A SERIES OF BIOLOGICAL HANDBOOKS UNDER THE GENERAL EDITORSHIP OF Professor J. Arthur Thomson, M.A., LL.D. THE BIOLOGY OF INSECTS UNIFORM fVITH THIS FOLUME Illustrated. Each \6s. net. THE BIOLOGY OF THE SEA-SHORE. By F. W. Flattely and C. L. Walton. With an Introduction by Professor J. Arthur Thomson. THE BIOLOGY OF BIRDS. By Professor J. Arthur Thomson, M.A., LL.D. THE BIOLOGY OF FLOWERING PLANTS. By Macgregor Skene, D.Sc. THE BIOLOGY OF FISHES. By Harry M. Kyle, M.A., D.Sc. /;/ Preparation. THE BIOLOGY OF MAMMALS. By James Ritchie, M.A., D.Sc. THE BIOLOGY OF SPIDERS. By Theodore H. Savory. PLATi: I A. Stick-insect {Caraiisius 7nofosi/s), Java, moulting while hanging from Oak-twig. Half size. [/. //. ]l'atso>t, thoio. ■|^BHHHHHHpB|fH| ■npr^ HjKr**?!^ n^H ^ ^IH IK'i^vJ Hite. < npl l»^^ w^^^^B^\ .' *'i^^H hphh H^^s Ik- ^^^HD^H '^ j4K>- .A^^iB IHB 1 H^Mr^B - - '^^1^1^ '^ jI^'IHH^BRlH R^HHIl ^JL HHRkt^k^ fe^^g ^hI n Biy^y ^H Hi ■ VhS V -nIcPIBHI^^s^ I^S I9h H B. Snowy-fly {Alcyrodcs vapoy'tariim) ON Leaf, x io. Frontispiece.'] l^^- Brittcu, photo. THE BIOLOGY OF INSECTS BY GEORGE H. CARPENTER, D.Sc. KEEPER OF THE MANCHESTER MUSEUM, UNIVERSITY OF ItANCHESTBR FORMERLY PROFESSOR OF ZOOLOGY IN THE ROYAL COLLEGE OF SCIENCE, DUBLIN * *. LONDON SIDGWICK & JACKSON, LTD, 1928 MADE AND PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS. LIMITED. LONDON AND BECCLKS. TO MY WIFE AFTER THIRTY-SIX YEARS OF COMRADESHIP PREFACE The writer on Insects in this series of volumes, setting forth various groups of creatures from the biological point of view, has the advantage of his subject in a class of animals comprising the largest number of diverse forms whence he may choose examples of life relations. He has also the disadvantage which follows inevitably from this richness of the available material ; many highly interesting features of insect life must be neglected or treated in- adequately if the book is to be kept within reasonable compass. He must therefore acknowledge himself charge- able with the offence of omitting many subjects which might be expected to appear in a survey of the Biology of Insects. In this volume structural features are described only so far as seems necessary for the understanding of function and behaviour, while questions of systematic entomology are discussed only as they bear on problems of ecology and evolution. The author's indebtedness to those who have studied and written upon the various aspects of Insect Biology is apparent, and, he trusts, duly acknow- ledged, both in the text and in the descriptions of those photographs and figures which have been borrowed or copied for illustration. He desires to express gratefully his appreciation of the help and encouragement accorded by his friend, Professor J. Arthur Thomson, the editor of this series, whose suggestions and criticisms have contributed ^ PREFACE largely to whatever value the book may have. It is a pleasure to acknowledge also the willing help of two friendly colleagues in the work of illustration: Miss R. A. Barr with her careful draughtsmanship and Mr. H. Britten with his taste and skill in nature-photography. Finally hearty thanks are due to the publishers for their patience and consideration through the prolonged fulfil- ment of the author's promise to take this part in their scientific enterprise. G. H. C. The University, Manchester, October, 1927- CONTENTS PAGE CHAPTER Preface "^^ Contents ^^ List of Plates ^^ List of Drawings in the Text xin ' I. Introduction , . * j'tr / ^ Structure and Function in Insects, i. Breathing and i< ced- ing, 3. Nerve-action, 7. Reproduction and growth, 10. Definition of the Class, 13. II. Feeding and Breathing • *, ^^ Living Cells and Protoplasm, 15. Jaws for Biting and Sucking, 17. Digestion, 25. Blood and Circulation, 32. Absorption, 36. Excretion, 38. Breathing, Air-tubes, 39. in. Movement . . . •-,..' ' c r ' fir ^'^ Muscles, 47. Motion of Legs, 50 ; of Wings, 54 ; ot Jaws, bi. IV. Sensation and Reaction • .69 Stimulation and Response, 69. Impulse and Reflex, 70. Touch, 72. Smell and Taste, 73. Equilibration and Hear- ing, 77- Sight, 83. V. Behaviour, Instinctive and Intelligent . . ^ . 94 Action Conscious and Purposive? 94. Tropisms, 95. Auto- matic and Routine Activity, 102. Experience and Adapta- tion, 103. Memory and Individuality, 106. Apparent Prevision, 107. Change of Behaviour, no. VL Reproduction and Heredity . . _„,.•..*_/ "^ Germ-cells: Sperms and Eggs, 114. Cell-division, ii5. FertiHsation and Maturation, 117. Inheritance ; Mendel s "Law," 121. Sex-determination, 125. Gynandromorptis and Intersexes, 131. Parthenogenesis, 137. Genital Organs, Ovipositor and Armature, 141- Secondary Sexual <^ha- racters, 147. VII. Growth and Transformation . • * ^ '. • V ^^° Embryo and its Development, 150. Germ-layers; Origin ot Endoderm, 153. Vestigial Appendages, 158. Hatching and Moults, 159. Direct Growth and Transformation, 102. Imaginal Buds, 173. Forms of Larvae, 176. Divergence m mode of Wing-growth, 186. Nature of Pupa, 191. ^irtli of Young, 195. Paedogenesis, 196. Polyerabryony, 197- IX 37304 X CONTENTS CHAPTBR PAOB VIII. Family Life 2C» Pairing, 200. Recognition by sight and scent, 201. Stridu- lation, 206. Courtship by feeding, 207. Egg-laying, 208. Parental Care, 210. Nests and Families, 215. IX. Social Life 218 Family Assemblies, 218. Social Beetles, 219. Family Societies : Wasps, 222. Bees, 225. Inquilines, 232. Ants, 237: Fungus-growers and Drivers, 239; Honey-ants, 241; Harvesters, 242 ; Slavers, 245. Guests, 248. Termites, 253. Society and Individual, 260. X. Adaptations to Haunts and Seasons .... 262 Geographical Range, 263. Plabit and Body-form, 268. Pro- tective resemblance, 272. Aquatic Insects, 273. Marine Insects, 282. Seasonal Adaptation, 298. Migration, 304. XI. Classification 306 Affinities and Relationship, 306. Wingless and Winged In- sects, 307. Exopterygota and Endopterygota, 308. Summary of Insect Orders, 309. Sub-orders and Families, 318. Genera and Species, 321. XII. Evolution 325 Natural Classification, 325. Descent with Modification, 326. Course of Insect Evolution, 327. Modification of Jaws, 329 ; of Wings, 330. Life-histories, 332. Fossil Insects and Geo- logical History, 336. Insects and other Arthropoda, 346. Factors of Evolution, 349. Heredity with Variation, 350. Continuous and Discontinuous Variation, 351. Use-Inherit- ance, 355. Germ-plasm Theory and Germ-track, 359. In- fluence of Environment and Food, 361, Natural Selection, 365. Protective Resemblance, 369. Mimicry, 371. "Dar- winism" and "Mendelism," 376. Isolation, 377. XIII. Insects and other Organisms 380 Plants as Food for Insects, 380. "Food-chains," 381. Leaf- eating Insects, 382. Wood-feeders, 384. Gall-makers, 385. Inquilines, 387. Migrant Plant-feeders, 389. Effects on Plants, 393. Insects and Flowers, 394. Insectivorous Plants, 395. Predaceous Enemies of Insects, 397. Insect Parasites of Animals, 398. Internal Parasites of Insects : Worms, 408 ; Protozoa, 409. Insects as alternate hosts, 412. External Parasites, 415. Transport of small animals by Insects, 416. XIV. Insects and Mankind 418 Pests and Allies of the Cultivator, 418. Cotton Insects, 419. Predaceous Insects, 423. Messmates of Man, 424. Destroyers of Buildings and Furniture, 427. Disease-carriers, 428. In- sects in Ancient History, 439. Insects as Human Food, 440. Domesticated Insects, 441. Services rendered by Insects, 445. Insect and Human Societies, 446. Parable and Purpose, 447. REFERENCES 449 INDEX 465 LIST OF PLATES Pl-ATE TO FACE PAGI I. A. Stick-insect moulting, b. Snowy-flies Frontispiece II. Eggs and Young Larvae of Warble-fly . . .108 III. Wanderings of Warble-maggots in Cattle . . .110 IV. Moths illustrating Mendelian Inheritance . . 122 V. A. Gynandromorph Emperor Moth. b. Poplar Hawk Moth and Eggs • 132 VI. Eggs and Young Larvae of Currant Sawfly . . 160 VII. Nest and Comb of Social Wasps 182 VIII. a. Nest of Tree-wasp cut open. b. Nest of Ground- wasp DUG OUT BY Badger 224 IX. Nest of Bumble-bee 226 X. A. LooPER Caterpillars protectively marked. B. Nymph Cuticles of Stonefly .... 272 XI. A. Parasitic Larvae on Caterpillar, b. Nest of Solitary Wasp , , 322 XII. Insect Fossils from Palaeozoic Rocks, Carboniferous AND Permian . . . . . . . . 336 XIII. British Lepidoptera (a) showing Regional Variation, AND (b) used to demonstrate MODIFICATION THROUGH Change of Food 356 XIV. African Butterflies illustrating Seasonal Forms and Mimicry 374 XV. Oak-apples and Root-galls of Alternating Seasonal Generations 386 XVI. Late Larvae and Puparia of Warble-flies of Cattle 404 DRAWINGS IN THE TEXT .^ FIG. I. 2. 3. 4- 5. 6. 7- 8. 9- 10. II. 12. 13- 14. 16. 17. 18. 19. 20. 21. 22. 23- 24. 25- 26. 27. 28. 29. 30. 31- 32. Section through Skin and Cuticle of Bluebottle Stonefly [Taeniopteryx) and Jaws Germ-cells (Sperm and Egg) of Insects Stages in Growth of Plant-bug Owl Moth {Agrotis) with Egg and Larva Mouth-parts of Earwig MOUTH-PARTS OF BrEEZE-FLY Jaws and Mouth of Shield-bug Maxilla of Owl Moth Digestive and Nervous Systems of Honey Bee Crop and Stomach of Honey Bee Blood and Air-tube Systems of Honey Bee Malpigkian Tube of Honey Bee Tracheal System of Termite Spiracle and Air-tubes of Louse Histology of Insect Muscle Leg of Cockroach with Muscles Wing-structure in Stone-fly and Dragon-fly Wing-muscles of Honey Bee Wings of Bee Wing-coupling Apparatus . Mandibles of Cockroach with Muscles Maxilla of Cockroach with Muscles Head of Plant-bug with Muscles Sense-organs in Feeler of Syrphid Fly Chordotonal Organs .... Tympanal Organ of Chrysiridia . Compound Eye of Honey Bee Details of Bee's Eye .... Trails of normal Bees in Directive Light Trails of Bees with one eye obscured . Diagrams of Cell-division and Maturation xiii PAGE 2 3 9 12 18 20 22 24 26 28 33 38 41 43 48 51 55 56 58 59 62 64 66,67 74 79 81 84 86 97 99 119 XIV DRAWINGS IN THE TEXT FIG. 33. Chromosomes of Lygaeus .... 34. Female Reproductive Organs of Hytoderma 35. OvirosiTOR of Hypoderma . 36. Ovipositor of Grasshopper 37. Male Reproductive Organs of Hypoderma 38. Male Armature of Hypoderma . 39. Embryonic Development of Moth 40. Embryonic Development of Grasshopper 41. Formation of Mid-gut 42. Forms of Vine Aphid .... 43. Apple Sucker with Eggs and Young 44. Cicad and Larvae .... 45. Caterpillar and Pupa of Moth 46. Wing-buds of Lady-bird Beetle 47. Ground-beetle, Larva and Pupa 48. Chafer and Larva .... 49. Bark-beetle and Larva 50. Larva of Honey Bee .... 51. Larva of Rhagoletis (Apple Maggot) . 52. Mayfly Nymph and Bristle-tail 53. Mealworm with external Wing-rudiments 54. Pupa and Puparium of Rhagoletis . 55. POLYEMBRYONY OF PlATYGASTER . 56. Scent-apparatus of Male Amauris . 57. Seashore Bug {^Aepophilus) . 58. Galleries of Ambrosia Beetles . 59. Nest of West African Wasp {.Belonogaster 60. Wasps { Vespa austriaca and V. rufa) 61. Driver Ants {Dorylus) 62. Larvae of Ants 63. Guest Beetle {Claviger) from Ants' Nest 64. Forms of North American Termites 65. Guest Beetles of Termites 66. Springtail {Achorutes viaticus) 67. Bed-bug and Dog-flea 68. Larvae of Aquatic Beetle {Donacia) 69. Structural Details of Donacia Larva 70. Seashore Springtail {Archisotoma beselsi) 71. Seashore ^'E.^t'le {Micralymma brevipenne),'LKKV\ k^d 72. Seashore Midge {Clunio), Larva and Pupa 73. Submarine Midge {Pontomyia) from Samoa 74. Marine Bug {Trochopus) .... PAGE 127 142 144 145 146 148 151 153 156 163 165 167 171 173 177 178 179 180 181 185 188 193 197 204 211 220 223 234, 235 240 244 250 254 259 265 270 27s 276 284 285 290 291 294 Pupa DRAWINGS IN THE TEXT XV FIG. 75. Pelagic Bug {Halobates) from Indian Ocean 76. Sawfly {Emphytus) with Larva and Pupa 77. Structural Details of Hymenoptera 78. Larvae of Scorpion-fly 79. Mandible and Hypopharynx of Mayfly Larva, Bristle tail and Isopod Crustacean .... 80. Germ-cells of Miastor 81. Variation in Tail-spines of Achorutes longispinus 82. Injury to Leaf-tissue by Hemiptera 83. Structure of Larvae of Hypoderma 84. Protozoan Parasites of Insects 85. Water-scorpion carrying small Bivalve . 86. Cotton Boll-weevil 87. CuLiciNE and Anopheline Mosquitoes with Larvae 88. Bee and Scarab Beetle on ancient Egyptian Stone Slabs PAGE 296 321 335 347 360 377 392 403 411 416 420 432 439 THE BIOLOGY OF INSECTS CHAPTER I INTRODUCTION The studies of an entomologist are often associated with long rows of dead, dried insects, set on pins or mounted on cards, arranged in a series of cabinets or storeboxes. Yet such a typical collection of beetles or butterflies is made up of creatures that once had life, and the opportunity which it affords for the examination and comparison of the forms of insect- bo dies may help towards an understanding of the living insects which once flew or crawled, fed and breathed, mated and bred in the brightness of a summer's day. In the succeeding pages the attempt will be made from diverse points of view to demonstrate insects as living organisms. In most groups of animals there are certain outstanding characteristics which determine to a great extent each creature's form of body and mode of life. The bird is feathered and flying, moulded as it were in clothing and manner of movement to the air. The typical worm is a crawler or burrower, compelled to seek shelter in soil or sand for its soft, ill-defended body. What then are the main features of an insect's body-structure which are correlated with its functions as a live being ? Curiously enough the collection of dry pinned or carded specimens suggests the beginning of an answer. Insects can be pre- served in this way because of the firmness of their outer I B THE BIOLOGY OF INSECTS body-covering ; they belong to that great race ^ of animals among whose members the living skin (Fig. i, e) forms out- side itself a more or less firm and resistant cuticle, composed of a horny substance called chitin (the chemical composition of which has been represented by the formula C30H5QO1QN4), and thickened segmentally in agreement with the marked segmentation of the body and limbs so as to build up a jointed exoskeleton. This coat of mail, as it may be some- what fancifully termed, affords the living insect much protection from its enemies and also enables it to achieve great precision and rapidity of movement, since the muscles of the trunk and limbs are attached internally to the hard firm parts of the exoskeleton, which being united by tracts of flexible cuticle, move readily on one another. Further, it is noteworthy that in this great group of animals the muscle- fibres when viewed with the microscope, show the same kind of cross-striping that charac- terises the body-muscles of vertebrates. Other classes of animals be- sides insects are built on the general plan just described — spiders, centipedes, millipedes, lobsters and crabs for example. But while most of these have many pairs of jointed limbs adapted for crawling or swimming, insects generally have the legs reduced in number to six, and the great majority of them display a highly characteristic feature in the presence of two pairs of flattened outgrowths arising from the dorso-lateral region of two adjacent, forwardly situated body-segments (Fig. 2). These outgrowths, jointed on to the body and capable of depression and elevation by the action of suitable muscles ^ The Arthropoda. See Classification, Chapter XI. Fig. I . — Section through cuticle (c) and skin or epidermis (e) of the leg-base of a Bluebottle (Calliphora) just after casting the pupal cuticle, h, sensory hair ; n, nerve-cell. X 300. INTRODUCTION 3 attached at or near their bases, are the wings — rigid enough, since they consist largely of firm cuticle, to support in the air the creature that bears them, so that insects are among the few groups of animals which possess the power of true flight ; like the birds they are moulded to the air. There are indeed many insects wingless or incapable of flight, the best known examples being such parasites as lice or fleas. But taking the class as a whole the wings are a dominant feature and flight is a characteristic activity, so that the vast Fig. 2. — A, Stonefly {Taeniopteryx pacified) North America, X 5 ; By mandibles ; C, maxilla ; Z), labium ; X 25. From E. J. Newcomer, Journ. Agric. Res. (U.S.D.A.) xiii, 1918. majority of insects may be regarded as segmentally built* armoured creatures, able not only with nice precision to walk or run, but to rise into and propel themselves through the air. Such strenuous activity calls for a constant supply of energy. In all animals the source of the energy dissipated in motion, in the radiation of heat, or in other channels, must be sought in the highly complex chemical constitution of the living body-substance the food materials combined or in association with which are broken down by oxidation- 4 THE BIOLOGY OF INSECTS processes analogous to combustion or explosion. The animal body requires therefore a supply of oxygen to support this combustion-process, and a supply of food to provide for the repair of waste and to act the part of fuel in the Hving heat-engine. An animal that continues to live must feed and breathe. Nov^ the organs and the method of breathing in insects are among the most markedly distinctive of the structural and vital features of this class of animals. In several classes of the Arthropoda — the great compre- hensive Group or Phylum to which insects belong — we find that the outer skin with its firm cuticle is pushed into the body in such a manner as to give rise to a series of branching air-tubes of which the firm cuticle necessarily forms the lining and prevents their channels from collapsing. As these air-tubes open to the outside through a set of breathing- holes or spiracles, air can pass in and come into touch with organs inside the creature's body so that the process of oxidation becom.es easy and direct ; the oxygen needs not the circulating fluid or blood to carry it from special breathing organs such as lungs or gills to the tissues where the combustion processes are constantly going on (Figs. 14, 15). Air-tubes like those just described are found in centi- pedes, in certain spiders, and in many mites as well as in insects ; but it is among the insects that we find them in the highest grade of development. In a typical aerial flying insect the spiracles are arranged segmentally in pairs along the sides of the body, while the tubes branch repeatedly, ending in minute ramifications with walls of exceeding delicacy which allow free gaseous exchange between the tissues and the atmosphere, oxygen being taken to support the combustion-processes while waste products — carbon dioxide and water- vapour — are given off. Such a flying insect, therefore, while it lives in the air and is bathed in it — using the aerial resistance for support and progress — is also permeated with air inwardly, illustrating as perhaps no other animal can do, the deeper meanings of that old-time INTRODUCTION 5 definition of living creatures : ** in which is the breath of Hfe." Yet, though insects as a group are typically aerial, they afford many and remarkable adaptations to life in water ; especially noteworthy are those members of the class — dragonflies and gnats for example — which pass the early or preparatory stages of their lives in streams or ponds, emerging into the upper air as they acquire the wings whose full development marks always the attainment by an insect of the adult or perfect state. The various modifications connected with such aquatic modes of life will be considered later. In turning from breathing to feeding, we fail to find a distinctive method of action for the class of insects generally. Indeed they are remarkable rather for the great variety displayed in the nature of their food and the means by which they procure it. With regard to the latter question, however, attention may suitably be drawn in this introductory sketch to a structural character common to insects and allied classes of the Arthropoda. This is the modification for feeding pur- poses of certain pairs of limbs that belong to those three or four segments of the body that make up the hinder region of the head (Figs. 2, By C, D ; 6, 7, 8, 9). These Hmbs become wonderfully adapted for testing or tasting, for biting or piercing, for licking up or sucking in the widely different substances — such as the tissues or fluids of live plants and animals, the decaying remains of dead organisms — from which insects of different type draw their food supply. One or two features in the digestive system of insects are also worthy of note. The outer skin with its secreted covering of cuticle is pushed inwards in such a way as to form the lining of two extensive tracts situated respectively at the front and hinder ends of the food canal ; these are the ** fore-gut " and *' hind-gut " of the arthropodous digestive tube. The " mid-gut," often comparatively restricted in length, is lined by a sheet of living cells uncoated with cuticle ; this cell-layer or epithelium elaborates the juices with their digestive ferments which act upon the 6 THE BIOLOGY OF INSECTS food (Figs. 10, ii). Such action may go on in the fore-gut or hind-gut also, by the transference thither of portions of food mingled with the juices, and the digested food-material may be absorbed through the thin cuticle lining those regions of the canal, and so traverse their wall, as well as the wall of the mid-gut, to pass into the blood for the general nourishment of the body. In most animals that have attained to a high or even moderate degree of organisation there is a circulating fluid or blood which serves as the medium of exchange between the general living tissues of the body and the organs of specialised function, bringing to the tissues food and oxygen, providing the raw material for their secretions, and taking up from them waste-substances to be carried to and eliminated by the organs of excretion. As a rule these exchanges are carried on while the blood flows through vessels minutely fine with exceedingly thin walls which afford opportunity for diffusion in either direction. But in insects, as among Arthropoda generally, the blood during much of its circulatory course flows through great spaces in the body, surrounding the digestive and other organs and bathing them on all sides ; thus instead of a typical body-cavity (coelom) avast blood-space (haemocoel) occupies most of an insect's inside. From this principal cavity where food-material is absorbed through the gut-wall, the blood passes up through a perforated membrane into a special blood-space — relatively long and broad but shallow — just beneath the dorsal body-wall where the narrow tubular heart is placed, gaining entrance to the heart by means of a series of paired slits (Fig. 12, h). The heart's rhythmical pulsations force the blood towards the front region of the insect where it passes from the system of closed tubes into the great blood-space already described. The elimination of nitrogenous w^aste-matters — among the most important end-products of the chemical changes (metabolism) always going on in a living body — is performed in insects by a set of organs highly characteristic of the class. These are elongate, narrow tubes (Fig. 10, Mt) which grow INTRODUCTION 7 out from the front end of the hind-gut and lie freely in the great blood-space that surrounds, as we have seen, the digestive system. The epithelium that forms the thin walls of these " Malpighian " tubes — so named in honour of Marcello Malpighi, the great seventeenth- century pioneer of insect anatomy — separates from the surrounding blood the waste matter in a state of solution, so that along the tubes it may pass into the intestine and out of the body. The working of the various systems thus briefly reviewed is co-ordinated by means of the nervous system, as ever the seat of general control of an animal's activities. An insect is essentially a segmented animal, and in each segment is present a pair of closely apposed nerve-centres or gangHa, usually situated just within the ventral body-wall. The ganglia of the successive segments are linked up by a pair of longitudinal nerve- trunks, and from each ganglion nerves pass to the muscles and other structures of its segment (Figs. 10, 12), so that each ganglion, while it controls the action of organs in its own segment, is capable of receiving nerve-impulses from or sending them to other segmental ganglia. But in insects there may often be observed a tendency for the ganglia of two or more successive segments to become fused together, resulting in an integration of the nerve-centres and a consequent centralisation of nervous control. Such integration is notably illustrated by the coalescence, in all insects, of the anterior three or four ganglia of the head to form a brain (Figs. 10, 6 ; 12, op), situated above or in front of the mouth, and linked to the ganglia behind it by paired nerve trunks passing one on either side of the gullet. From the brain the general activities of an insect are clearly directed, yet mutilation of the creature shows that considerable power of control resides in local centres, for the hinder region of the body may, after separation from the head, continue to move in a manner seemingly purposeful. In any animal a large number of its habitual motions may be shown to occur as definite responses to external stimulations of various kinds. Some irritable nerve-ending 8 THE BIOLOGY OF INSECTS receives an impression from the surroundings, resulting in the transmission of a nerve-impulse to the central system, its arrival at which may or may not result in a definite sensation, but will surely result in the transmission from the central system, to muscle-fibres or other tissue, of an impulse which will lead to suitable movement or other response — the whole operation aflFording an example of what physio- logists call a '' reflex." In insects and other arthropods the most characteristic feature of nervous action results from the fact that, the living skin being everywhere covered by the cuticle, impressions on the nervous system can be made only through some modified part of this cuticle. Although that envelope may be generally described as a protective armour defending its wearer from outside influences, yet it possesses hundreds or thousands of admirable modifica- tions adapted for the reception of stimuli. An insect's body or limb-segment often bears many " hairs " or ** bristles " — stiflF projections, each jointed to the general cuticular surface by a flexible basal region (Fig. i, A). Each hair is the secretion of a special cell of the skin, and if this cell in touch with the hair or bristle, be prolonged into a nerve-fibre running towards a ganglion, the hair is ** sensory " in function, the resulting sensation being com- parable to a sense such as touch, which we ourselves know by experience. Sensory hairs of this type are often especially numerous on certain appendages of the head — notably on the feelers, which are modified Hmbs belonging to one of the brain-segments, and on the jointed leg- like " palps " borne on some of those hinder head-limbs that sei-ve as jaws. Even the highly specialised senses of hearing and sight depend upon specialised cuticular areas. Thin, tense patches of the cuticle in many insects are capable of being thrown into vibration by impinging sound-waves, so that adjacent nerve-endings can be aflfected ; while in almost every insect the paired eyes, so prominent on the head, show a reticulated surface of transparent cuticular facets, each of which overlays a set of nerve- elements modi- fied from the skin and in connection by means of the INTRODUCTION 9 optic ganglia and nerve-fibres with the brain. For a -m on Fig. 3. — A, Spermatozoon of a Longhorn Beetle (Morimus) : h, head with nucleus ; w, middle-piece with centrosome ; /, flagellum. B, Ovum or Egg of a Midge (Chironomus), pr, protoplasmic layer ; y, yolk-globules ; on, egg-nucleus ; sn, sperm-nucleus (which has entered egg) ; pn, polar nucleus, separated from egg-nucleus. X 400. After Ballowitz and Ritter, Zeitschr.f. wissensch. Zool. 1, 1890. chemical stimulus to aff'ect nerve-cells so as to give rise to sensations like smell or taste, it is necessary that the cuticle 10 THE BIOLOGY OF INSECTS over the receptive organ should be exceedingly thin or perforated ; such sensory " pegs " or " pits " are found in large numbers on feelers and palps. Thus an insect's skeletal and protective cuticle is so modified as to ensure all needful correspondence with the outer world, while the high development of its central nervous system is correlated with behaviour often apparently intelligent. One further aspect of the life of insects remains for discussion in this brief preliminary sketch. All animals reproduce their kind, so that any living creature that may delight us as we watch its purposeful activity reaches always its full development through a process of growth and change from simple beginnings. The details of reproduction and growth among insects are of exceptional interest to the student. The two kinds of germ-cells or gametes — the small, motile sperms and the large passive eggs (Fig. 3) are developed, as is most often the case among animals, in two sets of individuals known respectively as males and females, which often present striking difference in their outward aspect and mode of behaviour. Many male insects when compared with their females afford examples of smaller size together with more highly elaborated sense-organs, brighter colours and greater activity ; it is not unusual, for example, to find flying male insects that have wingless mates, or chirping males that have silent females. The eggs of insects are relatively large with a liberal supply of food-material or yolk (Fig. 3, B, y). Fertilisation of the eggs may not follow immediately on pairing, as the sperms are received into a special female sperm-case (spermatheca) to be discharged when required as the eggs are laid. Not a few insects are derived from eggs never fertilised, pro- viding the best known examples among animals of virgin- reproduction (parthenogenesis). In the course of their growth, most insects undergo a remarkable process of change. The cuticle wherewith they are clothed, not being formed of living tissue, cannot grow and possesses only a limited capability of stretching ; hence it must, during the insect's growth, be periodically INTRODUCTION II renewed and shed, the casting or '' moult " of the exo- skeleton being called an ecdysis. Often the various instars (the forms assumed by an insect in the successive stages of its life-history) differ from one another, and this is, to some degree, inevitable, since, while the vast majority of insects are winged when adult ; no insect has wings when first hatched or born. The wingless young may be Fig. 4. — Successive instars of Meadow Plant-bug (Leptopferna dolobrata), Great Britain, from newly-hatched (A) to adult (F) ; in C, D, and E can be traced the growth of the wing-rudiments (tv). X 5. After A. Tullgren, 1919, and H. Osborn, jfourn. Agric. Res. (U.S.D.A.) XV, 1918. Strikingly Hke the winged adult as in the case of grass- hoppers and plant-bugs (Fig. 4), or to all outward seeming most unlike it, as may be seen by comparing a butterfly or moth with its caterpillar (Fig. 5), a bee with its grub, or a bluebottle fly with its maggot ; in such cases the transfor- mation exemplified in the insect's Hfe-history is profound. 12 THE BIOLOGY OF INSECTS And not only is this true as regards form ; it holds fre- quently also as to place of abode, method of feeding, life- relations in the wide aspect. By processes of gradual growth and, at times, of apparently sudden change is the winged creature moulded to the atmosphere in and around which it lives. A crawling larv^a dwelling in water, breath- ing dissolved air and biting solid food, may be transformed into a winged aerial being, which flits or poises itself with Fig. 5. — a, Owl Moth (Agrotis saiicia) Great Britain ; b, its caterpillar, side \'iew ; c, caterpillar rolled up ; d, caterpillar, dorsal view ; /, mass of eggs laid on twig. All natural size, e, single egg, X 20. From F. H. Chittenden (after Howard), Entoyn. Bull. 27, U.S.D.A. 1901. easy grace as it seeks for food in the nectar-stores of floral cups. Among the most fascinating aspects of the study of insects must be reckoned the modes of behaviour often associated with the function of breeding. The mother lays her eggs amid suitable surroundings, often on food substances which she herself, in her winged condition, never tastes, but which supply the nutriment needed by the newly hatched grub. In some cases great care is taken INTRODUCTION 13 of the eggs, or the young may be fed like the helpless nest- lings of birds. Thus family life of a kind is exemplified by many insects, and it is well known how, among the wasps, bees, and ants, for example, the size of the family becomes enormously increased so that the assemblage may not un- suitably be compared to a state with its " officers of sorts," and the nest to a city whose streets and habitations swarm with orderly and industrious crowds. Such " social " insects have from early times aroused the admiration of observant men who have sought to draw, for their own guidance, lessons of wisdom from the apparently intelligent behaviour of these small yet wonderfully organised fellow- creatures. Later we may have opportunity to discuss how far such comparisons between insectan and human societies may justifiably be carried. In bringing to a close our short survey of the field which it is proposed to cover in this volume, a summary — condensed and therefore necessarily technical in mode of expression — of the main features of insect structure in relation to life- conditions may perhaps be of service. Insects, then, are a class of the Arthropoda : segmented appendiculate animals, clothed with a chitinous cuticle which forms an exoskeleton giving attachment to striated muscle-fibres. The head, whose appendages are modified as feelers (one pair) and jaws (usually three pairs), is sharply distinct from the three-segmented thorax which bears six legs and usually four wings, the latter being dorso-lateral outgrowths of mesothorax and metathorax ; thus a variety of precise and rapid motions, including flight, may be possible. The breathing- organs are complex, branching air- tubes, Hned with spirally- thickened chitinous cuticle, facilitating direct gaseous exchange between tissues and atmosphere wherewith this tracheal system typically com- municates through a series of paired spiracles. The digestive tube has extensive anterior and posterior tracts lined with chitin. The perivisceral and pericardial spaces are haemocoels, the latter receiving blood through its 14 THE BIOLOGY OF INSECTS perforated floor and transmitting blood to the heart by means of paired ostia. The organs of nitrogenous excretion are elongate, tubular outgrowths of the hind-gut. The nervous system consists of a double chain of segmentally arranged ventral ganglia with longitudinal trunks connected with an anterior complex brain. Sensory nerve-endings are affected through modified cuticular areas or outgrowths. Insects are of separate sexes (dioecious), the female producing large eggs which may, in some cases, develop parthenogenetically ; secondary sexual characters are often conspicuous. Growth is necessarily accompanied by a series of ecdyses in which a less or greater degree of metamorphosis is connected with the acquisition of wings and other structures characteristic of the adult creature. CHAPTER II FEEDING AND BREATHING The diverse systems of organs that build up the body of any animal and the various functions that these perform are so closely inter-related with one another, that it is impossible to consider any one system or mode of activity entirely by itself. Yet for a detailed discussion of the biology or " life-knowledge " of a creature, it is necessary to attempt, in some kind of order, a survey of its various organs and their actions, though opinions may well differ as to what order is the most convenient and reasonable. In this discussion of the Biology of Insects it is proposed to begin with the functions of feeding and breathing and the important series of changes within the body associated with these familiar manifestations of life. All observers of nature have in mind a distinction, implied if not expressed, between living creatures and life- less objects. In this book it is assumed that such distinction is justified, and while questions about the ultimate meaning of life must be left to the philosopher, the visible manifesta- tions of life — the modes of behaviour of living creatures — are proper subjects of study for the naturalist. Among these, feeding is one of the most obvious, as well as one of great importance. In our introduction (p. lo above) we have noticed that not the whole body of an insect is alive ; we have dwelt, for example, on the distinction between the outer cuticle or exoskeleton, which is a horny, lifeless envelope, and the skin, a sheet of living cells beneath, by whose activity the cuticle is built up (Fig. i). The essential component of these cells, the substance 15 i6 THE BIOLOGY OF INSECTS of the creature that is truly alive, is known as protoplasm, the *' physical basis of life " as Huxley long ago called it. Protoplasm is a semi-fluid material composed of various elaborate nitrogenous chemical compounds known as proteins. These substances are of high complexity ; in addition to carbon, hydrogen and oxygen, nitrogen is always present as well as a small proportion of sulphur, and a protein is now regarded as " formed by the condensation of a number of molecules of various amino-acids " (W. M. Bayliss, 1924), different combinations of these '* building stones " being characteristic of different types of protein. The component molecules of protoplasm are in the colloidal state. To repair the waste which living protoplasm con- stantly undergoes, food in the form of protein must be obtained by the insect, and such food is found either in the tissues of plants, or in those of other insects or various animals which may serve as prey or *' hosts," or in waste or decaying organic substances. As mentioned in the introductory chapter, energy is constantly liberated in insects as in living creatures generally, becoming evident in movement and radiation of heat. The " fuel " needed to supply this energy is largely furnished by non-nitrogenous food-substances : carbohydrates such as starch and the sugars, and fats. The energy liberated in the activity of living tissues is due to the breaking down of chemically complex substances, either components or in- clusions of the cell-protoplasm. The renewal or rebuilding of these requires a constant supply of food material, which must be brought to the cells in a dissolved and absorbable condition. An insect or other animal is truly fed only as the living substance of its cells is fed, and the seizing and swallowing of food, which is what many unobservant persons understand by " feeding," is really only ingestion, the necessary prelude to the true feeding process. We may then suitably begin our study of the feeding of insects by considering some of the methods by which they procure and swallow or ingest their foodstuffs. A well- known feature characteristic of animals generally is the FEEDING AND BREATHINC? 17 possession of a digestive cavity v^ithin which the swallov^ed food-material undergoes the changes necessary to fit it for furnishing nourishment to the living tissues. This cavity may be regarded as a modified tube running from a forwardly situated opening, the mouth, to a hinder opening, the vent or anus, through which the useless remnants of the food are ejected. Insects, like many of the highly organised animals of various groups, have a definite head formed by the union of a number of primitive body segments, and in front of or beneath this head the mouth is situated. An insect's mouth is furnished with jaws for seizing, masticating, piercing, sucking, or otherwise dealing with suitable food-substances. The jaws are arranged in pairs, and it was long ago shown (Savign)^, 181 6) that they are modified appendages belonging to the series of paired jointed limbs characteristic of Arthropods generally. These jaw- limbs differ greatly in form in different groups of insects according as they are used for biting solid food material, or for piercing and sucking or licking up various fluids. We may conveniently introduce this study of insect jaws by examining those of a somewhat primitive type of biting insect such as a common earwig (Fig. 6). Below the face region (clypeus) of the earwig's head is hinged a median flap with straight lower edge and rounded margins. This is the upper Hp (labrum) which bounds the mouth in front ; it is to be regarded as part of the insect's head skeleton (Fig. 6, A). Just behind it and only in part hidden by it, lie the front jaws or mandibles (Fig. 6, B), stout, strong organs each consisting of a single finely moulded piece, the broad base articulating with the head skeleton by means of a knob-like condyle behind and a concave surface (ginglymus) in front, the outer edge evenly rounded, trending to the sharp, inwardly directed apical teeth, the inner edge approximately straight with a ridged grinding or molar area towards its base. The two man- dibles are arranged facing each other, they can be draw^n together by the action of adductor or apart by abductor muscles. When the mandibles are drawn together the c i8 THE BIOLOGY OF INSECTS apical teeth interlock and the molar areas are in contact. Thus pieces of leaves or blossoms can be seized and bitten up by the teeth, and then ground into small particles between the ridged molar surfaces which move over each other as the mandibles alternately are pulled inwards Fig. 6. — Mouth parts of Earwig (Forficula aiiricularia), A, labrum or upper lip (la) attached to face or clypeus (c), alongside base of feeler (/). B, left mandible (front view) : at, apical teeth ; w, molar (grinding) area ; c, condyle ; g, ginglymus ; ad, tendon of adductor and ar of abductor muscles. C, hypopharynx (h) with superlinguae (s) and supporting feet (p). D, left maxilla (back view) : c, cardo ; s, stipes ; /, lacinia ; g, galea ; p, palp. E, labium formed by conjoined hinder maxillae : gu, neck sclerite ; sm, submentum ; m, mentum ; gl, united galea and lacinia ; p, palp. X 50. towards the mouth by the slight contraction of their abductor muscles. FEEDING AND BREATHING 19 Behind the mandibles and placed farther apart from one another than they, we find a second pair of jaw-limbs, the maxillae (Fig. 65 D), wliich are somewhat complex in form. Each maxilla has a base composed of three pieces, two short ones arranged transversely forming a hinge (cardo) to which is attached a longitudinal axis (stipes), bearing an externally and forwardly directed jointed leg-like organ, the palp, and two lobes, of w^hich the outer (hood or galea) is evenly rounded and coated with soft hairs, while the inner (blade or lacinia) is provided with strongish apical teeth and a row of sharp, hard spines. A live insect with such maxillae as these may be seen often to test with the tips of its palps the surface over which it walks as though to " feel " whether the material is suitable for food. The maxillary hood serves as a cover for the blade with its sharp teeth or spines, and the blades are of service in further dividing the food, already roughly masticated by the mandibles, into finer particles. The earwig's mouth is closed behind by what is often termed its lower lip (Fig, 6, E) (labium). Examination of this organ soon convinces the student that it is composed of a pair of maxilla- like limbs joined together by their bases, for paired blades, hoods and palps, though smaller than their counter- parts of the maxillae, are evident, and the basal plates (mentum and submentum) of the labium can readily be compared with the fused hinges and axes of the maxillae. Situated between the two maxillae in the middle of the mouth is the tongue or hypopharynx (Fig. 6, C), a com- paratively soft and membranous organ, yet covered by cuticle beset with closely arranged hairs at the top, sup- ported by a pair of strong chitinous basal pieces, and having a pair of bristly lobes (superlinguae or paragnaths) attached to its front face. Beliind it — that is to say between it and the labium — opens the duct of the salivary glands, to be described later. The hairy tip of the tongue may be regarded as concerned with the appreciation of food taken into the mouth. Such jaws as these are adapted for seizing and masti- cating solid food which is passed on into the digestive canal 20 THE BIOLOGY OF INSECTS Fig. 7.— Mouth-parts of Breeze-fly (Therioplectes) A, labmrn- epipharynx. B, hypopharnyx {sr, salivary reservior) ; C, mandible. D maxilla (c, cardo ; s, stipes ; /, lacima ; g, galea ; p, palp), t, labium {la, labella with pseudotracheal channels for suction ; m, mentum). X 30. FEEDING AND BREATHING 21 in small lumps. Very different is the form of the jaws of insects that take their food in the liquid state. As an example we may consider a blood-sucking fly (Fig. 7) such as a breeze-fly (Tabanid). Here the median (unpaired) organs of the mouth — the labrum-epipharynx, and hypo- pharynx — are formidable dagger-like piercers (A, B) pro- jecting downwards from the head. The mandibles (C) are curved, with the base broad, the sharp edges tapering to a fine point, like the blade of a broadsword ; by the action of muscles these can be thrust out or pulled back. The maxillae (D) are straight, narrower than the mandibles, with their tips not only sharply pointed but armed with formid- able barbs so that when thrust into the skin of the animal whose blood is being sucked they hold firmly. Each maxilla appears to be composed of lacinia and galea closely united along their whole length. There is a short hairy palp (Fig. 7, D,^) consisting of a single elongate segment broad at the base and tapering to a point. The labium (E) is a thick leathery organ with conspicuous bi-lobed extremity along which run numerous channels strength- ened by chitinous transverse rings, through which the blood is draw^n into the hollow base of the labium, whose cavity leads onward to the mouth and gullet. Other insects that feed by suction may differ from the breeze-fly in the form and arrangement of their jaws and other mouth parts. Among bugs (Fig. 8) and their relations, such as " greenfly" for example, the labium is modified into a stiff jointed *' beak " (rostrum) with a groove extending along its front aspect. Over the base of this groove lies the short, acute, flexible lab rum, and within are found the mandibles and maxillae, slender, strong piercers, the tips of the former being barbed like those of the breeze-fly (Fig. 8, Mn, Ma). These piercers can be thrust out beyond the extremity of the beak, so as to puncture the tissues of a plant whence sap can be sucked, or of an animal whence blood or other nutritious fluid can be drawn. Such liquid food is sucked in through the exceedingly fine tubular channel (Fig. 8, B, 5c) formed by the concave inner surfaces of the piercers, and Fig, 8.— a, Jaws and mouth of Shield-bug (Tessaratoma) shown as dissected from the side with the labium (la) displaced backwards. X 22. B, Cross-section through piercing stylets of a shield-bug (Graphosonia). X 300. Mn, mandible ; Ma, maxilla ; c, clypeus ; Ir, labrum ; ph, phary^nx ; g, gullet ; p, salivary pump with muscle ; d, salivary duct ; scj suction canal, and ec, excretory canal between maxillae. After C. Bugnion, Arch. Zool. Exp. (5) vii, 191 1. FEEDING AND BREATHING 23 past the bases of these it is passed on into the mouth and gullet. Many sucking insects, however, depend upon liquids that can be reached without the necessity of piercing any plant or animal tissue. A common house-fly belongs to the same order as the breeze-fly just mentioned, and observant persons must notice how that familiar insect wanders over lumps of sugar in a bowl, applying to their surface the thickened tip of a leathery proboscis which appears to be hinged on beneath the head. This is a labium, corresponding closely with that of the breeze-fly, but with the bi-lobed tip (labella) broader, more elaborately formed and with more numerous channels. The house-fly has no mandibles and its maxillae are represented only by their basal regions and their palps ; the insect sucks its food on exposed surfaces of many kinds. Another type of sucking insect whose mode of feeding may often be observed is a butterfly or moth. Such a graceful creature is seen to rest on a flower-head or poise itself in front of a blossom, and unroll a deUcate flexible " trunk," which when not in use rests coiled up in a spiral beneath the head and between the prominent scaly labial palps. This trunk (Fig. 9) is composed of the elongate hoods or galeae of the maxillae, which, being grooved on their inner aspects, are modified into flexible half-pipes, provided with interlocking hooks or spines so that the pair of organs can be conjoined to form a tubular sucker whose tip can be stretched out to reach the nectar at the bases of the floral leaves. The butterfly's maxilla has no blade (lacinia) and its palp is reduced to a tiny scale-bearing process, while the mandibles are altogether wanting or represented only by minute vestiges. Thus, by means of the jaw-limbs and other feeding structures, variously modified, as has been seen, in diff"erent insects, the food is passed into the mouth and thence into the gullet or front region of the digestive tract (Fig. 10, oe). This, as well as the succeeding regions — the crop (cr) and the proventriculus (pv) — is derived from the fore-gut, an 24 THE BIOLOGY OF INSECTS inpushing of the skin at the mouth-region and consequently lined with a chitinous layer which is an extension of the outer cuticle. The details of structure of the gullet and Fig. 9. — Maxilla of an Owl Moth (Agrotis). c, cardo ; 5, stipes ; p, palp (vestigial) ; g, long flexible galea with groove (gr) on its inner face and coupling hooks and sensory organs towards its tip. X 50. crop differ — like those of the jaws — in insects that differ in the nature of their food. Thus in a cockroach, earwig, or beetle, which swallows smaller or larger solid lumps, the FEEDING AND BREATHING 25 narrow tubular gullet traverses the thorax and then, in the abdomen, opens out into a capacious ovoid crop in which a quantity of material can be stored awaiting digestion. In the bee the crop is similar in form and arrangement but relatively smaller (Fig. 10, cr.). A butterfly has the front end of the gullet expanded into a small spherical bulb situated in the head, this by the alternate expansion and contraction of its wall, induced by the action of suitably arranged muscles, sucks liquid nectar into its cavity and then propels it along the gullet into the crop. And the butter- fly's crop is not simply an expanded region in the course of the fore-gut, it is a sub-globular sac forming a lateral out- growth at the hinder end of the gullet, so that the liquid food may accumulate in it, and pass on later to the further regions of the digestive tube. In a two-winged fly of the common housefly or bluebottle type, the gullet is pro- longed backwards far beyond the front end of the abdomen, widening into the sub-globular or ovoid crop which, having no other outlet, serves in this case also as a reservoir for liquid food. The proventriculus (Fig. 10, pv) is the third or final section of the fore-gut. In a biting insect such as a cock- roach or beetle, this forms a short region of the digestive tract, sub-globular or hemispherical in shape, with the muscular coat of its wall very thick, and its internal chitinous lining raised into strong tooth-hke ridges which project into the cavity. Contraction of the muscle fibres tends to bring these teeth together in the mid-cavity of the organ which is frequently spoken of as a '* gizzard," under the assumption that its function is the crushing of solid food- material. Probably, however, the action of the proventi- culus is rather that of a strainer, preventing the passage of the food into the middle region of the food-canal until its treatment in the fore-gut has been brought to completion. In very many insects the hindmost region of the proventri- culus projects as a comparatively narrow tube into the front end of the stomach or mid-gut. This arrangement may be well seen in that familiar insect the hive-bee (Figs. 10, 11), 26 THE BIOLOGY OF INSECTS b Ph Fig. io. — Dorsal Dissection of Worker Honey Bee (Apis mellifica), showing Digestive and Nervous Systems, ph, pharynx ; oe, gullet ; cr, crop or " honey-stomach " ; pv, proventriculus ; v, ventriculus or chyle-stomach (mid-gut) ; il, ileum ; co, colon ; sg, salivary glands ; phg, pharyngeal glands ; Mt, Malpighian tubes ; (chyle stomach and intestines are displaced to left to expose the nerve-cords connecting the chain of ganglia (g) ; b, brain, the three simple eyes (ocelli) above it ; e, compound eye. X lo. Adapted from R. E. Snodgrass, Anatomy of Honey Bee (U.S.D.A.) 1910. FEEDING AND BREATHING 27 and a study of the working of the parts of this region of the food-canal shows clearly that the proventriculus acts as a strainer ; it was long ago named the " honey-stopper '* by the various writers on the anatomy of the hive-bee. Here the organ is quadrate in cross-section, lined with four prominent chitinous ridges which are separated by the contraction of longitudinal muscles or approximated by the contraction of circularly arranged fibres. This quadrate " stomach-mouth " can be seen to close and open rhythmic- ally when the digestive tube is slit up in a freshly killed bee. Pollen-grains are thus allowed to pass on into the stomach, the chitinous lining of the proventriculus bearing hair-Uke, backwardly directed outgrowths which prevent the return of solid particles into the crop. The liquid nectar or honey can, however, be forced in either direction, as the need for digestion, absorption, or storage may require. While in the crop, the food is mixed with saliva or spittle, secreted by the insect's salivary glands (Fig. 10, sg)y which lie on either side of the gullet and open by special tubes or ducts into the median tube which enters the mouth between the tongue and the labium. It is noteworthy that the salivary ducts, being outgrowths of the fore- gut, have a chitinous lining which is often spirally thickened as in an air tube. The cells of the glands contain large bent elongated bodies, the secretory capsules, which may be nuclear in their nature. The saliva is an alkaline fluid containing a diastatic ferment which acts on carbohydrate food materials. A reservoir for storing saliva between feeding-times is usually associated with these glands, which are very much larger in plant-eaters than in insects of prey. In bees they are remarkably numerous and complex, and their secretion is partly effective in transforming the nectar sucked from blossoms into honey, the cane sugar (sucrose) of the former being ** inverted " to levulose and glucose. But this action is mainly due to ferments or enzymes (sucrase) present in the gastric juice, which, formed in the middle gut or stomach, is passed forward into the crop. The nectar sucked by bees from flowers is stored in the crop, which is 28 THE BIOLOGY OF INSECTS therefore often termed the " honey-stomach," and after undergoing the digestive process therein, is regurgitated as honey for use in feeding the inmates of the hive or for deposit in the waxen chambers of the comb. From the proventricukis an insect's food passes on into Fig. II. — A, Crop (cr), proventriculus (pv), and chyle-stomach (v) of Honey Bee worker (Apis mellifica). B, The same with most of the crop- wall removed to expose *' stomach-mouth " (m). X lo. C, Longitudinal section through the same organs of a queen bee : 7no, stomach-mouth ; vCy proventricular valve ; Im, tm, longitudinal and transverse muscle- layers ; e, epithelium ; ct, cuticular lining ; g, gelatinous secretion whence peritrophic membrane is formed. X 40. After Snodgrass. the stomach (ventriculus or mid-gut (Figs. 10, 11, v), usually cylindrical in form. This is lined by a sheet of columnar cells, v^hich may have their free surfaces thickened to form an intima, or the cells may project as fine processes FEEDING AND BREATHING 29 into the cavity of the organ, but they never secrete a chitinous cuticle. This sheet of cells (or epithelium) is often thrown into ridges or prominences that project on the free, inner surface of the stomach, with intervening furrows or depressions (Fig. 11). Outside is a com- paratively thin m.uscular coat, the rhythmic contraction of whose fibres works the contained food-mass through the organ. At the front end of the stomach there are usually blind outgrowths — the pyloric caeca — which in different insects assume the form of elongate tubes or of comparatively short pouches. The cells which line these tubes or pouches — formed of necessity as an extension of the stomach epithelium — secrete the gastric juice which, besides diastatic ferments, contains a proteoclastic ferment whose function is to act on the protein constituents of the food. Gastric juice may also be formed in the pits or depressions already mentioned as often numerous on the lining of the stomach. In many insects the food contents of the stomach are not in direct contact with its wall. They form a rod-hke mass traversing the stomach from end to end, and sur- rounded by a delicate coat — the peritrophic membrane. The stomach contents of insects are alkaline, their reaction thus offering a contrast to the acid nature of the stomach contents of a vertebrate animal, and suggesting that the insectan gastric juice is analogous to the pancreatic rather than to the gastric juice of a vertebrate. The lining cells of the stomach that are glandular in function, Hberate their secretion into the cavity of the organ. The secretion is usually a fluid poured out over the general surface of the epithelium ; but in certain fly-larvae and other insects A. van Gehuchten (1890) and other investi- gators have shown that small bladder-like processes grow out from the stomach cells and become constricted off so as to float freely in the cavity ; these can convey the digestive juice to the central mass. F. W. Cragg has shown (1920) that the cells lining the stomach of blood-sucking breeze- flies (Tabanus) throw off their secretion as a mass of globules which become broken up. As the secreting cells wear out 30 THE BIOLOGY OF INSECTS they are replaced by basal cells which grow towards the surface and regenerate the lining layer. In the stomach cavity of bees a large number of small spherical clear, cell- like bodies can be distinguished, as Snodgrass (1910) and other observers have described ; these have a similar origin and function. The peritrophic membrane is beheved to be formed in many insects by the envelopes of such cells, thrown off by the epithelium ; but in other cases it appears to be due to a secretion (Fig. 11,^) formed in successive layers by the cells of the stomach, as described by R. E. Snodgrass (1925) for the Honey Bee. Brought thus into close touch with the central food- mass, the digestive juice mixes with the nutrient substances and the proteoclastic ferment acts on the proteins, breaking them up into their constituent amino-acids. In many plant-eating insects it has been shown that diastatic ferments in the juice act on the starch of the food converting it finally into a sugar (monosaccharide) with a relatively small molecule. In the stomach also it appears that fatty con- stituents of the food of many insects are emulsified, at least in part, and hydrolysed by the action of special (lipoclastic) enzymes, giving rise to glycerol and fatt}^ acids. Thus the various nutrient substances are reduced as regards the complexity of their chemical composition and prepared for absorption in the soluble state by the living cells that build up the insect's organs and tissues. This, as has already been emphasised (p. 16), must be regarded as the true feeding process. Absorption of the digested food-substances is carried on through the wall of the stomach. H. Jordan (19 1 3) proved that various insects, with whose food iron- compounds had been mixed, showed the presence of iron in certain cells of the stomach epithelium. The com- parative thinness of the stomach-wall would suggest that there absorption must be actively carried on. But there seems no doubt that in the terminal region of the food- canal — the hind-gut as it is called — absorption also takes place. The hind-gut or intestine of an insect has, like the fore- FEEDING AND BREATHING 31 gut, its inner sheet of cells formed by inpushing of the outer skin, and therefore lined with a chitinous sheet that is an extension of the cuticle. The hind-gut is readily divisible into several regions. The front portion is usually a cylindrical tube of narrower diameter than the stomach, and a section immediately behind the stomach may be distinguished — in a bee (Fig. 10, il), for example — as a small intestine or ileum. From the extreme front end of this are given off the elongate Malpighian or excretory tubes (Fig. 10, Mt), already briefly mentioned (p. 7), whose form and function will be discussed more fully later (pp. 38- 39). The food-canal widens behind into the large in- testine or colon (co). In these regions the cellular layer (epithelium) is regular and distinct, its inner surface pro- tected by a delicate but definite chitinous Hning, while the muscular coat is relatively thick, the contractions of its circularly arranged fibres driving on the contained food still undergoing digestion. The terminal portion of the hind- gut is the rectum, usually a capacious, bladder- like organ (Fig. 10, r) with relatively thin walls, often thrown into longitudinal folds that form ridges and furrov/s. Here the epithelial layer is largely degenerate, the cell boundaries being indistinct except in certain regularly arranged blind tubes or pouches — the so-called rectal glands — whose lining cells are large and columnar. The secretion of these glands is apparently not digestive but auxiliary to the ejection of the unusable residue of the food. Despite the cuticular lining of the hind-gut there seems to be no doubt that food-absorption takes place through its walls. S. Metalnikoff (1896) states that the cuticle of the cockroach's large intestine is porous and that the epithelium cells absorb particles of iron administered in the insect's food. In hive-bees the rectum often contains a quantity of pollen-grains, which constitute the source of the nitro- genous and fatty food-supply of these insects ; only on reaching the rectum is the digestion of much of the pollen concluded, so the high probability of the absorption of the digestive products there may be inferred. 