BIO-AGRICULTURAL LIBRARY UNIVERSITY OF CALIFORNIA RIVERSIDE, CALIFORNIA 92521 A NATURAL HISTORY OF THE BRITISH LEPIDOPTERA. A TEXT-BOOK FOR STUDENTS AND COLLECTORS J. W. TUTT, F.E.S., Author of " The British Noctuse and their Varieties," " Monograph of the British Pterophorina," "British Butterflies," "British Moths," etc. VOL. I. LONDON : SWAN SONNENSCHEIN & Co., Paternoster Square, E.G. BEBLIN : FBIEDLANDER & SOHN, 11, Carlstrasse, N.W. JANUARY, 1899. AGRICULTURAL LIBRARY UMVLRSITY OF CALIFORNIA ^,-ronc DPQPARCH CENTER AND PREFACE. In submitting this volume to the entomological public, the author rusts that the method of treatment will commend itself. The recent ork that has been accomplished in the classification of the Lepidoptera Chapman, Dyar, Packard and others, has rendered a radical re- angement necessary. In commencing with the more generalised, proceeding to the more specialised, superfamilies, the author con- rs that he has adopted a logical course that will meet with the oval of those best qualified to judge in this matter. It has been idered better to complete thoroughly a few superfamilies rather to attempt to deal with a large number superficially, and it is that the separate treatment of the main points in the life-history species dealt with, will be of advantage to the various classes of logists synonymists, systematists, biologists, and those that ae subject under its geographical, or any one of its philosophical arge part of a work of this description is necessarily more or less .pilation, and the author wishes here to express his obligation to 3 authors to whose works he is indebted for information, as well as ae very great number of entomologists (rather more than 200 in .nber) to whom he is indebted for local lists, and to those who have -plied him with other items of interest that have added to the Jness and completeness of the volume. These have always been hedged, he believes, in the body of the work. There are many, -, who have done much more than this. To Messrs. J. H. W. F. Kirby, L. B. Prout and Lord Walsingham, for their dealing with matters of " synonymy," to Messrs. A. Bacot, 3. Fletcher, Drs. T. A. Chapman and J. H. Wood, for the vast .t of information relating to the " life-histories " of the insects bed, to Mr. G. C. Bignell for notes on the "parasites" affecting , to Mr. F. Lemann for copious translations from German works, .. Oberthiir for the gift and loan of many rare Anthrocerids, and fr. C. Fenn for the generous use of his voluminous note-books, the jhor tenders his sincerest and grateful thanks. Although essentially a work on British Lepidoptera, it is trusted b it will have an interest for other than purely British lepidopterists. 3 chapters on each superfamily cover the whole fauna included in the 1 ^erfamily, and should, therefore, be of use generally to students of uese superfamilies. The " distribution " of each species, too, outside the British Isles, is considered separately from the recorded localities within the limits of our own country, and should be useful to students of geographical distribution in all parts of the world. The author is fully aware that in a book containing so much detail, there must necessarily be many sins of commission and omission. He can onfy hope that these are not serious, and assure his readers that he has taken the greatest care to eliminate them. The trouble to which the author has been put, and the hours of comparatively waste time that he has spent, in compiling the lists of localities, synonymic tables, distribution, etc., and in unearthing records of the rarer varieties and aberrations, owing to the incomplete and imperfect indexes of entomological magazines in general and works on Lepidoptera in particular, have led him to index every reference to super- families, families, genera, species, varieties, etc., mentioned in the book. It is trusted that this will be found of great time-saving value to all who have need to refer to the volume. The publication of a purely technical book of this description would be practically impossible but for the generosity of a section of the entomological public who take an author on trust, as it were, and practically guarantee him against any serious financial loss. . To the following ladies and gentlemen, therefore, who have in reality brought about the publication of this volume, the author tenders his heartiest thanks, and trusts that it will meet with their full approval and approbation. LIST OF SUBSCRIBERS I Adams, Herbert J., F.E.S'. Adkin, Benjamin W. Adkin, liobert, F.E.S. Alderson, Hope, F.E.S. Ames, (Mrs.) Sarah Arbuthnott, H. C. Ash. Rev. C. D., B.A. Auckland, Lord Bacot, A. Bankes, Eustace II., M.A., F.E.S. Barclay, Francis H., F.G.S., F.E.S. Barker, Reginald H. Bartlett, C. Bateson, William, M.A., F.R.S..F.Z.S., F.E.S. Beare, Professor T. Hudson, B.SC., F.R.8.E., F.E.S. Bishop, E. B. Blathwayt, Lt.-Col. Linley, F.L.S., F.E.S. Bostock, E. D. Bouskell, Frank, F.E.S. Bowell, E. W. W. (2 copies) Bower, B. A., F.E.S. Bowles, E. Augustus, M.A., F.E.S. (2 copies) Briggs, Thomas H., M.A., F.E.S. Brown, Capt. E. W. Brown, H. Rowland-, M.A., F.E.S. Buckmaster, Rev. C. J., M.A. Burnett, Basil Burrows, Rev. C. R. N. Butler, W. E. Butterfield, J. A., B.SC. Capper, Samuel J., F.L.S., F.E.S. Chamberlain, Neville Chapman, T. A., M.D., F.Z.S., F.E.S. Cheesman. E. M. Christy, W. M., M.A., F.E.S. Clark. J. A., F.E.S. Clarke, A. H., F.E.S. Clarke, H. Shortridge, F.E.S. Cotton, John, M.R.C.S. Cowl (Mrs.), M. E. Crabtree, Benjamin H., F.E.S. Crass, C. H. Dadd. E. Martin Dalglish, A. Adie Daltry, Rev. T. W., M.A., F.E.S. Daws, William Day, F. H. Day, G. 0., F.E.S. Dent, Hastings C., C.E., F.L.S., F.E.S. Dobson, H. T., F.E.S. Doncaster, Arthar (6 copies) Donisthorpe, H. St. John K., F.Z.S., F.E.S. Dupont, L. Eddrupp, Rev. T. B., M.A. Edelsten, H. M. Edwards, Stanley, F.L.S., F.Z.S., F.E.S. (2 copies) Elwes, Henry J., F.E.S., F.Z.S., F.E.S. Fenn, Charles, F.E.S. Fernald, Professor C. H., PH.D., M.A., F.E.S. Fletcher, W. H. B., M.A., F.E.S. Freer, Richard, M.B., M.R.C.S., F.E.S. Friedlander and Sohn (6 copies) Garland, G. R. Glenny, F. Gordon, Roger S., F.E.S. Greer, T. Griffith. A. F., M.A. Griffiths, G. C., F.Z.S., F.E.S. Gunning, Montague, M.D., F.E.S. Haggart, J. C. Hall, Thomas W., F.E.S. Hamm, A. H. Hanbury, Frederick J., F.L.S., F.E.S. Harrison, Albert, F.C.S., F.L.S., F.E.S. Harrison, Edward S. Head, H. W. Headly, Charles B., F.E.S. Helps, John A. Hepburn, Sir Archibald Buchan, F.K.S. Hewett, Rev. G. M. A., M.A. Heyne, Alexander Hill, Henry A., F.Z.S., F.E.S. Holland, Rev. W. J., D.D., PH.D., F.E.S. Holland, William Home, Arthur, F.E.S. Image, Selwyn, M.A., F.E.S. Jenner, J. H. A., F.E.S. Jones, A. H., F.E.S. Jones, F. B., M.R.C.V.S. Kane, W. F. de Vismes, M.A., M.R.I.A., F.E.S. Kaye, William J., F.E.S. Lemann, Frederick C., F.E.S. Llewelyn, Sir John T. D., Bart., M.A., M.P.. F.L.S. Lloyd, Alfred, F.C s., F.E.S. Lloyd, B. Wylie, F.E.S. Lowe, llev. Frank E., M.A., F.E.S. Lucock, Frank Luff, W. A. Marsden, H. W. ( 6 copies) Mason, Philip 13., M.R.C.S.. F.L.S., F.E.S. Massey, Herbert T., F.E.S. (2 copies) May, H. H. Maze, W. P. Blackburne-, F.E.S. Mclntyre, F. Merrifield, Frederic, F E.S. Moberly, J. C., M.A., F.E.S (2 copies) Moore, Harry, F.E.S. Morton, Kenneth J., M.A., F.E.S. Moss, Rev. A. M., M.A. Mousley, H., F.E.S. Nevinson, Basil G., M.A., F.Z.S., F.E.S, Newland, C. Bingham Nicholson, Charles, F.E.S. Nicholson, William E., F.E.S. Ovenden, Joseph Page, Herbert E., F.E.S. Pearson, (Mrs.) C. N. Peed, John Phillips, Hubert C., M.E.C.S., F.E.S. Pitman, M. A. Porritt, George T., F.L.S. , F.E.S. Prout, Louis B., F.E.S. Rao, (Mrs.) E. llajisom, E. llaynor, llev. Gilbert H., M.A. Beuter, Professor Enzio, PH.D., F.E.S. lleid, William, F.E.S. lleid, Capt. S. G., F.E.S. Riding, William S.. B.A., M.D., F.E.S. Robertson, Major R. B. Robinson, George Robson, John E., F.E.S. Rothschild, Hon. Nathaniel C., B.A., F.Z.S., F.E S. Rothschild, Hon. Walter, F.Z.S., F.E.S. (2 copies) Routledge, George B., F.E.S. Russell, A., F.E.S. Russell, S. G. C., F.E.S. Sheldon, W. G. Smetham, Henry Smith, W. Hawker- Snellen. Pieter Carl T., Hon. F.E.S. Steck, Dr. Theodor Studd, E. F. C., M.A., B.C.L., F.E.S. Tarbat, Rev. J. E., M.A. Thornhill. W. B. Thornthwaite, W., F.K.A.S. Tunaley, Henry, F.E.S. Tunstall, Wilmot, F.E.S. Tunstall, Thomas Turner, H. J., F.E.S. Waldegrave. The Rt. Hon. Earl Walker, Samuel Walsingham, The Rt. Hon. Lord. M.A., LL.D., F.R.S., F.E.S., etc. (2 copies) Watkins, C. J. Whittle, F. G. Wilkinson, G. Williams, H. Wolfe. J. J. Woolley, H. S., F.E.S. Wright, Dudley, F.B.C.S., F.E.S. Ashby, Herbert, F.E.S. Nonpareil Entomological Society. CONTENTS. PART I. CHAP. PAGE. I. THE OPJGIN OF THE LEPIDOPTEBA ... ... ... 1 II. THE OVUM OR EGG ... ... ... ... 6 III. EMBRYOLOGY OF A LEPIDOPTEROUS INSECT ... ... 16 IV. PARTHENOGENESIS OR AGAMOGENESIS IN LEPIDOPTERA ... 23 V. EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA ... 80 VI. INTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA ... 51 VII. VARIATION OF THE IMAGINES OF LEPIDOPTERA ... ... GO VIII. PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LEPIDOPTEROUS LARVAE ... ... ... ... 76 IX. CLASSIFICATION OF LEPIDOPTERA 102-112 PART II. THE SPHINGO-MICROPTERYGID STIRPS ... ... ... 113 THE MICROPTERYGIDES ... ... ... ... 129 THE NEPTICULIDES ... ... ... ... 162 THE COCHLIDIDES (OR EUCLEIDEs) ... ... ... 860 THE ANTHROCERIDES ... ... ... ... 888 INDEX .. 547-560 BRITISH LEPIDOPTERA. CHAPTER I- THE ORIGIN OF THE LEPIDOPTERA. FOE many years entomologists have attempted to work out the line of descent by which the Lepidoptera have been evolved. McLachlan, in 1865, and Speyer, in 1870, pointed out certain broad affinities between the Trichoptera (caddis-flies) and some families of the Lepi- doptera. Packard, in 1863, had also suggested a relationship between the two orders. The co-ordinal value of the two orders, however, was maintained by all these writers, and it was not until 1896 that Sharp, commenting on the pupa of a species of Microptery.c (probably aemipurpurella), stated that he "considered the pupa to be that of a Trichopterous insect," and that Micropteryx should be referred to that order, and that, if this course were not adopted, he felt clear that Trichoptera could not be maintained distinct as an order from Lepi- doptera. Chapman had previously described at length the pupal jaws of Micropteryx, and pointed out that they exhibited characters quite unique among the Lepidoptera. There is, however, a group of Lepidoptera, in some respects, pro- bably, more generalised than the MICROPTERYGIDES. These are the ERIOCEPHALIDES comprising the British species, calthella, seppella, aruncella, thunberyella and viansuetella. Walterf discovered that the imagines of K. calthella had maxillae constructed on the type of those of biting or mandibulate insects. Chapman has described! the way in which the jaws are used in eating the pollen. The generalised mouth-parts of Eriocephala consist of maxillary lobes, mandibles, etc., but not only do they diner from all other Lepidoptera in this par- ticular, but the thorax and abdomen of the imago are also more generalised. Both the MICROPTERYGIDES and the ERIOCEPHALIDES have the fore- and hind- wings united by a jugum, and, in this respect, as well as in the highly generalised condition of the neuration, they resemble "the Trichoptera. Another super-family with Trichopterygid affinities is the HEPIA- LIDES. Speyer, in a very interesting paper, refers to the similarity of * Trans. Ent. Soc. London, 1893, p. 263. f Jenaische Zcitschrijt, 1885. \ Trans. Ent. Soc. London, 1894, p. 338. Stett. Ent. Zeitung, 1870. 2 BRITISH LEP1DOPTEKA. the neuration of the Hcpialiilae and Coanidae, and remarks that they resemhle the Trichoptera no less than the Micropterygidae, though the JJcpiulidae exhihit other close analogies with the Trichoptera. lie also adds that the middle cell of the wing in the Phniijanciilai' is not fundamentally different from that of the He-pialidac, tVm///r/<> and Micropteryyidae, whilst the hind-wings of the Pxyckidae exhibit similar characters. This brief summary indicates the directions in which it has been suggested that the Lepidoptera are allied to the Trichoptera. The nature of the alliance has been variously discussed, but the general conclusions reached fall into one of two lines : (1) That the Lepi- doptera have descended by way of the MICROPTERYGIDES, HEPIALII>KS, and PSYCHIDES directly from the Trichoptera. (2) That the Trich- optera and Lepidoptera have developed from a common ancestor. To discuss this matter satisfactorily we must first consider the similarities between Lopidoptera and Trichoptera. The resemblance between their lame is very strong, their external structure being almost the same, the principal difference being that the lepidopterous larva possesses abdominal prolegs. These, however, are absent in Micropterygid larvae, as well as in other lepidopterous larvae whose habit it is to mine into their food-plants. The similarity of the pupa of Micropteryv to that of the Trichoptera has been already noticed. The abdominal segments of both are more or less freely movable upon each other. They form the " Pupfe Liberia " of Packard, whilst those generalised lepidopterous pupae, which have a considerable number of free (movable) abdominal segments, the " Pupa 1 Incompletae" of Chapman, are much nearer to the ancestral forms than the " Pupae Obtectse," which represent the more specialised forms. The lepidopterous pupa has been looked upon as presenting a sub- imaginal condition of a type midway between the ametabolous and metabolous orders of insects. This has been suggested by the con- dition of the pupal wing-cases, which are similar to those of metabolous nymphs, such as Dermaptera, Tertnitidae, Pnoi-iilac and Hemiptera. Spiiler has shown that the neuration of the lepidopterous pupa is almost identical with that of the Blattidae and Fulyoridae. Packard says that the " wings of the lepidopterous pupa may be said to be in the nymph stage of the ametabolous insects mentioned, since they are direct outgrowths from the tergites of the segments from which they arise." He further says that " if the wing-cases of any lepidopterous pupa, together with the meso- and meta-thorax are, before the larval skin is moulted, removed and spread out," it will be seen that " they bear, as Hpiiler shows, a striking resemblance to those of a beetle, I'l'i-nn-n, 1'mn-itit, or any hemipterous insect." He further points out that the pupal neuration, as well as the appendages maxillre, labium and legs are ancestral and phylogenetic, showing considerable differences when compared "with the corresponding structures in The more specialised imago. The importance o^tN^pupa, as bearing on the origin of the Lepi- doptera, is also very evident when the more generalised forms of tho lepidopterous pupa are compared with the more generalised forms of the dipterous pupa, as exhibited by the Bibinniilae, 'l'ipulia) rubi ; Bacot records the destruction of a whole batch of Arctia caia eggs by the same species, whilst Bignell states that he bred 2,100 imagines, of the same parasite, from 200 eggs of M. rubi, an average of more than ten to each egg ; Dimmock mentions the breeding of 30 hymenopterous parasites from a single egg of Snicrintltm t'i-<-. i/ein- iiiaria, 20 days; Hijbemia h-ucuphaearia, 38 days ; Lan-ntia cantata, 24 days, etc. In some species the length of time varies in different years, probably depending on meteorological conditions. Thus, JJiaton hirtaria may take from 17 to 37 days ; Hemerophila abniptaria, from 14 to 26 days; Selenia lunaria took 7 days in 1865, 12 days in 1861, and 15 days in 1886 all of the first brood. Selenia hi lunaria has the following record : 1880, first brood, 16 days; 1883, first brood, 28 days, second brood, 16 days; 1890 and 1891, second brood, 15 days. But different broods of the same species may vary in the same year ; thus, in 1865, one batch of Camptoyrammti flnriata took 5 days, another 10 days, and a third 21 days. Of those species which pass the winter in the egg stage, the time is so great that the combined larval, pupal, and imaginal periods are comparatively very short. Thus the egg stage of Kpione apiciaria lasts as long as 9f months ; of Ennomos aiitumnaria, 7f to 10 months ; of Hhin'ra pennaria, 5 months; of Opnrabia jilii/raiiniiaria, 4f months ; of ('idaria testata, 8 months ; of Clicaias xpa rtiata, 4 months. The egg stage of Theela n- -album and Zepliymx //w/vm- lasts from July to early May; of Theda j/riini, from June until late April; of Plebeius acijon, from July to April ; of Trichiura crataciji, from September to * Entom. Record, etc., iii., pp. 175-176 ; iv., p. 255 ; iv., p. 2\)2. THE OVUM OR EGG. 15 April ; of the Catocalids, from July and August to April, and so on. The condition of the egg during the hybernating period is very interesting. In some species, such as Aryynnis adippe, Pampldla comma, Parnasxiux apolln, etc., the fully formed caterpillar remains coiled up within the shell all the winter ; in others, the eggs appear to remain until spring, almost in the same condition, so far as the con- tents are concerned, as that in which they were laid. Buckler records that eggs of Sombyx mori, Trichiura crataeiji, Ennotnos (alniaria) tiliaria, E. quercinaria, Cheimatobia brumata, C. boreata, Scotosia vetulata, Ptilopkora plumiyera and Polio, chi, have been examined from time to time until the middle of January, and nothing but the faintest traces of the future larvae have been detected by a microscopic examination of their still fluid contents. In the case of Tiliacea (Xanthia) aurai/o, however, an egg was found to contain a partially developed larva on January 14th. It occasionally happens, as in the case of Polio, .rantJioiiiista var. nii/rocincta, that part of a batch of eggs, which should normally hybernate during the winter, hatches in the autumn, and the larvae attempt to feed up, whilst the remainder of the batch goes over normally. It is recorded, also, that in a batch of Oryyia antiqua eggs, the hatching takes place most irregularly, a few larvae appearing at a time, and the emergence of the whole brood thus spread over a long period. This happens also in Epione apiciaria, Laiiocampa tnfolii, Catocalia species, etc. The influence that temperature has on the hatching period, and on the vitality of lepidopterous eggs, has been well shown by Merriiield. lie has recorded that eggs of Selenia bilnnaria, and those of Selenia tetralwiaria, were quite uninjured by exposure to a temperature of from 80 F. to 90 F., their development, on the contrary, being greatly accelerated. Spring-laid eggs of 8. bilunaria began to have their vitality affected after being " iced " (at a temperature of 32 F., when they were in the central red stage), for 28 days, and none hatched after 60 days' icing. The result was even worse with spring- laid eggs of 8 den i a tetralwiaria, none of which survived 42 days' icing, and some summer-laid eggs of the same species, exposed to the same conditions, fared no better. In all the experiments, up to 60 days' exposure, nearly all the eggs, after being removed from the ice, matured so far as to admit of the formation of the young larva, which could be seen through the transparent shell. The failure was a failure to hatch. Standfuss has recorded that eggs of Arctia fasciata, Dasychim alictix, Odoneatis (Lasiocampa) pruni and Dendrolinnia pini, which were exposed to a temperature of 30 C. (93 F.), during the process of laying by the female, and up to the time of hatching, produced larvae in two-thirds or less of the normal time, and there emerged as perfect insects in the same year, i.e., without hibernation of the larva, in the case of A. fam-iata, 71 per cent.; of D. abietis, 90 per cent. ; of (.). jn-uni, 100 per cent. ; and of D.pini, 81 percent. The larvae and pupae of the broods were kept, as far as possible, at a mean temperature of 25 C. The eggs of the same females as those used in the above experi- ment, which had already been laid at a normal temperature (22 C.), and were left in this until hatched, afterwards remaining in the same mean temperature of 25 C., as the other larvae and pupae, produced a considerably smaller number of perfect insects, without hibernation of 16 BRITISH LEPIDOPTERA. the larvae, viz., A. fasciata, 23 per cent. ; D. abietis, 12 per cent. ; 0. pnmi, 64 per cent. ; D. pini, 28 per cent. It has been suggested that the sex of the imagines reared from eggs can be determined by the conditions in regard to abundance of food, or the reverse, under which the larvae are reared ; that, under a specially nutritious diet, lepidopterous larvae tend to produce female imagines, whilst a starvation diet tends to the production of males. This, of course, assumes a neutral condition as regards sex in the newly-hatched larva, but the experiments that are supposed to have proved this simply show that male larvae will stand more starving than those of females, or, in other words, that the minimum food which will allow male larvae to just pupate, is, in the same species, often insufficient to allow the process in female larvae, which die under such extreme treatment. The sexual organs of newly-hatched larvae are moderately well-developed. Another theory which has been assumed, viz., that eggs laid suc- cessively by the same female are of opposite sex, has been entirely disproved, and experiment has shown that the relative proportion of the sexes is subject to immense fluctuation on the separate dates on which eggs are laid. As regards eggs laid on any one day, the sexes generally succeed each other in little groups of irregular size. Ifc is further recorded that the pupae obtained from different batches of Vani'xxa io had a large proportion of a certain sex, some batches pro- ducing almost entirely males, others consisting almost entirely of females. The eggs of Lepidoptera are developed in the ovaries of the parent, whence they pass down the oviduct into the vagina. In connection with the vagina are one or more pouches called receptacula seminis, in which the spermatozoa are stored after copulation. As the egg passes along the vagina to the ovipositor, the spermatozoa, or sperm- cells, are released from the receptacula, and certain of them enter the egg through the micropylar tubes, one of which fertilises the egg. Fertilisation, then, takes place at the time that the egg is being laid, by the spermatozoa entering the micropylar pores at the time that the egg passes the pouches. It is sometimes noticed that the latest-laid eggs of a moth are infertile, a result probably due to the supply of spermatozoa being exhausted before all the eggs are laid. It is well- known that many Lepidoptera pair more than once. Anticlca ber- bcrata, Tei>lironia l>istortata, and various Zygronid species have been observed to do so repeatedly. No doubt, the habit is of common occurrence. CHAPTEE III. EMBRYOLOGY OF A LEPIDOPTEROUS INSECT. IT may be well now to briefly consider the changes that take place in the fertilised ovum or egg, and that have, as their result, the pro- duction of an individual resembling its parents. These changes are of the utmost importance, and the embryological studies made by various entomologists have done much to throw light upon the wider biological problems which embryology presents. EMBRYOLOGY OF A LEPIDOPTEROUS INSECT. 17 It is well known that all animals during their embryonic life undergo a series of remarkable changes, both in form and structure. The earliest embryonic appearance of widely different animals is such that it is difficult to say even to what class the embryo belongs, but as development proceeds, the characteristic features of the class are developed. When we come to consider the embryonic conditions of genera and species we find that the similarity of their early stages is much more pronounced, the likeness extending even to small matters of detail. It is possible to limit the study of the embryology of insects to the changes that take place within the egg, but it is well known that the larvre and pupte of lepidoptera are essentially embryonic conditions, leading up to the production of the imagines. At the same time, their independent life, their competition in the struggle for existence, and the different conditions of their environment, have led to the formation of habits, and given rise to peculiar characters, which more or less obliterate, as it were, their true embryonic characters. It is necessary, therefore, in dealing with these stages (larval and pupal) to bear in mind two points : (1) Whether the similarities which one sees are phylogenetic, that is, whether they are due to the transitory re-appearance of the characters of a bygone epoch in the ancestral history, or, (2) Whether they are oecological in their origin, and due to a similar relationship of the animals to their organic and inorganic environment. The characters manifested in the egg-state must almost of necessity belong to the first division ; those in the active larval (considered as an embryonic) condition may belong to the first or second. It will be seen, then, that such phylogenetic conditions as the embryological stages of insects offer, indicate the lines of descent through which the species have passed. The complete study of em- bryology must, in time, give us much more correct notions of actual relationships than any other line of enquiry ; for it is highly probable that the embryonic stages show us, more or less completely, the lines through which the ancestral form has been developed, to produce the present condition of its offspring. It is to embryology, therefore, that we must look to furnish the clues to the true relationships which exist between animals, and a true genealogical classification can only be formulated by the aid of the knowledge which it contributes. We aim at obtaining a " natural " system of classification of insects, i.e., an indication of the line of descent of the various species we study, and their connection with each other, and, hence, for this purpose, the structure of the embryo is often of more importance than that of the adult. Darwin says : "In two or more groups of animals, however much they may differ from each other in structure and habits in their adult condition, if they pass through closely similar embryonic stages, we may feel assured that all are descended from one parent form, and are, therefore, closely related. Thus, community in embryonic structure reveals community of descent ; but dissimilarity in embryonic development does not prove discommunity of descent, for, in one of two groups, the developmental stages may have been suppressed, or may have been so greatly modified through adaptation to new habits of life, as to be no longer recognisable. Even in groups in which the adults have been modified to an extreme degree, community of origin is 18 BRITISH LEPIDOPTERA. often revealed by the structure of the larvae As the embryo often shows us, more or less plainly, the structure of the less modified and ancient progenitor of the group, we can see why ancient and extinct forms so often resemble, in their adult state, the embryos of existing species of the same classes Embryology rises greatly in interest, when we look at the embryo as a picture, more or less obscured, of the progenitor, either in its adult or larval state, of all the members of the same great class." We may now look briefly at the embryonic life of a lepidopterous insect from the time of the fertilisation of the ovum, until the larva hatches from the egg. This can only be done by the aid of a micro- scope. A very simple instrument with two lenses, a f and , is sufficient for ordinary purposes, although, of course, many other accessories are exceedingly useful. To get eggs for this purpose, take an ordinary glass tube and enclose a few females of some common Tortricid moth. These moths will usually lay their eggs on the glass, and their eggshells are so trans- parent that the changes may be readily observed. Among the butter- flies, eggs of Pararge inegaera and Nemeobitis lucina are not at all unsuit- able for observation. It is sometimes inconvenient to study the embryological changes which go on in an egg under a microscope, at the time that they actually occur. Two very good methods have been described in detail, by which the eggs may be killed and preserved for future observation. One of these is the distribution of the eggs in phials, one phial to be filled with carbolic acid, an egg put in, and the phial stoppered on each day, until the final one contains the newly-hatched larva. The other is to kill by heating in water at 80 C., then puncture the eggs with a fine needle, and stain with " Grenachar's borax carmine " or " Czochar's cochineal." It is an established fact of science, that every living being is evolved from a single unicellular germ. The egg in insects is not the earliest condition of the creature, because the primitive ovule can be traced back to the ovariole, or even to the primitive ovary, before the ovariole is developed. There is no need here to enter into the development of an ovum from the primitive ovary, as it is fully de- scribed elsewhere. Suffice it to say, that the ovum at last is formed in the egg-chamber, and consists of a mass of yelk surrounded and embedded in protoplasm, and containing the female pronucleus, whilst at the time that the egg is laid, the main mass of it is made up of yelk-spherules. These spherules become granular, and the granules gradually replace the spherules, and are themselves again changed into yelk-cells, the probability being that they are thus changed in order to form suitable nourishment for the young embryo. At this time, the newly-formed blastoderm-cells begin to pass towards the circumference, leaving the degenerated yelk-cells in the centre. In addition to these yelk-spherules, the egg contains a homogeneous fluid, which has the ordinary composition of protoplasm, and consists essentially of the chemical elements, carbon, hydrogen, oxygen, nitro- gen, sulphur, phosphorus, lime, soda, potash, and other substances in minute proportions. The great characteristic of this protoplasmic * Entom. Record, vol. v., p. 212. EMBRYOLOGY OF A LEPIDOPTEROUS INSECT. 19 fluid is its vitality, its ability to break up and sub-divide, to develop cellular structure, and to build up tissue from the cells produced by cell-division. After fertilisation, the protoplasmic fluid inside the ovum remains in a homogeneous condition for a certain time ; this varies for different species, but is comparatively constant in the same species. The first change that the protoplasm undergoes is that of the ordinary yelk segmentation, but, once this is set up, development continues generally with more or less rapidity. The segmentation starts at a point on the surface of the yelk called the " first segmenta- tion nucleus," and this nucleus undergoes cell-division in such a manner, as to form a superficial blastodermic layer. Side by side with this process of segmentation, the yelk separates from the outside cell- wall, and appears to become enveloped in a sac. The blastoderm layer (or layer of segmentation cells) has an elongated ventral plate formed in it, and in this the development of the embryo commences. This ventral plate broadens anteriorly, but the posterior part is divided transversely into segments. This development is at once followed up by the formation of a longitudinal depression, the outer sac gradually enclosing this depression on either side, until, at last, the opposite sides of the epiblast, or outside layer of cells undergoing segmentation, unite over the depression, leaving it as a longitudinal tube. This becomes detached as a solid cellular mass, which splits into two longitudinal (mesoblastic) bands. At this period it would appear that the amnion is formed. Of this, Osborne says : " After the yelk has become surrounded by the growth of cells called the blastoderm, and, after the germinal stripe, or foundation of the embryo, has been differentiated along one side of this blastoderm, a double fold of the latter grows up all round the cir- cumference of the germinal stripe, and finally closes in over it, the edges of the fold fixing together, and the two layers (of blastoderm) of which it is composed, at the same time separating from one another. The inner of these, continuous with the embryo itself, and lying im- mediately over it, is the amnion ; the outer, continuous with the blastoderm surrounding the yelk, is the serous membrane. Two sacs are thus formed, the one within the other, and between them lies the yelk. In the lepidopterous egg, the yelk next finds its way into the space between the amnion and the serous membrane, flowing over the former and depressing it and the embryo beneath it, till both are completely submerged in yelk, and consequently hidden from view." After this the mesoblastic bands become divided into somites, and the first traces of the abdominal segments may be noticed, followed by the appearance of the three thoracic segments. The somites coalesce, and the common body-cavity thus enclosed, is called the coelom. The three thoracic segments bear legs. The head, which appears to be formed of four segments, and the eye-spots, of which there are two clusters (each made up of six ocelli), placed one on either side of the second segment of the head, reckoning from the front, are then developed, followed in" turn by the ventral prolegs. The inner part of the hypo- blast is absorbed to form the alimentary canal. The cells, now con- tained between the outside wall of the egg and the newly-formed alimentary canal, divide up into clusters, which are gradually differ- entiated into the various internal organs. The first of these to be formed is the dorsal vessel, which is so called because it is placed in 20 BRITISH LEPIDOPTKRA. the dorsal part of the larva ; this corresponds with the heart of the higher animals. The other organs gradually undergo differentiation, and the mouth organs also become developed. At this period of development faint pulsations of the dorsal vessel are discernible. The separation of the alimentary canal into an oesophagus, a widened sac or stomach, and another contracted tube or intestine is clearly discernible, whilst the outer proteid part of the egg-contents is probably absorbed by cutaneous endosmosis. The tracheae are developed from the spiracles inwards, but do not become visible until injected with air. Such are the broad outlines of the larval development within the egg. From a tiny mass of protoplasm in the yelk of the egg, we get a larva produced such as we know it when newly-hatched. The egg- shell of most of our larger species is too opaque to allow these changes to be seen, but they can be readily observed, as we have already stated, in the eggs of Tortricids or Pyralids, owing to the thinness of the walls of the eggs in these groups. During the first stages of embryonic development, the ventral side of the embryo is external, or lies along the inner concave side of the egg, development commencing (as is usual in the Articulata and Vertebrata) on the ventral side of the insect. As development proceeds, the embryo changes its position, on account of the turning of the anal segment and its gradual upward movement, and that of the growing segments behind it, along the venter. In this manner the ventral part of the embryo gets turned towards the centre of the egg, whilst the dorsal part is turned towards the outside. Our observations of these movements were made on the embryo of Peronca (Tortrix) fernujana. We found that when the embryo' begins to show traces of segmentation, the thoracic segments are seen to develop three pairs of jointed buds or legs. At this time the embryo occupies a somewhat curved position, with the head slightly bent round towards the anal extremity, but with the legs outside, i.e., the larva is bent back on itself so as to form a curve agreeing roughly with the curvature of the shell, with what afterwards becomes the ventral sur- face of the larva outside, and the dorsum towards the centre. The embryo then gradually changes its position, the anal segment curling round and being pushed by the growth of the preceding abdominal segments, slowly up the ventral surface of the larva; whilst the dorsum gets pushed out, as it were, towards the centre of the egg. During this process the embryo becomes shaped something like the letter the movement continuing until a complete reversal of the embryo has been effected. The next stage is that in which the head and anus are in contact, each half running almost parallel, and this again is followed by an almost circular position, in which the dorsal area is now outside, and the ventral surface (with the legs) on the inside. The head, during all this time, scarcely changes its position. Very little further change in position takes place, the embryo, by this time, occupying all the available space in the egg. With regard to the change in position that the embryo undergoes in the egg, Chapman says that at the time that the ventral surface is towards the margin of the egg, the dorsal surface, or rather dorsal aspect, is still applied to the yelk-sac. At this time the dorsal sur- face is still broken by the umbilical opening, but, when the latter closes, EMBRYOLOGY OP A LEPIDOPTEROUS INSECT. 21 the young larva is truly a larva, possessing no organic connection with the other egg structures, and may no longer be regarded as an appen- dage to the yelk-sac. The first use it makes of this liberty is to assume the S or pot-hook shape, continuing until at length its position is reversed, the dorsum being along the circumference of the egg and the venter being central. The head and tail sometimes merely meet (in the flattest eggs), sometimes slightly overlap, whilst in the dome-shaped eggs, the head so overlaps as to take, very often, a central position in the vertex of the egg, forming a dark spot there, as in Acronycta, Callimorpha, Hesperids, and many others. The essen- tial importance of this observation is that it shows that the em- bryonic position of the nervous system is the same in insects as in vertebrates, and since it must, therefore, be identical also in the mature animal, it follows that the venter of insects corresponds, ana- tomically, with the dorsum of vertebrates and vice versa. Another important point with regard to this movement is, that whilst the larva is still truly an embryo, i.e., attached to the yelk and egg-structures, it has the venter outwards, but when the embryo becomes free, it moves as it likes, although this particular movement goes on so slowly, and without any apparent voluntary or even muscular effort, that it appears to be due to the mere force of the growth and development of the larva. During all this time, the disappearance of yelk has been taking place, but just when the embryo has attained its full growth, voluntary efforts to swallow are apparent, and the remainder of the yelk dis- appears. The remaining fluid is either absorbed by the larva through the skin, or evaporates through the shell ; the tracheae become visible by becoming filled with air, and the larva usually begins soon after- wards to commence eating its way through the shell. It would appear from Jeffrey's observations that the tracheae come rather suddenly into view, at the time that they are first distended with air. He states that " the filling of the tracheae commenced in the posterior segments, a sort of cloud gathering at the band where it is close to the head and in a line with the eye." He says : " I saw an apparently dark flood start from this spot, and, creeping along with a spasmodic effort, filling the branches, in its course, till it reached the head, and the whole of the tracheae became conspicuously visible on that side of the body." The same observer describes how the dorsal vessel (heart) became visible in an embryonic Botijs hyalinalis, on the tenth day after incu- bation. The pulsations were at first (8 a.m.) very faint and feeble, taking place somewhat irregularly at long intervals of 20 and even 30 seconds ; but, after a few hours, they became more distinct, with shorter intervals between each beat, and became still more ac- celerated by the evening of the same day. Two days afterwards, a beautifully clear view of the heart and its action was obtained, the pulsations being timed at 40 per minute, increasing to 60 a few minutes before the larva escaped from the egg. The important part played by the blood-tissue in larval nutrition, together with the supposition, entertained for many years by certain eminent naturalists, that circulation of the blood did not take place in * Ent. Mo. May., vols. xxii. and xxiii. 22 URITISH LEPIDOPTERA. insects, has led to considerable discussion. The origin of the "blood- tissue " was worked out at length by Graber, who concludes that the whole of the structures forming this " tissue," viz., oenocytes (certain cell-masses), fac-body and blood-corpuscles, are ectodermic structures. He further finds that the oenocytes are metamorphosed into the fat- body, and that the blood corpuscles arise from the fat-body, and, probably, also directly from the oenocytes. Wheeler, f however, looks upon the fat-body as a thickened part of the inner coelomic wall, due to an accumulation of fat-vacuoles in the cytoplasm of the mesoderm-cells." He further concludes that the fat-body is not derived from the oenocytes, is of mesodermal, not ectodermal, origin, and concludes that there is no evidence for the origin of the blood from the oenocytes. Wheeler also remarks that " Few insects appear to be better adapted for tracing out the origin of the oenocytes than the Lepidoptera. This is especially true of the larger Bombycid moths. That the segmental cell-clusters arise by delamination from the ecto- derm was conclusively made out in the embryos of Platytamia cecropia and Telea polyphemus. Each cluster is several cell-layers in thickness, and lies just behind, and a little ventral to, an abdominal stigma. The succulent cells constituting the cluster are at first polygonal from mutual pressure, but, as the time for hatching approaches, they become rounder and more loosely united. I have not traced them through the larval stages, and merely record these fragmentary obser- vations because they completely confirm Tichomiroff 's and Graber's observation on the origin of the oenocytes from the ectoderm." The study of the lepidopterous embryo has given us many other interesting morphological particulars. Kowalewski found ten ab- dominal somites in the embryo of Swerinthus populi, all bearing pro- legs ; whilst Tichomiroff detected eleven abdominal somites in the embryo of Bombyx mori, all provided with prolegs except the first. Graber also found the abdomen of the lepidopterous embryo to consist of eleven true segments, and observed that the abdominal segments of Eutnchd (Gastropacha) quercifolia were at first devoid of appendages, and that, when they did appear, they developed only on those seg- ments on which they persist in the adult. The mode in which the earliest development of the generative organs in the embryo of insects takes place is very obscure, but it would appear that the primitive ovaries are composed of a mass of cells, produced by an infolding of the ectoderm. Some writers, however, consider them to be derived from the mesoderm, whilst others trace their origin back to certain so-called pole cells, which originate even before the blastoderm is formed. However this may be, it would appear that they are, in that early stage, quite indistinguish- able from the other blastoderm cells. As development proceeds, the great mass of cells become differentiated into various structures, which subserve a special purpose, or perform a certain function. Certain cells in the ovary, however, retain their primitive condition, and, with it, the power, under suitable conditions, of forming another in- dividual of the same species. On this subject, Woodworth writes : " About the time of the completion of the blastoderm, the already * " Ueber die embryonale Anlage des Blut- und Fett-gewebes der Insekten," BioL Centralbl., Bd. ii., Nos. 7-8, pp., 212-224. f Psyche, vol. vi., p. 255 et. seq. EMBRYOLOGY OF A LEPIDOPTEROtS INSECT. ao differentiated ventral plate infolds at a point on the median line about two-thirds from the upper end, and forms a very narrow pocket. The cells composing it look like the rest of the cells of the ventral plate at this time; they are almost round, and have a lining on one side, made of the grey matter which originally bordered the whole egg, but which became a part of the blastoderm cells. The pocket remains open but a short time, but there is a long depression at the upper end of the bunch of cells. The mass of cells is soon cut off from the ventral plate, and they are then free in the body cavity, but remain in contact with the ventral plate at the point where they were produced. Later stages show that these cells produce the generative organs. The generative organs thus appear to be pro- duced by an infolding of the ectoderm, or possibly of the blastoderm, before the ectoderm is produced, but from a portion which is later to become ectoderm. The general idea has been that the generative organs in insects are produced from the mesoderm, although Metsch- nikow, as early as 1866, showed for certain insects a different origin." Those further interested in the details of this subject would do well to refer to the writer's chapter on the " Embryology of a lepidop- terous insect," Ent. Record, vol. v., 1895. CHAPTEE IV. PARTHENOGENESIS OR AGAMOGENESIS IN LEPIDOPTERA. IT is generally necessary, among the Lepidoptera, that the two generative elements should unite before the fertilisation of the ovum can take place, and, since these elements are always developed in different individuals, it follows that copulation between the sexes is necessary for fertilisation, and for the subsequent production of young. It appears, however, that under certain conditions copulation is not necessary to ensure the production of young, since, occasionally, eggs will produce larvae without the union of the sexes, and larvae thus pro- duced have been recorded as developing in the ordinary course into fully matured and fertile imagines. It is a well-known fact that, under ordinary circumstances, the eggs of almost all lepidopterous insects undergo certain changes after being laid. Some of these are common both to fertilised and unfertilised eggs, and since they must be looked upon as the outward sign of a change that is taking place within the egg, it is probable that the first changes which take place in the egg, i.e., the very first stages of embryonic growth, are independent of fertilisation. The changes which take place in the unfertilised eggs of some species are much greater than those which take place in others, and there are, aa previously stated, cases on record in which development has proceeded so far, that the growth of the embryo has been completed, and a larva has hatched from the unfertilised egg. We see, then, that, under special conditions, nature produces progeny from virgin females without the intervention of the male, The production of such progeny among bees has long been known. 