The Cambridge Manuals of Science and Literature
THE LIFE-STORY OF INSECTS
CAMBRIDGE UNIVERSITY PRESS London: FETTER LANE, E.C. C.F. CLAY, MANAGER
Edinburgh: 100, PRINCES STREET London: H.K. LEWIS, 136, GOWER STREET, W.C. WILLIAM WESLEY & SON, 28, ESSEX STREET, STRAND Berlin: A. ASHER AND CO. Leipzig: F.A. BROCKHAUS New York: G.P. PUTNAM'S SONS Bombay and Calcutta: MACMILLAN AND CO., LTD.
GEO. H. CARPENTER
Professor of Zoology in the Royal College of Science, Dublin
Cambridge: at the University Press New York: G.P. Putnam's Sons 1913
Cambridge: PRINTED BY JOHN CLAY, M.A. AT THE UNIVERSITY PRESS
With the exception of the coat of arms at the foot, the design on the title page is a reproduction of one used by the earliest known Cambridge printer John Siberch 1521
The object of this little book is to afford an outline sketch of the facts and meaning of insect-transformations. Considerations of space forbid anything like an exhaustive treatment of so vast a subject, and some aspects of the question, the physiological for example, are almost neglected. Other books already published in this series, such as Dr Gordon Hewitt's House-flies and Mr O H. Latter's Bees and Wasps, may be consulted with advantage for details of special insect life-stories. Recent researches have emphasised the practical importance to human society of entomological study, and insects will always be a source of delight to the lover of nature. This humble volume will best serve its object if its reading should lead fresh observers to the brookside and the woodland.
I. Introduction 1
II. Growth and Change 8
III. The Life-stories of some Sucking Insects 16
IV. From Water to Air 23
V. Transformations, Outward and Inward 35
VI. Larvae and their Adaptations 49
VII. Pupae and their Modifications 79
VIII. The Life-story and the Seasons 89
IX. Past and Present—the Meaning of the Story 105
Outline Classification of Insects 122
Table of Geological Systems 123
LIST OF ILLUSTRATIONS
Stages in the Transformations of a Gnat Frontispiece
FIG PAGE 1. Stages of the Diamond-back Moth (Plutella 3 cruciferarum)
2. Head of typical Moth 5
3. Head of Caterpillar 5
4. Common Cockroach (Blatta orientalis) 12
5. Nymph of Locust (Schistocera americana) 13
6. Aphis pomi, winged and wingless females 19
7. Mussel Scale-Insect (Mytilaspis pomorum) 21
8. Emergence of Dragon-fly (Aeschna cyanea) 29-31
9. Nymph of May-fly (Chloeon dipterum) 33
10. Imaginal buds of Butterfly 39
11. Imaginal buds of Blow-fly 43
12. Carrion Beetle (Silpha) and larva 51
13. Larva of Ground-beetle (Aepus) 52
14. Willow-beetle (Phyllodecta) and larva 53
15. Cabbage-beetle (Psylliodes) and larva 54
16. Corn Weevil (Calandra) and larva 55
17. Ruby Tiger Moth (Phragmatobia fuliginosa) 61
18. Larvae and Pupa of Hive-bee (Apis mellifica) 65
19. Larva of Gall-midge (Contarinia nasturtii) 68
20. Crane-fly (Tipula oleracea) and larva 69
21. Maggot of House-fly (Musca domestica) 71
22. Ox Warble-fly (Hypoderma bovis) with egg, larva, and puparium 75
23. Pupa of White Butterfly (Pieris) 85
Among the manifold operations of living creatures few have more strongly impressed the casual observer or more deeply interested the thoughtful student than the transformations of insects. The schoolboy watches the tiny green caterpillars hatched from eggs laid on a cabbage leaf by the common white butterfly, or maybe rears successfully a batch of silkworms through the changes and chances of their lives, while the naturalist questions yet again the 'how' and 'why' of these common though wondrous life-stories, as he seeks to trace their course more fully than his predecessors knew.
Everyone is familiar with the main facts of such a life-story as that of a moth or butterfly. The form of the adult insect (fig. 1 a) is dominated by the wings—two pairs of scaly wings, carried respectively on the middle and hindmost of the three segments that make up the thorax or central region of the insect's body. Each of these three segments carries a pair of legs. In front of the thorax is the head on which the pair of long jointed feelers and the pair of large, sub-globular, compound eyes are the most prominent features. Below the head, however, may be seen, now coiled up like a watch-spring, now stretched out to draw the nectar from some scented blossom, the butterfly's sucking trunk or proboscis, situated between a pair of short hairy limbs or palps (fig. 2). These palps belong to the appendages of the hindmost segment of the head, appendages which in insects are modified to form a hind-lip or labium, bounding the mouth cavity below or behind. The proboscis is made up of the pair of jaw-appendages in front of the labium, the maxillae, as they are called. Behind the thorax is situated the abdomen, made up of nine or ten recognisable segments, none of which carry limbs comparable to the walking legs, or to the jaws which are the modified limbs of the head-segments. The whole cuticle or outer covering of the body, formed (as is usual in the group of animals to which insects belong) of a horny (chitinous) secretion of the skin, is firm and hard, and densely covered with hairy or scaly outgrowths. Along the sides of the insect are a series of paired openings or spiracles, leading to a set of air-tubes which ramify throughout the body and carry oxygen directly to the tissues.
Such a butterfly as we have briefly sketched lays an egg on the leaf of some suitable food-plant, and there is hatched from it the well-known crawling larva (fig. 1 b, c, d) called a caterpillar, offering in many superficial features a marked contrast to its parent. Except on the head, whose surface is hard and firm, the caterpillar's cuticle is as a rule thin and flexible, though it may carry a protective armature of closely set hairs, or strong sharp spines. The feelers (fig. 3 At) are very short and the eyes are small and simple. In connection with the mouth, there are present in front of the maxillae a pair of mandibles (fig. 3 Mn), strong jaws, adapted for biting solid food, which are absent from the adult butterfly, though well developed in cockroaches, dragon-flies, beetles, and many other insects. The three pairs of legs on the segments of the thorax are relatively short, and as many as five segments of the abdomen may carry short cylindrical limbs or pro-legs, which assist the clinging habits and worm-like locomotion of the caterpillar. No trace of wings is visible externally. The caterpillar, therefore, differs markedly from its parent in its outward structure, in its mode of progression, and in its manner of feeding; for while the butterfly sucks nectar or other liquid food, the caterpillar bites up and devours solid vegetable substances, such as the leaves of herbs or trees. It is well-known that between the close of its larval life and its attainment of perfection as a butterfly, the insect spends a period as a pupa (fig. 1 e) unable to move from place to place, and taking no food.
 The term larva is applied to any young animal which differs markedly from its parent.
Such, in brief, is the course of the most familiar of insect life-stories. For the student of the animal world as a whole, this familiar transformation raises some startling problems, which have been suggestively treated by F. Brauer (1869), L.C. Miall (1895), J. Lubbock (1874), R. Heymons (1907), P. Deegener (1909) and other writers. To appreciate these problems is the first step towards learning the true meaning of the transformation.
 The dates in brackets after authors' names will facilitate reference to the Bibliography (pp. 124-8).
The butterfly's egg is absolutely and relatively of large size, and contains a considerable amount of yolk. As a rule we find that young animals hatched from such eggs resemble their parents rather closely and pass through no marked changes during their lives. A chicken, a crocodile, a dogfish, a cuttlefish, and a spider afford well-known examples of this rule. Land-animals, generally, produce young which are miniature copies of themselves, for example horses, dogs, and other mammals, snails and slugs, scorpions and earthworms. On the other hand, metamorphosis among animals is associated with eggs of small size, with aquatic habit, and with relatively low zoological rank. The young of a starfish, for example, has hardly a character in common with its parent, while a marine segmented worm and an oyster, unlike enough when adult, develop from closely similar larval forms. If we take a class of animals, the Crustacea, nearly allied to insects, we find that its more lowly members, such as 'water-fleas' and barnacles, pass through far more striking changes than its higher groups, such as lobsters and woodlice. But among the Insects, a class of predominantly terrestrial and aerial creatures producing large eggs, the highest groups undergo, as we shall see, the most profound changes. The life-story of the butterfly, then, well-known as it may be, furnishes a puzzling exception to some wide-reaching generalisations concerning animal development. And the student of science often finds that an exception to some rule is the key to a problem of the highest interest.