32 THE BIOLOGY OF INSECTS The rectum opens at the vent (or anus) on the hindmost segment of the insect's abdomen ; through this the rejected remains of the ingested foodstuffs are passed out. The nature of this excrement varies necessarily with the insect's diet. Its aspect often proclaims the kind of solid food- materials that have been devoured ; the little pellets passed by a leaf-eating caterpillar are green and show the micro- scopical characters of leaf-tissue, while the ** frass " ejected by a wood-eating larva has the aspect of fine sawdust. Curious and interesting is the fact that the Uquid excrement of such sucking insects as greenfly, still containing a pro- portion of available carbohydrate, serves as an acceptable food to many kinds of ants which follow and tend the aphids in order to obtain it. Many insects while in the larval state pass no excrement from the food-canal. In the '' ant- lion " grubs of lace- wing flies the vent is closed, while in the grubs of wasps, bees, and ants there is no outlet from the stomach into the hind-gut until the close of larval life. On the other hand, there are many insects — for example mayflies, silk-moths, and botflies — which in the adult condition take no food at all. Their jaws are reduced and useless and their mouths closed, but the various regions of their food-canals are developed like those of their relations that feed until the end of their Hves. The con- sideration of details as to the immense variety of methods of feeding among insects must, however, be postponed until later chapters deaUng with their general habits and mode of Ufe. We have seen that the digested and mostly soluble food- constituents are absorbed by the living cells that form the epithehal Hning of the food-canal. Thence they are conveyed to all other tissues of the body in the circulating fluid well known as the blood. The blood and the circu- latory system therefore next demand our attention. In most insects the blood is colourless, consisting of a watery plasma with salts, sugar, proteins, and amino-acids in solution, in which float small cells with distinct nuclei and often a tendency to change periodically their shape, FEEDING AND BREATHING 33 Fig. 12. — Lateral Dissection of Worker Honey Bee (Apis mellifica), showing Nerve-cord (nc) and Circulatory and Respiratory Systems. ts, tracheal sacs connected with air-tubes, h, heart, with its paired slits (ostia) oSy surrounded by pericardial blood-space {pc) ; dd, dorsal and W, ventral diaphragms ; a, aorta.; o^, optic lobe of brain ; mJ, mandible. X lo. After R. E. Snodgrass, 1910. 34 THE BIOLOGY OF INSECTS resembling thus the leucocytes or " white corpuscles " of the blood of vertebrates. When, as is sometimes the case, an insect's blood is coloured yellow, green, or even red, the pigment is dissolved in the plasma, not, as in the red blood of vertebrates, concentrated in the specialised cells known as *' coloured corpuscles." The blood, being the circulatory agent in the body, must be kept moving in a constant and regular flow. The pro- pellant organ for this circulation is the heart (Fig. 12, h), which has already (p. 6) been briefly described as a narrow tube running fore and aft in the middle axis of the insect, just beneath the dorsal body-w^all, while below it a deUcate membrane is stretched from side to side, forming the floor of the shallow pericardial chamber (Fig. 12, pc) in which the heart lies. It will be remembered that this pericardial cavity is a blood-space. The heart is formed of a variable number of chambers arranged one behind the other in a series, agreeing more or less with the segmentation or serial repetition of similar parts that characterises the body of an insect or other arthropod. But while in a cockroach the heart extends the whole length of the body, with thirteen chambers corresponding to the thrte ' thoracic and ten abdominal segments of the insect, in a bee there are only four chambers all situated in the abdomen, while in a few insect types only a single chamber can be recognised. The wall of the heart is muscular, its hinder end is closed and the blood driven through the successive chambers towards the head, the wave of muscular contraction passing on along the tube from behind forwards. Each chamber is some- what wider at its hinder than at its front end, where the constricted portion is provided with a valve that allows the blood to flow forward but not to return towards the tail. Into the hinder widened end of each chamber open a pair of minute slits (ostia) provided with valves which allow the blood to pass into the heart from the surrounding peri- cardial blood-space, but not in the reverse direction. As, therefore, the rhythmic contractions of the tubular heart drive the blood constantly forward towards the head, a FEEDING AND BREATHING 35 steady flow passes into each chamber of the heart from either side of the shallow blood-space in which it lies. The front end of the heart passes into a narrow median tube, the aorta or main artery of the body (Fig. 12, a). Thus there is beneath the dorsal wall of an insect, a con- tinuous longitudinal blood-vessel, of which the hinder region is heart and the front portion aorta. The relative lengths of these two portions differ in different insects ; from what has already been stated it may be realised that while in a cockroach the aorta begins at the front end of the thorax, in a bee it begins towards the front end of the abdomen. Through the aorta therefore the blood is pro- pelled forsvards into the head. Now, in back-boned animals it is well known that the blood passes from the aorta into a well- developed system of branching arteries and then circulates through a network of minute vessels, the capil- laries (through whose excessively thin walls exchange of substances can go on with the fluid lymph that bathes the tissues), and is returned to the heart by a set of veins. But in insects there is no such *' closed " circulatory system. Here the blood streams out from the open front end of the aorta, bathes the brain, passes backwards through the thorax surrounding the large muscles that move the legs and wings, and flows into the abdomen, whose cavity has already been defined as a great blood-space surrounding the digestive tube. In many insects there is in the abdomen a definite ventral blood-space, in which the nerve cord lies ; this space is roofed by a delicate diaphragm whose con- tractions drive the blood upwards into the main cavity of the body. This cavity having the ventral diaphragm just mentioned as its floor, is roofed by the thin membrane already described as the floor of the pericardial space. Into this latter the blood streams from the large main cavity below, passing in some insects — cockroaches, for example — through minute holes or pores in the membrane, in others — such as beetles and bees — around the segmentally scalloped edges of the membrane. From the pericardial space, as already mentioned, the blood flows back into the 36 THE BIOLOGY OF INSECTS heart through the paired slits in its wall, and thus the course of the circulation is complete. In insects, as in the whole great group of the Arthropoda, the blood, streaming through the spacious cavities of the body, comes into direct contact with the tissues. Thus the absorbed products of digestion can diffuse through the wall of the stomach and intestine, so as to pass in solution into the blood. And as the blood directly bathes all the living tissues of the body it supplies continually to the celle whether of muscle or nerve, gland or skin, germ or tube- wall — the materials w^hich they need for building up the complex proteins that compose their living protoplasm, and the fuel that is required for the support of combustion and the Hberation of energy. There is reason to believe, from, analog}^ with what is now known to occur in verte- brates (Bayliss, 1924) that the proteins of the cell are built up of the amino-acids carried in the blood, so that in the economy of the living body the analytic action of the digestive organs is succeeded by a synthesis of protein. The carbohydrate " fuel-food " is absorbed by the intestinal wall, circulates in the blood, and is suppHed to the tissues in the form of monosaccharide. Insects have no special organ comparable to the liver of vertebrates and molluscs in v/hich the monosaccharide sugar is transformed into polysaccharide glycogen and so stored up, to be drawn on as required and reconverted into sugar circulating in the blood. But in many insects glycogen probably forms a carbohydrate store, and this may especially be expected to occur in larvae preparing for the great expenditure of energy involved in transformation. J. Straus has shown (191 1) that nearly a third by weight of the substance of bee-grubs after dessication consists of glycogen. Stored in the cells of the fat-body, the glycogen is utilised for the reconstruc- tion of tissue that marks the pupal period during which it is retransformed into sugar and disappears. Fat is absorbed in the form of an emulsion or dissociated into its component acid and glycerol, to be afterwards re-synthetised so that the minute fat-globules are passed FEEDING AND BREATHING 37 into the blood, and serve along with the carbohydrates as *' fuel-food." In very many insects comparatively large masses of fatty tissue lie around the food canal in the great blood-space, or on either side of the heart in the pericardial wall. Such a *' fat-body " may reasonably be regarded as a reserve of lipoid food-material on which the organism can draw for the needs of its life-activities. The fat-body of a mature larva is often relatively enormous, in view of the coming transformation. Besides fat-globules the cells of an insect " fat-body " contain protein granules, that serve as a store of nitrogenous food. It has been shown by G. A. Koschevnikov (1900) and others, by means of experi- ments on larvae provided with coloured foodstuffs, that the store of fat may be elaborated from nitrogenous food as well as from the sugar (carbohydrate) present in honey. Thus while much of the material on which an insect feeds goes rapidly to repair its constantly wasting living substance, or to supply the energy needed for its activities, much is stored in forms convenient for future utilisation. The transformations of energy that continually go on in the insect's tissue depend therefore on chemical reactions between the cells themselves. The contraction of muscle fibres, and the more delicate and obscure physical changes concerned in the production of the nerve-impulses to be discussed in some detail in our next two chapters, are due to the liberation of energy resulting from chemical dissocia- tion. It has been seen how these cells are supplied by the blood with the necessary materials for building protoplasm and with a constant stream of readily available fuel. These supplies themselves depend on the digestive processes carried on in the food- canal and due to the action of the digestive ferments or enzymes on the ingested food-sub- stances. The formation of the digestive juices is the work of the cells of the salivary and gastric glands. These cells are, like all other cells of the body, dependent for their " raw material " on the substances supplied to them by the blood, whence each gland extracts constituents which it needs, and elaborates them into its own appropriate secretion. 38 THE BIOLOGY OF INSECTS We are compelled to regard every cell — whatever be its specific function — as a minute but marvellous laboratory in which the characteristic chemical changes are continually wrought through the influence of unknown ferments formed within the cell, '* intracellular enzymes " as they have appropriately been called. The chemical processes thus briefly reviewed are neces- sarily accompanied by the production of waste substances which must be eliminated from the body ; thus we are led to consider the subject of excretion. The proteins of the living tissue, continually built up from the materials supplied in the blood, are continually broken down through the activities of the cells, and thus nitrogenous waste-products — of simple chemical composition when compared with the im- mensely complex proteins — are set free. Just as the blood resigns to the tissue- cells the needed food-ma- terials, so it takes from them the eflFete waste-sub- stances, carrying these to those organs of excretion whose function it is to eliminate them from the body. In most of the great groups of animals the organs of nitrogenous excretion are essentially tubular in structure, consisting of tubes or groups of tubules around whose walls blood may circulate — whether flowing through capillary vessels or freely bathing the cells — and into whose cavities the waste products may be passed, as the tubular excretory systems lead directly or indirectly to some outward opening of the body. In insects, as has already been mentioned, the organs of nitrogenous excretion are the Malpighian or kidney- tubes (Fig. lo, Mt) which grow out from the front Fig. 13. — Malpighian Tube of Honey Bee (Apis mellifica) shown in obHque cross-section, e, epithe- lium ; b, basement membrane ; tr, small tracheal tube with fine branches on surface of Malpighian tube. X 450. FEEDING AND BREATHING 39 ends of the hind-gut into the great blood-space. When a tube is examined microscopically its wall (Fig. 13) is seen to be composed of a single layer of glandular cells with a thin basement membrane surrounded in some insects — as shown by L. Leger and A. Dusboscq (1899) and by L. Eastham (1925) — by delicate muscular fibres, which by their contraction may assist the passage of the excretion towards the intestine. The tubes lie freely in the great blood-space, floating as it were in the blood currents and capable themselves of bending and extending movements. Chemical examination shows that the tubes contain such compounds as oxalates and urates besides amino-acids such as leucin. These substances — resulting from the disinte- gration of protein or due to excess of nutrient materials in the blood — are extracted from the blood by the cells of the tubules or elaborated by these cells from other related substances held by the blood in solution. An interesting feature in these excretory tubes of insects is the variation in their number. Moths and butterflies have but two, beetles four or six, produced apparently by a branching of the original pair, while cockroaches, grasshoppers, bees and w^asps have a large number, in some cases over a hundred, also derived by elaboration from the simpler condition. In various insect larvae as well as in some wingless springtails, the fat-body may contain concretions of urates, the waste material being thus segregated and stored, but not actually removed from the body. More than once it has been pointed out that the Hbera- tion of energy required for the performance of an animal's activities is connected with a combustion process constantly going on in the tissues. For the support of this com- bustion a supply of oxygen is necessary, while the products of combustion — water-vapour and carbon dioxide — must be ehminated from the body. This important work is the function of the breathing organs or respiratory system. The remarkably characteristic form and mode of action of the breathing organs in insects have already (pp. 4-5) been very briefly described. In the vast majority of animals the 40 THE BIOLOGY OF INSECTS blood acts as a carrier of oxygen to the tissues and of com- bustion-products from them, fresh oxygen being obtained and the waste water vapour and carbon dioxide given up, as the blood passes through the fine vessels or passages of the breathing organs, such as the lungs of terrestrial and aerial creatures, or the gills of aquatic animals that are dependent for their oxygen-supply on the air dissolved in the surrounding water. In all typical insects breathing is carried on by means of a set of branching air-tubes, a tracheal system ; the fine, thin-walled terminal branchlets lead into minute tracheoles whose deHcate walls are in closest contact with the various organs and tissues of the body. Thus gaseous exchange is effected directly between the insect's Hving substance and the air contained in the tracheoles, the oxygen passing in from the atmosphere and the carbon dioxide and water vapour passing out to it by means of a set of diffusion processes accompanying alternate intakes and expulsions of air. The air- tubes of an insect (Fig. 13, tr) are, as already mentioned, due to ingrowths of the skin, and are naturally therefore lined with an extension of the chitinous cuticle. It is surely suggestive that the oxygen of the free air which surrounds the insect's body should be brought into touch with the creature's inner tissues by a series of actual ingrowths of the skin with its overlying cuticle, each ingrowth dividing and branching repeatedly, pushing thus its way ever more deeply among the masses of living cells which vary greatly in form and function but are all aHke hungry for oxygen. For the greater part of their course an insect's air-tubes have their chitinous lining thickened spirally, so that when broken the wall of the tube shows the appearance of a par- tially unwound thread (Fig. 13, tr). Below the firm lining is the epitheHum or cellular layer which is the continuation of the skin ; outside this is a thin, supporting '* basement membrane." The spirally thickened chitinous lining of the air-tube has an important bearing on the function of breathing. It gives such firmness to the tube-wall that this cannot collapse, yet it yields to some extent to the pressure FEEDING AND BREATHING 41 Fig 14— Magnified Diagram of the Tracheal System in a young Termite, d, dorsal, sp, spiracular, and v, ventral longitudmal trunks ; c, commissure ; vs, visceral branches ; a, antennal branch, i, 2, 3 (on left side), air-tubes of legs, i, ii, thoracic spiracles ; 1-8 M right side) abdominal spiracles. Adapted from C. Fuller, Ann. Natal Mus. iv, 1919. 43 THE BIOLOGY OF INSECTS of surrounding tissues so that the calibre of the tube must be reduced when the insect's body contracts, while there is elasticity in the Uning which ensures expansion of the tube again when the pressure is withdrawn. The fine air- tubes lead into the minute tracheoles, through whose exceedingly thin walls the gaseous exchanges between air and tissues are carried on. Each tracheole (Fig. 46, tr) arises as a cavity in a cell of the tracheal epithelium, ultimately uniting with the cavity of the tube, the cell growing out as an elongate hollow thread. The insect's breathing system is adapted for bringing air-currents alternately into and out of these tracheoles, and the mechanism of the process is of great interest. The air-tube system may be regarded as growing inwards from the paired series of openings (spiracles or stigmata, Figs. 14, 15) that are found on the sides of most of the body- segments in the great majority of insects ; these spiracles indicate indeed where the air-tubes began to grow into the body from the outer skin (ectoderm) during the embryonic growth of the insect (see p. 154). Spiracles are commonly present on one or two of the thoracic and eight of the abdominal segments. In some primitive wingless insects — bristle-tails — the branching system of air-tubes arising from each spiracle remains distinct from all the others, but usually large longitudinal trunks run along the body connect- ing the successive spiracular tubes ; more slender longi- tudinal trunks run dorsally along either side of the heart and ventrally along either side of the nerve-cord, while transverse commissures link up the right and left trunks so that the whole tracheal system forms a complex network. Each spiracle is surrounded by a rim of strong thick chitin, and the aperture is often guarded by a series of inwardly directed hair-like or spine-like processes which hinder the access of foreign bodies to the system. Just within the spiracle is a valve wherein by means of a specially thickened chitinous *' bow," or a lever operated by suitable muscles the cavity of the spiracular tube can be closed (Fig. 15). The breath- ing of insects was well described by F. Plateau (1884) FEEDING AND BREATHING 43 and L. C. Miall (1886). The rhythmic contraction and relaxation of the spiracular muscles and the elasticity of their chitinous attachments cause the spiracular valves to open and close alternately. The contraction of the abdominal muscles, the details of whose arrangement vary in different groups of insects, reduces the capacity of the abdomen so that the organs press on the air-tubes ; when the abdominal muscles are relaxed, the maximum capacity is restored through the elasticity of the body- walls. Thus it comes to pass that with the expansion of the body a small volume of fresh air passes in through the open spiracles to supplement the residual air always present in the tubes. When the spiracular vahv^es close and the body contracts so that pressure is exerted on the tubes, air is forced into the tracheoles. The opening of the spiracles while the body is still contracting ensures the expiration of a certain quantity of air, to be followed— when the abdominal muscles relax— by the inspiratory action already explained. The passage in and out of the inspired and expired air-current is accompanied by a twofold diffusion of gaseous constituents, oxygen passing from the fresh air wherein its tension is high to the tracheoles and tissues with their low oxygen tension, while the excess of carbon dioxide diffuses from the tracheoles to pass out of the spiracles with the expiratory air-currents into the atmo- sphere with its low tension of carbon dioxide. Fig. 15. — Abdominal spiracle and its connec- tion with the tracheal system in a Louse {Haematopinus) . s, spiracle ; v, vestibule ; /, lever ; cm, closing muscle ; om, opening muscle ; st, spiracular trachea ; t, part of longitudinal tracheal trunk. X 200. 44 THE BIOLOGY OF INSECTS The combustion processes that go on in the tissues of insects tend to maintain a fairly high body temperature. This becomes evident when a number of active insects are crowded together ; the summer temperature of a beehive is over 90" F. and the winter temperature nearly 80"^. Experiments on the gaseous exchanges of a number of bees whose thoracic air-tubes were choked with small parasitic mites {Acarapis, see J. Rennie, 1921) show that the output of carbon dioxide is in such cases very much below the normal, while the wing- muscles, largely deprived of their needed oxygen, are often incapable of normal action so that the insects cannot fly. In insects of feeble or occasional flight, such as cock- roaches, the whole tracheal system may be described as tubular ; but in many insects of great activity and continual and powerful flight, capacious air-sacs of rotund or oval shape, are developed as enlargements of the main trunks or branches. Grasshoppers, flying beetles such as chafers, dragonflies, and bees are examples of insects with extensive air-sacs (Fig. 12, ts). The walls of these expansions are without the spiral thickening of the Hning that charac- terises the tracheal system generally. It may be concluded that they serve as reservoirs of air from which the finer branches may be replenished. Many and great modifications of the typical insectan breathing system are found to be correlated with special modes of hfe, particularly in the immature stages of those insects which pass through marked transformation in their life-history. These will be more appropriately considered in later chapters. But in this general account of the function of breathing an introductory reference is necessary to the means whereby insects living submerged in water obtain their supply of air. Many adult insects that frequently dive and swim under water — some aquatic beetles and bugs for example — carry down with them a supply of air, sealed beneath their firm forewings or as a bubble surrounding some part of the body giving access to the spiracles. Some aquatic larvae — like the familiar gnat-grub — have their hind- FEEDING AND BREATHING 45 most spiracles alone open, and these are situated on some outgrowth which can be thrust through the surface film of the water so as to make temporary contact with the upper air. But in other aquatic insect-larvae there is provision for using the air dissolved in the water by means of gills. In the grubs of mayflies and the slender dragon-flies (" damsel-flies ") one or more pairs of abdominal Hmbs — tubular or flattened in form with delicate cuticle and con- taining a network of branching air-tubes — serve for the passage of oxygen from the dissolved air into the tracheal system of the insect. These organs are distinguished as tracheal gills. In the larvae of caddis-flies and many midges there are delicate hollow finger- like outgrowths on the abdomen ; the cavities of these are prolongations of the great blood-space of the body and they are therefore known as blood-gills. In this case the blood-currents pass in and out of these gills, so that the blood serves as a carrier of oxygen to and of carbon dioxide from the tissues, as is usual in animals generally. The well-known '' blood- worm '* larva of the midge Chironomus (Miall and Hammond, 1900) has difl"used through its blood-plasma the same respiratory pigment (haemoglobin, a compound containing protein and iron) that is characteristic of vertebrates. These midge-grubs spend much of their time burrowing in mud, and the affinity of the haemoglobin for oxygen, which it holds in loose combination, enables it to retain a store of that *' vital " gas which can be renewed when the slender grubs rise towards the surface of their native pond, where the water is fairly well aerated. Some midges of the Chironomus family have, however, larvae without any air- tube system at all, and there are many aquatic grubs in which the special breathing organs become well- developed only late in the process of growth. Such are able to effect gaseous exchanges through the thin and delicate body- wall, as many worms do. Their small size and excessively thin cuticle render any respiratory system needless. The same condition is found throughout life, in most members of the order of wingless insects known as 46 THE BIOLOGY OF INSECTS springtails (Collembola). These live mostjjr in sheltered situations, under bark, in soil, beneath stones, where they can effect directly through the skin and cuticle gaseous exchange with the surrounding moist air. The whole body in such cases appears to act as a blood-gill. But we have still much to learn as to the method by which these exchanges are brought about. Enough has perhaps been stated to suggest how the living insect builds its tissues and obtains its sources of energy out of the materials supplied in its food, how it draws directly or indirectly on the atmosphere for its needed oxygen, and how it gives back to its surroundings the com- paratively simple waste substances such as carbon dioxide, water, urates — the end-products of those internal changes that are a necessary accompaniment of its life. CHAPTER III MOVEMENT One of the facts most easily to be noticed by the observer of living insects is that the vast majority of these creatures are constantly moving. According to its kind, its inborn habits, or its impressions received from the outside world, an insect walks or runs, creeps or swims, jumps or flies. These actions if carefully watched are seen to result from the beautifully co-ordinated movements of various parts of the exoskeleton. For example, when a beetle walks three of its six legs — the front and hind legs of one side with the intermediate leg of the other — are lifted and carried forward ; then the feet of these three limbs press the ground, and the other three legs are lifted and moved forward in their turn. If the movements of a single leg are watched it will be seen that the long stout segment or thigh is moved in relation to the body and that the adjoining more slender segment or shin moves in relation to the thigh, the angle between thigh and shin being alternately acute (when the knee joint is bent or flexed) and obtuse (when the knee joint is straightened or extended). Such visible movements on the part of an insect are due to the longitudinal contraction of the specialised tissue that forms the muscles which, situated inside the legs or other parts of the exoskeleton, are so attached to these that by their power of longitudinal contractiHty they can move the parts in relation to each other. The fibrous structure of insect-muscles is readily seen on examination with a hand-lens, and a small fragment of such muscle, suitably treated, shows that it is composed of an enormous number of 47 48 THE BIOLOGY OF INSECTS fibres — slender contractile threads extending in the direction of the length of the muscle, which may thus be considered as composed of many bundles of fibres. It has already been mentioned (p. 2), that the muscles of insects resemble the skeletal muscles of vertebrates in being composed of striped or striated fibres (Fig. 16). With a comparatively low degree of magnification it is seen that these fibres exhibit alternating darker and lighter transverse bands ; they are striated. On studying the structure of a single fibre with a high degree of magnification, it is seen in places to split lengthwise into Fig. 16. — Microscopic structure of mandibular Muscle of Hornet ( Fespa crahro). A, part of fibre treated with potash ; X 425. B, the same stained with haematoxylin. Sarc, sarcolemma ; Id., Dobie's Hne ; da, thickenings of fibrillae ; T, air-tube ; N, nerve end-plate. C, partial cross-section at Dobie's line showing radiating fibrils ; D, partial cross-section through median region of a sarcomere showing fibrillae and nuclei (nuc), X 850. From C. Janet, Etudes sur les Fourmis les Guepes et les Abeilles, xii, 1895. excessively fine threads or fibrils, while the cross-striation becomes so definite as to suggest that the fibre is built up of an innumerable series of discs arranged one on the other ; a very rough model of its apparent structure might be made with a pile of pennies and half-crowns arranged alternately, and such transverse dissociation of a fibre can be brought about by treatment with gold chloride. But the fibre never spHts naturally into discs as it does into fibrils, and the alternating darker and lighter transverse regions are indica- tions of physical differentiation in the substance of the fibre. MOVEMENT 49 The muscle substance is essentially protoplasm endued with the special power of contracting. Each fibre is bounded by a delicate structureless sheath, the sarcolemma (Fig. 16), inside which nuclei can be distinguished ; the fibre has therefore been formed by the coalescence of a large number of cells whose boundaries can no longer be detected. The striation of the fibres is comparatively coarse in the muscles of many insects and the structure has been minutely studied by various observers. According to the researches of C. Janet (1895) and E. Schafer (i89i)on the wing-muscles of ants, wasps, and beetles, each light stripe is traversed by a fine line (Dobie's line or Krause's ** membrane " (Fig. 16, Id)) which sometimes appears broken (a " dotted line "). The portion of a fibril between two such lines is a sarcomere, and the central dark region of the sarcomere appears as a band or disc ; the dark substance or sarcous element has a striped aspect because it is traversed by two sets of longitudinal pores which are open towards Dobie's line in either direction but closed towards the median plane of the sarcomere. The substance of the sarcous element, as well as the pores, becomes shorter and broader when the fibril contracts, longer and narrower when it relaxes ; in the latter condition a clear transverse line (Hensen's line) may be distinguished in the median plane of the sarcomere. The paler and presumably more fluid constituent (hyaloplasm) of the sarcomere is largely absorbed in the pores during con- traction and squeezed out of them during relaxation. An in- terstitial network with radially arranged filaments (Fig. 16, c) has been described in the neighbourhood of Dobie's line. The arrangement of the various constituents is such that contraction can take effect only in the longitudinal direction of the fibre. All the fibrils contract together and the amount of contraction of each fibril may be regarded as the sum of the contractions of the sarcomeres that compose it. Not that all the sarcomeres in a fibril contract simultaneously ; a wave of contraction whose velocity can be measured passes along each fibre, and therefore along the whole muscle from end to end. The muscles of insects are, as we have already E 50 THE BIOLOGY OF INSECTS seen, surrounded by blood which brings nutrient substances to the active tissue and removes waste products from it, and supplied with numerous fine branches (tracheoles) of the air-tube system through whose delicate walls oxygen passes into the muscle for support of the combustion processes that Hberate the energy needed for contraction, while the carbon dioxide resulting from the combustion is diffused out. In outer contact with the sarcolemma are numerous discoid nerve-endings (Fig. i6, A, N) associated with nerve fibres along which travel the impulses which excite the muscle-fibres to contract, as will be elucidated in the next chapter. The muscles of insects are not surrounded by tough connective-tissue sheaths like those of vertebrates ; the bundles of fibres can be readily separated by teasing out, wherever a muscle is exposed by removal of some part of the exoskeleton. In many cases these fibres are directly connected with the portion of the exoskeleton which is moved by their contraction, but generally the muscle works on the skeletal sclerite through a tendon or a number of tendinous cords. These are cuticular structures secreted by inpushed portions of the skin which grow towards and find contact with the muscles. The insertion of a muscle by means of a tendon results in the pull due to its contraction being exerted on a certain special small area of sclerite ; thus precision of action is ensured. In studying the mechanism of movement among insects, the arrangement of the firm regions of the cuticle which make up the exoskeleton is as important as the form and nature of the muscles themselves ; the admirable accuracy of an insect's observable actions depends on the correlation of these two sets of organs. As an introductory illustration of such action the exoskeleton and muscles of the legs, already briefly mentioned, may be profitably studied. The three pairs of legs of an insect are articulated respec- tively to the three thoracic segments. Each leg is seen to consist of a series of hard sclerites segmentally arranged with flexible tracts of cuticle at the joints between adjacent MOVEMENT 51 Abd. of coxa sclerites, which are thus capable of movement in relation to each other (Fig. 17). The segment of an insect's leg next the body is the conical or subconical haunch (coxa), the broad base of which adjoins the sternal region of its segment while it tapers distally to its junction with the small sickle-shaped trochanter which is succeeded by the long stout thigh or femur. In some in- sects, cockroaches for example, the cuticle of the haunch is partly transparent so that the strong muscles which, by their contraction, move the leg as a whole can be seen through it, their white colour contrasting markedly with the general brown hue of the exoskeleton. The leg-muscles in the cock- roach are described by Miall and Denny (1886). From the dorsal region of the thorax on either side muscles pass to the inner and outer edges of each haunch ; these are respectively the adductor and abductor coxae, as by contraction of the former the leg is drawn in towards the axis of the body, while by con- traction of the latter it is moved outwards. A very strong muscle with its basal attachment (origin) also on the thoracic skeleton traverses the haunch, its fibres converging to their distal attach- ment (insertion) by means of a tendon on the inner (convex) edge of the trochanter. This is the muscle whose contraction straightens the thigh in relation to the €xt. lib. R«tr. tan. Fig. 17. — Left intermediate Leg of Cockroach (Blatta orien- talis), showing segments and muscles, c, coxa; tr, tro- chanter ; /, femur or thigh ; ti, tibia or shin ; ta, tarsus or foot. Muscles : Add, adductor ; Abd, abductor ; Ext, extensor ; Fl, flexor : Retr, retractor. X3. After L. C. Miall and A Denny, The Cockroach^ 1886. 52 THE BIOLOGY OF INSECTS haunch, trochanter and thigh moving as one piece ; the muscle is therefore known as the extensor femoris. It is readily understood that by the straightening of the leg with the foot resting on the ground the body of the insect is thrust forwards ; hence these muscles are of great importance in such actions as walking or running. A smaller muscle {flexor femoris) has its origin along the outer aspect of the haunch and is inserted by means of a tendon on the outer (concave) edge of the trochanter ; when this muscle con- tracts the thigh is bent towards the haunch, a movement which lifts the foot from the ground and draws it forward in preparation for the subsequent repetition of the extending and propelling action. Within the thigh are two muscles whose fibres converge from their origin along the outer and inner edges of that segment to tendinous insertions on the corresponding edges of the base of the next leg-segment, the shin (tibia) articulated to the thigh at the flexible knee- joint ; these muscles are respectively the extensor ^nd flexor tibiae^ their function respectively being to straighten or bend the leg at the knee. Beyond the shin follows the foot (tarsus) with its five segments, the first the longest and the fifth longer than any of the three intervening. A muscle {flexor tarsi) originating along part of the inner edge of the shin has its fibres converging into a long tendon which traverses the successive segments of the foot and is inserted into its tip ; when this muscle contracts the foot is curled or bent by the flexion of all its segments in the same direction. There are also two other muscles in the shin, the protractor and retractor tarsi ; these are inserted at the base of the foot respectively in front and behind, and their action is to swing the foot backwards or forwards in relation to the rest of the leg. The backward swing of the foot clearly assists in the forward propulsion of the body in walk- ing or running. Each foot-segment has a patch of w^hite cuticle below its terminal region ; this gives some power of adherence to smooth surfaces, and the patch {pulvillus) at the tip of the foot is larger and more conspicuous than the others. The terminal foot-segment carries a pair of strong MOVEMENT 53 curved claws, which can be flexed by the tendon of the flexor tarsi muscle, and are of use in affording the insect a hold on rough surfaces. In the act of walking all these various muscles in the creature's six legs are brought into play, the contractions synchronising or alternating rhythmically so as to bring about the orderly movement of the legs in two sets of three as already mentioned Besides walking and running, many insects have other modes of movement by means of their legs, such as leaping, gUding, or swimming. In a grasshopper the hindmost legs are by far the longest and strongest of the three pairs. When such an insect is at rest these limbs are seen to be strongly flexed at the knee joint ; the sudden extension of the shin, brought about by contraction of the large muscles within the swollen thighs, propels the insect for a distance of several feet through the air. Fleas are notorious for their power of leaping ; in their legs the haunches are exceptionally long as well as stout, and the flea's jump is due to the extension of the thighs. Aquatic insects such as water-beetles which swim and dive have either the intermediate or hind legs or both pairs flattened so as to serve as oars or paddles. The thighs move on the haunches and there is little change in the relative position of thigh and shin, as the strong limbs are swept rhythmically forwards and backwards. It is note- worthy that in these jumping and swimming actions the two legs of a pair move together, not alternately as in a walk or run. By means of suitably arranged muscles many parts of the trunk can be moved in relation to one another. The dorsal sclerites (terga) of the abdomen are Knked up by longitudinal strands of muscle the abdominal tergals, as the ventral sclerites (sterna) of the same region by sheets of abdominal sternals, while vertically directed tergo-stemal muscles connect each tergumwith the sternum of its segment. By the contraction of the last-named muscles the dorsal and ventral walls are drawn together, the capacity of the abdomen being thus reduced and the pressure necessary for respiration exerted on the air-tubes (see pp. 42-3 above). 54 THE BIOLOGY OF INSECTS When both sets of the longitudinal muscles of the abdomen contract, that region is shortened by a partial telescoping of the segments ; contraction of the abdominal tergals tends to straighten the abdomen or to bend the tail upwards, while contraction of the sternals flexes the abdomen ventralwards. There are also short oblique tergal and sternal muscles inserted laterally, which when they contract bend the abdomen to one side or the other. In many insects during the early (larval) stages of their life-history, when the limbs are relatively small or even wanting, these muscles connecting adjacent segments are the main agents in locomotion. By stretching out the head region and then drawing the rest of the body after it by waves of contraction, the characteristic creeping move- ment of a grub or maggot (Plate II, D) is brought about. A moderately full account of the muscular system of any insect would be far beyond the scope of this book ; one of the famous past-time students of insect anatomy, Pierre Lyonet (1762), dissected out and described sixteen hundred and forty-six muscles in the caterpillar of the Goat Moth. These muscles, like the other organs of immature insects which have to pass through a transformation before attaining maturity, are arranged so as to be specially adapted for the insect's activities during the larval period of its life, and these are commonly widely different from the activities of the creature when adult. Most insects when fully grown have the power of flight, and the action of the wings is therefore among the most characteristic and remarkable of all the movements of insects. The wing of an insect of powerful flight, such as a dragon-fly, appears to be a sheet of firm, transparent membrane stiffened by tubular " veins " or nervures. Study of the method of wing-growth, however, shows that a wing arises as a pouch or outgrowth from the dorso-lateral region of the thoracic segment that bears it. This pouch becomes flattened and in the developed wing the two folds — '' roof" and '' floor" as they might be called — come into close apposition over the greater part of the surface, which is MOVEMENT 55 covered by firm cuticle. Air- tubes, however, grow into the pouch or wing- rudiment during its development, and these become surrounded by the thickened tubular structures which appear as the supporting nervures of the wing (Fig. 1 8). As development proceeds the cells of the skin (Fig. i8, B, e), which are active in forming the cuticle, become attenuated and at length disappear, so that the adult wing is entirely cuticular. A fully developed wing is jointed to the thoracic seg- ment from which it arises, the forewings belonging to Fig. i8. — A, Forewing of nymphal Stone-fly {Taeniopteryx) showing the tracheal trunks and their branches {sc, subcostal ; r, radial ; m, median; c, cubital; a, anal). The pale tracks indicate positions of developing nervures. X 12. B, cross-section through developing wing of a Dragon-fly nymph {Anax). cu, cuticle ; e, epidermis ; rt, radial trachea ; mt, median trachea. X 150. After J. H. Comstock and J. G. Needham. Amer. Nat. xxxii, xxxiii. the second (mesothorax) and the hindwings to the third segment (metathorax) of that region of the body. The base of the wing, its region of attachment, is relatively small and is supported by a series of sclerites which are connected by flexible cuticle to the thoracic wall, so that the wing is 56 THE BIOLOGY OF INSECTS free to move up or down, to spread so as to project outwards from the axis of the body, or to be drawn inwards so that jj p its length is approxi- mately parallel to that axis as well as to be partly turned on its axis so that its lower aspect is directed backwards. The spreading and indrawing move- ments are brought about respectively by extensor and flexor muscles in- serted at the wing base. The forward turn of the wing- surface is due to the action of pro?iator muscles (Fig. 19, B, pr) inserted into a sclerite (anterior parapterum) at the front costal wing- base, and also to the resistant action of the air on the flexible wing-membrane. But the large and powerful muscles which by their con- traction move the wings up and down are in the vast majority of insects attached not directly to the wing-base but to regions of the thoracic wall. The Fig. 19. — Principal Wing Muscles of male Honey Bee {Apis tnellijica), as seen in median section through thorax (A), and internal view of right pleuron of meso- thorax (B). el, elevators ; dp, depressors ; pr, pronator : Jfl, flexor muscles ; pa, para- pterum ; a, axillary sclerites. C2, C3, haunches of second and third legs. X 15. After R. E. Snodgrass. MOVEMENT 57 elevators of the wing are muscles with fibres running almost vertically (Fig. 19, el) between the ventral and dorsal aspects of the thorax ; when they contract they pull the dorsal wall down, the wing-base is necessarily pulled down and the wing- tip rises. The depressors of the wing (Fig. 19, A, dp) run obliquely along the thorax so as to pull part of its wall backwards ; this raises the dorsal region with the wing-bases, and so the wing generally is depressed. In dragon-flies the great wing muscles are attached to the sclerites of the wing- bases so that they act directly on the wings as on levers. Their arrangement, three elevator and five depressor muscles to each of the four wings, has been described in detail by R. von Lendenfeld (1881). H. R. A. Mallock (1919), commenting on this remarkable contrast in the working of an essential group of muscles in dragon- flies as compared with other insects, writes : " The question arises as to why has this compHcated and indirect method prevailed ? If the problem were set of designing a mechanism for flapping wings, the dragon- flies' solution would certainly be the first to suggest itself ; yet it evidently must have some dis- advantages since it has not been generally adopted." It may be suggested that the flight muscles, attached between various regions of the thoracic wall, though acting indirectly on the wings are perfectly correlated with the articulation of these to the thorax, and that such mode of attachment is less delicate and liable to derangement than the direct action which characterises the dragon- flies. In Locusts, Grass- hoppers and their allies there are direct elevators and depressors inserted at the wing-bases, as well as the indirect muscles attached to the thoracic wall. It is noteworthy that dragon-flies, while remarkably powerful, swift and accurate fliers, have no provision for coupling the two wings of one side together when flying ; the forewings and hind- wings are worked by independent sets of muscles and the co-ordination of the wings of the two pairs depends upon an internally placed lever connecting their basal sclerites, and correlation in nervous control. In the more primitive insects of other orders also, such as Orthoptera (Cockroaches 58 THE BIOLOGY OF INSECTS and Grasshoppers), Plecoptera (Stone-flies), Isoptera (Ter- mites), the wings of the two pairs are uncoupled and each wing-bearing segment has its own set of thoracic muscles. But, on the other hand, in the great majority of highly organised flying insects of other orders there is some form of wing-coupHng apparatus, so that the fore and hind wing of each side move in unison, following the rhythmic changes of outline of the thorax due to the alternate con- traction of the sets of flight-muscles. Thus among the bees. Fig. 20. — Right wings of Honey Bee {Apis), h, row of booklets on costa of hindwing which hold dorsal edge of forewing during flight. X 8. After Comstock and Needham and Snodgrass. wasps, and other insects of the order Hymenoptera a series of curved hook-like bristles along the front edge (costa) of the hindwing catch on the thickened hind-edge (dorsum) of the forewing (Fig. 20) ; it is easy to demonstrate this linkage by manipulating the wings of a large wasp or sawfly. In some of the Lacewing and Scorpion-fly groups (Neuroptera and Mecoptera), as R. J. Tillyard (1920) has pointed out, there is a lobe or process (jugum) at the hind- base of the forewing, and on the costa of the hindwing a corresponding humeral lobe bearing one or two stiff bristles (frenulum). The jugum is present also in many Trichoptera (Caddis-flies) and in a few Lepidoptera (Moths), though in MOVEMENT 59 the vast majority of families of this great order we find a well- developed frenulum composed of numerous bristles closely apposed, projecting from the costal base of the hind- , ^ - d Fig. 21. — Types of Wing-coupling Apparatus in various Insects. a, Mecoptera (Taeniochorista), X 80 ; b, Neuroptera {Archichauliodes), X 10; c, Lepidoptera (Micropterygidae : Sahatinca), X 80 : J, Lepi- doptera (Hepialidae : Charagia), X 14; Cy Lepidoptera (Noctuidae : Plusia), X 30. jl, jugal lobe ; jjugum ; &, jugal bristles (on forewing) ; h, humeral lobe, and F, frenulum (on hindwing) ; r, retinacular bristles on cubital nervure (r) of forewing. After R. J. Tillyard. (Proc. Linn. Soc. N.S.W. xliii, 191 8.) wing and fitting beneath a number of stiff hairs or scales under the base of the forewing (Fig. 21, e). Through such 6o THE BIOLOGY OF INSECTS arrangements the two wings of a side are coupled and present an extensive surface to the air, which is compressed as they are pulled downwards by means of the depressor muscles, and turned backwards by the action of the flexors and by the wing's elasticity. This atmospheric compression leads to the resistance which is the mechanical agent in supporting and propelling the insect during flight. Then in the great order of the two-winged flies (Diptera), the forewings alone are developed as organs of flight, while in the beetles and earwigs the hindwings only are efficient, the forewings being modified into firm sheaths (elytra). Our knowledge of the mechanism of insect flight is largely due to the researches of E. J. Marey (1895), who by obtaining tracings of the wing-tips of flying insects on the smoked cylinder and fastening a spangle of gold leaf to the tip of the wings, vibrating as in flight, of a wasp held by forceps in bright sunlight, demonstrated that the path (trajectory) of the wing- tip is a narrow and elongate " figure of eight." Marey 's work has been supplemented by the special studies of F. Stellwaag (1910) and W. Ritter (191 1) on the flight of the hive-bee and the blow-fly respectively. In the latter insect there are ten directly-acting muscles attached to the sclerites of each wing-base ; these though not working as the depressors and elevators of the wing, are of much importance in effecting the suitable tension of the wing, and possibly act in steering. Observation of insects during flight aflfords to the student abundant opportunity of noticing how the details of wing movement vary in diff"erent groups. How markedly, for example, does the heavy flapping flight of many of our larger butterflies contrast with the darting movement, alternating with the apparently motionless poising in the air, of a " Humming-bird " Hawk-moth or a hoverfly ! In the former case the number of wing-strokes per second might be roughly calculated by observation ; in the latter they can be determined only by tracings on smoked paper compared with those made by a tuning fork of standard vibration rate, or, if the wing- vibrations produce an audible hum, by verifying MOVEMENT 6i the musical note in unison therewith. Thus it has been found that a Common White Butterfly (Pieris) makes only nine wing-strokes a second as contrasted with the 190 strokes a second of the Hive-bee (Apis), or the 330 strokes a second of a Housefly (Musca). Relatively large insects with extensive wing- area and powers of prolonged, often soaring flight, have a rate of wing- motion intermediate between these extremes. A dragonfly, for example, may beat its wings twenty-eight times a second as it hawks through the air in pursuit of its prey. The different instances mentioned may serve to illustrate how, in every case, the special features of an insect's flight corre- spond with its peculiar way of life. Further consideration of flight in relation to the general behaviour and environment of insects may be deferred till a later chapter. Walking, running, leaping, crawling, and flying are obvious modes of movement which cannot be overlooked by the most casual observer of insects. But there are many other movements of equal importance to the creature's life, which, though less conspicuous, are all brought about by muscular action. An insect continually moves its feelers as though seeking to test the neighbourhood into which it advances ; these movements are due to the contraction of small muscles passing from the inside of the head-capsule to the feeler's basal segments. The feelers are, as already stated, modified Hmbs in series with the insect's legs. The three pairs of jaws (mandibles, maxillae, and labium) whose position and action were briefly described in the last chapter (pp. 17-23) in relation to the function of feeding, are likewise modified limbs. The muscles by which these jaws are worked are often numerous and always, in view of their use in nutrition, important. Some description of their arrangement may therefore be given first in a biting and then in a sucking insect. The jaws and their muscles in the cockroach, a typical biting insect, have been described in detail by J. Mangan (1908). The Cockroach's mandible with its apical teeth and molar area, its convex and concave articulations (condyle 62 THE BIOLOGY OF INSECTS Fig. 22. — A, Mandibles and tongue (hypopharynx) of Australian Cockroach {Periplaneta australasiae) , viewed from behind, c, condyle ; g, ginglymus ; mp, molar area ; la, lacinia of mandible. L, adductor muscle ; Ex, abductor ; S, short adductor, hy, hypopharnyx partly turned backwards ; x, y, z, its sclerites {z probably the superlingua or paragnath) ; In, protractor and V, retractor muscle of tongue, ten, tentorium (internal head skeleton) ; rcep, salivary receptacle. X 24. B, The mandibles closed with teeth interlocking. X 15. After J. Mangan {Proc. R. I. Acad, xxvii, 1908). MOVEMENT 63 and ginglymus) with the head-skeleton, resembles that of the Eanvig previously described (pp. 17-19). The mandible (Fig. 22) is pulled outwards by its abductor muscle (Ex), which consists of several bundles of fibres arising from the upper lateral region of the head capsule, and converging to a chitinous tendon inserted at the outer edge of the mandible beyond the condyle. The large adductor (L) wliich pulls the mandible inwards to meet its fellow of the opposite side is a much thicker and stronger muscle than the abductor ; its numerous bundles of fibres arise from the top and back of the head-capsule, and converge to a strong chitinous tendon wliich is inserted in the hind inner region of the jaw. Each mandible has also a short adductor (S) whose parallel fibres pass directly from the inner head skeleton (tentorium) to the inner face of the mandible's convex outer wall. The contraction of these two adductors pulls the mandible in towards its fellow, so that the pair of jaws meet, with their teeth interlocking, opposite the centre of the mouth (Fig. 22, B), unless some food substance happens to He bet^veen them ; in such case the teeth cut it, and the molar areas grind it into small fragments. The outer edge of the mandible has another muscle (In) of small extent whose fibres are fastened to its inner aspect. These converge into a long slender tendon which is inserted into the side of the tongue. This pair of muscles {levator es linguae) serve by their contraction to protrude the tongue. They are opposed by another pair of muscles (retractores linguae) which, passing from the tentorium on either side to the base of the tongue, pull that organ backwards into the mouth (Fig, 22, V). The complex structure of the maxillae (see previous chapter, p. 19, and Fig. 23) necessitates a corresponding elaboration in the muscular system by which their parts are worked. Inserted into each cardo is a strap-Hke depressor muscle (IF) with three sections arising from the tentorium. This depressor, by its contraction pulls the cardo and thus the w^hole maxilla downwards. An elevator or abductor (P) runs from the top of the head-capsule (epicranium) to the Fig. 23. — Maxillae of Australian Cockroach. For explanation, see opposite page. MOVEMENT 65 cardo, and raises the maxilla. From the tentorium a broad strap-like muscle (G) passes to the stipes which it pulls, when contracting, towards the centre of the mouth ; this is the adductor of the stipes (and necessarily of the lobes carried by it) which it moves on the hinged articulation with the cardo. From within the stipes arise two long, thick tapering muscles (A, B), one the adductor of the lacinia and the other of the galea to the bases whereof they are respectively attached, the action of the latter being aided by another muscle arising from the tentorium. Two smaller muscles (D, E) passing from the inner edge of the stipes to the upper and lower edge of the palpiger are respectively the adductor and abductor of the palp as a whole, while the individual segments of that leg-hke region are worked by slender muscles concealed within it. Thus the movements of the palp in testing objects over which the insect walks, as well as the action of the maxillary laciniae or blades in breaking up foodstuffs into finely divided particles are brought about. The arrangement and mode of working of the head muscles are necessarily strongly modified in the case of jaws adapted for piercing and sucking. As an example we may take the jaws of a plant-bug such as Lygus pabulinus described in detail by P. R. Awati (1914)- The mandibles and maxillae of insects of this order are elongate piercing stylets working to and fro in the dorsal groove of a jointed beak which is the modified labium. For the puUing back of these piercing jaws, there are retractor muscles which arise from chitinous rods in the dorsal region of the head-capsule, and are inserted into the swollen bases of the mandibles and maxillae respec- tively (Fig. 24A, r, rx). The jaws are thrust out, so Fig. 23. — Maxillae of Australian Cockroach {Periplaneta australasiae). A, right maxilla from behind ; B, left maxilla from front, car, cardo ; St, stipes ; la, lacinia ; ga, galea ; p, palp ; m, articulation of lacinia. ten, tentorium. Muscles : A and Q, adductors of lacinia ; B, adductor of galea ; D, adductor, and E abductor of palp ; H, L, N, T, segmental muscles of palp ; G, adductor of stipes ; P, elevator, and W, depressor of cardo. X 24. After J. Mangan {Proc. R. I. Acad, xxvii, 1908). F 66 THE BIOLOGY OF INSECTS mx Fig. 24A. — Cross-Section through Head of Plant-bug (Lygus pabu- linus) ; Ph, pharynx ; dy its divaricator muscle ; g, gullet ; ps, salivary pump ; mp, its muscle ; sd, salivary duct ; r, retractor, and p, protractor muscles of mandible {mn) ; rx and px, retractor and protractor of maxilla {mx) ; /, labial muscle ; b, brain ; s, sub-oesophageal ganglion ; «, nerve- commissure ; o, taste-organs. X 200. After P. R. Awati, P. Z. S. 1914. as to perforate the plant tissues and procure a supply of food-sap by the contraction of protractor muscles, which arise from the frontal ventral head-skeleton, and are in- serted, in the case of the maxillary protractors {px)^ directly into the swollen bases of the jaws, but for the mandibular protractors {p) into transverse chitinous rods attached to the bases of the mandibles, and therefore pulling these as by the agency of levers. The labial beak, in which the mandibles and maxillae are thus pulled in or out, can be moved forwards and backwards by elongate muscles which pass from the dorsal wall of the prothorax to the front and hind aspect of the base of the terminal segment of MOVEMENT 67 Fig. 24B, — Longitudinal Section through Head of Plant-bug For explanation, see Fig. 24A. the beak. The liquid food obtained by the piercing action of the jaws is sucked in by movement of the walls of the pharynx, the region of the food-canal into which the mouth opens and from which the gullet leads out. There is a group of divaricator muscles {d) passing from the inside of the facial head region to the front wall of the pharynx ; the contraction of these increases the capacity of the organ. They are antagonised by lateral pharyngeal muscles which arise from the head skeleton on either hand and are inserted into thickly chitinised paired regions of the pharyngeal wall. When these are pulled apart by the contraction of the lateral muscles, the dorsal and ventral walls of the pharynx are drawn closely together, and the cavity is almost obliterated. For the digestion of an insect's food, movement of the walls of the digestive canal are necessary, and the action of the muscular layer of the proventriculus in constricting its aperture has been discussed in the previous chapter 68 THE BIOLOGY OF INSECTS (p. 27). Similarly contractions of the thin muscular wall of the heart are effective in propelling the blood, while by the action of the abdominal muscles changes in the form and capacity of the body are brought about so that through varying pressure on the tracheal tubes air is drawn in and expelled, the spiracles being rhythmically opened and closed by means of their special musculature. Only a glance at the highly complicated muscular system of an insect has been possible, but it will be realised that the actions of the various parts of this system must be co-ordinated if the creature's movements are to have purposeful relation to the needs of its Hfe. Some discussion of the mode of such co-ordination will be found in the succeeding chapters. CHAPTER IV SENSATION AND REACTION The various movements of insects, some of which have been considered in the previous chapter, are often observed to be expressions of the creatures' responses to various kinds of stimulation due to the condition of their surroundings. A moth flies in through an open window on a summer night, steering a straight course for the lamp, so that the observer exclaims that the insect is " attracted by the light," and concludes that it possesses a faculty resembling at least to some degree his own power of vision. He may go several steps farther in equating the insect's behaviour and sensa- tions with his own by saying that " it prefers light to dark- ness," or that "it is dazzled by the glare of the lamp." Collectors of insects desirous of obtaining male specimens of certain rare moths often take into the open country or woodlands a hve female of the kind desired shut up in a box, and find that this miniature prison may be surrounded by large numbers of males which have directed their flight towards it. The captive female, though invisible, has allured them, and the observer may reasonably infer that these males have been thus " assembled " because they have received impressions accompanied by some