MnJOimr LUBBOCK /iU~->.RR,S. OF INSECTS ORIGIN AND METAMORPHOSES OF INSECTS. NATURE SERIES. ON THE ORIGIN AND METAMORPHOSES OF INSECTS.. BY SIR JOHN LUBBOCK, Bart., M.P., F.R.S. VICE-CHANCELLOR OF THE UNIVERSITY OF LONDON, ETC. ETC. IVITI/ NUMER 0 US ILL USTRA TIONS. MACMILLAN AND CO. 1874. Ri^ht of Translation and Rcproduciioii is reservcd.'\ LONDON : R. CLAY, SONS, AND TAYLOR, PRINTERS, BREAD STREET HILL, PREFACE. For some years, much of my leisure time has been devoted to the study of the anatomy, develop- ment, and habits of the Annulosa, and especially of Insects, on which subjects I have published various memoirs, chiefly in the Transactions of the Royal, Linnaean, and Entomological Societies : of these papers I subjoin a list. Although the details, of which these memoirs necessarily for the most part consist, offer little interest, excepting to those per- sons who are specially devoted to Entomology, still there are portions which, having reference to the nature of metamorphoses and to the origin of insects, are of a more general character. I have also briefly referred to these questions in a Molio- graph of the Collembola and Thysanura, recently published by the Ray Society, and in the Opening Address to the Biological Section of the British Association at Brighton in 1872. Under these Vlll PREFACE. circumstances, it has been suggested to me that a small volume, containing, at somewhat greater length, in a more accessible form, and with the advantage of illustrations, the conclusions to which I have been led on this interesting subject, might not be alto- gether without interest to the general reader. The result, which has already appeared in the pages of Nature, is now submitted to the public, with some additions. I am well aware that it has no pretence to be in any sense a complete treatise ; that the subject itself is one as to which our knowledge is still very incomplete, and on which the highest authorities are much divided in opinion. Whatever differences of opinion, however, there may be as to the views here put forward, the facts on which they are based will, I believe, be found correct. On this point I speak with the more confidence, on account of the valuable assistance I have received from many friends : to Mr. and Mrs. Busk and Dr. Hooker I am especially indebted. The papers above referred to are as follows : — 1. On Labidocera. — Annals and Magazine of Natural History, vol. xi., 1853. 2, On Two New Sub-genera of Calanidije. — Annals and Magazine of Natural History, vol. xii., 1853. PREFACE. 3. On Two New Species of Calanidse. — Annals and Magazine of Natural History, vol. xii., No. Ixvii., 1853. 4. On Two New Species of Calanidce. — Annals and Magazine of Natural History, vol. xii.. No. Ixix. 1853. . 5. On some Arctic Calanidoe.- — Annals and Magazine of Natural History, 1854. 6. On the Freshwater Entomostraca of South America. — Transac- tions of the Entomological Society, vol. iii., 1855. 7. On some New Entomostraca. — Transactions of the Entomological Society, vol. iv. , 1856. 8. On some Marine Entomostraca found at Weymouth. — Annals and Magazine of Natural History, vol. xx., 1857. 9.. On the Respiration of Insects. — Entomological Annual, 1857. 10. An Account of the Two Methods of Reproduction in Daphnia. — Transactions of the Royal Society, 1857. 11. On the Ova and Pseudova of Insects. — Transactions of the Royal Society, 1858. 12. On the Arrangement of the Cutaneous Muscles of Pygsera Eu- cephala. — Linnsean Society's Transactions, vol, xxii., 1858. 13. On the Freshwater Entotnostraca of South America. — Entomo- logical Society's Transactions, 1858. 14. On Coccus Hesperidum. — Royal Society Proceedings, vol. ix. 1858. 15. On the Distribution of Tracheae in Insects. — Linnsean Society's Transactions, vol. xxiii., i860. 16. On the Generative Organs and on the Formation of the Egg in Annulosa. Transactions of the Royal Society, 1861. 17. On Sphaerularia Bombi. — Natural History Review, 1861. 18. On some Oceanic. Entomostraca. — Linnsean Society's Transactions, vol. xxiii., i860. 19. On the Thysanura. Part I. — Linnaean Society's Transactions, 1862. 20. On the Development of Lonchoptera. — Entomological Society's Transactions, 1862. 21. On the Thysanura. Part 2. — Linnsean Society's Transactions, 1862. 22. On the Development of Chloeon. Part i. — Linnsean Society's Transactions, 1863. 23. On Two Aquatic Hymenoptera. — Linnsean Society's Transactions, 1863. 24. On some little-known Species of Freshwater Entomostraca. — Linnsean Society's Transactions, vol. xxiv. , 1863. 25. On Sphserularia Bombi. — Natural History Review, 1864. PREFACE. 26. On the Development of Chloeon. Part 2. — Linnsean Society's Transactions, 1865. 27. On the Metamorphoses of Insects, — Journal of the Royal Institu- tion, 1866. 28. On Pauropus.— Linnsean Society's Transactions, 1866. 29. Notes on the Thysanura. Part 3. — Linngean Society's Transactions, vol. xxiv., 1867. 30. Address to the Entomological Society. — Entomological Society's Transactions, 1867. 31. On the Larva of Micropeplus Staphilinoides. — Entomological So- ciety's Transactions, 1868. 32. Notes on the Thysanura. Part 4.— Linngean Society's Transactions, 1869. 33. Addresses to the Entomological Society. — Entomological Society's Transactions, 1867- 1868. 34. On the Origin of Insects. — ^Journal of the Linnsean Society, vol. xi. 35. Opening Address to the Biological Section of the British Associa- tion.— British Association Report, 1872. CONTENTS. CHAPTER I. THE CLASSIFICATION OF INSECTS.^ Introduction. — Stages in the Life of an Insect. — Classification of Insects. — Characters derived from the Wings ; from the parts of the Mouth ; from the Metamorphoses. — The Classes of Insects : Hymenoptera, Strepsiptera, Coleoptera, Euplexoptera, Orthoptera, Thysanoptera, Neuroptera, Trichoptera, Diptera, Aphaniptera, Heteroptera, Homoptera, Lepidoptera page i — 26 CHAPTER II. THE INFLUENCE OF EXTERNAL CONDITIONS ON THE FORM AND STRUCTURE OF LARV^. Larvae depend partly on the group to w^hich they belong. — Wood-eating Larvse. — Larvae of Lamellicorns. — Larvae depend also in part on mode of life.- -Larvae of Hymenoptera, of Sirex ; of Tenthredo ; of Ichneumons; of Bees. — Rudimentary legs of Bee Embryo. — Beetles, Weevils, Scolyhts, Crioceris, Sitaris, Metamorphoses of Pteromalidae. Platygaster, Folynema. — Influence of external conditions. — Deve- lopmental and adaptive Metamorphoses Page 27 — 41 CHAPTER HI. THE NATURE OF METAMORPHOSES. The life history of an Insect must be considered as a whole. — Vagueness of the term Larva. — Some larvae much more advanced than others. — Organs develope in different order, in different groups. — Sup- pressed stages. — Apod condition of Phryganea ; of Aphis; of Xll CONTENTS. Chrysopa. — Libellulidae. — Donacia. — Spiders. — Myriapods. — Apod stage of Homomorphous Insects once probably longer than now. — Suppression of embryonic stages. — Metamorphoses of Hydroida, Crustacea, Isopods, and Amphipods. — Echinoderms. — Variations in development induced by the influence of external conditions. page 41 — 62 CHAPTER IV. THE ORIGIN OF METAMORPHOSES. Origin of Metamorphoses. — Views of Messrs. Kirby and Spence. — Nature of the question. — Young animals often more similar than mature. — Views of Darwin, Herbert Spencer, Johannes Miiller, Fritz Miiller, and Agassiz. — Effect of size of egg. — Insects leave the egg in a more or less developed condition. — Consideration of pupal condition. — Quiescence of pupa. — Period of quiescence at each moult. — Changes not so abrupt as generally supposed. — Change in mouth-parts. — Difficulty in reference to Darwinian theory. — Mouth- parts of Campodea and Collembola, as intermediate between the mandibulate and haustellate types. — Change in mouth-parts as con- nected with pupal conditions. — Origin of wings. — Use of wings under water. — Connection of metamorphoses with alternation of generations.