24 BRITISH LEPIDOPTEKA. Virgil refers to it in the Georyics, and the old authors termed the phenomenon, " Lucina sine concubitu." It is now known as " aga- mogenesis " or "parthenogenesis." It must be confessed that scientific experiments, conducted with sufficient care, relating to this subject, have been rarely performed, and that the evidence rests largely on chance observations. Still, there can be no doubt that some of the experiments, at least, have been sufficiently accurate to necessitate a scientific explanation of the phenomenon. It would be out of place here to discuss the general question of reproduction in the lower Invertebrates, a brief summary of which may be found, Entom. Record., v., pp. 219 ct seq. It need only be mentioned that fission or cleavage, gemmation or budding, and encystation are the more general means by which it is effected. In the Hydrozoa, reproduction is carried on all the summer by gemmation, but in the autumn, sperm cells and germ cells are produced in the same individual, the former fertilising the latter, which then become ova, in which stage these creatures pass the winter. This method of sexual reproduction (i.e., with both sexes in the same individual) is very common in the lower animals, but among the higher invertebrates the sexes are usually differentiated in separate individuals, and, as a rule, coition is necessary for reproduction. This is the ordinary condition among insects. Among the Crustacea such species as Polyphemus oculus, Apiis can- criformis and Limnadia yiyas consist, Newman says, almost entirely of female individuals, the presence of a male being the exception. Daphnia has males as well as females, but, according to Lubbock, the females appear equally prolific in the absence of the males. Newman also states that in some Arachnids the fertility of the female is not dependent on coition with the male. He instances Epeira diadem a, which he states invariably produced fertile eggs without union with a male. Among insects, the agamic reproduction of Aphides has long been well understood. This, however, is rather different from the partheno- genetic phenomenon presented by Lepidoptera, Hymenoptera, etc. In the former, viviparous young are produced by the females ; in the latter, eggs are laid, and produce larvas in due course, without the usual intervention of the spermatozoa. Most of the records of the occurrence of parthenogenesis in Lepi- doptera are, from a scientific point of view, most unsatisfactory, and based on chance observation, rather than on specially devised experi- ments. This is, perhaps, due to the fact that those entomologists who inbreed insects in the largest numbers, do so in order to obtain fine specimens for collections, and, as a matter of course, pair the females with males in order to ensure the due fertilisation of the eggs. It must also be borne in mind that, so far as our observations have gone, those species that show a parthenogenetic tendency, only lay a very few eggs in an occasional batch, that will produce parthenogenetic young. A very large number of female motbs, therefore, would have to be sacrificed in order to obtain a very small number of parthenogenetically fertile eggs. This does not apply, however, to the Psychids, where parthenogenesis, in some species, appears to be the rule rather than the exception. This has been clearly shown by Jourdan in the case of Bombyx PARTHEXOGENESIS OR AGAMOGENBISIS IN LEPIDOPTERA. 25 mori fComptes Rendus Hebdomadaires des Seances de V Academic des Sciences, Paris, liii., 1861, pp. 1093-1096), where he remarks that it has long been customary, in the silk-producing countries of France, to regenerate a worn-out race by using "la graine vierge," i.e., eggs pro- duced from females that have not been paired with males. He details certain experiments made in 1851, which show the proportion of female moths that give fertile eggs parthenogenetically. From these experiments we learn that he had 300 yellow Milanese cocoons of a form of B. mori, that gives only one generation per year. The results work out as follows : June, 1851 300 cocoons selected, each cocoon placed in a small cardboard box covered with gauze, so as to com- pletely imprison the moth on emergence. The 300 cocoons produced 147 females and 151 males. The boxes containing males were re- moved and the females carefully preserved without being uncovered. Of the 147 females, six gave fertile eggs. Two gave 7 eggs each, two others 4 eggs each, one gave 5 eggs, and one 2 eggs. These 29 eggs, preserved in their respective boxes without being uncovered, to render error impossible, hatched May, 1852. Many other eggs, it is men- tioned, passed from the pale yellow (colour when newly-laid) to the slaty-grey hue, which replaces the former after some days in fertile eggs. The summarised results of this experiment worked out at : 147 females, laid about 58,000 eggs, of which 29 produced larvae, i.e., about 1 : 2,000. Another experiment was made by Jourdan, in July, 1851, on white cocoons from South China, of a form of B. mori, giving five or six successive generations in one year. Fifty cocoons were separately isolated, as in the last experiment. From these emerged 23 females and 26 males. Seventeen of these females gave completely fertile eggs. One gave 113, and the least productive 12. The total number of eggs laid was 9,000, of which 520 produced larvae. This gives a proportion of 1 : 17. They hatched seventeen days after being laid. Although these experiments proved conclusively that some virgin females of B. mori could reproduce their kind without copulation, it was evident from the results, that the parthenogenetic reproductive power was exceedingly feeble. Of the two different races experimented upon, that with five or six successive generations per year was much more productive, parthenogenetically, than that with a single generation. One of the earliest essays on this subject was that of Von Siebold (translated by Dallas), entitled: On a true parthenogenesis in moths and bees. Siebold was led into his enquiries by some observations made on the reproduction of a species of Psychid moth, which, he noticed,, propagated without copulation. He followed this up with observations on bees and B. mori, and found that the phenomenon of reproduction by virgin females was not at all uncommon. For this, he adopted the term "parthenogenesis," which had previously been applied by Owen to the phenomenon now known as " alternation of generations." According to Siebold, we learn that the oldest communication relative,to reproduction by female insects, sine concubitu, was made by Albrecht of Hildesheim, who (in 1701) relates that he found a brown pupa in a cocoon on a black-currant bush, and preserved it to see what moth would emerge from it. At the end of July, a moth of yellowish- white colour was disclosed, and in a few days laid a great number of eggs, and then died. In April of the following year, Albrecht was 26 BRITISH LEPIDOPTEBA. astonished to find young black caterpillars in the box, instead of the eggs. His communication to the Leopoldine Academy of Naturalists shows that he was satisfied that copulation had not taken place. In 1772, Bernoulli recorded that Baster had obtained fertile eggs from an isolated female of Gastropacha qiierdfolia, that had been bred from a caterpillar ; and further, that a caterpillar of Episema (Diloba) caendeo- cephala, having changed to a pupa, the latter was left in a closed box, and that, about fifteen days after, he was surprised, on opening the box, to find, besides the enclosed moth, a family of young caterpillars, which had already devoured the pupa-case of their mother, and a portion of their own egg-shells. Denis and Schiffermuller pointed out, in 1776 (Syst. Verz. der Schmett. der Wiener Gey end, etc., p. 293) that these cases were possibly errors of observation ; whilst Von Scheven considered that the larv were probably from eggs laid by another female moth, previously confined in the same box. Siebold, being very dissatisfied with what was known about the subject, turned his attention to the " case-bearers," Solenobia lichendla and S. triquetrella, and during the years 1850-1852 (the date of Jourdan's experiments on B. mori) he collected several hundred cases. None but females emerged from these cases, and they commenced almost immediately to lay eggs. They " possessed such a violent impulse to lay their eggs, that, when I removed them from their cases they let their eggs fall openly. If I had wondered at the zeal for oviposition in these husbandless Solenobia, how was I astonished when all the eggs of these females, of whose virgin state I was most positively convinced, gave birth to young caterpillars, which looked about with the greatest assiduity in search of materials for the manufacture of little cases ! " Parthenogenetic reproduction in Solenobia lichenella had also been observed by Wocke and Keutti. For many years the female of Apterona crenulella (Psyche helir] only was known, and Siebold, to make sure that none of the " wingless and footless moths " were males, dissected many. He satisfied himself that all were females, and their unfertilised eggs were found to develop larvae in the same year. In 1795, Constans de Castellet, general inspector of the silk industry in Sardinia, had reported to Reaumur that he had reared caterpillars from unfertilised eggs of Bombyx mori. " Ex nihilo nihil fit," was Reaumur's sceptical reply. Herold, in 1838, reported that amongst the unfertilised eggs of B. mori, some here and there passed wholly or partially through the same changes as fertilised eggs, although they failed to hatch, and he distinguishes (Dis. de anim. vert, caren. in ovo formatione, Fasc. ii., 1838, Tab. 7, fig. 31) between the fffitus developed from fecundated, and that developed from unfecundated eggs, the former escaping as a larva, whilst the latter perishes in the egg- shell. He distinguished readily, also, various degrees of the faculty of development of unfertilised eggs, which manifested themselves by infinite differences in the disposition, number, form, and strength of the coloured portions of the egg. Herold was able to extract a fo3tus from one of these unfertilised eggs in the middle of winter. According * The male of Apterona crenulella (Psyche helix) was re-discovered by Clauss. He described and figured the larval case of the male, the difference between the pupae of the sexes, and the male imago in Zeits. Weiss. Zool., xvii., p. 470. Until then it does not seem to have been noticed since the time of Reaumur. PARTHENOGENESIS OK AGAMOGENESIS IN LEPIDOPTERA. 27 to Herold, embryos were not developed in all the unfertilised eggs examined, nor did he know of any case in which such embryos emerged from the egg. As far back as 1669, it may be mentioned that Malpighi was well acquainted (Marc. Malp. Diss. de Bombyce, Lond., p. 82) with these differences. He also then knew that the eggs of Lepidoptera were not fertilised at the time of copulation, but that each one was afterwards fertilised separately. Siebold quotes, on the authority of Filippi, that Curtis had received an isolated chrysalis of Telea polyphemm from America, from which a female emerged, all of whose eggs developed, adding that he believed a similar occurrence sometimes took place in B. mori. Filippi relates that, in 1850, he observed the phenomenon in that variety of the latter species known as trerotini, which has three broods in a year. He also states that Griseri had also observed that many eggs of virgin females of B. mori developed. Siebold observes that various silkworm breeders in Breslau and Munich gave him similar information, and that he himself noticed exactly the same well-known change of colour, which took place in the fertilised eggs of this species, occurring in a large number of unfertilised eggs, although many stopped at various stages, only becoming reddish or violet, whilst only a very few went through the entire series of colour-change to slaty-grey. Siebold ob- tained no larvae from them, but, in 1854, he received unfertilised eggs from Schmid, which produced larvas. He tells us that he expected to breed only males, due to his having read Lacordaire's account of Carlier's observations, that " he obtained, without copulation, three generations of Porthetria (Liparis) dispar, of which the last gave only males, which naturally brought the experiment to an end." Siebold, however, bred both males and females, which copulated freely, and appeared to have the ordinary amount of vitality. Kipp had pre- viously recorded the rearing of both males and females from some unfecundated eggs of Smerinthus populi. A brief summary of what has been observed in this country (with a few incidental outside observations) may now be useful. Newman in 1856, gave a list of Lepidoptera in which the phenomenon of par- thenogenesis had been noticed up to that date. These were : Sphinx ligustri, Smerinthus populi, S. ocellatus, Porthetria dispar, Psilura monacha, Diloba caeruleocephala, Telea polyphemus, Saturnia pyri, S. pavonia, Orgyia gonostif/ma, 0. antiqua, Bombyx mori, Lasiocampa querciis, Arctia caia, A. villica, A. casta, Dendrolimus pini, Cosmotriche (Odonestis) potatoria, Eutricha (Gastropacha) quercifolia, Sterrhopterix hirsutella (Psyche Jusca), Apterona crenulella (Psyche helLv), Canephorfi unicolor (Psyche graminella), Fumea casta (Psyche nitidella), Solenobia triquetrella, S. clathrella, S. lichenella. The observations on which this list were based are sometimes of a very unsatisfactory nature, but others are more convincing, e.g., Tardy's experiments with L. querciis, in which three generations of perfectly vigorous and full-sized moths were reared without a single ^coition- having taken place. Mory of Basle (Ent. Eec., vi., p. 209) recently obtained larvse from unfertilised eggs of this species. A note in the Ent. Weekly Int., iii., pp. 175-176, states that parthenogenetic females of Solenobia inconspicuella had been bred, whilst in the Ent. Rec., vi., p. 89, Freer records the rearing of Talaeporia pseudobombycetta parthenogenetically. Douglas (Substitute, p. 78) states that he has bred 28 BRITISH LEPIDOPTERA. Fume a nitiddla from what he believed to be unfertilised eggs ; the evidence, however, is here very unsatisfactory. Newman (Entom., ii., p. 28) records larvae from unfecundated eggs laid by a female Phiyalia pedaria. These in due time became pupae, but no imagines were reared. Eaton (Entom., iii., p. 104) records an instance in which parthenogenetic progeny of Oiyyia antiqna were reared to the third generation. The details are : First generation. From a pupa found at Venn Hall, Sherborne, Dorset, in the autumn of 1864, a female imago emerged, which laid eggs. Second generation. Of the above- mentioned eggs, ten hatched in the spring of 1865, but of these larvae, one only, the largest from the first, came to maturity ; this produced a female which laid eggs. Third generation. Five larvae from these eggs attained the pupal state of development, and one of them pro- duced a female imago by the middle of October. Mo further details are given. It may now be well to summarise some of the records under the various families to which the species belong : SOLENOBIIDES. Solenobia inconspicuella, vide Ent. Rec., vol. vi., p. 5, where it is stated that par- thenogenesis in this species is well known. S. clathrella, mentioned by Newman in Phys. Characters in Classif., 1856. S. lichendla, Wocke anu Reutti (teste Siebold). S. triquetrella, Siebold's experiments on this species, and S. lichenella, have already been detailed, see Ent. liec., v., pp. 292-3. Talaepona pseudobombycdla, Freer, Ent. Rec., vi., p. 89, very many parthenogenetic larvae obtained. PSYCHIDES. Apte- rona crenulclla (Psyche helLv), was experimented upon by Siebold, and dissections of the parthenogenetic females were made. At this time, the male of this species was unknown. Canephora vnicolor (Psyche yramindla) and SterrhopterLv hirsutella (P. fusca) are both mentioned in Newman's list. Fumea casta (nitidella) rests as a par- thenogenetic species, on Douglas' unsatisfactory evidence. NOTODON- TIDES. Diloba caeruleocephala, Bernoulli, 1772, records the hatching of unfertilised eggs. Notodonta du-taroidi's. There is a record made by Alderson, as to the probability of this species being parthenogenetic, Entom. Bee., vol. i., p. 96. Cerura vinula. Alderson notices unfertilised eggs of this species hatching, Ent. Rec., i., p. 95. LIPARIDES. Porthetria dispar, Carlier (teste Lacordaire), records three generations without copulation having taken place. Pearce, Ent., xii., p. 229, obtained larvae May 6th, 1879, from unfertilised eggs. Weijenbergh, Archives Neerlandaises, v., 1870, pp. 258-264, records that fertile eggs of autumn, 1866, hatched April, 1867, and produced imagines, August, 1867 ; from these, without fecundation, eggs hatched April, 1868, and imagines appeared August, 1868 ; from these, again, without fecundation, eggs hatched in April, 1869, imagines in August, 1869 ; from these, without fecundation, eggs did not hatch in spring of 1870, but dried up. Ladia coenosa. Brown, Entom., v.,p. 395, an isolated female emerged, laid more than 60 eggs, which duly hatched and were sent to Hellins. Oryyia antiqiia. Eaton, Entom., iii, p. 104, data already given. O. ijonoKtiijma and Psilura vionacha. Mentioned by Newman, Essay I'liyx. Charac., etc. ARCTIIDES. Arctia caia, A. villica and A. casta, are mentioned by Newman in his Essay Phys. Charac. Spiloxoma mendica. Bowell, Ent. Ptec., i., p. 174, obtained a batch of ova from female just out of pupa case, of which 15 hatched. NOCTUIDES. Anarta myrtilli. Watson, Entom., xv., pp. 261-2, records putting a PARTHENOGENESIS OR AGA3IOGENESIS IN LEPIDOPTERA. 29 pupa in a closed box, that the latter was opened some time after, and contained a female imago and many young larvae, dead. LASIO- CAMPIDES. Lasiocampa quercm. Tardy (teste Newman) bred three parthenogenetic broods of perfectly vigorous and full-sized moths. Mory of Basle, Soc. Ent., April 1st, 1895, also records many larvae from unfertilised eggs. L. trifolii is recorded by Bouskell, Trans. Leices. Lit. Soc., iv., p. 422, as laying a few unfertilised eggs in 1896, which hatched in the following spring, and shortly afterwards died. Eutricha (Gastropacha) quercifolia. Baster (teste Bernoulli) ob- tained fertile eggs from an isolated bred female, Cosmotriche (Odonestis) potatoria and Dendrolimus pint, are both mentioned by Newman in his Essay Phys. Charac. BOMBYCIDES. Bomby.v mori. Many cases have already been dealt with at length. SATURNIIDES. Saturnia pavonia. Noticed by Newman, Essay Phys. Charac. ; also by Bouskell, Trans. Leic. Lit. Soc., iv., p. 422, who mentions that a female laid six eggs in her cocoon, she being unable to get out ; these all hatched. S. pyri is mentioned by Newman, Essay Phys. Charac. TeJea polyphemus. Curtis (teste Filippi) obtained fertile eggs from a moth that emerged from a single cocoon in his possession, and that had come from America. SPHINGIDES. Smerinthus ocellatm. Newman, Essay. Phys. Charac. Class. ; Brown, Entow.., v., p. 395 ; Headly, Trans. Leices. Lit. Soc., iv., p. 421, the latter mentions that of the eggs laid, 75 per cent, hatched. S. populi. Kipp (teste Siebold) reared both sexes from unfecundated eggs ; Newman, Essay Phys. Charac. Class. ; Bouskell, Trans. Leices. Lit. Soc., iv., p. 421, mentions twenty out of seventy eggs hatching. S. tiliae. Brown, Ent., v., p. 395, no data. Acherontia atropus. Geddes and Thompson, Evolution of Sex, no data given. Sphin.f lit/ustri. Newman, Essay Phys. Charac. in Classif. ; Nix, Entom., iv., p. 323, all eggs hatched in this brood. Clogg, Entom., v., pp. 356-7, fifty eggs hatched out of the brood. GEOMETRIDES. Phigalia pedaria. Newman, Entom., ii., p. 28, records the laying of many eggs by three unfertilised females at end of February, 1864, and states that, on April 17th, the cage was swarming with newly-hatched larvae. He failed, however, to get imagines. Although it may safely be assumed that parthenogenesis does occur in Lepidoptera, yet, as we have just said, it must be confessed that the material based on true scientific experiment is not large, and that many careful observations based on the most exact experiments are required. The elucidation of the peculiar phenomena presented, is worth all the patience with which the entomologist must attack this subject, and he would have the reward of knowing that he had helped to make clearer one of the greatest mysteries of insect life. The phenomenon of parthenogenesis appears to me to be explicable only by supposing that the potency of the male element is handed down generation after generation, and that former fertilisations affect the embryo, independently of the actual union which fertilises the ovum. The male element must be looked upon as possessing, not only a great and direct influence on the development of the eggs im- mediately fertilised by it, but also on the eggs of successive issues not directly fecundated. That this is probably so, is shown by the fact that the unfertilised egg often undergoes varying conditions of deve- lopment, short of the actual development of a perfect embryo. This was foreshadowed in our notes on " the ovum," where the variation 30 BRITISH LEPIDOPTERA. and change of the colour of the egg are dealt with. In cases of par- thenogenesis, the influence must be powerful enough to cause full development, not only for one generation, but for one or more genera- tions beyond the one normally reached, and in this way may be explained the phenomenon that some species, which usually do not multiply without sexual intercourse, occasionally produce partheno- genetic young, even in cases like Sphinx liyiistri, Bombyx ntori, etc., where it could scarcely be expected. It is remarkable that, in most orders of insects, the parthenogenetic progeny is usually male, but, in the Psychidae among Lepidoptera, helotoky, or the production of parthenogenetic females, alone takes place. I may mention, in conclusion, that the great difference that exists between parthenogenesis (1) in the Psychids, where it appears to be, in some species, the rule rather than the exception, as it is in some Cynipids and in bees (as regards male eggs), and (2) in all other Lepi- doptera, where it is a rare and occasional phenomenon, is somewhat striking and important. This difference, no doubt, is only one of degree, but so great a degree as to be parallel to a difference in kind. It is quite possible, too, by means of the Psychids, to ally the partheno- genesis that takes place in Lepidoptera with that known to occur in the Cynipidae, and the phenomena might perhaps be brought into connection with a more primitive method of reproduction, e.g., gem- mation. I am quite clear that the modus operandi of parthenogenesis in Lepidoptera is still as obscure as ever, and that the explanation I have offered does not help matters much. It, however, is the only logical explanation that has occurred to me, and must be taken for, and only for, what it is worth. CHAPTER V. THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROU S LARVA. AT the time that the lepidopterous larva escapes from the egg, it possesses true insect characters. Its body is composed of a series of segments, containing the muscular, digestive, circulatory, respiratory, and nervous systems. It breathes by means of tracheae, a series of fine tubes composed of an elastic membrane, and kept open by a spiral structure, which passes throughout their whole length. The four segments of the head are now welded into an almost inseparable whole, and, although the first three body-segments are assigned to the thorax, there is no well-marked separation between the thoracic and abdominal regions. The .skin of the newly-hatched larva is very soft, but it quickly becomes harder, owing to the solidification of the horny substance called chitin in the outer cuticle. Usually larvae have a somewhat colourless skin when just out of the egg ; but the harden- ing of the cuticle is frequently accompanied by the production of a difference in colour, and by the development of the distinct markings which are characteristic of the larval cuticle, so that an almost colourless larva may, within an hour of hatching, become almost black. This hardening does not affect the sutures, and the interseg- THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 31 mental membranes allow the segments to move freely upon each other. The body segments are usually sub-divided into subsidiary rings or sub-segments, which also move more or less freely upon each other. The sub-segments are divided again into still smaller solid portions, which have a certain amount of freedom, and are technically called sclerites. A general description of a typical caterpillar or larva now becomes necessary. We find that the caterpillars of Lepidoptera are usually long and cylindrical, being, however, somewhat flattened on the ventral surface. They may be considered as being composed of a head, thorax and abdomen, as in the imago or perfect insect, although the distinction between thorax and abdomen is not, as previously noted, distinctly marked. The cylindrical shape of the larva depends upon the fact that the larval skin contains fluid under considerable pressure. The head is a somewhat horny, compact, oval case, and is furnished with a number of appendages about the oral opening. It is made up of four (or more) segments, which, however, are not distinguishable after hatching. On each side of the head are, usually, six simple ocelli, arranged in lunular form on the cheeks. The mouth consists of a labrum, mandibles, maxillae (with maxillary palpi) and labium (with labial palpi). The thorax is composed of three segments (those following the head), which are known as the pro-thorax, meso-thorax, and meta- thorax (or post-thorax) respectively. In all larvae which burrow under- ground or feed internally, and in many others, which do not, the dorsum of the pro-thorax is protected with a hard, corneous plate, often, indeed, extending to the meso-thorax and meta-thorax. Each of the three thoracic segments bears on the ventral surface a pair of more or less horny legs (the true legs), which have five joints and terminate in a single claw. The abdomen consists of the last ten segments of the caterpillar's body. They are very similar, in general appearance, to the thoracic segments, but the tubercles, or little chitinous hair-bearing knobs which they carry, are usually somewhat differently arranged, and they never give rise to true legs. Some of the abdominal segments, however, bear on their ventral surface a pair of stout fleshy protuberances, called pro-legs or claspers ; these prolegs are really extensions of the integu- ment, and have, on their free surface, a number of hooks. The number and development of the prolegs, and the arrangement of their terminal hooks vary considerably, different patterns distinguishing the several families, and even genera. Very little use is made of the true legs , for purposes of progression, this being accomplished almost entirely by means of the prolegs. The terminal segment of the abdomen is known as the anal segment. On either side of the first thoracic, and of the first eight abdominal segments, is $, tiny opening called a spiracle. The spiracles are round, oval, or longitudinal in shape, and are the channels through which respiration is carried on. Regularly pjaced on certain parts of the body are to be found little chitinous, hair-bearing knobs (sometimes modified into fleshy elevations), which we have already said are called tubercles. These may give rise to single hairs, but sometimes to exceedingly close and dense fascicles. These tubercles are often strikingly modified at each successive ecdysis or change of skin. 32 BRITISH LEPIDOPTERA. The lepidopterous larva, by its active, independent existence, under- goes special modification and development, in order to protect itself from its various enemies. Hence the larvae of different species assume, by modification, a manifold variety of shapes, and of arrangement of the various external structures hairs, tubercles, etc. As the most specialised larvae present, therefore, such wide divergences from the original type from which they have sprung, it becomes necessary for us often to homologise the complex structures which they now bear with the simple structures from which they originated, and to do this a comparison must be instituted with those larvas which, from the exigencies of their environment, are but little changed from the more ancestral larvae. The newly-hatched larvae of many species, which are very specialised in their adult stages, have the specialised structures in a very simple condition, both as to form, structure and arrange- ment ; whilst many boring and case-bearing larvae are still more simple in the structure and arrangement of the tubercles, hairs and pro- legs, which are especially prone to be changed by external conditions. Larvae which show this simple arrangement of tubercles, hairs and prolegs, are often spoken of as generalised, in contradistinction to those in which the structures are complicated, and which are termed specialised, larvae. It must not be forgotten, however, that the most generalised of all lepidopterous larvae must be far in advance of the larvae of those insects (Orthoptera, etc.), with incomplete metamor- phoses. There can be little doubt that insects belonging to these orders, in which the metamorphoses are carried on within the very narrowest limits, and in which the various stages present but little change, inter se y are much more ancestral than the insects belonging to those orders in which the metamorphoses are distinct, and in which the various stages bear but little resemblance to each other. The Lepidoptera which have the most generalised form of larvae are the Eriocephalids, Micropterygids, Adelids, Tineids, Sesiids, Psychids, Hepialids, Zeuzerids and the Tortricids. A comparison of these with each other, and with larvae belonging to more specialised super-families, soon gives us a clue as to the lines on which modifica- tion has proceeded in the higher groups. Having glanced at the general structure of a lepidopterous larva, we may deal with a few of the organs in more detail. The head of a caterpillar is divided into two lateral halves by a suture, which divides, however, in the centre of the face, and leaves between its forked branches a triangular space. This frontal triangle is termed the clypeus, and is very often distinctly and characteristically marked. Just within, and parallel to the central facial suture, is a deeply-grooved furrow, which is the reverse of a ridge that faces in- ternally, and to which the muscles of the head are attached. The true sutural line is but little developed in newly-hatched larvae. Just below the clypeus is a short inconspicuous piece of chitin, welded to the clypeus. This is very distinct in some butterfly larvae (ix persists. This is the case in Hepialus. In the Sesiids, again, the circle of hooks is flattened antero-posteriorly, and is weak or wanting at the outer and inner ends, showing a relationship to Incurvaria. The anal prolegs very rarely have more than the anterior half developed. In Hepialus the circle is fairly complete. The Crambids have hooks of alternate size, like the Tortricids. Crambus often has three sizes of hooks alternated in one row." Attention is also drawn to the fact that the larvae of the Hesperids show, in their three rows of hooks, a persistence of Adelid (or, at least, very low) structure, whilst the adult larvae of the true butterflies have the same structure as the true " Macros." The same observer finally concludes that " the proleg seems to reach its full development with a complete circle of booklets. A higher development of the insect is not only accompanied by a fuller deve- lopment of the inner half of this circle, but also by tbe degeneration and disappearance of the outer half. This may often be followed out in ' Macros,' usually among the butterflies, where the young larva has ' Pyraloid ' prolegs, which often suddenly (at one moult), or more gradually (in two or three), assume, in the full-grown larva, the tHE EXTERNAL STRUCTURE OF fHE LEflDOPTEROtfS LARVA. 37 unilateral ' Macro ' type " (Trans. Ent. Soc. London, 1893). Prout has noticed that, in the Georaetrid genus Oporabia, the newly-hatched larva has a complete circle of hooks. We have already mentioned that the segments which usually bear the prolegs are the third, fourth, fifth, sixth and tenth abdominal. The Geometrids, however, usually have them only on the sixth and tenth abdominal segments. In the early stages of many Noctuid larvae, we find, however, only the merest traces of prolegs on the third and fourth abdominal segments ; these, however, usually develop com- pletely at the later ecdyses. The peculiar method of progression, characteristic of Geometrid larvae, is due entirely to the absence of the prolegs on the third, fourth and fifth abdominal segments, and those Noctuid larvae which do not develop prolegs on the third and fourth abdominal segments, until late in life, resemble the Geometrid larvae in their mode of progression, whilst a whole group of Noctuid moths, which never do develop them, retain the looping habit throughout, and have been called, on this account, by some entomologists, HEMI- GEOMETERS. In some Geometrid larvae, prolegs appear on other than the abdo- minal segments normally carrying them. The larva of Himera pennaria obtains a pair of ill-developed ones, on the fifth abdominal segment, at the first moult ; these persist after the second and third moults and disappear with the fourth moult. In larvae of Anisopteryx aescularia, prolegs are developed on the same segment, but these continue through- out the whole larval existence. The larva of an American moth, Layoa crispata, described as being like a hairy Limacodid (Heterogenea) larva, with the head retracted, the body short, and the legs so rudimentary as to impart a gliding motion to the caterpillar when it moves, has seven pairs of short abdominal prolegs, the second and seventh abdominal segments each bearing a pair of rudimentary prolegs, in addition to those which normally carry them. Burmeister found exactly similar prolegs on the second and seventh abdominal segments of Chrytopyga undulata. According to the figures of Kowalewski and Tichomiroff, the embryonic larvae of Sphinx and Boinbyx mori have, at first, a pair of prolegs on each abdo- minal segment, but half of these are absorbed again before the larva hatches. Some very peculiar methods of progression are to be noticed among the larvae of certain species of lepidoptera, none, however, is more peculiar than that of the Cochliopodids, of which our two British species, Hetero- ) genea cruciata (asella) and Apoda avellana (testudo) are very fair representatives. Resting on the upper surface of the leaves of their food-plants, with the body inflated to form a dome-like structure, they look very little like lepidopterous larvae, and bear, in fact, a strong resemblance to the pupaa of ladybirds (Coccinellidae). The almost evanescent character of the prolegs makes progression on the smooth upper surface of a leaf difficult, and Poulton has suggested that the remarkable undulatory movement by which the Cochliopodid larvae now progress was due originally to the larvae first walking " with adhesive claspers," that these gradually became shorter and broader, thus yielding increased support by extending the area by means of which they adhered. Finally the claspers, he considers, would be altogether lost, and the whole of the ventral surface, from which they formerly 38 BRITISH LEPIDOPTERA. projected, would take part in locomotion. The modification of the prolegs and the method of progression, is, without doubt, designed to enable the larva to move freely over the smooth upper surface of leaves, which it could not well do under ordinary conditions. The sticky condition of the abdominal surface supports this view, but there can be no doubt that they spin some small quantity of silk on which they walk, as do so many other lepidopterous larvae. Besides the tubercles, which have fairly fixed positions on the seg- ments, the skin has, scattered more or less regularly over the body, little elevations, resembling, somewhat, a fine pile or covering of minute hairs. This pile is a very common feature in butterfly larvae, is sup- ported by very minute papillae, and is generally distributed with con- siderable regularity, usually in a transverse, though sometimes in a longitudinal, direction. It is, however, occasionally scattered irregularly all over the body. When it is arranged transversely, it is usually some- what closely related to the subsegmental divisions into which the seg- ments are divided. Bacot says that this pile, which appears something like a clothing of short pointed spines, is very common in lepidopterous larvae in their first skin, and, in some, is so fine that a one-fourth lens (or even higher power) is required to detect it. The minute spines or hairs are often only visible at a certain angle, or when the edge of the dorsum is silhouetted against a bright background. In some larvae this coat is lost at the first, or at a subsequent, moult ; in others, it persists throughout the whole larval existence, becoming just a trifle coarser at each moult. The larvae of D icy da oo, Dianthoccia curpo- phat/a, and Taeniocampa pulverulenta fcruda), among many others, illustrate this phase of its development. Bacot is of opinion that primitive and secondary hairs are of different origin, the former arising from the primitive setae or tubercles, the latter from the minute hairs forming the pile just described. He is also of opinion that the bifid shagreen hairs of Smerinthus, the dense clothing of short secondary hairs in some Lasiocampids, the short pyramidal granulations of cer- tain Liparids, and the highly specialised secondary hairs of some butterfly larvae, are evolved from the minute hairs, which in their simplest condition, form the pile above described. That this pile is found rather generally among larvae is proved by the following, very incomplete, list furnished by Bacot. ZYG^SNIDES : Adscita statices and Anthrocera trifolii (both in first skin). LASIOCAM- PIDES : Trichiura crataegi. BOMBYCIDES : Boinbyx mori (very fine). GEOMETKIDES: Phorodesma smaragdaria (first stage, skin granular later). PLATYPTERYGIDES : Drepana cultraria. NOTODONTIDES : Leiocanipa (Pheosia) tremula(dictaea), black in first skin, no trace in second, except on horn, Diloba caeruleocephala (in first stage), Odontosia caniiclita (faint traces in third skin), Phalera bucephala (in first and second skins, (?) developed into secondary hairs later on). LIPARIDES : Dasy- cltira fascelina(in first skin), Demas coryli (strong in first, small in second to fourth skins), Orgyia antiqua (distinct but fine), Leucoma salicis&Q(iPsiliiramonacha(in first and second skins), Portliesia similis. ARCTIIDES: Spilosoma lubricipeda (first to third skins, small), S.fuligi- nosa (first and (?) third skins), Arctia villica (first to fourth skins), Callimorpha dominula (strongly developed), Euthemonia nissula (first to fifth skins), Kuclielia jacobaeae (first skin). N GLIDES : Nola cuculla- tella (in later stages rather granules than prickles). NOCTUIDES ; THE EXTERNAL STRUCTURE of THE LEPIDOPTEROUS LARVA. 59 Acronicta leporina (slightly in first skin), Cuspidia megacephala (in fourth skin, very noticeable and long, almost secondary hairs), 1'haretra euphorbiae var. myricae (first to third skins), Pachnobia leitco- ijrapha (weak in first skin, no trace after), Triphaena pronuba, T. comes, T. Jimbria, T. iantkina (just traceable in first skin, then absent), reridroma saticia and Ayrotis puta (first skin, very small), Dianthoecia carpophaga (to full-grown, very long), Taeniocampa miniosa (large and distinct in first skin, only traces after), T. yracilis (very fine, black, in first skin, no trace after), T . puherulenta (strongly marked throughout) , Calocampa exoleta (in first, no trace in fourth, skin), Aporophyla aiifitralis (absent, or exceedingly fine in first skin), Calymnia affinis (strong in first, small in third, skin, no trace later), Polia chi (slight traces in first skin), Dicycla oo (strongly marked throughout), Plusia feMucae (present in third skin). PAPILIONIDES : Zepliyrusquerciis (strong when, and not until, full-fed), Aylais urticae (strong, in early stages). Since the observations, on which this list is compiled, were made off- hand, and when studying other characters presented by the larvae, it can be readily understood how common an occurrence is the presence of this pile in lepidopterous larvae. Bacot says : Most of the Noctuids lose the character very early, yet in some it persists strongly throughout the larval life. Dianthoecia carpopliaija exhibits it from the youngest to adult stage, yet adult D. cucubali shows no trace of it. T. pulvendenta retains, but T. miniosa soon loses, it. Scudder believes that " the use of this clothing is tolerably clear, since this pile must prevent the too rapid evaporation of the heat from the surface of the body, for, although caterpillars are classed among the cold-blooded animals, they, nevertheless, have an internal heat above that of the surrounding atmosphere, which originates from the activities of the organs and the respiratory functions, and which they would lose more rapidly but for this investing pile." On the dorsum of the thoracic (and more rarely the abdominal) segments of the larva, a hard chitinous shield is found. This is par- ticularly noticeable in all wood-boring larvae, such as those of the Cossids, Hepialids and Sesiids, as well as in Crambids, Tortricids, and many Noctuids and Tineids. It is, however, more general and most marked on the pro-thorax, and hence it is often spoken of as the pro- thoracic shield. Since this structure is equally well-developed in the larvae of the Cerambycidae and other Coleopterous larvae which also bore into hard substances, it appears probable that this hard chitinous plate serves to protect the head, and parts of the body underlying the> shield, from injury. Its appearance, too, in larvae belonging not only to different families of the Lepidoptera, but also, to different orders, suggests that it has been developed in response to the external stimulus supplied by continual friction, an excess of chitin having been deposited (or developed) by the hypodermal cells of the tergal arch of the pro- thoracic segment. It is not unusual to find the shield, in some form of decadence, in larvae which now feed fully exposed, especially in certain Noctuids, and occasionally the shield is present in the first larval stage, but lost in the later ones. These occurrences generally take place in larvae some of whose allies have, or had, boring habits. The value of this shield to boring larvae for leverage purposes must also be very great, since it gives a solid fulcrum for the head. The excessive 40 BRITISH LEPIDOPTEBA. development of the dorsum of the pro-thorax in the larva of Cemra ap* pears to have no phylogenetic significance, nor any close connection with the chitinous pro-thoracic shields of boring larvae. It is certainly smooth and shining, but appears to have been modified independently, for protective purposes, in this particular genus. Still, its probable use for the moulding of its hard cocoon must not be altogether overlooked. The anal segment has caused much discussion as to its structure and homologies, especially with regard to the suranal plate, the infra-anal lobe, the paranal lobes and the paranal tubercles. The supra-anal, or "suranal," plate of Packard, is the "podex" of Kirby and Spence, and both in its shape and ornamentation would appear, especially in Bombycid and Geometrid larvre, to afford specific characters. It varies much, also, among the Notodonts and Satur- niids, and is especially well-developed in those larvae which constantly use the anal legs for grasping, while the front part of the body is more or less raised. It appears to be correlated with enlarged anal pro- legs. According to Packard, this plate, morphologically, appears to " represent the dorsal arch of the tenth or last abdominal segment of the body, and is the ' anal operculum ' or ' lamina supra-analis ' of different authors. This suranal plate is, in the Platyptericidae re- markably elongated, forming an approach to a flagellum-like terrify- ing appendage, and, in the larva of Aylia tau, forms a long, promi- nent, sharp spine. Its shape, also, in Centra caterpillars, is rather unusual, being long and narrow. In the Ceratocampidae, especially in Anisota, Dryocampa, Fades and Citherunia, this plate is very large, the surface and edges being rough and tuberculated, while it seems to attain its maximum in Sphingicampa, being triangular, and ending in a bifid point " (Bombycine Moths, p. 25). The " paranal lobes " are the " homologues of the two anal valves observed in the cockroach, and occur in all, or nearly all, insects," according to the same author. They are the " valvulfe " of Bur- meister, and the " podical plates " of Huxley. They are fleshy and papilliform in Geometrid larvae, and appear as if projecting backward from the base of the anal legs. In the larvae of the Dicranurids they are similar, and each ends in a seta. The "paranal forks" or "paranal tubercles" are two bristles arising from the end of .a papilla, directed backward. They are found in the larvae of most arboreal caterpillars, being especially well- developed in those of Notodonts and many Geometrids, whilst they are wanting in the larvae of Noctuids, Sphingids, Bhopalocera (?), and some Geometrids- and INCOMPLETE (Micro-lepidoptera). In the American Choerodes, they are very large ; so also are they in the larva of our common Uropteryx sambucaria, where they become papilliform and setiferous. Their use was discovered by Hellins. In his description of the larva of Cerura bifida, he writes of them : " At the tip of the anal flap are two sharp points, and another pair underneath, which are used to throw the pellets of frass to a distance." Packard has seen the frass pellets held by the two spines of the paranal tubercles in Centra borealis, whilst Dyar says that he has seen the caterpillars throw their pellets, with the aid of these spines, away from them, so as to strike against the side of a tumbler in which they were confined. The "infra-anal lobe" is described by Packard as a "thick conical fleshy lobe or flap, ending often in a hard chitinous point, and situated THE EXTERNAL STRUCTURE OF THE LEPlDOPTEROUS LARVA. 