During many centuries naturalists have bent their energies to explain the difficulties presented by insect transformations. Aristotle, the first serious student of organised beings whose writings have been preserved for us, and William Harvey, the famous demonstrator of the mammalian blood circulation two thousand years later, agreed in regarding the pupa as a second egg. The egg laid by a butterfly had not, according to Harvey, enough store of food to provide for the building-up of a complex organism like the parent; only the imperfect larva could be produced from it. The larva was regarded as feeding voraciously for the purpose of acquiring a large store of nutritive material, after which it was believed to revert to the state of a second but far larger egg, the pupa, from which the winged insect could take origin. Others again, following de Reaumur (1734), have speculated whether the development of pupa within larva, and of winged insect within pupa might not be explained as abnormal births. But a comparison of the transformation of butterflies with simpler insect life-stories will convince the enquirer that no such heroic theories as these are necessary. It will be realised that even the most profound transformation among insects can be explained as a special case of growth.
GROWTH AND CHANGE
The caterpillar differs markedly from the butterfly. As we pursue our studies of insect growth and transformation we shall find that in some cases the difference between young and adult is much greater—as for example between the maggot and the house-fly, in others far less—as between the young and full-grown grasshopper or plant-bug. It is evidently wise to begin a general survey of the subject with some of those simpler cases in which the differences between the young and adult insect are comparatively slight. We shall then be in a position to understand better the meaning of the more puzzling and complex cases in which the differences between the stages are profound.
In the first place it is necessary to realise that the changes which any insect passes through during its life-story are essentially accompaniments of its growth. The limits of this little book allow only slight reference to features of internal structure; we must be content, in the main, to deal with the outward form. But there is an important relation between this outward form and the underlying living tissues which must be clearly understood. Throughout the great race of animals—the Arthropoda—of which insects form a class, the body is covered outwardly by a cuticle or secretion of the underlying layer of living cells which form the outer skin or epidermis (see fig. 10 ep, cu, p. 39). This cuticle has regions which are hard and firm, forming an exoskeleton, and, between these, areas which are relatively soft and flexible. The firm regions are commonly segmental in their arrangement, and the intervening flexible connections render possible accurate motions of the exoskeletal parts in relation to each other, the motions being due to the contraction of muscles which are attached within the exoskeleton.
 The term 'hypodermis' frequently applied to this layer is misleading. The layer is the true outer skin—ectoderm or epidermis.
Now this jointed exoskeleton—an admirably formed suit of armour though it often is—has one drawback: it is not part of the insect's living tissues. It is a cuticle formed by the solidifying of a fluid secreted by the epidermal cells, therefore without life, without the power of growth, and with only a limited capacity for stretching. It follows, therefore, that at least during the period through which the insect continues to grow, the cuticle must be periodically shed. Thus in the life-story of an insect or other arthropod, such as a lobster, a spider, or a centipede, there must be a succession of cuticle-castings—'moults' or ecdyses as they are often called.
When such a moult is about to take place the cuticle separates from the underlying epidermis, and a fluid collects beneath. A delicate new cuticle (see fig. 10 cu') is then formed in contact with the epidermis, and the old cuticle opens, usually with a slit lengthwise along the back, to allow the insect in its new coat to emerge. At first this new coat is thin and flabby, but after a period of exposure to the air it hardens and darkens, becoming a worthy and larger successor to that which has been cast. The cuticle moreover is by no means wholly external. The greater part of the digestive canal and the whole air-tube system are formed by inpushings of the outer skin (ectoderm) and are consequently lined with an extension of the chitinous cuticle which is shed and renewed at every moult.
In all insects these successive moults tend to be associated with change of form, sometimes slight, sometimes very great. The new cuticle is rarely an exact reproduction of the old one, it exhibits some new features, which are often indications of the insect's approach towards maturity. Even in some of those interesting and primitive insects the Bristle-tails (Thysanura) and Spring-tails (Collembola), in which wings are never developed, perceptible differences in the form and arrangement of the abdominal limbs can be traced through the successive stages, as R. Heymons (1906) and K.W. Verhoeff (1911) have shown for Machilis. But the changes undergone by such insects are comparatively so slight, that the creatures are often known as 'Ametabola' or insects without transformation in the life-history. Now there are a considerable number of winged insects—cockroaches and grasshoppers for example—in which the observable changes are also comparatively slight. We will sketch briefly the main features of the life-story of such an insect.
The young creature is hatched from the egg in a form closely resembling, on the whole, that of its parent, so that the term 'miniature adult' sometimes applied to it, is not inappropriate. The baby cockroach (fig. 4 d) is known by its flattened body, rounded prothorax, and stiff, jointed tail-feelers or cercopods; the baby grasshopper by its strong, elongate hind-legs, adapted, like those of the adult, for vigorous leaping. During the growth of the insect to the adult state there may be four or five moults, each preceded and succeeded by a characteristic instar. The first instar differs, however, from the adult in one conspicuous and noteworthy feature, it possesses no trace of wings. But after the first or the second moult, definite wing-rudiments are visible in the form of outgrowths on the corners of the second and third thoracic segments. In each succeeding instar these rudiments become more prominent, and in the fourth or the fifth stage, they show a branching arrangement of air-tubes, prefiguring the nervures of the adult's wing (fig. 5). After the last moult the wings are exposed, articulated to the segments that bear them, and capable of motion. Having been formed beneath the cuticle of the wing-rudiments of the penultimate instar, the wings are necessarily abbreviated and crumpled. But during the process of hardening of the cuticle, they rapidly increase in size, blood and air being forced through the nervures, so that the wings attaining their full expanse and firmness, become suited for the function of flight.
 The convenient term 'instar' has been proposed by Fischer and advocated by Sharp (1895) for the form assumed by an insect during a stage of its life-story. Thus the creature as hatched from the egg is the first instar, after the first moult it has become the second instar, and so on, the number of moults being always one less than the number of instars.
The changes through which these insects pass are therefore largely connected with the development of the wings. It is noteworthy that in an immature cockroach the entire dorsal cuticle is hard and firm. In the adult, however, while the cuticle of the prothorax remains firm, that of the two hinder thoracic and of all the abdominal segments is somewhat thin and delicate on the dorsal aspect. It needs not now to be resistant, because it is covered by the two firm forewings, which shield and protect it, except when the insect is flying. There are, indeed, slight changes in other structures not directly connected with the wings. In a young grasshopper, for example, the feelers are relatively stouter than in the adult, and the prothorax does not show the specifically distinctive shape with its definite keels and furrows. Changes in the secondary sexual characters may also be noticed. For instance, in an immature cockroach both male and female carry a pair of jointed tail-feelers or cercopods on the tenth abdominal segment, and a pair of unjointed limbs or stylets on the ninth. In the adult stage, both sexes possess cercopods, but the males only have stylets, those of the female disappearing at the final moult.
Reviewing the main features of the life-story of a grasshopper or cockroach, we notice that there is no marked or sudden change of form. The newly-hatched insect resembles generally its parent, except that it has no wings. Wing-rudiments appear, however, in an early instar as visible outgrowths on the thoracic segments, and become larger after each moult. All through its various stages the immature insect—nymph as it is called—lives in the same kind of situations and on the same kind of food as its parent, and it is all along active and lively, undergoing no resting period like the pupal stage in the transformation of the butterfly.
One interesting and suggestive fact remains to be mentioned. There are grasshoppers and cockroaches in which the changes are even less than those just sketched, because the wings remain, even in the adult, in a rudimentary state (as for example in the female of the common kitchen cockroach, Blatta orientalis, see fig. 4 a), or are never developed at all. Such exceptional winglessness in members of a winged family can only be explained by the recognition of a life-story, not merely in the individual but in the race. We cannot doubt that the ancestors of these wingless insects possessed wings, which in the course of time have been lost by the whole species or by the members of the female sex. It is generally assumed that this loss has been gradual, and so in many cases it probably may have been. But there are species of insects in which some generations are winged and others wingless; a winged mother gives birth to wingless offspring, and a wingless parent to young with well-developed wings. Such discontinuity in the life-story of a single generation forces us to recognise the possibility of similar sudden mutations in the course of that age-long process of evolution to which the facts of insect growth, and indeed of all animal development, bear striking testimony.