— Parthenogenetic larvae of Cecidomyia. — In alternation of generations one form always agamic. — Dimorphism andDieidism. — Summary and Conclusions page 62 — 81 CHAPTER V. THE ORIGIN OF INSECTS. The Origin of Insects. — Mistaken views of Darwinian theory. — Natural 's,€i&z\XQ)VL vera catisa. — Application of Darwin's views to Insects. — Similarity of young Crustacea as compared with mature forms ; ditto in Insects. — Type of Insecta. — Two principal types of larvse : Hexapod and Apod. — Conclusions to be drawn from them. — Cam- podea the modern representative of the Insect-stock. — Campodea, perhaps derived from Tardigrade. — Vermiform or Apod type of larva. — Views of Fritz Miiller, Brauer, and Packard. — Represents a still earlier ancestor. — Modern representatives. — Notommata, Albert ia, Liiidia.—KoxMex forms difficult to trace. — Lowest forms of animal life. — Yolk-segmentation. — Embryology and Evolution. — Light thrown by the evolution of the individual on that of the species page 82 — 108 I DESCRIPTION OF THE PLATES. ' PLATE I. p. 7. FIG. 1. Cricket. Westwood, Intro, to the Modern Classification of Insects, vol. i. p. 440. 2. Earwig, Westwood, loc. cit. vol. i. p. 399. 3. Aphis. Packard, Guide to the Study of Insects, pp. 521, 522. 4. Scolytus. Westwood, loc. cit. vol. i. p. 350. 5. Anthrax. Westwood, loc. cit. vol. ii. p. 538. 6. Balaninus. 7. Cynips. Westwood, loc. cit. vol. ii. p. 121. 8. Ant (Formica). Westwood, loc. cit. vol. ii. p. 218. 9. Wasp. Ormerod, Nat. Hist, of Wasps, pi. i. fig. i. PLATE II. p. 8. FIG. 1. Larva of Cricket. Westwood, loc. cit. vol. i. p. 440. 2. Larva of Aphis. Packard, loc. cit pp. 521, 522. 3. Larva of Earwig. Westwood, loc. cit. vol. i. p. 399. 4. Larva of Scolytus. Westwood, loc. cit. vol. i. p. 350. 5. Larva of Anthrax. Westwood, loc. cit. vol. ii. p. 546. 6. Larva of Balaninus. 7. Larva of Cynips. Westwood, loc. cit. vol. ii. p. 121. 8. Larva of Ant (Formica). Westwood, loc. cit. vol. ii. p. 226. 9. Larva of Wasp. Newport, Art. Insecta, Todd's Cycl. Anat. and Phys., p. 871. PLATE III. p. 14. FIG. 1. Chloeon. Linn. Trans. 1866. 2. Meloe. Spry and Shuckard, Coleoptera Delineated, pi. 56. 3. Calepteryx. 4. Sitaris. Spry and Shuckard, loc. cit. pi. 56. 5. Campodea. Suites a BufFon. Apteres. xiv DESCRIPTION OF THE PLATES. FIG. 6. Acilius. Westwood, loc. cit. vol. i. p. lOO. 7. Termes. Westwood, loc. cit. vol. ii. p. 12. 8. Stylops. Duncan, Met. of Insects, p. 387 ; Packard, p. 482. 9. Thrips. Westwood, loc. cit. vol. ii. p. i. PLATE IV. p. 15. G. 1. Larva of Chloeon. Linn. Trans. 1863. 2. Larvla of Meloe. Chapuis and Candeze, Mem. Soc. Roy. Liege, 1853, pp. I, 7- 3. Larva of Calepteryx. Dufour, Ann. Sci. Nat. 1852. 4. Larva of Sitaris. Duncan, Met. of Insects, p. 309. 5. Larva of Campodea. Gervais' Suites a Buffon. Apteres. 6. Larva of Acilius. Westwood, loc. cit. vol. i. p. 100, 7. Larva of Termes. Duncan, loc. cit. p. 348. 8. Larva of Stylops. Westwood, Trans, Ent. Soc. 1839, vol. ii. pi. XV. fig. 13a. 9. Larva of Thrips. Westwood, loc. cit. vol, ii. p. i. PLATE V. p. 99. I — 5. Protamoeba, 6 — 9. Protamyxa Aurantiaca. Haeckel Beit. zur. Monog. der Mo- neren, pi. i. 10 — 18. Magosphoera planula. Haeckel, loc. cit. pi. v. PLATE VI. p. 105. FIG. I — 4. Yolk-seginentation in Laomedea. .^fter Allman. Mon. of Tubularian Ilydroids. Ray Society. 5 — 9. Yolk-segmentation in Filaria. After Van Beneden. Mem. sur les Vers Intestinaux. 10—13. Yolk-segmentation in Echinus. After Derbes. Ann des. Sci. Nat. 1847. 14 — 17. Yolk-segmentation in Lacinularia. After Huxley. J. of Mici Sci. 1853. 18 — 21. Yolk- segmentation in Purpura. After Koren and Danielssen. Ann. des. Sci. Nat. 1853. 22 — 25. Yolk-segmentation in Amphioxus. After Haeckel. Naturliche Schopfungsgeschichte, pi. x. 28 — 29, Yolk-segmentation in Vertebrate. After Allen Thompson Art. Ovum. Cyclop, of Anatomy and Physiology. DESCRIPTION OF THE FIGURES. 1. Larva of the Cockchafer (Melolontha) 2. Larva of Cetonia. 3. Larva of Trox. 4. Larva of Oryctes. 5. Larva of Aphodius. 6. Larva of Lucanus. 7. Larva of Brachytarsus. 8. Larva of Crioceris. 9. Larva of Sitaris humeralis. 10. Larva of Sitaris humeralis, in the second stage. I r. Larva of Sitaris humerahs, in the third stage. 12. Larva of Sitaris humerahs, in the fourth stage. 13. Pupa of Sitaris. 14. Larva of Sirex. 15. Egg of Rhynchites, showing the parasitic larva. 16. The parasitic larva, more magnified. 17. Egg of Platygaster. 18. Egg of Platygaster, showing the central cell. 19. Egg of Platygaster, after the division of the central cell. 20. Egg of Platygaster, more advanced. 21. Egg of Platygaster, more advanced. 22. Egg of Platygaster, showing the rudiment of the embryo. 23. Larva of Platygaster. — mo, mouth; a, antenna; kf, hooked feet; r, toothed process ; Ifg, lateral process ; f, branches of the tail. 24. Larva of another species of Platygaster, (The letters indicate the same parts as in the preceding figure.) 25. Larva of a third species of Platygaster. (The letters indicate the same parts as in the preceding figure. ) 26. Larva of Platygaster in the second stage. — -mo, mouth ; s.kj, oesophagus ; gsae^ supra-oesophagal ganglion : muscles ; bsm, nervous system ; gagh, rudiments of the reproductive glands. 27. Larva of Platygaster in the third stage.— wc, mouth; md, man- dibles ; gsae, supra-oesophagal ganglion ; slk, oesophagus ; ag, ducts of the salivary glands ; bnm, ventral nervous system ; sp, salivary glands ; msl, stomach ; im^ imaginal discs ; tr, trachece ; //', fatty tissue ; ed, intestine ; ga, rudiments of reproductive organs ; rM, wider portion of intestine ; ao, posterior opening. 28. Embryo of Polynema. 29. Larva of Polynema. — <7i'^-//, rudiments of the antennae; fisch, of the wings ; bsch, of the legs ; vfg, lateral projections ; gsch^ rudiments of the ovipositor ; fk, fatty tissue. DESCRIPTION OF THE FIGURES. FIG. 30. Egg of Phryganea (Mystacides). — mandibular segment ; O-O, maxillary, labial, and three thoracic segments ; Z>, abdomen. 31. Egg of Phryganea somewhat more t'.dvanced, — b, mandibles; c, maxillae ; cfs, rudiments of the three pairs of legs. 32. Egg of Pholcus opilionides, showing the Protozonites. 33. Embryo of Julus. 34. Colony of Bougainvillea fruticosa, natural size, attached to the under- side of a pi-ce of floating timber. 35. Portion of the same, more magnified. 36. The Medusa from the same species. 37. Larva of Prawn, Nauplius stage. 38. Larva of Prawn, more advanced, Zoea stage. 39. Larva of Echino-cidaris oequituberculata seen from above X to-- 40. Larva of Echinus X 100. — A, front arm ; arms of the mouth- process ; posterior side arm ; E^, accessory arm of the mouth- process ; a, mouth ; a}-, oesophagus ; stomach ; intestine ; posterior orifice ; d, ciliated bands ; ciliated epaulets ; disc of future Echinus, 41. Comatula rosacea. 42. I^arva of Comatula rosacea. 43. Larva of Comatula rosacea, more advanced. 44. Larva of Comatula rosacea, in the Pentacrinus state. 45. Larva of Starfish (Bipinnaria), X 100. 46. Larva of Starfish (Bipinnaria), X 100, seen from the side. — mouth ; b, oesophagus ; c, stomach ; r^, intestine. 47. Larva of another Bipinnaria, showing the commencement of the Starfish. — g, canal of the ciliated sac ; i, rudiments of tentacles ; d. ciliated band. 48. Larva of Moth (Agrotis). 49. Larva of Beetle (Haltica). 