41 directly below the vent. In appearance, it is somewhat like the egg- guide of the Acrydii, though the latter is thin and flat." Its use is, evidently, to aid in tossing the pellets of excrement away, so that they may not come in contact with the body. Packard, in an article describing the larvae of certain species of Cerura, gave it as his opinion, that the "stemapoda" or filamentous anal pro- cesses of these caterpillars were homologous with the anal prolegs of other Notodonts, and, to show this, figures the anal prolegs of Dasy- lophia angirina in its first larval stage. He points out, in his comparison, that it is intermediate in form between the normal anal proleg and the stemapod, and remarks that it "has no crochets, but the planta, of which the flagellum of Cerura seems to be the homologue, is re- tracted, and the retractor muscles, one of which is divided, are much as in the filamental legs of Cerura. It, however, is not the general opinion of British entomologists that the stemapoda are modified anal prolegs. Hellins regarded them as " dorsal appendages, somewhat after the fashion of the anal spines of the larvae of the Satyridae" Packard discusses this view, and concludes : " After repeated com- parisons of the filamental anal legs of Cerura with those of Macruro- campa marthesia, and comparing these with the greatly elongated anal legs of young Heterocampa unicolor, as figured by Popenoe, and taking into account the structures and homologies of the supra-anal and paranal flaps, one can scarcely doubt that those of Cerura are modified anal legs." There appears to be no doubt whatever that Packard is quite right, and that the view hitherto held by British entomologists, is a wrong one. The ancestral lepidopterous larvae probably lived, at first, on grasses and low growing plants, and the arboreal habit was possibly assumed at a comparatively late period of larval evolution. This view is fully borne out by the geological evidence, for it is generally considered that flowering plants and trees were probably developed in the Cretaceous or Tertiary periods, and that our present race of lepi- dopterous insects became evolved side by side with the great changes that then took place in the flora of the world. Many of the most highly developed groups of Lepidoptera most of the Noctuids, Arctiids, Pierids, Satyrids, etc. feed, even now, almost exclusively upon low plants, and we find that, amongst larvae with this particular habit, the caterpillar is usually devoid of spines, and smooth or covered with a short, dense, velvety pile, whilst the markings consist chiefly of longitudinal lines of various shades of green, grey, etc., running from the head to the anus, dorsally, ' laterally, and ventrally. There are, of course, many very hairy and spiny larvae that feed on low plants, but these live usually a more or less exposed life neither hiding under leaves (like the Satyrids) nor stones (Noctuids and Crambids) by day and the great development of hairs, pencils, spines and bristles, appears to be due often to the cater- pillars having changed their mode of life from a concealed to an exposed Condition, the change having frequently been accompanied by a move from a herbaceous to an arboreal feeding ground. Just as the caterpillars of grass-feeding larvae are green or grey in colour, and are chiefly ornamented with longitudinal lines of various shades, so the larvae of arboreal caterpillars Catocalids, Geometrids, etc. have their bodies usually of a grey or ash colour, ornamented 42 BRITISH LEPIDOPTERA. with dorsal and lateral humps, so that they may assimilate more readily with the colour of the bark of the tree upon which they rest, and to small twigs bearing leaf-buds, etc. But such larva as are particularly protected in this manner do not lead such exposed lives as do those which, by the modification of the tubercles and setae of the more generalised larv, have developed conspicuous spines, pencils of hairs, etc., or those which, by the development of bright warning colours, ocellated spots, etc., present an inedible, or even dangerous appearance to the avian, and numberless other, enemies which surround them on every side. Those larvae which live upon trees, and trust for their escape to their resemblance to pieces of stick, etc., are sometimes remarkably tuberculated. This is particularly noticeable in the Geometrids and Notodonts. On the other hand, those larvae which are arboreal, but which trust for their concealment to leafy abodes which they make and in which they dwell such as the Tortricids, Pyralids, etc. have retained, in many ways, much more generalised forms of larvae, both as regards colour, markings and tubercles. The adaptation of exposed larvae to their surroundings is also very remarkably illustrated in the case of many "plume " larvae. No better illustration is needed than the similarity of the dermal clothing of the larva of Aciptilia galacto- dactyla to the woolly covering of the underside of the leaves of burdock (Arctium lappa), whilst Miss Murtfeldt quotes a parallel case among the American " plumes," stating (Psyclw, iii., p. 390) that " there is a very close imitation in the dermal clothing of the larvae of Leioptilus sericidactylus to that of the young leaves of Vernonia, on which the spring and early summer broods feed." The inedible nature of hairs needs no demonstration. That many birds are able to eat hairy larvae is no detraction from the general principle. The fact that some birds do eat hairy larvae leaves un- answered the fact that there are numbers of birds that cannot ; and, undoubtedly, many small insectivorous birds that would eat a Tortricid larva with gusto, and make no objection to its simple setiferous hairs, would object to a larva of Arctia caia, or that of Acronicta leponna. We may take it for granted that the ultimate use of spines and hairs is for protection, and further, that they have been stimulated in their development by natural selection, indicating to insectivorous birds that the bristly armature is inedible ; yet it seems that we have hardly reached the bottom of the question, if we look upon the special develop- ment of the setae and spines as due to protective needs, arising either from the attacks of birds or parasitic insects, but that we yet require some explanation of the initial cause of the development of such spines and specially developed hair structures. Fritz Miiller, in 1864, maintained that the so called metamorphoses of insects, in which these animals quit the eggs as grubs or cater- pillars, and afterwards become quiescent pupae, incapable of feeding, was not inherited from the primitive ancestor of all insects, but was acquired at a later period. Brauer, in 1869, divided the larvae of insects into two groups, the " campodea " form and " raupen " form. In 1871, Packardf adopted these views, and gave the name of " eruci- * " Betrachtungen iiber die Verwandlung der Inseckten, etc.," Verh. K. K. Zool. bot. Get. Wien, 1869. f American Naturalist, September, 1871. THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 43 form larvae" to the cylindrical larvae of certain Coleoptera (weevils, etc.), as well as to those of Diptera, Lepidoptera, and Hymenoptera, con- sidering that the larvae of all these were the result of adaptation, and were " derivatives of the primary ' campodea ' type of larva." Lubbock practically adopted Brauer's views in 1873. In 1895, Packard consideredf that, " while the origin of the eruciform larvae of the Cerambycidae. Curctdionidae, Scolytidae, and other wood-boring and seed-inhabiting and burrowing coleopterous larvae in general, is plainly attributable to adaptation to changed modes of life, as contrasted with the habits of roving, carnivorous campodeiform larvae, it is not so easy to account for the origin of the higher rneta- bolous orders of Diptera, Lepidoptera, and Hymenoptera, whose larvae are all more or less eruciform." He supposes them all to have arisen independently from groups belonging to the Neuroptera (in the modern sense), or to some allied but extinct group. In 1895, we suggested! that the earliest forms of lepidopterous larvae were hidden, and probably internal feeders. This view is not shared by Packard, who suggests that the earliest type was " allied to some Tineoid which lived, not only on land, but on low herbage, not being a miner or sack-bearer." This conclusion is arrived at by his consideration of the remarkable changes in form of certain Tineoid mining larvae, described and figured by Chambers|j and Dimmock. These larvae were those of the Lithocolletids, Gracilariids, etc., and we quite agree that these apodous forms of mining larvae are the result of adaptation to their habits. Our own idea of the ancestral form was, and is, one more closely resembling those of Hepialus, Cossus or Zetizera, but the point matters little. What most authorities are agreed upon is that by the time the ancestral larva was essen- tially lepidopterous, it was provided with prolegs that bore terminal crochets or hooks, and with simple fleshy warts or tubercles bearing simple hairs. The various forms in which the crochets are now arranged on the prolegs, and the many modifications which one finds in the arrangement and character of the piliferous tubercles, must be looked upon as more recent developments. Meldola first suggested^]" that the green colour of many cater- pillars was due to the presence of chlorophyll in their tissues, and the matter was carried much further by Poulton in his experiments on the larvae of certain species of the genus Smerinthus. Packard thinks that the cuticle was at first colourless or horn-coloured, and suggests that " after habitually feeding in the direct sunlight on green leaves, the chlorophyll thus introduced into the digestive system, and into the blood and the hypodermal tissues, would cause the cuticle to become green," whilst, afterwards, " by further adaptation and by heredity, this colour would become the hue common to caterpillars." In view of Poulton's more recent experimentsf it would not do to labour this point too much, and we are inclined to agree with him, that the effect is rather " phytoscopic " than "phytophagic," inas- much as the colour of the surface of the leaf, rather than its substance, * Origin and Metamorphosis of Insects, 1873. f Bombycine Moths of America, 1895. J Entom. Record, etc., vii., p. 6. || American Naturalist, iii., 255-262 ; Psyche, ii., 81, 137, 227, etc. Psyche, iii., pp. 99-103. 1!" Proc. Zool. Soc. of London, 1873, p. 159. * Proc. Eoy. Soc. Lond., 1885, p. 269. f Trans. Ent. Soc. Lond., 1892, pp. 294 et seq. 44 BRITISH LEPIDOtTERA. acts as the stimulus, and this view has been materially strengthened by his experiments on larvae, such as Rumia luteolata, etc., which show so much initial variation in nature, that some are green and some brown. His observations on larvae of this species, as well as on those of Ennouws quercinaria (anyularia), Selenia lunaria, Crocallis elinyuaria, Pkiyalia pedaria, and, above all, Amphidasys betularia, show conclu- sively that the colour of some larvae is much affected by the surround- ing environment, and hence, as a general conclusion, we must assume, as far as our knowledge at present goes, that the general green colour of those larvae which essentially live among green leaves, is due rather to the influence of the particular environment surrounding them than to any direct action of the chlorophyll, which is consumed with their food. Commenting on these experiments, Poulton says : " Of the colour changes we must distinguish two main kinds : (a) Changes in the colour of the true animal pigments, leading to various shades of brown, grey, etc. (6) The change to a green colour modified from plant pigment, in the food. When such a change of colour is possible, the true pigments are always superficial to the green, and cannot be retained without concealing the latter, the degree of concealment depending on the amount and distribution of pigment. Thus, in Amphidasys betularia, the true pigments are chiefly placed in the epidermic cells, the green in the subjacent fat, whilst in many others, the former are in the superficial layer of the cuticle, the latter in the blood, or sometimes in the lower layers of the cuticle. But the appear- ance of the green is not merely the removal of a screen, although this must occur ; in some cases, at any rate, it also means the formation of the green colouring matter itself." Probably the first attempt at ornamentation in the lepidopterous larva consisted of longitudinal lines. These usually consist of (1) The dorsal or medio-dorsal line (a line running down the centre of the dorsurn, throughout its whole length). (2) Sub-dorsal lines (one on either side of the medio-dorsal line). (8) Supra-spiracular lines (one on either side: above the spiracles). (4) Sub-spiracular lines (one on either side below the spiracles). Sometimes there is a spiracular line running along and including the spiracles. The medio-dorsal line (as such) is probably, occasionally, due to the alimentary canal showing through the skin. It is certainly so in many transparent-skinned larvae (Ephestia kukniella, etc.), and it is just possible that, whatever form its modifications may now take, it originated in this manner. Weismann has concluded, from his studies of the Sphingids, that the sub-dorsal line arose before the spiracular, and Packard f shows how, after the sub-dorsal and spiracular lines are formed, others are rapidly introduced and some may as rapidly vanish, as necessary features of certain stages which, when they become useless, are discarded. Weismann, in his Studies in the Tlwory of Descent, has shown that the primitive markings of caterpillars were lines and longitudinal bands. He further shows that larval spots are formed by interruptions, " the serial atrophy," of the lines or bands. Packard says : The lines, bars, stripes, spots, and other colorational markings of caterpillars, by which they mimic the colours and shadows of leaves, stems, etc., have evidently been, in the first place, induced by the nature of the food * Tram. Ent. Hoc. Loud., 1892, pp. 458*459. Boinbycine Moths of America, p. 15. THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 45 (chlorophyll), by the effects produced by light and shade, by adaptation to the form of the edge of the leaf (as in the serrated back of certain Notodonts), by adaptation to the colours of different leaves and to the stems, since shades of greens, yellows, reds, and browns, are almost as common in the cuticle of caterpillars, as on the surface or cuticle of the leaves and their stems, or in the bark of the twigs and branches. He also adds that probably many have observed that the peculiar brown spots and patches of certain Notodonts do not appear until late in larval life, and also late in the summer, or early in the autumn, contem- poraneously with the appearance of dead and sere blotches in the leaves themselves. This phase of the subject will be dealt with at length in a later chapter. Tactile hairs, defensive setfe, locomotive setae, and spines of various kinds, occur in worms ; these, too, often arise from fleshy warts or tubercles. It is, therefore, not at all unlikely that the ancestral lepi- dopterous larva was provided with piliferous warts, and that many of the specialised spines, etc., now found in lepidopterous larvae, are modifications of these ancestral simple structures. It may be safely assumed that spines, hair-tufts, etc., serve to pro- tect the organism fi-om external attack, probably also to strengthen the shell or skin. That even the most complex spines are modifica- tions of the tubercular structure is evident if one examines the cast skin of a Vanessid larva when it has just been thrown off, and the pupal state assumed. Packard, in a long argument.* 3 suggests that " it is not improbable that tubercles, humps, or spines, may have in the first place been developed in a few generations, as the result of some change in the environment during the critical time attending or following the close of the Palaeozoic, or the early part of the Mesozoic age, the time when deciduous trees and flowers probably began to appear." The same author refers to Darwin's significant remarkf that " organic beings, when subjected during several generations to any change whatever in their conditions, tend to vary," further, that " variations of all kinds and degrees are directly or indirectly caused by the conditions of life to which each being and, more especially, its ancestors have been exposed" (p. 241) and again, that " changes of any kind in the conditions of life, even extremely slight changes, often suffice to cause variability. Excess of nutriment is, perhaps, the most efficient single exciting cause." Referring to the geological fact, that in the Cretaceous period, the forests consisted of oaks, maples, willows, beech, poplar, etc., Packard assumes that, in all probability, the low-feeding caterpillars of that time began to desert the herbaceous plants to feed on trees, and that they then experienced sufficient change to induce considerable variation, and that, to a great extent, tree-feeding necessitated isolation. He thinks, moreover, that the change from herbaceous to arboreal feeding, not only affected the shape of the body, causing it to become thick and fleshy, but also led to a hypertrophy of the piliferous warts, common to all lepidopterous larvae. We deal with this at length, not because we are inclined to agree with its assumptions, but because no other explanation of the actual origin of the cause of the modification has been offered. * Bombycine Moths of America, pp. 16 etseq. t Variation of Animals and Plants under Domestication, 2nd Edition, 1888, 46 BRITISH LEPIDOPTEKA. We find, in definite positions on the larval cuticle, small buttons of chitinous material called tubercles. These usually bear a structure, formerly termed a " hair," but to which the term " seta " is now usually applied, since the seta is not morphologically equivalent or homologous with the hairs of mammals. These setae arise through a modification and hypertrophy of the nuclei of certain cells of the cuticle. According to Dyar, the " primitive form of tubercle consists of a little chitinous button on the skin, bearing a single long hair. It is found in the less specialised groups of Lepidoptera, and exclusively in the JUGATE and the Psychids. When this form is present, there are, in general, no other hairs on the body." It would appear that in the phytophagous Hymenoptera (Tenthred- inidae), there are well-developed setiferous tubercles, apparently more generalised than those found in any Lepidoptera, but in the Lepidop- tera there appear to be, according to Dyar, two types of arrangement. (1) By far the more generalised, consists, on the abdominal segments, of five tubercles above the spiracle on each side, three in a transverse row about the middle of the segment and two behind, whilst below the spiracle are two oblique rows, containing respectively two and four tubercles. This type is found in Hepialus. (2) The second type con- sists of two dissimilar lines of modification of the first type, of which the fundamental arrangement consists of three tubercles on each side above the spiracle ; three more on each side, below or behind the spiracle and above the base of the leg ; and three (or four) on the base of the leg on the outside, and one on the inside near the mid- ventral line. As Dyar has made himself quite an authority on these setiferous tubercles, it may be well to glance at his nomenclature. Commencing from the dorsum, he calls the tubercles above the spiracles i, ii, iii, the three below, iv, v, and vif ; the group on the outside of the leg is known as vii, and the single one on the inside of the leg as viii. Tubercles vii and viii, Dyar says, are present also on the legless abdominal segments (1, 2, 7, 8 and 9), in a position corresponding to those on the segments bearing prolegs. On the last two abdominal segments (9 and 10) the number of tubercles is always less than the fundamental number, even in generalised larvfe. This is evidently due to the fact that these segments have been partly aborted, being without spiracles. The reduction of the ninth abdominal segment has taken place on the anterior portion, whilst the tenth abdominal has lost the lateral part (Classification of Lep. Larvae, pp. 196-7). Dyar's conclusions as to the relationship which the lepidopterous super- families bear to each other are based on (1) The position of the tubercles with regard to the sub-segments into which the abdominal segments are divided. (2) The tendency for tubercles iv and v (the post-spiracular and sub-spiracular tubercles) to coalesce or separate. As to their position, Dyar says that in the JUGATE (Hepialids) the three tubercles of the middle sub-segment are all present, and the upper and lower of the posterior sub-segment. In the Psychids, the three tubercles are retained on the middle sub-segment, but both are * i i= anterior trapezoidal, ii := posterior trapezoidal, iii = supra-spiracular. f This is a secondary tubercle, absent usually in the newly hatched (gene- ralised) larva of the higher families. Hence its importance is less valuable than Dyar afterwards insists, when discussing the Psychids and MICBO-FBENAT^E. THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 47 lost on the posterior one ; the sub-stigmatal tubercles are retained and approximated, the anterior one of the four on the base of the leg seems to have been moved up, forming tubercle vi, which is thus anterior ( = pre-spiracular). This explanation accounts for the possible formation of the pre-spiracular tubercle as such, for it will be observed that, whereas tubercle v of Dyar is the typical sub-spiracular tubercle of the more specialised families, tubercles iv and vi, typically originating below the spiracle, according to Dyar, become respectively the post-spiracular and pre-spiracular in special instances. In all the other families of the Lepidoptera, Dyar states that the middle tubercle of the three on the middle sub-segment is lost, but the upper on the posterior sub -segment is retained ; the two (iv and v) below the spiracle are also retained, as in the Psychids, but they are either approximated (sometimes even united to form a compound sub- spiracular tubercle, as is Margarodia), or separated so as to form two distinct tubercles, viz., the sub-spiracular and post-spiracular, whilst of the four tubercles at the base of the leg, the posterior one (not the anterior one, as is the case in the Psychids) is moved up to form tubercle vi. The tendency for tubercles iv and v to coalesce so as to form a compound sub-spiracular tubercle, appears to be characteristic of the larvae which comprise, in its broad lines, Comstock's MICROFRENAT.E or GENERALISED FRENAT^E, whilst the tendency for tubercles iv and v to separate and form post-spiracular and sub-spiracular tubercles, re- spectively, appears to be characteristic of his SPECIALISED FRF.NAT^:. Dyar notes, and if it held good it would be very curious, that " it is a striking fact that we do not find a series of intergrading forms between the single-haired tubercle and the many-haired wart, though both may occur in different genera of the same family," and he considers that this is explicable on the principle of discontinuous variation, which is insisted upon by Bateson. He says that in the lower (more generalised) families we have the simple and primitive form of tubercle ; in the more specialised families we find a modification, which consists in the tubercles becoming enlarged and many-haired. In these compound tubercles each hair arises from its own minute tubercle, and the whole are borne upon an enlarged base or wart. Modification then takes place in the higher groups, by a reduction in the number of tubercles, the reduction taking place : (a) By coalescence, (b) By unequal development and final obliteration of particular ones. (This is discussed later in chapter.) We have seen that in some of the more specialised larvae there is a general tendency to the reduction of tubercles, so that some may entirely disappear. In some cases, however, the bases of the tubercles are developed into long fleshy processes, carrying aborted setae, as in the case of certain larvae of the Nymphalids, Papilionids, etc. In other cases, the setae remain as glandular hairs, in some instances secreting an urticating (? odorous) fluid, or the hairs themselves become highly specialised, and greatly increased in number, forming brushes, tufts, plumes, etc., as in the larvae of Acronyctids, Liparids, Arctiids, etc. One of the most striking modifications of the tubercles is seen in the caudal horn of the SPHINGIDES. This is an unpaired dorsal process on the 8th abdominal segment. A figure of the larva of Deilephila eiifihorbiae (Weismann, Studies in tJie TJieory of Descent, PI. v., fig. 38) in its first skin, shows that the two setae of tubercle i are borne on the 48 BRITISH LEPIDOPTERA. apex of the caudal horn. This would point strongly to the conclusion that the horn represents the base of the unconsolidated pair of tubercles i, the tubercles themselves having disappeared. This disagrees with Poulton's view, for he looks upon the caudal horn as representing the consolidated pair of tubercles i of the Saturniids. The caudal horn of the remarkable genus of Plume moths, does not, according to Bacot, rise from the 8th abdominal segment, and bear the anterior trapezoidals of that segment, as in the Sphingids, but is situated on what is either a small 9th abdominal segment, or a large and distinct subsegment of the 8th abdominal, both the anterior and posterior trapezoidals of the 8th segment being in front of the horn, and in their correct position relative to the spiracle. The production of a central row of dorsal tubercles apparently un- paired, in certain families, is very remarkable. This is well seen in the medio-dorsal row of spines in the adult larvae of certain Vanessids, where, too, the real nature of the spines forming this row may be readily learned, by comparing the adult larvae with those in their earlier stages. They are formed by the union of tubercle i on each side, consolidating on the central line of the dorsum. A similar arrangement also occurs in the Saturniids. The modifications which tubercles and setas undergo have been tabulated by Packard. f His table reads as follows : A. TUBERCLES. a. Simple and minute, due to a slight thickening of the hypodermis, and a decided thickening of the overlying cuticle ; the hypodermis contains a large unicellular gland, either for the secretion of the seta or for the production of poison. 1. Minute piliferous warts (most Tineid, Tortricid and Noctuid larvce). 2. Enlarged smooth tubercles, bearing a single seta (many Geometrid and Bombycine larvae). 3. Enlarged spherical tubercles, bearing a number of setae, either radiated or subverticillate (Arctians, Lithosians). 4. High, movable, smooth tubercles, having a terrifying function (Schizura, Xylinodes, Notodonta, Nerice). 5. Low and broad, rudimentary, replacing the " caudal horn " (Choero- campa, Leiocampa (Pheosia) dictaea, and L. dictaeoides). b. More or less spinulose or spiny (disappearing in some Sphinges after Stage 1). 1. Long and slender, usually situated on the top of the eighth abdominal segment, with microscopic spinules in Stage 1. (Most Sphingidac and SmaJ). 2. Smooth subspherical warts (Chalcosia, East Indies) ; or elongated, but still smooth (Attacus atlas). 3. Subspherical or clavate spiny tubercles of many Attaci ; the spinules usually short. 4. - Spinulated spines or elongated tubercles of Ceratocampidae and Hcmi- liicidae (Automeris io and Hemileuca maia, etc.). 5. Spike-like hairs or spines (Samia cynthia, Anisota, Hypsa (E. Indies), Anagnia). G. Antler-like spines. Early stages of Heterocampa biundata, H. guttivitta and H. obliqua). * Trans. Ent. Soc. Loud., 1888, pp. 5G8-574. f Bombycine Moths of America, p. 21. } Packard does not use Sesia in the sense usually understood in Britain, i.e., for the true Clearwing moths, but as a synonym of Macroglotsa, THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 49 B. SET*: (HAIRS, BRISTLES, ETC.). 1. Simple, fine, short or long, macroscopic or microscopic setae, tapering hairs, scattered or dense, often forming pencils (many Bombyces, Zygaenidae,* Noctuo-Bombyees, Apatelae). 2. Glandular hairs, truncate, spindle-shaped or forked at the end, and secreting a more or less viscid fluid [many Notodonts in Stages 1 and 2 ; many butterfly larvae ; Pterophoridae (in last stages)]. 3. Long spindle-shaped hairs of Apatelodes (Apatela americana), and Tinolius ebarneigiitta. 4. Flattened, triangular hairs in the tufts, or on the sides of the body of Gastropacha americana , or flattened, spindle-shaped scales in the European G. qiiercifolia. 5. Spinulated or barbed hairs (most Glaucopides, Arctians, Lithosians, Liparids and many Bombycids). C. PSEUDO-TUBERCLES. 1. Filamental anal legs (stemapoda) of Centra and Heterocampa marthesia. 2. The long suranal spine of Platyptericidae. Before leaving our consideration of the hairs of larvas, it may be well to mention the spathulate hairs of Jocheaera alni. These are usually erect and conspicuous, but in the adult stage are spread some- what laterally. Chapman gives them as measuring, in the 4th larval skin : on pro-thorax, 8 mm., on 5th abdominal, 1 mm., on 9th abdominal, 2| mm. ; in the 5th larval skin, on the same segments 6, 3| and 4 mm. respectively, and in the 6th larval skin (extra moulter), 7," 4, and 4^ mm. respectively. The larva of Eutricha guercifolia and those of other species possess remarkable scale-like hairs, as mentioned above by Packard. The study of the newly-hatched larva is one of the most important factors in considering the phylogeny of the lepidoptera, for it happens that many species which have the most specialised adult larvae hatch in a very generalised condition, and hence, comparison of the tubercles in the newly-hatched larvae, with the more specialised structures that replace them afterwards, gives many valuable clues to the origin of the complicated structures of the adult. From this, it would appear, that the more primitive arrangement of the five chief tubercles and setae occurring on the abdominal segments, is such that the three tubercles above the spiracle exist as the anterior trapezoidal, posterior trapezoidal, and supra-spiracular tubercle, respectively, whilst the sub- and post- spiracular tubercles are both placed beneath the spiracle. Dyar remarksf : " Curiously enough, the most generalised condition is ex- hibited in the first stage of the butterflies (Rhopalocera). This is to be accounted for by the fact, which was brought out by a comparison of the first stage of such genera as Danais and Grapta, with their later stages, rfc., that the armature of the butterfly larva is not developed mainly from the primary tubercles, but almost entirely independent of them." This is certainly too sweeping an assertion to comprise the facts re- lating to the armature of the Vanessid and Argynnid larvae, and pro- bably some others. In many cases there can be no doubt that the armature is frequently developed from the primary tubercles, often, of course, with certain stages of the evolution left out. In some the process of development is comparatively simple, as may be seen, if the larva be * As used in America, this = our Euchromiidae, which are Arctiids, not the family British lepidopterists call Zygaenidae. \ " Additional notes on the classification of Lepidopterous larvse," Trans. New York Acad. Sci., xxv., p. 52. D 50 BRITISH LEPIDOPTERA. examined carefully at each ecdysis. The case of Aylais urticae and others occur to me. The horn which characterises the Sphingid caterpillars is, as we have seen, placed on the dorsum of the eighth abdominal segment, and it is remarkable that when it is absent in allied forms, it is replaced by a small, low and flattened tubercle, the segment itself being somewhat swollen. Many Noctuid larvae Amphipyra, Maniestra pemicariae, etc., have a prominent hump on this segment, so also have the larvae of the Agaristitlae, and others. In many Notodont larvae the first ab- dominal segment bears a conspicuous hump, sometimes forked, often ending in a seta. It would appear, from Packard's researches, that the three thoracic segments, and the first and eighth abdominal seg- ments, are those most usually characterised by tall fleshy tubercles, horns, etc. The same author shows that the first and eighth ab- dominal segments bear no prolegs, and that, when walking, these apodous segments are more raised than the others, and that, if it be true, as it appears to be, that these humps do frequently rise from the most elevated portions of the larva when crawling, then the move- ment of these conspicuous structures might tend to be of service in frightening away other creatures. He further suggests that the humping or looping of these segments may have had something to do with inducing the hypertrophy of the dermal tissues which enter into the formation of the tubercles or horns, whilst with regard to the mutant or movable tubercles, he suggests that the movement of these appendages would suffice to scare off an approaching ichneumon or Tachina. Lame are, of course, subject to the conditions involved by the struggle for existence, and to modification in relation to environment, and, hence, is due the modification of the setiferous tubercles, by which the larva is made to resemble different objects at different phases of its existence. Everyone knows how different is the larva of Jocheaera alni in its third skin, in what is known as the " birds'-dropping " stage, from the adult larva with its conspicuous bulbous-tipped hairs. This reference to a subject already discussed in a previous part of this chapter (p. 47) gives us a chance of explaining why we have thrown doubt upon Dyar's statement that " we do not find intergrading forms between the single-haired tubercle and the many-haired wart." He probably had in mind some such change as that occurring in the Anthrocerids, in which the simple single-haired tubercle of the first skin becomes a many-haired wart in the second, increasing in size at each subsequent moult. It happens, as a matter of fact, that intergrading forms are exceedingly common in many species of Lepidoptera, a single-haired tubercle in the first skin ac- quiring some hairs at each subsequent moult, until it becomes a wart. In the Acronyctid larva? there are various stages in different species, even in the first skin, the differences extending from a one-haired tubercle, two-haired tubercle, etc., to a many-haired wart, and such cases are not at all uncommon. In the case of Anthrocera, it is pos- sible that some stages in the evolution of the many-haired wart are now missed, but, in others, the intergrading forms are, as we have said, by no means unknown. * Chapman, Entomologist's Record, etc., vol. U-, p. 123. THE EXTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 51 The varied stages of development of the setiferous tubercles, some- times reached in allied genera in the egg, is of the highest significance, as is also their comparative development in the various stages of the larvae of allied genera, as in Ornithoptera and Papilio, in Af/lia and Citheronia; whilst Packard** states that the "tubercles of the adult larvae of Saturnia (pavonia and pyn) are on the same plane with the embryo, just before exclusion, of the more highly specialised forms of the group Attacinae," and, again, " whilst the late embryos of the Attacinae are, perhaps, paralleled by the fully-grown larva of Satumia, the fully-grown larva of the most, or one of the most, generalised of the Attacinae, Platysamia, is on the same plane of specialisation as the larva of Callosamia in its third stage." The larvae of a large number of Lepidoptera are provided with what may be fairly termed glandular setse. They are more especially abundant in young larvae, and occur in butterflies (Pierids and Satyrids), Geometrids (Ortholitha cervinata), Notodonts (Datana, Dasy- lophia), and many others. Packard describes the glandular hairs of newly-hatched larvae of Ceratosia tricolor as " flattened at the tip, which is slightly tridentate, with grooves passing down the shaft from the notches between the teeth." In the Pierids they form an open basin, fringed with cilia, supported on an exceedingly slender, hollow pedicel, the hairs looking as if tipped with dew. In a preceding part of this chapter (p. 40), we query the absence of the paranal forks in the Khopalocera. This is because Chapman has called attention to a well-known structure, called the " anal comb," which is possibly homologous with the paranal forks. It is found just under the anal flap in many Tortricid, Hesperid, and Pierid larvae. Scudder figures the anal comb in Colias (Eurymus) pMlodice, but does not seem to mention it in the text. This should, of course, have been mentioned directly after the paragraph referring to the " paranal forks." It has been repeatedly noticed that certain larvae, when confined, have a tendency to crawl upwards, and this is more particularly the case with some species than others. Larvae of the genus Coleophora, Aylais urticae, Vanessa io and others, might be instanced as always taking possession of the highest possible point of any receptacle in which they may be placed. Poulton suggests that this is due to the fact that the larvae in these movements are guided by an appreciation of the force of gravitation. That it is not always in order to seek food is evident, for the larvae will crawl over the food-plant in order to reach the highest available point. It is very possible that these movements are made in order to seek light, or air. At any rate, it is not yet at all clear how far the latter causes are factors in bring- ing about these movements, and how far the force of gravity has effect. Poulton further considers that the force of gravity has been potent in bringing about the characteristic "Sphinx-like "attitude that characterises the larvae of certain Sphingids, Aylia, etc. This atti- tude, he says, bears a distinct relationship to the position assumed by these larvae. The thoracic legs, in such larvae as adopt this attitude, are not used for the support of the body, and, hence, when * " Studies in the Transformation of Moths of the Family Saturniidee," Proc. Amer. Acad, Arts and Sciences, 1893. 52 BRITISH LEPIDOPTKRA. the larva is clinging as is its wont, the weight of all the parts of the body anterior to the third abdominal segment is only indirectly supported by means of the claspers. He further points out that the young larvae of all species which .exhibit this habit, habitually rest on the underside of leaves, and, therefore, have the dorsal area pointing downwards. Under these circumstances "the organism reacts upon the strain, and the muscular body- walls strongly contract upon their fluid contents in such a .manner as to produce compensating rigidity, and thus give to the body the curve which is characteristic of the attitude. The Sphinx-like attitude is to be explained as the com- bined effect of gravity and of muscular reaction upon the anterior un- supported parts of the body. The muscular arrangements, which are most favourable for counteracting these strains, are also made use of in the older larvae for the maintenance of a feebly marked Sphinx-like attitude, when the larva is seated on the upper side of a horizontal twig. The attitude is most strongly marked when the larva is resting on a vertical twig, because gravity tends to draw the anterior part of the body backwards as well as downwards. These large larvae habitually rest on vertical twigs, with the head uppermost, because the twig itself is approached from its base, and gradually stripped of leaves towards its apex. The essential dependence of the* attitude upon gravity is well seen, when a vertical twig, with a larva upon it, is carefully bent downwards, so that the strain is in the opposite direction, and tends to bend the anterior part forwards instead of backwards. Under these circumstances the larva begins to yield to the strain in a few minutes (Trans. Ent. Soc. Lond., 1888, p. 675). An interesting subject of enquiry is the evolution of the Geometrid form. The fact that this form is found, in a more or less modified condition, in certain Noctuid larvae, has suggested an alliance between the two groups. It seems very probable, however, that this similarity has been brought about by somewhat similar needs, the Geometrid form being, in many respects, a very specialised one. Many Noctuid larvae that have the full number of prolegs when adult, are more or less Geometrid in form when young. It appears probable that this form has been developed in order to give these larvae a greater reach (1) to obtain their food, (2) to travel from one twig to another. The Geometrids are essentially herbaceous and arboreal in their habits, remaining on their food-plants the whole of the day, so also are the Plusias and other Geometriform Noctuids. The Noctuids that have a Geometrid form of progression when young, also, at this period of their lives, remain on their food-plants, but when they gain the hitherto absent prolegs, they climb down the plants and hide at the roots, or under the ground by day, ascending the plant again to feed by night. The comparatively low-feeding Geometrid larvae are, as a rule, small species, and the bushy herbs on which they feed, bear to their power of reach much the same proportion as the larger trees bear to the reaching power of the larger larvae. Another view of the matter suggests itself, viz., the necessity of Geometrid larvae to travel more quickly than other tree-feeding larvae. The Sphingids, Saturniids, Lasiocampids, Dicranurids, Catocalids, etc., are specially protected by spines, hairs, etc. The Geometrid larva is naked, usually only protected by the resemblance of its colour to its environment, and by its power to remain rigid and motionless. When THE EXTERNAL STRUCTURE OF TflE LEPlDOfTEROUS LARVA. 