recognised sensation analogous to our sense of smell. Reactions of insects to their surroundings, such as the two just men- tioned, can be readily observed, and by study of the nervous system and its connection with the various parts of the body it is possible to understand, at least in part, the mechanism by which the various reactions are brought about. But when we ascribe to an insect sensations, or, 69 70 THE BIOLOGY OF INSECTS still more, intentions and apprehensions like our own, we quickly pass into regions of speculation, as we can no longer be guided by carefully tested fact. An insect may react to its surroundings in somewhat the same way as a man reacts, but it is not justifiable to infer from this that the insect's state of consciousness is like the man's. The questions thus raised are interesting, even fascinating, despite their difficulty. Before attempting to discuss them it will be well to consider in detail some examples of the working of stimulation and response in the insect's body. A brief general sketch of the insectan nervous system has been already given in the first chapter (pp. 7-10) of this book, and it was there mentioned that movements such as the moth's flight towards a lamp or towards a female of his kind shut up in a box, movements clearly to be regarded as responses to stimulation from without, are called reflex actions. In a reflex action some nerve-ending, usually near the body surface and capable of being stimulated through a specialised region of the cuticle, is aflFected so that it transmits through a nerve-fibre an impulse to a nerve-centre such as part of the brain or a ventral ganglion. From the nerve- centre the impulse is then reflected along other nerve-fibres, whose endings in contact with muscle-fibres, impel these to contract and thus give rise to visible movement. Such movement is the result of, and to the observer affords evidence for, the transmission of the nerve-impulse to and from the nerve-centre. Impulses towards the centre, as well as the fibres along which they travel, are usually defined as aff"erent, while impulses from the centre to the muscles and also the nerve-fibres conveying them are called efferent. An insect's nervous system may be regarded as a vast complex of living cells from each of which processes branch in various directions, many of these processes passing into the axes of nerve-fibres. In every nerve-centre a number of cells are grouped, and the fibres, whose axes are the prolongations of these cells, are bound into the white thread-like cords, evident on dissection, which are called nerves : every nerve is a bundle of many fibres. A nerve- SENSATION AND REACTION 71 impulse passes from cell to cell along the path of the connecting fibres, a fibre-axis at its extremity coming into touch by means of exceedingly minute branches with relatively short finely branched outgrowths from other cells ; thus a nerve impulse is passed on through a series of " cell-relays." An insect's apparently purposeful move- ments are therefore dependent on the co-ordinated trans- mission of a large number of impulses through the various parts of its nervous system, from receptive nerve-endings to the ganglia and thence to the muscles. The nerve-endings of insects that can be affected by outward stimulation are of many kinds, adapted for the reception of varied types of influence. Attention has been repeatedly directed to the horny cuticle which covers an insect's body, and mention has been made of the hairs and spines — often long and conspicuous- — which are specialised portions of the cuticle, jointed on to the main surface by basal flexible membrane (Fig. i). Many hairs are hollow, and into their cavities pass fine thread-like processes of nerve-cells lying in the skin, whence nerve-fibres lead to groups of cells, situated in some ganglion. Such nerve- endings as these are clearly adapted for receiving tactile impressions. The slightest contact of any outside object with the hollow hair must lead to stimulation of the included thread-Hke process. Thus the skin-cell is affected, and from the cell the impulse passes along the nerve-fibre to a ganglion cell. Various objects that may be touched will necessarily cause many nerve-endings to be affected, and thus a number of impulses will simultaneously travel to a gangHon, where, through the linkage of the receptive cells, they will be co-ordinated and the impulses passed on to motor cells, whence by efferent fibres they will be reflected to the muscles whose contraction will result in the reaction appropriate to the stimulation received. The feelers carried on an insect's head, the palps con- nected with the jaws (maxillae and labium) as well as certain abdominal appendages such as the tail-feelers (cerci) are often richly provided with tactile organs, so that 72 THE BIOLOGY OF INSECTS the creature is exceedingly sensitive to impressions derived from objects touched. Sensory hairs and spines are also often present on the cuticle of various segments of the body, especially in the case of those insects whose exoskeleton is relatively weak, so that prompt response to impressions from all quarters is essential to secure safety. Many insects, as they walk or run, may be seen to keep the tips of their feelers in continual swaying movement, which has the effect of bringing the sensory hairs on the segments of the feelers in such relation v^dth surrounding objects that there are abundant possibilities of varied points of contact. And the response of the whole insect to the impressions thus received depends upon the positions of the various sensory hairs where these impressions are started through contact with external objects. The nerve- endings affected through such hairs are termed tactile, and their mode of action suggests that insects possess a '' sense of touch." It is certain that these nerve- endings are affected when the hairs within which their extremities He come into contact with outside objects, and the insect's reaction to such stimulation, due to muscular contraction, shows that a reflex impulse travels to and from the central nervous system. So far there is analogy between what happens in the insect and in a human being. But what we understand by a sensation which leads us to distinguish an object as hard or soft, rough or smooth, is essentially a conscious experience, accompanying certain nerve-impulses, but not identifiable with them. We are not, therefore, able to assert that an insect *' feels " as we do. Indeed, the insect's tactile nerve-endings are in form very different from those of vertebrates and are affected in a different manner. A tactile corpuscle of the human finger, for example, is situated beneath the outer skin (epidermis) and is affected indirectly by variations of pressure or resist- ance acting at the surface of the skin yet able to " irritate " the nerve-ending, though the latter lies under many layers of cells. The insect's tactile nerve-ending is, as we have seen, prolonged so as to rest within a narrow hair ; it is SENSATION AND REACTION 73 therefore directly pressed or bent when the hair touches an external object. As far as physiological mechanism can guide us, we may conclude that the insect's reception of and reaction to tactile stimulation are far more delicate than our own. But we have no evidence that an insect's " sense of touch," experienced in whatever form of consciousness the creature may possess, is more vivid than ours, or indeed that it is like ours at all. Besides these " sensory " hairs and spines whose function is undoubtedly tactile — that is to say, the nerve-ending can be stimulated only if the hair actually touches some external object — the feelers and parts of the jaws of many insects carry blunt peg-like outgrowths of the cuticle in which the chitin is markedly thinner than in typical tactile hairs. Into the cavity of those also pass delicate processes of nerve- cells in the skin. Most students of the sense-organs of insects — for example, A. Forel (1908) — have regarded these as concerned with smell or taste, considering that the cuticle covering the nerve- endings is sufficiently thin and delicate to permit the permeation of vapour and thus allow the nerve-endings to be stimulated chemically. Such peg-like organs are abundant on the feelers, maxillae and labium of many insects, and experiments made by Forel and others show conclusively that insects can be affected from a distance by pungent substances, the sensibihty diminishing or disappearing if the feelers be cut off. N. E. Mclndoo (1916), on the other hand, has pointed out the improba- bility of delicate chemical stimulation, such as that which results in sensations like smell or taste, acting through cuticle even though it be thin and delicate ; therefore he regards these peg-hke organs as tactile Hke the slender and more rigid hairs or spines. The feelers and palps of insects, however, are often provided richly with sensory nerve-endings of apparently another type, which have been described by K. M. Smith (19 1 9) as well as by several previous writers. The positions of these are evident through the presence of pits or pores on the surface of the cuticle. Study of sections of the organs 74 THE BIOLOGY OF INSECTS (Fig. 25) shows that the normal thick cuticle is interrupted, and that just in or beneath the aperture is a long, narrow- ended spindle-shaped cell or a group (s) of such cells ; often delicate processes (p) from the cells project towards the pore, and these may possibly be exposed to the outer atmo- sphere, though they do not protrude through the pore so as to be liable to contact with external solid objects. Such nerve-endings are not tactile, but they are adapted to receive chemical stimulation by vapour diffused through the air from some odoriferous substance in food- material or in the body of another in- sect, or from minute floating particles. In many, if not all, of these sensory pits, however, the nerve- endings are covered by extremely thin and delicate cuticle, through which odori- ferous or tasty sub- stances might readily be absorbed if dissolved in the fluid secreted by the glandular cells said to be often associated with these organs (A. Berlese, 1909). Nerve-endings of closely similar form and arrangement to those of the sensoiy pits of insects are well known in vertebrate animals Fig. 25.— Section through Feeler of Syrphid Fly ( Volucella bombylans) , showing cuticle {c) with pore {a) leading to sense-organ con- sisting of processes {p) arising from sensory cells {s) of the skin or epidermis {e). n, nerve to brain. X 600. After K. M. Smith {Proc. Zool. Soc. 1 919). SENSATION AND REACTION 75 as the organs whence start those nerve-impulses which, on arrival in the brain, result in sensations of smell or taste. In many insects sense-organs of this type are found in large numbers on the feelers, and any increase in the com- plexity of those appendages renders possible an increase in the number of such olfactory nerve-endings. For example, in the male moths mentioned above, that can be attracted from a distance towards an imprisoned and con- cealed female, the segments of the feelers are drawn out into relatively elongate processes in which the special nerve- endings are present in large numbers. The number and arrangement of these organs enable an insect possessing them not only to appreciate the vicinity of a female of its own kind, but also to detect the Hne of advance along which she may be found. The sense associated with nerve- impulses starting from these organs is clearly similar to our own sense of smell, while the large number of the organs, their arrangement on the antennal processes, and the mobility of the feelers all combine to give distinct indication as to the direction of the source whence comes the odorous vapour. Many beetles have the segments of the terminal region of the feeler thickened or greatly expanded and flattened, so that the feeler becomes clubbed (clavate) or plate-like (lamellate) in form. The surface provided by the enlargement of these segments are the seat of large numbers of sense-organs of this olfactory type. The males of such beetles often have antennal structures more elaborate than those of the females, and their highly developed sense of smell facilitates their detection of possible mates. But in both sexes of these insects the olfactory sense serves as a guide to their food which consists commonly of strongly smelling material such as dung or decaying flesh, on which substances the females also lay their eggs. Similar sensor}^ pores are also found abundantly on the mouth parts of various insects, such as the maxillary palps, the epipharynx, the tip of the labium, or in some cases the mandibles. These have usually been regarded as organs of taste, as by F. Will (1885), A. Forel (1908), and other 76 THE BIOLOGY OF INSECTS students. N. E. Mclndoo (191 6) has, however, given some reasons for doubting whether the so-called gustatory sense of insects can really be distinguished from their sense of smell. Such discussions illustrate the uncertainty which must surround our conceptions of an insect's conscious sensations, and indeed many physiologists regard our own discrimination of flavours as referable to smell rather than to taste. But there can be no doubt that the minute structure of sense-organs of this type indicates clearly that they are adapted for receiving chemical stimuU such as those which give rise to normal sensations of smell and taste, and there are many recorded experiments and observations on the reactions of various insects which afford support to this opinion. On several occasions A. Forel (see 1908, pp. 73 f.) performed experiments which appear to prove the presence of definite olfactory nerve-endings in insects' feelers. Ants recognise members of their own communities by smell, and as they approach one another their feelers are constantly moving ; when these appendages are removed ants were found to distinguish no longer between sisters and strangers. Female flies of the bluebottle group lay their eggs on the flesh of dead animals, and there is no doubt that the approach to a carcase for the purpose of egg-laying is a reaction to the olfactory sense. Forel found that if the feelers of such flies be removed they lose the power of locating the objects on which they lay their eggs, while a fly, thus mutilated, and actually placed on a putrid mole did not at once lay eggs thereon as a normal female would have done. From such results it appears that the approach to a suitable breeding place, and the laying of the eggs thereon, are alike actions reflex to the reception of olfactory stimulation. Burying beetles (Silphidae) not only lay their eggs on dead animals, but themselves feed on such carrion. In these beetles the terminal segments of the feelers are thickened so as to form a club, and the removal of the clubbed tips of the feelers rendered the beetles incapable of finding their food and breeding-material. Further, a male Silkworm Moth {Bombyx SENSATION AND REACTION 77 mort) failed to find its mate when deprived of its feathered feelers, though in its normal state it would run from a considerable distance straight towards a female across the floor of a room. That many insects are able to discriminate between foods of various composition has been shown by several observers and experimenters, whether the sense by which the differences are appreciated be defined as taste or smell. Mclndoo (19 1 6) describes the result of mixing various substances with honey or candy offered as food to bees. Wliile as many as 35 or 40 out of a hundred bees ate the pure sweetmeat, none would take candy to which oil of peppermint or carbolic acid had been added, though 22 per cent, partook of honey contaminated with whisky and 29 per cent, were attracted by cane sugar with a small amount of cider vinegar. When the bees were offered the alternatives of pure cane sugar, cane sugar and quinine, or cane sugar and strychnine, 47 per cent, ate the first, nearly 6 per cent, the second, and 4*6 per cent, the third ; but when the only alternative was between the sugar treated either with quinine or strychnine, 49 per cent, showed preference for the quinine flavour while only 4 per cent, would take the strychnine. The experimenter was espe- cially impressed by this last result, as he failed himself to distinguish by tasting between quinine and strychnine. Pure cane sugar was markedly preferred to sugar containing various salts of sodium and potassium. No bees would eat sugar and potassium cyanide, but sugar and potassium ferrocyanide attracted 33 out of a hundred bees to whom nothing else was offered to eat. Reference has been made to the precision of movement which most insects display whether they creep, walk, run, or fly, and observers of their movements have often inferred that they possess some kind of sense of direction. In many cases the adjustment of motion to direction may be most reasonably supposed to work through vision, but many facts suggest that insects possess a definite equilibrating sense analogous to that which is associated with the semi- 78 THE BIOLOGY OF INSECTS circular canals of a vertebrate's ear, the motions of the insect being responses to nerve-impulses that result from its position with respect to a plane vertical to the surface of the earth. Insects belonging to the large order of the Diptera or two-winged flies have the hindwings reduced to small drum- stick like rods known as halters. Many experimenters have shown that if one or both of these organs be cut away from a fly's thorax, the insect is no longer able properly to control its flight. Microscopical investigation of the base of the halters made by B. T. Lowne (1890) and others shows that they contain numbers of the remarkable sensory structures which have been observed and described by V. Graber (1882) and many subsequent observers, in various parts of insects of diverse orders, and called chordotonal organs. We may therefore regard it as highly probable that the function of some chordotonal organs at least is to receive impressions from the movements of the surrounding blood which lead to " equilibrating " sensation when passed on to the central nervous system. Recent investigations into the minute structure of these organs by W. N. Hess (19 1 7) have shown that they consist essentially of elongate spindle-shaped sense-cells whose axes are prolonged as aflFerent fibres while their distal ends are in contact with peg- like scolopales enveloped by accessory cells, the whole organ surrounded by a delicate outer sheath (Fig. 26). The tips of the scolopales are surrounded by cover-cells, and these delicate sensory endings may either float freely in the blood-spaces of the insect's body or be connected by a terminal ligament with some part of the cuticle (Fig. 26, A, Z). In either case the direction and pressure of the surrounding fluid in contact with the chordotonal organ will vary according to the position of the insect's body in relation to the horizontal and vertical planes, so that such organs afford the necessary mechanism for inducing equilibrating sensations. But chordotonal organs have generally been regarded as connected with the sense of hearing, and there can be no SENSATION AND REACTION 79 doubt that in many cases of their occurrence this belief is well-founded, because they are associated with some ^:i Fig. 26. — A, Cross-section of hind shin of Red Ant (Myrmica rubra) showing chordotonal organ (ch) with its scolopales (s), ganglion cells (nc) and nerve-fibres («). e, epidermis ; ai, cuticle. X 120. After C. Janet (" Observations sur les Fourmis," 1904). B, Scolopophore from the chordotonal organ of a Longhorn Beetle Larva (Ergates). I, ligament ; c, cap-cell ; s, scolopale ; v, vacuole ; e, enveloping cell ; 7ic, ganglion cell ; n, nerve-fibre. X 500. After W. N. Hess (Ann. Entom. Soc. Amer. x, 1917). 8o THE BIOLOGY OF INSECTS specialised region of the cuticle, thin, tense and delicate, evidently serving as an " ear-drum " which can be thrown into vibration by sound-waves impinging on it. Further, it is to be noted that in most kinds of insects which possess such auditory organs, there are found also special structures for producing sounds. The best known of insect ears are probably those found on the first abdominal segment in ordinary grasshoppers and locusts. On either side of this segment in a locust the sub-circular or ovoid membranous ear-drum is easily seen without the help of a magnifier ; close in front of it is the spiracle of the segment, and over the lower area of its inner surface are spread the fibres of the tensor muscle whose contraction increases its tightness. Attached to the inner face of the drum-membrane, near its centre, is a nerve ganglion in connection with the auditory nerve going to the central system. This ganglion, often called Miiller's organ from its discoverer of a century ago (Johannes Miiller, 1826), contains a number of large nerve- cells, which receive impulses from groups of chordotonal organs, whose distal ends are bound to the drum membrane, and transmit the impulses through the fibres of the auditory nerve to the segmental ganglia of the ventral trunk-cord. These remarkable " ears " on the abdomen in grass- hoppers and locusts are among the best known of insect sense-organs. Only recently, however, have we become acquainted v^th ears of a somewhat similar nature on the abdomen in certain moths of the Geometridae, Uraniidae, and allied families, and on the thorax of Noctuidae, Arctiidae and their allies, mainly through the researches of F. Eggers (19 1 9) and H. Eltringham (1923). The tympanal organs of a geometrid or uraniid (Fig. 27) belong to the second abdominal segment, not the first as in a locust. The cuticle of this segment is inpushed on either side to form a hollow vesicle, often much shallower in the male than in the female moth, and at the front aspect of this vesicle the delicate ovate drum-membrane (t) is evident. Connected with the strong ridge surrounding the tympanic membrane are radiating bands of muscle fibres serving apparently to regulate the SENSATION AND REACTION 8i Fig. 27. — A, Lateral view of metathorax {th) and the first three abdominal segments (i, 2, 3) of male Uraniid moth (Chrysiridia ripheus), scale-pad removed to expose vesicle (v) and tympanum (t) of ear ; s, spiracles. X 6. B, Internal view of tympanum [t), with muscle bands (w), chordo tonal organ (ch) and ganglion (g) ; v, cut edge of vesicle. X 20. C, Thoracic ganglion (G) and nerve-cord (n) with branches to tympanal organs (to), X 8. D, Chordo tonal organ of female, with scolopales (s), ganglion cells (nc) and tracheal sheath (ir). X 400. After H. Eltringham (Trans. Entom. Soc. 1919). G 82 THE BIOLOGY OF INSECTS tension of the membrane, the inner surface of which, formed by a modified air-tube, is in touch with a ganglion and tympanic thread surrounded by a sheet of tissue derived from the air-tube system and containing scolopales character- istic of chordotonal organs (Fig. 27). In a noctuid or other moth in which these organs are thoracic, they are present on the metathorax, or hindmost segment of that region of the body. Eltringham summarises their structure by stating that each organ '* consists essentially of a modified tracheal vesicle carrying two drums. One of these is the true tympanum with its chordotonal thread and the other would seem to be a kind of resonator." The tympanum lies, according to Eggers, in a lateral and ventral position with regard to the " resonator." The nature of these organs with their drums and nerve- cords leaves little doubt that their function is auditory, and with regard to the grasshoppers and locusts, this conclusion is supported by the well-known fact that the males of such insects produce a shrill chirping noise by rubbing certain wing-nervures over blunt pegs on their hind thighs. Noises produced by moths, though less familiar, have been detected by reliable observers ; it is remarkable that the Death's-head moth (Acherontia) whose mouse-like squeak has often been described, belongs to a family (Sphingidae) in no member of which have tympanal organs been detected. In long-horned grasshoppers and crickets it is well known that there are ears situated in the front shins, paired inpushings of cuticle below the knee-joint being in contact with an expanded air- tube, so that the vibrations can affect a set of nerve-endings arranged along an extended ridge. The organs have been well described by N. von Adelung (1892) and others, and J. Regen (19 12) has shown that females of a cricket Liogryllus campestris, normally attracted by the chirping of the males, were no longer influenced after their auditory organs had been removed. It is possible that in many insects such organs as these may be responsive to vibrations too rapid for appreciation by means of the human ear. SENSATION AND REACTION 83 There remains for consideration the sense of vision, or of some dimmer perception due to the stimulation of nerve- endings by the impact of light- rays, referred to at the opening of this chapter. An insect's eye is a sense-organ of considerable complexity, but it is, like all the various types of sense-organ, due to a modification of the skin with its overlying cuticle and the connection of certain special sensory organs in the skin with the underlying nervous system. The cuticle covering an eye may be relatively thick, but it must be transparent to allow rays of light to pass through and stimulate the nerve-endings beneath. In many insect larvae groups of small circular or sub-circular simple eyes (ocelli) are readily observed on either side of the head, while many adult insects have two or three ocelli on the vertex (Fig. 10, 6). Their surface presents a glassy aspect, and the deeper structures imperfectly seen through the transparent cuticle (cornea) appear black on account of the dark pigment which is contained in or associated with the visual cells of animals generally. Microscopical examination of the simple eye or ocellus of an insect shows that the trans- parent cornea is convex on the upper surface, and possibly also on the lower, beneath which the living cells of the skin extend in a continuous sheet ; the cornea has of necessity been formed by the activity of these cells. Below them come the receptive cells which form the retina of the eye ; they are elongate and taper towards their inner ends which are produced into the axes of fibres. These retinal cells, w^hich are derived immediately from the skin, may contain a quantity of dark pigment or the pigment may be accumu- lated in special cells intercalated between groups of retinal cells, the retina being thus divided into a number of small cell-groups (retinulae). The structure of such a simple eye suggests that the circular convex cornea acts as a lens, the rays of light converging on the sensitive retina whence nerve impulses pass on through the fibres to the brain. Eyes essentially of the same type as these insect ocelli are found also among worms, molluscs, and other animals. Further discussion of their function may be postponed until we 84 THE BIOLOGY OF INSECTS have described the main features of the compound eyes which are particularly characteristic of insects and of their relations the Crustacea. The pair of convex, dark compound eyes present, in most insects, a conspicuous feature of the head. Each eye may occupy a comparatively small round or oval area as in a bug or beetle, or may assume an extensive protruding sub-spherical form as in a breeze-fly or a dragon-fly. Fig. 28. — Section through Compound Eye of Honey Bee (Apis mellifica). c, cornea ; o, ommatidia ; b, basement membrane ; p, peri- opticon ; e, epiopticon ; op, opticon. X 100. After E. F. Philips {Proc. Acad. Nat. Sci. Philadelphia, 1905). Examining with low magnification the surface of such a compound eye it is seen to be made up of many small, usually hexagonal areas which may be numbered by hundreds or even by thousands. Examination of a section cut through the eye (Fig. 28) vertical to its surface demon- strates that the elements of which it is composed are arranged so as to converge from the inner face of the cuticle (where each element is in contact with one of the small hexagonal SENSATION AND REACTION 85 areas or corneal facets) towards the underlying optic ganglion. It is therefore evident that such an eye is adapted for the transmission to the ganglion of a number of nerve impulses due to the impact of light waves from many quarters of the insect's surroundings, these impulses travelling along con- verging paths towards the central system, and giving rise to visual impressions from a wide range of direction. A fly with two such eyes, large and sub-globular, occupying the greater part of the surface of the head, should evidently be able to receive impressions through the eyes from before, above, and beneath as well as from either side, and to some extent from behind. While we recognise thus the great scope of an insect's vision as regards direction, we remember that the corneal area of the compound eyes, consisting of a modified portion of the head-cuticle, cannot be moved, and we realise that insect vision must differ greatly from our own and that of most vertebrate animals, in that neither both eyes nor one can be brought to bear especially on a com- paratively small region of the environment, the close examination of which might be desirable. Even a super- ficial view of an insect's compound eye inclines us therefore to the opinion that the sight of such a creature is by no means like ours, and we may surmise that what the insect gains in range of direction it may lose in perception of distant objects as well as in clearness of definition. Before pursuing this line of discussion it will, however, be advisable to consider in some detail the structure of the elements that compose an insect's compound eye. The eye of the well-known Honey Bee has been well described by E. F. Phillips (1905), and his account is summarised in R. E. Snodgrass's works on that insect (1910, 1925). Each corneal facet is the outer area of a thickened transparent section of the cuticle which may be regarded as a lens (Fig. 29, cr). Beneath each lens is a transparent crystalline cone due to the modification of four special cone-cells derived from the skin (Fig. 29, c). Beneath each cone is a clear rod (rhabdom, r) due to the modification and elongation of four other cells. Around 86 THE BIOLOGY OF INSECTS each rod are grouped eight elongate cells which compose the retinula (rt) of the element, and each retinular cell is prolonged into a fibre that perforates the inner basement Fig. 29. — Details of Ommatidia of Compound Eye of Honey Bee, A, Longitudinal section of an ommatidium ; B, Surface view ; C, Oblique transverse section through apices of cones and outer region of retinulae ; D, Transverse section through retinulae and rhabdoms. cr, corneal facet and lens ; c, crystalline cone ; r, rhabdom ; rt, retinula ; />, pigment cells ; «, nerve-fibre. A and B X 450. C and D X 1000. After E. F. Phillips. membrane of the eye and passes on towards the optic ganglion. Each series of lens, crystalline cone, rhabdom SENSATION AND REACTION 87 and retinula makes up an element or o?nmatidium of the compound eye and the ommatidia (Fig. 28, o, 29) which correspond in number with the corneal facets, are isolated from each other by pigment cells (Fig. 29, p). As the ommatidia converge inwards from the surface of the eye it is evident that each of them is concerned with the reception of impressions coming along a path represented by a con- tinuation of its axis, since rays of light not closely parallel with this axis, which pass through the lens, will be absorbed by the surrounding dark pigment. The transparent cones allow the light- rays to reach the underlying rods and thus to excite the retinular cells that surround them. From the retinulae nerve impulses pass along the fibres to the optic ganglion, through a series of cell-stations the first of which form a periopticon (Fig. 28, p) lying within the basement membrane of the eye, and the second an epiopticon {e) in the outer region of the optic ganglion. From this latter the impulses pass to the central ganglionic opticon (op), a mass of nerve-cells in connection with the central brain. The nerve fibres passing between the various cell-stations undergo extensive crossing-over (decussation), so that some impulses started in the upper elements of the eye pass to the lower, and those from the lower elements to the upper. Such a scheme for the reception of the impulses in the brain ensures a certain amount of co-ordination of the multitudinous impressions that aflfect the central nerve-masses through the two large compound eyes of a highly organised insect. We may now pass to consider the nature of an insect's sight. The eyes are clearly adapted to receive visual im- pressions, but it is by no means certain that an insect sees as a vertebrate animal sees, and such questions as " Can an insect clearly discern the form of surrounding objects ? " or " Can it distinguish between various colours ? " have often been asked, and have been very differently answered by different students. In the simple eyes of insects, such as the three ocelli on the crown of a bee's head, or the group of such organs on either side of the head in caterpillars and other larvae, we 88 THE BIOLOGY OF INSECTS have seen that the cuticle is thickened so as to form a biconvex lens beneath which lie the retinal cells. Here, therefore, is an arrangement by which rays of light coming from surrounding objects may be brought to a focus on the retina so as to produce there an inverted image. From the analogy of structure of a vertebrate's eye and our own experience of seeing, there seems no reason to doubt that an insect's ocellus is adapted for the appreciation of definite form. But in contrast to the vertebrate's eye, the ocellus has no provision for accommodation, and the usual high convexitN' of its lens renders it available for seeing only objects that are close at hand, wliile the extent of its field of vision must be greatly restricted. A compound eye consists of a number of elements each consisting of a set of transparent structures — cornea, lens, cone, rhabdom — the last-named surrounded by the receptive retinular cells connected by fibres with the ganglionic centre towards which the closely arrayed elements converge. J. Miiller long ago (1826) pointed out that the general visual sensation induced through such an eye must be regarded as the sum of the multitudinous sensations due to the individual elements which, owing to the surrounding dark pigment are more or less insulated from one another. Such a hypo- thetical '' built up " impression suggested to Miiller the term " mosaic vision " from the analog}' of a picture made up of a number of small apposed pieces ; and this term has been generally accepted as being suitably descriptive of an insect's seeing with its compound eyes. But how far is such vision definite ? Miiller denied that any clear image could be formed in a compound eye and suggested that an insect receives no more than the impression of as many spots of light as there are elements, the slender pencil of rays traversing each element being concentrated at the apex of the cone and the intensit}^ of light appreciated through each element varv'ing with the source whence it comes. H. Grenacher (1879) recognised that the lens of each element is adapted for the production of an image, but believed that the position of the image is such as to render SENSATION AND REACTION 89 impossible its appreciation by the retinular cells. Recently, however, good reason for regarding compound eyes as capable of appreciating a distinct image has been given in an extensive study of the subject by S. Exner (1891). His work has been well expounded by H. Eltringham (19 19) who by observations and experiments of his own has carried it farther. Exner describes how in the common glowworm — the wingless larva-like female of the beetle Lampyris noctiluca, an insect active in feeble light — the pigment cells may move outwards towards the corneal surface, so that the deeper regions of the elements are no longer completely isolated, and some of the rays traversing a cone may affect not only its own retinula but also neighbouring retinulae. The result is the formation of a set of erect images partly super- imposed on each other, and it follows that the glow-worm may thus obtain clear and definite vision of a portion of its near surroundings. These superposition images are believed by Exner to be characteristic of the vision of those insects, such as moths, which fly in the dusk or at night time and have therefore to make the best use of feeble light. Eltring- ham has studied the images formed by the eyes of day- flying insects such as butterflies, dragonflies, and blue- bottles. In the case of the two latter groups, in which the cone-forming cells are imperfectly changed into the com- pletely " crystalline " substance, he agrees with Exner that the general apposition image induced may be best repre- sented to our imagination as *' a mosaic of light spots." But in the compound eyes of butterflies which possess fully developed transparent cones, ** there is at the apex of the cone a tiny erect image of that part of the field appertaining to each facet unit." Hence it may be inferred that an erect image of the whole field of vision may be appreciated by means of such an " eucone " eye. The reinversion of the images by the cone, so as to make them ultimately erect, disposes of the difficulty arising from the conception of an extensive general image, made up of a large number of minute individual images each regarded as reversed. But the clarity of an insect's vision as regards form is 90 THE BIOLOGY OF INSECTS presumably less than that of a vertebrate's because the retinal structure of an insect's eye is distinctly coarser. And it has already been mentioned that the lack of any provision for accommodation results in the insect being very short-sighted. *' We know," remarks Eltringham, " how readily one white butterfly will pursue and investigate another to see if it is a suitable mate, but I have never seen this kind of flirtation begin from a distance of more than a few feet." The compound eyes of insects, therefore, while giving their possessors a wide range of vision as regards direction, are ineffectual in perceiving objects at a distance, and in some cases only can they receive from objects close at hand any definite indications of form. A subject of great interest in connection with the sight of insects is their appreciation of colour. Information as to this cannot be obtained by microscopic examination of the eye structure, but approximate certainty has been reached by means of careful observ^ations and experiments. Reference has been made above to the fact that a butterfly often pursues one of its own species ; such action is due to visual recognition, because the attraction is also exercised by a dead dried specimen or even by a coloured model, as Eltringham has demonstrated. He found that a common " fritillary " (Brenthis euphrosyne) dipped directly to a spot on the ground where was lying a wing accidentally broken off from an insect of its own kind long dead. This observa- tion affords convincing evidence that the butterfly could recognise the characteristic colour of the wing, for any specific scent must have been for a long period absent from such a dried fragment. Many investigators have concluded that insects of various kinds distinguish the colours of flowers on account of the apparent preference which they show for certain hues. Eltringham watching Vanessid butterflies on a bed of asters, white, pink, and purple, found that of 427 visits, 47 were to white, 135 to pink, and 245 to purple flowers, though the purple blossoms were only three- quarters as many as the pink. H. Mliller (1878) was long ago led to the opinion that the colours of flowers serve as SENSATION AND REACTION 91 an attraction to insects that visit the blossoms to obtain nectar, and the doubts that some subsequent observers have thrown on the existence of a true colour sense in insects do not seem to be justified. An experiment of J. Lubbock (1882) often quoted, and several times subsequently verified, is especially valuable in this connection. He placed honey on a coloured paper disc and thus trained bees to associate a particular colour (red or blue, for example) v^ith the location of food ; then he found that a bee which had once found honey on a blue disc would return to a disc of that colour even though the food were now not on a blue but on a red one. This experiment shows that in such cases the attraction by colour must be more powerful than the attraction by scent, and confirmation has been afforded by Forel and others who, having varnished or removed the feelers of various insects, found that reaction to the colour stimulus remained unaflFected. We may conclude, therefore, that the eyes of many insects enable the creatures to distinguish the forms and colours of near objects ; Eltringham estimates that butterflies recognise members of their own species when about a yard away. Those with large compound eyes can see in many directions at once. Beyond the restricted distance for which the lenses and cones of its eyes afford suitable focus, an insect can appreciate changes in the intensity of the light falling on its corneal area. Any one who passes his hand above the station of a resting fly can demonstrate that an insect perceives a moving shadow, readily if the motion be rapid, less distinctly if it be slow. If the observer is trying to catch the fly with his moving hand, he may go farther in his interpretation of the insect's behaviour, and conclude that the approaching shadow " startles " or " frightens " it when he sees it dart suddenly away. This reference to the possible sensations or mental experiences of insects recalls the opening remarks of this chapter on an insect's response to stimulation from a source of light, and the inferences that may be drawn from the creature's behaviour. It may be advisable to return to such 92 THE BIOLOGY OF INSECTS questions before bringing tliis section of our study to a close. Manv insects, like the proverbial moth, fly directly towards a lamp, and the obser^'ations of J. Loeb (1905) and others show that a definite relation of the creature's body axis to the source of light is always brought about. The hght seems, one may say, to exert a pull on the insect, inciting it first to turn head on towards the light, and then to move its body as a whole along the path of the luminous rays in the direction opposite to their course, so that at length it flies against the lamp. Such a response is usually called a tropism, because the insect is ine\'itably turned in a certain direction and along a certain path under the impact of the stimulus. As this response draws the insect towards the source of light it is defined as " positively phototropic." The same reaction is shown by many lars'ae as well as by winged insects. Caterpillars commonly move towards the light. " If they are obKquely placed," writes E. L. Bouvier (1922), " on a plane surface opposite to the source of light they quickly turn their heads, follo\^'ing the axis of the light ray and then the rest of the body moves in this direction." On the other hand, there are insects on which the influence of light is the exact reverse of this ; they turn away from it, as may be observed by W. B. Herns (191 1) in the maggots of flies or bluebottles exposed to bright daylight or strong lamp- light. These lan-ae first direct their heads from the source of light and then move away towards the gloom. So experimenters call them " negatively phototropic." It is important to realise that such terms as these, while con- venient for analytical definitions of behaviour, offer no explanation of such beha\*iour ; they simply express in tw^o or three long words that some insects appear to seek and others to shun the light. But the use of the word tropism to define such responses of insects to stimulus is generally taken to imply that the response is automatic and not necessarily accompanied by any distinct sensation or experience ; the term was long ago applied to describe the behaviour of plants in response to luminous and other SENSATION AND REACTION 93 stimulation from their surroundings, and nobody has seriously suggested that the growth of a plant shoot towards the light is accompanied by sensation in the organism. According to Loeb's interpretation of phototropic reaction, the incidence of light on one side of the sensitive head region of the insect leads to a corresponding one-sided action of the body-muscles so that the creature is necessarily turned in the direction of the rays of light. Then the '' luminous intensity being the same on both sides, there is no reason for the animal to turn from this direction either to the right or to the left." It must now directly approach or recede from the source of light. It is noteworthy that the fly- maggots taken as examples of negative phototropism have no eyes, and there is no reason for attributing any definite visual sensation to them. But insects Hke moths that fly into a lamp have eyes, and even though their reaction to the light be automatic and inevitable, it does not necessarily follow that they have no appreciation of the light towards which they are irresistibly drawn. The apparently suicidal action of a moth flying at last into a flame may be due to the " pull " of the light in rapid alternation on either side as at close quarters the insect turns to and fro. Other responses to stimulation mentioned in this chapter may be recognised as referable to the group of tropic actions. The male moth " allured " by the caged female, or the female fly attracted by carrion to her egg-laying, is positively chemotropic, while the black aphids which climb as high as possible on the bean-shoots where they feed are negatively geotropic ; they turn from the earth with its gravitational attraction. The consideration of these and similar actions leads naturally to the wide subject of behaviour. CHAPTER V BEHAVIOUR, INSTINCTIVE AND INTELLIGENT In the preceding chapter our consideration of the sense- organs of insects and the reactions associated with their working has led to comparisons between the sensations and responses of insects and those of back-boned creatures, including our own race. The behaviour of insects — especially of those groups which exhibit a highly developed family or social life — has been repeatedly used by earnest teachers of mankind to stimulate their fellows to more strenuous habits of life and work. " Go to the ant, thou sluggard ; consider her ways and be wise " was the call of the Hebrew seeker after wisdom to the men of his day, and such calls have been echoed since through the ages. Survivors from the nineteenth century remember how in their youth they were ** exhorted to virtue " in the verse of Isaac Watts :— " How doth the Httle busy bee Improve each shining hour, And gather honey all the day From every opening flower 1 " But some years before that century closed F. Anstey had suggested that the " modern child " of the period might be expected to reply to such exhortation in this wise : — " How doth the little bee do this ? Why, by an instinct blind. Cease then to praise good works of such An automatic kind." We will return to the specialised activities of ants and bees in a later chapter ; the points of view contrasted in the two verses just quoted are, however, among those 94 BEHAVIOUR 95 applicable, in the opinion of different students, to the behaviour of insects generally, and indeed to that of animals lower or higher in the scale of life than they. Already comment has been made on the tendency to credit insects, when they show response to stimulation of various kinds, not only with sensations but with states of consciousness — pleasure, discomfort, terror — comparable to those which we realise in our own experience. The food-gathering activi- ties of ants and bees look like conscious efforts directed intelligently to an obvious purpose. That is the assumption underlying the belief that such insects display industry and foresight in their work. On the other hand, Anstey's verse suggests that the purposeful activities of insects may be carried on without design or knowledge on their part, and such an outlook on the subject has become increasingly popular during recent years. It is undeniable that a large proportion of the actions of insects are reflexes resulting directly from various stimulations from outside, which call forth inevitable responses through the nervous system of the insect acting on its muscles. The term " instinct," often used somewhat loosely to describe the causes of actions not due to intelligence on the part of the agent, has at its root the idea of " need " or '' urge." But such urge or impulse arises as the response of the organism to stimulation, and instinct was therefore defined by Herbert Spencer as " com- pound reflex action," and a creature's instinctive behaviour has been regarded as the sum of its responses to environ- mental influence. Stimulations are being continually received by means of the various sense-organs, and the insect's nervous system is of such a nature that certain responses follow in each case. The simplest form of response is some kind of tropism such as was described at the close of the previous chapter (pp. 92-3), and some recent distinguished students of insect behaviour — ^A. Bethe (1898) and J. Loeb (1905), for example — believed that all the life-activities of the creatures can be explained as a complex of tropisms. Light, gravity, contact with particles of soil, odorous plant secretions, all affect the insect and the 96 THE BIOLOGY OF INSECTS result in each case follows certainly, so that the behaviour of a bee among blossoms may on this interpretation be strictly compared to ** the behaviour of iron filings in the magnetic field." On this view the observable actions of an insect depend altogether upon the kinds of stimulation to which it may be subjected, along with the innate tendency of its tissues and organs to respond to such stimulations in particular ways. A moth flies into the candle flame that destroys its life, because it must needs react to a source of light by way of direct approach. And on the other hand, a female housefly makes for a heap of refuse and lays her eggs therein, because the smell of the refuse is an attraction, and the nerve impulses pass on to the centres that control the actions of the genital ducts and ovipositor, so that the laying of eggs is part of the bundle of reflexes, which in this case leads not to the curtailment of the insect's life but to the perpetuation of its race. Such insect activities as can be explained by this simple ** tropic " formula may therefore be either harmful or beneficial to the species. What impresses the thoughtful observer is that if strikingly adaptive actions are the result of simple reflexes, there remains the further question why the combined reflexes work out to a beneficial end ? While the study of tropisms undoubtedly helps the naturalist to analyse the activities of insects, it appears that the explanation of all kinds of insect behaviour as due to these reactions maybe deceptively simple. The observer who attributes to an insect motives and states of consciousness like his own, because he sees a moth fly towards a lamp or an ant scuttle into a burrow, is misled by false analogies of behaviour. But such an observer at least recognises that the insect that he watches is alive, while the rigid follower of the tropic explanation, who is certain that all the creature's actions are inevitably determined by the nature of the stimulations from without that affect its nervous system or act directly on its various tissues, seems to overlook those stimulations which may originate within the organism and give rise to aspects of behaviour un- foreseen and inexplicable. BEHAVIOUR 97 Some recent experimental work by D. E. Minnich (1919) on the reaction of Hive-bees (Apis mellifica) is of importance Fig. 30. — a-f. Two trails of each of six normal Honey Bees {Apis) in directive light, g, h, i, three successive records of trail of another normal bee in directive light, the first two {g and h) being altogether contrary to expectation. After D. E. Minnich {Journ. Exper. Zool. xxix, 1 919). H 98 THE BIOLOGY OF INSECTS in this connection. Bees are markedly phototropic, stimu- lated by light, flying or crawling towards a definite source of illumination. Minnich shows, however, that individual variations may be apparent in their responses. Bees generally turned rapidly so as to face, then crawled towards, a '* directive " light, but only about a quarter of the number whose course was carefully traced, made for the lamp in an approximately straight Hne, the rest show^ed more or less dcN-iation, and in a few cases the path was markedly indirect. Out of seven bees tested on one occasion, six responded by making directly for the source of light (Fig. 30, a-f). The seventh on its first trial wandered around and away ; fourteen minutes later it went by a less circuitous route away from the light ; but a minute later, starting from a point 30 cm. nearer to the lamp, it made for it though not in a straight Hne (Fig. 30,^, //,/). "This example shows," comments Minnich, '' that even the constant response of the bee to directive illumination is not free from abrupt and apparently inexplicable departures." Experiments with bees allowed to wander in a uniformly illuminated area (" non-directive light ") demonstrated much variation in their individual beha\iour. '' The animal may turn markedly towards a given side in one trial, and in the next turn quite as markedly towards the opposite side. Again, the direction of turning may be completely changed several times in the course of a single trial." It seems clear, therefore, that the reaction of at least some hive-bees to visual stimulation is not so fixed as to be absolutely predictable. The bee may comince the obser\^er that she is alive by behaving sometimes in an unexpected manner. A set of experiments performed by Minnich on bees with one eye blackened over so as to be impervious to light or almost so, is of ver}' special interest. Many observers have tried similar experiments with various insects, and find that their reaction to directive light is to " loop " towards the side on which the eye is still functional. INIinnich showed that some of the bees subjected to such treatment learned by experience to modify this reaction so that their BEHAVIOUR 99 course came to approach that of a normal bee with both eyes in use (Fig. 31). Through a number of trials some of these insects were observed gradually to reduce the number Fig. 31. — /, ni, two trails of a Honey Bee with right eye obscured ; n, 0, two trails of a Bee with left eye obscured, showing tendency to turn constantly towards the side of the functional eye. p, r, two trails of a Bee with left eye obscured, which behaved as a normal insect approach- ing the source of light. After D. E. Minnich {Jotirn. Exper. Zool. xxix, 1919). In Figs. 30 and 31 the small circle represents the source of light and the thin straight line the direction of the rays. and frequency of their *' circus " movements, until after twelve or twenty trials, they made for the source of light in 100 THE BIOLOGY OF INSECTS an approximately straight line, as normal bees would do. It is noteworthy, on the other hand, that some of the bees used in these experiments never seemed able to modify their response from that at first incited through their *' one- eyed " condition ; they were far less ready than others to learn by experience, more completely dominated than their fellows by the abnormal reflex arising from their abnormal condition. Reference has been made to the definition of instinct as " compound reflex action," and Lloyd Morgan in his classical discussion on Animal Behaviour (1900) remarks that " instinct supplies an outline sketch of behaviour to which experience adds colour and shading." In the changed behaviour of Minnich's bees we have an example of the modification of a simple " instinctive " response through experience acquired by certain individuals subjected to unusual conditions in the course of their lives. The fact that some creatures become more easily adapted than others to new and strange conditions suggests the possibility of our learning a little as to the origin of new modes of behaviour among insects from such methods of investigation. These results further lead us to conclude that an insect may be able to escape from the domination of an abnormal tropism which prevents or retards the creature's normal response to a stimulation. A bee with both eyes in action directs the axis of its body towards, and then moves towards a source of light, whereas a bee with the left eye blinded loops towards the right under the stimulation of a bright light. The fact that it may learn how to correct this abnormal reflex and revert to its former mode of behaviour suggests that in performing the normal reflex action the insect experiences at least a gleam of consciousness akin to the satisfaction which we realise in ourselves as we perform many normal reflexes, and which we may infer is realised to some extent at least by the higher vertebrates generally. With insects, however, it can hardly be doubted that the consciousness accompanying reflexes of this kind is weaker by far than with vertebrates ; the '* works " of the *' little BEHAVIOUR loi busy bee " are indeed mostly " of an automatic kind." The segmentation of the central nervous system in insects is a structural feature which suggests imperfect integration of nerve-action, and consequent imperfect individuality, and observ^ations following mutilation, whether deliberate or accidental, convince us that there is considerable exercise of independence between various parts of an insect's body. In a trisected wasp, for example, the jaws and mouth con- tinue to feed and the thorax to walk, while stimulation of the abdomen by a careless observer may lead to unpleasant demonstration that the sting will act as the result of a reflex conducted through the abdominal nerve-centres without possibility of reference to the brain. Insects of various orders have the habit known generally as " death-shamming " ; when touched or handled they lie motionless on the back with the limbs strongly flexed. This is a reaction to contact with certain objects, whereof a human hand is one. H. H. P. and H. C. Severin (191 1), who have carefully studied such reactions in aquatic bugs (Nepa and Zaitha), conclude that the contact-stimulus works so as to bring the muscles into a state of intense contraction — a kind of prolonged tetanus. " Nepa, while feigning death, may be taken by any tibia or femur and held in a position so that the weight of the entire body is borne by the extensor muscles of a single segment of one leg." Here the head nerve-centres appear to exercise an inhibitory power over the muscles while the reaction lasts, for if a death-feigning Nepa be beheaded, the muscles immediately relax, while if the insect be bisected across the thorax the hmbs in front of the cut remain flexed while those behind it relax. This term *' death-shamming " has been commonly applied to this habit because it has often been regarded as a voluntary and purposeful method of behaviour adopted by the insect in the presence of some recognised danger with the object of securing safety. There seems no reason, however, for regarding it as anything beyond a simple automatic reflex, which may in some cases serve as a pro- tection to the creature that adopts it. 102 THE BIOLOGY OF INSECTS No doubt can be felt that a very large proportion of an insect's normal activities are instinctive, in that they result from a set of complicated reflexes, while they serve to ensure the survival either of the individual or of the race. This is especially evident in modes of behaviour concerned with reproduction and growth among insects whose manner of life when adult diflfers from that in the early stages of their life-history. The actions that accompany egg-laying by a female moth or digging- wasp, for example, are all directed towards the provision of environment and food suitable for the larva. The moth, herself a feeder on the nectar of flowers which she sucks, lays her eggs on the leaf of a plant — often of some one definite species — which the caterpillar will devour. Either the plant provides a stimulus to egg- laying through the senses of smell or sight and the act is instinctive, or the moth remembers how she fed when she was a caterpillar and provides for her young accordingly. No student of insect behaviour would seriously suggest the second alternative as probable, and we feel compelled to accept the first. A digging-wasp makes a nest, usually by excavating a pit in the ground, and either before or after this labour, hunts for prey to bury along with her eggs so as to ensure a supply of food for her grubs. In most cases that have been carefully observed, the behaviour of the mother insect is so uniform that she may be said to follow a definite routine, each step in the process apparently suggesting the next, so that nothing is done out of its regular order. Some observations of G. W. and E. G. Peckham (1898) are of great interest in this connection. Some American species of Pompilus, studied by them, capture spiders to serve as food for the grubs ; the female wasp paralyses the spider with her sting, then places it with its waist in the fork of a plant-shoot, so that it will not fall, and then proceeds to dig the hole for her nest. Wishing to observe carefully how the wasp stings the spider, the Peckhams on one occasion removed from its place on a bean plant the paralysed spider which a female Pompilus had just put there, and substituted an uninjured spider. The BEHAVIOUR 103 wasp, after digging her nest, returned to the bean-plant in search of the paralysed prey, saw the uninjured substitute, but would not touch it. After several fruitless searches, the wasp went away, caught and stung another spider, placed it in the usual position on the bean-plant, and then dug another nest, although the first made one was ready and empty. The break made in the insect's usual routine by the observer's act, resulted in the whole process being started again from the beginning. Apparently the wasp recognised that the spider she found had not been stung, but she did not attempt to deal with a normal spider already in place on the plant, she was impelled to go and hunt for another. Then having stung this, she proceeded to the usual next step of digging a nest ; the available empty nest that she had made shortly before was neglected because the work of nest- making always follows, in the instinctive cycle, the stinging of the prey. The hitch in the work due to the removal of the first victim led to a repetition of the whole process from the start ; the wasp showed no adapt- ability to unusual conditions by making a change in the usual sequence of her actions. Her nervous system appears to be so attuned to the various stimulations and experiences that each of