50. Larva of Saw-fly (Cimbex). 51. Larva of Julus. 52. Agrotis sufifusa. 53. Haltica. 54. Cimbex. 55. Julus. 56. Tardigrade. 57. Larva of Cecidomyia. 58. Lindia torulosa. 59. Prorhynchus stagnalis. 60. Egg of Tardigrade. 61. Egg of Tardigrade, after the yolk has subdivided. 62. Egg of Tardigrade, in the next stage. 63. Egg of Tardigrade, more advanced. ON THE ORIGIN AND METAMORPHOSES OF INSECTS. CHAPTER I. THE CLASSIFICA TION OF INSECTS. About forty years ago the civil and ecclesiastical authorities of St. Fernando in Chili arrested a certain M. Renous on a charge of witchcraft, because he kept some caterpillars which turned into butterflies.^ This was no doubt an extreme case of ignorance ; it is now almost universally known that the great majority of insects quit the egg in a state very different from that which they ultimately assume ; and the general statement in works on entomology has been that the life of an insect may be divided into four periods. Thus, according to Kirby and Spence,^ " The states through which insects pass are four : the ego^^ the larva, the pupa, and the imagoT Burmeister,^ also, 1 ,Darwin's " Researches into the Geology and Natural History of the Countries visited by H.M.S. Beagle,'" p. 326. ^ Introduction to Entomology, vi. p. 50. 3 Manual of Entomology, p. 30, ON THE ORIGIN AND [chap. says that, excluding certain very rare anomalies, *'we may observe four distinct periods of existence in every insect, — namely, those of the egg, the larva, the pupa, and the imago, or perfect insect." In fact, however, the various groups of insects differ widely from one another in the metamorphoses they pass through : in some, as in the grasshoppers and crickets, the changes consist principally in a gradual increase of size, and in the acquisition of wings ; while others, as for instance the common fly, acquire their full bulk in a form very different from that which they ultimately assume, and pass through a period of in- action in which not only is the whole form of the body altered, not only are legs and wings acquired, but even the internal organs themselves are almost entirely disintegrated and re-formed. It will be my object, after having briefly described these changes, to throw some light on the causes to which they are due, and on the indications they afford of the stages through which insects have been evolved. The following list gives the orders or principal groups into which the Class Insecta may be divided. I will not, indeed, here enter upon my own views, but will adopt the system given by Mr. Westwood in his excellent Introduction to the Modern Classification of Insects," from which also, as a standard authority, most of the figures on Plates I. to IV., when not other- wise acknowledged, have been taken. He divides insects into thirteen groups, and with reference to eight of them it may be said that there is little difference of opinion among entomologists. These orders are by far the most numerous, and I have METAMORPHOSES OE INSECTS. placed them in capital letters. As regards the other five there is still much difference of opinion. It must also be observed that Prof. Westwood omits the parasitic Anoplura, as well as the Thysanura and Collembola. ORDERS OF INSECTS ACCORDING TO WESTWOOD. Hymenoptera Strepsiptera COLEOPTERA Euplexoptera Orthoptera Thysanoptera Neurottera Trichoptera DiPTERA . Aphaniptera Heteroptera H)M0PTERA Lepidoptera Bees, Wasps, Ants, &c. Stylops, Zenos, &c. Beetles. Iiarwigs. Cirasshoppers, Crickets, Cockroaches, &c. Thrips. Ephemeras, &c. rhryganea. Flies and Gnatb. Fleas. Rugs. Aphis, Coccus, &c. Butterflies and Moths. Of these thirteen orders, the eight which I have placed in capital letters — namely the first, third, fifth, seventh, ninth, eleventh, twelfth, and thirteenth, are much the most important in the number and variety of their species ; the other five form comparatively small groups. The Strepsiptera are minute insects, parasitic on Hymenoptera : Rossi, by whom they were discovered, regarded them as Hymenopterous ; Lamarck placed them among the Diptera ; by others they have been considered to be most closely allied to the Coleoptera, but they are now generally treated as an independent order. The Euplexoptera or Earwigs are only too familiar to most of us. Linnceus classed them among the B 2 4 ON THE ORIGIN AND [chap. Coleoptera, from which, however, they differ in their transformations. Fabricius, Olivier, and Latreille regarded them as Orthoptera ; but Dr. Leach, on account of the structure of their wings, considered them as forming the type of a distinct order, in which view he has been followed by Westwood, Kirby, and m.any other entomologists. The Thysanoptera, consisting of the Linnaean genus Thrips, are minute insects well known to gardeners, differing from the Coleoptera in the nature of their metamorphoses, in which they resemble the Orthop- tera and Hemiptera. The structure of the wings and mouth-parts, however, are considered to exclude them from these two orders. The Trichoptera, or Caddis worms, offer many points of resemblance to the Neuroptera, while in others they approach more nearly to the Lepidoptera. According to Westwood, the genus Phryganea " forms the connecting link between the Neuroptera and Lepi- doptera." The last of these small aberrant orders is that of the Aphaniptera, constituted for the family Pulicidae. In their transformations, as in many other respects, they closely resemble the Diptera. Strauss Durck- heim indeed said that la puce est im diptere sa7is ailer " Westwood, however, regards it as consti- tuting; a separate order. As indicated by the names of these orders, the structure of the wings affords extremely natural and convenient characters by which the various groups may be distinguished from one another. The mouth- parts also are very important ; and, regarded from 1.] METAMORPHOSES OF INSECTS. 5 this point of view, the Insecta have been divided into two series — the Mandibulata and Haustellata, or mandibulate and suctorial groups, between which, as I have elsewhere shown,^ the Collembola (Podura, Smynthurus, &c.) occupy an intermediate position. These two series are : — Mandibulata. Haustellata. Hymenoptera, Lepidoptera. Strepsiptera. Diptera. Coleoptera. Aphaniptera. Euplexoptera, Hemiptera. Orthoptera. Homoptera. Trlchoptera? Thysanoptera ? Again — and this is the most important from my present point of view — insects have sometimes been divided into two other series, according to the nature of their metamorphoses: "Heteromorpha," to use the terminology of Prof. Westwood,^ " or those in which there is no resemblance between the parent and the offspring ; and Homomorpha, or those in which the larva resembles the imago, except in the absence of wings. In the former the larva is generally worm- like, of a soft and fleshy consistence, and furnished with a mouth, and often with six short legs attached in pairs to the three segments succeeding the head. In the Homomorpha, including the Orthoptera, Hemiptera, Homoptera, and certain Neuroptera, the body, legs, and antennae are nearly similar in their form to those of the perfect insect, but the wings are wanting," 1 Linnean Journal, vol. xi. 2 Introduction to the Modern Classification of Insects, p. 17. 