53 moving, therefore, it is helpless, and must travel from place to place with as much speed as may be possible. Every observer knows that the tree-feeding larvae of the other groups mentioned above are extremely slow in their movements. It is essential, above all things, that a tree-feeding larva should hold very firmly, and this it is enabled to do by spinning silken threads and ladders, and by the possession of remarkably strong and well-developed prolegs. The large Saturniids, arboreal Sphingids, Lasiocampids, etc., cling with amazing tenacity, but, at the same time, they walk with extreme slowness. With them, the opening and closing of their prolegs is a remarkably complex operation, in which a whole army of muscles is brought into play. The Geometrid larva has to cling as tightly as these. At the same time it has to move more rapidly, hence it has reduced its prolegs to the smallest possible effective number, and has, especially, anal ones of the very best kind. Thus it is able to obtain a long stretch for each step, and is able to progress with comparative speed. The young Noctuid larva, too, has often a considerable amount of travelling to do in search of food (eggs being often laid away from the food-plant, etc., ante, p. 13), and a certain amount of looping increases its activity by lengthening the step ; and this is, perhaps, much more important in the young state when the larvfe have an arboreal habit. It may be, therefore, that rapidity and facility of progression is a great part of the object in view. An Arctiid larva, when travelling rapidly, hardly uses the prolegs at all, but progresses by a rapid looping movement, the ordinary progression, segment by segment, being altogether too slow for its needs. Every field naturalist has observed how a Geometrid larva will maintain its hold upon a twig and eat a leaf, and, for this, reach is also required. The difference between the way in which a tree- feeding Geometrid larva and a Sphingid larva will attack a leaf is remarkable. The Geometrid stretches itself out to its full length, and eats as much as it can reach without moving, often beginning near the tip and devouring the whole leaf. The powerful Sphingid larva pulls the leaf towards itself, and thus does by greater strength what the Geometrid larva does by greater reach. The Geometrid form, therefore, appears to be correlated with habits of (1) greater reaching or stretching power, (2) greater speed. It does not seem to have any important phylogenetic significance. In a previous part of this chapter, we have referred to the fact that lepidopterous larvas have a certain number of ocelli on each cheek. Landois considers that these do not essentially differ from compound eyes, and states that if many of them were grouped together they could hardly be distinguished from compound eyes. In each ocellus, he says, the cornea is divided into three lenses, each corresponding to three nerves, each with a separate terminal enlargement, forming the so-called crystalline bodies. Each ocellus, therefore, might be re- garded as Joeing, in reality, composed of three. On the other hand, the three arches of the cornea are so closely connected together, that they give the impression of forming a simple cornea. The three lenses are also very closely pressed, and the three nerves unite into one. Under these circumstances, Landois regards the ocelli of caterpillars as a connecting link between simple and compound eyes, and proposes for them the name of " ocelli compositi." Chapman says : That 54 BRITISH LEPIDOPTERA. the larval ocelli are descended from compound eyes, or are per- sistent from the embryonic form of compound eye, is undoubted. They often occupy a definite tract on the head, which probably repre- sents the area of the compound eye, of which some ocelli only are developed (MI litt.). There has not, we believe, as yet, been any attempt to locate an organ of hearing in the larvae of Lepidoptera, although various authors have done so in the imago. Swinton summarises (Ent. Mo. May., xiv., p. 121) the various notes that have appeared on the aural apparatus of Lepidoptera. There is direct evidence that some larvae, at least, show considerable sensitiveness to sound waves. We have noticed that larvae of many species Aglais urticae. Callimorpha dominula, Xenieophila plantaginis, and Lasiocampa querciis, among others throw their bodies violently from side to side, if one speaks in a loud tone, when in their vicinity. CHAPTER VI. THE INTERNAL STRUCTURE OF THE LE PIDOPTE RO U S LARVA. THE external characters of the lepidopterous larva are, owing to the division of the body into segments, each with its own special organs and appendages, easily described, and the position of these structures located. The location of the internal organs is, however, more difficult, for they are not restricted to certain segments, but run longi- tudinally through the body, frequently extending from the thorax forward into the head, or backward into the abdomen. It is necessary, therefore, in dealing with the internal organs, to consider each separately, both as regards its position and function. The movements of the body are of the first importance, and we find that larvae have undergone great modifications in order to enable them to vary their movements according to their needs. Move- ment is dependent upon the muscular system, and by the muscles, then, the changes that take place in the external framework and appendages are brought about. The nutrition of the various parts is carried on by food, and to understand this we must study the digestive system. The absorption of the digested food into the blood and its carriage to all parts of the body necessitate a circulatory system, whilst the oxygenation of the blood introduces us to the respiratory system. This latter is so intimately connected with the excretion of waste, that one is insensibly led to consider the excretory system, whilst the organs, by which the whole of these various systems are governed, comprise what is known as the nervous system, and this has to be considered both in its relation to volition and sensation. These various systems comprise, then, the different organs (and their functions), by means of which the life of an insect is carried on, and their external results, as exemplified by their movements, etc., are the outward sign of their vitality. The reproductive system, which is not, however, matured in the larval stage, must take the highest place in relation to the continued life of the species. Closely related, THE INTERNAL STRUCTURE OF THE LEPIDOPTEROUS LARVA. 55 too, with the digestive, is the cellular, system, by means of which the caterpillar is able to store up large quantities of surplus material for use in the later stages of its metamorphoses. The voluntary muscular system of the caterpillar is that by means of which it is enabled to move about in order to obtain its food. The muscular fibres are usually arranged in the form of flat ribbons, or conical bundles. The latter make up almost the whole structure of the head, are fastened chiefly to the head walls, and end as fine tendinous cords, attached to the various organs which the insect is thus enabled to move. In this way, certain muscles reach down into the mandibles, which they close when they contract ; whilst the mandibles are opened by muscles which are attached to their outer bases and to the head, just below the ocelli. Other fine flat retractor muscles draw the labrum inwards, whilst extensor muscles work in the opposite direction. A series of contiguous muscular cords, often forming a double band of simple, longitudinal muscular fibres, runs from one end of the body to the other, on each side, just under the skin, between the spiracles and the ventral area of the body. Mus- cular bands, too, run transversely and obliquely in the front of each seg- ment, and are attached to the medio-ventral line farther back in the segment. Above the spiracles on each side are other longitudinal bands, made of three layers, whilst between these and the skin, at the front of each segment, a transverse muscular belt encircles the body, passing at the spiracular region over the longitudinal tracheal vessel, which unites the contiguous spiracles, and straps it to the integument. The flexor muscles of the true legs arise just beneath the longitudinal straps, previously described as running between the spiracles and the ventral area, and extend to the opposite wall of the segment in which they take their rise. The muscles of the prolegs are somewhat different, flat bands forming, as it were, a muscular coating to the walls of the legs just beneath the skin. Usually, these pass directly down, narrow- ing as they go ; the muscular fibres, too, appear not to cross to opposite sides of the leg. The involuntary muscular system is principally connected with the digestive and the circulatory organs. The ossophagus is provided with fine longitudinal bands of muscular fibres, and also with less well- developed transverse encircling bands. The inner coating of the stomach is enclosed in delicate strips of muscular fibre, crossing each other diagonally ; besides these, longitudinal muscles run throughout its length, and well-developed transverse muscles encircle the stomach similarly to those found in the ossophagus. The arrangement of the muscular tissue in the intestine, in longitudinal and transverse bands, is very similar to that in the other parts of the alimentary canal, but, in this, the longitudinal bands are often thick, white and glistening, whilst near where the small intestine joins the stomach, the walls are plentifully supplied with short -longitudinal muscles. The diagonal bands found in the stomach have also their representatives here. The ajimentary canal is held in its place by a series of muscular bands attached to the body wall, one set passing round that portion of the intestine where it is connected with the stomach, another set being attached to, and supporting, the posterior end of the small intestine, these muscles stretching horizontally from the middle of one side of the 8th abdominal segment to the opposite side. 6 BRITISH LEPIDOPTERAi, The mouth opens into a long narrow tube (the ossophagus), into which several long tubules pass. These represent the salivary glands of the higher animals, and secrete a fluid, which is discharged into the oesophagus, and which is swallowed with the food. It dissolves the starch and cellulose of the food, and fits it to soak through the walls of the alimentary canal, so that it., can enter the system. The oeso- phagus is composed essentially of muscular tissue, and expands into a crop (or food receptacle), and then into a gizzard. This is provided with hard plates, that help to grind up the food, which, after being so ground up, is passed through another short tubular passage into the stomach. The walls of the stomach secrete a fluid resembling the gastric juice of the higher animals ; this changes the insoluble proteid of the food into a soluble peptone, which is readily absorbed by the walls of the stomach and intestine. The stomach opens into the intestine, the upper end of which is connected with a number of tubular glands. These are supposed to represent the liver of the higher animals. The intestine ends in a chamber called the " cloaca," in which the waste matters are collected, and from which they are expelled through the anus. In vertebrates, the nervous system is placed dorsally, and the circulatory and respiratory systems ventrally, in relation to the ali- mentary canal. These positions are exactly reversed in insects, the nervous system being placed ventrally, the circulatory and respiratory systems dorsally, the alimentary canal being still placed between them. It has, however, been shown that this difference is more apparent than real, the dorsum of the insect being really analogous with the venter of the vertebrate, but the position of the limbs is reversed. In the upper part of the body, and directly under the dorsal integument, is a longitudinal organ, somewhat like a long tube, which is known as the dorsal vessel. This corresponds with the heart of the vertebrates, and it consists essentially of only one chamber, although this is divided into 8 or 9 sacs, the latter, with openings along the sides, called ostia. It is composed chiefly of muscular tissue, and is connected with the roof of the body by short stout muscles, which keep it in position. It opens towards the head into a kind of arterial trunk. As the dorsal vessel contracts from behind forwards, the blood, which consists of plasma, or fluid, and colourless corpuscles, is driven forward into the trunk. The latter subdivides into smaller vessels, which are soon lost, the walls gradually becoming inseparable from those of the ordinary lacunae, or depressions found between the tissues, and which are lined in many places with epithelium. As the blood passes through these lacunae, it is brought into contact with the tracheal branches and aerated. At the same time the nutritious parts of the food, which soak through the walls of the stomach and intestine, enter the blood in the lacunae found near these organs. The great difference that exists between the blood of insects and that of vertebrates, is such that one feels that it is a great mistake to call two so dissimilar fluids, with different functions, by the same name. The blood of insects varies with the species, sometimes even with the various stages of the same insect. Its function is to carry the nutritious matters to the tissues, and to feed, as it were, the tissues it bathes. It is frequently filled with somewhat crude fatty matters, and Graber calls it " a refined or distilled chyle." THE INTERNAL STRUCTURE OP THE LEPIDOPTEROUS LARVA. 57 Beneath the dorsal vessel, a fine membrane is stretched in such a manner as to separate the dorsal vessel from the surrounding organs, and, at the same time, leave a cavity around the dorsal vessel itself. This cavity is called the pericardia! cavity or sinus. The membrane itself is incomplete, and when certain delicate muscles connecting it with the body-wall contract, they pull it down tightly upon the tissues below, and this, of course, at once increases the size of the sinus. The tissues thus pressed upon are full of chyle and blood, and the fluid is squeezed from these structures through the incomplete mem- brane, into the pericardial chamber, and from thence it re-enters the dorsal vessel again. The number of contractions of the dorsal vessel varies remarkably. They may amount to as many as a hundred per minute ; they may cease altogether without death ensuing. It is recorded as pulsating from 48 to 52 times per minute in the larva of Triaena (Acromjcta) psi, and 44 times per minute in the larva of Brotolomia meticulosa. In spite of the fact that Swammerdam, Reaumur, Bonnet, De Geer, and others, all speak of blood-currents, of fluids moving in the body, of pulsations of the heart or dorsal vessel, and of circulation, Kirby and Spence record their emphatic opinion that there is no circu- lation in insects. The idea of circulation taking place in the lacunae of the tissues does not appear to have suggested itself, and the early authors appear to have thought that definite tubes with definable parietes were necessary for circulation. Bowerbank, and others, placed the matter beyond dispute, and it is only necessary to refer to it here, because many entomologists still seem inclined to accept the state- ment of Kirby and Spence. The fat-body is a very prominent part of the structure of the lepi- dopterous larva. It consists of fat masses of various size, loosely connected together, and enveloping most of the organs. It varies in colour and appearance in almost every species of insect, and appears to consist of a reservoir of reserve material, which increases in the larval stage, when the insect is busily engaged in feeding, and upon which the insect can draw in the future, when it is unable for along period to take food, e.c/., such periods as occur at each exuviation of the larval skin, and also at the more exhausting periods of metamorphosis. It must also be looked upon as a storehouse on which the insect can draw when in the more quiescent pupal stage. The respiration of the Lepidoptera has been partly dealt with in the preceding chapter, and we have seen that air is conveyed into all parts of the body by means of the tracheae. The tracheae are elastic tubes, held open by an inner chitinous layer, and they are all intimately connected. Large tubes connect the spiracles longitudinally, others pass from one side of the body to the other, whilst a set of tracheae in the lower part of the body is connected with another set in the upper part by ascending tubes. These main branches give out small branches, which fork in all directions, and hence the body is supplied most plenteously with air. The tubes have a white glistening appearance, and hence can be detected in a freshly killed insect without difficulty. [In insects of strong flight, there are air-sacs connected with the tracheae, and capable of holding sufficient air to decrease, when distended, the specific gravity of the insect.] The finest tracheal tubes are supposed to penetrate cells, but it is not known Whether they terminate with open or closed extremities. 58 BRITISH LEPIDOPTEBA. The activity of the respiratory system of the Lepidoptera may be readily surmised from the rapidity with which they are affected by agents, such as ammonia or chloroform, yet the exact manner in which breathing is carried on is unknown. Rapid movements of contraction and expansion of various parts of the body, accompanied by the opening and shutting of the spiracles, are often observed, and are supposed to be respiratory, but it is generally believed that, al- though the tracheae must supply the tissues with oxygen, they do not carry off the carbonaceous waste from the tissues. Many consider that some of these waste matters are passed from the skin, and this is more probable than any other explanation yet offered. It is well-known that caterpillars, shut up and with insufficient air, throw off waste products most freely from the skin, the process being popularly known as " sweating." Some entomologists consider that the skin is built up from within, and since chitin is composed largely of carbon and nitrogen, it is possible that certain of the waste matters may be used in the formation of chitin, and finally passed off when the larva exuviates or casts its skin. The Malpighian tubes, a number of coiled filaments found in the dorsum of the larva, used to be considered analogous with the liver of vertebrates, and were supposed to secrete a substance somewhat analogous with bile. They are now known to be excretory organs, and to remove various compounds from the system. It is not yet known how the tubes are emptied, but the material contained in those of some of the Lasiocampid and Saturniid moths, is supposed to be mixed with the silk of the cocoon, and to be used for the purpose of hardening the latter. It certainly seems to be so used in Malacosouia (Clisiocawpa), Erioyaster, etc. The substance excreted is generally in the form of oxalate of lime, or some allied compound. Lepidoptera, in common with many other insects, have a very complicated nervous system, which may be conveniently considered as consisting of three divisions : (1) The cephalic system. (2) The ventral or ganglionic chain. (3) The accessory sympathetic system. These divisions are, of course, very intimately connected. The cephalic system consists of two masses. One is large, and placed above the O3sophagus, and, hence, is termed the supra- 03sophageal ganglion ; the other is smaller, and placed below the oesophagus, and, hence, is termed the infra-oasophageal ganglion. These are united with nerve fibres, passing round the oesophagus, and forming what is often termed the oasophageal ring or collar. These cephalic ganglia are often spoken of as the brain, and, in these, the nerves which supply the eyes, antennae and tongue originate. The ventral chain consists of a series of ganglia. These are small masses of nerve substance, placed longitudinally along the ventral side of the insect. They are arranged in pairs (theoretically one pair in each segment, although often various pairs of ganglia are united), and the ganglia are connected with the ganglia preceding and suc- ceeding by longitudinal nerve fibres or commissures. From these ganglia the motor nerves of the body are distributed to the muscles in the various parts of the body. In the larva of Tischeria anyustico- lella, the paired ganglia are very distinct in each of the thoracic seg- ments, and in the abdominal segments 1-6. Scudder says that they are usually found in the lepidopterous larvae as far as the 7th ab- THE INTEENAL STRUCTURE OP THE LEPIDOPTEROUS LARVA. 59 dominal segment, in which there is a pair of ganglia, and here the nervous cord terminates. The nerve ganglia of Tischeria are placed very nearly to the front of each segment. [In the lepidopterous imago the union of the ganglia in adjacent segments is sometimes very com- plete. In different families there appear to be sometimes two, at other times three, thoracic ganglia, but always four abdominal ganglia, with the exception of the Hepialids, which appear only to have three.] The sympathetic system consists of a median nerve cord, dilating at intervals into ganglia, and placed above the ventral system, with the commissures of which it is connected by nerve fibres. The nerves from this system are distributed to the various organs of the body connected with alimentation, circulation and respiration. It must be remembered that, although apparently so different, the development of the nervous system in the embryo is analogous with that in vertebrates, and that, although the nervous system of insects is apparently ventral, whilst that of vertebrates is dorsal, the ventral part of an insect corresponds with the dorsal part of a vertebrate, i.e., in reality, opposite parts of the body are placed ventrally in insects and vertebrates respectively, owing to the limbs being turned in opposite directions in the two cases. It used to be a generally accepted belief that the lepidopterous larva had no sexual organs, and this, in spite of the fact that Reaumur, a century and a half ago, stated that he had discovered eggs in the larva of Porthetria ilixpar, and that Malpighius found them in the larva of Bombyx mori. The reproductive organs, however, are not difficult to observe in some larvas, and can usually be obtained by a little careful dissection. The testes and ovaries are placed just beneath the skin of the 5th abdominal segment. They exist in pairs, one on either side of the dorsal vessel, just above the position of the alimentary canal. The testes form two lobes of a not very distinctly reniform shape, whilst the ovaries, which are only to be seen with a lens, and then in comparatively few species, are much smaller, and consist of tubes. The testes are generally much more readily observed than the ovaries, being, usually, yellow or brown, and may be seen distinctly in the larvae of those species which feed internally, or which have fairly transparent skins. Weniger detected the blind terminations of the ducts from the sexual organs in the larvae of Antheraea yama-mai, A. pernyi, Actias selene and Samia cecropia, " on the underside of the last segment that bears a spiracle " (8th abdominal). In the female of the first of these species is a black blotch, with a yellow central spot, whilst in the male is a similar black blotch, with a dark green central spot. Herold represented, as long ago as 1815, the changes which the essential reproductive glands undergo in the larva and succeeding stages of Pieris brassicae, but up to the present time there appears to have been no external openings, in connection with the sexual organs, discovered in any lepidopterous larva. Certain statements which have been made on this subject are mentioned here only in order to draw attention to them, in the hope that they will be disproved or confirmed. De Geer states that the brown larvae of Triphaena pronuba produce males, and the green larvae, females. Doncaster says that the same larval colour distinction, as to sex, holds good in the Satyrid butterflies. He also states that the male larvae of Oryyia antiqua and 0. yonostiyma have yellow dorsal brushes, the female larvae, brown. Suckow distin- 60 BRITISH LEPIDOPTERA. guishes male Dendrolimus pini larvae from female larvas : (1) By the smaller size. (2) By the lighter, almost smoky-grey colour. (3) By a black- brown band situated beneath the second pair of prolegs. (This band is said to be only obscurely marked in the female). Jackson says that the larval ovaries are situated in the 5th abdo- minal somite, and close to the dorsal middle line. Their proximal or attached extremities are approximated, and they diverge from one another posteriorly. The colour gets deeper during the quiescent period preceding pupation. Four opaque white lines, the future ovari- oles, traverse the larval ovaries lengthwise and converge towards their hinder extremities, from which the larval oviducts spring. The latter are very delicate filaments, and difficult to make out. Bessels gives the following table of species in which the larval testes and ovaries are dissimilar in colour : SPECIES. OVARY. TESTIS. FAT-BODY. Porthetria dispar ... yellow . flesh-red . white Ccsmotriche potatoria yellow Deilephila euphorbiae yellow Pier is brassicae ... yellow Cossus ligniperda ... white yellow reddish violet white white yellow white white Jackson adds that, in these particulars, the larvffi of Sphinx liyustri and Phalera biicephala agree with Cossus. In Pier is brassicae the fresh fat-body posteriorly to the 6th segment is greenish or olive-yellow, anteriorly to it opaque yellow or green on the dorsal aspect, but on the ventral aspect white. The fat-body of the larva of Vanessa io is yellow, and becomes orange in the pupa (Trans. Linn. Soc. Loml., Zool., vol. v., p. 159). With regard to the point of development reached by the sexual organs in the lepidopterous larva, it would appear that they have developed as far as that reached by the adult (imago) Ephemerid (May- flies). In the imagines of the Lepidoptera, the two oviducts unite, and form a single tube down which the egg passes. In the adult Ephemerid, the two oviducts remain separate. In the larva of Vanessa io, the oviducts are separate, as in the Ephemerid imago, but by the time that the butterfly is matured, the oviducts have united to form a quite typical ovipositor. Such a line of evolution, however, suggests that the oviduct of the Lepidoptera passed through a stage similar to that which is to be observed in the Ephemera at the present time, before it reached its present high stage of development. CHAPTER VII. THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. THE variation in the colours of insects is so patent to every observer of these interesting creatures, that there is no need for one to attempt to show that variation exists. Superficially examined, we find that the individuals of a given species are very similar to each other, yet the eye of an expert sees minute differences in these individuals, and he knows that just as no two men or women are exactly alike, so no two THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. 61 insects are, in any of their stages, precisely similar. Variation is general throughout every stage of an insect's existence, i.e., in the egg, larval, pupal and imaginal stages. Every living animal seems to exist for two distinct purposes to eat and to be eaten. Nature provides everything with a means of offence or defence, or both. Among insects, weapons of offence are rare, and, generally speaking, their safety lies rather in their defensive characters. These are usually of the most inactive kind, and consist essentially of various disguises, by means of which, when in repose, they bear a strong resemblance to the various objects on which they rest the bringing into harmony, as it were, the colours of insects with their environment, so that they may agree in tint with the object on which they rest, or that they may bear a close resemblance in hue and shape to some object common upon their resting-place. This bringing into harmony presupposes the possibility of a change in the colours of insects, in order that they may respond to the varying con- ditions under which they may be placed, and in which they have to live. This further presupposes a plastic condition of the colours them- selves, otherwise they would not be able to respond to differences of environment. These differences are so many and so varied, that we find variation in the colours of insects occurring under a multitude of different conditions, and to be presented in a variety of ways. In these notes we shall confine ourselves to the brief consideration of a few of the principal phases of variation exhibited by the imagines of certain Lepidoptera. The colours of the wings of butterflies and moths are due largely to the scales found on the wing membrane, and, in a less degree, to the colours of the wing membrane itself. The scales themselves are hollow chitinous cells, united by a ball and socket joint to the mem- brane of the wing. They are epithelial expansions, which, having attained the size and shape peculiar to the species, become hardened externally by a chitinous deposit. In the process of their develop- ment, they go through a regular series of changes. They are at first transparent, then they become whitish, then a secretion from the pupal haemolymph, called " pigment factor," enters the scale, and it becomes yellow ; lastly the pigment-factor is elaborated, and the scales assume the coloration that they will have in the wing of the perfect insect. These changes, of course, all take place in the pupa, before the imago emerges, and no development takes places afterwards ; any change that then occurs being due to exposure, the influence of light, etc. There can be no active response, whatever, in the perfect lepidopterous insect, to any change of environment, i.e., no change can occur in its colora- tion once the insect has emerged from the pupal state. Ordinary white light can be decomposed. Popularly, we say, it can be broken up into a number of differently coloured lights red, orange, yellow, green, blue, indigo and violet, and we call these the colours of the solar spectrum. These colours, in fact, represent the effect produced on the optic nerve by the variable rate of vibration of the constituent waves, of which white light is really composed. If a substance has the power of absorbing some of the light waves, from the white light which ordinarily falls upon it, and of reflecting others, only the reflected portion can possibly affect the optic nerve. If the red rays only be reflected, then the colour of the substance appears to 62 BRITISH LEPIDOPTERA. us to be red, if blue, then the colour appears to be blue, and so on. Substances which are thus able to select certain light waves for absorption, and to reflect others to our eyes, are termed pigments, and the fact that most scales of Lepidoptera contain substances that can do this, causes us to term the colours thus produced pigmentary colours. But colours are also obtained by the refraction, interference and diffraction of white light. Scratched and striated surfaces diffract light. Diffraction breaks up the bent part of a ray of light into its component parts, and, dispersing the waves, gives, on the edge of each bright space between the slits or striations, a fringe of colour. The exposed surface of the scales of many Lepidoptera are striated, both longitudinally and transversely, hence these produce surface colours by diffraction. One of the best-known examples of this kind of coloration in British insects is the purple of the male of Apatura iris. Such colours as these are usually termed in entomological magazines, non-pigmentary colours. Having thus briefly stated the phenomena by means of which, practically, all the colours of the scales of butterflies and moths are derived, we see that the colours are due either to the selective power of the pigment contained in the scales or membrane of the wing, or they are due to the peculiarities of structure and form of the scale. We have already stated that variation is general in all insects, no two butterflies or moths of the same species being exactly alike. Sometimes this general variation in a particular species is so marked and conspicuous, that the most casual observer notices the fact. Such species are then said to be polymorphic. In a less degree, however, it may be accepted as a general fact that all species of insects are polymorphic. The enemies of butterflies and moths are very numerous insec- tivorous birds, reptiles, mammals, other insects and as they have practically no weapons of offence, their safety lies in their resemblance to their surroundings. Danger, to them, is probably more real when they are at rest, hence, when at rest in a natural attitude, one is at once struck by the marvellous resemblance which most butterflies and moths bear to the surface (or to some common object on the sur- face) on which they rest. With the initial general variation which we have observed to occur in all insects, it is pretty certain that some individuals will be more readily detected than others, some peculiarity of tint, some mark or spot of colour, maybe, rendering them a little more conspicuous. These will fall a more ready prey to the enemies that are searching for them, and they are, as a rule, the first eaten. Those that are best protected are most likely to be left, the laws of heredity step in, and a larger proportion of well-protected specimens results in the progeny. Of course, the general variation which must exist in all broods, and between all individuals, the tendency to atavism, and similar causes, will always result, even then, in producing some less favoured individuals. Still the general result will be that a well protected race, suited to the particular environment by which it is surrounded, will be developed. It is evident, when we consider the different habits of insects, that the particular habit and environment of each species, will determine THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. DO the main general lines on which the variation of the species will proceed. Butterflies sit with closed wings, hence it is the undersides of butterflies that are then exposed, and, therefore, the undersides take such form, colour and markings, under the influence of natural selection, as will best protect the individual, e.g., the marbled green and white underside of Eucldoe cardamines, which rests on umbelli- ferous flowers, the dark undersides and jagged wing margins of the Vanessids, which hybernate in hollow trees, and exactly resemble dead leaves, when at rest. Then there are the " reed " moths, which, be- longing to many different super-families NOCTUIDES, LIPARIDES, CRAM- BIDES, TORTRICIDES, TINEIDES, ZsuzERiDEs sit by day on the reeds, their bodies closely appressed to the reed, their wings folded partly round it, so that each insect represents a gentle swelling of the stem, culminating in an apparent node on the culm, where the insect's head is situated. The colour of all these moths is a very pale wainscot the tint of a dead or dying reed with very fine longitudinal striations, agreeing absolutely with the colours and markings of the reed stem. Another large group of moths chiefly GEOMETRIDES have the habit of resting on tree-trunks, where their general grey hue, marbled with transverse wavy lines, gives them a very close resemblance to the bark on which they rest. Again, in hilly and mountainous districts particularly, a large number of species rest upon the rocks, when their colour usually assimilates closely to that of the rocks upon which they rest, and these, too, are generally covered with transverse wavy lines, which cause them to be very inconspicuous so long as they remain immovable upon their resting-places. Some moths that rest on walls, rocks, or trees, are marked with green and yellow. Such are the species of Folia, Bryopldla and Cleora, Larentiaflavicinctata, and others. These, when at rest, are scarcely to be distinguished from the lichens which grow upon the rocks on which they sit. Then there are the green and yellow moths the Emeralds, Thorns and Sallows which hide among the leaves of trees, or the lower herbage, and resemble, in hue, dead or living leaves so exactly, that they are scarcely to be detected, whilst those that rest among the roots of grass and low her- bage, generally, are of various shades of grey, or buff, or brown, which make them very inconspicuous near or upon the surface of the ground. It is quite clear that, in all these general cases, and in many special ones, natural selection has produced races, particularly well suited in the case of each species to the environment in which it is placed, also that the more conspicuous individuals become a ready prey to enemies, whilst inconspicuous individuals are more often left to carry on the race. One of the most interesting special phases of variation exhibited by British Lepidoptera is that of melanism and melanochroism, the former term being applied to those individuals which exhibit a tendency to develop a greater proportion of black in the ground colour than is exhibited by the type, the latter, when the ground colour is intensified, but not in the direction of becoming blacker. The ab. doubledayuria (popularly known as the "Negro") of Amphidasys betularia may be cited as an example of the melanic class. The ab. ochracea (of a deep ochreous or buff tint) of Spilosoma menthastri, which is white in its typical form, is a very good example of those insects which exhibit melanochroic tendencies. These tendencies are noticed to be much 64 BRITISH LEPIDOPTERA. more generally developed in species that rest on fences, the trunks of trees, the faces of rocks, or on the ground, than in other species. It may, of course, be assumed that those usually found upon fences were originally confined, more or less, to tree-trunks, and that the influences acting upon one are equally potent on the other. It has been observed that, in a great number of species of moths that rest on fences and tree-trunks, and are more or less abundant in the London district, the individuals are darker in colour than those of the same species, captured a few miles outside the metropolis. This is clearly observable in Triaena psi, Haneropkila abruptaria, Acidalia I'irynlaria, Eupitkecia rectangulata, ^Iclanippe fluctuate, Boarmia yem- maria, Hybernia defoliaria, H. marginaria, H. leucophaearia, Oporabia dilutata, Diurnaca fagella, Tortrix podana, Hedya occllana and many other species. There can be no doubt that in the suburbs of London, fences and tree-trunks are generously covered with soot. (Those who have green- houses, and attempt to keep the white paint clean, will understand how completely they are covered). The tree-trunks have become darker during the last fifty years, and the depth of colour is gradually increas- ing in what were then suburban districts. The pale grey and ochreous specimens of the insects just named used to be well protected 'on their then clean resting-places. Such specimens are now exceedingly con- spicuous when they occur, which they only occasionally do, for the selec- tion of the darker specimens for preservation by nature, has resulted in the permanent darkening of the race. But it is in the manufacturing districts in Yorkshire, Lancashire, Cheshire, Derbyshire, Notts, Staf- fordshire, South Wales, etc. where thick smoke is poured from number- less chimneys, and where the fences, tree-trunks, and even the surface of the ground are begrimed with soot, that the most marked cases of what may be termed protective melanism occur. There we get the "Negro" aberration (ab. doubledayana) of Amphidasys betularia, the ab. niijra of Tephrosia creptuctdaria (biundularia), the ab. fmcata of Hybernia marginaria, the ab. obscura of Epunda riminalis, the ab. niijra of Boarmia repandata, whilst many other species give absolutely black aberrations, which are rarely observed elsewhere. These black aberra- tions, it is well-known, have practically come into existence during the last half-century, and their range is rapidly extending. So completely, too, are many of these dark aberrations supplanting the type that, in some localities, the pale typical forms are almost unknown. These moths are nearly all essentially grey that is, black and white in their typical forms. The gradual darkening of the tree-trunks, etc., by the deposition of soot, has resulted in the better protection of the darker specimens, and hence their better preservation, and, as we have just hinted, the trunks and fences have become so blackened that, in some districts, the absolutely black specimens comprise the best protected form of the species. Parallel, if not absolutely identical, with this form of melanism is that exhibited by those species that rest on rocks. Certain Alpine species exhibit this form of melanism in a most marked manner, both in the mountains of Europe and N. America. Certain species that rest on peat are black, wherever they may be found, and however different may be the meteorological conditions of the various districts they inhabit. On the peat bogs in the New Forest, Gnopkos obscwata THE VARIATION OF THE IMAGINES OF THE LEPIDOPTEEA. 65 is black, so also is it on the dark rocks of Perthshire ; in Sussex, on the chalk, it is white, and the response of this moth, in ground colour, to the colour of the rocks on which it rests, is very remarkable. The black specimens found on peat in the New Forest, and on the dark rocks of Perthshire, have a similar melanic appearance, the colour evidently having been induced under such entirely different environments, by a similar process of selection. But it is in the wet, mountainous, and western districts of the British Islands, where the rocks are blackened with moisture, and, even in summer, do not lose one lot of wet until they have received another, that we find the most striking cases of melanism. Thus, on the coasts of Scotland, the Isle of Man and Ireland, we find black races of Af/rotis lucemea, an insect that is quite pale on the chalk rocks of the Isle of Wight. In the Isle of Man the dark ab. manani of Dianthoecia caesia, quite unlike the mottled Continental type, occurs. The aberrations nigra and in- fuscata of Xijlophasia monoylypha, an insect which rests upon the ground, are found in all districts where the rocks are naturally dark, or where there is a heavy rainfall. On the west coast of Ireland, melanic forms of Camptoyramma bilineata are found resting on the rocks, and contrasting greatly with the beautiful golden specimens that hide on the undersurfaces of leaves in our gardens, whilst the aberrations suffusa, intermedia, ochrca and obliterae of Dianthoecia conspersa are found on our northern and western coasts, and respond so perfectly to the rocks upon which they rest, that the professional collectors can tell almost the exact localities in various parts of the Shetlands and Hebrides, from which individual specimens have come. In Shetland, again, the little whitish Emmelesia albulata of our southern pastures and meadows, becomes of a deep unicolorous leaden colour. In all these cases, moisture plays an important, if indirect, part. In the first case, it brings down, in manufacturing districts, the soot in the air, which, when evaporation takes place, is left behind and forms a coating on the tree-trunks, fences, or rocks on which the insects hide. In the second, it permanently dark'ens the rocks in mountainous districts, and more or less so in the western areas, where there is a heavy rainfall. It makes, therefore, the work of natural selection in the direction of producing melanic aberrations exceedingly easy. This aspect of melanism has been already worked out at con- siderable length. There have been occasionally general statements made to the effect that insects from high latitudes are usually melanic. This is so, if only the coast districts and areas with a heavy rainfall be taken into account ; but if the open areas of high latitudes be considered, we find that, although there is a general suffusion of markings and a tendency to ill- developed pigment, due probably to the extreme conditions under which development takes place, yet, as a rule, melanism is rare. Mr. Merrifield has, however, shown us two cases in which temperature tends to pro- duce melanic forms. These are remarkable from the fact that the exposure of the pupa to a low temperature in one case, Eiujonia poly- chloros, produces a melanic form ; in the other, Chrysophanm phlaeas, exposure of the pupa to a high temperature produces a somewhat similar result. These, and parallel cases, are not difficult of explanation. * Tutt, Melanism and Melanochroism in British Lepidoptera, 1891. OD BRITISH LEPIDOPTERA. The pupae are exposed to the low and high temperature respectively, at the period when the scale-pigments are undergoing differentiation in the scales, from the haemolymph of the pupal blood. There is a point at which this elaboration is carried on at a normally healthy rate. At a temperature considerably above or below this normal point, the pigment is developed abnormally, maybe never reaches its normal condition (chemically), or, maybe, overshoots it. In either case, abnormal conditions are produced, and, in these two instances, the abnormality results in a melanic appearance of the insects. There are, of course, other forms of melanism which probably have nothing in common with the cases already cited. One of these is well represented by the ab. ralesina of Dnjas paphia, by the ab. suffiisa of Aryynnis aijlaia, etc., which are probably survivals of the old form of the Argynnid female (vide, Entom. Rec., L, pp. 29-31). The production of albinism in Lepidoptera is not of very frequent occurrence, still it occurs sufficiently often for the phenomenon to be worthy of mention. It occurs in a more or less perfect manner in species that rest on the ground, and which vary in tint according to the colour of the soil upon which they rest. In Gnophus obm-urata, almost purely white specimens are often found in districts where the insects rest upon the bare chalk, and the same is true of Eubolia bi- punctaria, which has almost similar habits. These insects are, in their typical forms, grey, i.e., their scales are some black, others white. The process of natural selection has weeded out the more conspicuous (darker) examples in these localities, until a more or less white race has been produced. It may be urged that these are not truly albinic specimens, but they are exactly parallel in their mode of development with some of the melanic forms to which we have previously referred. True albinic specimens, we take it, are such as those of Calli- morjiha hera, Triphaena promtba, Catocala nupta, and other species that have been recorded, in which the yellow or red pigment has failed, and the scales have become white. In dealing with these specimens it is evident we have a result based directly on physiological processes, for the scales contain no pigment, the normal elaboration of the haemolymph ma- terial having been largely or entirely suspended and the scales filled with air. In our collection are specimens of Hemerophila abruptaria and Hybernia aurantiaria exhibiting this phenomenon, and we believe that the specimens of Sesia culiciformis in which the normal red (or yellow) pigment of the abdominal belt is occasionally white, afford a similar instance. Not very different is the cause which gives rise to the xanthic aberra- tions, which are often included under the same head. In a paper, " The genetic sequence of insect colours," we long since pointed out that many instances of white coloration were due to an unstable pig- ment in the cells, and that certain instances of black coloration were also the result of highly differentiated pigment. These " whites " are very rapidly changed to ochreous or buff under the influence of am- monia, but regain their chemical equilibrium quickly on exposure to the air. The embryonic scale is apparently filled with a secretion from the hremolymph, which, in its first stage, becomes of a milky- white coloration, afterwards changing rapidly to buff and ochreous- * British Noctuae and their I'uriftics, vol. ii., pp. i.