them, as it is felt or completed, becomes an incitement to the " doing of the next thing." Yet these facts are no justification for denying that the insect may be a conscious being, even though its conscious- ness be dim and feeble as compared with that of a vertebrate. It is reasonable to believe that the insect's instinctive routine has, to quote Lloyd Morgan, *' a psychological aspect of awareness and desire." Though the " outline sketch of behaviour " which is drawn, as it were, for the insect by its long-inherited instinct, is rigid and unvarying, there may be opportunity, at least in some instances, for the addition by experience of " colour and shading," as well as clear evidence of individual memory. For example, the Peckhams described in detail the behaviour of a female pompilid wasp Aporus fasciatus which had captured a spider larger than herself and left it on a melon leaf while she sought a suitable 104 THE BIOLOGY OF INSECTS place for nest-making. *' At length she went under a leaf that lay close to the ground and began to dig." The observers removed the leaf so as to watch the wasp's actions more closely ; after ten minutes she flew away to where she had left the spider, and then returned to seek the un- finished nest, but ** it was evident that some landmark was missing — when she reached the spot she did not recognise it. At last we laid the leaf back in its place over the opening, when she at once went in and resumed her work, keeping at it steadily for ten minutes longer." But this nest was never finished ; if we may judge from the wasp's behaviour its position did not satisfy her, for she filled it up and started four others in succession, each to be in its turn abandoned and filled up. At the sixth trial the nest was completed and the spider dragged in. Lloyd Morgan is justified in his comment on this description " that it shows an amount of apparent fastidiousness which is quite irreconcilable with the hypothesis that the behaviour is merely instinctive." It shows also that the wasp recognised the position of her first nest by the leaf lying over its mouth ; her behaviour suggests inevitably that she remembered that leaf so that it served as a guide to the nest. Somewhat similar results were obtained by C. Ferton (1905) in his observations on certain kinds of bees (two species of Osmia) which make their nests in empty snail- shells. After depositing honey and pollen and laying an egg, the bee closes the mouth of the shell with fragments of leaves worked up with her spittle. One species, Osmia rufohirta, has the habit of rolling the shell, after provisioning the nest and laying her egg, to some sheltered spot, returning to the latter with the covering of leaf- fragments. A female of this species was observed by Ferton to go and return between the place where she was leaf-gathering and the hiding-place of the shell by way of the station at which she had first found the latter ; she travelled over her former tracks, a recognised path. While the bee was engaged on the journeys necessary for the completion of her task Ferton removed the shell from its hiding-place to a position BEHAVIOUR 105 some six inches away. The Osmia failing to find the shell where she had left it, proceeded to hunt about until she recovered it, and then for several subsequent journeys made her flight to its new station by way of its two former resting- places. Afterwards, however, she was seen to go directly to the new station : " Little by little the images of the former places of her nest are effaced in the memory of the insect." E. L. Bouvier (1920) remarks on these observations that such insects " know how to manage things and to meet the most unexpected situations. They do not act as automata ; the memory that guides them in these circumstances seems indeed from its essential characteristics to belong to the same degree of psychism as the human memory." It may be doubted, however, whether such behaviour, remarkable and instructive though it is, can be regarded as of '' the same degree " as the memor}^ of man. Of all the unusual modes of behaviour by insects of which we have reliable record few have appealed more strongly to the imagination of naturalists than the work of the North American digging-wasp Ammophila urnaria, that captures, stings, and buries caterpillars as a food supply for her grubs. Most of the species of Ammophila dig their nests in the soil before hunting for prey to deposit in them, and a completed nest may harbour four or five paralysed caterpillars with as many Ammophila eggs. One of these females covers the mouth of her burrow with pellets of earth, but normally never fails to find the nest on her successive returns from the hunting. When the nest is fully stored and the eggs are all laid, the wasp finally closes up the mouth. The Peckhams (1898, pp. 6-32), who paid especial attention to the habits of these wasps, comment on the difference in behaviour shown by individual females in the work of nest-closing, some being much more careful in performing their task than others. " Of two wasps that we saw close their nests on the same day, one wedged two or three pellets into the top of the hole, kicked in a little dust, and then smoothed the surface over, finishing it all within five minutes. . . . The other worked . . . for an hour. io6 THE BIOLOGY OF INSECTS first filling the neck of the burrow with fine earth which was jammed down with much energy . . . and next arranging the surface of the ground with scrupulous care and sweeping every particle of dust to a distance." Finally, after unsuccessful trials with a small stone and a lump of earth, she laid a leaf over the closed mouth of her burrow. But it is another individual Ammophila that the Peckhams have made famous by observing how, after filling the hole with loose earth and ramming it down with her head, and continuing this process until the hole was full of soil to the ground level : '' she brought a quantity of fine grains of dust to the spot and picking up a small pebble in her mandibles, used it as a hammer in pounding them down with rapid strokes, thus making this spot as hard and firm as the surrounding surface." In connection with this very remarkable incident, it is instructive to note that females of Ammophila sometimes bring small stones to serve as stoppers for the mouths of their burrows. The behaviour of the individuals that have been thus seen to use such stones for pounding down earth may perhaps be regarded as a further advance in the intelligent use of materials serving for nest- making, something beyond the usual habits of their kind. Digging wasps paralyse their prey — whether spiders or caterpillars — by stinging the victims repeatedly, frequently applying the sting along the line of the ventral nerve-cord beneath the body. It has often been stated that the opera- tion of stinging is always performed in the same way by all females of the same kind, and that it results in the paralysis of the victim, so that it cannot move, but not in its death, so that the wasp's grubs when hatched will find fresh food in a living though helpless prey. In order to bring about this result it is necessary that the wasp should sting the caterpillar accurately along the ventral nerve-cord so as to pierce the series of ganglia wherein are the nerve-centres that control the movements of the various segments. These considerations have led some enthusiastic naturahsts to imagine that the wasp must have a knowledge of the anatomy of the caterpillar and of the functions of its nervous system. BEHAVIOUR 107 The facts, as shown by the careful observations of the Peckhams, are that the part of the body where the cater- pillars are stung varies with different females of the same species of wasp ; only occasionally does the wasp inflict a series of stings along the mid ventral line. Moreover, the victims found in the wasps' nests are often dead, and the wasp grubs can feed and grow if supplied with a dead or even decomposing victim. '* We believe that the primary purpose of stinging is to overcome resistance and to prevent the escape of the victims, and that incidentally some of them are killed and others paralysed." Such considerations warn us that the student of insect behaviour who does not allow for individual variations in the habits of the same species may readily fall into the error of regarding all insects as utterly unconscious automatic machines, or into the opposite error of assigning to them intelligence and fore- sight of a degree far beyond that warranted by the facts of their behaviour. The habits of egg-laying and providing in advance for the needs of offspring as yet unhatched, some of which have been briefly considered, are clearly to a large extent in- stinctive ; most details of an insect's complex behaviour directed to these ends result from her inborn tendency to respond in certain ways to suitable surroundings and stimu- lation. It is instructive, in this connection, to turn to some examples of behaviour in larval insects which appear to suggest prevision of the needs involved in the future final transformation into the adult form. Such prevision is of course impossible ; a highly imaginative observer might convince himself that a butterfly laying eggs knows what will happen to her offspring after hatching because she was once a caterpillar herself, but he could hardly be persuaded that a larva foresees its transformation into a winged insect, and knows the conditions, often seemingly difficult, that will have to be overcome in connection therewith. Yet numerous larv^ae of different orders follow specialised modes of behaviour which are related to their future needs, and thus afford striking example of the part played in the life io8 THE BIOLOGY OF INSECTS of insects by the sets of activities which we generally call instinctive. It is well known that during the pupal stage which intervenes between the larval and the adult condition in the vast majority of insects, the creature remains, as a rule, quiescent and does not feed. In such a resting condition the insect needs protection, and its behaviour when nearing the close of larval life is largely concerned with this coming need. Before the moult or casting of the last larval cuticle which reveals the pupa (see Chap. VII, p. 172) the full-grown larva often spins a silken cocoon, as do the silkworm and many other moth- caterpillars, the cocoon in some species being strengthened with the larval hairs or bristles, or with foreign substances such as chips of wood or particles of soil. Such cocoons serve as shelters for the pupae resting within them. Or the larva, if it does not make a cocoon, often seeks shelter by burying itself in the ground as many hawk-moth and owl-moth caterpillars do, or by creeping beneath a loose piece of bark, like the small cater- pillar of the codling-moth that has fed within a growing apple on a tree. But if a pupa lies enclosed in a cocoon, the perfect winged insect, when developed, has to make its way out of the cocoon after undergoing the moult that sets it free from the pupal cuticle. Often it is found that some provision for this need also is made beforehand by the larva. The silk- worm's cocoon is left comparatively thin and weak at the head end where the moth will have to come out, so that in this region it is readily weakened and partly dissolved by a fluid which the moth, when developed, discharges from its mouth. The same provision is found in the behaviour of the remarkable caterpillar of the Puss Moth {Cerura vinula) which forms a hard and dense cocoon, but leaves in front a weak area which is readily acted on by the strong alkaline fluid that the emerging moth discharges from its mouth, as O. H. Latter (1895) has shown. Still more remarkable, perhaps, is the behaviour of larvae which feed in some object or substance out of which PLATE 11 A. Row of Eggs on Cow's Hair B. Two Eggs cleared to show unhatched Larvae, {three are hatched). X 5. X 45. C. Fiist-stage Larva, showing Mouth-armature D. Second-stage Larva, and Air-tuljes. X 50. showing Muscles. X 6. ' Eggs and Larvae of Hypodcnna lincatum. To face p., o^.-] [T. Price, photo. BEHAVIOUR 109 the perfect insect has ultimately to make its way ; in such cases the larva before pupation has the habit of coming close to the surface. The large caterpillar of the Goat Moth (Cossus), for example, feeds for more than a year on wood as it tunnels through the timber of some tree. When fully grown it usually comes to the bark before it makes its cocoon of chips of wood fastened together by its silky secretion. This is an insect whose pupa works its way partly out of the cocoon before undergoing the final moult which releases the moth, and as the cocoon has been formed close to the surface of the tree trunk or branch, the greater part of the pupa's body projects into the air, so that there is, after that moult, no obstacle to the free emergence of the moth. The small larva of a seed-beetle (Bruchus), when hatched from the egg laid on a leguminous blossom, bores into the carpel and enters the developing seed, where it tunnels and feeds until fully grown. Before pupation, however, it makes its way to the circumference of the cotyledon just within the seed-coat (testa) where it pupates. Thence in due time the beetle emerges, though in several species of this group it may remain resting there for several months before it bites its way through the skin in preparation for egg-laying on the blossoms of the succeeding spring. In many cases the objects within which insect larvae feed are the living bodies of other insects or of some larger animal at whose expense they carry on their parasitic existence. The modes of behaviour of insect parasites are as remarkable as their structure, and often have a definite bearing on the later stages of their own life-histories. As an example we may sketch briefly the wanderings of the maggots of the Warble-flies (Hypoderma) through the bodies of the cattle wherein they live. The female flies lay their eggs on the hairs of the legs or the lower parts of the bodies of grazing cattle. Immediately after hatching the tiny maggots crawl along the hair and bore into the skin. Once beneath the skin they work their way upwards, their behaviour suggesting that they have the " negatively geo tropic " reaction. But their migrations are too com- no THE BIOLOGY OF INSECTS plicated to be explained as due to one tropism only, for they find their way to the sub-mucous coat of the gullet, where they rest or wander to and fro for several weeks, and whence they travel backwards by way of the dorsal muscles or the diaphragm and the vertebral canal to a position just beneath the skin of the host-animal's back. Their final resting-place is never far from the line of the beast's backbone, which suggests that negative geotropism may still be one factor governing their behaviour, and not long after their arrival there they perforate the beast's skin. This action ensures a supply of fresh air to the spiracles at the tail-end of the body directed towards the hole during the later weeks of lars'al life, when the maggots have attained a considerable size, and also provides for the maggot when it is ripe a way out of the host's body ; it works through the skin, falls to the ground, seeks shelter, and pupates (Plates II, III, XVI). Many more examples might be given, but these are sufficient to indicate how often an insect's behaviour at some period of its life has reference, not only to its present need for food and possibly for shelter, but also to future contingencies in its gro\\th and development which it cannot possibly foresee. These actions are remarkably purpose-like, but the creature that performs them can have no knowledge of their purpose. The tendency to react in certain ways to the environment and its stimulation is part of the insect's inherited nature ; it is so bound up with the stor}^ of the race, that many thinkers on these questions who realise that the memor}- or foresight of the individual can play no part in the appointed process, do not hesitate to idealise that process by suggesting that it implies a '' racial memor>^ " so impressed on the species that the appropriate Unes of behaviour are followed by adult and larva as one generation follows another. The behaviour of social insects is a subject of special interest, and some details of this will be discussed in a later chapter. In closing this general sketch of insectan behaviour it may be noted how in many cases a large insect population of the same or of allied species, not practising social life PLATE III A. Skin of Cow's Thigh (Hair clipped) with Entrance Holes of First-stage Larvae of Hypodenna lincatum. a B. Second-stage Larva of II. lineatum IN Sub-mucous Coat of Gullet of Bullock, [a to a, Mucous Coat cut away.) To face p. no.] C. Final Stage Larvae of H. bovis IN '* Warbles " beneath Skin of Bul- lock's Back. [A and B, H. Britie?i, photo. [C, T. Price, photo. BEHAVIOUR III in the accepted meaning of the term, may undergo marked changes of habit on account of some new factor in their sur- roundings. Such new factors are often due to man's uncon- scious intervention, and the resulting changes of behaviour among insects may therefore be noticed by those naturalists who pay especial attention to those insects that feed on plants cultivated as farm or garden crops, or that live as parasites on domestic animals. There is a beetle, a small weevil, Orchestes fagi^ that is very common on beech trees in this country, the adults eating the leaves into holes, and their grubs mining between the upper and lower leaf-skins. During recent years both in England (F. V. Theobald, 191 2) and in Ireland (G. H. Carpenter, 1920) it has been observed that large numbers of these beetles, blown off beech trees into apple orchards, take to feeding on the growing fruitlets — a new kind of tree and a different part of the plant as compared with the normal feeding place of their kind. Also a sucking insect, a plant bug Plesio- coris rtigkolliSy whose normal food is the sap of willow leaves, has been noticed repeatedly since the observations of F. R. Fetherbridge and M. A. Husain (19 18) piercing the skin of young apples to suck their juice. It is highly probable that these are new modes of behaviour adopted by members of these and other species that have been accidentally brought into touch with a new plant on which they are able to feed. Similar changes in habit are undoubtedly new departures when they result from a newly introduced factor in the surroundings of the insects that display them. Since the beginning of this century tobacco has been grown in Ireland on a small scale, and a crop in Co. Kilkenny was found to be severely injured by multitudes of springtails feeding on the leaves. These on examination proved to belong to a north European species Isotoma tenella, never before recognised in these countries. The minute insects had certainly not been introduced with the tobacco, which was raised from seed, and taking all the facts into consideration, no doubt can remain that the presence of a large number of plants of a 112 THE BIOLOGY OF INSECTS kind new to the country on a comparatively small area, had provided so favourable an environment for the springtails that a species, formerly so scarce as to be overlooked, forced itself on the attention of the cultivators by multiplying so fast as to become a " pest." Here the insects in their thousands all responded in the same way to the stimulation of their new surroundings by feeding on the plants with which they were thus for the first time brought into touch. Another similar case of a more remarkable kind has been furnished through the introduction of great flocks of sheep into Australia during the last century, and the reaction of this introduction on some Australian flies during the last thirty years. It is well known that in the British Islands and in other countries of Western Europe a greenbottle fly {Liicilia sertcata), belonging to a group the usual behaviour of whose members is to lay eggs on carrion, has the habit of depositing eggs on the wool of live sheep into whose skin and flesh the maggots when hatched eat their way. (See G. H. Carpenter, 1902, and R. S. Macdougall, 1909.) Such abnormal behaviour is a response which this species has been making for centuries in England to the presence of thousands of sheep in the fields and on the hills ; the female fly is attracted by the odorous secretions and excrements of the sheep, and her approach to the animals for the purpose of egg-laying may be regarded as a typical chemotropic action. The student of insect behaviour cannot but wonder why one and only one kind out of the scores of nearly related flies should commonly adopt this habit, at once horrifying and interesting. In Australia, where the sheep-grazing areas have been constantly extended with the advance of human settlement into the interior of the continent, a precisely similar response to the presence of flocks of sheep has been made by at least five or six of the native species of the bluebottle and greenbottle group of flies. W. W. Froggatt (191 5-18) has found that besides Lucilia sertcata and L. caesar, presumably introduced from Europe, four Australian species of Calliphora, an Ophyra, and a Sarcophaga act as '' Sheep-maggot flies " on the vast BEHAVIOUR 113 grazings of New South Wales. In connection with the special aspect of insect behaviour now under discussion it is noteworthy that only since 1895, or thereabout, has this habit ** become a real menace to sheepowners." Here the introduction into the surroundings of certain Australian insects of large flocks of sheep has led to a response by myriads of female flies which has resulted in their larvae feeding on living instead of dead flesh, and has incidentally aff"ected seriously an important activity of mankind. These examples of changes in behaviour on a large scale by myriads of individual insects of the same kind serve to emphasise the fact that the habits of these creatures may be remarkably plastic as they are subjected to changes in their environment. While the experimental methods of laboratory study often enable the student to predict with confidence how an insect will react to various kinds 0/ stimulation, the wider survey of insect behaviour as it can be observed by the naturalist in the open country aflfords evidence that the creature's actions are not universally and unalterably stereotyped. The races of the insect world have reached their present conditions of form and activity through the long and changeful history of many generations, and it is fascinating to be afforded clear glimpses of changes in behaviour which assure us that by careful study of the activities of the living insects around us we can learn at least something of the course of that still unfinished story. CHAPTER VI REPRODUCTION AND HEREDITY Our discussion on Behaviour in the previous chapter has suggested that many of the actions performed by insects depend upon the constitution which they inherit through their parents and ancestors ; the insect comes into the world with its nervous and other systems " set " in special ways, so that its behaviour is as a rule like that of its parents under similar conditions. It is also obvious and generally recognised that in^sects — like other living creatures — resemble their parents in form and structure even to minute details ; yet beginners in the study of such insects as the moths belonging to certain groups quickly realise that even among a family reared from the same batch of eggs there may be a considerable degree of individual variation. All living creatures arise as the offspring of pre-existing living creatures. It is of interest to recall that less than two centuries ago many learned men beHeved and taught that such insects as maggots might be ** bred " or *' spontaneously generated " by dirt or carrion, although these beliefs had long before been confuted by F. Redi (1671), who showed that the maggots which feed in flesh develop into flies, and that no maggots can appear in flesh which is carefully screened so that flies cannot lay eggs on it. The genetic chains of living creatures around us, and including indeed our own race, are so familiar as part of our accepted world that we readily overlook the problems that present themselves when we consider the means by which inherited characters, whether of form, appearance, or behaviour, are passed on through the successive generations 114 REPRODUCTION AND HEREDITY 115 of a race. These problems have claimed much attention from naturalists in recent years, and the partial solution of some of them has been reached through studies on insects by various methods of research, such as the careful micro- scopic examination of their germ-cells and the tracing out of factors in their inheritance by means of experimental breeding. Among living beings there are two well-known methods of reproduction, firstly by means of small living units known as germ-cells, secondly by the strong outgrowth of a portion of the organism which may be termed a bud, the bud often separating subsequently from the parent-body. Any kind of reproduction that can be regarded as budding is very rare among insects ; they '' increase and multiply " as a rule directly from the germ-cells, to which there- fore attention may now be directed. It has already been mentioned in our Introduction (Chap. I, p. 10) that the germ-cells of insects, as of animals generally, are of tv/o kinds : minute sperm-cells, or sperma- tozoa, active and mobile ; and comparatively large egg- cells or ova, passive in behaviour and containing more or less food material or yolk (Fig. 3). The sperms are found in the sets of animals that we call males and the eggs in females ; hence the two kinds of germ-cell are sometimes termed respectively male and female cells, but this mode of expres- sion may lead to misunderstanding. It is convenient to have in use a term which can be applied either to a sperm or to an egg, as many features of essential importance in reproduction and inheritance are common to both ; such a term is provided in the word gamete. Many animals — earthworms and snails, for example — have both kinds of gamete developed in the same individual, which is therefore neither exclusively male nor female but hermaphrodite. Among insects, however, hermaphrodites are extremely uncommon. Reproduction by means of gametes is known as sexual, and for normal sexual reproduction there must be union between the nuclei of gametes of either kind, that is between ii6 THE BIOLOGY OF INSECTS an egg-nucleus and a sperm-nucleus. This process is often called the fertilisation of the egg, and fertilisation is usually a necessary preliminary to reproduction. As the egg is relatively large and passive and the sperm is very small and active, the latter moves towards the former ; this it can do because it possesses, besides the ovoid or rod-Hke head in which the nucleus lies, a long vibratile tail or flagellum by means of which it can swim through fluid. When pairing (copulation) takes place between a male and a female insect a large number of sperms are passed by the former into the reproductive system of the latter. They are stored in a special receptacle (spermatheca) so that they may subse- quently fertiHse the eggs before these are laid. Among insects, therefore, fertilisation may not occur until some time after pairing. Fertilisation — the union of two gamete- nuclei to form a zygote-nucleus — is the essential starting- point for normal sexual reproduction. From this brief account it will be realised that the germ-nuclei must play a very important part in the pro- cesses of reproduction and inheritance, and these nuclei are bodies of very small size. A male insect's sperm of average dimensions is about 3^0 rnn^- (= ihi inch) in length, and its head, enclosing the nucleus, may be no more than one- tenth the length of the tail or flagellum. As the head, with the adjacent '' centre-piece," is all that usually enters the egg Sit fertihsation, it is clear that this body must contain all the material necessary for the transmission to an individual of the next generation of the characters, either of structure or habit, that are inherited through the male parent. The formation of the mature flagellate sperms results from a process of development that may often be traced back to an early stage in the Hfe of a male insect. Some essential facts about this development (spermatogenesis) may now be profitably considered as an introduction to our study of inheritance. Sperm-formation is the result of a special type of cell- division, and it is well known that the development, growth, and maintenance of any living body are due to a series of REPRODUCTION AND HEREDITY 117 cell-divisions in which the cell-nuclei take an important part. The microscopic examination of suitably stained sections reveals in the nucleus of any cell the presence of a substance (chromatin) which takes up the microscopist's stains very readily, and therefore becomes conspicuous. The chromatin often appears in the form of knotted or looped threads, but when a cell is going to divide it becomes segregated in a definite number of minute ovoid or rod- like bodies (chromosomes) each of which splits into two halves — " daughter- chromosomes " — one of these going to form part of the nucleus of either of the '* daughter- nuclei " resulting from the cell-division (Fig. 32, a). As the growth and repair of all the tissues of the body depend upon an enormous series of such cell- divisions it follows that the number of chromosomes must remain constant throughout the body-cells of any creature, and in a large number of insects (as in other animals) a definite number is character- istic for each kind of creature. A full account of the forms and behaviour of these bodies may be found in the work of L. Doncaster (1920) or E. B. Wilson (1925) ; only the most essential points can be discussed here, but the student of Insect Biology may profitably remember that a vast amount of information about cell-structure and behaviour which has an important bearing on the Hfe and development of animals generally, has been obtained through the study of insect tissues and germ-cells. Now, we have seen that in the fertilisation of an egg, the essential process is the conjugation of the two germ- nuclei, egg- nucleus and sperm-nucleus. The number of chromosomes in the zygote-nucleus of the fertilised egg must be clearly the sum of the numbers in the two gamete- nuclei. But if these numbers were the same as in the body- cells of the insect, they would be necessarily doubled after every reproductive pairing, they w^ould be repeatedly doubled throughout successive generations and the creature's nuclear constitution would become impossibly complex. This condition is prevented and the number of chromo- somes is kept constant through successive generations by ii8 THE BIOLOGY OF INSECTS " reduction divisions " during that maturation of the germ- cells in both sexes which is a prelude to fertilisation. Sperm-cells are produced by the divisions of cells known as spermatocytes which are the offspring of the primitive germ-cells of the male insect. The primary spermatocytes, like the primitive germ-cells, have the number of chromosomes normal to the male of their kind of insect. But when a primary spermatocyte is preparing to divide into two secondary spermatocytes its chromosomes become associated in couples (synapsis) and the members of a couple, instead of splitting, separate from each other, one passing into either nucleus of the two secondary spermatocytes (Fig, 32, b). Thus each daughter- nucleus receives only one chromosome of a pair and the chromosome number is reduced to half that normal for the species. Each secondary spermatocyte divides to form two sperma- tids, the chromosomes splitting so that their reduced number is maintained, this splitting being indeed in some cases apparent before the completion of the first sperma- tocyte division. Four spermatids are therefore formed as the offspring of each primary spermatocyte, and up to this stage in sperm-formation all the cells are minute globular bodies. Then each spermatid becomes transformed into an active sperm- cell (spermatozoon) with its compact nuclear head and long vibratile tail or flagellum (Fig. 3, A) ready to play its part in fertilising an egg-cell. The tgg has also a maturation process which it must undergo in preparation for fertilisation. Immature eggs (or primary oocytes) are the offspring of primitive germ- cells not differing in aspect from the sperm-forming germ- cells of the male. But the oocyte is much larger than the spermatocyte because within it a quantity of yolk is stored up to serve as food-supply for the growing embryo. As regards the nucleus with its chromatin the maturation- process of the tgg is essentially similar to that of the sperm. The nucleus of the primary oocyte undergoes a reducing division (meiosis), by which the nuclear material and the number of chromosomes are reduced to half what they REPRODUCTION AND HEREDITY 119 were in the original nucleus. But as each primary oocyte has only enough cell-substance to make one egg, the two Fig 32 —Diagrams of (a) normal cell-division (mitosis) with eight chromosomes shown after splitting to form daughter-nuclei ; (6) matura- tion of sperms (spermatogenesis), the primary spermatocyte divides into two secondary spermatocytes (sc) with " reduced nuclei, and each of these divides into two spermatids (s) ; (c) maturation ot egg (oogenesis),/), ist polar body and pr 2nd polar body mature egg ready for fertilisation, d, Maturation in a case where the full (diploid) number of chromosomes is uneven, the sex chromosome {x) having no partner. 120 THE BIOLOGY OF INSECTS secondary oocytes are extremely unequal in size, one of them, the maturing egg, keeping nearly all the cell- protoplasm and yolk, while the other is a minute *' polar body " (Fig. 32, c). This division is followed by a second in which (as in the formation of spermatids) each chromosome is split. The egg, now mature, again keeps nearly all the cell-substance, though its nucleus corresponds to that of a spermatid or spermatozoon, while its sister- nucleus, sur- rounded by a small mass of protoplasm, forms a " second polar body." If, as sometimes happens, the first polar body divides into two daughter- cells it is clear that there are four reduced nuclei (corresponding to the four mature sperm-nuclei), but only one of them becomes the nucleus of a mature egg ; the other three have cell-bodies so minute that they can perform no reproductive function, and they were formerly regarded merely as " extrusions " from the ripening egg. The nucleus of the mature egg has how- ever been reduced, so that when the sperm-nucleus enters at fertilisation, the conjoint or zygote-nucleus becomes quantitatively double that of either gamete and the number of chromosomes is restored to that normal for the body- cells of the creature to be developed from the zygote (Fig. 32). In very many insect eggs the polar-nuclei remain within the egg-substance, or the polar bodies after extrusion are reabsorbed by it. These processes connected with maturation and fertilisa- tion, the more essential features of which have been briefly sketched, have become known through investigations carried out during the last half-century. They were first elucidated by E. van Beneden (1883), T. Boveri (1899), and many students of various animals other than insects ; full details of these and subsequent discoveries may be found in the recent text-book of E. B. Wilson (1925). It is evident from the appearance of a definite number of chromosomes in succes- sive generations of creatures of the same kind that there is real individuality and continuity in these bodies and that they are closely connected with the transmission of inherited characters. Half the chromosomes present in the zygote- REPRODUCTION AND HEREDITY 121 nucleus are derived through either parent, and from these are derived by repeated fission all the chromatin in the cells of the body whose development starts from the fertilised egg. Clearly therefore the chromosomes are to be regarded as furnishing the " mechanism of heredity " ; it is somehow through them that " like begets like." But at the beginning of this chapter we reminded our- selves that offspring may not resemble their parents exactly and that members of the same family may differ from each other. The process of reducing division in the maturation of the germ-cells enables us, partially at least, to under- stand how such differences, collectively known as variation, are frequently an accompaniment of heredity, because the behaviour of those minute chromosomes within the germ- cell nuclei corresponds with facts that can be observed when the characters of members of the families of successive generations are compared. It is well known that this last- named line of inquiry, pursued with regard to hybridised varieties of plants by J. G. Mendel as long ago as 1865 (see W. Bateson, 1909), has been eagerly followed since the beginning of the present century by many students of various groups of animals, among which certain insects have yielded most important results. A simple introductory example is afforded by the " Orange " Moth {Anger ona prunaria) a common British insect of the Geometrid family, the male of which has orange and the female yellow wings, with delicate darkish streaks scattered over the surface, a dark line in the middle of the disc of each wing, and a series of dark dots along the hinder edge or termen. There is a form of this moth, known as the variety sordiata^ in which the scattered dark streaks are reduced or absent but there is on the forewing a great extension of the dark scaling so as to form two conspicuous bands, one along the termen and the other across the wing base (Plate IV, A). It has been shown by L. Doncaster and G. H. Raynor (1906) that if a pair of these moths, one of the pale type and the other of the variety, be bred together, the offspring will show the 122 THE BIOLOGY OF INSECTS dark banding of sordiata together with the streaky scaling of prunaria ; these offspring are hybrids, and their hybrid character is recognisable in their appearance although on the whole they resemble sordiata rather than prunaria. If now such hybrids breed among themselves in numbers, a count of the families of the next generation will show approximately a quarter of the population typical, speckled, unhanded prtmaria, a quarter unspeckled, markedly banded sordiata, while half will be speckled and banded like their parents of the first hybrid family. There is no doubt that these colour patterns depend on inherited characters, and we have seen that there is much reason for believing that the chromosomes in the germ-cell nuclei are concerned with the inheritance of such characters ; as it is now usually expressed, the germ-cells carry the '* factors " or " deter- minants " for them. Recalling the behaviour of the germ- cells in maturation, we remember that at the reducing division the chromosomes are first paired together (one of each pair being derived from either parent), and then separated so that they pass into different mature nuclei. Now it is conceivable that the two chromosome partners in the pairing (synapsis) may be alike, or they may differ in certain factors which they contain ; just so will the mature germ-cells resulting from the reducing division be alike or different as regards those factors. The results of such breeding experiments as those with the Orange Moth agree exactly with the indications afforded by the maturation processes. Call the germinal factor that induces the banding of the wings (in sordiata) S and the factor for the unhanded wings (typical prunaria) s. In a strain of pure bred sordiata all the germ-cells carry the factor S, and in pure bred prunaria all carry s ; in each case all the germ- cells are alike as regards one of these alternative characters and the individuals of such strains are '' gametically pure " or homozygous. When, however, an tgg of sordiata carrying the factor S, is fertilised by a sperm of prunaria carrying s, the zygote nucleus must contain both factors, and it is from such a PLATE IV m a .^^ A. Males of {a) Angerona prunaria, {b) its variety sordiata, and [c) prunaria-sordiata Hybrid. [After Doncaste?- &" Raynor, P.Z.S. 1904. -^•gf'-tXj^ m B. Male and P^emale of Amphidasys hetnlaria (left) and of its variety dotibledayaria (riglit). Two-thirds size. To face /. 1 22.] [^- i^' i^^'^"^ P'^'^^'"' REPRODUCTION AND HEREDITY 123 zygote that the hybrids produced in the cross-breeding just described are developed. Being hybrids they are hetero- zygotes, their composition as regards the characters under discussion being represented by Ss, pure prunaria being ss, and pure sordiata SS. In these pure strains all the mature germ-cells produced in any individual carry the factor for one character or the other. But in the reducing divisions which result in the maturation of the hybrid's germ- cells, the pairing of the chromosomes brings S always alongside s ; then they part and go into different gametic nuclei (see Fig. 32, b, c). So the hybrid produces two kinds of germ- cells each of which must carry either S or s but cannot carr}* both. Therefore the pairing of such hybrids may be represented graphically by S 5 and if each individual produces the two kinds of gametes in equal numbers any gamete will have two chances of meeting in fertilisation a gamete of the opposite kind to a single chance of meeting one of its own. Hence, in families of the second generation resulting from the pairings of hybrids among themselves, half the population may be expected to be hybrid, a quarter to be of one pure strain, and a quarter of the other. And this is as a rule closely approximate to the actual result of breeding experiments. The same principles of inheritance are illustrated, though with an interesting difference in detail, in another common British geometrid, the " Pepper and Salt " Moth {Amphi- dasys betularia). This insect (Plate IV, B) derives its common name from the pale grey colour of its wings, traversed by interrupted black bands which vary considerably in extent and degree in different individuals. The species has, how- ever, a well-marked variety (doubledayarid) in which the black scaling is so heavy that the wings present a con- tinuously sooty aspect. It is found that w^hen typical betularia is mated with doubledayaria all the hybrid 124 THE BIOLOGY OF INSECTS offspring are of the latter variety ; they are indistinguishable from their dark-winged parent and display no evidence of their hybrid nature. The dark wing-colour is '' dominant " to the pale ; but when hybrids breed among themselves, the resulting families, if sufficiently numerous in individuals, have three-quarters of their members sooty and a quarter pale. The pale character, hidden in the first generation and reappearing in the second, is known as '* recessive." The dark-winged moths of the second generation look all alike, as all resemble their dominant parent, but if the nature of their germ-cells be tested by a study of their offspring, it will be found that a third of them — that is, a quarter of the whole second generation — are pure (homo- zygous) as to the dominant character, and their inbred descendants will continue to be pure douhledayaria^ while the other tvvo-thirds — that is, half of the whole second generation — are hybrids (heterozygous) and will therefore give, if bred among themselves, the same proportion of a quarter pure dominants, half hybrid dominants, and a quarter pure recessives as before. The characters of the alternative forms, douhledayaria and hetularia may be indicated respectively by D and d. Then the zygote com- position of a pure strain of the former will be DD, and all the gametes must carry the D factor, while in a pure betularia strain the zygote composition will be dd and the gametes will all carry d. The hybrids of the first generation must all have the zygotic composition Dd, and half of their gametes will carry either D or d, just as in the hybrid sordiata-prunaria the zygotic composition is Ss and any individual gamete must carry either S or s. In breeding experiments of this kind it is often the practice to make what is called a " back- cross," that is, to mate a hybrid with one of its parent forms. If hybrid douhledayaria be paired with betularia it is found that in a large enough population half the offspring resemble either parent, as a consideration of their germinal constitution would lead one to expect. For since the hybrid has the composition Dd, and the betularia-psirQnt is dd, it is clear REPRODUCTION AND HEREDITY 125 that an approximately equal number of Dd and dd conjuga- tions may be expected. It was this phenomenon of the dominance of one of a pair of alternative characters over the other (recessive) in inheritance that led Mendel in his classical experiments sixty- two years ago with garden peas, so to analyse the facts of inheritance as to conclude that individual mature germ- cells might carry the factor for one or the other alternative character, but not the factors for both. The processes of maturation, not observed by microscopists until long afterwards, supply the mechanism by which these results are brought about. One most important conclusion that may be drawn from the family histories of these moths is, that while the nature of the offspring may not always agree with the nature of their parents (as in the family of a pair of hybrid douhledayaria)^ it does depend on the nature of the mature germ-ceils (gametes) which those parents carry. This is one of many indications of the importance in inheritance of the material which A. Weismann (1893, 1904) called '* germ-plasm," a material which we have every reason to believe resides in the nuclear chromatin. It seems at least clear that the factors carried by the chromo- somes of the germ-cells determine certain characters of the body. Many other examples, furnished by insects, of the inheritance of alternative characters and of variation might be given, and further reference to these important subjects will be made in later chapters. There is, however, one question which may be considered most suitably here, as it concerns very closely the whole subject of reproduction ; that question is the nature and determination of sex. Great uncertainty still surrounds many aspects of the problem, but some remarkable advances towards partial solution have been made during recent years, and much of our knowledge of sex- determination has been reached through investigations on insects. In many insects as in other animals of which statistical studies on the question have been made, the numbers of males and females in a large population are approximately 126 THE BIOLOGY OF INSECTS equal. It is also evident that in any kind of animal whose members are of one sex or the other, maleness and female- ness may be regarded as alternative characters which might be compared with such characters as dark or pale wings. And it has just been indicated how, in a number of unions between hybrid dominants and pure recessives, we may expect approximately half the offspring to resemble either kind of parent. Further, sex is, in most cases, a truly inborn character. It seems likely, therefore, that the determination of sex follows from what are called '' Mende- lian " factors ; if this be so one sex-factor must be dominant over the other, and members of the one must be hybrid (heterozygous), those of the other pure (homozygous) for the sex-factors. For example, if all males have the zygotic composition M/", there will be two kinds of sperms with factors for one sex or the other, while all females will have the composition ff and all the eggs will be alike carriers of the feminine factor. Such an tgg if fertilised by a male- determining sperm will develop into a male, but if by a sperm of the other kind, into a female. Clear evidence that sex is indeed thus determined in various insects has been afforded by studies of the germ- cells and by breeding experiments. From what has been already stated about reducing- divisions (pp. 117-118) it will have been realised that the number of chromosomes in a zygote-nucleus and in the nuclei of body-cells is commonly an even number — half thereof derived from the gamete contributed by either parent. Many years ago, however, H. Henking (1891) observed that in a bug (Pyrrhocoris) two kinds of sperms are produced, distinguished by one kind having a chromosome fewer than the other. Subsequent work by E. B. Wilson and his colleagues (see his text-book, 1925) showed that such a condition occurs in many bugs (Hemiptera) and other insects, whose body-cells have one chromosome less in males than in females, a difference of three or four in the chromosome number of the two sexes {e.g. 26 male to 30 female) being sometimes noted. Usually, however, in the body-cells of such differentiated insects the REPRODUCTION AND HEREDITY 127 female is characterised by an even number of chromosomes («) and the male by an odd number, one less (w — i). All the mature eggs have - chromosomes, while the ripe sperms have either - or - — i , the latter being clearly male, and the former female- determining. At the pairing of chromo- somes for the maturation-process in males of such insects tCMM* ••.. fill ••» • MMM V Fig. 33. — Diagram of Chromosomes of Bug {Lygaem). a, Diploid group of male insect in pairs, one " sex-chromosome " (.r) being much larger than the other {y). d. Diploid group of female, no such differ- ence apparent, c, Maturation division in sperm formation .v and v separating to the two daughter-groups shown in polar \-iew at t/, " male- producing " sperm with y, "female-producing" ^^•ith x. After E. B. Wilson {Journ. Exper. Zool. ii, 1905). there is a sex-chromosome which has no partner ; there- fore in the reducing division half the sperms will receive a chromosome more than the other half (Fig. 32, ^). In such cases where the male- determining sperm is distinguished by the absence of a sex-chromosome, it seems inappropriate to state that the male-factor is *' dominant," though the result is comparable to dominance in ordinar}' Mendelian inheritance. It is also of great interest to know that in other insects 128 THE BIOLOGY OF INSECTS the tvvo kinds of sperms are distinguished not by the total absence of a certain chromosome, but by a marked difference in size or shape between the two of a pair. Thus, in a bug {Lygaens) investigated by Wilson, the full (zygote) number is fourteen ; in a female the two " sex-chromosomes " are equal in size, while in a male (Fig. 33) one of them {x the " female-producing ") is much larger than the other (y). The two members of this unequal pair go into different daughter-cells at any reducing division, and fertilisation by an .Y-bearing sperm will result in a female-producing zygote, or by a jy-bearing sperm in a male-producing one. These facts establish the conclusion that sex- determina- tion is closely dependent on certain factors situated in certain definite chromosomes of the germ-cells. That the same is true for other inherited characters is shown by what is called " sex- linked " inheritance — a term applied to cases in which some readily observed feature in body- structure or appearance is inherited by means of a factor carried in the same chromosome that bears the factor for sex (either male or female). The insects in which this condition has been most thoroughly studied are small Fruit-flies {Drosophila) which breed very quickly and prolifically under observation in the laboratory, and are thus particularly suitable for studies in heredity, as a number of generations can be observed in a comparatively short time. Investigations on these flies have been carried on through several years by T. H. Morgan and his colleagues (191 6, etc.) ; only a brief summary of some of the results can be given here. The normal colour of the eyes of some of these flies is red, but a male variety appeared in which the eyes were white. Breeding experiments show that the normal red eye-colour is dominant to white, and that males are hybrid and females pure as regards the sex-factors. The crucial test of sex-linkage is furnished by pairing w^hite-eyed females with hybrid red-eyed males ; the off- spring of such a cross are half white-eyed males and half red-eyed females. If we indicate graphically the con- stitution of the parents. REPRODUCTION AND HEREDITY 129 Male Rw Female ww ?(^ ?? the two kinds of offspring will be represented by Males WW Females Rw and consideration of possibilities convinces us that the sex and eye-colour of the progeny depends on the linkage of the factor for red eyes with that for femininity in half the ripe sperms of the hybrid males. This can only mean that those two factors reside in the same chromosome, which at a reducing division passes into one or other of a pair of sperm-nuclei. Study of chromosome behaviour in the cells of Droso- phila is comparatively simple because there are only four pairs of those bodies — one pair minute and round, two pairs bow-shaped and clubbed at their extremities, and a pair of which in the female both members are straight and rod- like, while in the male one is straight and the other sharply bent towards one end. The members of this last pair are the sex- chromosomes, the bent member of the male pair being visibly the y chromosome ; this carries the factors which determine the male sex and also that which brings about absence of pigment in the eyes. The factor for redness of eyes, if borne in a male Drosophila, is linked with that for femaleness in the x chromosome. More than a hundred such sex-linked characters have been studied by Morgan and his colleagues in their breeding experiments with Drosophila, and the way in which these are grouped in inheritance correspond with the number of chromosomes, four pairs, present in the zygote nuclei. Mention must be made of a few exceptional results shown by these experiments which, when analysed, are found to throw further light on the part played by the chromosomes in heredity. In the families resulting from the union of hybrid red-eyed male Drosophila with pure white-eyed females, the offspring are not always all white- 130 THE BIOLOGY OF INSECTS eyed males and red-eyed females ; sometimes this " criss- cross " inheritance fails to work in some 5 per cent, of the population which appear just like their parents — red-eyed males and white-eyed females. C. B. Bridges (1916), by careful breeding observations confirmed by microscopical study of the cell-nuclei, has demonstrated that these exceptions arise from *' non- disjunction " of the sex- chromosomes in the maturation divisions of the egg-nuclei ; the members of a pair of white-bearing, female- determining (x) chromosomes may not part company with each other but may both pass into the same daughter- nucleus, so that a ripe egg, instead of carrying the normal one x chromosome may carry two or none. Bridges has proved that if the former kind of egg be fertilised by a sperm with the male factor the offspring will be the exceptional white-eyed female, while if the egg without any sex- chromosome be fertilised by a sperm v^th the female and red-eye factors the offspring will be the exceptional red-eyed male. These conclusions are startling when compared with the con- ditions that apparently determine sex in the normal results of cross-breeding already considered (pp. 128-9). In these exceptional families the white-eyed flies are females although there is a male-factor (which is commonly regarded as " dominant ") in the zygote nucleus. But then there are two female factors, so that we might conclude that the quantity of the respective determinant in the nucleus helps to decide the resulting sex ; the single male-factor may be dominant over one female factor but has to give way to a " double dose " of femininity ! The constitution of the exceptional red-eyed male, however, is still more surprising, for this insect has no chromosome with the male factor ; his only sex-chromosome is female-producing. In view of these facts several students of these subjects, including R. Goldschmidt in his recent (1923) critical discussion of the question of sex- determination, believe that the normal heterozygotic male creature (fM or xy) is a male, not because he possesses the *' male " factor (y), but because he has only one chromosome v^dth the " female '* factor (x). REPRODUCTION AND HEREDITY 131 When two x chromosomes are present the creature must be female. Sex-chromosomes of the same essential type as those of Drosophila have now been demonstrated in other insects of the same order (Diptera) as well as in grasshoppers (Orthoptera) and beetles (Coleoptera). In several types of moths (Lepidoptera), the males are pure and the females hybrid for sex- characters : this was worked out by L. Doncaster (1908, 1914) in his famous studies on the breeding of the Magpie Moth {Abraxas grossulariata) with its pale variety lacticolor, known only from female specimens, and affording an example of sex-linked inheritance con- trasting with that of Drosophila. In bees, wasps, and most Hymenoptera that have been studied there is a remarkable difference between the two sexes in the chromosome- number, females having twice as many as males ; the meaning and results of this condition will be discussed later in this chapter. The facts set forth in the preceding pages might be thought to suggest strongly that an insect's sex is a definite and irrevocable character determined by the constitution of the germ-cells of the creature's parents. Yet, there are other facts well known to students of insects which warn us that by resting in such a conclusion we miss part of the truth of the matter. Insects are hardly evenjl truly herma- phrodite, but abnormal individual specimens in which the characters of the two sexes are more or less combined are well known ; such creatures are called gynandromorphs. In certain moths in which the male differs from the female in wing colour or pattern and has more elaborately developed feelers than she, such gynandromorphs can be very easily recognised. In the simplest case a moth may, for example, be male on the left and female on the right side, Hke the *' Emperor " Moth (Saturma pavonia) depicted on Plate V, the difference affecting not only the visible (" secondary ") characters of wings and feelers, but frequently also the essential organs of reproduction, so that there is on the 132 THE BIOLOGY OF INSECTS left of the body a testis with sperms and on the right an ovary with eggs. In such a case it is likely that in the first division of the egg nucleus one of the daughter x- chromosomes failed to pass into the cell whence the left side of the body developed, therefore the left side is *' male " and the right '' female." An alternative explanation is due to L. Doncaster (19 14), who noticed that some eggs of moths are binucleate and that both nuclei may be fertilised ; if one were male and the other female in composition, each might give rise to the half of a gyn- andromorph. K. Toyama (1906) described gynandromorph moths of the common silkworm (Bomhyx mori), hybrids from a cross of two races, and the caterpillars whence they developed showed a bilateral distinction in the colours (white and dark- spotted) of the larvae of the parents. Instead of the axial division of an insect into right and left halves showing respectively the characters of either sex, gynandromorphs sometimes occur in which the two sets of characters form an apparently irregular " mosaic " pattern. This con- dition might be explained also by the irregular division, though at a later stage, of the sex-chromosome in those cells where the abnormal tissues and organs arose in development. In the course of the breeding experiments with Drosophila many cases of gynandromorphism were observed, and described by Morgan and Bridges (1919). These abnormal insects were often the offspring of cross- breeding and consequently showed divergent body- characters (length of wing, eye-colour, etc.) as well as some of the external (" secondary ") features characterising the two sexes. Many of these Drosophila gynandromorphs were bilaterally male and female, while others displayed a more or less irregular " mosaic," the eye, for example, on one side being divided between the alternative colours that are sex-linked in inheritance. Analysis of the results together with knowledge of the nuclear constitution of the various forms, enabled the investigators to demonstrate that all the characters displayed in such abnormal insects are sex- PLATE V V \ ^^^^^Bl ' I niai^f ^' ^^BhHB^^SlOik ^\ A. Emperor Moth {Saiuniia pavonia) Gynandromorph. [/. Afjnifage, photo. B. Poplar Hawk Moth {Smeri)ithus popiiU), Female and Eggs. Half size. To face p. 132.] [H. Britten, phoio.