-reared moths,
18-8 per cent, of them were laid on blackthorn and 81-2 per cent,
on hawthorn. Furthermore, out of 4,293 eggs laid by the
blackthorn race, 69 per cent, were on blackthorn and 31 per cent,
on hawthorn. Thorpe's results, therefore, lend strong support to
the host-selection principle and they raise the practical point as
to whether it is necessary for the apple grower to deal with
infestations on surrounding hawthorn hedges. It would appear
from the data already given that such an infestation is unlikely,
under normal conditions, and that treatment of the hedges would
not be a sound practical proposition.

R.A. ENTOMOLOGY.

10

290 THE PRACTICAL APPLICATION OF ECOLOGY

The discovery by Lai (1934), in Edinburgh, of two biological
races of the common hemipterous insect Psylla mali, is of con-
siderable interest. The forms P. peregrina of the hawthorn and
P. mali of the apple have hitherto been regarded as two separate
species. Lai has shown that the imagines of both forms are
identical and the only differences are those shown by the nymphs.
In peregrina the last instar nymph has a brown streak on the
developing wings which is absent in mali. Also, in peregrina
the seta? surrounding the abdominal margin are all of one kind —
long and spatulate : in mali they are of two types, long, pointed,
variously curved setae and small setae of spear-like shape. In
peregrina the dorsal surface bears round tubercles which are
replaced in mali by very minute setae, which cover the body of the
latter. The two forms are closely confined to their individual
food-plants : they do not interbreed and neither will oviposit on
the host of the other. The nymph of either, when transferred to
the food-plants of the other, does not survive long. In view of the
foregoing evidence, it is pointed out that P. mali Schmidt has a
race peregrina Forst which has previously been regarded as
another species. This example, wherein similar adults betray
nymphal differences, affords an instance of what Giard termed
poecilogony, which is a rare phenomenon among insects.

So far the subject of biological races has only been considered
with reference to their expression in differences of food-preference.
Evidence of such races with very different characteristics requires
mention. Thus, the moth Ephestia kuhniella appears to occur as
two strains or races with marked different rates of development.
Payne (1933) found that the slow strain was able to complete its
developmental cycle between 8° and 34° C. and the fast strain
between 10° and 32° C. A number of observers had previously
noted marked differences in the rates of development of larvae
hatched upon the same day. The most recent worker, Ahmad
(1936), has confirmed Payne's observations and points out that
this insect is only susceptible to parasitisation by Nemeritis
during the last larval and pupal stages, but the mixture of strains,
he states, results in the continuous presence of the host in the right
condition for parasitisation.

RACES OF MOSQUITOES 291

An extensive literature has grown, during the past few years,
around the subject of biological races in mosquitoes. The pre-
valence of " anophelism " without malaria in many districts has
formed the basis of the idea that races may exist which vary in
ability to transmit the malaria Plasmodium and in diverse other
biological features. Thus Anopheles maculipennis labranchice lays
a maculated grey egg, is especially prevalent in human habitations
and does not undergo complete hibernation. This race occurs
wherever malaria is prevalent in Europe, and its larvae are stated
to tolerate brackish water. Anopheles maculipennis messece lays
eggs marked with two transverse black bars : it frequents stables,
undergoes hibernation and is the predominating form in non-
malarial districts. Its larvae are stated to show preference for
fresh water. Differentiation of two races upon the foregoing
criteria as stated by Hackett, Martini and Missiroli (1932) has
been confirmed in the main, but according to Roubaud and
Gaschen (1933), the so-called maxillary index or, in other words,
the number of serrations on the maxillary stylets is a better
criterion than the egg-characters already alluded to. More
recent work by Missiroli, Hackett and Martini (1933) has led to
the conclusion that in addition to the two races mentioned, three
or possibly four others also occur. These differ, among other
features, in their feeding habits : thus, labranchice and sacharovi
are the important malaria carriers : messece and atroparvus
normally resort to animal hosts, but attack man in the absence
of the latter, while melanoon and typicus have no significance in
connection with malaria transmission.

The researches of Roubaud and his co-workers, published in a
number of papers between 1929 and 1932, have shown that in
France there are two definite biological races of Culex pipiens.
These races are characterised as follows : (1) C pipiens pipiens,
which undergoes cyclical hibernation, does not oviposit without
blood-meals (anautogeny) and requires a large area for pairing
(eurygamy). (2) C. pipiens autogenicus, which does not hibernate,
but breeds continuously throughout the year. Females of this
race lay fertile eggs, without partaking of any food as adults, the
ova maturing at the expense of food-reserves accumulated during

292 THE PRACTICAL APPLICATION OF ECOLOGY

larval growth (autogeny) and fertilisation occurs in confined
spaces such as very small vessels (stenogamy). A number of
other workers have confirmed the existence of two such races,
but differing in various ways from the two French races just
referred to. The extensive literature on this subject is discussed
by Tate and Vincent (1936), and these authors have maintained
three autogenous strains of C. pipiens, and the general result of
their work confirms the conclusions of Roubaud and others that
there are two distinct races of C. pipiens in Europe. They point
out that confusion has arisen from the fact that Roubaud described
the autogenous race as an urban form, whereas it occurs in rural
areas also. It is true that it may breed freely in towns where
conditions are not suitable for the anautogenous race, but is not
confined to such habitats. The result of cross-mating experiments
show that interbreeding may take place in nature where the two
races co-exist. They find that the English autogenous race is not
subject to obligatory, cyclical hibernation. Prolonged illumination
has an activating influence on the females, both hibernating and
laboratory bred, and greatly stimulates gorging. With the aid of
artificial light during the winter months this race has been kept
breeding in the laboratory throughout the year for eleven genera-
tions without occurrence of cyclical hibernation. They think
that the reduced daylight may have an important influence in
inducing hibernation in autumn, since if given extra illumination
they remain active and feed. The females, whether they be of
hibernating or active generations, laid normally after one blood -
meal. Unfed females lived for a maximum of nineteen days and
never developed fat-body autotrophically. If fed on carbohydrate
food, in the form of fruit juice, for five weeks they accumulated
sufficient reserves to enable a small proportion of them to survive
fifteen weeks of subsequent starvation. Pairing always begins in
the air, but may be completed on the ground. The females do
not lay without a blood-meal and they show little tendency to
bite man, but attack birds voraciously.

The strains of the autogenous race have been kept breeding
without blood-meals for several years and through forty-five to
forty-nine generations. This had no deleterious effects, a fact

EACES OF MOSQUITOES 293

which is at variance with Roubaud's conclusion that a gradual
decrease in the number of females produced in successive genera-
tions followed prolonged continuous autogeny. The feeding of
the females with fruit juice is apparently without influence on
egg-production, but after a blood-meal the number of eggs in a
raft is about twice as many. The males pair with resting females,
and it is evident that mating can occur in very small vessels or
restricted areas. The females are voracious blood-suckers of
both birds and man. Tate and Vincent regard the term autogeny
as being unfortunate, owing to its being a variable character,
since on an average only 60 per cent, of the females lay autogenous
eggs and under unfavourable conditions very few may do so.
They consider that eurygamy and stenogamy are more constant
characters for differentiating between the two races. Roubaud
(1935) has described differences in the micro-structure of the egg-
floats of the two races, and Marshall and Staley (1935) found
differences in the chsetotaxy of the larvae.

Reciprocal cross-mating between individuals of the sexes of the
two races was easily obtained. The experiments show that
autogeny and stenogamy are genetical characters transmitted to
the progeny. Stenogamy always appears in the F^ generation,
but autogeny may not appear until the Fg generation or later.
Much further genetical work is necessary, however, before the
inheritance of the racial characters can be discerned with
confidence.

The question naturally arises as to the evolutionary significance
of biological races. Do they, in fact, afford evidence of evolu-
tionary divergence which, if carried far enough, will ultimately
lead to the production of new species ? A number of workers have
answered this question in the affirmative. If, at any rate, some
of the races are species in the making, and where their characters
are germinally fixed there seems good evidence for this view, we
have to conclude there is some degree of isolation of a physiological
or behaviouristic kind which ensures their continuance. None of
the evidence so far presented provides sufficient data to carry
full conviction in this matter. It does, however, suggest a
profitable field for further enquiry, and for a discussion of the

294 THE PRACTICAL APPLICATION OF ECOLOGY

subject the reader is referred to the general review by Thorpe
(1930).

The Adoption of New Food-plants. A certain amount of
experimental evidence exists with reference to the adoption of
new plant-hosts by various insects. One of the earliest experi-
ments of this kind appears to be that of Schroeder (1903), with
reference to the leaf beetle Phyllodecta (Phratora) vitellince. The
larvae of this insect skeletonise the leaves of Salix fragilis so
as to leave only the upper epidermis intact. Young larvae
transferred to Salix viminalis which, unlike ^S*. fragilis, has its
leaves densely pilose beneath, adapted themselves to this new
food-plant. By cutting and pushing aside the hairy covering
they reached the central parenchyma and one larva excavated a
mine in the leaf-tissue. After four generations upon S. viminalis
the larvae became adapted to leaf-mining, and in each generation
the proportion of adults which selected this food-plant in pre-
ference to the original host steadily increased from 9 per cent, to
42 per cent.

Experiments conducted by Marchal (1908) with the scale insect
Euclecanium persicce involved the transfer of the eggs of individuals
living upon peach to false acacia (Rohinia pseudacacia). Only
certain of the young scales resulting from these eggs were able to
adapt themselves to the new host, but in the following year their
progeny had so completely accepted the unusual food-plant that
the transfer of individuals back again to peach resulted in their
failing to reach sexual maturity. The individuals reared upon
Rohinia, it may be added, closely resembled in their large size,
form and colour the race or species E. rohiniarum of that same
host. Since the Rohinia species was unknown in France prior to
1881, Marchal was of opinion that it was a recent derivation from
the species E. persicce. Evidence in support of this contention is
not entirely conclusive, since the Rohinia is a tree imported from
America, where the scale insect E. rohiniarum is also known to
occur.

Sladden (1934), in experiments dealing with induced food-
preference in the stick-insect Carausius, came to the conclusion
that individuals reared upon ivy showed a greater ability to feed

HOST-SELECTION 295

on that plant, or even a greater preference for it, than their
parents or controls of the same generation derived from privet-
fed stock. In a later paper (Sladden, 1935) this conclusion is
confirmed and the experiments show that the offspring of each
subsequent generation accepted ivy more readily than did their
parents, and even exhibited an increased preference for it.

Harrison (1927) published the results of certain field experiments
with the gall-forming saw-fly Pontania salicis. He removed a
strain of this insect which is locally adapted to Salix phylicifolia
to a district where various species of Salix grew, and the insects
were allowed to oviposit unhindered in the open. They showed
a marked preference to adopt Salix andersoniana to the compara-
tive neglect of the very closely allied original host S. phylicifolia,
a feature that was particularly evident during two successive
years. In other experiments he found that a strain of the same
saw-fly from Salix andersoniana brought to another locality
where only Salix rubra was available became adapted to that
host. Later, when plants of S. andersoniana were established in
that locality in proximity to S. rubra the former species of willow
remained entirely free from the galls of this saw-fly. He further
mentions that the latter, in its reaction to a changed food-plant,
shows corresponding colour differences. Mention needs also to
be made of Pictet's experiments (1911) with larvae of Lasiocampa
quercus, which feed upon deciduous trees and bushes. When
placed upon Pinus many of them died and those which survived
fed by biting into the parenchyma at the extremities of the pine
needles. In the second generation the insects became adapted
to the new diet to the extent that they either starved when
offered leaves of deciduous trees, or they only attacked the apices
of the leaves, which they hollowed out in a manner similar to that
performed by larvae of the previous generation.

It will be gathered from these experiments that it is possible
for an insect to become rapidly adapted to a new food-plant to an
extent that it may partially or completely reject its original host.
It is reasonable, therefore, to conclude that similar behaviour may
occur in a state of Nature, particularly if the normal food-plant be
lacking. Harrison's experiments, already alluded to, indicate that

296 THE PRACTICAL APPLICATION OF ECOLOGY

such change of habit may be equally rapid under natural conditions.
There exists in entomological literature many records of insects
adopting food-plants which they were previously unknown to
affect, but apart from the actual observation of the facts little
else is known with respect to this behaviour. In one or two
instances, however, well-attested evidence points clearly to the
change having occurred within recent years, and appears to be
leading to the evolution of special biological races. A well-
known example is familiar to North American entomologists in
the case of the Trypetid fly Rhagoletis pomonella. The larvae of
this insect have long been known to live in the blueberry, but in
the north-eastern states its more recent spread to apple is a
matter of direct record. According to Woods (1915) it has
developed into two distinct races distinguished by the small size
of both larvae and adults derived from blueberry as compared
with those from apple. Attempts to transfer the apple race to
blueberry and vice versa were unsuccessful, the two forms being
seemingly independent. What appears to be a parallel example
in England is afforded by the Capsid Plesiocoris rugicollis which
has now become a widespread and serious pest of the apple. It is
practically certain that it was not an apple pest in England
prior to 1900, and since that time a race appears to have developed
which has adopted this tree as its host in preference to S alloc,
which is its usual wild food-plant. This change seems to have
involved some readjustment in the life-cycle of the insect since,
when living upon willow, it usually occurs from June to August ;
whereas, upon apple, eclosion from the eggs takes place from the
middle of April onwards, and adults occur from about the first
of June until the end of July. The whole life-cycle, it appears,
has been advanced by about one month. There is need for
exhaustive experiments with reference to the behaviour of this
insect ; under artificial conditions it can be readily transferred
from willow to apple, or even to new food-plants such as plum.
In a state of Nature, however, it shows a tendency to remain
upon the host-plant on which hatching took place. In certain
cases willow trees were found with their branches actually
touching or interlacing those of apple, yet although the willows

HYBRIDISATION 297

were extensively infested, no transfer to apple took place. For
further information and references to the principal literature on
Plesiocoris vide Fryer (1929).

Hybridisation Experiments. Evidence of the nature of the
food-preferences of crosses between biological races and sub-
species is of a very scanty nature. In this connection the
experiments of Gdschen (1913), therefore, are of particular
interest. The European Sphingid moth Celerio euphorbice is
represented by two races or sub-species named euphorbice and
mauritanica respectively. In both races the usual food-plant is
Euphorbia and neither will accept Salix, Goschen found that the
hybrid mauritanica S X euphorbice $ and known as wagneri, and
also the reciprocal cross, both can be reared to maturity without
difficulty upon Salix. Similarly the hybrid known as C
kinderwateri, produced by the cross euphorbice S X galii $, feeds
readily upon Salix, a food-plant upon which neither of the parents
naturally occur. On the other hand, observations made by Field
(1910), in North America, appear to show that host-selection is not
necessarily affected by hybridisation. Field's observations bear
upon the feeding propensities of the butterflies, Basilarcha arthemis,
B. astyanax and B. proserpina, the last named being regarded as a
natural hybrid between the first two. Larvae produced from a
captured proserpina female refused birch, sallow and poplar, the
usual food-plants of arthemis, and were reared upon wild cherry,
which is one of the favoured plant-hosts of astyanax. These
same larvse finally produced all the three forms of Basilarcha
named. He suggests that the constitution of the original
parent was proserpina (hybrid) x arthemis (colour recessive),
since astyanax does not prevail in the district where the insect
was captured. The unexpected preference of the astyanax food-
plant is explained as having been inherited from a grandparent
of that form. Unfortunately, neither Goschen nor Field were
able to continue their observations to later generations of the
species mentioned.

The Interpretation of Host-selection. It has been seen that
the results of recent research afford definite support in favour
of Hopkins' host-selection principle in a considerable proportion

298 THE PRACTICAL APPLICATION OF ECOLOGY

of the cases studied. The phenomenon is most in evidence
where the species concerned are able to exercise freedom of
choice in the selection of their food-plants. In the absence of
such freedom, under experimental conditions, it has already been
shown that host-preference may be insufficiently habituated to
preclude an insect from adopting new hosts. In cases where
host-preferences are most strongly developed they appear to have
led to the evolution of definite biological, or phytophagic, races
which are now known to occur in widely separated groups of
insects. Such races may further exhibit differences of form,
colour and behaviour.

The factor or factors involved in the exercise of host-selection
are, as yet, very little understood, and various hypotheses have
been advanced to account for the repetition of similar habits
in this respect from generation to generation. There is
considerable evidence in favour of the contention that for many
insects polyphagy is phylogenetically the older habit, oligophagy
and monophagy being more recent developments. Both Hering
for Lepidoptera and Mordwilko for Aphididas, who have specially
considered this problem, hold that it applies to the groups referred
to. A very large number of records occur in recent entomological
literature of insects resorting at times to unusual, or hitherto
unknown, food-plants. It may be argued that, in certain cases,
this sporadic habit is to be interpreted as a physiological mutation
in the species concerned. On the other hand, it appears equally
probable that such instances may be examples of the exercise of
a latent, more or less vestigial, former polyphagic habit. If
this latter explanation be the true one, it accounts for the varying
degree of potential polyphagy found in so many insects. Little
is known with respect to the host-preferences exercised by
different individuals of a true polyphagic species — whether they
do exhibit selection of the specific hosts upon which they them-
selves have developed in preference to all others. The beginnings
of oligophagy are initiated by preference of this kind, and the
more the oligophagic habit is developed, the more the species
loses its former trophic plasticity or power of accommodation to
new hosts, The repetition of the habit of specific host-preference

THE LOCUST PROBLEM .299

appears to be explainable as the result of a chemical influence,
derived from the food-plant, acting upon each successive
generation. Specific host-preference and its repetition in each
generation seems to afford an analogy with what Pavlov has
termed a conditioned reflex. The conditioning takes place during
the developmental cycle of the insect and influences the chemo-
tropic responses of the female during oviposition. It determines
the preference for selecting the same host-species upon which
her own developmental stages were passed. The faculty for
conditioning, which appears to be present in all insects to a
greater or lesser degree, is almost certainly inherited. The
chemical influence derived from the food-plant, on the other
hand, seems to determine the direction along which the con-
ditioning shall take place. It has already been shown that there
is some evidence that specific host-selection may become intensified
in successive generations, or, in other words, intensified specific
conditioning may occur. The question naturally arises as to
whether this is due to the passing on and accumulation in
successive generations of some chemical agency derived from the
food-plant ? There is, again, experimental evidence showing
that if an insect be transferred to a different food-plant, the
effects of this conditioning may become lost and a new conditioning
substituted. If, on the other hand, a specific conditioning be
repeated through a sufficient number of generations as to become
unalterable, it would appear to have directly affected the germ
cells of the insect. Up to the present, however, we have not
adequate experimental data in order to determine whether a
permanent effect can be so induced and reflected in a rigid host-
selection. The available evidence will hardly be considered by
most biologists to be sufficiently well grounded to demonstrate a
principle of inheritance so open to controversy.

Ecological Aspects of the Locust Problem

The more or less irregular periodicity of locust outbreaks has
long attracted attention, and various explanations have been
advanced to account for the phenomenon. There seems to be
very little doubt that the periodicity is the result of certain

300 THE PRACTICAL APPLICATION OF ECOLOGY

environmental influences acting either directly or indirectly upon
the insect. If this be correct it follows that a thorough ecological
study of each locust species in relation to fluctuations in external
conditions, over a period of years, would be the most profitable
way of arriving at an understanding of the factors regulating out-
breaks of these insects.

While the menace of locust swarms can be better counteracted
by the prompt application of artificial control measures than has
happened in the past, these measures still remain hopelessly
inadequate when combating great invasions. Being remedial in
effect their application may alleviate pressing needs, but the
menace of subsequent swarms is left unaffected. In other words,
they leave the main problem unsolved, i.e., the causes of locust
outbreaks and the areas of their origin.

In 1921 Uvarov advanced what has subsequently become
known as the phase theory of locusts. His views were first put
forward as a working hypothesis, but later observations conducted
both in the laboratory and in the field by Faure (1932) and other
workers have so fully corroborated its main contentions that the
phase theory of locusts is regarded to-day as an established
biological phenomenon.

The Phase Theory. The starting point of the theory is that
the gregarious species of Acrididae or true locusts are polymorphic.
Such species are not constant in all their characters, but are
capable of producing a series of forms differing from one another,
both morphologically and biologically. In their notable discussion
on polymorphism, Uvarov and Zolotarevsky (1929) propose a
standard phase nomenclature which, although referring more
especially to the well-known species, Locusta migratoria, will
probably prove applicable to the other species also. According
to their interpretation a locust can exist in three unstable biological
phases, namely, a solitary one, phasis solitaria ; a gregarious one,
phasis gregaria ; and a transitional phase between these two,
which they term phasis transiens. These phases differ from each
other in morphological and colour characteristics, on the one
hand, and in biological features (mainly behaviour) on the other.
The phases gregaria and solitaria are often so strikingly distinct

PHASES OF LOCUSTS

301

that they have hitherto been regarded as different species, which
has naturally resulted in confusion.

Taking the African migratory locust (Locusta migratoria
migratorioides) as an example, the essential characteristics of the
phases may briefly be diagnosed as follows (sexual differences
being excluded).

The phasis gregaria is the swarming or migratory phase which is
so highly destructive. It is characterised in the nymphal instars
by the great predominance of black and deep orange-yellow
coloration which develops quite independently of the nature of
the environment. The adults have the pronotum somewhat
concave, rather short and prominently constricted : there is
clearly no marked dorsal carina. Biometrical studies show
difference in the size and proportion of the tegmina and hind
femora as compared with those organs of individuals in the
solitary phase, the femora being relatively shorter and the tegmina
relatively longer. Colour differences are of minor importance in
diagnosing adult individuals of this phase, and the most reliable
characters are afforded by ascertaining the values of the following
ratios.

Table XVII

Ratio.

Gregaria.

Solitaria.

Length of legmen
Length of femur

Equal to or greater
than 2.

Always less than 2.

Length of pronotum
Width of head

Very near 1.

Notably more than 1 ;
about 1-58.

Height of pronotum
Width of head

Very near 1.

Notably more than 1 :
about 1-47.

Width of pronotum at
constriction

Notably less than 1 :
that is, 0-6 t8 0-8.

1, or slightly below 1.

Width of head

The phasis solitaria, or solitary non-swarming phase, is relatively
harmless. It is characterised by the nymphs being extremely
plastic in their coloration, which shows a strong tendency to

302 THE PRACTICAL APPLICATION OF ECOLOGY

THAN SI ENS

simulate that of the immediate environment. The adults have
the pronotum somewhat arched with an evident longitudinal
carina : there is no marked constriction and the pronotum is
relatively longer than in the gregarious phase. The hind femora
are also relatively longer and the tegmina shorter. The diagnostic
ratios are given in the table above.

The phasis transiens is not represented by any definite form,
but by a continuous series of transitional forms between the
solitary and gregarious phases. Such a series may be observed
either {a) when the transformation is from the solitary phase
towards the gregarious phase, when it may be termed phasis
congregans ; or (b) when the tendency is in the opposite direction,
when it is termed phasis dissocians. These two phases are

essentially of a biological nature,
but it appears that it may be
possible to distinguish them also by
minor details of structure and colour.
In the accompanying diagram the
GREGARiA general scheme of cyclic inter-
relations between the four phases
just described is represented. It
must be pointed out that the phasis
transiens may be represented by
forms morphologically and bio-
logically quite close to either p. solitaria, or p. gregaria, into both
of which the series merges almost imperceptibly.

Under experimental conditions, given the right combination of
temperature and humidity, plenty of sunlight and adequate
suitable food, it is easy to rear locusts in one or other of their
phases. If the newly hatched nymphs are reared under crowded
conditions, they show all the main features of the phase gregaria,
although it is doubtful whether absolutely typical gregaria are
always produced under these conditions. Under less crowded
conditions the locusts develop into phase transiens, while if each
early nymph be reared in a separate container it develops into
solitaria. Faure (1932) showed clearly that in this phase the
individual nymphs assume a coloration more or less resembling

SOLITARIA

PHASES OF LOCUSTS 303

that of the background of the cages in which they are reared.
Hertz and Imms (1937) have explored this subject further and
analysed the relation between this phenomenon and the spectral
and other qualities of the light rays that are involved. It was
also shown by Faure that green nymphs are produced as the result
of humid conditions and an abundance of moist green food
irrespective of the nature of the background. In the production
of the green nymphs, Hertz and Imms came to the conclusion
that high humidity alone is involved and that the threshold
of humidity in this connection appears to differ for different
individuals. Below this threshold green forms are not produced
and the insects assume a coloration in conformity with the nature
of the background. A green background, however, produces
yellowish forms and never green. It would appear that this
faculty of colour adaptation may serve a protective function of
rendering the developing insects less conspicuous in their environ-
ment. The green nymphs in a state of nature, it may be added,
occur among an abundance of green herbage when the humidity
would presumably be relatively high. Physiological differences
between the two extreme phases have not so far been explored
in any detail. Butler and Innes (1936), however, have shown that
the oxygen uptake of individuals in the gregarious phase is
markedly greater than those in the solitary condition. Strel'nikov
(1936) states that careful body-temperature readings indicate that
the black and orange hoppers of the gregaria phase absorb more
radiant heat than individuals (green) in the solitaria phase. In
the shade the body temperature of hoppers of both phases was
more or less alike and slightly above that of the surrounding
air. After six minutes in the sun the temperature of gregaria
rose by 15° C. and that of solitaria by only 10-2° C. On being
shaded, gregaria individuals lost heat more rapidly than solitaria.
Under crowded conditions the nymphs become extremely
active and this naturally affects the rate of metabolism of the
organism. Faure believes that more excretory products are
produced than are liberated by the Malpighian tubes and that
these, or their derivatives, become deposited in integument, thus
imparting the black and orange coloration that is an invaluable

304 THE PRACTICAL APPLICATION OF ECOLOGY

attribute of nymphs in the gregarious phase. Faure gave the
tentative name of " locustine " for the products of the excessive
metaboHsm which lead to the development of the gregaria
coloration. He points out that the black coloration appears in
nymphs of the gregaria phase an hour or less after hatching. It
could not, therefore, he says, be possibly due to the activity of
the nymphs themselves and concludes that it is due to the passing
on of locustine from the female to the embryo in the nutritive
yolk of the eg^. The interesting problem of the biochemical
nature of the pigments concerned is as yet unexplored. The
morphological characters of the thorax which differentiates the
two extreme phases are also an expression of this activity. Strains
and stresses accompanying muscular exertion appear to exercise
an influence which moulds the thorax into the form characteristic
of gregaria — the more typical or ancestral form being retained
by individuals in the solitaria phase. The hypothesis that the
characteristic pigmentation of the gregaria hoppers is produced
as a by-product of high metabolic activity receives support from
recent experiments by Husain and Mathur (1936). These authors
found that when hoppers of the locust Schistocerca gregaria, in
the solitaria phase, were forced to crawl a certain distance daily
in revolving gauze drums and cylindrical cages that rotated
vertically at intervals, they assumed the black coloration
characteristic of hoppers in the gregaria phase. The pigmentation
of the latter phase was also induced when individual solitaria
hoppers were placed in cages crowded with hoppers of Poecilocerus
pictus and adults of Chrotogonus.

In the phase solitaria the nymphs are relatively sluggish :
they sit about the food-plants quietly eating and undergo little
movement unless disturbed. We are still in the dark, however,
as to why mere crowding exercises an influence which appears to
activate a gregarious instinct — whether the actual stimulus
involved is, for example, contact, visual or odour, etc. Within
the limited space available it is not possible to discuss at all
adequately the immense and rapidly growing literature dealing
with the different aspects of locust life. The reader is referred
to the useful abstracts which appear monthly in the Review of

SWARMING OF LOCUSTS 305

Applied Entomology, Series A, for further information and
references to the hterature concerned.

Factors Governing Swarming. The first impulse for a locust
to increase in numbers appears to be provided by favourable
meteorological conditions. This is automatically followed by
evolution towards the swarming phases. Once swarming has
taken place a crisis is reached, a decline supervenes, and this
usually leads to the solitary phase.

In the case of Locusta migratoria, sub-sp. migratorioides, the
solitary phase has an enormous area of distribution, whereas the
gregarious phase only occurs in certain regions. The areas where
the latter phase can be produced are characterised by the extensive
development of reeds, bamboos and similar tall plants. It would
appear that such vegetation provides the most suitable food, but it
is also possible that the connection between the gregarious phase
of this insect and reed-beds may be due to the climatic conditions
associated with the latter. Such a uniform environment, providing
an abundance of food and shelter, is very suitable for mass
development of the locust, and the appearance of large numbers
of hoppers in a favourable year leads to transformation into the
swarming phase. The stimulus to increase in numbers appears to
be provided by dryness in tropical countries, while in temperate
lands a high temperature is of great importance in this respect.
Once the dense migrating swarms have left their breeding grounds,
the latter remain almost free until a fresh increase in numbers
results. Migrating swarms usually settle on lands very different
from whence they came. Difference of conditions appears to
reduce the numbers of the generation which results from their
egg-laying, and transformation into the solitary phase supervenes.
The activities of parasites, birds, and other natural enemies
increase at this period and further accelerate the transformation.
The hoppers of the solitary phase do not differ in their general
behaviour from those of other solitary grasshoppers, and do not
exhibit any tendency towards gregariousness.

The actual stimuli which promote the migratory flight are,
according to Uvarov, physiological. Lack of food can be dis-
missed, since the vast stretches of reed-beds in their breeding

306 THE PRACTICAL APPLICATION OF ECOLOGY

ground afford an inexhaustible supply upon which the winged
swarms make relatively little impression. There appears to be
no doubt that migration is intimately connected with the
development of the genital products, as happens also in the
case of the nuptial flights of termites and of ants. With locusts
the flight acquires a regularity and duration of such a character
that it goes on so long as physiological causes impel it. It may
terminate or only be interrupted owing to unfavourable environ-
mental conditions, but physiological causes such as the exhaustion
of the fat-body by oxidation, and the rise of body temperature,
resulting from flight, accelerating the maturity of the sexual
products, appear to be of greater importance. It is further
noteworthy that the available evidence indicates that the
emigration is not concerned with the quest for suitable new
areas for egg-laying. The settling of swarms in places totally
unsuited either for feeding or oviposition has often been recorded
and is totally opposed to this, at one time, prevalent belief.

There is obviously much that remains obscure in the phase
theory, and the various factors involved require deep and
searching investigation with respect to different species of locusts.
The theory, although an established fact, is still very largely
hypothetical as regards fundamental causes. It lends, however,
a new and essentially ecological interpretation of locust outbreaks.
By directing research along a defined course, it will co-ordinate
investigation, hitherto largely prosecuted blindly in the hope
of eventually alighting upon some solution of the problems
concerned. It opens up the possibility that mass transformation
of locusts into the gregarious phase may be foretold, while
discovery of the locations of the favoured breeding grounds may
lead to the destruction, by aeroplane dusting, of immense numbers
of potentially harmful individuals. Further, ecological study
opens up the possibility of altering the character of such breeding
grounds, either by cultivation or other means, and thereby
rendering them less potent sources of the invasion of other lands.

The Species of True Locusts. The species of true migratory
locusts are relatively few in number. In the light of the phase
theory their number has been about halved, since forms hitherto

SPECIES OF LOCUSTS

307

known under separate specific designations have been shown to
be merely phases of one or other kind. This has naturally involved
changes of nomenclature, and many old specific names must now
be replaced by standard phase designations {vide Uvarov, 1928).
1. The Migratory Locust {Locusta migratoria L.).

Phasis gregaria = migratoria L.

Phasis solitaria = danica L.
Although confined to the Old World, this insect enjoys a wider
range of distribution than any other species. Its range is
practically universal between latitude 60 N. and 20 S., besides
extending into New Zealand. Recent studies by Uvarov and
Zolotarevsky (1929), Zolotarevsky (1933) and Uvarov (1936)
have shown that, although this species is a very definite entity in
itself, it is divisible into a number of sub-species of which five are
at present known, each being characteristic of a geographical
area. The characters differentiating these sub-species are of a
relatively minor order and require expert determination : their
biometrical features are given in the table below. (1) The
Palaearctic Locust or sub-sp. migratoria L. (sen. str.), South-East
Russia. (2) The Central Russian Locust or sub-sp. rossica Uv.
and ZoL, Central Russia and Western Europe. (3) The African
Locust or sub-sp. migratorioides R. and F., Africa and Western
Asia. (4) The Madagascar Locust or sub-sp. capito ZoL,
Madagascar. (5) The Oriental Locust or sub-sp. manilensis
Mey., Malaysia, East Indies, Philippines, China.

Table XVIII. Average Measurements and Ratios of the ph.
gregaria in Different Sub-species of Locusta migratoria (from
Uvarov).

Males.

Females.

Sub-species.

E.

F.

E/F.

P/C.

H/a

M/C.

E.

F.

E/F.

P/C.

H/C.

M/C.

Rossica .

3-8

1-9

20

1-30

1-22

0-88

41

2-3

1-78

114

1-21

0-86

Migratoria

51

2-4

2-13

110

116

0-85

5-35

2-5

2-11

117

M6

0-87

Manilensis

4-26

2-08

211

1-11

MO

0-81

4-58

213

216

111

1-08

0-83

Migratorioides .

4-71

2-18

2-17

0-99

1-06

0-75

4-98

2-25

2-22

0-94

107

0-74

Capito

4-64

2-22

209

104

107

0-72

4-83

2-27

213

1-00

111

0-72

Note. E, length of elytron ; F, length of femur ; P, length of pronotum ; H, height of pronotum
M, width of pronotum at the constriction ; C, maximum width of the head,

308 THE PRACTICAL APPLICATION OF ECOLOGY

Each of these sub-species develops into the phases already
alluded to. The very wide range of this species, as a whole,
refers more especially to the limits to which odd specimens from
swarms extend and transformation into the phase solitaria takes
place. In the palaearctic region its breeding grounds, as at present
known, are centred more especially in middle Asia among the
vast reed-beds located in the vicinity of the Caspian and Aral
Seas and of Lake Balkhash. In Western Europe permanent
breeding grounds (swarm centres) no longer exist, most probably
owing to the spread of cultivation and its effects upon the original
ecological conditions of such areas. The Central Russian Locust
has its breeding grounds in small areas of sandy soil with sparse
vegetation. The African Locust has a most important breeding
ground in reedy areas in the Middle Niger region of the French
Sudan, from which all the recent invasions of that insect took
origin. The breeding grounds of the Madagascar Locust in
South Madagascar are in arid country with light soil and liable
to be flooded during rains. The Oriental Locust has diverse
types of breeding grounds : in China they resemble those of
the African Locust, while in Borneo, etc., they appear to be in
areas of humid tropical forest. It needs to be stressed that
comparatively little is yet known as to these breeding grounds
of each sub-species : their discovery and study forms one of the
most important aspects of the locust problem.