THE LIFE-STORIES OF SOME SUCKING INSECTS
We may now turn our attention to some examples of the remarkable alternation of winged and wingless generations in the yearly life-cycle of the same species, mentioned at the end of the last chapter. Cockroaches and grasshoppers belong to an order of insects, the Orthoptera, characterised by firm forewings and biting jaws; in all of them the change of form during the life-history is comparatively slight. A great contrast to those insects in the structure of the mouth-parts is presented by the Hemiptera, an order including the bugs, pond-skaters, cicads, plant-lice, and scale-insects. These all have an elongated, grooved labium projecting from the head in form of a beak, within which work, to and fro, the slender needle-like mandibles and maxillae by means of which the insect pierces holes through the skin of a leaf or an animal, and is thus enabled to suck a meal of sap or blood, according to its mode of life. In many Hemiptera—the various families of bugs both aquatic and terrestrial, for example—the life-history is nearly as simple as that of a cockroach. It is the family of the plant-lice (Aphidae) that affords typical illustrations of that alternation of generations to which reference has been made.
 See outline classification of insects, p. 122.
The yearly cycle of the common Aphids of the apple tree has been lately worked out in detail by J.B. Smith (1900) and E.D. Sanderson (1902). In late autumn tiny wingless males and females are found in large numbers on the withered leaves. The sexes pair together, and the females lay their relatively large, smooth, hard-coated black eggs on the twigs; these resistant eggs carry the species safely over the winter. At springtide, when the leaves begin to sprout from the opening buds the aphid eggs are hatched, and the young insects after a series of moults, through which hardly any change of form is apparent, all grow into wingless 'stem-mothers' much larger than the egg-laying females of the autumn. The stem-mothers have the power, unusual among animals as a whole, but not very infrequent in the insects and their allies, of reproducing their kind without having paired with a male. Eggs capable of parthenogenetic development, produced in large numbers in the ovaries of these females, give rise to young which, developing within the body of the mother, are born in an active state. Successive broods of these wingless virgin females (fig. 6 a) appear through the spring and summer months, and as the rate of their development is rapid, often the whole life-story is completed within a week. The aphid population increases very fast. Later a generation appears in which the thoracic segments of the nymphs are seen to bear wing-rudiments like those of the young cockroach, and a host of winged females (fig. 6b) are produced; these have the power of migrating to other plants. We understand that wings are not necessary to the earlier broods whose members have plenty of room and food on their native shoots, but that when the population becomes crowded, a winged brood capable of emigration is advantageous to the race.
 Such virgin reproduction is termed 'parthenogenesis.'
Many generations of virgin female aphids, some wingless, others winged when adult, succeed each other through the summer months. At the close of the year the latest brood of these bring forth young, which develop into males and egg-laying females; thus the yearly cycle is completed. Variations in points of detail may be noticed in different species of aphids. The autumn males and egg-laying females are, for example, frequently winged, and the same species may have constantly recurring generations of different forms adapted for different food-plants, or for different regions of the same food-plant. But taking a general view of the life-story of aphids for comparison with the life-story of other insects, three points are especially noteworthy. Virgin reproduction recurs regularly, parthenogenetic broods being succeeded by a single sexual brood. A winged parent brings forth young which remain always wingless, and wingless adults produce young which acquire wings. The wings are developed, as in the cockroach, from outward and visible wing-rudiments.
A family of Hemiptera, related to the Aphidae and equally obnoxious to the gardener, is that of the Coccidae or scale-insects. These furnish an excellent illustration of features noticeable in certain insect life-histories. In the first place, the newly-hatched young differs markedly from the parent in the details of its structure. A young coccid (fig. 7 c) is flattened oval in shape, has well-developed feelers (fig. 7 d) and legs, and runs actively about, usually on the leaves or bark of trees and shrubs, through which it pierces with its long jaws, so that it may suck sap from the soft tissues beneath. After a time it fixes itself by means of these jaws and the characteristic scale or protective covering, composed partly of a waxy secretion and partly of dried excrement, begins to grow over its body. The female loses legs and feelers, and never acquires wings, becoming little more than a sluggish egg-bag (fig. 7 e). The male on the other hand passes into a second larval stage in which there are no functional legs, but rudiments of legs and of wings are present on the epidermis beneath the cuticle, as shown by B.O. Schmidt for Aspidiotus (1885). The penultimate instar of this sex in which the wing-rudiments are visible externally lies passively beneath the scale, its behaviour resembling that of a butterfly pupa. The adult winged male (fig. 7 a) leads a short, but active life.
Another family allied to the Aphidae is that of the Cicads, hardly represented in our fauna but abundant in many of the warmer regions of the earth. Here also the young insect differs widely from its parent in form, living underground and being provided with strong fore-legs for digging in the soil. After a long subterranean existence, usually extending over several years, the insect attains the penultimate stage of its life-story, during which it rests passively within an earthen cell, awaiting the final moult, which will usher in its winged and perfect state.
In the life-histories of cicads and coccids, then, there are some features which recall those of the caterpillar's transformation into the butterfly. The newly-hatched insect is externally so unlike its parent that it may be styled a larva. The penultimate instar is quiescent and does not feed. But while the caterpillar shows throughout its life no outward trace of wings, external wing-rudiments are evident in the young stages of the cicad. In the male coccid we find a late larval stage with hidden wing-rudiments, the importance of which, for comparison with the caterpillar, will be appreciated later.
FROM WATER TO AIR
Insects as a whole are preeminently creatures of the land and the air. This is shown not only by the possession of wings by a vast majority of the class, but by the mode of breathing to which reference has already been made (p. 2), a system of branching air-tubes carrying atmospheric air with its combustion-supporting oxygen to all the insect's tissues. The air gains access to these tubes through a number of paired air-holes or spiracles, arranged segmentally in series.
It is of great interest to find that, nevertheless, a number of insects spend much of their time under water. This is true of not a few in the perfect winged state, as for example aquatic beetles and water-bugs ('boatmen' and 'scorpions') which have some way of protecting their spiracles when submerged, and, possessing usually the power of flight, can pass on occasion from pond or stream to upper air. But it is advisable in connection with our present subject to dwell especially on some insects that remain continually under water till they are ready to undergo their final moult and attain the winged state, which they pass entirely in the air. The preparatory instars of such insects are aquatic; the adult instar is aerial. All may-flies, dragon-flies, and caddis-flies, many beetles and two-winged flies, and a few moths thus divide their life-story between the water and the air. For the present we confine attention to the Stone-flies, the May-flies, and the Dragon-flies, three well-known orders of insects respectively called by systematists the Plecoptera, the Ephemeroptera and the Odonata.
In the case of many insects that have aquatic larvae, the latter are provided with some arrangement for enabling them to reach atmospheric air through the surface-film of the water. But the larva of a stone-fly, a dragon-fly, or a may-fly is adapted more completely than these for aquatic life; it can, by means of gills of some kind, breathe the air dissolved in water.
The aquatic young of a stone-fly does not differ sufficiently in form from its parent to warrant us in calling it a larva; the life-history is like that of a cockroach, all the instars however except the final one—the winged adult or imago—live in the water. The young of one of our large species, a Perla for example, has well-chitinised cuticle, broad head, powerful legs, long feelers and cerci like those of the imago; its wings arise from external rudiments, which are conspicuous in the later aquatic stages. But it lives completely submerged, usually clinging or walking beneath the stones that lie in the bed of a clear stream, and examination of the ventral aspect of the thorax reveals six pairs of tufted gills, by means of which it is able to breathe the air dissolved in the water wherein it lives. At the base of the tail-feelers or cerci also, there are little tufts of thread-like gills as J.A. Palmen (1877) has shown. An insect that is continually submerged and has no contact with the upper air cannot breathe through a series of paired spiracles, and during the aquatic life-period of the stone-fly these remain closed. Nevertheless, breathing is carried on by means of the ordinary system of branching air-tubes, the trunks of which are in connection with the tufted hollow gill-filaments, through whose delicate cuticle gaseous exchange can take place, though the method of this exchange is as yet very imperfectly understood. When the stone-fly nymph is fully grown, it comes out of the water and climbs to some convenient eminence. The cuticle splits open along the back, and the imago, clothed in its new cuticle, as yet soft and flexible, creeps out. The spiracles are now open, and the stone-fly breathes atmospheric air like other flying insects. But throughout its winged life, the stone-fly bears memorials of its aquatic past in the little withered vestiges of gills that can still be distinguished beneath the thorax.