6 ON THE ORIGIN AND [chap. Heteromorpha. HOMOMORPHA. Hymenoptera. Strepsiptera, Coleoptera. Trichoptera, Diptera. Thysanoptera. Orthoptera. Hemiptera. Homopteia. Euplexoptera. Aphaniptera. Lepidoptera. Neuroptera. But though the Homomorphic insects do not pass through such striking changes of form as the Heteramorphic, and are active throughout life, still it was until within the last few years generally (though erroneously) considered, that in them, as in the Heteromorpha, the life fell into four distinct periods ; those of (i) the egg, (2) the larva, charac- terized by the absence of wings, (3) the pupa with imperfect wings, and (4) the imago, or perfect insect I have, however, elsewhere ^ shown that there are not, as a matter of fact, four well-marked stages, and four only, but that in many cases the process is much more gradual. The species belonging to the order Hymenoptera are among the most interesting of insects. To this order belong the gallflies, the sawflies, the ichneu- mons, and, above all, the ants and bees. We are accustomed to class the Anthropoid apes next to man in the scale of creation, but if we were to judge animals by their works, the chimpanzee and the gorilla must certainly give place to the bee and the ant. The larvae of the sawflies, which live on leaves, and of the Siricidse or long-tailed wasps, which feed on wood, are very much like caterpillars, i Linnean Transactions, 1863 — "On the Development of Chloeon." METAMORPHOSES OF INSECTS. 7 having three pairs of legs, and in the former case abdominal pro-legs as well : but in the great majority of Hymenoptera the larvae are legless, fleshy grubs (Plate II., Figs. 7-9) ; and the various modes by which the females provide for, or secure to, them a sufficient supply of appropriate nourishment constitutes one of the most interesting pages of Natural History. The species of Hymenoptera are very numerous ; in this country alone there are about 3,000 kinds, most of which are very small. In the pupa state they are inactive, and show distinctly all the limbs of the perfect insect, encased in distinct sheaths, and folded on the breast. In the perfect state they are highly organized and very active. The working ants and some few species are wingless, but the great majority have four strong membranous wings, a cha- racter distinguishing them at once from the true flies, which have only one pair of wings. The sawflies are so called because they possess at the end of the body a curious organ, corresponding to the sting of a wasp, but which is in the form of a fine-toothed saw. With this instrument the female sawfly cuts a slit in the stem or leaf of a plant, into v/hich she introduces her egg. The larva much re- sembles a caterpillar, both in form and habits. To this group belongs the nigger, or black caterpillar of the turnip, which is often in sufficient numbers to do much mischief. Some species make galls, but the greater number of galls are formed by insects of another family, the Cynipidse. In the Cynipidse (Plate I., Fig. 7) the female is provided with an organ corresponding to the saw of PLATE IT.— LARV^ OF THE INSECTS REPRESENTED ON PLATE L Fig:, r, Larva of Cricket: 2, Larva of Aphis; 3, Larva of Earwi,s:; 4, Larva of Scolvtus (Beetle): 5, La-va of Anthrax (Fly); 6, T arva of Balaninus (Nut Weevil) ; 7, Larva of Cynips , 8, Larva of Ant ; 9, Larva of Wasp. lO ON THE ORIGIN AND [chap. the sawfly, but resembling a needle. With this she stings or punctures the surface of leaves, buds, stalks, or even roots of various plants. In the wound thus produced she lays one or more eggs. The effects of this proceeding, and particularly of the irritating fluid which she injects into the wound, is to produce a tumour or gall, within which the egg hatches, and on which the larva, a thick fleshy grub (Plate II., Fig. 7), feeds. In some species each gall contains a single larva ; in others, several live together. The oak supports several kinds of gallflies : one produces the well-known oak-apple, one a small swelling on the leaf, resembling a currant, another a gall some- what like an acorn, another attacks the root ; the species making the bullet-like galls, which are now so common, has only existed for a few years in this country; the beautiful little spangles so common in autumn on the under side of oak-leaves are the work of another species, the Cynips longipennis. When the larva is full-grown, it eats through the gall, falls to the earth, and turns into a chrysalis. One curious point about this group is, that in some of the commonest species the females alone are known, no one yet having ever succeeded in finding a male. Another great family of the Hymenoptera is that of the ichneumons ; the females lay their eggs either in or on other insects, within the bodies of which the larvae live. These larvae are thick, fleshy, legless grubs, and feed on the fatty tissues of their hosts, but do not attack the vital organs. When full- grown, the grubs eat their way through the skin of I.J METAMORPHOSES OF INSECTS. II the insect, and turn into chrysalides. Ahiiost every kind of insect is subject to the attacks of these Httle creatures, which are no doubt useful in preventing the too great multiplication of insects, and especially of caterpillars. Some species are so minute that they actually lay their eggs within those of other insects (Figs. 15, 16). These parasites assume very curious forms in their larval state. But of all the Hymenoptera, the group containing the ant, the bee, and the wasp is the most interesting. This is especially the case with the social species, though the solitary ones also are extremely remark- able. The solitary bee or wasp, for instance, forms a cell generally in the ground, places in it a sufficient amount of food, lays an egg, and closes the ceil. In the case of bees, the food consists of honey ; in that of wasps, the larva requires animal food, and the mother therefore places a certain number of insects in the cell, each species having its own special prey, some selecting small caterpillars, some beetles, some spiders. Cerceri's biipresticida, as its name denotes, attacks beetles belonging to the genus Bu- prestis. Now if the Cerceris were to kill the beetle before placing it in the cell, it would decay, and the young larva, when hatched, would find only a mass of corruption. On the other hand, if the beetle were buried uninjured, in its struggles to escape it would be almost certain to destroy the egg. The wasp has, however, the instinct of stinging its prey in the centre of the nervous system, thus depriving it of motion, and let us hope of suffering, but not of life ; consequently, when the young larva leaves 12 ON THE ORIGIN AND [chap. the egg, it finds ready a sufficient store of whole- some food. Other wasps are social, and, like the bees and ants, dwell together in communities. They live for one season, dying in autumn, except some of the females, which hibernate, awake in the spring, and form new colonies. These, however, do not, under ordinary circumstances, live through a second winter. One specimen which I kept tame through last spring and summer, lived until the end of February, but then died. The larvae of wasps (Plate II., Fig. 9) are fat, fleshy, legless grubs. When full-grown they spin for themselves a silken covering, within which they turn into chrysalides. The oval bodies which are so nu- merous in ants' nests, and which are generally called ants' eggs, are really not eggs but cocoons. Ants are very fond of the honey-dew which is formed by the Aphides, and have been seen to tap the Aphides with their ante'nnae, as if to induce them to emit some of the sweet secretion. There is a species of Aphis which lives on the roots of grass, and some ants collect these into their nests, keeping them, in fact, just as we do cows. One species of red ant' does no work for itself, but makes slaves of a black kind, which then do everything for their masters. Ants also keep a variety of beetles and other insects in their nests. That they have some reason for this seems clear, because they readily attack any un- welcome intruder ; but what that reason is, we do not yet know. If these insects are to be regarded as the domestic animals of the ants, then we must admit that the ants possess more domestic animals than we do. I.] METAMORPHOSES OF INSECTS. 