2. The Desert Locust (Schistocerca gregaria Forsk).

Phasis gregaria = gregaria Forsk.

Phasis solitaria = flaviventris Burm.
The Desert Locust has likewise a very wide distribution in
the Old World. It covers nearly the whole of Africa, besides
Arabia, Palestine, Syria, Persia and Western India. It is
definitely connected with dry desert and semi-desert conditions
and, ecologically, it differs markedly from the preceding species.
In point of destructiveness it deserves first place among locusts
of the Old World, and records of its ravages extend far back
into history, including Biblical times. A very extensive literature,
which is rapidly growing, exists with reference to this species, and
in particular the memoirs by Bodenheimer (1929) and by Ballard

SPECIES OF LOCUSTS 309

and collaborators (1932) may be mentioned. In the latter work
an illustrated account of the phases and their chief characteristics
will be found. Very little definite information is available as to
the permanent breeding grounds of this species owing to the vast
extent and difficult nature of the countries wherein they are
believed to be located. A good deal of evidence points to the
source of recent outbreaks as being in lands bordering the Red
Sea (Sudanese-Arabian region). The localisation of these breeding
grounds by aerial surveys and other means is one of the most
important of the schemes now being undertaken by international
arrangement regarding anti-locust investigation.

3. The Moroccan Locust (Dociostaurus maroccanus Thnb.).

Phasis gregaria = maroccanus Thnb.

Phasis transiens = degeneratus Brnov.
The area of distribution of this locust comprises dry, hilly, and
upland regions of those countries which form the Mediterranean
littoral and the greater part of Southern Russia. Its permanent
breeding grounds are the stony slopes of mountains and hills
where vegetation is somewhat scanty. From these regions
swarms invade the plains of neighbouring territory and the insect
is more especially prevalent in Spain, Cyprus, Transcaucasia and
Turkestan.

4. The South American Locust (Schistocerca paranensis 'Burm.).

Phasis gregaria = paranensis Burm.

Phasis solitaria = ? americana F.
Knowledge of South American locusts is still very incomplete,
and it has been claimed that several species are involved, but
the evidence on the whole points to there being only one of
true migratory habits. It is further probable that the North
American species S. americana is only the solitary phase of the
locust under consideration. The ^range of the South American
locust is from Argentina north of the fortieth parallel up
through parts of Brazil to Venezuela and Central America.
Migrating swarms further extend to Mexico and the West
Indies. Relatively little is known as to its bionomics and
periodicity, and its breeding grounds have not so far been
located.

310 THE PRACTICAL APPLICATION OF ECOLOGY

5. The Brown Locust (Locustana pardalina Walk.).

Phasis gregaria = pardalina Walk.

Phasis solitaria = solitaria Uv.
This insect is the chief destructive locust of South Africa,
where its swarms have been recorded at various times since 1653.
It has been studied in both the laboratory and the field by Faure
(1932), who found that the behaviour of the insect in the solitary
phase exhibits many of the phenomena also found in L. migratoria.
The nymphs exhibit similar colour adaptations to their environ-
ment and, likewise, green forms appear as the result of humid
conditions. Important breeding grounds are located in the
Karoo and in South- West Africa.

6. The Red Locust (Nomodacris septemfasciata Serv.).

Phasis gregaria = septemfasciata Serv.

Phasis solitaria = coangustata Luc.
The range of this locust includes the whole of South Africa,
together with the Southern Sudan and Abyssinia. In its solitary
phase only it is also known from Madagascar, Reunion and the
Comoro Islands. Two of its important breeding grounds are in
parts of Tanganyika and Northern Rhodesia, while the southern
shore of Lake Chad, along with other areas, are regarded as being
potential centres for swarm development.

7. The Rocky Mountain Locust (Melanoplus mexicanus Sauss.).

Phasis gregaria = spretus Walsh.

Phasis solitaria = atlanis Riley.
There is strong reason to suspect that the common North
American grasshopper Melanoplus (atlanis) mexicanus has an
intimate relation with the well-known but now extinct Rocky
Mountain Locust (M. spretus) which was formerly so dreaded a
pest in the United States. The differences between the two are
relatively small and suggest that spretus was the long winged or
gregarious phase of mexicanus. Faure (1933) conducted some
experiments in which the solitary form (mexicanus) was
transformed into the gregarious form or phase (spretus). There
seems in consequence every probability that the transformation of
this Acridian from one phase into the other occurs in Nature in the
presence of suitable ecological conditions. The disappearance of

GRASSHOPPERS 311

spretus may possibly be traced to changes in its former breeding
grounds on the slopes of the Rocky Mountains from whence
swarms invaded the American plains. It is generally maintained
that the spread of agriculture has altered these breeding grounds
to an extent which has led to this loss of capacity to develop
into the gregarious phase. The form mexicanus, however, still
occasions severe damage in Canada and the United States. During
outbreaks it shows a definite tendency to form loose swarms and
to become semi-migratory or gregarious. The conditions which
might favour this kind of behaviour to an extent that would
lead to the development of a true migratory phase are unknown.
It is, however, noteworthy that specimens collected in the field
under outbreak conditions showed a strong tendency to vary in
the direction of spretus as regards wing length and in size. It
would seem, therefore, that the species mexicanus occupies an
intermediate position between the true solitary grasshoppers and
migratory locusts in that it becomes at times exceedingly abundant
and develops incipient swarms.

Phase Development in Grasshoppers. It has been generally
believed that the existence of phases is a peculiarity belonging to
locusts and one not betrayed among non-migratory Acrididae, or
ordinary grasshoppers. Recently evidence has been brought
forward by Rubtzov (1935) showing that an essentially similar
phenomenon occurs in grasshoppers, but that the amplitude of
such phase differences is less pronounced than in locusts. Rubtzov's
important observations were carried out in Siberia, where he was
impressed by the colour variations found among individuals of
the same species of grasshoppers. Since some of the species in
which the phenomenon has been found occur often commonly in
Britain, a new field for study is opened up for workers in this
country. Certain species of grasshoppers were found by Rubtzov
to differ in coloration more or less in accordance with the
population density per unit area of territory. Individuals
collected from relatively dense associations, with up to 300
grasshoppers per square metre, were notably dark coloured,
larger, and had longer wings than those distributed in the
proportion of one, or fewer, over the same unit area. These

312 THE PRACTICAL APPLICATION OF ECOLOGY

differences are especially marked in Mropus sibiricus and
Chorthippus alhomarginatus, while a number of other species can
be arranged in descending order of their tendency to show marked
colour differences. The most pronounced differences are found
in A, sibiricus, which is an active species, tending to form dense
aggregations of individuals. These differences are almost equally
well shown in C. albomarginatus. In all cases the dark coloured
forms occurring collectively are interpreted as representing the
phase gregaria, while the pale examples found singly and sparsely
are regarded as representing the solitaria phase. This conclusion
is supported by such experimental evidence as is available.

Thus, individuals of C. albomarginatus, reared in isolation,
showed a tendency to develop into the pale coloured solitaria
form, while those reared gregariously, in crowded conditions,
produced a very definite gregaria type. Apart from the colour
differences alluded to, the supposed phase differences are also
betrayed in : (1) solitaria individuals being smaller, with femur
and tegmen shorter ; (2) shortening of the tegmen in solitaria
being more pronounced than that of the femora, the ratio being
greater in solitaria than gregaria ; and (3) variability in solitaria
being always greater than in gregaria — as can be seen by comparing
the maximum and minimum figures. Rubtzov's data are
accompanied by coloured figures portraying supposed phase
differences, and they unmistakably suggest that we have, among
non-swarming grasshoppers, clear evidence of the existence of those
same phases that feature so markedly in the economy of locusts.

A further phenomenon, discussed by Rubtzov, can only be
briefly alluded to here. He shows that a homologous series of
colour variations reveals themselves through many species of
Acrididae. These are inheritable variations and, according to
him, each such race possesses the potentiality, to a greater or
lesser degree, to exhibit such phase characteristics. The latter,
as it were, are super-added to the former in response to the
extreme conditions of individual life. In practice, it resolves
into the necessity, fully recognised by this author, of clearly
determining which race of a given species is being utilised in all
studies designed to explore the phase idea.

LITERATURE 313

Literature

A. Regulation of Crop Growth

Andrews, 1923. " Factors Affecting the Control of the Tea Mosquito Bug."

Indian Tea Association.
CuNLiFFE, 1928. Ann. App. Biol., XV., 473.

1929. Ibid., XVI., 135.

1936. Ibid., XXIII., 822.
CuNLiFFE and Fryer, 1924. Ibid., XL, 465.
CuNLiFFE, Fryer and Gibson, 1925. Ibid., XII., 516.
Davidson, 1934. Commomvealth of Australia, Dept. Sci. and Indus. Res.y

Bull. 79.
Flint and Larrimer, 1928. Illinois Nat. Hist. Survey Bull., XVII., Art. XI.
Frew, 1924. Ann. App. Biol., XL, 175.
Froghopper Committee (Trinidad and Tobago), 1925, et seq. Minutes and

Proceedings.
Fryer and Collin, 1924. Ann. App. Biol., XL, 448.
Hold AW AY, 1927. Commomvealth of Australia, Dept. Sci. and Indus. Res.,

Pamphlet 4. Melbourne.
Hopkins, 1919. Scientific Monthly, June, 496.
Jepson and Gadd, 1926. DejJt. of Agric, Ceylon, Bull. 78.
MacLagan, 1932. Bull. Ent. Res., XXIIL, 101, 151.
NouGARET and Lapham, 1928. U.S. Dept. Agric., Tech. Bull. 20.
Williams, 1921. Mem. Dept. Agric, Trinidad and Tobago, No. 1.
WiTHYCOMBE, 1926. Ann. App. Biol., XIIL, 64.

B. Resistant Varieties

Crane, 1936. Ann. App. Biol., XXIV., 188.
Davidson, 1922. Ibid., IX., 135.
MiSRA, 1920. Agric. Journ. India, XV., 627.
Monzen, 1926. Internal. Congr. Ent., Zurich, II., 249.
Parnell, 1925. Empire Cotton Growing Rev., IL, 330.
Roach, 1936. Ann. App. Biol., XXIV., 206.

C. Host-Selection and Biological Races

Ahmad, 1936. Journ. Anim. Ecol., V., 67.

Craighead, 1922. Journ. Agric. Res., XXIL, 189.

Field, 1910. Psyche, XVIL, 87.

Fryer, 1929. Trans. Internal. Congr. Ent., Ithaca, IL, 229.

GoscHEN, 1913. Zeits. iviss. Insektenbiol., IX., 72.

Hackett, Martini and Missiroli, 1932. Am. Journ. Ilyg., XVI., 137.

Harrison, 1927. Proc. Roy. Soc. B., 115.

Hopkins, 1916. U.S. Dept. Agric, Progr. of Work, 1917, 353.

Lal, 1934. Ann. App. Biol., XXL, 641. -

Larson, 1927. Ann. Ent. Soc Am., XX., 37.

Marchal, 1908. C. R. Soc. Biol., Paris, LXV., 2.

Marshall and Staley, 1935. Parasitology, XXVIL, 501.

Missiroli, Hackett and Martini, 1933. Rev. Malariol., XIL, 1.

Payne, 1933. Biol. Bull., LXV., 187.

PiCTET, 1911. Arch. Soc Phijs. Nat. Geneve (4), XXXL, 561.

RouBAUD and Gaschen, 1933. Bull. Soc Path. Exot., XXVL, 965.

RouBAUD, 1935. Ibid., XXVIIL, 443.

Schroeder, 1903. Verh. deutsch. zool. Ges., 158.

314 THE PRACTICAL APPLICATION OF ECOLOGY

Sladden, 1934. Proc. Roy. Soc. B., CXIV., 441.

1935. Ibid., CXIX., 31.
Tate and Vincent, 1936. Parasitology, XXVIII., 115.
Thompson and Parker, 1928. Bull. Ent. Res., XVIII., 359.
Thorpe, 1929. Journ. Linn. Soc. ZooL, XXXVI., 621.

1931. i6irf., XXXVII., 489.

1930a. Biol. Rev., V., 177.
Woods, 1915. Bull. 244, Maine Agric. Exp. Sta., 249.

D. The Locust Problem

Ballard and others, 1932. Min. Agric. Egypt, Bull., 110.
Bodenheimer, 1929. Zeits. angew. Ent., XV., 1.
Butler and Innes, 1936. Proc. Roy. Soc. B., CXIX., 296.
Faure, 1932. Bull. Ent. Res., XXIII., 293.

1933. Journ. Econ. Ent., XXVI., 706.
Hertz and Imms, 1937. Proc. Roy. Soc. B., CXXII., 281.
HusAiN and Mathur, 1936. Ind. Journ. Agric. Sci., VI., 591.
RuBTZOv, 1935. Bull. Ent. Res., XXVI., 499.
Strel'nikov, 1936. Vide Rev. Appl. Ent. A., XXIV., 312.
UvAROv, 1921. Bull. Ent. Res., XII., 135.

1928. " Locusts and Grasshoppers." London.
UvAROv, 1936. Bull. Ent. Res., XX., 261.
UvAROV and Zolotarevsky, 1929. Ibid., XXVII., 91.
Zolotarevsky, 1933. Ann. Epiphyt., XIX., 47.

CHAPTER XII
PARASITISM

Introductory Remarks, p. 315. Host-selection, p. 318. Phases
OF Parasitism. Multiparasitism, p. 322 ; Superparasitism, p. 325 ;
Hyperparasitism, p. 326.

Introductory Remarks

In discussing parasitism among insects it is necessary, in the
first instance, to define the Hmits of the subject. A generally
accepted definition describes a parasite as a living organism
which takes up its abode temporarily or permanently on or within
some other living organism (its host), and at whose expense it
obtains its nourishment. This definition is generally held to
imply that the host has developed tolerance to the presence of
the parasite, and consequently suffers no fatal effects from its
activities.

(a) Among insects there exists a considerable number of species
whose behaviour is covered by the above definition. They are
true parasites and include the kinds discussed in text-books of
medical entomology and parasitology. Such parasites are almost
always insignificant in size as compared with their hosts which,
it may be added, are almost always vertebrates. Anoplura, (Estrid
larvae, Aphaniptera, Streblidse, Nycteribiidge and Hippoboscidae
are familiar examples.

(h) There are, however, a much larger number of insect parasites
which exhibit a different behaviour. They belong to various
families of Diptera and Hymenoptei^a, but are very rare outside
those orders. They are parasites only as larvae, while the adults,
into which they transform, are highly developed, free-living
creatures. Their hosts are nearly always members of their own
class, and are nearly always destroyed as the result of the
parasitism. Furthermore, such parasites are relatively large in
size in comparison with their hosts, and this fact appears mainly

315

316 PARASITISM

accountable for their fatal activities. At the outset they behave
as true parasites, deriving their sustenance mainly from the blood
of their hosts, which suffer little ill-effects and continue to feed
and grow. Sooner or later the fat-body and less vital organs are
attacked, and the parasites become, in the biological sense,
internal predators. Growth of the hosts at this stage virtually
ceases, but they continue to live, and often feed voraciously.
As the parasites near maturity their carnivorous propensity
intensifies, and all food-yielding parts of their hosts are ultimately
consuined until only the exoskeleton remains. As their final
economy they, as often as not, conduct their pupal transformation
within the husks of their hosts, and thus sheltered in these grim
cells they are able to dispense with cocoons. " Fatal " parasites
of this kind have been termed by O. M. Reuter, and subsequently
by W. M. Wheeler, " parasitoids." Among other entomophagous
parasites the Strepsiptera are to be regarded as non-fatal
parasitoids, since their hosts are not usually destroyed as the
direct result of the parasitisation ; in this respect they are
annectant between a and b.

(c) Parasitoids may live either externally or internally in
relation to their hosts. It seems tolerably clear that the
ectoparasitic life is the simpler condition of the two, and that the
endoparasitic method is a development from it. There is every
probability that the parasitoidal mode of life originated from
simple predatism. A predaceous insect is one which seeks out
and devours a succession of living organisms or prey until its
needs are satisfied. It follows, therefore, that a true predator
must have its locomotory and sensory organs sufficiently developed
to equip it for this mode of existence. In the eventuality of a
single individual prey being large enough to provide a carnivorous
insect with all the food it needs, there would seem to be little
difference between a predator and an " ectoparasitoid." Since,
however, the larval life of the latter is spent in close and continuous
association with a single host, and it exhibits the degeneration
correlated with such an existence, it shares the characteristics of
a parasite rather than of a predator, and this distinction is the
one commonly adopted.

TYPES OF PARASITISM 317

(d) A third type of parasitism is met with among those social
insects which have themselves adopted a life of parasitism in the
colonies of other social insects. They parasitise a colony instead
of an individual of a species and obtain their food and that of
their larvae at the expense of supplies brought in by the members
of the colony. Social parasites occur among ants, bees, wasps
and termites, and their behaviour has been fully discussed in
numerous contributions by W. M. Wheeler.

(e) Among solitary wasps and bees there is found an inter-
grading series between parasitism on the one hand and inquilinism
on the other. In many cases it is difficult to draw any hard
and fast lines between the several types of behaviour. In the
primitive Vespoid families Scoliidae, Thynnidae and Mutillidae
the parent insect seeks out insect larvae living in concealed
situations ; these they immobilise by stinging and then lay their
eggs upon them, and their larvae are parasitoids living externally
to their hosts. Most other solitary wasps store up in their nests
one or more insects which they have previously paralysed, and
attach one of their eggs to the body of the prey. The wasp
larva upon hatching behaves as an ectoparasite (or predator) and
devours the prey ; or, if there be several, it consumes them in
succession. This behaviour differs but little from that of the
Scoliid and other wasps except that the parent insect transfers
the prey from its natural habitat into the nest. Many parasitic
bees lay their eggs in the cells of other bees of allied genera,
among the stores of honey and pollen. The parasitic larva,
upon emergence from the egg, finds itself in close proximity to
the normal occupant of the cell. In cases studied by Graenacher
the parasitic larva has more powerful jaws than its rival, and
easily succeeds in destroying it. It then proceeds to consume the
food-supply and casts its enlarged mandibles with the first ecdysis.
There are, again, other so-called parasitic wasps whose larvae are
solely nourished upon the food-stores of the host, the parent
removing and destroying the egg or young larva of the host
preparatory to the emergence of its own offspring. In Sapyga
quinquepunctata and ♦S'. similis, which live at the expense of solitary
bees, both Fabre and Nielsen have shown that the long ovipositor-

318 PARASITISM

like sting is used for piercing the cell and depositing an egg within.
The first instar Sapyga larva devours the egg of the host and
subsequently feeds solely upon the vegetarian stores within the
cell.

The foregoing remarks emphasise the diversity and intergrading
exhibited by parasitism among insects. Growth of recent
knowledge of the subject is too extensive to allow of adequate
treatment of all its phases within a restricted compass. For this
reason discussion is mainly limited to the parasitoid Hymenoptera
and Diptera (Tachinidae).

Host-selection

Our knowledge of the fundamental aspects of parasite specificity
admits of but few generalisations. Notwithstanding the mass of
information which exists relative to the hosts of entomophagous
parasites, we are completely in the dark as to the determining
causes which lead certain parasites constantly to select certain
host-species, while others utilise hosts of different taxonomic
relationships. The vast majority of parasites are but seldom
restricted to a single host-species, not many are confined to an
individual genus, and those so accredited are becoming fewer in
number with increasing knowledge.

Among Hymenopterous parasites examples of more or less
restricted host specificity are numerous. Thus, the Proctotrypid
Embidobia is only known to develop in the eggs of Embioptera ;
the Proctotrypid Polygnotus selects Cecidomyid larvae as its
hosts ; the whole family Scelionidse are egg-parasites, while
the species of Scelio appear to attack only locust eggs ; the
Braconid Dachnusa parasitises Dipterous larvae of the family
Agomyzidae ; the Cynipid Ibalia and the Ichneumonid genera
Rhyssa and Thalessa are parasites of Siricidae ; Aphidius and its
allied Braconids attack Aphididae, while another Braconid genus,
e.g., Perilitus, confines itself almost exclusively to adult Coleoptera ;
and so on.

Among Tachinid parasites specificity of host-selection is, as a
rule, far less restricted than in Hymenoptera, but the exceptions
present certain puzzling problems. Thus the species of the

HOST-SELECTION 319

Phasiine group mainly attack Pentatomidse, but, as Thompson
and Parker (1927) point out, the genera, Frercea, Weberia and
Rondania are parasites of Coleoptera. Of four related Dexiine
genera, Thelaira, Fortisia, Cyrillia, and Thrixion, their hosts are
Lepidopterous larvae, the Chilopod Lithobius, the Isopod
Metaponorthus and certain Phasmidae respectively. The relations
of Tachinids parasitic upon woodlice (Oniscidae) have formed the
subject of a detailed study by Thompson (1934). The available
data appear to suggest that the parasitisation of woodlice depends,
in the main, on the habits of the latter. Those Oniscidge living in
large colonies, under loose bark of logs, are the kinds most usually
parasitised, for the reason that they frequent situations favourable
to attack.

Among the many parasites of polyphagous habits it will suffice
if two examples be quoted. The Tachinid Compsilura concinnata,
which was introduced from Europe into the United States in
1906, has been reared, according to Webber and Schaffner (1926),
from over ninety species of American insects representing three
orders — Lepidoptera are mostly selected, and no less than eighteen
families are included in their host list. There appears to be no
one distinctive character common to all its hosts — neither colour
nor colour-pattern ; hairiness or absence of hairs ; low-feeders,
bush-feeders or tree-feeders ; habit, whether the hosts be
gregarious or solitary, diurnal or nocturnal ; specific food-plants,
whether they be Gymnosperms, Monocotyledons, Dicotyledons or
lichens ; odour or absence of odour ; presence or absence of
urticating hairs, etc. — as Thompson points out, all these categories
are represented among the host-species, and from whatever
standpoint this varied assemblage be considered its members have
in common only the fact that they are all hosts of Compsilura.
Among Hymenopterous parasites a parallel example is afforded
by the Ichneumon Hemiteles areator, which has been reared from a
remarkable range of hosts comprising many Lepidoptera, various
Hymenoptera, including other Ichneumonidse, besides several
Coleoptera and Diptera.

In the two foregoing examples the insects mentioned are
endoparasites, and it appears probable that at least some of the

320 PARASITISM

innate determining factors are physiological. Certain experiments
recorded by Thompson and Parker (1927) are significant in this
connection, who mention that although larvae of the Tachinid
Sturmia scutellata develop freely in caterpillars of Porthetria
dispar they die immediately after penetrating into those of
Hemcrocampa leucostigma and Notolophus antiquus, which belong
to the same family. On the other hand, Sturmia larvae develop
freely in caterpillars of Malacosoma, which belong to a different
family — the Lasiocampidae. Webber and Schaffner mention that
one of the numerous hosts of Compsilura concinnata is also
Porthetria dispar, but yet this Tachinid rejects caterpillars of
Malacosoma, which are acceptable to the Sturmia. The opinion
has been advanced that the activities of the phagocytes of the
hosts constitute a defensive mechanism against endoparasites. It
is claimed that species of the latter are destroyed by this agency,
and only those kinds specially adapted by physiological or structural
features are able to survive and become regular parasites of specific
hosts. The evidence in support of this contention is scanty, and
the theory is discussed in a recent paper by Thompson (1934, p. 433).
The host-preferences of ectoparasites present several peculiar
features. The ectoparasitic mode of life is unknown, for example,
in the whole family of the Tachinidae, whereas it is frequent
among diverse groups of Hymenopterous parasites. The factors
that determine whether a species lives within or external to its
host are evidently fundamental, since the endoparasitic mode of
life involves a complete change in methods of respiration. In this
connection it is remarkable that certain parasite genera, notably
the Chalcid Aphelinus, include among their species some which are
endoparasites and others that are ectoparasites. It is open to
speculation whether the internal physiological reactions of the host
are a determining factor which force, as it were, certain parasites to
live externally, while better adapted, but closely related, species
are able to pursue an endozoic existence. Whether such reactions
are a potent factor in this case or not is scarcely amenable to proof,
and, at any rate, the capacity of a species to undergo the correlated
respiratory changes involved in an endoparasitic mode of life is
at least as important.

HOST-SELECTION 321

The outstanding feature with respect to host-selection by
ectoparasites is that they very rarely affect hosts living openly
and exposed. This peculiarity is very constant and it may
possibly have originated as an adaptation whereby the exposed
parasites are enabled to secure a measure of protection from the
attacks of hyperparasites. Among parasites exhibiting this kind
of host-preference the Chalcid Melittohia acasta affords an
instructive example, since it selects a wide range of victims, which
are distributed among the orders Diptera and Hymenoptera.
As Thompson and Parker point out, in all cases the hosts are
enclosed in cells or nests of mud or " paper," in silken cocoons or
in chitinised puparia. Also, in their analysis of features common
to these hosts they find that the latter are at least twice the size
of the parasite and are surrounded by a protective envelope which
is separated from the body of the host by an intervening space.
Genieys (1925) has made an experimental study of the Braconid
Habrobracon brevicornis, which is an ectoparasite of naked and
more or less sedentary Lepidopterous larvae living in galleries in
plants or immersed in stored vegetable products. His efforts
to induce this species to parasitise hosts that live exposed or are
densely hairy usually failed ; in such cases oviposition either did
not occur or the young parasitic larvae generally failed to develop
after hatching. Certain pupal parasites of Cyclorrhaphous
Diptera are sensu stricto ectoparasites, since they live attached to
the outsides of the pupae just beneath the puparial shell. The
larval stages of several species of the Staphylinid genus Aleochara
and of the Chalcid Mormoniella vitripennis (Nasonia brevicornis) y
for example, are passed in this manner.

In a few exceptional instances, exposed, free-living hosts are selected.
Certain species of the Chalcid genus Euplectrus are ectoparasites of
Noctuid larvae which feed openly on Jow plants. This genus is also
exceptional in that it has acquired the habit of cocoon-formation
which is wanting in other Chalcids {vide Thomsen, 1927). The cocoon
substance is a product of the Malpighian tubes which is discharged
through the anus, and it would seem that its utilisation in this manner
has been evolved as a protection for the vulnerable pupae. In another
Chalcid, Schizonotus sieboldii, Cushman has shown that its larvae are
ectoparasites of the exposed pupae of the beetle Melasoma interruptum.
The Ichneumon Paniscus parasitises Noctuid and other caterpillars

K.A. ENTOMOLOGY. 11

322 PARASITISM

which usually form their cocoons shortly afterwards ; Polysphincta is a
parasite of spiders and the Vespoid Rhopalosoma pceyi is similarly an
ectoparasite of the tree cricket Orocharis.

Phases of Parasitism

Multiparasitism. Multiparasitism is the simultaneous infesta-
tion of the same individual host insect by two or more species of
primary parasites. The phenomenon is more usually known as
superparasitism, but it is proposed to limit the use of the latter
expression to instances in which an individual host is attacked by
tw^o or more parasites pertaining to the same species.

The prevalence of multiparasitism is frequently dependent
upon whether or not the female parasites, at the time of oviposition,
are able to distinguish between unattacked healthy hosts and
those already parasitised. In other words, its occurrence would
appear to be dependent upon errors of instinct. It also obtains
in other cases as the result of pressure of competition induced by
a high ratio of parasites to host-population, when oviposition
in at least some proportion of already parasitised individuals
becomes unavoidable. Irrespective of the inducing cause, multi-
])arasitism plays an important part in the economy of parasites,
but it has so far been subject to very little exact experimental
investigation. The manifestations of the phenomenon are various,
and the following classification covers the phases which prevail.

1. Multiparasitism Involving the Survival of a Single Parasite
Species. This occurrence appears to be very prevalent and the
survivor may bring about the death of the other parasite or
parasites in one of two ways, viz., either by direct attack or by
the indirect effects of its presence. The studies of Pemberton
and Willard (lois) on the inter-relations of fruit-fly parasites in
Hawaii have shown that when the Braconids Diachasma tryoni
Cam. and Opius humilis Silv. occur together in the same individual
fruit-fly larva? the last-mentioned species was usually killed. It
succumbs as the result of wounds and lacerations inflicted upon
it by the long, curved, sickle-like mandibles of the newly hatched
Diachasma larvae. They mention that from a total of 627 fruit-
fly puparia examined, 143 contained larvae of the two species of
parasites, and that in 133 of these the Opius larvae were found to

MULTIPARASITISM

323

have been destroyed. Pierce (1910), in his paper on insect enemies
of weevils, enumerates a number of examples of multiparasitism
involving competition between two species of parasites ; the
surviving species is recorded where known, but whether it gained
the ascendancy by direct attack upon its competitor or by indirect
nutritional effects was apparently not determined.

In other cases of multiparasitism a passive struggle for existence
results, and the death of one or other parasite supervenes upon
exhaustion of the food-supply by the survivor. As a rule those
parasites which attack a given host earlier reach the destructive
feeding stage first and are consequently survivors in the struggle.
Thus, Tothill (1922), in discussing the parasites of Hyphantria
cunea, mentions that one caterpillar which he dissected contained
twenty-two parasitic larvae belonging to four species, and it is
doubtful whether any would survive to complete their trans-
formations. Sometimes the larvae of the Ichneumon Campoplex
succumbs as the result of the presence of the Tachinid Ernestia, and
sometimes the opposite occurs. When another Tachinid, Therio7i,
enters into competition it stands little chance of survival, since it
is the latest to attain the destructive feeding stage. It will be
observed in the accompanying table that the incidence of the
different parasites affecting the Hyphantria is more or less in a
sequence which consequently reduces the possibilities of
multiparasitism to a considerable degree.

Table XIX. Sequence of Parasites of Hyphantria in 1913
(after Tothill).

Egg Stage.

Larval Stages.

Pupal Stages.

Parasites.

Fresh.

Old.

l3t.

2nd.

3rd.

4th.

5th.

6th.

7th.

Pre-
pup.

Fresh.

Old.

Adult.

Apanteles .
Meteorus
Campoplex 1
Campoplex 2
Ernestia
Therion
Pimpla, etc..

324 PARASITISM

Since the Apanteles attacks only very young caterpillars, and the
other parasites do not, it is only subject to competition in its
later instars — when it is most likely to prove the survivor. Pimpla,
on the other hand, attacks only pupae, and for that reason is
largely out of range of severe competition.

Another noteworthy example is instanced by Muesebeck (1918)
in his studies of the parasites of the brown-tail moth {Euproctis
chrysorrhosa L.). As a result of the dissections of over 13,000
host caterpillars it was found that whenever larvae of the Braconid
Apanteles lacteicolor enter into competition with those of Meteorus
versicolor Wesm. or of the Tachinid Zygobothria nidicola Town.,
or with both species, these latter are killed, evidently as the result
of some toxic secretion induced by the Apanteles.

2. Multiparasitism without parasite mortality. This phase is
probably by no means infrequent, especially with regard to the
parasites of Lepidopterous larvae and of the Lecaniine scale-insects.
In the case of Coccus caprece of the hawthorn, multiparasitism is
a common event, the parasites involved being the Chalcids
Blastothrix sericea (hritannicd) and Aphycus punctipes (melano-
stomatus). In localities in Britain where these two parasites are
abundant, a dozen or more examples of both species may sometimes
be reared from the same individual host. Pierce mentions an
instance among weevil parasites in which nine examples belonging
to two species of Aphiochceta were reared from a single host, and
cases where two species of ectoparasites were reared under similar
circumstances.

The effects of the competition upon the insects concerned are
usually in proportion to the numbers present. The most obvious
result is the development of individuals stunted or reduced in size,
and according to Fiske material reduction in their reproductive
capacity is also common.

3. Multiparasitism involving the death of the parasites.
Instances of this kind have been seldom recorded, but it appears
that the fatal consequences are the result of exhaustion of the
food-supply or of the premature death of the host. In his study
of the parasites of the white-marked tussock moth L. O. Howard
states that, when the Ichneumon Pimpla inquisitor oviposits in

SUPERPABASITI8M 325

caterpillars already infested with Tachinid larvse, both types of
parasites succumb in every case, with but one exception. Pierce
found with weevil parasites that survival was rare when more
than two species of parasites were involved.

Superparasitism. Superparasitism is but a phase of multi-
parasitism in which two or more parasites pertaining to the same
species occur in the same individual host. In many cases it is a
normal and regular occurrence, and results in no fatal effects
upon the parasite individuals involved. It is especially prevalent,
for example, among Braconids and Chalcids, as records of rearings
of these parasites testify. In the polyembryonic Chalcids super-
parasitism is obligatory, and as many as nearly 3,000 individuals
of a single species have been reared from one individual Plusia
larva. In other cases its occurrence is irregular, and is the
result either of errors of instinct or of pressure of competition
induced by scarcity of suitable hosts. In such instances fatal
consequences commonly supervene. Thus H. S. Smith (1912)
records finding up to five planidia ^ of the Chalcid Perilampus
in a single individual host, but in no instance out of many hundreds
has more than one adult Perilampus emerged. It appears that
the individual most advanced in its development gains the
ascendancy by starving out its confreres, either through devouring
the host, or by rendering it unsuitable as a source of food.
E. W. Wheeler (1923) states that from one to eight larvae of
the Braconid Aphidius phorodontis may occur within the same
individual aphid host, but never more than one attains maturity.
This same fact has been noted by Spencer and other observers
of Aphidiine parasites. Timberlake (1912) has induced super-
parasitism under obligatory experimental conditions. In one
experiment five females of the Ichneumon Limnerium validum
were placed in consort with ten caterpillars of the brown-tail
moth. All the hosts were heavily attacked, and succumbed before
the parasitic larvae had grown to any appreciable extent, as the
result of the parasitisation.