-xvi. THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. 67 yellow, the further coloration of the pigment giving rise to the special coloration of the scales. Xanthic patches, sometimes extending to a whole wing, at others to the whole insect, are exceedingly common in Satyrid and Argynnid butterflies, and appear to be due to an arrest of the development of the pigment in the whitish or ochreous stage. The instances of protective resemblance already referred to, remind us at once of the particular case which has been brought under our notice by Wallace, Weismann, Niceville, and others, w^., that affecting the "leaf butterflies " of the tropics. In the instances we have hitherto con- sidered, the resemblance of the insect has been to the surface upon which it rested, gaining its protection by its resemblance to that sur- face as a whole. In the case of the " leaf butterflies," the resemblance is to a special definite object, viz., the leaf of the particular tree on which the insect rests. These butterflies are Nymphalids, and belong to the oriental genus Kallima, the Indian species paralekta, inachis, and philarcJnts being nearly four inches in expanse, while the African species, K. rumia, is smaller. These butterflies are conspicuous objects when flying, but when they alight upon a twig, the wings raised over the back, and the fore-wings thrown well forward, the pattern and colour of the undersurface are such that they make a perfect resem- blance to a leaf. The mid-rib consists of a coloured stripe crossing both wings, which, taking its rise at the apex of the fore-wing, is con- tinued over the hind-wing, and terminates in a tail-like extension of the latter, the extension just reaching the twig, and thus resembling the petiole of a leaf. But the similarity of colouring between some of the unprotected Pierids and the nauseous Nymphalids, is, perhaps, more remarkable. It is well known among entomologists that many of the latter are specially protected from the attacks of birds, and other insect-eating animals, by the production of various scents, which make them distasteful and objectionable as articles of food. On the other hand, the Pierids of which our common white and yellow butterflies are good examples appear to be particularly subject to the attacks of numerous enemies. Bates, Trimen, and other observers have noticed that in the tropics, the Pierids, flying with the Nymphalids, frequently lose the ordinary Pierid coloration and type of markings, and become orange-coloured, and marked on the same general lines as the common Nymphalids. So similar are the colour and markings in some instances, that even specialists have been, for a time, deceived, and have failed at first to recognise them, not only as belonging to different families, but even as distinct species. That this similarity served the purpose of protec- tion to the Pierid was first propounded by Bates, and it soon became generally accepted as an explanation of the facts, that the Pierids, owing to their similarity to the nauseous Nymphalids, were less likely to be attacked by birds and other insectivorous animals, which had learned by experience that insects of a certain colour were objectionable as articles of diet. We frequently find that when the sexes of a given species vary much in Habit, there is considerable difference in the colour, and less often in the markings, of the sexes. Sometimes, too, there is con- siderable sexual diversity when the habits are very similar. Many species have the males brilliantly coloured in comparison with the females ; frequently the sexes are almost identical in tint, but the 68 BRITISH LEPIDOPTERA. instances, among insects, in which the female is more brilliantly tinted than the male must be exceedingly rare. Scudder mentions one South American genus where this is so ; perhaps Zephynis qwrcus, Z. betulae and Thecla ilicis may also be cited, but the cases are com- paratively few. Darwin considers that the excessive beauty on the part of the males is due to sexual selection, the females having, through a long period of time, selected the more attractive males. He further thinks that the various forms of beauty originated as casual variations, and that the special characters were then intensified by the selection exercised by the female. Wallace, on the other hand, con- siders that the sober colours of female insects are due to natural selection, and have been the means of their preservation, since the operation of natural selection has eliminated those individuals of the latter sex that are most gay, and, therefore, conspicuous to their enemies. Darwin starts from inconspicuous forms, from which he derives the conspicuous ones, whilst Wallace starts from conspicuous forms, and from them derives the inconspicuous ones. We have al- ready shown that, among the Lepidoptera, facts distinctly bear out Wallace's view. There is no need to give any special examples of sexual dimorphism, for, as a matter of fact, it would be difficult to find among our British lepidoptera many species that do not exhibit this phenomenon to a greater or less extent. Another marked form of variation that occurs in certain species is that known as seasonal dimorphism. In those countries which have a very distinct difference between the summer and winter temperatures, certain species produce one form of the imago in the spring, after the pupa has been exposed for some months to the climatic conditions of winter, and another form of the imago in the early autumn, after the pupa has been exposed for only a few weeks, or even days, to the climatic conditions of summer. The differences between these two broods are usually marked in two ways (1) Size. (2) Colour. It frequently happens that the summer or autumn-emerging brood is the smaller, and this is undoubtedly due to the difference in the quantity of food eaten, since the larval state of this brood lasts a much shorter time than that of those that emerge in the spring, the larvae, indeed, often missing a moult in order to come to maturity more quickly. The difference in colour is probably due, in different species, to two distinct causes : (1) The less energy at disposal for the purpose of pigment formation in the quickly- feeding individuals. (2) The direct influence of the temperature on the pigment during its formation. Standfuss asserts (Causes of Variation, etc., p. 5) that, in some ex- periments that he made on lepidopterous larvae, the more the period of larval feeding was shortened by the raising of the temperature, the better marked was the reduction in size of the imago. This was the regular, and almost invariable result. A pair of Kutricha querd folia, of which the male measured 58 and the female 89 mm. across the wings, produced offspring of which, after a sojourn of 70-85 days in the larval, and 12-15 days in the pupal, condition, the males measured only 35-37 and the females 36-39 mm. across the wings. Ai-ctia fi, male 46 mm., female 48 mm. across the wings (from pupae British Xoctuae, etc., vol. Hi., pp. xvii. ct seq. THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. 69 collected in the open air), produced three females, measuring 36-39 mm., after a larval stage of 68-87, and a pupal of 15-20 days. Callimorpka dominula var. romanovitf , of 59 mm. and C. var. persona $ of 55 mm., gave rise, after 65-71 days of larval feeding and 14-19 days in the pupal stage, to a form measuring only 35-38 mm. across the wings, in more than a dozen examples. Although in the following experiments no hybernation of the larvae occurred, yet, in contrast to the instances just given, individuals of A. fasciata were reared from eggs of the same pair as above, after 142-163 days of larval, and 25-31 days of pupal, existence, which measured 55-57 mm. in expanse ; and eggs of Dendrolimus pini (male 59 mm. and female 74 mm.), yielded descendants expanding 65-68 mm. in the male, and 84-86 mm. in the female, after 150-172 days of larval feeding and 25-37 days in the pupal condition. In tropical countries where there is less marked difference in the temperature at various seasons, but where there is a marked difference in the rainfall, i.e., in countries where the seasons are known as " wet " and " dry," we get what are known as " dry seasonal forms " and " wet seasonal forms," i.e., that a larva that feeds up during the wet season produces an imago different from that produced from a larva of the same species that has fed up during the dry season. The difference sometimes occurs in the shape of the wing, sometimes it is a difference of colour, more often of ocellation. So marked are the differences, that the forms have, in almost all instances, been described as distinct species, but Doherty succeeded, in the early part of the dry season, in the Island of Sumbawa, in breeding both Melanitis leda and M. ismene, from the eggs of M. leda, the two having been previously considered to be distinct species. This was done by separat- ing a batch of larvae of M. leda, and rearing one part under natural conditions, which produced, in due course, the dry season form, ismene, the other part being reared in a box, in which a wet sponge was kept, in order to retain a damp atmosphere. The imagines produced from the larvae reared under the latter conditions were the wet season form, M. leda. De Niceville, Marshall, and others, have also reared the one form of various species from eggs laid by another. It has been clearly shown that temperature and moisture are two important factors in bringing about changes in the appearance of insects, i.e., they act in such a manner as to exert a marked influence in producing variation. Our climate, of course, is not suitable for the production of " dry" and ' wet " seasonal forms, as it does not afford the necessary conditions. It does, however, afford the condi- tions (although in a much less marked degree, than some parts of Continental Europe) for the production of " spring " and " summer " seasonal forms. Yet, since the action of moisture is so distinctly the predominant feature in bringing about the phenomena of " dry " and " wet " seasonal dimorphism, it must be conceded that moisture is an effective external factor in influencing the larval life in such a manner as to leave a marked impress on the resulting imago, and, in a moist climate like ours, there can be no doubt whatever that moisture is an effective external factor in determining variation, and that its general effect may be considerable, especially in those species * Journal of the Asiatic Society of Bengal, vol. Iviii., pt. ii., No. 1, 1889. tO BKITISH LEPIDOPTERA. which are not particularly affected by differences of temperature. On the other hand, temperature, possibly, has some effect as a factor in determining general variation in tropical countries, where it is, how- ever, overshadowed by moisture. There are, probably, many other external factors besides "moisture and drought," and " high and low temperatures," which react on insects in such a manner as to deter- mine variation, but these are the factors which have, at present, been most studied. It may now be understood how a species, which exists under two very different environments, may produce two very different-looking imagines, so different, indeed, that their specific identity may be doubted. In the European fauna, this is well illustrated by species that exist both at the sea-level and also on mountains at a high elevation. The specimens of Melitaea aiirinia from the plains are large, brightly coloured and am pie- winged. Those from the mountains are small, ill-pigmented and narrow-winged, and are known as var. nieropc. That the factor that determines this change of size is food, we think very possible ; that moisture and temperature have also something to do with the matter, is exceedingly probable. In this manner we get " lowland " and "alpine" forms of the same species; "northern" and "southern" forms similarly occur when a species is spread over several degrees of latitude; "eastern" and "western" forms, when specimens of a species caught in Western Europe are compared with specimens of the same species taken in Japan, and so on, differing often in size, shape or colour, or even all combined, and such races known as geographical races are often so distinct, that an expert can often tell at a glance the exact area or district from which the specimens have come. The differences that mark these various races have an important bearing on the question of the origin of species. We believe that these differences are often correlated with variations that exist in the organism itself ; sometimes, indeed, that they are the manifestations of such variation, and, if the conditions which are thus set up, and in- sisted upon by the environment year after year, be intensified, as, by the localisation and isolation of these races, they must be, the differ- ences may often become permanent and acquire specific value. Differences in habitat altitude for example often permanently alter the time of appearance of the insect in the imaginal state, and thus the life-cycle is modified, the particular form is isolated, and its special features become, as it were, more and more fixed. Thus far we have dealt only with the external manifestations of variation, as they are presented to our observation. Dixey has shown us by what sequence of modifications the patterns of the winga of the Nymphalid butterflies have been formed from a more primitive type. He has also shownf us how the mimicking Pierids have attained the markings by means of which they so closely resemble the nauseous Nymphalids they mimic ; but he has given us no clue as to the phy- siological processes underlying these changes. Starting from the basis that every portion of an insect's wing has in it, from the germ, endless possibilities in the direction of variation, Weismann argues that utility determines the particular form of variation which will be acquired by the individual. We have already criticised, J * Trans. Ent. Soc. Land., 1890, pp. 89 et seq. f Ibid, 1894, pp. 249 et seq. J Entom. Record, etc., vol. viii., pp. 1 et seq. THE VARIATION OP THE IMAGINES OF THE LEPIDOPTERA. 71 at length, the details of this theory as to the origin of the protective coloration of the leaf-butterflies of the genus Kallima. We consider that the response to environment, as exemplified by the form, colour, shape, etc., of insects, is the outward expression of certain variable factors, which, arising within the organism, are directed as to what exact lines they shall ultimately take by the external conditions of life, i.e., by utility. Weismann asserts that selection and utility originate the colour-patterns, which is true so far, and only so far, that actual colour-patterns do not exist until natural selection forms them out of the crude material at its disposal. It appears to us that there are two processes of selection engaged, before the ultimate production of any colour-pattern (1) Selection (internal) among the biophors them- selves, and dependent on the vital processes. (2) Natural selection (external), by means of which the variously coloured scales are formed into a pattern useful to the insect. The scale determinants, it seems, are subject to the same physiological laws as those of other organs. They are guided in the course of their development by various con- siderations, and whilst their general characters are due to internal forces, the special peculiarities of their arrangement are determined by natural selection ; the particular variable factors which are useful for the preservation of the species being chosen for the purpose of building up the required patterns. The fundamental difference (apart from detail) between the old conception of selection, and that more recently propounded by Weis- mann, is that by Darwin, variations were considered to be fortuitous, and that selection had to wait for one of these chance occasions to occur, whilst Weismann considers that every portion of the organism contains within itself, from the first, an indefinite number of varia- tions, some of which are almost sure to be in the direction required. Selection chooses those required, and, by the process of intra-selec- tion, compels them, as it were, to overcome their competitors, and utilises them to produce those results which shall be of service to the organism. At present, however, we have not got to the all-important factor in the study of variation, -viz., what are the physiological factors that decide which of Weismann's theoretical " determinants " shall be developed, and which be extinguished. Weismann tells us that, even in the germ, every part of an insect a wing or a scale however large or small, is composed of theoretical molecules called " determi- nants." These, by intra-selection, i.e., by the competition of the molecules themselves, become non-existent as one absorbs the other, the predominant " determinant " finally deciding the nature and character of the part. He, therefore, considers that the particular form of the part, say a scale, is determined at a comparatively early stage of the insect's existence, i.e., once the predominance of a particular " determinant " is assured. Although the assumption of " determinants " enables us to explain certain phenomena, it does not bring us any nearer to the actual physio- logical activities which result in variation, nor does it explain to us how certain external factors result in variation. To say that a scale of an insect originally has in it the potentialities of becoming white, yellow or red, and that when the scale finally emerges red, to explain it by saying that the red " determinant " was successful over the white 72 BRITISH LEPlDOPTERA. and yellow " determinants," does not help us much, and we un- hesitatingly affirm that whether the scale is finally red, yellow or white, depends primarily on the conditions under which the organism carries on its existence. Let us hark back a little. The wing of a butterfly is present in the embryo caterpillar before it hatches from the egg. The wing develops with the caterpillar, and with the pupa, and only ceases to develop with the stretching of the wing following the emergence of the perfect insect. Supposing the animal to have been supplied with a sufficient quantity of suitable nutritious food until the moment of pupation, to have been kept under the most perfect conditions of health throughout its larval and pupal existence, as a result we shall have an imago normal and perfectly typical in size, shape, colour and markings. On the other hand, let the food supply be short and innutritious, or the conditions under which it is reared unhealthy, or let the pupa undergo its final metamorphosis under adverse conditions, and we shall get a specimen small in size, stunted, crippled, maybe failing more or less in colour and modified in markings. Everyone who has bred insects in large numbers knows that these results are certain. These facts are only mentioned to show that these aberrations are outward manifestations of the vital activities of the insect. Let us go a little more minutely into the subject. When the pupa of an insect is formed, the tissues (except those connected with the reproductive system) undergo histolysis. They are reduced and changed in character, and, from -the material resulting in the de- gradation of the tissues as it were, new tissues are built up by the process of histogenesis. Among others, the scales are formed from epi- thelial cells, and they are filled with a secretion from the haemolymph, known as "pigment-factor," containing the chemical constituents for pigmentation. The pigmentary matter is deposited on the inside of the scales, the hsemolymph secretion is withdrawn from the scales, and air takes its place. The pigment in the scales of insects is of an excretory nature, i.e., it is a product derived from the pnpal blood, and, in a pupa, the larva of which has been reared under typically healthy conditions as to food and environment, this material will be normal ; but if the pupa be weak, due to the unhealthy conditions of food and environment to which the larva has been subjected, will not the material from which the pigment is elaborated suffer with the other tissues, and will not this weakness tend to result in a departure from the normal, i.e., produce an aberration ? If it be granted that these conditions are a possible cause of variation, it may readily be surmised that less pronounced changes in the life of a species may produce a less pronounced change in the general appearance, colour, and markings of the individual. It may be that the changes are severe enough to influence, but yet not severe enough to seriously affect, its health. If the change be permanent, then the outward manifestations of the changed vital processes will be exhibited permanently possibly in the colour and markings. In this manner the differences existing between local races of the same insect living under different environmental conditions may possibly be explained. The predisposing factor to the change may have been food, moisture, heat, cold, or one of many other things, but the factor THE VARIATION OF THK IMAGINES OF THE LEPIDOPTERA. 73 acting on the organism has brought about the result we see. If the result produced by these internal activities be such as to cause positive injury to the race, by rendering the individuals conspicuous, more palatable, etc., then natural selection will direct the variations that exist in the newly-formed race, into such lines, as will necessarily be of advantage to it. But it may happen that a larva may exist under very distinctly favourable conditions until pupation takes place, and that, then, the pupa may be subjected to unusual conditions. We have already seen that the pupal period is that in which the wing-scales are entirely formed, and their contained pigment entirely elaborated. It is well- known that the pigmentary material goes through a regular (and for each species, fixed) genetic sequence before the mature colour is reached. It is evident, therefore, that changes of colour, due either to modification of scale-structure, or pigment, will be more readily effected in this stage than any other. The most powerful factor in bringing about a direct change has, thus far, been temperature. By exposing pupae to tem- peratures to which they are not normally subjected, Merrifield has produced definite changes in the colours of certain species. These changes have been particularly marked in those species which have normally two seasonal spring and summer forms, varying in colour. Other species have, however, responded to the stimulus somewhat readily. No general results, however, have yet been deduced. In some instances (Kutjonia polycldoros), as we have already seen, a low temperature produces a darkening of the normal colour ; in others (Chrysoplianw phlaeas), a high temperature produces a similar effect. What is the nature of the difference in the pigment-factor of the scales of these two insects that makes them thus respond in such similar ways to different stimuli ? It is well-known that, within the area of distribution of a species, there is a certain part in which the environment is more perfectly fitted than in the remainder, by food supply, climatic conditions, etc., for the development of the species in its most vigorous form. Outside this limited area the species exists under less completely favourable conditions ; the food-supply may partially fail, the climatic conditions, or other external factors of environment, may be less suitable, and, as a result, the insect produced may be less vigorous, less highly developed, either as regards size or colour, and may altogether show considerable difference from individuals developed under the most favourable conditions. The darkness (or brightness) of Chrysopkanus phlaeas seems to be due to climatic (temperature) conditions. It is well known that, as regards acclimatisation, some species succeed better in cold and others in hot, some in wet and others in dry, seasons. Whether this be due to the fact that some insects have spread to us from more northern, others from more southern latitudes, or to other causes, it is rendered highly probable that the same amount of heat may act prejudicially on one insect and advantageously on another. In the latter case, increased heat may be expected to produce effects that show an increase of vitality, whilst in the former, cold will produce the same result. A great excess of either heat or cold would, of course, be injurious to any species. To an insect that exists in Britain, at a mean temperature, say of 54, but prefers 60, any decrease of tempe- rature will be injurious, whilst increased temperature will affect it 74 BRITISH LEPIDOPTERA. beneficially, until it reaches 60, and probably will not affect it preju- dicially until it exceeds 66. It is a fact that the largest, most vigorous, and most brightly coloured specimens of Ckrysopkanut phlaeas are obtained in the temperate parts of the Palaearctic area, and that, as we pass south, the insect becomes less brilliant, darker, and often smaller. This tends to show that it is one of those species that prefer-an environment more like that of our temperate climes, and that a higher temperature affects it more or less prejudicially. The most easily-marked evidence of this prejudicial action appears to be seen in the scaling, for, even in Britain, a very hot summer, like that of 1893, always produces a fair proportion of dark specimens, even in those localities where, in cooler seasons, the colour is most brilliant. This is sufficient to prove that the range of variation in the determinants of the scales is such as may enable the insect to be either black or of a bright ruddy golden colour, and the external stimulus which brings one or other of these extreme con- ditions to the fore, appears to be that of temperature. If we apply the simplest elementary laws of vital force to our con- sideration of the development of the pupa, we find that the following facts hold good : (1) The pupa, when first formed, has a certain amount of inherent vital force, by means of which, both the process of histolysis, and that of histogenesis, are carried on in it. (2) That pupa which has the nearest approach to the normal amount of vital force will undergo the most perfect histolysis and histogenesis, and will produce an imago most nearly conforming to the natural type, that is, to the form produced under the most healthy and satisfactory conditions. Conversely, the pupa whose amount of vital force is removed from the normal (whether by excess or defect) is the one in which histolysis and histogenesis will be least perfect, and the imago produced therefrom will be farthest removed from the normal type. (3) That individual which has been best fed, and which has enjoyed the most perfect health in the larval stage, will enter pupal life under the most satisfactory conditions, and will (the pupal conditions being equally satisfactory) emerge therefrom as the best specialised product, while the converse of this must also be true. Another important point appears also to depend on an elementary principle. The vital force of the pupa is converted into energy ; the energy at the disposal of the pupa is most probably directed, first, to the building up of the vital and reproductive organs, afterwards to the secondary organs or tissues, or such as are not necessary to life. There- fore, any excess of energy in a pupa will be expended, as a rule, on secondary structures rather than on vital ones, and so we find that a weak or diseased pupa fails first in regard to non-vital tissues, such as pigment, scales, wing membrane, etc. It would appear therefore that, as a general rule, pigment, scales, etc., are well- or ill-developed in proportion to the amount of material and energy available for the purpose. As a result, such insects as pass through their metamorphoses at the normal temperature, pro- duce the form which is normal for the district ; that is, they undergo the normal processes of histolysis and histogenesis, and, in a state of health, have at their disposal the energy requisite to give them the normal wing-expanse, scaling and colour. If an increase or decrease of temperature lowers the vitality of the pupa, it lessens the available THE VARIATION OF THE IMAGINES OF THE LEPIDOPTERA. 75 energy. The insect, therefore, does not develop under such favour- able conditions ; it needs what energy it possesses to build up its vital organs, and so fails in perfectly building up the secondary tissues. This failure is in direct proportion to the degree in which the vitality is lessened. If the temperature during the period of active develop- ment be below a certain degree, the vital force ceases to act at all, and death results. Heat, greater than that to which the insect is normally subjected, instead of reducing the vitality to the lowest ebb at which life can be sustained, affects the histolysis and histogenesis, usually, in a di- rectly opposite manner. Under its influence the vital processes are carried on at express speed, Energy is expended at the fastest rate possible, and the tissues are developed without having sufficient time to mature, as they would under normal conditions (we may here suppose these to be those that are most beneficial to the species) ; the surplus material is rapidly utilised, with the result that as marked an abnor- mality is produced under the one condition as under the other, although in an opposite direction. It is conceivable that to insects which normally mature at a low temperature, a moderately high temperature might be fatal, and that the pupal tissues would not form at all. It is clear, however, that all changes in the environment of the pupa must necessarily produce some effect on its development.. If the change be sufficiently extreme, then the effect is death ; anything short of such an extreme will produce an effect proportioned to its magni- tude. If a pupa be thoroughly acclimatised to a given range of temperature, then excessive heat or cold must be injurious. The fact that an increased temperature produces dark specimens of Chrysophamts phlaean, must be looked upon as simply a fortuitous circumstance, inasmuch as it appears to be largely due to the dark ground coloration of the scales, for, as we have seen, Ewjonia poly- chloros becomes darker by the subjection of its pupa to a low tem- perature. Probably the physiological result is much the same in both cases ; heat in the case of C. phlaeas, cold in the case of E. polychloros, being detrimental to the development of the most highly specialised individuals of these species. We consider, therefore, that within the limits of existence, the possibilities of the germ are such, that the determinants of the scales (under the influence of intra-selection), present a range of variation within the extreme limits possible to the species, and that external influences determine, through their action on the organism, which of the three before-mentioned factors shall come to the fore in the final production of the scales. There can be no doubt that it is in this direction that experiment and observation are particularly wanted, if we are to obtain any real, as apart from a theoretical, knowledge of the factors underlying variation. These problems relate rather to vital activities, and to physiological phenomena, than to anatomical structures, or the external guiding influence exerted by natural selection, and it is on these lines, it appears, that the laws governing variation will finally have to be worked out. 76 BRITISH LEPIDOPTERA. CHAPTER VIII. THE PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LEPIDOPTEROUS LARVJE. We have already, incidentally, referred to the fact that the exposed life led by many lepidopterous larvae, renders them very liable to be attacked by ichneumons, and to be preyed upon by various mammals, birds, reptiles, and carnivorous insects. We have also shown that they have undergone considerable modification, both for the purposes of con- cealment and defence. For the former, we have seen (ante, p. 43) that larval colours are often modified, probably by phytoscopic influences, and that natural selection has produced specialised patterns, by means of which the larvae are suitably coloured to escape detection in the environment in which they are placed. For the purpose of protection, we have noticed how the simple hairs of the generalised setae (ante, p. 45), and the fine clothing of the skin (ante, p. 38), may be respectively modi- fied into dense hairy fascicles and a thick clothing of hair, which render the larvae quite inedible to many entomophagous animals. We know, also, that the tubercles themselves may be modified into chitinous, prickly spines (as in the Vanessid, Saturniid, and many otherwise widely different, larvae), which serve as a defence for the caterpillars against many of their enemies. It may be here remarked that, since the larva leads an independent and entirely different mode of existence from that of the pupa and imago, it often happens that for the successful continuation of a species, a high degree of specialisation is necessary in the larval stage, and that the necessary modification has taken place without a corresponding specialisation of the other stages egg, pupa, or imago. Conversely, it is conceivable that a highly specialised condition may be necessary in either of these stages, whilst the larva remains in a more generalised condition. It is, however, our intention to exclude the consideration of these latter contingencies at present, and to confine ourselves to those specialisations, defensive and offensive, which larvaa have de- veloped for the purpose of protection. We may premise, then, by stating that larva? are protected in a variety of ways. They have, in many cases, adopted various habits of concealment when not feeding. Hesperid and Tortricid larvae twist up leaves, Crambid larvae make silken galleries ; many gregarious larvae (such as those of Eriogeuter, Malacosoma, Cnethocampa, etc.) spin silken webs, and these live therein whilst others (as ]>n>/>ana, Pyrameis, etc.) spin leaves together, and dwell in the tent thus formed ; others, again, feed only by night, some drop to the ground, and others throw themselves about violently, when disturbed. We have already noticed (ante, pp. 43-44) by what processes they have become specialised in colour, so as to resemble the leaves among which they rest. One of the most general forms of ornamentation of grass- feeding larvae, is a series of longitudinal lines. These produce an effect resembling the lines of light and shade resulting from the illumination of a grass leaf, and due to the parallel venation of the leaf. Similarly, other larvae are specialised in colour to resemble the PROTECTIVE COLOKATION AND DEFENSIVE STRUCTURES OF LABV.3E. 77 twigs and other objects upon which they normally rest. We have already seen (ante, p. 42) that the specialisation of hairs and spines makes them unsuitable for the food of insectivorous birds. Other lines of specialisation by means of which they are protected are by eversible glands (sometimes taking the form of flagella), acid excretions, obnoxious odours, dangerous-looking spines, and horns, and spots ; even remarkable attitudes help to swell the sum total of the defensive possibilities of larvae. That larvae are protected by having a habitation into which to retire, and that they thus gain an advantage in the struggle for existence, appears certain. Niceville mentions (Butterflies of Sumatra, p. 394) that the larva of a large Skipper butterfly (Hidari irava) and that of a Nymphalid butterfly (Amatkusia phidippiis) live, at the same time, on the leaves of Cucos nucifera. He says that, owing to their general abundance, the two species often have a severe struggle to live together, in which the more robust Hesperid, which secures a shelter for itself by spinning the leaves together, is generally victorious. The various means by which larvae are protected, owing to their similarity to some part of their food-plant, or by their resemblance to some object common upon it, is well-known. Some larvae resemble structures on the leaves ; thus, whilst the larva of Apoda avellana assimilates to the surface of an oak leaf, that of Heterof/enea cruciata has been compared with a gall. Packard also says that the larva of Lithacoiles faaciola and those of Packardia are entirely green, oval in form, and might easily be mistaken for a fold or bend in a leaf. The greater part of the Geometrid larvae resemble twigs, whilst arboreal Noctuid larvae are either coloured so as to suit their envi- ronment, or otherwise resemble some portion of the tree sufficiently well to escape detection, whilst ground-feeding larvae resemble, in tint, the ground on or under which they rest by day. Elliott says that the larvae of the American Heteroyenea flexuosa and H. testacea&re wonderfully similar to the red dipterous or aphidid galls on oak and other leaves. Packard, too, notes the resemblance between these larvae and the small reddish-green galls, which appear late in summer on the leaves at the time when the larvae themselves become fully grown. He then adds : These forms being thus protected from observa- tion and harm, do not need the armature of the larvae of the other group (of this superfamily), and the tubercles and spines have disappeared through simple disuse ; while being without poison-bearing spines, they have also lost by disuse the bright colours and conspicuous spots of the armed genera. On the other hand, the larvae of Odoneta, Entpretia, Em-lea, and allied forms, with their remarkably bright colours and markings, and poison-bearing (? urticating) tubercles, feed conspicuously, the warning colours and showy ornamentation repelling the attacks of birds. We are inclined to the belief that the armed slug-worms were the earlier, from the probability that, in the Coleop- tera, the earliest and most generalised groups were the Sta/thylinidae and the "carnivorous Cai-abidae, and their allies; while the later, most extremely modified forms were the weevils and Scolytitlae, in which the larvae are footless. In the Diptera, also, it is not improbable that those families with the most perfectly developed larvae, such as the Culicidae and Tipulidae, were the earliest and most generalised types, while the 78 BRITISH LEPIDOPTERA. Muscidae, with their apodous maggots, present the extreme of modifica- tion though not of specialisation, and so with other apodous insects and apodous Arthropods in general (Proc. Amer. Phil. Soc., xxxi., pp. 84-85). The resemblance of many lepidopterous larvae to a bird's dropping is well known, and the same form of resemblance is often adopted by many lepidopterous imagines (Ant'tthesia salicana, Cilix ylaucata, etc,). So marked is this resemblance when the larva of Jocheaera alni is in its fourth skin, that it is commonly known as the " bird's dropping " stage. The young larva of Papilio machaon is similarly protected, and Niceville says that " the young larvae of P. polytes, like those of P. memnon, P. helenus and P. nephelits, have a strong superficial resemblance to a bird's dropping, which doubtless greatly protects them." When we see a Geometrid larva stiff and rigid on a twig, we are at once attracted by the peculiar structure which enables it to maintain its shape, simply by the pressure of the body-walls on the contained fluids. We also observe how liable such a structure is to danger, and thus, while we note how suited the lepidopterous larva is to exert a great motive force at any movable point of its body-surface by means of its fluid contents, we recognise also that its liability to injury must necessitate some very successful expedients for its protection, if it is to fight its way through the hosts of enemies which surround it. When we examine a number of larvas, we find how rarely they are provided with offensive structures, and, as a rule, lepidopterous larvae rely on a purely passive defence, the most common of which is their resem- blance to some part of their food-plant, such resemblance being their sole protection, and ensuring their escape. We will now examine a few of the special cases in which larvas resemble their food-plant so closely that they can only with difficulty be detected when at rest, and, for this purpose, almost any Geometrid, and numberless other, larvae offer excellent illustrations. The young larvae of lodis vemaria hatches in July or August, is green in colour, rests on the stems of the food-plant (Clematis), stretching straight up, holding on merely by the hind claspers. It has a bifurcate hump on the pro-thorax, standing forward over the head, and its resemblance to a broken leaf-stalk, or tendril, is most remarkable. It is a hyber- nating larva, and in the late autumn, when the leaves and stems of the Clematis turn brown, the larva moults, turns brown with them, and exactly assimilates in colour with the stems of the plant. This brown hue it retains until the spring, and then, when it commences to feed, the brown skin is discarded with the first moult, and it becomes green again like the growing plant, retaining the' green colour until pupation takes place. (The pupa, in a cocoon among the leaves, is also green, and the moth is green). If disturbed, the larva drops by a thread, remaining quite rigid, and looks just like a tiny piece of stick. Somewhat similar to the changes occurring in the larva of 7. ver- naria are those of Geometra papilionaria. In this species, the young larva, which rests chiefly on the branches of alder and birch, is of a pale-brown colour, with, according to Poulton, some power of colour adjustment to the twigs of its food-plant. The larva? remain brown during the winter, but, in spring, moulting produces dimorphism in them, some individuals becoming green, whilst others retain their PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LARV^. 79 brown hue. The larva also becomes stout, and comparatively short, and its resemblance at this stage to the catkins of the birch is very striking, the green larvae resembling the younger, the brown larvae the older, catkins. Harwood says that the brown larvae mature later, and that the larvae found on hazel are somewhat different in appearance from those found on birch. Equally peculiar as to the change of colour, only in this case the change accompanies a change of habit, is that of the larva of Emme- lesia nnifasciata, which feeds within the seed-pods of Bartsia odontites when young and the pods are green, and is itself, at that stage, of a green tint, corresponding with that of the seed-pods. It, however, changes its habit by feeding outside, when almost mature, and con- temporaneously with its last change of skin, it changes its tint and ornamentation. Miss Gould says that the resemblance of the larva of Eumia luteolata, in shape as well, as in colour, is extremely protective, the angular attitude of the larva at rest, rendering it almost indistinguishable from the twig. In the case of larvae with green surroundings, this likeness is greatly heightened by the touches of red, which exactly match the thorns and one side of the stem of the young hawthorn shoot. Poulton also, referring to this species, says that the resemblance of the larva to a twig of its food-plant is most striking, for the dorsal tubercles which are to be found near the middle of the larva represent very faithfully a superficially similar structure upon many side twigs of the food-plant, and, he further notices, that not only do these pro- jections occur towards the middle of the length of the twigs, but they are situated on the angle of a slight bend, a character which is also produced in the larval form. He further points out that the different forms of tha larvae are coloured in almost the same manner as the varying tints of the hawthorn twigs. He considers that the remarkable specialisation of the form and colours of certain larvae to a special food-plant, gives a strong clue to the ancestral food-plant of a species, whose larva now feeds on more than one plant. Barrett notes the resemblance that the larva of Eupitliecia extemaria bears to its food-plant, Artemisia marithna, and says : The stems and leaf-stalks of the plant are furrowed and clothed with white down, in such a manner that all appear striped, with alternate green and dull white, and this larva is similarly ornamented with longitudinal stripes of the same colours and of the same width ; the young flower-buds of the plant are tipped with brown, and the front of the head of the larva is coloured in the same manner ; the segments of the leaves are some- what tumid at the tips, and the anal legs or claspers of the larva are swollen or rounded into precisely the same shape. This last adaptation would appear superfluous, if it were not for a curious trick which the larva has, at times, of raising its posterior end stiffly out while holding on by its thoracic legs thus apparently standing on its head. The special resemblance that the full-grown larva of Hybocampa mil/iameri bears to a curled oak-leaf, partly eaten and abandoned by a Tortrix (mridana /) larva has been well described (Entom., xxiii.,p. 92) by Chapman. He says : By chance 1 one day brought in with the food for some larvae of this species, so exact a resemblance of the full- grown larva, that there could not be any doubt as to the meaning of all its curious outlines and markings. This was a curled oak-leaf, 80 BRITISH LEPIDOPTERA. eaten and abandoned by a Tortriv (viridana ?