The adult dragon-fly (fig. 8 d) is specialised in such a way that it captures its prey—flies and other small insects—on the wing, swooping through the air like a hawk and feeding voraciously. The head is remarkable for its large globular compound eyes, its short bristle-like feelers, and its very strong mandibles which bite up the bodies of the victims. The thorax bears the two pairs of ample wings, firm and almost glassy in texture, and its segments are projected forward ventrally, so that all six legs, which are armed with rows of sharp, slender spines, can be held in front of the mouth, where they form an effective fly-trap. The abdomen is very long and usually narrow.
A female dragon-fly after a remarkable mode of pairing, the details of which are beside our present subject, drops her eggs in the water, or lays them on water-weeds, perhaps cutting an incision where they can be the more safely lodged, or even goes down below the surface and deposits them in the mud at the bottom of a pond. From the eggs are hatched the aquatic larvae which differ in many respects from the imago. The dragon-fly larva has the same predaceous mode of life as its parent, but it is sluggish in habit, lurking for its prey at the bottom of the pond, among the mud or vegetation, which it resembles in colour. The thoracic segments have not the specialisation that they show in the imago; the abdomen is relatively shorter and broader. The larval head has, like that of the imago, short feelers, and the eyes are somewhat large, though far from attaining the size of the great globular eyes of the dragon-fly. But the third pair of jaws, forming the labium, are most remarkably modified into a 'mask,' the distal central portion (mentum) being hinged to the basal piece (sub-mentum) which is itself jointed below the head. The mentum carries at its extremity a pair of lobes with sharp fangs. Thus the mask can be folded under the head when the larva lurks in its hiding place, or be suddenly darted out so as to secure any unwary small insect that may pass close enough for capture. Dragon-fly larvae walk, and also swim by movements of the abdomen or by expelling a jet of water from the hind-gut. The walls of this terminal region of the intestine have areas lined with delicate cuticle and traversed by numerous air-tubes, so that gaseous exchange can take place between the air in the tubes and that dissolved in the water. The larvae of the larger and heavier dragon-flies (Libellulidae and Aeschnidae) breathe mostly in this way. Those of the slender and delicate 'Demoiselles' (Agrionidae) are provided with three leaf-like gill-plates at the tail, between whose delicate surfaces numerous air-tubes ramify. These gill-plates are at times used for propulsion. Thus air supply is ensured during aquatic life. But occasionally, when the water in which the larva lives is foul and poor in oxygen, the tail is thrust out of the water so that air can be admitted directly into the intestinal chamber. The aquatic life of these insects lasts for more than a year, and F. Balfour-Browne (1909) has observed from ten to fourteen moults in Agrion. Outward wing-rudiments are early visible on the thoracic segments; when these have become conspicuous the insect, beginning in some respects to approach the adult condition, is often called a nymph. In an advanced dragon-fly nymph, H. Dewitz (1891) has shown that the thoracic spiracles are open, and, as the time for its final moult draws near, the insect may thrust the front part of its body out of the water, and breathe atmospheric air through these. Thus before the great change takes place the nymph has foretastes of the aerial mode of breathing which it will practise when the perfect stage shall have been attained. The emergence of the dragon-fly from its nymph-cuticle has been described by many naturalists from de Reaumur (1740) to L.C. Miall (1895) and O.H. Latter (1904). The nymph climbs out of the water by ascending some aquatic plant, and awaits the change so graphically sketched by Tennyson:
A hidden impulse rent the veil, Of his old husk, from head to tail, Came out clear plates of sapphire mail.
'From head to tail,' for the nymph-cuticle splits lengthwise down the back, and the head and thorax of the imago are freed from it (fig. 8 a), then the legs clasp the empty cuticle, and the abdomen is drawn out (fig. 8 b, c). After a short rest, the newly-emerged fly climbs yet higher up the water-weed, and remains for some hours with the abdomen bent concave dorsalwards (fig. 8 d), to allow space for the expansion and hardening of the wings. For some days after emergence the cuticle of the dragon-fly has a dull pale hue, as compared with the dark or brightly metallic aspect that characterises it when fully mature. The life of the imago endures but a short time compared with the long aquatic larval and nymphal stages. After some weeks, or at most a few months, the dragon-flies, having paired and laid their eggs, die before the approach of winter.
The life-story of a may-fly follows the same general course as that just described for the dragon-flies, but there are some suggestive differences. In the first place, we notice a wider divergence between the imago and the larva. An adult may-fly is one of the most delicate of insects; the head has elaborate compound eyes, but the feelers are very short, and the jaws are reduced to such tiny vestiges that the insect is unable to feed. Its aquatic larva is fairly robust, with a large head which is provided with well-developed jaws, as the larval and nymphal stages extend over one or two years, and the insects browse on water-weeds or devour creatures smaller and weaker than themselves. They breathe dissolved air by means of thread-like or plate-like gills traversed by branching air-tubes, somewhat resembling those of the demoiselle dragon-fly larva. But in the may-fly larva, there is a series of these gills (fig. 9b) arranged laterally in pairs on the abdominal segments, and C. Boerner (1909) has recently given reasons, from the position and muscular attachments of these organs, for believing that they show a true correspondence to (in technical phraseology are homologous with) the thoracic legs. One feature in which the larva often agrees with the imago is the possession on the terminal abdominal segment of a pair of long jointed cerci, and in many genera a median jointed tail-process (see fig. 9) is also present, in some cases both in the larva and the imago, in others in the larva during its later stages only. The prolonged larval life in may-flies often involves a large series of moults; Lubbock (1863) has enumerated twenty-one in the life-history of Chloeon. In the second year of aquatic life wing-rudiments (fig. 9 a) are visible, and the larva becomes a nymph. When the time for the winged condition approaches the nymphs leave the water in large swarms. The vivid accounts of these swarms given by Swammerdam (1675), de Reaumur (1742) and other old-time observers are available in summarised form for English readers in Miall's admirable book (1895). May-flies are eagerly sought as food by trout, and the rise of the fly on many lakes ushers in a welcome season to the angler.
The nymph-cuticle opens and the winged insect emerges. But this is not the final instar; may-flies are exceptional among insects in undergoing yet another moult after they have acquired wings which they can use for flight. The instar that emerges from the nymph-cuticle is a sub-imago, dull in hue, with a curious immature aspect about it. A few hours later the final moult takes place, a very delicate cuticle being shed and revealing the true imago. Then follow the dancing flight over the calm waters, the mating and egg-laying, the rapid death. The whole winged existence prepared for by the long aquatic life may be over in a single evening; at most it lasts but for a few days.
From Miall and Denny after Vayssiere.]
In the development of the may-flies, then, we notice not only a considerable divergence between larva and imago, both in habitat and structure; we see also what is to be observed often in more highly organised insects—a feeding stage prolonged through the years of larval and nymphal life, while the winged imago takes no food and devotes its energies through its short existence to the task of reproduction. Such division of the life-history into a long feeding, and a short breeding period has, as will be seen later, an important bearing on the question of insect transformation generally, and the dragon-flies and may-flies afford examples of two stages in its specialisation. The sub-imaginal instar of the may-fly furnishes also a noteworthy fact for comparison with other insect histories. In two points, however, the life-story of these flies with their aquatic larvae recalls that of the cockroach. All the larval and nymphal instars are active, and the wing-rudiments are outwardly visible long before the final moult.
TRANSFORMATIONS,—OUTWARD AND INWARD
We are now in a position to study in some detail the transformation of those insects whose life-story corresponds more or less closely with that of the butterfly, sketched in the opening pages of this little book. In the case of some of the insects reviewed in the last three chapters, the may-flies and cicads for example, a marked difference between the larva and the imago has been noticed; in others, as the coccids, we find a resting instar before the winged condition is assumed, suggesting the pupal stage in the butterfly's life-story.