In many cases it is noteworthy that the female parasites are
guided by their own instincts so as to avoid hosts which they
^ For definition, see p. 333.

326 PARASITISM

themselves, or other members of their species, have already
parasitised. Thus, Haviland (1922) has observed that the
Braconid Dacnusa areolaris Nees never attacked larvae of the
leaf-miner Phytomyza angelicas if they were already parasitised,
and in the considerable quantity of material examined two
parasites were never found in the same individual host.

Salt (1934-37) has shown that the Chalcid Trichogramma
evanescens is able to distinguish between parasitised and
unparasitised hosts. Before parasitising a host (which is always
an insect ^gg) the female walks repeatedly over it and in course
of that process an odour is left on the egg-surface which can be
perceived by another female coming later. If this odour be
removed by washing the host-egg the second parasite is unable
to detect that the egg has already been attacked. After inserting
the ovipositor, however, the female becomes aware of the egg
being already parasitised and soon goes away without having
oviposited. If hosts are relatively few in number superparasi-
tisation occurs : this does not happen immediately, since the
Trichogramma female is able to withhold her eggs for a period
rather than lay them in parasitised hosts. As the density of the
parasites in relation to a fixed population is increased the more
intensive the superparasitisation becomes : when it is severe there
is either no resulting progeny, or the individuals of the latter which
emerge are dwarfed or imperfect. When given a choice of hosts
Trichogramma selects the larger of the eggs : also, when the
number of hosts is limited the larger ones are selected for super-
parasitisation. It further transpired that the females are unable
to distinguish between suitable and unsuitable non-parasitised
hosts and will select the latter for oviposition if of larger size than
the former. When given false hosts of various kinds such as flour
particles, tiny glass globules, sand particles. Lobelia seeds, etc.,
the insect attempted to oviposit in these objects, where they were
of larger dimensions than the true hosts which were available
along with them. Whether this error of instinct, so evident under
experimental conditions, often prevails in a state of Nature is
unknown.

Hyperparasitism. Probably comj^aratively few primary para-

HYPERPARA8ITISM 327

sites, of a given insect, are immune from attacks by their own
specific parasites, which are termed hyperparasites or secondary
parasites. Hyperparasitism, consequently, is a significant factor
in the maintenance of the balance between insect species in Nature,
and the biological association of hosts, primary parasites and
secondary parasites is a highlj^ adjusted complex. It is probable
that hyperparasitism is a development of multiparasitism and,
in some cases, it is difficult to distinguish between the two
phenomena, since one may intergrade into the other.

By far the largest number of hyperparasites are members of
the Ichneumon sub-families Cryptinse and Ophioninae, and of
various families of Chalcids, but it is among the latter group
that they are the most numerous. The Proctotrypoidea and
Cynipoidea also comprise a considerable number of hyperparasites.
Outside Hymenoptera secondary parasitism is very rare, and the
few known examples include certain Bombyliidse among Diptera
and two records among Coleoptera [vide Muesebeck and Dohanian,
1927).

In frequent cases it is not possible to distinguish, absolutely,
certain species as primary parasites or as hyperparasites, since
their behaviour is plastic and largely determined by the types of
hosts adopted. Timberlake (1913), for example, found that the
Chalcid Coccophagus lecanii Fitch is a common primary parasite
of the soft scale (Coccus hesperidum), but, when associated with
other Chalcid primary parasites (Microterys or Aphycus), may
becomes hyperparasitic.^ The Chalcids Monodontomerus cereus
Walk, and Eupteromalus nidulans (Foerst.) occur as primary
parasites of the gipsy moth and brown-tail moth respectively, but
both species also frequently become hyperparasitic through
Braconid or Tachinid parasites of those hosts. Many examples of
this character are known, and further references will be found in
the excellent bulletin by Muesebeck and Dohanian.

When two or more secondary parasites occur in the same

^ It may be mentioned that Coccophagus is normally parthenogenetic and
thelyotokous, producing females generation after generation, but when it is
hyperparasitic, Timberlake states that the resulting progeny are all males.
This unique phase has been ascertained to be prevalent both in the field and
under laboratory conditions.

328 PARASITISM

individual host super- or multiparasitic relations are involved.
An example of this kind is instanced by Haviland (1922a) with
respect to hyperparasites of aphides through the Braconid
Aphidius. The Aphidius may be attacked by the endoparasitic
Cynipid Charips or by the ectoparasites Lygocerus (Proctotry-
poidea), Asaphes or Pachycrepis, the two latter being Chalcids.
The Charips, along with its Aphidius host, perishes as the result of
attack by one or other of these ectoparasites. The latter, on
their part, frequently suffer from mutual competition. Thus,
Haviland states that when two of these hyperparasites occur
together, and they may belong to the same or to different species,
only one survives or both may perish consequent upon the
exhaustion of the nutritive supply.

Hyperparasites in general are far less restricted in their host-
selection than most Hymenopterous primary parasites. They
are able to adapt themselves to a wider range of hosts, especially
in the absence of preferred species — a trait which explains the
heavy attacks certain introduced primary parasites undergo in
countries where the normal and regular hyperparasites are absent.
The Chalcid Dibrachys cavus (houcheanus Ratz.), for example, will
attack an immense range of hosts, including many Lepidoptera and
Diptera and several Coleoptera ; as a rule it behaves as a hyper-
parasite, but in other instances it is a primary parasite,
especially of the larvae of Microlepidoptera. In many cases, when
Hymenopterous parasites exercise a wide and apparently indis-
criminate range of host-selection, they will prove to be secondary
rather than primary parasites.

The vast majority of hyperparasites live externally to their
immediate hosts, which are either concealed within the original
host or enclosed in puparia or cocoons. The internal feeding
habit is consequently rare ; it obtains in Charips, already
mentioned, and also as a special adaptation in certain other
instances.

Parasites of a tertiary order are often referred to in literature,
but proof that this type of parasitism is a constant and obligatory
feature of the species concerned, and not merely a phase of
multiparasitism, is in many cases non-existent. As an example

HYPERPARASITISM 329

of what appears to be true tertiary parasitism, the observations
of Muesebeck and Dohanian on the Chalcid Pleurotropis are
noteworthy. Out of several hundred cocoons from which the
species P. tarsalis Ashm. emerged, in 97 per cent, it had developed
at the expense of Hymenopterous secondary parasites. Only in
very few instances the Pleurotropis itself behaved as a secondary
parasite of the host concerned (Apanteles). Since the Pleurotropis
is an internal parasite, which pupates within the larval or pupal
covering of the host it destroys, its actual biological relationships
can be readily determined. Reputed quarternary and even
quinquenary parasites are on record, but it is extremely doubtful
whether their behaviour is of the order ascribed to them.

CHAPTER XIII

PARASlTlSM—coiiiinued

Types of Life-Cycle. Group A, p. 331 ; 1. digestion of Parasite
Eggs by the Host, p. 331 ; 2. Types of Migratory Primary Larvce, p. 333 ;
3. Non-Migratory Primary Larvce, p. 336 ; Group B, p. 337 ; Group C,
p. 339 ; Group D, p. 340. Host Relations of Parasites, p. 343.
Hymenoptera, p. 344 ; Diptera, p. 351. Polyembryony and Related
Phenomena, p. 353.^ — (a) Chalcidoidea, p. 354 ; (b) Proctotrypoidea,
p. 356 ; (c) Ichneumonoidea, p. 358 ; {d) Vespoidea, p. 358 ; Origin of
Polyembryony, p. 358 ; The Embryonic Membranes, p. 359 ; Origin of
Mixed Broods of Parasites, p. 361. Adult Parasites. Feeding
Behaviour, p. 363 ; Phoresy, p. 364. Literature, p. 365.

Types of Life -Cycle

Probably no branch of entomology has grown so rapidly in
recent years as our knowledge of the biology of entomophagous
parasites. It has revealed an amazingly wide diversity of phases
in the life-cycle and of adaptations to different modes of life.
This subject has become too extensive to admit of discussion from
every aspect in a limited space, and it is only intended to lay the
more important facts before the reader in a summarised and
co-ordinated form.

In a broad general way the life-cycles of parasitoid Diptera and
Hymenoptera may be grouped into four categories.

Group A. The eggs, or more rarely the larvae, are not
deposited on the hosts, but in situations favourable to the
latter becoming infested. Post-embryonic development may
be either ecto- or endoparasitic.

Group B. The eggs are deposited externally on the hosts
and an ectoparasitic life supervenes.

Group C. The eggs or larvae are deposited externally on
the hosts, but the subsequent development is endoparasitic.
Group D, The whole parasitic life is endoparasitic from

330

TYPES OF LIFE -CYCLE

331

the outset, the eggs or larvae as the case may be being inserted
within the bodies of the hosts.

Each of these types of hfe-cycle and the adaptive features
involved will be briefly discussed.

Group A. Since the eggs, or, in other cases, larvae, are not
actually deposited on the hosts it is obvious that special provision
is necessary whereby the parasites are enabled to obtain access
to the latter. Such provision is forthcoming either (1) by very
high fecundity and the deposition of minute eggs adapted to be
ingested by the host during the ordinary process of feeding ; or

(2) by the primary larval stage being essentially a migratory phase
enabling the parasite to discover its hosts by its own efforts ; or

(3) by the primary larva developing an erect stationary habit and
awaiting an opportunity of

attaching itself to the first
suitable host coming within
reach.

1. The Ingestion of Parasite
Eggs by the Host. This habit
is peculiar to certain of the
Tachinidge and its prevalence
was made known by the earlier
researches of Townsend (1908)
and especially of Pantel ( 1 909 ) ;
more recently Thompson (1924) has considerably extended our
knowledge of the subject. These observers have shown that
species which lay very large numbers of minute more or less
fusiform eggs (microtype eggs of Pantel) deposit them upon the
food-plant of the host. The larvae within these eggs are fully
formed, and only await a favourable opportunity for emergence.
The species Gonia capitata, a para^site of the prairie cutworm
(Porosagrotis orthogonia) in Canada, has been studied by Strickland
(1923) and may be taken as an example. It lays upwards of
4,000 small black eggs, measuring 0-24 mm. long X 0-15 mm.
wide and 0-12 mm. high, which do not normally hatch until
they are swallowed by the cutworm larva along with its food
(Fig. 75). Upon coming in contact with the digestive juices the

Fig. 75. Eggs of a Tachinid fly (Gonia
capitata) laid on blade of wheat.
X 50. a, Adhesive material.

(After Strickland, Tech. Bull. 26,
Dept. Agric.y Canada,)

332

PARASITISM

chorion speedily ruptures and the larva issues forth. After a
variable period in the mesenteron it bores its way through the gut
wall and migrates to the supra-oesophageal ganglion. Within this
nerve centre it feeds apparently very little and subsequently
escapes into the body-cavity of the host. Here it does most of its
active feeding, establishes a respiratory connection with the
exterior and completes its development (Fig. 76). Clausen (1931)
states that certain Trigonalidse lay their eggs on foliage and they

Fig. 76. Gonia capitata De G. A, First instar larva on salivary
duct of host. B, Unfed larva surrounded by phagocytic cyst
(p.c.) ; f.b., fat-body, C, Larva in right lobe of supra-oesophageal
ganglion. D, Late larva attached to host-pupa by respiratory
funnel (r./.). (After Strickland, loc. cit.)

are swallowed with the food by caterpillars of saw-flies and of
Lepidoptera. The eggs hatch into spinose larvae, which require as
their true hosts either Ichneumon or Tachinid larvae. Those
Trigonalid larvae which occur in unparasitised caterpillars
presumably perish.

Thompson (1928) has studied the behaviour of the Tachinid,
Rhacodineura antiqua, a parasite of the common earwig in England
and other parts of Europe. It likewise lays minute eggs (0-13
mm. X 0-10 mm.), which are ingested by the earwig and hatch
in the mid-intestine. The subsequent history is incomplete,

MIGRATORY LARVJE

333

but it appears that having left the gut the larva bores through
the neck region of the host in order to establish respiratory
connection with the exterior. In this position it becomes
enveloped in an integumentary sheath and, when nearing maturity,
the sheath is ruptured and the larva becomes free in its host,
mainly feeding upon the fat-body.

2. Types of Migratory Primary Larvae. In many cases where
the parasitic insect deposits
her eggs or larvae on leaves,
buds, in soil or other situa-
tions, but not actually on the
host itself, the primary larvae
assume a form very different
from that found in subsequent
instars, and hypermetamor-
phosis results. Under such
conditions the young larvae
are active migratory organisms
well adapted for the business
of seeking their hosts. This
kind of behaviour has been
independently acquired among
certain members of no less than
five orders of insects, which
will be discussed separately.

(a) In the Chalcid family,
Perilampidae, the first stage
larvae of a number of species
of Perilampus are minute,
elongate, darkly coloured objects about 0-1 to 0-16 mm. long.
The dorsal integument is heavily . sclerotised, there is a well-
defined head followed by twelve body-segments and a pair of
cerciform caudal setae ; ventrally the body is provided with an
armature of spines and spinous processes which aid in locomotion.
Such a larva is termed a planidium (Fig. 77). H. S. Smith (1912)
has shown that the planidium of Perilampus hyalinus bores into
caterpillars of Hyphantria textor, of which it is a hyperparasite.

Fig. 77. Planidium of Perilampus. «,
Ventral ; 6, dorsal. (After H. S. Smith.)

334 PARASITISM

The planidium undergoes no further development until it meets
within this host one of its parasites, i.e., the Ichneumon Limnerium
validum or the Tachinid Varichceta aldrichi. It then enters the
larva of one or the other, but remains there in a quiescent state
until either host pupates. Once the latter phase is assumed, the
planidium gnaws its way out and becomes an ectoparasite and
subsequently transforms into an inert, maggot-like larva.

In the related family of the Eucharidse, whose members
parasitise certain ants, the first stage larva is likewise a planidium.
It was in the species Orasema viridis, which parasitises the Texan
ant Pheidole instabilis, that Wheeler first discovered this peculiar
larval type in 1905. The most complete observations are those
of Clausen (1923) on Schizaspidia tenuicornis, whose host is the
Japanese carpenter ant Camponotus japonicus. It appears that
the Schizaspidia lays large numbers of minute eggs within the
buds of certain trees, and the planidia, upon emergence, await
opportunity for attaching themselves to the Camponotus workers
which roam about the trees attending aphides. On meeting an
ant the planidium attaches itself to the hairs of the tarsi and thus
gains entry into the nest. It then becomes an ectoparasite of
an ant larva, upon whose back it settles down and perforates the
cuticle behind the head, on the thorax. Two instars follow, during
which the planidium becomes a vermiform grub of the usual
Hymenopterous type and, after the host has pupated, the parasite
regains its hold following the ecdysis. It reattaches itself on the
metathorax beneath the wings or legs and finally pupates within
the host's cocoon. When the parasite becomes an imago it leaves
the nest and flies away. The effect of the extraction of the body-
juices of the host by the parasite results in its inability to become
an adult, although death may not supervene until some time
afterwards.

{h) In the Dipterous family Tachinidae there are certain
species whose primary larvae emerge from the egg as active
migratory organisms somewhat resembling planidia. In
Digonochceta setipennis the eggs are laid in the immediate vicinity
of resting earwigs, which serve as the hosts. They hatch almost
immediately into deeply pigmented larvae, 0-6 mm. long, armed

MIGRATORY LARVJS

335

with plates, bands and spines, and the tracheal system is meta-
pneustic (Thompson, 1928). Such a larva moves rapidly towards
an earwig and, having secured itself, bores its way through the
neck, or between the thoracic segments, until only its caudal
extremity is exposed to the exterior (Altson). Certain other
Tachinids deposit their eggs in the soil, and the first stage larvae
burrow through that medium in search of
Scarabseid larvae which serve as their
hosts. Thus, both Dexia ventralis Aid.
and Prosena sibirita Fab. betray this type
of behaviour. Their first stage larvae are
elongate and vermiform and consequently
adapted for burrowing ; the tracheal
system is metapneustic and the body is
armed either with segmental setae or with
a caudal pair only (Fig. 78). Upon
meeting a host they bore their way
within ; the larva of the Prosena subse-
quently becomes attached by its spiracular
extremity to one of the main tracheae of
the host, while that of the Dexia maintains
its caudal end in the original perforation
hole and thus breathes the outside air
(Clausen, King and Teranischi, 1927).

In another Dipterous family, the
Cyrtidae, King (1916) has shown that the
species Pterodontia flavipes deposits her
eggs on tree-trunks, and the first stage
larvae are strikingly planidium-like. Its
hard integument is armed dorsally and ventrally with segmental
bands of powerful spines and pectinate scales. At the caudal
extremity it bears a sucker flanked on either side by a long seta.
These active larvae bore their way into Lycosid spiders, within
whose bodies they transform into maggot-like endoparasites.

An active first stage larva is also known in the parasitic families
Nemestrinidae and Bombyliidae.

(c) Among the few known parasitoid Coleoptera the first stage

Fig. 78. A, First instar
larva of Prosena
sibirita, and B, its
mouth-parts. C, First
instar larva of Centeter
cinerea, and D its
mouth-parts. (Adapted
from Clausen, King and
Teranischi, U.S. Dept.
Agric. Bull., 1429.)

336 PARASITISM

larva is of a typical, active, campodeiform type. Thus, in certain
species of the Staphylinid genus Aleochara the primary larva is
engaged in seeking out Anthomyiid puparia, and having discovered
its host, it bores its way within and transforms into an inert
cruciform parasite, its obvious degeneration being correlated with
the changed mode of life {vide Wadsworth, 1915 ; Kemner, 1926 ;
Voris, 1934). Silvestri (1904) has shown that the Carabid Lebia
scapularis, whose host is the Chrysomelid beetle Galerucella
luteola, passes through a very similar type of hypermetamorphosis.

The triangulin or first stage larva so characteristic of the order
Strepsiptera, and present also in the Meloidae and Rhipiphoridse
among Coleoptera, is comparable with the campodeiform larva
of Aleochara and Lebia just alluded to, and entomologists are
familiar with the campodeiform primary larva of Mantispa
among Neuroptera. There is little doubt that in these examples
the first stage larva is the primitive one, and the adoption of a
parasitic mode of life is responsible for the highly modified instars
that supervene. Comparison with the ontogeny of members
of allied families which lead a normal non-parasitic existence
supports this contention. It is noteworthy that the general
similarity, both in structure and behaviour, of the Hymenopterous
planidium and these campodeiform primary larvns has been
commented upon by several observers. Wheeler is inclined to
pursue the analogy further and to regard the planidium as an
archaic, though often considerably modified stage which has been
suppressed in other families of Hymenoptera. In the Tachinid
Diptera, however, evidence points to the migratory phase being
an adaptive one secondaril3^ intercalated in the ontogeny.

3. Non-migratory Primary Larvae. This habit appears to be
infrequent, but prevails among certain members of the Tachinidse
which betray leaf oviposition. It is probable that true larvi-
position does not occur, as has been sometimes stated, but that
the larvae issue from the eggs almost immediately after the
latter have been laid. The behaviour of Bonnetia compta
(Strickland, 1923) and of Ernesiia (Panzeria) rudis (Prell, 1915)
which parasitise caterpillars, may be regarded as typical.
Oviposition occurs on the food-plants of the host, and the young

NON -MIGRATORY LARVJE 337

larva breaks through the upper or cephahc pole of the egg. It

gradually works the remains of the chorion and vitelline membrane

over its body to the anal extremity, where they persist as a

crumpled annular base of attachment to the leaf. With the

approach of a moving object, the larva stands erect within the

cup-like base referred to, and its whole body rotates vigorously

in the efforts made by the parasite to obtain a hold on the object

in its vicinity (Fig. 79). If it be a cutworm, the parasite speedily

bores through the integument, and in a few minutes it has passed

within its host with the exception of its spiracular extremity,

which rests in the original perforation, in free

communication with the outside air. An

integumental sheath grows around the larva

and, maintained in this position, it completes

its development at the expense of the blood

and the fat-body of the host.

This stationary habit prevails in the group

Echinomyiinse whose members are essentially

parasites of caterpillars {vide Thompson, 1923).

The primary larvae are well adapted to remain ~ ^^'~

on the host food-plant for considerable periods, Fig. 79. Newly

J j.i_ij-£j' j.-ii. r hatched larva of

and are notably different m structure from Panzeria rudis

those which live within their hosts at a attached to

correspondingly early stage. The integument [^^"rjFronrpiell!)

is tough, resistant and deeply coloured with

an armature of scale-like plates and spines. Thus equipped they

are not only able to resist desiccation, but are also enabled to hold

on to and bore a way rapidly into their hosts.

According to Strickland the larva of Bomietia may leave its
cup-like basal pad and migrate a short distance to assume again
its erect position. This apparently insignificant feature is
noteworthy, since species with sedentary primary larvae intergrade
with others like Digonochcuto (p. 334) whose larvae are actively
migratory.

Group B. The habit of depositing eggs externally on the body
of the host, which hatch into ectoparasitic larvae, is of common
occurrence among Chalcids and Ichneumonoids. It is, however,

338 PARASITISM

unknown among Tachinids and rare in other parasitoidal Diptera,-^
a feature which contrasts with its prevalence among Hymenoptera.
Adaptations to an exclusively ectoparasitic life of this kind are
not markedly different from those exhibited by most non-parasitic
Hymenoptera, whose larvae likewise pass an inert existence in
contact with ample nutriment. In view of the uniformity of
conditions prevailing during the whole parasitic life, it is scarcely
surprising that the metamorphoses of such ectoparasites do not
involve the striking diversity of larval forms so frequent among
the endoparasitic members of the same order.

Upon emergence from the egg the ectoparasitic larva punctures
the skin of its host with its mandibles and imbibes the body-
fluids through the rupture thus formed. It rarely moves from
the position originally taken up, and this method of nutrition
commonly prevails throughout the larval existence. In cases
where predatory instincts subsequently develop the larva may
become more active and finally consume its host bodily.

Structurally, ectoparasites present few constant features
which are unquestionably correlated with their mode of life
and at the same time undeveloped in some or other of the
endoparasitic forms. The most important character is afforded
by the respiratory system and, as a rule, open spiracles are present
from the first instar onwards. In some cases, as in Aphelinus
mytilaspidis, for example, the full complement of spiracles is
present at the time of eclosion from the egg (which never obtains
in endoparasitic Hymenoptera), in others it is not acquired until
later. It should be pointed out, however, that information
respecting the tracheal system of the early instars is very
incomplete and generalisations relating thereto may require
subsequent modification. In the primary larvae of many
ectoparasites the head is relatively larger and more firmly
sclerotised and the mandibles more powerful than in subsequent
instars, characters which are correlated with the maintenance of
a firm hold upon the host and with the initial perforation of the
integument. The body frequently exhibits a definite chsetotaxy,

1 Instances of this type of life-cycle are found in the Bombyliidae among
Diptera.

OVIPOSIT ION 339

little developed in some forms and more evident in others. In
the Chalcid genera Eupelmus, Eurytomus, Torymus and Leucospis,
bands of slender macrochaeta^ are very characteristic. The
foregoing characters disappear in the later instars and finally
there remain few or no evident structural differences whereby
an ectoparasite can be distinguished from the later instars of
most endoparasites. This subject is also discussed by Evans
(1933) with reference to certain Hymenopterous parasites of
Diptera.

Group C. External oviposition or larviposition, followed by an
endoparasitic life, is exemplified in many Tachinidae which form
Groups I., III. and VI. of Pantel (1909).

In Pantel's Group I. those Tachinids with a macrotype egg
{i.e., relatively large and of a short, somewhat flattened form)
oviposit on the bodies of the hosts, fastening the eggs in position
by means of a glutinous secretion.
The embryos in such eggs are but
little developed at the time of
oviposition, and are covered with
a tough, durable chorion. Emer- Fig. 80. Empty egg (e) of T/^rmon
gence from the eggs may take ^:S^%^^^^„t
place in one of two ways. Either shown the passage eaten by the

the chorion dehisces, but aids in }^^^^ JT ^^ff""^ ^^'^ host.
' ^ _ (From Pantel.)

maintaining the larva in position as

it bores its way through the integument of the host ( Winthemyia
4<-pustulata and Tricholyga major) ; or, the chorion does not
rupture and the larva eats its way through that part of it which
is in contact with the integument of the host, entering the latter
without exposure to the exterior (species of Tachina, Gymnosoma,
Thrixion (Fig. 80), Centeter, etc.). In Centeter cinerea Clausen and
his co-workers (Fig. 78) mention that the eggs are laid on the
hard integument of the Japanese beetle, and the first instar
larva has its mouth-parts armed with a dentate rasper ^ which
enables it to bore through the cuticle of the host. Owing to errors
of the oviposition instinct, many larvae perish on account of the

1 This structure is frequently present in larvae which bore through the
integument of the host-

340 PARASITISM

eggs being lodged on parts of the integument too thick for
penetration.

Pantel's Group III. comprises those Sarcophagidge which deposit
larva) on the bodies of their hosts. They are true viviparous
flies with many parasitic species and are now often regarded as a
sub-family of the Tachinid?e. A good example of supra-cutaneous
larviposition is afforded by Sarcophaga kellyi which deposits its
larvae beneath the folded wings or on the abdomen of Melanoplus
grasshoppers. The young parasites quickly bore their way through
the inter-articular membranes and lead an endoparasitic life
(Kelly, 1914).

There are again other Tachinids, forming Group VI, of Pantel,
which are ovo- viviparous. Their eggs are thin-shelled, and the
larva? issue either in the uterus of the parent or almost immediately
after oviposition. They differ from those mentioned in Group A
in that the integument is devoid of chitinous armature, and they
are ill-adapted for prolonged exposure. Many species of Tachinids
come under this category, but it is probable that actual larvi-
position occurs less frequently on the host itself, and more often
eggs are deposited which hatch almost immediately. Townsend
has pointed out that certain Dexiine species parasitise insects
living in concealed situations, such as leaf-miners, wood-borers,
etc. The flies penetrate into the burrows of these concealed
hosts depositing their larvae in their immediate proximity. Such
unarmoured parasites, finding themselves in confined spaces,
have only short distances to traverse before discovering their
hosts. Among ovo-viviparous Tachinids, species of Sturmia,
Exorista, Thelaira, Bucentes, Plagia, Voria, and other genera are
enumerated by Baer (1920) and by Thompson (1926).

Group D. In this group are included the greater number of
the parasitic species of Hymenoptera and certain of the Diptera,
including the Pipunculidse, Conopidse, the Agromyzid genus
Cryptocho'tum and a few of the Tachinidae.

Among endoparasitic Hymenoptera the eggs may be inserted
either within the eggs, larvae, pupae or adults of the hosts con-
cerned. The Mymaridae, Trichogrammidae, Scelionidae and many
Platygasteridae are essentially egg-parasites completing their

OVIPOSIT ION 341

whole development in the eggs of various insects. Certain of the
Encyrtidas and Eulophidae parasitise eggs of Lepidoptera or
Coleoptera, but complete their development in the larval stages of
their hosts. An enormous number of Hymenoptera parasitise
larvae only, or subsequently complete their development within
the pupae, while certain others are exclusively pupal parasites.
Parasites confined to imagines are less frequent, but the Braconid
genera Perilitus and Dinocampus are well known to parasitise adult
Coleoptera.

The ovipositor is an instrument often used with extreme
accuracy ; as a rule the eggs are placed in the general body-cavity
of the host, but in other cases intra-organic oviposition prevails,
the eggs being laid in the fat-body, gut, salivary glands, brain or
other situations. The eggs as a rule are poor in yolk, or are almost
devoid of that material, while the chorion is thin or even mem-
branous or may be absent. After deposition the eggs frequently
increase markedly in size until eclosion of the larva supervenes.
Thus in the Ichneumon Meteorus dimidiatus the newly laid egg
measures 0-14 mm. x 0-04 mm., but it finally attains a maximum
size of 1-2 mm. x 1-5 mm. (Fig. 84) ; in Dinocampus rutilis,
another Braconid, Jackson states that the egg may increase in
cubic content over 1,200 times. The primary larvae frequently
emerge in a very early stage of ontogenetic development, they
assume great diversity of form and hypermetamorphosis prevails
{vide Chapter III.). The significance of the morphological features
thus exhibited is very obscure ; certain of them appear to be
inherited ancestral survivals, while some are most likely adaptive
in significance. In the final instars, however, the apodous maggot-
like type common to all Hymenoptera-Apocrita is assumed. The
chief features of endoparasitic life are discussed on pp. 343-353,
and consequently will not be referred to here.

Among Tachinidae both subcutaneous oviposition and larvi-
position occur, and the species exhibiting the habit form groups
VII., VIII. and IX. of Pantel. The female fly is provided with an
instrument for perforating the integument of the host, and in its
most usual form, as exhibited in Compsilura concinnata, it consists
of a curved acutely pointed piercer which is grooved dorsally to

342 PARASITISM

receive the tubular larvipositor (or ovipositor). The two structures
act in unison, and the introduction of the larvipositor follows upon
the action of the piercer in perforating the host. The ventral
region of the abdomen is strongly carinate, and bears backwardly
directed spines together with a prominent backwardly directed
process of the fifth sternum. These structures evidently aid the
fly in maintaining a firm grip on the host while the puncture is
being made. In Neocelatoria ferox which parasitises the beetle
Diabrotica Walton (1914) concludes that the host is gripped, as it
were, by the " forceps " formed by the piercer and the sternal

process while the parasite makes
the perforation. In this species
the sternal process is borne on
the second segment, and is
strongly dentate (Fig. 81).

In Compsilura the instrument
is a true larvipositor, and the
larvae are present in the uterus

Fig. SI. Neocelatoria ferox Walt. 'Lett: of the parent prior to
abdomen of female. Right : figure , ... -^ , . , i

showing probable functioning of deposition. In certain other
appendages. a, Ovipositor. b, Tachinidse (species of Alophora,
Piercer. c, Process of second ^t i • j t^ . \ t» . i

abdominal segment, d. End view Hyalomyia and Xysta) Pantel
of host beetle {Diabrotica). (From concludes that oviposition takes
Walton.) , , ,

place by an analogous pro-
cedure. The presence of eggs but little advanced in development
when in the uterus, the absence of an incubatory pouch, and
the existence of a complex piercing ovipositor support his
contention. He also places the Conopidae provisionally in this
same group.

In the Agromyzid Cryptochcetum a highly specialised piercing
ovipositor is present : it appears to afford a striking example of
adaptation to a parasitic mode of life, since the eggs are inserted
within the body-cavity of coccids of the sub-family Monophlebinae.
The ovipositor has peculiar structural features which seem to
preclude its origin as a specialisation of the rasping type of organ
found among other Agromyzidae (Thorpe, 1934).

HOST RELATIONS 343

Host Relations of Parasites

The behaviour of entomophagous parasites in relation to their
hosts reveals many striking and remarkable features. As a general
rule their activities result in the death of the insects attacked,
which usually succumb very soon after the parasites have left
their bodies. In other cases the fatal effects of parasitisation
may reach this climax more slowly, and the host is capable of
limited reproduction before its premature death supervenes. This
regularly happens with the adult females of the brown scale
{Coccus caprece) when parasitised by certain Chalcids, while some
aphides are capable of restricted viviparous reproduction when
parasitised by Aphidius if they have passed the third ecdysis.
Instances are known of adult Coleoptera which have remained
active, and laid eggs, after larvae of the Braconid Dinocampus
have issued from their bodies. In this connection it is noteworthy
that Timberlake (1916) was able, under experimental conditions,
to rear two generations of D. americanus from a single individual
Coccinellid before death of the latter supervened. Such survivals
are probably exceptional and the result of relatively light parasi-
tisation affecting robust and well-developed host individuals.
Furthermore, the parasites feed upon the blood and fat-body,
sedulously avoiding injuring the more vital organs. Many
haemophagous or steatophagous parasites bring about the death
of their hosts largely on account of the nutritive drain which their
feeding activities entail upon the latter. In cases when the hosts
are adult insects, the partial or complete suppression of the
reproductive functions of the parasitised individuals (parasitic
castration) is a common feature. This occurs with the European
earwig when parasitised by the Tachinid Digonochceta setipennis,
with many Homoptera parasitised by Dryinidae and in both the
Homopterous and Hymenopterous hosts of Strepsiptera. As
previously mentioned, the activities of the Strepsiptera rarely
result in the death of the hosts, which continue to live and feed
after the male parasites have left their bodies.

During their early life endoparasites lie within their hosts
surrounded by the blood or other serous fluid from which they

344

PARASITISM

derive their nutriment. Respiration, under these circumstances,
has much in common with that of aquatic insects, and the necessary
oxygen is obtained either from the surrounding medium or by
estabhshing a direct or indirect connection with the outer air.
With many parasites the demands of growth cannot be fulfilled in
this manner throughout life and, as the body-fluids of the host
tend to become exhausted, the parasites turn to the more solid
tissues as their source of food and become, as it were, internal
predators. Their feeding activities entail
the laceration of many of the smaller
tracheal vessels, which results in the
liberation of air in the body of the host.
Direct respiration by the parasites now
becomes possible, and correlated changes
in the respiratory system result. Hymen-
opterous parasites develop at this stage
a peripneustic tracheal system, and
Dipterous parasites may no longer need
the special adaptations previously
acquired as a means of breathing the
outside air. Finally, when larval de-
velopment is completed, the parasites
either pupate within their host or gnaw
their way through its integument and
transform outside its body. This, in a
few words, is the course of life followed
by most entomophagous parasites, but
the behaviour of different groups exhibits
so great a diversity of phases that few generalisations are possible.
It will be convenient, therefore, to discuss Hymenopterous and
Dipterous parasites separately.

Hymenoptera. In its first instar a Hymenopterous parasite
is characterised by the absence of spiracles, the tracheal system is
rudimentary or wanting, and the body integument is a delicate,
flexible membrane. At this stage its diet consists solely of the
blood or serous fluids of its hosts ; the mouth-parts are represented
by a pair of sharp, piercing mandibles, while the remaining trophi

Fig. 82. Metapneustic first
instar larva L of Blasto-
thrix attached to chorion
of egg whose pedicel P
protrudes through the
integument I of the host.
S, Spiracles.