13 But it has not been shown that the beetles produce any secretion of use to the ants; and yet there are some remarkable species, rarely, if ever, found, ex- cepting in ants' nests, which are blind and apparently helpless, and which the ants tend with much care. M. Lespes, who regards these blind beetles as true domestic animals, has recorded ^ some interest- ing observations on the relations between one of them {Claviger Duvalii) and the ants (Lashis nigcr) with which it lives. This species of Claviger is never met with except in ants' nests, though on the other hand there are many communities of Lasius which possess none of these beetles ; and M. Lespes found that when he placed Clavigers in a nest of ants which had none of their own, the beetles were immediately killed and eaten, the ants themselves being on the other hand kindly received by other communities of the same species. He concludes from these obser- vations that some communities of ants are more ad- vanced in civilization than others : the suggestion is no doubt ingenious, and the fact curiously resembles the experience of navigators who have endeavoured to introduce domestic animals among barbarous tribes ; but M. Lespes has not yet, so far as I am aware, published the details of his observations, without which it is impossible to form a decided opinion. I have sometimes wondered whether the ants have any feeling of reverence for these beetles ; but the whole subject is as yet very obscure, and would well repay careful study. 1 "Surla Domestication des Clavigers par les Fourmis." Bull, de la Soc, d' Anthropologic de Paris, 1868, p. 315. PLATE III.— MATURE INSECTS, g. I. Chloeon ; 2, Meloe (after Shuckard) ; 3, Cabpteryx ; 4 Sitaris (after Shuck- ard) ; 5, Campodsa (after Gervais) ; 6, Acilius ; 7, Termes ; 8, Stylops (female); g, Thrips. i6 [chap. The order Strepsiptera are a small, but very re- markable group of insects, parasitic on bees and wasps. The larva (PI. IV., Fig. 8) is minute, six-legged, and very active ; it passes through its transformations within the body of the bee or wasp. The male and female are very dissimilar. The males are minute, very active, short-lived, and excitable, with one pair of large membranous wings. The females (PL III., Fig. 8), on the contrary, are almost motionless, and shaped very much like a bottle ; they never quit the body of the bee, but only thrust out the top of the bottle between the abdominal rings of the bee. In the order Coleoptera, the larvae differ very much in form. The majority are elongated, active, hexapod, and more or less depressed ; but those of the Weevils (PI. II., Fig. 6), of Scolytus (PI. IL, Fig. 4), &c., which are vegetable feeders, and live surrounded by their food, — as, for instance, ifi grain, nuts, &c., — are apod, white, fleshy grubs, not unlike those of bees and ants. The larvae of the Longicorns, which live inside trees, are long, soft, and fleshy, with six short legs. The Geodephaga, corresponding with the Linnsean genera Cicindela and Carabus, have six-legged, slender, car- nivorous larvae ; those of Cicindela, which waylay their prey, being less active than the hunting larvae of the Carabidae. The Hydradephaga, or water-beetles (Dyticidae and Gyrinidae), have long and narrow larvae (PL IV., Fig. 6), with strong sickle-shaped jaws, short antennae, four palpi, and six small eyes on each side of the head ; they are very voracious. The larvae of the Staphylinidae are by no means unlike the perfect insect, and are found in similar situations ; their jaws I.] METAMORPHOSES OF INSECTS. 17 are powerful, and their legs moderately strong. The larvae of the Lamellicorn beetles (Figs. 1-6) — cock- chafers, stag-beetles, &c. — feed on vegetable sub- stances or on dead animal matter. They are long, soft, fleshy grubs, with the abdomen somewhat curved, and generally lie on their side. The larvae of the Elateridae, known as wireworms, are long and slender, with short legs. That of the glowworm (Lampyridae) is not unlike the apterous female. The male glow- worm, on the contrary, is very different. It has long, thin, brown wing-cases, and often flies into rooms at night, attracted by the light, which it probably mis- takes for that of its mate. The metamorphoses of the Cantharidae are very remarkable, and will be described subsequently. The larvae are active and hexapod. The Phy- tophaga (Crioceris, Galeruca, Haltica, Chrysomela, &c.) are vegetable feeders, both as larvae and in the perfect state. The larvae are furnished with legs, and are not unlike the caterpillars of certain Lepidoptera. The larva of Coccinella (the Ladybird) is some- what depressed, of an elongated ovate form, with a small head, and moderately strong legs. It feeds on Aphides. Thus, then, we see that there are among the Cole- optera many different forms of larvae. Macleay con- sidered that there were five principal types. I. Carnivorous hexapod larvae, with an elongated, more or less flattened body, six eyes on each side of the head, and sharp falciform mandibles (Carabus, Dyticus, &c.). \ C i3 ON THE ORIGIN AND [chap. 2. Herbivorous hexapod larvae, with fleshy, cyHn- drical bodies, somewhat curved, so that they he on their side. 3. Apod grub like larvae, with scarcely the rudiments of antennae (Curculio). 4. Hexapod antenniferous larvae, with a subovate body, the second segment being somewhat larger than the others (Chrysomela, Coccinella). 5. Hexapod antenniferous larvae, of oblong form, somewhat resembling the former, but with caudal appendages (Meloe, Sitaris). The pupa of the Coleoptera is quiescent, and the parts of the future beetle are plainly perceivable, being encased in distinct sheaths ; the head is applied against the breast ; the antennae lie along the sides of the thorax ; the elytra and wings are short and folded at the sides of the body, meeting on the under side of the abdomen ; the two anterior pairs of legs are entirely exposed, but the hind pair are covered by wing-cases, the extremity of the thigh only appearing beyond the sides of the body." ^ In the next three orders — namely, the Orthoptera (grasshoppers, locusts, crickets, walking-stick insects, cockroaches, &c.), Euplexoptera (earwigs), and Thy- sanoptera, a small group of insects well known to gardeners under the name of Thrips (PL I. and H., Figs. I and 2)— the larvae when they quit the egg already much resemble the mature form, differing, in fact, principally in the absence of wings, which are more or less gradually acquired, as the insect in- creases in size. They are active throughout life. 1 WfStwcod's Introduction, vol. i. p 36. ■ I.] METAMORPHOSES OE INSECTS. 19 Those specimens which have rudimentary wings are, however, usually called pupae. The Neuroptera present, perhaps, more differences in the character of their metamorphoses than any other order of insects. Their larvae are generally active, hexapod, httle creatures, and do not vary from one another in appearance so much, for in- stance, as those of the Coleoptera, but their pupae differ essentially ; some groups, namely, the Psocidae, Termitidae, Libellulidae, Ephemeridae, and Perlidae, remaining active throughout life, like the Orthoptera ; while a second division, including the Myrmeleonidae, Hemerobiidae, Sialidae, Panorpidae, Raphidiidae, and Mantispidae, have quiescent pupae, which, however, in some cases, acquire more or less power of locomotion shortly before they assume the mature state : thus, that of Raphidia, though motionless at first, at length acquires strength enough to walk, even while still enclosed in the pupa skin, which is very thin.