The various insect orders whose members exhibit no marked divergence between larva and imago (the Orthoptera for example) are often said to undergo no transformation, to be 'Ametabola.' Those with life-stories such as the dragon-flies' are said to undergo partial transformation, and are termed 'Hemimetabola.' Moths, caddis-flies, beetles, two-winged flies, saw-flies, ants, wasps, bees, and the great majority of insects, having the same type of life-story as the butterfly, are said to undergo complete transformation and are classed as 'Metabola' or 'Holometabola.' Wherein lies the fundamental difference between these Holometabola on the one hand and the Hemimetabola and Ametabola on the other? It is not that the larva differs from the imago or that there is a passive stage in the life-history; these conditions are observable among insects with a 'partial' transformation as we have seen, though the resting instar that simulates the butterfly pupa is certainly exceptional. It has been pointed out by Sharp (1899) that the most important indication of the difference between the two modes of development is furnished by the position of the wing-rudiments. In all Ametabola and Hemimetabola these are visible externally long before the penultimate instar has been reached; in the Holometabola they are not seen until the pupal stage.
Attention has already been drawn to the contrast in outward form between a butterfly and its caterpillar. As in the case of dragon-fly or may-fly, the larval period is essentially a time for feeding and growth, and during this period the larval cuticle is cast four or five, in some species even seven or eight times. After each moult some changes in detail may be observable, for example in the proportions of the body-segments or their outgrowths, in the colour or the closeness of the hairy or spiny armature. But in all main features the caterpillar retains throughout its life the characteristic form in which it left the egg. From the tiny, newly-hatched larva to the full-fed caterpillar, possibly several inches in length, there is all along the same crawling, somewhat worm-like body, destitute of any outward trace of wings. When however the last larval cuticle has split open lengthwise along the back, and has been worked off by vigorous wriggling motions of the insect, the pupa thus revealed shows the wing-rudiments conspicuous at the sides of the body, and lying neatly alongside these are to be seen the forms of feelers, legs, and maxillae of the imago prefigured in the cuticle of the pupa (fig. 1 e). The pupa thus resembles the imago much more closely than it resembles the larva; even in the proportions of the body a relative shortening is to be noticed, and the imago of any insect with complete transformation is reduced in length as compared with the full-fed larva. Now these wings and other structures characteristic of the imago, appear in the pupa which is revealed by the shedding of the last larval cuticle. From these facts we infer that the wing-rudiments must be present in the larva, hidden beneath the cuticle; and until the last larval instar, not beneath the cuticle only, but growing in such-wise that they are hidden by the epidermis. For if they were growing outwardly the new cuticle would be formed over them, so that they would be apparent after the next moult. But it is clear that only in the pupa, forming beneath the cuticle of the last larval instar, can they grow outwards.
Anatomical study of the caterpillar at various stages verifies the conclusions just drawn from superficial observation. A hundred and fifty years ago P. Lyonet in his monumental work (1762) on the caterpillar of the Goat Moth (Cossus) detected, in the second and third thoracic segments, four little white masses buried in the fat-body, and, while doubtful as to their real meaning, he suggested that their number and position might well give rise to the suspicion that they were rudiments of the wings of the moth. But it was a century later that A. Weismann in his classical studies (1864) on the development of common flies, showed the presence in the maggot of definite rudiments of wings, and other organs of the adult—rudiments to which he gave the name of imaginal discs. We will recur later to these transformations of the Diptera. For the present, we pursue our survey of changes in the life-history of the Lepidoptera and can take to guide us the excellent researches of J. Gonin (1894).
Careful study of the imaginal discs of the wings in a caterpillar (fig. 10) made by examining microscopically sections cut through them, shows that the epidermis is pushed in to form a little pouch (C, p) and that into this grows the actual wing-rudiment. Consequently the whitish disk which seems to lie within the body-wall of the larva, is really a double fold of the epidermis, the outer fold forming the pouch, the inner the actual wing-bud. Into the cavity of the latter pass branches from the air-tube system. In its earliest stage, the wing-bud is simply an ingrowing mass of cells (fig. 10 A) which subsequently becomes an inpushed pouch (B). Until the last stage of larval life the wing-bud remains hidden in its pouch, and no cuticle is formed over it. When the pupal stage draws near the bud grows out of its sheath, and projecting from the general surface of the epidermis becomes covered with cuticle to be revealed, as we have seen, after the last larval moult, as the pupal wing. Thus all through the life of the humble, crawling caterpillar, 'it doth not yet appear what it shall be,' but there are being prepared, hidden and unseen, the wondrous organs of flight, which in due time will equip the insect for the glorious aerial existence that awaits it.
As mentioned above, this hidden growth of the wing-rudiments, in butterflies, beetles, flies, bees, and the great majority of the winged insects, has been emphasised by Sharp (1899) as a character contrasting markedly with the outward and visible growth of the wing-rudiments in such insects as cockroaches, bugs, and dragon-flies. The divergence between the two modes of development is certainly very striking, and a conceivable method of transition from the one to the other is not easy to explain. Sharp has expressed the divergence by the terms Endopterygota, applied to all the orders of insects with hidden wing-rudiments (the 'Metabola' or 'Holometabola' of most classifications) and Exopterygota, including all those insects whose wing-rudiments are visible throughout growth ('Hemimetabola' and 'Ametabola'). Those curious lowly insects, belonging to the two orders of the Collembola and Thysanura, none of whose members ever develop wings at all, form a third sub-class, the Apterygota (see Classificatory Table, p. 122).
Not the wings only, but other structures of the imago, varying in extent in different orders, are formed from the imaginal discs. For example, de Reaumur and G. Newport (1839) found that if the thoracic leg of a late-stage caterpillar were cut off, the corresponding leg of the resulting butterfly would still be developed, although in a truncated condition. Gonin has shown that in the Cabbage White butterfly (Pieris brassicae) the legs of the imago are represented, through the greater part of larval life, only by small groups of cells situated within the bases of the larval legs. After the third moult these imaginal discs grow rapidly and the proximal portion of each, destined to develop into the thigh and shin of the butterfly's leg, sinks into a depression at the side of the thorax, while the tip of the shin and the five-segmented foot project into the cavity of the larval leg. Hence we understand that the amputation of the latter by the old naturalists truncated only and did not destroy the imaginal limb. In the blow-fly maggot, Weismann, B.T. Lowne (1890) and J. Van Rees (1888) have shown that the imaginal discs of the legs (fig. 11—1, 2, 3) grow out from deep dermal inpushings. Simple at first, these outgrowths by partial splitting, become differentiated into thigh and shin.
Similarly the feelers and jaws of the butterfly are developed from imaginal discs, and this fact explains how it comes to pass that they differ so widely from the corresponding structures in the caterpillar. The larval feelers (fig. 3 At) are short and stumpy, those of the butterfly long and many-jointed. The maxilla of the larva (fig. 3 Mx) consists of a base carrying two short jointed processes; in the butterfly a certain portion of the maxilla, the hood or galea, is modified into a long, flexible grooved process, capable of forming with its fellow the trunk through which the insect sucks its liquid food (fig. 2). Nothing but some such provision as that of the imaginal discs could render possible the wonderful replacement of the caterpillar's jaws, biting solid food, into those of the butterfly sipping nectar from flowers.
A curious segmental displacement of the imaginal discs with regard to the larva is noticeable in some Diptera. In the larva of the harlequin-midge (Chironomus) as described by Miall and Hammond (1900) the brain is situated in the thorax, and the imaginal discs for the head, eyes, and feelers of the adult lie in close association with it, though they arise from inpushings of the larval head. These rudiments do not appear until the last larval stage has been reached. In the gnats Culex and Corethra, on the other hand, the imaginal discs for the head-appendages retain their normal position within the larval head, and appear in an early stage of larval life. Among the flies of the bluebottle group (Muscidae) the brain (fig. 11 B) is situated, as in Chironomus, in the thoracic region of the legless maggot, which is the larva of an insect of this family, and the imaginal discs for eyes and feelers (fig. 11 e, f) lie just in front of it. Here, the imaginal buds of the legs (fig. 11—1, 2, 3) and wings (fig. 11 W, w) are deeply inpushed, retaining their connection with the skin only by means of a thread of cells. As the larva is legless and headless its outer form is not affected by the discs and it is not surprising to learn that they appear early. It has indeed been suggested that the pharyngeal region of the larva, in connection with which the imaginal head-discs are developed, should be regarded, though it lies in the thorax, as an inpushed anterior section of the larval head. In any case this region is pushed out during the formation of the pupa within the final larval cuticle, so that the imaginal head with its contained brain, its compound eyes, and its complex feelers, takes its rightful place at the front end of the insect.