^ REPRODUCTION AND HEREDITY 133 linked, and are due to the absence of one .r-chromosome in the cells producing those parts of the body which show male features. It follows from this that the gynandro- morph must be considered as originally a female with the two ;c-chromosomes ; if in the repeated cell-divisions leading to development one of these be accidentally " dropped out," male features will appear. It is note- worthy, however, that in a bilateral g\^nandromorphic Drosophila only the outward male characters are present. Internally there are right and left ovaries. Where the distribution of the two sex-characters in a gynandromorph Drosopliila is mosaic and irregular, the female areas always predominate ; this confirms the conclusion that the cell division or divisions in which one of the .y- chromosomes " dropped out " came later in development than the primary division of the egg, and it also confirms the originally female nature of the gA'nandromorph. Various abnormal hybrid moths afford examples of a mixture of male and female characters which differ in more important respects from the cases hitherto mentioned, as the most remarkable forms show a tendency to trans- formation from one sex to the other ; it is therefore con- venient to distinguish them from g\'nandromorphs, and they have become generally known in recent years as " intersexes." The best known work on such insects is that of R. Goldschmidt (1916-17, 1923), who has made extensive breeding experiments with various races of the Gipsy Moth {Porthetria dispar). As is implied in its specific name, this insect shows very marked sexual differ- entiation, the female having whitish wings with black cross- markings and feebly pectinate feelers, while the male's feelers are strongly pectinate and his wings show a dark brown ground colour on which the black banding is relatively inconspicuous. The species has a ver^^ wide range from Western Europe to Japan (the indigenous British race is extinct), and shows well-marked geographical forms. In Goldschmidt's experiments, various Japanese varieties were cross-bred with the European forms of the 134 THE BIOLOGY OF INSECTS modi or whk one aoodier. In the earlier e]q)criments Eui Dpc m males ^nere crossed with Japanese females and the lAfiiig ^pcre all normal as regards sex, but when males wa^ bred with European females the off- grew into rkormal males and into females with a blend of male characters. Later, when a number of DCS bad been tested, it was found that " the of intersexnafity is definite and ty|Mcal fcH- a particular ' so that the intersexes could be sorted into low, and btgh grades. Extreme members of the last- group were actually males in appearance and bAafkmr , and oncrosoopic cxaminatiofi o£ the refHoductiTe Ofgans sl wfiMed d^eoerate egr^ tne orarian tissue becoming dKp la rrd by spcmuz:. it sperms being produced. It is powahlc dbc - tersexes tending to derf^r feminine : _ .ess completehr than :!" ri or tr^ f rust described. nsects like the G - _, ___c ^icri^^c. L-.C sex cannot be t.j it: by the x or _y chromosomes. In r-AVc 5.een, the male is homozygous (xx) for - - : rs- Goldschmidt seeks an explanation for r rTr rs bv asM i iiBi i g the existeiice erf a factor for femin- 7) in addrdoci to the normal male sex-factor that : in die X dutHDosomes. The hypothetical female *• purely maternal in inheritance," may be supposed "^ egg-cytc^lasm or in the v chromosome, : are believed to be produced when the rength c: tie —.z'e factors in the fertilised -.7'* - hig-h : .": zT the tendency to -^~ ^ : 2 - 7 ' ' ' "emale is actually tr_ 7 7 . results appear in : v»er of one parent- A: that these results show the 7 ordinary tyi>e of sex- -iparoit doubt that such i}-s brought about by means 3C Irr- _ %. i^. l?»5- *• r: -^ "inje zssrrpg ir lEe- ■•rr>r Scr-^:_ -^ ~-^ :a£^ g-T?»rnrgri?g ir x TlBa^tl^gT HT r. xt "■ mmucnr -rC!s see iT^rarig: XZL psp^ arTTkr SfflUllS" i^ 3/- rx i 136 THE BIOLOGY OF INSECTS very nearly complete half-and-half example." According to Poulton's interpretation these effects can be produced because the formative pupal tissues retain to some extent the embryonic potentialities : *' the pupal factors which determine the secondary characters of sex are in a condition analogous to that of a Mendelian heterozygote . . . and the underlying characters are revealed by a correctly timed mechanical shock." The shock is apparently able to affect even in the developing pupa these factors, which are continuous, we may believe, with the factors of the germ- cells. That the chromosomes supply the mechanism through which inherited characters are passed on through successive generations seems to be well established by the facts that we have considered ; the linkage of various factors in the same chromosome affords strong support for this view. It is interesting to notice that further evidence has been deduced from exceptions to the ordinary linkage phenomena that have been observed by students of heredity in many groups of animals, and notably by Morgan and his colleagues in Drosophila (1922). They beheve that they have suc- ceeded in locating the factors for a large number of varying characters in one or other of the four chromosomes in the gametes of this fly, and finding that the inheritance of these characters is not always according to expectation — for , AbcD. ^ ABCD example -57^7 mstead of , , — they conclude that, when the chromosomes are paired for the reducing division and more or less closely twisted or looped in certain regions of their length, there may be a fracture and exchange of parts between the two chromosomes of a pair ; the frequency of this occurrence, known as ** crossing over," can be observed, and from subsequent calculations the investigators believe that they can '' map " the chromosomes so as to determine approximately how the factors of the various characters are arranged along their length. From the preceding statements and discussion one REPRODUCTION AND HEREDITY 137 important limitation of the evidence must be admitted. All the characters whose inheritance by means of the chromosomes has been clearly demonstrated, are detailed characters — varietal or specific. It is extremely likely that the factors for the more fundamental characters — those, for example, which distinguish a fly from a moth or a bee — are also situated in the chromosomes, but some students of the problems of heredity believe it possible that these may reside rather in the cytoplasm of the germ-cells, especially perhaps in the egg-substance. In connection with this possibility, it is of interest to note that definite bodies granular or rod- like (chromidia) and also reticulate (" Golgi- bodies "), have been now observed in the body-cells and germ-cells of many insects as of other animals. In some cases these bodies appear to undergo a definite and regular process of division when the cells are dividing, so that their individuality and continuity may be inferred. Possibly, as some students of the subject have supposed, they are nuclear in origin, and may be the agents by means of which the nucleus influences the substance of the cell. For information on these bodies reference may be made to the writings of Wilson (1925) and J. B. Gatenby (1917-19). So far we have considered the working of inheritance and reproduction among insects along the usual lines of the sexual process common to the great groups of animals generally, the new individual developing from the fertilised egg. It is well known, however, that in many animals cases of development from an unfertilised egg occur, some- times exceptionally, and sometimes as part of the regular life-cycle of the creature. Study of the reproduction of insects shows some of the most remarkable and interesting examples of this virgin- generation (parthenogenesis) that the animal kingdom affords. Not a few female moths that had never paired with a male have been known on occasion to lay eggs from which caterpillars were hatched to be in due course transformed into moths ; the Common Silkworm {Bomhyx mori) and 138 THE BIOLOGY OF INSECTS the Gipsy Moth {Portheiria dispar), aheady referred to in this chapter, are examples of species as to which perfectly reliable observations have been made. Such exceptional instances of virgin reproduction are of much interest because there can be no doubt that, in the history of animal groups, regularly occurring parthenogenesis is a condition secondarily derived from reproduction through normal sexual union, and that as a starting point for what is now regular parthenogenesis, we must look to an originally exceptional appearance of this mode of development. There are insect species of various orders in which parthenogenesis is the usual method of reproduction, males being exceedingly rare in some cases and altogether unknown in others. Of greater interest, however, are those specialised insects in which virgin reproduction alternates definitely in the Hfe-cycle with the usual sexual method. The Aphids (plant- lice or " green-fly ") afford the best known example of this cyclical or seasonal parthenogenesis. Among most aphids there are males and females which pair in autumn, and the females lay fertilised hard-shelled eggs which carry the race over the winter. From these eggs females only are hatched, " stem-mothers " as they are called, whose eggs without fertilisation develop within the oviducts of the female so that active young are born. Successive generations of such ** viviparous " virgin females follow each other through the spring and summer, those of the latest brood giving birth to males as well as to the sexual females of the autumn. Here there are many generations in the course of the yearly life-cycle. The Cynipidae or Gall-flies, a well-known family of the Hymenoptera, have not more than two alternating generations in the year, usually a summer sexual brood, and a winter or spring brood consisting of virgin females only. Among many at least of the social Hymenoptera (ants, wasps, and bees) the mother or " queen " insect may lay either fertilised eggs from which females are normally developed, or unfertilised eggs which as a rule only produce males. In these insects, therefore, the occurrence of parthenogenesis seems REPRODUCTION AND HEREDITY 139 to be strangely connected with the determination of sex. It appears at first sight anomalous that in an ordinary community of hive-bees all the male members (" drones ") should be without any inherited characters derived through a male parent, but as the queen-bee develops from a fertilised egg, each drone has a maternal grandfather. In many of these cases the facts of virgin-reproduction have been proved to correspond with some abnormal mode of nuclear division among the germ-cells. Thus in Aphids and their allies the Phylloxerans, T. H. Morgan has shown (1909) that the eggs of the parthenogenetic females mature without reduction ; only one polar body is extruded and the number of chromosomes remains at the full " diploid " complement (zn). In the sexual broods of aphids which produce the winter eggs, while the females have the full double number (zn) the males have one or two fewer (zn— 1 or 2^2 — 2). This is brought about by a partial reduction during the maturation of the male- producing eggs of the virgin females of the last generation, one or two chromosomes passing undivided into the polar- nucleus, so that the ripe- egg nucleus has one or two fewer than the full double number. It has been mentioned that the fertilised winter eggs, which the sexual aphids produce, all develop into female insects, when hatched the next spring. In the spermatogenesis of the autumn males the usual reduction-division takes place, but only those sperma- tocytes whose nuclei contain the full single (" haploid ") number of chromosomes (n) develop so as to give rise to active spermatozoa ; all those without the ;c-chromosome (« — I or « — 2) are much below the normal size and cannot produce functional sperm- cells. Hence it follows that every fertilised egg has the full diploid number of chromo- somes and develops into a female insect. Among the Hymenoptera, as has already been mentioned, the number of chromosomes in the cell-nuclei of a male insect is half the number that characterises the female^s nuclei ; in bees, for example, a male's nucleus has sixteen HO THE BIOLOGY OF INSECTS and a female's thirty-two chromosomes. F. Meves (1901) and others have shown that in the female egg maturation pursues its usual course so that the ripe egg has only six- teen. In spermatogenesis, however, the first maturation division is abortive, a small cytoplasmic body being divided off from the spermatocyte which retains all sixteen chromo- somes in its nucleus. Then in the succeeding division these are, as usual, split so that each of the two resulting spermatids has sixteen chromosomes. Hence it follows that an unfertilised egg (with sixteen chromosomes) will develop into a male or " drone " bee, while a fertihsed egg (with thirty-two chromosomes) will develop into a female bee, either a *' queen "or a '' worker " ; which of these two latter alternative results will be produced depends upon the treatment and feeding of the larva, a striking illustration of the co-operation of the factors of heredity and environment — of " nature " and " nurture " — in bringing about the final result of the reproductive process. The origin of females from fertilised and of males from virgin eggs has long been recognised as the general rule among the bees, wasps, and their allies. Some careful breeding experiments by W. Newell (191 5), who crossed yellow Italian bees with members of a grey race, afford confirmation of the accepted view. When yellow queens were mated with grey drones all the offspring were yellow, the colour factor for this being dominant to that for grey ; but when grey females were mated with yellow drones the workers were yellow but the drones were grey ; these latter clearly had no inheritance through a male parent. But from the eggs of hybrid yellow females either grey or yellow drones might be produced. Statements have, however, often been made that drone bees may be developed from fertilised eggs by special feeding of the larvae ; if such statements really represent the facts, cases of intersexuality (PP- i33~5) niight be supposed. R. W. Jack (19 16) has brought forward evidence that cannot be lightly set aside for the occasional development of worker bees from the REPRODUCTION AND HEREDITY 141 unfertilised eggs that may be produced in the rudimentary ovaries of workers, and this might be explained by a sup- pression of the reduction division in maturation. R. Gold- schmidt (1923), in a discussion of such alleged exceptional modes of reproductive behaviour, calls attention to the possibility of non-disjunction (p. 130) in maturation as the cause. " If such an abnormality occurs among the Hymenoptera, it would be possible for ripe eggs to be produced containing two ^-chromosomes as well as ripe eggs with none. The first would give females partheno- genetically whilst the latter would, after fertilisation, give males." Thus far we have considered the germ-cells and the hereditary factors borne, as we believe, in their chromosomes; these must be regarded as the essential elements in repro- duction. They require, however, for their action, many accessory structures and processes which, in the pairing and breeding of insects, are often conspicuous, characteristic, and noteworthy. The eggs of a female insect are developed in her ovaries (Fig. 34) — a pair of organs each consisting of a number of tubes to which the contained eggs as they ripen give a '* beaded " aspect. Each egg needs to accumulate a store of food-material (yolk) for the nourishment of the embryo which, it may be hoped, will grow from it ; this food- stuff is in many insects obtained at the expense of other cells in the ovarian tube, '' nurse- cells " as they are called, which lie in groups between successive eggs, or form a single group at the fine terminal end of the tube, the eggs keeping contact with them by means of protoplasmic threads. Where there are no nurse-cells the " follicular cells," which form a sheet or follicle closely enveloping each egg, supply food material, and each foUicle secretes on its inner surface the shell of the egg. This is a relatively hard protective envelope of characteristic shape in various families of insects — globular (Plate V,? B), cylindrical, elongate, and rounded at either end (Plate VI, A), flat and disc-Hke — often adorned with sculptured markings which may indicate the 142 THE BIOLOGY OF INSECTS outlines of the cells that formed it, provided with an opening or micropyle through which a sperm can make its way into Fig. 34. — Female Reproductive Organs of Warble-fly {Hypoderma bovis), lateral view, the right ovary (ov. r) displaced ventralwards ; ov. /., left ovary; od, paired oviducts, uniting to form median oviduct (od') which passes into vagina (va) ; sp, spermathecae whence three ducts pass to vagina ; ag, accessory glands ; op, ovipositor ; in, intestine ; r, rectum with its glands (rg.). X 8. After Carpenter and Hewitt (Sci. Proc. R. Dublin Soc. xiv, 19 14). the living egg-substance, and often with a lid through which the young insect can emerge when the time of hatching arrives. REPRODUCTION AND HEREDITY 143 The number of ovarian tubes and the rate at which the eggs ripen varies enormously in different insects. Some female beetles have only two tubes to each ovary ; most moths have four ; other beetles like the chafers have six ; cockroaches eight ; while a queen-bee has about a hundred and fifty and a queen-termite over two thousand tubes on each side of the body. The ovarian tubes of either side open into an oviduct and the two oviducts lead, in most insects, into a median external passage, the vagina, which has a lining of cuticle being formed by an inpushing of the outer skin. The reproductive passages of insects always open towards the hinder end of the body, and the vaginal aperture is usually situated on or immediately in front of the eighth abdominal segment. Just behind it is the opening of the spermatheca or reservoir into which the sperm-cells pass when the female pairs with a male insect. This spermatheca may be a simple ovoid or sub-globular chamber provided with a short duct, or consist of two or three chambers (Fig. 34, sp) with relatively long ducts ; the whole apparatus is lined with cuticle. The typical insect ovipositor consists of three pairs of processes, one of which belongs to the eighth and two to the ninth abdominal segment ; these acting as a forceps hold the egg that is being laid (Fig. 35). The processes of the ovipositor may be relatively short, and usually unseen because retracted into a pouch formed by the inpushing of the hinder abdominal region, or they may project conspicuously at the tail-end of the insect as in ichneumon-flies and many grass- hoppers (Fig. 36). In most two- winged flies (Diptera), the processes of the ovipositor are very short, but owing to a great development of the intersegmental cuticle in the hinder part of the abdomen, that region can be extended in a ** telescopic " manner when eggs are being laid and retracted again when the organ is out of use (Fige 35). The well-known sting of wasps, bees, and their allies is a highly specialised ovipositor modified into a formidable weapon of defence or attack, but in some cases retaining still its original function as an egg-laying organ. ^pfC0^^^^ Fig. 35.— a, Dorsal and B, Ventral view of hinder abdominal segments and ovipositor of Hypoderma bovis, fully extended ; the segments numbered v-ix are connected by long intersegmental mem- branes. X 7. C, the same v^^ith segments retracted ("telescoped"), the ovipositor holding an egg. X 20. D, Lateral, E, dorsal, and F ventral views of terminal segments with ovipositor. The numbers and parts of the eighth, ninth, and tenth abdominal segments are indicated, T, terga ; St, sterna, P, processes of ovipositor, X 60. After Carpenter and Hewitt. REPRODUCTION AND HEREDITY 145 In every case we find the form and action of an insect's ovipositor suited to the position in which eggs have to be placed. The long ovipositor (Fig. 36, A) of a female phasgonurid grasshopper enable her to bury her eggs deep in the ground, and the long tapering telescopic abdomen of a female crane-fly or carrot-fly enables her to achieve the same result on a smaller scale. The serrated processes of a sawfly's or a cicad's oviposi- tor serve to cut in- cisions in plant tissues , while the dart-like egg-laying organ of » an ichneumon fly pierces the body- wall of a caterpillar, and prepares for the life of her larva as an internal parasite. Be- sides laying the eggs the female often fixes or protects them by a hardened fluid Fig secretion of glands opening into the vagina. To such pro- tective action further reference will be made a later chapter m 36. — A, Hinder Abdominal Segments (7-10) and Ovipositor of Longhorned Grasshopper (Conocephalus), lateral view. X 3. B, Diagram of Hinder Abdominal Segments (6-10) and developing Ovipositor of a typical female insect, c, cerci ; v, vulva ; ga, processes (gonapophyses) of eighth segment ; gb, inner and gc, outer processes of ninth segment. After R. E. Snodgrass (Anatomy of the Honey Bee, (p. 302). ■'°"- The sperm-cells are developed, as already mentioned, in the testes of the male, whose abdomen contains on either side in the dorsal region a testis which corresponds to the female's ovary, and is composed, like that organ, of a number of tubes which open into a duct called, in the male, the L 146 THE BIOLOGY OF INSECTS V.d. vas deferens. The paired vasa deferentia lead into a median chitin-lined ejaculatory duct, which corresponds to the female's vagina, and has associated with it a seminal vesicle, wherein the sperms often complete their development and await the act of pairing, as well as accessory glands ; these secrete a fluid by which, when partially evaporated, the sperms are united in the bundles ready for transference to the sperm reservoir of the female (Fig. 37). In many insects, the drone hive-bee, for example, these accessory glands are of relatively enor- mous size. In order to ensure trans- ference of the sperms to the female's spermatheca, male insects are furnished with a cuticular genital armature, corresponding to the female's ovipositor, and used for grasp- ing the female's abdomen in the act of pairing. The ejaculatory duct terminates in an intromittent organ (aedeagus) which may be simply tubular as in bristle- tails, or provided with a basal bulb and a set of paired outgrowths as in bees and flies. Pairs of processes on the After Carpenter eighth and ninth abdominal segments or on one of these, serve as claspers working laterally, while the terminal dorsal and ventral plates of the abdomen may be modified into vertically disposed claspers. It is instructive to Fig. 37. — Male Reproductive Or- gans of Warble-fly {Hypoderma bovis). Te, testis; V.d., vas de- ferens ; A.G., accessory glands; £).e., ejaculatory duct; 5. e., ejacu- latory sac ; 6, 9, 10, terga of the sixth, ninth, and tenth abdominal segments. X 8 and Hewitt. REPRODUCTION AND HEREDITY 147 compare the simple, primitive armature of a male bristle-tail such as a Machilid with the complex apparatus of a hive-bee, or the still more elaborate structures found in a moth or a muscoid fly (Fig. 38). These outer organs of repro- duction, the action of which ensures the fertilisation of the eggy are obviously of great importance to the life of the individual and of the race. It is found that the details of their structure and form are remarkably constant among insects of the same kind differing in definite features from those of nearly allied kinds (in Fig. 38 compare A, B with D, E); the m^ale's structures are thus adapted to fit or interlock with those of his mate so as to transmit the sperms into her reservoir, whence as previously explained, they are discharged as required for fertilisation of the eggs when these are laid. The foregoing descriptions and discussions suggest that the processes of inheritance and reproduction are closely connected with the vital fact of sex-differentiation, and reference has already been made in this chapter to some of the differences of appearance and behaviour, apart from those directly connected with the reproductive system, that are often conspicuous in male and female insects respec- tively. Well-known examples of such " secondary sexual characters " are furnished by the brighter, richer, or more vivid colours of many male butterflies, dragon-flies, and other insects as compared with their mates, by the presence of wings in a number of male cockroaches, grasshoppers, moths, and Hymenoptera whose females are wingless, and by the greater relative size and elaboration of sense-organs in the male as compared with the female, as shown in the larger compound eyes of male bees and many flies, the elaborate feathered feelers of many male moths and gnats, the immensely enlarged plate-like feelers, and the strange head and body-outgrowths of many male chafers. Much support is afforded by such facts to the well-known con- tention of P. Geddes and J. A. Thomson (1889) that male animals express in their organisation the active nature of St 8 Tio Fig. 38. — Terminal i Abdominal Segments and Male Genital Armature of Warble-flies, A, B, C, Hypoderma hovis; D, E, H. lineatum. A and D are lateral, B and E ventral, and C postero-ventral views. X 35. Terga (T), Sterna [St.), paired Processes (P), and Gonapophyses (G., G.i., internal, and G.e., external) of the various segments are numbered (6,7, 8, or 9). D.e.y ejaculatory duct.; S.e.y ejaculatory sac, with apodeme ; Ap., great apodeme ; T/?., Theca of aedeagus ; S., its median spine ; Th' ., its lateral processes ; Ae., Aedeagus ; /)., anterior ridge-procss of ninth segment; ep., epipleuron of ninth segment. After Carpenter and Hewitt. REPRODUCTION AND HEREDITY 149 their sex as exhibited in the sperm- cells — an innate tendency towards rapid motion and dissipation of energy, while the larger, quieter, less conspicuous, less aggressive female follows the tendency of the egg to grow excessively and to store up food. It is, however, noteworthy that in the vast majority of insects secondary sexual characters are not conspicuously developed, and the problem remains why these outward differences should be so unequally evident in various members of the same order or family with regard to sex ? We have seen reason for concluding that the sex of an individual insect depends normally on the germinal constitution of the egg (fertilised or unfertilised) whence it has developed, while the existence of the gynandromorphs and intersexes warns us that the normal development may, on occasion, be disturbed or side-tracked through some irregular behaviour of the multiplying cells. Secondar}^ sexual characters and modes of activity lead us naturally to the subject of that behaviour before actual pairing, wbJch is generally known as courtship. It will, however, now be convenient to turn immediately to the manner of growth of an insect from egg to adult, and then to pass on to aspects of courtship in connection with a general discussion on the family life of insects. CHAPTER VII GROWTH AND TRANSFORMATION In the previous chapter we have discussed the behaviour of the germ-cells in maturation and fertilisation, and the power of determination exercised by their germinal constitution on the nature of the insect that may be developed from either the fertilised or the virgin egg. It is well known that between such an egg and the adult of the next generation there intervenes a longer or shorter process of growth and change of form. Some of the more important features of this process must now be described and discussed. In tracing the complete life-history of an insect from egg to adult it is convenient to discriminate first of all between the embryonic development that goes on within the egg- shell up to the appearance of the young insect in the outer world, and the post-embryonic development which occurs after hatching and brings about the growth of the newborn or newly hatched creature into the mature insect capable of reproduction. As the second of these periods of development is a markedly characteristic feature of insect life, it will be considered in greater detail than the embryonic growth which is common, in some form, to animals generally. Yet the development of an insect embryo presents many features of interest which must not be altogether neglected in our survey of the subject. Reference has already been made to the relatively large size of an insect's egg^ and the amount of food-yolk that is 150 GROWTH AND TRANSFORMATION 151 stored in it. So great is this that the egg-protoplasm necessarily forms centrally a diffuse network, with a con- densed circumferential layer within the envelope or shell of the egg ; in the central network the yolk spheres, relatively large transparent bodies, are found, as well as a Pig. 39.— a, B, C, Embryonic Development of a Tortricid Moth {Eudemis naevana), as shown in transverse sections of the egg-shell with embryo at successive stages. X 50. g, germ band ; a, amnion ; s, serosa ; y yolk • V, vitellophags (yolk-absorbing cells) ; ec, ectoderm ; en, inner- cell mass (endoblast or ventral plate). After L. H. Huie {Proc. R. Soc. Edinh. xxxviii, 1918). number of minute corpuscles. The egg thus contains a diffused mass of living protoplasm capable of division and growth, and a quantity of food-material for nourishing the developing embryo. Such a relatively large, heavily yolked egg is common in the great comprehensive group of animals 152 THE BIOLOGY OF INSECTS — the Arthropoda — to which insects belong. The egg shell in most insects is elongate in shape with rounded ends — the eggs {" fly-blow ") of a bluebottle furnish familiar examples, but the most varied forms — spherical, discoidal, cylindrical, flask-shaped — may be seen, and the outer surface of the shell is often marked with ridges and furrows, presenting to the observer a beautifully sculptured aspect. The embryo (a common term for the unhatched or unborn young) is built up by an orderly process of division (segmentation) of the egg. The zygote-nucleus (p. 117 above) divides in two and the daughter-nuclei divide repeatedly, so that their number rapidly increases. Each nucleus becomes the centre of a protoplasmic mass or cell, though cell-boundaries may not, in the earlier stages of the process, be very evident. As the segmentation of the egg thus proceeds the numerous cells arrange themselves for the most part around the outer region, enclosing a central mass which consists of yolk spheres and corpuscles with a few yolk-cells. Thus there is formed a definite cell-layer or blastoderm surrounding the yolk. Then the blastoderm on one long face of the egg becomes thicker than on the other owing to the deepening of its component cells. The thickened portion or germ-band marks the ventral, the thinner the dorsal aspect of the growing embryo, the inception of which is marked by the insinking of a mass of cells (" middle plate ") along the axis of the germ-band to form a lower layer (endoblast), while the rest of the germ- band overgrows it and becomes the outer embryonic layer (ectoderm) ; thus the embryo assumes a definitely two- layered condition (Fig. 39). Then by uprising of the thin blastoderm around the germ-band or by the inpushing of the latter into the former a protective layer (amnion) is formed over the embryo (Fig. 39, B, C). Meanwhile the ventral embryonic region becomes definitely segmented owing to the successive appearance of a series of transverse inter- segmental grooves. The early embryo may be regarded as consisting of a primitive head and tail ; the remaining segments of the body are formed in succession from before m- mx. 1- I- ab— Pig ao.— a, B, C, D, Stages in Embryonic Development of Grass- hopper " (X/p/uWmm) ; surface views of germ-band or embryo frorn ventral aspect, resting on yolk, pc, head lobes ;e eye; s, mouth [stomodaeum) v, ventral groove ; /feeler -m, mandible ; mx maxilla I labium • I 2, ^, legs ; ab, first abdominal segment ; 7, ». 9. lo. 7tti x'oth Tdiminal^'segm^nts ; .i>, -stigial abdominal appendages ;g^ gonopophyses ; c, cercopods. Magnified. After W. M. Wheeler (jfourn. Morph. viii, 1893). 154 THE BIOLOGY OF INSECTS backwards between these two, so that the tail is pushed farther and farther from the head until the whole series of head, thoracic, and abdominal segments have been formed ; and on some of them the rudiments of limbs bud out. This body-segmentation with the series of appendages can be observed on the surface of the developing embryo (Fig- 40)- But deeper investigation by examination of serial sections shows the origin and elaboration of various sets of organs. The outer layer (ectoderm) gives rise not only to the skin (epidermis) but, as in animals generally, to the nervous system, whose rudiments sink in as a pair of elongated segmented ridges whence are formed the series of ganglia with their connecting cords. Fine paired inpushings of ectoderm on most of the segments mark the position of the spiracles and furnish the rudiments of the tracheal or air-tube system. Median inpushings of the ectoderm near the front end of the embryo and at the tail indicate the future mouth and vent, and grow into the fore-gut and hind-gut respectively ; these, it will be remembered (pp. 5, 23-8, 30-1), as well as the air-tubes, are lined with cuticle in the developed insect. As growth proceeds, the mouth, originally in front of or between the feelers, moves back- wards so that it comes to lie between and behind the mandibles. The lower layer is necessarily situated between the ectodermic structures just mentioned and the yolk ; most of it grows to form a series of segmental cell-masses (meso- dermal somites) each with a pair of cavities (coelomic spaces). From these cell-masses the muscles and connective tissues of the body are formed, regions of many of them growing out into the developing limbs as they arise. Near the walls of certain of the coelomic spaces the primitive germ-cells appear, and these spaces themselves become the cavities of the reproductive organs and their ducts (ovaries and oviducts in the female, testes and vasa deferentia in the male). The mesoderm grows dorsalwards on either side beneath the ectoderm, forming masses of loose tissue, from this the GROWTH AND TRANSFORMATION 155 heart is formed, and the spaces which arise in the spongy mesoderm (mesenchyme) become extended and coalesce to produce the enlarged blood-containing cavity (haemocoel, see pp. 6, 35) characteristic of insects and of arthropods generally. There is one important feature of the embryonic growth as to which many insects seem to show a remarkable di- vergence from animals generally. We have seen that the germ-band early becomes differentiated into outer and inner cell-layers. Of these, the outer (ectoderm) gives rise to the skin and the nervous system as is the case in the vast majority of animals. As a general rule the inner layer of cells (endoderm) becomes the lining of the primitive digestive cavity, and the mesoderm whence the muscular and connective tissues are developed, appears as a derivative of the endoderm close to its junction with the ectoderm. Now in insects, and indeed in most Arthropods, the digestive portion (mid-gut) of the food-canal is much restricted, occupying only a relatively small section of the alimentary tract, the extensive fore-gut and hind-gut being derived from inpushed ectoderm and Hned with cuticle. It is certain that in all insect embryos by far the greater part of the inner cell-layer must be regarded as mesoderm, since from it the muscles, and connective and blood tissues are formed, as is typical in animals generally. The feature of insectan embryolog}^ still most imperfectly understood is the origin of the mid-gut, and it is not established if any definite rudiment in an insect embryo can be identified as certainly comparable to the endoderm among animals generally. This is a surprising gap in our knowledge of development, since the endoderm is one of the two " primary " germ- layers, usually recognisable as a definite entity in a very early stage of animal growth in the egg. Three principal alternative interpretations of the origin of the mid-gut in insects have been given by the numerous investigators of the subject. Many early students of insect embryology regarded the yolk-cells as representing the endoderm and giving rise to the mid-gut ; but in more 156 THE BIOLOGY OF INSECTS Fig. 41.— Formation of the Mid-gut (mesenteron) in various Insects. A, Longitudinal section through front region of Embryo of Mason Bee (Chalicodoma). B, Similar section through same region of Embryo of Leaf-beetle (Donacia). C. Longitudinal section through central part of digestive tract of Larval Dragon-fly (Epitheca). ec, ectoderm ; m, mesoderm ; en, endoderm ; er (in B), rudiment of mid-gut (regarded as ectodermal) ; b (in A), boundary of endoderm and mesoderm ; v (in C), vitellophags (yolk-devouring cells) : s, stomodaeum (forming mouth and fore-gut ; me, mJd-gut ; p, proctodaeum (forming hind-gut). X about 500. A, after J. Carri^re and O. Biirger (Nova Acta Leopold-Carol. Akad. Ixix, 1897). B, after K. Friederichs {lb. Ixxxv, 1906). C, after H. yon Tschuproff {Zool. Ans. xxvii, 1903). GROWTH AND TRANSFORMATION 157 recent work upon this subject the only reliable support afforded to this view is from R. Heymons' research (1897) into the development of the bristle- tail Lepisma, and H. von Tschuproff's account (1903) of the origin of the germ- layers in certain dragon-flies (Fig. 41 , C). In the latter case the central portion, in the former the whole of the mid-gut is said to arise from yolk-cells. Most recent workers in this field state that the mid-gut arises from two rudiments which grow from the inner ends of the fore-gut and hind- gut, respectively backwards and forwards, till they meet ; the origin of these rudiments appears to prove that they are ectotermal. Such is the interpretation given by R. Heymons (1895) of the development of cockroaches and crickets, by A. Lecaillon (1898) and P. Deegener (1900) of that of various beetles (see Fig. 41 , B), and by K. Toyama (1902) of that of the silkworm. But many other investi- gators of insectan embryology have described the mid-gut rudiments as arising from two cell-masses of the inner layer towards the front and hinder ends of the germ-band, so that they can be fairly regarded as endoderm. Such is the interpretation of W. M. Wheeler (1893) from his studies of the germ-layers in cockroaches and beetles, of K. Heider (1889) in his classical research on the water beetle Hydro- philus, of K. Escherich (1900) working at fly embryos, of J. Nusbaum and B. Filinski (1906, 1909) from researches on crickets and cockroaches, of J. Hirschler (1909) from investigations on the development of the beetle Donacia, and of J. A. Nelson (191 5) in his account of the embryology of the hive-bee, confirming the early account given by B. Grassi (1884) on the development of the same familiar, insect and that of J. Carriere and O. Burger (1897) in their full description of the embryology of the Mason Bee (Chalicodoma). These last observers state that the two endodermal rudiments which give rise to the mid-gut in the mason bee embryo appear in those regions of the germ- band v/here the two ectodermal inpushings produce later the fore-gut and the hind-gut (Fig. 41 , A). It may perhaps be possible to find reconciliation between the second and 158 THE BIOLOGY OF INSECTS third views, as set forth here, by supposing that the cell- groups at the inner ends of the two ectodermal inpushings are originally endodermal rudiments which are pushed inwards by the growth of the fore- and hind-gut rudiments of which they appear to form part. It is well known to all students of animal form and development that, largely through the influence of E. Haeckel (1877) and others very great stress has been generally laid on the *' germ-layer theory," according to which the two primary germ-layers (ectoderm and endoderm) give rise throughout the animal kingdom to certain definite regions and systems of the embryo. As regards the ectoderm and its derivatives insects appear to follow the general rule, but on any possible inter- pretation of the facts, there is some abnormality in the origin of the endoderm and its derivatives ; this important layer can be recognised only with great difficulty if at all. Heymons, who finds the endoderm in the yolk cells of the embryo of the primitive bristle- tail, believes that in insects generally the original endodermal mid-gut has been replaced by a new one of ectodermal origin. The dragon-fly embryos, in which TschuproflF states that the central region of the mid-gut is derived from yolk-cells and the front and hind regions from ectoderm, might be regarded as indicating a transition between the. old and the new conditions. As the embryo continues to develop, the tissues and organs which in fully formed insects are dorsal in position, tend to grow away from the primitively ventral germ-band, and the yolk becomes surrounded by the cells that form the lining of the gut. The series of appendages : feelers, jaws, legs, assume to some extent their characteristic shapes before the young insect is hatched ; at their origin all are much alike, but they become diff"erentiated for various functions as growth proceeds (Fig. 40). It is of much interest to notice that in many insect embryos (that of the grasshopper Xiphi- dium, for example) small paired limbs appear on many of the abdominal segments, but vanish before hatching (Fig. 40, C, ap). There may be small or vestigial head-appendages between the feelers and mandibles, and rarely also (J. W. GROWTH AND TRANSFORMATION 159 Folsom, 1900) between the mandibles and maxillae. The bearings of such facts on speculations about the racial history of insects will be discussed in later chapters. The embryonic development of an insect is closed by the experience of hatching (Plate VI, B) which introduces the young creature into the outer world A necessary pre- liminary to hatching, as a rule, is the rupture of the egg- case or shell. This may be brought about by inflation of the cuticle of the neck region just behind the head, so as to burst the egg- case, as in the emergence of the young grass- hopper, or by the young insect biting a hole through the shell with its mandibles or, as in the case of certain fly- maggots, with its mouth- hooks or with a hard and sharp mouth-spine ; or special spinose processes or ridges — " hatching spines " — may be present on the head or prothorax. Such structures, the purpose which they serve, and the manner in which they work are striking instances of what is constantly noticed by students of the life of insects — an apparent prevision of the needs of the creature in succeed- ing stages of its development, so that it finds itself furnished beforehand with the instruments needed for the next act in its life-drama. The whole course of embryonic develop- ment, which has been briefly sketched in the preceding pages, may be regarded as a series of successive events each leading on and preparing for that next to follow, and tending to the construction of the young insect which is to be hatched in due season. Remembering that the process is essentially one brought about by the division and specialisa- tion of cells, we are able to catch some glimpse of the mechanism of the process as we remember that all these cells are derived from the fertilised egg- cell with its complex of inherited factors that render possible the continuity of the racial characters through an innumerable series of generations. The nature of the germ-plasm is such that it can serve as the means for ensuring that the embryo passes through a series of progressive stages that are generally in the same order and of the same nature as those through which its parents and its ancestors passed, each fresh step i6o THE BIOLOGY OF INSECTS in the process following as a response to some internal or external stimulus. The course of embryonic development, like the programme of instinctive behaviour illustrated in a previous chapter (p. 107), suggests to many students the thought of a '' racial memory." As already mentioned in this chapter (p. 150), the mode of growth of insects after hatching or birth is an especially characteristic feature of their life, and the transformation (metamorphosis) which in many cases is an accompaniment of this post-embryonic growth, has, from early times, arrested the eager attention of observant people. A brief introductory survey of the subject has been given in our first chapter (pp. 10—12) ; now it requires to be discussed with some fullness and in sufficient detail for appreciation of the essential problems that it presents to the student. It has been noted that an insect's cuticle, being an outer secretion of the skin and not a sheet of living tissue, cannot grow with the creature's growth ; therefore it must be periodically shed and renew^ed. Thus the life-history of an insect is marked by a series of '' moults " (ecdyses), which divide the period of its growth into a series of stages. Before the actual moulting process the skin sinks away from the old cuticle, pouring out a '' moulting fluid," the secretion in some cases at least, as J. Gonin (1894) and W. L. Tower (1906) have shown, of special large unicellular glands of the skin. Thus there is a fluid-filled space between skin and cuticle. The body now grows quickly, the body- wall, as it is still im- prisoned in the cuticle, being of necessity thrown into ridges and furrows to a greater or less degree. Then the skin begins to secrete a new cuticle beneath the old one ; this is at first soft and flexible, and necessarily follows the folds and wrinkles of the skin. By the pressure of the accumulated fluid the old cuticle is burst open, generally along a median suture or " line of weakness " in the dorsal region of the thorax ; and through the slit thus made the creature in its new cuticle emerges ; as a rule the dorsal region of the thorax comes first, then the head with its ap- pendages, then the legs, and lastly the abdomen (Plate I, A), PLATE VI A. Eggs of Sawfly {Nemattis ribcsU) laid beneath Currant Leaf. X 4. B. Young Larvae of N. ribesil Hatching, x 3. To face p. 160.] lA. G. Britten, photo. GROWTH AND TRANSFORMATION i6i the old cuticle may slip off backwards and in some cases be turned partly inside out. The cuticular linings of the fore- gut, hind-gut, and air-tubes are shed along with the exo- skeleton. After emergence the parts of the creature's body in its new form have room to expand ; the sclerites of the cuticle become firm and darkened so as to assume their characteristic colour, and the successive layers of secondary cuticle begin to form on the surface of the skin beneath the first-formed or primary cuticle. When no longer confined in its " old husk " the insect undergoes a process of expansion through the smoothing out of the folds and wrinkles in the body- wall, so that while it often assumes a form like that which it had borne in the previous stage of its life, it shows soon after the moult a considerable increase in size. Thus through its series of castings and renewals of the cuticle it may grow to a bulk many times greater than it possessed at hatching. Growth accompanied by a number of moults is a necessary feature in the life after hatching of the great majority of insects, and they share this feature with members of other classes of the great group of Arthropoda. But most insects, when adult, are strikingly distinguished from other arthropods by the possession of wings and the power of flight ; the wings furnish a most distinctive and characteristic feature in the insect's structure. Now, while many insects, such as cockroaches, grasshoppers, and plant-bugs, are hatched in a form generally resembling their parents, displaying the same general build of body, shape, and proportions of legs, nature and function of jaws, there is no trace of wings to be seen on these newly hatched young. Every one knows that such a familiar insect-larva as a butterfly's caterpillar exhibits no trace of wings, and the same condition is noticeable in all insects newly hatched or born. So we see that while insects, like arthropods gene- rally, pass through a series of moults in the course of their growth, the development of the wings that are distinctive of the vast majority of insects is carried on during this post- embryonic period, and that no insect has its wings already formed when it first appears in the outside world. M i62 THE BIOLOGY OF INSECTS In the simple direct type of insect- growth, where the young, after hatching, resembles the adult in all essential features, as in grasshoppers, cockroaches, and bugs (Fig. 4), for example, the only marked changes observable in the successive stages are those due to the development of the wings and, in some insects, to the appearance of the female's ovipositor and other processes or appendages connected with reproduction. An insect's wings arise as hollow paired outgrowths of the second and third thoracic segments, their cavities are continuous with the great blood-space of the body and sets of air-tubes grow into them. After the first or second moult in the life-history of a young cockroach or grasshopper the wing-rudiments can be seen as rounded lobes projecting at the hinder corners of the mesonotum and metanotum, and after each successive moult they become larger than before, displaying in some cases, on the surface of the cuticle, branching tracks which indicate the courses of the air-tubes within. The wing-rudiments, which may be regarded as flattened pouches, become more markedly flattened as their areas extend so that they approximate to the condition in the adult insect. But even in the stage before the final moult, the wing rudiments of an immature cockroach, grasshopper, or bug are shorter by far than the wings of the adult insect. It is necessary, therefore, as part of the preparation for this moult, that the wings should grow extensively and rapidly beneath the separated cuticle and thus something of a crisis is apparent at this stage of develop- ment. When the last moult has taken place the newly exposed wings are seen to be greatly folded or crumpled, a necessary condition of their extensive growth beneath the cuticle of the penultimate instar. When exposed they unfold and flatten from the base outwards, the cuticle becomes hard and firm, and the floor and roof of the hollow outgrowth of the body from which the wing arises become approximated together except along the courses of the air- tubes where the thickened cuticle forms the supporting tubular nervures or " veins " of the wing. In such life-histories as those that have been briefly GROWTH AND TRANSFORMATION 163 described, the young insect, all through its growth in general aspect, and indeed in many details of structure, resembles the adult, and its wing- rudiments appear at an early stage as outgrowths of the two thoracic segments that carry the wings. Such insects undergo no marked trans- formation ; a young grasshopper has the long leaping legs, and a young cockroach the flattened body and rounded Fig. 42. — Forms of Vine Aphid {M acrosiphum vittcola) , Amenca . A, newborn young, X 70 ; B, nymph with wing-rudiments, X 20; C, winged virgin female, X 25 ; D, adult wingless female, X 35. After A. C. Baker (jfotdrn. Agric, Res. U.S.D.A. xi, 1917). pronotum of their respective parents. In a study of the biology of insects it is noteworthy that the similarity of form between adult and young goes along with the similarity in the mode of life ; young cockroaches in various stages of growth may be found along with adults sheltering in cracks of walls or lurking beneath hot-water pipes in houses, while young grasshoppers and locusts walk or leap among herbage 1 64 THE BIOLOGY OF INSECTS and devour leaves as their parents do. A very familiar example of this likeness of the young to the adult in habit and in form is afforded by the Aphids or " greenfly " (Fig. 42). In the spring and summer virgin female broods of these abundant insects, the mother may be seen on a leaf of her food-plant surrounded by her large family of newly or lately born young. They have the same general aspect as their parent, the same tapering abdomen with its prominent, paired cornicles, and they feed in just the same way by piercing the plant-tissues and sucking thence a continual supply of sap. Newly born aphids are, of course, all wingless, and it is interesting to find that in a large pro- portion of the spring and summer females (Fig. 42, D) — produced in a series of virgin generations — ^wings are never developed, so that the adults, never wandering far from their birthplace over their native plant, remain wingless like the new-born young. In many of the aphid summer females, however, wings are developed (Fig. 42, B, C) from outward rudiments that increase in size after each moult, and these winged individuals can fly away to other plants so as to extend their feeding-ground. Among the aphids the absence of wings accompanies passivity of habit, and the same connection is still more strikingly shown by whole groups of insects that pass their lives, from egg to adult, on the bodies of animals, deriving thence their food- supply, such as the Anoplura or blood-sucking lice, and the Mallophaga or biting lice which nibble at the hairs or feathers or bite the skin of their mammalian or bird hosts. These groups are entirely wingless throughout life and afford interesting examples of the association of winglessness with the parasitic habit among insects. Here, as might be expected, the adult differs from the newly hatched louse in little except size and the development of the organs of reproduction. As no wings appear, the series of moults through which the insects pass is marked by the smallest possible change of form. Where, however, the insect in its earlier stages lives among surroundings or under conditions differing from GROWTH AND TRANSFORMATION 165 those of the adult, we find, as a rale, more or less difference in structure, and such difference involves transformation of a greater or less degree in the course of the life-history. For example, there are two families of sucking-insects — the Fsyllidae and the Cicadidae — related to the aphids mentioned above. Aphids undergo little or no change of form in the Fig. 43. — Apple Sucker {Psylla mali) . a, female, X 8; 6, egg, X 80 ; c, first-stage larva (ventral view), X 100 ; d, nymph, fifth instar (dorsal view with legs shown on left, feelers and wing- rudiments on right), X 20. After G. H. Carpenter {Econ.Proc. R. Dublin Soc.i, 1909). process of their growth, and young or adult aphids usually live, as we have seen, under much the same conditions. But in the life-cycle of a psyllid or a cicad striking changes of form are to be noted, the young of these insects living in conditions quite different from those of the adult. Psyllids i66 THE BIOLOGY OF INSECTS or ** suckers " are active little insects with firm cuticle, a body of considerable depth dorso-ventrally, relatively long feelers and legs, and well-developed wings (Fig. 43, a) ; they fly and leap on the shoots of plants whose sap they suck for food. These insects in their young stages are found between the leaves of partially opened buds or clinging to the under surface of the foliage, or in cavities due to the folding or crumpling of leaves apparently resulting from a response evoked by the irritation of the insect's presence, or in definite galls arising in the same manner. In corre- spondence with such environment, the young sucker has a flattened body relatively broader than that of its parent, rounded in front and behind. The early stages of the Apple Sucker {Psylla malt) have been described by Carpenter (1910) and by Awati (191 5). In the newly hatched young the feelers and legs are short with fewer segments than in the adult, and the head seems to be fused with the thorax, as there is no division dorsally between crown and pronotum, while ventrally the beak lies between and behind the bases of the fore-legs. The cuticle is relatively soft and thin, and the young insect diff"ers so markedly from its parent that it may be called a larva — the general term (meaning literally a *' mask ") applied to any young creature which has to undergo transformation before it reaches the adult state. The larval apple sucker (Fig. 43, c), after hatching from its curiously stalked egg, wanders to a blossom-bud outside which it waits until the scale-leaves open sufficiently to allow it to crawl inside. Then amid the soft, crowded, developing young leaves it finds shelter and abundance of food, so that it grows quickly, passing through five stages within the bud. In the second of these, as in the first, no wing-rudiments are evident, but in the third and subsequent stages wing-rudiments appear on the thorax ; the creature has now become a nymph whose general body-form is still markedly different from its parent, the prothorax not yet marked off from the head, though the feelers and legs begin to approach the adult proportions. Each of the first three stages lasts about a week, the two later nymph stages GROWTH AND TRANSFORMATION 167 (Fig. 43, d), in which the wing- rudiments are well-developed and the cuticle is relatively firm, each last for ten days or a fortnight. After its fifth moult the wings are fully developed and the sucker assumes the adult form. Cicads are much larger insects than suckers or aphids ; Pig, 44. — Seventeen-year Cicad {Tihicina septendecim) , North America, a, first-stage larva, X 20 ; b, fourth-stage larva, X 5 ; c, feniale, natural size. After C. L. Marlatt (Entom. Bull. 71 , U.S.D.A. 1907). they form a prominent feature in the insect fauna of most warm and tropical regions, but they are represented in England by a single species, found only in the south, and there but rarely. They have robust bodies, and broad i68 THE BIOLOGY OF INSECTS heads with large eyes and short feelers ; the front legs are stouter than those of the middle and hind pairs. They have ample wings and fly about in the woodland regions alighting on trees from whose leaves they suck sap. The habits and life-history of the common North American Tibicina septen- decim (Fig. 44) have been well described by C. L. Marlatt (1907) and R. E. Snodgrass (1921). The female cicad possesses a long cutting ovipositor wherewith she excavates slit-like cavities or " nests " in the twigs of trees and deposits her eggs therein. The young, when hatched, live for a time crawling about the branches of their native tree and then drop to the ground and burrow into the soil. The newly hatched cicad-larva (Fig. 44, a) differs from its parent by its soft, pale cuticle, its elongate ovoid head with relatively long feelers and small eyes, its regularly segmented body with little differentiation in the various regions, and its very broad and powerful fore- legs with the spinose tip of the sliin adapted for burrowing. The successive stages of the life-history are passed underground, and during these the foot (tarsus) of the fore-limb, well developed in the newly hatched young, becomes greatly reduced. In the two well-marked races of Tihicina septendecim it is well known that the growth and transformation of the individual insect is carried on through a period of twelve and sixteen years respectively." Such excessively lengthened life-cycles afford extreme instances of a condition often to be noticed in insect development — a prolonged period of preparation through most of which the creature feeds and grows, leading up to a comparatively brief adult existence the beginning of which usually marks the end of growth and change, and which, whether the creature's way of life and manner of feeding are like or unlike those of its early stages, is strikingly curtailed in its relation to the prolonged period of immaturity. In their manner of development cicads display yet another feature of much general interest. After the long underground life of years during which the creature feeds by sucking sap from roots which it pierces with its needle- GROWTH AND TRANSFORMATION 169 like jaws, there follows before the emergence of the winged adult, a long period of quiescence. The nymph in its last stage, with its front feet again developed, and displaying prominent wing-rudiments, rests motionless in its earthen burrow for several days or weeks, taking no food while it awaits the time of its final moult. All immature insects suspend their activities for a while when preparing to shed the cuticle, and we have already noticed that the final moult, as it precedes the full development of the wings, marks always something of a crisis in the life-history ; the stage leading up to this may therefore be regarded as naturally suitable for prolongation into a definite resting stage. It is noteworthy that such a quiescent interlude is to be seen in the life-histories of insects of several families allied to the cicads and classed in the same sub-order (Homoptera) as they. Thus the Coccidae (mealy bugs and scale insects) and the Aleyrodidae (Plate I, B) pass through a resting stage which may indeed begin at a comparatively early period of the development, before the last moult but one or the last but two ; in these insects, however, though motion ceases, feeding by suction may go on. Scale insects and snowyflies exhibit a remarkable modification of the moulting process in connection with these resting phases. The old cuticle, after separation from the skin, is not cast off, but becoming hard and firm serves as a protective case for the creature in the comparatively thinly-coated condition that characterises it during the next stage ; an emerging snowy- fly may thus have to make its way to the outer world through three successive discarded cuticles. A striking feature of the life-history of many Coccidae is that the female scale insect never emerges at all ; while the male develops wings the female not only remains wingless when adult (like the summer aphids already mentioned in this chapter), but passes the rest of her life and lays her eggs under the shehering '' scale," which consists of the last-shed cuticle strengthened and enlarged by waxy secretions of the skin. In the insect life-histories so far sketched in this chapter we have seen that the newly hatched young may resemble 1 70 THE BIOLOGY OF INSECTS generally or differ markedly from its parent, as a rule, according as its mode of life is the same or diverse, and that, in either case, the process of development is largely concerned with the acquisition of wings, while before the final perfect- ing of these organs there may be a prolonged resting period. We pass on now to consider some life-histories of another type which prevails among the great majority of the insect families. After comparison of the two types it should be possible to appreciate the essential difference between them. It is a matter of common knowledge that most insects during their life-histories pass through a marked trans- formation (metamorphosis) ; the change of a caterpillar into a butterfly, for example, is familiar to every one, and hardly less familiar is the fact that maggots feeding in dead flesh or carrion are the offspring of bluebottles, and that into bluebottles they will in due course be changed. The caterpillar displays many conspicuous features of divergence from its parent butterfly, and the maggot is still more dis- similar to the bluebottle. In the process of transition from the one to the other there must evidently be a considerable amount of reconstruction, and it is therefore not surprising that in the stage preceding the adult, the insect is a quiescent pupa, remaining usually motionless and taking no food. We have already seen that in several groups of insects — cicads, scales, thrips — whose growth exhibits far less marked change of form than the growth of a butterfly or bluebottle, there is partial or complete passivity during the penultimate stage. The quiescence of the pupa, then, is not the essential feature that distinguishes what is generally called '' com- plete " from '' incomplete " metamorphosis ; it is necessary to seek farther for the true distinction. The caterpillar hatched from the egg of a butterfly or moth differs conspicuously from its parent, but the differences are in details of structure, not in the funda- mental plan of the body. A caterpillar (Fig. 45) has the typical insectan head with all its appendages and organs present, though several of them are simplified or speciaHsed as compared with those of most adult insects ; thus the feelers GROWTH AND TRANSFORMATION 171 are very short and inconspicuous, the compound eyes are replaced by a few ocelli, and the maxillae have very short palps and reduced lobes, but the labium though small has its central ligula drawn out into a spinneret whence the silken thread formed as a secretion of the specialised salivary or silk glands, is passed out. The three segments behind the head obviously make up the thorax of a typical insect, as each carries a pair of jointed legs ; yet the caterpillar's leg is very short compared with that of the butterfly, its foot-segments are undifferentiated and it has only one claw Fig. 45. — a, Dorsal, and 6, lateral view of Caterpillar of Diamond- back Moth (P/wfe/Za crwa/erarww) ; c, pupa (ventral view). X 6. From Carpenter (journ. Dept. Agr. Ireland, I). The abdomen of the caterpillar is composed of the ten segments usually recognisable in the hind body of an insect. In most caterpillars five of these (the third, fourth, fifth, sixth, and tenth) carry