} larva. This particular leaf was, in detail, exactly imitated by the larva of //. wilhaweri. There was a curled portion of leaf with the outline of the body of the larva, the netted green texture of the leaf like the small markings on the surface of the larva, whilst a brown decayed mark or two were similar to those found on it ; the extremity was eaten off on lines partly following a rib, so as to imitate the truncate aspect the larva has, however viewed ; whilst the secondary ribs of the leaf, being eaten between, projected laterally from the roll, just like the dorsal spines of the larva, and in about the same size and order ; the tall one on the 5th segment, the dwindling ones on the 6th-10th, and the taller bifid one on the 12th, this one resembling points from both edges of the leaf. Most curious, perhaps, of all, the little backward projecting points at the tips of the spines (or humps), apparently so super- fluously complicated in the larva, were exactly represented in the leaf ; the Tortriv larva, in eating the substance of the leaf between the secondary ribs, had eaten these down to some extent also, but stuck fast just at a tertiary branch, the small remaining portion of which precisely represented the backward process of the larval spine. I never met with another rolled leaf that happened, in exact number, size and position, to represent all the processes of the larva as this one did, but almost any rolled and abandoned leaf bore a very close resemblance to the larva. The resemblance of the larva of Smerinthtts oceHatw to a willow or curled apple leaf, is most remarkable, and the larvae of both our other British Smerinthid species similarly resemble the curled leaves of their respective food-plants. The lateral stripes give an idea of light and shadow on the supposed leaf, and the similarity to a willow leaf is often increased in the case of certain larvae of S. ocellatus, in which extra red lateral spots resemble very closely the little red galls on the willow leaves. In America, it has been found that the red blotches on the larvae of S. mi/ops are not at all uniform in number, and are much more frequently found on examples of the late brood, although some of them are entirely green. These red spots correspond exactly in colour with similar spots found on the leaves of the wild cherry (the food-plant of the species) at that seasen. Poulton has given (Tram. Knt. Soc. Land., 1887) a detailed 'account of the remarkable manner in which the larva of Deilepldla hippophavs is specialised in regard to its colour and markings, so as to resemble even in minute detail the peculiarities of its food-plant, tJippophaei rhannwides, and he states that not only are the colours of the leaves faithfully carried out, but the characteristic orange berries are represented by an orange spot at the base of the caudal horn upon each side. Holland notes that he picked up a full-fed larva of Sfauroptu fayi on the path of a beech wood, which very closely resembled a curled - up beech leaf, like those beside it on the path. Poulton says, that when at rest and undisturbed, the larva is difficult to detect, and is protected by its resemblance to a withered beech leaf irregularly curled up, the body, which is often held asymmetrically, representing the leaf, being of about the appropriate diameter, colour and length, whilst the two caudal processes, modified from the last pair of prolegs and always applied together when at rest, represent the leaf-stalk. The second and third pairs of thoracic legs, folded in the middle of their length, PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LARV.E. 81 hang down, and resemble a bunch of the stipules of the foliage leaves of the beech. Among American larvae we find many illustrations of this nature. Packard says that the larvae of the Sckizurae exactly imitate a portion of the fresh, green, serrated edge of a leaf, including a sere-brown withered spot, the angular serrate outline of the back corres- ponding with the serrate outline of the edge of the leaf ; and, as the leaves only become spotted with sere-brown markings by the end of the summer, so the single-brooded caterpillars do not, in the northern States, develop so as to exhibit their protective coloration until late in the summer, i.e., by the middle and end of August. The larva of Schizura leptinoides is of the same shape and colour as a sere-brown, more or less twisted portion, of a serrated leaf, such as that of beech, hornbeam, and similar trees. The larva of S. unicornis is pale, with much glaucous colour about the back, and with certain shades of purple-brown, flesh-brown, olive, and pale green, that make it very similar to the tints found on the withering leaves and canes of the blackberry bushes. Miss Payne writes (Amer. Entom., ii., p. 341) : I think this caterpillar furnishes a wonderful instance of mimicry. The green segments just behind the head resemble a small portion of the green leaf, and the other parts admirably counterfeit the brown and russet tints of the dead leaf, whilst the form of the animal in its various postures aids the deception, by its resemblance to a leaf partly living and partly dead, the green mostly eaten, and the brown torn. Eiley writes : The mimicry of the larva, when on the blackberry, either stem or leaf, is perfect, and the imitative resemblance of the moth when at rest, to the bark of a tree, is still more striking. Hudson records (Entom., xxiii., p. 55) that while gathering some small branches from a birch-tree, on the table-land of Mount Arthur (New Zealand), he discovered a beautifully variegated larva imitating exactly the delicate hues of the lichen-covered twigs. After feeding for a few days, it pupated, and, on June 7th, the imago emerged as a very grey form of Dedana roccoae. He states that he had often before seen the larva of this species around Wellington, where, however, it does not in the least resemble the curious caterpillar found on the table-land. A very similar instance is to be found in the larva of our British species, Cleora lichenaria. So variable is this larva, that it exhibits some peculiarity in almost every locality in which it occurs, the peculiar tint, etc., causing it to closely resemble the particular lichens on which it is feeding. Many years ago, Moller noticed a general tendency for the larva of Amphidasys betularia to be yellowish- green when living on the birch, ashy-grey when on oak, yellowish- brown when on elm, yellowish-green, clouded with rust colour, when on willow or poplar. This general tendency in nature has been found by Poulton to be paralleled in confinement, under varying conditions of environment, and the great amount of colour-variation artificially obtained by this experimenter, shows how valuable all such colours may be under certain possible natural surroundings, or on certain plants th5t the species is known to affect. Thus, Poulton correlates the whitish larvae of this species with trees and shrubs having white pubescent or glaucous shoots. The green larvae he connects with rose, the green shoots of sallow, broom, and Kibes americana. The brown forms are associated with cherry, oak and birch, whilst Sidgwick has 82 BRITISH LEPIDOPTEBA. noticed a difference between the dark larvae beaten from oak and birch corresponding with the difference between the twigs on which the larvae rest in the two cases. Wilson beat a larva of this species from a lichen-covered food-plant, that so exactly resembled the lichen, that he thought the larva must be that of another species, until the moth appeared. The longitudinal lines of the larva of Panolis piniperda make it almost invisible when hiding among the needles of the Scotch fir. The larva of Anarta myrtilli, with its intricate crossing and recrossing of lines, is scarcely discernible when resting on a heather twig. The larva of Anticlea cucullata (sinuata) is scarcely to be detected on the seed- heads of Galium veruin, nor that of Cidaria sayittata on those of Tka- lictnim Jlavum, so close is their resemblance to their respective food- plants, when at rest. Even the large hairy larva of Eutricha qut'iri folia, when motionless on its food-plants sallow, hawthorn, blackthorn and buckthorn is so difficult to detect, that the usual way of finding it, on Wicken Fen, where the species is abundant, is to run the hand down the stems to feel for it. Besides instances, such as those just quoted, of special protective resemblance between a larva and its own particular food-plant, a general protective resemblance, due to a general harmony between the object and its surroundings, is often to be observed. It seems almost impossible to understand how there can be a general harmony between some large and apparently conspicuous larvae and their food-plants, when one considers them apart from each other, but when one sees for the first time, the larva of Deilepliila enjiJiorbiae or J'apilio machaon on its food-plant in a state of nature and surrounded by those plants that make up its natural environment, one is no longer struck with the difficulty often experienced in a first search for the larva, and recognises that, in the blending of the various tints of the plants around it, the fitness of the colours of the caterpillar, for its effectual concealment, is very evident. In such cases as these, the larva does not resemble any one particular piece of the food-plant, but the general character of the larva mimics or resembles the general environment, whilst special parts of the larva represent special objects in the en- vironment. One of the best examples of this general protective mimicry is afforded by the larva of Charaj-es jasiw. This larva rests on the upper surface of a leaf of Arbutus unedo (or on a bunch of leaves fastened together with silk), basking in the sun by day, and always fully exposed. One might suppose from this, that the larva would be very conspicuous, yet, on the contrary, it is difficult to detect. Chapman says (tint. Record, ix., p. 193) : The larva at rest, seen from whatever direction, exactly imitates some aspect of leaves or buds under the different effects of light and shade, and it is thus possible for an untrained eye, in many instances, to look at it, and for it, for some time before seeing it. The yellow lateral line resembles the mid-rib of the leaf seen from above or below, according to light ; the colour and apparent texture of the skin are the same as those of many leaves. The extraordinary head, with its coloured jaws and spines, suggests in many aspects, the little group of buds at the extremity of the branches. One has often to look a second time at certain leaves and branches, as well as at the buds, to be sure that they are parts of the tree, and not a larva. The PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LARVAE. 88 curiously coloured circles on the back of the abdominal segments 3 and 5, which are more brilliant with their blue and yellow than any- thing on an Arbutus leaf, nevertheless produce exactly the effect of certain little rings of fungus or decay, that are very common on the leaves. In the mimicry of larvae, then, it is to be noticed that many of them do not so often exactly imitate the thing mimicked, as some particular aspect of it under certain illuminations, and so, in tidae and Lasiocampidae. V. The SPHINGINA, including the Sphingidae. VI. The EHOPALOCEBA, including the families usually associated under this term. * American Naturalist, 1895, p. 803. t Trans. Ent. Soc. Land., 1896, pp. 129 et seq. and Entom. Record, vol. vii., pp., 268 et seq. t " Relationship of Pyralidffi and Pterophoridaa from the larvae," Ent. Neivs t Feb., 1895. 108 BRITISH LEPIDOPTERA. There was sufficient material here for the basis on which to con- struct the broad lines of a natural genealogical tree, if used in con- junction with the tables given us by Chapman and Hampsonf. But the desiderated clue as to the actual details of such was not obtained until the publication of Chapman's valuable paper, " The phylogeny and evolution of the Lepidoptera from a pupal and oval standpoint." In this we had a factor that could- -be applied in the way desired, and that showed us, not which were specialised and which generalised superfamilies, but which of the specialised and which generalised superfamilies of the various stirpes were related to each other. This paper showed that the form of egg found in each different super-family is very constant, and that thereappears to be no rapid tran- sition from one form to the other among the Lepidoptera. There are, broadly, among the higher Obtect families, two forms of egg, the flat and the upright egg, the former being divisible into the Geometrid and the Bombycid. The Geometrid egg is generally marked by a greater roughness and by coarser ribbing or network; the Bombycid is smoother and more polished, although there are many striking excep- tions to this otherwise pretty general rule. Chapman is inclined to derive these two forms of flat eggs from distinct origins, very low down in the evolutionary scale, but thinks it probable that the various forms of the upright egg (Noctuid, Papilionid, etc.) had a common origin, though very low down. He is supported in this conclusion by the pre- sence of the chin-gland (ante, p. 94), which is found only in Papilionids, Noctuids, Notodonts and other superfamilies with upright eggs, but nowhere among those with flat eggs, and we may accept Chapman's conclusion that, however widely the butterflies are separated from the Noctuids, and the evidence of the Hesperid pupa shows that the butterfly stirps separated from the Noctuid stirps a very considerable way below any Noctua-like form usually placed with the Macros, the evidence of the egg and the presence of the larval chin-gland, suffice to show that they jointly separated from the Geometrids and Bombycids still lower down. The evidence of the egg, too, shows that the Noctuids and Papilionids were not derived, as Meyrick suggests, from any Pyralid form, as the Pyralids are, in some respects, of a higher type than the Hesperids, and yet the former still belong very markedly to one of the flat-egged stirpes. No very clear indica- tion has yet been obtained to show where the upright egg branched from the flat egg. The most probable point is between the Cossids and Zeuzerids. These superfamilies are, in many respects, somewhat closely allied. The former has an upright, the latter a flat, egg, and Chapman considers that we have here, probably, the point where the two forms are still unfixed and capable of easy variation. The alli- ance (by pupa) of Castnia with Cossus, would perhaps point to this also as being somewhat near the origin of the butterfly stirps. Accepting the principles here laid down, there can be no doubt that the flat egg is the ancestral form, and the upright egg a more specialised structure. Examination of a large number of eggs of species belonging to several superfamilies, shows that the upright eggs which characterise the Notodonts, Noctuids, Lithosiids, Euchro- * Trans. Ent. Soc. Land., 1893, pp. 118-119. f Annals and Magazine of Natural History, 1894, pp. 258-259. CLASSIFICATION OF LEPIDOPTEBA. 109 miicls, Lymantriids and Papilionids, are modifications of one and the same structure. If now we turn back to Dyar's group IV, the NOCTUINA (ante, p. 107), and take the superfamilies with upright eggs therefrom, we have left a series of families of which the Tlu/atiriilae (Cyniatophoridae), C-eoinct- ridae and Drepanidae are the most important. These show also a close alliance, not only inter se, but also with the Pyralids and Crambids, since they possess essentially the same type of egg. Here, then, is clearly a dichotomous division in Dyar's NOCTUINA, one branch showing relationship with his MICRO-LEPIDOPTERA through the Cossids, the other through the Pyralids. Below these super- families (Cossids and Pyralids), however, the egg proves of very little value, but other characters of the larva, pupa, and imaginal neuration show that these MICRO-I.EPIDOPTERA belong to one or other of the main stirpes above indicated. Dyar's NOCTUINA (specialised), and MICRO-LEPIDOPTERA (generalised), therefore, divide into the two following groups : I. The NOCTUO-HEPIAUD STIRPS. HEPIALIDES, ZEUZERIDES, TORTRICIDES, COSSIDES (generalised superfamilies), leading up to : (1) NOTODONTIDES, NOCTUIDES, NYCTEOLIDES, ARCTIIDES (with the Lithosiids), LYMAXTRIIDES. (2) CASTNIIDES, HESPERIIDES, PAPILIOXIDES (the specialised superfamilies). II. The GEOMETRO-ERIOCRANIID STIRPS. ERIOCRANIIDES, ADELIDES, TINEIDES, etc. (generalised superfamilies), leading up to BREPHIDES, CYMATOPHORIDES (TIIYATIUIDES),DREPAXULIDES (PLATYPTERYQiDEs),and GEOMETRIDES (the specialised superfamilies). This arrangement practically absorbs three of Dyar's main divisions, leaving only the ANTHROCERINA, BOMBYCINA and SPHINGINA. It is very evident here, from an examination of the eggs, that these all belong to one stirps, and that Dyar has rightly diagnosed and divided these, his ANTHROCERINA representing the generalised, and his BOMBYCINA and SPHINGINA two specialised, branches of the same stirps. The latter works out thus : III. The SPIIINGO-MlCROPTERYGID STIRPS. MlCROPTERTGIDES, NEPTICULIDES, EUCLEIDES, MEGALOPYGIDES, HETEROGYNIDES, ANTHROCERIDES, PSYCHIDES, PTERO- PHORIDES (the generalised superfamilies), leading up to the LASIOCAMPIDES, EUPTEROTIDES, E.NDROMIDES, BoMBYCIDES, SATURNHDES and SPHINGIDES (the specialised superfamilies). We are inclined to attach but little importance to the pre-spiracular tubercle of the Sphingids ; the whole of the other essential characters, both of egg and larva, showing considerable affinity with the Endro- mids and Saturniids. One of the most puzzling points in the taxonomy of the Lepidoptera is the affinity of the Pterophorids. The imagines of this superfamily have, in common with the Orneodids (Alucitids), " plumed " wings, and therefore our more superficial investigators place them somewhere near each other. Of their utter want of relationship Chapman speaks with no uncertain sound. He says : Epermenia and Omeodea are typical members of the Pyraloid-Micropterygid (i.e., Pyraloid-Erio- craniid) series ; Pterophorus is not a member of this series, etc. Again, he notesf : There is no relationship between the pupa of Orneodes and that of Pteroplwrus. The latter has not followed the line towards the Macros that has been taken by the PYRALIDES, but has struck out an en- tirely separate line of its own, and still retains nearly all the features of a Micro pupa. The only point that interests us here, in connection with * Trans. Ent. Soc. Land., 189(5, p. 145. f Ent. Record, vol. vii., No. 11, 18%. 110 BRITISH LEPIDOPTERA. Orneodes, is that the one Micro character which Orneodcs has preserved and exaggerated (the large cephalic dorsal plate) happens, in Pterophortu, to have taken precisely the contrary direction. In Pterophonta it hardly exists, and is difficult to see ; yet it does exist, and that so effectually that, as in nearly all Micros, it carries the eye-cover with it on dehiscence. Dyar is quite clear as to the larva, and associates the Pterophorids with the Anthrocerids, and as we have examined many of the larvae of this super- family, we can the more readily acquiesce in his claim for this alliance. The smooth Pterophorid egg, too, supports very strongly the suggested alliance with the Anthrocerids. Yet there are strange peculiarities about the Pterophorids as a member of this stirps, not the least of which is the development of a cremaster, by which it attaches itself after the fashion of Hijpcrcallia, Zonosoma, and the Papilionid pupe. By the develop- ment of this cremastral attachment, the Pterophorid pupa has lost the character of motility, practically universal among the superfamilies with Incomplete pupre. On this account, Chapman considers that the Pterophorids are one of the highest of the superfamilies in the INCOMPLETE. On the other hand, the abdominal prolegs of the Anthrocerid larva are so completely of the typical Macro form, i.e., have hooks only on the inner side of the foot, that this character places the latter super-family also high among the INCOMPLETE. There is difficulty in detecting the eye-collar (which is remarkably well-developed in the lowest superfamilies of the Sphingo-Micropterygid stirps) in the Pterophorid,and, to a less extent, in the Anthrocerid, pupa, another character that places them moderately high in their stirps. Chapman considers that, besides the Micropterygids and Erio- craniids, the Hepialids, Nepticulids, and possibly, even the Psychids, have been derived directly from the Palrco-lepidoptera, without any intermediate forms that we now possess to indicate the special lines they took, and that the lines of their evolution, therefore, diverged from this low point. With regard to our distribution of the MICROPTERYGIDES, ERIO- CRANIIDES and HEPIALIDES, as the bases of the Sphingo-Micropterygid, Geometro-Eriocraniid and Noctuo-Hepialid stirpes respectively, we may at once state that each of these might equally well be put at the bottom of any stirps. The Micropterygids are, presumably, the lowest form we have, and, therefore, might be placed as a base for all the stirpes. This superfamily and the Nepticulids, however, show more interesting parallels with the generalised superfamilies (Eucleids, etc.) of the Sphingo-Micropterygid stirps, than with those of any other, and hence are best treated here. Chapman has shown clearly that the Eriocraniids have no very close alliance with the Micropterygids, but he also pointsf out that they show distinct relationship with the Adelids (by way of Incurraria), and through them with the Tineids, Pyraloids and Pyralids. He further points out that the Hepialids, whilst preserving many Tineid characters, show distinct alliance with the Cossids, and that the latter have all the essential characters of the Tortricid, as distinguished from the Tineid, stirps. We have, there- fore, selected those of these superfamilies nearest to the lower forms of each of our three stirpes, and propose to bring them into exami- nation with the stirps to which they have been respectively attached. * Trant, Ent. Soc. Lond., 1894, pp. 336 et seq. f IMd., 1896, p. 132. CLASSIFICATION OF LEPIDOPTERA. Ill The Micropterygids and Nepticulids, therefore, will be considered in connection with the Sphingo-Micropterygid group, although, as we have just shown, they might almost as well have been treated with all the lower superfamilies together. It is, however, necessary to break down the idea that there is a great hiatus between the lower (Incom- plete) forms, and the higher (Obtect) forms on the same stirps, or a similar hiatus between the smaller species and the larger. For the same reason, although we place the Eriocraniids at the base of the Geometro- Eriocraniid stirps, they might equally well be placed almost at the bottom of all Lepidoptera. The Hepialids, owing to their relationship with the Zeuzerids and Cossids, are placed at the bottom of the Noctuo- Hepialid stirps. There are one or two points in which we think the sum total of characters shows that Dyar and Hampson have erred in their group- ings. Thus the PSYCHIDES belong rather to the Sphingo-Micropterygid than to the Geometro-Eriocraniid stirps. The LASIOCAMPIDES most dis- tinctly belong to Dyar's BOMBYCINA, i.e., to our Sphingo-Micropterygid stirps, whilst the NOTODONTIDES as certainly belong to the Noctuo- Hepialid stirps. We have already shown that Dyar's NOCTUINA is heterogeneous on the egg characters, and the two main branches included must be differentiated on other larval characters than those already used, so as to separate the true Noctuids from the Geometrids. We observe that Hampson finds neurational characters to make this separation. We are totally unable to accept the conclusions reached in Dyar's " Synopsis of the Families of Bombycides " (Proc. Bos. Soc. of Nat. History, vol. xxvii., pp. 129-130) as being of any real classificatory value. In this we find the Notodontidae, Pseudoipsidae, Noctuidae, AfHttclidae, Arctiidae, Pericopidae, Nolidae, Litlto.nidae, Euchromiidae, Ltjinantriidae. (of the Noctuo-Hepialid stirps), Brephidae, Geowetridae, T/tyatiridac&nd Drepanidae (of the Geometro-Eriocraniid stirps), united with the Bombyddae, Eupterotidae and Lasiocampidae (of the Sphingo- Micropterygid stirps), to make up the " Higher Bombyces." To explain away the position of the Eupterotids, in which Dyar recognises that tubercles iv and v do not satisfy his definition of the group, we learn that " warts iv andv are degenerate, and have come to assume a generalised position, probably secondarily " (p. 128). Concerning the Bombycidac, Dyar states, what is a fact, that " the warts are small and degenerate, but they are true warts of the typical Lasio- campid pattern" (p. 140). Kegarding the warts of the Lasiocampids, we read that "on the abdominal segments (of Tolype velleda), v is smaller than iv, and all except i and vi are greatly reduced. These, two warts alone persist in the adult " (p. 144). The reduction of iv is quite characteristic of the higher branches of the Lasiocampids, Bombycids, Eupterotids, etc., and entirely different from the well- defined post-spiracular iv of the Noctuids, Arctiids, etc. In its broadest lines, then, our scheme of classification assumes three main general evolutionary branches, along which the various superfamilies of Lepidoptera have developed, two of these being flat- 'egged and one an upright-egged stirps. These, with the main super- families included in each, have already been given (ante, p. 109), so that there is no need to repeat them. * Annals and Magazine of Natural History, Oct., 1894, p. 259, 112 BRITISH LBPIDOPTEEA. That the details of such an arrangement as this will be modified by further observation is highly probable, but that this will form a sound basis for future work we feel convinced. We shall find, for example, in future schemes, no derivation of generalised from specialised super- families, nor a flat-egged family from an upright-egged one, the former giving rise again to another upright-egged family, as repeatedly occurs in the work of Packard, Dyar and Meyrick. As an illustration of this point we may give the following : Packard derives Lithosiidae (an Arctiid form, with upright egg and Arctiid larva) from Tineina (with flat egg), and then derives the Geometridae (another flat-egged group) from the Lithosiidae. In his work the following series occurs : Tineina (flat egg), Notodontidae (upright egg), Ceratocampidae (flat egg), etc., i.e., a reversion from the " upright " to the " flat " egg form in every alternate stage of the evolution. Dyar, too, obtains the following series in one of his phylogenetic trees : Notodonts (upright egg), Eupterotids (flat egg), Lymantriids (upright), Bombycids and Lasiocampid (flat eggs), an impossible combination. The diagram (Plate I) which we have added to illustrate this chapter will show roughly our views as to the evolution of the three main stirpes at their base, and the details of the evolution of the Sphingo-Microp- terygid stirps. * American Naturalist, 1895, p. 803. A60ISTIDES COCHUDIOB MfGALOPYC'IDES PLATE I. PHYLOGENETIC TREE ILLUSTEATING THE DEVELOPMENT OF LEPIDOPTEBA FKOJI A HYPOTHETICAL BASE. THE SPHINGO-MICROPTERYGID STIEPS. 114 BRITISH LEPIDOPTEBA. Stirps I : SPHINGO-MICROPTERYGIDES. THE Sphingo-Micropterygid (or as we have sometimes termed it, the Sphingo-Bombycid) stirps is so-called from two of the most characteristic superfamilies it contains, the SPHINGIDES and the MicnopTERYGiDEs(Eriocephalids),the former, one of the most specialised, the latter, one of the most generalised, of the superfamilies, not only of the stirps, but of all Lepidoptera'. Although our knowledge is at present very incomplete, there appears to be good ground for including on the same evolutionary line with these superfamilies, several others of considerable size and im- portance. These are all more or less characterised by the following structural peculiarities : (1) The possession of a flat egg (i.e., with the long micropylar axis horizontal, and with a short vertical axis). (2) The maintenance of tubercles iv and v, as sub-spiracular tubercles (except in Sphingids, where v becomes pre-spiracular*) ; a tendency for iv and v to become united into a single sub-spiracular wart ; a tendency for i to form a many-haired dorsal wart, and to form, with iii and iv -f v, on either side, a transverse row of warts on each segment ; ii tends very strongly (in some families) to become atrophied. The Micropterygids (i.e., the Eriocephalids of Chapman) are so remarkable, that they have been separated by Packard into a sub-order equal in value to all other Lepidoptera combined, and thus we get : Sub-order I : LEPIDOPTERA-LACINIATA including only the MICROPTERY- GIDES. Sub-order II : LEPIDOPTERA-HAUSTELLATA :- 1. PAL^O-LEPIDOPTERA (Pupse-liberae) including only ERIOCRANIIDES. 2. NEO-LEPIDOPTERA (PuptB-incompletse, and Pupae-obtectae) in- cluding all other Lepidoptera. This, however, represents only the separation of what we may call the stranded remnants of the ancestral lepidopterous fauna, and since Chapman remarks! that "the Zyyaenidae (Anthroceridae), Lhnacodidae (Eucleidae) t and Micropterygidae (Eriocephalidae) form a group which, though the last member is as low as the lowest Tineina and the first as high as the butterflies or Noctuids, has, nevertheless, been evolved on its own lines, from a common source, as a separate branch of Heterocera," we feel quite justified, in spite of the vast gulf that separates them, in retaining these as superfamilies of this stirps, for there are, of course, almost inconceivable breaks between the superfamilies, even of the same stirps, represented (1) in time, by icons of years, and (2) in evolutionary development, by the extinction of thousands of connecting groups, which once surrounded the existent groups, and of which we have now no trace, and can only vaguely surmise either their character or relationships. Roughly, then, and bearing in mind what has just been said, we may divide the superfamilies of this stirpsj into two groups according to the amount of specialisation they have undergone. We should then get : * In Agdistis iv becomes post-spiracular, and v sub-spiracular, thus differing from any other Plume larvae known to us. f Trans. Ent. Soc. Land., 1894, pp. 335 ct seq. \ We are well aware that many other exotic superfamilies may belong to this stirps, but having no special knowledge of the early stages of the species of such superfamilies, they have been excluded. THE SPHINGO-MICROPTERYGIDES. 115 I. GENERALISED SPHINGO - MICROPTERYGIDES. MICROPTERYGIDES. NEPTICULIDES, EUCLEIDES, MEGALOPYGIDES, HETEROGYNIDES, ANTHROCERIDES, PSYCHIDES and PTEROPHOHIDES. II. SPECIALISED SPHINGO - MICROPTERYGIDES. LASIOCAMPIDES, EUPTEROTIDES, ENDROMIDES, BoMBYCIDES, SATURNIIDES and SPHINGIDES. Some of these superfamilies are well-defined, but others have not yet been very clearly separated (by our authorities) from the super- families of the other stirpes that have undergone parallel development in the imaginal condition. The most important fact to bear in mind when considering the affinities of the generalised groups, is that the species of some of the superfamilies are more specialised (or at least more modified) in one stage than in the others, thus the Eucleids have a somewhat specialised larva, and yet the pupa is among the most generalised (with that of the Nepticulids) of all Lepidoptera, and we have just seen that the more or less generalised ANTHEOCERIDES are considered, by Chapman, to be, in some respects, as high as Papiiionids or Noctuids. The MEGALOPY- GIDES are mainly separated from the EUCLEIDES, owing to the presence of seven pairs of abdominal prolegs (on the 2nd-7th and 10th abdominal segments). The EUPTEROTIDES may be Lasiocampids, in the broadest sense, but are here restricted to the " processionary " moths, Cnetho- campa and its allies. The BOJIBYCIDES, similarly, are restricted to the group of which Bombyx mori is the type, and do not include the ENDROMIDES, as suggested by Kirby, nor do they include the Notodonts, Noctuids, and other superfamilies belonging to a quite distinct stirps, as recently insisted upon by Dyar and Grote. There is a tendency to split the PSYCHIDES into Tineid and Bombycid portions, but at present we have no information supporting this view. With regard to the larvae of this stirps, it may be worth while to recall attention here to a character that appears to be of some structural importance. In almost all larvae belonging to the generalised super- families of the Lepidoptera, tubercles iv and v are normally placed below the spiracles, i.e., both are sub-spiracular. In this stirps, the larvae of the generalised superfamilies follow the usual formula in this respect, but there is a strong general tendency for tubercles iv and v to approximate, and (especially after the first moult) to form a many-haired wart, a character that is carried on also to many of the specialised super- families. Nor, in those superfamilies (LASIOCAMPIDES, etc.), in which there is a distinct tendency to the obliteration of the warts, owing to the development of a hairy coat from the ordinary pile of the body, does iv move up to form a distinctly post-spiracular tubercle, a line of evolution very, general in the Noctuo-Hepialid stirps, and probably also in the Geometro-Eriocraniid. On the other hand, this movement is said to take place in Agdistis, at present classified with the PTEROPHORIDES. The diagramatic representation (PI. I) will illustrate roughly what we consider the probable lines of development taken by this stirps, and its connection with the other stirpes. It will be noticed that we have attempted to avoid the method of deriving one superfamily from an existent superfamily. The main line, we consider, carries on many of the ancestral features of egg, larva, pupa and imago, some, maybe, not much modified, whilst others are exceedingly modified. Many characters have, of course, been entirely lost. From this main stem, each branch has carried on certain broad characters, which have become modified into those more special characters which mark the superfamily. The break between the generalised and specialised 116 BRITISH LEPIDOPTERA. superfamilies of the stirps is a very great one, and the reason thereof is not difficult to understand : (1) Larvae are specialised for protec- tive and defensive purposes in such a manner as to obscure the primary structural characters. (2) The obtect pupa (which charac- terises the specialised superfamilies, not only of this but also of each of the other stirpes) is very similar throughout all the Lepidoptera, the ancestral traces have largely disappeared and structural characters are very uniform. (3) In the imago, the neuration is largely modified by the peculiar structure and particular flying habits of each super- family. In all stages, of course, secondary characters are at the mercy of the environment. The relationship of the Micropterygids (Eriocephalids) to this stirps is worked out at length by Chapman. He bases his conclu- sions on the pupal structure, the slug-like form and habit of external feeding of the larva, the homology existing between the extra abdo- minal prolegs and those of the Megalopygids, and between the latter and the abdominal suckers of the Eucleids ; the parallelism between the disposition and structure of the spines of the newly-hatched Eucleid larva, Apoda ardlana (testudo), with the similar arrangement in the larva of the Micropterygids. With regard to this latter point, Chapman says : The spines of the newly-hatched larva of Apoda avdlana (testudo) are parallel in disposition and structure with nothing known, except the similar arrangements in Microptery.c (Erioce/ihala). The relationship of the Nepticulids with the Eucleids is also dealtf with at length by Chapman. He finds the eggs very similar, an agreement in the apodous condition of their larvae (although it must not be forgotten that the former is a miner), and a similarity amounting almost to identity in the pupae, both superfamilies pre- senting the incomplete pupa in its most extreme form, the segments and appendages being quite free in both of them. This latter factor appears to suggest that our treatment of MICROPTERYGIDES, as a superfamily of this stirps, is preferable to that of Packard, whose definition of " Pupae-liberae " is evidently applicable to other pupae, besides those of the Micropterygids. The generalised superfamilies of this stirps (excluding the Nepti- culids) form Dyar's ANTHROCERINA, which he characterises! from the larval characters as follows : Tubercles with single seta, or converted into warts, or absent ; i and ii, ns well as iv and v, approximate or consolidated. Includes the families, Pterophoridae . Anthroceridae (and Pyromorphidac), Merialopygidae and Eucleidae. I. Body cylindrical, prolegs normal, setae single or con- verted into warts Pteroplioridac. II. Body more or less flattened ventrally. 1. Tubercles converted into warts; iv and v distinct. a. Prolegs normal ; warts reduced ... Anthroceridae. b. Two additional pairs of prolegs without hooks ; warts hairy Megalopygidac. 2. Tubercles converted into spinous processes or absent ; iv and v aborted ; abdominal prolegs re- placed by sticky ventral surface ... ... ... Euclcldnr. * Trans. Ent. Hoc. Land., 1894, pp, 335 et seq. f Trans. Ent. Soc. Land., 1893, pp. 115 et seq. \ Additional notes on the Classification of Lcpidopterous Larvae," Trans. New York Acad. Kci., vol. xiv., 1895, p. 54. We would here observe that after the first skin iv and v are as much con- solidated as i and ii in Anthroccra. THE SPHINGO-HICROPTERYGIDES. 117 We are unable to accept Dyar's distinction of " Body cylindrical " for the I'terophoridae and " Body more or less flattened ventrally " for the Anthroccridae, as being of real value, for such larvae as those of Aciptiliayalactodactyla, A.spilodactyla, etc., are most distinctly flattened ventrally. There is also considerable variation in the character of the Pterophorid prolegs, and also in the character of their warts. Our own summary of these related generalised superfamilies based on Dyar's lines would rather be : I. Body cylindrical ; case-bearers ; third pair of true legs enlarged ; tubercles with simple seta, and very small ; tubercles i, ii and iii, often in an almost direct line above spiracle; iv and v approximate ... ... ... ... PSYCHIDES. II. Body more or less flattened ventrally ; prolegs, variable ; tubercles, rarely with single seta, former usually con- verted into warts ; iv and v usually approximate (except ? Agdistis). 1. With tubercles i and ii more or less separated ; tubercles with simple seta, or changed into warts ; prolegs, variable (from few hooks to almost full circle)... PTEKOPHORIDES. 2. With highly specialised prolegs. a. Tubercles forming many-haired warts ; i and ii united into large wart, also iv and v. ANTHROCERIDES. b. Tubercles simple, with single seta ... HETEHOGYNIDES. 3. With prolegs modified. a. Seven pairs of abnormal prolegs, 5 pairs normal, with usual hooks, those on ab- dominal segments 2 and 7 without hooks ; warts hairy MEGALOPYGIDES. b. Prolegs forming sticky sucking-discs on abdominal segments 2 7 ; tubercles con- verted into spinous processes, or absent ; iv and v aborted EUCLEIDES. In many respects the HETEROGYNIDES, a superfamily not included in this tabulation by Dyar, shows very strong Anthrocerid affinities. It has a somewhat Anthrocerid-looking egg, but the eggs are laid within the pupa-case (a common Psychid habit). The wingless female, by a vermiform movement, re-enters the pupa-case after fertilisation ; the apterous condition of the female is another Psychid affinity ; but the larva and pupa are distinctly Anthrocerid, although the tubercles bear a single seta, and tubercles iv and v are distinct, and not warted as in Anthrocera ; the male is very like a Procrid, with some suspicion of a Psychid ; the soft, fluffy, silken cocoon being quite sui generis, yet approaching Antlirocera. Dyar suggests for Heterogyna an affinity with COSSINA (Adelids, Psychids, Cossids, Pyralids, Tortricids, Sesiids, Tineids, Orneodids and Lacosomids), but this must not be taken too seriously, for, later in the same paper (p. 25), the author defines the COSSINA in terms so general, that it absorbs the ANTHRO- CERINA (Eucleids, Megalopygids, Anthrocerids and Pterophorids), and the term (COSSINA) becomes merely a name, in which the generalised families of all the different stirpes are included. Chapman says that, among the Anthrocerids, he has failed to detect any structures in the young larvae examined at all parallel with the spines of Apoda acdlana, and must still rely on the structure of the egg, the form and * " Combination of two classifications of Lepidoptera," Journal New York Ent. Soc., 1895, p, 20. 118 BRITISH LEPIDOPTEKA. habit of the larva, and the very primitive form of the pupa for its alliance with this section (Trans. Ent. Soc. Lund., 1894, p. 348). The Psychids have been spoken of recently by authors as if they were Tineids of the very lowest type. This is not so. They form a generalised superfainily, but not only is the larva of a rather high type (for a generalised group), the pupa also has advanced much more than has that of Anthrocera since they separated from the main stem. Although we consider that the common features of the Anthrocerids and Psychids were not derived the one from the other, but were obtained from a common ancestor remote from both, yet, taking into account the respective specialisation of the two superfamilies, it is evident that such points of resemblance as exist in Heterogynids and Psychids have been derived by the latter from the former, rather than vice versa. The Pterophorids, as Chapman has proved , have no alliance structurally with the Orneodids (Alucitids) and Pyralids, with which they have been associated, and Dyar and others have shown their alliance with the Anthrocerid stirps. Their larvae exhibit a very wide range of variation, some possessing quite simple tubercles, with a single seta, and having tubercles, i, ii and iii, arranged above each other, and above the spiracle, almost exactly as in certain Psychids. Others have the tubercles converted into many-haired warts, very similar to those of Anthrocerid larvae ; in many, i and ii are united into a large many- haired wart, iii forming a second, and iv and v a third, similar wart, thus forming on each segment a ring of prominent hairy warts, so characteristic of this stirps. One genus, Aydistis, has tubercles bearing a simple seta on all the abdominal segments except the 9th, on which there is a caudal horn, reminding one of the Sphingids. This genus, too, is said to have iv and v developed as post- and sub-spiracular tubercles respectively, a most aberrant condition, if true, for this stirps. That the Pterophorids thus show, inter se, a wide range of larval characters is evident. Some of these characters, too, are largely correlated with the habits and mode of life of the larvae, those with simple tubercles being borers. The near approach of the larvae which bear many-haired warts, in their habits and structure, to those of the Anthrocerids is very marked. It must be admitted, in spite of this, that the affinities of the Pterophorids are more puzzling than those of any other of the generalised superfamilies of the stirps. The pupal attachment by a cremaster, in this group, is also remarkable. The difficulty of placing the Pterophorids at all satisfactorily is, perhaps, sufficient warrant for following Chapman and Dyar in this respect. The former concludes that they might be placed with his Micros whose larvae are exposed feeders. He saysf : Dyar places these with the Anthrocerids and Limacodids, and, both in structure and habits, the larva falls into that division as readily as into any other ; at any rate, it is almost certainly not a member of the Adelid series. Further, Chapman states that there is extreme divergence between the pupa of Pterophorids and that of Pyralids, the latter having a pupa that is a true Macro in dehiscence, with the 4th and 7th abdominal segments fixed in both sexes and possessing no Micro characters, except a dorsal headpiece (a character that goes very high up), maxillary palpi, and, * Trans. Ent. Soc. Land., 1896, pp. 129 et seq. t " Notes on Pupae," etc., Trans. Ent. Soc. Land., 1896, pp. 136-137. THE SPHINOO-MICROPTERYGIDES. 