H08T RELATIONS

345

are reduced to the condition of fleshy lobes forming the boundary
of the suctorial mouth. Salivary glands and Malpighian tubes are
usually well developed, and the mid- intestine is a capacious sac,
but without direct continuity with the proctodaeum. Respiration
is cutaneous, and the necessary oxygen is derived from the blood
of the host. It is true that little is definitely known with respect
to the oxygen-carrying
capacity of insect blood, but
all the evidence strongly
indicates that it is from the
latter source that endopara-
sites first obtain the oxygen
they may require. Species of
certain Chalcid genera, in-
cluding Blastothrix, Encyrtus,
Microterys and Phcenodiscus
(Fig. 83), which parasitise
Coccidse, are exceptional in
being metapneustic in the first
instar (Imms, 1918 ; Silvestri,
1919 ; Maple, 1937). In these
genera the egg is provided with
an elongated pedicel, whose
apex protrudes through the
integument of the host to the
exterior. The anal extremity
of the larva, in such cases, is
closely grasped by the per-
sistent chorion of the egg in
such a manner that the
spiracles are able to inspire the outside air through the pedicel,
which serves as a respiratory funnel (Fig. 82). The chorion in
this region is specially modified and is freely permeable to air.
During their later life these parasites free themselves from all
connection with the egg pedicel, and Acquire a peripneustic
respiratory system. Special organs, either in the form of a
caudal appendage or of an anal vesicle, are present in diverse

Fig. 83. Two metapneustic larvae of
Phcenodiscus emeus attached by their
egg pedicels to integument of host
{Sphcerolecanium). (After Silvestri.)

346

PARASITISM

groups of parasitic Hymenoptera. These structures are found
only in the early instars, and become gradually resorbed as
growth proceeds and spiracular respiration becomes established.

HOST RELATIONS 347

This fact, along with the absence of such organs in all
ectoparasitic forms, suggest that they exercise a respiratory
function. A caudal appendage has been acquired in certain
genera among all the great groups of parasitic Hymenoptera.
Among Ichneumons it is prevalent in Exochilum (Anomalon),
Limnerium, Mesochorus, Campoplex, Mesoleius, Thersilochus and
others. In Braconidic (Fig. 83) it is found in Dmocampus and
Adelura ; it is exceptionally well developed in Meieorus, but
greatly reduced in Praon and in Aphidius. A similar organ is
present in several Chalcid genera {vide Parker, 1921) ; it is always
prevalent in the primary larvae of the Figitidse, and in some
Proctotrypoidea. There appears to be nothing in the structure of
this organ to contradict the view that it is respiratory in function,
since it is a thin-walled tube lined by a delicate hypodermal layer
and filled with blood, and in some cases it contains trachse also.
It appears to be a kind of blood gill which serves to increase the
general respiratory surface of the body in order to fulfil the
breathing requirements of the young growing larva. This
exj^lanation, which has been accepted in some quarters, is con-
troverted by the researches of Thorpe (1932). By the use of
biological indicators it was shown where, over the body-surface of
endoparasitic larvae, oxygen absorption was going on while carbon
dioxide output was investigated mainly by means ofpH indicators.
Parasitic larvae were dissected out from their hosts, washed free
from the tissues of the latter and examined under a microscope
in salt or Ringer's solution. Use was made of certain Protozoa, of
which Polytoma was found most suitable, and advantage taken of
the fact that such organisms migrate to a region where the oxygen
tension is lower than that in water saturated at atmospheric
pressure. When a culture of these organisms is run under the
coverslip these creatures at once arrange themselves in a zone
around the parasitic larva where oxygen is being absorbed from
the fluid. As the oxygen tension falls below the optimum, owing
to respiratory activity of the parasite, the Protozoa retreat
further away from the region of the parasite concerned. By using
luminous bacteria (B. phosphor escens), which were introduced into
the solution in the same way, examination was made in complete

348

PARASITISM

darkness. As the oxygen became used up by the parasite the
luminous zone gradually spread outwards until luminosity was
only betrayed at the edges of the coverslip, where oxygen was
passing into solution from the outside air. By means of this
technique it was shown that the tail of Ichneumon larvae performs
no apparent function in respiration, the Protozoa grouping
themselves entirely in relation to the body of the parasite only.
The absence of any proper blood circulation in the caudal
appendage appears also to argue against this organ performing a
respiratory function. It is probable also that the " tail " of
Proctotrypid and Cynipid larvae is likewise devoid of a respiratory
function. With regard to the caudal vesicle of Braconids, referred

Fig. 85.

Longitudinal section through larva of Apanteles. d, d, Dorsal,
vessel, in, Intestine. (Adapted from Grandori.)

to below, Thorpe's experiments show that where this vesicle is
large and supplied with a good blood circulation it is undoubtedly
of importance as a respiratory organ. It is not, however, the sole
or even the most important organ of respiration, since gaseous
exchange takes place over the whole surface of the larva. It is
concluded that in Apanteles and Microgaster the vesicle, when at
its maximum development, cannot be responsible for more than
about one-third of the total respiration. In other Braconids
(Orgilus) its respiratory function is less. The caudal vesicle found
in Apanteles, Microgaster, Microplitis, and some other Braconid
larvae has been much discussed by various investigators (vide
Tothill, 1922). It is evidently the everted proctodaeum which
becomes gradually withdrawn into the body during later growth

HOST RELATIONS

349

to form the hindmost region of the intestine (Grandori, 1911 ).i
The greater part of the cavity of the vesicle is occupied by the
much enlarged terminal chamber of the heart, which suggests that
the organ mainly functions as a specialised type of blood gill.
According to Grandori, the character of its lining epithelium
indicates that it also plays some part in excretion (Figs. 85 and 87).

Fig. 86. Fifth instar larva of Encyrtus infelix dissected from its
host while still enclosed in its sheath, s, together with the
longitudinal tracheal trunks, t, of the host attached, sa, anterior
spiracle ; sp, posterior spiracle ; t', longitudinal trachea of a
parasite ; ta, tracheal attachment. (From Thorpe.)

In all endoparasitic Hymenopterous larvge there is no through
passage between the mid-intestine and the exterior, an arrangement
which ensures against the contamination of the host by faecal
products. It is only at the last ecdysis, prior to pupation, that
direct continuity becomes established and the contents of the

^ A somewhat similar view has been expressed by Gatenby (1919), who,
along with Tothill, was not aware of Grandori's important memoir on
Aaputeles,

350

PARASITISM

intestine discharged. The waste matter usually assumes the form
of a chain of faecal pellets which is commonly found in close
proximity to the pupa.

A new type of respiratory inter-relationship between a
Hymenopterous parasite and its host has been recently described
by Thorpe (1936). The parasite, Encyrtus infelix (Chalcidoidea),
utilises the Coccid SaiseUia hemisphcrrica as its host. In its first
three larval instars the Encyrtus is metapneustic, the spiracles
being in communication with the atmospheric air by means of the
protruding pedicel of the egg in the manner already described in

the case of certain other
Chalcids. In the fourth and
fifth instars the mode of
respiration of the parasite
becomes completely altered.
The parasite now becomes
dependent upon a special
structural development of its
host. It turns round within
the body of the latter and
^/ " becomes invested with a

Fig. 87. A, diagram of anal extremity closely fitting membranous
of young larva of Apanteles. B The j^ ^j j^- j^ • ^ ^ .
same of fully grown larva. e, Blmd -^

extremity of gut. m, Mid-intestine, the phagocytes and tracheal
m.^. Malpighian tubes. v, Caudal branches of the host. The
vesicle.

sheath becomes attached to

the main lateral tracheal trunks of the host in four or six places
in the vicinity of the spiracles of the parasite (Fig. 86). An
actual continuity becomes developed between the lumen of the
host trachea at these points and the narrow space between the
sheath and the parasite. In this manner the larval parasite is
able to breathe air from the tracheal system of its host.

Another peculiar type of host relationship, of a different nature,
has been recorded by Flanders (1936) with reference to the
Chalcid Coccophagus. In the species studied the progeny of
parthenogenetic females consists, as is often the case, of males
only. Fertilised females produce females only, and these develop

HOST RELATIONS 351

entirely as primary parasites of Lecaniine scale insects and mealy
bugs. The males, on the other hand, only develop as hyper-
parasites of larvae of its own or other species of parasites. It
would seem that the stimulus of fertilisation results in a marked
change in host-selection by the female Chalcid. Prior to mating
these oviposit freely in the body-cavity of a parasitic Hymen-
opterous larva, but after mating egg-laying only occurs within a
scale insect. This peculiar and apparently unique phenomenon
requires further investigation.

Diptera. Endoj^arasitic Dipterous larvae differ from those of
almost all Hymenoptera, which lead a similar mode of life, in two
important features : (1) When they issue from the egg they are
provided with a metapneustic tracheal system ; and (2) they
nearly always establish a respiratory connection with the outside
air during some stage of their development. Exceptions to these
generalisations are relatively few. In cases where the early
parasitic life is passed in the body-cavity of the host or within
some specific organ of the body the spiracles, although well
developed, appear to be non-functional, or they may sometimes
be vestigial or wanting. Also, in a small number of species,
chiefly Pipunculidse and Sarcophagidse, the whole larval existence
is passed within the bodies of the hosts without acquiring any
respiratory connections with the exterior. The. vast proportion of
the parasitoid Diptera belong, however, to the family Tachinidse
and almost all the species of this extensive group respire free air,
either by means of a perforation of the body-wall of the host or
by acquiring a secondary connection with its tracheal system. In
either event the parasitic larva becomes enclosed in a sheath
(" gaine de fixation " of Pantel), which may be of a primary or
secondary character. Whichever way it is formed it benefits host
and parasite alike ; it maintains the parasite in intimate connection
with the necessary oxygen supply, but it is also a protective
reaction on the part of the host against irritation and microbic
infection induced by the presence of the parasite.

A primary sheath is always cutaneous in origin, and is formed
as an integumentary ingrowth from the margin of the initial
perforation made by the larva in entering the host. The perfora-

352

PARASITISM

tion functions as a respiratory pore (" soupirail " of Pantel) and
the parasite is maintained in the body-cavity of the host, with its
spiracular extremity firmly wedged in the aperture. The sheath
consists of the same layers as the integument, its lining being
continuous with the external cuticle. It forms a stout funnel,
which grows around the hindmost segments of the parasite, and
its walls often become thickened by the incorporation of fat cells,
dead phagocytes and other tissues which become bound together
by a copious infiltration of chitin produced by the hypodermal
layer. Distally the sheath is membranous, and may entirely

enclose the larva and, under
such circumstances, it
appears to be permeable to
the body-fluids of the host
which form the sole
nutriment of the parasite
within. The first two
ecdyses commonly occur
while the parasite is thus
enclosed, and the exuviae
cast off within the
of the sheath.
Ultimately the parasite
either frees itself completely
from this investment and
becomes actively carni-
vorous among the organs of the host ; or, it ruptures the
membranous bag, but continues to respire the outer air with its
spiracular extremity still retained within the basal funnel of the
sheath (Fig. 88).

A secondary sheath may be either cutaneous or tracheal in
origin (Fig. 89), and is formed with respect to species whose early
life is passed either free in the body-cavity of the host, or within
an individual organ of its body, usually a nerve centre or the
mid-intestine. It is formed at a stage when the respiratory needs
of the growing parasite demand a more adequate supply of oxygen
than can be derived from the body-fluids. The parasite fulfils its

are

Fig. 88. Section through a larva of cavitv
Thrixion attached to cutaneous respira-
tory funnel /. a. Primary entrance hole
through integument e of host. sh,
Integumental sheath enclosing parasite.
s, Spiracle. (Adapted from Pantel.)

POLY EM BRYONY

353

owii requirement either hy perforating the integument or by
rupturing one of the larger traeheye or an air-sac of its host. It
uses its armoured spiraeular extremity to make the necessary
perforation in the integument or trachea as the case may be, and
the wound thus caused leads to the formation of a sheath similar
to the primary sheath already described. Many Tachinids
maintain this respiratory adaptation until ready to pupate, while
others, such as Centeter cinerea, free them-
selves after attaining the third instar, and
proceed to devour the viscera of their hosts.
In the case of the endoparasitic genus
Cryptochwturn, of the family Agromyzid^,
Thorpe (1931) has studied the economy of
the species C. iceryw in detail, while that of
C grandicorne formed the subject of a later
communication (Thorpe, 1934). The larvae
of these two species occur in the body-
cavity of certain Coccidae, the host being
Icerya purchasi in the case of the first-
mentioned species and Guerinia serratulce in
the case of the second species. The larvae
are provided with a pair of filamentous

prolongations of the last abdominal segment Fig. 89. Tachinid larva
, . , . . 1 . 1 . . • . attached to a tracheal

which increase in length in succestive instars

trunk t by means of a
secondary respiratory
funnel c. I, Limits of
tracheal sheath. s,
Spiracle. (From
Pantel.)

in conformity with the needs of the growing

parasite. These filaments contain blood

and tracheae and ramify among the organs

of the host. Experiments of a similar kind

to those referred to on p. 347 indicate that these caudal filaments

serve to increase the surface available for respiratory exchange

between the larva and the blood of the host. When the tracheal

system becomes in communication with the outside by means of the

spiracles, the caudal filaments shrivel and are no longer functional.

Polyembryony and Related Phenomena

Polyembryony is rare in the whole animal kingdom, and among
insects it is almost exclusively confined to certain of the parasitic

i.A. EXTOMOLOQY.

354 PARASITISM

Hymenoptera,^ where it exhibits a remarkable sequence of
biological processes. In view of the immense number of a species
of parasitic Hymenoptera already described, and the very small
number whose economy has been adequately studied, it is likely
that the incidence of polyembryony will eventually be found to
be more frequent. The researches of Martin (1914) in Germany,
Silvestri (1906) in Italy, and of Leiby and Hill (1923-24) and of
Patterson in America — which followed upon the earlier funda-
mental discoveries of Marchal in France — have led to the
elucidation of the essential features of the phenomenon. It is
known to have arisen independently in three families belonging
respectively to the Chalcidoidea, Proctotrypoidea and Vespoidea,
and its phases will be briefly considered in each of these groups
separately.

(a) Chalcidoidea. It was in the species Ageniaspis (Encyrtus)
fuscicollis that Marchal definitely proved in 1898 the occurrence
of polyembryony in insects. It has only so far been observed
within the single family Encyrtida?. where it prevails in species
belonging to a group of closely related genera. Polyembryony has
been recorded in species of Ageniaspis, Encyrtus, Copidosoma,
Litomastix, and Berecyntus, whose hosts so far as known are
Lepidoptera.2 In all cases the eggs of the parasites are deposited
within those of their hosts which duly hatch into larvae, and the
developing parasites are contained within their bodies. The
main features of the embryonic development are very similar in
the individual cases studied, irrespective of whether the eggs arc
fertilised or not, vide Fig. 90. Development commences by
the oocyte nucleus dividing twice to form two polar bodies, and
the first formed of the latter may also undergo division (Fig. 90, A).

^ In 1923, Noskiewicz and Poluszynski published a very brief preliminary
communication on the occurrence of polyembryony in the Stylopid
JIalictostylops, which parasitises the bee, Ilalictus simplex. Their much
condensed account is difficult to interpret in the absence of any illustrations,
but it appears that two types of embryos are formed. In those destined to
develop polyembryonically the blastoderm becomes divided into groups of
cells which form separate vesicles, and the latter ultimately give rise to the
embryos. It is stated that the number of embryos produced may vary from
two to more than forty.

2 These Lepidopterous hosts pertain to the families Gelechiida?,
Hyponomeutid.TC, Coleophoridae, Gracilariidae, Tortricidae and Noctuidae.

POLYEMBRYONY

355

The cytoplasm around the nucleus then becomes separated off from
the rest of the egg, which contains the polar nuclei, and conse-
quently the egg becomes divisible into an embryonic area and a
polar area (Fig. 89, B). A little later, the embryonic nuclei are
seen in division, and the

. -S— P,

polar nuclei have given rise
to the paranucleus or nucleus
of the future trophamnion
(Fig. 91). The developing
egg then soon becomes
enclosed in a cellular adven-
titious sheath derived from
the surrounding tissues of
the host ; the embryonic
nuclei increase in number,
the paranucleus also under-
goes division, and the original
polar area comes to form a
definite covering, or troph-
amnion, surrounding the
whole of the embryonic area
of the egg (Fig. 91). The
first indication of poly-
embryony is the division of
the embryonic area into
small groups of cells, or
morulse, which are seen to
lie in cavities within the
trophamnion. The whole egg
increases in size, and mean-
time both the adventitious

04^

c>

s

g

e-
B

tr

Fig. 90. Encyrtus fuscicollis, early poly-
embryonic development. A, Egg with
first and second polar bodies p^, p^.
B, Division into embryonic area e and
trophamnion tr. C, Transverse section
of polygerminal mass, a, Adventitious
sheath. e, e. Embryos, g, Germ-cell
determinant. n, Oocyte nucleus.

p. Paranucleus. tr, Trophamnion.

, . , J . 1 • (Adapted from Martin.)

sheath and trophamnion ^

grow simultaneously and maintain a complete double investment

(Fig. 90, C). The further development of the morulas need not

concern us here, and it will suffice to add that each subsequently

becomes a separate embryo, destined to transform into an

individual parasite. The number of embryos which originate

12—3

356 PARASITISM

from a single egg varies in different species ; in Ageniaspis
testaceipes 13 are produced (Marchal) ; in Copidosoma gelechice
from about 163 to 191 are recorded (Patterson, Leiby), while in
Litomastix truncatellus an average of about 1,500 embryos results
(Silvestri). It appears that, as the number of embryos derived
from a single egg increases, the polyembryonic process becomes
more complicated. The first-formed morula may each divide
into two, as in G. gelechice, while in L. truncatellus secondary
morulas divide to form tertiary bodies which finally develop into
larvae. AVhen fully formed the individual larvae break through
the investing membranes and become free within the body of
the host.

[h) Proctotrypoidea. Polyembryony occurs in this grouj) in
the genera Polygnotus and Platygaster, belonging to the family
Platygasteridse. Their hosts in all cases are Cecidomyidse, which
are parasitised either as eggs or as young larvae. Polygnotus
minutus utilises the Hessian fly (Mayetiola destructor) and the oat
midge (M. avence) as its hosts and deposits its egg in the mid-
intestine of the young larva, where it ultimately produces usually
10 to 12 embryos. Since the egg is bathed in the chyle of the gut
no adventitious sheath is formed around the embryonic mass.
Platygaster hiemalis, according to Lieby and Hill, also parasitises
the Hessian fly, attacking the eggs or, more occasionally, the
young larvae. It is of particular interest because the egg may
develop either monoembryonically or polyembryonically. From
4 to 8 eggs are laid in an individual host ; some develop into
single embryos, others give rise to two embryos, and the remainder
degenerate. The embryos become dispersed in the body-cavity
of the host embryo or larva, as the case may be, and become
invested by certain of the tissues which form an adventitious
sheath. The embryonic nucleus divides to form four nuclei
and subsequently the whole embryonic mass divides, and the
two daughter masses become separated by an extension of the
trophamnion between them. The cells of each embryo thus
formed then undergo further division until a hollow 16-celled
blastula is formed (Fig. 91). Without following the development
further, it may be added that finally the resulting larvae feed upon

POLY EM BRYONY

357

the host tissues, which they wholly consume. In eggs which
develop monoembryonically, the 4-celled stage gives rise to a
single blastula, but otherwise the development is similar in the
two cases. We have in P. hiemalis the simplest possible type of
polyembryony, and it appears that it has arisen from an already
specialised method of monoembryonic development, which is also
found in other Platygasterids. In another species, P. vernalis,
which is also parasitic upon the Hessian fly, these same observers

Fig. 91. Platygaster hiemalis, early polyembryonic development.
A, Egg with four embryonic nuclei. J5, Division into two embryos
surrounded by a common trophamnion. C, Transverse section of
an embryo in early blastula or 16-celled stage, of which seven are
shown. (Adapted from Leiby and Hill.) Lettering as in Fig. 90.

state that it lays a single egg in the egg of its host. The parasitic
egg is deposited in such a manner that it eventually comes to lie
in the mid-intestine of the host larva and consequently does not
become invested by an adventitious sheath. During early
development the embryonic nucleus undergoes division until
16 nuclei are produced. A cell membrane, enclosing a small
amount of cytoplasm, becomes formed around each nucleus, and
the cell thus constituted is recognised as a germ. Each germ maj"
be the progenitor of either one or two larvae, depending upon

358 PARASITISM

whether it undergoes division in the morula stage or not. Since
a number of the embryos abort during development, the brood
finally results on an average of only eight parasites being produced
from one Qgg.

{c) Ichneumonoidea. Among the ichneumonoid flies poly-
embryony occurs in the Braconid genus Macrocentrus, which
parasitises larvse of the Lepidoptera. According to Parker (1931)
the species M. gifuensis, which has been introduced into North
America in connection with the control of the European corn
borer, produces eggs each of which may yield up to eight or ten
larvae. In this species there appears to be no proliferation of the
host tissue to form an adventitious sheath. Polyembryony is also
stated to occur in M. ancylivorus, which parasitises the oriental
fruit moth larva among other hosts. According to Daniel (1932)
only a single polyembryonically produced individual survives in
each host.

(d) Vespoidea. The only example of polyembryony in solitary
parasitic wasps is afforded by Aphelopus thelice, a member of
the family Dryinidae. According to Kornhauser (1919) its host
is Thelia bimaculata, one of the largest and commonest of the
Membracidae found in north-eastern America. The parasite
usually lays a single egg in each host, selecting individuals which
are in a nymphal instar. The full details of the polyembryonic
process are not given, but it is evident, from Kornhauser's account,
that a trophamnion is present surrounding the developing poly-
germinal mass. The latter becomes converted into branching
chains of embryos which ultimately result in the production of
about 50 parasites. When the larvai become fully grown, the
host is either in its last nymphal instar, or has become an adult,
and is killed as the result of the larvae boring their way out to
pupate beneath the soil.

Certain features of the polyembryonic process require further
mention, viz., the origin of polyembryony, the function of the
embryonic membranes, and the significance of mixed broods of
parasites.

Origin of Polyembryony. Among parasitic Hymenoptera
polyembryony occurs where the eggs are immersed in a fluid

EMBRYONIC MEMBRANES 359

medium — either the blood or gut contents of the host — and, as
will be mentioned later, a certain amount of the fluid is absorbed
to nourish the developing embryos. As development proceeds
and the embryonic area becomes enclosed in a cavity within the
trophamnion, fluid accumulates in the narrow intervening space
(Fig. 90, C). It is possible that, owing either to the osmotic action
of this fluid or to pressure exerted on it during growth of the egg,
the embryonic mass becomes constricted with the result that it
finally becomes separated into two groups of blastomeres, each
capable of development into an individual parasite. The simple
condition described in Platygaster kiemalis appears to be explain-
able on this hypothesis and that the more complex cases of
polyembryony are highly specialised phases derived from it.
Marchal lays stress upon the influence of relatively sudden changes
of osmotic pressure in promoting embryonic dissociation, and
points out that a parallel is afforded by the experimental
results obtained by Loeb and by de Bataillon with the eggs of
sea urchins and of the lamprey. By using fluids exerting
different degrees of osmotic pressure, they were able to induce
separation of the blastomeres and their development into distinct
individuals.

The Embryonic Membranes. In recent years many instances
have been brought to notice, among endoparasitic Hymenoptera,
in which the embryo is surrounded by a cellular membrane. Its
existence has been observed in members of all the main parasitic
groups, but in some cases only in its last stage as a remnant still
attached to the newly hatched larva. Authors have variously
referred to it as the trophic membrane, trophamnion, amnion,
serosa or pseudoserosa. Its method of origin, however, has only
been determined in a few instances, and there appears to be little
doubt that it is not the same in all cases. The fact that an
embryonic membrane of this type is developed in eggs very
deficient in yolk, or almost devoid of it, has led to the conclusion
that it plays an important part in nourishing the developing
embryo. It is believed to absorb and elaborate food material
from the host tissues and act as a means of transferring such
products to the embryo with which it is in close association,

360 PARASITISM

hence the term trophamnion. In the eggs just alluded to the
chorion is either wanting, or is shed at an early stage. The
type of membrane which, as already described, originates from
the polar area of the egg occurs in poly embryonic forms
irrespective of their taxonomic affinities, and also in certain
closely related monoembryonic species. In other examples an
embryonic membrane is described as originating either as the result
of simple delamination of the blastoderm, or by the extrusion of
cellular material to the exterior, and its ultimate fusion to form a
complete investment of the egg. The process of delamination is
described by Henneguy (1892) in Chalcis (Smicra) clavipes and
by Silvestri (1908) in Encyrtus aphidivorus and Oopthora semblidis,
these three species being Chalcids. Spencer (1926) states that in
the Braconid Diccretus rapce the embryonic membrane is formed by
the delamination or extrusion of cells from the ectoderm early in
development. A unique condition is described by Silvestri (1906)
in Litomastix truncatellus in which two embryonic membranes are
present, viz. an outer membrane or true trophamnion, and an
inner membrane formed by delamination during the morula stage
of development. It is evident that the actual method of origin
of the second type of envelope needs fuller investigation, but it is
probable that functionally both types behave alike and their
subsequent history is very much the same. The term trophamnion
is best confined to a membrane derived from the polar bodies of
the egg, while in those cases where the fate of the polar bodies is
different, and the embryonic membrane is formed by delamination
or by cellular extrusion, such a membrane may be termed a
pseudoserosa. The use of the terms amnion or serosa is open to
objection, since these envelopes have a different origin and
significance.

Either just before, or a short time after, the larva has assumed
independent existence the embryonic membrane (either troph-
amnion or pseudoserosa) usually distintegrates ; its cells separate
apart or adhere in small aggregations and become liberated into
the body-cavity of the host. Here they may continue to grow
in size, but undergo evident nuclear changes, and it is probable
that they usually serve as food for the growing parasitic larva.

EMBRYONIC MEMBRANES 361

Both Spencer (1926) and Jackson (1928) have definitely shown
that they serve this function, and in Perlitus rutilis Jackson
mentions that they absorb fatty material from the host and
increase notably in size ; in cases where the parasitic larva happens
to die the unconsumed membrane cells increases from an original
size of about 60/Lt to a maximuin dimension
of 580/x.

According to Parker (1931), " pseudo-
germs " or dissociated fragments of the
trophamnion in Macrocentris are liberated
into the body-cavity of the host. They
grow, and also divide by fusion, into many
smaller entities and eventually they are
consumed by third and fourth stage larvae.

In some cases the membrane remains
intact for a time after the young larva has
emerged from the egg. In the Ichneumon
Therion morio, a parasite of the caterpillars
of the fall webworm in Canada, Tothill
(1922) states that the embryonic membrane
completely invests the first stage larva from
August until the following June. He terms
this phase the feeding embryo, since the
young Therion grows, and absorbs the
body-fluid of its host, while thus enclosed.
Grandori (1911) found that in Apanteles ^%!:^^,'^%Zira^^
glomeratus the whole envelope encloses the invested by its

larva throughout its life. At first it is a embryonic envelope,

° whose cells have

complete closed membrane, but subse- assumed the form of a

quently lacunae form and it assumes the reticulum. (Adapted
^ , : , . from Grandori.)

condition of a protoplasmic i-eticulum

(Fig. 92). When the larva is fully grown the protoplasmic

prolongations break down, and the cells themselves become free,

but retain their vitality up to the time the parasite bores its

way out 6f the host.

Origin of Mixed Broods of Parasites. The factors governing

the production of broods of polyembryonic parasites, containing

362 PARASITISM

individuals of both sexes, have been much discussed. Evidence
points very strongly to the conclusion that an egg laid by a
fertilised female gives rise to an exclusively female brood of
parasites, while an unfertilised egg produces males. It has
frequently been observed that among the individual parasites
emerging from a host both sexes are represented, and this fact is
accounted for on the basis of more than one egg having been laid
originally. Giard noted that nearly 3,000 individual Litomastix
may issue from a single Plusia larva ; both sexes were represented,
and it is certain that so large a number of parasites were not
derived by the division of one egg. Patterson records instances
where only a single male has been reared among broods of 919 and
1,550 individuals respectively, and it is believed by Leiby (1926)
that in such cases the male resulted from a monoembryonic egg.
Patterson (1919), on the other hand, believed that individuals of
both sexes can, and do, develop from a single fertilised egg on the
basis of Bridges' idea of somatic non-disjunction of the sex
chromosomes in a fertilised egg ; if certain of the blastomeres
received only a single X-chromosome those blastomeres so affected
ultimately give rise to one or more male embryos. In view of the
absence of cytological evidence supporting this hypothesis the
first-mentioned explanation is the one generally accepted. Leiby's
studies on Platygaster are based on carefully controlled experiments
supplemented by microscopic examination, and it appears that
a female may deposit several eggs in one host individual
and mixed broods result. Some of the eggs are inseminated
while others are not, and unfecundated eggs are deposited
along with those containing sperms by a female known to
have been fertilised. In controlled experiments with Copidosoma
gehchiw, in which a fecundated female was only allowed to
oviposit once in each host, either pure male or pure female
broods resulted. Ten broods resulting from eggs laid by
virgin females were comprised of males only. His studies of
various factors involved in these experiments have led him
to conclude that insemination of the eggs is apparently under
the control of the female, as is believed to obtain in the hive
bee.

PARASITE BEHAVIOUR 363

The Behaviour of Adult Parasites

Feeding Behaviour. Little definite information exists relative
to the feeding habits of many groups of adult parasites ; there
are a number of species which may be reared under experimental
conditions, where they freely mate and deposit their eggs in the
normal manner without food, owing perhaps to the exigencies
of the reproductive instinct. The vast majority of the
Ichneumonoidea, and probably all the Tachinidae, occur freely
in the open and are very constant visitors to umbelliferous
and other flowers. On the other hand, there are many parasitic
Hymenoptera which never appear to frequent flowers, and
probably subsist upon honey-dew and other organic material
available.

A phenomenon first noted by Marchal in 1905 in the Chalcid
genus Tetrastichus, and frequently recorded in other Hymenoptera
by subsequent observers, is the habit of feeding upon the blood
of the same, or another, individual of the species which is destined
to serve as the host for the offspring. In these cases the actual
feeding takes place at punctures made by the ovipositor through
which the blood of the victim exudes. It is possible that
nitrogenous food of this kind is necessary to ensure complete
maturation of the eggs, but whether or not this is the correct
explanation, the habit is prevalent in many species. The punctures
may be made in the eggs, larvae, pupae or adults of the species
attacked and, when expressly made for feeding purposes,
oviposition often takes place in another individual of the same
species. Not infrequently a number of individuals may be
punctured in succession without any deposition of eggs taking
place. Muesebeck and Dohanian (1927) are among the most recent
observers of this behaviour, and they state that it prevails among
nearly all the parasites of Apanteles melanoscelus, a primary'
parasite of the gipsy moth. The habit is sometimes so extensive
that the entire fluid content of the Apanteles may be consumed,
and in other instances, where oviposition has followed, the young
hyperparasitic larvae perished through insufficient nutriment being
afforded by the primary parasite. These observers consider

364 PARASITISM

that the habit increases the capacity for destruction by the
hyperparasites concerned, since the primary parasites thus fed
upon always fail to develop. This peculiar method of feeding -^
has been mainly recorded among Chalcids, with but few examples
among the Ichneumonoids, and it has also been noted in Vespoid
parasites of the family Bethylidse.

In 1921 Trouvelot published observations, which were subse-
quently extended in a later paper (1924), upon a remarkable
development of the same type of behaviour in the Braconid
Habrobracon johannseni, which punctures the cocoons of its host,
the potato tuber moth. It appears that in piercing the cocoon,
in order to reach the moth larva within, the Braconid secretes at
the same time a viscid fluid which forms into a minute gelatinous
tube, thus establishing a connection between the body-fluid of
the larva and the exterior. After completion of the operation,
the parasite withdraws its ovipositor and applies its head to the
mouth of the tube and sucks the blood of the host through this
improvised funnel. In 1921 Lichtenstcin also recorded an
identical method of feeding in the Chalcid Hahrocytus cionicida
which oviposits in pupae of the weevil Cionus. Its ovipositor is
inserted through the cocoon of the host until it enters the pupa
within ; the organ is retained in this position for about thirty
minutes, until the secretion hardens around it, after which it is
withdrawn and feeding takes place. Faure (1924) has described
the same process in considerable detail with respect to two Chalcids
of the genera Pteromalus and Eurytoma, which parasitise Apanteles
glomeratus after the latter has formed its cocoon. He concludes
that the habit is probably frequent as a method of extracting
nutriment from hosts enclosed in cocoons, or in other receptacles,
which offer difficulties to feeding by the more direct method
already described. Other instances of this remarkable phase of
behaviour are alluded to by Faure which render it unnecessary
to make further mention here.

Phoresy. According to Howard (1927) P. Lesne proposed the
word phoresie in 1896 with reference to cases where one animal

^ Many recorded instances are enumerated by Wheeler, The Social Insects,
London, 1928, p. 52.

LITERATURE 365

acts as a vehicle of transportation to another, and this term in its
anglicised form, phoresy, as suggested by Howard is adopted here.
The habit has so far been little investigated and appears to be
mainly developed among parasites, where it is evidently an
adaptation enabling them to obtain access to their hosts. Several
examples are instanced by Ferriere (1926), which are quoted
along with certain others in the article by Howard just alluded
to. It is well known that the triungulin larvae of certain blister
beetles attach themselves to the legs and hairs of bees in order to
reach the nests of these insects, where they complete their
transformations. Very similar habits are displayed by the Chalcid
Schizaspidia studied by Clausen (1923), whose primary larvae, or
planidia, attach themselves to the legs of passing ants and are
carried by them to their nests, where they become parasites of the
brood {vide also p. 334). Phoresy, in the case of adult parasites,
is a recent discovery of particular interest in that it leads to
the beginning of parasitism by the imago. Certain Scelionid
egg-parasites of Orthoptera attach themselves as adults to
female grasshoppers and remain until those insects deposit their
egg-masses, whereupon the parasites utilise the latter as hosts for
their own eggs. An advanced type of similar behaviour is
described by Chopard (1922) with reference to the species Riela
manticida Kieff., whose development takes place in the eggs of the
common praying mantis. Upon emergence the adult parasites
make their way to the mantids, upon whose bodies they settle
down. In this situation they cast off their wings and lead a true
ectoparasitic life, since Chopard asserts that they undoubtedly
feed upon their mantid vector. In cases where the latter is a
female, and has commenced oviposition, the Riela moves to the
genital region in order to lay its eggs in the viscid mass of the
ootheca as it is being formed. Individuals which settle upon
male mantids are stated to be short-lived and perish with their
hosts.