each a pair of short cylindrical pro- legs armed with circles or crescents of spines ; these pro- legs are of value in regard to the caterpillar's special mode of life, as they enable the creature to cling to or crawl along 172 THE BIOLOGY OF INSECTS a twig or even a leaf-edge of its food-plant. The body- segments are still all much alike, the cuticle is usually thin and flexible, and the general aspect of the caterpillar may be described as worm-like ; it is essentially a " creeping thing." The newly hatched caterpillar is very small, but it feeds voraciously and grows quickly, passing through its successive stages and undergoing four or five moults before it attains its full size. The caterpillar in its last stage is enormous compared with what it was when it left the egg, but it does not differ in any essential feature of outward form. Head, body-segments, jaws, legs, and pro-legs appear after each moult much as they did before it, and at no stage of larval life is there any trace of outward wing- rudiments. This last feature is, as D. Sharp (1898) pointed out, by far the most important of the readily observable distinctive cha- racters of the type of life-history illustrated by the trans- formation of the caterpillar into the butterfly ; we have seen that in the growth of cockroaches, bugs, aphids, and cicads, there are evident wing-rudiments at an early stage of growth after hatching, and the same condition is found in the aquatic nymphs of stone-flies, may-flies, and dragon-flies. But in the development of the butterfly no trace of wings is apparent until the last larval cuticle has been shed and the pupa revealed ; on the pupa (Fig. 45, c) the wings may readily be seen at either side of the body, so closely adpressed indeed that they do not stand out, but quite recognisable as to their shape, as are also the legs and feelers, elongate like those of the adult, and sometimes also the slender, flexible maxillae which will enable the butterfly to feed by suction, the biting mandibles of the caterpillar used for feeding on solid plant tissues having vanished. The pupa, then, resembles the adult insect much more closely than the larva, and this can be seen more clearly than in the case of the butterfly if we study the pupa of a beetle (Fig. 47, b) or a bee. For in these insects the pupal wings and legs are not closely adherent to the body as in the '* obtect " butterfly chrysalid, but stand out in the manner GROWTH AND TRANSFORMATION 173 characteristic of the '' free " pupa, as such a type is called. The obvious presence, when the pupa is revealed, of wings and other organs characteristic of the " imago " or perfect insect confirms our impression that the pupal stage of the life-history indicates a period of reconstruction and com- paratively rapid change. But a knowledge of the processes of animal growth in general leads the student to infer that such profound changes could not be brought about without previous preparation. So that we are led to expect that wing-rudiments must be somewhere present in the cater- FiG. 46. — A, B, C, Stages in development of wing-bud (b) of a Lady-bird Beetle (Hippodamia) , shown in section, c, cuticle (shown in A only) ; e, epidermis ; s, sensory hair-cell ; i, trachea; fr, tracheoles. X 100 (approx.). After Comstock and Needham, " Wings of Insects." pillar or other insect larva. They were indeed observed more than a century and a half ago when P. Lyonet in his great treatise (1762) on the caterpillar of the " Goat " Moth (Cossus) sav/ two pairs of small white bodies lying in the fatty tissue of the second and third thoracic segments. He did not certainly recognise them as wing- rudiments, but he pointed out that their number and position suggested that such might be their nature. They are indeed the wing-buds of the insect, lying hidden beneath the body- 174 THE BIOLOGY OF INSECTS wall. These '' imaginal discs " as they are now called, have been detected in all metamorphic insects whose develop- ment has been carefully traced. J. Gonin, for example, has shown (1894) ^^^^^ i^ ^^^ White Cabbage Butterfly (Pteris) they arise as thickenings beneath the skin, and grow inwards as they increase in size in such a way as to form little flattened hollow pads lying in thin-walled pouches continuous with the skin whence they originate ; thus, although they are situated within the body, they retain their primitive connection with its outer wall. Branches from the air- tube system grow into them, prefiguring the main features of the nervures in the developed wing. After the last larval cuticle has separated from the skin in prepara- tion for the final moult of the caterpillar, these wing-buds grow very quickly and are thrust out from their pouches ; thus projecting from the surface they become covered with cuticle, and so the wings are apparent when, the moult completed, the pupa is revealed. A similar mode of wing- growth has been traced in the larva of a Lady-bird Beetle {Hippodamia) by Comstock and Needham (1899), some of whose drawings are reproduced here (Fig. 46). Wing- buds of essentially the same type can be demonstrated in the grub of the Honey Bee (Fig. 50,/, h). Not only the wings, but all the organs of the winged adult that become apparent in the pupa, arise in the larva as imaginal discs, often sinking within the body but remain- ing connected by strands of tissue with the skin whence they first develop. Up to the third larval stage the leg of a caterpillar may be cut off without damage to the corre- sponding limb of the adult, but if such mutilation be perpetrated later in the course of development, the tip of the shin and the foot of the imaginal leg will be removed, as these then project into the cavity of the larval leg, though the basal region of the limb is sunk in a lateral depression of the body. The transformation of the internal organs differs in nature and degree in the various systems of organs and in the various orders of insects. Generally it may be stated that the nervous system, the heart, and the ovaries GROWTH AND TRANSFORMATION 175 or testes of the adult arise directly from the corresponding structures present in the larva with changes as to size and elongation, as well as separation or fusion of segmental structures such as nen^e-ganglia. On the other hand, the digestive system and associated structures undergo dissolu- tion at the close of the last larval stage and the corresponding organs of the winged adult are developed from special imaginal cells, which may be recognised during larval life, either appearing as small scattered units among the larger normal cells of the larv^al digestive epithelium, or forming aggregated groups at definite regions of the larval food- canal. Such small imaginal groups of cells, relatively few in number, give rise to the greater part of the internal organs of the adult, so that these organs may be regarded as new formations during early pupal life, while the larval structures, now no longer needed, are broken down by the chemical action of special enzymes, the effete products of the disso- lution process being devoured by active, wandering ** amoeboid " cells like white blood-corpuscles. This destructive process is known as *' histolysis," and the re- placement due to the growth of the imaginal discs as " histogenesis." For a detailed account and discussion of these processes reference may be made to the treatise of L. F. Henneguy (1904). The muscles and the air- tubes also undergo, like the digestive system, dissolution and reconstruction more or less profound according as these systems in the imago differ from those in the larva. The tissue of the larval fat-body serves as a reservoir of food- material which is drawn upon for the energy needed in the rapid processes of growth and change that go on at the crisis of transformation in the insect's life-history. These internal changes, the nature of which were in part elucidated by the work of A. Weismann (1864) on the metamorphosis of flies, and have been traced in full detail by subsequent students, are no less surprising than the changes in outward form which led some of the earlier naturalists to regard the formation of the pupa and the subsequent emergence there- from of the winged adult as a veritable new birth. All the 176 THE BIOLOGY OF INSECTS various imaginal rudiments, however, are formed from organs or from groups of cells in the larva whose tissues have arisen from the segmenting egg in the course of embryonic development. The comparison of a pupa to a *' second egg " is therefore fanciful, and the life-history of the butterfly or bee after hatching must be regarded as a specialised and curiously modified form of growth. In all such insects the hidden development of the wing-buds, lying apparently within the body, affords, as D. Sharp (1898) pointed out, a definite character by which these metamorphic or holo- metabolous insects (Endopterygota) may be separated from those hemimetabolous insects, undergoing little or com- paratively sHght transformation and exhibiting outward wing- rudiments early in their life-history (Exopterygota). We find here the essential distinction between *' incomplete " and " complete " metamorphosis among insects. The comparison of a butterfly with its caterpillar demonstrates that the adult insect or imago diflfers from its larva not only in structure but also in its manner of Hfe, and this aspect of the study of insect transformation appeals to those interested in questions of Hfe-relations and the organism's adaptations to meet the conditions under which it has to exist. The winged adult flies while the larva crawls ; the butterfly sucks nectar from flowers while the caterpillar eats solid pieces of plant tissue, leaves, roots, or wood, which it bites off with its strong mandibles. As the student of the biology of insects reviews a series of larvae belonging to various types, he becomes convinced that in each case the form of the larva is adapted to its habitation and manner of feeding ; its differences from the adult correspond with different life-relations. It has already been pointed out that although the cater- pillar differs markedly in aspect from its parent butterfly, its body is built essentially on the same general plan. There is agreement in the number of segments, in the regions (head, thorax, and abdomen) into which they are grouped, in the relative positions of the various systems of organs. The difference between a newly-hatched caterpillar and a GROWTH AND TRANSFORMATION 177 butterfly may easily be exaggerated. It is necessary to remember that such a young insect, rightly called a larva because in it the aspect of the adult is to a considerable extent masked, must not be regarded as an embryo hatched before its time. It presents, for example, a great advance in its stage of development on the young larvae of many Crustacea (such as water-fleas, certain shrimps, and crabs) in which there are but a few segments and limbs apparent. Fig. 47. — c, Ground-beetle (Chlaejiius bioculatus), India, a, larva; h, pupa. X 5. From T. B. Fletcher {Bull. 89, Agr. Res. Inst. Pusa, 1919). the greater number of these appearing only after hatching. Still more does it display a contrast to the early larvae of starfishes and their allies, which are veritable precociously hatched embryos, comparable at most to those earliest stages in the development of insects that follow the segmentation of the egg. Study of a series of grubs belonging to different orders of insects, or even to members of the single order of beetles (Coleoptera), furnishes examples of lars^ae some of wliich N 178 THE BIOLOGY OF INSECTS differ from their adults less than a caterpillar differs from a butterfly while others differ much more. The grub of a ground-beetle (Chlaenius), Fig. 47, has the terga of the segments strongly chitinised so that the body is well armoured, and its feelers and legs are relatively long, while Fig. 48. — Chafer (A?iomala bengalensis) , India, a, larva (side view); b, female. X 3. From T. B. Fletcher {_Bidl. 89, Agr. Res. Inst. Pusa, 1919). the mandibles are powerful and provided with strong sharp teeth ; such a grub feeds, like its parent beetle, on weaker insects which it captures and devours. The well-known '* wireworm," or larva of a click-beetle (Agriotes), has also a strongly armoured body ; it is, however, narrow and GROWTH AND TRANSFORMATION 179 elongate with very short legs adapted for working its way through the soil where it spends its relatively long life of two or three years feeding on roots of plants. A chafer grub (Fig. 48) also feeds on roots, but does not wander as the wireworm does ; only its head and its relatively long legs are firmly chitinised, the cuticle of the body-segments remains pale and flexible, the tail region being somewhat swollen, so that the grub looks like a fat caterpillar without pro-legs. It spends much time resting in an earthen chamber some distance underground feeding on adjacent roots. The larvae of beetles of the " death-watch " group (Anobium, etc.) live and feed in tunnels which they make in wood, or among stored dried food - materials ; they resemble somewhat minia- ture chafer- grubs, but their heads are smaller and their legs much shorter. From these we pass naturally to the larvae of weevils (Curculionidae) and bark-beetles (Scolytidae), Fig. 49, in which the body- cuticle is white, flexible, and wrinkled, while legs are altogether wanting ; such grubs live in concealed situations in the soil, or in plant tissues, mining leaves or timber, or in galleries beneath the bark. These larvae clearly differ from their parent-beetles more than a caterpillar diflFers from a butterfly. Other orders of insects show still greater divergence between larva and imago. A wasp or bee-grub (Fig. 50) is legless and pale like a weevil's, but its cuticle is smoother and more delicate and its head much smaller. Fig. 49. — Pine Bark-beetle {Dendroctonus hrevi- comis). North America, o, larva (side view) ; 6, male. X 9. From J. L. Webb {U.S.D.A. Ent.Bull. 58, 1906). i8o THE BIOLOGY OF INSECTS Among the two-winged flies (Diptera) all the larvae are destitute of true legs, and in the house-fly and bluebottle group (Muscoidea) the head-region becomes so much Fig. 50. — Larva of Honey Bee (Apis mellifica). A, side view, X 4 ; zv, wing-buds seen through skin ; s, spiracles. B, ventral body- wall with nerve-cord (w) exposed by dissection, X 12; 6, brain ; l^, /g, I3, imaginal buds of legs, /, forewing and //, hind wing buds, in their pouches; g, 8, 9, developing gonapophyses (processes of ovipositor) . After J. A. Nelson {Journ. Agr. Res. U.S.D.A. xxviii, 1924). reduced as to be hardly recognisable. The maggot (Fig. 51) of such a fly tapers from the broad tail to the narrow GROWTH AND TRANSFORMATION i8i vm Fig. si.— Larva ("Apple Maggot") of Trypetid F\y (Rhagoletts pomonella) . A, lateral view, X lo ; B, diagrammatic longitudinal section through anterior region; C, pharynx and mouth-hooks, lateral view, X 30; D, section through anterior region of maggot, X 30. i, pro- thorax; 2, mesothorax; 3, metathorax ; /, first abdominal segment ; F///, eighth abdominal segment ; ^ , anterior lateral plate of pharyngeal skeleton; ah, antennal bud in frontal sac; An, anus; yl 6^, anterior spiracle ; Atr, atrium, anterior, part of larval pharynx resu ting troni involution of original head of larva; B, posterior lateral plate ot pharyngeal skeleton ; Br, brain ; C, dorsal or wing plates of pharyngeal i82 THE BIOLOGY OF INSECTS skeleton ; c, ridge on plate C of pharyngeal skeleton ; D, bridge plate of pharyngeal skeleton in roof of atrium ; DiMcl, dilator muscles of pharynx; DP, dorsal pouch of atrium, divided beyond base into two wings containing plates C of pharyngeal skeleton and leading to roots of frontal sacs ; DPMcl, dorsal protractor muscles of pharynx ; EMcl, extensor muscles of oral hooks ; FMcl, flexor muscle of oral hook ; FS, frontal sacs, containing imaginal buds of antennae and compound eyes ; GC, gastric caecum ; Gng, ventral ganglionic nerve mass ; g, h, sensory papillae of snout of larval head; Hk, mouth-hooks; Lj, Lg, L3, leg buds ; LbB, imaginal buds of labium ; LH, larval head ; LMy labrum ; LPMcl, lateral protractor muscles of pharynx; Mth, larval mouth; ofe, imaginal bud of compound eye ; (E, gullet ; Phy, lumen of larval pharynx ; PSp, posterior spiracle; Pvent, proventriculus ; SalD, salivary duct; »Sa/G/, salivary gland ; F^f, stomach. From R. E. Snodgrass (Jowrw. Agr. Res. U.S.D.A. xxviii, 1924). front end, where there are paired sensory tubercles, and strong mouth-hooks used for tearing, which can only with much doubt be compared with typical insect mandibles. An interesting peculiarity of the muscoid maggots is the restriction of spiracles to a large pair at the tail-end of the body and a small pair on the prothorax which can have but a very restricted function (Fig. 51, PSp, ASp). In some of these maggots several pairs of the lateral spiracles have been detected in a vestigal condition, their connect- ing air-tubes excessively slender and solidified by internal deposition of cuticle. These larvae of wasps, bees, and muscoid flies, which differ — especially the last — so profoundly from their parents, are adapted each to its characteristic mode of life. The wasp-grub rests in its paper chamber (Plate VII) in the nest where it is fed on insect-fragments by its sisters the worker- wasps, and the bee-grubs in the waxen chambers of the comb are provided by the worker-bees with floral food materials such as honey and pollen. The soft defenceless cuticle, the small head and relatively weak jaws are enough for creatures that are protected and provided for, have no need to flee from enemies nor to wander in search of food. A remarkable feature of these hymenopterous grubs is that throughout the larval stages the hind intestine is closed and no waste matter passes from the food-canal until just before pupation ; this seems a suitable adaptation in view PLATE VII Nest of Tree-wasp ( Vespa 7iorvegica). One-sixth size. Envelope partly cut away Comb of. Wasp-nest (F. vulgaris), seen from below. One-ihird size. To face p. 182.J [//. Britten, plioto. GROWTH AND TRANSFORMATION 183 of the highly nutritious food of these grubs and their pro- longed period of residence in a crowded nest or hive. Similarly the muscoid maggot, with its front region tapering towards the head armed with strong mouth-hooks, is excel- lently adapted for burrowing into the mass of its foodstuff — the bluebottle's larva, for example, into soft flesh, the house- fly's into horse-dung or garden refuse, the cabbage-root fly's or the mangel fly's into its appropriate plant-tissue ; and in many cases it is easy to recognise the advantage of restricting the functional spiracles to a pair of large ones at the tail-end which remains nearest to the free surface of the food-mass within which most of the maggot's body is buried. In the larvae of several groups of flies such as the gnats (Culicidae) and the drone-flies (Eristalis) these tail spiracles are found at the end of a very short or elongate hinder outgrowth of the body, enabling the grubs, which live under water, to obtain contact with the atmosphere through the surface-film and thus breathe the upper air while they feed in a ditch or puddle which is possibly most foul. The study of such a series of insect-larvae as we have rapidly passed in review brings out clearly the striking adaptation of each to its own manner of life during the period of immaturity and growth. It also suggests that the adaptations have been brought about by the divergence in a less or greater degree of each larva from the form and con- ditions of its parent. Those young insects such as grass- hoppers, cockroaches, and bugs, which resemble their parents closely in form, usually live in the same surroundings as the adult and on the same kind of food. The fact of larval adaptation to special life- conditions in conjunction with the fact that an insect-larva's structure is comparable with that of an adult rather than with that of an embryo, suggests most strongly that the creatures during the early stages of their life-history have diverged from the primitive parental type, in many cases by degeneration, while the adults have diverged by specialisation and elaboration. This view is confirmed by consideration of our series of i84 THE BIOLOGY OF INSECTS young insects, which clearly illustrates increase in divergence of larva from imago. It also helps to explain that feature of insect metamorphosis according to which the trans- formation becomes most profound in the most highly specialised groups. It is well known that among animals generally, marked transformation in the course of the life- history characterises creatures of comparatively primitive organisation, w^hich live in the sea and as a rule produce eggs of small size. This combination is illustrated by the pro- found transformations undergone by starfishes and other echinoderms, or by the marked change of form in growth after hatching to be observed in most fishes, compared with the young of terrestrial reptiles and birds hatched from large-yolked eggs in a condition already well-developed. Insects form a class of creatures essentially terrestrial and aerial, whose eggs are of relatively large size. Yet the young of most insects pass through marked changes after hatching and the greatest degree of change is shown by members of the most highly specialised orders. From the facts surveyed in this chapter it will be apparent that the insect larva, even of a type so degraded as the muscoid maggot, is not a precociously hatched embryo, but a modifi- cation of the type of structure displayed by the developed insect. Whenever the larva differs markedly from the imago we find that it lives and feeds differently from the latter, and we conclude that there must have been specialisa- tion not in one direction only, but in two. The imago shows high elaboration in the form of jaws, wings, sense- organs, while the larva, even if degenerate, is also itself specialised in correspondence with a mode of life widely divergent from that of the winged insect. The degree of divergence between the adult and the larval structure and life necessitates a corresponding degree of reconstruction at the crisis of development marked by the pupal stage, involving a resting period in the life-history during which the reconstructive processes can be carried out. In the series of gnibs, belonging to beetles and other insects, illustrating increasing divergence between larva and GROWTH AND TRANSFORMATION 185 imago we noticed (pp. 177-8) that there is a type of larva with cuticle predominantly firm so that the body is well armoured and provided with relatively long feelers and legs. When such a larva is slender in build and furnished with a Fig. 52.— a, Nymph of Mayfly {Chloeopsis dipterd) , w, wing rudi- ments ; g, abdominal gills . X 8 . After Vayssiere and Eaton . B , Bristle- tail (Petrobius maritimus) y female (ventral view), a5, abdominal stylets ; 0^, ovipositor, x 5. In part after J. T. Oudemans. pair of tail appendages (cerci), as in the familiar mayfly nymph (Fig. 52, A), it presents, as F. Brauer (1869) pointed out, rather strong likeness to a bristle-tail (Thysanuran) such as a Machihd (Fig. 52, B) or Campodea, hence it is i86 THE BIOLOGY OF INSECTS often distinguished as a " campodeiform " larva. As the wingless bristle- tails are the most primitive of living insects, it has been suggested that the campodeiform is the primitive type of insect larva, and that larvae such as caterpillars or chafer-grubs — the *' cruciform " type of A. S. Packard (1898) — are to be regarded as more strongly modified in correlation with their habits, which differ more markedly from those of their adults. In the maggot of a muscoid fly we see a still more profoundly modified, in fact degraded, " vermiform " type of larva. Confirmation of the opinion that these types indicate an increasing degree of divergence between larva and imago, is afforded by those insect life- histories which exhibit more than one larval type in the course of development. For example, the young of many oil and blister beetles are hatched from the egg as tiny active, armoured campodeiform larvae which seek to attach themselves to the body of a bee, and if successful, are carried to the nest, where, after the first moult, they become changed to soft-coated grubs feeding on the stored honey. The campodeiform precedes the cruciform type in the one life- history, and the inference is drawn that the former con- dition is exceptionally retained because the parent beetle cannot enter the bee's nest so as to lay her eggs where the grubs will spend the greater part of their term of existence, and the active long-legged, armoured form of larva is the best adapted for making its way thither. In the case of the vast majority of metamorphic insects which place their eggs upon, within, or adjacent to the material whereon their grubs will feed, the young insect, as soon as hatched, conforms to the cruciform or the vermiform type. These considerations throw light on the nature of the primitive immature insect and on the problem of the origin of the more specialised and degraded larval types. It may now be advisable to pass to another problem which con- fronts the student of insect development : the relation of the open (exopterygote) to the hidden (endopterygote) manner of wing- growth. We have seen that the former is GROWTH AND TRANSFORMATION 187 characteristic of the more generalised and the latter of the more specialised orders of winged insects. From this, as well as from the probability that there has been mutual divergence between larva and imago among the meta- morphic insects, it may be presumed that the hidden method of wing-growth has in these orders superseded the primitive open method. It is by no means easy, however, to under- stand why the wing-rudiments which are evident on the outer aspect of a young grasshopper early in its life-history, do not in a beetle or moth pupa become externally visible until the last larval cuticle has been shed and the pupa revealed. We have seen, from the origin and growth of the imaginal wing-buds in a caterpillar, how this state of things is brought about ; the problem that confronts the student is to find a reason why they sink into apparently internal pouches instead of growing outwards. The fact that they grow outwardly in the more primitive orders of insects indicates that the hidden type of growth must be regarded as secondary ; the problem may therefore be stated as the mode of derivation of the one type of wing-growth from the other, the origin of the endopterygote from the exopterygote life-history. In elucidation of this problem it may be instructive to notice examples of the abnormal appearance of outward wing-rudiments on the larvae of certain metamorphic insects. This was observed in " mealworms," grubs of the beetle Tenehrio molitor^ by R. Heymons (1896), and has recently been studied with some detail in that same species by H. Singh-Pruthi (1924). Abnormal outward wing-rudi- ments on mealworms (Fig. 53) have been usually noticed on well-grown specimens, but Singh-Pruthi has demonstrated them on comparatively early larvae and has shown that their appearance is facilitated by submitting the insects to a high temperature, which has the effect of retarding or preventing the final transformation into beetles. Most of these abnormal mealworms fail to pupate ; the pupae that do result are also abnormal, and those individuals that succeed in reaching the pupal stage rarely develop into beetles. 1 88 THE BIOLOGY OF INSECTS Apparently these abnormal insects at all stages experience a difficulty in shedding the cuticle : " Sometimes ... an individual resembles neither a pupa nor a larva, as it has the head and thorax of the former and the abdomen of the latter." A fact of much interest discovered by Singh-Pruthi is that in the abnormal mealworms only a part of the wing- rudiment is external ; the remainder Hes as usual within Fig. 53. — A, Mealworm (Larva of Tenebrio molitor) , with abnormal external wing-rudiments iw) , partly dissected to expose food-canal (5, stomach; i, intestine; r, rectum); a, fat-body; t, testis. X 3. B, Transverse section through thorax. X 20. e, epidermis ; c, c' , old and new cuticle ; 5, stomach ; m, muscles ; b, wing-rudiment, part external and part in pouch {p) . After H. Singh-Pruthi (Proc. Camb. Phil. Soc. Biol, i, 1924) . the inpushed pouch of the skin (Fig. 53, B,^). This indicates that a portion only of the wing-bud was everted during the preparation for the preceding moult. During the last twenty years somewhat similar observations have been made on the larvae of other beetles and of certain moths ; of these the most noteworthy is the case of the ground-beetle Lehia GROWTH AND TRANSFORMATION 189 scapularis described by F. Silvestri (1905), who considers that in this species the final larva with external wing- rudiments is a normal prepupal stage in the Ufe-history. The outward appearance of part of a wing- rudiment on the thoracic segment of a beetle or other metamorphic insect may be most reasonably interpreted as a reversion towards the primitive condition found in young exopterygote insects, indicating that from such conditions the metamorphosis has been elaborated by the postponement of the outward appearance of the wing-buds until successively later stages of the life-history. This postponement is clearly correlated with the structural divergence between larva and imago to which reference has already been made. It has also been noted that such divergence is commonly associated with difference in feeding-habits. Comparison of larva and imago from this point of view furnishes an interesting and instructive study. Among beetles and the Alderfly group of the Neuroptera, the larva as well as the imago bites soUd food by means of typical insectan mandibles ; in the details of its feeding, however, the larva usually differs from the adult, devouring roots, for example, while the latter eats leaves, or during its life in pond or stream pursuing aquatic prey while the perfect insect attacks inhabitants of the land and air weaker than itself. In most families of Neuroptera as well as the carnivorous Water-beetles (Dyticidae) the perfect insect has normal biting mandibles, while in the larva the slender curved jaws are modified for piercing the bodies of insects which serve as prey and sucking their juices. The very remarkable divergence as regards feeding shown by the Lepidoptera, among which the caterpillar has strong biting mandibles, while butterflies and nearly all moths have vestigial mandibles and elongate flexible maxillae adapted for sucking nectar or other fluid, is familiar to all students of insect life. In connection with our contention as to the mutual divergence of imago and larva among metamorphic insects it is note- worthy that the Micropterygidae, the most primitive of all moths, have still, when adult, small functional mandibles, 190 THE BIOLOGY OF INSECTS while their maxillae retain the typical form with slight modification, the lacinia or blade, absent in Lepidoptera generally, being well developed. The contrasts in feeding habit between insects of the same kind during the immature and adult periods of their lives suggests the mention of the still greater contrast afforded by many insects which do not feed at all after com- pleting their transformations and acquiring their wings. In most insect life-histories the preparatory stages extend over a far longer time than the duration of adult life. Dragon- fly larvae often spend several years under water before emerging into the air in readiness for the winged insect's flight of a few weeks or months, while the underground life of the cicad already mentioned in this chapter is pro- longed for thirteen or seventeen years, the winged adult dying before the winter of the year in which it comes up. In most metamorphic insects with a yearly life-cycle the life of the imago is much shorter than that of the larva, the former to be reckoned usually in weeks and the latter in months. We have seen that the larval period of the life- history aflFords an opportunity for eating and digesting food and storing it up in the tissues, so that there may be ample supply for the extensive re-making of the creature at the pupal period. Apart from the exceptional precocious modes of reproduction to be considered later in this chapter, pairing and egg-laying are unknown until the insect has reached its adult condition, so that the imago may be regarded as essentially performing the function of repro- duction. It is not surprising, therefore, from this point of view to find that in a number of insects — ^whether exo- pterygote as the Mayflies, or endopterygote as the Silkworm Moths (Bombycidae and Saturniidae) and the botflies (Oestridae) — the imago when developed has the jaws so excessively reduced that it is incapable of taking food, and its activity is entirely concerned with breeding, the feeding necessary for the accomplishment of its life-purpose having already been performed during the larval stages. It has often been remarked that the power of flight, acquired by GROWTH AND TRANSFORMATION 191 the vast majority of insects when adult, has an important if indirect bearing on reproduction, as it facilitates a wide range over localities suitable for egg-laying, and thus tends to bring about an increase of the area occupied by the species. It has been noticed that in the growth of insects generally there is something of a crisis at the penultimate stage of the life-history, and this becomes especially evident in the development of those insects, the vast majority of the class, that undergo complete transformation with a resting pupal stage between the end of the larval and the beginning of the adult hfe. The nature and meaning of the pupa has always presented a fascinating problem to students of the biology of insects. The Greek philosopher Aristotle regarded the insect pupa as a second egg, and William Harvey (1666), taking a similar view, suggested that the amount of food- material in a butterfly's egg is insufficient for the building up of so highly organised a being as the parent, and so only the imperfect caterpillar can be hatched from it ; the cater- pillar after weeks of feeding stores up the necessary amount of food and then reverts to the condition of a second egg (the pupa), whence the butterfly in due time may be hatched. A superficial examination of the hard, egg-shaped puparium of a bluebottle or the brittle cocoon wherein rests the pupa of an " eggar " moth might be thought to afford countenance to such a view. But even in the obtect pupa of a butterfly, with its wings and appendages closely adherent to the body, many of the organs of the perfect insect can be clearly recognised, and much more is this the case in the " free " pupa of a beetle, lacewing, or wasp, in which the wings and limbs stand out from the body in much the same way as they do in the adult. The envelope of the actual pupa, therefore, is clearly the cuticle of the insect itself, even though, in the case of an obtect pupa, it is specially modified in correspondence with what is predominantly a passive stage in the life-history. Examples have already been given of exopterygote insects such as cicads, and scale-insects, in which the penultimate 192 THE BIOLOGY OF INSECTS instar is quiescent, and in the transformation of the last- named family we notice that the wing-rudiments of the male are formed beneath the preceding larval cuticle. These conditions suggest an approach on the part of certain Exopter^'gota towards the pupa of metamorphic insects. The Mayflies (Ephemeroptera) are of especial interest in this connection because they combine the open method of wing-growth with a very wide divergence between larva and imago. The larval mayfly (Fig. 52, A) might almost be described as a bristle-tail adapted for aquatic life, since in its long feelers and tail-cercopods and its crustacean man- dibles, it closely resembles a thysanure, while its paired abdominal tracheal gills must be compared, as R. Heymons (1896) and C. Borner (1909) have shown, with a bristle-tail's short abdominal limbs. There is a long aquatic larval life with very many moults, outward wing- rudiments becoming conspicuous in the later stages. The mayfly has little obvious likeness to its larva except in its elongate abdomen bearing terminal cercopds, for the feelers are very short and the mouth parts reduced to mere vestiges, so that the insect in its winged state cannot feed and its very name implies the rapid passing of its life. Its aerial existence nevertheless presents a feature of very great and exceptional interest. When the ripe nymph has come out of the water and shed its cuticle, the instar revealed, though possessing developed wings, is not the true adult, but a " sub-imago " which has to undergo another moult before the insect reaches the imaginal state and becomes capable of repro- duction. In no other group of insects does a moult occur after the power of flight has been acquired. The existence of the majrfly's sub-imago suggests the fascinating idea that a moult after the development of functional wings was possibly of general occurrence among the primitive winged insects of past ages. In connection with this view it is worthy of notice that in many of the less specialised meta- morphic insects of to-day — Coleoptera and Neuroptera, for example — the pupal wings are relatively of large size, and that in the transformation of several Hymenoptera, including GROWTH AND TRANSFORMATION 193 the familiar hive-bee, the casting of the last larval cuticle, whereby a '' prepupa " with conspicuous wing- rudiments Fig. 54. — Pupa and Puparium of Rhagoletis pomonella. X lo. A, early-stage pupa enclosed in puparium and shedding prepupal cuticle (lateral view) ; B,the same, ventral view ; C, later pupa within puparium and separated prepupal cuticle (lateral view) ; D, still later pupa removed from puparium (lateral view) ; E, pupa shortly before emergence (ventral view) . An, anus ; Anty antennal lobes ; ASp, anterior larval spiracle ; Atr, atrium (anterior part of larval pharynx) ; E, compound eye; Mthy larval mouth; A^i, pronotum ; ATj, mesonotum ; pm, pupa- rium; ppu, prepupal larva (fourth larval instar, inside of puparium) ; P'-6, proboscis ; PSp, posterior larval spiracle : Pt, ptilinum ; Pu, pupa : PuSp, pupal dorsal spiracle of pronotum; tra, tracheal linings of preceding instar; w, subocular lobe; W2, wing; W3, halter. X 10. From R. E. Snodgrass (jfourn. Agr, Res., U.S.D.A. xxviii, 1924) . O 194 THE BIOLOGY OF INSECTS is revealed, is followed by another moult ushering in what is regarded as the true pupal stage. R. E. Snodgrass (1924) has described in the small muscoid dipteron Rhagoletis, a ** prepupal " cuticle which is formed within the puparium and envelops the pupa (Fig. 54). This instar, however, resembles the contracted maggot in form and has no wing- rudiments. Further, in connection with the biology of the pupa, it is noteworthy that among the metamorphic insects there is a great range of variation in the creature's power of movement during this stage of its life. The house-fly pupa lies quiescent within its hard protective puparium — the shrunken and condensed larval cuticle — out of which the fly has to make its way after emerging from the cast pupal coat. The pupa of a butterfly or of a moth belonging to one of the highly organised families can move only a few of its abdominal segments. But among the more primitive Lepidoptera the pupa, provided with rows of locomotor spines on its abdominal segments, works its way partially out of its cocoon or from the earth in which it lay buried ; the empty pupa coat of the Goat Moth (Cossus) may be seen partly protruding from a tree wherein the caterpillar fed, that of a Swift Moth (Hepialus) from the surface of the soil in which the larva devoured roots. Among the more primitive Diptera the same tendency to pupal activity may be noticed in cases where the life-conditions render it appropriate ; the pupa of a Crane-fly (Tipula) raises the front half of its body out of the ground, and gnat pupae swim actively through the water making use of the surface film to obtain atmospheric air for breathing by means of paired '* trumpets *' on the thoracic region of their bodies. From the foregoing examples it may be realised that while insects practising the open method of wing-growth are as a rule active, and those practising the hidden method passive in the penultimate stage, there is no absolute devia- tion in this respect between the two great types of insect life-history. The pupa or its corresponding instar seems GROWTH AND TRANSFORMATION 195 to display just as much activity as may correspond to its manner of life or to the necessity of preparation for the final moult. The details of structure and habit are largely adaptive, and the course of the life-history of insects as a whole suggests a great degree of plasticity in correspondence with biological relations. The same conclusion as to a plasticity in the details of development and correspondence to environmental needs is suggested by many facts which confront the student as startling exceptions to the normal progress of insect trans- formation ; some examples of such may fitly close this chapter. Reference has already been made to the summer genera- tions of greenfly (Aphids) in which the eggs develop within the mother's body so that the young are not hatched but born. Many two- winged flies (Diptera) give birth to active maggots instead of laying eggs. The Sheep-fly {Oestrus ovis), for example, usually deposits tiny larvae in the nostrils of the sheep, though sometimes according to the weather conditions, as W. E. CoUinge (1906) has shown, she lays eggs, whence later the maggots are hatched. The big Flesh-fly {Sarcophaga carnaria) is constantly ** larvi- parous," and so are many of the Tachinid flies whose maggots feed as parasites in the bodies of other insects. The female of a species of Compsilura is provided with a sharp " larvipositor " by means of which she pierces the body-wall of a caterpillar and thus places her offspring safely inside, where they invade the wall of the stomach and begin to feed. More rarely does the larva undergo most of its growth within its mother's body ; the dreaded African Tsetse-flies (Glossina) as well as certain Dipterous insects (Hippoboscidae and Melophagidae) which suck blood from mammals and birds, bring forth mature larvae that pupate immediately after birth ; hence these last named insects are often called '* pupiparous." Some of these in correlation with their parasitic life are wingless, and there are two most remarkable types of wingless Diptera, living as '' guests " in the nests of termites in the African and 196 THE BIOLOGY OF INSECTS Eastern tropics — known as Termitoxenia and Termitomyia — from whose life-history the whole larval and pupal stages are omitted, for according to their discoverer, E. Wasmann (i90i),the former lays a relatively enormous egg whence a developed adult is hatched, while the latter gives birth to a single offspring already in the adult form. It is remarkable that such extreme abnormalities of life-history as these should occur among insects of that order (the Diptera) in which the ordinary course of transformation has become most elaborated with the most profound difference between adult and larva. While in these insects the preparatory stages are largely or wholly omitted from the life-cycle, there are other Diptera in which young may be produced by lar/ae or pupae, so that insects not adult have the power of reproduction ; these furnish examples of " paedogenesis " or precocious parenthood. More than half a century ago O. Grimm (1870) saw female pupae of the midge Chironomus lay eggs which gave rise to active larvae ; we have seen that a typical insect pupa is closely like an adult in essential features of form, so this exceptional occurrence might be regarded as a somewhat surprising instance of virgin reproduction. Five years earlier, however, N. Wagner (1865) had noticed that within a grub of certain gall-midges (Cecidomyidae) a number of smaller larvae might be seen, these ultimately making their way out to free life through the body-wall of their larval mother. It is now known, as stated by W. Kahle (1908), that these abnormal young are developed from eggs which break loose from the ovaries already present in the parent hrva, and float in the body-cavity, where they segment and form embryos, which develop into the larvae that burst out of the parent's body. After a succession of these larval families young are produced that complete their transformation into pupae and adults, so that the whole cycle is made up of a sexual generation alternating with a number of virgin larval generations such as is normal in several groups of animals, notably in certain parasitic worms. GROWTH AND TRANSFORMATION 197 Of all aberrant cases of insect development perhaps the most remarkable is the '' polyembryony '' of certain small Fig c5._Polyembryonic development of Platygaster vernalis a, 6, egg undergoing maturation divisions (m, m') ; />', first polar body, s spfrm-nucleus; ., d, conjugation of egg and sperm nuc ei o 5 to form zygote nucleus (n) ; polar-bodies {p) mcreasmg m f^l^\>J^l^ stage when the polar bodies have given rise to nu ntive ^dls (^a) the trophamnion, surrounding the eight embryonic cells (« derived from the zvgote-nucleus, X i350 ;/, 5^, stdl later; stages >,with embryo e) developing from each embryonic cell. / X 700^ ^.X^^^^; ^*'^' R. W. Leiby and C. C. Hill {Joum. Agr. Res., U.S.D.A. ig^S-) 198 THE BIOLOGY OF INSECTS parasitic Hymenoptera. In this strange mode of develop- ment, as observed in Encyrtus of the Chalcid family by P. Marchal (1904), the female lays, in the egg of a moth, her minute egg so that it becomes enclosed in the body of the growing embryo as this develops into the caterpillar. The Encyrtus egg undergoes a curious kind of development, the polar nuclei persisting and multiplying at one end, while the egg nucleus, which may be fertilised or not, segments in the hinder region of the egg-substance and forms blastomeres. Ultimately the polar cells give rise to a nutrient capsule which spreads around the embryonic cells ; these by a process akin to budding form a large number of embryos, in some cases over a hundred resulting from a single egg. Growth is slow during the winter while the host larva develops in the egg-shell, but after hatching, when the caterpillar begins to feed, the embryonic mass of the parasitic chalcid increases rapidly in size, and assumes the form of a sinuous thread extending through the caterpillar's fat-body, the nutrient membrane being now enclosed in a sheath derived from the host's tissues. At length the Encyrtus grubs become free in the caterpillar's body- cavity and finally eat their way out through its dried skin and cuticle to pupate and assume the adult form. Allied forms, which have been found to undergo a similar course of development, are described by F. Silvestri (1908) and R. W. Leiby (1922). The result is to bring about a hundredfold multiplication between the single egg laid in the egg of the moth by the tiny chalcid fly and the enormous family at the close of the completed transformation. R. W. Leiby and C. C. Hill (1923, 1924) have shown that in species of Platygaster (belonging to the Proctrotrupidae), parasitic on gall-midge (" Hessian Fly ") larvae, there may be the usual direct development of the egg into one larva, or a poly- embryonic development resulting in the production of six or eight parasitic grubs (Fig. 55). This condition, suitable to a host-larva of small size, suggests an early stage towards the abnormal fecundity of Encyrtus. In face of such facts as these, the student cannot but feel GROWTH AND TRANSFORMATION 199 convinced that the various types of growth exhibited by various members of the class of insects are the resuh of modification and specialisation in the course of a long racial history, and that the creatures show at all stages of their growth an adaptive plasticity which may respond to changing conditions in ways that are strange and new. CHAPTER VIII FAMILY LIFE In the two preceding chapters we have sought to follow the processes of reproduction and growth among insects ; now we turn to consider the behaviour of the creatures in con- nection wdth these processes. The insect in its final winged condition has as its essential function the perpetuation of its race, and the activities of an adult insect are, to a great degree, obviously concerned with breeding in its various aspects. Pairing of the sexes is a necessary preliminary to the fertilisation of eggs, and the prospective mother must place her eggs in situations suitable for their development if the young are in their turn to grow^ to maturity. Her egg-laying may be her only and sufficient contribution to the welfare of these young, but not a few female insects feed or otherwise tend their offspring after hatching. In some cases the members of a family remain in association for a shorter or a longer period of larval Hfe ; when the association is preserved after the adult condition has been attained, the family may be said to pass into a community and the life of such insects becomes definitely social. The pairing of the sexes may naturally be considered first among the various activities concerned with repro- duction and the rearing of the young, and of especial interest are certain aspects of behaviour preHminary to pairing, which may be regarded as comparable, at least in some degree, to the courtship practised by many back- boned animals. Insects have diverse ways of attracting members of their own kind but of the opposite sex. Some of these are clearly simple responses to sense stimulation, FAMILY LIFE 201 while others involve behaviour that suggests selection or choice. The recognition by an insect of a possible mate often depends upon the sense of sight. In a previous chapter (p. 90) evidence has been given that butterflies may be attracted v^hen they see a wing of one of their own kind lying on the ground so that they stoop towards it. In most insect families it is the male that seeks the female, as is the case among animals generally, and the distinctive colour-pattern of the wings in such insects as butterflies apparently serves as an attraction when it is recognised. Occasionally the female is attracted by the male ; this is the method of courtship in the Swift Moth Hepialus humuliy a species known as the '* Ghost," because the male's wings are of a sheeny white above while the female's are, like those of both sexes in related species, brownish in hue. In the dusk of the midsummer evenings the white male hovers above the damp pasture or marsh-land ; a female attracted by the white wings collides with him and the two then drop among the herbage and pair. It is of interest to notice that in the most northerly districts of its range, including Shet- land, where at midsummer it is never really dark at night, the male Hepialus humuli is of the same brownish aspect as the female. The obvious conclusion is that the conspicuous white colour of the common British form is a special adapta- tion to aid courtship and hasten pairing ; the male has become modified in correspondence with the special breeding habits of these insects. It is well known that in many butterflies of various families the male is adorned with bright colour while the female is comparatively plain ; several of our British " Blues " (Lycaenidae) and the '* Orange-Tip " {Euchloe cardamines) among the Pieridae afford examples of this. The characteristic blue colours of the male Polyotn- matus tear us, Argiades corydon, and A. bellarguSy respectively, may be regarded as facilitating recognition by their several mates ; but there is no convincing evidence that female butterflies or other insects choose their mates in a '' brilliance competition," as suggested by C. Darwin (1871) in his well- 202 THE BIOLOGY OF INSECTS known theory of sexual selection. The actual pairing of butterflies usually takes place while the insects are in the air, and during the nuptial flight one partner carries the other, whose wings remain closed in the usual resting position, the upper surfaces meeting over the back. Darwin pointed out that, as a rule, the male butterfly carries the female, except in those exceptional cases where the latter sex is the more brightly coloured ; then the female carries the male. This generalisation has been to a great extent confirmed by the extensive obser^^ations on tropical African insects made by G. D. Hale Carpenter (1920). A courting male butterfly often strokes the wings of a desired mate with his fore-feet ; if his attractions prove ineflFectual he flies away and leaves her, if she accepts his advances he carries her off. At least some of the females just mentioned as more brightly coloured than their mates, are more active than they in the courtship, so that examples are aflForded of a complete reversal of the parts commonly played by the two sexes in the drama of pairing. Dragon-flies, which vie with butterflies in the brilliance of their colours, comprise many species in which the male displays a brighter or more conspicuous appearance than his mate. Thus in our two common large British '' damsel- flies," Calopteryx virgo and C. splendens^ the wings of the female are uniformly russet or hyaline, while those of the males are respectively suflPused with deep metallic blue, or each traversed by a broad dusky or blue patch. It is likely that such conspicuous distinctions may serve as recognition- marks to the females. In some dragonflies definite acts of courtship have been observed. R. J. Tillyard (191 7) describes how in the small green Australian Hemiphlehia mirahilis^ the abnormally long white terminal ** inferior appendages " of the male *' are displayed as a sign to the female, by raising the abdomen and bending it slightly sideways while walking up the reed stem." The female answers this signal '' by moving the whitened end of her abdomen from side to side in a peculiar manner." After a dance-like flight together the couple mate with one FAMILY LIFE 203 another. The method of pairing in Dragon-flies differs most strikingly from that prevalent among insects generally ; the tail processes of the male on the tenth abdominal segment do not clasp, as usual, the hinder region of the female's abdomen, but her neck. The actual copulatory apparatus of a male dragon-fly, exceedingly complex in structure, is situated toward the front end of the abdomen, on the second and third segments. To a central vesicle in this region the sperm-masses are transferred by the male flexing his abdomen ventrally so as to bring the opening of the ejaculatory duct on the ninth abdominal segment into contact with the cavity of the vesicle. Then, in the actual process of pairing, after the male has seized the female by her neck and prothorax, she flexes her abdomen strongly forward so that the spermathecal opening on her eighth abdominal segment is brought against his genital armature. Dragon- flies, whose feelers are very small and poorly provided with sense-organs, appear to make little or no use of scent perceptions in their courtship and pairing. Among Lepidoptera, however, the sense of smell is often of great importance as a sex attraction, as has been mentioned in a previous chapter (p. 69), where reference was made to the " assembling " of male moths around a captive female ; moths thus attracted have usually complex feelers with sense-organs abundantly developed. The Swift Moths, whose recognition of mates through vision has just been described, have also the attraction of scent ; the male of Hepialiis humuli emits from the bases of the hind-legs an odour that has been compared to that of almonds, and this probably acts as an auxiliary to his conspicuous white wings for an allurement to the female. In the smaller H. hectus both sexes are alike in their wing-colour ; the male's hind- legs are strongly swollen and the skin glands within these limbs secrete a fragrant fluid whose vapour carries a scent like that of the pine-apple. In these insects, therefore, the male is provided with attractive appearance or perfume or both ; but there is no definite evidence of choice being 204 THE BIOLOGY OF INSECTS QOO qM Fig. 56. — Scent-apparatus of male Amauris ntavius. a, scales of general wing-area; b, scales of glandular patch, X 160. c, scales (out- line dotted) in relation to perforated chitinous projections, X 350. d, section through wing showing upper (s') and lower (s) scales and glands C?), X 450. e section of abdominal " brush " showing filaments arising from cells, and retractor muscle (w), X 100. /, cells of brush with bases of filaments, X 200. 