^ One of the most remarkable families belonging to this order is that of the Termites, or white ants, They abound in the tropics, where they are a perfect pest, and a serious impediment to human develop- ment. Their colonies are extremely numerous, and they attack woodwork and furniture of all kinds, generally working from within, so that their presence is often unsuspected, until it is suddenly found that they have completely eaten away the interior of some post or table, leaving nothing but a thin outer shell. Their nests, which are made of earth, are sometimes ten or tv/elve feet high, and strong enough 1 Westwood's Introduction, vol. ii. p. 52. C 2 20 ON THE ORIGIN AND [chap. to bear a man. One species, Termes hicifugiLS, is found in the South of France, where it has been carefully studied by Latreille. He found in these communities five kinds of individuals — (i) males; (2) females, which grow to a very large size, their bodies being distended with eggs, of which they sometimes lay as many as 80,000 in a day ; (3) a form described by some observers as Pupae, but by others as neuters. These differ very much from the others, having a long, soft body without wings, but with an immense head, and very large, strong jaws. These individuals act as soldiers, doing apparently no work, but keeping watch over the nest and at- tacking intruders with great boldness. (4) Apterous, eyeless individuals, somewhat resembling the winged ones, but with a larger and more rounded head ; these constitute the greater part of the community, and, like the workers of ants and bees, perform all the labour, building the nest and collecting food. (5) La- treille mentions another kind of individual which he regards as the pupa, and which resembles the workers, but has four white tubercles on the back, where the Avings afterwards make their appearance. There is still, however, much difference of opinion among entomologists, with reference to the true nature of these different classes of individuals. M. Lespes, who has recently studied the same species, describes a second kind of male and a second kind of female, and the subject, indeed, is one which offers a most promising field for future study. Another interesting family of Neuroptera is that of the Ephemerae, or Mayflies (PL III., Fig. i), so well I.] METAMORPHOSES OF INSECTS. 2J known to fishermen. The larvae (PI. IV., Fig-, i) are semi-transparent, active, six-legged little creatures, which live in water ; having at first no gills, they re- spire through the general surface of the body. They grow rapidly and change their skin every few days. After one or two moults they acquire seven pairs of branchi^, or gills, which are generally in the form of leaves, one pair to a segment. When the larvse are about half grown, the posterior angles of the two posterior thoracic segments begin to elongate. These elongations become more and more marked with every change of skin. One morning, in the month of June, some years ago, I observed a full-grown larva, which had a glistening appearance, owing to the presence of a film of air under the skin. I put it under the microscope, and, having added a drop of water with a pipette, looked through the glass. To my astonishment, the insect was gone, and an empty skin only remained. I then caught a second specimen, in a similar condition, and put it under the microscope, hoping to see it come out. Nor was I disappointed. Very few moments had elapsed, when I had the satisfaction of seeing the thorax open along the m.iddle of the back ; the two sides turned over ; the insect literally walked out of itself, unfolded its wings, and in an instant flew up to the window. Several times since, I have had the pleasure of wit- nessing this wonderful change, and it is really won- derful how rapidly it takes place : from the moment when the skin first cracks, not ten seconds are over before the insect has flown away. Another family of Neuroptera, the Dragon-flies, 22 ON THE ORIGIN AND [chap. or Horse-stingers, as they are sometimes called, from a mistaken idea that they sting severely eroiigh to hurt a horse, though in fact they ere quite Irirnless, also spend their early days in the water. The larvae are brown, sluggish, ugly creatures, with six legs. They feed on small water-animals, for which they wait very patiently, either at the bottom of the water, or on some aquatic plant. The lower jaws are attached to a long folding rod ; and when any unwary little creature approaches too near the larva, this apparatus is shot out with such velocity that the prey which comes within its reach seldom escapes. In their perfect condition, also, Dragon-flies feed on other insects, and may often be seen hawking round ponds. The so-called Ant-lions in many respects resemble the Dragon-flies, but the habits of the larvae are very dissimilar. They do not live in the water, but prefer dry places, where they bury them- selves in the loose sand, and seize with their long jaws any small insect which may pass. The true Ant- lion makes itself a round, shallow pit in loose ground or sand, and buries itself at the bottom. Any inatten- tive little insect which steps over the edge of this pit immediately falls to the bottom, and is instantaneously seized by the Ant-lion. Should the insect escape, and attempt to climb up the side of the pit, the Ant-lion is said to throw sand at it, knocking it down again. One other family of Neuroptera which I must mention, is the Hemerobiidae. The perfect insect is a beautiful, lace-winged, very delicate, green creature, something like a tender Dragon-fly, and with bright, green, touching eyes. The female deposits her eggs I.] METAMORPHOSES OE INSECTS. 23 on leaves, not directly on the plant itself, but attached to it by a long white slender footstalk. The larva has six le-^^s and powenYu jaws, and makes itself very useful in destroying tiie Hop-fly. The insects forming the order Trichoptera are v/ell known in their larval condition, under the name of caddis worms. These larv^ are not altogether unlike caterpillars in form, but they live in water— which is the case with very few lepidopterous larvae — and form for themselves cylindrical cases or tubes, built up of sand, little stones, bits of stick, leaves, or even shells. They generally feed on vegetable substances, but will also attack minute freshwater animals. When full grown, the larva fastens its case to a stone, the stem of a plant, or some other fixed substance, and closes the two ends with an open grating of silken threads, so as to admit the free access of water, while excluding enemies. It then turns into a pupa, which bears some resemblance to the perfect insect, " except that the antennae, palpi, wings, and legs are shorter, en- closed in separate sheaths, and arranged upon the breast." The pupa remains quiet in the tube until nearly ready to emerge, when it comes to the surface, and in some cases creeps out of the water. It is not therefore so completely motionless as the pups of Lepidoptera. The Diptera, or Flies, comprise insects with two wings only, the hinder pair being represented by minute club-shaped organs called halteres." Flies quit the egg generally in the form of fat, fleshy, legless grubs. They feed principally on decaying animal or vegetable matter, and are no doubt useful 24 OA' THE ORIGIN AND as