Literature

Baer. 1920. Zeits. Angew. Ent., VI., 185.
Chopard, 1922. Ann. Soc. Ent. Fr., XCI., 249.

366 PARASITISM

Clausen, 1923. Afin. Ent. Soc. Am., XVI., 195.

1931. Proc. Ent. Soc. Washington, 72.

1936. Ann. Ent. Soc. Am., XXIX., 20\.
Clausen, King and Teranischi, 1927. Vide p. 419.
Daniel, 1932. Tech. Bull, 187, N.Y. Agric. Exp. Sta.
Evans, 1933. Bull. Ent. Res., XXIV., 385.
Faure, 1924. Rev. Zool. Agric. et AppL, XXIII., 225.
Ferriere, 1926. Mitt. Schweiz. ent. GeselL, XIII., 489.
Flanders, 1936. Journ. Econ. Ent., XXIX., 468.
Genieys, 1925. Ann. Ent. Soc. Amer., XVIII., 143.
Grandori, 1911. Redia, VII., 363.
Haviland, 1922. Parasitology, XIV., 167.

1922a. Quart. Journ. Micros. Sci., LXVL, 322.
Henneguy, 1892. C. R. Acad. Sci. Paris, CXIV., 133.
Howard, 1927. Ent. Neivs, XXXVIII., 145.

1929. Ann. Ent. Soc. Am., XXII., 341.
Imms, 1918. Quart. Journ. Mic. Sci., LXIII., 293.
Jackson, 1928. Proc. Zool. Soc. London, Pt. II., 597.
Kelly, 1914. Journ. Agric. Res., II., 435.
King, 1916. Ann. Ent. Soc. Am., IX., 309.
Kornhauser, 1919. Journ. Morph., XXXVII., 195.
Leiby, 1926. Ann. Ent. Soc. Am., XIX., 290.
Leiby and Hill, 1923. Journ. Agric. Res., XXV., 337.

1924. Ibid., XXVIII., 829.
Lichti:nstein, 1921. C. R. Acad. Sci. Paris, No. 17, October.
Maple, 1937. Ann. Ent. Soc. Am., XXX., 123.
Marchal, 1898. Arch. Zool. Exp. et Gen., Ser. 4, II., 257.

1904. Ibid., CXXVL, 662.

1906. Ibid., IV., 485.
Martin, 1914. Zeits. iviss. Zool., CX., 419.
Muesebeck, 1918. Journ. Agric. Res., XIV., 191.
MuESEBECK and Dohamax, 1927. Bull. 1487, U.S. Dept. Agric.
Pantel, 1909. La Cellule, XXVI., 27.
Parker, 1924. Ann. Soc. Ent. Fr., XCIII., 261.

1931. Tech. Bull. 320, U.S. Dept. Agric.
Patterson, 1919. Journ. Heredity, X., 344.

1927. Quart. Rev. Biology, II., 408.
Pemberton and Willard, 1918. Journ. Agric. Res., XII., 285.
Pierce, 1910. Joitrn. Econ. Ent., III., 451.
Prell, 1914. Zeits. angeiv. Ent., I., 172.

1915. Ibid., II., 57.
Salt, 1934. Proc. Roy. Soc. B., CXIV., 450.

1935. Ibid., CXVII., 413.

1936. Journ. Exp. Biol., XIII., 363.

1937. Proc. Roy. Soc. B., CXXII., 57.
SiLVESTRi, 1904. Redia, II., 68.

1906. Ann. R. Scuola Agr. Portici, VI., 1 .

1908. Boll. Lab. Zool. Gen. e Agrar., Portici, III., 29.

1915. Ibid., X., 66.

1919. Ibid., XIII., 70.
Smith, H. S., 1912. U.S. Bur. Ent., Tech. Ser. Bull. 19, Pt. IV.

1929. Bull. Ent. Res., XX., 141.
Spencer, 1926. Atm. Entom. Soc. Am., XIX., 119.
Strickland, 1923. Dept. Agric. Canada, Tech. Bull. 26 (N.S.).
Thomsen, 1927. Vidensk. Medd. fra Dansk. naturh. Foren., LXXXIV., 73.

LITERATURE 367

Thompson. 1923. Ann. des Epiph., X., 137.

1924. Ann. de Parasit., II., 185, 279.

1926. Ibid., IV., Ill, 207.

1928. Parasitology, XX., 123.

1934. Ibid., XXVI., 378.
Thompson and Parker, 1927. Ibid., XIX., 1.
Thorpe, 1931. Proc. Zoo. Soc, 929.

1932. Proc. Roij. Soc. B., CIX., 450.

1934. Quart. J. Mic. Sci., LXXVII., 273.

1936. Parasitologif, XXVIII., 517.
TiMBERLAKE, 1912. U.S. BuT. Ent., Tech. Ser. Bull. 19, Pt. IV.

1913. Journ. Econ. Ent., VI., 293.

1916. Canad. Ent., XLVIII., 89.
ToTHiLL, 1922. Tech. Bull. 3 (N.S.), Dept. Agric, Ottawa.
TowNSEND, 1908. U.S. Bur. Ent., Tech. Ser. Bull. 12, Pt. VI.
Trouvelot, 1924. Ann. des Epiph., X., 1.
VoRis, 1934. Trans. Acad. Sci. St. Louis, XXVIII. (8), 242.
Wadsworth, 1915. Journ. Econ. Biol., X., 1.
Walton, 1914. Proc. Ent. Soc. Washington, XVI., 11.
Webber and Schaffner, 1926. Bull. 1363, U.S. Dept. Agric.
Wheeler, E. W., 1923. Ann. Ent. Soc. Amer., XVI., 1.

CHAPTER XIV
BIOLOGICAL CONTROL

Biological Control of Insect Pests, p. 869. I. Parasite Intro-
ductions. The Introduction of Vedalia into Ccdifornia, p. 370 ; The
Hawaiian Islands, p. 371 ; Later Work in California, p. 376 ; The Gipsy
and Brown-tail Moths in New England, p. 379 ; Other Biological Control
Problems in the United States, p. 381 ; Neiv Zealand, p. 385 ; Canada,
p. 386 ; Fiji, p. 387 ; Experiments in England, p. 389.

In its natural or original habitat a species of animal or plant is
ordinarily maintained in a condition of equilibrium ^ by the
interaction upon it of biological and other agencies. If, on the
other hand, any factor or factors disturb this state of equilibrium,
an animal or plant may be able to maintain a supernormal
ascendancy for a variable period. Similarly, one or the other
may find a home in a foreign country, affording it a favourable
environment where those agencies which control it in its original
habitat are ineffective or wanting. Under either of these conditions
the probability is that the animal or plant, as the case may be, will
multiply and spread so as to assume the status of a pest.

Among the natural agencies exercising a controlling influence
upon insect life the complex of meteorological factors constituting
climate is ever present. Simultaneously another complex of a
biological nature is exerting its influence, and represents the sum
total of the activities of bacterial, fungal and other diseases, of
insectivorous birds and mammals, and of parasites and predators.
The restraint exercised by these several biological agencies is
known as biological control, and in so far as insect life is
concerned, the influence of parasites and predators alone is of
immense significance. Plant life is likewise affected by a complex
of biological agencies including fungal, bacterial, and other

^ Equilibrium, in the strict meaning of the word, is not implied, but a
condition involving relatively small fluctuation within narrow limits.

368

BIOLOGICAL CONTROL 360

diseases, together with the depredations of those insects directly
dependent upon vegetation for their sustenance. From the
entomological viewi)oint, therefore, the problems of biological
control may be divided into those affecting insects and those
affecting plants.

The economic significance of insect parasites and predators,
which live at the expense of other members of their class, has led
during the last forty years to their increasing practical utilisation
as natural agents in pest repression.

Biological Control of Insect Pests

Biological control of insect pests may be considered under
two categories : {a) introductions of parasites into countries
where they did not previously exist, and (b) the utilisation of the
indigenous parasites of a country. The remarkable developments
which the application of this method has undergone in recent years
are almost exclusively with reference to parasite introductions.

The chief principles underlying the theory and practice of
biological control are outlined by Thompson (1930), and a general
text-book of the subject by Sweetman (1936) has recently appeared.
The mathematical theories vmderlying certain of these principles
have been discussed by several writers, and of these the most
recent are Nicholson (1933), H. S. Smith (1935), and Nicholson
and Bailey (1935). The last named have brought into prominence
the fact that if there is a balance in the population of a particular
insect some of the factors causing mortality must be dependent
upon host density and destroy a greater proportion of the
individuals of a specific host when the density of the latter is
high than when it is low. In the case of parasites and predators,
it is assumed that they search for their hosts at random, and it is
shown that the mortality which they cause will be dependent upon
host density. Since climate and other factors operate irrespective
of such density, parasites and predators are regarded as the main
factors bringing about balance. The reader who is specially
interested in this side of the subject is referred to these two
papers, which form a notable contribution to the theoretical
aspects.

370 BIOLOGICAL CONTROL

I. Parasite^ Introductions

Biological control has been notably successful in certain cases
where injurious insects have become established in lands they did
not previously inhabit, and have there developed into persistent
pests. In such instances the noxious species have been free from
the attacks of the parasites which restrain them in their country
of origin. Biological control aims at restoring the condition of
natural equilibrium by the introduction of the missing parasite
factor. The first definite experiment of this character appears to
have been made in 1873, when Planchon and Riley introduced
an American predatory mite (Tyroglyphus phylloccerce Riley) into
France for the purpose of endeavouring to secure a measure of
natural control over the Phylloxera of the vine. Although the
mite became established, no appreciable restraint over the pest
resulted, and control to-day has to be exercised by other methods.
In 1883 Riley attempted the control of the cabbage white butterfly
(Pieris rupee L.) by importing into North America one of its chief
European enemies, the Braconid Apanteles glomeratus. This
parasite has become well established.

The Introduction of Vedalia into California. Biological control
received its first stimulus in the subjugation of the fluted or
cottony-cushion scale (leery a purehasi) in California. In the
year 1886 the leery a became so virulent a pest in the orange and
lemon groves of that state that the matter attracted national
attention. Accidentally introduced on plants imported from
Australia, it multiplied to a degree which threatened the industry
concerned with extinction. The mission of A. Koebele, on behalf
of the U.S. Dept. of Agriculture, to study the natural enemies of
the fluted scale in its original habitat has often been recounted.
Suffice to say, he discovered in the small Coccinellid beetle,
Vedalia ^ cardinalis, an extraordinarily efficient predator. Its
successful transmission from Sydney to Los Angeles was speedily
followed by its liberation in the open, and in less than five years
the Icerya was so completely destroyed that the survivors became

^ As a convenience, the word " parasite " is often used in this chapter in a
general sense which implies both parasites and predators.
- Now known as Rodolia cardinalis.

HAWAIIAN ISLANDS 371

numerically insignificant. The entry of the Icerya into other
regions of the globe subsequently led to the introduction of the
predatory beetle into Florida, the Hawaiian Islands, New Zealand,
South Africa, Portugal, Italy, Syria, Egypt, the South of France
and Malta, with the same satisfactory results.

The remarkable efficiency of the Vedalia is to be ascribed to a
combination of biological features. The insect is very active
both as larva and adult, which easily seeks out the sedentary
fluted scales on the trees. Also, it has at least two generations
in the time its prey is passing through but a single generation.
Furthermore, the Vedalia appears to be free from natural enemies
of its own, an important and remarkable fact. The rapid and
complete establishment of the Agromyzid fly Cryptochcetum
iceryce, which was introduced from Australia along with the
Vedalia, has also proved an important factor in the control of the
cottony-cushion scale. It has become evident that the success of
the Vedalia cannot wholly be dissociated with the activities of this
fly. The widespread publicity accorded to the success of the
Vedalia experiment was, in a sense, a pit-fall for the growers, who
soon came to believe that the expense involved in insecticidal
treatment against other pests was no longer necessary ; all that
was required was to obtain the natural enemies of a pest and
subjugation of the latter would follow. As Howard remarks, by
blinding people to other and assured methods of repression, the
predominant success of the Vedalia actually retarded the general
progress of insect control in California.

The Hawaiian Islands. The sugar planters of these islands
have set an example to the rest of the world in the reliance they
have placed on the application of scientific knowledge to the
cultivation and protection of their crops. Among the severest
factors they have had to contend with have been the insect
enemies of cane. The latter are almost entirely immigrant species
which have become established in the Islands unaccompanied by
parasites or predators, which control them to a large extent in
the lands from whence the pests originally came (Imms, 1925;
Timberlake, 1927). The fact that artificial methods of repression
generally proved expensive, and gave poor results, had much

372 BIOLOGICAL CONTROL

to do with the adoption of biological control. Encouraged by
the early success of Vedalia in California, followed by the
subjugation of the cottony-cushion scale in the Islands in 1890
by similar means, the growers turned to biological methods of
relieving them of the immense damage wrought in the cane fields
by the leafho}:)per Perkinsiella saccharicida. The introduction in
1905-1906 by Perkins and Koebele of Paranagrus optabilis Perks,
from Queensland, and of Ootetrastichus heatus Perks, by Koebele
from Fiji, led to a rapid decline in the damage occasioned by the
leafhopper. Both these insects are small Chalcids attacking the
eggs of their host and, valuable as they proved to be, they scarcely
gave the degree of control required by the growers. It became
evident that some robuster type of leafhopper enemy was needed
which would, at the same time, withstand the bad conditions
occasioned by heavy rains. In 1920 Muir introduced the
predaceous Capsid bug Cyrtorhinus mundulus (Bred.) from P'iji
and Queensland. The habits of this insect are exceptional in that
it preys upon the eggs of the leafhopper, whereas other Capsidae
are almost exclusively plant-feeders. Muir's observations showed
that the habit was a fixed one, and the possibility of its varying its
behaviour under new surroundings and becoming a plant pest was
happily not fulfilled. After some difficulty the insect became
securely established in the Islands, and it is now generally conceded
that the Cyrtorhinus has completed the subjugation of the leaf-
hopper. The latter, which was once so abundant and so destructive,
requires searching for in the cane fields to-day.

Muir's success with the Cyrtorhinus followed his long journeys
in the Orient for the purpose of investigating other sugar cane
pests. The cane borer weevil Rhabocnemis obscurq (Boisd.) was
one of the pests which demanded a measure of control. Search
for the weevil in Southern China, the Malay States, Java and
other of the East Indian Islands brought no success, but eventually
Muir found the insect in Amboina, and furthermore discovered
that it was attacked by the Tachinid Ceromasia sphenophori Vil.
Difficulties in transporting living material of the fly to Honolulu
proved too great, and it became necessary to discover some other
locality from whence it would be easier to effect shipment. After

HAWAIIAN ISLANDS 373

visiting Ceram, Muir proceeded to New Guinea, which yielded
both the weevil and its parasite. Owing to the shortness of the
life-cycle of the Tachinid and difficulties attending the breeding
of it in small cages, it was found desirable to establish intermediate
breeding stations in Australia and in Fiji. After further difficulties
and delays a supply of parasites reached Fiji and sufficient stock
was bred from them to be conveyed to Honolulu. The first
examples reached Oahu in 1910, and by 1913 the Tachinid was
firmly established in the Islands, and noticeable reduction, not
only in the number of borers in the field, but also in loss of the
crop harvested, resulted. At the present day it would seem that
the Tachinid has attained its maximum efficiency, and Swezey
(1928) states that losses from the borer are negligible over the
greater part of the sugar cane area. In certain districts, which
have heavy rainfall and high winds coupled with a variety of cane
that lodges, the effectiveness of the Tachinid is much less marked.
In such areas the borer gains the ascendancy, and notwithstanding
efforts made in other lands to discover further enemies of the borer
beetle, the Tachinid remains as its most efficient controlling agency
{vide Williams, 1931).

In the case of the Lamellicorn beetle, Anomala orientalis
(Waterh.), a remarkably complete degree of control has been
achieved. It was probably introduced along with potted plants
prior to 1910, but fortunately the infestation had only assumed
local proportions before it was checked. When abundant this
species nearly, or quite, kills out the cane, and was causing an
annual loss of £10,000 to two plantations. Attempts were made
to control the beetle by various parasites, but success was not
achieved until Muir obtained the solitary wasp Scolia maniliw
Ashm. in the Philippines. Introduced in 1916, it increased so
rapidly that within two years the damage caused by the Anomala
was reduced to a remarkable extent. The Scolia appears to have
effectively checked the spread of the beetle find there are now only
occasional local outbreaks, not of a very serious character.

Another remarkably complete example of biological control
within a short interval of time is afforded by the avocardo
mealybug {Pseudococcus nipce Mask.) which affected, besides

374 BIOLOGICAL CONTROL

avocardo, fig, mulberry, guava, coconut and other trees. Through
the efforts of H. T. Osborn, the Encyrtid parasite Pseudaphycus
utilis Tinib. was introduced from Mexico into Oahu in 1922. In
about two years the mealybug has been practically exterminated
and the rapidity of spread and the completeness of the parasitism
exercised by this Chalcid have been truly remarkable.

The army worms Cirphis unipuncta (How.) and Spodophera
mauritia (Boisd.) have been troublesome for many years, especially
in sugar plantations adjacent to grass land. A number of parasites
have been introduced from time to time until the army worms
were w^ell controlled. Outbreaks to-day seldom occur in most of
the Islands, and not so often, or so extensively, on the island of
Hawaii, where they have been the most prevalent (Swezey). The
introduction of the Chalcid Euplectrus platyhypence How. and the
Tachinid Archytas cirphis Cur. by H. T. Osborn from Mexico in
1923 and 1924 had much to do with achieving this result. Several
other parasites, including the egg-parasite Telenomus nawaii Ashm.
from Japan, were also contributing factors.

The Mediterranean fruit-fly, Ceratitis capitata Wied., was
discovered in the Islands in 1910, and had probably made its
entry a year or two previously from Australia. It spread rapidly
and severely attacked nearly all kinds of fruits of importance,
excepting banana and pineapple ; peach, mango and guava were
the worst infested, and sound fruits of these kinds became hard to
obtain. In 1912 Professor F. Silvestri was engaged by the
Hawaiian Board of Commissioners of Agriculture and Forestry
to search for fruit-fly parasites. Silvestri mainly concentrated
his efforts in Africa, but also visited Australia, and of the species
of parasites he brought to Honolulu, the Braconids Opius humilis
Silv. from South Africa and Diachasma tryoni Cam. from Australia
became effectively colonised. In 1914 the Territorial Government
commissioned D. T. Fullaway and J. C. Bridwell to visit West
Africa in order to endeavour to obtain other parasites. The result
of their efforts was that the Braconid Diachasma fullawayi Silv.
and the Chalcid Tetrastichus giffardianus Silv. were successfully
established. All four of the above-mentioned species are now
widely distributed through the Islands, and on an average from

HAWAIIAN ISLANDS 375

36-2 per cent, to 56-4 per cent, of the fruit-fly larvae are parasitised.
It is recognised that, although a sufficiently high degree of control
from the commercial standpoint has not been obtained, the
parasites have decreased the infestation to a marked extent.
While coffee cherries and guavas have been largely freed from
infestation, other fruits, like the mango, which have a thick pulp,
are much more severely attacked, since the parasites are unable
to reach many of the Ceratitis larvae within them. With the
object of effecting a more adequate measure of biological control
of this insect, entomologists are now engaged in studying the
possibility of further parasite introductions to this end being
effected from abroad.

It would require too much space here to recount all the
examples of biological control attempted or achieved in these
islands : a chronological list of them up to about 1930 is given in
the book by Williams (1931). At the present time the horn-fly
pest of cattle is still uncontrolled, and the same applies to the
pineapple mealybug, and some other pests. It must not be
assumed that success in biological control in these Islands has
come easily, or that every beneficial insect introduced has become
established in its new surroundings. A history of the subject
would have to record a number of failures, and the expenditures
of strenuous effort and large sums of money. Thus, the complete
control of the cane leaf hopper only resulted after the introduction
of about a score of different parasites and after about twenty
years of work. On the other hand, the control of the Anomala
beetle and the avocardo mealybug was achieved in a remarkably
short interval of time.

Viewed as a whole, biological control has achieved a degree of success
in the Hawaiian Islands not so far attained elsewhere. The majority of
the important pests have become subjugated by this method, and there
is little necessity for the application of insecticides. At the same time
it must be remembered that the limited area of the terrain, its isolation,
and its warm, equable climate, all favour measures of this character.
Beneficial insects can be introduced at almost any time of the year, and
are able to pass through uninterrupted cycles of development owing
to the absence of a dormant season. It is true that such conditions
conduce to the multiplication of pests, but they are at least as
favourable to parasites and predators. Furthermore, the unique

376

BIOLOGICAL CONTROL

character and simplicity of the indigenous fauna with its paucity of
hyperparasites and other competing forms, provide a biological
association exceptionally favourable to the introduction of beneficial
species which meet with but few obstacles in the struggle for existence.

Later Work in California. Following the introduction of
Vedalia cardinalis, already alluded to. biological control in

Fig. 9.'}. A typical insectary used for rearing Crijptohvmns. San
Gabriel Valley Pest Control Association, Lamanda Park,
California. (Photo supplied by courtesy of Professor H. S.
Smith.)

California had, for a time, a somewhat chequered record. Grave
mistakes were made by unscientific enthusiasts, large sums of
money expended with no appreciable results, and world travel
undertaken by agents of the State with disappointing consequences.
But as H. S. Smith remarks (1926), this is past history, and the
enthusiasm of the public for biological control is now tempered
by a far better knowledge of its limitations. The worst pests of
the citrus groves are Coccidae, and following the control of the

LATER WORK IN CALIFORNIA Zll

Icerya scale, mealybugs (P seudococcus spp.) next demanded
attention. The introduction of the Coccinellid Cryptolcemus
montrouzieri by Koebele from Australia about 1892 is familiar
to entomologists, and the most interesting feature respecting
this predator is the fact that in areas where it has been unable
to maintain itself, it can be induced to dominate its prey by
continuous laboratory breeding and distribution. It seems certain
that the citrophilous mealybug {P, gahani Green), a citrus pest
of first importance, has been largely kept under control by this
method. There are a number of special insectaries (Fig. 93),
located in infested counties or districts, which are given over to
the rearing and distribution of the Cryptolcemus, an enterprise
which has been developed on a commercial basis {vide Smith and
Armitage, 1931). In these structures the mealybugs are reared on
shoots of potato previously sprouted in the dark, and when a
sufficient supply of the host is available the Cryptolcemus is
introduced and allowed to multiply freely under favourable
conditions of constant temperature. During 1924, 4,038,238
Cryptolcemus were reared and distributed at an inclusive cost of
$48,750, and the total citrus area infected with mealybugs was
approximately 20,000 acres. In comparison with this method it
has been pointed out that the cost of a single spraying for that
acreage would amount to about 8200,000, and several spray
applications per year are necessary, in order to ensure control of
the pest. Notwithstanding the destruction exercised by the
Cryptolcemus, the citrophilous mealybug has steadily increased its
range in Southern California ; in 1928 the infested area increased
to about 50,000 acres, and some 40,000,000 of the predators were
being bred for distribution. Believing that the efficacy of biological
control would be strengthened if additional enemies, especially
internal Hymenopterous parasites, _^ were available, efforts have
been made to locate the country of origin of the pest. In 1916 and
1917 Clausen explored Japan, China, Formosa and the Philippines
without success. Further search was made by F. Silvestri, who
was already in the Orient, in the countries visited by Clausen,
and also in Indo-China and the Malay Archipelago without
revealing its presence. In 1927 Compere located the insect in

378 BIOLOGICAL CONTROL

Australia, where he also discovered that it was attacked by
important enemies. Of the latter two Chalcid parasites (Cocco-
phagus gurneyi and Tetracnemus pretiosus), which were introduced
along with other enemies, have proved remarkably successful.
The Pseudococcus has become reduced to negligible numbers and
the saving of large annual expenditure has resulted. The
contention that the observed results have been due to the
activities of these two species of parasites is based upon the fact
that, without exception, the presence or absence of serious
infestations has been positively correlated with the absence or
presence of the parasites referred to (Compere and Smith, 1932).
The outcome has been that the Cryptolcemus has now become
relegated to second place and its propagation no longer has its
former importance.

Among other mealybug parasites that have been introduced
into California the Chalcid Leptomastidea abnormis (Gir.) is
noteworthy. Established in 1915 from stock colonies obtained
in Sicily, this insect has played an appreciable part in the reduction
of Pseudococcus citri. Attempts have also been made to augment
its effectiveness by mass breeding in the county insectaries
referred to, and over one million examples were liberated in 1924.
In 1922 the Coccinellid Scymnus bincevatus was successfully
introduced from Cape Town, and is now an accessory in the
control of Pseudococcus gahani. The most important of all
Calif ornian citrus pests is the black scale (Saissetia olece) and the
biological control of this insect is one of the main projects of the
Citrus Experiment Station at Riverside. The ladybird Rhizobius
ventralis, introduced by Koebele many years ago, and the Chalcid
Scutellista cyanea brought in from South Africa in 1902, destroy
great numbers of this scale insect, but their combined activities
fail to exercise the standard of control demanded by the growers.
In 1919 the Chalcid Metaphycus lounsburyi was introduced from
South Africa, and it has since become well established in the
coastal zone of the State. In this region the life-cycle of its host
is such that the phases of development, upon which the Metaphycus
is able to breed, exist at all seasons of the year. On the other
hand, in the interior of the State, owing to different climatic

GIPSY AND BROWN-TAIL MOTHS 379

conditions, the necessary phases of the host are not available for
many months, and the parasite consequently becomes scarce.
After a few years, wherein this parasite effected marked control of
its host, its activities became largely neutralised by the secondary
parasite Quylea whittieri. Notwithstanding the combined activities
of the entomophagous insects already mentioned, together with
others of lesser importance, the black scale maintains its rank as
the worst of citrus pests.

The red scale (Chrysomphalus aurantii) and the purple scale
{Lepidosaphes beckii) likewise have not so far been subject to
satisfactory biological control. In 1925 a search for further
parasites of the first-mentioned species was undertaken on behalf
of the State by F. Silvestri in China. Various new parasites and
predators were located, and shipments made to California, but
owing to the civil war and other circumstances work in the Orient
has had to be discontinued for the time being.

For detailed information respecting the application of biological
control in California, and the various hosts and their parasites
that are involved, reference should be made to the work by Essig
(1931).

The Gipsy and Brown-tail Moths in New England. The most
prolonged and extensive campaign of biological control so far
waged in any quarter of the globe is that centred round the
gipsy moth {Porthetria dispar L.) and the brown-tail moth
{Euproctis chrysorrhoea L.) in New England. Both insects came
originally from Europe, and appeared first in Massachusetts ; the
gipsy moth in 1867 and the brown-tail in 1897. Some twenty
years elapsed before the gipsy moth developed into a serious pest,
and it was not until 1905 that any campaign was formulated for
introducing natural enemies of either of these insects. Previous
observations had shown that the aggregate effectiveness of
indigenous parasites in controlling these species was wholly
insignificant. On the other hand, both species were found to be
heavily parasitised in Europe, and it seemed further evident that
whatever degree of restraint is exercised by natural enemies in
Europe, it is the result of a sequence of species attacking the
eggs, larvae and pupae respectively. Through efforts initiated by

380 BIOLOGICAL CONTROL

L. O. Howard, the work of parasite introduction has now gone on
for about twenty-five years, and after exploring European countries
Federal entomologists have also directed their attention to Japan.
Some 47 species of parasites and predators have been brought into
the United States, where many millions of individuals have been
reared and liberated (Burgess and Crossman, 1929). This vast
experiment is governed by the contention that only the strongest
possible sequence of enemies is likely to exercise any appreciable
influence over the pests in question now that they are so firmly
established. Continuous effort is consequently being made to
assemble all likely parasites from whatever country the hosts
inhabit. Up to 1929 fifteen species of introduced parasites had
become established in New England, and of these the following are
now abundant : the beetle Calosoma sycophanta L., a predator of the
larva? and pupa? ; the Tachinid parasites Compsilura concinnata
Meig. and Blepharipa scuteUata R.D., and the Braconid Ajxinteles
melanoscelus Ratz., which attack the larva; ; and the Chalcid
egg-parasites Anastatus hifasciatus Fonsc, and Schedius kuwance
How. It is extremely difficult to evaluate the effects of these
parasite introductions, owing to the existence of a number of other
controlling agencies. Among the latter climatic variations, wilt
disease, the reduction of favoured shade trees, arsenical spraying,
and the wholesale destruction of egg-clusters, all have to be taken
into account. It is noteworthy that the average collective
parasitism was at its highest in 1923, and in the summer of 1924
gipsy moth infestation reached its lowest level for twenty years.
Since then a decline in parasitism has been observed, and in several
parts of the infested area this has been followed by increased
abundance of the pest. Fluctuations of this character will most
likely recur periodically, and at present there is no reliable
indication as to whether the gipsy moth or its enemies will
ultimately gain control. Those who are inclined to believe that
parasite control has been a failure in this particular instance are
in the difficult position of having to prove that the infestation
to-day would not have gained materially, either in intensity or
distribution, if parasite introductions had not been made.

The brown-tail moth, on the other hand, gives every indication

OTHER PROBLEMS IN UNITED STATES 381

of tolerably successful control, and it now only occurs in destructive
numbers in limited areas. This result is credited to the establish-
ment of a more effective sequence of parasitism, but owing to the
multiplicity of other factors involved absolute proof of the
contention is scarcely attainable.

Other Biological Control Problems in the United States. Among
noxious insects which have accidentally entered the United
States in recent years, the alfalfa weevil, Hypera variabilis Hbst.
(Phytonomus posticus Gyll.), the European corn-borer, Pyrausta
nubilalis Hiibn., Oriental fruit moth, Cydia molesta Busek., and
the Japanese beetle, PopilUa japonica Newm., are especially
noteworthy. They continue to increase and spread to a degree
which affords conclusive evidence of their escape from effective
control. As W. R. Thompson observes (1928), the only factor
present in the native home of most introduced species and absent
in their new terrain is the parasitic factor. The probability,
therefore, is that the absence of parasites and predators is the real
cause of their increase. Attempts to restore the loss of balance
occasioned by the absence of the parasite factor form an important
part of the programmes of control with respect to the four pests
mentioned.

The alfalfa weevil appears to have been first reported in
America in 1904, when it appeared in Utah ; by the end of 1912
it had also invaded Idaho and Wyoming, and by 1926 it had
spread into Washington, California, Montana and Nevada : since
then further States have come under infestation. The introduction
of several European parasites has resulted in one important
species, e.g., the Ichneumonid Bathyplectes curculionis Thoms.
becoming well established. During the years 1912 to 1914
liberations were made in the fields of Utah and colonisation
appears to have been easily effected, although the area of
dispersion remained extremely limited (Chamberlin, 1926).
Later the parasite spread rapidly, and after about eight years
its distribution was about coincident with that of the weevil. In
1925 it parasitised weevil larvse to the extent of 85 per cent, near
Salt Lake City, and to an even higher degree in some other places.
On the other hand, it attacks only the older larvae and is not

382 BIOLOGICAL CONTROL

uniformly effective during the whole period when these are
available as hosts. Apart from Bathyplectes, only one other
introduced parasite of the weevil is stated to have become properly
established. Control of this pest by biological means does not, at
any rate, at present seem to afford more than an accessory
measure. Cultural and other methods of control remain the most
important weapon in combating it.

The European corn-borer is proving to be one of the most
formidable of all foreign pests that have become introduced into
North America. Discovered in Massachusetts in 1917, when it
was already infesting a known area of about 100 square miles, it
has since extended its range with great rapidity, and is still
continuing to do so. In 1924 it had infested territories in Southern
Ontario and the United States amounting to 42,773 square miles,
and by the end of 1926 this area was increased to about 93,000
square miles. In 1934 the total area infested was stated to amount
to over 312,000 square miles. Substantial commercial losses,
however, only occurred in scattered districts over this wide
expanse of territory. In 1919 an investigation of the status and
controlling factors of the insect in Europe was commenced, and
this has led to the accumulation of a large amount of ecological
and other data. It appears that, with the exception of certain
parts of central Europe, the corn-borer is controlled by a complex
of climatic, cultural and parasitic factors, with the result that
it seldom attracts attention. These several factors vary both
quantitatively and qualitatively in different ecological zones
inhabited by the insect, and also from year to year. In so far
as parasitism is concerned, it is clear that it forms but one of the
many factors operating against the corn-borer. At least nineteen
species of parasites had been reared from this host up to 1928, but
their relative frequency varies enormously in different regions.
Thus in the vicinity of Paris an average parasitism of over 36
per cent, has been noted, but only four species of parasites are
present, while in south-western France the parasites number nine
species, yet their collective parasitism averages less than 15 per
cent. Thompson (1928) has computed the relative importance
of the different components of the parasitic factor in the control

OTHER PROBLEMS IN UNITED STATES 383

of the corn-borer in different zones or areas, and in a bulletin
written in conjunction with H. L. Parker (1928) he remarks as
follows : " Since control in any given region results from the
action of all the factors working together, it seems reasonable to
assume that the absence of any constant cause of mortality will
ordinarily disturb the natural equilibrium, and permit an
inevitable, though perhaps gradual, increase in numbers from
generation to generation until great economic injury results."
The introduction of parasites of the corn-borer has been in
progress since 1920. European shipments are received at the
special laboratory located at Arlington (Mass.), where stocks are
bred up in sufficient numbers prior to liberation. At the present
time no imported parasite is sufficiently numerous in the field to
be an appreciable factor in corn-borer control, but in view of the
magnitude of the area to be covered a more optimistic result would
be a premature expectation. It is evident, however, that from
among about twenty introduced species of parasites two are of
special importance, viz., the Ichneumonid Inareolata jmnctoria
and the Tachinid Lydela stahulans (Masicera senilis). These latter
show considerable promise as controlling agencies, but success in
a project of this kind is a matter of continued efforts spread over
what will probably prove to be a very considerable length of time.
The Oriental Fruit Moth Cydia {Grapholitha) molesta Busek. as
a larva bores into the shoots and later the fruit of especially the
peach. It was probably an accidental immigrant into the United
States from Japan about 1913. First detected in the environs of
Washington city, it has spread to every peach-growing district in
the eastern states and also further west. It is common in the
Ontario province of Canada and appears to be still extending its
range in America. In the absence of uniformly satisfactory
artificial measures of repression, biological means of control have
come to the fore in recent years in regard to this pest. Notwith-
standing the importation of foreign parasites, the most effective
natural agent has proved to be the Braconid Macrocentrus
ancylivorus — an indigenous parasite which had been previously
known to utilise the strawberry leaf-roller as its host. The
Macrocentrus can be bred in large numbers and liberated where

384 BIOLOGICAL CONTROL

required, and in this way it has been widely disseminated wherever
it did not previously occur. It has now become to be relied upon
in preference to any other natural enemies, since it often attains
90 per cent, or more parasitism of the fruit moth larvae. Its chief
value is as an accessory, since artificial control usually has to be
maintained also in the chief centres of infestation. The promise
shown by Macrocentrus has led to its introduction into other
countries where the fruit moth is prevalent, notably into Italy and
Japan.