5-, fragments of filaments, X 700. After H. Eltringham, Trans. Ent. Soc, 191 3. FAMILY LIFE 205 exercised by the female for one special male among a number of others. In many male butterflies of the Danaine group there are noticeable dull patches on the wings (either the fore or the hind pair) known as '' brands " ; these are clothed with scales, smaller than those clothing the general wing-area, and overlying little circular or ovoid *' scent-cups " each covered by a cuticular lid with a minute central pore ; beneath each of these is a multinucleate gland which secretes the odorous substance peculiar to the insect (Fig. 56, c, d). At the hinder end of the abdomen, in connection with the genital armature are paired '' brushes " formed of elongate scales usually white or pale in colour. Each brush is carried in an extensible membranous bag ; when this is everted by fluid pressure, the brushes appear as a conspicuous tuft at the male butterfly's tail-end. These remarkable structures on wings and abdomen have been well described by H. H. Freiling (1909) and by H. Eltringham (191 3). The " brush- bag " in Amatiris niaviuSy described by Eltringham, con- tains special groups of cells " which produce numerous delicate chitinous filaments, these having the property of breaking up transversely into innumerable tiny particles, thus forming a kind of dust " (Fig. 56, e,f). The butterfly, provided with this apparatus, brings the abdominal brushes into contact with the scent-brand on the wings, and then by everting them, scatters the perfume around, the '' dust " apparently helping to diffuse the scent. The details of these structures vary in diflFerent members of the family. '* Neither wing-glands nor dust-producing devices are invariably present ; the brush itself and not the wing may produce the scent material . . . whilst the dust may be produced by the wing and not by the brush, and in the pupal instead of in the imaginal state." The scent emitted by these organs may be certainly regarded as an attraction to the opposite sex, but according to an observation made by Hale Carpenter (1920) its effect is not always immediately successful. A male Amauris in Uganda was *' flying about after a female, which presently alighted on a dead flower- 2o6 THE BIOLOGY OF INSECTS spike. She . . . remained perfectly still while the male hovered a few inches above her head with a peculiar flutter causing him to rise and fall a little." The male displayed the " large, white brush-like structure . . . most energetic- ally protruded and as rapidly withdrawn." But at length " the female suddenly flew away as if the performance had not appealed to her and the male followed." The reader of this unfinished story may imagine, if he please, that the courtship was finally successful. Besides vision and scent, there is reason to believe that the females' power of hearing sounds produced by male insects of a few groups is an important factor in courtship. Reference has already been made (Chap. IV, pp. 80-82) to the stridulating organs on the legs and wings of male grass- hoppers and crickets which produce the familiar chirping song of those insects, and the ears in the first abdominal segment or near the front knee-joint with which they are provided. It was also mentioned that some female crickets from which the ears had been removed were no longer attracted by the chirping of the males. Some positive observations on the value of chirping and hearing in the courtship of several species of European grasshoppers are due to E. B. Poulton (1896). Some of the males appeared to chirp in rivalry, and even to fight with each other by means of kicking or biting. The power of stridulation ** seemed almost without exception to be exercised with direct reference to females, or in rivalry to other males in the presence of a female." In a species Pezotettix pedestris in which, the wings being underdeveloped in both sexes, stridulation is impossible, the male practises nothing that can be regarded as courtship, but jumps suddenly on a female and captures her as his mate. It is likely, even certain, that many insects produce sounds inaudible to us but appreciated by the auditory organs of their own species, and the perception of such excessively rapid vibrations may be of service in courtship. For example, the beautifully formed ear known as Johnson's organ in the base of the feeler of many male gnats and midges may enable these FAMILY LIFE 207 insects to hear the high-pitched hum of the females and to direct their flight toward them. Few insects display more remarkable habits in courtship than some predaceous two-winged flies of the family Empidae, which have been studied by M. Howlett (1907) and A. H. Hamm (1908-9). Empis borealts, a fairly large species with russet brown wings, is common about mid- summer in our hill-districts, and a number of females may often be seen in " dancing " flight over the water of a stream. A male with an insect such as a stonefly or a mayfly captured as prey and carried in his legs, approaches, and after flying up and down beneath one of the females secures her and flies to some convenient plant-shoot. When observed there, it is seen that the prey has been transferred to the female by whom it is sucked during the process of pairing. " The male," writes Howlett, '' usually hung by the front pair of legs to a twig, or blade of grass, supporting thus the whole weight of himself and partner ; the middle legs clasped the thorax of the female, while the hind pair of feet supported the prey in position beneath her proboscis, the apical part of the femora meanwhile firmly compressing her upturned abdomen. The hind legs of the female hung idle while with the two front pairs she manipulated her prey, kneading it as one who sucks an orange dry, and every now and then turning it about to insert her beak in a fresh spot." The males were never observed to suck the prey which they caught nor did the females appear to catch any insects for themselves. Hamm describes the methods of capture practised by Empis tessellata : " the male sits in wait upon a leaf or grass stem, darting upon any fly coming near enough. If successful he immediately proceeds to hang by the tarsal claw of one of the anterior legs to the edge of a leaf or twig, the other five legs being tightly clasped round the struggling victim. He then proceeds to feel with the tip of the proboscis over the thorax of the fly, finally reaching and immediately piercing the junction between head and thorax. The proboscis was withdrawn after a few seconds^ the victim being apparently paralysed 2o8 THE BIOLOGY OF INSECTS and only showing slight movements of the body or limbs. These observations seem to point to the conclusion that it is the central nervous system that is acted upon." Evidently these insects go through a complicated series of actions, the ** dancing " movements of the females incite the males to catch prey and offer it to their desired mates, and the sucking of the victim's juices may stimulate the reproductive functions of the females. The act of pairing is followed after a shorter or longer interval by that of egg-laying, in which we see the first, and in the case of many insects, the only manifestation of parental care. Some reference has already been made (pp. 75, 112) to the attraction of various female insects by chemical stimulus to substances suitable for the feeding of their young larvae or nymphs after hatching. A few further examples of the working of this function, extremely important for the life of the race, may be given. The dragonflies, already mentioned in this chapter, afford an interesting diversity in the manner of their egg- laying. These insects live during their prolonged pre- paratory stages submerged in the water of ponds and streams, and the majority of females of the order drop their eggs " while flying over the surface of the water, merely by striking the tip of the abdomen from time to time against the water, and so washing off the steady flow of exuding egg-masses " (Tilly ard, 191 7). These masses of eggs are surrounded by a gelatinous substance, which may dissolve in the stream, '' so that the eggs spread out on the river bed." In some cases, however, the effect of the water is to coagulate the gelatinous envelope which may then form a " rope," enclosing hundreds or thousands of eggs, twisted around the twigs of some aquatic plant. The females of the slender ** damsel-flies " (Zygoptera), however, as well as those of the large, elongate Aeschnines, have two pairs of the processes of the ovipositor strongly developed as cutting organs with saw- like edges. By means of these, incisions are made in the stems of reeds or other aquatic plants, and the eggs FAMILY LIFE 209 deposited therein singly or in small groups. The dragon- flies that provide such shelters produce egg-shells of the elongate form usual among insects, while the shells of eggs dropped into the water are rotund or shortly oval in shape. Some of the dragon-flies that cut incisions descend beneath the surface of the water in order to lay their eggs. It is remarkable that this provision of shelter in plant-tissue should be made, for the larvae when hatched are little beasts of prey, catching and feeding on weaker denizens of the water. In many dragon-flies the male continues his attend- ance on his mate throughout her egg-laying activities, so that both parents appear concerned in preparation for their offsprings' future. In previous chapters reference has been made to the laying of eggs by female insects of various orders within or alongside some substance — plant- tissue, animal body, refuse, or carrion — that will serve as food suitable for the grubs after hatcliing. In many such activities the prospective mother in her egg-laying is reacting to an appropriate stimulation through her sense-organs of vision or smell, and it is perhaps dangerous to assign any psychic element to her behaviour. Yet her action tends definitely towards the provision of food for her young. Quite a number of insects, however, go beyond this indirect pro- vision, and take trouble to collect food for their larvae before laying their eggs. The most striking examples of this practice are to be found among the Hymen optera ; the hunting and nest-provisioning habits of the digging-wasps have been mentioned in Chapter V (pp. 105-7), ^^^ some features in the activities of wasps, bees, and ants will be discussed in the next chapter on Social Life among Insects. For the present the food-providing habits of a common European dorbeetle {Geotrupes typhaeiis), as described in one of the famous memoirs of J. H. Fabre (1907), may serve as an example of behaviour which cannot but suggest parental care. A male and female of G. typhaeiis pair in the early spring and excavate a cylindrical tunnel running vertically down from the surface of the ground to a depth 210 THE BIOLOGY OF INSECTS of as much as five feet in some cases. The female does the digging with her strong fore-limbs, while the male hoists the displaced soil to the surface, a work in which the three sharp processes on his prothorax prove of much use by holding the fragments of earth. He then collects sheep- dung, which in the upper part of the tunnel he works into pellets with his thoracic spines and front legs, breaks into large fragments, and lets fall to the lower part of the tunnel where the female reduces these fragments to a fine state of division and arranges them in the form of a *' sausage " or ** long cylindrical loaf." The egg is laid a short distance below the food-mass, to reach which the grub after hatching " will have to demolish and pass through a ceiling of sand some millimetres thick." From such provision of food-supply by father and mother for their young, we may pass to actual care for the family after hatching as well as for the eggs. This is illus- trated by the habits of our Common Earwig {Forficula auricularia) and other members of the same lowly family. More than a century and a half ago C. De Geer (1773) observed the female of the Common Earwig brooding over her eggs, and M. T. Goe (1925) has lately stated that the eggs will not hatch unless this incubation has been practised. The incubation period lasts for about a fortnight, and after hatching, the young earwigs are often tended for some time by their apparently careful mother. A pleasing sight is pre- sented to the naturalist, lifting a partly sunken stone beside a hedgerow in winter or spring, by a female earwig with her eggs or her tiny pale youngsters, already strikingly like her in general aspect, but with the forceps-limbs relatively slender and weak. The larger shore-hunting earwig Anisolahis maritima has, according to C. B. Bennett (1904), similar breeding habits. The female, in preparation for egg-laying, hollows out beneath a log or stone, a " little chamber " an inch wide and half as deep, carrying away the excavated soil between her jaws. '' The chamber is made perfectly clean ; no sticks or bits of wood or pebbles are allowed by the more careful females to remain." The eggs FAMILY LIFE 211 themselves, after being laid are carefully rolled and cleaned in the mother's mouth ; then they are watched and guarded until hatched, as are the young also for at least a few days. The female Anisolabis does not, however, maintain for long her reputation as a good mother, for when the family '* had once left her to seek food for themselves they could not safely return lest she should endeavour to eat them." Similar but more prolonged care for offspring is shown by a common European and British Shield-bug Acanthosoma griseum^ whose habits were, like those of the earwig, observed in the eighteenth century by De Geer. His observations have been confirmed by several naturaHsts whose notes are conveniently sum- marised in the recent work of E. A. Butler (1923). E. Parfitt watched how the mother shield-bug " came to the rescue " of a young- ster touched by him with a twig, " putting her antennae down to the Httle thing and drawing them over it." J. Hellins saw a family of twenty newly-hatched young bugs beneath a birch-leaf covered, together with the empty egg-shells, by their mother's body. At a later stage of their development he writes that she " was now quite in a state of fuss ... if I attempted to touch her brood she fluttered her wings rapidly ... at night when the wind blew roughly, the mother contrived to get them under her, and sat covering them as at first." The point at which the story ends suggests the progress of many other families ; the young were seen *' just setting off on their travels," then " busy exploring," while " the mother ran from place to place feeling for them." There is also some evidence for maternal Fig. 57. — Seashore Bug (Aepophilus bonnairei) coasts of South Britain, Ireland and W. Europe. X 10. 212 THE BIOLOGY OF INSECTS care in the small bug Aepophilns bonnairei (Fig. 57) which lives on the sea-shore between tide-marks. A number of young beneath a stone, kept for observation by J. H. Keys, arranged themselves in a circle facing the mother in the centre. When the stone was lifted, ** the adult would almost instantly alarm the young with a rapid tap with each antenna alternately, and the whole troop would scamper round to the other s'de of the stone with great speed." Undoubtedly one of the most remarkable of all recorded cases of family life among ** non-social " insects is that of the European horned dung-beetles (Copris) whose habits are described in a well-known memoir of J. H. Fabre (1897). These insects do not roll balls of dung about as i their allies the '* sacred " beetles (Scarabaeus) do. For their own food-supply they excavate beneath lumps of excrement lying on the surface of the land, and '' here is engulfed without definite shape, an enormous supply of victuals, bearing eloquent witness to the insect's gluttony." In the breeding season, which comes in May or June, however, a pair of Copris work together, digging out a spacious ovoid chamber, within which sheep- dung, collected by the male, is comminuted and kneaded by the female into an egg-shaped mass. This is later subdivided into several pellets on each of which the female carefully forms a shallow basin- like cavity, lays an egg therein, and covers it by judiciously applied pressure. The male of Copris hispanns leaves all this work to his mate, but in C. lunatus the father remains underground and as a result of his assistance the pellets are twice as numerous as in the other species. In the former case the mother, in the latter both the parents, keep guard for several weeks over these pellets within which the grubs are developing. In due time the larvae pupate, and at length the young beetles are perfected and emerge ; they make their way to the surface of the ground accompanied by their parents, who thus have the privilege — very rare among insects outside the *' social " groups — of seeing the members of their family reach the adult state. It is this unusual condition which makes FAMILY LIFE 213 Copris of especial interest to the student of insect biology from the comparative point of view. It is believed that all through the weeks during which the young are developing, the parent keeps guard, fasting in the underground chamber — her behaviour contrasting markedly with the " gluttony " in which she indulged before the breeding season. Nearly related to the Scarabs and their allies are the Passalidae — a group of large flattened beetles, black or brown in colour, distributed through the tropics and warmer regions of both hemispheres. These also display a family life of quite remarkable interest, which has been elucidated by F. Ohaus (i 899-1 900) and W. M. Wheeler (1923). They live in galleries eaten out in decaying timber. The parents guard the eggs after laying, and prepare food for the larvae after hatching by breaking up the wood with their jaws and partly digesting it. This is necessary because the grub's jaws are too weak for direct attack on timber ; the larvae *' are therefore compelled to follow along after their tunnelling parents and pick up the prepared food." In the darkness of their burrows, the Passalid beetles and their grubs communicate with each other by audible signals, the beetles stridulate by rubbing toothed surfaces below the wings over similar surfaces on the abdomen, while the grubs scrape the strong denticles carried by their very short, unjointed " paw- like " hind legs over striated areas on the sides of the thorax. The care of the parent Passalids is said to be continued through the pupal stage of the offspring until they assume the condition of mature beetles. The order of the Hymenoptera is well known as exhibit- ing the most striking examples of parental care among insects. In a previous chapter (V) some account was given, in connection with a discussion on insect behaviour, of the egg-laying, nest-making, and provisioning habits of various digging wasps. Such creatures provide beforehand for the needs of their offspring, but the mother does not survive to see the hatching of the grubs and tend them as a fannily. Many female Hymenoptera, however, have this relation- ship with their young, and examples are afforded by the 214 THE BIOLOGY OF INSECTS comparatively lowly family of the Sawflies (Tenthredinidae) whose larvae are caterpillars (Fig. 76, b) feeding on leaves. The habits of Australian species of Perga have been described by R. H. Lewis (1836) and W. W. Froggatt (1891, 1918). The female Perga lewisii lays about eighty eggs in an incision cut between the two surfaces of a gum-tree leaf, and rests on the leaf until the eggs are hatched ; after this she follows the young caterpillars about as they feed, " sitting with outstretched legs over her brood, preserving them from the heat of the sun, and protecting them from the attacks of parasites and other enemies.'* When fully grown the larvae crawl down to the ground-level and spin cocoons for pupa- tion in the soil. In the later stages of larval growth, these caterpillars are no longer guarded by their mother, but they continue to feed and move in companies so that they may still be regarded as a family living to some extent at least a common life. Such gregarious habits, often resulting from the limited space available on the food-plant, are dis- played by many sav^y caterpillars, as well as by caterpillars of moths and butterflies (Lepidoptera). The local migra- tions of swarms of larvae of the Antler Moth (Chareas graminis) or the Vapourer {Orgyia antiqua) are impressive. Members of such communities move together, apparently guided by contact, their behaviour suggesting that they should be regarded less as a family than as a flock. The family association among untended larval insects seems most apparent in cases where the young creatures by their united labour spin a silken web over the twigs and leaves of their food-plants, and live together on this shelter, a kind of nest not provided by the parent but made by members of the family. The caterpillars of the Peacock Butterfly {Vanessa to) afford illustrations of this habit in their younger stages, while the caterpillars of the Lackey Moth (Clisiocampa neustria) and the Small Ermines (various species of Hypono- meuta) practise it throughout larval life. The silken cobweb-like habitations of numerous families of the last- named group are often so close together that the plants on which they feed, a hawthorn hedge, for example, appear FAMILY LIFE 215 covered by a continuous sheet of fine threads, and the families of caterpillars are merged in a great, if unorganised, society. Returning to the Hymenoptera that store in their nests provision for their grubs by burying or immuring paralysed or dead caterpillars and other prey, it is of interest to find traces of the development from this common habit to that of actual care for the grubs after hatching. In his most instruc- tive work on *' Social Life among Insects," W. M. Wheeler (1923) draws attention to the habits of certain African solitary wasps (Synagris) as described by E. Roubaud (191 6) and J. A. Bequaert (191 8). These wasps make mud- nests on such surfaces as the thatch of huts, and the female '' under normal conditions, when food is abundant, lays an egg in her mud cell, fills it in the course of a few days with small paralysed caterpillars, and then closes it." But when prey is scarce the mother-wasp guards the egg, and after it hatches, collects a few caterpillars at intervals so as to provide food for the grub during the greater part of its period of growth. When the larva is about three-quarters grown, the mother " immures it in its cell with the last supply of provisions." In such species therefore we see actual transition from the storing of food for grubs which the mother will never see to actual care and feeding of the young. This has apparently become the normal habit of one species, Synagris cornuta, as the female feeds her off- spring '' from day to day with pellets made up of a paste of ground-up caterpillars." The habit of bringing food to the larvae through their period of growth is well known in the Sphecoid digging-wasps of the genus Bembex. The habits of the North American B. spinolae have been vividly described by the Peckhams, who dwell on the female's " habit of feeding her young from day to day or rather from hour to hour as long as it remains in the larval state." These insects catch two-winged flies which are placed in the nest after having been killed by stinging. Wheeler remarks how the number or size of the victims is increased ** as they are needed by the growing and increasingly voracious larva." 2i6 THE BIOLOGY OF INSECTS Like the wasps the great majority of bees provide for their young by storing food — honey and pollen — in the nests wherein they lay their eggs, the nests being made in tunnels excavated in the soil, by species of Andrena, CoUetes, and Megachile for example, or in the twigs of plants as by the well-known Osmiae, or in dry wood as by the " carpenter '* bees (Xylocopa), or in remarkably firm structures of stone- fragments and cement as by the '* mason " bees (Chalico- doma). Among all these the nest- chamber is sealed up with the egg a- d the store of food, and the mother bee never sees her offspring. The habits of Osmia tridentata, as described by J. H. Fabre (1891), afford an example of what may be regarded as family relationship. The female of this species lays eggs in a series of chambers along a hollow bramble-stem, each chamber with a provision of food for the grub after hatching, and the grub when fully grown spins a cocoon and pupates. When a young bee emerges from the pupal coat, it bites its way through the cocoon, and then through the partition closing the chamber in which it has been reared. Should the young insect be in the last- formed chamber next the opening of the hollow twig, it comes out at once into the open and begins its active aerial life. If, however, it finds the way to liberty blocked by the cocoon of a younger sister it waits for her emergence or tries to press a way between her cocoon and the wall of the chamber ; it is stated that a young Osmia never injures other members of the family in attempts to escape from its nest. Although the last-laid egg is nearest to the outlet and the first-laid in the nest farthest from it, the young do not necessarily complete their development in the regular order that might be expected ; the older offspring may have to wait comparatively long for the emergence of their younger sisters, or these latter may complete their transformations more rapidly than those hatched from the earlier laid eggs and so get out of the way in good time. It is of interest to note that the perfect insects among the Hymenoptera commonly take food of the same kind as they provide for their larvae. Bees as well as bee-grubs FAMILY LIFE 217 feed on honey and pollen. Wheeler has pointed out how an ichneumon fly sometimes licks up the blood of insects that she has pierced with her ovipositor, and that this organ may thus be used for self-feeding as well as for reproduction. Somewhat similarly a digging- wasp, seizing a caterpillar between her mandibles, may bite the neck of the victim, and take a portion of liquid food from it for herself, before depositing it in her nest as a food-store for her offspring. This possibly close association of the feeding with the reproductive instincts is believed by Wheeler to have been of great importance in the development of true social life among insects from such ordinary family relationships as have been considered. Some of the wonderful and fasci- nating facts and problems of insect societies may now there- fore demand our attention. CHAPTER IX SOCIAL LIFE The social life of insects may be fairly regarded as the central subject in our presentation of the biological study of these fascinating creatures, since the habits of such societies as those of the bees and ants are commonly known in their main features, and have attracted during many centuries the admiring notice of mankind. A community of bees, ants, wasps, or termites is a family, of which the individual members are greatly multiplied and their associa- tion for common activity so highly organised that the individuality of the single insect becomes merged in the wider individuality of the complex, social organism. Such insect societies as these are by far the best known, but there are others in which the community consists not of one huge family, all the offspring of a single abnormally fertile mother, but of an assemblage of families living together in such a way as to promote mutual protection and provision. W. M. Wheeler, whose comprehensive treatise (1923) on the social insects has been already referred to, enumerates as many as twenty-four different groups of insects among which a common way of life has become more or less com- pletely adopted. It is, however, doubtful if all of these can be regarded as having developed so far beyond the simple family relation as to attain a truly social state. Wheeler himself designates fourteen of his twenty-four groups as '' incipiently social or subsocial." It hardly needs to be stated that very many insects, locusts, dragonflies, butterflies, midges, for example, which are occasionally or habitually gathered into flocks, cannot be reckoned even SOCIAL LIFE 219 among the " sub-social " groups ; the members of such a flock keep near their companions, but in their activities they show no such mutual co-operation as characterises a true insect community. Examples of insects which live in societies made up of an assemblage of many families are afforded by several groups of beetles. The family life of the Passalidae was described in the last chapter (p. 213), and the conditions of their existence — parents and offspring feeding in galleries excavated in timber — afford a starting-point for the more distinctly social habits of the *' ambrosia " beetles, which are akin to the destructive " bark-beetles " of our forests. Most of these Scolytidae (or Ipidae) eat bark or v/ood, both in their larval and adult stages, but the " ambrosia " beetles have developed a more elaborate method of feeding and a simple type of social life. The habits of various European and North American species of Gnathotrichus, Xyleborus and Platypus have been described by H. G. Hubbard (1897) and others whose accounts are summarised by Wheeler (1923). While most of the Scolytidae make galleries at the inner surface of the bark, the ambrosia beetles burrow deeply into the wood ; it is not wood, however, on which they feed, but a fungus — the " ambrosia " — specially cultivated on a '' bed " prepared from woody material which has passed undigested through the insects' food canals. Beetles of both sexes work together, but most of the parental care devolves on the mother, who excavates a series of circular pits along the tunnel, laying an egg in each and depositing fragments from the " bed " with some growing fungus as food for the grubs when hatched. Each grub as it grows increases the size of its '* cradle " (Fig. 58, G', /) by biting and swallowing wood which is not digested but, being ejected in pellets from the intestine, is removed by the mother and used for fungus-bed. " The mouth of each cradle is closed with a plug of the food fungus, and as fast as this is consumed it is renewed with fresh material." The females of Platypus lay eggs in groups of ten or twelve at intervals along the 220 THE BIOLOGY OF INSECTS galleries, and the grubs when hatched *' wander about in the passages (Fig. 58, P) and feed in company upon the ambrosia which grows here and there upon the walls." Probably each kind of ambrosia beetle " grows its own peculiar fungus in a pure culture," and each female starts a new culture for her offspring by carrying away from the cradle or passage in which she was herself reared a mass of Fig. 58, — Galleries of Ambrosia Beetles. Platypus (P) and Gnatho- trichus (G) in pine trunk. North America. The small circles along the course of the galleries, as seen in cross section of trunk, indicate position of the " cradles," which are shown in side-view at G' (e, egg; /, larvae in various stages ; p, pupa ; a, adults waiting for emergence). About natural size. After J. M. Swaine. (Canad. Dept. Agric. Ent. Bull. 7, 1914.) spores, either on her head, in her jaws, or in the front region of her stomach, whence they are regurgitated when she has reached a site suitable for the foundation of a new family. Most remarkable perhaps of all the social beetles are two forms of the family Cucujidae recently found by Wheeler living in the hollow leaf-stalks of young Tachigalia trees of SOCIAL LIFE 221 the Guiana forests. These are small, narrowly elongate insects with clubbed feelers and short legs, which, after entering the hollow stalks, live and feed along strands of especially nutritive tissue ; after a time their excrement accumulates in longitudinal ridges adjacent to the ** food- grooves." Following the beetles, large numbers of small mealy bugs (Pseudococcus) invade the hollow stalks and begin also to feed along the nutrient strands ; then the beetles and also their larvae go to these mealy-bugs for nourishment, stroking with their feelers the little white sucking insects and inciting them to discharge from their intestines the sweet honey- dew. When two or more beetles," writes Wheeler, '' or two or more larvae or a group of beetles and larvae happen to be engaged in stroking the same mealy-bug, they stand around it like so many pigs around a trough, and the larger or stronger individual keeps butting the others away with its head." From this account it seems that the " self- regarding " instincts are not wholly eliminated in the social life of these beetles of the Tachigalia trees. Wheeler, in his account of these insects, lays stress on the fact that they are found in the leaf- stalks only so long as the trees are young enough to form part of the forest undergrowth. '* The older trees . . . have all their petioles inhabited by viciously biting or stinging ants." These latter invade the leaf- stalks as the tree grows, driving out the beetles, but preserving the mealy-bugs and adapting these '' cattle " to their own use. It is well known that many societies of insects, and especially ants, harbour a miscellaneous assemblage of '* guests," some of which are clearly of service to their *' hosts." It is of much interest to trace in the succession of insect inhabitants of the leaf- stalks of these tropical American trees the varying relations between the plants, the mealy-bugs, and the two strikingly diverse types of communities, first beetles and then ants, which successively make use of other organisms for obtaining shelter and food. We may now pass on to consider the social life of those 222 THE BIOLOGY OF INSECTS well-known groups of Hymenoptera, the wasps, bees, and ants. As previously mentioned, the societies of these insects are in reality large, often immense, families, and the family life of some wasps and bees has been described in the preceding chapter. In a typical insect community belonging to one of these groups, the vast majority of the population are those modified females, known as " workers," usually infertile and often in some way specialised for carrying on the essential activities of the society. The fertile females in a nest, actual or potential mothers, are known as " queens." Among the greater number of genera of wasps and bees, such as those mentioned in the last chapter, the individual insects are all normal, fertile females and males. There are no workers, and these insects are commonly termed *' solitary " — a term that may be considered not altogether appropriate, for though the family does not grow into a great society, a number of females of the same kind often make their nests close together, as with Synagris among the wasps and Andrena among the bees ; but although scores or hundreds of such wasps and bees have their nests close together there is no co-operation between the diiferent mothers or their families. The hymenopterous society being a very large and more or less specialised family, we expect to find some indication of its mode of development from an ordinary family, and Wheeler suggests that the habits of certain wasps indicate stages of transition between the " solitary " and " social " way of life. F. X. Williams (19 19) has shown how some species of the eastern tropical Stenogaster are solitary while others approach the social habit, as they construct nests of several chambers, sometimes enclosed in an envelope, the mother feeding a number of larvae at the same time, sealing up the chambers when they are fully fed, and sharing the nest with her daughters when these have attained the adult condition. None of the Stenogaster females, however, appear to be infertile or to be in any way specially modified as workers ; the adult inhabitants of the nest are relatively few in number, and it is doubtful how far the daughter- SOCIAL LIFE 223 wasps help the mother in feeding the younger grubs. The societies of the African Belonogaster (Fig. 59) are larger, but again all the daughters of a family community are fertile, and parties of them, at intervals, leave the parent nest together to found fresh societies, each of which consequently possesses as many potential mothers as there are foundresses. Fig, 59. — Nest of Wasp (Belonogaster juttceus) , West Africa. Above chambers with eggs, lower with larvae, at bottom closed cocoons. Adult wasps, males and females, the latter attending and feeding the larvae, both obtaining the larval salivary secretion. Drawn from a photograph, E. Roubaud (Ann, Set. Nat. Zool. (10) i. 1916). In allied South American wasps occurs a differentiation between the normal fertile females or queens and the smaller females (workers), with reduced or vestigial ovaries which if capable of producing eggs, lay only unfertilised ones which may all be expected (see p. 140) to develop into males. These American insects '' regularly form new 224 THE BIOLOGY OF INSECTS colonies and nests by sending off swarms of workers with one or two dozen queens." The plurality of mothers C polygynous " condition) results in their nests becoming crowded with " hundreds or even thousands of individuals." Most of the tropical social wasps are similarly polygynous, whereas the widespread Vespa (represented by seven species in Great Britain) has only a single queen-foundress for each nest. After surviving the winter she starts a new family and habitation in the spring. Of course, a proportion of the young female insects reared in one of our native wasps' nests develop into queens, but the vast majority are workers. It is of great interest to find that females of an intermediate type may occur, smaller than a queen, larger than a worker, and with ovaries reduced yet functional. O. H. Latter (1904) points out that these " fertile workers " are developed as the effect of an especially rich food-supply being available for the grubs in certain seasons. The workers of Vespa feed the grubs on fragments of captured insects, bitten up, malaxated by moistening with saliva, and moulded into small pellets. The wasp larvae in the chambers of the comb thrust out their heads, as Wheeler remarks, " like so many nesthng birds, and when very hungry may actually scratch on the walls of the cells to attract the attention of their nurses." It is now, however, well established that the feeding activities in a wasps' nest are by no means one-sided. P. Marchal (1896), C. Janet (1903), E. Roubaud (1916), and other observers have found that the parent or nurse-wasps obtain from the larvae which they tend sweet saliva often in large quantity, and Wheeler (1923) supports the opinion that this drain on the larval food supply is a potent physiological factor in preventing the development of the young insect's reproductive system, so that it becomes a sterile worker instead of a queen. Wasps in a nest engaged in tendance of the grubs stimulate the mouths of these by contact, or even by seizure of their heads between their own mandibles in order to incite the secretion and flow of the desired fluid. In some cases at least it has been estimated " that there is a flagrant disproportion between the quantity of nourishment PLATE VIII 4 9^^^^^ k -Jt^ 'j^^^^^H Xlst of Tree-wasp [Vespa noi-vigica). One-tenth size. Nest of Ground-wasp {V. gei-manica), dug out by a Badger. One-tenth size. To face p. 224.] \h. Britten, p/wto. SOCIAL LIFE 225 distributed among the larvae by the females and that of the salivary liquid which they receive in return." This com- parison is made by Roubaud, who does not hesitate to accuse the nurses of " actual exploitation of the larvae.'* The males or " drones " in the nest though they bring no food to the grubs take from them the sweet fluid (see Fig. 59). Roubaud and Wheeler believe that the evolution of the family society is to a great extent a result of the reciprocal feeding ('* trophallaxis ") ; there " naturally follows a tendency to increase the number of larvae to be reared simultaneously in order at the same time to satisfy the urgency of oviposition and to profit by the greater abundance of the secretion of the larvae." The bees, which must now be considered, have, it is calculated, only about 500 out of their 10,000 species living in true social communities whereof specialised worker- females form the great majority of members. The wasps are predominantly predaceous, though many of them feed on vegetable substances occasionally or habitually. The bees, however, are as a group dependent on the products of flowers — on the nectar which after digestion in the insects' stomachs becomes honey, and on pollen ; the broad shin and basal foot-segment and the feathered hairs of bees, among the most striking of their structural features, are connected with the habit of gathering pollen from the floral anthers, while the tip of the labium is elongated to form a beautifully efficient organ for licking up fluid from the nectaries of blossoms. " SoUtary " bees, among which no workers occur, provision their nests, as we have seen, with honey and pollen, usually mixed to form " bee-bread," partitioning the nest into chambers each of which contains at first an egg, later a grub with its appropriate store of food. In certain species of Halictus, which nest in burrows in the ground, the mother closes up the chambers, but guards the nest afterwards and survives till the development of the young is complete ; there is, however, not time enough for the realisation of social life between her and her offspring. Wheeler quotes some very interesting recent Q 226 THE BIOLOGY OF INSECTS observations made by H. Brauns on South African species of Allodape which nest, like our native Osmia, in hollow plant stems, but make no partitions between the successively laid eggs so that the nest is not divided into chambers. In some species a ** food-packet " is provided for each grub at the time of egg-laying. In others the mother arranges the eggs in such a way that the grubs, when hatched, direct their heads towards the entrance of the nest, and she brings to them at intervals lumps of " bee-bread " on which they feed in common. A further stage of development in be- haviour is attained by species of Allodape whose females feed their grubs individually, and produce an *' affiliation of the offspring with the mother to form a co-operative family." No workers, however, are produced among these incipiently social bees, and it must be admitted that the three truly social groups — the bumble-bees, the stingless bees, and the Hive-bee with its wild allies — stand markedly distinct from all the rest of the family. A feature of struc- ture in relation to life conditions that characterises them is the possession of abdominal glands that secrete the well- known bees-wax used by the insects for building the chambers of their combs. The stoutly built hairy bumble-bees (Bombus and allies) are well known to every country rambler, and the habits of our British species have been excellently described by F. W. L. Sladen (19 12). The societies of these insects in temperate climates are in most respects parallel to those of the social wasps, in that a queen, reared and paired in the previous summer, survives the winter in some sheltered spot and starts a new family in the spring. The young queen chooses in autumn a bank with a northern aspect on which to seek a burrow for hibernation ; on a south- facing slope she might be awakened by the midday sun too early in the year and perish in a sudden frost. For the site of her nest (Plate IX) she seeks some convenient under- ground cavity, often the deserted burrow of a field-mouse, and begins her building work by moulding a lump of moistened pollen on which she erects a waxen wall enclosing PLATE IX Nest of Bumble-bee {Bo/iibus musconmi). To face p. 226.] [/• '^^ Dixon, photo. SOCIAL LIFE 227 a cup-shaped hollow where the first eggs may be laid ; she constructs also a small round-mouthed, waxen honey-pot. She broods over the eggs to incubate them and covers them with a thin layer of wax. The young grubs '' devour the pollen which forms their bed, and also fresh pollen which is added to the lump by the queen," who also feeds her offspring with a mixture of pollen and honey which is squirted into the larval chamber through a small hole bitten by the queen in the waxen covering. While the larvae remain small they are fed collectively, but " when they grow large each one receives a separate ingestion." The chamber within which they are developing increases in size, and when, less than a fortnight after hatching, they are fully grow^n, each spins round itself a thin, tough, paper- like cocoon and pupates. All the early bees of the family are small infertile females or workers, which when they become active, take on the work of nest-making and grub-tending, while the queen confines her attention to laying eggs. As with the wasp communities, the population of a bumble-bee nest grows through the summer, though it rarely exceeds a final total of a few hundreds. The un- developed condition of the workers may be explained as due to poverty of feeding ; later in the season, when the total number of larvae in a nest becomes diminished, young queens are reared, as well as males, the latter arising (see p. 140) from unfertilised eggs. As with the wasps again, neither workers nor males survive the winter ; only the young queens hibernate so as to renew the race in the suc- ceeding spring. In tropical regions, however, the bumble- bee community survives through a number of years, and over-population is relieved by means of a succession of swarms. The Stingless Bees (Meliponinae) are a small group confined to tropical countries and most abundant by far in South America. They are of small size and the stings of the females are so far reduced as to be useless as weapons. They are structurally more specialised than bumble-bees, as the queens have narrow shins and proximal foot-segments 228 THE BIOLOGY OF INSECTS — a degenerative modification — while the workers except for their sterility represent " the typical female of the species." Wax ** is produced by the males as well as by the workers — the one case in which a male Hymenopteran seems to perform a useful social function," as Wheeler remarks. In the communities of stingless bees the activities of the nest are carried on by the workers, the queen's functions being from the beginning confined to repro- duction ; hence for the foundation of a new community a swarm comprising a queen with some workers is essential. These insects construct nests usually with layers of waxen comb consisting of hexagonal chambers with the openings on top. Their habits are curiously primitive, for the workers, according to Wheeler's account, '* put a quantity of honey and pollen into each cell, and after the queen has laid an egg in it, provide it with a waxen cover, so that the larva is reared exactly Hke that of a solitary bee." But they have the habit of storing food in special receptacles ; H. W. Bates (1863) described the nest of a Brazilian Melipona which contained *' about two quarts of pleasantly-tasted liquid honey." The Hive-bee group (Apis) are the most highly specialised of the family, and the exceedingly ancient domestication of the common species by mankind has rendered it one of the most familiar of all insects. Some aspects of the association of bees with our own race will be discussed later in this volume (Chap. XIV) ; for the present it is sufficient to suggest that the conditions under which domestic bees are reared, in straw skep or wooden hive, are modified from the manner of life of their wild ancestors. The few wild species of Apis all inhabit south- eastern Asia. Apis dorsata and A. florea build typical waxen combs in a single layer with two series of chambers arranged back to back and opening horizontally, suspended openly from the branches of trees. A. indica (hardly separable as a species from the domestic A. mellifica) makes a number of similar combs hanging side by side in hollow trees, and this is the habit of communities of A. mellifica SOCIAL LIFE 229 which, escaping from the tutelage of mankind, have become feral, reverting to an independent mode of life. Sometimes, however, feral nests of the hive-bee are found hanging without protective covering from the branches of trees. All these facts combine to indicate that the genus Apis originated in the tropics of the Eastern Hemisphere, and that it has become adapted to life in cooler regions by the habit of nesting in shelters either discovered or provided. In their mode of life, shown among the truly wild as well as in the domesticated species, the hive-bee group is distinguished by a marked *' division of labour " between workers and queen and by the progressive feeding and tendance of the larvae throughout Ufe. Details of the economy of hive-bees have been frequently described, and no more need here be attempted beyond the indication of a few leading facts of especial biological interest. These insects store honey in some of the chambers of their comb, and thus ensure a constant and convenient supply for the growing grubs as well as for the adult bees in the nest. The habit renders possible also the persistence of the com- munity from year to year in those northern countries with cold winters where hive-bees have been for centuries domesticated. New communities are established by swarms, consisting of a queen with a crowd of workers, which leave the old habitation. Swarming in the domestic bee-com- munities follows the emergence of a daughter- queen — an event controlled by the workers. The daughter remains in the hive, and the swarm is " led off " by the mother-bee. The worker bees take over all the activities necessary for the rearing of the young : comb-building, food-gathering, and distribution, and they also control by collective action the reproductive function of the queen, who may be regarded as " a mere egg-laying machine entirely dependent on her worker-progeny." Her hind legs are destitute of those pollen-gathering adaptations so elaborately developed in the workers, though among the bumble-bees these are common to queens and workers. As an egg-laying machine, how- ever, the queen-bee is highly efficient, as she may produce 230 THE BIOLOGY OF INSECTS 1,500,000 eggs during her life. As mentioned previously (Chap. VI), female Ilymenoptera are developed normally from fertilised and males from unfertilised eggs, and it is well established that the growth of a bee-grub hatched from a fertilised egg either into a queen or into a worker depends on the nature of its food. Larvae destined to become queens are reared in large ovoid *' queen-cells " situated at the edge of the comb, and fed throughout their growth with " royal jelly," believed to be a secretion of the worker's pharyngeal glands. Worker-grubs, on the other hand, receive this food only for the first four days of their development after hatching ; during the final two days of larval life they are fed by the workers on honey and pollen. The food of the queen-larvae is relatively very rich in fat, that of the worker- larvae in sugar. That the kind of female developed from a fertilised egg is determined by feeding has been often proved experimentally by transferring very young grubs from one kind of chamber to the other. The male bees are developed from unfertilised eggs which the queen lays in '' drone- cells " of hexagonal shape like the worker-cells but of larger size ; when laying these the queen-bee releases no spermatozoa from her sperm reservoir, as she does when laying in cells provided for the rearing of female bees. This is doubtless the general mode of procedure, but reasons have already been given (p. 140) for allowing the possibility of the exceptional origin of male bees from fertilised and of females from unfertilised eggs. Besides the '' brood-comb " in which the grubs are reared many series of chambers form a '' honeycomb " for the storage of food. It is from this store that the grubs are supplied, and it has long been known that the first period of a worker-bee's life after emergence from the pupal coat, is spent within the hive, and that expeditions for the gather- ing of nectar and pollen are not undertaken until she has attained a certain age. It has been recently stated by G. Roesch and K. von Frisch (1925) that there is a routine of duties through which all the workers of a hive succes- sively pass. The first work of a newly emerged bee is to SOCIAL LIFE 231 clean out chambers of the comb ready for egg-laying ; her next to remain in chambers where eggs have been laid, apparently for the purpose of assisting incubation. During the early period she is fed by her sisters, but when about three days old she begins to collect pollen and honey from the store, some of which she uses herself, but passes on most to the older grubs. When about a week old, her own digestive system becomes fully active and she spends the next week in feeding the young grubs that require her secretions for their proper nourishment. Then she begins to take up a certain amount of outdoor work, trying her powers in exploring flights, receiving nectar and pollen from the older foraging bees and removing dead comrades and refuse from the hive. The foraging bees on returning to the nest " dance " on the honeycomb with a circular turning motion, and during this performance the younger workers surround them with the apparent object of appreciating and remembering the scent of the flowers which the foragers have been visiting, so that these younger members when they in their turn leave the hive on foraging flights are guided to the same sources of food supply in special kinds of flower. The rapid, changing activities of a worker-bee at the height of the honey-season soon wear out the insect's constitution, and her life may last no longer than three weeks. Worker-bees, unlike the wasps, derive no food- substance from the grubs that they tend. The habit of storing a large reserve of honey and pollen in the nest ensures an abundant supply for adult insects as well as grubs, so that there is no exploitation of the latter by their nurses. The life of the worker-bee is devoted to the service of the society whereof she is a member, and her varied activities, briefly sketched in these pages, which follow one another in regular order as responses to successive stimula- tions, indicate to how great an extent her behaviour is regulated by inherited instinct. Our wonder at the matter is increased as we recollect that neither the remarkable structures on which her activities depend nor the instincts themselves are characteristic of the parents, queen and 232 THE BIOLOGY OF INSECTS drone, through which the inherited characters are trans- mitted. Most insect communities harbour a number of other insects which Hve during a part or the whole of their lives as '* messmates," ** guests," or parasites. The large popu- lation of a wasps' nest or a bee-hive may contribute to render such association easy and profitable to the invaders if not to the ** hosts." It is possible only to refer to some of these invaders that are definitely related to the social life of the host-insects. The large two- winged (Syrphid) flies of the genus Volucella are conspicuous British insects, whose curiously formed maggots, provided with elongate processes on the body segments, live, in the case of Vohicella inanis in the nests of social wasps, and in the case of V. bombylans in those of bumble-bees. The former fly has a yellow-banded abdomen like a wasp, and the latter is hairy like a bumble-bee. Both enter freely into the nests of their respective hosts in order to lay their eggs, and the maggots, after hatching, live on the combs, where they appear to act as scavengers by devouring refuse, thus performing a definite service to the community. There is no evidence in support of statements that have been made to the effect that the Volucella maggots devour the grubs of the wasps and bees ; the position in the nest w^here the syrphid larvae live indicates that they feed on refuse. It is possible that the bee or wasp-like aspect of the Volucella may serve as a protection against an attack by the workers of the com- munity. Sladen has noticed that the female Volucella bombylans continues to lay eggs after receiving fatal injury, and has suggested that this pow^r of partial survival may enable an invading fly, detected and stung by a bumble- bee, to ensure, notwithstanding her own speedy death, the development of her offspring. It is well known that a very large proportion of the nests of solitary wasps and bees are invaded by females of other wasps and bees for the purpose of egg-laying, and the grubs, as soon as hatched, begin to feed on the store provided by the rightful owner for her own offspring, which may conse- SOCIAL LIFE 233 quently die of starvation. Such insects, thus taking ad- vantage of the labours of others, are known as inquilines or '* cuckoo-parasites." In some cases the inquiline larva devours the grub of the host, behaving like a beast of prey, having by its mother's action been insinuated into the habitation of its victims. The origin of these types of parasitism among the sting-bearing Hymenoptera (Aculeata) has been suggestively discussed by W. M. Wheeler (191 9), who gives good reason for believing that the parasitic habit arose in all cases among members of the same species, whereof in times of '* scarcity of prey or food . . . individuals . . . might find it as easy as advantageous to steal the pro- visions of other individuals." Wheeler suggests further that the urgency of the egg-laying reflex would reinforce the stimulus due to scarcity in tending to establish the parasitic habit. He insists that many of the inquiline Hymenoptera now distinguishable specifically and often generically from their hosts, are nevertheless closely related to the latter, and the habit as developed among social bees and wasps affords strong evidence in support of this view. In communities of bumble-bees it has been shown by F. W. L. Sladen (1912) that occasional or incipient para- sitism is fairly common. One queen may enter the nest of another of the same kind, kill her and install herself in the vacant place. Among our British species are two nearly related, Bomhus terrestris and B. liicorum, the former of which not infrequently preys on the latter, " killing the lucorum queen and getting the lucorum workers to rear her [own] young." This habit is, of course, abnormal and occasional, but it might easily become the starting-point for a definitely parasitic race. Its further development is illustrated by the peculiar inquiline bumble-bees of the genus Psithyrus, which have no worker-caste and whose queens are destitute of the characteristic pollen-gathering structures on the hind-legs. An over- wintered female of Psithyrus enters a young nest of Bombus in the spring after the first set of the workers have been developed. She seeks " to ingratiate herself with the inhabitants, and in this 234 THE BIOLOGY OF INSECTS Fig. 6oa. — Queen-Wasp (Vespa amtriaca), X3. she succeeds so well that the workers soon cease to show any hostility towards her." After a time she kills the Bombus queen, thereby cutting off any further production of workers, but those already in the nest are enough for the tendance of the intruder's eggs to which and to the resulting larvae the workers " soon get reconciled " so that " they feed and tend the Psithyrus brood with as much devotion as if it were their own species." Sladen insists that the re- semblance of a Psithyrus to a Bombus " is not merely superficial but extends to nearly all the important details of structure, so that it is impossible to avoid the conclusion that Psithyrus has sprung from Bombus, and this at quite a recent period in the history of Hfe." Here we find the inquiline specialised in such a way as to bring the invading queen into definite social contact with the host-workers, so that these — not through any motive that can be truly called '* devotion," but as a result of their normal, inherited reflex tendencies — feed her young as though they were their own sisters. Among the social wasps (Vespa) there are several species SOCIAL LIFE 235 vm. Fig. 60B. — Vespa rufa and V. austriaca. i, Face of typical male V. rufa\ I., face of typical male V. austriaca ; 2-5, series of male rufa faces approaching austriaca-. II.-V., series of male austriaca faces approaching rufa, X 4. 6, Male armature of V. rufa (ventral view) ; VI., of V. austriaca; {st, stipes; sa, sagittae ; a, internal, b, terminal process of stipes), X 8. 7, internal stipital process of male V. rufa\ VII., of V. austriaca^ X 16. 