The mention of the brain suggests a few brief remarks on the changes in the internal organs during insect transformation. There are no imaginal discs for the nervous system; the brain, nerve-cords and ganglia of the butterfly or bluebottle are the direct outcome of those of the caterpillar or maggot. More than seventy years ago, Newport (1839) traced the rapid but continuous changes, which, during the early pupal period, convert the elongate nerve-cord of the caterpillar with its relatively far-separated ganglia into the shortened, condensed nerve-cord of the Tortoise-shell butterfly (Vanessa urticae) with several of the ganglia coalesced. In many Diptera, on the other hand, the nervous system of the larva is more concentrated than that of the imago.
The tubular heart also of a winged insect is the directly modified survival of the larval heart.
Similarly the reproductive organs undergo a gradual, continuous development throughout an insect's life-story. Their rudiments appear in the embryo, often at a very early stage; they are recognisable in the larva, and the matured structures in the imago are the result of their slow process of growth, the details of which must be reckoned beyond the scope of this book. For a full summary of the subject the reader is referred to L.F. Henneguy's work (1904) containing references to much important modern literature, which cannot be mentioned here.
On the other hand, the digestive system of insects that undergo a metamorphosis, passes through a profound crisis of dissolution and rebuilding. This is not surprising when we remember that there is often a great difference between larva and imago in the nature of the food. The digestive canal of a caterpillar runs a fairly straight course through the body and consists of a gullet, stomach (mid-gut), intestine, and rectum; it is adapted for the digestion of solid food. In the butterfly there is one outgrowth of the gullet in the head—a pharyngeal sac adapted for sucking liquids; and another outgrowth at the hinder end of the gullet (which is much longer than in the larva)—a crop or food-reservoir lying in the abdomen. The intestine of the butterfly also is longer than that of the larva, being coiled or twisted. Towards the end of the last larval stage, the cells of the inner coat (epithelium) lining the stomach begin to undergo degeneration, small replacing cells appearing between their bases and later giving rise to the more delicate epithelium that lines the mid-gut of the imago. The larval cells are shed into the cavity of the stomach and become completely broken down. J. Anglas (1902), describing these microscopic changes in the transformations of wasps and bees, has shown that the tiny replacing cells can be recognised in sections through the digestive canal of a very young larva; they may be regarded as representing imaginal buds of the adult gastric epithelium. In the transformations of two-winged flies of the bluebottle group, A. Kowalevsky (1887) has shown that these replacing cells are aggregated in little masses scattered at different points along the stomach and thus corresponding rather closely to the imaginal discs of the legs and wings.
The gullet, crop, and gizzard of an insect, which lie in front of the stomach, are lined by cells derived from the outer skin (ectoderm) which is pushed in to form what is called the 'fore-gut.' Similarly the intestine and rectum, behind the stomach, are lined with ectodermal cells which arise from the inpushed 'hind-gut.' The larval fore- and hind-guts are broken down at the end of larval life and their lining is replaced by fresh tissue derived from two imaginal bands which surround the cavity of the digestive tube, one at the hinder end of the fore-gut, and the other at the front end of the hind-gut. The larval salivary glands in connection with the gullet are also broken down, and fresh glands are formed for the imago.
A large part of the substance of an insect larva consists of muscular tissue, surrounding the digestive tube, and forming the great muscles that move the various parts of the body, and of fat, surrounding the organs and serving as a store of food-material. Very many of the muscle-fibres and the fat-cells also become disintegrated during the late larval and pupal stages, and the corresponding tissues of the adult are new formations derived from special groups of imaginal cells, though some muscles may persist from the larva to the adult. Similarly the complex air-tube or tracheal system of the larva is broken down and a fresh set of tubes is developed, adapted to the altered body-form of pupa and imago.
The destruction of larval tissue and the development of replacing organs from special groups of cells, derived of course from the embryo, and carrying on the continuity of cell-lineage to the adult, are among the most remarkable facts connected with the life-story of insects. The process of tissue-destruction is known as 'histolysis'; the rebuilding process is called 'histogenesis.' Considerable difference of opinion has existed as to factors causing histolysis, and for a summary of the conflicting or complementary theories, the reader is referred to the work of L.F. Henneguy (1904, pp. 677-684). In the histolysis of the two-winged flies, wandering amoeboid cells—like the white corpuscles or leucocytes of vertebrate blood—have been observed destroying the larval tissues that need to be broken down, as they destroy invading micro-organisms in the body. But students of the internal changes that accompany transformation in insects of other orders have often been unable to observe such devouring activity of these 'phagocytes,' and attribute the dissolution of the larval tissues to internal chemical changes. The fact that in all insect transformation a part, and in many a large part, of the larval organs pass over to the pupa and imago, suggests that only those structures whose work is done are broken down through the action of internally formed destructive substances, and one function of the phagocytes is to act as scavengers by devouring what has become effete and useless.
LARVAE AND THEIR ADAPTATIONS
Among the insects that undergo a complete transformation, there is, as we have seen in the preceding chapter, an amount of inward change, of dissolution and rebuilding of tissues, that varies in its completeness in members of different orders. It is now advisable to consider the various outward forms assumed by the larvae of these insects, or rather by a few examples chosen from a vast array of well-nigh 'infinite variety.'
In comparing the transformations of endopterygote insects of different orders, it is worthy of notice that in some cases all the members of an order have larvae remarkably constant in their main structural features, while in others there is great variety of larval form within the order. For example, the caterpillars of all Lepidoptera are fundamentally much alike, while the grubs of beetles of different families diverge widely from one another. A review of a selected series of beetle-larvae will therefore serve well to introduce this branch of the subject.
Beetles are as a rule remarkable among insects for the firm consistency of their chitinous cuticle, the various pieces (sclerites) of which are fitted together with admirable precision. In some families of beetles the larva also is furnished with a complete chitinous armour, the sclerites, both dorsal and ventral, of the successive body-segments being hard and firm, while the relatively long legs possess well-defined segments and are often spiny. Such a larva is evidently far less unlike its parent beetle than a caterpillar is unlike a butterfly. Perhaps of all beetle larvae, the woodlouse-like grub (fig. 12 b) of a carrion-beetle (Silpha) or of a semi-aquatic dascillid such as Helodes shows the least amount of difference from the typical adult, on account of the conspicuous jointed feelers. The larval glow-worm, however, is of the same woodlouse-like aspect, and in this case, where the female never acquires wings, but becomes mature in a form which does not differ markedly from that of the larva, the exceptional resemblance is closer still. In all beetle-grubs the legs are simplified, there being only one segment (a combined shin and foot) below the knee-joint, whereas in the adult there is a shin followed by five, four, or at least three distinct tarsal segments. The foot of an adult beetle bears two claws at its tip, while the larval foot in the great majority of families has only one claw. In one section of the order, however, the Adephaga comprising the predaceous terrestrial and aquatic beetles, the larval foot has, like that of the adult, two claws. Some adephagous larvae, notably those of the large carnivorous water-beetles (Dyticus), often destructive to tadpoles and young fish, have completely armoured bodies as well as long jointed legs. More commonly, as with most of the well-known Ground-beetles (Carabidae), the cuticle is less consistently hard, firm sclerites segmentally arranged alternating with considerable tracts of cuticle which remain feebly chitinised and flexible. Most of the adephagous larvae (fig. 13) have a pair of stiff processes on the ninth abdominal segment, and the insect, from its general likeness to a bristle-tail of the genus Campodea, is often called a campodeiform larva (Brauer, 1869). From such as these, a series of forms can be traced among larvae of beetles, showing an increasing divergence from the imago. The well-known wireworms—grubs of the Click-beetles (Elateridae)—that eat the roots of farm crops, have well-armoured bodies, but their shape is elongate, cylindrical, worm-like; and their legs are relatively short, the build of the insect being adapted for rapid motion through the soil. The grubs of the Chafers (Scarabaeidae) are also root-eaters, but they are less active in their habits than the wireworms, and the cuticle of their somewhat stout bodies is, for the most part, pale and flexible; only the head and legs are hard and horny. Usually an evident correspondence can be traced between the outward form of any larva and its mode of life. For example, in the family of the Leaf-beetles (Chrysomelidae) some larvae feed openly on the foliage of trees or herbs, while others burrow into the plant tissues. The exposed larvae of the Willow-beetles (Phyllodecta, fig. 14) have their somewhat abbreviated body segments protected by numerous spine-bearing, firm tubercles. But the grub of the 'Turnip Fly' (Phyllotreta) which feeds between the upper and lower skins of a leaf, or of Psylliodes chrysocephala (fig. 15), which burrows in stalks, has a pale, soft cuticle like that of a caterpillar.