The Japanese beetle was introduced in shipments of Japanese
iris in 1911, and in the course of ten years it has invaded about
14,000 square miles of territory, including the entire state of New
Jersey and portions of Connecticut, New York, Pennsylvania and
Delaware (L. B. Smith, 1928). At the present time, however, it
covers a still wider area and has reached the Middle West, while
its entry into Canada is being closely watched. The heaviest
infestations still occur in New Jersey and adjacent territory. Its
chief depredations are caused by the adult beetle attacking the
fruit and foliage of apples and peaches, but it also affects a great
variety of other fruits, together with vegetable crops, shade trees,
etc., while the larvae, it may be added, are destructive to lawns
and pastures. In 1920 work on the insect enemies was commenced
in Japan, where parasites are important agents in controlling the
insect. The fact that at least five species of parasites participate
in reducing the abundance of the beetle in Japan led the United
States Bureau of Entomology to undertake the problem of building
up an ecological association of parasites similar to what obtains
in that country (King, 1928). In 1922 investigations were
extended to Korea, where the study of parasites of closely allied
beetles was commenced, and later the work embraced also
parasites of allied beetles found in China and India. Since this
expansion has taken place, parasites have been transmitted to
America from each of the countries mentioned. From among
about a dozen or more possible species, those established in the
United States up to 1927 included three Tachinids and two solitary
wasps of the genus Tiphia. Of the Tachinids, Centeter cinerea
parasitises the adult beetles, while Prosena siherita and Dexia

NEW ZEALAND 385

ventralis are larval parasites. The two wasps, viz., Tiphia
popillivora and T. vernalis, are ectoparasites of the beetle larva?
(Clausen, King and Teranishi, 1927). Other species of parasites
have been introduced since that year, but the total area so far
colonised is too restricted to warrant any definite forecast being
made as to the final outcome of these biological measures.

New Zealand. In so far as forest pests are concerned the
introduction into New Zealand of the Coccinellid Rhizobius
ventralis from Australia in 1900 for the purpose of controlling the
eucalyptus scale (Eriococcus coriaceus) has proved exceptionally
successful. The eucalyptus leaf weevil (Gonipterus scutellatus) is
now present in greatly reduced numbers in the North Island owing
to the activities of the Mymarid Anaphoidea nitens which was
introduced from Australia in 1927. The golden oak scale
(Asterolecanium variolosum) is stated to be under adequate control
in most localities owing to the introduction of Hahrolepis dalmanni
from the United States. The control of the " Chermes," Pineus
pini, was attempted in 1932, when the Agromyzid fly (Leucopis
sp.) was introduced from England, but so far no satisfactory
results have been obtained. The introduction of the large
Ichneumon Rhyssa persuasoria in 1928 for the purpose of
controlling the wood wasp {Sirex noctilio) has apparently resulted
in its becoming established, but the economic outcome of the
project cannot yet be forecasted. The Cynipid Ibalia, introduced
about the same time, proved a failure.

As regards fruit pests, the reduction of the cottony-cushion
scale to the status of an occasional and minor pest of citrus, pear
and apple by the efficacy of the Coccinellid Vedalia cardinalis is
now well known. The control of the woolly aphis of apple by the
introduction of Aphelinus mall from the United States in 1921 is
an outstanding example of the biological method. The Aphelinus
is now successfully established throughout New Zealand and its
host has been reduced to the condition of a minor pest.^

1 The importance of Aphelinus mali as an agent in the control of the woolly
aphis of the apple has attracted almost world-wide attention {vide Howard,
1929). In several countries of South America it is now well established, and
exercising a high degree of control ; in Argentina the control is stated to have
exceeded all expectations. In Italy it has been introduced from Uruguay and

R.A. ENTOMOLOGY. 13

386 BIOLOGICAL CONTROL

Among field crop pests the accidental establishment of Pieris
rapce in 1929 led to the introduction of the Chalcid Pteromalus
puparum. This scheme is proving strikingly successful and the
parasite is stated to be destroying over 90 per cent, of the pupae
of this butterfly.

A number of other introductions have been made for the purpose
of controlling various insect pests : some of these have proved
failures, but as regards the larger number of cases the process of
colonisation is either too recent or not sufficiently known to
warrant definite conclusions being drawn. The next few years,
however, should teach much with regard to the outcome of these
experiments.

Canada. The enormous damage entailed by the European
larch saw-fly {Lygceonematus erichsoni Htg.) to Canadian forests
led C. G. Hewitt to seek to relieve the situation by parasite
introduction. In 1912-13 importations of cocoons of the saw-fly,
collected in the larch woods of England, were received in Canada.
The distribution of these cocoons in a heavily infested larch area
in South Manitoba led to the establishment of the Ichneumon
})arasite Mesoleius tenthredinis Motl. Collections of saw-fly cocoons
made in 1916 revealed that 19 per cent, contained the parasite,
and in succeeding years the proportion steadily increased from 40
per cent, in 1919, to 66 per cent, in 1920, and again in 1926 ; while
in 1927 it averaged at least 75 per cent., rising as high as 88 per
cent, in some localities (Criddle, 1928). In the course of these
years the parasite has withstood the rigours of the climate and
spread across a patch of open prairie not less than six miles in
extent to a larch swamp twenty miles across, penetrating to its
utmost limits. The parasite has become well established and.
according to Swaine, it has been recovered as far as 200 miles from

France, and is proving an efficient parasite, destroying its host at all seasons
of the year. It has also been introduced from New Zealand into parts of
Australia, and re-introduced into South Africa ; in New South Wales it has
become as completely successful as in New Zealand, but in South Africa it has
so far proved less efficient. Introductions have also been made into France,
Germany, Holland, Spain, and some other countries. Although established,
it has not, up to the present, given evidence of exercising appreciable control
in Germany, or Holland, while in France it appears to be rather more promising
in the south than elsewhere.

FIJI 387

the original place of liberation. While there is no doubt that the
Mesoleius has played an important part in the decline of the larch
saw-fly it is possible that certain indigenous parasites have also
participated in the process.

The introduction from the Imperial Institute of Entomology
Laboratory at Farnham Royal, England, of the Chalcid
Blastoihrix sericea into British Columbia has proved strikingly
successful. According to Glendenning (1933), the trouble arose
in connection with the scale insect Eulecanium coryli, which had
become one of the most destructive insects in the coastal region
of British Columbia. The city of Vancouver constantly had to
adopt oil-spraying in order to preserve its shade trees. Greater
Vancouver, it appears, is now freed from an objectionable pest.

Several of the biological control projects in Canada are organised
in collaboration with Federal or State entomologists of the United
States. This arrangement obtains in connection with the gypsy
and brown-tail moths, European corn-borer. Oriental fruit moth
and other pests.

Fiji. The Zyggenid moth Artona catoxantha Hamps. has been
known as a coconut pest in Malaysia for many years, but it was
not until 1914 that it was discovered to be affected by parasites.
Subsequent observations showed that the Tachinid fly Ptychomyia
remota Aid. was an important enemy of the moth in question. Out
in Fiji, another Zygaenid, Levuana iridescens Beth. Baker, was
causing serious depredation in coconut plantations, and it led the
Government of those islands to explore the possibilities of biological
control. Several entomologists have been concerned in this work,
and the ultimate outcome has been the successful colonisation of
the Tachinid in Fiji. In the Malay States its efficacy is largely
restrained by the activities of hyperparasites, but, notwithstanding
this drawback, it is able to parasitise its host to the extent of 30
to 40 per cent, or more. In Fiji the Tachinid bore out previous
expectations in parasitising the Levuana as readily as it affects
the Artona. The fly was reported to be definitely established in
1925, when 75 per cent, parasitism of the Levuana caterpillars was
being commonly attained, and even 90 per cent, was occasionally
reached. The parasite has now spread to all infested parts of Fiji,

13—2

388 BIOLOGICAL CONTROL

and the control attained appears to be of a permanent nature.
The whole history and the biology of the host and parasite are
described in an elaborately produced book by Tothill, Taylor and
Paine (1930).

The success achieved by the introduction of Ptychomyia led to
the further application of biological methods of control in the
Islands. The scale insect Aspidiotus destructor is another severe
enemy of coconut besides attacking bananas, avocardo and other
plants. It is not known how, or exactly when, this insect
became introduced into Fiji, but several attempts have been made
to control it by parasite introductions without success. It was
not until 1928 that certain predators of the family Coccinellidae
were introduced from Trinidad, where they had been observed to
exercise notable control over the same host scale in that island.
The Coccinellid Cryptognatha nodiceps soon proved itself superior
to the other species that were introduced along with it. Only nine
months after liberations were made in Fiji the scale insect was
brought under control in all the more important islands of the
group. In his full account of this campaign Taylor (1935) states
that the scale has become reduced to negligible proportions and
an apparently permanent check has been maintained upon it.

The complete success of the introduction of C. nodiceps is
attributable to a combination of factors. First, it breeds
continuously with a high rate of multiplication throughout the
year in Fiji. It is a voracious predator, both as larva and adult,
and in the latter phase it is long-lived. Secondly, it has remarkable
powers of dispersal. Thirdly, it has no serious enemies in Fiji.
Finally, the Cryptognatha is able to survive even when the
Aspidiotus has become reduced to a condition of great scarcity.
In this connection, the fact that it has an alternative host in
another scale insect, namely, Diaspis pentagona, which is not a
serious pest in Fiji, is important.

An even more recent success is that achieved in the repression
of the leaf-mining Hispid beetle of the coconut, Promecotheca
reichei, an account of which has been published in detail in a
special volume by Taylor (1937). The introduction of the Chalcid
Pleurotropis parvulus from Java in 1933 led to the rapid and

ENGLAND 389

effective suppression of this pest. There is adequate reason for
beheving that the good results will be maintained after the
manner that prevails in connection with the two previously men-
tioned examples. The details of the campaign and the factors
involved are discussed at length in Taylor's well-illustrated
monograph, to which the reader is referred. It is thus evident
that the results achieved in Fiji rank among the most strikingly
successful examples of biological control under restricted or
islandic conditions.

Experiments in England. (1) Encarsia formosa. Several
species of the Chalcid genus Encarsia are known to parasitise
Aleurodidse or " White-flies " in certain parts of the world,
including North America, the Orient, Africa and Barbados. The
first occurrence of a member of the genus in England was in 1914,
when a species, probably E. partenopea Masi, was noted by Fox
Wilson in a tomato house at Wisley (Surrey). In 1926 E. formosa
Gahan was found in a small greenhouse at Elstree (Herts.), and
an account of this insect is given by Speyer (1927, 1928). Its
country of origin is uncertain, but it appears to be a tropical
species, possibly imported on plants from India. It parasitises
the common white-fly Trialeurodes vaporariorum (Westw.),
attacking and destroying the pupae. Under suitable conditions
of temperature a number of generations can be bred under glass
each year, and stocks of the insect distributed among tomato
and cucumber houses have resulted in clearance of the white-fly
to an extent which rendered fumigation unnecessary. Owing to
its intolerance of cold, it is desirable to maintain the parasite in a
specially heated house over winter, so that large numbers are
available for distribution the following spring. The maintenance
and distribution of this parasite have been undertaken by the
Experimental and Research Station at Cheshunt (Herts.), and
the outlook for its utilisation as^ permanent measure appears to
be promising. (2) Aphelinus mali. The encouraging results
derived from the introduction of Aphelinus mali into New Zealand
and to a lesser degree in some other countries led J. C. F. Fryer, of
the Ministry of Agriculture, to explore its possibilities of controlling
the woolly aphis under English conditions. In 1923 colonies

390 BIOLOGICAL CONTROL

derived from stock originally of American descent were received
from France. They multiplied with great rapidity in captivity,
and in 1924 liberations were made in the open in a number
of widely separated localities. Notwithstanding the prevailing
inclement weather, coupled with an abnormal scarcity of the host,
the Aphelinus survived in certain orchards until the following
year. At the present time it appears to be maintaining itself in a
few places in small numbers, but without spreading or increasing
to any appreciable extent (vide also Thompson, 1934).

CHAPTER XV
BIOLOGICAL CONTROL— continued

II. Factors Governing Parasite Introductions, p. 391. Geo-
graphical Location, p. 392 ; Multiparasitism, p. 396 ; Hyperparasitism,
p. 398 ; The Sequence Theory of Parasitic Control, p. 401 ; Other Factors,
p. 402. III. The Utilisation of Indigenous Parasites, p. 405.
— Conservation of Parasites, p. 406 ; Direct Increase of a Parasite
Population, p. 407 ; Transference of Parasites to New Areas, p. 409.
Biological Control of Noxious Weeds. Principles Involved,
p. 410 ; Control of Lantana, p. 412 ; Control of Prickly Pear, p. 413 ;
Control of Prickly Pear in Other Lands, p. 416 ; Control of Clidemia,
p. 418 ; Problems in New Zealand, p. 418. Literature, p. 419.

II. Factors Governing Parasite Introductions

In the foregoing section a number of the more important
parasite introductions are discussed. They are by no means
exhaustive, and many other experiments have been undertaken
in various parts of the world. Certain of these have yielded
beneficial results, but in a much larger number of cases no
successful outcome has been recorded ; in other instances the
introduced species failed to establish themselves or the experiments
were doomed to failure from the outset owing to the haphazard
nature, lack of requisite knowledge and other causes.

In judging success or failure it is, of course, important to bear
in mind that the primary objective is to bring about an economic
reduction in the abundance of the pest concerned. If, for example,
introductions are attempted with ten different species of parasites
and only two of these succeed in estabhshing themselves, and these
are effective in restraining the pest in question, the experiment
has been successful. It would be wide of the mark to assert
there had been eight failures, since the objective was pest control,
and once this is achieved it matters little as to what becomes
of the ineffective parasites. Also it needs to be pointed out that
biological control aims at bringing about a permanent measure

391

392 BIOLOGICAL CONTROL

of pest repression ; in the most successful cases it may result in
artificial methods being no longer necessary, while in others
biological control may serve to supplement insecticidal or other
treatment. It is only in rare instances and in the most favourable
environment that almost complete suppression of a pest results,
and in most cases a successful outcome is regarded as having been
achieved once an appreciable degree of permanent control over a
given pest has been accomplished. Thus if an injurious insect A,
feeding within the stems of a cereal crop and exempt from the
effects of any practicable artificial method of control, is affected
by biological measures to an extent of reducing the damage which
it occasions by an average of 20 per cent., taken over a period of
years, few would assert that the experiment was not justified. Or
again, if a scale insect B, infesting fruit trees to an extent which
demands spray treatment several times annually, is reduced by
biological control so that spraying becomes necessary, on average,
only once every three years, immense benefit will have resulted
to growers in the saving of expenditure incurred.

It is desirable at this point to discuss the more important
factors and principles which have a definite bearing upon the
success or failure of parasite introductions.

Geographical Location. Up to the present time some of the
most spectacular and complete examples of the application of
biological control have resulted from parasite introductions
made under insular conditions. The successes achieved in the
Hawaiian Islands, New Zealand and Fiji are examples in question.
In continental areas the most encouraging results have been
achieved where parasite introductions have been made, (a) with
respect to certain tracts which are, ecologically, separate islands,
e.g., California ; and (h) with respect to the control of pests
affecting a localised crop, e.g., citrus fruits, mulberry, etc. In
the case of widespread pests affecting crops or trees whose range
of distribution extends over a vast continental area, involving
varied climatic and other conditions, no indubitable success has
so far been achieved.

(a) Insular Areas. The most favourable type of physical and
biological environment afforded by insular conditions is

PARASITE INTRODUCTIONS 393

exemplified in the Hawaiian group. These conditions may be
summarised as follows, and they probably hold good for most
oceanic islands :

(1) Warm and equable climate allowing parasites to multiply
unchecked by seasonal factors.

(2) An indigenous fauna of a peculiar and restricted character,
evolved as the result of long isolation. Introduced parasites meet
with relatively little competition from indigenous forms, and the
parasitic element is but poorly developed.

(3) The area to be covered by parasite colonisation is
circumscribed and, as there are only few main crops, adequate
organisation and centralisation of control experiments are greatly
facilitated.

In most islands of continental origin parasite introduction has
to contend with a different type of environment. In Trinidad,
for example, conditions (1) and (3) hold good, but (2) is entirely
different. Owing to the geological formation of the island, and
its close proximity to the mainland of South America, the fauna
is essentially that of the neighbouring continent. Introduced
parasites consequently have to contend with a high degree of
environmental resistance of a biological character. Among a
fauna already rich in the parasite element new introductions, if
they are to be effective from the economic standpoint, require to
be species capable of forcing their way into and exerting influence
in already complexly adjusted biological associations. On the
other hand, in New Zealand, although (1) is far less ideal than in
Hawaii, the climatic conditions are more favourable than those
from which the introduced pests have come, and from where
parasite introductions have been made. Condition (2) has many
characteristics of an insular fauna, and although rich in many
dominant groups of insects, other groups are wanting or but poorly
represented in New Zealand, and the indigenous parasite and
hyperparasite element is not highly developed.

In New Zealand there are, for example, according to Tillyard, no
indigenous Thysanoptera, only thirteen species of Odonata, and the
same number of Neuroptera ; Hemiptera number about 300 species
(Australia 1,970), and Hymenoptera include 316 species (Australia

394 BIOLOGICAL CONTROL

6,250). It is further noteworthy that among the Hymenoptera there
are only two species of saw-flies, and less than 250 species of the parasitic
groups. The paucity of the latter is well emphasised by comparison
with Great Britain, where there are about 1,500 species of Chalcidoidea
alone.

In Great Britain none of the three groups of conditions
previously mentioned appears to favour parasite introduction. A
variable, and on the whole, cool humid climate ; a complex and
highly adjusted continental insect fauna ; and a diversity of
crops and cultural methods all appear to militate against the
possibility of any striking economic results being achieved through
the application of biological control. An environmental resistance
of this character has stood the country in good stead, since it is
almost free from introduced insects affecting agriculture. It is
evident, therefore, that parasite introductions would naturally
have to be framed to cope with one or other of the major indigenous
pests. In the future it may be found feasible to explore the
possibilities of north temperate countries with this object in view.

(b) Localised Continental Areas. Biological control in con-
tinental areas has proved successful in certain cases where the
crops involved occupy more or less circumscribed tracts of country
and require climatic conditions of a warm and equable nature.
This applies especially to mulberry cultivation in Italy and to the
cultivation of citrus fruits in various parts of the world. These
same conditions, it would appear, favour parasite introductions,
and a partial explanation may be found therein of the success
attending the importation of Prospaltella berlesei and of Vedalia
cardinalis into many countries. The effective control exercised
by Aphelinus mali has so far only been evident under continental
conditions in areas of a more or less localised character, areas
where the climate is favourable and apple cultivation is prosecuted
on a large scale. ^

The successes achieved in California are to be ascribed to a
combination of special conditions. Bounded partly by the
Pacific Ocean and partly by barriers of mountain and desert, the

^ As mentioned previously, other contributing factors appear to be the
rapid multiplication of the parasites in question, and the fact that they attack
hosts which are largely sedentary in behaviour.

PARASITE INTRODUCTIONS 395

State is physiographically isolated to an exceptional degree.
Much of the country was originally desert, and has been brought
under cultivation through the development of irrigation, which
has established sub-tropical conditions in its southern territory.
The insect fauna in such parts is consequently largely introduced
and built up of immigrants from adjoining territory with certain
accidental importations from abroad. Compared with the rest
of the continent, the climate is wonderfully equable in the long
tract of country between the mountains and the sea. The area
where parasite introductions have been made covers the zone of
citrus cultivation in the south, which is exceptionally uniform
ecologically and circumscribed by physical conditions.

Conditions apparently analogous to those found in California
occur in other parts of the world. As Tillyard has observed,
Australia is made up of a number of diverse areas separated from
each other by mountain barriers or great stretches of desert.
Looked at from this point of view. Western Australia, the elevated
apple lands of South Queensland and other parts, should offer
favourable conditions for the application of biological control.
Egypt, limited as it is by desert and sea, Mesopotamia, Palestine
and other lands would also appear to afford ideal conditions for
parasite introductions, provided judicious selection, based upon
sound preliminary investigation, be exercised (vide also Myers,
1928).

(c) Wide Continental Areas. Biological control applied to
imported pests menacing crops widely distributed over vast
continental areas has to contend with manifold difficulties.
Competent investigators have expressed serious doubts whether
the method is likely ever to prove really efficient under conditions
of this character. The United States suffers more than any other
region of the world from the depredations of pests now disseminated
over immense tracts of territory. In the case of the gipsy and
brown-tail moths parasite introductions have been going on for
over a quarter of a century. More recently attention has also
been directed to the alfalfa weevil, the European corn-borer and
the Mexican bean beetle. The corn-borer, for example, now
infests well over 300,000 square miles of territory, while the areas

396 BIOLOGICAL CONTROL

covered by the alfalfa weevil and Mexican bean beetle are already
of great extent.

The problem of parasite introduction in continental areas is
discussed by Thompson (1928), who points out that the effects
which follow the establishment of foreign pests in such areas
constitute a clear demonstration of their escape from control.
The evidence indicates that the factor whose absence has allowed
of increase and spread is the parasitic factor. In any attempt
to establish equilibrium over so vast an area, he is of opinion
that we have to rely upon biological methods if permanent and
inexpensive control is to be hoped for. Under such conditions no
one parasite or predator is likely to effect appreciable control.
With the European corn-borer the parasitic enemies do not exist
all together in any individual zone yet discovered in Europe.
The composition of the parasitic fauna differs in every region,
and in view of the differences of climate obtaining over the great
area of North America affected by this insect, it is most probable
that only a limited quota of the introduced parasites will colonise
a given zone. In order to establish the best results, species of
parasites failing to become colonised should be reintroduced as
the corn-borer reaches new areas differing climatically from those
now infested. (Thompson and Parker, 1928.) In other words,
re-establishment of natural equilibrium appears to stand the best
chance of being accomplished on an ecological basis.

Multiparasitism. The factor of a multiparasitism is discussed on
p. 322, and a well-known example is afforded by the parasites of the
Mediterranean fruit-fly (Ceratitis capitata Wied.) in the Hawaiian
Islands (p. 374). When first introduced, the Braconid Opius
humilis Silv. was the most efficient of the imported enemies of
this insect. In 1915 it parasitised 31-5 per cent, of all the Ceratitis
larvae developing in fruits about Honolulu ; since then its efficiency
has declined, and in 1923 it was as low as 4-1 per cent. In 1924
and 1925 its effectiveness increased, when it parasitised 14-5 per
cent, of the larvae each year. Among coffee cherries in Kona the
Opius parasitised from 59 to 97 per cent, of the fruit-fly in 1915,
but its effectiveness gradually declined, and in 1926 not a single
example was reared from the larvae from 18,955 coffee cherries

MULTIPARASITISM

397

(Willard, 1926). Its decline has been traced to the fact that
0. humilis frequently occurs within the same individual host
larva, which is also harbouring either of the other Braconids,
Diachasma tryoni Cam, and D. fullawayi Silv. When this
happens the multiparasitism invariably results in the death of
the Opius.

The view has been expressed that had the biology of these
parasites been fully studied prior to their liberation, their mutual
relations would have been disclosed and the desirability of liberat-
ing 0. humilis only, in the first instance, might have been apparent.
It is possible, therefore, that this single species of parasite might

Fig. 94. Diagram representing multiparasitism, showing percentages
of hosts attacked and destroyed by Diachasma and Opius and
percentage of overlaps, which in most cases produced Diachasma.
From data of Pemberton and Willard. (After H. S. Smith,
Bull. Ent. Res., XX., 1929.)

have proved more efficacious than the combined activities of the
four species already liberated.

H. S. Smith (1929) has critically analysed this problem and
concludes that it is far from proven that the effects of the multiple
parasitism would have been less beneficial than the activities of
the Opius alone. He stresses the superior " balancing " effect of
several species of parasites over a single one in that they are not
all likely to suffer to the same degree from deleterious factors.
Several species would tend to stabilise a parasite population,
whereas a single species would tend to fluctuate in efficiency
annually. In Fig. 94 one phase of the problem is represented
graphically, and Smith claims that even if the 17 per cent, of the

398 BIOLOGICAL CONTROL

puparia affected by multiparasitism were attacked by the Opius
alone, it is very uncertain whether it would have raised the
efficiency of the latter to more than 65 per cent.

Hyperparasitism. The incidence of secondary parasites attack-
ing a primary parasite of an injurious insect has a very definite
bearing upon biological control. There is direct evidence that
the efficiency of certain introduced parasites has been enormously
reduced or even nullified by the activities of secondary parasites.
The perplexing and involved problems centred around the inter-
relations of hyperparasites and their different hosts have already
been alluded to (p. 327), but in illustration of their practical
bearing certain concrete examples may be quoted. A primary
parasite of one species of insect may function as a secondary
parasite of another, as is well exemplified by the Chalcid Mono-
dontomerus cereus Walk., which was introduced into the United
States in connection with the control of the gipsy and brown-tail
moths. Although a primary parasite of the pupge, and especially
those of the brown-tail moth, it has also proved to be a secondary
parasite. Among imported primary parasites which this species
utilises as its hosts are the Braconid Apanteles lacteicolor and the
Tachinid Zygobothria nidicola. This dual behaviour has rendered
it extremely difficult to determine whether its occurrence as a
secondary parasite outbalances its value as a primary agent in
control. Another example is afforded by Schedius kuvance How.,
a valuable egg-parasite of the gipsy moth, introduced from Japan.
Later observations have shown that this species also behaves as
a secondary parasite, since it occasionally attacks the Braconid
Apanteles melanoscelus Ratz. The latter species, it may be added,
is a native of Europe, and was also introduced for controlling the
gipsy moth.

A striking example of the importance of hyperparasitism is to
be found in California, where the Chalcid Metaphycus lounshuryi
was introduced from South Africa for purposes of controlling the
black scale of citrus fruits (p. 378). This introduction gave
promise at first of an almost complete control of the scale insect
in the coastal areas. Gradually, however, its efficiency has
become checked by a secondary parasite, Quaylea whittieri (Gir.),

HYPERPARA8ITI8M 399

which had been introduced at an earlier date and without adequate
knowledge of its host-relationships. This same hyperparasite
also attacks Scutellista cyanea, and there seems little doubt that
its activities constitute a most serious inhibitory factor in the
achievement of adequate biological control over the scale insect
in question. Although due care may be exercised in the exclusion
of hyperparasites when parasite introductions are made, it is
obviously impossible to guard against the contingency of sub-
sequent attacks by indigenous secondary parasites, as has
happened with certain of the parasites of the gipsy moth after
their introduction into the United States.

In the Hawaiian Islands the earlier attempts to control the
sugar cane leafhopper involved the introduction of various species
of Dryinidae. These introductions proved unsuccessful partly
owing to the prevalence of hyperparasites. Some of the latter
were probably native species which, owing to their minute size,
had previously remained undetected, while others were chance
immigrants brought in from the United States. It is noteworthy
that in England the common earwig Forficula auricularia (L.) is
feebly parasitised by the Tachinid Digonochceta setipennis ;
observations conducted by A. M. Altson on earwigs collected
from various localities failed to reveal more than a 7 per cent,
parasitism. At the same time the Digonochceta is subject to attack
by the Chalcid Dibrachys cavus {b ouche anus Hatz.) and by the
Ichneumon Phygadeuon scaposus Thoms. In attempts to establish
this Tachinid in New Zealand in the expectation that in a new
and favourable environment it may prove a more efficient element
in earwig control, it is important that these hyperparasites be
excluded, since they are unknown in that country. On the other
hand, it is problematical whether the exclusion of the Dibrachys
from consignments of the Tachinid which have been introduced
for earwig control into the Pacific side of North America will result
in any advantage for the reason that the hyperparasite is already
widely distributed over that continent.

Proper (1934) is one of the most recent writers who discusses
the subject of hyperparasitism from a considerable body of
evidence. In dealing with the secondary parasites of certain

400 BIOLOGICAL CONTROL

Lepidopterous larvae he arrives at the conclusion that about
one-third ^ of the cocoons and puparia of the primary parasites
are destroyed in this way. Since the Lepidopterous hosts are
notable defoliators, etc., and their primary parasites are important
factors in their control, the activities of the hyperparasites function
to a great disadvantage. His observations support the conclusion
that one of the chief causes of the reduction in numbers of the
parasites imported into New England to aid in controlling the
gypsy moth, the brown-tail moth and other Lepidoptera has been
the activities of hyperparasites.

Taylor (1937) is of opinion that the importance of secondary
parasites has been greatly exaggerated in some cases. He remarks,
" Many secondaries are wholly incapable of hindering extensively
the primaries which they attack, mainly because the percentage of
primary individuals which they succeed in finding increases
directly as the concentration of the primaries and the percentage
cannot approach 100 until the primaries are already sufficiently
plentiful to suppress the pest. In other cases, inferior egg-
capacity, or other factors of this nature, may ensure economic
negligibility in the secondaries. For these reasons it is unwise to
refrain from introducing a parasite which seems otherwise
desirable, and to search for another, merely because the former
is certain to be attacked by secondaries which are already present.
The elimination of the secondaries from the consignment of
primaries introduced is an obviously desirable and usually simple
precaution, but in many cases the introduction of the former
would not be harmful to an appreciable extent."

The available facts lead to the conclusion that hyperparasitism
is most likely to supervene where parasite introductions are made
into continental areas, especially where a sequence involving
numerous species is aimed at. Increasing knowledge may to
some extent reduce the possibility of species liable to betray this
habit being introduced. On the other hand, in view of the
general behaviour and widespread activities of hyperparasites, a

^ In the case of some of the primary parasites, the average mortality, over
a four-year period, was only from 5 to Hi per cent., while in others it ranged
from 44 to 84 per cent.

THE SEQUENCE THEORY 401

greater or lesser degree of influence from indigenous species will
usually have to be contended with in such projects.

In insular areas, if due precautions be taken, the problem of
hyperparasitism is frequently capable of circumvention, and this
appears to be an important contributory factor in the successes
attained with parasite introductions under such conditions.

The Sequence Theory of Parasitic Control. The sequence theory
of the application of parasitic control is due to Fiske, and its
practical importance has been stressed by Howard in various
publications (vide Howard, 1924). The theory applies especially
to the control of Lepidoptera, in which the component stages of
the life-cycle differ in activity, habit, seasonal appearance, and
other factors. In a state of nature such species are almost
invariably attacked by a succession of different parasites affecting
the eggs, larvae and pupae respectively. When, therefore, an
insect like the gipsy moth, for example, has become accidentally
established in another country the theory postulates the necessity
of reproducing as nearly as may be possible its original parasite
association. It is claimed that there exists a delicately adjusted
condition of equilibrium between the host and each of its parasites.
This involves a limitation of the activities of the latter by environ-
mental and other factors which preclude the percentage of the.
hosts it destroys from rising above a certain mean figure. The
theory maintains for this reason that a single species of parasite
is incapable of destroying as many of a host population as is
destroyed by a succession of different enemies. It is claimed
that it is only when a certain percentage of the eggs have been
destroyed that parasites of the larvae will be able to reduce a
given pest to a sufficient degree that the final pupal parasites will
bring the species measurably under control. According to Fiske,
the gipsy moth has a six-fold rate of increase per generation, and
consequently a total parasitism of 88-33 per cent, is necessary in
order to attain effective control.

The theory is ably criticised by Thompson (1923), to whose
paper the reader is referred for the detailed arguments. He points
out that insufficient evidence exists to prove that a sequence of
parasites is essential to the biological control of a given pest ;

402 BIOLOGICAL CONTROL

or that, if the sequence be incomplete, the pest would not be held
in check. According to Thompson's view, control depends
primarily upon the ratio between the respective rates of repro-
duction of host and parasite. If the rate of reproduction of
one species of parasite is equal to, or greater than, the rate of
reproduction of the host, this parasite could, in many cases,
increase to the point when complete control of a host is achieved
by its unaided activities. The time taken for the attainments of
effective control of the host by the parasite will depend, not only
upon the reproductive ratios just mentioned, but also upon the
ratio between the number of individuals of a parasite introduced
and the number of individuals of the host at the time the intro-
duction is made. Thompson concludes that, although there are
cases where the percentage of parasitism of a given species may
never rise above a certain maximum, instances must be rare in
Nature where all the parasites of a host are thus limited in their
activities, and it will usually not be found true of most members
of a parasite sequence. According to him the theory is invalid as
a general theory of parasitic action, since it can only apply to a
limited number of special cases. On the other hand, Thompson
supports the general idea of the introduction of a sequence of
parasites for a reason entirely different from the one implied
by the theory. It is simply because by introducing a number
of species of parasites the probability is greatly increased of
selecting one whose reproductive potentialities are equal to, or
greater than, those of the host concerned and which, for this
reason, offers possibilities of exercising the desired degree of
control.