scavengers. Other species, as the gadflies, deposit their eggs on the bodies of animals, within which the grubs feed, when hatched. The mouth is generally furnished w^ith two hooks which serve instead of jaws. The pupae of Diptera are of two kinds. In the true flies, the outer skin of the full-grown larva is not shed, but contracts and hardens, thus assuming the appear- ance of an oval brov/nish shell or case, within which the msect changes into a chrysalis. The pupae of the gnats, on the contrary, have the limbs distinct and enclosed in sheaths. They are generally inactive, but some of the aquatic species continue to swim about. One group of Flies, which is parasitic on horses, sheep, bats, and other animals, has been called the Pupipara, because it was supposed that they were not born until they had arrived at the condition of pupae. They come into the world in the form of smooth ovate bodies, much resembling ordinary dip- terous pup^, but, as Leuckart has shown,^ they are true, though abnormal, larvae. The next order, that of the Aphaniotera, is very small in number, containing only the difl"erent species of Flea. The larva is long, cylindrical, and legless ; the chrysalis is motionless, and the perfect insect is too well known, at least as regards its habits, to need any description. The Heteroptera, unlike the preceding orders of insects, quit the egg in a form, difiering from that of the perfect insect principally in the absence of wings, which are gradually acquired. In their metamor- ^ Die Fortpflaiizung und Entwickelung der Pupipai-en. Von Dr. R. Leuckart. Halle, 1848. I.] METAMORPHOSES OE INSECTS. 25 phoses they resemble the Orthoptera, and are active through hfe. The majority are dull in colour, though some few are very beautiful. The species constituting this group, though very numerous, are generally small, and not so familiarly known to us as those of the other large orders, with indeed one exception, the well-known Bug. This is not, apparently, an indi- genous insect, but seems to have been introduced. The word is indeed used by old writers, but either as meaning a bugbear, or in a general sense, and not with reference to this particular insect. In this country it never acquires wings, but is stated to do so some- times in warmer climates. The Heteroptera cannot exactly be said either to sting or bite.. The jaws, of which, as usual among insects, there are two pairs, are like needles, which are driven into the flesh, and the blood is then sucked up the lower lip, which has the form of a tube. This peculiar structure of the mouth prevails throughout the whole order ; conse- quently their nutriment consists almost entirely of the j;uices of animals^ or plants. The Homcptera agree with the Heteroptera in the- structure of the mouth, and in the metamorphoses. They differ prin- cipally in the front wings, which in Homoptera are membranous throughout, . while in the Heteroptera the front part is thickened and leathery. As in the Heteroptera, however, so also in the Homoptera, some species do not acquire wings. The Cicada, celebrated for its chirp, and the lanthorn fly, belong to this group. So also does the so-called Cuckoospit, so common in our gardens, which has the curious faculty of secreting round itself a quantity of frothy 26 ON THE ORIGIN AND [chap. fluid which serves to protect it from its enemies. But the best known insects of this grorp are the Aphides or Plant-Hce ; while trie most I'scful ijelong to the Coccidae, or scale insects, from one species of which we obtain the substance called lac, so exten- sively used in the manufacture of sealing-wax and varnish. Several species also have been used in dyeing, especially the Cochineal insect of Mexico, a species which lives on the cactus. The male Coccus is a minute, active insect, with four large wings ; while the female, on the contrary, never acquires wings, but is very sluggish, broad, more or less flattened, and in fact, when full grown, looks like a small brown, red, or white scale. The larvae of the order Lepidoptera are familiar to us all, under the name of caterpillars. The insects of this order in their larval condition are almost all phytophagous, and are very uniform both in structure and in habits. The body is long and cylindrical, con- sisting of thirteen segments ; the head is armed with powerful jaws ; the three following segments, the future prothorax, mesothorax, and metathorax, each bears* a pair of simple articulated legs. Of the pos- terior segments, five also bear false or pro-legs, which are short, unjointed, and provided with a number of hooklets. A caterpillar leads a dull and uneventful life ; it eats ravenously, and grows rapidly, casting its skin several times during the process, which generally lasts only a few weeks ; though in some cases, as for instance that of the goat-moth, it extends over a period of two or three years, after which the larva changes into a quiescent pupa or chrysalis. 11.] METAMORPHOSES OF INSECTS. 27 CHAPTER II. THE INFLUENCE OF EXTERNAL CONDITIONS ON THE FORM AND STRUCTURE OF LARV.E. ' The facts recapitulated briefly in the preceding chapter show, that the forms of insect larvae depend greatly on the group to which they belong. Thus the same tree may harbour larvae of Diptera, Hymenoptera, Coleoptera, and Lepidoptera ; each presenting the form typical of the family to which it belongs. If, again, we take a group, such, for instance, as the Lamellicorn beetles, we shall find larvae extremely similar in form, yet very different in habits. Those, for instance, of the common cockchafer (Fig. i) feed on the roots of grass ; those of Cetonia aurata (Fig. 2) inhabit ants' nests ; the larvae of the genus Trox (f'jg- 3) found on dry animal substances ; of Oryctes (Fig. 4) in tan-pits ; of Aphodius (Fig. 5) in dung ; of Lucanus (the stag-beetle, Fig. 6) in wood. On the other hand, in the present chapter it will be m}^ object to show that the form of the larva depends very much on the conditions of its life. Thus, those larvae which are internal parasites, whether in animals 2.8 ON THE ORIGIN AND [chap. or plants, are vermiform, as are those which Hve in cells, and depend on their parents for food. On the other hand, larvae which burrow in w^ood have strong jaws and generally somewhat weak thoracic legs ; whilst those which feed on leaves have the thoracic legs more developed, but less so than the carni- vorous species. Now, the Hymenoptera, as a general rule, belong to the first category : the larvae of the Ichneumons, &c., which live in animals, — those of Fig. I, Larva of the Cockchafer (IMelolonthaV (Westwood, Int. to the Modern Classificaiion of Insec s vol. i. p. 194 ) 2, Larva of Cctonia, 3, Larva o^ Irox 4, Larva of Oryctes. 5, Larva of Aphodius (Chapnis and Candeze, Mem. Soc. Roy. Liege, 1853). 6, Larva of Lucanus (Packard, Guide to the Study of Lisccts, Fig. 403). 'the Cynipidae, inhabiting galls,— and those of ants, bees, wasps, &c,, which are fed by their parents, are fleshy, apodal grubs ; though the remarkable fact that the embryos of bees in one stage of their de- velopment possess rudiments of thoracic legs which subsequently disappear, seems to show, not indeed that the larvae of bees were ever hexapod, but that bees are descended from ancestors which had hex- II.J METAMORPHOSES OF INSECTS. 