In the larvae of the little timber-beetles and their allies (Ptinidae), including the 'death-watches' whose tapping in old furniture is often heard, a marked shortening of the legs and reduction in the size of the head accompany the whitening and softening of the cuticle. This shortening of the legs is still more marked in the larvae of the Longhorn Beetles (Cerambycidae) burrowing in the wood of trees or felled trunks; here the legs are reduced to small vestiges.
Finally in the large family of the Weevils (Curculionidae, fig. 16) and the Bark-beetles (Scolytidae), the grubs, eating underground root or stem structures, mining in leaves or seeds, or tunnelling beneath the bark of trees, have no legs at all, the place of these limbs being indicated only by tiny tubercles on the thoracic segments. Such larvae as these latter are examples of the type called eruciform by A.S. Packard (1898) who as well as other writers has laid stress on the series of transitional steps from the campodeiform to the eruciform type afforded by the larvae of the Coleoptera.
A fact of much importance in the transformations of beetles as pointed out by Brauer (1869) is that in a few families, the first larval instar is campodeiform, while the subsequent instars are eruciform. We may take as an example of such 'hypermetamorphosis' the life-story of the Oil or Blister-beetles (Meloidae) as first described by J.H. Fabre (1857), and later with more elaboration by H. Beauregard (1890). From the egg of one of these beetles is hatched a minute armoured larva, with long feelers, legs, and cerci, whose task is, for example, to seize hold of a bee in order that the latter may carry it, an uninvited guest, to her nest. Safely within the nest, the little 'triungulin' beetle-grub moults; the second instar has a soft cuticle and relatively shorter legs, which, as the larva, now living as a cuckoo-parasite, proceeds to gorge itself with honey, soon appear still further abbreviated. Later comes a stage during which legs are entirely wanting, the larva then resting and taking no food. The last larval instar again has short legs like the grub of the second period. In connection with this life-history we notice that the newly-hatched larva is not in the neighbourhood of its appropriate food. Hence the preliminary armoured and active instar is necessary in order to reach the feeding place; this journey accomplished, the eruciform condition is at once assumed.
In all cases indeed we may say that the particular larval form is adapted to the special conditions of life. A few examples from other orders of endopterygote insects will illustrate this point. The campodeiform type is relatively unusual, but most of the Neuroptera have larvae of this kind, active, armoured creatures with long legs, though devoid of the tail-processes often associated with similar larvae among the Coleoptera. Such are the 'Ant-lions,' larvae of the exotic lacewing flies, which hunt small insects, digging a sandy pit for their unwary steps in the case of the best-known members of the group, some of which are found as far north as Paris. In our own islands the 'Aphis-lions,' larvae of Hemerobius and Chrysopa, prowl on plants infested with 'green-fly' which they impale on their sharp grooved mandibles, sucking out the victims' juices, and then, in some cases, using the dried cuticle to furnish a clothing for their own bodies. Among these insects, while the mouth of the imago is of the normal mandibulate type adapted for eating solid food, the larval mouth is constricted and the slender mandibles are grooved for the transmission of liquid food.
Turning to eruciform types of larva, we find the caterpillar (fig. 1 b, c, d) distinguished by its elongate, usually cylindrical body with feeble cuticle, short thoracic legs and a variable number of pairs of abdominal pro-legs, universal among the moths and butterflies forming the great order Lepidoptera, and usual among the saw-flies, which belong to the Hymenoptera. The vast majority of caterpillars feed on the leaves of plants and their long worm-like bodies with the series of paired pro-legs, are excellently adapted for their habit of clinging to twigs, and crawling along shoots or the edges of leaves as they go in search of food. Of great importance to a caterpillar is its power of spinning silk, consisting of fine threads solidified from the secretion of specially modified salivary glands whose ducts open in the insect's mouth at the tip of the tubular tongue which forms a spinneret.
On the same bush caterpillars of moths and of saw-flies may often be seen feeding together. The lepidopterous caterpillar, in our countries at least, has never more than five pairs of pro-legs, situated on the third, fourth, fifth, sixth, and tenth abdominal segments; each of these pro-legs bears a number of minute hooklets, arranged in a circular or crescentic pattern, which assist the caterpillar in clinging to its food-plant. The saw-fly caterpillar, on the other hand, may have as many as eight pairs of pro-legs, the series beginning on the second abdominal segment; here, however, the pro-legs have no hooklets. Among the Lepidoptera, we notice a reduction in the number of pro-legs in the 'looper' caterpillars of Geometrid moths. Here only two pairs are present, those on the sixth and tenth abdominal segments. Consequently, as the caterpillar can cling only by the thorax and by the hinder region of the abdomen, the middle region of the body is first straightened out and then bent into an arch-like form, as the insect makes its progress by alternate movements of stretching and 'looping.'
Caterpillars, with their relatively soft bodies, feeding openly on the leaves of plants, are exposed to the attacks of many enemies, and the various ways in which they obtain protection are well worth studying. A clothing of hairs or spines is often present, and it is interesting to find that many species of our native Tiger and Eggar Moths (Arctiadae and Lasiocampidae) which pass the winter in the larval stage, have caterpillars with an especially dense hairy covering (fig. 17). Experiments have shown that hairy and spiny insects are distasteful to birds and other creatures that prey readily on smooth-skinned species, a conclusion that might well have been expected. Certain smooth caterpillars however appear to be protected by producing some nauseous secretion, which renders them unpalatable. Many of these, as the familiar cream yellow and black larva of the Magpie Moth (Abraxas grossulariata), are very conspicuously adorned, and furnish examples of what is known as 'warning coloration,' on the supposition that the gaudy aspect of such insects serves as an advertisement that they are not fit to eat, and that birds and other possible devourers thus learn to leave them alone. On the other hand, smooth caterpillars which are readily eaten by birds are usually 'protectively' coloured, so as to resemble their surroundings and remain hidden except to careful seekers. Many such caterpillars are green, the upper surface, which is naturally exposed to the light, being darker than the lower which is in shadow. When the caterpillar is large, the green area is often broken up by pale lines, longitudinal as on the larvae of many Owl Moths (Noctuidae) or oblique, as on the great caterpillars of most Hawk Moths (Sphingidae). Such an arrangement tends to make the insect less easily seen than were it to display a continuous area of the same colour. The 'looper' caterpillars mentioned above afford remarkable examples of 'protective' resemblance, for many of them show a marvellous likeness to the twigs of their food-plant, tubercles on the insect's body resembling closely the little outgrowths of the plant's cortex. It has been shown by E.B. Poulton (1892) that many caterpillars are, in their early stages, directly responsive to their surroundings as regards colour. Usually green when hatched, they remain green if kept among leaves or young shoots of plants, while they turn red, brown, or blackish if placed among twigs of these respective hues. This effect appears to be due to a direct response of the subcutaneous tissue to the rays of light reflected from the surrounding objects. The sensitiveness dies away as the caterpillar grows older, since little or no change of hue in response to a change of environment could be induced after the penultimate moult.
 The 'hairs' of an insect are not in the least comparable to the hairs of mammals, being in truth, modified portions of the cuticle, secreted by special cells.
Among those families of the Lepidoptera which are usually regarded as low in the scale of organisation, caterpillars are very generally protected by the habit of feeding in some concealed situation. For example, the great larvae of the Goat Moth (Cossus) and the whitish caterpillars of the Clearwing Moths (Sesiidae) burrow through the wood of trees, eating the timber as they go. The little irritable caterpillars of the Bell Moths (Tortricidae) roll leaves, fastening the edges together with silk, and thus make for themselves a shelter; or they bore their way into seeds or fruits, like the larva of the Codling Moth that is the cause of 'worm-eaten' apples, too well-known to orchard-keepers. Very many small caterpillars mine between the two skins of a leaf, eating out the soft green tissue, and giving rise to a characteristic blister in form of a spreading patch or a narrow sinuous track through the leaf. The caterpillars of the Clothes-moths (Tineidae) make for themselves garments out of their own excrement, the particles fastened together by silk. In such curious cylindrical cases they wander over the wool or fur, feeding and indirectly supplying themselves with clothing at the same time.