Other Factors. Among other factors which may have a direct
bearing upon parasite introductions, the following will be briefly
discussed : —

1. The subject of host-parasite relationship is discussed in
Chapter XIII, and little can be usefully added here. The bulk
of the information relative to this factor is of an empirical nature.
No means exist of predicting the results of parasite introductions
for the reason that it is not possible to forecast with any certainty
how a particular parasite will behave towards its possible hosts

VARIOUS FACTORS 403

in a new area or country. Investigations conducted in the
country of origin of a parasite are of prime importance, but actual
experiment alone can provide the solution of its behaviour in a
fresh environment.

The introduction of a multivoltine parasite to attack a univoltine
host may be selected as an example of a particular type of host-
relationship. In an eventuality of this kind success may largely
depend upon the availability of other suitable hosts at whose
expense the parasite will be able to maintain itself during the
period when the requisite stage of the intended host is no longer
in evidence. A comparison of possible hosts indigenous to the
country of introduction with the known hosts in the country of
origin of the parasite will obviously provide tentative information,
but only experiment will determine the extent to which any
forecast of this kind will be realised.

2. The effects of artificial control measures upon the progress
of introduced parasites can only be studied in a broad general
way when reduced to mathematical terms. No data exist which
permit of this problem being considered from its experimental
aspect. In the theoretical analyses given by Thompson the
initial numbers of reproducing parasites and hosts present at the
beginning of a given experiment, the effective rates of reproduction,
and sex ratios of the two species and the proportions of each
eliminated by artificial control before the period of reproduction
are taken into consideration. It is shown that if artificial con-
trolling measures eliminate at least as many hosts as parasites,
precautions for conserving the latter are unnecessary. If, on the
other hand, conditions are such that artificial measures of control
result in the destruction of appreciably more parasites than hosts,
they may exercise a disastrous effect upon the progress of an
introduced parasite and delay its gaining effective control for
many years. For a full discussion of the theoretical aspects of this
subject the reader is referred to Thompson's original paper
(1927a), and also to a recent paper by Clausen (1936).

3. What may be termed the time factor is of particular import-
ance, and lack of its due recognition may lead to erroneous
conclusions respecting the supposed failure of parasite activities

404 BIOLOGICAL CONTROL

in certain instances. For example, Howard (1924) mentions the
case of the Hessian fly parasite, Pleurotropis epigonus, which was
introduced into the United States from England in 1894. After
liberation in the field this species was not recovered until twenty-
two years later, and since then it has become one of the dominant
parasites of its host over a large area in the eastern States.
Thompson (1927) has discussed the rate of progress of introduced
parasites from the mathematical standpoint and, although his
conclusions are necessarily theoretical, they serve to show the
bearing of the relationship between the ratios of the rates of
reproduction of parasite and host, together with the initial
numbers of parasite and host, and the time involved before a
given parasite could exercise a dominant influence. Thus, when
the rates of reproduction of parasite and host are equal and the
host population is very large in relation to that of the parasite, the
advance of the latter will be so slow as to be barely recognisable.
Periodical field investigations may reveal the presence of the
parasite, but for many years there will be no evidence of increase
in its numbers. In cases where a parasite reproduces less rapidly
than its host it will never subjugate the latter unless its initial
population is much larger in relation to that of the host than is
ever the case in any practical undertaking. In certain cases where
a multivoltine parasite with a high rate of fecundity is able to
attack a univoltine host, or one with a much lower reproductive
rate, control has been achieved under favourable conditions in a
remarkably short interval of time. This is exemplified in the case
of the control of the cottony-cushion scale in various parts of the
world, of the mulberry scale in Italy, and of the woolly apple
aphis in New Zealand.

4. Finally, the capacity of a parasite to discover its host is
a factor often overlooked and one difficult to evaluate except in
a very general way. This capacity appears to be most often
inversely proportional to its fecundity, and it affords examples
of Nature's prodigality or economy as the case may be. When
the chances of a parasite discovering its host are small, the
progeny are numerous {e.g. Stylops, leaf-ovipositing Tachinidae,
etc.), and where the chances are great {e.g. Scoliid wasps) the

INDIGENOUS PARASITES 405

progeny are relatively few in number. The efficiency of Scolia
manilice in Hawaii is largely attributed by Muir to its highly
developed faculty for discovering its host. Furthermore, a
fragile and minute parasite with restricted powers of locomotion
may prove highly effective when hosts are abundant, but play
an insignificant part in control when they are scarce. The converse
would appear to be the case with certain of the more robust
parasites of low fecundity.

III. The Utilisation of Indigenous Parasites

The relationship between a given indigenous insect and its
parasites is one involving a complex condition of ecological
equilibrium which, moreover, is subject to seasonal fluctuation.
The component factors, together with their individual and
collective effects, which govern these fluctuations, have not so far
been subjected to other than the most elementary analysis, and
we are hardly in a position to assert whether adequate analysis is
at all possible. In the absence of fundamental information of this
kind the feasibility of the utilisation of indigenous parasites can
only be answered by actual experiment. It needs to be recollected
that the subject involves somewhat different principles from
those concerned with parasite introductions. In the latter case
the building up of a condition of natural equilibrium is aimed at,
whereas the utilisation of indigenous parasites is largely concerned
with efforts to modify a condition of equilibrium usually already
highly adjusted. In order to secure practical results it is necessary
to maintain a permanently readjusted equilibrium against
tendencies which are constantly operating in the opposite
direction.

Suggestions have been frequently made and, in certain cases,
actual attempts carried out, with a view to making use of
indigenous parasites as auxiliary agents in pest control. Such
operations consist either of conserving or increasing the numbers
of a parasite or predator in a given area with the idea of obtaining
a higher degree of control over some individual species of pests,
or of attempting to colonise such enemies in an area of a country
where they did not previously exist.

406 BIOLOGICAL CONTROL

Conservation of Parasites. The principle involved in parasite
conservation consists of either altering the host-parasite ratio
by the adoption of methods which allow of more hosts than
parasites being destroyed ; or, the abandonment of measures
which tend to reduce the existing parasite population. The
earliest application of any method directly aimed at conserving
indigenous parasites appears to have been made by F. Decaux
in 1880 with regard to the insect enemies of the apple blossom
weevil (Anthonomus pomorum) in Picardy. By collecting infested
or " capped " blossoms from about 800 trees and placing them in
gauze-covered boxes, he claimed to have evolved a procedure far
superior to destruction by burning. All that was necessary
afterwards, he stated, was to raise the gauze from time to time
to allow of the parasites escaping. This process was repeated
for a second year, and since the orchards were isolated in the
middle of cultivated land all serious damage from the weevil
was stated to have been held in abeyance for ten years. A
similar procedure was advocated by Berlese in 1902, with
reference to the conservation of the parasites of the grape-vine
Cochylis.

With regard to the Hessian fly, Marchal (1907) pointed out
that the destruction of wheat stubble, if carried out a little late
after the harvest, results in the annihilation of great numbers of
parasites, since the Hessian flies emerge before their parasites.
Similarly, Kieffer has shown that one of the measures for con-
trolling the wheat midge, viz., burning the debris after threshing,
has only an injurious effect, for the reason that the healthy pupae
occur in the soil, while those found in the debris are parasitised.
In Louisiana it has been pointed out by T. E. Holloway and
others (1928) that the practice of burning the trash or debris left
in the fields after the cutting of sugar cane is not an effective
method of decreasing infestation by the moth-borer Diatrcea
saccharalis Fab. The trash, it appears, is the hibernating place
for large numbers of parasitic insects. Following this contention
the collected trash was left unburned at the sugar experiment
station near New Orleans for a period of years, while the State
planters, as a whole, burned their trash. A comparison of losses

INDIGENOUS PARASITES

407

at the experiment station with those represented by the average
for the State as a whole is given below.

Percentage of Loss due to Biatrce.

1915

1916

1917

1918

1919

1921

Trash unburned
Trash burned (State
average)

12
24

13
30

7
17-5

19
21

6
9

18
25

With increasing knowledge of host and parasite behaviour,
instances where analogous methods may appear to be applicable
will undoubtedly be multiplied. On the other hand, it is difficult
to prove that any observed effect is due to the method in question.
Also, such methods are not likely to prove effective unless they
be carried out by general agreement among cultivators over a
wide area ; under more restricted circumstances only a fraction
of the insect population would be affected, and no appreciable
result is likely to supervene.

Direct Increase of a Parasite Population. It has been contended
on theoretical grounds that the alteration of the host-parasite
ratio in specific cases by the artificial quantitative breeding of
suitable parasites is an evident possibility. A species of parasite
whose behaviour appears to lend itself favourably to the application
of the idea is the Chalcid Trichogramma minutum Riley, one of
the best known of all egg-parasites. It has a wide geographical
range, while its hosts include over 50 species of Lepidoptera,
including such notable pests as the sugar cane borer {Diatrcca
saccharalis Fab.), the European corn-borer (Pyrausta nubilalis
Hub.), and the codling moth (Cydia pomonella L.). Early in
the season this parasite is relatively scarce, while its spread is
slow owing to its limited power of distribution, but it is maintained
that if intensive infestation can be induced at the right time
potential outbreaks of certain pests might be largely counteracted.
At the present day attention is being centred on this parasite in
many parts of the world, and it promises to be the first species of

408 BIOLOGICAL CONTROL

its kind likely to be propagated much along the same lines as the
introduced predator Cryptolamus (p. 377). The largest scale
experiments are being conducted in Louisiana and California, with
reference to the control of the cane-borer and codling moth
respectively. According to S. E. Flanders the Trichogramma can
complete its life-cycle in eight days at a temperature of about
83° F. It is readily bred in captivity upon the eggs of the
Angoumois grain moth (Sitotroga cerealella Oliv.), which likewise
is capable of many generations in the year, and the whole technique
for the mass-rearing of the parasite on this host has become a
matter of skilfully standardised routine. With comparatively
simple and inexpensive technique it has been possible to rear
200,000 individual Trichogramma a day at certain periods, while
numbers approaching one million per day are stated to have been
attained subsequently. Whether the reduction of the host
population can be regularly attained to a degree which justifies
the expenditure of money on projects of this kind has given rise
to a good deal of doubt. While encouraging results are reported
by Tucker, in Barbados, on carefully analysed evidence, satis-
factory data are wanting elsewhere. The use of Trichogramma for
the purpose mentioned is, it may be added, inspiring less confidence
than formerly.

Another experiment of this kind concerns the Coccinellid,
Hippodamia convergens, in California. This insect is generally
distributed in the areas involved by the experiment, but is
ineffective in controlling the great abundance of aphides on truck
crops, etc. It has been estimated that an increase of about
30,000 individuals per acre might be expected to give the result
desired. Watch was therefore kept in autumn in the canyons of
the Sierra Nevada Mountains where the Hippodamia congregate
in masses prior to hibernation, and the particular assembling
places marked. Before the snows melted in spring the hibernating
lady-birds were collected in enormous numbers and transferred
to storage plant in Sacramento, where they were retained at a
suitable temperature until required.

On reports being received from farmers concerning aphid
outbreaks, shipments of the Coccinellids were made to the desired

INDIGENOUS PARASITES 409

localities, where they were liberated. Notwithstanding the
considerable mortality which occurred while in artificial storage,
upwards of 75,000,000 have been distributed among farmers in
a single season. The operation was carried on by the State
Horticultural Commission for seven years in succession before the
subject was given expert study as to its value as a commercial
undertaking. In 1918 and 1919 W. M. Davidson, of the United
States Bureau of Entomology, made a series of experiments in
order to ascertain the distribution of the Coccinellids after
liberation. Large numbers were collected and sprayed with
" silver " or " gold " aluminium paint which rendered them
recognisable, and caused only slight mortality. It was found that
the insects distributed themselves over large areas so rapidly that
they were of little value in controlling local aphid outbreaks.
Only one out of every 65,000 sprayed individuals was recovered
(Davidson, 1924).

Transference of Parasites to New Areas. The conception of the
transfer and colonisation of parasites, from an area of a given
country where they are prevalent into a district where they are
apparently non-existent, dates from about 1872. Theoretically,
the method appears feasible if physiographical barriers have
precluded the entry of a parasite into a given area. On the other
hand, if its absence is to be ascribed to climatic causes, the chances
of its successful colonisation would seem to be remote.

Three examples of the application of this method may be
instanced. The incidence of the Braconid Macrocentrus ancylivorus
has already been referred to on p. 383 in connection with the
control of the Oriental fruit moth. It appears that the New Jersc}^
strain of this parasite has a higher efficiency in bringing about host
mortality than examples derived from other localities. The result
has been that this so-called race has been widely disseminated
artificially in many states where it has been subsequently bred in
large numbers followed by liberation. Its colonisation is now an
established fact, and it is regarded as the most effective biological
agent in the control of its host so far discovered.

The colonisation of Aphelinus 7nali, another North American
indigenous parasite, in the Pacific North-West, has, it is stated,

410 BIOLOGICAL CONTROL

resulted in complete control of the woolly aphis (Eriosoma lanigera)
in that region. Owing to some unexplained factor, or factors, the
natural spread of the host in the area mentioned had been accom-
plished without the usual accompaniment of the Aphelinus. In
the third example the experience of Hazelhoff (1929) may be
quoted. He states that the transfer of the Chalcid Encarsia
scutellum, from the older or established cane-fields in Java into
newly planted sugar cane areas, has proved definitely successful
in maintaining the aphid Oregina lanigera under control. The
powers of dispersal of this parasite, it appears, are very restricted,
and consequently its incidence in new cane -fields is greatly
delayed if the insect be left to disseminate itself by its unaided
efforts.

Biological Control of Noxious Weeds

The application of biological control to weeds presents itself
as a possibility where alien plants have invaded and colonised
land to the extent of rendering it useless. It is well known that
in certain parts of the world introduced weeds have resisted all
attempts at cultural, chemical or other methods of control, or
such methods have proved to be economically impracticable.
It is in circumstances of this character that insect control of the
noxious plants concerned has been attempted in several different
countries.

Principles Involved. In any scheme to control noxious weeds
by the introduction of insects living at their expense certain
guiding " principles " require emphasis (vide also Imms, 1929).

1. It has already been pointed out that there is abundant
evidence that alien insects which have entered countries
unaccompanied by their natural parasites can become firmly
established and cause great depredations to cultivated crops.
On the other hand, there appears to be no valid reason why
suitable types of insects deliberately introduced, free from their
natural enemies, should not likewise exercise a destructive effect
upon noxious alien weeds, given an environment favourable to the
colonisation of such insects. It is this principle that is fundamental
to the problem concerned.

NOXIOUS WEED CONTROL 411

2. The second principle involved is that any insects utihsed
must be species known to exercise a controlHng influence over
the weeds concerned in their own habitat. At the same time
they must be kinds which are known to be specific to the host
concerned or confined to a very restricted range of host-plants
and, unlikely in a new environment, to infest important economic
plants.

With certain plant-hosts, such as Opuntia, the possibihty of
insects which feed upon them resorting to other kinds of plants is
remote. Insects feeding upon plants of such unusual characteristics
and endowed with acrid juices, resistant cuticle, along with other
specialised morphological and physiological properties, exhibit so
peculiar and rigid an adaptation to that particular mode of life
that they are unlikely to infest plants with entirely different
physico-chemical attributes. With plants of the natural order
Rosacese, for example, on the other hand, there is always the
looming possibihty that in a new environment insects which
normally feed upon a single genus or species of that order will
infest one or other of its numerous cultivated members.

3. In contingencies where a pest-plant has close allies of
economic importance, insect control requires exceptionally
cautious procedure. In such instances the method is beset with
manifold difficulties, and lack of foresight may only too easily
result in the remedy proving worse than the disease.

If insect control of such types of pest-plants commends itself as
a feasible economic measure, recourse should be rigidly confined,
in the writer's opinion, to the introduction of species of specialised
habits and behaviour. Root-borers, stem-borers and internal
seed- or fruit-feeders are preferable to leaf-feeders, since their
economy is usually more delicately adjusted to their specific
hosts, and for this reason they are probably less liable to seek
other hosts in a new environment. ^

4. Any insects deemed likely to prove valuable in the biological
control of weeds require exhaustive testing with reference to the
possibility of their feeding upon plants other than the particular
host concerned. Such experiments require to be carried out
firstly in the native country of the insects, and, if the tests prove

412 BIOLOGICAL CONTROL

negative, further experiments are necessary under the new environ-
mental conditions wherein it is proposed to colonise them.

5. It will be readily understood that, under the conditions
demanded by noxious weed control, parasites assume the opposite
role to the one they perform in the repression of insect pests. In
other words, parasites require to be rigidly excluded in all cases
where insects are utilised in weed repression. Under a new and
favourable environment the unforeseen entry of such agents might
well result in the activities of a valuable phytophagous insect
being permanently nullified to a degree which, in an extreme case,
might almost amount to totality.

6. It needs to be recollected, as W. R. Thompson has pointed
out, that, although insects entail a vast amount of destruction to
economic plants, they rarely cause sufficiently vital damage which
will affect the survival of wxll-established species to a marked
degree. Measured in economic terms, such damage frequently
assumes vast proportions, but in the biological sense it is not
necessarily of the same importance. Furthermore, the recupera-
tive powers of plants are so great that they are much less liable
to be killed outright as the result of insect attack than insects
themselves, which almost always succumb from the effects of
their parasites.

In the pages which follow an account is given of the principal
examples of the biological control of pest-plants.

Control of Lantana. The first attempt to control a noxious
2)lant by its insect enemies was made in the Hawaiian Islands.
Nearly seventy years ago Dr. Hillebrand, the distinguished
botanist of the Islands, introduced the thorny species of Lantana,
L. camara, for ornamental purposes. This shrub subsequently
became a veritable scourge to the pastures of low-lying regions,
especially on the leeward side of the Islands. It is a native of
the warmer parts of America, and during a visit made by Koebele
to Mexico in 1898 it was observed in its habitat, and some insects
were bred from Lantana seeds. In 1902 Koebele again visited
Mexico, this time with the object of securing any promising
insect enemies of the plant for purposes of introduction into the
Islands. The result of his work was that 23 species of insects

CONTROL OF LANTANA 413

were sent to Perkins in Honolulu, and of these 8 were successfully
established (vide Perkins and Swezey, 1924). Each of these 8
species attacks the Lantana in a somewhat different way. The
most effective appears to be the Tortricid moth Crocidosema
lantana Busck., whose larvae bore into the flower stems and
consume the flowers and fruit. The larvae of the seed-fly Agromyza
lantance Frogg. attack large numbers of the ripening berries,
either destroying the seeds or causing the berries to shrivel up on
the bushes. The Tingid bug Teleonemia lantance Dist. destroys
the young leaves so effectively that extensive areas of Lantana
are checked in growth and fail to blossom. The larvae of the
Trypaneid gall-fly Eutreta xanthochwta Aid. to some extent check
the growth of new shoots. The larvae of the plume moth Platyptilia
pusillidactyla Walk, destroy many of the flowers in very much
the same manner as those of the Tortricid already referred to ;
on the whole it appears to be less efficient than the latter insect.
The Tineid leaf-miner Cremastobomhycia lantanella Busck. is of
minor importance, but its larvae when abundant cause the death
of the leaves, and to some extent restrain the vigour of the plant.
Finally, two species of Lycaenid butterflies Thecla echion L. and
T. bazochii God. are important, since their larvae destroy a great
many of the flowers, and thus restrain the production of the
berries. Unfortunately the Chalcid Pantarthron flavum Perk,
parasitises the eggs of the Theclas to some extent, while, on the
other hand, the larvae of T. echion have proved not entirely
harmless to garden plants, being occasionally found feeding on
fruits of the egg-plant. The most valuable of the insects just
enumerated are those which inhibit seed production, and the
result has been that the spread of the plant has been greatly
checked, and it does not so readily re-infest land from which it
had been previously eradicated. Although the work of repressing
Lantana has been materially lightened by these introductions, the
control of that plant can only be regarded as having been partially
accomplished, and it is possible that the colonisation of other
insects will be attempted in the future.

Control of Prickly Pear. In Australia the area infested by
prickly pear (Opuntia spp.) is estimated to amount to 60,000,000

414 BIOLOGICAL CONTROL

acres in Queensland and New South Wales, and in 1925 the
plant was stated to be spreading at the rate of a million acres a
year, but this increase is no longer being maintained. In the
main the affected areas embrace natural grazing country, where
the land is worth less than £3 per acre. Under such conditions
chemical or mechanical control is impracticable except in lightly
infested country. In their native terrain in North and South
America about 350 species of Opuntia are known, but none is a very
serious enemy, yet of the few kinds introduced into Australia, at
least four are to be regarded as major or minor pests. In America
insects, diseases, and other agencies keep the prickly pear within
reasonable bounds, whereas in Australia such natural controlling
factors are wanting, and there is little to check the reproduction
and spread of the pest. The Prickly Pear Board of Australia is
concerned with an attempt to bring about a condition of biological
equilibrium by the introduction of insects and plant diseases
likely to act as natural checks ; the control aimed at depends
upon the introduction of a complex of organisms working together
in destructive unison (vide Dodd, 1927, 1929). It needs to be
recollected that the two chief pest pears in Australia are Opuntia
inermis and 0. stricta, while several others are of minor importance.
This fact complicates the problem, for the reason that a particular
species of insect may prove effective against one kind of prickly
pear, and yet be of little value with respect to other kinds of those
plants. Officers of the Board have been engaged in studying the
insects affecting Opuntias in their native surroundings. They
have covered widespread cactus areas in North America, where
field stations have been set up, and have also visited South
America and the West Indies. In work of this character it is
important to study on the spot, not only the insects actually
attacking prickly pear, but also the natural parasites of such
insects. The exclusion of the parasitic forms from Australia is of
prime importance if their host insects are to multiply freely and
vigorously attack the prickly pear. At the Board's Station at
Urvalde, in Texas, extensive biological work has been prosecuted,
and the most promising cactus-feeding insects bred under caged
conditions ; furthermore, their life-histories have been worked out.

CONTROL OF PRICKLY PEAR 415

such insects were tested relative to the possibilities of their
attacking economic plants, and freedom from parasites being
ensured. The material received from America was transferred to
quarantine buildings at Sherwood, near Brisbane, where they
were bred through one or more generations as an additional
safeguard against the accidental introduction of their parasites.
At Sherwood, also, further tests were conducted with respect to
the possibility of the introduced insects attacking crops and other
useful plants. From Sherwood the insects were eventually
forwarded to acclimatising and breeding centres where, as a rule,
the first liberations were also carried out. The acclimatisation of
North American insects in a country where they are faced with
opposite seasonal conditions naturally presents considerable
difficulties. Generally it has been found that repeated shipments
of a species of over a period of one or more years have been
necessary before it has become established, but in a few cases
efforts have failed altogether.

A number of species have become acclimatised to Australian
conditions, and among them the moth Cactohlastis cactorum, whose
larvae tunnel through the tissues, is the most important. The
cochineal, Dactylopius tomentosus, became distributed almost
throughout the infested areas, and the plant bug, Chelinidea
tubulata, spread in countless millions in many localities. The red
spider, Tetranychus opuntice, also covered many thousands of
square miles. A. P. Dodd, in his 1929 Report on this problem,
stated that the established complex of insect enemies was already
bringing about a considerable degree of control of this noxious
pest. In the heart of the prickly pear country it was possible to
travel for 100 miles without seeing any healthy plants. In his
1936 paper Dodd emphasises that the Board's policy was based
upon the conception that biological control offered best chance of
success if a carefully selected group Xtf species working more or less
in association was established. It was unforeseen that the
outstanding success, evident in 1936, would have been effected
by the agency of a single species of insect in the space of a few
years. Nevertheless, this is what actually has happened, and the
insect in question is the Phycitid moth, Cactohlastis cactorum. The

416 BIOLOGICAL CONTROL

fact is all the more remarkable for the reason that only 2,500 eggs
(from the Argentine) of the insect were introduced into Australia,
yet between 1928-30 about three thousand million eggs, laid by
descendants of insects issuing from the original batch, have been
distributed in the great prickly pear areas. The eggs are laid by the
moth in " sticks," averaging seventy-five eggs in each : these sticks
are readily collected and artificially attached to the cladodes of the
host-plants. The resulting larvae are gregarious, internal feeders
which tunnel in companies through the tissue of the plant, thus
also providing for the ingress of disease organisms. In this way
the prickly pear ultimately becomes so completely destroyed that
it is reduced to a rotting mass of pulp. The various insects,
established prior to the Cactoblastis, have either been largely
suppressed or their activities nullified, owing to competition with
its larvae. It is only locally, and in relation to a few species of
Opuntia of lesser importance, that the Cactoblastis has proved
more or less ineffective. Such problems, however, are being dealt
with through the operations of other phytophagous insects,
including cochineal (Dactylopius) and Ccrambycid beetles. To-day,
the enormous rate of increase of the prickly pear has been arrested
and less than 10 per cent, of the former great body of infestation
survives : the whole of the primary pear in Queensland and much
in New South Wales has broken down and collapsed. Approxi-
mately 25 million acres of good land are now" cleared and are being
developed and brought under production.

Control of Prickly Pear in other Lands. The spectacular success
attending the control of prickly pear in Australia has rather
masked the efforts made to check Opuntias in other lands. In
Ceylon an extensive area of the Northern Province was formerly
infested by Opuntia monocantha, but was ultimately almost
completely exterminated by the cochineal Dactylopius ceylonicus,
which is stated to have come from Madras. This insect, it may
be added, has exercised a high degree of control over 0. monocantha
in many parts of India for a number of years past. According to
Jepson (1930) the cochineal insect has not maintained the same
efficiency in controlling this species of prickly pear in the Southern
Province of Ceylon. Another Opuntia, viz. 0. dillenii, has become

CONTROL OF PRICKLY PEAR 417

very prevalent in recent years in the Northern Province, and it is
believed to be an accidental introduction from Madras. Dactylopius
ceylonicus does not appear to be capable of exercising any con-
trolling influence over this plant, a fact which led to the
introduction in 1924 of D. tomentosus (opuntice) from Australia.
According to Jepson, the experiment has proved remarkably
successful. A large area of scrub in the neighbourhood of
Trincomalee was cleared by its agency in eighteen months. It
has also given equally satisfactory results in other districts, land
which was previously occupied by Opuntia scrub, 6 to 10 feet high,
is now under cultivation.

In 1926-27 consignments of Dactylopius tomentosus were
forwarded to Mysore and to Tuticorin with the object of
endeavouring to control Opuntia dillenii. The insects have
become established and it is stated that an area of 40,000 square
miles is now covered by the cochineal in southern Madras. Very
large tracts of land have been cleared through its agency, and land
so liberated has now come under cultivation for the first time for
many years. It is further claimed that the whole prickly pear area
of 114,000 square miles offers prospects of being cleared by 1940,
judging by the progress so far achieved. As Jepson remarks, if
this forecast be fulfilled, the experiment promises to become a
classic example of the control of weeds by biological methods.

Several species of Opuntia have become troublesome pests in
Mauritius, notably 0. tuna and 0. monocantha. The introduction
of Dactylopius ceylonicus by d'Emmerez de Charmoy resulted in
rapid destruction of Opuntia monocantha, and in his 1929 report
on the subject it is stated that within fifteen years this plant has
become completely controlled on the Island. The introduction of
D. tomentosus from Ceylon in an attempt to control 0. tuna is a
much more recent experiment. Its establishment has proved
successful and this prickly pear has been cleared from large areas.

In Madagascar it appears that rather an anomalous state of
affairs exists. R. Decary states that Opuntia dillenii was formerly
prevalent in a wild state on uncultivated land and was a local
fodder plant. Since 1924 the commercial cochineal Dactylopius
coccus entered Madagascar under accidental circumstances, it is

K.A. EXTOMOLOQY. - 14

418 BIOLOGICAL CONTROL

believed from the neighbouring Island of Reunion. The insect
spread so rapidly under the prevailing conditions that, in about
two 3"ears, the Opuntia was virtually destroyed and the cochineal
insect itself became no longer evident. Decary has more recently
suggested the introduction of some spineless species of Opuntia
suitable to take the place of 0. dillenii, and which, at the same
time, is resistant to the Dactylopius.

Control of Clidemia. An experiment has been carried out in
Fiji for the purpose of attempting to control the noxious shrub
Clidemia hirta. In Trinidad the plant is not regarded as a pest,
and is kept under restraint by a combination of various factors, of
which phytophagous insects are but one of them. Studies carried
out in Trinidad showed that the Thrips, Liothrips urichi, is common
on this plant throughout the Island. It apparently causes its host
to become stunted and to flower less freely. It therefore appeared
to be a suitable insect for introduction into Fiji, provided that it
were proved to exhibit no tendency to attack plants of economic
importance. In October, 1929, W. H. Simmonds visited Trinidad
and studied the Thrips on the spot. It was found to be attacked
by a number of natural enemies which had to be excluded in any
effort made to introduce the insect into Fiji. He was able to send
a large consignment of the species to Suva, where a proportion of
its individuals were released in the field in 1930 and the remainder
kept for breeding further stock. The Liothrips has taken readily
to Fijian conditions and has continued to spread. To-day it is
claimed that it has brought Clidemia definitely under control in
large areas. This has been attained not by killing the plant, but
by so inhibiting its growth that it is no longer able to compete with
dominant local plants.

Problems in New Zealand. A campaign was inaugurated in
1927 in order to attempt the control of four kinds of pest-plants
— gorse {Ulex eiiropoeus), blackberry (Ruhus fruticosus), ragwort
(Senecio jacobea') and piri-piri (Acama spp.). In the case of gorse
the plant is by no means wholly injurious, and for this reason only
control of its ultimate spread through seeding was aimed at. In
this connection the introduction from England of the pod-
infesting weevil, Apion ulicis, shows promise of success and its

LITERATURE 419

establishment in New Zealand is now becoming achieved. With
blackberry the preliminary testing of the stem-boring Buprestid
beetle (Coroebus rubi) from South Europe has given unfavourable
results, since both the larva and adult gave evidence of capability
for attacking economically useful plants. The outcome has been
the discarding of the project of biological control of blackberry,
since no other insect was deemed safe for the purpose. Ragwort
control is being attempted by the introduction of the Hypsid
moth, Tyria jacobece, from England, but the results obtained are
not very promising. This insect has become established in various
parts of New Zealand, but the activities of parasites of indigenous
insects and unknown influences are proving deleterious factors.
Owing to these difficulties, a second ragwort-feeding insect is being
studied with the object of being utilised in place of Tyria. The
insect in question, Pegohylemia jacobece, appears to be an effective
seed destroyer in its larval stage and is a member of the Dipterous
family Anthomyidse. No liberations appear to have yet been
made, and its possible success is still in the balance. In the case
of Accena, the burrs become entangled in the wool of the grazing
sheep and the market value of this product is estimated to suffer
an annual loss of £250,000. In this connection the introduction of
the saw-fly, Antholcus varinervis, from South Chile offers promise of
checking the plant in question, but no liberations have yet been
made. The flea-beetle, Haltica virescens, has also been tested out
as a potential beneficial insect, but it had to be abandoned owing
to the revelation of polyphagous propensities.

The present situation, therefore, as regards weed control in
New Zealand is that, although the project has not fulfilled the
promise hoped for by Tillyard, who originated the scheme, in
another ten years' time a useful degree of control will possibly
have been achieved in certain cases (vide Miller, 1936).

Literature

Burgess and Crossman, 1929. Tech. Bull. 86, U.S. Dept. Agric.

Chamberlin, 1926. Journ. Econ. Ent., XIX., 302.

Clausen, 1936. Ann. Ent. Soc. Am., XXIX., 201.

Clausen, King and Teranishi, 1927. Bull. 1429, U.S. Dept. Agric.

Compere and Smith, 1932. Ililgardia, VI., 585.

420 BIOLOGICAL CONTROL

Criddle, 1928. Canad. Ent., LX., 51.

Davidson, W. M., 1924. Trans. Amer. Ent. Soc, I., 163.

DoDD, 1927. Bull. 34, Council for Sci. and Indus. Res., Melbourne.

1 929. Progress of Biologieal Control of Prickly Pear in Australia, Brisbane.

1936. Bull. Ent. lies., XXVII., 503.

EssiG, 1931. '' History of Kntomologv," New York.

Flanders, 1929. Journ. Econ. Ent., XXII., 245.

Glendenning, 1933. Canad. Ent., LXV., 169.

Hazelhoff, 1929. IV. Internat. Cong. Ent., Ithaca., II., 55.

HOLLOWAY, Haley and Loftin, 1928. Tech. Bull. 41, U.S. Dept. Agric.

Howard, 1924. Proc. Ent. Soc. Washington, XXVI., 27.

Imms, 1926. Ann. App. Biol., XIII., 402.

1929. IV. Interned. Cong. Ent., Ithaca., II., 10.
Jepson, 1930. Trop. Agriculturist, LXXV., 63.
King, 1928. In U.S. Yearbook of Agric, 1927, 403.
Marchal, 1907. Ann. de Vlnst. Agronom., Ser. 2, VI.
Miller and Others, 1930. N.Z. Journ. Sci. Tech., XVIII., 579.
Myers, 1928. Empire Cotton Groiving Bev., V., 1.
Nicholson, 1933. Journ. Anim. EcoL, II., 132.

and Bailey, 1935. Proc. Zool. Soc, 551.
Perkins and Swezey, 1924. Ent. Bull. 16, Ilaxvaiian Sugar Plant. Exp. Sta.
Proper, 1934. Journ. Agric Pes., XLVIIL, 359.
Smith, H. S., 1926. Joiirn. Econ. Ent., XIX., 294.