29 apod larvae, and that the present apod condition of these larvae is not original, but results from their mode of life. On the other hand, the larvae of Sircx (Fig. 14) being wood-burrowers, possess well-developed tho- racic legs. Again, the larvae, of the Tenthredinidae, which feed upon leaves, closely resemble the cater- pillars of Lepidoptera, even, to the presence of abdominal pro-legs. The larvae of most Coleoptera (Beetles) are active, hexapod, and more or less flattened : but those which live inside vegetable tissues, such as the weevils, are apod fleshy grubs, like those of Hymenoptera. PI. II., Fig. 6, represents the larva of the nut-weevil, Bala- ninus (PI. I., P'ig. 6), and it will be seen that it closely resembles PI. II., Fig. 5, which represents that of a fly (Anthrax), PI. I., Fig. 5, and PI. II., Figs, 7, 8, and 9, which represent respectively those of a Cynips or gall-fly (PI. I., Fig. ;), an ant (PI. I., Fig. 8), and wasp (PL I., Fig. 9). Nor is Balaninus the only genus of Coleoptera which afibrds us examples of this fact, Thus in the genus Scolytus (PI. I., Fig. 4), the larvae (PI. II., Fig. 4), which, as already mentioned, feed on .. 7 ■ Fig. 7, Larva of Brachy tarsus (Ratzeburg, Forst Insecten). 8, Larva of Crioceris ' (Westwood, loj. cit.). the bark of the elm, closely resemble those just de- scribed, as also do those of Brachytarsus (Fig. 7). On the other hand, the larvae of certain beetles feed on 30 ON THE ORIGIN AND [chap. leaves, like the caterpillars of Lepidoptera ; thus that of Crioceris Asparagi (Fig. 8)— which, as its name denotes, feeds on the asparagus — closely resembles the larvae of certain Lepidoptera, as for instance of Thecla spini. From this point of view the trans- formations of the genus Sitaris (?J. III., Fig. 4), which have been very carefully investigated by M. Fabre, are peculiarly interesting.^ Fig. 9, Larva of Sitaris Rum'eralis (Kabre, Ann. des Sci. • Nat., ser. 4, tome vii.). lo, Larva of Sitaris iiumerili<, in the sepond ."-tage. J I, Larva of Sitaris humeralis, in the third stage. 12, Larva of Sitaris humeralis, in the fourth stage. 13, Pupa of Sitaris. The genus Sitaris (a small beetle allied to Can- tharis, the blister-fly, and to Meloe, the oil-beetle) is parasitic on a kind of Bee (Anthophora), which ex- cavates subterranean galleries, each leading to a cell. The eggs of the Sitaris, which are deposited at the entrance of these galleries, are hatched at the end of September or beginning of October ; and M. Fabre not ^ Ann. des Sci. Nat., ser. 4, tome vii. See also Natural History Rroiezv, April 1862. METAMORPHOSES OF INSECTS. 31 unnaturally expected that the young larvae, which arc active little creatures with six serviceable legs (Fig. 9), would at once eat their way into the cells of the An- thophora. No such thing : till the month of April following they remain without leaving their birthplace, and consequently without food ; nor do they in this long time change either in form or size. M. Fabre ascertained this, not only by examining the burrows of the Anthophoras, but also by direct observation of some young larvae kept in captivity. In April, however, his captives at last awoke from their long lethargy, and hurried anxiously about their prisons. Naturally inferring that they were in search of food, M. Fabre supposed that this would consist either of the larvae ' or pupae of the Anthophora, or of the honey with which it stores its cell. All three were tried without success. The first two were neglected, and the larvae, when placed on the latter, either hurried away, or perished in the attempt, being evidently unable to deal with the sticky substance. M. Fabre was in despair : " Jamais experience," he says, " n'a eprouve pareille deconfiture. Larves, nymphes, cellules, miel, je vous ai tous offert ; que voulez-vous done, bestioles maudites } " The first ray of light came to him from our countryman, Newport, who ascertained that a small parasite found by Leon Dufour on one of the wild bees, and named by him Triungulinus, was, in fact, the larva of Meloe. The larvae of Sitaris much resembled Dufour's Triungulinus ; and acting on this hint, M. Fabre examined many specimens of Antho- phora, and found on them at last the larvae of his Sitaris. The males of Anthophora emerge from the 32 ON THE ORIGIN AND [chap. pupse sooner than the females, and M. Fabre ascer- tained that, as they come out of their galleries, the little Sitaris larvae fasten upon them. Not, however, for long : instinct teaches them that they are not yet in the straight path of development ; and, watching their opportunity, they pass from the male to the female bee. Guided by these indications, M. Fabre examined several cells of the Anthophora : in some, the egg of the Anthophora floated by itself on the surface of the honey ; in others, on the egg, as on a raft, sat the still more minute larva of the Sitaris. The mystery was solved. At the moment when the egg is laid the Sitaris larva springs upon it. Even while the poor mother is carefully fastening up her cell, her mortal enemy is beginning to devour her offspring : for the egg of the Anthophora serves not only as a raft, but as a repast. The honey which is enough for either, would be too little for both ; and the Sitaris, therefore, at its first meal, relieves itself from its only rival. After eight days the egg is con- sumed, and on the empty shell the Sitaris undergoes its first transformation, and makes its appearance in a very different form, as shown in Fig. lo. The honey which was fatal before is now necessary ; the activity which before was necessary is now use- less ; consequently, with the change of skin, the active, slim larva changes into a white, fleshy grub, so organized as to float on the surface of the honey, with the mouth beneath, and the spiracles above the surface : "grace a I'embonpoint du ventre," says M. Fabre, " la larve est a I'abri de I'asphyxie." In this state it remains until the honey is consumed ; then the II.] METAMORPHOSES OF INSECTS. 33 animal contracts, and detaches itself from its skin, within which the further transformations take place. In the next stage, which M. Fabre calls the pseudo- chrysalis (Fig. 11), the larva has a soHd corneous envelope and an oval shape ; and in its colour, con- sistency, and immobility reminds one of a Dipterous pupa. The time passed in this condition varies much. When it has elapsed, the animal moults again, again changes its form, and assumes that shown in Fig. 12 ; after this it becomes a pupa (Fig. 13) without any remarkable peculiarities. Finally, after these wonderful changes and adventures, in the month of August the perfect Sitaris (PI. III., Fig. 4) makes its appearance. On the other hand, there are cases in which larvae diverge remarkably from the ordinary type of the group to which they belong, without, as it seems in our present imperfect state of information, any suffi- cient reason. Thus the ordinary type of Hymenopterous larva, as we have already seen, is a fleshy apod grub ; although those of the leaf-eating and wood-boring groups, Tenthredinidae and Siricidse (Fig. 14), are caterpillars, more or less closely resembling those of Lepidoptera. There is, however, a group of minute Hyrhenoptera, the larvae of which reside within the eggs or larvae of other insects. It is difficult to understand why these larvae should differ from those of Ichneumons, which are also parasitic Hymenoptera, and should be, as will be seen by the accompanying figures, of such remarkable and grotesque forms. The first known of these curious larvae was observed by De Filippi,^ who, 1 Ann. and Mag. of Nat. Hist. 1852. 5 D 34 ON THE ORIGIN AND [chap. having collected some of the transparent eggs of a small Beetle {Rhynchites betuleti), to his great surprise found more than half of them attacked by a parasite, which proved to be the larva of a minute Hymenopterous insect belonging to the Pteromalid