The case-forming habit of the Clothes-moth caterpillars leads us naturally to consider the similar habit adopted by their allies the Caddis-larvae which live in the waters of ponds and streams, for the Caddis-flies (Trichoptera) have much in common with the more primitive Lepidoptera. The caddis-larva is as a rule of the eruciform type, but with well-developed thoracic legs, and with hook-like tail-appendages; by means of the latter it anchors itself to the extremity of its curious 'house.' It is of interest to note that in the earlier stages of some caddises lately described and figured by A.J. Siltala (1907), the legs are relatively very long, and the larva is quite campodeiform in aspect. Some of these caddis-grubs retain the campodeiform condition and do not shelter permanently in cases, as their relations do. Different genera of caddises differ in their mode of building. Some fasten together fragments of water-weeds and plant refuse, others take tiny particles of stone, of which they make firmly compacted walls, others again lay hold of water-snail shells, which may even contain live inhabitants, and bind these into a limy rampart behind which their bodies are in safe hiding.
The silk with which the 'caddis-worms' fasten together the materials for their houses is produced from spinning-glands which like those of the Lepidoptera open into the mouth.
The survey of the various types of beetle-larvae enumerated above (pp. 50-56) concluded with a short description of the legless grub, which is the young form of a weevil or a bark-beetle. This is a larva in which the head alone has its cuticle firm and hard; the rest of the body is covered with a pale, flexible cuticle, so that the grub is often described as 'fleshy.' This type of larva is by no means confined to certain families of the beetles, it is frequently met with, in more or less modified form, in two other important orders of insects, the Hymenoptera and the Diptera. Among the Hymenoptera this is indeed the predominant larval type. We have just seen that a caterpillar is the usual form of larva among the saw-flies, but in all other families of the Hymenoptera we find the legless grub. A grub of this order may usually be distinguished from the larva of a weevil or other beetle, by its relatively smaller head and smoother, less wrinkled cuticle; it strikes the observer as a feebler, more helpless creature than a beetle-grub. And it is of interest to note that this somewhat degraded type of larva is remarkably constant through a great series of families—gall-flies, ichneumon-flies, wasps, bees (fig. 18), ants—that vary widely in the details of their structure and in their habits and mode of life. Almost without exception, however, they make in some way abundant provision for their young. The feeble, helpless, larva is in every case well sheltered and well fed; it has not to make its own way in the world, as the active armoured larva of a ground-beetle or the caterpillar of a butterfly is obliged to do.
Among those saw-flies whose larvae feed throughout life in a concealed situation, we find an interesting transition between the caterpillar and the legless grub. For example, the giant saw-flies (so called 'Wood-wasps') have larvae that burrow in timber, and these larvae possess relatively large heads, somewhat flattened bodies with pointed tail-end, and very greatly reduced legs. The feeble legless grub, characteristic of the remaining families of the Hymenoptera, is provided for in a well-nigh endless variety of ways. The female imago among these insects is furnished with an elaborate and beautifully formed ovipositor, and the act of egg-laying is usually in itself a provision for the offspring. Gall-flies pierce plant-tissues within which their grubs find shelter and food, the plant responding to the irritation due to the presence of the larva by forming a characteristic growth, the gall, pathological but often regular and shapely, in whose hollow chamber the grub lives and eats. Ichneumon-flies and their allies pierce the skin of caterpillars and other insect-larvae, laying their eggs within the victims' bodies, which their grubs proceed to devour internally. Some very small members of these families are content to lay their eggs within the eggs of larger insects, thus obtaining rich food-supply and effective protection for their tiny larvae. In Platygaster and other genera of the family Proctotrypidae, M. Ganin (1869) showed the occurrence of hypermetamorphosis somewhat like that already described as occurring among the Oil-beetles (Meloidae). The larva of Platygaster is at first rather like a small Copepod crustacean, with prominent spiny tail-processes; after a moult this form changes into the legless grub characteristic of the Hymenoptera, among which larvae even approaching the campodeiform type are very exceptional. The species of Platygaster pass their larval stages within the larvae of gall-midges.
Wasps, bees and ants, have the ovipositor of the female modified into a sting, which is often used for the purpose of providing food for the helpless grubs. Thus the digging wasps (Sphegidae and Pompilidae) hunt for caterpillars, spiders, and other creatures which they can paralyse with their stings, and bury them alongside their eggs to furnish a food-supply for the newly-hatched young. The social wasps and many ants sting and kill flies and other insects, which they break up so as to feed their grubs within the nest. It is well known that the labour of tending the larvae in these insect societies is performed for the most part not by the mother ('Queen') but by the modified infertile females or 'workers.' Other ants and the bees feed their grubs (fig. 18), also sheltered in well-constructed nests, on honey elaborated from nectar within their own digestive canals. In all cases we see that the helplessness of the grub is associated with some kind of parental care.
From the Hymenoptera we may pass on to the Diptera or Two-winged Flies, an order of which the vast number of species and in many cases the myriads of individuals force themselves on the observer's notice. F. Brauer (1863) divided the Diptera into two sub-orders; of the first of these a Crane-fly or 'Daddy-long-legs' may be taken as typical, of the second an ordinary House-fly or Bluebottle. All the larvae of the Diptera are legless, those of the Crane-fly group have well-developed hard heads, with biting mandibles, but in the House-fly section the larva is of the degraded vermiculiform type known as the maggot, not only legless, but without a definite head, the front end of the creature usually tapering to the mouth, where there are a pair of strong hooks, used for tearing up the food. A few examples of each of these types must suffice in the present brief survey. A few pages back (p. 66) reference was made to the production of galls on various plants, through the activity of larvae of the hymenopterous family Cynipidae. Many plant-galls are due, however, to the presence of grubs of tiny dipterous insects, the Cecidomyidae or Gall-midges. A cecid grub (fig. 19) has an elongate body with flexible, wrinkled cuticle, tapering somewhat at the two ends. The head, if rather narrow, is distinct, and beneath the prothorax is a characteristic sclerite known as the 'anchor process' or 'breast bone.' Along either side of the body is a series of paired spiracles, each usually situated at the tip of a little tubular outgrowth of the cuticle; the hindmost spiracles are often larger than the others. These little grubs live in family communities, their presence leading to some deformation of the plant that serves to shelter them. A shrivelled fruit or an arrested and swollen shoot, such as may be due respectively to the Pear-midge (Diplosis pyrivora) or the Osier-midge (Rhabdophaga heterobia), is a frequent result of the irritation set up by these little grubs. In a larva of the crane-fly family (Tipulidae, fig. 20) living underground and eating plant-roots, like the well-known 'leather-jacket' grubs of the large 'Daddy-long-legs' (Tipula) or burrowing into a rotting turnip or swollen fungus, like the more slender grub of a 'Winter Gnat' (Trichocera), the student notices a somewhat tough cuticle, a relatively small but distinct head, and frequently prominent finger-like processes on the tail-segment. Further examination shows a striking modification in the arrangement of the spiracles. Instead of a paired series on most of the body-segments, as in caterpillars and the vast majority of insects whether larval or adult, there are two large spiracles surrounded by the prominent tail-processes, and a pair of very small ones on the prothorax, the latter possibly closed up and useless. This restriction of the breathing-holes to a front and hind pair (amphipneustic condition) or to a hind pair only (metapneustic type) is highly characteristic of the larvae of Two-winged flies.
 Known as the Orthorrhapha and the Cyclorrhapha; these terms are derived from the manner in which the larval or pupal cuticle splits, as will be explained in the next chapter (p. 88).
Turning now to the maggot, characteristic of the House-fly section (fig. 21) of the Diptera, we see the greatest contrast between the larva and the imago that can be found throughout the whole class of the insects. The Bluebottle's eggs, the well-known 'fly blow' laid in summer time on exposed meat, not unnaturally arouse feelings of disgust, yet they are the prelude to one of the most marvellous of all insect life-stories. The fly—with its large globular head, bearing the extensive compound eyes, the highly modified feelers with their exquisitely feathered slender sensory tips, and the complex suctorial jaws; with its compact thorax bearing the glassy fore-wings alone used for flight, though the hind-wings modified into tiny drumstick-like 'halters' are the organs of a fine equilibrating sense—is perhaps the most specialised, structurally the 'highest' of all insects. Yet in a week or two this swift, alert, winged creature is developed from the degraded maggot, white, legless, headless, that buries itself in putrid flesh, 'feeding on corruption.'