1929. Btdl. Ent. Res., XX., 141.
1935. Journ. Econ. Ent., XXVIII., 873.

Smith, H. S., and Armitage, 1931. Univ. Calif. Exp. Sta. Bull., 509.
Smith, L. B., 1928. In U.S. Yearbook of Agric, 1927, 398.
Speyer, 1927. Bull. Ent. Res., XVII., 301.

1928. In Thirteenth Ann. Rep. Exp. and Res. Station, Cheshnnt, 75.
Sweetman, 1936. " The Biological Control of Insects," Ithaca.
Swezey, 1923. Proc Ilaivaiian Ent. Soc, V., 299.

1928. Journ. Econ. Ent., XXL, 669.
Taylor, 1935. Bull. Ent. Res., XXVI., 1.

1937. " The Biological Control of an Insect in Fiji," London.
Thompson, 1923. Ann. Ent. Soc Amer., XVI., 115.

1927. Biill. Ent. Res., XVIL, 273.
1927a. ift/rf., XVIIL, 13.

1928. Parasitology, XX., 90.

1930. Ann. App. Biol., XVIL, 306.
1934. Parasitology, XXVL, 449.

Thompson and Parker, 1928. Tech. Bull. 59, U.S. Dept. Agric.
TiMBERLAKE, 1927. Proc Ilcnvaiian Ent. Soc, Ml., 529.
Tothill, Taylor and Paine, 1930. " The Coconut Moth in Fiji," London.
WiLLARD, 1926. Proc. Ilcacaiian Ent. Soc, VI., 505.

Williams, 1931. " Insects and other Invertebrates of Hawaiian Sugar Cane
Fields," Honoluhi.

INDEX OF AUTHORS

Abbott, C. E., 137, 150, 169

Ahmad, T., 290, 313

Ainslie, C. G., 165, 170

Albro, H. T., 70

Alpatov, W. W., and Pearl, R., 217,

259
Anderson, A. L., 137, 169
Andrews, E. A., 273, 313
Aschner, M., 250, 259 ; and Ries, E.,

250, 259
Atzler, M., 202, 203

Back, E. A., and Cotton, R. T., 220,

237, 259; and Pemberton, C. E.,

252, 259
Bacot, A. C, and Harden, A., 244, 259
Baer, W., 293, 365
Baier, L. J., 150, 169
Ballard, E., 165, 170; and others,

308-9, 314
Barber, G. W., 221, 259
Barnes, T. C, 146, 169
Barrows, W. M., 160, 163, 170
Becker, R., 28, 50
Belehradek, J., 206, 259
Berland, L., 238, 259
Berlese, A., 51, 70, 126, 168, 249, 259
Bertholf, L. M., 114, 121, 168
Bodenheimer, F. S., 230, 259, 308,

314
Bodme, J. H., 210, 259
Bonier, C, 28, 33, 50
Bounhoil, J., 68, 70
Bozler, E., Ill, 168
Brecher, L., 200, 201, 203
Brindley, M. Haviland, 189, 203. See

also Haviland.
Brues, C. T., 229, 259; and Glaser,

R. W., 249, 259
Buchner, P., 247, 248, 260
Buddenbrock, W. von, 118, 122, 154,

168, 170 ; and Schultz, E., 122, 168
Bugnion, E., and Popoff, N., 132, 169
Burgess, A. F., and Grossman, S. S.,

380, 419
Butler, C. G., and Inne§, J. M., 303, 314

Buxton, P. A., 220, 228, 234, 238, 260 ;
and Lewis, D. J., 233, 260; and
Mellanby, K., 234, 260

Carothers, E. E., 211, 260

Carpenter, F. M., 81, 102

Caspari, E., and Plagge, E., 68, 70

Chamberlin, T. R., 381, 419

Chapman, R. N., 245, 260

Chopard, L., 365

Christophers, S. R., 47, 50

Chrystal, R. N., 51, 71

Clausen, C. P., 334, 365, 366, 403, 419 ;

King, J. L., and Teranischi, C, 335,

366, 384, 419
Cleveland, L. R., 246, 260
Coad, B. R., 238, 260
Cockayne, E . A . , 203 . See also Mottram .
Cockerell, T. D. A., 93
Compere, H., and Smith, H.S., 378, 419
Comstock, J. H., 49, 55, 71 ; and

Needham, J., 17, 18, 49
Cook, W. C, 163, 170, 213, 214, 260
Craighead, F. C, 287, 313
Crampton, G. C, 11, 33, 37, 50
Crane, M. B., 280, 313
Criddle, N., 165, 170, 386, 420
Crow, S., 137, 169
Crumb, S. F., 161, 170
Cunliffe, N., 264, 283, 313; Fryer,

J. C. F., and Gibson, G. W., 264, 313 ;

and Fryer, J. C. F., 282, 313

Daniel, D. M., 358, 366

Davidson, J., 236, 240, 254, 260, 266,

273, 277, 313
•Davidson, W. M., 409, 420
Davies, W. M., 230, 260
Dean, G. A., Cotton, R. T., and Wagner,

G. B., 221, 260
Dendy, A., and Elkington, H, D., 220,

221, 260
Denny, A., 50
Dodd, A. P., 415, 420
Durken, B., 201, 203

421

422

INDEX OF AUTHORS

Eastham, L. E. S., 5, 49, 51, 71

Eggers, F., 144, 170

Ellsworth, J. K., 112, 168

Eltringham, H., 108, 138, 168, 169

Essig, E. 0., 379, 420

Evans, A. C, 245, 260, 339, 366

Ewing, H. E., 22, 50

Faure, J. C, 187, 203, 300, 302, 310,

314, 364, 365, 366
Ferriere, C, 365, 366
Field, W. L. W., 297, 313
Flanders, S. E., 350, 366, 408, 420
Flint, W. P., and Larrimer, W. H., 267,

313
Folsom, J. W., 11, 49; and Bondy,

F. F., 167, 170
Forbes, W. T. M., 83, 102
Fowler, R., 157, 170
Fraenkel, G., 69, 71
Frew, J. G. H., 218, 260, 266, 269, 313
Friedrich, H., Ill, 168
Frisch, K. von, 113, 130, 137, 168
Frost, S. W., 157, 170
Fryer, J. C. F., 256, 297, 313 ; and

Collin, J., 264, 313. See also Cunliffe.

Gaffron, M., 119, 168

Genieys, P., 321, 366

George, C. J., 47, 50

Gerould, J. H., 185, 196, 204

Giersberg, H., 187, 202, 203

Glaser, R. W., 110, 131, 168, 169, 188,

203, 243, 260
Glendenning, R., 287, 420
Goschen, F., 297, 313
Grandori, R., 261, 349, 366
Griswold, G. H., and Crowell, M. F.,

233, 260
Gryse, J. J. de, 51, 71
Gunn, D. L., 233, 260

Hackett, L. W., Martini, E., and
Missiroli, A., 291, 313. See also
Missiroli.

, 49, 102
93, 102
10, 22, 27, 28, 30, 49, 50

Handlirsch, A.
Handschin, E.
Hansen, H. J.,
Hanstrom, B., 8, 49
Hardy, A. C, 239

Harrison, J. W. H., 198, 201, 203, 295,
313 ; and Garrett, F. C, 198, 204

Harukawa, C., Takato, R., and

Kumashiro, S., 167, 170
Hasebroek, H., 191, 192, 199, 204
Haviland, M. D., 326, 328, 366. See

also Brindley.
Hazelhoff, E, H., 410, 420
Headlee, T. J., 230, 260
Hefley, H. F., 231, 260
Henneguy, L. F., 360, 366
Hertweek, H., 146, 170
Hertz, M., 116-20, 168; and Imms

A. D., 187, 204, 314, 363
Hess, W. N., 170
Heymons, R., 28, 50
Hobson, R. P., 245, 250
Holdaway, F. G., 274, 313
Hollande, A. C, 187, 190, 204
Holloway, T. E., and others, 406, 420
Hopkins, A. D., 266, 287, 313
Hopkins, Sir F. G., 193, 204
Hoskins, W. M., and Craig, R., 2, 49,

163
Howard, L. 0., 364, 366, 401, 420
Howlett, F. M., 161, 170
Hungerford, H. B., 189, 204
Hunter, W, D., and Pierce, W. D., 207,

260
Husain, M. A., and Khan, A. W., 165,

170 ; and Mathur, C. B., 304, 314

ILSE, D., 114, 117, 169

Imms, A. D., 3, 49, 51, 71, 221, 345, 366,

371, 420; and Husain, M. A., 155,

160, 170. See also Hertz.

Jackson, D. J., 341, 361, 366
James, H. C, 51, 71
Jepson, F. P., 417, 420; and Gadd,
C. H., 271, 313

Kaestner, H., 192, 204
Keilin, D., 189, 204
Kelly, E. 0. G., 340, 366
Kemner, A., 336
King, J. L., 335, 366, 384, 420
Knight, H. H., 185, 204
Knoll, F., 114, 169
Koebele, A., 370, 372, 412
Koller, G., 70, 71
Kopec, S., 68, 71
Kornhauser, S. I., 358, 366
Krogh, A., 209, 218, 219, 260

INDEX OF AUTHORS

423

Kugler, F., 114, 169

Kuhn, A., 69, 71, 114, 115, 169

Lal, K. B., 290, 313

Lameere, A., 19, 50, 102

Larson, A. 0., 287, 313

Leiby, R. W., 356, 362, 366 ; and Hill,

C. G., 354, 366
Lichtenstein, J., 364, 366
Linden, Grafin G. von, 190, 204
Lloyd, LL, 158, 167, 170, 242, 260
Loeb, J., 122, 168
Ludwig, D., and Cable, R. M., 215, 260

MacLagan, D. S., 274, 313

Mansour, K., and Mansour-Bek. J. J.,

248, 260
Manton, S. M., 8, 49
Maple, J. D., 345, 366
Marchal, P., 57, 71, 294, 313, 354, 363,

366, 406, 420
Marcovitch, S., 236, 260
Marshall, J., 131, 163, 169
Marshall, J. F., and Staley, J., 293, 313
Martin, F., 354, 366
Martynov, A. B., 80, 102
Mason, G. W., 173, 204
Mayer, A. G., 148, 170
Mclndoo, N. E., 123, 124, 128, 130, 133,

137, 139, 155, 158, 163, 168, 169, 170
McKenny-Hughes, A. W., 198, 204
Mehta, D. R., 43, 47, 50
Mellanby, H., 6, 49
Mellanby, K., 229, 260
Metcalfe, M. E., 43, 47, 50
Meyer, P. F., 187, 204
Michelson, A. A., 179, 204
Miller, D., 419, 420
Minnich, D. E., 129, 133, 149, 169
Misra, S., 280, 313
Missiroli, A., Haekett, L. W., and

Martini, E., 291, 313. See also

Haekett.
Monzen, K., 277, 280, 313
Mottram, J. C, and Cockayne, E. A.,

203
Muesebeck, C. F. W., 324; and

Dohanian, S. M., 327, 363, 366
Muir, F., 372, 373
Miiller, E., Ill, 169
Myers, J. G., 152, 170, 395, 420

Nabert, a., 66, 71

Nel, R. I., 50

Nelson, J. A., 71

Neustadt, E. A. von, 165, 170

Newton, H. C. F., 128, 169

Nicholson, A. J., 369, 420 ; and Bailey,

V. A., 369, 420
Northrop, J. H., 219, 235, 260
Noskiewicz, J., and Poluszynski, G., 354
Nougaret, R. L., and Lapham, M. H.,

275, 313
Nuttall, G. H. F., 167, 221

Onslow, H., 173, 191, 204

Palmer, L. S., 184, 204 ; and Knight,

H. H., 184, 188, 196, 204
Pantel, J., 331, 366
Parker, H. L., 347, 358, 361, 366. See

also Thompson.
Parker, J. R., 210, 260
Parnell, F. R., 285, 313
Patterson, J. T., 354, 362, 366
Payne, N. M., 223, 260, 290, 313
Peairs, L. M., 208, 219, 260
Peniberton, C. E., and Willard, H. F.,

322, 367
Perez, C, 71
Perkins, R. C. L., 372 ; and Swezey,

0. H., 413, 420
Peterson, A., 157, 161, 170; and.

Heussler, G. J., 121, 169
Philiptschenko, J., 11, 49
Pictet, A., 237, 295, 313
Pierantoni, U., 249, 260
Pierce, W. D., 323, 366
Ping, C, 74, 92, 102
Pirsch, G. B., 220, 260
Poulton, E. B., 183, 199, 204
Power, F. B., and Chesnut, V. K., 159,

170
Prell, H., 28, 50, 336, 366
Priebatsch, J., 121, 169, 202, 204
Proper, A. B., 399, 420
.Przibrara, H., 186, 192, 201, 204
Pumphrey, R. J., and Rawdon-Smith,

F., 148, 152, 170

Ramsay, J. A., 229, 260
Redtenbacher, J., 17, 19, 50
Richardson, C. H., 161, 170, 243, 244,
260

424

INDEX OF AUTHORS

Richmond, E. A., 159, 170

Ries, E., 250, 260

Ripley, L. B., and Hepburn, G. A., 163,

169
Ripper, W., 248, 260
Roach, W. A., 282, 313
Robinson, W., 222, 261
Roonwal, M. L., 6, 30, 37, 49, 229, 261
Roubaud, E., 291, 293, 313; and

Gaschen, H., 291, 313 ; and Veillon,

R., 161, 170
Rubtzov, I. A., 311, 314

Sacharov, N. L., 224, 261

Salt, G., 326, 366

Salt, R. W., 226, 261

Sander, W., 120, 169

Sanders, J. G., and Fracker, S. B., 167,

170
Sayle, M. H., 218, 261
Schlegtendahl, 115, 169
Schlieper, C, 115, 169
Schmalfuss, H., 192, 204 ; and

Barthmeyer, H., 192, 204
Schnauer, W., 259, 261
Schopf, C, and Wieland, H., 193,

204
Schrader, F., 248, 261
Schroeder, C, 294, 313
Severin, H. H. P., and H. C, 163, 170
Shelford, V. E., 209, 216, 261
Shiraki, T., 165, 170
Sihler, H., 104, 128, 168
Silvestri, F., 11, 49, 336, 345, 354, 360,

366
Simmons, P., and Ellington, G. W., 241,

261
Singh Pruthi, H., 43, 47, 50
Sladden, D. E., 294, 314
Slifer, E. H., 146
Smith, H. S., 325, 333, 366, 369, 397,

420 ; and Armitage, 377, 420. See

also Compere.
Smith, L. B., 384, 420
Snodgrass, R. E., 1, 9, 22, 27, 28, 34, 37,

49, 50, 104, 126, 168
Spencer, H., 360, 361, 366
Speyer, E. R., 160, 171, 389, 420
Stear, J. R., 157, 171
Stickney, F. S., 37, 50
Strel'nikov, I. D., 303, 314
Strickland, E. H., 336, 366
Suffert, F., 173, 182, 204
Sweetman, H. L., 369, 420

Swezey, 0. H., 373, 420. See also
Perkins.

Taylor, I. R., 219, 261

Taylor, T. H. C, 388, 400, 420. See also

Tothill.
Theile, R., 280
Theobald, F. V., 165, 171
Thompson, D. L., 188, 204
Thompson, W. R., 293, 319, 331, 332,

337, 369, 381, 396, 401, 403, 404, 420 ;

and Parker, H. L., 287, 314, 319, 320,

396, 420
Thomsen, M., 321, 366 ; and Lemche,

H., 199, 204
Thorpe, W. H., 286, 288, 314, 342, 347,

350, 353, 367
Tillyard, R. J., 18, 50, 85, 102
Timberlake, P. H., 325, 327, 343, 367,

371, 420
Tothill, J. D., 348, 361, 367 ; Taylor,

T. H. C, and Paine, R. W., 388, 420
Townsend, C. H. T., 331, 367
Trager, W., 247, 261
Trouvelot, L., 364, 367
Turner, W. L., 165, 171

UiCHANCO, L. B., 248, 261

Urban, F., 122, 169

Uvarov, B. P., 234, 237, 239, 243, 248,

261, 300, 307, 314 ; and Zolotarevsky,

B. N., 307, 314
Uzel, H., 49

Valentine, J. M., 131, 169

Verguin, J., 252, 261

Verne, J., 190, 204

Verschaffelt, E., 158, 171

Vigier, P., 107, 169

Vignon, P., 21, 50 ; and Seguy, E., 21,

50
Vogel, R., 103, 127, 144, 168, 170
Voris, R., 336, 367

Wadley, F. M., 236, 261
Wadsworth, J. T., 336, 367
Walker, E. M., 37, 46, 50
Walsh, B., 286
Walton, W. R., 342, 367
Webber, R. T., and Schnaffner, J. V.,
319, 367

INDEX OF AUTHORS

425

Weber, H., 1, 28, 49

Weed, I. G., 68, 70, 71

Weinland, H., 218, 261

Weis, I., 137, 169

Werner, E., 247, 261

Wever, E. G., and Bray, C. W., 152,

170
Wheeler, E. W., 325, 364
Wheeler, W. M., 110, 127, 168, 169
Wieland, H., and Schopf, C, 193, 204
Wiesmann, R., 5, 49
Wigglesworth, V. B., 3, 49, 66, 70, 71,

154, 170, 193, 204, 229, 250, 261
Willard, H. F., 297, 420
Willcocks, F. C, 165, 171

Williams, C. B., 166, 171, 269, 313
Williams, F. X., 373, 375, 420
Withycombe, C. L., 269, 313
Wolf, E., 119, 169
Wolsky, A., 153, 170
Woods, W. C, 296, 314
Worrall, L., 284

YoTHERS, M. A., 156, 171

Zerrhan, G., 117, 169

Zolotarevsky, B. N., 307, 314. See also

Uvarov.
Zwolfer, W., 253, 261

GENERAL INDEX

(Generic names only are given in italics)

abdominal appendages, 41

Acerentulus, 28

Ageniaspis {Encyrtus), 354

air currents, 238

Aleochara, 336

alfalfa weevil, 381

Anomala, 373

Anomochorista, 78

Anopheles, races of, 291

antennae, 40

anthocyanins, 187

Anurida, 11

Apanteles, 186, 348, 361, 363, 370, 398

Aphelinus, 320, 338, 385, 389, 409

Aphelopus, 358

aphides, 188, 231, 236

Aphidius, 325, 328, 343

Aphis, 240, 241, 266, 277

Aphycus, 324

Apion, 418

Apis, 7. See also bee.

apodous larvae, 60

Archipanorpa, 90

Archotermopsis, 93

Aristopsyche, 90

Artona, 387

Aspidiotus, 388

Asterochiton, 167

atmometer, 235

atmospheric pressure, 237

Australia, biological control in, 413

Balisarcha, 297

Balhyplectes, 381

bee, 137, 220 ; colour vision, 114

Belmontia, 86, 99

bioclimatic law, 266

biological control, 368 ; factors

involved, 391 ; weed repression,

410 ; literature, 419
biological races, 286
biramous limbs, 22
blackberry control, 418

Blastothrix, 324, 344, 387

Blatella, 7, 44, 45

Blatta, 45, 233 ; sensillae, 128

Bonnetia, 336

" bound " water, 222

Brodia, 96

brown-tail moth, 379

Bruchus, 230, 287

Cadoblastis, 415

Calandra, 220

California, biological control in, 370,

376, 394, 408
CaUiphora, 69, 136, 230, 319
Calosoma, 380
Campodea, 7, 13, 23, 24, 28
Canada, biological control in, 386
Carausius, 5, 7, 12, 294 ; colour

phenomenon in, 121, 187, 202
carbon dioxide output, 214, 219
Carboniferous insects, 86
carotin, 184
Centeter, 339, 384
Ceralitis, 252, 374, 396
cerci, 42
cercoids, 43

Ceylon, prickly pear control in, 417
Chalicodoma, 7
Charips, 328
chemoreceptors, 123
chemotropism, applications of, 155
chlorophyll, 183
'Chlorops, 266,269

chordotonal organs, 140 ; functions, 145
Cicadidse, tympanal organs, 144, 152
Cimex, 66
Clonus, 197
claspers, 45, 49

Clidemia, biological control of, 418
climate, 251 ; and weather, 257
Coccophagus, 327, 350
codling moth, 123
cold, effects of, 221

426

GENERAL INDEX

427

cold hardiness, 222-7

Coleoptera ; trochanters, 25 ; gula, 38 ;

genitalia, 47 ; fossil forms, 93 ;

coloration, 179, 181 ; parasitic forms,

335, 336
Colias, larval colour, 186, 196
Collembola, 11, 28, 230 ; fossil forms, 86
Colorado potato beetle, see Leptinotarsa,

159
coloration, 172 ; literature on, 203
colours, structural, 172 ; pigmentary,

183 ; combination colours, 194
colour vision, 113
compound eyes, 107
Compsilura, 319, 341, 342
continental areas, biological control in,

395
Copidosoma, 356, 362
Coroebus, 419
corpora allata, 65
cotton boll weevil, 139, 207
cotton Jassid, 284
coxites, 41, 45, 46
coxosternum, 41
Cretaceous insects, 92
crops and insect attacks, 263
Cryptocercus, 247
Cryptochoetum, 342, 353, 371
Cryptognatha, 388
Culex, races of, 291
cyclopoid larva, 57, 58
Cydia molesta, 156, 381, 383
Cydia pomonella, 140, 156. See also

codling moth.
Cyrtidse, 335
Cyrtorhinus, 372

Dachnusa, 326
Dactylopius, 416
Dacus, 162
Datana, 150
Dendroctonus, 287
developmental units, 216
Devonian insects, 86
Dexia, 335

Diaclmsma, 322, 374, 397
Diceretus, 360
Diatrcea, 406
Dibrachys, 328, 399
Dieconeura, 76
diffraction colours, 174
Digonochceta, 399
Dinocampus, 341, 343

Diptera, 21 ; genitalia, 47 ; larvae, 60 ;
fossil forms, 91, 94 ; chemotropic
responses, 160, 161 ; parasitism in,
351

Dissosteira, 13

Ditaxineura, 96

Dociostaurus, 309

" dopa," 191

double ocelli, 110

dragon-flies, see Odonata.

Drosophila, response to light. 111 ;
chemotropism, 160 ; development,
215 ; duration of life, 217 ; COg out-
put, 219; effects of light, 235;
nutrition, 243

Dunbaria, 89

Dytiscus, 25

ectoparasites, 320, 339

embryonic phases, 52

Empire Cotton Growing Corporation,

284
Encarsia, 389, 410
EncyHus, 349, 350, 355
England, biological control in, 389, 394,

399
Entinms, coloration,' 178, 179
Eosentomon, 28, 42

Ephemeroptera, 15, 20 ; fossil forms, 97
Ephestia, 244, 290
Ephialtites, 91
Eriosoma, 410
Ernestia, 336 .
Euclecanium, 387
Eugereon, 82
Eulecanium, 294
Euplectrus, 321
Euproctis, 324
European corn borer, see Pyrausta.

Figites, polypod larva, 58

Figitidffi, 58

Fiji, biological control in, 387, 418

flavones, 188

fluorescence, 203

food plants, change of habits with

respect to, 294
Forficula, 7, 399
fossil insects, 72 ; sequence of, 86, 101

literature on, 102
frit-fly, see Oscinella.

428

GENERAL INDEX

genitalia, 43, 49

geographical distribution, 251

geological history of insects, 86

gipsy moth, 379, 399, 401

Glossina, 233, 251

Gonepteryx, 193

Gania, 321

gorse, control of, 418

Gout fly, see Chlorops.

grasshoppers, phase development, 311

development, 211
Grylloblatta, 46, 47
Gryllus, 7, 104, 105, 148
gula, 37
gustatory sense, 132

Habrobracon, 192, 321, 364

Habrocytus, 364

Hadejia {Polia), 158, 242

haemoglobin, 189

halteres, 154

harpagones, 45

harpes, 47

Hawaiian Islands, biological control in,

371,393,399,412
head, segmentation of, 4
heat, effects of, 220
Helopeltis, 273
Hemiptera, gula, 38 ; genitalia, 47 ;

fossil forms, 98 ; gustatory organ,

132; colours of, 184, 188 ; symbionts

of, 248
Hemiteles, 319
Hessian fly, see Mayetiola.
Heterojapyx, 14
Hippodamia, 408

hormones and metamorphosis, 64
host -selection, 286, 318 ; interpretation

of, 297
host-selection principle, 287
hot springs, 220
house-fly, 243
humidity, 228
hybridisation, 297
Hymenoptera, trochanters, 26 ; larval

types, 57, 58, 60 ; fossil forms, 91, 94 ;

parasitism in, 344
Hypera, 61

hypermetamorphosis, 62
hyperparasitism, 326, 398
Hyphantria, parasites of, 323
Hyponomeuta, 288
hypopharynx, 14, 15

Ibalia, 58, 59

Icerya, 353, 370

India, prickly pear control in, 416

indigenous parasites, utilisation of, 405

intercalary segment, 12

interference colours, 175

islands, biological control in, 392

Isoptera, 35 ; fossil forms, 93

Isotoma, 28, 29

Japanese beetle, see Popillia.
Japyx, 7, 10, 24, 39, 40, 41
Johnston's organ, 141, 143
Jurassic insects, 90

Kennedya, 88, 96

labium, 35

labrum, 5, 7

Lachnosterna, 166

Lantana, biological control of, 412

larch sawfly, 386

larvae, 54, 55 ; types of, 56

Lebia, 336

leg, 23 ; primitive structure, 31

Lemmatophora, 79

Lepidoptera, genitalia, 47 ; fossil forms,
93; colour vision, 114; tympanal
organs, 144 ; sound perception in
larvae, 149 ; melanism, 196 ;
structural colours, 177 ; pigmentary
colours, 183 ; fluorescence, 203

Lepisma, 4, 7, 24, 28

Leptinotarsa, 159, 255

Leptomastidea, 378

Levuana, 387

light-compass orientation, 122

light, reactions to, 121, 164 ; action of,
235

Limnerium, 325

Liothrips, 418

Lithobius, 14, 26, 32

Litomastix, 356, 362

Locusta, 6, 7, 12, 30, 37, 152, 229;
coloration, 187 ; phases of, 300, 307

Locustana, 310

locustine, 304

locusts, 299 ; swarming of, 305
species of, 306

Lucilia, 112, 208, 245

Lygceonematus, 380

Lymantria, 68

GENERAL INDEX

429

Machilis, 15, 25, 28, 44

Macrocentrus, 358, 361, 383, 409

Madagascar, prickly pear control in, 417

mandibles, 38

Mantis, 7

Mastotermes, 35, 44, 45, 93

maxilla, 32

maxillulse, 10, 11

Mayetiola, 267, 356

Mecoptera ; fossil forms, 92, 99

Mediterranean fruit fly, see Ceratitis.

Megasecoptera, 77, 95

Megatypiis, 90

melanin, 191

melanism, 197

Melanoplus, 68, 70, 110, 146, 310

Melittobia, 321

Mesolenis, 386

Mesotitan, 90

metamorphosis, 51

Metaphycus, 378, 398

Meteorus, 341, 346

micro-organisms and insects, 245

Micropterygidae, 93

Miomoptera, 80

Mischoptera, 78

Morpho, coloration, 174, 177, 178, 194

multiparasitism, 322, 396

mushroom bodies, 9

mycetocytes, 248

naiad, 55

Necrobia, 242

Nemeritis, 290

Neocelatoria, larvipositor of, 342

Nesomachilis, 14

New England, biological control in, 379,

393
New Zealand, biological control in, 385,

399,418
Nicoletia, 41, 48, 49
Nomodacris, 310
nun moth, 253

nutrition, the general subject, 239
nymphs, 54, 55 ; and pupse, 63

ocelli, 109

Odonata, 15, 21 ; trochanters, 27

fossil forms, 88, 96 ; vision, 107
oenocytes, 70
olfactometers, 158, 163
olfactory sense, 124 ; sensillse, 126
oligopod larvae, 60

Oniscidse, parasites of, 319

Ootetrastichns, 372

Opius, 322, 374, 396

Opuntia, biological control of, 411, 413

oriental peach moth, see Cydia molesta.

Orthophlebia, 92

Orthoptera, 20, 36 ; ovipositor, 45 ;

genitalia, 46 ; tympanal organs, 143,

146
Oryssns, 61
Oscinella, 263, 282
ovipositor, 43, 49

Palaeodictyoptera, 19, 74, 94

palaeontology, 72

Panorpoid complex, 18, 20

Parabehnontia, 86

paragnaths, 11

Paramecoptera, 85, 99

parameres, 48

Paranagrus, 372

parasites, 315 ; host-selection in, 318 ;

life-cycles of, 330 ; host relations,

343 ; phoresy, 364
parasitism, 315 ; phases of, 322 ;

sequence theory of, 401
parasitoids; 316
Paratrichoptera, 85
Paroryssus, 91
Pediculus, 147, 221, 250
Pegohylemyia, 419
penis, 47

Perilampus, 325 ; planidium of, 333
Perillus, 184, 185
PeripUneta, 33, 131, 146, 148, 229
Perkinsiella, 372
Permian insects, 87
Permochorista, 99
Permosialis, 89
Permotipula, 85
Phcenodiscns, 345
Phmnoserphus, 59
phases of locusts, 300
Philonthus, 25
phoresy, 364
Phormia, 135
Phyllodeda, 294
Phylloxera, 274
phylogeny of insect orders, 94
physiology of insects, 2
Pieris, 5, 191, 193, 199, 200, 370, 386 ;

chemotropism in, 129, 158 ; sound

perception, 150
Pimpla, 324

430

GENERAL INDEX

Piophila, 242

planidium, 333

Platyedra, 165

Platygaster, 356 ; protopod larva, 57

Plecoptera, 21, 36 ; fossil forms, 97

Plesiocoris, 296

pleuron, 28

Pleurotropis, 329, 386, 404

poly em bryony, 353

Polygnotus, 356

polypod larvae, 58

Polysphincta, 60

Pontania, 295

PopiUia, 159, 381, 384

Porosagrotis, 212

postmentum, 35

pre-antennse, 12

prementum, 35

prepupa, 64

pretarsus, 23

prickly pear, biological control of, 413

Promecotheca , 388

Prosena, 335, 384

Protagrion, 80

Protelytron, 84

Protelytroptera, 83

Protereisma, 89

Protoblattoidea, 76, 95

Protocoleoptera, 82

Protocoleus, 83

Protodiptera, 85

Protodonata, 79

Protohemiptera, 81, 97

" Protohymenoptera," 78

Protomecoptera, 99

Protoparce, 231

Protoperlaria, 80, 97 ; nymph, 81

Protophasma, 11

protopod larvae, 56

Protorthoptera, 75, 95

Protozoa and termites, 246 ; and

Cryptocercus, 247
Protura, 28, 42
Pseudapkycus, 374
Pseudococcus, 373
Psylla, races of, 290
Pteromalus, 386
Ptychomyia, 387
pupae, 63

Pyrameis, 133, 137, 138
Pyransta, 221, 287, 381, 382

Quaylm, 379, 398

ragwort, control of, 418

rectal papillae or " glands," 229

respiration in endoparasites, 347

Rhabocnemis, 372

Bhacodineura, 332, 334

Rhagoletis, 296

Rhodnius, 6, 7, 12, 132, 154, 230, 250

hormones in, 66, 69, 70
Rhyniella, 86
Rhyssa, 385
Riela, 365
Rodolia, see Vedalia.

Saissetia, 378

Schistocerca, 304, 308

Schizaspidia, 334, 365

Schoenobnis, 165

Scolia, 373, 405

scolopale, 105

Scolopendra, 1, 14, 32

Scutellista, 378, 399

Scut ig era, 14

Scutigerella, 28, 32

Selenia, melanism, 198

sense organs, 103

sense-rods, 105

sensillse, 103

sequence theory of parasite control, 401

shot -hole borer, 271

sinigrin, 158

Sitodrepa, 216, 247

Sitotroga, 230, 408

Sminthurus, 230, 254, 273

soil conditions and insect attack, 267

standard metabolism, 218

Stenobothrus, 7

Stenodictya, 76

Stenoperlidmm, 89, 97

stimulatory organs, 153

subcoxa, 27

sugar cane froghopper, 268

superparasitism, 325

symbiosis, 245

Symphyla, 24, 28

Tachinidae, parasitism in, 251

tactile sense, 139

tarsal perception, 133

Telea, 222

temperature, 206 ; alternating, 210 ;

summation, 216
Tenebrio, 47, 131, 209, 219, 228
tentorium, 14, 15

GENERAL INDEX

431

termites and Protozoa, 246

Tertiary insects, 92

Tetrastichus, 363, 374

Therion, 361

Thrixion, 339, 352

Thj^sanura, 11, 23, 28, 36; abdominal

appendages, 41 ; genitalia, 48
Tibicina, 30
Tineola, 233
Tiphia, 385
transtarsus, 23
Triassagricni, 90
Triassic insects, 90
Triassolestes, 90
Triassopsy chops, 90
Triatoma, 66
TriboUum, 245
Trichogramma, 326, 407
Trichlosoba, 87
Trichoptera, 91
trochanter, 25, 26, 27
trochantin, 29
trophamnion, 360
tropisms, 122
tsetse flies, see Glossina.
tympanal organs, 143 ; functions, 151
Tyria, 419
tyrosin, 191, 193
tyrosinase, 191, 193, 201

ultra-violet perception, 114, 121
United States, biological control in, 370,

376, 379, 381, 395
Urania, coloration, 176
uric acid, 193

vacuum, effects of, 237

Vanessa, 109, 133, 149, 153, 190, 192,

199
van't Hoff's law, 210, 219
varietal resistance to insect attack, 276
Vedalia, 370, 385
velocity curve, 209
venation, 16
vision, 108; by ocelli, 110; colour

vision, 113
vitamins, 243

weeds, biological control of, 410'
white-fly control, 389
Winthemia, 231
woodlice, parasites of, 319
woolly aphis, 277, 279

Xiphidium, protopod phase, 57
Xyleboms, 271

PRINTED IN GREAT BRITAIN BY THE WHITEFRIARS PRESS LTD.
LONDON AND TONBRIDGE

\