119 in some families, appendages that project beyond the 4th abdominal segment. The Pterophorid pupa is a true Micro in dehiscence, has the 4th abdominal segment free in both sexes (and the 7th also in the male), the dorsal headpiece is evanescent, and it has lost the maxillary palpi. As Chapman concludes that " it is impossible for one of these forms to be derived from the other," we consider ourselves quite justified in following Dyar, and linking the Pterophorids with the lower superfamilies of the stirps under consideration. Having briefly discussed the general relationship of the lower superfamilies of this stirps to each other, it may now be of advantage to examine, in more detail, the characters offered by the egg, larva and pupa. The Sphingo-Micropterygid stirps, we have already stated, is characterised by the possession of a flat egg, i.e., an egg with the micropyle at the end of the long, horizontal axis. The egg has three axes of different lengths, of which the micropylar is the longest, and the vertical the shortest. The surface is generally smooth, the sculpturing very slight and shallow, the shell thin, the texture tending to be delicate, and it has both ends of pretty equal size, not being more narrowed towards the micropylar than it is at the opposite end. The Anthrocerid, Pterophorid, and Sphingid eggs, are, however, much more delicate than those of the Lasiocampid and allied branches of the stirps. Although the Eucleid egg is hardly typical for this group, being soft, scale-like, and somewhat flattened, and, in these respects, resembling closely the Tortricid egg, Chapman says that " the type of egg described above, as characteristic of the Sphingo- Micropterygid stirps, is so similar to that obtaining in the Anthrocerids (Zygaenids) and Megalopygids (Cochliopodids), that the assumption may be made that the stirps originated in these families." The egg of the Meijalopytjiilae (teste Packard) does not appear to us to be so different from what one might have supposed to belong to the most generalised form of this stirps. As represented by Latfoa crispata, the micropylar length : the other horizontal diameter : : 3| : 1, the height is, however, reduced to the least possible dimensions. Still, it is essentially a flat egg in all its characters, with three unequal axes ; of which the one representing the thickness is the least. We have, elsewhere, remarked on its similarity-to the Anthrocerid egg, and on its being covered with silky hairs, as in some Lasiocampids Tnchiura crataei/i, Eriogaster lanestris, etc. , There is no typical larva for this stirps, that of every superfamily having been specialised (or modified) in its own particular direction. The Micropterygid larva (like the egg) is quite sui yeneris, but exhibits, as already detailed, some remarkable parallels with that of the Eucleids. The Nepticulid larva, feeding on the parenchyma of the leaves, is a mining whitish-coloured grub, with nine pairs of hookless discs on the abdominal segments. The Eucleid (Cochliopodid) larva has been specialised in many ways, the most remarkable of which is, however, in the substitution of crawling discs for prolegs. The Anthrocerid and some Pterophorid larvfe have been specialised in the union of tubercles i + ii and iv-fv into large hair-bearing warts similar to iii, so that there are three large warts on either side of each segment. The Atfilistis larva is further specialised by the development of a caudal horn on abdominal segment 9 (not 8, as in the Sphingids), 120 BRITISH LEPIDOPTERA. The Lasiocampid larva, in many ways generalised, is specialised in the direction of a thick hairy coat developed from the secondary or skin (not tubercular) hairs, whilst the normal tubercles have become much atrophied or flattened, although their position is similar to that found in the more generalised superfamilies. The Bombycid and Eupte- rotid larvre show affinities tending to lessen the distance between the Lasiocampids and Sphingids, the former inclining to the Sphingids, the latter to the Lasiocampids. The Endromid larva is remarkable for its general Lasiocampid structure (warts, etc.) in the 1st skin, and general Sphingid appearance in the later skins, whilst the Sphingids and Saturniids present to us some of the most highly specialised of lepidopterous larvae ; the former with its bright, oblique, lateral stripes, and prominent caudal horn, the latter, with its remarkable spines and hairs presenting to us larval types that have lost almost all traces of the general characters that distin- guish the least modified superfamilies. Chapman states that larvre have not yet been found, that present characters to bridge over the dis- tance between the Cochliopodids and Lasiocampids, but Packard hints that the Megalopygid larva (Lirjoa) is connected with the Saturniids. The observations, however, made by Chapman on the spines of the larva of Apoda and Eadca, as well as on those of Sphingids and Saturniids, tend to show that all are branches of the same stirps. The observa- tions of Poulton and Weismann on the larvre of A/iin.c, one half of which disappeared before hatching, thus leaving the normal number. Tichomiroff found prolegs developed in the embryo of Boiubyx mori on the abdominal segments 2-10, but those on segments 2, 7, 8 and 9 were absorbed again in a later embryonic stage, and Graber notices that on all the abdominal segments, except the 9th and 10th, of the early embryos of the same species, faintly marked knob-like elevations are to be seen, which may be considered as the first indications of rudimentary appendages. The same author considers that in B. mori " the stage of pantopody has only a very ephemeral duration." What value these ephemeral structures have in Bombycid and Sphingid embryos, and what meaning is to be attached to them is not quite clear. We only draw attention to the fact that they appear in the embryos of two specialised super- families, where probably homologous structures still occur normally in the larvae of one of the most generalised superfamilies of the same stirps. It may be noticed here that the mining larvae of the Nepticulids have nine pairs of abdominal prolegs that bear no terminal crochets or hooks. The sluggish habits of the larvae of many of the superfamilies in- cluded in this stirps, are probably due to their large size, to the highly developed condition of the prolegs, and to the complex mecha- nism by means of which walking is accomplished. Anyone who has watched a Saturniid or Sphingid larva walk knows that it cannot hurry. It is probably on account of this slow method of movement that the various larvae are so remarkably protected by spines or hairs, some of which are of a most complex character. We are inclined to connect these sluggish habits with an arboreal (i.e., as opposed to a ground-feeding) habitat, and have already shown how the larvae of the large Geometrid superfamily have met the difficulty by protective coloration and a decrease in the number of prolegs, which gives them greater speed and lessens the necessity of other special defensive structures, and yet have maintained their arboreal environment. Packard also connects these sluggish habits with an arboreal condi- tion of life, and asserts that such larvae are surrounded by a purer air, * Bombycine Moths of America, pp. 32 and 83. THE SPHINGO-MICROPTERYGIDES. 127 freer circulation, and more equable temperature, this combination of favourable circumstances causing them to eat more. He says : " The fat, overgrown slugworms (Eucleids) may be compared with the over- fed, high-bred pig, which eats voraciously, has little need of rooting, and takes but little exercise. Where, as among cave animals, there is a deficiency of food, we have a constant tendency to slimness, to an attenuation of the body. This is seen in the blind cave Arthro- pods compared with their allies which live under normal conditions." The generalised superfamilies of the stirps present us with a free, or with an incomplete, pupa, i.e., Pupa-libera and Pupa-incompleta of the earlier authors. Pupae of the former kind occur in the Microp- terygids (?), Nepticulids, Eucleids and Megalopygids, of the latter, in the Anthrocerids, Psychids and Pterophorids, whilst the remaining superfamilies have obtect pupae, i.e., the Pupa-obtectaof the early authors. These two divisions, therefore, fall into Chapman's two sections, INCOMPLETE and DETECTS respectively. The free pupa of the Microptery- gids (?), Nepticulids, and Eucleids, probably represents one of the most generalised of all existing forms of lepidopterous pupae, having all the abdominal incisions movable, and the appendages free, i.e., not soldered. Those of the Anthrocerids, Psychids and Pterophorids, are but little more specialised. They have traces of the "eye-collar" (maxillary palpi), a character almost entirely confined to the most primitive pupal forms, and well-developed in the Nepticulids and Eucleids. In the Anthrocerid pupa the free abdominal segments are 3, 4, 5, 6 and 7 in the male, and 3, 4, 5 and 6 in the female. The Psychid and Pterophorid pupae are more specialised, having only 4, 5, 6 and 7 free in the male, and 4, 5 and 6 in the female. That of the latter is remarkable also from the fact that it has a well-developed cremaster. It has not, however, a silken central body girth. When obtect rank is reached there are few structural characters that can be used for differentiation, owing to their uniformity, but it may be noticed that the Lasiocampid pupa has the dorsal head-piece, a character that suggests this as the lowest of the obtect superfamilies of the stirps. The delicate pupal integument is probably another vestigial character. Strangely the obtect pupa of the Endromids has retained the remarkable Micro character of pupal locomotion. This is a peculiar habit exhibited by the pupae of the more generalised Eucleids, Anthrocerids, etc., by means of which they leave (more or less com- pletely) the cocoon before the emergence of the imago. The same phenomenon is exhibited by certain Sphingid (Ckoerocampa) pupre, whilst that of Macrothylacia rubi travels to and fro in its long cocoon in order to take advantage of the heat of the sun. As might be expected, the diverse habits of the imagines of the various species comprised in the superfamilies of which this stirps is composed, have resulted in a marvellous difference in the imaginal forms, some of which are extremely specialised, each in its own par- ticular direction. Without going into detail, we have the heavy-bodied Eucleid^, Lasiocampids, Endromids and Saturniids, the males with strongly pectinated antennae, dashing about with exceeding swiftness in the hottest sunshine, the females lethargic by day, and flying heavily by night when ovipositing. There are the microscopic sun-loving Nepticulids, and the dusk-loving Pterophorids. The Anthrocerids 128 BRITISH LEPIDOPTERA. (both sexes) boom along heavily only in the hottest sunshine, and are entirely inactive by night ; on the other hand, the Sphingids fly swiftly by night, many of the species migrating vast distances, whilst by day they rest on posts, tree-trunks, and similar objects, to which the colour of their roof-shaped wings assimilates. The shapes of the wings of the various superfamilies follow, to some extent, the habits of the insects. Thus, the Sphingids have long pointed wings that carry them swiftly forward in their long flights ; the females of the Eucleids, Lasiocampids, Bombycids, Saturniids, etc., have large wings that enable them to carry their huge egg-laden bodies when ovipositing. No one has yet told us the special value of the peculiar shape of the wings of the plume moths, but the resemblance of some to tiny pieces of dried grass and stick is remarkable, when they have their wings folded up and are at rest, but these will be dealt with at length when we consider the superfamilies individually. The variability of the habits of the imagines of the different super- families of this group is, perhaps, less remarkable than the difference between the habits of the sexes of the same species. Thus in the Eucleids, Lasiocampids, Bombycids, Endromids and Saturniids, as we have said, the male flies swiftly in the hottest sunshine, whilst the female is comparatively sluggish and rarely seen. In the late afternoon or evening the males of almost all these species "assemble" freely to the females (a habit also indulged in during the daytime by the An- throcerids), the female flying much later for the purpose of oviposition. The huge abdomina of the females of these moths explain the differ- ence, for, in spite of the increase of wing area in this sex,*in these superfamilies, the weight of the body prevents the species flying very fast, or to any very great distance. Packard associates this sluggish- ness with the habit of the females laying their eggs near their cocoons. He says : " When the ancestors of the moths belonging to the Bombycid stirps, became arboreal feeders, the species tended to become segre- gated. For example, the females of the Attaci and their allies, as well as the Cochliopodids may, at first, have had larger wings and smaller bodies, or been more active during flight than their descend- ants. Their present, heavy, thick bodies and sluggish habits are evidently secondary and adaptive, and these features were perhaps induced by the habit of the females ovipositing directly upon leaving their cocoons, and cocoon-spinning moths are, perhaps, more slug- gish and heavy-bodied than those which enter the earth to transform, as witness the Ceratocampidae, compared with the cocoon-spinning J>< nnly j- mori and the Attaci. Spinning their cocoons among the leaves at a period of the earth's history when there was no alterna- tion of winter and summer, and probably only times of drought (as in the dry season of the Tropics at the present time), the females may have gradually formed the habit of depositing their eggs immediately after exclusion, and on the leaves of the trees forming their larval abode. The females thus scarcely used their wings (as in Callosamia pro- methca), the males with their larger wings, lighter bodies, broadly pectinated antennae, and consequently far keener sense of smell could fly to a greater or less distance in search of their mates " (Bombycine MotJis of America, p. 19). Among our British species the general principle involved in the above may be largely true, but as a matter of detail, it is open to criticism, for although the females may not THE SPHINGO-MICROPTERYGIDES. 129 wander far to lay their eggs, yet they do wander, and the females of Endromift rersicolor, Saturnia pavonia, Eutricha quercifolia, Lasiocampa qitercits, Cosmotriche potatoria, etc., do not lay all their eggs at one time, nor in one place. Most of these pair where they emerge, lay a batch of eggs near this spot, and then fly a short distance (and pro- bably pair again) before laying another batch. Still, there is, owing to the sluggish habits of the female, a tendency to segregation in all these species. With regard to the relative age of the various superfamilies belonging to this stirps, Packard says that he has always regarded the Bombycids (the superfamily of silkworm moths) as a very ancient one, which has lost many forms by geological extinction. This accounts for the many gaps between the genera. Both the larvae and imagines differ structurally inter se, much more than do those of the Geometrid and Noctuid moths, and the number of species is less. The completeness of the two latter superfamilies suggests that their species have been, to a great extent, developed since, or contempo- raneously with, the early Tertiary period. On the other hand, Packard supposes that the Bombycids originated previous to Tertiary, and probably in Cretaceous, times, and he suggests that the plasticity of the Bombycid larval forms, especially in the more generalised families, is due to the great changes in their environment during the Cretaceous and Tertiary periods. In like manner, Packard says, the great gaps in the genera of our existing Bombycids are probably due to geological extinction, and also to the great plasticity or marked difference in the larvse, as compared with the homogeneousness of the imagines, these being due to the widespread changes in the environment which took place during the late Mesozoic arid Tertiary periods, and which reacted on the insects in their early rather than their later stages. Packard further says : " Were fossil Bombycids ever to be found in Europe, we should expect to discover among them representatives of the Cochlio- podidae, of the Attacine families (Saturniidae and Ceratocampidae), now characteristic of North and South America, or of the tropical regions of Asia, and perhaps of Africa." He bases this view on the theory that these groups have, to a great extent, become extinct in Europe, but still remain characteristic of the American fauna. He says : " Where a family or subfamily is equally developed both in the Old and New Worlds, we are inclined to suppose that it has been a recently evolved group. It is well known that America has lagged behind Europe, geologically speaking, although America is the older continent as such ; the process of specialisation, and then of extinction, has gone on more rapidly in the Old World, or at least the western portion of it " (Bomby- cine Moths of America, p. 82). Superfamily I : MICROPTEBYGIDES. In Hiibner's Verzeichniss, etc. (1826), the genus Micropterix (Microp- tery.v) was founded to receive three species, mucidella, Hb., podevinella, Hb. ( = anincella, Scop.), and pusilella, Hb. ( = calt/iella, Linn.) The first species being an Elackista, it leaves anincclla and calthella as repre- sentatives of Micropteryv, Hb. In 1839, Curtis separated (Brit. Ent., xvi.) the British Micropterygid and Eriocraniid species from Lampronia * This terra is used by Packard to include the Saturniids and Lasiocampids, as well as the Bombycids proper. I 130 BRITISH LEPIDOPTEKA. under the name of Eriocephala, and cited " calthdla " as the type. This, of course, made Eriocephala, Curfc.=Jfwrropteri#, Hb. for Stephens (1835) had constituted calthdla tpye of Micropteryx. The name Eriocephala was maintained in 1850 by Stainton, in his paper entitled " A monograph of the British species of the genus Micropteryx of Zeller," the species then included in the genus Eriocephala being the six now known as British, viz., calthella, aruncella, scppella, aureatella (alliondla}, thunbergella (rubrifasciella) , and mansuetella. Stainton, however, de- scribes (p. 27) seppella as aruncella, correcting this error, and adding the true aruncella in a " Supplemental note " (p. 39). He also included Phylloporia bistriyella (Tinea subammanella) in the genus. Stainton's diagnoses (corrected) of the species of this genus read as follows : I. Head ferruginous. 1. Calthella. Anterior wings golden-green, with the base entirely purple. 2. Seppella. ? Anterior wings golden-green, with the base purple on the costa. )'y.r, and included (Linn. Entomclogiea, v., pp. 322-3) in this genus calthella, aruncella, eximidla, myrtetella, paykulldla, alliondla, subammanella, rubrifasciella, ayladla, and mansuetella ; whilst he placed the " purpurella group" in the newly-created subgenus Eriocrania, evidently on the ground that Hiibner's Micropteryx and Curtis' Eriocephala are identical, as indeed they are. This division is main- tained by Snellen in De Vlinders tan Nederland, 1882, pp. 1065 et set], Kirby properly sinks (Lloyd's Nat. Hint., Lep., v., p. 815) Erioce- phala, Curt., as a synonym of Micropteryx, Hb., but says that the type of the latter is aureatella, a species not included in Hiibner's genus. Following Stephens, it is evident that " calthella" should be the type of Micropteryx and " purpurella " of Eriocrania, Zell. It is quite clear, therefore, that Meyrick's use of the name Eriocephala for the former group, and Micropteryx for the latter, is erroneous. In this work then we shall use the name MICROPTERYGIDES to include calthdla, seppella, aruncella, aureatella, mansuetella and thunbergella. We shall also use the name ERIOCRANIIDES for purpurella, subpurpurella, semipurpurdla , ftparrmanella, unimaculdla, salojnella, sanrjii, etc. Our knowledge of the MICROPTERYGIDES is derived almost entirely from Walter and Chapman, the former of whom first made known the remarkable structure of the headparts of the imago, whilst the latter worked out their life-history, and has told us all that is yet known of their early stages. To Packard is due, in large measure, the assertion of the great value of the facts that Walter and Chapman made known. * Tram. Ent. Soc. Lond., 1850, pp. 26 et seq. THE MTCROPTERYGIDES. 181 So marvellous were the facts brought to our notice by Chapman, and so far-reaching were the suggestions made as to the relationships of the insects, that it seems difficult even now for us to recognise that insects so different in size, shape, and general appearance, should be related, to the exclusion of others, to families with which one could not suppose they bore any relationship. The presence of characters common to the Microp- terygids, Eucleids and Anthrocerids, and absent (so far as is known) in all other Lepidoptera, bespeaks an affinity, in spite of the number- less links that have been extinguished in the course of their evolution. The MICROPTERYGIDES, then, form a superfamily containing some of the most ancestral of all Lepidoptera. The species have no near relatives, although the ERIOCRANIIDES and HEPIALIDES have been united with them, but the alliance has little more in it than the fact that these three superfamilies are amongst the, if not the, most ancestral of all Lepidoptera, and have had no real connection since a geological time which is almost inconceivable. Through all these ages they have retained certain ancestral characters, and whilst thousands of other forms have come, given rise to new forms, and then dis- appeared, leaving us only here and there a group that has been able to withstand the climatic and other changes of so vast a geological period, these have gone on, modified, of course, to a great extent, but retaining many of the features that distinguished them, probably as far back as the Carboniferous or Silurian periods. It is possibly this cause that has made the vast gap between the generalised and specialised families of each stirps, for there can be no doubt that many of the latter (e.g., the Geometrids, Noctuids, etc.), have been evolved in recent times, probably in the Tertiary, certainly one would suppose not before the Cretaceous, period. When, therefore, one wonders at the inclusion of the MICROPTERYGIDES in a stirps of which the highest superfamilies are the Saturniids and the Sphingids, it must not be forgotten that the former are just a little branch of a stem that has divided endlessly, and given rise to a multiplicity of forms under an almost inconceivable variety of conditions, whilst all this time this little superfamily itself has been the sport of the same varying conditions, and yet has retained those characters which enable us to judge of its antiquity. It is not easy at once to uproot one's cherished associations so as to separate the Micropterygids (Eriocephalids) from the Eriocraniids, with which superficially they appear to have so much in common, especially in size, colour, and neuration ; but Chapman's comparison of the two superfamilies (Trans. Ent. Soc. Land., 1894, p. 336), shows us that Mey rick's attempt (Handbook, etc., pp. 802-805) to keep them as genera of the same family is not at all in accordance with the facts at our disposal. There seems, therefore, not only a necessity to place them in different superfamilies, but practically on different stirpes in the classification we have adopted in this work. So far as we at pre- sent know, the Palaearctic species belonging to the superfamily MICROPTERYGIDES are not only referable to the same family, but also to the same genus Micropteryx, Hb. ( = Eriocephala, Curt.). The charac- ters of the genus Microptenjx are givenf by Chapman as follows : * " Micro-Lepidoptera whose larvne are external feeders," Trans. Ent. Soc. Lond., 1894. pp. 335 et seq. t " Some notes on Micro-Lepidoptera," etc., Trans. Ent. Soc. Lond., 1894, p. 336. Revised tn lift., May 5th, 1898. 132 BRITISH LEPIDOPTERA. OVUM : Spherical, or where ovoid the ends are alike ; opaque ; covered with a snow-like coating ; laid externally (not in the substance of leaves). LARVA : Short, square, and angular, with ten rows of globular appendages, and eight pairs of abdominal legs of special structure ; an anal sucker ; two setse on last segment (possibly cerci) ; long antennae ; feeding exposed (i.e., not under a web) on moss. PCPA: (Probably not unlike a Nepticitln, and in a cocoon above ground. Only the head and antenna-piece seen). IMAGO: Six-jointed maxillary palpi, used as feeding-hands; well-developed, serviceable jaws ; ovipositor simple, tubular, of three pieces ; last abdominal seg- ment the seventh. Packard also summarises the characters of the MICROPTERYGIDES, which, as we have already stated, he erects into a suborder, called LEPIDOPTERA-LACINIATA or PROTO-LEPIDOPTERA, equal in value to the whole of the rest of Lepidoptera, called LEPIDOPTERA-HAUSTELLATA, as follows : I. IMAGO : Maxilla, with a well-developed lacinia and galea, arising, as in mandi- bulate insects, from a definite stipes and cardo ; the galeae not elongated, nor united and differentiated into a haustellum, each being separate from its fellow. The maxillary palpi enormous, six-jointed ; mandibles large, scarcely vestigial, with a broad-toothed cutting edge, and with three apparently functional hinge processes at the base, as usual in mandibulate insects. Hypopharynx well developed, some- what as in Diptera and Hymenoptera. The second maxillae divided into a mala exterior and mala interior, recalling those of mandibulate insects ; palpi three- jointed. Thorax with prothorax very much reduced ; metathorax very large, with the two halves of the scutum widely separate. Neuration highly generalised ; both fore- and hind-wings with tbe internal lobe or jugum, as in Trichoptera ; nervures as in Micropteryx (i.e., Eriocrania, Zell.), and showing no notable dis- tinction compared with those of that genus ; scales generalised ; fine scattered setae present on costal edge and on the nervures ; abdomen elongated, with the male genital armature neuropteroid, exserted ; the dorsal, lateral and sternal appendages very large. II. PUPA : Libera (?). III. LARVA: Highly modified in form, compared with that of Micropteryx (i.e., Eriocrania, Zell.), with large four-jointed antennae and very large three-jointed maxillary palpi ; no spinneret (?) ; no abdominal prolegs, their place supplied by a pair of tubercles ending in a curved spine on abdominal segments 1 8; a sternal sucker at the end of the body. IV. EGG : Spherical. Meyrick diagnoses the imago of Micro]tten/.v (ErincejJiala) as follows : Mandibles developed. No tongue. Labial palpi obsolete. Posterior tibiae with spurs placed in groups of bristles. Fore-wings : nervure 7 to costa, 11 connected by bar with 12, 12 giving rise to an additional nervure (13) about middle. Hind-wings as fore-wings, but 18 usually absent (Handbook, etc., p. 805). The taxonomic importance of this group is so great that it must be our excuse if we enter somewhat in detail into the characters which it presents. We have already said that for our knowledge of the egg, larval and pupal states, we are indebted entirely to Chapman, whilst we owe our knowledge of the imaginal mouth-parts (which has given so much material for study) mainly to Walter, Chapman having worked out some few details in this direction independently. With regard to the bearings of the discoveries of these observers on the taxonomy of the Lepidoptera, Packard says : " The presence of two maxillary lobes, homologous with the galea' and lacinia of the Mecoptera (Panorpidae) and Neuroptera (Corydahts, Mynneleoti) as well as the lower orders, Dermaptera, Orthoptera, Coleoptera, etc., in what in other important respects also is the " lowest " or most * Bombycine Moths of America, p. 61, THE MICROPTERYGIDES. 138 primitive genus of Lepidoptera, the lacinia being a rudimental, scarcely functional, haustellum or tongue, and not merely a vestigial structure, is of great significance from a phylogenetic point of view, besides affording a basis for a division of the Lepidoptera into two grand divisions or suborders, for which we would propose the name LEPIDOPTERA - LACINIATA and LEPIDOPTERA - HAUSTELLATA " (Bombycine Moths of America, p. 58). The imaginal mouthparts are perhaps the most unusual struc- tures presented by these remarkable insects, and show most strikingly Mecopterid and Neuropterid affinities. It will be well, therefore, to examine these in detail. After stating that the mouth-parts of the lower Micropterygina (i.e., the Micropterygids) exhibit several most primitive characteristics, Walter shows that the maxilhe are Constructed on the type of those of biting or mandibulate insects, i.e., with an inner lobe (galea) and an outer lobe (lacinia) besides the palpi. He writes of the first pair of maxilla? as follows : "In the first pair of maxillae of Mtcropteryx calthella, aruncella, andersckella, and aureatdla, cardo and stipes are present as two clearly separate pieces. The former in M. calthella and M . aruncella, in comparison with the latter, is larger than in M. anderschella and M. aureatella. In the last two species, the cardo is still tolerably broad, but reduced. The stipes is considerably longer than the cardo in the last two species, whilst it is of the same thickness. From the stipes arises the large G-jointed palpus maxillaris, folded two or three times, and con- cealing the entire front of the head, and all the mouth-parts. At its base, and this is unique among all the Lepidoptera, two entirely separate maxillary lobes arise from the stipes. The external represents the most primitive rudiment (anlage) of a lepidopterous tongue." With regard to this first pair of maxillae, Packard writes : " It is evident from Walter's figures and description, that this structure is not a case of reduction by disuse, but that it represents the primitive con- dition of this lobe, the galea of the maxilla, and this is confirmed by the presence of the lacinia, a lobe of the maxilla not known to exist in any other adult lepidopterous insect, it being the two galeae, which become elongated, united, and highly specialised, to form the so-called tongue, haustellum, or glossa of all Lepidoptera above the Micropteryyidae (Erio- cepkalidae) , which we may therefore regard as the types of the LEPIDOPTERA - LACINIATA. Another most important feature correlated with this, and not known to exist in LEPIDOPTERA-HAUSTELLATA, is the presence of two lobes of the second maxillae, besides the three-jointed labial palpi, and which correspond to the ' mala exterior ' and the ' mala interior ' of the second maxillae of Dermaptera, Orthoptera, Platyptera, Perlidae, Termitidae, and Odonata, and also, as Walter states, to the ligula and paraglossre of Hymenoptera. In this respect the laciniate Lepidoptera are more generalised insects than the Trichoptera or Mecoptera " (American Naturalist, 1895, p. 637). Walter describes the second pair of maxilla?, each of which con- sists of two lobes, the outer and inner mala as follows : " Within and -at the base of the labial palpi is a pair of chitinous leaves provided with stiff bristles, being the external second lobes of the underlip, formed by the consolidation of the second pair of maxilla), and which * Jenaisch. Zeitsch.f. Naturwiss., v., 18 (1884) ; Ibid, v. (1885), pp. 751-807. 134 BRITISH LEPIDOPTERA. reach, when extended, to about the second third of the length of the second palpal joint. Its inner edge is directly connected with the inner lobe (inala interna). The latter are coalesced into a short, wide tube, which, by the greater size of the hinder wall, opens exter- nally on the point, also appearing as if at the same time cut off obliquely from within outward. The outer exterior edge of the tube forms a strongly chitinous semicircle, which, becoming thinner, finally passes into the delicate membranous hinder wall. Anteriorly a deli- cate membrane also appears to cover the chitinous portion. We have here in opposition to the weak, naked underlip, represented by a triangular chitinous plate in other Lepidoptera, a true ligula formed by the coalescence of the inner lobes of the second maxilla? into a tube, as in many Hymenoptera, and with free external lobes which corre- spond to the paraglossas of Hymenoptera." With regard to a paired structure which he considers to be the hypopharynx, Walter states : "A portion of the inner surface of the tube-like ligula is covered by a furrow-like band which, close to the inner side, is coalesced with it, and, in position, shape, as well as its appendages or teeth on the edge, may be regarded as nothing else than the hypopharynx." He then continues : " In the Microptery- gids (Eriocephalids) the furrow is here coalesced within with the inner side of the labium, and though I see in the entire structure of the head the inner edge of the ligula tube extended under the epipharynx as far as the mandible, I must also accept the fact that here also the hypopharynx extends to the mouth-opening, as in all other sucking insects with a well-developed underlip, viz., the Diptera and Hyme- noptera." Walter further shows that the mandibles also exist in the form of true gnawing jaws, like those of the biting insects. They possess powerful chitinous teeth on the opposed cutting edge, twelve to fifteen on each mandible and also the typical articulating hook-like processes by which they are joined to the gena, and fit in corresponding cavities in the latter. In Eriocrania, and other of the more generalised moths, the mandibles in a very reduced form have survived as functionless vestiges of the condition in Micropteryx (Eriocephala). Kellogg, in a paper that would have been much more useful than it is, had it not been for the mixing up of the Micropterygid (Erio- cephalid) and Eriocraniid species in his examination, states that in M. anderschella, it appears to him to be the outer lobes of the maxillae which seem to be free, while the inner ones go to form the very rudimentary proboscis referred to by Walter. However, he is not at all satisfied (nor clear) on the matter, and prefers to leave it open. As to the use to which the remarkable mouth structures of the imagines may be put, Chapman says that they use their great claw-like maxillary palpi with sharp knife points to scrape and tear at both the pollen of the stamens and the surface of the petals, in the latter case, perhaps, collecting fallen pollen. They certainly do something very like eating as regards this pollen, and digest and use it, as would appear from two circumstances : firstly, that very slender moths get very fat, and lay many eggs, and, secondly, their dejecta are very abundant. Moths will live in confinement for three weeks if * " The mouth-parta of the Lepidoptera," American Naturalist, 1895, pp. 546 et seq. THE MlCROPTERYGIDfcS. 135 fed in this way, and kept damp enough. The imagines pair readily, and apparently do so more than once. Packard finds other primitive characters in the head and trunk. He says : " The head is of moderate size, as well as the body, with small compound eyes, and with two ocelli. The occipital region is well developed, as is the epicranium ; the clypeus and labrum are of moderate size. The generalised nature of the thorax is especially noteworthy. The prothorax is seen to be very much reduced, the two tergites being separate and minute, not readily seen from above. The rest of the thorax is very long, exhibiting but little concentration. The mesothorax is but slightly larger than the metathorax. The meso- scutum is very short ; the scutellum rather triangular than scutellate. The metathorax is but little shorter and smaller than the mesothorax, and remarkable for the widely separated halves of the scutum, a neuropterous character (compare Ascalaphus and Corydalm), in which it differs from Micropteryx (i.e. , Enocrania) . The shape of the scutellum is that of a low flattened triangle. As regards the abdomen, attention should be called to the disparity in size and shape between the sexes ; also to the male genital armature, which is very large, and completely exserted, and reminds us of that of Corydalus, in which, however, the lateral claspers are much reduced ; and also of that of certain Trichoptera (Sencostoma, Tinodes, Stenophylax, Hydropsyche, etc.). The neu ration of both pairs of wings is much as in Micropteryx (i.e., Erio- crania] " [Bombycine Moths, etc. p. 59] . Kellogg has shown that in the Micropterygids (mansuetella, thun- bcrtjclla, seppella and anderschella), there are, in common with the Eriocraniids (unimaculella, sparnnaiiella,fastuosella, semipurpurella),a,nd. Hepialids (sylvinus, gracilis, humuli, argentata, hecta, purpurascens, etc.), besides the specialised lepidopterous scales arranged in regular rows or tiers over the membrane, a covering of very fine hairs, differing radically from the true scales, in size, arrangement, and mode of attachment to the membrane. These minute hairs have not yet been discovered in any superfamily other than the three referred to the JUGATE. This clothing of the wings is considered to be essentially that of the Trichoptera, only in a more specialised condition. He also statesf that the well-known scale-hairs of the Trichoptera are simply the true lepidopterous scale in a generalised condition, and that there are many instances among the caddis-flies (Setodes,Mystacides, etc.), of the presence of well-developed scales. Chapman states that the Micropterygid (Eriocephalid) larva is the only other " Micro " larva (not mining or feeding internally or under a web) having a similar form to the larvae of Anthrocerids and Eucleids (Limacodids). He further considers that the suckers on the first eight abdominal segments of the Eucleid larva, are probably homo- logous with prolegs, and also with the eight pairs of abdominal legs of Micropteryx (Eriocephala). Chapman also says that the long spines (described later) which develop rapidly in the larva of Apoda avellana at the,period of hatching, correspond in position with no larval processes known, except those of Micropteryx. In the former, the dorsal series on one side, though consisting of one spine on each segment, has * " Classification of the Lepidoptera," American Naturalist, 1895, pp. 250 et seq. . t "Affinities of the Lepidopterous wing," American Naturalist, 1895, pp.709 et seq. 136 BRITISH LEPIDOPTERA. them placed as though they were the double rows of Microptt>r;/.r, with alternate spines omitted, i.e., the inner spine is wanting on the first abdominal segment, the outer one on the second, and so on alternately ; laterally the spines appear to belong to the lower row of the lateral series of Micropteryx, but the upper row is represented on the thorax by two spines. A further point of resemblance in these spines to those of Micropteryx is in their peculiar spiculate apex, which reminds one much of the peculiar stellate spiculae of the knobs of Micropteryx. With regard to the structure of the larva, Dyar says : " The setae are highly modified, and their arrangement has been much specialised, as shown by the fact that the last two thoracic segments are like those of the abdomen. This is the case in no generalised type, and has only been so perfectly attained in some of the highest lines of development in the Frenatae The curious abdominal legs are unique in the Lepidoptera. Probably they have been derived secondarily, and have no homologues elsewhere." This, as will be seen from the preceding paragraph, is not Chapman's opinion, since he homologises the abdominal legs of Micropteryx (Eriocephala) with the suckers of the Eucleid larva. Chapman considers that the two long setae on the 10th abdominal segment of the larva are homologous with cerci rather than with any ordinary tubercles or processes of lepidopterous larvse, and, comparing these larvae with those of Panorpa, he points out that, in the disposition of the tubercles and of the abdominal prolegs, the resemblance between them is very close. The idea that the bristles on the last abdominal segment are cerci, requires fuller investigation. It seems impossible to correlate them with any of the ordinary appendages of lepidopterous larvae, since they are only bristle-like appendages, and are quite different from the ball appendages of the Eriocephalid larva that pro- bably represent the usual tubercles in other lepidopterous larvae. It is to be remembered that, though called bristles, they are of very large size (for bristles) in comparison with the larva itself, and their struc- ture hasnot yet been investigated (Chapman, in lift., March 25th, 1898). The Micropterygid egg is very remarkable. Those of M. calthclla and M. seppella are spherical, but eggs of M. ainwanella, Hb. (= an/. calthella, due, perhaps, to the fact that the flowers they frequent are smaller and more numerous, so that one single flower does not harbour so many moths ; the moths are generally taken by sweeping, and occur at about the same time as M. calthella. Stainton says that it is a common species, that he found it in abundance in MICROPTERYX SEPPELLA. 151 a wood near Ambleside, and in a lane near Coniston, in June, 1846, and on a grassy bank, near Carron, in 1874, when several pairs were captured in copula. Haram takes it commonly in damp places through- out the Beading district, comprising parts of the counties of Berks, Hants, and Oxon, by brushing lightly amcfag the long grass. South found it in North Devon, resting in flowers of V. chamaedrys, on sunny banks, often several females in a flower, the male very rarely met with. TIME OF APPEARANCE. This species is on the wing throughout June. In the Eeading district, June 6th is a fair average date (Hamm). The following are recorded dates : July 6th, 1858, at Flamborough Head (Horton) ; June 2nd, 1850, at West Wickham; June 4th, 1849, at Carron ; June 5th, 1850, at Sanderstead ; June llth, 1850, at Sanderstead ; June 23rd, 1853, at Lewisham ; July 7th- 10th, 1879, at the Bridge of Allan (Stainton) ; June 13th, 1887, June 16th, 1891, May 25th, 1893, June 6th, 1895, at Greenhithe ; June 18th, 1892, at Eltham ; June 24th, 1892, at Sanderstead ; June 19th, 1893, at Chattenden ; June 28th, 1894, near Farningham ; June 18th, 1895, at Bexley (Bower) ; May 27th. 1890, at Ashtead (Cansdale) ; Jane 19th, 1853, at Castle Eden Dene (Sang) ; occurs rather later in the year than E. calthella, sometimes at end of May, but generally during June, and even as late as beginning of July, common June 28th, 1S97 (Atmore) ; June 29th- July 8th, 1887, July 10th, 1888, at Portland, and July 7th-9th, 1891, in Bloxworth Woods (Kichardson) ; June 4th, 1893, at Chippenham Fen ; June 21st, 1893, roadsides near Cambridge (Farren) ; June 9th-14th, 1881, June 16th-30th, 1882, June 7th-18th, 1883, June 15th-24th, 1884, June 6th-26th, 1885, June 15th-28th, 1887. June 16th-July 2nd, 1888, June 12th-24th, 1889, June 13th-26th. 1890, June 15th-30th, 1891, June 13th-28th, 1892, June llth-25th, 1893, June 14th-29th, 1894, June 16th-27th, 1895, June 12th-27th, 1896, June 15th-28th, 1897, at King's Lynn (E. A. Atmore); June 12th, 1889, near Doncaster; June 22nd, 1893, at Wheatley Woods (Corbett) ; June 16th, 1877, at Windermere (Threlfall) ; May 30th, 1883, two specimens, ex larra, g and ? , then supposed to have been bred from Carex, emerged at Merton (Walsingham) ; March 25th, 1888, at Merton (Durrant). LOCALITIES. ABERDEEN : Pitcaple (Eeid). BERKS : Reading (Hamm), Bulmershe (Holland). CAMBRIDGE : Cambridge and Chippenham Fen (Farren). CHESHIRE: Birkenhead (Stainton), Bidston and Tranmere (Brockholes), New Brighton (Gregson), Knutsford (Chappell). CUMBERLAND : Lake District (Stainton). DERBY : Henhurst (Brown). DEVON: Chudleigh (Stainton), North Devon (South). DORSET; Portland, Bloxworth (Kichardson), Purbeck (Bankes), Wareham (Cam- bridge). DURHAM ; Darlington (Stainton), Castle Eden Dene (Sang), Hesleden Dene (Gardner). EDINBURGH: Edinburgh district, abundant (Evans). GLOUCESTER : Bristol (Stainton), Leigh Woods (Mason). HANTS: Brockenhurst (Farren), north part of county (Hamm). HEREFORD: Tarrington (Wood),Leominster(Hutchinson). KENT: Lewisham and West Wickham (Stainton), Greenhithe, Eltham, Chattenden, near Farningham and Bexley (Bower). LANCASHIRE : Manchester and Coniston (Stainton), Grange and Preston (Threlfall). LEICESTER: Market Harboro' (Matthews). NORFOLK : King's Lynn (Atmore), Merton (Walsingham), Norwich (Barrett). NORTHUMBERLAND: Newcastle (Stainton). OXON: Hardwick (Holland), part of county in Heading district (Hamm). SOMERSET: Castle Cary (Macmillan), Clevedon (Mason). STIRLINGSHIRE: Carron and Bridge of Allan (Stainton). SUFFOLK: Tuddenham (Warren). SURREY: Coombe Wood (Curtis), Haslemere (Barrett), Ashtead (Cansdale), Sanderstead (Bower). SUSSEX: Common in the county (Fletcher), Lewes (Stainton), Cissbury and Clapham Woods (Fletcher), Hastings and St. Leonards (Bloomfield), Abbotts Wood, Chailey, Brighton Downs (Vine). WESTMORLAND: Ambleside (Stainton), Windermere (Threlfall). WICKLOW: Wicklow (Birchall). WILTS: Marlborough (Meyrick). YORKSHIRE: York and Scarborough (Stainton), Flamborough Head (Horton), Huddersfield (Porritt), Wheatley Wood and Sprotboro 1 Woods, near Doncaster (Corbett). DISTRIBUTION. M. seppella is generally distributed in England, and extends into Scotland as far north as Aberdeenshire. It is of general occur- rence in Ireland (Birchall) . It is so mixed up with M. aruncella on the con- tinent that its range is practically unknown. Staudinger and Wocke give ; 152 BRITISH LEPlDOPTERA. England, France and Italy. Stainton's collection contains specimens labelled " N. Germany," and others from Glogau, in Silesia. Stainton records it from the Forest of Soignies, nr. -Brussels. Constant has specimens from Burgundy. Mann notices it from Sicily ; Curo from Tuscany. It is also recorded from France : Saone-et-Loire, Indre ; Germany : Brandenburg, at Havelland and Pomerania. MICROPTERYX MANSUETELLA, Zell. SYNONYMY. Species : Mantuetella, Zell., " Bericht des Schles. Tausch.-Vereins,* v., p. 16 (1844) ; "Linn. Ent.," v., p. 337 (1851); Sta., " Monog.," p. 33, fig. 6 (1850) ; "Ins. Brit.," p. 43; " Man.," ii., p. 303 ; H.-Sch., " Sys. Bear.," v.. jr. 391 (1856) ; De Graaf, " Tijds. v. Ent.." 1870, p. 220 ; Stdgr. and Wocke "Cat.," p. 340 (1871); Wocke, in Hein., "Schmett. Deutsch.," p. 775 (1876); Snellen, " Vlinders," etc., p. 1067; Meyrick, "Handbook," etc., p. 806. Ammi- nella, Wood (necRb.), " Index," p. 231, no. 1602, pi. 50 (1839) (name preoccupied). ORIGINAL DESCRIPTION. Aehnlichund so gross wie calthella, schwars- kopfig, mit verloschenen hellen Binden auf der Vorderfliigeln, bei Glogau in einen Erlenbruche zu gleicher Zeit mit calthdla auf Bliithen von Sorbus auciipariae in Menge, seltener an Spiraea ulmaria u. Kietgrasbliithen (Zeller, Bericht des Schles. Tausch.