8, Terminal stipital process of typical V, rufa ; VI 1 1., of F. austriaca ; 9 and 10 of rufa approaching austriaca ; IX. and X., of aw^fnaca approaching rw/a, X 28. After H. G. Cuth- bert {Irish Nat. vi. 1897) and Carpenter and Pack-Beresford {lb. xii, 1903)- of which no workers are certainly known, such as V. austriaca (Figs. 60A and 6ob) in Europe including the British Islands, and V. arctica in boreal North America. It has been well known, since the observations of J. W. Robson (1898), that queens and males of V. austriaca are reared in nests of the common V. rufa, to which it is generally believed to stand in the relation of inquiline to host. The American V. arctica is 236 THE BIOLOGY OF INSECTS reared, according to Wheeler and others, by workers of V. diabolica. The systematic relationship between these apparently " cuckoo " wasps and their hosts is exceedingly close, much closer than that between Pysithrus and Bombus. G. H. Carpenter and D. R. Pack-Beresford (1903) showed from a study of the male armature and the variation of coloration and structural features in V. austriaca and V. riifa that these tsvo species are much more nearly related mutually than either is to any other British wasp, and that each of the two forms varies towards the other as regards these characters (Fig. 60B, i-io and I -X.). From a census of the population of a nest in which the old queen was an austriaca and the latest emerged members workers and males of rufa, these observers concluded that the two might reasonably be regarded as alternative forms of one variable species, since an austriaca queen was apparently the mother of individuals which were clearly referable to rufa. Yet the occasional inquiline habit of one race of Bombus on another, mentioned already in this chapter, lends support to the accepted view that Vespa austriaca frequently plays the part of a cuckoo-parasite on V. rufa. Certainly all the facts combine to confirm Wheeler's generalisation that inquilines among the aculeate Hymenoptera are always nearly related to and genetically derived from their hosts. Turning from wasps and bees to ants, we come to con- sider the niost remarkable of all social insects — in some respects indeed the most remarkable of social animals, for the ant-communit}^ may be organised and specialised, within the limits imposed by the structural and psychic conditions of its members, to a degree that challenges comparison with the human commonwealth, so that the activities of ants have been held up by moralists as an example to encourage industrious effort among men. Modern students of the ways of ants have the great advantage of being able to consult the comprehensive treatises of W. M. Wheeler (1910, see also 1923 and 1926) and A. Forel (1921), while the habits of our British species have been well described by H. K. Donisthorpe (1927). SOCIAL LIFE 237 Among ants all the species are social, and members of the worker caste are sharply differentiated by complete winglessness ; while they show the highly developed structural features of their order, with modifications fitting them for the specialised activities of their lives, wings are never developed in the course of their transformation. Hence the worker-ants are more strongly differentiated from their fertile sisters than worker wasps and bees are ; and in many groups there may be recognised, among the wingless infertile females, two or more distinct forms divergent from each other in size and often in structure, while among the fertile females (queens) as well as the males, separate castes may in some cases be distinguished. The winglessness of worker-ants is matched by the females of many non-social Hymenoptera, species of gallflies and members of groups with parasitic larvae, for example. Among the ScoHoids, believed by Wheeler and others to represent the ancestral stock of true ants, there are two small famiHes — the Thynnidae and Mutillidae — whose females are wingless. It seems, therefore, that this condition may have arisen independently in a number of groups among the Hymenoptera, and that it has become a fixed character among all the ant workers. It is suggestive to remember that in the workers of certain species of ant, vestiges of wings may occasionally be detected. Quite a number of abnormal forms of queens tending to resemble workers and also forms of worker more or less resembling queens have been described and provided with special names for which reference may be made to Wheeler's book (1910). Some of these worker- like ('* ergatoid ") queens are wingless like the workers. It is well known that all ant- queens after the nuptial flight shed their wings ; only by the persistent bases of these can the queen in a nest be recognised as an insect once able to rise in the air. The flight-muscles of the wingless queen, now no longer required, undergo a rapid degenerative process described in detail by C. Janet (1907). These muscles, " the most voluminous of all the organs of the body, experience a precocious senescence," 238 THE BIOLOGY OF INSECTS suffer interruption of their innervation, and undergo a solvent action by ferments present in the blood. Among ants generally a new community arises, as among social wasps and bumble-bees, by a young queen starting a new nest, and herself doing all necessary work, rearing the grubs and feeding them with her own spittle, as she does not leave the nest until the first brood of workers are developed ; then these take on the labours of the society. In some cases, however, daughter- queens may return, after the nuptial flight, to their native nest which thus comes to harbour a number of fertile females all helping to increase the size of the family. Again such a many-mothered (polygynic) society may give oif colonies consisting of young queens " each accompanied by a detachment of workers." The variety of method shown by ants in pro- ducing new family societies is of great interest, suggesting that their behaviour is more plastic and less stereotyped than that of most other social insects. The wingless condition of the worker-ants is correlated with their prevalent habit of nesting in the soil ; the vast majority of members of this family may be reckoned among those insects which have abandoned aerial life for an exist- ence mainly terrestrial. With this mode of life are associated many adaptations of structure well known to observers at all familiar with the aspect of worker-ants. Their eyes are usually small, while their elongate feelers well provided with tactile and olfactory sense-organs are in constant use. Their long slender legs are well suited for the rapid running which goes far to compensate for the loss of flight. Wheeler dwells on the *' very long and intimate contact with the soil " which " has made the ants singularly plastic in their nesting habits," while A. Espinas (1877) pointed out that the loss of the power of flight among worker-ants is not without compensation in the increased intelligence partly at least attributable to their terrestrial life. " On the earth . . . there is not a movement that is not a contact and does not yield precise information, not a journey that fails to leave some reminiscence." SOCIAL LIFE 239 Among many ants of the same species the workers of a community differ among themselves so that two or more definite types or castes are recognisable. These are usually adapted for performing special functions in the social life so that the communities practise a " division of labour." For example, there may be workers with abnormally large heads armed with powerful mandibles ; such in some societies are the so-called '' soldiers," '' policemen or defenders of the colony," while in certain harvesting or hunting ants their work is " to crush seed or the hard parts of insects so that the softer parts may be exposed and eaten by the smaller individuals." Among the well-known " leaf- cutting ants " of the American tropics (Atta and allied genera), whose object in collecting and dividing the leaves was first detected by T. Belt (1874) as the provision of a soil on which to raise in the underground chambers of their nests the fungi on which they feed, there are in many species large-headed workers which carry out the spectacular raiding expeditions in the forests and bear away the leaves into the nests, where smaller workers with heads of normal form act as ** mushroom- gardeners " in the deeply situated cavities where the food of the community is actually grown. A striking feature noticeable in most ant communities is the close correspondence between the modifications of the workers and the nature of the food. Among the " legionary " or " driver " ants — which include the African and Oriental Dorylus (Fig. 61) and Anomma, and the tropical American Eciton and allies — the communities range over the country often in huge swarms attacking insects, spiders and sometimes even large vertebrates which furnish their food. The workers are blind, guided on their forays through their senses of touch and smell, varying greatly in size, the larger castes with great broad heads, armed with long trenchant mandibles. The picturesque account of T. S. Savage (1847) ^^^ been confirmed by later observers. He described " an arch for the protection of the [smaller] workers constructed of the bodies of their largest class," whose '* widely extended jaws, long slender wings and 240 THE BIOLOGY OF INSECTS projecting antennae intertwining form a sort of network." He watched these driver ants hang together so as to form " festoons or lines of the size of a man's thumb," reaching from the lower branches of trees to the undergrowth. Across such living bridges other ants can pass to and fro, up or down. Savage saw ** one of these festoons in the act of formation . . . ant after ant coming down from above, extending their long limbs and opening wide their jaws gradually lengthening out the living chain until it touched the broad leaf of a Canna below." As the festoon of ants Fig. 61. — South African "Driver" Ants, a, Dory lus fulviis, male, X 3 ; 6, Z). fimbriatiis, female (queen after casting wings), X 3 ; c, D.fulvus, worker, X 2. After G. Arnold (Amu S. Afr. Mus. xiv, 1916). swung in the wind, the lowest ant tried to lay hold of the leaf with feet or jaws without success, whereupon a large worker climbed up on the leaf from below and " fixing hind- legs with the apex of the abdomen firmly to the leaf," reached upwards with her front legs and opening wide her jaws, seized the lowest comrade on the chain, '' and thus completed the most curious ladder in the world." While the large workers or " soldiers " attack, seize, and bite up the prey, the small members v^dth short mandibles carry the grubs and pupae when on migration. The nests of these driver ants are to a considerable extent temporary. Wheeler SOCIAL LIFE 241 quotes observations on the East African Anomma molestunty a community of which '' occupies the same nest until it has destroyed all the available prey in the locality," an operation v^hich may occupy " some eight or ten days," then '* the colony migrates to a new nest." The fertile females of most Doryline ants are blind and wingless, resembling workers in their structure much more closely than queens. They are, however, when compared with the workers, of relatively enormous size as are also the heavily built males furnished v^th wings and eyes (Fig. 61). The feeding habits of these " drivers " seem crude and primitive when compared with the elaborate leaf-cutting and fungus-growing performances of the species of Atta previously mentioned. Wheeler believes that the hunting, pastoral, and agricultural modes of life have succeeded one another among ants as they are commonly believed to have done among men. Those ants whose staple food is honey or " honey- dew " illustrate the pastoral stage of society, because their sweet nourishment is obtained largely from the intestines of aphids, scales, and other insects which suck sap from plants. They are often tended and protected by the ants whose behaviour in connection with the '' guests " of their societies will be discussed later in this chapter. Many ants, however, go directly to plants and obtain supplies of sweet sap " from small glands or nectaries situated on their leaves or stems, where it is eagerly sought and imbibed " (Wheeler, 1923), or lick up the honey-dew which has been voided by aphids and spread over foliage and shoots Worker-ants that collect honey or honey- dew have the habit, on returning to the nest, of regurgitating from the crop a portion of what they have swallowed, allowing drops of it to be licked up by those workers which remain in the nest and act as nurses to the larvae, to whom a share of the liquid food is passed on. Ants are incited to disgorge this liquid food when touched or stroked by the feelers of their com- rades. Such workers tend therefore to act as temporary reservoirs of food material, and in many groups with this habit (Camponottis, Lasiiis, and Prenolepis, for example) the R 242 THE BIOLOGY OF INSECTS insects' abdomens become capable of considerable dis- tension. This condition is brought to extreme development by the oft-described Mexican honey-ants (Myrmecocystus)^ in the underground chambers of whose nests special workers with immensely swollen abdomens may be found hanging from the roof by their feet, back downwards. These bloated creatures (" repletes ") incapable of movement, are fed by the foraging ants in order that they may serve as " honey-pots " for the community as a whole. The abdomen of a replete assumes approximately a globular form, the pale intersegmental cuticle becoming greatly expanded and the normal dark abdominal sclerites appearing on its area as narrow transverse bars. H. McCook (1882), who first carefully investigated the nests of these honey- ants, believed that the repletes are workers definitely adapted for their strange function by the structure and character of their abdominal cuticle and the wall of their food-canal. Wheeler, however, considers that there is no inherited difference between the foraging and the honey-pot workers ; the latter are modified if they assume the part of reservoirs when sufficiently young, while their tissues are still plastic, but *' thoroughly hardened workers of the ordinary form . . . are no longer able to become repletes." The most specialised form of behaviour among ants that feed on vegetable substances is probably exhibited by the fungus-gardeners (Atta) already mentioned in this chapter (p. 239). The harvesting ants, however, whose activities are celebrated in the writings of Hebrew sages and Greek and Roman poets, display purposive habits hardly less surprising. The ancient accounts of these insects were indeed regarded with no little doubt until the careful observations of T. C. Jerdon (1854) on Indian species of Pheidole and SolenopsiSy and of T. J. Moggridge (1873) on the South European Messor harharus and allied forms, convinced naturalists that the workers of these ants do indeed collect the seeds of plants suitable for food and store them in the underground chambers of their nests. Germination of the harvested grain is prevented by the ants SOCIAL LIFE 243 biting off the radicle, and bringing the seeds in damp weather out of the nest to dry them in the sun. When, as sometimes happens through neglect of these precautions, some of the stored seeds begin to sprout they are carried out of the nest and placed on the surrounding soil to form what may be termed a refuse heap. From such rejected seeds plants may grow up, and the presence of these close to the ants' nests has led some students to the mistaken inference that the insects have deliberately sown the seeds there in the anticipa- tion of a harvest ! Much recent information on these fascinating creatures will be found in Wheeler's great book (1910). All worker- ants feed the larvae of their nests, but Wheeler has recently (191 8, 1923) laid stress on the fact that among ants, as among wasps, the feeding is reciprocal. Some ant-grubs, like the wasp-larvae, supply a salivary juice from the mouth to appease the workers that attend on them, but the common habit of the ant-larva is to " sweat a fatty secretion through the general integument of the body." The licking of grubs by the female ants (whether queens or workers) is not therefore to be interpreted as a sign of affection or solicitude but as a method of obtaining attractive liquid food. In ant communities the practice of mutual feeding (trophallaxis) is thus almost universal, not only among the developed adults, but between these and the helpless grubs which they tend. The larvae of ants are, like those of bees and wasps, legless grubs, but while in the latter groups the larval cuticle is smooth and bare, that of the ant-grubs is usually covered with hairs, which may be simple, forked, hooked, branched, or sawlike and, in some cases, borne on distinct tubercles. The hairiness of ant larvae is definitely suited to their manner of life, as by means of these cuticular out- growths, the grubs are kept from direct contact with the damp walls of earthen nest galleries, while they are anchored to the walls or to the under surface of covering stones. The hairs of many neighbouring grubs may interlock so as to * hold the young larvae together in packets," and enable 244 THE BIOLOGY OF INSECTS the workers *' to transport large numbers with little effort," as Wheeler remarks. It is likely also that the hairy covering of an ant-grub protects its body from injury when seized by the jaws of a worker. Most ant-grubs are fed by the workers with disgorged liquid food, and in such the man- dibles are feeble. In some groups, however, the grubs are fed on insects and are provided with fairly strong jaws. Fig. 62. — a. Full-grown larva of Ant, Tetroponera tessmatini, West Africa. Side view, b, young larva of Pachysima aethiops, West Africa. Side view (e, exudatorium) . c, full-grown larva of P. aethiops, ventral view. Magnified. After W. M. Wheeler {Proc. Amer. Phil. Soc. Ivii, 1918). Wheeler has described (191 8) how in certain American Ponerine ants the larvae are fed lying on their backs, the workers depositing bitten-up insects on the ventral surface of the body of each grub, which then pours out a copious salivary secretion ; this serves to digest the grub's own meal and to provide a nutritious draught for its nurse. As already mentioned, most ant-larvae exude from the body- surface a sweet fluid which the workers lick up. Wheeler SOCIAL LIFE 245 describes how in the larvae of some African species of Pachysima (Fig. 62) there are on the ventral region of the thoracic and first abdominal segments thin-walled finger- like outgrowths in which fat- cells lie near the base while the distal portion of the " exudatorium " contains the clear fluid which, forced out through the body wall, can be conveniently imbibed by the worker as she feeds the grub, the grub's head being so situated as to be surrounded by the curious outgrowths ; these become relatively smaller in the later stages of larval life (Fig. 62, b, c) when the food consists of pellets of insect fragments, the grubs in earlier stages being fed on disgorged fluid. While the general behaviour of the members of ant communities is mutually helpful, there are occasions when individual self-assertion becomes evident. In times of scarcity workers, especially the larger castes among poly- morphic ants, may devour their comrades. Workers, short of food, may eat grubs and pupae instead of tending and protecting them. Between ants of different communities or of different kinds there are various highly interesting possibilities of relation. The driver or legionary ants (p. 240 above) are as ready in their raids to prey upon other ants as on cockroaches, grasshoppers, spiders, or vertebrates. " Certain small but aggressive ants,'* writes Wheeler, " which secure at least a portion of their sustenance by way- laying the foraging workers of another species and snatching away their food, deserve the name of brigands. Such ants naturally make their nests near those of the species they plunder." Some predaceous species raid the nests of more pacific ants, kill adults and carry off larvae and pupae to serve as food. From such habits as this has arisen, in the opinion of Wheeler and others, the oft- described " slave- making " of Formica sanguinea which ranges over all the north temperate regions of the globe, its societies harbouring a number of workers of the allied F. fiisca. From estab- lished communities the sanguinea workers go forth to raid nests of fusca whence they bring back grubs and pupae, some of which are not killed, but preserved to grow into 246 THE BIOLOGY OF INSECTS workers that share in the labours of the captors. According to Wheeler's observ^ation a young scmguinea queen is in- capable of establishing a nest of her own ; she therefore enters a fusca nest, and seizes a number of worker-pupae, killing any fusca workers that seek to interfere with her. The gang of workers she has annexed begin as soon as they have emerged to feed her and tend the grubs hatched from her eggs. The sanguinea workers reared from these have the inherited instinct to raid nests of fusca and capture larvae and pupae ; thus a mixed community is formed, the workers of the two species sharing in the common labour. Communities of Formica fusca suffer also from raids by '' Amazon Ants " (Polyergus), oppressors more formidable than F. sanguinea. There are several species of Polyergus, most of them, like the well-known P. rufescens, bright red in colour, provided with sharp, slender curved mandibles " perfectly adapted for fighting but of no use for digging in the earth or capturing food." The main facts about the behaviour of these slave-making ants were described more than a century ago by P. Huber (1810), who bestowed on them the suggestive title of '' Amazon." Extensive observa- tions on the European forms were subsequently made by Forel (1874), while C. Emery (191 1) has given an account of the foundation of a community by the young queen Polyergus. She invades a weak nest of Formica fusca, kills its queen by biting into her head, and then is adopted by the fusca workers, which tend and feed the grubs hatched from the amazon's eggs. The Polyergus workers which develop from these have neither the structure nor the instincts to enable them to do the work of the nest or to procure food. Their part is to make raids on other fusca nests, where they kill the adults so far as may be necessary for their purpose of carrying off the fusca larvae and cocoons to their own nest, the " slave " population of which is thus kept up to the necessar}- level. The Amazons are, as has frequently been remarked, '' absolutely dependent on their slaves " for the maintenance of the community and the survival of the race. Wheeler comments on their " two SOCIAL LIFE 247 contrasting sets of instincts. While in the home nest they sit about in stolid idleness or pass the long hours begging their slaves for food or cleaning themselves and burnishing their ruddy armour ; but when outside the nest on one of their predatory expeditions they display a dazzling courage and capacity for concerted action compared with which the raids of sanguinea resemble the clumsy efforts of a lot of untrained militia." The Amazons make their raids always in the afternoon hours and Forel actually observed forty- four raids on thirty afternoons during a period of seven weeks after midsummer. He estimated the number of amazon workers at 1000, and of the pupae captured by them during the summer at 40,000. But only a small proportion of these develop into slaves ; many are killed and eaten by the Amazons while others are accidentally and fatally injured in transport. Contrasted with these slave-making instincts are the ways of certain ants which Wheeler defines as temporary or permanent " social parasites." Communities of the former group arise through the adoption of a young queen in a nest of the host ants. F. Santschi (1920) describes how a newly hatched female of the North African Bothriomyrmex de- capitans is '' arrested " by workers of Tapinoma 7iigerrimum when she approaches their nest, and dragged inside. If any denizen threatens attack, she gets among the host- larvae or on the back of the host- queen ; in such situations her own characteristic odour is masked by that of the native insects. The invader while on the back of the Tapinoma queen may employ herself in biting off the latter's head (hence the specific name, decapitans). The Bothriomyrmex grubs are tended and fed by the Tapinoma workers ; these ultimately die off, and as there is no host- queen left, the mixed community is succeeded by a society composed entirely of Bothriomyrmex. This cannot be started without the help of Tapinoma, but when established it can carry on independently, so that the parasitism is temporary. Among the permanent social parasites there is no worker caste. The European Afiergates atratulus inhabits nests of 248 THE BIOLOGY OF INSECTS Tetramorium cespitum ; its habits are described by Forel (1874), J^i^et (1897), Wheeler (1910), and others. The males are wingless, and pairing must therefore take place in the nest of Tetramorium, whence the young winged females emerge in summer, and after casting their wings, enter other Tetramorium nests, where they are received by the workers, which ultimately kill their own queen and devote themselves to attendance on the Anergates grubs. The Anergates queen displays, as her eggs develop, a greatly swollen abdomen, on which the sclerites become widely separated by tracts of pale flexible cuticle, as in a '* replete " worker honey-ant or a queen-termite. The Tetramorium community that harbours Anergates must ultimately die out since the workers have assassinated their mother. The conditions of the permanent social parasitism among ants are most remarkable both as regards the degeneration of the parasite, and the apparently unnatural behaviour of the host- workers. Wheeler agrees with Emery in considering the mode of life of the temporarily parasitic ants to have been derived from the slave-raiding habit, which seems itself to have arisen as a specialisation of predaceous feeding. The degeneration of habit noticeable in the " slavers " is emphasised among the temporary parasites, while the per- manent parasites have no longer a worker caste. Wheeler calls attention to the extreme rarity of species of the last group ; they are " so very scarce that they must be on the very verge of extinction — a fact which shows that parasitism, so far as race is concerned, is anything but a promising or profitable business." It is noteworthy that nearly all the slavers and social parasites among the ants are closely related to their hosts, as the parasitic wasps and bees are, so that for the ants also that practise such habits we may infer a common origin with the creatures which they oppress or exploit. Ants, more than all other insects, furnish examples, numerous and varied, of association with insects of other orders and creatures of other animal classes which inhabit their nests and share in their social life as guests of the com- SOCIAL LIFE 249 munity. Such " myrmecophiles " have been extensively studied in recent years, and for details as to their relations to ants the writings of E. Wasmann (1920) may be advantage- ously consulted. Many of the ant guests contribute nutrient material on which the ants feed. Of these the aphids (*' greenfly ") and some allied Hemiptera are the best known. As previously mentioned in this chapter (p. 241), the aphids, sucking sap in great quantity from plants, void from their intestines drops of the surplus food substance, of which they can use but a small proportion for their own sustenance. The fluid evacuated by the aphids is therefore not excreted waste-matter but digested fluid food in which much of the sugar has undergone inversion. Aphids therefore furnish an extensive and con- venient source of food supply to the large proportion of ants that live principally or entirely on '' honey-dew." To obtain the liquid ants may follow the " greenfly " as they feed on plant-shoots, or, in the case of root-sucking aphids, harbour them in their nests. Ants in attendance on aphids may be observed stroking them with feelers or fore-legs, and the aphids in response exude drops of honey- dew which the ants swallow. Besides aphids, scale insects and mealy bugs (Coccidae) and sucking insects (Homoptera) of other allied families are harboured by ants for the sake of the honey-dew voided from their intestines. There are many records of the care taken by the various kinds of worker- ants of young, new- born aphids, root-sucking kinds being carried by the ants to fresh rootlets especially soft and succulent. While such underground aphids are herded by their ant guardians within specially constructed earthen " pens," some aphids and coccids that feed on the shoots of plants are gathered into droves by the ants, which build over them covers of silky or papery substance beneath which they are protected and sheltered. Wheeler points out that certain features of structure and behaviour observ^able in these ants that feed on honey- dew and in the sucking insects that supply their food, confirm the opinion that the relation between the 250 THE BIOLOGY OF INSECTS two kinds of insect is mutually beneficial. Ants that use aphids in this way, as " cattle," never devour or attack them ; they rather protect them and drive off threatening enemies or carry away the aphids to some safe refuge. The aphids, on their side, never try " to escape from the ants . . . but accept the presence of these attendants as a matter of course." A. Mordwilko (1907) states that in some aphids habitually associated with ants, there is a ring of stiff hairs surrounding the vent ; these hairs hold the drop of honey- dew until the feeding ant has swallowed it, so that they appear to be related rather to the latter than to the aphid itself. Among the most remarkable of all ant-guests are various kinds of beetles that spend their whole lives in ants' nests. They belong to several distinct families, such as Staphylinidae or rove- beetles and their allies the Psela- phidae (Fig. 63), as well as the curious Paussidae. Wasmann has described and discussed at length the relations between the ants and these guest-beetles which are fre- quently reddish in colour, their bodies adorned with tufts of yellow hairs surrounding the openings of glands which secrete an aromatic volatile fluid licked The beetles are themselves fed by the liquid ; the guests solicit this by Fig. 63. — Pselaphid Beetle (Claviger testaceus) a * ' guest " of the British and European Yellow Ant (Lasius flavus). X 12. up by the ants, ants on disgorged stroking with their feelers or fore-legs the heads of their hosts. Their jaws are often modified for the reception of this liquid food, so different from the solid nutriment devoured by the vast majority of beetles. Not the adult beetles only but their grubs also live in the ants' nests, where they are tended and fed by the workers, though they have often been observed to attack and devour the ant larvae, It has not unnaturally been suggested that the ants SOCIAL LIFE 251 which hart)our these guests " care more " for the beetle grubs than for their own. But this mode of expression attributes to the ants motives for behaviour which do not necessarily follow from the observed actions, for it is very doubtful how far the worker- ants, whose responses are mostly made to tactile and olfactory stimulations, dis- tinguish between the various inmates in the nest with which they come into touch. A creature, be it sister-ant or guest-beetle, which gives the tap on the head to which the normal response is regurgitation of food, is fed as a matter of course by any worker in the nest. This conclusion is supported by the relation between several species of Lasius and certain tiny mites (Anten- nophorus) which are carried about and fed by the worker- ants although, unlike most of the guest-beetles, they furnish the ants with no food substance in recompense. Janet (1897) has described how a worker may be observed to carry along the galleries of an underground nest of Lasius, three of these mites, one with its back downwards clinging to the ant's neck with its three hinder pairs of feet, and the other two holding on one on either side of her abdomen. The long front legs of the Antennophonis are used to tap the head of an ant so as to obtain in response a drop of honey ; evidently the mite carried beneath the ant's head can obtain the boon readily and directly from its bearer, while those mites which cling to an ant's abdomen depend for their supplies of food on other ants, which they touch in the course of their journeys through the nest galleries. It might be imagined that the ants would not carry about and feed these useless guests unless some feeling of satisfaction were to result from the act, comparable, for example, to the gratification many human creatures seem to derive from carrying about and feeding useless small dogs and kittens. But these mite-harbouring ants have really no goodwill towards their tiny guests, for when a mite first attaches itself the ant- carrier tries to shake it off, and the act of feeding in response to the tap of the little creature's foot, is a simple and inevitable reflex, 252 THE BIOLOGY OF INSECTS Another method of obtaining food is practised on Lasius mixtns by a small bristle-tail (Atelura) found in numbers in the ants* nests. Janet (i 896) describes how when one worker- ant is disgorging honey to feed a comrade, the Atelura thrusts itself between the two, *' raises its head, snaps up the droplet, and makes off at once as if to escape merited pursuit." This action might be naturally described as thieving ; the little bristle-tails lurk in the nest where they find shelter and take any opportunity of seizing food. A more specialised method of exploiting ants is practised by the maggots of a Texan fly (Metopina) described by Wheeler as *' messmates " in the nest of a species of Pachycondyla, which feeds its larvae with fragments of insects, these being placed by the workers on the concave ventral surface of the grub within reach of its jaws. Each maggot of Metopina coils itself around the neck-region of the ant-grub, and whenever the latter receives its allowance of insect frag- ments, the fly-larva ''' uncoils its body and partakes of the feast." Both " host " and '' guest " become full fed about the same time, and the latter, enclosed in the former's cocoon, retires for pupation to the tail end of the ant-grub, which completes its transformation before the fxy-maggot, and on emergence from its cocoon leaves an opening through which the fly when subsequently developed can make its way out. Many more examples might be given of creatures of other kinds that share the life of the ant communities. The nature of the association varies immensely. On the one hand, we notice the mutually beneficial activities of ants with aphids or with various caterpillars belonging to blue (Lycaenid) butterflies ; these produce from a dorsal gland opening on the sixth abdominal segment, a sweet fluid acceptable to the ants which follow^ the caterpillars about on plants or harbour them in their nests. On the other hand, there are mere thieves like the bristle-tails, mites, and Metopina maggots, or " insect jackals " like certain rove-beetles that devour dead, decrepid and feeble ants, or attack and prey on solitary active ones. Wheeler well SOCIAL LIFE 253 describes them as a ** perplexing assemblage of assassins, scavengers, satellites, guests, commensals, and parasites." Wasps, bees, and ants, among which are included the vast majority of social insects, belong to the Hymenoptera, one of the most highly specialised of all the orders of the insect class. It is, however, of great interest to find that an elaborate social life, depending on the growth of an enormous family, is characteristic of a lowly organised group — the Termites, which though often called *' white ants," have no near relationship to true ants, but belong to a comparatively primitive order, the Isoptera. Among the Embiidae, a tropical and sub-tropical family closely akin to the termites, there is to be noticed an incipiently social habit analogous to that characterising some of the '* solitary " wasps and bees. Many male embiids are winged ; the wingless females tend their eggs and young much as the mother-earwig (p. 210) cares for her brood. A. D. Imms (191 3) describes how in the Himalayan Embia major, " when the young larvae are hatched, they remain around the parent female, who conceals them, so far as she is able by means of her body." These insects spin, from glands situated in the fore-feet, an abundance of silken thread, and thus construct extensive webs and galleries in which their families carry on a primitive community life. All the adults among the Embiidae are normal fertile males or females ; there is no worker caste. In a termite society, on the other hand, the immense majority of members are infertile, wingless *' workers " (Fig. 64, d) with small heads and jaws, together with a smaller though considerable number of larger-headed, long-jawed '* soldiers " (Fig. 64, c). While in the bee or ant community all the infertile insects are females, the worker and soldier castes among the termites may belong to either sex. The differences between the various members of a termite society were formerly believed to be induced by differences in feeding, but the researches of E. Bugnion (1913) and Caroline B. Thompson (1916) have demon- strated that, at least in some cases, the caste characters of Fig. 64. — Forms of Termites, North America, a, Lai-va of Leuco- termesflaripes, X 8, after C.L. Marlatt (U.S.D. A.Ent. Bull. 4) ; b, male of Amttertnes tubiformans {left v/ings cast) , X 10; t, soldier, and d, worker, of A. arizonensis, X 10 ; e, "queen" of Leucotermes flaripes, X 16; /, second form, and g, third form female of L. tibialis, X 8. After N. Banks and T. E. Snyder (Bull 10 U.S. Nat. Mus. 1920). SOCIAL LIFE 255 termites are definitely inherited, as among the newly hatched young two sets of individuals may be detected, some with smaller brain, eyes, and reproductive organs destined to become workers or soldiers, while others with those struc- tures normal, develop into fertile insects. It is therefore not unlikely that the caste of any termite is determined by the nuclear constitution of the egg whence it arises. Workers and also soldiers of the same species may differ in size, and many termite communities have three distinct forms of fertile males and females. The '' kings " (Fig. 64, b) and ** queens " are winged insects with firm dark cuticle ; swarms of them when mature leave their native nests and, after flying for some distance, come to the ground and shed their wings. Many of such swarms are devoured by birds and other creatures ; the survivors associate in couples, a male and female, excavating, by their common labour, the rudiment of a new nest in form of a small underground chamber where they pair and start the foundation of a community. In members of the second fertile caste (Fig. 64, /) the wings remain undeveloped though recog- nisable in a rudimentary condition, while the cuticle of the body is feeble and pale. The third fertile caste (Fig. 64, g) is characterised by a ver}.' pale cuticle and the total absence of wings, features which recall the condition of the workers. It is doubtful if these second and third castes of fertile termites ever leave their native nests ; they have been termed " substitution royalties," under the impression that they are kept as " understudies " for the " royalties " in case of disaster to the latter. Termites are among those more primitive insects in which there is no marked transformation in the course of growth, such as is so conspicuous in all the social Hymenoptera, with their pale legless grubs. The newly hatched termite displays all the essential features of its parent, and the adult worker, wingless, soft-coated and pale, with its reproductive system undeveloped, may be regarded as retaining to a great extent the characters of the young. The same view may be fairly taken of the third- form fertile (" neoteinic ") termite in which no traces of 256 THE BIOLOGY OF INSECTS wings appear. The soldier termites also remain wingless, and are in that respect undeveloped and youthful in their character ; but their heads are highly modified, large with firm, brown capsule, either bearing extremely prominent trenchant mandibles or prolonged into a snout-like process with a repellant gland opening at its tip. They are the defenders of the termite society. It is believed that from eggs of the third-form females, fertile insects like themselves as well as workers and soldiers can be developed. The second-form fertile termites may be parents of members of their own and of the three " lower " castes. Only the kings and queens can give rise to all the varied forms of their kind. When exceptionally a worker or a soldier becomes fertile, it can reproduce its own caste only. The fertile female termites, in whose bodies numerous eggs are developing, tend to become swollen in the abdominal region, tracts of pale cuticle showing between the darker sclerites, as the integument is stretched. In most " queens " (Fig. 64, e) this process is carried to an extreme degree, the swollen abdomen becoming seven or eight times as long as the rest of the body, with its area almost entirely composed of tense whitish cuticle. Thus, while a pair of termites starting a new nest are of the same size, it comes to pass that in some species the " physogastric " queen is four times as long and a hundred times as bulky as her mate, and perhaps fifteen times as long and three thousand times as bulky as her worker offspring. K. Escherich, in his excellent account (1909) of the termites, reckons that a queen of the tropical African Termes hellicosus lays about thirty thousand eggs a day, a rate of reproduction which would work out to ten million eggs a year, and to a hundred million eggs in the average ten-year life of one of these insects. He concludes, therefore, that such termite queens must be regarded as the most fruitful females in the whole animal kingdom. The staple food of termites is wood, and the damage which these insects do to timber structures is too well known to dwellers in tropical and sub-tropical regions. The SOCIAL LIFE 257 workers swallow and digest food and then disgorge it in order to feed the " royalties " and young in their nests. Termites also devour each other's excrement. They build earthen tunnels over exposed surfaces of wood on wliich they are feeding, and most species are, like ants, pre- dominantly subterranean in habit. Hence particles of soil, as well as wood, are constantly used as food. It is of great interest also to find that many termites, like the leaf-cutting ants already mentioned in this chapter (p. 242), are " fungus- gardeners " and carry on this very specialised method of feeding to as great an extent as the true ants. Wheeler remarks (1923, p. 270) that while the ant mushroom- gardeners '' are all confined to a single Myrmicine tribe and are exclusively American, the fungus-growing termites all belong to a few genera . . . and are confined to the Ethiopian and Indo-Malayan regions." In the chambers of their nests the worker termites construct spongy '' mush- room beds " consisting of wood and other vegetable material which has been broken up and passed through the insects' food canals. The '* fungus-gardens," remarks Wheeler, " are really the nurseries of the termitarium, and are full of just hatched young, which crop the food-bodies like so many little snow-white sheep." The cultivated fungus is not eaten by the developed workers and soldiers ; it is a special food provided for the growing young, for the " royalties " and for other fertile members of the community. Termites, like ants and wasps, practise extensively mutual feeding. Besides the disgorged, digested food and excrement already mentioned, these insects produce exudate substances, derived from the blood and fat-body, which, permeating the thin-body wall and cuticle of the abdomen, can be '' licked up by other members of the colony." This habit was observed by N. Holmgren (1909) and K. Escherich (191 1), the latter stating that the large swollen queen produces more abundant and richer exudate than any other members of the community, and that her attendant workers in order to obtain the delicacy take the liberty of biting through the royal skin. Wheeler dwells on the importance s 258 THE BIOLOGY OF INSECTS of this " trophallaxis " among the termites, as among the ants and \vasps, in promoting social life. The habit is even more elaborately developed in termite than in ant and wasp communities ; the termites *' may be said to be bound together by a circulating medium of glandular secretions, fatty exudates, and partly and wholly digested food, just as the cells of the body of a higher animal are bound together as a svntrophic whole by means of the circulating blood." Termites of different kinds show much varietv' of habit in the construction of their nests. The small communities of more primitive forms live in irregular galleries or tunnels excavated in wood or soil. A. D. Imms (19 19), in his account of the Himalayan Archotermopsis, describes the tunnels made by these insects in fallen trunks and logs of deodar, and comments on " the complete absence of any- thing in the nature of a true nest or termitarium." Many of the " white-ants " that are notorious as destroyers of timber buildings or furniture, such as the American Leucotermes ffaripes, inhabit cavities eaten out in dry wood. The " concentrated " nests of more highly organised termites, begun in the underground chamber excavated by the royal pair, are developed through the excavation of surrounding galleries and small chambers by the workers, and completed through the up-building of a broad sloping mound or a steep conspicuous " liill-nest," which may attain a height of fifteen or eighteen feet in the case of several tropical African species, while the nests of some Australian termites, t\venty or twent}'-five feet high, have been claimed as the largest of all animal dwellings, if the work of human builders be left out of account. The material of the above- ground structure of these hill nests, like that of the earthen tunnels built by many termites over the stems and branches of trees and shrubs, is soil moistened with the termites* spittle, or disgorged or evacuated after being swallowed, and thus brought into condition suitable for use as building material. *' On drying," remarks Wheeler, '' the sub- stances employed, especially the saliva-impregnated earth, become almost as hard as cement." Wood also, after SOCIAL LIFE 259 treatment wdth the digestive juices, is used by many termites for the construction of their nests, which, in such case, assume a carton-like consistency. These are often found suspended from the branches of trees, they are particularly characteristic of the tropical American forests, " and vary from the size of a football to that of a barrel." The nests of termites in their multifarious modifications have been lately described in detail in the treatise of E. Hegh (1922). Fig. 65. — Rove-beetle {Termitoptocinus australiensis). a, dorsal, and 6, lateral views, X 10; c, larva, dorsal view, X 12. "Guests" of Eutermes fumipennis, N. Australia. After F. Silvestri {Boll. Lab. Portici, XV, 1021). Like the societies of true ants, the termite communities are remarkably interesting on account of the number and variety of alien insects which they harbour as " guests." Our knowledge of these is largely due to the researches of E. Wasmann (19 10-12). Many of them are rove-beetles (Staphylinidae), which, like the fertile termites, have assumed the *' physogastric " condition, the abdomen being greatly swollen and covered for the most part with soft flexible 26o THE BIOLOGY OF INSECTS cuticle, while membranous areas of the thorax grow out into bladder-like or finger-like processes (exudatoria). This curiously degenerative condition, due to overgrowth of the fat-body, brings the beetles into direct feeding relation with their termite hosts, as the insects can all obtain and swallow exudations from each others' swollen bodies. It is note- worthy that the larvae of these " termitophile " rove- beetles (Fig. 65, a) are of a primitive relatively long-legged insectan type, resembhng in aspect the young termites (Fig. 64, a) whose quarters they share. Very remarkable among the termites' guests are certain abnormal flies (Diptera) like the beetles with swollen abdomens and with their wings reduced to strap-like vestiges. Of these the African and Indian Termitoxenia and Termitomyia are described by Wasmann (1900), the African Ptochomyia (1920) and the Brazilian Termitomastus (1901) by F. Sil- vestri. These are clearly related to well-known families of Diptera (Phoridae and others), of normal structure and with well-developed wings, but all have undergone degenera- tive modification in correspondence with their dependent life in the termites' nests. Wasmann beheves that in Termitoxenia the larval and pupal stages have been elimi- nated from the life-histor}^ and that an imago is hatched from the egg. Early in this chapter (p. 218) it was suggested that in an advanced insect community the individuality of the single bee or ant might be regarded as merged in a greater individuality of the society. This view has been forcibly advocated by Julian Huxley (19 12) in a general discussion on the Individual in the Animal Kingdom. '' Communities of ants and bees are," according to him, '* undoubted in- dividuals " ; the single insects are so modified as to exhibit a differentiation of structure and function corresponding to the " division of labour " among the organs of an animal body ; one single ant or bee apart from her comrades is incapable of prolonged survival, and owing to the develop- ment of the insectan nervous system the ant society is *' an individual . . . whose parts, though not contiguous in space, SOCIAL LIFE 261 are yet bound together as fast as the cells of a sponge or the persons of a Siphonophoran." On this view of the matter a single ant cannot be defined as an " actual individual," though *' morphologically and historically equivalent " thereto. The members of the community are now " functioning as parts, but descended from ancestors that functioned as wholes." A similar line of argument has been advanced also by W.N. Wheeler (191 1) in his discussion on the " Ant Colony as an Organism." He dwells on the community as an organic system in relation, as a whole, with its environment, and in this social organism, the fertile members stand for the germ-cells, the workers and soldiers for the body. H. Bergson (1907) claimed that a bee- community is " really and not metaphorically a unique organism." This manner of regarding an insect society is analogous to the personification of such a human society as the city or state. In our own communities, however, it is very rarely possible to forget the true individuality of the single member. This may be overlooked in the ants' nests or the bee-hive, because the behaviour of each single insect is so largely determined by inherited reflexes all tending to the maintenance of the community-life, that the single insect ceases to count. The general perfection of this pre- determined behaviour makes the communistic ant or wasp less plastic and originative than the ** solitary " members of her family often are. Yet worker insects, confronted with unusual conditions, have been observed to behave in a manner demonstrating some power of initiative and adapta- tion, and the morphologist, considering the problem, will find it hard to deny the true individuality of any one member of an ants' nest, even if he is willing to call the whole society a " super-organism." The parallel and divergent con- ditions of the insect community as compared with human society present many fascinating problems for consideration ; but the discussion of these must be deferred to our closing pages. CHAPTER X ADAPTATIONS TO HAUNTS AND SEASONS Repeatedly in the previous chapters of this book, attention has been directed to the adaptation of various kinds of insects in the successive stages of their growth to the surroundings and conditions of their lives. Singly and as a whole, they are admirably fitted to their environment in the wide meaning of that term. We have seen, for example, that the form of their bodies, their legs and wings and the muscles that move these, are adapted to bring about their characteristic movements whether walking, running, leaping, swimming, or flying. Their jaws and digestive canals are suited to the nature of their food whether solid or liquid. They are provided with beautifully constructed sense- organs, and their nerve-centres are so arranged that the impressions received through these organs lead directly to reflex actions appropriate to the conditions under which the creatures find themselves at the time. In our discussion of the growth and transformation of insects after hatching we saw that the differences so frequent and remarkable between adult insects and their larvae may be to some extent explained as modifications which fit the immature creature for life- conditions markedly different from those of its parent. We noticed also that the form and behaviour of an insect in one stage of its development may often suggest a prevision of the conditions of the succeeding stage. The creature is adapted not only for its immediate present needs ; it often prepares in advance for the future events of its Ufe. The subject of adaptation is of such importance and 262 ADAPTATIONS TO HAUNTS AND SEASONS 263 interest to the student of insect biology, that it seems advisable to devote, at this stage, a special chapter to the subject, with illustrations of some of the ways in which insects of different groups, in varying stages of their develop- ment are found to be definitely fitted for certain haunts or places of abode, and are enabled to survive the seasonal changes of the year. We may well begin by considering the haunts of insects from a wide viewpoint — that of the geographer. A com- parative study of the distribution of various kinds and groups of insects over the surface of the earth shows that while very many are adapted by special modifications of form and habit to a special and restricted environment, others seem capable of adapting themselves to the most diverse conditions so that they range widely over vast areas. The '' Painted Lady " Butterfly (Pyrameis cardui), for example, may be found in the most widely separated regions. In many seasons it is abundant in our islands, the large butterflies with their handsome russet, black and white wings flitting along lowland hedgerows or swooping in bold flight over the bare tops of North British and Irish hills. Swarms of the insects migrate northwards in May and June from the Mediterranean district ; these lay their eggs on thistles and other plants and the spring caterpillars feed through the summer, transforming in August and September into a second generation of butterflies. These, however, cannot survive the winter in the climate of northern and north-western Europe ; thus the butterfly, though its powerful flight and the variety and wide range of the plants on which its larvae feed, enable it every summer to invade thousands of square miles of northern territory and there produce progeny, can never establish itself as a true resident outside those warmer regions, the conditions of which allow it to carry on a succession of three or four life-cycles each year. The species ranges eastward far across Asia into India and Japan, and is found also abundantly in many parts of America. It is therefore abundant, dominant, endued with great power for wide dispersal, but limited in its 264 THE BIOLOGY OF INSECTS adaptability for permanent residence in northern latitudes through its intolerance of the cold or damp of winter. Other butterflies of the same family (Nymphalidae), however, such as the " Peacock " (Vanessa to) and the Small Tortoiseshell (Aglais urticae), whose caterpillars feed on nettles, are not only commonly conspicuous members of the British insect fauna, but permanently resident species, because the butterflies of the second or third brood, which emerge from the pupa in autumn, are able to survive the winter in various shelters whither they betake themselves at the onset of cold weather. So strong is the adaptation of Aglais urticae to severe climatic conditions that it is a member of the Arctic fauna, a resident in Greenland. In such a case the presence of the insect in the far northern portion of its range is a demonstration of its power to endure extreme severity of climate. Dragon-flies, as a group, are also remarkable for their strong flight, and many of them occupy a wide range of territory. One of our common British species, Lihellula quadrimaculata, has a range extending all round the northern hemisphere, and the immense migratory swarms in which it sometimes appears must be an important factor in its distribution. R. J. Tillyard (19 17) points out that the large Pantala flavescens and some species of the allied Tramea '' travel far and wide and have overspread the whole of the tropics." He records how an AustraHan dragon-fly, Hemi- cordulia tau, " has recently colonised Tasmania across a strait two hundred miles wide." Very different from such strong flying insects as butter- flies and dragon-flies are the tiny, lowly wingless springtails (Collembola), many kinds of which may be found in our country beneath stones, under bark, among fallen leaves, or feeding on soft plant-tissues or on decaying vegetable or animal substances. Some of these, despite their small size and feeble cuticle, have an enormously wide range. There is a dark, almost black, species Achorutes viaticus (Fig. 66), not more than j^ in. in length, found commonly all over our islands and able to adapt itself to what seem the ADAPTATIONS TO HAUNTS AND SEASONS 265 most varied surroundings. Large numbers may often be seen in garden rubbish-heaps and on farm-land around decaying organic matter. Literally myriads have been observed at the sewage outfalls of large towns — whether on the sea-coast as at Dublin, or on the sewage farms around Manchester and other populous cities of northern England. At Edinburgh and elsewhere it has appeared in multitudes on the water drawn from street hydrants, and Fig. 66. — a, Sprlngtail Achoriites viaticus, side view, X 36 ; 6, right group of ocelli and post-antennal organ, X 240; c, hind foot, X 300; d, dens and mucro of spring, X 240. such occurrences, which have naturally attracted the attention of those responsible for the public health, may be explained by the sweeping in of colonies of the springtail in times of flood from waterside haunts in the catchment area of the town's supply. Achorutes viaticus is often common on the sea-shore, sometimes below high- water mark ; the little insects crowd around the rich food- store in a putrid starfish. Similarly in most European 266 THE BIOLOGY OF INSECTS countries it is known to abound, and naturalists collecting insects in the far north find it common on the coasts of Spitsbergen, in Novaya Zemlya and in Greenland. Probably it inhabits all the continents of the globe, as it has been recorded from Tierra del Fuego, from New Zealand, and from small, sub-antarctic islands to the southward. Many other species of Achorutes have a known range nearly as wide as this, while a closely allied genus (Gomphiocephalus) and an obscure Isotomine springtail are the only non- parasitic insects as yet discovered by explorers of the great Antarctic Continent. G. Taylor has described (1914) how at Granite Harbour in South Victoria Land Gomphio- cephalus swarmed on the surface of a small pool or clustered in a film of ice : " as one turned a pebble to the sun they would thaw out and crawl around for exercise." Delicate white and blind springtails (Onychiurus) are among the commonest members of our " soil fauna," often congregating in hundreds on soft plant tissues and decaying vegetable matter. One of this group — Onychiurus armatus — has nearly as wide a range as the Achorutes just mentioned, and is among those insects recorded by E. Handschin (1924) as present in the Swiss Alps to a height approaching 10,000 feet. Several species of springtail have long been known to disport themselves on the surface of the high Alpine snow- fields where masses of darker coloured Achorutes may appear as blackish patches conspicuous on the pure white back- ground. And while such members of the order live far up the mountain heights, a considerable number of the springtails form a relatively large section of the fauna of deep caves. Among the white species mentioned above as living in the soil, is Onychiurus hiermis ; this insect is found commonly in the deep galleries of caves excavated in the Carboniferous Limestone districts of Great Britain and Ireland. All the species of Onychiurus are eyeless, and most of the springtails inhabiting caves are white and blind, even if they belong to groups whose members are normally provided with eyes. While some of the British and European cave springtails — such as Heteromurus margari- ADAPTATIONS TO HAUNTS AND SEASONS 267 tatus — are not known to live anywhere except in cave galleries, others like Pseudosinella caver narum and Arrhopaltes caecus are found also in less profound dark dwelling-places, such as ants' nests, moles' nests, and quarry- tunnels or under large, deeply imbedded boulders. These distributional facts about springtails have been mentioned in order to emphasise the exceedingly wide range of these frail, lowly insects, whether one considers the order as a whole or many of its component species. If we pass on to inquire the reason of this remarkable adaptability to surroundings often apparently so diverse, the answer seems to be furnished by the small size and comparative simplicity of form and function in these insects, which enable them to survive, increase, and multiply in haunts all of which afford a sufficient degree of shelter and an adequate and easily obtained food-supply, while the conditions as regards humidity allow the necessary gaseous exchanges to go on through the delicate body-wall. Springtails have undergone profound racial changes which may be regarded as indicating degenerative specialisation. Among these is the loss — total or extensive — of the typical insectan system of air-tubes for breathing ; these insects have reverted to a primxitive method of breathing through the general surface of the skin such as is practised by the earthworms and other lowly organisms. Another degenerative change is the disappearance of the compound eyes ; throughout the order these are replaced by sets of simple eyes (ocelli) eight at most on either side of the head (Fig. 66, h), and in many species there are no eyes of any kind. Such blind springtails have been mentioned as characteristic denizens of caves. They have often been regarded as blind on account of their residence in darkness through a long succession of generations ; but, on the other hand, caves and other similar dark places furnish haunts which are tolerated by creatures that cannot see, while those with well-formed eyes generally have the reaction of approaching any perceived source of light and will not therefore remain in darkness if, after wandering or being transported thither, they find themselves free to make a way 268 THE BIOLOGY OF INSECTS to the outer world. Those students of life-relations who doubt that such insects are blind because they dwell in caves might be willing rather to believe that they are found in such haunts because they are blind. Though seeing Httle or nothing, springtails are, however, often provided with organs for receiving other kinds of sense-impression. Their bodies are frequently clothed with long tactile bristles ; the impression conveyed by means of these must be of value in guiding the steps of insects living in darkness and obscurity. On the feelers are peg-Hke or bladder-like organs probably adapted to receive chemical stimulation and to guide the creatures in the search for food. On either side of the head, between the eyes and the base of the feeler, there is found in many springtails a problematical '' post-antennal " organ con- sisting of a set of deUcate areas or processes of cuticle, often arranged in form of ^ circle, an ellipse, or a rosette. To this goes a branch of the optic nerve, the fibres whereof may receive through it impulses due to vibration or chemical stimulation (Fig. 70, h). Springtails as a group are very small, and their body- structure is remarkable among insects, because the number of abdominal segments is reduced from the usual ten or eleven to six. This contraction and shrinkage brings about a decrease in size w^hich enables the insects to subsist on a relatively small food-supply ; hence they can survive and propagate their race in apparently unpromising underground, arctic and alpine haunts where larger and more elaborately organised insects would speedily perish. They are able to hold their own under conditions which would be fatal to the large, strongly built and dominant butterflies and dragon-flies that we were previously considering as occupiers of wide territory. It is now time to turn to examples of insects adapted to special kinds of environments, and the first feature for discussion in this connection is the general form of the creature. An insect's body is made up of a number of segments, arranged in series, one behind the other, and the ADAPTATIONS TO HAUNTS AND SEASONS 269 width of each segment seen in end view is often greater than its depth, the ventral surface of the cuticle is, as a rule, markedly convex, while the dorsal aspect is less convex or flattened. Such a body-form is seen in earwigs, cock- roaches and the great majority of beetles, which are adapted for life on the ground and move principally by walking or running. A relatively broad body is clearly best adapted for this mode of progression ; the same general form is also seen in many insects of comparatively feeble flight. In insects of large size which live under stones or among leaves, the dorso-ventral flattening is carried so far that the body becomes very much wider than deep. This is apparent in most cockroaches which live in the warm countries where they abound in forests among fallen leaves or under bark. In the familiar cockroaches which have been introduced into our cooler regions, we notice that this flattened form enables them to creep into warm shelters between the bricks of stove or oven settings, and behind or beneath hot- water pipes. In another household insect — smaller and more unpleasant than the cockroaches — the blood-sucking " Bed- bug," the flattening is carried to an extreme so great that the breadth of the abdomen is twelve times as great as its depth and the upper surface becomes concave. Here we see a form of body suited to a creature that Hves as a parasite closely adjacent to the skin surface of its host, and finds shelter in the crevices of wooden furniture (Fig. 67, a, h). It is suggestive, however, to remember that among insects of the order (Hemiptera) to which the Bed-bug belongs, a flattening of the body, similar if less extreme, is a common feature. Among insects that are pre-eminently aerial in their mode of hfe we notice an increase in the depth of the body in proportion to its width. This is often particularly evident in the thorax, where provision has to be made for space and attachment for the powerful flight muscles, but a deepening and relative narrowing of the abdomen may often be noted. In many Hymenoptera and Diptera the abdominal segments are distinctly broader than deep, but 270 THE BIOLOGY OF INSECTS in ants, and bees, and wasps the vertical dimension of the abdomen approaches, equals, or exceeds the horizontal, as it does among the more highly organised members of the various groups of two- winged flies — gnats, midges, and the house-fly group. Similar proportions characterise most of the Lepidoptera (moths and butterflies). In dragon-flies also the body is, as a rule, deeper than broad ; these insects have elongate abdomens which being narrow offer Httle resistance to the air in swift flight ; it is interesting to r«^' Fig. 67. — a, Bed-bug (Cimex lectularius) , dorsal view, X 5 ; 6, diagram- matic cross-section of Cimex to show dorso-ventral flattening ; c. Dog- flea {Ctenocephalus cams), lateral view, X 12 (from L. O. Howard); d, diagrammatic cross-section of Flea to show lateral flattening. remember that their grubs crawling on the bottom of pools and streams have bodies often distinctly broader than deep. In contrast to the flattened cockroaches, the grasshoppers and locusts, belonging to the same order (Orthoptera) as they, have bodies distinctly deeper than broad. These insects are often strong fliers, and instead of running along the ground, they spend much of their time crawling on the stems and leaves of plants, whence by the action of their ADAPTATIONS TO HAUNTS AND SEASONS 271 long and powerful hind-legs, they leap into the air, not alighting again until a great distance has been cleared. The compression of the body that characterises them is clearly suited to their movement by great vertical jumps. It is interesting to notice the same modification of form carried to an extreme degree in those well-known parasites the fleas — wingless but obviously related to winged insects — whose bodies are so compressed as to be three times as deep as broad (Fig. 67, c, d). They also are active vertical leapers, their agility affording them the chance of getting on to fresh hosts, and in this respect they afford a most interesting because independent parallel to the grasshoppers, and a striking contrast to the bugs, which are like them- selves parasites. A form of body cylindrical or approximately so, is familiar in many insect larvae such as caterpillars and maggots. In a caterpillar (Fig. 45) this shape together with the support afforded by legs and pro-legs all along the body-length is well-suited for crawling and feeding along twigs or leaf-edges, as well as on leaf-surfaces. Among maggots and grubs generally the rounded form of body, whether cylindrical or tapering, is adapted for burrowing in the plant tissues or refuse which furnish the food supply of so many insect larvae of diverse orders. Among adult insects the cylindrical body is especially characteristic of wood-borers. The well-known " shot- hole " borings in old furniture and timber roofs indicate the presence of Ptinids (" death-watch ") beetles which live in tunnels excavated in the wood and feed on the material thus bitten away. These beetles are all approximately cyHndrical in body form, and a similar shape is noticeable in beetles of other families which practise the same manner of life. Of these the Bostrychidae are akin to the Ptinidae, but the Scolytidae, or bark-beetles (Fig. 49), like those in general aspect, are distinguished by many important structural characters, and belong to a distinct group of the order. We notice here again, therefore, independent and parallel modification of form in two or three different families in 272 THE BIOLOGY OF INSECTS correspondence with their dwelling-places and with the conditions of their lives. The adaptation of many insects to their haunts, in such a manner that they resemble in appearance, to a greater or less extent, the objects among which they live, offers a subject of great interest. Examples have been, however, frequently described and discussed, and some contribution to the question will be attempted in a later chapter (XII), when the fascinating problems of insect evolution generally will be considered. For the present, therefore, it may suffice to recall a few cases illustrative of these relations between insects and their surroundings. Caterpillars feed- ing on leaves are frequently green in colour because the pigment of the plant tissues within the food canals passes into the fat-body and becomes apparent through the translucent skin and cuticle. But dark pigment may be developed in the skin, and as this happens to an increasing degree, the caterpillar tends to become darker and may approx