-Vereins, v., p. 16). In 1851, Zeller gives the following diagnosis : " Capillis atris ; alis anterioribus nitidis aureo-purpureis, fasciis duabus, altera prope basin, altera media, obsolete aureis " (Linn. Ent., v., p. 337). To which, after a reference to his former note in the Schles. Schmett., he adds Stainton's diagnosis : " Capillis atris ; alis anticis aureo-viridibus, macula basale ad costam, fascia ante medium, apiceque, rufescentibus, obsoletis " (Monograph, p. 83, fig. 6). [One might suppose from these descriptions that Zeller and Stainton were describing two different species, but whilst one has taken the golden-purple tint as the ground-- colour, the other has taken the golden-green, and described the purple markings as reddish. Their insects are identical.] IMAGO. Head black (that of other species reddish or yellowish). Fore-wings 8-9 mm. ; shiny, light golden green, with the base, a trans- verse fascia before the middle, and the apex of the wing, purplish. Hind-wings dark grey, apex purplish ; cilia ashy. COMPARISON BETWEEN M. MANSUETELLA AND M. THUNBERGELLA. Stainton writes : M. mansuetella is readily distinguished from every other known British species of the genus by its deep black head ; but, independently of this character, it may be recognised by being of the size of M. calt/iella, with the markings on the anterior wings some- what resembling those of M. thunbenjella, only more indistinct. HABITAT. Healy, early in May, 1861, observed the imagines of this species in numbers flying over and settling upon plants of Mcrciirialis perennis, and although driven away by his hand, they quickly returned. Barrett records it as swarming in a damp place by the railway embankment at Haslemere, on the blossoms of Luzula pilosa, in company with M. calthella, and writes : " At a short distance from Haslemere, there is a swampy copse, consisting principally of alder, sallow and birch bushes, with plenty of bramble, rushes, and high tussocks of sedge, and cut up in every direction with drains. Here, on May 18th, 1866, I found M. mansuetella and M. alliomlla pretty commonly, with M. calthella in abundance, all flitting among and settling upon the rushes and culms of sedge, keeping generally in the shadow of the high bushes, not in the sunshine. At the end of the MlCROPTERYX SIANSUETELLA. 5$ month, these species were found settling upon the sedge blossoms, where they were joined by M. thunberyella." Hoclgkinson notices it as occurring about streams, and in swampy places, at Windermere ; whilst Threlfall says that in the latter district it is local, but very abundant where it occurs. Near Grantham it is particularly attached to flowers of Mercurialit percnnin, and to those of Care.v in woods, and flies in sunshine (W. A. Atmore). Abundant in blossoms of king's-cup ( Caltha faltutrit] in May, and may be obtained by sweeping other herbage, but always in swampy places, about Bloxworth (Cambridge). teller writes : " I discovered this species in a boggy alder swamp in the woods, near Glogau, among low bushes, settling on the Carex blooms in company with Ji. calthclla. After that I found it on the flowers of the Surbu* bushes, whether growing in the open or surrounded by other trees. The insects seek their favourite flowers, often 10 or 12 feet above the ground, from which it' is difficult to dislodge them. If disturbed, they conceal themselves among the blossoms, or creep under the leaves. On some of the umbels a large number were found, and when the bushes were shaken, some were afterwards discovered resting on the grass below." Bankes says that it is found commonly in some wet and mossy spots in woods and coppices near Bloxworth. TIME OF APPEARANCE. Gregson obtained this species, by sweeping, at Windermere, towards the end of May, 1870. Healy obtained it in early May, 1861. Eaton captured it on June llth, 1880, in Portugal, by the streamlet near Cea. Stainton gives the end of May and June as its time of appearance. Other recorded dates are : May 18th-31st, 1866, at Haslemere (Barrett) ; June loth, 1857, at West Wick- ham (Stainton) ; May 28th, 1887, near Grantham (W. A. Atmore) ; May 5th, 1890, May lst-18th, 1892, very abundant at Bloxworth (Cambridge) ; May 22nd, 1896, at Aberfoyle (Evans) ; June 6th, 1876, June 16th, 1877, May 29th, 1878, at Winder- mere (Threlfall) ; May 22nd-25th, 1888, June 2nd, 1891, May 2-th, 1898, abundant, at Merton (Durrani) ; May 22nd. 1850 and May 20th-25th, 1851, at Glogau (Zeller, tcste Durrant) . LOCALITIES. CAMBRIDGE : Cambridge Fens (Warren). CUMBERLAND : Lake District (Stainton). DORSET: Bloxworth, very abundant (Cambridge), Glanvilles Wootton (Bankes). GLOUCESTER: (Stainton). HEREFORD : Tarrington (Wood). KENT : Pembury and West Wickham (Stainton). LANCASHIRE : Manchester (Stainton). LINCOLNSHIRE : Grantham (W. A. Atmore). NORFOLK: Norwich (Barrett), Merton, Eanworth (Atmore). PERTHSHIRE: Aberfoyle (Evans). SUFFOLK; Stowmarket (Stainton). SURREY; Haslemere (Barrett). SUSSEX: Abbotts Wood (Vine). WESTMORLAND: Braithwaite, Windermere (Gregson). YORKSHIRE: York (Stainton). DISTRIBUTION. Zeller gives Glogau and the Silesian mountains (near Warmbrunn). Snellen records it from Holland, near Arnhern ; Stainton received specimens from Staudinger labelled " N. Germany," and Constant has others from Bavaria ; Staudinger and Wocke give Silesia. Meyrick writes: England (local), Germany. Eaton adds: Portugal (testa Stainton, E.M.M., xvii., 246). Other records are Germany : Wiirteiiiburg (Hoffmann), Landsberg, Friedland, Stettin, Hanover (Sorhagen), Prussia (Krause) ; Austria (Schleicher) ; South Norway (Wallengren). MICROPTERYX THUNBERGELLA, Fab. SYNONYMY. Species : Tunbergella (rect. Thunbergella), Fab., "Mant.," ii., No. 131, p. 253 (1787) ; Sta., " Ins. Brit.," 44 (1854) ; " Man.," ii., p. 303 ; H.-Sch., " Sys. Bearb.," v., p. 392 (1856) ; Frey, " Die Tineen," etc., 49 ; Stdgr. and Wocke, " Cat.," p. 341 ; Wocke, in Hein., " Schmett. Deutsch.," p. 776 ; Snellen " Vlinders," etc., p. 1067 ; Meyrick, " Handbook, etc.," p. 806. Rulirifasdella, Haw., "Lep. Brit.," no. 41, 572 (1828) ; Sta., " Monog.," p. 32, tig. U (1850); 154 SfclTlSH LEt>IDOt>TERA. Zell.,"Linn. Ent.," v., p. 334 (1851). Anderschella, Hb.. " Samm.," pi. 51, 352 (1816) Tr., "DieSchmett.." ix.. 2, 126 (1833); Dup.," Hist. Nat.," xi.. p. 403 (1838); Zell., "Isis." 183'J, p. 185. Hcllu-u,dla, Stphs., " 111.," iv., 360 (1835) ; Wood, " Index," p. 231, no. 1597. Depictclla, H -Sch,, " Schmett. Eur. Micr.," pi. i., fig. 7 (1851). Paykullelki, Thnb., " Diss. Ac. Upsal.," iii., 101 (1794). ORIGINAL DESCBIPTION. Tinea alls auratis, maculis fasciaque media purpureis. Parva. Corpus nigrum. Antennae breves nigra\ Alae anticre aurere, uitidas. Maculse duae purpurese ante medium ad marginem crassiorem. In medio fascia lata purpurea ad marginem crassiorem bifida, versus apicem denique adhuc macula purpurea. Alae posticae auratfe, immaculatfe. Habitat Kilije (Kiel), Dom. Daldorf (Fabricius, Mantissa Insectoruin, No. 131, p. 253). IMAGO. Fore- wings 8-9 mm. ; shiny, light golden-green, with red costal spot near the base, an oblique fascia before the middle, another beyond the middle forked on thecosta and joined to a costal spot towards apex (the latter spot is sometimes absent). Hind- wings pale grey, with a purplish tint towards apex. VARIATION. Zeller describes (Linn. Ent., v., p. 334) two forms of this species as follows : 1. Capillis ferrugineis ; alls ant. saturate 'aureis nitidis, huraero, macula transversa costse prope basim maculaque maxima postica costam ter tangente purpureis = rubrifasciellti, Haw. Haworth's original description reads as follows : " Tinea rubrlfasciella (The red-barred Gold), alls aureis fasciis quatuor rubro- purpureis. Expansio alarum 4 lin. Ahe anticaj saturate aurea metallic fasciis quatuor tequidistantibus rubro-purpureis ; prima basi : secunda, aliquantulam arcuata ante, tertia pone medium et ad costam bifida ; quarta limbi postici, juncta ad tertiam juxta marginem tenuiorem. Cilia fusca. Posticae fusco-cupreee, prsecipue pone medium. Plabitat in Cantio, at valde infrequens " [Lep. Brit., No. 41., p. 572(1828)]. 2. Macula postica in duas soluta. priore costam bis tangente. Treitschke describes his anile -rschella, also considered a form of this species, as : 3. Alis anticis purpureis, maculis saepius confluentibus aureis = ander- schella, Tr. One form given by Zeller as a var. of 3/. aurcatella is referred here. This is : /. Ut a (=allionella, F.), sed macula postica ad marginem posticum usque producta ^ paykullella, Thunberg, described as " Alis purpureis, fasciis 3 aureis " (Diss., iii., p. 101). HABITAT. The habitats of this species vary exceedingly. In some places it abounds on the chalk-hills, as at Cuxton in Kent, in others it is equally abundant in fenny districts. In Bohemia it occurs on the sand-hills, and in Livonia, in the bushy districts on chalk-hills. Farren says that in Chippenham Fen, in 1891, the species was to be found swarming about the bird-cherry and fir cones. (It was so abundant that he boxed fifteen specimens in one large chip-box, off a single cone.) He states that it is more abundant on a close, warm day, soon after noon ; the firs on which this species was found were growing on the outside of a belt of trees in the Fen, the bird-cherry among the less thick parts of the ordinary fen growth. Hodgkinson reports it as swarming under the shade of a yew-tree, at Grange, in 1873 ; whilst, at the same place, Shuttleworth records it as flying in the partial shade afforded by beech and fir trees in 1882. Bower has found it flying in the afternoon sun, and has taken it by beating and searching fences and tree-trunks in various localities in Kent and Surrey. At Grantham, Atmore has obtained it by beating hawthorn, the moth MICROPTERYX THUNBERGELLA. 155 resting on the blossoms and flying by day. Kichardson says that the insect is common in many places on the downs near Worthing, flitting about near the ground. Cambridge notes it at Wareham, as being abundant on oak trees and underwood in May ; and Butterfield as being common in Scotch fir woods at Wilsden. Holland notes that the species flies around birch, nut, and other trees, yet it appears to be always beneath or near oaks, and a sweep of the net will some- times result in the capture of a score. Yet it can have no real con- nection with the oak, except for food or shelter. Farren notices it as resting on trunks of firs and oaks at Brockenhurst. South observes that in North Devon, on May 26th, 1882, he brushed the low branch of an oak-tree, in a small oak copse, and from it came quite a cloud of J/. thtmbergella. The insect, however, was on this day confined to the foliage of the tree first .shaken, although at the end of the month, single specimens were beaten out of various trees and the undergrowth. Barrett says that at Haslemere it visits the sedge-blossoms with M. calthella, Mm mansuetella and M. allionclla. Vine beats it out of furze bushes (Ulex europaeus) about the second week of May, near Brighton, and Durrani sweeps it from young birch trees (? feeding on catkins), at Merton. TIME OF APPEARANCE. Usually this species is to be found in May and early June, and Madam Lienig gives the first days of May as the time of its appearance in Livonia. Other recorded dates are as follows : May 28th, 1850, at West Wickham ; May llth, 1852, at Dartford Heath (Stainton) ; May 20th, 1886, at Witherslack ; May 31st, 1873, at Grani;e-over- Sands (Hodgkinson) ; May 13th, 1882, at Grange (Shuttleworth) ; June 15th, 1887, at Dartford Heath ; May 16th, 1888, at Headley Lane ; May 24th, 1888, June 5th, 1891, Greenhithe ; May 28th, 1891, June 5th, 1894, April 30th and May 27th, 1897, at Bexley ; June 14th, 1892, at Addington Park ; April 18th, 1893 and May 10th, 1895, at Box Hill (Bower) ; June 4th, 1887, at Cuxton (Tutt) ; June 26th, 1853, June 8th, 1856, June 6th, 1857, June 7th, 1873, June 2nd, 1875, at Richmond, Yorks. (Sang) ; in May, at Granthani (W. A. Atmore) ; end of May and beginning of June, at Chippenham Fen (Farren) ; May 23rd, 1887, May 16th, 1889, near Arundel, (Fletcher) ; May 26th. 1882, in North Devon (South) ; end of May, 1866, at Haslemere (Barrett) ; May 25th, 1874 and May 26th, 1877, at Witherslack ; May 14th, 1875, at Grange, and May 29th, 1878, at Windermere (Threlfall) ; May 20th, 1891, at Chingford, beaten from oak (Prout) ; May 26th, 1891, at Merton (Durrant) ; May lst-3rd, 1898, at Oxton (Studd). LOCALITIES. ABERDEEN : Pitcaple district (Reid). BERKS : Reading (Hamm) ; Sulham, Tilehurst, Padworth and Aldermaston (Holland). CAMBRIDGE ; Chippenham Fen (Farren). CUMBERLAND : Lake District (Stainton). DERBY : Henhurst (Brown). DEVON : North Devon (South). DORSET : Kimpton (Curtis), Wareham (Cambridge), Bloxworth, Glanvilles Wootton and Purbeck (Bankes). DURHAM; Darlington (Stainton). ESSEX: Chingford (Prout). GLOUCESTER : Bristol (Stainton). HANTS': New Forest (Fletcher), Pamber Forest (Holland), Brockenhurst (Farren). HEREFORD: Tarrington (Wood). KENT: Cuxton (Tutt), West Wickham (Stainton), Dartford Heath, Greenhithe and Bexley (Bower). LANCASHIRE : Manchester (Stainton), Croxteth Wood (Gregson), Grange (Threlfall). LEICESTER: Market Harborough (Matthews). LINCOLNSHIRE: Gran- tham (W. A. Atmore) . NORFOLK: Merton (Walsingham). NORTHUMBERLAND: New- castle (Stainton). SOMERSET: Castle Gary (Macmillan), Brislington (Sircom), Clevedon district, abundant (Mason). SUFFOLK: Tuddenham (Warren). SURREY: Haslemere (Barrett), Headley Lane, Addington Park and Bax Hill (Bower). SUSSEX: Thinly but widely distributed in the county (Fletcher), Lewes (Stainton), Goring Woods, Sompting (Fletcher), downs near Worthing (Richardson), Hastings and St. Leonards (Bloomfield), Moules Combe, near Brighton (Vine). WESTMOR- LAND : Windermere and Witherslack (Threlfall). YORK: Richmond (Sang), York (Prest), Scarborough (Stainton), Huddersfield 'Jnchbald), Sedbergh (Warren). DISTRIBUTION. Occurs in England and Scotland ; in Austria and in 156 fefclTiStf LEPIDOPTERA. Bohemia (Treitscbke) ; near Lauban, in Silesia (Wiesehiitter) ; Livonia, in the Duna district (Lienig), Kiel (Fabricius) ; Forest of Soignies, Belgium (Stainton). Staudinger and Wocke give : Livonia, Galicia, Germany, Switzerland and England. Constant adds : Burgundy. Meyrick writes : England (in woods, common) ; Central Europe. The following is a more complete list : Austria : Bohemia (Treitschke), Vienna (H.-Schaffer), E. Galicia. Belgium : Soignies (Stainton). Denmark (B.-Haas). France : Saone-et-Loire (Constant), Cher and Indre (Sand). Germany : Landsberg, Brandenburg, Stettin, Hamburg (Sorhagen), Wiirtemberg (Hoffmann), Riesengebirge (Zeller), Strelitz, Ratisbon (H.-Schaffer), Nassau (Rossler), Thuringia (Knapp). Russia : Livonia, Duna dist. (Lienig), Bielsteinshof (Nolcken). Sweden : Vestrogothia (Thunberg), Scania, Bahusia (Wallengren). Switzerland : nr. Zurich (Frey), Lausanne (Laharpe), nr. Schiipfen (Rothenbach). M1CBOPTERYX AUREATELLA, Scop. SYNONYMY. Species : Aitreatella, Scop., " Ent. Cam.," pp. 254-255, no. 662 (1763); Goze, "Beitr.," iii. (4), 149, no. 192 (1783); Stdgr. and Wocke, "Cat.," p. 108 (1861) ; Const., " Cat. Lep.," p. 311, no. 1128 (18(56) ; Stdgr. and Wocke, "Cat.," p. 340 (1871); Nik., Lp. Fn. Estl.," etc., pp. 495-7, no. 426 (1871) ; B -Haas, " N.H. Tdsk.," x., 47 (1875); Wallengren. " Bih. Svensk. Vet.," etc.. iii.. 82, no. 430 (1875) ; Hein. and Wk., " Schmett. Deutsch.," 774, no. 6 (1876) ; Zell., ' Stett. Ent. Zeit ," xxxix., pp. 161-2 (1878); Sand. "Cat. Lep. France," p. 203, no. 3099 (1879) ; Snellen, Vlinders," etc., p. 1064 (1882) ; Curo, " Nat. Sic.," i.. p. 89 (1882) ; Wallengren, " Ent. Tids.," 1883. 214 ; Sorhgn., " Kleinschm. M. Brandenbg.," p. 314 (1886); Meyr., "Handbook," etc. p. 806. Allionella, Fab., " Ent. Syst.," hi., pt. 2, p. 321, no. 148 (1794) ; Zell., "Isis," 1839, 185; Lienig. "Isis," 1846, p. 274; Tgstrom., "Not. Sallsk. Fn.," etc., i., 115 (1847); Sta.. Cat. Br. Tin.," p. 9 (1849); " Tr. Ent. Soc. Lond.," 1850, p. 31. pi. iii., fig. 7 ; Zell., " Stett. Ent Zeit.," xi., p. 63 (18oO) ; " Lin. Ent." (rars. a and b only), v., pp. 3;iO-l (1851) ; H.-Sch., " Schmett. Eur. Microp.," pi. i., fig. 6 (1851) ; Ghil., " Fn. Ent. Ital.," 78 (1852) ; Sta., Ins. Br.," pp. 43-4 (1854) ; H.-Sch.," Schmett. Eur.," v., p. 392 (1856) ; Frey, " Die Tineen," etc., pp. 411-50 (1856) ; Sta , " Ent. Ann.." 1857, 123; "Man.," ii.. p. 303 (18o9) ; Frey, "Ent Ann.," 1858, 142; Eosslr., " Jahrbuch Nassau," etc., xix.-xx., no. 1324 (1866) ; Sta., " Tin. S. Eur.," 56, no. 27, 91, 284 (1869); Sta., "Ent. Ann.," 1874,6 ; Frey, "Lep. Schweiz," 427 (1880) ; Porritt, "List Yorks. Lep.," p. 137 (1883); Walter, "Jena Zts.," xviii., pp. 751- 807 (1885). Amoeiiella, Stphs., "Cat. Br. Ins.," ii., p. *27, no. 7582 (1829). A)iiiii(indla, Curt . "Guide," 188, no. 1037, 19 (1831); Stphs., "Ill.,"iv., p. 362 (1835); Zett., "Ins. Lapp.," p. 1008, no. 10 (1840); Koch, "Isis," 1848,950; Hdnrch., " Lp. Eur. Cat.," p. 80, no. 139 (1851) ; Westwood and Humph., " Br. Moths " (2 Ed.), ii., p. 254 (1854). Tricinctella, Costa, " Fauna Napoli," p. 11, pi. ii., fig. 2 (1834). Paijkullelld, Werneburg, " Stett. Ent. Zeit.," xx., p. 67 (1859). ORIGINAL DESCRIPTION. Long. lin. 2. Alfe anticse rubro-aurata) ; fascia maculisque aureis. Macula aurea in limbo et alia ad basim, fascia intermedia. Caput rufis pilis pubescens. Oculi fusci. In sylvestris locis (Scopoli, Ent. Camiolica, pp. 254-255). IMAGO. Fore-wings 8-11 mm. ; purple, with a pale golden fascia near the base, another in the middle, and a pale golden costal spot near the apex. Hind-wings pale ashy, with a purple tinge towards the apex. VARIATION. This species, so far as may be judged from purely British specimens, would be considered a constant species with two golden fasciae, one basal, the other central, and a costal spot towards the apex. Continental specimens, however, are referred to this species, which frequently have an extra golden spot within the normal costal apical one, sometimes small, at other times much larger. There is also some slight difference in the direction of the central fascia, which is straighter in our British examples. Our own experience of the latter form was obtained at Cannes, on April 19th, 1898, when Dr. Chapman MICROPTERYX AUREATELLA. 157 .collected a long series. Examination of these, both alive and dead, tended to make us suppose that they represented a species distinct from the British examples. Lord Walsingham, however, refers them to M. aureatdla, and writes : " I have examined very long series of this species, ranging from Cannes through Switzerland and the Tyrol to North England. Although at first sight the British specimens appear to be distinguishable, I find variations, especially among the Bergiin examples, which seem to me to connect the two forms. Many of the foreign ones possess a strong golden-green suffusion about the termen (not on the dorsum), and the extra spot is usually (not always) present. Some British examples possess this extra spot, and some foreign specimens are without the terminal gloss. I am driven to the same conclusion as Snellen, that M. aureatella, which certainly occurs at Cannes (as entirely distinct from J/. ammandla and J/. aijlaelhi) is inseparable from our British specimens known as allionella, Fab." He also adds : " A/, ainmandla, Hb. ( = an(lersclidla, H.-S. in error), is easily distinguished from J/. aureatdla, Scop., by the golden-green gloss on the dorsum from the base to the fascia, whilst M. ayladla, Dup., is a much smaller species." Meyrick notes (Ent. Mo. Ma;/., vol. xxvii., p. 58) that this species is variable in Algeria, but it is possible that Meyrick is here referring toM. alyeriella, Rag. Zeller gives (Linn. Ent., v., p. 330) a table of six different forms which he refers to this species, of which, however, well-defined, connected with lower margin of cell by a bar near base, 2 and 3 from point of angle, transverse vein sometimes (chalcophmies) obsolete between 3 and 4, forked parting-vein well-defined, rising out of lower margin of cell near base, sometimes (chalcophanes) connected with upper margin of a bar near base, terminating in 4 and 5, between which the tranverse vein is absent, 7 and 8 stalked, 7 to hind margin, secondary cell well-defined, 9 and 10 out of its upper margin, 11 from i of cell, giving rise to an additional vein and connected with 12 by a bar above (chryg(injyru) or below (dialcopliancx) connected with upper margin of cell at base, giving rise to an additional vein above in middle, and sometimes (clialcophanex) a second near base. Hind-wings rather narrower than fore-wings ; ovate-lanceolate, cilia f , neuration identical with that of fore-wings, except as follows : 1 I rising out of upper fork of 1 a, not connected with cell, 2 GENERA OF EXOTIC MICROPTERYGIDES. 161 J 1 .nd 3 remote, transverse vein between 3 and 4 well-defined, the four main veins lot connected at base of wing, 11 from middle of cell, 11 and 12 without additional jranches" [Trans. New Zealand Inst., xviii., pp. 180-181 (1886)]. Comparing this genus with the typical genus Micropteryx, Meyrick says it " differs by the stalking of veins 7 and 8 in both wings, and the additional branch of 11 in fore-wings." He further adds [2V. New Zeal. Inst., xx. (1888), p. 91] that " in P. doroxena veins 7 and 8 of both wings are separate and the generic definition should be widened to include this case. The genus remains distinct from Micropteryx by the presence of the additional branch of vein 11 of the fore-wings." MNESARCHAEA. Imago : Head loosely haired, somewhat rough ; tongue obsolete, ocelli present. Antennae if, stout, filiform, in male simple, basal joint moderate, with- out pecten. Labial palpi moderately long, straight, porrected, clothed with long loose scales, forming a dilated terminal brush. Abdomen, in male, with uncus and valves well developed, and two long linear internal processes. Posterior tibise thinly clothed with long bristles, middle and posterior tarsi with whorls of projecting bristles at apex of four basal joints. Fore-wings with vein 1 simple, 2 almost from angle of cell, 6 out of stalk of 7 and 8 near base, 7 and 8 stalked, 7 to hind margin, 11 absent. Hind- wings of fore-wings, lanceolate, cilia rather over 1; neuration 'exactly as in fore-wings, except that vein 6 is separate from 7 [Trans. New. Zeal. Inst.. xviii., p. 180 (1886;]. To this diagnosis Meyrick adds (Ibid., xx. (1888), p. 90) : " In M. lo.i-oacia and M. hemadelpha, the tongue is well- developed, and vein 6 of the fore-wings is separate ; in all other respects the structure is identical with that of M. paracosma. The antennae in all the species are clothed with loose hair-scales, arranged in whorls at the joints ; the spurs of the middle tibiae are well developed." The comparison of the neuration of Palaeomicra with that of certain Trichopterygids led to a most interesting statement by Meyrick, who writes : " The nearest of these (Trichopterygids) to Palaeomicra is Il/n/acopfiila (Rhyacophittdtu) ; Cyrnus and Holocentropus (Hydropsy - chidae) also approximate closely, and Diplectrona and Hydropsyche in the same family, less nearly ; Calamoceras (Leptoceridae) is rather more remote. In the fore-wings of Rhyacophila the only important difference is the existence of an additional vein rising out of 4 ; but in the hind-wings one observes, with interest, that this very difference has disappeared, this additional vein being absent ; throughout these genera it seems that, in the tendency to a progressive simplification of structure, the hind-wings took the lead, with the result that in the finally established lepidopterous type the hind-wings have permanently four veins less than the fore-wings. lihyacophila shows no other essential difference from Palaeomicra ; the other points of difference consist in the position (whether above or below the furcations) of the transverse bars, or their partial obsolescence. Palaeomicra chalcnphanes is especially interesting, as being at present the only Lepidopteron known which shows the basal trifurcation of vein 1 a of the fore-wings, common to all the above-mentioned genera of Trich- optera ; and the same species possesses the second (basal) branch of vein 12 of the fore-wings, which is shown in Rhyacophila, but not in any of the others mentioned, except Hydroptyche, which does not, however, show the other or median branch. I may add that this basal branch is, perhaps, rather to be regarded as a transverse bar connecting vein 12 with the costa, than as a true branch. It appears to me that the type of neuration of the Trichoptera consists of five K 162 BRITISH LEPIDOPTERA. simple veins, variously fused towards the inner margin ; and seven apically furcate veins, variously fused towafds the base, and connected by a series of transverse bars." The genus Epimartyria is intermediate between Palaeomicra, Meyr. and Micropteryx, Hb. It appears to be more archaic, in some respects, than Microptery.t; and more closely allied to Palaeomicra. It is diagnosed by Walsingham as follows : EPIMARTYRIA. Imago : Antennae (f ) moniliform, each bead with a fringe of long hairs from the circle of its widest diameter, projecting forward as far as the middle of the bead above it. Mandibles developed. Haustellum absent. Labial palpi (? absent). Maxillary palpi strongly developed, 5-jointed, scaled, hairy on the basal joints, curved inward. Maxillae distinct. Ocelli present. Head and face very rough. Fore-wings with the costa somewhat excavate at the base, thence slightly arched, apex rounded ; somewhat widened across the middle. Neuration : 1ft furcate at base, connected by a bar with Ic, which has a small fork at the extreme base ; vein 2 out of 3 immediately beyond angle of lower cell, 2 + 3 (cubital) con- tinued to near the base, but becoming furcate before the bar from Ib to Ic ; the lower limb of the fork is almost obsolete, and goes to Ic ; the upper is distinct and goes to the median (4 + 5 + 6) ; 4 furcate, one limb going to cubital before origin of 2, the other limb to median at about ^, at which point 5 is bent down to unite with it ; 5 and 6 separate and almost parallel ; 6 furcate at base connecting median with radial ; 7 and 8 stalked (7 to slightly above apex) and continued through cell to about midway between 11 and 10 ; the transverse veins joining 10 to 9, 9 to 7 + 8. and 7 + 8 to 6 are weak, and that between 9 and 10 is furcate, enclosing a small cell between its limbs on vein 10 ; 11 and 12 connected by a transverse bar, before which vein 12 throws out a branch to the costa ; costa chitinised at base, outwardly sharply defined by a short humeral vein ; jugum developed ; anal margin of the wing chitinised. Hind-wings as broad as the fore-wings, apex rounded ; with flat scales. Neuration as in the fore-wings, but Ib not furcate at base, Ic not connected with cubital, and cubital not connected with median towards base ; transverse vein between 9 and 10 not furcate, 11 absent. Abdomen : genital armature, consisting of four lateral plates and strong bifid uncus, the lower plates with a tooth at their apex. Legs with hind tibiae not hairy above, median spurs moderate, apical very small ; middle legs without spurs (Entom. Record, etc., pp. 161-162, July, 1898). The beads of the antennae have longer stalks, and thus are more distinctly separated than those of Microptcryx (calthella) ; moreover, in M. calthella they are shaped like a small conical bullet with the base outwards, whereas, in E. pardella, they are more spherical, and are vase-shaped, with the mouth straight, and wider than the stem, the middle portion bulged, and almost flanged. No such structure is observable in the normally cylindrical and closely compressed joints of Eriocrania, Z. ERRATUM. p. 160, line 5 from bottom, read " or below (chalcophanes) the additional vein, 12 sometimes (chalcophanes) connected with upper margin," etc. Superfamily II : NEPTICULIDES. This superfamily includes the smallest known of all Lepidoptera, the imagines averaging from 3-8 mm. only in wing expanse. The rough head and face, and folded palpi are very characteristic of the Nepticulid species, whilst the antennae are short, and not unlike those of the Micropterygids. The anterior wings are short and broad, and the scales comparatively large for the size of the moths. The eggs are laid on a leaf or leafstalk, the young larvae, on hatching, imme- diately boring into the leaf-substance directly beneath the egg. The larva lives in the parenchyma, between the upper cuticle and the median vascular structure of the leaf, and makes a mine, the character of which is generally very marked for each individual species. Whether the egg be laid on the upper or underside of the leaf, the larva makes its way (with a very few exceptions) to the upperside, and confines its THE NEPTICDLIDES. 163 operations almost entirely to that portion. The larva is without any true chitinous legs, certain of the thoracic and abdominal segments bearing membranous prolongations, analogous in structure with the prolegs of other lepidopterous lame, but having no terminal crochets or hooks. De Geer says that there are 9 pairs (Wood allows but 8) of these modified legs of which the third pair is very ill -developed. The larva usually quits its mine in order to spin its cocoon, but those of some species, such as N. iceaceri, N. septembrella, N. agriinoniai', etc., make their cocoons in the mine itself. The cocoons vary much in size, shape and colour, and the pupa, in common with those of most INCOMPLET/E, usually protrudes its head and anterior segments before the emergence of the imago. The pupa itself is a " Pupa libera," with the segments un- fixed, and the appendages unsoldered to the rest of the pupal structure. The minute imagines fly freely in the sun, each species having its own particular time of activity ,, after which they rest in the crannies of the tree-irunks or branches, or sun themselves upon the leaves. In windy weather, they seek the shelter of fences, etc., near their haunts, and we have seen the crannies on the trunks of the oak trees in Chattenden woods, swarming with incredible numbers of N. subbimacidella (and smaller numbers of other species) on such days. It is a remarkable fact that, when the leaves containing the mines of these insects fall to the ground in autumn, the part of the leaf containing the mine resists decay long after the rest of the leaf has become withered, the part containing a larva remaining green after the other parts have changed colour. In 1771, De Geer wrote an excellent detailed description of the life history of N. anomalella. In 1793, Fabricius described and named a species of the genus, N. aurella. Hiibner figured and named two species, but at present they have not been recognised. Haworth, in 1828, gave good descriptions of 10 species, one of which is the Fabrician X. aurdla, and diagnosed many others, which he treated, however, as aberrations of the species he named. Zeller, in the his of 1839, diagnosed 8 species, of which 5 were identical with those described by Haworth, whilst in 1848, in the Linnaea Entomoloyica, vol. iii., the same author established 13 species, of which 3 were new. In 1851, Stainton published his Systematic Catalogue of the Tineidae, and this contained 18 species of the genus. This was followed (1855) by the Natural History of the Tineina, vol. i., in which 33 species were enumerated. In the same year, Herrich-Schaffer in his tiystematische Bearbeituny der Schtiiett.ron Europa was able to describe 48, whilst Frey, in the Linnaea Entomoloijica, vol. xi. (1857), monographed 58 species. In 1862, Stainton published the Natural History of the Tineina, vol. vii., in which the total number of species is placed at 74 certain, 4 others doubtful, and mines of two unknown species from South America. By 1871, Staudinger and Wocke in their Catalog, etc., were able to list 111 species then known to inhabit the Palsearctic area. Since then, several other species have been added, of which 6 are British. To Wood's papers on this superfamily (Ent. Mo. Mag., xxix., pp. 197 et seq.) we are greatly indebted for much practical information concerning the habits of some of the more obscure species. It is a speaking monument to Stainton and his colleagues, that the life-histories of the species comprised in this large superfamily are, perhaps, better known in England and Germany than those of 164 BRITISH LEPIDOPTERA. any other group of equal size, included in the so-called Micro- Lepidoptera. Besides the large genus Nepticula, at least two other genera belong to this superfamily, viz., Tnfurcula and Scoliaula (lioliemannia). When more is known of the early stages of other genera, probably also these, too, will have to be admitted. The impedimenta required for collecting the mines of these species consist only of a small pocket lens, and a few tin boxes in which to carry the leaves containing the mines when found. The specimens in collections are almost all bred from mines thus obtained. The species are mostly double- or continuously-brooded, and the mines may, there- fore, be collected in the summer month's, although autumn is the time of year usually occupied with this pursuit. The whole existence of one of these mining larvre is almost always spent within the confines of a single leaf, and hence the area of study is, so to speak, circum- scribed within these narrow limits. There are various means of determining, in the larval state, all our British species. The points which aid in this determination are : (1) The food-plant. (2) The position of the egg. (3) The form of the mine. (4) The arrangement of the frass. (5) The structural characters of the larva. The food-plants of almost all our British species are well-known. The Uosaceae, Salicineae and the Ciqndiferae supply, between them, the food-plants of about five-sixths of the British species. The birch supports at least 8 (? 9) species, the oak 5, hawthorn 6, and apple 5, 5 feed on rose, 6 on brambles, at least 3 (? 4) on the willows, and 4 on pear. The Vacciniaccae support 2 species, the Urticaceae (elms) 3, whilst 6 other natural orders (including the Leyuminosae and Labiatac] contain but one species of plant each, which is known to be acceptable to the Nepticulid larvae, and each of these is tenanted by only a single species ; the single Hypericwn species, however, is not so particular in its choice of a single species for food-plant. The total number of the British species (including doubtful ones) obtained from the lioaaceae and CupnUferae amounts to 58, whilst only 20 are obtained from plants of all other natural orders. Many species are confined to a single food-plant, but others are not so limited. Both the nut species, N. flodactella and N. microtheriella, are also found on Carpinus betulus. N. aeneofasciella feeds on Potentilla tormentiUa and Agriwonia eupatoria ; N. oxyacanthella occurs on apple and pear, as well as on mountain-ash, Cotoneaster ajfinis and hawthorn ; N. atricollis on apple and hawthorn, as well as on pear; N. angulifa&ciella on Poterium sanyuisorba as well as rose ; whilst Wood gives the palm, in this respect, to N. aurella, which, he says, feeds on bramble, strawberry, agrimony and Spiraea. He also considers that N. gei, from Cfeum, is only a form of this species. Whether the palm should net be given to N. oxyacanthella is doubtful. Fletcher says that he has never found N. aurella in the leaves of any- thing but evergreen bramble (Rubtis fruticosus), and he suspects that the larvae on strawberry, agrimony, Spiraea and Geum are N. gei, which he has found in leaves of all these plants. As to N. aurella and X. gei being different species, Fletcher writes : " N. aurella and N. gei are very much alike, still I can always breed them pure. The mines are somewhat different, and the larvre occur at different seasons." He also notes : " N. xplendidissimdla is quite distinct ; the mine is more distorted, tends to be more in the outer angle of the leaf, and has a less tendency to THE 165 run along a vein of the leaf than that of IV. yei. The larva feeds in dew- berry and Rubus corylifoliits, is local, and is fond of leaves in cover of woods, etc., in Sussex and Lincolnshire, also among rough grass on sandhills at Mablethorpe. I have bred it freely from Rubus chamnetnorus, from Scotland, but from no other plants. The moth is much smaller than either IV. aurella or N. gei (the comparison is made on 60 specimens, picked from a very much larger number of all three species). IV. splendidissimella has conspicuously black head, contrast- ing sharply with the colour of the eyecaps, and always suggests strongly to me that it wears ' gig-lamps ' " (in litt., May 18th, 1898). The following list of the British species of the genus Nepticula, arranged according to the natural orders on which the larvae feed, will help to illustrate the large number of species that feed on plants of the natural orders, Rosaceae, Cupuliferae and Salicineae. The species (which are not arranged with any view to relationship, and often have other food-plants besides those mentioned) are as follows : 1. Feeding on ROSACE.E : FOODPLANT. FOODPLANT. N. anomalella, Goze 11. canina, etc. N. splendidissimella, II. csosius, R. N. fletcheri, n. sp. llosa arvensis H.-Sch. corylifolius, etc. N. angulifasciella, Sta. 11. canina ? N. tengstromi, Nolck. Rubus chamoe- ? N. hodgkinsoni, Sta. Rosa ? sp. morus N. centifoliella, Zell. It. rubiginosa, N. arcuatella, H.-Sch. Fragaria vesca ? subsp. mi- and ? Poten- crantha tilla f ragarias- N. pygmteella, Haw. CratiEgus oxya- trum cantha ? N. dulcella, Hein. F. vesca N. ignobilella, Sta. C. oxyacantha N. serella, Sta. Potentilla tor- N. atricollis, Sta. C. oxyacantha, mentilla Pyrus malus, N. agrimoniffi, Frey Agrimonia eu- P. communis patoria N. gratiosella, Sta. C. oxyacantha N. teneofasciella, H.- A. eupatoria N. regiella, H.-Sch. C. oxyacantha Sch. and Potentilla N. oxyacanthella. Sta. C. oxyacantha, tormentilla P. malus, P. N. fragariella, Heyd. Fragaria vesca, communis, P. Agrimonia eu- aucuparia and patoria Cotoneaster ? N. gei, Wocke Geum urbanum, atfinis Rubus coryli- N. pomella, Vaugh. Pyrus malus folius N. desperatella, Frey P. m ( alus 2. Feeding on CUPULIFKKJE : N. malella, Sta. P. malus N. atricapitella, Haw. Quercus robur N. pulverosella, Sta. P. malus N. ruficapitella, Haw. Q. robur N. pyri, Glitz N. minusculella,H.-Sch P. communis P. communis N. basiguttella. Hein. Q. robur N. subbimaculella.Haw. Q. robur N. aucupariffi, Frey P. aucuparia N. quinquella, Beddell Q. robur N. nylandriella, Teng. P. aucuparia N. castanella, Edles. Castanea sativa N. sorbi, Sta. P. aucuparia N. turicella, H.-Sch. Fagus sylvatica N. torminalis, Wood P. torminalis (=tityrella,coll.Ang.) N. prunetorura, Sta. Prunus spinosa N. basalella, H.-Sch. F. sylvatica N. plagicolella, Sta. P. communis, (= fulgens,coll.Ang.) myrobalana, N. floslactella, Haw. Carpinus betu- sinensis lus and Cory- N. filipendulsj, Wocke Spiraea nlipen- lus avelliina dula N. microtheriella, Sta. C. betulus and N. poterii, Sta. Poterium san- C. avellana guisorba N. alnetella, Sta. Alnus glutinosa N. rubivora, Wocke Rubus csBsius N. glutinosae, Sta. A. glutinosa N. aurella. Fab. 11. fruticosus N. betulicola, Sta. Betula alba, ? B. N. auromarginella, 11. fruticosus nana llich. N. woolhopiella, Wood B. alba 166 BRITISH LEPIDOPTERA. FOODPLANT. N. argentipedella, Zell. B. alba N. confusella, Walsm. B. alba and Wood N. continuella, Sta. B. alba N. luteella, Sta. B. alba N. lapponica, Wocke B. alba N. distinguenda, Hein. B. alba ? N. bistriinaculella, B. alba Heyd. 3. Feeding on SALICINE^E : N. argyropeza, Zell. Populus tremula (apicella, Sta.) N. subapicella, Sta. ignota (argyopeza, Sta.) N. assimilella, Zell. N. trimaculella, Haw. N. intimella, Zell. N. vimineticola, Frey N. salicis, Sta. ? N. diversa, Glitz P. tremula P. nigra Salix russelliana S. alba S. caprea, S. aurita and S. cinerea S. caprea 4. Feeding on UBTICACE.E : N. ulmivora, Fologne Ulmus carnpes- tris N. viscerella, Sta. U. campestris N. marginicolella, Sta. U. campestris FOODPLANT. 5. Feeding on SAPINDACE.E : N. sericopeza, Zell. Acer campestris 6. Feeding on TILIACE.& : N. tiliffi, Frey Tilia parvifolia 7. Feeding on RHAMNEJE : N. catharticella, Sta. Khamnus cath- articus 8. Feeding on VACCINIACE.E : N. weaver!, Sta. Vacciniumvitis- idsea N. myrtillella, Sta. V. myrtillus '9. Feeding on HYPERICINEJE : N. septembrella, Sta. Hypericum quadrangu- lum, H. per- foratum 10. Feeding on POLYGONACE*: : N. acetosse. Sta. Eumex acetosella 11. Feeding on LABIATE : N. headleyella, Sta. Prunella vul- garis 12. Feeding on LEGUMINOS.E : N. cryptella, Sta. Lotus cornicu- latus The Palfearctic species of the genus, not yet discovered in Britain, are as follows : N. samiatella, H.-Sch. Quercus robur N. mespilicola, Frey Sorbus aria, N. uniformis, Hein. Salix caprea Amelanchier N. nitidella, Hein. Crataegus oxya- vulgaris cantha N. dewitziella, Sorhgn. Salix caprea N. paradoxa, Frey C. oxyacantha N. strigilella, Thbg. ignota N. subnitidella, Zell. ignota N. gilvella, Rossi. ignota N. rhamnella, H.-Sch. Rhamnus cath- N. lemniscella, Zell. Ulmus articus N. ilicella, Wlsm. Quercus ilex N. lonicerarum, Frey Lonicera capri- N. trifolii, Sorhgn. Trifolium folium N. nobilella. Wocke ignota N. tristis, Wocke Betula nana N. angustella, Hein. ? Tormentilla N. sanguisorbee, Wock 3 Sanguisorba and Wocke or Fragaria officinalis N. suberoidella, Wlsm. Quercus ilex N. stettinensis, Hein. Pyrus commu- N. rubescens, Hein. Alnus glutinosa nis N. lediella, Schleich Ledum palustre N. pyricola, Wocke Pyrus commu- N. inaequalis, Hein. Fragaria vesca nis N. occultella, Hein. Tormentilla N. ilicivora, Peyr. Quercus ilex erecta, Poten- N. aceris, Frey Acer campestris tilla anserina N. latifasciella,H.-Sch ignota N. potentillce, Glitz T.erecta.Poten- N. pretiosa, Hein. Geum urbanum tilla anserina N. bollii, Frey Kubus fruticosus N. hiibnerella, Zell. ignota N. tormentillella, Tormentilla N. dimidiatella, H.-Sch. ignota H.-Sch. erecta N. geminella, Frey Poterium san- N. dryadella. Hoff. Dryasoctopetala guisorba N. nitens, Fologne Agrimonia eu- N. diffinis, Wocke Sanguisorba patoria omcinalis N. comari, Wocke Comarum |pa- N. ulmariffl, Wocke Spiraea ulmaria lustre N. zelleriella, Snellen ? Salix fusca N. penicillata, Wocke ignota N. latifoliella, Mill. Phillyrea lati- N. speciosa, Frey Acer pseudo- folia platanus N. hemargyrella, Zell. F. sylvatica N. flexuosella, Fologne ignota THE NEPTICULIDES. 167 FOODPLANT. FOODPLANT. N. freyella, Heyd. Convolvulus se- N. cistivora, Peyr. Cistus monspe- pium, C. arven- liensis, C. sis salvisefolius N. schleichiella, Frey Sanguisorba N. albifasciella, Hein. Quercus officinalis N. promissa, Stdgr. Pistacia lentis- N. aterrima, Wocke Cratsegus oxya- cus, Ehus cantha cotinus N. obliquella, Hein. ? Salix ' N. turbidella, H.-Sch. Populus nigra, N. suberivora, Sta. Quercus suber P. alba N. suberis, Sta. Quercus suber N. hannoverella, Glitz Populus pyra- N. carpinella, Heyd. Carpinus betu- midalis lus N. simplicella, Hein. ignota N. fagella. H.-Sch. ?Fagus sylvatica N. euphorbiella, Sta. Euphorbia den- N. helianthemella, Helianthemum droides H.-Sch. vulgare N. alpinella, H.-Sch. ignota N. decentella, H.-Sch. Acer pseudo- N. aureocaputella, Mill ignota platanus N. argyrostigma, Frey ignota N. rubiella, Rag. (m.s.) Kubus N. wockeella, Hein. Salix alba The above forms a moderately complete list of the known Nepticulid species inhabiting the Palsearctic area, with the food-plants of the larvae. The superfamily has, however, been so little worked outside Europe, that a list, sent to me by Mr. Durrani, appears to be worth reproducing, as it will serve to show not only how cosmopolitan is its range, but what a possibility of discovery there is for any one who will work at the group systematically in those countries from which species have been recorded. This list reads as follows : From NOBTH AMERICA : Nepticula amelanchierella, Clem. N. anguinella, Clem. N. apicialbella, Chamb. N. badiocapitella, Chamb. N. belfrageella, Chamb. N. bifasciella, Chamb, N. bosqueella, Chamb. N. castanenefoliella, Chamb. N. carysefoliella, Clem. N. cillicefuscella, Chamb. N. clemensella, Chamb. N. corylifoliella, Clem. N. cratsegifoliella. Clem. N. dallasiana, F. and B. N. fuscocapitella, Chamb. N. fuscotibiseella, Clem. N. grandisella, Chamb. N. juglandifoliella, Chamb. N. latifasciella, Chamb. N. maculosella, Chamb. N. maximella, Chamb. N. minimella, Chamb. N. nigriverticella, Chamb. N. ochrefasciella, Chamb. N. ostryaefoliella, Clem. N. platanella, Clem. N. platea, Clem. Neptieula populetorum, F. and B. N. prunifoliella, Clem. N. pteliacella, Chamb. N. quercicastanella, Chamb. N. quercipulchella, Chamb. N. resplendensella, Chamb. N. rostefoliella, Clem. N. rubifoliella, Clem. N. saginella, Clem. N. serotinseella, Chamb. N. thoracealbella, Chamb. N. unifasciella, Chamb. N. villosella, Clem. N. virginiella, Clem. From SOUTH AMEBICA : Nepticula molybditis, Zell. (Colombia). N. johannis, Zell. (Bogota) From CHINA: Nepticula mandarinella, Wlsm. (Chang-hai). M.S. From NEW ZEALAND : Nepticula tricentra, Meyr. N. ogygia, Meyr. N. propaloea, Meyr. Currant further writes : " I have seen a species of Nepticula from St. Thomas (Danish West Indies), and Lord Walsingham has un- described species (more than one) from Australia." Hering, quoting a letter from Hedemann, also notes (Stett. Ent. Zeit.} the occurrence of two unnamed species of the genus in the West Indies. 168 BRITISH LEPIDOPTERA. The Nepticulid egg is rather large for the size of the moth, of the " flat " type, ovate in character, roundish-oval in outline, some- what domed above and flattened beneath, and, in spite of its gene- ralised nature, not unlike that characterising certain superfamilies belonging to the Sphingo-Micropterygid stirps. Chapman states that it is not unlike that of Heterorfenea cruciata (asellus). The larva does not, as is usual with lepidopterous larvae, eat its way out of the micropylar end, or the upper side of the shell, but bores directly through the base into the leaf below on which the egg is laid. As a result of this the empty egg-shell is usually of a black or brown colour, due to the presence of frass which the larva deposits in it as it bores its way into the leaf. The position in which the egg is laid is usually very constant for each species. Of 41 species, observed by Wood, 37 exhibited habitually some preference as to the position selected. Sometimes this is merely the selection of one side of a leaf in preference to the other, at other times this preference extends to a particular part of the leaf such as the extreme edge, the shelter of a projecting rib, or other position. The egg of N. intimella is placed on the petiole of a leaf of SalLc ruxselliana or on the upper surface of the midrib of .S'. caprca for the larva burrows at first into the stalk or midrib, and only, in the latter part of its life, mines the lamina of the leaf. The egg of N. ar